Comparing Romanizations·
로마자 표기법의 비교
(Note: This article is undergoing constant revision as new, arcane bits of information come to light. This article's permanent location is http://21cseonbi.blogspot.com/2005/05/comparing-romanizations.html.)
This article compares three commonly used Korean romanization systems: McCune-Reischauer Romanization [1], Revised Romanization of Korean [2], and Yale Romanization of Korean [3]. The comparison is purely analytical, and I will withhold from passing judgement on the perceived merits or demerits of any particular system.
Many systems for romanizing Korean are currently in use or have been used in the past. I have already discussed two such systems (McCune-Reischauer and Revised Romanization) at moderate length elsewhere. One that is frequently used but often overlooked is the Yale Romanization of Korean. This system is almost universally used by linguists when not writing in Korean, and deserves some attention. Unlike the Revised Romanization of Korean and the McCune-Reischauer System—both of which are transcription systems that attempt to approximate the pronunciation of Korean—Yale Romanization is a transliteration system that denotes the Hangul spelling of Korean words [4]. Thus, there is a one-to-one correspondence between Hangul and Roman letters that deliberately disregards the sound changes that consonants frequently undergo in Korean.
Without further ado, I herewith present a comparison table of the McCune-Reischauer, Revised, and Yale Romanization systems [5]. All of the notes below apply to the table:
- Letters and letter combinations are listed in the order in which they appear in most dictionaries.
- The column heading "Initial" means "at the beginning of a word or after a consonant." McCune-Reischauer distinguishes between initial and non-initial consonants.
- "Medial" means "between vowels" (or more properly, "after a vowel and before the consonant ㅇ").
- "Final" means "at the end of a word or before a consonant (other than ㅇ)." Both McCune-Reischauer and Revised Romanization distinguish between final and non-final consonants. The pronunciation of final consonants is defined by the "final consonant rule" (patch'im kyuch'ik; 받침 규칙; ——規則), which assimilates all final consonants to one of seven permissible sounds: "k" (ㄱ), "n" (ㄴ), "l" (ㄹ), "m" (ㅁ), "b" (ㅂ), "t" (ㅅ), and "ng" (ㅇ). (When ㅎ appears as a final consonant, the sound of the consonant that follows it is modified, transforming the entire combination into a medial sound.)
- The double consonants (toensori; 된소리; "intensified sound") ㄸ, ㅃ, and ㅉ never occur in the final position. The double consonants ㄲ and ㅆ can occur in any position. (When ㄲ and ㅆ occur in the final position, they are called ssangbatch'im (쌍받침; 雙——; "twin final consonants.")
- Consonant clusters (kyôppatch'im; 겹받침; "double final consonants")—the non-double consonant pairs ㄳ, ㄵ, ㄶ, ㄺ, ㄻ, ㄼ, ㄽ, ㄾ, ㄿ, ㅀ, and ㅄ—never occur in the initial position, but can occur in either the medial or final position.
- Improbably, kyôppatch'im are not dealt with at all in the official document that defines the Revised Romanization of Korean. The spellings given in the table below are based on results obtained from a program created by the Ministry of Culture and Tourism for converting Hangul words into their Revised Romanization transcriptions. Note that for medial ㄳ, ㄽ, and ㅄ, the ㅅ is transcribed as a double "s" ("ss").
- The kyôppatch'im ㄺ and ㄼ have complicated pronunciation rules, which are shown in the table.
- In the official 1984 South Korean romanization system based on McCune-Reischauer, the consonants and consonant combinations ㄳ, ㄽ, ㅄ, ㅅ, and ㅆ are normally written as "ks," "ls," "ps," "s," and "ss" respectively, but as "ksh," "lsh," "psh," "sh," and "ssh" respectively when followed by "i" or "wi." The change of "s" to "sh" is not applied in the original 1939 McCune-Reischauer system itself, however—except for the single case of "shwi" (쉬)—and is not shown here.
