Observations on discounts and predictiveness

I was remembering something from my trip to Seoul in October, which then got me on to other things. The word for woman, or at least the important syllable when it comes to choosing the right bathroom (i.e. not the one that says ‘woman’ in my case) is 야 (ya). I thought of this because a friend who is studying hanja asked about 肉 which is 욕 (yok) in Korean (as far as the hanja is pronounced) but nyo’ in Wu, yuk in Cantonese and にく (niku) in Japanese¹. So basically I figured Mandarin r- becomes Korean hanja y-, though it’s ny- in Wu and Cantonese. Turns out I stopped one step too soon. the y- in Korean is actually only half the story. If the syllable is the first in the phrase, then it is in fact y-. Beyond that, however, it picks up an n-, making ny-. So the 肉 in 鸡肉 would actually be closer to nyok, bringing it almost perfectly in line with many Wu dialects.

A couple years back I’d read that Korean was used in the phonetic reconstruction of Middle Chinese², but didn’t quite see how that was the case. Now I know.

In studying Wu/Shanghainese I had (mistakenly, i later found out) thought that perhaps I could apply a phonetic matrix to what I knew of Mandarin and just go from there. That’s not the case, of course, as Wu has more tones (except in Shanghainese where it practically has roughly zero tones³) and has voiced initials not found in Mandarin. So with much of the distinction lost in Mandarin that is preserved in Wu, you can’t reasonably just convert.

But what about Korean? Could one take a more fundamental collection of details about Sinitic languages (including an 8-tone scheme and voiced intials) and accurately predict how a given character would be pronounced in Korean as hanja? Of course there are some minor issues, as 金 is 김 only when being used as a family name (Kim, e.g. Jong-il) and 금 when referring to the metal (or the day of the week, 金曜日 금요일, or the planet, 金星 금성). But maybe there’s a system to that as well. I’d be interested to know if anyone had encountered or formulated such a system.

Only slightly related, there’s a part of me thinks that the feeling of making progress in studying Korean would be much greater if words like “napkin” and “allergy” weren’t simply “napkin” (냅킨) and “allergy” (알레르기).

– – –
1. At least for the reading borrowed from Sinitic
2. And perhaps speaking to the value in early Chinese to Korean heritage, Paul Rouzer includes a Korean Pronunciation Index in his “New Practical Primer of Literary Chinese”
3. Roughly. It can be argued that Shanghainese, unlike other Wu dialects, is actually using a pitch-accent system and not tones proper. In fact only 2 of the 5 tones actually carry any semantic value anymore

6 responses to “Observations on discounts and predictiveness”

  1. Louis Platz says:

    Underlying such an idea is the assumption that Chinese words were borrowed into Korean in a regular way. There is some truth to this idea, as Korean readings of rù-shēng characters show correspondences to those seen in southern dialects. (I won’t go as far as to make any claims of correspondences between the qièyùn system and Korean character readings, although it would be interesting to see a comprehensive comparison of the two.)

    You might be interested in studying the records the Koreans made on some Chinese language(s) during the Ming dynasty. These language(s) are related to the work that later western missionaries did on Chinese, which geographically correlate to the Nanjing area. The language the Koreans were working on don’t look to be closely related to that shown in the Zhōngyuán Yīnyùn. Perhaps this is because the Zhōngyuán Yīnyùn reflects a northern dialect. (See Coblin. 2000. “A Brief History of Mandarin”. Journal of the American Oriental Society 120.4: pp.537-552.)

    There is a doctoral dissertation, from the University of Washington, that focuses on the work Koreans did on Chinese during the Ming dynasty; “A phonological study of Middle Mandarin: reflected in Korean sources of the mid-15th and early 16th centuries” by Kwangjo Kim.

