Kanji
Chinese characters used in Japanese writing
From Wikipedia, the free encyclopedia
Kanji (/ˈkæn.dʒi, ˈkɑːn.-/;[1] Japanese: 漢字, hiragana: かんじ, Katakana: カンジ, pronounced [kä̀ɲ̟́.ɟ̟͡ʑí, kä́ɲ̟́.-] ⓘ,[2][3] 'Han characters'[4][5]) are logographic Chinese characters, historically adapted from Chinese writing scripts, used in the writing of Japanese.[6] They comprised a major part of the Japanese writing system during the time of Old Japanese and are still used today, along with the subsequently derived syllabic (phonographic) scripts of hiragana and katakana.[7][8] Kanji may have one or more readings, conventionally grouped into two categories: kun'yomi, based on native Japanese words associated with the characters' meanings, and on'yomi, based on Chinese pronunciations adapted to Japanese at different periods. Some characters, known as kokuji, were created in Japan using components and principles derived from Chinese characters.[9]
| Kanji | |
|---|---|
 "Kanji" written in kanji with furigana | |
| Script type | |
Period | 5th century CE – present |
| Direction | Vertical right-to-left, left-to-right |
| Languages | Old Japanese, Kanbun, Japanese, Ryukyuan languages, Hachijō |
| Related scripts | |
Parent systems | |
Sister systems | Hanja, zhuyin, traditional Chinese, simplified Chinese, chữ Hán, chữ Nôm, Khitan script, Jurchen script, Tangut script, Yi script |
| ISO 15924 | |
| ISO 15924 | Hani (500), Han (Hanzi, Kanji, Hanja) |
| Unicode | |
Unicode alias | Han |
After the Meiji Restoration, Japan made its own efforts to simplify kanji, producing the forms now known as shinjitai (新字体; 'new character forms'). Since the 1920s, the Japanese government has periodically issued lists to guide character use and education. The principal lists comprise 2,136 jōyō kanji for general use and 863 additional jinmeiyō kanji permitted in names; a further character, 勒, is scheduled to become name-permitted in December 2026.[10][11][12]
The term kanji is a direct borrowing and phonetic reading (on'yomi) of the Chinese word hanzi (traditional Chinese: 漢字; simplified Chinese: 汉字; pinyin: hànzì), which is one of the formal terms used when referring to Chinese characters.[13][14] The significant use of Chinese characters in Japan first began to take hold around the 5th century CE and has since had a profound influence in shaping Japanese culture, language, literature, history, and records.[15] Inkstone artifacts at archaeological sites dating back to the earlier Yayoi period were also found to contain Chinese characters.[16]
Although many characters retain related meanings across Japanese and Chinese, their pronunciations and some of their meanings have diverged. For example, 誠 means 'sincerity' in both languages but has the Japanese readings makoto and sei, alongside Mandarin chéng. Japanese-created characters and words formed from Sino-Japanese morphemes have also sometimes been borrowed into other East Asian languages; Japanese-coined Sino-Japanese words are called wasei-kango.[9]
History
Chinese characters first reached Japan on imported seals, coins, mirrors, swords and other inscribed objects. The best-known early example is the King of Na gold seal, presented by Emperor Guangwu of Han to an emissary from Wa in 57 CE.[17] Chinese coins and possible writing implements have also been found at Yayoi period sites, but widespread production and use of writing in Japan developed later, especially from the fifth century onward.[15] The Kojiki and Nihon Shoki preserve the traditional account that the scholar Wani came from Baekje during the reign of Emperor Ōjin, bringing Confucian texts and knowledge of Chinese writing; modern accounts treat the episode as a literary tradition rather than a precisely datable event.[18][19]
The earliest Japanese documents were probably written by bilingual Chinese or Korean officials employed at the Yamato court.[15] For example, the diplomatic correspondence from King Bu of Wa to Emperor Shun of Liu Song in 478 CE has been praised for its skillful use of allusion. Later, groups of people called fuhito were organized under the monarch to read and write Classical Chinese. During the reign of Empress Suiko (593–628), the Yamato court began sending full-scale diplomatic missions to China, which resulted in a large increase in Chinese literacy at the Japanese court.[20]
From the seventh century, large numbers of inscribed wooden tablets called mokkan (木簡) document administrative records, labels for goods, correspondence and writing practice. Their survival provides direct evidence for the spread of literacy and bureaucratic writing, although earlier inscriptions are known from other objects.[19][21]
Chinese characters initially served to write Sinitic texts in Japan. By the Nara and Heian periods, reading practices now called kanbun kundoku used glosses and diacritical marks to render such texts with Japanese word order, particles, and inflections.[22]
