American English pronunciation of Uyghur proper nouns

« previous post | next post »

The discussion following my original post ("A Little Primer of Xinjiang Proper Nouns", 7/13/2009) has proven quite edifying, at least to me.  One thing that I realized from the lively comments is that I forgot to give an indication of how the name Xinjiang itself should be pronounced.  There's also the question of what sort of "pronouncers" or "respellings" to provide for speakers of American English who need to pronounce these names but cannot be expected to render them exactly the way a native speaker of Uyghur would.

I must preface the following remarks by stating that I'm probably not the best person to offer standard American English readings of these names because I'm a fluent speaker of Mandarin and know a bit of Uyghur. Consequently, when I want to say these words as an American would, I'm afraid that my Mandarin and my Uyghur get in the way. Still, I will make an honest effort to separate the three modes and offer useful guides for speakers of American English.

Xin1jiang1 新疆 The Uyghur pronunciation of this Mandarin term is Shinjang شىنجاڭ.   I think that it's all right for an American simply to say the name Xinjiang about the way the Uyghurs do: SHIN-jahng (I put the "h" in there to lengthen the "a"). Being acculturated to Mandarin, I find it hard to avoid saying something like SHIN-jeeahng, even when I'm trying to speak as an American.

Recordings of Xinjiang:

1. Mandarin   

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

2. American English

a. unmodified

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

b. modified

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

"Sinkiang", the Postal Map Romanization of the name, represents a time / place / dialect (Shanghai, 1906) before the palatalization of the velars (compare Peking and Beijing).

As for how the Uyghurs themselves refer to Xinjiang (Eastern Central Asia to me), depending upon what part of it they come from or the circumstances in which they are speaking, they will call it variously Uyghur Eli (country of the Uyghurs), Uyghur Diyari (Uyghur land), Uyghur Rayoni (Uyghur Region), Sherki Türkistan (Eastern Türkistan), or Uyghuristan, and so forth.


We've gone over the variant spellings of Uyghur in my earlier post, so now we have to confront the diverse pronunciations that have been applied to this name. First of all, I don't see how anyone can get "weewer" out of what the Uyghur speaker twice says very clearly in the recording. Due to severe tinnitus, my ears are very bad, but I can distinctly hear a guttural "gh" sound at the beginning of the second syllable. Now, it is very difficult for an American to pronounce this sound the way the Uyghurs do, and I don't think it's necessary for someone speaking English to attempt it. On the other hand, we needn't resort to something like "wigger" either. I think wee-GOOR / WEE-goor and WEE-gur are both fine for American English. Most people I know do say WEE-gur, as the NYT had it, and that seems quite natural, although wee-GOOR / WEE-goor more nearly approximates the Turkic sounds. Now, when I'm in the presence of Turkic speakers, even in America, I usually aspirate the 'g', so it would be something like wee-GHOOR / WEE-ghoor or WEE-ghur, even though I know that's a pretty pathetic approximation of the sound the Uyghurs make at the beginning of the second syllable.

American English pronunciations of Uyghur:

a. unmodified: 

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

or

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

b. modified: 

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.


Finally, the name of the capital of the region, Urumchi (and numerous other spellings), has a broadly accepted pronunciation in English as oo-ROOM-chee. Again, when I'm in the presence of cognoscenti, I'll put umlauts on the "u's".


American English pronunciations of Urumchi:

a. unmodified: 

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

b. modified: 

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

To sum up, for the standard American renderings of these three names, I recommend the unmodified pronunciations: SHIN-jahng, WEE-gur or WEE-goor / wee-GOOR (emphasis on the first syllable is preferred), and oo-ROOM-chee.



52 Comments

  1. Vijay John said,

    July 15, 2009 @ 11:45 am

    Interesting that you aspirate the "g" instead of making a velar(?) fricative. It makes you sound rather Indian :-D

  2. Doc Rock said,

    July 15, 2009 @ 12:08 pm

    Just to note the obvious, the "Rayoni " of "Uyghur Rayoni (Uyghur Region)" is a loan into Uyghur from Russian.

  3. Tlönista said,

    July 15, 2009 @ 1:07 pm

    Sigh. I listened to the sound clip of "Uyghur" in the previous post again, but still can't hear anything but "wee-wer". The second w does sound distinctly different from the first, but it's still definitely a w.

    [(myl) Maybe your eyes can help convince your ears. Here's a wide-band spectrogram of the first rendition from the previous post:

    The second formant (the sort of inverted-U thing that rises to a peak of about 2700 Hz in the first syllable) falls in a way that indicates the labialization of the medial consonant. And the fact that the second formant continues through the medial consonant shows you that it's not a stop. But the amplitude of the medial consonant is lower than we'd expect it to be for a medial /w/ — and most important, look at the frication noise between about 4 and 16 kHz.

