Nondifferentiation of -n and -ng

« previous post | next post »

In Shanghai, Tom Mazanec recently came across a listing for a kind of tea called Tiě Guāngyīn 铁光阴 (second from the bottom in the photo), which he thought might be a knockoff of the famous Tiě Guānyīn 铁观音. The picture was taken at a restaurant near Fudan University called Xiǎo Dōngběi 小东北 (the name of the restaurant [Xiǎo Dōngběi sīfang cài 小东北私房菜, at the top of the menu] is rather endearingly translated as "The small northeastern dishes").

As explained in Wikipedia:

The tea is named after the Chinese Goddess of Mercy Guanyin, who is known in Japan as Kannon and in Korea as Guam-eum. Guanyin is a female embodiment of Avalokiteśvara Bodhisattva. Other spellings and names include "Ti Kuan Yin," "Tit Kwun Yum," "Ti Kwan Yin," "Iron Buddha," "Iron Goddess Oolong," and "Tea of the Iron Bodhisattva." It is also known in the abbreviated form as "TGY."

None of her names has a -g at the end of the first syllable.

Guānyīn 观音 is commonly understood to mean "hearing the sounds (i.e., cries) of those who are suffering", whereas guāngyīn 光阴 means "time" (not a specific time, but the idea of the passage of time), so there's no semantic overlap between guānyīn 观音 and guāngyīn 光阴.  The confusion between the two must be purely phonological.  Except for the final -g of the first syllable of the second rendering, phonologically the two versions are identical, right down to the tones.

So here's what happened.  Many speakers of Mandarin do not distinguish between -n and -ng.  We've had it happen right here on Language Log recently; see the beginning of this comment.

What happened in this case is that the person who wrote the menu was thinking guānyīn 观音 but pronouncing guāngyīn.  Consequently, to match the sound in their head, they wrote guāngyīn 光阴 instead of guānyīn 观音.  This phenomenon of writing words with the wrong characters because of topolectal pronunciation differences has existed for as long as the script has existed.  It is a verification of the primacy of speech over writing.

Some of my students from Taiwan and the mainland, including those who have graduated from the best schools, routinely mix up -n and -ng.  It's a very common error among Mandarin speakers of diverse backgrounds.  I even know excellent teachers of Mandarin who occasionally mix up these two endings.  Usually I don't want to embarrass them by pointing out the confusion, but in very special circumstances when I do mention it, they can't tell the difference.


  1. K. Chang said,

    May 25, 2015 @ 11:11 pm

    But that's almost like them speaking Mandarin with a bad accent!

    Reminded me when I got asked a question "why does this girl's 'zero' sounded so different from all the other examples I heard?"

    Then I had to explain that this girl has a heavy regional accent and is hardly speaking "standard" Mandarin.

    It'd be much like learning "American English" by listening to Brooklyn accent.

  2. John Rohsenow said,

    May 25, 2015 @ 11:39 pm

    When I was a TA teaching ESL at the U of Michigan about 100 yrs ago, and just starting Chinese, I had a student from Taiwan who could not differentiate between final -n and -ng in English. I rapidly ascertained that she also did not make the distinction in her Mandarin either. I finally asked her to pronounce the two symbols for these sounds in the National Phonetic Alphabet "Zhuyin Fuhao", which she easily did. It turned out that the only time she made the distinction was in pronouncing these two symbols in the NPA and nowhere else. She thereafter used this to correct
    her pronunciation in English, but as near as I could tell, it didn't affect her Mandarin pronunciation at all.

  3. K. Chang said,

    May 26, 2015 @ 4:12 am

    @Victor Mair — I have perfect Mandarin pronunciation. I just suck at transliterating / pinyin.

  4. Edward Lindon said,

    May 26, 2015 @ 7:07 am

    "It is a verification of the primacy of speech over writing." Isn't that a stretch? At most, isn't it a verification of the primacy of speech over writing for *some individuals*, namely those whose writing exhibit such mistakes, which makes this "verification" no more than a sign for itself.

    As for the reconstruction of the writing process, couldn't the author simply have mistyped the pinyin (which I often do, regardless of whether I am being misled by my lousy pronunciation) and then failed to catch the typo at the proofing stage (because in all likelihood there wasn't one)?

  5. joanne salton said,

    May 26, 2015 @ 8:15 am

    I find Chinese students have this n/ng problem to the extent that they will fairly commonly make mistakes when writing out their own name in pinyin. That would tend to suggest that it is more of a writing issue than a topolect issue – but as to why it is so difficult to write the correct one I am not sure.

  6. languagehat said,

    May 26, 2015 @ 9:29 am

    rather endearingly translated as "The small northeastern dishes"

    Actually, "The small northeast dishes."

    /professional copyeditor

  7. Victor Mair said,

    May 26, 2015 @ 11:13 am

    @John Rohsenow

    Brilliant demonstration of how someone may be physically capable of making the -n / -ng distinction but that it is absent in her brand of Mandarin! Thanks!