Chinese characters also came to be used to write texts in the vernacular Japanese language, resulting in the modern kana syllabaries. Around 650 CE, a writing system called man'yōgana (used in the ancient poetry anthology Man'yōshū) evolved that used a number of Chinese characters for their sound, rather than for their meaning. Man'yōgana written in cursive style evolved into hiragana (literally "fluttering kana" in reference to the motion of the brush during cursive writing), or onna-de, that is, "ladies' hand",[23] a style strongly associated with women's writing at the Heian court. Major works of Heian-era literature by women were written in hiragana. Katakana (literally "partial kana", in reference to the practice of using a part of a kanji character) emerged via a parallel path: monastery students simplified man'yōgana to a single constituent element. Thus the two other writing systems, hiragana and katakana, referred to collectively as kana, are descended from kanji. In contrast with kana (仮名, literally "borrowed name", in reference to the character being "borrowed" as a label for its sound), kanji are also called mana (真名, literally "true name", in reference to the character being used as a label for its meaning).[24][25]
In modern Japanese, kanji are used to write certain words or parts of words (usually content words such as nouns, adjective stems, and verb stems), while hiragana are used to write inflected verb and adjective endings, phonetic complements to disambiguate readings (okurigana), particles, and miscellaneous words which have no kanji or whose kanji are considered obscure or too difficult to read or remember. Katakana are mostly used for representing onomatopoeia, non-Japanese loanwords (except those borrowed from ancient Chinese), the names of plants and animals (with exceptions), and for emphasis on certain words.[26]
Orthographic reform and lists of kanji
Debate over the place and number of kanji has a long history. The Edo-period scholar Kamo no Mabuchi criticized the burden imposed by a large character inventory and contrasted it with the smaller kana syllabaries; proposals to restrict or replace kanji became especially prominent in modern language reform.[27]
After the Meiji Restoration and as Japan entered an era of active exchange with foreign countries, the need for script reform in Japan began to be called for. Some scholars argued for the abolition of kanji and the writing of Japanese using only kana or Latin characters. These views were not widespread.[28]
However, the need to limit the number of kanji characters was understood, and in May 1923, the Japanese government announced 1,962 kanji characters for regular use. In 1940, the Japanese Army decided on the "Table of Restricted Kanji for Weapons Names" (兵器名称用制限漢字表, heiki meishō yō seigen kanji hyō) which limited the number of kanji that could be used for weapons names to 1,235. In 1942, the National Language Council announced the "Standard Kanji Table" (標準漢字表, hyōjun kanji-hyō) with a total of 2,528 characters, showing the standard for kanji used by ministries and agencies and in general society.[29]
In 1946, during the Allied occupation, the Japanese government adopted major orthographic reforms intended to simplify education and written communication. Although devised in the occupation context, the 1946 character policy was a Japanese reform and was not imposed by the occupation authorities.[30]
The number of characters in circulation was reduced, and formal lists of characters to be learned during each grade of school were established. Some characters were given simplified glyphs, called shinjitai (新字体). Many variant forms of characters and obscure alternatives for common characters were officially discouraged.[10]
The official character lists are guidelines for general social use rather than a ban on other characters. Characters outside the jōyō list are known as hyōgaiji (表外字).[10][31]
Kyōiku kanji
The kyōiku kanji (教育漢字; lit. "education kanji") are the 1,026 characters allocated across the six grades of Japanese elementary school. The grade-level breakdown is the gakunen-betsu kanji haitōhyō (学年別漢字配当表), maintained through the national curriculum standards of the Ministry of Education.[32]
Jōyō kanji
The jōyō kanji (常用漢字; regular-use kanji) list contains 2,136 characters, including all 1,026 elementary-school characters. It is an official guideline for kanji use in general social life, not a prohibition on other characters. The list was introduced in 1981 with 1,945 characters, replacing the 1,850-character tōyō kanji (当用漢字; general-use kanji) list of 1946, and was expanded to 2,136 in 2010. The 2010 additions include characters used in Japanese prefecture names.[10]
Jinmeiyō kanji
The jinmeiyō kanji (人名用漢字; kanji for use in personal names) are the supplementary characters that, together with the 2,136 jōyō kanji, may be used in children's registered names. The list began with 92 characters in 1952 and has been expanded repeatedly. It currently contains 863 characters; 勒 is scheduled to become the 864th in December 2026.[11][12]
Hyōgai kanji
Hyōgai kanji (表外漢字; "unlisted characters") are kanji not contained in the jōyō kanji list. They may include characters permitted for names; in ordinary publications they are often printed in traditional forms, although extended shinjitai forms also exist.[33]
Japanese Industrial Standards for kanji
The Japanese Industrial Standards define coded character sets used in Japanese information processing. The principal kanji standards have been revised or supplemented over time:[34]
- JIS X 0208, the main legacy two-byte standard, contains 6,355 kanji.