    It's definitely a voiced fricative — though the voicing becomes quite weak, and maybe even stops for a period or two in the middle:

    I have to say that the evidence for velar stricture is not very strong in this particular rendition — it might simply be an aspirated [w] or a labialized [ɦ]. ]

    I can kind of mimic it by combining [x] (or [χ]) and [w] to make a kind of khw sound. But it's hard because ɣ simply doesn't exist in any language I speak. My roommate (who has English as a first language and speaks some Arabic and Japanese) took a look at the Wikipedia chart and pronounced it [x] and [g], variously.

    I give up! Wee-gur it is.

  4. Victor Mair said,

    July 15, 2009 @ 1:20 pm

    From Peter Golden, a Turkologist at Rutgers: "Uighur is, of course, 'oooyghurr,' and I must confess that Weegur grates on my ears. A Turk would have some problems pronouncing that."

    Peter adds that "Someone should put together a brief, but serious reference work on the toponyms and ethnonyms associated with Xinjiang, including Han and Tang-era reconstructed pronunciations of the Chinese names, as well as Khotanese, Tokharian and of course Arabic and Persian forms."

    I'm all in favor of Peter's suggestions. One of the banes of those who study Central Asian history is reconstructing the original place names and personal names on the basis of their transcriptions and transliterations in other languages. Fittingly, I just happened upon this illustration of how to use the word "bane" in one of its senses (source of persistent annoyance or exasperation): "The spellings of foreign names are often the bane of busy copy editors" (Norm Goldstein). And that reminded me of Geoffrey Pullum's posts on copy editors:

    http://languagelog.ldc.upenn.edu/nll/?p=1565
    http://languagelog.ldc.upenn.edu/nll/?p=1561

  5. Tim said,

    July 15, 2009 @ 1:29 pm

    When I listened last night to the clip in the previous post, I definitely heard "weewer", as well. Since I knew that couldn't be right, I switched from speakers to headphones and listened very carefully and could just barely hear the correct sound. Although, it still seemed to be overlaid by a /w/, as if the speaker were forming that sound with their lips at the same time.

  6. john riemann soong said,

    July 15, 2009 @ 1:52 pm

    Is the palatalization that recent?

    If a 1920s textbook reflects the near-modern pronunciation, I do wonder.

    Are there any papers on the lexical diffusion of this sound change?

    10-20 years seems awfully rapid, a fraction of the time it took for the Great Vowel Shift.

  7. vanya said,

    July 15, 2009 @ 1:55 pm

    I must confess that Weegur grates on my ears

    Why does Golden say that? As far as I know Turks from Turkey don't use a velar or aspirated g in that word either, I think they say "Uygur (i.e. "Wee-goor"). If it's fine for Anatolian Turks, should be OK for Americans.

  8. Ellen said,

    July 15, 2009 @ 2:00 pm

    It's not so much that those of us who "hear" is as weewer are hearing a w in the middle. Rather, we are mapping what we hear onto our w phoneme.

    (I think -oor/-ur versus -er is due to imagining that as -er said with an accent, as I would if it were an English language word.)

  9. Boris Blagojević said,

    July 15, 2009 @ 2:01 pm

    Are [ʊɪ̯] or [ʊj] really that hard for Americans to pronounce?
    This is genuine interest, I'm not questioning anyone's mental abilities… I just wouldn't have expected that forcing /j/ to the coda position, or pronouncing a diphthong similar to the already existing ones would be such a harsh violation of phonotactic rules, that it would be necessary to fall back to /wi/ or something similar (perhaps Oygur or Oygoor?).

  10. john riemann soong said,

    July 15, 2009 @ 2:16 pm

    What's the difference between /ʊj/ and /wj/ anyway? (Or w^j)?

    Sometimes I see "house" transcribed as /ha^ws/. (where ^w denotes labialisation.)

  11. Sili said,

    July 15, 2009 @ 2:18 pm

    Not being an American I can't map that middle consonant to /w/, but I have to admit that I have trouble identifying it as anything in my own limited inventory – /x/ or /ʁ/ seem to be closest (yes, I'm aware that one is voiced and the other is not – one of the many reasons I'll never become a linguist or learn to speak English without an accent).

  12. Faldone said,

    July 15, 2009 @ 2:34 pm

    I hear something like wee-wer, too, but the second W is kind of like the Wh in some dialects' pronunciation of words like "what" except that instead of an H it's a ɣ. I suspect the W sound many of us hear is like the ʊ I hear in Russian ТЫ. People who know more about Russian than I do say it's not there and I have generally interpreted it as a transitional sound between the Т, which I take to be pronounced in the front of the mouth, and the Ы, which I take to be pronounced in the back of the mouth.

  13. bulbul said,

    July 15, 2009 @ 3:30 pm

    As chance would have it, I was watching BBC's "Real Chinese" and at one point, the word "Uighur" cropped up. What I heard the presenter say was [wiːgɚ]. Had I not read the previous thread, it would have been one huge whiskey-tango-foxtrot moment.

