The History of Kanji: From Oracle Bones to the Jōyō List
The history of kanji spans roughly 3,000 years. It begins with divination inscriptions in Bronze Age China, crosses the Korean peninsula into Japan in the 5th and 6th centuries, accumulates three layers of borrowed Chinese readings, and ends with a single Unicode codepoint a learner can type on any device.12 Modern Japanese uses three concurrent scripts; kanji is the only one that did not originate inside Japan.12
Overview
Why this history matters to a learner
Every kanji a learner studies is an artifact of that long arc. The shape on the page is a Tang-era 楷書 (kaisho, "regular script") form with a 1946 Japanese simplification applied on top.34 The two or three on-yomi attached to it are the residue of three borrowing waves separated by centuries.25 The same character renders correctly on a phone, in a textbook, and on a Japanese website because of an encoding standards chain that ran from 1978 to 1992.678
Understanding the lineage explains the three on-yomi strata, the shape of the jōyō list, and why every kanji a learner meets is typeable. None of it is arbitrary.
A one-paragraph timeline
The whole arc fits into nine dated points.
- c. 1200 BCE. Oracle-bone inscriptions in the late Shang capital near modern Anyang; the earliest dated Chinese writing yet found.91011
- 202 BCE to 220 CE. Clerical script (隷書) matures as the practical administrative script of the Han empire.1012
- 5th century CE. Kanji first attested in Japan in dated inscriptions; the Inariyama Sword carries a 471 CE inscription.131
- 538 CE. Buddhism formally transmitted from Baekje, accelerating literate use of kanji (the Nihon Shoki gives the traditional date as 552 CE).141
- 7th to 9th centuries. Japanese envoys to Tang China (遣唐使) bring back 漢音 readings, which the Heian court endorses as the correct pronunciation.25
- 16 November 1946. The Japanese cabinet adopts 当用漢字表 at 1,850 characters, with lightly simplified shapes.15316
- 10 October 1981. 常用漢字表 replaces it at 1,945 characters; the legal framing shifts from restriction to recommendation.173
- 30 November 2010. The jōyō list is revised to 2,136 characters (+196, −5).1819
- 1978 to 1992. JIS X 0208 fixes 6,355 kanji as a Japanese national encoding; Unicode CJK unification then folds JIS and the other regional standards into a shared codepoint space.67208
What this article does not cover
This article walks the chronological arc, not the classification system. The six 六書 categories (pictographs, ideographs, phono-semantic compounds, and the rest) belong to a separate article on this site.
The detailed mechanics of on-yomi and kun-yomi also live elsewhere. This article surveys the three on-yomi strata at the level of "where each came from and one example pair," not the predictive rules for which stratum surfaces in which compound.
The script-evolution arc: oracle bone to regular script
The shape of every kanji a learner studies is the endpoint of a five-stage evolution from the late Shang dynasty to the Tang. The stage names are still used today, both as historical terms and as calligraphic styles.