- JIS X 0212, a supplementary set containing 5,801 additional kanji, was not available in the common Shift JIS encoding and consequently saw limited implementation.
- JIS X 0213, first issued in 2000 and subsequently amended, extends the JIS X 0208 repertoire and defines Shift-JIS-compatible encodings for the enlarged set.
- JIS X 0221:2020, the Japanese adoption of the ISO 10646/Unicode standard.[35]
Gaiji
Gaiji (外字; literally "external characters") are kanji that are not represented in existing Japanese encoding systems. These include variant forms of common kanji that need to be represented alongside the more conventional glyph in reference works and can include non-kanji symbols as well.
Gaiji can be either user-defined characters, system-specific characters or third-party add-on products.[36] Both are a problem for information interchange, as the code point used to represent an external character will not be consistent from one computer or operating system to another.
Gaiji were nominally prohibited in JIS X 0208-1997 where the available number of code-points was reduced to only 940.[37] JIS X 0213-2000 used the entire range of code-points previously allocated to gaiji, making them completely unusable. Most desktop and mobile systems have moved to Unicode negating the need for gaiji for most users. Historically, gaiji were used by Japanese mobile service providers for emoji.
Unicode allows for optional encoding of gaiji in private use areas, while Adobe's SING (Smart INdependent Glyphlets)[38][39] technology allows the creation of customized gaiji.
The Text Encoding Initiative uses a <g> element to encode any non-standard character or glyph, including gaiji. The g stands for gaiji.[40][41]
Total number of kanji
There is no definitive count of kanji characters, just as there is none of Chinese characters generally. The Dai Kan-Wa Jiten, which is considered to be comprehensive in Japan, contains about 50,000 characters. The Zhonghua Zihai, published in 1994 in China, contains about 85,000 characters, but the majority of them are not in common use in any country, and many are obscure variants or archaic forms.[42][43][44]
The official 2,136-character jōyō list provides a guideline for general social use, while many additional characters occur in names, historical texts and specialist vocabulary. Japanese encoding standards therefore cover far more characters than are taught for routine use; the number depends on which JIS or Unicode repertoire is counted.[10][45]
Readings
Individual kanji may write one or more different words or morphemes, leading to different pronunciations or "readings". Context—including the intended word, whether the character stands alone or forms part of a compound, and surrounding grammatical material—normally determines the reading. For example, 今日 may be read kyō for "today" or konnichi in expressions referring to the present day. Furigana can specify rare, literary or otherwise non-standard readings.[9][46]
Readings are conventionally categorized as kun'yomi (訓読み) ("meaning readings"), based on native Japanese words, or on'yomi (音読み) ("sound readings"), derived from Chinese pronunciations. A character may have readings in both categories, in only one, or—especially in names and compounds—an irregular reading that applies in a particular word.[9][47]
Some common characters have many officially listed readings. The jōyō table, for example, gives 生 the on readings sei and shō and ten kun readings or inflected forms: i-kiru, i-kasu, i-keru, u-mu, u-mareru, ha-eru, ha-yasu, nama, ki, and o-u.[47]
On'yomi (Sino-Japanese reading)
The on'yomi (音読み; [oɰ̃jomi], lit. "sound(-based) reading") are Sino-Japanese readings based on Chinese pronunciations that were adapted to Japanese phonology. A character may have several on readings because borrowing occurred at different times, from different regional varieties, and with different adaptation strategies.[48] Although most Japanese-created characters (kokuji) have no on reading, there are exceptions: 働 "to work" has kun hatara(ku) and on dō, while 腺 "gland" is conventionally read sen; their phonetic components are 動 and 泉, respectively.[49][50]
Kun'yomi (native reading)
The kun'yomi (訓読み; [kɯɰ̃jomi], lit. "meaning reading") is based on a native Japanese word, or yamato kotoba, associated with the meaning of a character. A single character can have several kun'yomi, while others have none.[9]