  14. Tadeusz said,

    July 15, 2009 @ 4:45 pm

    Interestingly enough, the pronunciation dictionary by John Wells, highly regarded by those whose English is second language (and who speak British English) suggests as THE pronunciation of Uighur exactly what the BBC person said:
    ˈwiːɡə
    with the "better" pronunciation as the second one
    ˌuːiˈɡʊə.
    I suppose the problem is with phonotactics, not with skills. Speakers of English (not only those of American English) do not have the sequence /ɔɪ/ at the onset of a syllable, i.e., those speakers who have not had any phonetic training.

  15. Lazar said,

    July 15, 2009 @ 5:37 pm

    @Tadeusz: You mean [ʊɪ], right? On a related note, I've noticed that I tend to pronounce the interjection "ew" (one of disgust) as [ɪʊ], a diphthong that appears nowhere else in my speech. Does anyone else do this?

  16. Victor Mair said,

    July 15, 2009 @ 6:02 pm

    Justin Jacobs reminded me that a more precise transliteration of the Uyghur orthography itself would look something like "Oygur," with an umlaut over the O and a sign of gutturality added to the g, but no one (except a linguist) would know how to pronounce that.

    Juha Janhunen pointed out to me that the reason why Turkologists prefer a "y" over an "i" after the initial "U" is because it better corresponds to the phonotactics of Uyghur itself, where /y/ is a consonantal sound.

    Since I waffled on several different "pronouncers" for Uyghur, I feel obliged to say that my favorite recommendation for Americans would probably be wee-GHOOR. The second half forces English-speakers to get an approximation of the guttural in there, and avoids a "-ger" sound. The Uyghurs really don't like the sound of "-ger" at the end, and the WEE-gur the NYT had often results in just that.

    Simon ( seconded by Language Hat) recommends that we follow a standard reference, and actually cites the Webster transcription.

    \ˈwē-ˌgu̇r\

    You can hear the Webster version by clicking on the red speaker here:

    http://www.merriam-webster.com/dictionary/Uighurs

    This comes from a friend who specializes in Uyghur political and legal affairs: "My favorite pronunciation for 'Uyghur' (spotted in an article a few weeks ago) is 'WHEE-gur.' It just makes all things Uyghur seem like so much fun."

    To further lighten the spirits of those who care about the Uyghurs and their culture, here is an announcement about an exhibit of Uyghur art at the R St Gallery in Dupont Circle, Washington DC:

    http://www.rstgallery.com/id17.html

  17. Dan Velleman said,

    July 15, 2009 @ 6:32 pm

    It's not so much that those of us who "hear" is as weewer are hearing a w in the middle. Rather, we are mapping what we hear onto our w phoneme.

    This.

    I thought this was Ling 101 stuff. People without special training can't distinguish phonemes that aren't distinguished in their native language(s). A monolingual English speaker isn't just going to have trouble pronouncing [ɣ], they're going to have trouble hearing it as distinct from both [g] and [w]. Some map it onto their /g/ phoneme because it's velar and unrounded like /g/, some onto their /w/ because it's velar and sonorant like /w/.

    Why is this surprising? Why all the animosity towards the "weewer"-hearers in the last thread — who were behaving exactly the way English phonology predicts they should have?

  18. Mr Fnortner said,

    July 15, 2009 @ 6:37 pm

    I have cracked the code, and I am sure that no one gives a flying flip, but here's how to get the gh to sound like a w (or vice versa). This should satisfy just about everyone except the most constipated linguist purist. Imagine yourself a ventriloquist. Imagine that you must smile and not purse your lips to make a w sound. Imagine you must say "weewur" for your dummy. Imagine that you slip up and make the first w with your lips, but recover and make the second w with the back of your throat. You can do it if you try. If you can say the Spanish j or the German ch you can make this sound, only not harsh. Just clench and release while forming "wur." No "gur" allowed. Practice.

  19. Noetica said,

    July 15, 2009 @ 7:37 pm

    To sum up, for the standard American renderings of these three names, I recommend the unmodified pronunciations: SHIN-jahng, …

    What interests me is the American tendency to stress an early syllable. In Australia we would say "shin-JAHNG", and take "SHIN-jahng" as diagnostic of American. Compare our "Hong-KONG" versus "HONG-kong" (the first American pronunciation choice, according to M-W Collegiate).

    All of this is fascinating, but nothing in it is as noisome as the much-discussed non-affricate pronunciation /beɪˈʒɪŋ/.

  20. Lazar said,

    July 15, 2009 @ 8:29 pm

    @Noetica: As an American, I've always pronounced "Hong Kong" with the stress on the second syllable, and I'm surprised to see that dictionaries seem to list the other version first (although dictionary.com lists it with two primary stresses). That said, there are many words where American English prefers the early stress version, but at the same time, there are a lot of French-derived words (e.g. cafe, buffet, massage, garage, and French surnames in general) where American English prefers final stress and British English prefers initial stress.