甲骨文 (oracle bone script), c. 1200 BCE
The earliest dated Chinese writing yet found is 甲骨文: divination inscriptions carved into ox scapulae and tortoise plastrons during the reign of King Wu Ding of the late Shang dynasty. The inscriptions were excavated at the ruins of his capital Yin, near modern Anyang.910
Inscriptions associated with Wu Ding's reign have been radiocarbon-dated to 1254–1197 BCE (±10 years), in agreement with regnal dates independently reconstructed from received texts.11 An estimated 150,000 inscribed oracle-bone fragments have been recovered, the overwhelming majority from the Yinxu site at Anyang.10
The script was first identified as ancient Chinese writing in 1899 by Wang Yirong (1845–1900), Director of the Imperial Academy in Beijing.219 He recognized characters inscribed on "dragon bones" sold as a malaria cure at a Beijing apothecary. Later investigations traced the bones to Anyang.219
The forms are pictographic and irregular in size. They are the seed of every later Chinese script.10
金文 (bronze script), c. 1000 to 300 BCE
金文 ("metal writing") names inscriptions cast into ritual bronzes during the Western and Eastern Zhou dynasties. The shapes are more standardized than oracle-bone forms but still recognizably pictographic.109
The bronze-script corpus is the bridge between divination-era writing and the empire-scale administrative scripts that emerge under the Qin.10
篆書 (seal script), c. 220 BCE
Around 220 BCE the First Emperor of Qin ordered the standardization of weights, measures, currency, and writing across the newly unified empire. His chancellor Li Si implemented the writing reform. Li systematized small seal script (小篆) by eliminating regional variants.1022
Li Si is associated with the Cangjiepian (倉頡篇), a primer listing approximately 3,300 standardized small-seal characters.1022
Small seal script (篆書) remains in use today for personal seals (印鑑), formal stone inscriptions, and decorative calligraphy. The strokes are tall, balanced, and uniformly thin.22
隷書 (clerical script), Han dynasty, 202 BCE to 220 CE
隷書 emerged in the late Warring States period and matured during the Han dynasty as the practical administrative script of the Han empire. The mature form is also called 八分 (happun).1210
Clerical script breaks the visible pictographic continuity of earlier scripts. Characters become flatter and more angular, with distinct horizontal modulation and downward sweeping strokes. The form is faster to write than seal script.12
This is the shape revolution that produced forms recognizable to a modern reader of Chinese or Japanese. The transition from 篆書 to 隷書, known in Sinological scholarship as 隷変, is the main turning point in the visual history of Chinese writing.1210
楷書 (regular script), matured by the Tang dynasty, 618 to 907 CE
楷書 ("model script") descends from clerical script through the Three Kingdoms and Jin periods. Its style stabilized by the Tang.10
This is the form printed in modern dictionaries, rendered by digital fonts, and taught in every modern Japanese textbook. The semi-cursive 行書 and cursive 草書 are calligraphic derivatives of 楷書, not separate scripts.10
What this means for the kanji you read today
Every kanji on a JLPT vocabulary list is a 楷書 form, usually with a Japanese-specific 新字体 simplification applied on top of the Tang shape.34
The line from oracle bone to a Japanese textbook is unbroken in principle. But for most characters, the modern shape no longer resembles the original picture. The Han-era 隷変 erased the visible pictographic link for almost everything that followed.1012
The transmission to Japan, 5th to 6th centuries
Kanji did not pass directly from China to Japan. The script crossed the Korean peninsula on its way east, carried by scribes and Buddhist monks. That route shaped which Chinese pronunciations Japan borrowed first.
Before kanji arrived, Japan had no script
Old Japanese existed as a spoken language, but it had no native writing system.231 The earliest dated inscriptions of Japanese names and titles in kanji are 5th-century swords and mirrors.231
The Inariyama Sword (稲荷山古墳出土鉄剣), excavated in 1968 from the Inariyama Kofun in Saitama, carries a 115-character gold-inlaid inscription. It is dated to the 辛亥 year of the sexagenary cycle, a 60-year calendar cycle. The majority scholarly view places that year at 471 CE.1323
The Inariyama inscription names King Wakatakeru (獲加多支鹵大王), identified with Emperor Yūryaku. It uses kanji partly logographically and partly phonetically to spell Japanese names, which is an early step toward man'yōgana.1323
The route: continent to Korean peninsula to Japan
Sustained transmission of kanji to Japan in the 5th and 6th centuries came through the Korean kingdoms of Baekje (百済) and Goguryeo (高句麗), not through direct contact with the Chinese imperial center. Scribes and Buddhist monks travelling east from Baekje were the main carriers.12314
Both the Kojiki (712 CE) and the Nihon Shoki (720 CE) record that a scholar named Wani (王仁) was dispatched to Japan from Baekje during the reign of Emperor Ōjin, bringing the Analects (論語) and the Thousand Character Classic (千字文).231
Modern scholarship treats Wani as semi-legendary. The Thousand Character Classic is conventionally dated to the early 6th century, which post-dates Emperor Ōjin's traditional reign, and "Ōjin's 16th year" cannot be securely placed on a calendar. The Wani story is best read as a representative narrative of Baekje-mediated transmission rather than as a single historical event.123
Why Buddhism was the vehicle
The Nihon Shoki records that King Seong of Baekje (聖明王) sent King Kinmei of Yamato a gilt-bronze image of Śākyamuni, ritual banners, and Buddhist sūtras. The Nihon Shoki gives the traditional date as 552 CE. Contemporary scholarship more commonly accepts 538 CE, based on alternative early sources such as the Gangō-ji Engi and the Jōgū Shōtoku Hōō Teisetsu.141
Large-scale literate transmission required a reason to copy long texts. Buddhism created that demand: monasteries needed to reproduce sūtras written in Classical Chinese.142
This seeded both the first Japanese reading-aloud tradition (漢文訓読, kanbun-kundoku) and the first systematic borrowing of Chinese pronunciations.142 The 呉音 stratum, discussed below, is the linguistic deposit of this 6th-century Buddhist transmission.
