Script and language policy in East Asia

The entry reads 上親制諺文二十八字 — “the King personally created twenty-eight letters of the vernacular script” — and names them Hunminjeongeum. This is the earliest record that the script existed. Nothing was published and nothing was released; the letters are documented, not announced.

What is contested The record is placed at the last day of the twelfth lunar month. That is the Sillok convention for events logged without a day, not a claim about the date of invention.

The Sillok records 訓民正音成 — “Hunminjeongeum was completed”. What was completed is the book: Sejong’s preface, the rules of use, the technical commentary explaining how the letters were designed, worked examples, and Jeong In-ji’s postscript. The commentary was written by the Hall of Worthies scholars at the King’s direction, not by the King. Its colophon dates it to the first ten days of the ninth month.

What is contested This is the entry that “1446 promulgation” refers to, and the word 반포 (promulgate) does not appear in it. The National Institute of Korean Language states there is no record of any official promulgation ceremony having been held. Hangul Day’s 9 October is a modern conversion of the colophon’s ten-day window, calculated after the manuscript was rediscovered in Andong in 1940 — an anniversary of a book’s completion, not of a proclamation.

Article 14: laws and edicts shall take 국문 — the Korean script — as their base text, with a Chinese translation attached, or shall use mixed Korean-Chinese script. After five centuries in which the language of the state was literary Chinese, this is the instrument that gives Korean script formal standing in law.

What is contested Two things are commonly overstated. The edict permits hanja-annotated and mixed-script versions alongside the Korean base text; it does not mandate hangul-only documents. And its date is lunisolar — Korea did not adopt the Gregorian calendar until 1896 — so “21 November 1894” as usually printed is a lunar date in Gregorian dress. No converted equivalent is asserted here. The format itself was revised at least twice more, in mid-1894 and again in 1895.

Article 16 fixed, for elementary-school instruction, a single kana shape per sound — collapsing the many pre-modern variant forms (変体仮名) into what is essentially today’s standard hiragana — together with a standard on-reading kana usage, and directed that kanji be reduced in number to those of broad application. The article sets no numeric limit on kanji. The ordinance itself is no longer in force; the syllabary it fixed is still the one in use.

What is contested A widely circulated account attaches a 1,200-character kanji cap to this reform and cites 文部省令第24号. Both details are contradicted by the sources read here: the National Diet Library’s law index gives Ordinance No. 14, and the transcribed text of Article 16 contains no numeric cap. Treat the 1,200 figure as wrong rather than merely unconfirmed. Which later instrument formally repealed this ordinance could not be established, so no repeal date is claimed.

1,850 characters. The preface states that the list “shows the range of kanji to be used in laws, official documents, newspapers, magazines and general society”, and that words which cannot be written with them are to be replaced with other words or written in kana. A cabinet decision of 5 November 1946 approved its submission.

Issued the same day as the kanji list, and the larger change of the two. Moved kana spelling off historical orthography and onto modern pronunciation, for colloquial writing specifically — classical texts excepted. ゐ and ゑ go to い and え; を survives only as the object particle; くわ and ぐわ become か and が; ぢ and づ become じ and ず with compound exceptions; は and へ keep their historical spellings only where they are particles. Superseded in 1986 by 現代仮名遣い, cabinet notification No. 1.

One substantive sentence, in force the day it was promulgated — a date chosen to fall on Hangul Day: “The official documents of the Republic of Korea shall be written in hangul. Provided that, for some time, hanja may be used alongside where necessary.” It is the first statute of the new republic to legislate on script, and among the earliest it passed at all. Repealed in 2005 on the enactment of the Framework Act on the Korean Language.

What is contested This act is routinely described as banning hanja. Its own text does the opposite: the second clause expressly permits concurrent hanja use. It defines neither “official document” nor “for some time” nor “where necessary”, sets no penalty and names no enforcement mechanism — which is why Korean legal scholarship treats it as a declaratory statement of linguistic independence rather than an operative prohibition.