    I've mentioned this before on Language Log, but I'll repeat it for those who may not have been around when I did so. Namely, I've met plenty of people who have terrible Mandarin pronunciation, people that both of us know (e.g., Apollo Wu [Cantonese accent], Yin Binyong [Sichuan accent], Zhou Youguang [Wu accent]), but who have perfect Pinyin. Indeed, Zhou was the inventor of Pinyin, and Yin Binyong did more than anyone to establish the orthography of Pinyin as a working script (one half of a digraphia).

    Very few people speak "perfect Mandarin", and what is "perfect Mandarin" anyway? There is almost always local (topolectal) and personal (idiolectal) variation.

    Then there's the famous story about Y. R. Chao, one of the greatest Chinese linguists of the 20th century, being "The world's only speaker of standard Mandarin in 1923".

    From Steven Owyoung, a specialist on the history of tea:


    Taken literally, the notion of guangyin 光阴 is rather fascinating.
    Since learning Mandarin, I have noticed that some of my Chinese teachers, friends, and colleagues often pronounced lin as ling. I have also heard some pronounce wei as vei, substituting the w with a v. Does that strike you as odd? Interestingly, the same thing happened in the Hawai'ian language as well: v as w and thus Havai'i.


    The "W" at the beginning of my given name in Chinese is the result of this substitution: Méi Wéihéng梅维恒. The teacher who gave it to me pronounced "wéi" as "véi", meaning for it to stand for "Victor". I think she was a Wu speaker, or from somewhere near the Yangtze estuary. She didn't realize she was saying "véi" when she pronounced 维. She thought she was speaking it in the standard way, and she romanized it as "wéi".

    A couple of years ago in my "Language, Script, and Society in China" class, I did an experiment with all the native speakers (about 20 students), and it turned out that, for 英 about half of them said "yīng" and about half said "yīong". Those who said "yiong" were from Beijing and the northeast, and they insisted that "yīong" was standard. Of course, who wouldn't think that Beijing speech — which is supposed to be the basis for Putonghua (Modern Standard Mandarin [MSM]) — is standard?

    Here are a couple of comments from a student who was in that class and from one who was not.

    From Wei Shao, who is from Liaoning province in the northeast:


    That yīong pronunciation seems so weird to my husband [VHM: he's a Minnan speaker] (or any other friends from the south) that he laughed out loud the first time he heard it. But it is pretty common in the north in daily communication. I suppose Beijing people in general consider their accent to be "standard," because that's how Mandarin Chinese is defined. However, I have also noticed that when it comes to TV news and broadcast, where the authentic standard Mandarin is used, yīng is what I usually hear. To say yīong in a formal occasion, for example, if you were a master of ceremony, is considered a little "土." That's how I understand the difference of yīong and yīng.


    From Jing Wen, a real Beijinger, born and bred:


    I would pronounce it [ ji:ŋ ] with a the velar nasal. I think people in Beijing and Northeast pronounce ying with an emphasis on the velar nasal (like the ng at the end of a word in French and German), which sounds like yiong. In some parts of southern China, people have difficulty pronouncing the velar nasal. For them, ying (英) and yin (阴) have the same pronunciation. I know this because some of my college classmates from the south cannot pronounce my name (jing) correctly. They simply call me jin.


    Comment by Jiahong Yuan, a phonetician with LDC (Linguistic Data Consortium) who comes from the Northeast (Dongbei [Dalian, if I recall correctly]):


    That was an interesting find. I must also have heard of the pronunciation quite often because it makes sense to me from your description. My observation/impression is as follows:
    1. It is generally viewed that in 汉语拼音方案 /ong/ is 合口呼 (ipa [uŋ]) and /iong/ is 撮口呼 (ipa [yŋ]). But I don't quite agree with these transcriptions. I think there is an /o/ as the main vowel in the rimes.
    2 If you compare the finals in 英(yīong) and 雄 (xiong),they seem to the same rime but the "介音“ is different.
    3. So my transcription would be: /ong/ -> [oŋ], /iong/ -> [yoŋ], /yīong/ -> [ioŋ].
    4. Another interesting sound is "翁“. It is paired with "风“ in 汉语拼音方案 (ipa [uəŋ]), but it's often pronounced as more rounded, something like [uoŋ], paired with "东".
    5. If all the above are correct, then we have [oŋ] (/ong/), [yoŋ] (/iong/), [uoŋ] (sometimes for "翁“), and [ioŋ] ("yīong", sometimes for "英“). Then the position of /ong/ in 汉语拼音方案 should be 开口呼, which may or may not make sense.

    In short, I'd transcribe "yīong" as [ioŋ] and "iong" as [yoŋ].

