Carmen in Korean and Cantonese

« previous post | next post »

Reader Jean-Michel found an odd example of a Sinographic typo and it's got him stumped. This has to do with the Korean Blu-ray release of "As Tears Go By," the 1988 debut feature by Hong Kong director Wong Kar-wai.

In Chinese the film is known as Wàngjiǎo kǎmén 旺角卡門 ("Mongkok Carmen") after the Bizet opera (though the resemblances are very superficial). What is strange, however, is that the Korean Blu-ray art, as illustrated below, initially gave the characters as Wàngjiǎo xiàwèn 旺角下問.



I originally found these photographs here, but they don't seem to be available at that site any longer.

Jean-Michel reflects:

My first thought was that maybe they changed it for Korea because "下問" sounds like "Carmen" in Korean and "卡門" doesn't. But in Korean "下問" is 하문 "Hamun" (not "Kamun") and both 門 and 問 have the same pronunciation, so that wouldn't explain the replacement. Then I found that the Korean title of the film is actually completely different from the Chinese:《열혈남아》/《熱血男女》"Hot-Blooded Men and Women." Looking closely at the Blu-ray art, I can see the Hangul for this title in small type below the "Chinese" title.

So it appears this is a genuine mistake, and one I suspect will be fixed for the final release. But now the question becomes: What kind of mistake transforms "卡門" into "下問"?

VHM notes:

1. 卡門 is purely a transcription of "Carmen"; it is the standard way to refer to Bizet's opera in Chinese.

2. 下問 is probably just a typing error for 卡門, but it might conceivably mean something like "asking downward" (?), as in bùchǐ xiàwèn 不恥下問 ("not feel ashamed to ask [and learn from] one's subordinates"), though that would hardly make any sense in this context.

3. The usual Korean transliteration for "Carmen" is 카르멘 (Kaleumen).

4. The romanization of the Korean title of Wong Kar-wai's film, 열혈남아, is yeolhyeolnam-a (RR) / yŏrhyŏllam-a (MR).

5. The equivalent Hanja (Chinese characters) for 열혈남아 are 熱血男兒 ("hot blooded young man").

6. The Korean transcription of 下問 is 하문 hamun (RR) / hamun (MR).

7. The Korean transcription of 卡門 (though I'm not sure about the reading of the first character) would be 잡문 (jammun: RR, chammun: MR) OR 가문? (gamun: RR, kamun: MR).

8. The Cantonese pronunciation of 旺角卡門 (Mongkok Carmen) is Wong6gok3 Kaa1mun4. The name of the Hong Kong place name, Mongkok or Mong Kok is in accordance with Cantonese, not Mandarin (Wàngjiǎo). The surface signification of Wong6gok3 is "Lively / Flourishing / Prosperous corner", but the name has a deeper etymological history and was formerly written with other characters, which you can read about here.

9. The Cantonese pronunciation of 旺角下問 is Wong6gok3 Haa6man6.

10. I leave it so to someone who is better informed about the history of Cantonese phonology than I to explain why Wong6gok3 is transcribed as Mongkok in English.

I close by repeating Jean-Michel's final question: "What kind of mistake transforms '卡門' into '下問'?" Orthographic? Phonological? Was it the result of a particular shape-based inputting system that makes it easy to confuse 卡門 with 下問?

[Thanks to Bill Hannas, Bob Bauer, Haewon Cho, Abraham Chan, and Stephan Stiller]


  1. David Morris said,

    April 29, 2014 @ 4:34 am

    I speak with a basically non-rhotic accent. I have noticed a number of English words translated in Korean with the 르 added (카르멘 here and also 모차르트 when I was in Korea, that I can think of). On the other hand, a few weeks ago here we discussed 참ing biscuits, which doesn't have the 르. Is there a rule governing this?

  2. Jean-Michel said,

    April 29, 2014 @ 4:38 am

    5. The equivalent Hanja (Chinese characters) for 열혈남아 are 熱血男兒 ("hot blooded young man").

    Ouch, that's embarrassing…No particular mea culpa for my error, I'm just not fluent or literate in Korean and got 아 a confused with 여 eo (which I don't think would be the correct pronunciation of 女 in that context anyway) without bothering to check.

