[KS] Romanization systems

Stefan Ewing sa_ewing at hotmail.com
Thu Jun 2 04:10:29 EDT 2005

Dear KS readers:

First of all, I would again earnestly like to apologize for my flood of 
posts 24 hours ago.  Please accept as my excuse my naive excitement at 
discovering the KS list, recognizing by name alone so many famous (in the 
world of KS, anyway) Koreanologists whose books or articles--or about 
whom--I have had the privilege of reading over the years.

Thank you to Gari Ledyard and Ross King for your informative replies.  Now 
that both Dr. Ledyard and Edward J. Shultz have mentioned the same _Korean 
Studies_ article, I will definitely read it.  Alas, old back issues are not 
online as is the case for _Korea Journal_, so I guess I'll have to go out to 
UBC (hello, Drs. Baker, King, et al.) to find it.  (Good: I have an excuse 
to dive into the all the wonderful books at the Asian Library again!)

Drs. Ledyard's and King's mention of the shift from /pu/ to /pwu/ in the 
1930s is borne out by what McCune and Reischauer mentioned in their RASKB 
paper, written at a time when such events were still fresh in people's 
memories.  And thanks to Dr. King for pointing out the difficulties of using 
anything but the Yale system for pre-1933 orthography.

On the general matter of romanization, I agree with Gill Goddard's and Jim 
Hoare's comments.  In disseminating information on Korean language, culture, 
history, etc. to a general audience--be it through books, journal or 
newspaper articles, or online media--there is a need to transcribe or 
transliterate Korean in the target language.  Indeed, as Ms. Goddard has 
pointed out, there may even be technical reasons that make transcription (or 
transliteration) unavoidable.  (As is the case whenever I attempt to type 
Hangul in Hotmail, alas, the retransmitted output being completely illegible 
in any character encoding.)

In fact, I raised these arguably arcane issues not out of any mere idle 
curiosity, but because I want my own writing on Korea's language and culture 
to be broadly accessible as possible (my penchant for grammatical arcana 
notwithstanding), making romanization a necessity.

The challenges are in (a) choosing a system to use, and (b) ensuring that 
whichever system one chooses, words are rendered faithfully.  One has no 
desire to lead readers astray by, for example, claiming to use 
McCune-Reischauer romanization and then writing "shi" for "si" (the thorny 
issue of palatalization aside) or purporting to use the NAKL Revised system 
and then writing the egregious "euiweon" for "uiwon."  Unless modifications 
are deliberate and clearly motivated (as in the case of Dr. King's example 
of writing "wu" after labials), the relationship between Korean orthography 
and pronunciation is so complex and so easily encourages romanized 
misspellings that readers should not be encouraged to perpetuate them 
through misleading examples!

The most problematic and contentious issue is, of course, which romanization 
system to use, as we are certainly spoiled for choices.  I somewhat like the 
current NAKL system, but correct use of it (especially the poor, 
much-derided /eo/ and /eu/) almost demands some knowledge of Hangul 
pronunciation and word formation.  I would agree from first-hand observation 
with Ms. Goddard that /eo/ is quite consistenly mispronounced in exactly the 
same way ("ee-oh" or "ee-uh") by English speakers.  On top of that, in 
English, /g/ followed by /e/ or /i/ is generally pronounced as an affricate 
("soft g") as opposed to a velar ("hard g"); thus, the sequence /geo/ (Yale 
/ke/)--so frequent in Korean--is in particular especially problematic for 
English speakers (I can't say with certainty for other languages).  Quite 
apart from other, trickier matters (like the sound quality of initial 
consonants), issues like this lead one to conclude that McCune-Reischauer is 
somewhat better than the Revised system at accurately representing Korean 

On a final note, I would submit that while Hangul is clearly the best medium 
for writing Korean and quite elegant in its design and appearance, Hangul 
usage is not without its complications.  Modern Hangul orthography places a 
premium on preserving morphemic structure--which is tremendously helpful in 
discerning patterns of grammatical construction and word formation--at the 
cost of a sometimes elaborate relationship between spelling and punctuation. 
  Certainly, there are the regular, well-documented rules that beginning 
students learn early to surmount this problem, such as the changing of the 
final sound /k/ to /ng/ in front of /n, l, m/.  But then there are more 
obsure rules, such as the pronunciation of (Yale) /namuq.iph/ (?) as 
(McC.-R.) /namunnip/.  And what of the intensification of mid-word initial 
consonants?  Here, exceptions and irregularities seem to abound.  Thus, (to 
use McC.-R.) /ch^onja/ ("transliteration") and /hancha/ both have the same 
sequence -n'j- in Hangul, which is pronounced -nj- in the former case but 
-nch- (or -ntch- if one consults the Hangul pronunciation gloss in a Korean 
dictionary) in the latter.