The first Japanese uses of kanji
The earliest substantial Japanese-context uses of kanji are administrative and epigraphic: the Inariyama Sword, the Eta Funayama Sword, and the Suiko-era inscriptions.2324
The first sustained literary works are the Kojiki (712 CE, compiled by Ō no Yasumaro) and the Nihon Shoki (720 CE). These texts use kanji in three ways at the same time:
- For Chinese-language passages, written and read in Classical Chinese.
- For Japanese words by meaning, the seed of kun-yomi.
- For Japanese sounds, the man'yōgana that will eventually become hiragana and katakana.
A measurable difference in phonetic stratum already separates the two works. The Kojiki man'yōgana skews toward 呉音-based readings. The Nihon Shoki man'yōgana, compiled only eight years later, skews toward 漢音-based readings. The shift reflects the brief window in which the Tang-stratum readings began to displace the older Wu stratum at court.52
The three on-yomi strata, 5th to 16th centuries
When a kanji has two or three on-yomi, those readings are almost never arbitrary alternatives. They are the residue of distinct borrowing waves. Each wave is tied to a specific century and a specific Chinese pronunciation system.
| Stratum | Period | Source region | Vehicle | Survives in |
|---|---|---|---|---|
| 呉音 (Go-on) | 5th to 6th c. | Lower Yangtze (Jiankang / Nanjing) | Buddhist transmission via Korean peninsula | Buddhist vocabulary, some everyday words |
| 漢音 (Kan-on) | 7th to 9th c. | Tang capital Chang'an | 遣唐使 envoys; Heian court endorsement | Academic, governmental, modern Sino-Japanese |
| 唐音 (Tō-on) | 12th to 16th c. | Song / Yuan / Ming China | Zen monks, trade contact | A small, locked set of compounds |
呉音 (Go-on, Wu reading), 5th to 6th centuries
The 呉音 stratum derives from Early Middle Chinese pronunciations of the Lower Yangtze region around Jiankang (modern Nanjing), the capital of the Southern Dynasties.25 It reached Japan via the Korean peninsula along with early Buddhism.25
The name 呉音 is retrospective and slightly polemical. Tang-era speakers in Chang'an called their own readings 秦音 (shin'on, "Qin sound") and dismissed southern readings as 呉音 ("Wu sound"). The Heian court inherited that framing.2
呉音 survives most heavily in Buddhist vocabulary and in a number of everyday Sino-Japanese words.25
漢音 (Kan-on, Han reading), 7th to 9th centuries
The 漢音 stratum derives from Late Middle Chinese pronunciations used in the Tang dynasty capital Chang'an (modern Xi'an). It was brought back to Japan by 遣唐使 envoys and by Japanese students who studied in China.25
The Heian court formally endorsed 漢音 as the correct reading, in part because envoys returning from Chang'an treated the contemporary capital pronunciation as prestigious.25 This is why 漢音 is the most common on-yomi a modern learner meets.