    Oddly, if you look on the Internet Movie Database entry for the film, it gives "Rexie naner" as an alternate title, i.e. the Mandarin reading of 熱血男兒. I doubt the movie has ever been known by that title in Mandarin, so most likely someone saw the Korean title in Hanja (on this DVD cover, for example) and assumed it was an alternate Chinese title. The same title was recycled a few years ago for a Korean film and also seems to be the name of a porno mag (friendly advice: don't GIS that title at work…).

    Another question I have that I didn't include in my email: Is the character 卡 at all familiar to Koreans, even those with a solid grounding in Hanja? In Chinese it's now primarily a phonetic transcription character, something Korean doesn't have much use for nowadays. The original meaning seems to be something like "to stick" and it still occurs in some compounds like 卡住 "stuck," but I don't know if such vocabulary has been carried over to Korean.

  3. Jean-Michel said,

    April 29, 2014 @ 5:04 am

    @David Morris: According to the writeup on "참img", 참 not only sounds like "charm" but also means "genuine" in Korean, and apparently connotes "attractive" as well. That strikes me as a pretty good reason to go with 참 for "charm" instead of trying to approximate a rhotic pronunciation.

  4. Victor Mair said,

    April 29, 2014 @ 5:35 am

    What a goldmine of translatese in the AKA section of the Internet Movie Database entry for the film provided by Jean-Michel in his first comment!


    Also Known As (AKA)
    (original title) Wong Gok ka moon
    (original title) Wong Gok ka moon
    Carmen of the Streets
    (literal English title) Mongkok Carmen
    Rexie Naner
    China (Mandarin title) Wang Jiao ka men
    Germany (DVD title) As Tears Go By
    Spain El fluir de las lágrimas
    Finland As Tears Go By
    Greece (transliterated ISO-LATIN-1 title) Kathos kyloun ta dakrya mas
    Hong Kong (Cantonese title) Wong Gok ka moon
    Hungary Ahogy peregnek a könnyek
    South Korea Yeolheol nama
    Poland Kiedy lzy przemina
    Portugal (working title) Ao Sabor da Ambição
    Russia Пока не высохнут слезы
    Sweden As Tears Go By
    World-wide (English title) As Tears Go By

  5. Victor Mair said,

    April 29, 2014 @ 6:12 am

    Quick comments by a colleague:


    For the two earlier spellings Wikipedia lists, everyone would say 望角 as mong6gok3 (though at least one source has mong4 as a variant pronunciation for 望) and 芒角 as mong4gok3.

    If the explanation on the Chinese Wikipedia is correct (the English Wikipedia is confusing here), the Tanka (HK's boat people, one of the early populations of HK) pronounced 芒 (Cantonese: mong4) "as 望" (that would normally be mong6 in Cantonese, but if it was another dialect, we don't have precise knowledge, and the tone does not matter in this context anyways). But I'm thinking that the British could have taken this from any Cantonese (Yue) pronunciation. The Tanka speak Yue. I don't have time to research Hakka pronunciations of 芒 etc, so I don't want to post this (also I'm short on time these days). The point may be that 芒 doesn't sound like "mong" in Hakka, though note that Hakka seems to be close to Yue. Feel free to just add this to your original post.

    The context for what the Hakka have got to do with it is the second paragraph here:


    As for structural input methods, the codes are:

    So I doubt it was an input method typo.

  6. Kuiwon said,

    April 29, 2014 @ 6:53 am

    卡 is very unfamiliar to Koreans. It doesn't even show up in the Hanja Proficiency Test rankings:

    At least one Korean blog thinks that this must have been a typo:

  7. Jongseong Park said,

    April 29, 2014 @ 4:06 pm

    The RR transcription for 카르멘 /kʰaɾɯmen/ is Kareumen, not Kaleumen. I see a number of little transcription errors here and there. For example, 여 /jʌ/ is yeo, not eo (which corresponds to 어 /ʌ/).