    For those who are interested in variety within Mandarin and within Sinitic more generally, here are a few relevant Language Log posts:

    "'Chinese' well beyond Mandarin" (5/10/13)

    "Only 7% of people in China speak proper Putonghua: PRC MOE" (9/24/14)

    "Pushing Pekingese" (10/10/13)

    "Pekingese vs. Putonghua"(3/15/15) (includes many links to related posts)

  8. Victor Mair said,

    May 26, 2015 @ 11:20 am


    That's why we need copy editors.

  9. Mark Dunan said,

    May 26, 2015 @ 12:23 pm

    "It'd be much like learning "American English" by listening to Brooklyn accent."

    In which case, you would of course be learning the best and most beautiful form of English. ^_^; (Yes, I am Brooklyn-born.)

    As a resident of Japan, this discussion of 観音 reminds me of what has happened with this word in Japanese. It was adopted into the language as Kwannon くゎんのん over a millennium ago, with the n in kwan pushing itself onto the next vowel, which once was normal (creating words like 天皇 tennou and 因縁 innen). With the w dropping out about 150-200 years ago, it is now pronounced Kannon.

    But since Japanese people don't really think about that historical change, and because katakana are syllabic, when writing ruby characters next to the word to show how to pronounce it, you typically see かん (kan) with 観 and のん (non) with 音, leading people to falsely think that the kanji 音 (normally in or on) can also be read non by itself, and so you have parents naming their daughters and trying to pretend that 花音 and 華音 can be read as "Kanon".

    Bad enough that historical sound change defiled the first half of the Bodhisattva Kuan-yin's name. Now my generation is using their misreading of it to give their kids names that make purists sputter.

  10. Rachel said,

    May 26, 2015 @ 12:26 pm

    When I was teaching EFL in Taiwan, I'd often do a unit on romanization. A lot of my students would have strong disagreements on how to transliterate 新生, a fairly major road in Táiběi. Some students would say it was Xingsheng, some that it was Xinshen, and many would say that they couldn't hear a difference. I usually had to resort to my poor imitation of Běijīng accent to differentiate the two for students. They could then hear the difference, but were still often uncertain when thinking of their own pronunciation. (Also worth noting, I think this shows how little the students were relying on MPS and how much they were relying on their own pronunciation, because the difference between -n and -ng is pretty clear in MPS.) (The lookup tables I gave the students for Hànyǔ Pīnyīn used MPS as the source, for ease of reference.)

  11. Guy said,

    May 26, 2015 @ 12:38 pm

    I used to have a Japanese-Hawaiian-American friend in Taiwan (fluent in both English and Japanese) and this pretty much drove him crazy.

    He couldn't understand why so many native Mandarin speakers in Taiwan seemed incapable of distinguishing between n/ng when he, a non-native speaker, could easily tell the difference.

    Plus not only do we not pronounce it properly, but a lot of people literally can't hear it either. Does anyone have a reason why native speakers can't hear it?? I can hear the difference myself, but I grew up bilingual, so I'm probably not a good example.

    Funnily enough, it doesn't seem to affect people's ability to input correctly in bopomofo very much.

  12. JS said,

    May 26, 2015 @ 1:03 pm

    The colloquial northern pronunciation of PY -ing is clearly a centering diphthong, but I might write (IPA) [iɨŋ] or [iəŋ] rather than using /o/. More extreme examples do seem to have a labial gesture, but I'm not aware of mixture of the type (PY) yingyong, xingxiong (yet).

  13. Victor Mair said,

    May 26, 2015 @ 2:12 pm

    In a personal note to me, JS mentioned that "the pronunciation you describe applies to the whole PY -ing rhyme."

  14. Victor Mair said,

    May 26, 2015 @ 2:20 pm

    Rebecca Fu, from Harbin:


    I guess "yeng"?


    Notes from San Duanmu:


    Besides dialectal differences, pinyin input could be a source of error, too. For example, if you use abbreviated input, you can type 'gy' (first letter of each syllable for 'guan-yin') and make a bad choice among the suggested words.

    In Beijing, the IPA transcription of ying 英 is [jəŋ], where the main vowel is not [i] but [ə]. The syllable also rimes with deng 灯 [təŋ]. In contrast, 音 [in] and 根 [ən] are not a good riming pair.

    So it is better to spell 英 as yeng or yieng, but not yiong, because it does not rime with gong.


  15. shubert said,

    May 26, 2015 @ 2:32 pm

    I was not sure about 音 vs.影, but I have found a way to explain now.

  16. P'i-kou said,

    May 26, 2015 @ 3:31 pm


    Sounds [iəŋ] to me too (and that agrees with informal descriptions of 'pinyin' -ieng). Discussion e.g. here.

    Mixing or collapsing the nasal codas is especially common in areas where the local topolect has only one possible nasal ending (e.g. in the Wu-speaking region), but pronouncing specifically -in as -ing seems to be common even among speakers who still distinguish, say, -en from -eng (I suppose it helps that in such cases the quality of the vowel is typically also different). This seems to be especially common in Taiwan.

