Exactly 9,389 entries make up South Korea's list of hanja allowed in a child's name, and the total rebuilds to the last entry from three lists: the 1,800 basic hanja for middle and high school, 7,272 entries in the Supreme Court's first table and 317 variant forms in its second, by our count. An entry is one character with one reading, so a character allowed under two readings counts twice. The school list, 900 characters for middle school and 900 for high school, was set in 1972 and last changed its membership in 2000. On 29 April 2026 the Constitutional Court upheld the limit on name hanja; four justices dissented.
A list from 1972 still serves as the school standard
South Korea's Ministry of Education announced the 1,800 basic hanja for 한문, classical Chinese, in August 1972, by the ministry's history record. On 24 November 2000 Hankook Ilbo reported a proposal to replace 44 characters from spring 2001. 李, 朴, 革 and 舌 would move from the high-school 900 to the middle-school 900, and 44 others would leave the list. That was a draft for comment, so we checked the outcome in Unicode's data, which tags each character on the current list. Of the 44 the article named as additions, 43 carry that tag, and the 44th, 隸, carries it under its variant form 隷. None of the 44 named for removal does.
In 2007 the ministry replaced some popular character forms on the list with standard ones, and a Supreme Court rule change in January 2010 moved 7 of the dropped forms into its name list's second table. Primary schools have no list. In December 2016 the ministry picked 300 characters that grade 5–6 textbooks could gloss, and on 27 December 2017 it dropped them, Hankook Ilbo reported.
Every allowed reading and variant counts as its own entry
The scope is set by Article 37 of the Supreme Court's family-registration rules, which allows the 1,800 school hanja, the characters in the court's first table, and only those variant forms printed in its second. A character later dropped from the school list is treated as added to the first table, and all 44 characters dropped in 2000 do carry Unicode's name-list tag. The 9,389 therefore counts entries rather than distinct characters. 疋 was already on the list read 필, and the 2024 change added it again read 아.
The table has grown five times since 2013, by the court's record of amendment reasons.
| In force | Added | Total |
|---|---|---|
| July 2013 | 307 | 5,761 |
| January 2015 | 2,381 | 8,142 |
| December 2018 | 137 | 8,279 |
| February 2022 | 40 | 8,319 |
| 11 June 2024 | 1,070 | 9,389 |
The court's 2024 announcement named 㖀 (률), 疋 (아) and 䬈 (태) among the additions, which its attached list splits 1,069 to the first table and 1 to the second. Length limits are in our guide to name registry rules.
Hangul takes over as the record for a name outside the list
If a name contains a character outside that range, Article 37 says the register records the name in Hangul. A parent challenged that after a daughter's name was recorded in Hangul only, and in case 2023헌마282 the Constitutional Court held that the limit does not infringe the freedom to name a child. Four justices dissented. The decision quotes a registration rule under which a character added to the list later can be added to the record by a supplementary report, 추후보완신고, without a change of name. The Supreme Court's own name-hanja search answers whether a given character is on the list.
Traditional and Japanese forms often show up as separate code points
Of the 5,401 characters in Big5's frequent block, the basis of Traditional-Chinese computing in Taiwan and Hong Kong, 555 have no Korean reading in Unicode's Unihan database (version 17.0). At most 117 have a Korean form under another code point: 研 against 硏, 為 against 爲, 值 against 値. KS X 1001, South Korea's standard character set, encodes the second of each pair. The rest, such as 吧, 嗎 and 啤, are characters Korean does not use. Of Japan's 2,136 Jōyō kanji, 239 have no Korean reading, though for 95 of them a variant form does.
From the Korean side, by our matching method, 1,379 of the 1,800 school hanja (76.6%) share a code point with a Jōyō kanji and 1,112 (61.8%) with a character on mainland China's level-1 list of 3,500. Counting old forms, simplifications and variant links, at least 1,649 (91.6%) have a Jōyō counterpart and at least 1,765 (98.1%) a level-1 one. Unicode leaves some pairs unlinked, 尙 and 尚 among them, so the true shares are higher.
Our converter treated three school readings as word-initial forms
The data behind our Hanja to Hangul converter is identical to Unihan 17.0 for all 8,525 characters with a Korean reading, and it defaulted to the right school reading for 1,797 of the 1,800 school hanja. Until 2 October 2026 it treated the school readings of the other three, 怒, 異 and 諾, as the word-start forms of 로, 리 and 락, so 差異 came out 차리, 憤怒 분로 and 承諾 승락, for the standard 차이, 분노 and 승낙. It also printed ? for 研 and 為, wrote 困難 as 곤난 and gave 自動車 as 자동거.
It now gives all 1,800 their school reading. A character with no reading of its own is read through its Korean form, and the page says so: 研究 comes out 연구, read as 硏, which leaves 465 of the 5,401 Big5 frequent characters unread. 車 keeps its school reading 거 except in nine dictionary words such as 自動車, now 자동차, and the page now applies the spelling rules below.
Customary readings such as 시월 and 곤란 rank as standard spelling
South Korea's spelling rules, kept by the National Institute of Korean Language, move a hanja's syllable with its place in a word. Articles 10 to 12 cover the start of a word: 李 is 리 inside a word and 이 at the start, 女 is 녀 in 남녀 and 여 in 여자, and 률 after a vowel or ㄴ is written 율, as in 비율 and 백분율.
Article 52 accepts customary sounds where usage has moved: 「본음으로도 나고 속음으로도 나는 것은 각각 그 소리에 따라 적는다」. So 十月 is written 시월, though 십일 keeps the original sound of 十; 困難 is 곤란, 六月 is 유월 and 木瓜 is 모과. 諾 keeps its original sound in 승낙 and takes the customary 락 in 수락 and 허락. In numbers written by the decimal system, each digit after a place value starts a word, so 六千六百六十六 is 육천육백육십육.
Sino-Korean words qualify as most of the dictionary's headwords
The National Institute of Korean Language counted the 422,890 headwords of its Standard Korean Language Dictionary, 표준국어대사전 in May 2022. Of those, 235,173 (55.6%) are Sino-Korean, 75,520 (17.9%) native Korean, 23,819 (5.6%) loanwords and 88,378 (20.9%) hybrids that mix the types. Each headword counts once, however rare, so the figure describes the dictionary, and says nothing about how often the words turn up in speech or print. The often-repeated claim that 60 to 70 percent of Korean vocabulary is Sino-Korean cites no primary source on any of the pages we checked.
View author profile → · Editorial policy