    @David Morris: I have noticed a number of English words translated in Korean with the 르 added (카르멘 here and also 모차르트 when I was in Korea, that I can think of). On the other hand, a few weeks ago here we discussed 참ing biscuits, which doesn't have the 르. Is there a rule governing this?

    The short answer is that Carmen and Mozart are not originally English words, but Spanish/French and German respectively.

    In general, rhotic consonants in foreign languages are transcribed as 르 reu in front of consonants or word-finally. Carmen is thus regularly 카르멘 Kareumen whether it is taken as a Spanish or a French name. Mozart is 모차르트 Mochareuteu as it is a German name.

    However, in transcribing English, /r/ is not transcribed unless followed by a vowel. Either a non-rhotic pronunciation is taken as the basis of transcription, or even if a rhotic pronunciation is used, the fact that an English /r/ is an approximant in most dialects keeps it from being treated the same way as the rolled or tapped /r/ in other languages.

    So in addition to 참 cham "charm" and 차밍 chaming "charming", you have
    서비스 seobiseu "service"
    아트 ateu "art"
    컴퓨터 keompyuteo "computer"
    콘서트 konseoteu "concert"

    There are just a few exceptions, such as 멜버른 Melbeoreun "Melbourne" which would have been *멜번 Melbeon if it followed the usual rules.

    German is a special example. As a concession to German vocalization, word-final /r/ is written as 어 eo. So whereas we have 모차르트 Mochareuteu for Mozart and 베를린 Bereullin for Berlin, Mahler is 말러 Malleo and Kantor is 칸토어 Kantoeo. There are ambiguities in how to treat compound and derived words, and also plenty of exceptions.

    Because Carmen is so familiar in the spelling 카르멘 Kareumen due to the opera, Koreans will often use it for Anglophone Carmens instead of basing the transcription on the English pronunciation and using 카멘 Kamen, which would be considered more correct.

  8. Jongseong Park said,

    April 29, 2014 @ 4:39 pm

    Koreans usually input Chinese characters by typing the Korean reading of the character and then selecting the desired character from the menu that appears after pressing either a dedicated hanja key on the keyboard, some other keyboard shortcut, or some other method depending on the software. A bit cumbersome, but Chinese characters are used very rarely when writing Korean anyway.

    There are a couple of character recognition methods that are used when one doesn't know the reading of the character, but knows how it is written. One writes the character with the mouse or stylus on a window, and the software suggests characters that match.

    旺角卡門 is absolute gibberish to most Koreans, who aren't familiar with either the Hong Kong place name or Chinese name for Carmen. 下問, however, is a term that is actually used in Korean to describe asking something to a social inferior. So I think this usage, though rare, has got to have something to do with the typo. Perhaps there was even some sort of autocorrect failure involved.

  9. Stephan Stiller said,

    April 29, 2014 @ 4:54 pm

    @ Jongseong Park
    I would transcribe Korean 어 as [ɔ] even in a broad transcription. I've also seen people transcribe 어 as [ə], which I disagree with too.

  10. Jongseong Park said,

    April 29, 2014 @ 5:55 pm

    @Stephan Stiller, I have heard similar suggestions from American English speakers. Due to traditional transcriptions of English vowels, English speakers are used to /ʌ/ representing something a bit different from its cardinal value, so seem a bit resistant to its use for Korean 어.

    Many American English speakers have CAUGHT vowels that are at least partially unrounded, so for them Korean 어 is closer to the English vowel generally written /ɔː/ than to the English vowel generally written /ʌ/. For many British speakers, the CAUGHT vowel is somewhat more raised, so the LOT vowel is a closer match; John Wells uses /ɒ/ in the Longman Pronunciation Dictionary to transcribe Korean 어.