Ateji
In the narrow sense, ateji (当て字) are kanji used for their established sounds without regard to their meanings, as in phonetic spellings of names or loanwords. In broader reference usage, the term may also include meaning-based spellings whose reading belongs to the word as a whole rather than to the individual characters; these are more specifically called jukujikun.[51]
Gairaigo
Long foreign words have occasionally been assigned as meaning-based readings of single characters or symbols. For example, 糎 was coined for センチメートル (senchimētoru, "centimetre"), although the unit is now normally written with the Latin abbreviation cm.[52]
Mixed readings
Kanji compounds can mix on'yomi and kun'yomi. An on-kun compound is called jūbakoyomi (重箱読み), after jūbako (重箱; a tiered food box), while a kun-on compound is called yutōyomi (湯桶読み), after yutō (湯桶; a hot-water vessel). The names are themselves examples of the patterns. Other mixed compounds include basho (場所; "place", kun-on), kin'iro (金色; "golden", on-kun) and aikidō (合気道; the martial art Aikido", kun-on-on).[53][54]
Mixed or otherwise irregular readings also occur in phonetic place-name spellings. Sapporo, for instance, has an Ainu-derived name represented by the kanji 札幌; the characters provide a conventional written form rather than the etymology of the name.[55]
Special readings
Gikun (義訓) and jukujikun (熟字訓) are meaning-based readings in which a word's pronunciation does not correspond directly to the ordinary readings of its individual characters. In jukujikun, an established word is represented by a multi-character spelling chosen for its overall meaning.[51][56]
Gikun assigns a context-specific reading according to the meaning of a character or character sequence rather than using its fixed ordinary reading. Such readings may be indicated with furigana, allowing the written form and the supplied pronunciation to contribute different information.[57]
In jukujikun, the word is read as a whole. For example, 今朝 "this morning" is normally kesa, and 今日 "today" is normally kyō; neither reading divides into the regular readings of the two characters. The alternative konnichi for 今日 occurs in expressions referring to "the present day" and in the greeting konnichi wa.[56][58]
Jukujikun are used for native words such as 大和 (Yamato) and for some older borrowings, including 柳葉魚 (shishamo) from Ainu and 煙草 (tabako) from Portuguese. Such words are also commonly written in kana.[51][59]
A case where jukujikun is used for Sino-Japanese is the word kyōdai, which, prototypically, means "brothers" and is spelt 兄弟 ("big and little brothers"). However, the meaning has been expanded to "siblings" in general, and can assume such spellings as 姉妹 ("(big and little) sisters", alternatively pronounced shimai), 兄妹 ("big brother and little sister") and 姉弟 ("big sister and little brother"). It is also possible to say otoko kyōdai ("male siblings; brothers") and onna kyōdai ("female siblings; sisters").[60]
Because the reading belongs to the whole written word, the number of characters need not match the number of morae or syllables. 啄木鳥, for example, has the short reading kera as well as kitsutsuki and other historical forms.[61]
The written form of a jukujikun is tied to a particular word, and the same characters may also have a regular Sino-Japanese reading. For example, 馴鹿 can be read tonakai for "reindeer", alongside the rarer Sino-Japanese reading junroku.[62]
The word represented by jukujikun is often native Japanese or an old borrowing. Most are nouns, but established readings also occur in other word classes. Their histories can involve contraction or fusion; 今日 kyō, for example, developed from an older form kefu.[63]
Inflected words can also have whole-word readings. Examples include 可愛い (kawai-i, "cute"), 相応しい (fusawa-shii, "appropriate") and 流行る (haya-ru, "to be in vogue"). Their okurigana identifies the inflection even though the kanji sequence is not read character by character.[64][65]
Typographically, furigana for a whole-word reading may be centred over the entire compound rather than aligned character by character, reflecting that the reading belongs to the word as a unit.[66]
Single character gairaigo