  21. vanya said,

    July 15, 2009 @ 9:50 pm

    Dan,

    Animosity is overstating it. But I think the thread got a little more heated when one of the "weewer" supporters said that if you can hear a "g" of any sort then you must be "hallucinating it", which was a little ridiculous. I suppose people are very sensitive about being told their hearing is "defective." I'd still like to hear whether Uyghurs find Turkish pronunciation offputting, if anyone knows.

  22. Noetica said,

    July 15, 2009 @ 11:59 pm

    Lazar:

    … at the same time, there are a lot of French-derived words (e.g. cafe, buffet, massage, garage, and French surnames in general) where American English prefers final stress and British English prefers initial stress.

    O, certainly; the French phrase-final stress. American pronunciation of foreign words generally preserves original stress better than Commonwealth pronunciation. SOED's first pronunciation for oregano (from Spanish orégano) stresses the third syllable, and that is absolutely standard in Australia. M-W Collegiate gives it "correctly" on the second. This American tendency towards accurate preservation must sometimes compete with the tendency to push stress to the left. Cigarette might be a good case of this.

    American pronunciation of all sorts of foreign "o" is another matter; so is representation of these sounds in American dictionaries. (Don't get me started.)

  23. John Cowan said,

    July 16, 2009 @ 6:31 am

    It's true that English oregano is a direct import from Spanish, but the Spanish word is an obvious Latinism; in Italian it's origano, which is at least semi-popular in form. I have noted that those Americans who do say oregano with penultimate stress are Italian by descent.

    In any case, the name has lots of forms in older English: origanum, organum, origan, and even organ.

  24. Cameron said,

    July 16, 2009 @ 9:28 am

    What still baffles me is why the recommended "pronouncer" for the first syllable should be WEE- and not OOY-, or UY-, or even OY-

    Can Americans really not say OOY-ghoor? I'm sure they could if they tried.

  25. Franz Bebop said,

    July 16, 2009 @ 9:48 am

    @Boris: Are [ʊɪ̯] or [ʊj] really that hard for Americans to pronounce?

    In my opinion, yes, it's very difficult.

    …I just wouldn't have expected that forcing /j/ to the coda position, or pronouncing a diphthong similar to the already existing ones would be such a harsh violation of phonotactic rules, that it would be necessary to fall back to /wi/ or something similar (perhaps Oygur or Oygoor?).

    If those two vowels are adjacent, then my own instincts (for what they are worth) are to pronounce them as two syllables. If you want one syllable, then it has to be "oy" or "wee." It's like "boy" and "buoy" — my instinct is to pronounce "buoy" with two syllables, rhyming with "phooey."

  26. Ellen said,

    July 16, 2009 @ 10:03 am

    Cameron, regarding OY-, I would read that as like in boy, which I'm pretty sure isn't what anyone here is going for. And UY- doesn't make any sense at all as a "pronouncer". It doesn't clearly express a pronunciation. OOY- would work, I think. I think the WEE- would express a more americanized (or anglicized) pronunciation.

    As for the GH, there is no GH sound in contemporary American English. (And, for far as I know, contemporary English period.) If you want a g sound, just put -goor, no need for the H. If you want some sound that doesn't exist in American English, then your expectations are unrealistic, in my opinion.

  27. Franz Bebop said,

    July 16, 2009 @ 10:18 am

    @Cameron Can Americans really not say OOY-ghoor? I'm sure they could if they tried.

    Well, sure they can. Americans who have taught themselves Uyghur can pronounce the ɣ sound correctly, too. Americans who have learned Mandarin can also pronounce 维吾尔 without trouble.

    Remember, the question is not "what is correct Uyghur pronunciation" but rather "what is the anglicized pronunciation." I can pronounce the sound [ʊj] very easily. But when I am speaking English and I'm talking about Uyghurs, I wouldn't use the [ʊj] sound, instead I would say "WEE-gur" or "OY-gur."

  28. Noetica said,

    July 16, 2009 @ 10:37 am

    John Cowan:

    I have noted that those Americans who do say oregano with penultimate stress are Italian by descent

    Forgive my scepticism. Italian origano, like Latin origanum and Spanish orégano, is stressed on the second syllable. I have never heard any Italian pronounce it with the stress shifted from there; nor, by the way, any American.

    Those Italian–Americans you refer to, if they do shift the stress, might conceivably do it in order to distance themselves from their linguistic origins; as with Al Capone, who apparently was dangerous to be with if you pronounced his final "e". But that would be strange, because they would hear Americans all around them stressing the word as Italian does, anyway.

  29. J. W. Brewer said,

    July 16, 2009 @ 12:05 pm

    Since I had touted my own perhaps reactionary preference for the spelling "Sinkiang" on the other thread, I was especially pleased to learn from Prof. Mair that it was a reasonably phonetic spelling (for the right dialect, and subject to the skepticism expressed by J. R. Soong in his comment) at the time it came into use a century ago. But English of course rather notoriously tends not to update its orthography to reflect changes in pronunciation. So it seems particularly odd to expect us to change either the spelling or pronunciation of a toponym because the pronunciation or spelling conventions of the language it was originally borrowed from have changed in the interim.

