Government Announces New System of Romanization
for the Korean Language

The Ministry of Culture and Tourism of the Republic of Korea formally proclaimed a new system of Romanization for the Korean Language on July 4, 2000. The new Romanization system is the result of five years of effort and changes the previous official government Romanization system proclaimed in 1984. All geographical names will follow this new system without exception. Previously used Romanizations of personal and company names may continue to be used. However, the Ministry of Culture and Tourism will strongly encourage people to follow the new system when Romanizing personal names for the first time. Since all road signs cannot be changed at once, time will be given until the end of 2005 to make this change. The McCune-Reischauer System of Romanization is widely used overseas, but confusion has often resulted as the breve and apostrophe are frequently abbreviated. This and other difficulties in usage made revision of the official government system of Romanization unavoidable. The Government of the Republic of Korea will work consistently to see that this new system of Romanization for the Korean Language comes to be used widely around the world.

 

The Why and How of the New Romanization System
for the Korean Language

I. Contents of the New Romanization System

  • When ¤¡, ¤§, ¤², and ¤¸ appear as the first letter of a word, they will be represented by g, d, b, and j instead of k, t, p, and ch. This means that ±¤ÁÖ, ´ë±¸, ºÎ»ê, and Á¦ÁÖ become Gwangju, Daegu, Busan, and Jeju.
  • ¤», ¤¼, ¤½, and ¤º are Romanized as k, t, p, and ch instead of k', t', p', and ch'. ÅÂÆò·Î, Æ÷Ç× and ûÁÖ become Taepyeongno, Pohang and Cheongju.
  • The vowels ¾î, À¸, ¿©, and ÀÇ are Romanized as eo, eu, yeo, and ui instead of , , y and i. ÀüÁÖ, ¿µµîÆ÷ and Á¦ÁÖ become Jeonju, Yeongdeungpo and Jeju.
  • ½Ã is written si instead of shi. ½Å¶ó and ÀÓ½Ç become Silla and Imsil.
  • ¤¹ is represented by jj.
In short, the proposal attempts to maintain the larger framework of the old Romanization system. Firstly, more emphasis was given to how Korean is pronounced than to how it is spelled in Hangeul, this out of consideration for persons who do not speak Korean. Secondly, the Romanization for the consonants ¤¡, ¤§, ¤² differ from the old system when they appear as the first letter of a word, but they remain unchanged from the old system when they appear as the last letter in a word. Thirdly, all vowels other than ¤Ã, ¤Ñ, ¤Å, and ¤Ò are to be Romanized in the same fashion as the old system.

II. The Need for a New Romanization System

The Ministry of Culture and Tourism has worked since 1995 on a proposal for a new system of Romanization for the Korean language. The old official Republic of Korea Romanization system is based on, and almost entirely similar to, the 1939 McCune-Reischauer Romanization for the Korean Language. Unfortunately, the breve ( , ) and apostrophe (k', t', p', ch') in the McCune-Reischauer and old official Romanization system have greatly impeded accurate and consistent usage.

The breve that goes over the Roman characters o and u to form and is included in the old Romanization system because the Korean language makes clear distinction between the Korean vowels ¾î and ¿À, and À¸ and ¿ì, even though there may often seem to be little difference to native speakers of languages that afford no sensitivity to these sounds. The same is the case with k', t', p', and ch'. While to speakers of languages which are not especially sensitive to the differences between ¤¡/¤», ¤§/¤¼, ¤²/¤½, and ¤¸/¤º, the characteristics of the Korean language make the distinction essential, not only because it is possible to make this distinction in a Romanization system, but because without this distinction, much confusion arises as a result.

The greatest problem of the old system is that, while it is followed on road signs and in some academic writings, in a considerable number of cases it is not followed with consistency. The breve and the apostrophe are often abbreviated completely.

In the case of personal names, perhaps the single most frequently Romanized of all subjects, the breve and apostrophe are ignored completely. The breve and apostrophe are not allowed in passports. As a result there exist two kinds of people; those who ignore diacritical marks but otherwise follow current Romanization policy where possible, and those who use Roman characters in any which way they may feel led, without consistency or any attempt to follow any previously established system of Romanization.

Let us consider what happens when people "abbreviate" the breve and apostrophie. In Korean, entire words, an endless number of words, differ according whether they include the vowel ¾î and ¿À, Some easy examples would be Shinch' n (½Åõ) and Shinch'on (½ÅÃÌ), nyang (¾ð¾ç) and Onyang (¿Â¾ç), Y ngshik (¿µ½Ä) and Yongshik (¿ë½Ä), and Y ngsu (¿µ¼ö) and Yongsu (¿ë¼ö). The Korean language can simply not afford to confuse the vowels ¾î and ¿À because of the frequency of both. What happens when the apostrophe is left out as is more often the case? ¤¡, ¤§, ¤², ¤¸ each become confused with ¤», ¤¼, ¤½, ¤º respectively. Taey ng (´ë¿µ) and T'aey ng (Å¿µ), Taeshik (´ë½Ä) and T'aeshik (ŽÄ), Taesu (´ë¼ö) and T'aesu (żö) are a few simple examples of the confusion that arises.