- In many cases, when a consonant (chaûm; 자음; 子音) comes immediately before or after another consonant in a word, one or both consonants undergo a sound change (chaûm tonghwa; 자음 동화; 子音同化). For the most part, such changes are regular, follow well documented rules, and are preserved in both McCune-Reischauer and the Revised system. Due to their complexity, however, this table does not show such sound changes.
- In addition to regular sound changes, there are many irregular consonantal sound changes (or non-changes) as well, which are reflected in McCune-Reischauer. For example, the sequence -ㄴㅈ- is normally transcribed as "-nj-," but as "-nch-" in the word hancha (한자 [pronounced 한짜; 漢字]).
- Sound changes only apply to consonants. Vowels (moûm; 모음; 母音) sometimes do change their sounds somewhat depending on their position in a word or the sound values of the adjacent consonants, but such changes are not generally significant, nor are they governed by rules as consonantal sound changes are. Thus, vowels and vowel combinations do not change their spellings in any of the three Romanization systems.
- In Yale Romanization, when a consonant at the end of a syllable is followed by a vowel or consonant, the combination of which could lead to confusion, a period (.) is inserted after the first consonant. For example, 자연 (自然; "nature") is written as ca.yen, since the undifferentiated spelling "cayen" could be mistaken for 재언 (再言; "speaking again"). McCune-Reischauer uses an apostrophe (') when disambiguation is necessary, and Revised Romanization uses a hyphen (-).
- In Yale Romanization, there are special rules for:
- The initial letter of the first syllable in a word when that syllable represents a hanja (Sino-Korean character) that begins with a ㄹ (governed in Korean by the "two-sound rule" (tuûm pôpch'ik; 두음 법칙; 頭音法則));
- The reinforcing sound that occurs between certain pairs of syllables and is sometimes represented by a ㅅ (called in Korean the "interval sound" (saissori; 사잇 소리), with the ㅅ itself being called "interval siot" (sai siot; 사이 시옷), from the Korean name of the letter); and
- Cases in which a particular letter is included in South Korean spelling but excluded in North Korean spelling, or vice versa.
As time permits, I will elaborate upon each of these three rules.
- The initial letter of the first syllable in a word when that syllable represents a hanja (Sino-Korean character) that begins with a ㄹ (governed in Korean by the "two-sound rule" (tuûm pôpch'ik; 두음 법칙; 頭音法則));
- For technical reasons, the McCune-Reischauer letters ô and û are written with circumflexes in the table below. The circumflexes should properly be replaced with breves, which look like rounded, inverted circumflexes.
The Table
(Note: This table is still under construction. A question mark (?) denotes a spelling that I have to verify.)
| Comparison Table of McCune-Reischauer, Revised, and Yale Romanizations of Korean | ||||||
|---|---|---|---|---|---|---|
| Consonants · 자음 | ||||||
| Hangul | McCune-Reischauer | Revised | Yale | |||
| Initial | Medial | Final | Initial/ Medial | Final | ||
| ㄱ | k | g | k | g | k | k |
| ㄲ | kk | kk | k | kk | k | kk |
| ㄳ | — | ks | k | kss | k | ks |
| ㄴ | n | n | n | n | n | n* |
| * I will add a note on the Yale treatment of ㄴ later. | ||||||
| ㄵ | — | nj | n | nj | n | nc |
| ㄶ | — | n | n | n | n | nh |
| ㄷ | t | d | t | d | t | t |
| ㄸ | tt | tt | t | tt | t | tt |
| ㄹ | r | r | l | r | l | l* |
| * I will add a note on the Yale treatment of ㄹ later. | ||||||