In rare cases, a character has a meaning-based reading borrowed from a modern foreign language, although modern loanwords are normally written in katakana. Examples recorded in dictionaries include pēji (頁、ページ; page), botan (釦、ボタン; button), zero (零、ゼロ; zero), and the unit character mētoru (米、メートル; metre). Such readings are conventionally shown in katakana rather than the hiragana used for native kun readings.[67]
Nanori
Some kanji have name readings called nanori (名乗り). These may preserve uncommon or historically specialized readings and are used principally in personal names; place names likewise contain conventional readings not necessarily predictable from ordinary vocabulary.[68]
When to use which reading
Although compounds favour on'yomi and independent native words often use kun'yomi, these are tendencies rather than mechanical rules. Many characters have several readings, and the intended word must be identified from context; names and place names are especially unpredictable.[9][69]
A character followed by okurigana often has a kun reading, as in 赤い akai "red" and 見る miru "see", but this is not absolute: 感じる kanjiru "feel" has a Sino-Japanese stem. Okurigana can distinguish 食べる taberu from 食う kuu, both "eat", while 開く remains ambiguous between aku and hiraku, both "open".[9][10]
Multi-character compounds commonly use on readings, but mixed and all-kun compounds are well established. Examples include 牛肉 gyūniku "beef" (on-on), 豚肉 butaniku "pork" (kun-on), and 手紙 tegami "letter" (kun-kun). Conversely, a character used as a word by itself often has a kun reading, as in 北 kita "north", but on-reading words such as 愛 ai "love" and 禅 zen are also independent. Only the lexical and sentence context determines whether 金, for example, is kin "gold" or kane "money; metal".[9][47]
Multiple readings create written homographs. 上手 can be jōzu "skilful", uwate "the upper hand", or kamite "stage left"; 明日 may be ashita, asu, or the more formal myōnichi, all referring to tomorrow. The reverse problem occurs in speech: homographs such as 市立 and 私立, both normally shiritsu, may be distinguished as ichiritsu and watakushiritsu.[70][71][72][73]
Legalese
Certain words have conventional specialist readings in legal contexts. The following contrasts are recorded in major Japanese dictionaries:[74][75]
Ambiguous readings
Where context alone may not distinguish homophones, speakers sometimes substitute a mixed or otherwise distinctive reading. Dictionaries explicitly label a number of such disambiguating forms.[74][75] Examples include:
| Ambiguous reading | Disambiguated readings |
|---|---|
| baishun | baishun (売春; "selling sex", on)
kaishun (買春; "buying sex", yutō)[75] |
| jiten | kotobaten (辞典; "word dictionary", yutō)[75]
kototen (事典; "encyclopedia", yutō)[75][74] mojiten (字典; "character dictionary", irregular, from moji (文字; "character"))[75] |
| kagaku | kagaku (科学; "science", on) |
| karyō | ayamachiryō (過料; "administrative fine", yutō)[75][74] |
| Shin | Hatashin (秦; "Qin", irregular, from the alternative reading Hata used as a family name)[75][74]
Susumushin (晋; "Jin", irregular, from the alternative reading Susumu used as a personal name)[75][74] |
| shiritsu | ichiritsu (市立; "municipal", yutō)[75][74] |
There are also cases where the words are technically heterophones, but they have similar meanings and pronunciations, therefore liable to mishearing and misunderstanding.
Different kanji for different meanings
Japanese orthography sometimes distinguishes homophonous or historically related words by assigning different kanji spellings according to meaning.[10]
The polysemous adjective yasui [jɑ̈̀sɨ́ì] is spelt with different kanji depending on its meanings: 安い・廉い[77] for "cheap," with the kanji 廉 representing the Chinese word for "cheap;" and 易い[78] for "easy; simple," with 易 representing the Chinese word for "easy; simple." Its more archaic meaning, "calm; at ease; peaceful; at peace; in peace," is preserved in the imperative form 安かれ (yasukare; 'may you/they/it be at peace (alive); may you/they/it rest in peace (dead)'),[79] which explains the kanji 安 which represents the Chinese word for "calm; peaceful."
The adjective atsui [ɑ̈̀t͡sɨ́ì] is spelt as 熱い[80] for tangible things hot to the touch, and as 暑い[81] for hot weather or climate. 熱 and 暑 are distinct words in Chinese, although the former can also be used for hot weather or climate. Atsui (厚い; [ɑ̈̀t͡sɨ́í]) meaning "thick" is unrelated.