Stefan Ewing

>From: gkl1 at columbia.edu
>Reply-To: Korean Studies Discussion List <Koreanstudies at koreaweb.ws>
>To: Korean Studies Discussion List <Koreanstudies at koreaweb.ws>
>Subject: RE: [KS] Romanization systems
>Date: Wed, 01 Jun 2005 16:08:48 -0400
>I guess it's necessary to remind folks that the reason we have
>romanization at all is to help people who do not know Korean, such
>as foreign visitors to Korea who need to know the names of towns
>and other information on highway signs, in hotels, etc., and to
>make it convenient for scholars, journalists, novelists, etc., to
>write about Korea in English. No one seriously believes that Korean
>should be taught using romanization beyond perhaps the first two or
>three lessons. That was absolutely not the intention of
>McCune-Reischauer or Samuel Martin (who created Yale). Much less
>helpful is the suggestion that people simply learn Hangul and be
>weaned from romanization. We also see it opined in a posting that
>the International Phonetic Alphabet, with all its diacritics and
>special symbols, should be a single, universal romanization for all
>languages. I'm afraid not. In recent days we've even seen
>suggestions that the term "romanization" itself, as applied in any
>non-Romance language, is a usurpatious affront. Those Romans were
>terribly undiscriminating to let all those northerners use their
>    Admittedly, Americans of English speech have pioneered the older
>systems, and in most respects McC-R favors English phonetics. But
>there's no law saying that McC-R must be international, and nations
>or individuals who would be comfortable with systems that favor,
>say, French, German, or Spanish, are certainly free to invent
>methods comfortable for them. On the other hand, some scholars from
>each of those countries have of their own volition used McC-R, even
>when writing in their own langauges, and that's also their right.
>    I'll turn now to the more sober and practical questions raised by
>Stefan Ewing. He asks, 'Is it possible that the [McC-R] system has
>"officially" changed since 1939, or is this [si/shwi] practice due
>to a misconception?' McC-R, in use now for 66 years, has by its
>sheer history earned something like the status of
>"standard," but of course it has its problems and I'm sure people
>will never stop trying to create something better. Fine. But it is
>not an "official" system in the sense that some governing body
>makes rules or revisions in them. The only rules are the ones that
>McCune and Reischauer published in 1939. They made them in order to
>be of help to us all. They were duly humble before the difficult
>problems involved. They made no demands. Various people have
>suggested more or less minor revisions and useful
>refinements(notably and most competently, a group led by Sam Martin
>and the late Robert Austerlitz, who conducted a romanization
>workshop in 1980, the report of which was published in <Korean
>Studies> (Hawaii), vol 4, pp 111-125). Although the senior status
>of the scholars at that conclave gave them a certain authority,
>they only made suggestions and recommendations. Nobody owns McC-R.
>Its basic principles have been maintained and should be maintained,
>but within that limitation flexibility in practice has been the rule
>from the beginning. Thank heavens it is not an "official" system. On
>the one hand, people are trusted to be free; on the other a sensible
>and tolerant consensus prevails. Bureaucrats now control (or attempt
>to) the system that committees appointed by Korean government
>agencies have created. There's no one in the world that can say
>they have no right to do this, but all the rest of us have the
>right to accept or reject it on our own choice. As for the Yale
>System, I suppose its creator Sam Martin could change its details
>if he wanted to, but after many years it hasn't changed. I don't
>think he would claim any "official" status for it.
>    On the specific question of Mc-R <si> and <shwi> vs. <shi> and
><shwi>, there has also been much debate in the past. My own view is
>that neither is ideal. Both si- and shi- are off the actual Korean
>sound, at least in the mouths of most English speakers. But I don't
>think the basic problem lies in s- or sh-. I think it lies with the
>vowel -i-, which is highly fronted and palatalized. I would favor
><syi-> and <sywi->, adapting the vowel rather that the consonant.
>I've championed this in earlier romanization debates without
>attracting much if any support for the idea. Faced with that, my
>practice, faute de mieux, is to stay with McC-R's awkward
><si->/shwi->. One might review the thread on this list in the
>archive for April 1996. (There are also other extensive list
>discussions on romanization during 1997 and 1999.)
>    As for the Yale prescription that the vowel <wu> (the 7th vowel
>in the Standard Korean order, <u> in Mc-R) following a labial
>consonant (<m,p,pp,ph>) should drop the w-, that is indeed the
>practice and it has not been changed. It would seem that this rule
>erases the distinction between Yale's <wu> and <u> (the 7th and 9th
>vowels in the Standard Korean order, <u> and <^u>) in Mc-R), but
>Yale's point is that after these labial consonants there is in fact