| ㄺ | — | lg | k, l* | lg | k, l* | lk |
| * Normally "k," but pronounced "l" when followed by "k," as in 맑고 (pronounced "말꼬") [6]. | ||||||
| ㄻ | — | lm | m | lm | m | lm |
| ㄼ | — | lb | l, p* | lb | l, p* | lp |
| * Normally "l," but pronounced "p" for the verbs 밟다, 넓죽하다, and 넓둥글다 [6]. | ||||||
| ㄽ | — | ls | l | lss | l | ls |
| ㄾ | — | lt' | l | lt | l | lth |
| ㄿ | — | lp' | p | lp | p | lph |
| ㅀ | — | r | l | r | l | lh |
| ㅁ | m | m | m | m | m | m |
| ㅂ | p | b | p | b | p | p |
| ㅃ | pp | pp | p | pp | p | p |
| ㅄ | — | ps | p | pss | p | ps |
| ㅅ | s | s | t | s | t | s, q* |
| * Normally "s," but written as "q" when denoting sai siot (사이 시옷), which I shall discuss later. | ||||||
| ㅆ | ss | ss | t | ss | t | ss |
| ㅇ | — | ng* | ng | ng* | ng | ng* |
| * Not shown if ㅇ is at the beginning of a syllable. I will add a note on the Yale treatment of ㅇ later. | ||||||
| ㅈ | ch | j | t | j | t | c |
| ㅉ | tch | tch | t | jj | t | cc |
| ㅊ | ch' | ch' | t | ch | t | ch |
| ㅋ | k' | k' | k | k | k | kh |
| ㅌ | t' | t' | t | t | t | th |
| ㅍ | p' | p' | p | p | p | ph |
| ㅎ | h | (h)* | h | (h)* | h | h |
| * Normally not shown in the medial position | ||||||
| Vowels · 모음 | ||||||
| Hangul | McCune- Reischauer | Revised | Yale | |||
| ㅏ | a | a | a | |||
| ㅐ | ae | ae | ay | |||
| ㅑ | ya | ya | ya | |||
| ㅒ | yae | yae | yay | |||
| ㅓ | ô | eo | e | |||
| ㅔ | e | e | ey | |||
| ㅕ | yô | yeo | ye | |||
| ㅖ | ye | ye | yey | |||
| ㅗ | o | o | o | |||
| ㅘ | wa | wa | wa | |||
| ㅙ | wae | wae | way | |||
| ㅚ | oe | oe | oy | |||
| ㅛ | yo | yo | yo | |||
| ㅜ | u | u | (w)u* | |||
| * Normally "wu," but "u" after "m," "p," "pp," or "ph." | ||||||
| ㅝ | wô | wo | we | |||
| ㅞ | we | we | wey | |||
| ㅟ | wi | wi | wi | |||
| ㅠ | yu | yu | yu | |||
| ㅡ | û | eu | u | |||
| ㅢ | ûi | ui | uy | |||
| ㅣ | i | i | i | |||
| Copyright © 2005 Stefan Ewing. All rights reserved. | ||||||
Footnotes
[1] As defined in George M. McCune and Edwin O. Reischauer: "The Romanization of the Korean Language" (PDF file; download and open directly in Acrobat, then select View → Rotate View → Clockwise). Transactions of the Korea Branch of the Royal Asiatic Society, vol. XXIX. Seoul: Royal Asiatic Society Korea Branch, 1939. I have also consulted the official South Korean romanization system first published in 1984—based on the McCune-Reischauer system—and superseded in 2000 by the Revised Romanization of Korean. (Korean Romanization (Kugôûi Romaja P'yogibôp; 『국어의 로마자 표기법』). Seoul: Ministry of Education, 1987.)
[2] As defined in The Romanization of Korean (Kugôûi Romaja P'yogibôp; 『국어의 로마자 표기법』). Seoul: Ministry of Culture and Tourism, 2000.
[3] A detailed description of the Yale Romanization is provided in Samuel E. Martin: A Reference Grammar of Korean. Rutland, VT: Tuttle, 1992.
[4] The terms "transcription" (chônsappôp; 전사법 [pronounced -뻡]; 轉寫法) and "transliteration" (chônjappôp; 전자법 [pronounced -뻡]; 轉字法) illustrate a critical distinction between romanization systems. The same distinction is examined in detail in Korean-language article by Ik-seop I (이익섭): "Characteristics of Romanization" (Romaja P'yôgibôbûi Sônggyôk; 「로마자 표기법의 성격」). New Korean Life (Sae Kugôo Saenghwal; 『새 국어 생활』), Vol. 7 No. 2. Seoul: National Academy of the Korean Language, 1997. (Only the table of contents of the aforementioned issue is available online; if you wish to search for it in a library outside of Korea, some catalogues will return results for "sae kugo saenghwal".)