    A prominent Taiwanese businessman whose name in pinyin would be Cài Yǎnmíng 蔡衍明 is known as Tsai Eng-Meng instead. I don't know if there's any non-Mandarin motivation for that (nor do I know any Minnan but cursory googling suggests 衍 ends in an alveolar there as well).

    At the very least, mispinyinning the name of Tianjin is (anecdotally) not uncommon among residents of that fair city. As far as I know, Tianjin hua does differentiate the two nasals just fine.

    Two recent Tianjings on domains.

    Turning to wild speculation, some topolects with separate -n and -ng could perhaps have them in a different distribution than in Mandarin. MSM has a few examples of 'irregular' -n from Middle Chinese -ng (with a regular -ng elsewhere in Sinitic/Sinoxenic).

    The following are the most common examples of Mandarin irregular -n (in pinyin, Middle Chinese (Baxter transcription; X means rising, H departing tone) and Sino-Korean).

    kěn MC khongXgeung
    pìn MC pjiengHbing
    yùn MC yingHing
    zhēn MC trjengjeong
    xīn MC sjenghyeong

    Of these, all except 孕 seem to end in -ng in Cantonese as well, and at least 肯, 聘 and 馨 have Sino-Japanese readings that also correspond to -ng.

  17. K. Chang said,

    May 26, 2015 @ 4:10 pm

    Now try to romanize 阴影 dark shadow

  18. Victor Mair said,

    May 26, 2015 @ 4:48 pm

    For an amazingly clear demonstration of the pronunciation of ying (here in the third tone) that I was talking about above, click on the speaker button near the top left of this page: (that's for 阴影 / 陰影)

  19. P'i-kou said,

    May 26, 2015 @ 6:13 pm

    At the risk of falling off topic – the menu contains another example of -n/-ng fluidity, namely 听 tīng 'tin/can'. The word is a borrowing from English tin, possibly along a Yangjingbang/Pidgin English route. (The language most relevant to such an exchange is Shanghainese, a Wu topolect that merges the sources of Mandarin -in and -ing, but a direct borrowing into Mandarin might have given the same result as there's no syllable *tin.)

  20. Bespectacled Jewish Scholar said,

    May 26, 2015 @ 7:42 pm

    Yup – the phenomneon finds its way even into 红楼梦 – the 秦 (qin2) in 秦可卿 makes more sense as a pun for 情 (qing2) to southern speakers who can't differentiate between "n" and "ng"

  21. Victor Mair said,

    May 26, 2015 @ 9:16 pm

    From a graduate student who grew up in Beijing, but whose parents are from elsewhere to the south:

    Regarding the pronunciation of 英, I feel people who said "yiong" don't hear the difference between "ying" and "yiong." Most of them are probably from the north, but it may not hold true for all northerners.

    Personally I think "yiong" sounds unpleasantly casual (吊兒郎當)–I rarely hear it among my middle/high school classmates except the few whose families have lived in Beijing for generations.

    (My parents dislike northerners, so I've been instilled since little that Beijing topolect sounds uncultured and vulgar–so my view may be prejudiced.)

  22. Jongseong Park said,

    May 27, 2015 @ 1:05 am

    I might mention the late Lee Kuan Yew 李光耀. The character 光 which corresponds to "Kuan" is supposed to be pronounced with -ng in all of the major Sinitic varieties—guāng in Mandarin, gwōng in Cantonese, kong in Hokkien, kuang/keung/kng in Teochew, and kông in Hakka (I've mixed up several different romanization traditions so "g" and "k" both represent voiceless unaspirated [k(ʷ)] in all these instances). His parents or whoever gave the spelling "Kuan" for 光 probably did not consistently distinguish -ng and -n.

  23. Tom said,

    May 27, 2015 @ 3:20 am

    Interesting suggestion by San Duanmu that it could be the result of typing the abbreviation "gy." However, when I try that out on my computer, I find gōngyuán 公园, gǎnyìng 感应, guānyú 关于, and gōngyè 工业 before I even get to Guānyīn 观音, much less guāngyīn 光阴.

    I also noticed the use of tīng 听 for tin and was wondering about its history.

  24. Victor Mair said,

    May 27, 2015 @ 8:43 am

    I should mention that, when I did that experiment with having the students in my "Language, Script, and Society in China" course pronounce yīng 英, I ran through the whole class fairly quickly, one student at a time, one row after another, from the front of the room to the back. No one thought anything was amiss or unusual. I had even written "yīng 英" on the board.

    When I reached the last student, I said to the class, "Did you notice anything remarkable about the way you pronounced "yīng 英"? No, nobody noticed anything worth remarking on about the way they pronounced "yīng 英". They all thought that they were saying "yīng 英". I told them, "Some of you were saying 'yīng' and some were saying 'yīong'."

