The introduction of the Chinese script and the origins of kana
Ancient Japan did not have its own writing system. The Chinese script was imported from China but the way it was used was distinctive.
On and kun readings
In Japanese, Chinese characters (kanji, literally, “Han letters”) are read in several different ways, which famously makes life difficult for learners of Japanese. The main two pronunciations are the “on” (Chinese-style) reading and the “kun” (Japanese-style) reading. However, as a result of the complicated history of character pronunciation in Japan, many kanji have more than one “on” reading. In early times, Buddhist monks popularized the so-called Wu reading (J. go-on), or the pronunciation in use in the south-eastern Chinese kingdom of Wu. In the 7th and 8th centuries, the kan-on (“Han” pronunciation) reading, which was in use in the area of the Tang capital Chang’an, was introduced, and after the 13th century the so-called tō-on (“Tang” reading) was introduced. Because they were associated to specific areas of knowledge, ways of pronouncing and reading kanji were never standardized and, still today, Wu readings are used for Buddhist terminology, Han readings are common in legal language, and so forth. For example, the character 経 can be read kyō (go-on), kei (kan-on), and kin (tō-on) in different contexts; the character 行 can be read gyō (go-on), kō (kan-on), and an (tō-on).
Because Japan did not have its own writing system, there was no choice but to use the Chinese script. Even texts that dealt with Japan were written in Chinese (by translating them first into Chinese), but naturally there was also a desire to record the Japanese language as it was. This was achieved by using Chinese characters sometimes for their meaning (e.g. 山 = mountain) and sometimes for their sound, to represent the sounds of the Japanese language according to both their “on” and “kun” readings.
Because this system was used in the 8th-century poetry anthology Man’yōshū (Collection of Ten Thousand Leaves, c. 759), it is often called man’yōgana (“Man’yōshū-style kana”).
Fig.2 Example of Man’yōgana: Man’yōshū (Click to take a closer look)
As an example, let us look at one of the chronologically newer poems in the Man’yōshū, poem no. 4291 (Book 19) by Ōtomo no Yakamochi (c. 718-785), who is thought to have been the final compiler of the anthology:
和我屋度能 伊佐左村竹 布久風能 於等能可蘇気伎 許能由布敝可母
This string of characters is read: wa ga yado no / isasa muratake / fuku kaze no / oto no kasokeki / kono yūbe kamo (By my house / the bamboo grows in little clusters, / softly the leaves rustle / as the wind rises / on this quiet evening). Most of the characters are used phonetically according to the “one character per sound” principle, except for 村竹 (muratake) and 風 (kaze), which are read according to the kun reading. Thus, a method to write down the Japanese language was devised. However, because of the characteristics of the Japanese language, with this method sentences tended to become very long, and some of the characters were too elaborate as representations of single phonemes, making reading, and especially writing, too time-consuming. The hiragana and katakana syllabaries were developed as solutions to these problems.
Hiragana and katakana syllabaries
The hiragana letters are simplified or cursivized Chinese characters: the graphあ (a) is a simplification of the character 安; the graphい (i) is a simplification of 以; う (u) is a simplification of宇;お([o]) derives from 於, and so forth. By contrast, katakana graphs were created by taking elements or parts of Chinese characters. For example, the graph ア (a) is the left-hand part of the character 阿, the graph イ(i) is the left-hand part of the character 伊, the graph ウ(u) is the crown of 宇, the graph エ(e) is the right-hand part of the character 江, and the graph オ(o) is the left-hand part of the character 於. If we place the hiragana and katakana graphs next to the original Chinese characters, we can clearly see that, in both cases, the aim was to create a simplified phonetic script:
Katakana chart One thing that did not change after the invention of the kana syllabaries is that the same sound could be written in several different ways using different graphs. Despite slight variations by period, throughout the pre-modern period such variety was the norm. It was only in 1900 that the system was standardized, and only one way to write each sound or syllable was established. The obsolete characters are today called hentaigana, or, “non-standard kana.” There are now digital applications for learning how to read traditional kana (The Hentaigana App by Waseda University and University of California Los Angels; KuLA [iOS or Android] by Osaka University). If you are interested, they are very helpful tools to learn the traditional characters.
The word kana is written with the characters 仮 (ka, a contraction of kari, temporary) and 名 (na, names, signs). The term kana, therefore, means “temporary” or “provisional names.” The “hira” in hiragana means “simple,” whereas the kata in katakana means “section of”, “part of,” “incomplete.” To differentiate them from these new, simplified writing systems, the kanji were regarded as the “proper” way of writing and called “mana” (“real names”).
Because Heian sources sometimes refer to hiragana as “the woman’s hand” (onna-de) and to kanji based scripts as “the man’s hand” (otoko-de), it is thought that it was primarily women who used kana. One text that shows the special link between women and hiragana is the Tosa nikki (The Tosa Diary), a journal written in 935 by Ki no Tsurayuki, which describes his return voyage to the capital from Tosa (present-day Kōchi prefecture) and which is famously written from the perspective of a woman, in hiragana.
As the use of hiragana spread, it replaced man’yōgana as the script of choice to write down waka poetry, and, as waka was a courtly genre and served as a medium of communication among members of the aristocracy, male aristocrats also began to use it. However, throughout the entire pre-modern period, formal documents and male courtier’s diaries continued to be written in the kanji-based script, which remained the more prestigious variety.
By contrast, katakana, which was first devised as a notation system to render Chinese texts into a form of Japanese, was primarily used by, and remained traditionally associated with, men and the clergy. Because it was not commonly used to write new texts but to annotate existing ones, there are much fewer examples of texts written in katakana compared to hiragana. The rare extant examples of monogatari and waka collection written in katakana were most probably the work of members of the clergy.
Fig.3 Example of katakana text: Kokinwakashūjo-chū (Annotated preface to the Kokinwakashū). (Click to take a closer look)
The katakana is thought to have been developed at the beginning of the 9th century and the hiragana during the second half of the 9th century. By the end of the 9th century, therefore, three different scripts were in existence. The kana syllabaries were sometimes used on their own, but a mixture of kana and kanji was the more common choice.
Fig.4 Example of text mixing kana and kanji: Kokinwakashūjo-chū. (Click to take a closer look)
As a general rule, hiragana and katakana were not used together in the same text, and in the rare occasions that they do appear together, they show that they were perceived as different and used differently.
© Keio university