    But /ʌ/ is a perfectly legitimate choice for 어 as we're not required to cater to what English speakers expect phonetic symbols to represent. 어 is a non-front, mid vowel that patterns with other unrounded vowels like 으 /ɯ/ and 아 /a/, and cardinal /ʌ/ is quite close to a typical realization of 어 in modern Korean. Phonetically 어 may feature some rounding, but using /ɔ/ would imply that the difference with 오 /o/ is only that of height, which is not the case (except perhaps in Northern dialects, but I'm not even sure about that). Schematically, the hypothesis of a height distinction between two mid vowels might be more plausible if the 에 /e/ – 애 /ɛ/ distinction wasn't moribund in Modern Korean. I doubt any dialect of Korean featured simultaneous height-based distinctions of /o/-/ɔ/ and /e/-/ɛ/.

  11. Jongseong Park said,

    April 29, 2014 @ 6:41 pm

    I should add that in the traditional standard Korean dialect, long 어 /ʌː/ had a noticeable difference in quality from short 어 /ʌ/, being realized as a central vowel not very far from [əː]. Length distinctions are moribund in standard Korean, but [ə] will still be understood as a possible realization of 어.

  12. Stephan Stiller said,

    April 30, 2014 @ 4:07 pm

    @ Jongseong Park

    Thanks for your reply. Your reasoning is good and you are right that [ʌ] is conventionally used in ways for English where another symbol ([ɐ] with a diacritic or one of the other central-vowel symbols) might be more appropriate.

    That said, there is a question about the degree of rounding of 어; I'll look into that at some point. The symbol [ɒ] (which you say John Wells used for 어 in some place) certainly denotes a round vowel. (There's also a convention to stick with more common symbols if possible, but it's unclear whether that would favor [ɔ] over [ʌ] (since both are cardinal vowels), and this would apply only to broad transcriptions.)

  13. Jongseong Park said,

    April 30, 2014 @ 10:36 pm

    I would say that today, 어 is partially rounded, so we could narrowly transcribe it as [ʌ̹] or [ɔ̜] with diacritics indicating that it is more rounded or less rounded respectively. So on pure phonetic grounds either symbol would work.

    It seems that the older value of 어 was closer to central [ə] before it became a back vowel in contemporary standard Korean (perhaps to fill the gap left by the disappearance of the vowel represented by the archaic letter ㆍ, which was a back vowel). Similarly, 으 was more central, often transcribed as [ɨ] in older texts, but nowadays it is a back [ɯ]. So given the choice between /ʌ/ and /ɔ/, I prefer the former since 어 was traditionally an unrounded vowel in all reconstructions I've seen.

  14. David Morris said,

    May 1, 2014 @ 3:39 am

    @Jeongseong –
    Thanks. Informative, as ever.
    This explains the other 르 word which I couldn't think of when I first posted, which is 아르바이트 (from German). I had students in Korea who simply didn't believe me when I told them that that word a) wasn't English, and b) wouldn't be understood in the context of 'part-time job' by 99.99% of English speakers.

  15. Jongseong Park said,

    May 1, 2014 @ 7:24 am

    @David Morris,
    Regarding 아르바이트 areubaiteu, it is also interesting to look at its informal abbreviation is 알바 alba, which turns the tap into a lateral.

    This seems to me to have been influenced by the mistaken assumption that this is originally an English word. Although the official transcription policy treats English as non-rhotic, Korean speakers sometimes map the approximant R in English (not followed by a vowel) to [l] in Korean.

    For example, many Koreans say (and to a lesser extent, write) 싱가폴 Singgapol for Singapore, although it is officially written 싱가포르 Singgaporeu, following a rule that writes -ore in South Asian toponyms as 오르 oreu (Singapore is not treated as an English name in this case, at least not a typical one).

    Scorsese is often seen written as 스콜세지 seukolseji, though the official transcription is 스코세이지 seukoseiji.

    Again, this is because the English R is usually not a trill or a tap, but an approximant. Compare also the fact that the official transcription for Chinese rhotic approximant coda written er in pinyin is 얼 eol.

    So people who think 아르바이트 areubaiteu comes from English might think it natural to convert 르 reu into ㄹ l as many people do for 싱가포르 Singgaporeu / 싱가폴 Singgapol. Even those who are aware that Arbeit is German wouldn't necessarily know about the pronunciation differences in English and German rhotic sounds.

RSS feed for comments on this post