Passed at the State Council’s 23rd plenary session. Table 1 — 230 characters — became mandatory in printed matter from 1 February 1956. Table 2, of 285 characters, and a table of 54 simplified radicals were issued for trial. The resolution directs that traditional forms “should stop being used in printed matter”, excepting reprints of classical texts.

What is contested The figure of 1,055 variant characters abolished is often attached to this scheme. It belongs to a different instrument — the First Table of Variant Characters, December 1955 — and is not folded in here.

Approved at the fifth session of the first National People’s Congress. The resolution adopts Pinyin “as a tool for helping people learn Chinese characters and promote Putonghua”, directs that it be taught first in teacher-training and in primary and secondary schools, and that it be extended gradually into publishing.

What is contested Pinyin is widely said to have been intended to replace Chinese characters. The 1958 resolution does not say that — it adopts Pinyin as an auxiliary tool. Zhou Enlai’s separate report of 10 January 1958 states explicitly that Pinyin “is not a phonetic script meant to replace Chinese characters”, but that is his report, which the resolution approves only in principle, and not the enacting text. A long-range romanisation current did exist within the script-reform committee; it is not what this instrument enacts, and the popular claim is misattributed rather than invented. The NPC’s own hosting of the resolution could not be retrieved — the connection fails at the TLS handshake — so the wording rests on convergent reproductions.

Consolidated the 1956 scheme into three tables: 352 characters not used as components, 132 usable as components together with 14 simplified radicals, and 1,754 derived by analogy from those components. Not a State Council instrument — a publication of the script-reform agency, which is why its legal weight differs from the 1956 resolution it consolidates.

What is contested The component figures sum to 2,238. Secondary sources almost universally give 2,236, explained as two characters double-listed — but the table’s own footnote already states its 1,754 figure is deduplicated, and the pair invoked to reach 2,236 is not the pair that footnote describes. The 1964 gazette is not online. No winner is picked here.

Moved the target for removing hanja from schooling and administration forward from 1973 to 1970, directed that hanja be taken out of textbooks, and called for the 1948 Act to be amended. A curriculum ordinance of September 1969 then removed the legal basis for hanja instruction, and elementary Korean-language textbooks reverted to hangul-only in March 1970.

What is contested The change is popularly attributed to a presidential special directive of 1 March 1970. The National Archives’ own chronology shows March 1970 only as the month the textbooks changed, within the 1968–69 chain above, and no separate 1970 directive could be located.

Fixed 1,800 characters for school instruction — 900 for middle school and 900 for high school — after an August 1972 amendment to the Education Act’s enforcement decree restored a legal basis for teaching hanja. Instruction resumed the following month. The set of 1,800 has proved durable and still governs school hanja teaching.

What is contested Called “the 1972 reversal”, but 1972 supplies only an enforcement-decree amendment and this character list. The binding curriculum ordinances that made classical Chinese an independent required subject came later — ministry ordinance No. 325 of 31 August 1973 for middle schools, No. 350 of 31 December 1974 for high schools. The Ministry’s own archive page for this announcement would not load; the date and the figure rest on the National Archives chronology and search metadata.

Seven general rules, each with a principle and its exceptions, plus an appendix of set forms. Its own preface calls it a よりどころ — something to go by — applying to ordinary writing in general social life and expressly excluding technical and specialist terminology and proper nouns. It is the clearest statement in this chronology that a cabinet notification is a reference standard rather than a mandate. Revised on 1 October 1981 to track the new jōyō kanji table.

What is contested Legal drafting follows a separate and narrower Cabinet Legislation Bureau practice, not this notification; the two are frequently merged in secondary accounts. The instrument number for the 1981 revision could not be confirmed and is omitted.

Published in the People’s Daily and the other central papers. Table 1, of 248 characters, entered trial use the next day; Table 2, of 605, was circulated for discussion and never implemented. Trial use was discontinued by July 1978 — about seven months — after the new forms proved too ambiguous to read. Formally abolished in 1986. It is the only reform in this chronology that failed.