    Even if one were to reluctantly acquiesce in the dominant market share now attained by Pinyin, however, Xinjiang is a good example of why Pinyin ought to be at best a starting point rather than a finishing point for arriving at a sensible English spelling/pronunciation of Chinese-derived toponyms etc. The word-initial x- does *not* fit the xerox/xylophone pattern, which is the only one extant in English for that location of the letter, and is more of a troublemaker than the word-initial j- in Jugoslavija, which we quite sensibly refused to carry over into English. For "Shinjiang," I might be willing to abandon Sinkiang, if I had some assurance that the English spelling and pronunciation was intended to remain stable regardless of future changes of pronunciation or political fashion in the PRC.

  30. Cameron said,

    July 16, 2009 @ 12:24 pm

    Franz Bebop wrote Remember, the question is not "what is correct Uyghur pronunciation" but rather "what is the anglicized pronunciation." I can pronounce the sound [ʊj] very easily. But when I am speaking English and I'm talking about Uyghurs, I wouldn't use the [ʊj] sound, instead I would say "WEE-gur" or "OY-gur."

    But why "WEE-"? Why introduce that initial consonant? If [ʊj] is tricky for phonetically challenged Americans, I'd venture that "EE-gur" would make more sense as an approximation of "ÖY-ghur" or "OOY-ghur" than anything beginning with "W-".

  31. Tim May said,

    July 16, 2009 @ 12:44 pm

    Regarding perceptibility of the /ɣ/, the other Tim above is right: I can hear the fricative much more distinctly through headphones than through my speakers. So the tendency for some readers to hear it as [w] may depend on their audio equipment as well as on their experience in phonetics.

    Still, it is very soft, & I might mistake it for [w] if I didn't know what I was listening for. Certainly it sounds more [w]-like than the sample [ɣ]s here and here, which are more what I expect a [ɣ] to sound like.

  32. Victor Mair said,

    July 16, 2009 @ 12:49 pm

    The French romanization of Uyghur is Ouïgour, which looks like a reasonable way to make a French approximation of the name ([uigu:R] or [uiguR]), but how would a French speaker actually pronounce that? Is the "ouï-" at the beginning significantly different from "oui" ("yes"), which I pronounce as WE or WEE (I'm not sure if that's the exactly correct pronunciation for that ubiquitous French word, but that's how it sounds to Anglophone me).

    This "ouï-" also occurs in the native French words "ouïe" ("[sense of] hearing"), "ouïr" ("hear"), and "ouï-dire" ("hearsay").

    Would a French speaker please enlighten me as to whether there is any difference between the pronunciation of "ouï-" and that of "oui"?

    Finally, to mollify all of those who fall into the WEE-wur camp, as Roderick has pointed out on the other Uyghur thread, the name is transcribed as WEI2WU2'ER3 in Mandarin, so that puts Mandarin speakers in the same camp as those who believe that they hear WEE-wur. But that's not what I hear.

  33. Cameron said,

    July 16, 2009 @ 1:04 pm

    @Victor Mair: Would a French speaker please enlighten me as to whether there is any difference between the pronunciation of "ouï-" and that of "oui"?

    "ouï-" would be pronounced differently from the word "oui". "ouï-" would come out as OO-EE, with the two vowels quite distinctly pronounced.

  34. Bloix said,

    July 16, 2009 @ 1:07 pm

    Victor- is the gh in Uyghur the same as the gh in yoghurt? The French for yoghurt is yaourt, which has no intermediate consonant at all.

  35. Nick Z said,

    July 16, 2009 @ 3:20 pm

    @John Cowan, @Noetica. For what it's worth, British English pronounces oregano with penultimate stress.

    @cameron. Surely it's better to have roughly the right segments, in the 'explainer' for Uyghur, and thus start with WEE-, than to lose half the initial diphthong with EE-? Even if that means that you end up with an initial rising diphthong instead of a falling.

  36. Ken Brown said,

    July 16, 2009 @ 3:45 pm

    Dan Velleman said: "Why is this surprising? Why all the animosity towards the "weewer"-hearers in the last thread — who were behaving exactly the way English phonology predicts they should have?"

    I don't think its animosity, though it might be surprise. Maybe what its showing us is how different our Englishes are from each other. That I (& apparently about 3/4 of the other posters) hear a "gh" but some people hear a "w" might say something about the toolkit of sounds we use, or are used to hearing, in English.

    Also, although the English "gh" is all but obsolete in modern English its not something we never hear. (Not that it is the the same as the middle consonant in "Uyghur" any more than "w" is of course)

    /x/ still exists in some Scots accents (and possibly a few northern English ones) in words like "night" and "bright". It is used in English in Scottish (& maybe Irish) Gaelic place-names, and the word "loch" (complete with voiceless velar fricative) is productive in English as spoken in Scotland – you can say "look at that loch over there" even if "loch" is not part of the place-name of said body of water.