The words oji "uncle" and oba "aunt" can be spelt differently depending on the age of the aunt or uncle relative to that of one's parent. If he or she is older than one's parent, the spellings are 伯父[82] and 伯母;[83] if he or she is younger, they are 叔父[84] and 叔母.[85] These spellings are taken from Chinese, although 伯母 and 叔母 are specifically the wives of one's paternal uncles rather than just "aunts" as in Japanese.
Alteration of spelling with homophonous kanji
Homophonous kanji have also sometimes replaced older spellings, whether through orthographic reform or conventional substitution.
The character 和 (wa, "harmony") came to be preferred over the homophonous 倭 in names for Japan; the latter had acquired disparaging associations in some Chinese usage. See Wa (name of Japan).[86]
The kanji 才 (sai) originally means "talent," but due to its simple calligraphy as well as homophony with 歳 (sai), it is often used as a counter for "year of age,"[87] although the former has not completely displaced the latter. These characters are not homophonous in other languages, where 才 is cái in Mandarin, 재 (jae) in Korean and tài in Vietnamese, while 歳 is suì, 세 (se) and tuế.
The word for telephone, 電話(機) denwa(ki) in Japanese, was spelt with the Chinese characters for "lightning/electricity," "speech" and "machine" in that order, together literally meaning "electric-speech machine", a homophonous alteration of the original name for the device, 伝話機 (or 傳話機 in kyūjitai), literally "speech-transmitting machine," whereby 伝 meant "transmit." The original spelling was first attested in 1878, two years after the telephone was introduced into Japan,[88] and the later spelling came into competition with it not long after, as shown in an 1885 legislative document which pointed out the inconsistencies in terminology in a bill between 傳話 (with 傳 as in "transmit"), 電話 (with 電 as in "lightning") and even テレボン (terebon, supposedly a misspelling of terehon, an early rendition of the English word telephone).[89] One 1892 English-Japanese dictionary uses the spelling 傳話機 for the "telephone" entry whose definition contains the word "electricity", but 電話機 for the "telephonic" entry right underneath.[90] Ultimately, only 電話 survived and was adapted into Sino-Xenic vocabularies as diànhuà in Mandarin Chinese, 전화 (jeonhwa) in Korean and điện thoại in Vietnamese.[91] Likewise, the word denshin(ki) (電信(機); 'telegraph'; lit. 'electric-signal machine'; earlier 伝信機 lit. 'signal-transmitting machine', first attested in 1860)[92] was adapted as diànxìn, điện tín and 전신 (jeonsin). Compare denpō (電報; 'telegram'; lit. 'electric information') which has always been spelt as such,[93] adapted as diànbào, 전보 (jeonbo) and điện báo. The homophony between 電 "electricity" and 伝 "transmit" only occurs in Japanese and Korean, but not in Mandarin and Vietnamese, where 電 is diàn and điện, while 伝 is chuán and truyền.
The word kansū ('mathematical function'), a Chinese borrowing, was originally spelt 函数, with the kanji for "box/contain" and "number," thus "containing number," which refers to the fact that functions "contain" variables.[b] The term is first attested in Dàiwēijī Shíjí (代微積拾級 'Ten Levels of Algebra and Calculus'), a seminal 1859 translation into Chinese of Elias Loomis's Elements of Analytical Geometry and of the Differential and Integral Calculus, resulting from a joint effort between the Chinese mathematician Li Shanlan and the British missionary Alexander Wylie.[94][c] Due to the nonstandard status of the kanji 函 (kan), homophonous alternatives such as 干 (kan) were proposed,[95] but ultimately 関 (kan, but historically kwan) won out.[96] This homophony only occurs in Japanese, and other languages adopted their own readings of 函 as in Mandarin hánshù, Korean 함수 (hamsu) and Vietnamese hàm số, as opposed to their readings of 関 which are guān, 관 (gwan) and quan.