    They didn't believe me, and at that time, I didn't know what the regional or topolectal breakdown for the distinction was, but the difference in pronunciation was stark to me.

    Then I went through the whole class again, and had each of them pronounce "yīng 英" once more, and as they did so, I pointed out into which category they fell. Those who said "yīong" couldn't believe that they were saying "yīong", and even most of the other students in the class, the ones who said "yīng", had a hard time telling the difference.

    Finally, I went through the class yet again, much more slowly, and with careful analysis of each student's pronunciation. As I did so, I asked them where they were from in China. I still remember clearly the surprised looks on the faces of some of the students who said "yīong", when they realized that's what they were really saying. It was like a light bulb was going off in their head. The reaction of one girl in particular, He Ziwei, a true Beijinger, stands out. I had her say the syllable several times, and she just couldn't believe that she was saying anything different from "yīng". When it finally dawned on her — with the help of several classmates — that she was saying "yīong", not "yīng 英", she giggled and put her hand over her mouth bashfully.

    BTW, I had a particular affection for He Zewei because of her charming English. I'm often amazed by how wonderful the English of students coming from China is in recent years. Mostly they have American accents, but sometimes I detect unusual blends of British, Australian, or other features. When I asked them how they got to be so good and why they have these mixtures of traits, I found out that it has more to do with the films and shows they watched and the performers they listened to than to particular teachers they had, though in some cases where they were fortunate enough to have a native speaking teacher from abroad, that could have a decided influence on their pronunciation. But mostly they develop idiosyncratic mixtures of characteristics, and I find almost all of them genuinely appealing. I never cease to marvel at the incredible progress in English language ability that has been made by students from China during the last decade.

    Back to the "yīng" / "yīong" experiment. After everybody in the whole room could clearly hear the difference between "yīng" and "yīong" and could clearly tell which camp they fell into, we then did a regional analysis of the distinction, and it was obvious that the "yīong" pronunciation was centered on Beijing and the northeast. But what at first puzzled me is that, within Beijing speakers, some clearly fell into the "yīng" camp and some into the "yīong" camp, even for those who were born and grew up in the capital. That really mystified me, as it did all of the students. Finally, after talking things over for quite a while, we figured out that the parents of "yīng" speakers came from elsewhere, whereas the "yīong" speakers were from families who had deep roots in Beijing. Indeed, as is reflected in some of the previous comments in this thread, the outsiders / newcomers (even if they were born in Beijing and grew up there) often look down on Pekingese speech mannerisms, and their parents may be sharply disapproving of talking like a local Beijinger.

    One of the commenters in this thread, a very fine student who was born in Beijing and grew up there, but whose parents came from elsewhere, described Beijing speech mannerisms as diào'erlángdāng 吊儿郎当 ("slovenly; overly casual; undisciplined, careless"). Often, residents of Beijing with shallow roots will admit that they (and especially their parents) are prejudiced against Beijing speech mannerisms.

    The same kind of experimentation and analysis that I employed with the "yīng" / "yīong" distinction could also be applied to the "-n" / "-ng" distinction, the "n-" / "l-" distinction, the "zhi, chi, shi" / "zi, ci, si" distinction, and so forth. Now that every year I have a couple of dozen students from China, Taiwan, Hong Kong, Singapore, Malaysia and elsewhere who speak a wide variety of Sinitic languages and topolects, there are plenty of opportunities for this type of research every day.

    That was one of the best teaching days in my life, one of those days when you walk out of the classroom exhilarated and with a big smile on your face.

  25. Victor Mair said,

    May 27, 2015 @ 9:33 am


    "听 tīng 'tin/can'"

    Good catch! And I'm glad that Tom noticed it too.

    To expand on this, 听 is the simplified character for tīng 聽 ("hear; listen [to]; heed; obey; let") — 22 strokes (do you all spot the radical for "ear" among them?). Writing 听 to replace the complicated 聽 has been a common practice for the semiliterate and people in a hurry since the Yuan (Mongol) period about eight centuries ago. This same character, 听, incidentally, has another reading, yīn, reflecting a completely different morpheme signifying "mirthful appearance". This latter usage goes back more than two thousand years ago, so it is far older than using this character as the simplified replacement for tīng 聽 ("hear; listen [to]; heed; obey; let").

    Aside from its usage as the simplified character for tīng 聽 ("hear; listen [to]; heed; obey; let"), tīng 听 is also used to transcribe the sound of the English word "tin". It is used both for the metal ("tin") and in the sense of "can" (i.e., "tin can"). I can find no evidence of this usage before the 20th century.

  26. Rodger C said,

    May 27, 2015 @ 11:06 am

    I'm told that the late S. Y. Teng, who pronounced his name as English when speaking English, used to explain "Teng" to students by holding up his ten fingers.