    And there are similar sounds in some Northern English accents (especially Scouse) where standard English has /k/ (Imagine someone from Liverpool saying "cookbook")

    Maybe those sounds are much more familiar in UK than in the USA so we're more likely to hear them for "gh" than Americans are. Or maybe that's a just-so story I just made up. I'm not sure!

  37. Lazar said,

    July 16, 2009 @ 4:47 pm

    @JW Brewer: I think the difference between "Sinkiang" and "Xinjiang" is merely the result of choosing one or the other schemes of romanization (Wade-Giles or hanyu pinyin), not of any phonological change in Mandarin Chinese.

  38. dr pepper said,

    July 16, 2009 @ 5:06 pm

    I pronouce "buoy" with a slightly elongated "oy", not quite a dipthong but definitely different from "boy". I'm not sure where, but that's the way i heard it somewhere in early childhood. Most of the people i hear saying that word now, either say "boy" or "booey" Of course, as i don't go onto water or live that close to it, i rarely hear the word at all.

    Otoh, i'm heaing "uighur" a lot lately. Most of the commentators say "weeger", so that's what i've been saying too. I can pronounce the "ch" as in "loch" or "ich", but i'm sure that in my case at least, saying "weecher" wouldn't make me sound any closer to the native pronunciation. As for the "gh" as in "Van Gogh", i just end up gargling. This journal has really made me realize how really limited my phonemic repertoire is.

  39. Vijay John said,

    July 16, 2009 @ 7:37 pm

    @Lazar: it really is not a simple matter of romanization scheme. Note that in Wade-Giles, Xinjiang would be Hsinchiang. "Sinkiang" is Postal Map Romanization, which is partly based on Wade-Giles but also involves some additional (phonological) complications, as mentioned above.

  40. Bronwyn said,

    July 17, 2009 @ 11:21 am

    @Cameron Can Americans really not say OOY-ghoor? I'm sure they could if they tried.

    Another person chiming in, this time with anecdotal evidence.

    I did some work on the Uyghur language this spring, and learned the English pronounciation as having an initial diphthong [uj]. Since then I've talked to many friends and family about Uyghur, and almost all of them have said some variant of: "Sorry, what's the language called?" Then, when I repeat it, and they try to say it themselves, they all say something like "wee" for the first syllable.

    I was really surprised by this until I realized that [uj] doesn't really exist in English. There's buoy (mentioned above by Franz), but most people I know have trouble saying [buj], and instead pronounce buoy exactly like boy.

  41. Victor Mair said,

    July 17, 2009 @ 1:53 pm

    From Juhan Janhunen:

    I have also been following with some interest the evolution of the English usage, which earlier used to have "an Uighur" implying /ui/ (as in ruin), while today "a Uighur" seems to be more common implying /wii/ (as in weed). It may be that the latter pronunciation is actually better suited for English phonotactics. Initial /ui/ seems to be problematic or non-existent.

    There is a similar vacillation between "a Uralic (language)" and "an Uralic (language)", implying that some English speakers pronounce the name "Urals/Uralic" with an initial vowel (as in ooze). There seems, again, to be a phonotactic problem with short /u/ in initial position in most forms of English, but /uu/ and /yuu/ are fine.

    As far "Uighur" is concerned, there is also the vacillation between -g- and -gh-, as well as -i- and -y-. Many Turkologists like to write the name (in English) as "Uyghur". This corresponds to the phonotactics of Uighur itself, where /y/ is a consonantal sound, but I myself have always preferred the spelling "Uighur".

  42. Bloix said,

    July 17, 2009 @ 2:02 pm

    Any English speaker can say Huey, Dewey and Louie. Is that the initial sound that "Uy" is supposed to represent?

  43. Victor Mair said,

    July 17, 2009 @ 3:30 pm

    A letter sent to Radio Free Asia by a friend who wishes to remain anonymous. It's about the rendering of the name of the most famous Uyghur on the world stage before Rebiya Kadeer (AKA Rabiyä Qadir, Rabiya Kadir, Re4bi3ya3 Ka3de2'er3, etc.), Örkesh Dölet (AKA Wu'er Kaixi).

    ====

    In the article (I cite just one representative case) "1989 Leader Slams Crackdown" (http://www.rfa.org/english/news/china/wuerkaixi-07072009144343.html) there is a gentleman referred to as 'Wu'er Kaixi' and thereafter as 'Wu'er' — as if 'Wu'er' were his surname, and 'Kaixi' were his given name — both names being Han Chinese of course.

    This is much as if in an article appearing in English, Bill Clinton were referred to as 'Ke Lindun' and Mr. 'Ke'.