Since it is so difficult for most people to write ¾î as , many write ¾î as u or eo. The Korean vowel À¸ ( in the old system) presents the same difficulty, and many either leave the breve off and or write it u or as eu. People use Roman letters to write every which way they feel led precisely because the old system is so difficult to follow at the most practical level. People had little choice but to not follow the old government system of Romanization. If we are ever to expect any consistency in the Romanization of Korean, the breve and apostrophe must be omitted.

Many non-native Korean speakers do not feel the need to differentiate between the consonants ¤¡/¤», ¤§/¤¼, ¤²/¤½, and ¤¸/¤º, and the vowels ¾î/¿À. It is often these individuals who suggested leaving out the breve and apostrophe out of the old Romanization system, but that the old system should have been otherwise left untouched. Unfortunately this is an avoidance of the issues at hand.

It is not difficult to find brothers who's names are, for example, Y ngsu (¿µ¼ö) and Yongsu (¿ë¼ö). When the breve is eliminated, both become Yongsu. This is simply unimaginable to native speakers of Korean. Even for those who are non-native Korean speakers or speak no Korean at all, writing Y ngsu and Yongsu, Shinch' n and Shinch'on, and nyang and Onyang exactly the same way causes a considerable inconvenience.

The consonants ¤¡/¤», ¤§/¤¼, ¤²/¤½, and ¤¸/¤º simply must be differentiated. The question then becomes how to make the distinction. Some foreigners, Westerners in particular, may not feel the need to distinguish between these consonants. However, there are numerous cases where names and words become the same when something is not done to make this necessary distinction. Taey ng (´ë¿µ) and T'aey ng(Å¿µ) both become Taeyong. Changny (Àå³à) and ch'angny (â³à) both become changnyo. This makes it unacceptable to use t to represnt both ¤¡ and ¤», and to use ch to repreesnt both ¤¸ and ¤º. What alternatives remain? When a native speakers of Korean hears most foreigners pronounce g, d, b, and j, it sounds most close to ¤¡, ¤§, ¤², and ¤¸, and the foreigner's pronunciation of k, t, p, ch sounds most close to ¤», ¤¼, ¤½, and ¤º. The best possible answer, then, becomes clear.

We recognize that to most Westerners, when ¤¡, ¤§, ¤², and ¤¸ are pronounced as they are when they appear as the first letter in the first syllable of a word, they usually sound like they should be written as k, t, p, and ch. The problem with writing ¤¡, ¤§, ¤², and ¤¸ this way is that, unless the apostrophe is included with consistency and precision, there then becomes no way to tell the difference between these and the consonants ¤», ¤¼, ¤½, and ¤º. When the first letter in a word, ¤¡, ¤§, ¤², and ¤¸ may sound close to k, t, p and ch, but when they are represented in the first letter position as g, d, b, and j, it then becomes possible to differentiate between these consonants and ¤», ¤¼, ¤½, and ¤º.

In addition, the old Romanization system required ¤¡, ¤§, ¤², and ¤¸ to be written as k, t, p, and ch when they appear as the first letter in a word, but then required that the same characters be written as g, d, b, and j when ¤¡, ¤§, ¤², and ¤¸ appear between two voiced sounds. This makes consistent Romanization extremely difficult for native speakers of Korean, who usually cannot see why these same and frequent consonants must be written differently. Romanization is not exclusively for those who cannot speak Korean. Native speakers of Korean themselves frequently have reason to Romanize Korean as well. It is our hope that non-native speakers of Korean might understand how utterly difficult it is for native speakers of Korean to know to Romanize these Han-geul characters according to whether they are voiced or non-voiced.

We would hope also that it be widely understood that this new official government Romanization system was not made with consideration given only to native speakers of Korean. It would be much easier for Koreans to transliterate one Hangeul letter for one Roman letter. For example, it would be easiest for native Korean speakers to Romanize with consistency if they wrote Sinla for ½Å¶ó, and Soglisan for ¼Ó¸®»ê. But since the end results of Romanization are more frequently referred by non-native speakers of Korean, consideration for the convenience of foreigners was given in the way that the new system respects pronunciation more than consistency with how a word is spelled in Hangeul.

