The History of Kanji in Japan

Before kanji: a spoken language with no script

Japanese was spoken long before it was written. Early Japan had no writing system of its own, so when writing arrived it came from outside, from China, the dominant literate culture in East Asia. The word kanji itself, 漢字, means Han characters, the same two characters as the Chinese word hànzì. Our guide to kanji vs Chinese characters covers how the two systems differ today.

The earliest known example of Chinese characters connected with Japan is a gold seal. Chinese records say that in 57 AD Emperor Guangwu of Han gave it to an envoy from Wa, the old Chinese name for Japan. At this stage, characters were objects from abroad, not something Japanese people used to write their own language.

How kanji arrived: Korea, scholars and Classical Chinese

Much of the knowledge of Chinese writing reached Japan through the Korean peninsula. Japan's oldest chronicles, the Kojiki (712) and the Nihon Shoki (720), tell of a semi-legendary scholar named Wani, sent from the Korean kingdom of Baekje in the early 5th century, who brought Chinese learning with him. Historians treat the details as legend, but the general route, from China via Korea, is widely accepted.

One of the oldest inscriptions made in Japan is on the Inariyama sword, found in a burial mound in Saitama. X-ray analysis in 1978 revealed a gold-inlaid inscription of more than 100 characters, including Japanese personal names, and it is commonly dated to 471 AD. Writing Japanese names was the first hint of a problem: Chinese characters were designed for Chinese, a language with very different grammar and sounds.

For centuries, formal writing in Japan, including government records and Buddhist texts, was done in Classical Chinese. Japanese readers developed kanbun kundoku, a way of reading Chinese text as Japanese by adding small marks that show how to reorder the words and which Japanese endings to add.

Man'yōgana: using kanji for sound

To write Japanese words and poetry as they sounded, scribes began using kanji only for their pronunciation and ignoring their meaning. This system is called man'yōgana, after the Man'yōshū, a large poetry anthology compiled after 759 AD. Use of characters this way is attested from around the mid-7th century.

Man'yōgana worked, but it was slow. Each syllable needed a full kanji, several different kanji could stand for the same syllable, and a reader could not always tell whether a character was meant for its sound or its meaning. The solution was to simplify the sound characters into new, faster symbols.

From kanji to kana: hiragana and katakana

Two sets of kana grew out of man'yōgana, each from a different simplification:

• Hiragana came from cursive, flowing forms of whole man'yōgana characters. 安 became あ and 以 became い. Hiragana was widely used in personal writing and poetry, and it is closely linked with women writers of the Heian period.
• Katakana came from fragments of characters, used by Buddhist monks as a shorthand when annotating Chinese texts. 阿 gave ア (its left side) and 伊 gave イ (its left side).

The result is the mixed writing system Japanese still uses: kanji for the core of words, hiragana for grammar and endings, and katakana for loanwords and emphasis. Our guide to hiragana, katakana and kanji shows how the three work together in a modern sentence, and the endings written in kana are explained in okurigana explained.

Why many kanji have several readings

Kanji were not borrowed all at once. Chinese pronunciation changed over the centuries and differed by region, and Japan borrowed characters, with their Chinese sounds, at different times. Each wave left its own layer of onyomi (sound readings). On top of that, Japanese attached native words to many characters, which became their kunyomi. The basics are covered in onyomi vs kunyomi.

Japanese linguists name three main layers of onyomi. Go-on was already established in Japan before the second wave and survives especially in Buddhist vocabulary. Kan-on was brought back in the 7th and 8th centuries by envoys and student monks returning from Tang China, and reflects the pronunciation of the capital, Chang'an. Tō-on (also called sō-on) came from the Kamakura period onward with Zen monks and traders, and appears mainly in Zen terms and names of imported goods. The same character can carry all three:

You do not need to know which layer a reading belongs to. It is still useful to know why the readings exist: they are not random, they are history preserved in vocabulary. In everyday words one reading usually dominates, so learning kanji through common words, not reading lists, is the practical approach.

Japan also went beyond borrowing. Over time it created characters of its own, called kokuji, by combining existing parts, such as 働 (はたら(く), to work) and 畑 (はたけ, field). It also attached whole native words to kanji pairs, which gave readings like 今日 (きょう), explained in jukujikun and ateji.

Modern reforms: tōyō and jōyō kanji

Before the reforms, educated writing used many more kanji than today, often in complex traditional forms. After World War II, the government reformed the script. On 16 November 1946 the Ministry of Education announced the tōyō kanji list of 1,850 characters for official documents and general writing. It adopted simplified forms (shinjitai) that were already in common use, so 國 became 国 and 學 became 学.

In 1981 the tōyō list was replaced by the more flexible jōyō kanji list of 1,945 characters, described as a guide rather than a strict limit. The current version, issued in 2010, has 2,136 characters, and it is the list Japanese schools teach by the end of compulsory education. If you want to know how many of them you need, and when, see how many kanji are there.

Today the history keeps going in a quieter way. Most Japanese writing is typed, and a keyboard converts kana into kanji automatically, so people recognize far more characters than they can write from memory. For learners, that makes reading, recognizing kanji and knowing their common words the first priority, with writing practice on the most frequent characters. Apps such as Moji, which groups kanji by JLPT level from N5 to N1 and adds guided stroke-order writing, are built around that order.