    The individual in question however is no Han but Uyghur, and his name is Örkesh Dölet. Perhaps you can tell me why the media (or the non-Turkic media anyway) continues to this day to refer to him by a name transliterated into Chinese characters and then transcribed back into Hanyu Pinyin?

    In any case, even as a transcription of 'Örkesh', it ought to be a single unit — 'Wu'erkaixi', not 'Wu'er Kaixi'. Is this not so?

    [END OF LETTER TO RFA]
    ———————————————————————————————————————————————————————–

    Naturally in their pages in Uyghur, RFA gets it right; here's one in latinisation:
    http://www.rfa.org/uyghur/xewerler/qisqa_xewerler/orkesh-xitaygha-barmaqchi-04042009081813.html?encoding=latin

    I rather suspect that he prefers his name being rendered, in English, etc, as 'Wu'erkaixi' ( or even 'Wu'er Kaixi'), since that's how he got famous.

    ====

    It is interesting that "Wu'erkaixi" is content to be known internationally by the Pinyin romanization of his given name, not bothering with his Uyghur name Örkesh, much less his surname Dölet.

  44. Simon said,

    July 18, 2009 @ 2:11 pm

    I personally think that the the velar fricative is rounded, based on the voice samples here.

    If I'm not mistaken, it's quite common for the voiced velar fricative to be accompanied by lip rounding in world languages, so those hearing the velar fricative and those hearing the bilabial semivowel are likely both (semi) right in this case. There may be an allophonic variation with the lip rounding as well.

    The loss of the velar feature explains why there are so many diachronic instances in world languages of the voiced velar fricative becoming a bilabial semivowel. For instance, English "tallow" and German "Talg" come from the same source, where English has undergone the same develarization and German has not, and modern Danish has seen the voiced velar fricative die out entirely in just the past generation (ca. 50 years), becoming semivowels ([j], [w]) or zero, depending on context, in the speech of all but the very oldest living Danes.

    Anyway, those hearing "WEE-woor" are probably missing the velar feature, and those hearing "WEE-ghoor" may not be noticing the co-occurring lip rounding.

  45. Simon said,

    July 18, 2009 @ 2:15 pm

    Also, as for [uj] in English, which doesn't occur, I think "buoy" isn't pronounced as [buj] (one syllable) but as [bu.i] (two syllables) in most native speakers' speech; when people do pronounce it with one syllable, they say [boj] since [uj] doesn't occur.

  46. abd_ul_waajid said,

    July 18, 2009 @ 4:18 pm

    Listening to the audio sample for "Uighur" back in http://languagelog.ldc.upenn.edu/nll/?p=1576 , I can hear the "gh" (perhaps because I knew beforehand to listen to it), but it sure does sound like [wiɣur] (or [wiʁur]) to me — or even [weɣur] — rather than [ʊjɣur]. That is, it sounds like the vowel is a front unrounded while the glide is a back rounded. Is it possible that the conventional IPA transcription is influenced by the Uyghur-Arabic spelling and corresponding transliterations, moreso than the actual renderings of the words by Uyghur speakers? Alternatively, are we completely sure that the speaker recorded in the first piece is a native Uyghur speaker with a fairly standard accent? Her pronunciation of "uy" does sound a lot like what I would expect to hear from a Chinese speaker influenced by the Mandarin pronunciation of 维.

    Based on her pronunciation, I guess I'm going to have to stop trying to say [ʊjɣur] when talking to people in English, and just give in and use [wigɚ].

  47. abd_ul_waajid said,

    July 18, 2009 @ 4:21 pm

    I have [boj] for "buoy", and I regard [bu.i] as an acceptable variant. I wondered about the variable pronunciation since childhood, but now it makes sense that both are adaptations to an underlying [buj], which is impossible in English phonotactics. I would also like to point out that the common American English prounciation of the word "muy" (the sort of smattering-of-Spanish word that many Americans know) is two syllables: [mu.i]. I surprised to later learn that it is supposed to be one syllable in Spanish.

    Bloix,

    "Huey", "Louie", and "Dewey" are each two syllables in English; at least, they are when I say them: [j̥u.i], [lu.i], [du.i] (or perhaps [uwi] instead of simply [u.i]). If you tried to cram both vowels into one syllable, you would probably come up with something pretty close to [uj].

  48. abd_ul_waajid said,

    July 18, 2009 @ 5:20 pm

    I think it would be nice if everyone would standardise to using Uyghur Latin Yéziqi for Uyghur transliteration. Its design goals (see http://www.uyghurdictionary.org/excerpts/An Introduction to LSU.pdf) seem quite reasonable, among them "Close correspondence of the chosen Latin letters with their common international phonetic value". Not that it's necessarily "perfect" — using "e" for [æ] is a bit funny, but apparently [æ] is a much more common sound than [e] in Uyghur, so it made sense to use a letter with no diacritic for the former — but the results generally seem workable. "Rebiya Kadeer" (Chinese pinyin) would be "Rabiye Qadir". "Ürümqi" (Uyghur pinyin) would be "Ürümchi". ULY was developed inside the PRC, but apparently without a lot of governmnet interference, so its use should be seen as more or less politically neutral.