[5] For detailed comparisons of some or all the above systems (McCune-Reischauer, Revised, and Yale) to other romanization systems not covered in this article, see the PDF file Report on the Current Status of United Nations Romanization Systems for Geographical Names [Korean], Version 2.2 (UNGEGN Working Group on Romanization Systems, Jan. 2003) or the Korean-language article by Jeong-hui Won (원정희): "Comparison of Successive Important Romanization Systems" (Yôktae Chuyo Romaja P'yogibôp Pigyo; 「역대 주요 로마자 표기법 비교」). New Korean Life (Sae Kugô Saenghwal; 『새 국어 생활』), Vol. 7 No. 2. Seoul: National Academy of the Korean Language, 1997. (Only the table of contents of the aforementioned issue is available online; if you wish to search for it in a library outside of Korea, some catalogues will return results for "sae kugo saenghwal".)
[6] "Phonetic Changes" (Ûmunûi Pyôndong; 「음운의 변동」). High School Grammar (Kodûng Hakkyo Munpôp; 『고등 학교 문법』). http://woorimal.net/language/moonbub/phonology-2.htm. May 27, 2005.


19 Comments:
Woah... don't have time right now to read that post but in response to your question:
Yes... I am from Canada originally, but am now in sunny California.
Hello, I'm just random visitor.
I like your place. ^-^
I'm a high school student in America and I'm from South Korea.
I like people who are interested in knowing about Korean culture,language, and so on.
Will you come to my website,too?
www.ilovejennyjung.blogspot.com
I will be visiting here often to see what's going on.
It was nice visiting you.
Bye~ ^-^
Hi, Jenny:
Thanks for your kind compliments. I'll check out your site!
That's a serious post! And a very useful reference resource I hope - maybe you should submit something like this for Wikipedia or somewhere similar.
Still awaiting your table for Yale romanisation of Middle Korean. (only joking ^^)
Hah! I'd be way in over my head if I tried tackling Middle Korean!
I like Wikipedia in principle, but the problem is that because of its GFDL license, there are literally hundreds of clone sites out there, all with the same information. It's nice to see one's work in a prominent place, but it's a bit disconcerting to see it everywhere.
In fact, the biggest issue is simply that many of these clone sites don't mention that their material comes from Wikipedia (although they are supposed to), or put it in very, very small print at the bottom of each page.
So it is often not easy to trace an article back to its Wikipedia source (if the user doesn't know where it came from), and copied articles don't have an edit history, so it's not possible to know who's worked on it.... Well, I could go on.
I'll just hope that people somehow find this place.
Good show. For Middle Korean, you'll need Yale, only Yale and always Yale.
ㅜ is always spelled wu, and ㅗ wo, because o is 아래아
Of course, it is a little more complex than that – but not terribly so.
In my [ahem] unfinished PhD I used ywo and ywu for ㅛ and ㅠ, because one of my surmises was that yu (ㅣ + ㅡ) and yo (ㅣ + `, known still in Cheju dialect as ``) existed in the standard language until a little before the publication of 訓民正音 (it's even mentionned, in a round about way, there).
Keep up the good work!
Salut, dda:
Thanks for your comments and compliments!
You make some very interesting points, but just to be clear, in regards to "wu," you are describing romanization of Middle Korean, ne-c'est pas?
Martin's Reference Grammar of Korean clearly (and rather surprisingly) states that after "m," "p," "pp," or "ph," ㅜ should be written as "wu" and not "u." (The Reference Grammar is of course for Modern and not Middle Korean.)
I should mention at this point that the UN Romanization of Geographical Names paper (PDF file; mentioned in the footnotes) gives variant Yale spellings for some of the Hangul letters: your "ywu" for ㅠ and "wuy" for ㅟ, both of which are more consistent with how other vowels or digraphs are rendered than is the case for "yu" and "wi." (The dropping of "w" in "wu" is not mentioned.)