  2. Zrv says:

    Sino-Korean readings do in fact show a high degree of regularity of correspondence with Middle Chinese categories (and, in particular, with Late Middle Chinese).* These have been well documented in numerous studies. (I can provide references in Korean if you like.) It is not entirely clear to what degree this regularity is due to a relatively compact period of initial borrowing, and to what degree it is due to artificial ex-post-facto regularizations. In any case, because of this regularity, and because of the high degree of regularity between Sinitic language pronunciations (Mandarin, Wu, what have you) and Late Middle Chinese phonological categories, it is certainly possible to come up with regular rules of correspondence between modern forms of Chinese and Sino-Korean readings.

    But you have to be aware that these correspondences will be multiple and overlapping, and without a good knowledge of Middle Chinese they are difficult to fully recognize and understand. To look at the example you cited: the Middle Chinese palatal nasal (traditionally called rìmǔ 日母) has developed in most cases into Mandarin r- (but in a few syllable types to Mandarin er). In Sino-Korean it is preserved as ny in word-medial position, but as a result of native Korean phonological restrictions, comes out as y- in word-initial position. (Korean does not permit word-initial n- before high front vowel sounds.) So there is a general correspondence between Mandarin r- and Korean y-/ny-. However:
    1) Some instances of Mandarin r- do not come from MC palatal nasal. In particular, many Mandarin morphemes pronounced róng do not. (They are the result of a late Mandarin change of yóng to róng.)
    2) There are other Middle Chinese initials that surface as SK y-, and of course these won’t correlate with Mandarin r-. Some of them end up as ny in word-medial position due to complex internal Korean morphophonological rules.

    These kinds of complexities exist with all the general correspondences, which means that if you undertake a simple comparison of Chinese and Korean without regard to historical developments, your correspondence patterns will look highly irregular and unpredictable.

    And that’s setting aside the genuine exceptions that have multiple historical causes (of the type you note with the surname Kim)!

    * The exception is aspiration. Mismatches in aspiration between Middle Chinese and Korean initials occur with very high frequency, possibly because aspiration as a distinct feature in Korean phonology was still developing at the time of borrowing.

    ** A few typos in your post: Sino-Korean for ‘woman’ is 여 yeo (not ya); for ‘meat’ is 육 yuk (not yok).

  3. Zrv says:

    P.S. Surely for a native English speaker like you, the existence of words like “napkin” (냅킨) and “allergy” (알레르기) should make learning Korean easier, not harder!

  4. when living in korea for about a year, i spent nearly all day reading and trying to figure out the gazillions of signs around town. i’d say when it comes to shop signs, you’ll find a particularly high proportion of chinese, which helped me a lot to try and guess the meanings. i never put together a stringent system of chinese ↔ sinokorean correspondences; rather, i immersed myself, as it were, until i had developed a feel for the correspondences.

  5. Louis Platz says:

    朴庆松. 1999. “和谐律与《切韵》音系四个“等”的转变规则”. 语言研究第36期.

    Here’s part of the abstract:

    …本文通过考察韩国汉字音的语音分类和读音,系统地研究了它与《切韵》音系的对应关系,发现“等”的系统与十五~十六世纪韩语的元音和谐律有成系统的对应:阳性元音[ʌ, a, o]对应于《切韵》的一、二等元音,阴性元音[ɨ, ɘ, u]对应于三等见系字元音,中性元音[i]或中性元音加阴性元音[ɘ, u]对应于四等元音和三、四等韵已合流的三等非见系字.

    I haven’t read the article, but the abstract reflects what Zrv said.

  6. Baruch says:

    Hi! I am also interested in and have experience with so-called phonetic chain shifts between Korean Japanese and Chinese, so I am enjoying this article and these comments very much. just a couple of lame corrections on your Korean: Woman in Korean is 여자 (Yeoja) so the toilet you presumably used was probably marked 여 not 야 , and the word for uncooked flesh is 육 (yuk/yook) not 욕 see http://en.wikipedia.org/wiki/Yukhoe.. there are some interestingly consistent correspondences between Korean and Japanese like (J)ren = (k) Yeon = 連 (j)Renraku/(K)Yeollak (連絡 communicate) and Renkon/Yeongeun (蓮根 lotus root)… I would like to know if there are some tables of these correspondences published anywhere….

Leave a Reply to ^love*encounter~flow