  27. John Rohsenow said,

    May 27, 2015 @ 11:50 am

    Educated "immigrants" to Beijing look down on local Bjg 'tu hua' the same way educated speakers of "RP' in London look down on Cockney?

  28. Brendan said,

    May 27, 2015 @ 11:59 am

    I'd want to check this with native speakers, since I'm going by nothing but my non-native gut instincts here, but my sense is that people will produce both "-ing" and "-iong" ([jəŋ] sounds right to me for this) in different contexts, depending in large part on where the syllable is: 英 can turn into "yiong" pretty easily in isolation, but I don't think anyone would pronounce 英國 as "Yiōngguó;" likewise I can't picture 應該 being pronounced as "yiōnggāi," but have an easier time imagining someone saying, say, 有求必應 "yǒuqiú-bìyiòng." 行 xíng, used to indicate agreement, is often pronounced "xíəng" in Beijinghua, but I don't think the 行 of 自行車 zìxíngchē would be.

  29. K Chang said,

    May 27, 2015 @ 12:03 pm

    @Victor Mair — wonder if the waveforms on a computer would look different if you do it one way vs the other? Like a spectrogram with something like Audacity?

    My personal experience is that a lot of Mainland Chinese speakers tend to add a bit of "er" sound in their pronunciations. That example of "zero" I cited earlier would probably be the most egregious example. I can "do" that accent, but it's not the way I usually speak.

  30. K Chang said,

    May 27, 2015 @ 12:17 pm

    @Brendan — it can shift in a phrase as a part of "regional accent". By shifting the pronunciation slightly, and shift the tone slightly, you'll give the impression of a regional accent. I don't have training in accent identification, but I can imitate accents quite easily.

    In your example, 行 of 自行車 zìxíngchē , if you did pronounce it xi-er-ng (as one syllable) you'll still be understandable. Most listeners automatically filter the accent out.

    It obviously has the connotation that people who don't speak the unaccented version are backward simpletons. The CCTV hosts and news anchors are required to speak the "pure" version, of course.

  31. K Chang said,

    May 27, 2015 @ 12:27 pm

    Wonder if a discussion on Chinese tongue twisters 繞口令 (raokouling) appropriate here?

    Something like: 門外有四十四隻獅子,不知是四十四隻死獅子,還是四十四隻石獅子。

  32. Eidolon said,

    May 27, 2015 @ 5:06 pm

    @John Rohsenow Beijing "tu hua" is associated with the denizens of the hutongs – effectively the city ghettos – and therefore with lower socioeconomic class. It is stereotyped as the language of local taxi drivers, peddlers, and gangsters. So yes, I do think the analogy with cockney is apt.

  33. David Marjanović said,

    May 27, 2015 @ 5:51 pm

    A lot of my students would have strong disagreements on how to transliterate 新生, a fairly major road in Táiběi. Some students would say it was Xingsheng, some that it was Xinshen, and many would say that they couldn't hear a difference.

    It's Xīnshēng…

    I wonder what happens to the vowel. The syllable shen is supposed to have [ɛ], while sheng has [ɤ̃].

  34. julie lee said,

    May 27, 2015 @ 9:04 pm

    @David Marjanović:

    新生 xinsheng "new student" is pronounced "xingsen" in the Mandarin spoken in Hankow (Hankou), in Hubei, and many other places in Hunan and Hubei provinces. I grew up with the XINGSEN pronounciation but heard speakers of Beijing Mandarin in Taiwan University (when I was there many years ago) say XINSHENG. Here's the difference:

    XIN SHENG "new student" (Beijing Mandarin)

    XING SEN "new student" (my Mandarin)

    Similar differences in other words made my brother (in high school) and I (at university) ashamed of our Mandarin. Both of us made heroic efforts to rectify our defects. We had learned Mandarin from our parents, and now I thought my parents had a bad ear.

    Then, decades later, I met a librarian working at the Circulation Desk of Stanford Library who said XING SEN "new student" and other words just like my parents. A native of Hankow, she was new to America. She spoke just like my parents—said NIN "cold" instead of LENG "cold" (in Beijing Mandarin), and so on. I realized my parents didn't have a bad ear after all. They had simply moved from Hankow to Beijing, still keeping their Hankow Mandarin. So Victor Mair's question "And what exactly is standard Mandarin?" is a poignant one for me. In fact, it was after I read a paper by Prof. Mair with the title "Your Mother Tongue" (I can't remember the exact title) in praise of Chinese topolects and dialects that I stopped being ashamed of my native, mother tongue. I have gone back to saying XINGSEN "new student", NIN "cold", etc., the standard Mandarin of Hankow I learned as a child.

  35. julie lee said,

    May 27, 2015 @ 9:15 pm


    If I remember correctly, the paper by Victor Mair was "How to Forget Your Mother Tongue and Remember Your National Language".