III. Appeal

Given the possibilities for systems of Romanization, the differing needs of those who must Romanize Korean, and the differing sensitivities of native speakers of Korean, non-native speakers of Korean, and those who do not speak Korean at all, there will always be differing positions about what directions the Korean government's official Romanization system should take. If a Romanization system respects exclusively the position of most non-native Western speakers of Korean, then it becomes so difficult and inconvenient for native speakers of Korean that it is virtually impossible to follow with consistency (such as is the case with differentiating between voiced and non-voiced consonants). A Romanization system would be easiest for native speakers of Korean if it was a system of transliteration, one that requires the same Roman character for the same Hangeul character in all positions. Such a system would never succeed, since it would be mostly ignored by persons who do not speak Korean. The designing and choosing of a Romanization system, then, cannot exclusively follow the preferences of either major position. Without concessions from both positions, the problems of inconsistency will remain unsolved forever. The National Academy of the Korean Language believes that this revision proposal is an excellent compromise, and that it is the best proposal imaginable.

 

The Romanization of Korean
  1. Basic Principles of Romanization

    1. Romanization is based on standard Korean pronunciation.

    2. Symbols other than Roman letters are avoided to the greatest extent possible.

  2. Summary of the Romanization System

    1. Vowels are transcribed as follows:

      simple vowels
      ¤¿¤Ã¤Ç¤Ì¤Ñ¤Ó¤À¤Ä¤Ê¤Ï
      aeooueuiaeeoewi

      diphthongs
      ¤Á¤Å¤Ë¤Ð¤Â¤Æ¤È¤É¤Í¤Î¤Ò
      yayeoyoyuyaeyewawaewoweui

      Note 1: ¤Ò is transcribed as ui, even when pronounced as ¤Ó.
      Note 2: Long vowels are not reflected in Romanization.

    2. Consonants are transcribed as follows:

      plosives(stops)
      ¤¡¤¢¤»¤§¤¨¤¼¤²¤³¤½
      g,kkkkd,ttttb,pppp

      affricates
      ¤¸¤¹¤º
      jjjch

      fricatives
      ¤µ¤¶¤¾
      sssh

      nasals
      ¤¤¤±¤·
      nmng

      liquids
      ¤©
      r,l

    Note 1 : The sounds ¤¡, ¤§, and ¤² are transcribed respectively as g, d, and b when they appear before a vowel; they are transcribed as k, t, and p when followed by another consonant or form the final sound of a word.(They are Romanized as pronunciation in [ ].)
    [Example] ±¸¹Ì Gumi ¿µµ¿ Yeongdong ¹é¾Ï Baegam ¿Áõ Okcheon ÇÕ´ö Hapdeok È£¹ý Hobeop ¿ù°ù[¿ù°ð] Wolgot º¢²É[¹ù„Œ] beotkkot Çѹç[ÇѹÞ] Hanbat

    Note 2 : ¤© is transcribed as r when followed by a vowel, and as l when followed by a consonant or when appearing at the end of a word. ¤©¤© is transcribed as ll.
    [Example] ±¸¸® Guri ¼³¾Ç Seorak Ä¥°î Chilgok ÀÓ½Ç Imsil ¿ï¸ª Ulleung ´ë°ü·É[´ë°ý·É] Daegwallyeong

  3. Special Provisions for Romanization

    1. When Korean sound values change as in the following cases, the results of those changes are Romanized as follows:

      1. The case of assimilation of adjacent consonants
        [Example] ¹é¸¶[¹ð¸¶] Baengma ½Å¹®·Î[½Å¹®³ë] Sinmunno Á¾·Î[Á¾³ë] Jongno ¿Õ½Ê¸®[¿Õ½É´Ï] Wangsimni º°³»[º°·¡] Byeollae ½Å¶ó[½Ç¶ó] Silla

      2. The case of the epenthetic ¤¤ and ¤©
        [Example] Çп©¿ï[Ç׳à¿ï] Hangnyeoul ¾Ë¾à[¾Ë·«] allyak

      3. Cases of palatalization
        [Example] ÇØµ¸ÀÌ haedoji °°ÀÌ[°¡Ä¡] gachi ¸ÂÈ÷´Ù[¸¶Ä¡´Ù] machida

      4. Cases where ¤¡, ¤§, ¤², and ¤¸ are adjacent to ¤¾
        [Example] ÁÁ°í[Á¶ÄÚ] joko ³õ´Ù[³ëŸ] nota ÀâÇô[ÀÚÆì] japyeo ³ºÁö[³ªÄ¡] nachi
        However, aspirated sounds are not reflected in case of nouns where ¤¾ follows ¤¡, ¤§, and ¤², as in the examples below.
        [Example] ¹¬È£ Mukho ÁýÇöÀü Jiphyeonjeon