漢音 dominates academic, governmental, and Sino-Japanese compounds outside the Buddhist register. Classical Chinese reading instruction at the Daigakuryō (大学寮) was conducted in kan-on.224
唐音 (Tō-on, Tang reading), 12th to 16th centuries
The name 唐音 is misleading. Despite "唐" ("Tang"), the stratum reflects Old and Middle Mandarin pronunciations from the Song, Yuan, and Ming dynasties, transmitted long after the Tang dynasty had fallen.2 In this term, "唐" is used in the loose sense "Chinese," not as a dynasty marker.2
唐音 was carried by Zen monks (especially of the Rinzai and Sōtō traditions) and by trade contact with Song and Ming ports.25 It remains a small stratum, often locked into specific compounds.25
A single kanji can carry readings from all three strata side by side.
行者2
"ascetic practitioner" (呉音 reading)
行動2
"action, behavior" (漢音 reading)
行燈2
"paper lantern" (唐音 reading)
Why this matters to a learner
When a single kanji carries two or three on-yomi, the most common explanation is that the same character was borrowed at two or three different historical moments from two or three different Chinese pronunciation systems.25 The multiplicity has a chronology.
The predictive rules for which stratum appears in which compound (Buddhist 呉音, academic and modern 漢音, Zen and decorative-noun 唐音) belong to a separate article on this site.2 What matters at this level is that the readings are dated artifacts of a 1,000-year sequence of borrowings, not arbitrary noise.
The simplification debate after WWII
The 1946 kanji reform is usually remembered as a single decisive cabinet order. But the argument behind it had been running since the early Meiji period. The post-war moment supplied a political opening, not a new idea.
The pre-war argument
From the early Meiji period (late 1860s onward), some prominent Japanese intellectuals argued that kanji was a barrier to mass literacy and economic modernization.253
The proposals were varied. Maejima Hisoka (前島密) petitioned in 1866 for the abolition of kanji in favor of hiragana.253 Fukuzawa Yukichi (福澤諭吉) argued in Moji no Oshie (1873) for restricting daily-use kanji to roughly 1,000 characters, and demonstrated the principle in a children's primer using only 928 different kanji.253 Mori Arinori proposed replacing Japanese altogether with a simplified English; Nishi Amane proposed romaji in 1874.253
A romaji-promotion society (羅馬字会) reportedly had over 10,000 members by 1888.253 None of these proposals produced a binding national policy in the pre-war period. The debate ran in periodicals and learned societies for decades.253
The post-war pressure
The Allied Occupation (SCAP, 1945–1952) prioritized mass literacy and democratic publishing as part of its educational reform agenda.163 Inside the Ministry of Education, an SCAP-commissioned 1948 national literacy survey was used to argue that the existing kanji inventory was unworkable for compulsory education. The use of the survey was controversial in some accounts.163
The reform faction inside the Ministry of Education and the Kokugo Shingikai (国語審議会) won the resulting policy fight.315 The 当用漢字表 was published as a cabinet directive within sixteen months of surrender.315
Three possible directions, one chosen
The post-war debate ran along three axes.325
- Abolish kanji entirely: the kana-only and romaji proposals.
- Simplify the shapes radically: the path mainland China would take in 1956.
- Cap the inventory and lightly simplify a subset: the path Japan took.
The 16 November 1946 cabinet decision picked the third option.325 This is why modern Japanese 新字体 (shinjitai) is more conservative than mainland Chinese 简体字 (jiantizi). Japan capped the list and trimmed roughly 130 shapes; the PRC reform of 1956 simplified substantially more.34
The kanji lists: 当用漢字 to jōyō to 2,136
The 1946 list, the 1981 replacement, and the 2010 revision are usually presented as three separate inventories. It is clearer to read them as one continuous policy lineage, with the same regulatory structure and changing legal framing.