Place and personal names
Japanese place names draw on on readings, kun readings, phonetic spellings and name-specific conventions. Familiar examples with predominantly on readings include 東京 Tōkyō, 京都 Kyōto, 本州 Honshū and 九州 Kyūshū, while examples with native readings include 大阪 Ōsaka, 青森 Aomori and 箱根 Hakone. Historical place-name spellings may represent meaning, sound, or a mixture of the two, so their pronunciation cannot always be predicted from the ordinary readings of their characters.[9][97]
Abbreviated names may select a reading that differs from the full place name. 阪神 Hanshin, for example, combines 阪 from 大阪 Ōsaka and 神 from 神戸 Kōbe. 京成 Keisei similarly combines characters from 東京 Tōkyō and 成田 Narita.[98][99]
Personal names may use ordinary on and kun readings as well as nanori, and the same written name can have more than one possible pronunciation. Since 2025, Japanese family registers have recorded the furigana of names; the registered pronunciation generally must be recognizable as a reading of the written name, subject to provisions for established existing usage.[100][101][102]
Established Chinese names in Japanese have traditionally been written in kanji and read with Sino-Japanese pronunciations, while less familiar names are often accompanied or replaced by katakana approximating the source-language pronunciation. Postwar official guidance retained internationally conventional Japanese forms for a limited set of well-known place names.[103]
| English name | Kanji | Katakana | Romanization |
|---|---|---|---|
| Beijing | 北京 | ペキン | Pekin |
| Shanghai | 上海 | シャンハイ | Shanhai |
| Hong Kong | 香港 | ホンコン | Honkon |
| Taipei | 台北 | タイペイ / タイホク | Taipei / Taihoku[104] |
The same character can appear twice with different surface forms of its reading. Most commonly, the second occurrence undergoes voicing (rendaku), as in 人々 hitobito "people"; other repeated-character words may contain unrelated readings fixed by the word's history.[9]
Pronunciation assistance
The intended pronunciation of kanji can be shown with ruby characters known as furigana—small kana placed above horizontal text or to the right of vertical text, as in 振仮名. Furigana are used for unfamiliar or non-standard readings, in materials for children and language learners, and when a writer deliberately pairs a written expression with a different spoken form.[9][105]
Spelling words
Because many words and characters share a pronunciation, a kanji cannot always be identified from its reading alone. In speech, a character may be specified by a familiar word that contains it, by its native reading, or by describing its components; dictionaries instead provide standardized indexes by reading, radical and stroke count.[9][34]
Dictionaries
General Japanese dictionaries normally give word readings in kana. Character dictionaries conventionally distinguish on readings in katakana from kun readings in hiragana, and may separate the part represented by the character from following okurigana; for example, 食べる taberu may be shown as た.べる. Character entries also commonly provide compounds and indexes by radical and stroke count.[9][34]
Local developments and divergences from Chinese
Kanji descend from Chinese characters, but their Japanese use has developed independently over many centuries. Differences from modern Chinese usage include characters coined in Japan, characters assigned specifically Japanese meanings, and postwar Japanese simplifications known as shinjitai.[9][19] These differences include:
- the use of characters created in Japan,
- characters that have been given different meanings in Japanese, and
- post-World War II simplifications (shinjitai) of the character.
Mainland Chinese and Japanese simplification policies also produced different standard forms for some historically corresponding characters.[106]
Kokuji
Characters created in Japan according to Chinese structural principles are called kokuji (国字; national characters) or wasei kanji (和製漢字; Japanese-made kanji). They are often semantic compounds and ordinarily have only native readings, although exceptions such as 働 have an on reading derived from a phonetic component. Other familiar examples include 峠 tōge "mountain pass", 畑 hatake "field", and 辻 tsuji "crossroads".[107]
Comparable locally coined characters occur in other writing traditions derived from Chinese characters, including Korean gukja, Vietnamese chữ Nôm, and Zhuang sawndip.[108]
Kokkun
In addition to kokuji, existing Chinese characters have sometimes acquired specifically Japanese meanings. These are not newly created characters; the Japanese usages are called kokkun (国訓). Authoritative Japanese dictionaries give examples such as 鮎 ayu "sweetfish", where the corresponding Chinese character historically denoted a catfish.[109] Further examples include:
Types of kanji by category
Han-dynasty scholar Xu Shen, in his 2nd-century dictionary Shuowen Jiezi, classified Chinese characters into six categories (Chinese: 六書 liùshū, Japanese: 六書 rikusho). The traditional classification is still taught but is problematic and is no longer the focus of modern lexicographic practice, as some categories are not clearly defined, nor are they mutually exclusive: the first four refer to structural composition, while the last two refer to usage.[114][115]
