's first romanization system for Korean, published 1832 Possibly the earliest romanization system for Korean was an 1832 system by the German doctor
Philipp Franz von Siebold, who was living in
Japan. Another early romanization system was an 1835 unnamed and unpublished system by missionary
Walter Henry Medhurst that was used in his translation of a book on the
Chinese,
Korean, and
Japanese languages. Medhurst's romanization scheme was otherwise not significantly used. In 1874, the Dallet system was introduced; it was based around French-language
phonology. It was the first to use the digraphs
eo and
eu, and the first to use
diacritics for Korean romanization; it used the
grave and
acute accents over the letter "e". The first system to see significant usage was the Ross system, named for
John Ross, which was designed in 1882. It saw adoption by missionaries. In 1897,
James Scarth Gale introduced his system in his work
A Korean-English Dictionary. This system went on to achieve some adoption; it was reportedly adopted by the
Harvard Journal of Asiatic Studies. In spite of this, some scholars found issues with these early systems. More systems by Westerners emerged, based on English, French, and German phonology. Japanese scholars also developed their own romanizations for Korean, many of which were built on the work of Siebold and Dallet. In 1933, the first romanization system developed by Koreans, which was appended to the
Unified Han'gŭl Orthography System, was promulgated by the Korean Language Society. In 1935, published "The International Phonetic Transcription of Korean Speech Sounds". Systems continued to be developed to address various perceived shortcomings in other systems. By 1934, according to Japanese linguist
Shinpei Ogura's count, there were at least 27 extant systems. Whereas
Hepburn romanization had already become the widely accepted standard for the
romanization of Japanese by the 1930s, Korean continued to lack such a standard. This led to significant diversity and inconsistencies in romanizations, not only between scholars but reportedly even within the writings of individual authors.
Challenges for developing a standard romanization The task of developing a standard romanization scheme for Korean was complicated by a number of factors. Even into the 20th century, there were significant variations in the pronunciation and spelling of the Korean language and Hangul respectively, often due to the
dialects of Korean. Attempts were made to standardize the Korean language, but these efforts were made by multiple authorities. Two rivaling societies for standardizing Korean emerged: the
Korean Language Society () and the
Chŏson Ŏhak Yŏn'guhoe (); they published separate guidances. Eventually, the Korean Language Society's standard became the basis for the standards of both North and South Korea. Other references for spelling included those used in Gale's dictionary, guidances from the
Government-General of Chōsen, and a French dictionary. Other challenges were fundamental to properties of the Korean language and script, which make the language not easily mappable onto the Latin script. McCune and Reischauer claimed in 1939 that there are eight to ten vowels in Korean (this topic was still debated by that point). As there are only five vowels in the Latin script, the other vowel sounds had to be rendered either using multiple letters in the form of
digraphs (e.g.
eo for ) or by using
diacritics. Also, in many cases, pronunciation does not exactly match what is written in Hangul; similar phenomena occurs with all other major scripts as well. For example, due to
linguistic assimilation, the state
Silla is written in Korean as (), but pronounced . Some challenges were social and geopolitical. Reportedly, early scholars often wrote about Korea from
Sinocentric or Japanese perspectives; Korean place names were often rendered using pronunciations from the Chinese or Japanese languages. Furthermore, after
Korea went under Japanese rule, the "official" names of many places were considered to be those in the Japanese language. In addition, the
Japanese colonial government implemented various restrictions on the use of the Korean language around the mid-1930s; the
Korean Language Society was also persecuted in one incident. Regardless of romanization systems, many Koreans chose and continue to choose to spell their names in Latin script in an
ad hoc manner. For example,
이/리 (李) is variously romanized as
Lee,
Yi,
I, or
Rhee. In some cases, single families romanized their surnames differently on
South Korean passports. For example, within a single family, a father's surname was rendered as "Shim" and the son's as "Sim".
McCune–Reischauer McCune–Reischauer (MR) is a system that was first introduced in 1939, in the journal
Transactions of the Royal Asiatic Society Korea Branch. It is named for
George M. McCune and
Edwin O. Reischauer; the two developed the system together in consultation with Korean linguists
Choe Hyeon-bae, , and . The system has proved controversial with especially native Korean speakers. It had been developed mainly for use in Western academia, and reflected pronunciation rules that many Koreans were not consciously aware of, as they are not reflected in Hangul. Linguist Robert J. Fouser argued that another point of contention was related to nationalism; some disliked that the system had been developed by foreigners during the Japanese colonial period, and wanted a natively developed alternative.
