Thoughts on the evolution of the Japanese writing system, and implications on learning and using 漢字

The information in this article is based on Yaeko Sato Habein, The History of the Japanese Written Language, University of Tokyo Press (1984) ISBN 4-13-087047-5.

It may be hard to believe, but Japanese was an oral-only language (with no written form) until probably the 6-7th century AD. In contrast, writing in China originated from as early as the latter half of the 2nd millenium BC (and perhaps even earlier). Chinese writing was introduced into Japan either via trade (incriptions on mirrors, swords, etc.) or religion (Buddhist literature). Therefore, it is no surprise that when the Japanese first started writing, they used Chinese characters.

Because Chinese is a different language from Japanese (with different syntax, grammar and vocabulary), the only way to write was to translate Japanese into Chinese, and then write out the Chinese characters. This is called 漢文 かんぶん (Chinese style of writing) and the earliest example dates from 604 ( 憲法十七条 けんぽうじゅうしちじょう ). 純漢文 じゅんかんぶん is genuine Chinese writing, and a Chinese reader will have no difficulty reading it. Even today, 漢語 かんご (words borrowed from Chinese) are quite common in modern Japanese, although the use of 漢語 no longer differentiates the educated from the illiterate.

Soon thereafter, some Chinese characters were borrowed phonetically to represent Japanese personal names and names of places (which were otherwise untranslatable into Chinese). This usage is called 万葉仮名 まんようがな . Also, the use of 漢文 began to deviate from authentic Chinese to include the use of Japanese honorifics and syntax (such as word order), this style is called 変体漢文 へんたいかんぶん . However, word inflections and particles were still not included even in 変体漢文.

The use of 万葉仮名 expanded in the eighth century until entire poems could be written in this style. Chinese characters were used either in 借音 しゃくおん style (sound borrowing): approximating Chinese pronunciation of those characters (with complete disregard of meaning), or in 借訓 しゃくくん style (by borrowing the first syllable of the Japanese word corresponding to the meaning of those characters.

Even in 借音 style, however, one needs to be careful as there are multiple ways to pronounce the Chinese characters, either by approximating the Chinese pronunciation used in the T'ang capital ( 漢音 かんおん reading, broadly similar to modern day Mandarin) or the pronunciation brought in via Korea ( 呉音 ごおん reading, which is somewhat similar to modern day Cantonese).

Also, inflections and particles are beginning to be incorporated into the 宣命書き せんみょうがき style of writing by using two different sizes for writing characters: the larger font was used for meaning ( 正訓 しょうくん ) but the smaller font was used for particles/inflections.

By the ninth century, two Japanese syllabaries have evolved, corresponding to modern day カタカナ and ひらがな. カタカナ evolved from priests studying Buddhist texts written in Chinese during the Heian period. As the priests were required to read aloud these texts in Japanese, they developed a set of diacritical marks to be written next to the Chinese characters as an aid in the reading process. One set of marks, called okototen, denoted when to add particles such as を, こと, の, は, に, て and also inflections such as す and む (for verbs). The other set of marks were a pronunciation guide to reading the characters themselves. Originally, 万葉仮名 was used, but over time these characters were abbreviated and simplified (due to lack of space between the Chinese characters in the text) to become カタカナ. From this a style of writing known as 仮名交じり文 かなまじりぶん (containing a mixture of カタカナ and Chinese characters written in Japanese syntax) evolved.

ひらがな also developed from the cursive form of 万葉仮名 ( 草仮名 そうがな ) but originally for the purposes of calligraphy and writing poems. It was also called 女手 おんなで or 女文字 おんなもじ because it was used by women (who were excluded from the study of Chinese). It was instrumental in the development of Japanese literature ( 和文 わぶん ) and entire texts were written in ひらがな (the most famous of which is The Tale of Genji [ 源氏物語 げんじものがたり ] by 紫式部 ムラサキシキブ ).

By the Kamakura period, both 和文 and 仮名交じり文 were used together for documenting narrated tales in a style that became known as 和漢混淆文 わかんこんこうぶん .

It is important to note that the various styles of writing more or less coexisted till the twentieth century, and therefore there was not one written form of Japanese but many, with various degrees of alignment to the spoken language as well as alignment to Chinese. These various forms were used depending on the situation: 純漢文 continued to be used for official or royal communication, whereas 変体漢文 tend to be used for personal diaries and communication between educated men. The other styles continuted to be used for narration, prose/verse writing and religious studies.

However, as the spoken language evolved over the years, the writing styles did not and by the 室町 period essentially the written and spoken forms of the language has become quite different. The 明治 period, which reintroduced Japanese intellectuals to European languages, also brought upon the desire to reunify the written and spoken languages. Even then, 口語文 こうごぶん (written style in the spoken language) which is the basis for modern day written Japanese did not become pervasive until after World War II (along with restricting the number of 漢字 for general use, and various spelling reforms). The book Nanette Gottlieb, Kanji Politics, Language Policy and Japanese Script, Kegan Paul International, London and New York (1995) offers a fascinating account of how hotly debated language policy and script reform (国語国字問題, national language and script problem) is, even till today.