Shōkei moji (象形文字)
Shōkei (Mandarin: xiàngxíng) characters are pictographic representations of objects. Examples include 目 "eye" and 木 "tree". Their modern forms may differ greatly from the earliest inscriptions, and pictographs make up only a small fraction of the character inventory.[115]
Shiji moji (指事文字)
Shiji (Mandarin: zhǐshì) characters, often called simple indicatives, use graphic relations to represent abstract concepts; examples include 上 "above" and 下 "below". They make up a very small fraction of modern characters.[115]
Kaii moji (会意文字)
Kaii (Mandarin: huìyì) characters, often called associative compounds, combine components interpreted together for meaning. A traditional example is 休 "rest", analyzed from 亻 "person" and 木 "tree". The Japanese-created 峠 "mountain pass" similarly combines 山 "mountain", 上 "up" and 下 "down".[115][114]
Keisei moji (形声文字)
Keisei (Mandarin: xíngshēng) characters are phono-semantic compounds and form by far the largest structural category. Typically one component suggests a broad semantic class and another approximates the character's pronunciation at the time the compound was formed. Historical sound change and the borrowing of Chinese readings into Japanese can make the phonetic relationship less transparent in modern on'yomi; it does not predict a character's native kun'yomi.[115][114]
Tenchū moji (転注文字)
Tenchū (Mandarin: zhuǎnzhù) characters have variously been called "derivative characters", "derivative cognates", "mutually explanatory", or "mutually synonymous" characters. It is the least clearly defined of the six traditional categories, and proposed examples depend on how the category is interpreted.[115]
Kasha moji (仮借文字)
Kasha (Mandarin: jiǎjiè) are rebuses, sometimes called "phonetic loans". A graph originally associated with one word was borrowed to write a different, similar-sounding word. For example, 來 originally depicted wheat but was borrowed to write the homophonous word "come"; a differentiated graph, 麥 (modern Japanese 麦), was then used for "wheat".[115]
Related symbols
The iteration mark (々) indicates repetition of the preceding kanji, functioning similarly to a ditto mark in English. It is read as though the kanji were written twice, for example iroiro (色々; "various") and tokidoki (時々; "sometimes"), and also appears in names such as Sasaki (佐々木). Its origin is uncertain; dictionaries note that it is sometimes described as a cursive or simplified form related to 仝, a variant of dō (同; "same").[116][117]
Another abbreviated symbol is ヶ, in appearance a small katakana ke, but actually a simplified version of the kanji 箇, a general counter. It is pronounced ka when used to indicate quantity (such as 六ヶ月, rokkagetsu "six months") or ga if used as a genitive (as in 関ヶ原 sekigahara "Sekigahara").[118]
Collation
Kanji dictionaries commonly use radical-and-stroke sorting. Characters are assigned a principal radical, grouped under that radical, and then ordered by their remaining or total stroke counts according to the dictionary's convention. For example, 桜 "cherry" is indexed under the four-stroke radical 木 "tree".[34]
Other indexing methods include Jack Halpern's SKIP system, which classifies characters by their visual component pattern and stroke counts.[119]
Modern general-purpose Japanese dictionaries, as distinct from character dictionaries, normally arrange headwords by their kana readings in gojūon order.[9]
Kanji education
Japanese elementary-school pupils are taught the 1,026 kyōiku kanji according to the grade-by-grade allocation table: 80 in the first grade, followed by 160, 200, 202, 193 and 191 in grades two through six. The curriculum also introduces the reading and writing of additional jōyō kanji in lower secondary school, but the official list is a guideline for written communication rather than a definition of complete literacy.[32][10]
Learners of Japanese as a foreign language use a range of approaches, including repeated handwriting, component analysis, vocabulary-based study and verbal or pictorial mnemonics. Research on kanji learning emphasizes that strategy choice varies with proficiency, prior knowledge of Chinese characters and the learning context.[120]
The Japan Kanji Aptitude Testing Foundation provides the Kanji kentei (日本漢字能力検定試験 Nihon kanji nōryoku kentei shiken; "Test of Japanese Kanji Aptitude"), which tests the ability to read and write kanji. The highest level of the Kanji kentei tests about six thousand kanji.[121]
See also
- Chinese influence on Japanese culture
- Braille kanji
- Hanja (Korean equivalent)
- Chữ Hán (Vietnamese equivalent)
- Han unification
- Chinese family of scripts
- Japanese script reform
- Japanese typefaces (shotai)
- Japanese writing system
- Kanji of the year
- List of kanji by stroke count
- Radical (Chinese character)
- Stroke order
- Table of kanji radicals
- Rōmaji – method of writing Japanese with the Latin alphabet
- Cangjie – legendary inventor of Chinese characters