Has the Yale Romanization undergone revision since it was first developed, or since Samuel Martin wrote his Reference Grammar? Or are these variants the consequence of someone's not reading the rules carefully? (I would appreciate any insight you can provide on this!)
Anyhow, I suspect writing only "u" after "p," etc. has is in some way connected to something McCune and Reischauer mention in their 1939 paper, which is that around the time of their work—when the Hangûl Match'umbôp was being formed—there was a deliberate shift from in spelling from, for example, 브 to 부. Of course, in the otherwise highly unambiguous Yale Romanization, this can lead to confusion for someone who doesn't know the language, since of course "u" is also used for ㅡ: is "pappuda" 바쁘다 (yes) or 바뿌다 (no, but theoretically possible)?
By the way, M & R's paper is here, as a PDF file. (It's best to download it, open it directly in Acrobat, then select View -> Rotate View -> Clockwise.)
I'll find the citation for the 브 → 부 when I have a chance. (I may have to read through the entire article again to find it!)
Sewing... hope you could help a brother out.
I'm at the point in my studies where I think it is most effective if I just pound myself full of verbs.... problem is I can never find the gospel when it comes to rules (or a simple list would work).
Because I'm learning I just want to pack in the "Standard-Present Tense Form." For example, 하다 is useless for me to learn (in my opinion), because who really learns the infinitive form of that verb while studying a second language. Regardless, 해요 is much more useful for me while I'm on the ground in Korea.
Anyway, do you know of a list.. so far I've only been able to find a list of about 15.
Or the definative rules:
Something like (and I am not sure if I am right at all.. just giving an example of what I'm looking for):
1) 쁘다 (ie: 기쁘다) verbs become 뻐요.
2) 하다 (ie: 좋아해다) verbs become 해요.
3) etc etc...
I'm sure there exceptions and irregular verbs.. but I just need to load up here.
If you can help, thanks.
Hi, Shaun:
There are rules, and fairly clear-cut ones at that. Fortunately, if you leave apart the issue of speech levels and honorifics, verb conjugation is a fairly simple matter when compared to, say, French.
You have now given me the precise idea for my next post. It'll be nice to get back to something concrete and practical. I fear I haven't been helping you, Nathan, or others out by dwelling on arcane issues like era names and Yale romanization!
I'll post something the first moment I get to put anything new up here.
Thanks man. After I get those in I'll start worrying about the honorifics.. but I guess expressing myself a little bit would be be better than flapping my wings in front of them trying to communicate in English.
Sewing, I see you've managed to stir up the old hornets' nest of the romanisation issue over at the KS list. I think the list has been through quite a bit of the MR v. NAKL thing before, but there obviously seems to be quite a bit of mileage left in it yet.
Personally, I'm a partisan of the MR system, but I can't really get excited enough to get into a major debate over it - I'll just continue using the one I like. Funnily enough, I used to work on a Korean government publication so had to enforce the new romanisation when it first came in. Can't say I found it that traumatic.
Kotaji:
Thanks for your comments. Yes, I was a bit surprised that even some of the old veterans have jumped into the debate. Fortunately, it hasn't degenerated into the stuff that I see when I trawl through the archives from 1999-2000!
I'm sympathetic to the new system and would be all for it if it were just not for the issue of "eo" and "eu." I can pronounce them fine, but I still have to read carefully when I see a word like, say, "eomeoni."
Oh wow, what a 'cheem' post on romanisation !
BTW, 'cheem' is Singlish slang for 'complicated' / 심 / 深.
Personally, I prefer to use the government's romanisation rules of 2000.
But I still need to know the old rules bec Korean names (incl those of celebrities) are romanised that way.
Hi, Huangsy:
Thanks for the comments!
(Hmmm, I guess I should check for new comments on my old posts more often....)
Don't you receive an email alert whenever a reply is posted at this blog ?
Don't you receive an email alert whenever a reply is posted at this blog ?
Hi, huangsy:
No, sorry, as you can see, I don't. I switched that option off. As a result, here I am, 2 months later, replying to your last comment....
Post a Comment
Links to this post:
Create a Link
<< Home