1946: 当用漢字 (tōyō kanji), 1,850 characters
On 16 November 1946 the Japanese cabinet adopted the 当用漢字表 (昭和21年内閣告示第32号), capping the kanji intended for general use, official documents, and compulsory education at 1,850 characters.153
The same reform package introduced lightly simplified shapes (新字体) for roughly 130 characters and a kana-spelling reform (現代仮名遣い, gendai kanazukai) for spoken-language–aligned hiragana usage. The 当用漢字字体表 follow-up was issued in 1949.263
The list's framing language is restriction, not recommendation. 当用 literally means "for current use." The introductory text explicitly limits kanji outside the list in official, journalistic, and educational contexts.153
Representative shinjitai simplifications:
| Traditional | Shinjitai | Stroke change |
|---|---|---|
| 學 | 学 | 16 → 8 |
| 國 | 国 | 11 → 8 |
| 體 | 体 | 23 → 7 |
| 廣 | 広 | 15 → 5 |
| 圖 | 図 | 14 → 7 |
The replacement forms are largely drawn from pre-existing handwritten or scribal variants (略字), not newly invented characters.34 This is the structural difference from the PRC 1956 reform, which created many simplifications wholesale.34
1981: 常用漢字 (jōyō kanji), 1,945 characters
On 10 October 1981 the Ministry of Education replaced 当用漢字 with the 常用漢字表 (昭和56年内閣告示第1号), which grew the inventory to 1,945 characters.173
The transition was a net expansion. Relative to the 当用漢字 list, 95 characters were added, no characters were removed, and the shape of one character was changed.17
The legal framing shifted from "restriction" to "guideline." 常用 ("for ordinary use") defines what compulsory education must teach and what newspapers and government documents normally use. It does not forbid other characters.173
2010: jōyō revised to 2,136 characters
On 30 November 2010 the Council for Cultural Affairs (文化審議会) published the revised 常用漢字表 (平成22年内閣告示第2号), which added 196 characters and removed 5, for a net total of 2,136.1819
The five removed characters are 勺 (a measurement unit), 銑 (pig iron), 脹 (to swell), 錘 (a spindle weight), and 匁 (a measurement unit).1819 All five were judged obsolete in contemporary usage.1819
The revision criteria centered on frequency in contemporary text (especially newspapers and the web), word-forming productivity, and reading efficiency on screens.193 The additions disproportionately reflect characters in heavy contemporary use in personal and place names that the 1981 inventory had excluded.193
The 2010 revision also adjusted readings: 28 characters gained new readings and 3 lost obscure readings.1819
What sits outside the jōyō list
The 人名用漢字 (jinmeiyō, "name-use") list, maintained by the Ministry of Justice under the Family Registration Law (戸籍法), adds 863 further characters approved for personal names.2728
The boundary moves with the jōyō list. After the 2010 jōyō expansion, 129 characters were transferred from the jinmeiyō list into the jōyō list, and the jinmeiyō list was rebalanced to 861.2728 Later additions (巫 in 2015 and 渾 in 2017) brought the total to 863.2728
Everything outside jōyō and jinmeiyō is 表外漢字 (hyōgai, "outside the list"). These characters are still widely encountered in newspapers, literature, and signage, but they are not required in compulsory education.34 The boundary is policy, not language.34
The largest comprehensive kanji dictionary, the Dai Kan-Wa Jiten (大漢和辞典), contains roughly 50,000 characters.4 Treat this as the practical upper bound of "kanji that have ever been used in Sino-Japanese text," not as a learner target.4
The encoding history: JIS to Unicode
The kanji a learner types is a Unicode codepoint. It can move through fonts, files, and networks without losing its identity. That state was reached through a roughly fifteen-year standards chain.