      Note: Tense (or glottalized) sounds are not reflected in cases where morphemes are compounded, as in the examples below.
      [Example] ¾Ð±¸Á¤ Apgujeong ³«µ¿°­ Nakdonggang Á׺¯ Jukbyeon ³«¼º´ë Nakseongdae ÇÕÁ¤ Hapjeong ÆÈ´ç Paldang »ûº° saetbyeol ¿ï»ê Ulsan

    2. When there is the possibility of confusion in pronunciation, a hyphen '-' may be used.
      [Example] Áß¾Ó Jung-ang ¹Ý±¸´ë Ban-gudae ¼¼¿î Se-un ÇØ¿î´ë Hae-undae

    3. The first letter is capitalized in proper names.
      [Example] ºÎ»ê Busan ¼¼Á¾ Sejong

    4. Personal names are written by family name first, followed by a space and the given name. In principle, syllables in given names are not separated by hyphen, but the use of a hyphen between syllables is permitted.
      [Example] ¹Î¿ëÇÏ Min Yongha (Min Yong-ha) ¼Û³ª¸® Song Nari (Song Na-ri)

      ¨ç Assimilated sound changes between syllables in given names are not transcribed. [Example] ÇѺ¹³² Han Boknam (Han Bok-nam) È«ºû³ª Hong Bitna (Hong Bit-na)
      ¨è Romanization of family names will be determined separately.

    5. Administrative units such as µµ, ½Ã, ±º, ±¸, À¾, ¸é, ¸®, µ¿, and °¡ are transcribed respectively as do, si, gun, gu, eup, myeon, ri, dong, and ga, and are preceded by a hyphen. Assimilated sound changes before and after the hyphen are not reflected in Romanization.
      [Example] ÃæÃ»ºÏµµ Chungcheongbuk-do Á¦ÁÖµµ Jeju-do ÀÇÁ¤ºÎ½Ã Uijeongbu-si ¾çÁÖ±º Yangju-gun µµºÀ±¸ Dobong-gu ½ÅâÀ¾ Sinchang-eup »ïÁ׸é Samjuk-myeon Àοո® Inwang-ri ´ç»êµ¿ Dangsan-dong ºÀõ1µ¿ Bongcheon 1(il)-dong Á¾·Î 2°¡ Jongno 2(i)-ga Åð°è·Î 3°¡ Toegyero 3(sam)-ga

      Note: Terms for administrative units such as ½Ã, ±º, À¾ may be omitted.
      [Example] ûÁֽà Cheongju ÇÔÆò±º Hampyeong ¼øÃ¢À¾ Sunchang

    6. Names of geographic features, cultural properties, and man-made structures may be written without hyphens.
      [Example] ³²»ê Namsan ¼Ó¸®»ê Songnisan ±Ý°­ Geumgang µ¶µµ Dokdo °æº¹±Ã Gyeongbokgung ¹«·®¼öÀü Muryangsujeon ¿¬È­±³ Yeonhwagyo ±Ø¶ôÀü Geungnakjeon ¾È¾ÐÁö Anapji ³²Çѻ꼺 Namhansanseong È­¶û´ë Hwarangdae ºÒ±¹»ç Bulguksa ÇöÃæ»ç Hyeonchungsa µ¶¸³¹® Dongnimmun ¿ÀÁ×Çå Ojukheon Ã˼®·ç Chokseongnu Á¾¹¦ Jongmyo ´Ùº¸Å¾ Dabotap

    7. Proper names such as personal names and those of companies may continue to be written as they have been previously.

    8. When it is necessary to convert Romanized Korean back to Hangeul in special cases such as in academic articles, Romanization is done according to Hangeul spelling and not pronunciation. Each Hangeul letter is Romanized as explained in section 2 except that ¤¡, ¤§, ¤², ¤© are always written as g, d, b, l. When ¤· has no sound value, it is replaced by a hyphen may also be used when it is necessary to distinguish between syllables.
      [Example] Áý jib ¤ jip ¹Û bakk °ª gabs º×²É buskkoch ¸Ô´Â meogneun µ¶¸³ doglib ¹®¸® munli ¹°¿³ mul-yeos ±»ÀÌ gud-i ÁÁ´Ù johda °¡°î gagog Á¶¶û¸» jolangmal ¾ø¾ú½À´Ï´Ù eobs-eoss-seubnida

Additional Provisions

  1. This system of Romanization becomes effective on the date it is formally proclaimed.
  2. Signs using the previous system of Romanization (road signs, official large-scale notices, information posted at cultural sites, etc.) when this system of Romanization becomes effective must follow this system before December 31, 2005.
  3. Publication such as textbooks using the previous system of Romanization must follow this system by February 28, 2002.