After the liberation and division of Korea With 1945 came the
liberation of Korea, as well as
its division.
Both Koreas began to develop separate language standards. South Korea adopted MR in 1948. According to Reischauer, McCune "persuaded the American
Army Map Service to adopt [the McCune–Reischauer system], and through the Korean War it became the foundation for most current Romanizations of Korean place names". Just after the 1950–1953
Korean War, romanization was seen as a minor concern, compared to improving domestic literacy in Hangul. Meanwhile, romanization systems continued to emerge; by 1997, there were more than 40 romanization systems. In 1956, North Korea became the first of the two Koreas to design a new official romanization system. This system combines features of the Dallet and 1933 Unified systems. It was revised in 1986. In 1959, the published a romanization system, which has since been dubbed the
Ministry of Education system (MOE). The system received immediate backlash, mainly from foreigners. Fouser evaluated the system as prioritizing use for Koreans; it had a one-to-one correspondence from Hangul to Latin script, and did not account for the pronunciation changes that Hangul itself did not reflect. The system also tended to produce romanizations that bore superficial resemblance to words in English, some of which were seen as odd or humorous, such as
Dogribmun (), which superficially evokes the
ribs of
dogs. The Ministry of Education met in 1978 and 1979 and drafted several alterations to the system, although these did not come to pass allegedly because of political turmoil around that time. Eventually, the South Korean government began considering whether to use a more foreigner-friendly system in anticipation of the
1988 Summer Olympics, which were to be held in Seoul. Various attempts were made to measure objective and subjective metrics of the various systems, for example how frequently systems deviated from expected pronunciation or which systems produced the most accurate pronunciations by foreigners. In 1984, a slightly revised version of McCune–Reischauer was adopted, to pushback from Koreans. In 1968,
Samuel E. Martin introduced a system that has since been dubbed the
Yale system. The system became widely adopted by the international academic linguistics community, although few others adopted it. Fouser argues that while the system allowed for
reversibility, it is "unsightly", is suited to those who already know Hangul, and does not adequately communicate pronunciation, even in comparison to the MOE system.
Computer age With the spread of
computers and the
Internet by the 1980s and 1990s, complaints about MR grew. The breves used in MR are not easily accessible on a standard keyboard. Some took to replacing the breve with alternate characters or simply omitting it altogether; the diversity of practice and ambiguity if breves were not used led to confusion. In 1986, the
International Organization for Standardization (ISO) requested both North and South Korea to work together on developing a standard romanization. The two countries held a series of meetings, during which they failed to reach a consensus. Some created new systems and others proposed reverting to previous systems. In 1991, the South Korean
National Academy of the Korean Language (NAKL; ) proposed its own new system. Concurrently, Bok Moon Kim produced . Despite Kim's advocacy for his system, it never saw widespread adoption; some of the romanizations it produced (e.g. "Dongnipmoon") were mocked in the press for seeming humorous. In 1997, the South Korean government began moving to revise or switch romanization systems. The topic was hotly debated in South Korean press and foreigner communities. One point of concern was on the estimated expenses needed to repaint all road signs with new romanizations. In 1997, the National Academy of the Korean Language System was proposed. It was jointly proposed by the National Commission of Romanization of Korean and the
Academy of Korean Language. The system is transliteral in nature; journalist Choe Yong-shik of
The Korea Times alleged that the system was designed without the input of non-Koreans and mostly meant for ease of use for Koreans. Under that system,
Tongnimmun is rendered
Dogribmun.
Revised Romanization of Korean On July 7, 2000, the NAKL and
Ministry of Culture and Tourism announced that South Korea would adopt a new system: Revised Romanization (RR). Road signs and textbooks were required to follow these rules as soon as possible, at a cost estimated by the government to be at least US$500–600 million. In a 2020 book, linguists Sungdai Cho and John Whitman argued that RR's lack of diacritics has "helped it gain widespread acceptance on the Internet". == Romanization systems of the Soviet Union ==