  49. Victor Mair said,

    July 18, 2009 @ 8:46 pm

    Bloix asks a good question about yoghurt that is directly relevant to our discussion of how to pronounce "Uyghur / Uygur / Uighur / Uigur," since "yoghurt / yogurt" is also a Turkish word and also has a -g- or -gh- in the middle. Although there is ambivalence whether the medial is -g- or -gh-, the spelling with -g- is nearly four times more popular than that with -gh-. Much less common than yogurt or yoghurt is yoghourt.

    The pronunciations of the first two spellings may be heard here: http://media.merriam-webster.com/dictionary/yogurt They sound identical to me.

    It is interesting to trace (following the OED) the different spellings of the word since it entered the English vocabulary: 1625 Yoghurd; 1687 Yogourt; 1837 yahourt; 1883 yaghourt; 1912 yoghourt; 1925 Yoghurt; 1934 yoghourt; 1955 yoghourt; 1970 yogurt; 1980 yoghourt.

    Here I shall quote a couple of online reference works concerning the etymology and pronunciation of yogurt / yoghurt and then draw some conclusions based on them and on the information adduced above:

    ====

    Online Etymological Dictionary: a mispronunciation of Turk. yogurt, in which the -g- is a "soft" sound, in many dialects closer to an Eng. "w." The root yog means roughly "to condense" and is related to yogun "intense," yogush "liquify" (of water vapor), yogur "knead."

    ==

    Wikipedia:

    Etymology and spelling

    The word is derived from Turkish yoğurt, and is related to yoğurmak 'to knead' and yoğun "dense" or "thick". The letter ğ was traditionally rendered as "gh" in transliterations of Turkish, which used to be written in a variant of the Arabic alphabet until the introduction of the Latin alphabet in 1928. In older Turkish the letter denoted a voiced velar fricative /ɣ/, but this sound is elided between back vowels in modern Turkish, in which the word is pronounced [joˈuɾt]. Some eastern dialects retain the consonant in this position, and Turks in the Balkans pronounce the word with a hard /ɡ/.

    In English, there are several variations of the spelling of the word. In the United States, "yogurt'" is the usual spelling and "yoghurt" a minor variant. In the United Kingdom "yoghurt" and "yogurt" are both current, "yoghurt" being more common, and "yoghourt" is an uncommon alternative. Canadian brands typically use "yogourt" as it is correct in both official languages [English and French], however "yogurt" is used as well and is common among English speakers; in Australia and New Zealand "yoghurt" prevails.

    Whatever the spelling, the word is usually pronounced with a short o (/ˈjɒɡərt/) in the UK, with a long o (/ˈjoʊɡərt/) in North America, Ireland, Australia and South Africa, and with either a long or short o in New Zealand.

    ====

    The impression one gains from the above data is that the medial consonant of the original Turkish word for yogurt / yoghurt is and was pronounced differently in various dialects and at various times, sometimes being elided altogether. Likewise, English speakers at various places and at various times interpreted differently what they heard Turkish speakers saying. Eventually, however, the pronunciation of the medial consonant of yogurt / yoghurt in English has become standardized as /g/.

    Everything that I have just written about yogurt / yoghurt in the preceding paragraph applies, mutatis mutandis, to Uighur / Uyghur / Uigur / Uygur. I believe that the entire drift of this thread and the entire drift of the other Uyghur thread (http://languagelog.ldc.upenn.edu/nll/?p=1576) also support these conclusions.

  50. John Cowan said,

    July 19, 2009 @ 2:40 am

    Well, Noetica, I can only suppose that the pronunciation with penultimate stress reflects either a variety of Italian other than the standard, or else reflects what anglophones of Italian descent have reconstructed as the "proper" pronunciation of a language their parents or grandparents ceased to speak. English is, after all, a penultimate-stress language by default, albeit with a huge number of lexical exceptions.

  51. Noetica said,

    July 19, 2009 @ 7:41 am

    Yes, John Cowan. Who knows the detailed aetiology of some of these pronunciations?

    As for default stressing in English, the opposing forces mentioned earlier are well and truly in place – an American push to the "left", versus an equally American tendency to respect foreign stress, even when these push to the "right". Increasingly these affect other Englishes. Certainly American stresses infiltrate the speech of young Australians. We'll check the situation in another ten years, all right?

  52. Dan T. said,

    July 22, 2009 @ 2:06 pm

    The spelling "yogurt" makes more sense to me, since there isn't any "hurt" in it!

    It's interesting that this spelling apparently only goes back to 1970 or so; the earliest I remember hearing of yogurt is in the early '70s as a kid, and in that time it still had some airs of exoticness as an unusual imported food; it wasn't as commonplace as it is now.

RSS feed for comments on this post