  36. Brendan said,

    May 28, 2015 @ 10:45 am

    @K Chang — I think this is more than just a regional accent. Some speakers may drawl more than others, of course, but I'm pretty sure that "-ing" -> "-iəng" happens only in certain contexts, and that "zixingche" 自行車 isn't one of them.

    Going solely by the model of Beijinghua that I have in my head, I think this happens only with syllables that haven't got anything after them: "Yiəng" for 英 on its own, but never "Yiəngguo" for 英國 or "yiəngxiong" for 英雄; "xíəng" for 行 as in "Okay," but never "zixiəengche" for 自行車; "guóqíəng" for 國情, but not "qíəngrén" for 情人; maybe "bàomíəng" for 報名 and "wánrmìəng" for 玩兒命, but never "míəngzi" for 名字 or "mìəngyùn" for 命運 — whereas it would sound (to me) perfectly natural for 命令 to come out as "mìnglìəng." (This only holds true for natural speech: if someone is speaking loudly and slowly and emphasizing every syllable, as if to a foreigner or another mental deficient, then the "-ing" -> "-iəng" thing might happen.) I'd also been wondering whether or not this might be limited to stressed syllables, but I think the syllable's position in the word is the main constraint.

    Non-native speaker disclaimers apply even more strongly than usual here, obviously, but I'm pretty sure that this or something like it is the case.

  37. julie lee said,

    May 28, 2015 @ 10:50 am


    NIN "cold" should have been in Pinyin romanization "NEN" (rhyming with English "nin" as in "tannin").

  38. Victor Mair said,

    May 28, 2015 @ 12:44 pm

    From Tom Bartlett:


    This is interesting. What you wrote reminds me of the sound that I've heard my wife's nephew, a lifelong BJ resident, occasionally use for 行, pronounced as "xi-eng", with the initial x plus medial i component very perceptibly distinct from the main vowel e plus ending ng component. Between medial i and ending ng, the weak main vowel should be so compressed as to become inaudible, which is why the letter e drops out of all romanizations in such cases. When pronounced like this, with all elements audibly articulated in two component parts, it sounds like the first and second stages of the process that Taiwan students go through when saying aloud the 注音符號 spelling of 行: "xi, eng, xing". (Sorry, I don't have 注音符號 symbols on my iPhone.)
    I have also heard this pronunciation in films, particularly by young women, giving an effect that is apparently meant to be stylish. This was once a topic of discussion among my colleagues at Harvard. As I remember, Yang Ying, a native of BJ, said it originated among BJ high school girls in a certain sector of the city. I got the sense they did it consciously, for the fun of being ever so slightly, yet distinctively different, a familiar teenage sport.

    [VHM: Beijing Valley Girl talk]

    As for what you encountered, 英 spoken as "yiong", I haven't heard this articulation of the vowel used on a syllable with zero initial. Does your spelling mean that the syllable sounded like 用? Or are you using "o" to represent the weak middle vowel e that's not normally heard or written in this syllable, but is being (eccentrically?) articulated by some speakers?


    In reply to Tom Bartlett's last two questions: the sound I'm trying to describe in general is probably more like what he represents as "-ieng" or what some linguists who have been commenting on this post represent as "-iəng".

    My experiment and citations were all of "ying" in isolation, which in the speech of some Beijingers and northeasterners always came out as "yiong" (or "yieng" or "yiəng"). Brendan O'Kane's observations about the use of -ing syllables in actual speech show that context makes a difference in the actualization of the "yiəng" / "yieng" / "yiong" phenomenon.

    Much of the time, for many individuals, it's probably somewhere between "-iəng" and "-iong", so Tom's "-ieng" will do too. On the other hand, He Ziwei, whom I quoted in the original post, just sent in this note: "Sorry for not relying in time. I think you're right cause I also say 'yiong' for 英." And she sure did say "yiong", as she herself recognizes, not "ying", "yiəng", or "yieng". He Ziwei is a true Beijinger, but doesn't play up to all the stereotypes (slurring, extreme erization, etc.). She is a cultured, well-mannered, soft-spoken girl, but she has a very strong "yiong" realization of MSM "ying".

    In the end, I think that in Beijing and the northeast, there is a range of realizations for "ying", from "ying" to "yiəng" to "yieng" and "yiong", with some speakers probably intentionally putting their own stamp on their particular brand of this sound (greater emphasis on the ending). Reactions from other speakers to the various gradations range from not noticing, to mild annoyance or discomfort, to outright disapproval when the realization is more strongly and emphatically to the "yiong" side, thinking that it sounds too tǔ 土 ("earthy; colloquial") or even, in some cases, zuòzuo 做作 ("affected; put on"). For someone like He Ziwei, however, the "yiong" pronunciation is entirely natural. Indeed, before I called it to her attention a couple of years ago, she was completely unaware that she was doing it. He Ziwei has not an ounce of affectation in her, and she is certainly not tǔ 土 ("earthy; unfashionable"). In fact, I was just as impressed by her unique, tasteful wardrobe as I was by her elegant, distinctive English. "Yiong" was just her way of saying what most speakers of MSM pronounce as "ying".

