Chinese dialects, Korean, and Japanese all use Chinese characters in their writing systems, at least in some capacity. Chinese trivially so, Japanese through Kanji, and Korean through Hanja. As to be expected, each language has developed their own logographic simplifications as well as they differ in their pronunciation of various characters.

I cannot speak for Japanese, but Korean's pronunciations seem to be related to the Chinese pronunciations of some older form of Chinese but both languages have undergone sound changes meaning the differences are sometimes non-trivial.

I have noticed (and previously described here) that some Korean words, mostly proper nouns, can only be explained if they are translated literally from their written form, like some special form of a calque.

For example, the Korean word for "USA", 미국 (mikuk), is derived from the Hanja 美國. These are the same characters Mandarin uses for "USA". This etymology for the Mandarin 美國 (měiguó) shows that it must have originated in Mandarin (or some form of Chinese) before being adopted into Korean:

America -> 亞美利加 (yàměilìjiā) -> 美國 (měiguó)

(I've heard arguments that 미국 is actually derived from Chinese by way of Japanese, but this would not affect the phenomenon I am describing)

If Koreans had adopted 美國 through a phonological process alone, one would expect something like 메궈 (mekweo). The Korean transliteration for "America" is 아메리카 (amelikha). As such, it's hard to explain the presence of (mi) through any type of phonological adoption alone.

However, we can easily predict the appearance of (mi) from the character (literally meaning "beauty") which is consistently pronounced as (mi) even in situations not referring to the USA. For example, 미용실 (美容室, miyongsil, a beauty parlour). Similarly, is consistently pronounced (kuk) even in situations not referring to the USA. For example, 국회 (國會, kukhoy, the National Assembly). This is why I feel that the Korean pronunciation 미국 (mikuk) had to be adopted from the Chinese written form, not spoken.

However, there are alternatives. There is evidence that Korean sound change was more conservative when it came to the pronunciation of Sino-Korean vocabulary than some Chinese dialects. It is entirely possible that the original pronunciation of 美國 was closer to mikuk even in Chinese before sound change resulted in modern Mandarin's meiguo. Unfortunately I cannot rule this out as I do not know when 美國 entered the Chinese vocabulary but subjectively the mei in meiguo is a closer transliteration than the mi in mikuk and the difference in pronunciation can easily be explained by Hanja pronunciation rules (I am aware the subjective judgements are not evidence).

A clearer example is the Korean word for "Germany", 독일 (dokil). This was derived from the Japanese transliteration of "Deutsche" (as in "Deutschland"), ドイツ (doitsu). If Koreans had transliterated it themselves we would expect to see 도이치 (doichi). Similarly, if Korean had adopted it from the Japanese through phonological methods alone we would expect either 도이치 again or possibly 도이즈 (doiceu), again making it hard to explain the Korean pronunciation through phonological means. However, like before we can easily explain the Korean pronunciation using Hanja character pronunciations:

Deutsche -> ドイツ (doitsu, Katakana) -> 独逸 (Kanji) -> 獨逸 (Kyūjitai/Hanja) -> 독일 (dokil)

Note that the Katakana, Kanji, and Kyūjitai forms are pronounced identically and I am not implying they are discrete steps in the derivation just that I include them as evidence for the relationship between ドイツ and 독일. Also note that the Mandarin word for "Germany" uses unrelated characters and shows a different etymology.

This phenomenon even applies to names of people:

李连杰 (lǐ liánjié, Simplified Chinese) -> 李連杰 (Traditional Chinese/Hanja) -> 이연걸 (iyeonkeol)

In summary, Korean has an internally consistent system for pronouncing Chinese characters (called Hanja in Korean). All Sino-Korean words have a written form using Hanja characters which can be found in any Korean dictionary (although non Sino-Korean words do not have Hanja forms). In some cases it is almost impossible to explain the Korean pronunciation of a borrowing though phonological processes alone but trivial if one considers the borrowing's Hanja written form. This means that it's the written form that affects the word's pronunciation instead of the expected opposite.

Is there a name for this type of phenomenon, specifically with respect to borrowings? Does this phenomenon occur in any language other than Korean? (I suspect it occurs in all languages which relied heavily on Chinese script at any point in their history, but I do not have any examples) Is this phenomenon unique to Chinese script due to its ideographic nature and other geopolitical factors?

Is there any special term for a word's written form causing its spoken form to be reevaluated (even language-internally)? I would appreciate any examples of such.

  • I'm not too interested in examples like when English speakers purposefully pronounce the "j" in Spanish words like "jalapeno" for comedic effect. However, if such a change were systematically adopted by a large group of speakers, that would be OK. – acattle Apr 22 '13 at 7:05
  • Also, the scope of this phenomenon in Korean appears to be changing as even though Koreans recognize city names generated this way, such as 동경 (tongkeong) from 東京 (Kanji for Tokyo), most speakers (and airports) use more faithful phonetic adoptions. In this case, 도교 (tokyo) from the Japanese とうきょう (tokyo). – acattle Apr 22 '13 at 7:12
  • I'm not sure I follow, possibly because I know very little about CJ or K. Surely the "...own pronunciations and simplifications for these Chinese characters." are simply the J and K versions of the C pronunciations? In which case they are simply borrowing words and adapting them to their own phonology, ie normal borrowing? – Gaston Ümlaut Apr 22 '13 at 9:05
  • @GastonÜmlaut I've edited the question to try to be clearer about my thought process and my evidence. Unfortunately to fully understand this phenomenon you need to have some understanding of CJK languages and specifically how they use Chinese characters. The gist of my argument is that given certian Sino-Korean words, the pronunciations of the Hanja characters (which can be found in any Korean dictionary) easily and accurately explain the pronunciation of the word itself in situations where phonological processes alone would predict different a different form. – acattle Apr 22 '13 at 10:50
  • 1
    There are plenty of examples of spelling pronunciation in English. Some that come immediately to mind are: 'forehead' and 'waistcoat', for both of which the spelling pronunciation (as opposed to the traditional pronunciation) is now the dominant one. There are lots of examples on WP. Is this the phenomenon you are asking about? – Gaston Ümlaut Apr 22 '13 at 23:50

The 'x' in Mexico is pronounced 'ks' in English (and many other languages). This is because that's how the x is usually pronounced in English. However, in Spanish the x stood for a 'sh' sound or, more recently, as the spanish 'j' as we know now (as in jalapiño or mojito).

This messup is purely based on orthographical reasons.

| improve this answer | |
  • So that explains the occasional /me.hi.ko/ pronunciation of "Mexico". Any chance you could provide a source for this? – acattle Apr 22 '13 at 12:00
  • Like the X in quixotic – Double AA Apr 25 '13 at 19:31
  • This answer on Spanish.SE addresses this phenomenon in Spanish. – Flimzy May 8 '13 at 8:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.