>no distinction in the spoken pronunciation, and Yale, on the
>principle of simplicity, eliminates the w-. Phonemically that makes
>perfect sense, but unfortunately the unification upsets Yale's
>advantage in always enabling the restoration of the Hangul
>orthography, one of its many virtues. My own view would be, in
>cases where restorability of the original Hangul is a priority,
>that one "violate" Yale's rule and stay with <wu> even after the
>labial consonants. The original Middle Korean orthography wrote
>words like <puk>, "north," using the 9th vowel symbol (McC-R <^u>,
>and correctly so in terms of the historical Sino-Korean phonology.
>This practice lasted pretty much down to the beginning of the 20th
>century. During that period, as the "arae-a" gradually disappeared
>as a discrete vowel, the <^u> vowel moved higher and came to be
>confused with the 7th vowel symbol (<u>) after labials. The 1933
>orthographic reform mandated spelling words analagous to <puk> with
>the latter, and it has been so ever since.
>    On the problem of <wi> and <yu> vs. <wuy> and <ywu>, I wasn't
>aware of it. One can understand the rationalization, but for
>general purposes I see no reason to change from the former.
>    In addition to the useful discussion of romanization history and
>issues in Austerlitz et al., 1980, cited above, one might refer to
>an interesting little publication of the Royal Asiatic Society,
>Korea Branch (Seoul), <Guide to Romanizing Korean>, compiled by
>John Holstein and John Harvey, 1999. Those in Korea should find it
>easily obtainable.
>Gari Ledyard
>Quoting Stefan Ewing <sa_ewing at hotmail.com>:
> > Hi, everyone:
> >
> > In the absence of responses, I'll ask some specific questions
> > that are still
> > up in the air.  I would appreciate any and all replies that this
> > esteemed
> > crowd can provide.
> >
> > 1.  As I recall reading in McCune and Reischauer's 1939 RASKB
> > paper, "s" is
> > only to be written as "sh" before "wi."  Thus, "si," for example,
> > should
> > *not* be written as "shi."  The 1984 South Korean
> > romanization--which is
> > based on McCune-Reischauer--does however advise that not only "s"
> > but also
> > "ks," "ls," "ps," and "ss" should take an "h" before *both* "wi"
> > *and* "i."
> > (This rule apparently does not extend to "ya," "yae," "yo-breve,"
> > "ye,"
> > "yo," or "yu"; thus, "syop'ing" (shopping) should evidently be
> > written as
> > such and not as "shop'ing.")
> >
> > As a result of the 1984 system's influence, it appears that some
> > believe M-R
> > also advises writing "si" as "shi."  Is it possible that the
> > system has
> > "officially" changed since 1939, or is this practice due to a
> > misconception?
> >
> > 2.  Samuel E. Martin's _Reference Grammar of Korean_ states that
> > "wu" should
> > be written as "u" after "m," "p," "pp," or "ph."  I can see the
> > rule in
> > front of me in the book with my own eyes, but can anyone confirm
> > that this
> > practice is adhered to?  (I suspect--without anything to back me
> > up--that
> > this practice is somehow related to something M & R wrote in
> > their 1939
> > paper, which mentions that around the time of the formation of
> > the _Hangul
> > Match'umbop_, there was a shift in Hangul orthography from, for
> > example "p
> > u-breve" to "p u.")
> >
> > Also, the same book states quite clearly what "wi" and "yu"
> > should be
> > written as such.  In a recent paper by the UNGEGN Working Group
> > on
> > Romanization Systems (a UN committee), however, these are written
> > as "wuy"
> > and "ywu" respectively.  Has the Yale system in fact changed, or
> > would this
> > have been a case of someone's not reading the rules very
> > carefully, and
> > assuming that these two vowels or vowel combinations should
> > follow the same
> > pattern as analogous vowels or vowel combinations (e.g., "ay,"
> > "ya")?
> >
> > 3.  Improbably, the 2000 Revised Romanization of Korean makes
> > absolutely no
> > provision for the treatment of kyoppatch'im: the non-twinned
> > double
> > consonants ("ks," "nj," "nh," "lg," ..., "lh," "ps") that appear
> > at the ends
> > of some syllables.  Thus, I have only been able to make educated
> > guesses on
> > what the spelling of these should be in the medial and final
> > positions in
> > Revised Romanization, using the Revised Romanization
> > transcription rules for
> > other letters and McCune-Reischauer's rules for kyoppatch'im (and
> > also
> > following the Korean "final consonant rule" (patch'im kyuch'ik)).
> >  I have
> > put in a question to the National Academy of the Korean Language
> > (they
> > usually reply within a couple of days).  In the meantime, can
> > anyone provide
> > any insight on this matter?  (I have downloaded a program I found
> > for
> > converting Hangul to Revised Romanization, so I'll see if that
> > answers my
> > question.)
> >
> > Eagerly awaiting your replies,
> > Stefan Ewing
> >

Scan and help eliminate destructive viruses from your inbound and outbound 
e-mail and attachments. 
  Start enjoying all the benefits of MSN® Premium right now and get the 
first two months FREE*.

More information about the Koreanstudies mailing list