What is contested A draft (草案), not a promulgated standard — a materially weaker instrument than every other entry here, and listed as one.

1,945 characters, replacing the 1,850 of 1946 — and replacing its character too. Where the 1946 list set a range outside which words were to be rewritten, this one is framed as a guide (目安). The restriction becomes a recommendation.

What is contested Two dates circulate for this table and only one is the instrument. The National Language Council reported on 23 March 1981; the Cabinet promulgated on 1 October 1981. Chronologies pick whichever they encountered first. Separately, the commonly quoted breakdown of 95 characters added and none removed could not be confirmed — it traces to ministry pages that would not render. The 1,945 total and the notification number are confirmed on the Agency for Cultural Affairs’ own site.

Fixed 4,808 standard forms for common use. This is the instrument by which Taiwan standardised traditional characters rather than simplifying them — the same problem the mainland answered by simplification, answered instead by fixing which of several received forms is correct. A companion table of less-common characters was promulgated in the same window.

What is contested The companion table’s count differs by revision rather than by disagreement: 6,341 at the 1982 promulgation, 6,343 in the 1993 kai-script revision, 6,329 after a 2017 variant-character revision. Three states of one table. The ministry document number for the 1982 announcement could not be confirmed and is omitted.

Ended use of the 1977 draft from the date of the notice, nine years after it was issued and eight after it stopped being used in practice. The Complete List of Simplified Characters was reissued in a new edition dated 10 October 1986, un-merging four characters — 叠, 覆, 像, 囉 — that had been treated as variant equivalents, taking Table 1 from 352 to 350 and the analogy table from 1,754 to 1,753.

What is contested The revised components sum to 2,235. Secondary sources repeat “2,274 characters and 14 radicals”, which does not reconcile with the table’s own stated components under any combination. 2,235 is the figure that adds up; the origin of 2,274 could not be traced, so it is reported rather than dismissed.

Removed the diacritics that could not survive a keyboard. The breve goes — ㅓ becomes “eo” rather than “ŏ”, ㅡ becomes “eu” rather than “ŭ” — and the apostrophes marking aspiration go with it. It is the only instrument in this chronology whose stated problem is computer input, and the reason Korean place names are spelled as they now are in every airport and on every road sign.

What is contested It replaced the South Korean government’s own 1984 standard — Ministry of Education notification No. 84-1, itself an adaptation of McCune–Reischauer — and not the McCune–Reischauer system of 1939, which remains in scholarly use and was never Korea’s to repeal. Amended in turn by notification No. 2014-42 of 5 December 2014, which is the version now in force.

2,136 characters. Technically a fresh promulgation superseding the 1981 notification rather than an amendment to it. The Council for Cultural Affairs reported on 7 June 2010; the Cabinet promulgated on 30 November. The revision is the first to be shaped by the fact that most Japanese writing is now typed rather than handwritten — characters readers recognise but few could write by hand were admitted on that basis.

What is contested Two framings circulate and both are correct: 196 characters added and 5 removed (勺, 錘, 銑, 脹, 匁 — moved to the names-use list rather than dropped), or a net increase of 191. 1,945 + 196 − 5 = 2,136 = 1,945 + 191. Quoting one figure without the other is what makes two accounts look contradictory. The gross figures rest on a ministry PDF that could not be rendered; the 2,136 total, the date and the notification number are confirmed on the issuing body’s own page.

8,105 characters in three levels: 3,500 for general literacy, a further 3,000 by frequency, and 1,605 covering surnames, place names, scientific terms and classical characters used in textbooks. The notice directs that general usage follow this table and that previously relevant tables cease to be used. Released publicly on 19 August 2013. It is the instrument currently in force.

What is contested The notice states the supersession in general terms. It could not be confirmed that it names the 1986 Complete List by title — the tables are published as images rather than extractable text — so the repeal is recorded here as implicit rather than express.