1978: JIS X 0208, the first Japanese national kanji encoding
The original standard was published in 1978 as JIS C 6226 and renamed JIS X 0208 on 1 March 1987. It fixed 6,355 kanji and 524 non-kanji characters as a two-byte encoded national standard.67
The kanji are partitioned into Level 1 (第一水準, 2,965 high-frequency characters) and Level 2 (第二水準, 3,390 long-tail characters), arranged on a 94×94 grid addressed by "kuten" (区点) coordinates.67
This standard made Japanese word processors and the first generation of personal computers practical for Japanese-language text.6 It solved, at the national-standard level, the dictionary-encoding bottleneck that had limited typewriters.6
1983, 1990, 1997: JIS X 0208 revisions and supplementary standards
The 1983 revision (the so-called 83JIS) changed the printed shapes of 22 characters and rearranged some codepoints.67 It remains controversial because pre-existing files referencing the affected codepoints would render differently.67
The 1990 and 1997 revisions made further minor adjustments and re-unified variants that the 1983 revision had split; the 1990 revision added two characters.67
Supplementary standard JIS X 0212-1990 (補助漢字) added 5,801 auxiliary kanji and 266 non-kanji to address coverage gaps.67 In practice, the Shift-JIS encoding dominant on Windows and most workstations could not represent X 0212. That made the auxiliary set effectively inaccessible to most users.67
JIS X 0213 (2000, revised in 2004 as "JIS2004" and again in 2012) added 3,685 further kanji as Levels 3 and 4 and was designed to be Shift-JIS-encodable. It largely replaced X 0212 as the practical supplementary standard.67
1991 to 1992: Unicode and CJK unification
The Unicode Consortium released Unicode 1.0 on 1 October 1991.208 The BMP (Basic Multilingual Plane) code-point range U+4E00–U+9FFF was reserved for CJK Unified Ideographs. It contained no character data in Unicode 1.0 itself, pending finalization by the CJK Joint Research Group (CJK-JRG, later IRG).208
The first CJK-JRG meeting was held in Tokyo in July 1991.208 The Unified Repertoire and Ordering (URO) Version 2.0 was completed on 27 March 1992 and published as part of The Unicode Standard, Version 1.0, Volume 2 in 1992; the same repertoire was published by ISO as ISO/IEC 10646-1:1993.20829
The initial URO 2.0 contained 20,902 unified Han ideographs. It folded the major national standards of China, Japan, Korea, and Taiwan (including JIS X 0208 and JIS X 0212) into a shared codepoint space.208
Subsequent Extension blocks (Extension A in Unicode 3.0, Extension B in 3.1, on through Extension I) have grown the CJK total well past 90,000 characters.8
What CJK unification means in practice
Han unification was the editorial decision that characters judged to share an abstract identity, but differing only in regional or stylistic shape, should receive a single Unicode codepoint.8 Locale-appropriate rendering is handled by the font selected at display time.8
A learner types a kanji on a Japanese-localized IME, the character is stored as a Unicode codepoint, and that codepoint renders with the Japanese shape from a Japanese font. The same codepoint on a Simplified-Chinese-localized system renders the PRC shape.8
The Han unification problem is the criticism that abstract identity is the wrong notion for characters with culturally distinct regional shapes. It is a real disagreement among Sinologists and Japanologists, but it is largely invisible to a learner using standard fonts.8 Variation Selectors (U+E0100 to U+E01EF) provide a fine-grained override when needed.8
Why this history shapes the kanji you study today
Five long threads converge on one object: the kanji that appears on a page in front of a learner.
The shapes are 楷書 forms with Japanese 新字体 simplifications
The kanji a learner studies is, by default, a 1946-era 新字体 variant of a Tang-era 楷書 form.34 This is why Japanese 学 differs from traditional Chinese 學 and from the older Korean 學 (still in scholarly use). In this particular case, the PRC 简体字 reform happens to produce the same 学.34
Where the PRC simplification went further than the Japanese reform, the divergence visible in dictionaries comes directly from the 1946-versus-1956 policy split. 廣 became Japanese 広 but PRC 广. 龍 was retained in Japan as 竜/龍 but reduced in the PRC to 龙. 圓 became Japanese 円 but PRC 圆.34
Two or three on-yomi per kanji is a dated artifact
A kanji that entered Japanese in the 6th century retained its 5th-century Wu pronunciation (呉音). It then accumulated a 7th-century Tang Chang'an pronunciation (漢音), and sometimes a Song- or Ming-era pronunciation (唐音).25 The apparent chaos in a dictionary entry has a chronology.
The 2,136 jōyō list is a literacy target, not a complete inventory
Newspapers honor the jōyō boundary in editorial style, or use furigana when crossing it.34 Literature and personal names cross it freely.34
Finishing the jōyō list is the threshold for "able to read newspaper text without lookup," not for "able to read literature."