In summary then:

• Japanese originally evolved purely as an oral language.
• Initial attempts at writing involved translating Japanese into Chinese, and then writing in Chinese.
• Even after phonetic syllabaries were developed, the written form of Japanese differed substantially from the spoken form until relatively recently (after World War II).

This has a number of important implications for non-native learners of Japanese, particularly in the study of 漢字.

It is important to realise that 漢字 is not the "heart and soul" for representing Japanese, as many non-native beginning learners assume (based on the high usage of 漢字 in modern Japanese writing) but a set of foreign characters borrowed from another language (Chinese) and initially only used to represent words in Chinese (later on, many of these words were then assimilated into Japanese and became known as 漢語).

The usage of 漢字 in Japanese can be very different from usage of these characters in Chinese, due to the practices of 借音 and 万葉仮名 in the past. Therefore the usage of 漢字 may not be related to the original Chinese meaning or pronunciation of these characters. A good example is 風呂 (Japanese hot bath) - the meaning of the word bears no relationship to the individual characters that form the word (風 meaning 'wind', and 呂 meaning 'spine'). The word 風呂 is written that way because the 音読み pronunciations of the individual characters happen to correspond to the pronunciation of the native Japanese word ふろ. Similarly, the use of 漢字 in names of people and places may bear no relationship to the original meanings and pronunciations of the individual characters.

Many beginning students of Japanese naively assume that 漢字 are ideograms or logograms (believing that these characters have meanings or stand for parts of words) - this is further complicated by some scholars, teachers and native Japanese also believing this. After all, most of the words taught to beginners are either words or parts of words, and are usually pictographs (leading to another common beginner misunderstanding - that all 漢字 originated from pictographs). Examples include 山 meaning "mountain," 日 meaning "sun" etc.

In fact, 漢字 is more often used for sound rather than meaning, and "... the function of kanji in modern Japanese orthography is simply to replace strings of hiragana according to a vast tradition of accumulated conventional rules." (J Marshall Unger, Literacy and Script Reform: Reading Between the Lines, Oxford University Press 1996). There's research that suggests that when native Japanese speakers read 漢字, they are replacing the characters with readings (sounds) rather than meaning (see, for example, Sachiko Matsunaga, The Role of Phonological Coding in Reading Kanji, Technical Report #6, Second Language Teachingt and Curriculum Center, University of Hawaii at Manoa, 1995).

To make matters worse, native Japanese words often have a broad usage that do not map well to the relatively narrower meanings of Chinese characters. A good example is the classical Japanese word よし which was represented by several characters including 淑, 良, 吉, 好, 艻 depending on the specific nuance. In modern Japanese, we find common verbs such as あく (open) having different forms including 開く vs 空く depending on nuance (open an object, open an empty space). Similarly, for なく, 鳴く is the cry of an animal, but 泣く represents human tears. It would seem that あく (as well as なく) was originally a single verb that has now bifurcated into two distinct words that sound exactly the same (homophones) but written differently (reflecting different nuances of meaning). Some words can have multiple alternative 漢字 forms: one example that strikes to mind is きれい which can be written as either 綺麗 or 奇麗.

Given the above then, it would seem that maximising the use of 漢字 is not a good idea. For example, きれい (pretty, clean) for example is best written in hiragana, rather than either 綺麗 or 奇麗. Native Japanese words are probably better written in ひらがな rather than being pedantic and using the different 漢字 forms with different nuances in meaning. 漢字 is probably best reserved for 漢語 (words originally borrowed from Chinese). There's even a recent trend to use カタカナ for 漢語, to highlight that the word is 'borrowed' from another language and not native to Japanese. Certainly, words borrowed from a language other than Chinese but with a 漢字 form (such as 煙草) should just be written in カタカナ (eg. タバコ instead of 煙草). Some grammatical forms, such as ～てみる (try to ~) should never be written in 漢字 (ie. ～て見る). Arguably, even ～にいく (go to do ~) should not be written like ～に行く even though some people write it like this.

Likewise, overuse of 漢語 may not be a good idea. For many years, given that Chinese was only taught to the aristocrats or the literate elite, frequency of usage of 漢語 was what differentiated a literary person from the uneducated, and the also as an indication of the level of literacy. You will still find that older people in Japan tend to use archaic 漢語 more often than young people, so if a Japanese student overuses 漢語 they may sound somewhat anachronistic.

But if 漢字 is so foreign, why bother retaining it in modern written Japanese? Indeed, many have argued exactly this as part of the script reform debate - the use of 漢字 should be eliminated and everything should be written in either ひらがな or カタカナ (or even ローマ字!). Indeed, the original intention of limiting the use of 漢字 as well as limiting the use of special readings of 漢字 in general purpose communication was the hopeful gradual elimination of 漢字.

However, it also seems that many Japanese have sentimental reasons for preserving the use of 漢字. In addition, there seems to be valid reasons for retaining 漢字 - the characters aid in parsing written text, and help define boundaries between words. 漢字 also greatly helps in clarifying the usage of homophones, of which there are many in Japanese (due to the restricted sound set). In any case, one of the reasons for advocating the elimination of 漢字 (to assist in the use of technology and communicating Japanese to a world that largely depends on the Latin alphabet) seems less relevant, now that computers are powerful enough and sophisticated enough to handle non-Latin character sets (even mobile phones have no problems allowing users to enter 漢字).