  39. S said,

    May 28, 2015 @ 12:53 pm

    Taiwanese Mandarin speakers don't have a problem distinguishing -an and -ang at all, whether in comprehension or production. I'm not sure where you get that idea. I am one, and my ama and agong are rural folk with little education while my parents got educations and moved to the city, so I am acquainted with a wide variety of variously basilectal, "standard", etc. pronunciations. It's en/eng and in/ing only. After some consideration of this phenomenon, I had previously come to the tentative conclusion that it was because the vowels were too different in -an and -ang, but who knows. The sure thing is that they are NOT collapsed into one pronunciation.

  40. S said,

    May 28, 2015 @ 12:56 pm

    (To clarify, I don't mean the written Roman alphabet sequences a, n and a, n, g, but the lowish back vowel plus alveolar nasal or low back vowel plus velar nasal. Romanization is a different beast entirely.)

  41. S said,

    May 28, 2015 @ 2:29 pm

    Is there a reason that qualifying "all n/ng can be routinely mixed up" merits deletion?

  42. Rachel said,

    May 28, 2015 @ 3:14 pm

    It's Xīnshēng…

    I know that. The question is whether or not my students did.

    Put another way, Xīnshēng is the official MSM pronunciation, but there's a wide variation of local pronunciations for this.

    I believe, though don't remember clearly, that I had students do all four possible distributions of the -g: Xinsheng, Xingsheng, Xinshen and Xingshen.

  43. John said,

    May 29, 2015 @ 10:27 am

    S: I would say that just about all Taiwan Mandarin speakers could differentiate an/ang in comprehension, but certainly not in production. There's a reason that 南港展覽館 (a station in Taipei's metro system) is often considered a tongue twister. Look at this comic:

    (Standard Mandarin version: 「南港展覽館要搭哪一條?」「板南線啊。」「板南線是哪一條?」「藍色那條啊。」

  44. Rachel said,

    May 29, 2015 @ 11:29 am

    Re-reading what I wrote, I should've said: "I believe, though don't remember clearly, that I noticed students using all four possible distributions of the -g: Xinsheng, Xingsheng, Xinshen and Xingshen." That's hopefully a less ambiguous phrasing.

  45. K Chang said,

    May 29, 2015 @ 12:19 pm

    All this discussion is making me sad. I've been told that I'm an accent chameleon… I tend to sound like people I'm speaking to.

    @John — To "get" the joke of that cartoon, one almost have to speak Minnanyu-accented Mandarin. :D I can almost sound like that, but I my Minnanyu vocab is lacking. :D But I may be able to fake it by throwing in "Wa tong li gong!" (bad romanization of "I am telling you!" in Minnanyu) occasionally.

    The comments are actually more illustrative. The author wrote "字典裡沒有ㄢ" (no ㄢ in their dictionary) though technically he screwed up. It should have been 巷 instead of 線 

  46. julie lee said,

    May 29, 2015 @ 12:22 pm


    Yes, there is a wide variety of local pronunciations of Modern Standard Mandarin. I found that out among my schoolmates in Taiwan University years ago (when most of them were refugees from the Mainland). That is why I've gone back to Hankow Mandarin. It's well understood by MSM speakers, and if it's good enough for millions and millions of speakers in Hubei and Hunan, it's good enough for me.

    @Victor Mair's post

    My in-laws from Shandong say -iong/-yong for all words ending with -ing in Mod. Stand. Mandarin:
    So: YING "English" is "yiong/yong, JING "scripture" is "jiong/jyong", PING "level" is "piong/pyong", MING "bright" is "miong/myong".

    This North China "-iong/-yong" sound has even gone to Korea: Pyongyang 平壤 in North Korea is
    "Ping Rang" in Modern Stand. Mandarin and "Pyong Yang" in the Shandong speech of my in-laws. (Their initial r- 's [MSM] are all initial y- 's),. This also made me realize why the name Tokyo
    東京 "Eastern Capital", "Dong Jing" in MSM, became "To Kyo". 京 JING would have been KYONG in Shandong speech, and then the final -ng disappeared, giving us "-kyo".

  47. Akito said,

    May 30, 2015 @ 9:10 am

    This also made me realize why the name Tokyo 東京 "Eastern Capital", "Dong Jing" in MSM, became "To Kyo". 京 JING would have been KYONG in Shandong speech, and then the final -ng disappeared, giving us "-kyo".

    The dropped final /ŋ/ in Sino-Japanese is retained in the spelling as う (following /o/) or い (following /e/), both lengthening the preceding vowel.

  48. julie lee said,

    May 30, 2015 @ 12:29 pm


RSS feed for comments on this post