Anything you can read, you can type
Every kanji a learner encounters is a Unicode codepoint that any Unicode-aware device can render and a Japanese IME can produce.678 This state is the endpoint of the JIS-to-Unicode chain. It was not the default.
In 1975, typing 非常用漢字 (non-jōyō kanji) was a hardware-engineering problem. After 1992 it became a font-selection problem. From a learner's point of view, it was no longer a problem at all.
Good to know
"Kanji" literally means "Han characters"
The word 漢字 is itself a Sino-Japanese compound: 漢 names the Han dynasty (and by extension China as a civilization), and 字 means "character." The Korean cognate is 한자 (hanja), the Vietnamese cognate is Hán tự, and the Mandarin cognate is hànzì 汉字.14 All four words are the same compound across four East Asian languages, a small monument to the shared origin of the script.
The Korean peninsula was the actual transmission corridor
Popular accounts often say "kanji came from China." More precisely, kanji reached Japan from China via Baekje and Goguryeo on the Korean peninsula, carried by scribes and Buddhist monks.12314 The 5th- to 6th-century pronunciations Japan first borrowed (the 呉音 stratum) therefore reflect Korean-mediated transmission of southern Chinese readings, not direct northern-Chinese borrowing.12314
Japanese 新字体 and Chinese 简体字 are independent reforms
A common misconception is to treat Japanese 学 and mainland Chinese 学 as products of the same simplification. They are not.34
Japanese 新字体 (1946, roughly 130 shapes touched) and PRC 简体字 (1956, substantially more shapes touched) are independent reforms drawing on a shared stock of scribal abbreviations. The cases where they agree (国, 学, 体) coexist with cases where they disagree sharply: Japanese 広 versus PRC 广 (both from traditional 廣), Japanese 竜/龍 retained versus PRC 龙 (from 龍), and Japanese 円 versus PRC 圆 (both from 圓).34
"Tōyō kanji" is a historical term
The 当用漢字表 was superseded in 1981. The term 当用漢字 refers to the 1946–1981 policy artifact and is appropriate only in historical contexts.317
A learner who says 当用漢字 to mean "the kanji I am studying" is using a term about 35 years out of date. The standard term is 常用漢字 (jōyō kanji).317
"Ghost characters" (幽霊文字) in JIS X 0208
When the 1978 JIS C 6226 standard was being compiled, source documents for place and surname kanji were sometimes transcribed under poor conditions. Folds in paper were read as strokes, and smudges were read as radicals. Roughly twelve characters in the resulting standard have no known prior attestation in any document.306
They survive every JIS and Unicode revision because removing them would shift the codepoints of every later character. The affectionate term is 幽霊文字 (yūrei moji, "ghost characters"). Even the digital kanji inventory is a historical artifact.306
Tō-on is not Tang-era
The name 唐音 ("Tang sound") is misleading. The readings date mostly to the Song, Yuan, and Ming periods (10th to 16th centuries), not to the Tang dynasty (618–907).25 In this term, 唐 is used in the loose sense "Chinese," not as a dynasty marker.25
Dates worth remembering
A learner who internalizes nine dates can place every other detail in this article on the timeline.
- c. 1200 BCE: oracle bone inscriptions.
- 202 BCE: clerical script in the Han administration.
- 471 CE: Inariyama Sword, earliest dated kanji inscription in Japan.
- 712 CE: Kojiki.
- 1946: 当用漢字, 1,850 characters.
- 1981: jōyō, 1,945 characters.
- 2010: jōyō, 2,136 characters.
- 1978: JIS X 0208.
- 1991 to 1992: Unicode CJK unification standards arc.
Nine dates, the whole arc.
See also
- What Is Kanji? A Complete Beginner's Introduction
- How to Count Kanji Strokes (画数): The Eight Basic Strokes Plus the Corner, Hook, and Enclosure Rules
- Kanji Stroke Order: The General Rules Behind Every Character
- Writing Kanji by Hand: Is It Still Worth It?
- Secondary School Jōyō Kanji (中学校 + 高等学校): The 1,110-Character Set Beyond Elementary
- Wago, Kango, Gairaigo, Konshugo: The Four Vocabulary Strata of Japanese