Writing System

How Japanese writing looks like?
 One of the most significant properties of the Japanese writing system is it's representation by a mixture of different characters. These characters are known as Kanji (Chinese characters), Hiragana and Katakana (two phonetic syllabaries). Kanji was adopted from Chinese language in ancient days, although it has been developed differently to some extent in both countries, therefore Kanji and Chinese characters are not identical in modern usage. In Chinese, one character of Kanji usually represents one word, therefore Kanji is often called an ideogram or logorgam, which refers to symbols and which represents whole words or concepts/ideas. Hiragana and Katakana were created by modifying Chinese characters. Theoretically, Japanese could be written with either Hiragana or Katakana alone, although normal writing consists of Kanji 20-30% (1) as well as Hiragana and Katakana. Hiragana usually appears with the combination of Kanji in normal writing (2), on the other hand, Katakana(2) is basically used to represent words and names borrowed from other languages such as テニス tenisu 'tennis', バッグ baggu 'bag, and ロンドン rondon 'London'. It could also represent sound, that is onomatopoeia such as バタン batan 'slam' and ブンブン bun-bun 'buzz'. These days, Katakana is often seen in the form of advertisement, names of goods, items in a menu and so on, therefore beginning learners might feel it is more useful to learn Katakana first rather than Hiragana.


The nature of Kanji
 Kanji in Japanese is most distinctive and could be complicated with its relationship among three basic features: shape, sound and meaning(4), in contrast with other scripts such as the alphabet, which has only shape and sound. One Kanji could have more than one pronunciation depending on whether it is used to represent a word of Chinese or Japanese origin(5). The Kanji 山, for instance, which means 'mountain', could be pronounced san in the word of Chinese origin and yama in the one of Japanese origin.


How many Kanji should we learn?
 It is said if learners have the reading competence of approximately 2,000 kanji, it might facilitate their reading of Japanese articles in general(1). 2,000 Kanji is equivalent to the grade one level of Nihongo Noryoku Shiken (Japanese Language Proficiency Test)


Kanji classification
 To simplify Kanji learning, some ways to arrange them systematically have been developed, although there is no absolute way to classify Kanji. The most common way is to take the Kanji apart into elements.

 The structural size of one Kanji is generally uniform, though Kanji usually consists of some basic elements, the number of which varies among characters. One element sometimes constructs one Kanji by itself, for instance, 人 hito 'person', 大 oo(-kii) 'big' and 五 go 'five'. Whereas, some Kanji are constructed by several elements, for instance, 松 matsu 'pine tree', 時 toki 'time' and 国 kuni 'country'.

 In the Kanji 拍 HAKU 'beat', for instance, the right part 白 represents the sound but it has nothing to do with the Kanji's meaning as 白 means 'white' originally. While the left part, which is originally formed from 手 te 'hand', shows that this Kanji is related to hand. The part providing information about the meaning of the Kanji is often called 'radical', although radicals are not always related to the meaning of the Kanji.

 There are some methods to group Kanji by structural features such as pictographic (Shookee), indicational (Shiji), semantic (Kaii) and phonetic (Keesee) (8). Some books for Kanji learning are edited based on these methods.


Further information:
1. How many Kanji is used in Japanese?
 The Japanese writing system depends heavily on Kanji, although the total number of Kanji is controversial. Dai-kanwa-jiten (the great Kanji-Japanese dictionary), which is considered to be the biggest Kanji-Japanese dictionary, contains approximately 50,300 Kanji (1). According to research (2), the ratio of Kanji content to the whole script in one text varies from about 20% to 30%, depending on the type of the text. Although the Kanji content tends to be reduced these days (3), Kanji is still an essential part of the language which learners should face.

 Regarding the number of Kanji in magazines and newspapers, 3,132 Kanji and 3,328 Kanji are used respectively (4). All of these Kanji are not used equally in terms of frequency. If those Kanji are ordered by frequency, 500 of the most frequently used Kanji account for 79.4% of the total number of Kanji that appear in newspapers, 74.5% in magazines. 1,000 Kanji account for 93.3% and 90.0%, 1,500 Kanji account for 98.4% and 96.0% and 2,000 Kanji account for 99.6% and 98.6%, respectively (5). Thus, if learners memorised 2,000 Kanji, they should have little difficulty reading Japanese texts of those sorts.

2. The Usage of Katakana
 Katakana is also used to vary nuances of meaning. Katakana could sound more modern, stylish or emphasised and make the word somewhat special. This might be the reason why it is often used in the form of advertisement, names of goods, items in a menu and so on. The examples of Katakana for nuanced word are:

ボクは 街へ 出かけた。
Boku-wa machi-e dekaketa. 'I went out to the town centre.'
* The person seems stylish this way.
突然 名前を 呼ばれてビックリした。
Totsuzen namae-o yobarete bikkurishita.
I got surprised to be called my name suddenly.
* 'Surprised' is emphasised.


3. How Kanji works in written Japanese?
 Written Japanese looks like a long string of words, as a sentence does not have any spaces between words in the standard way of writing. In such a language, it seems difficult to distinguish each word to read. However, the combined use of Kanji and Hiragana facilitates reading Japanese properly. In normal writings, free morphemes, which represent a meaning by itself are usually written by Kanji and bound morphemes, which must be attached to another element are written by Hiragana. That is, Kanji represents the main part of a word and Hiragana usually represents the part which could be inflected and called okuri-gana.

 In the word 見るmi-ru 'look', for instance, the Kanji 見 represents the meaning and it is not changeable. The Hiraganaる, in contrast, indicates that this part can be replaced by inflection like 見-た mi-ta 'looked' (-ta indicates past tense). In some cases such as 少ない sukuna-i 'few, little', however, the unchangeable part is represented by the combination of Kanji 少 and Hiraganaな, and only the part which is represented by Hiragana い is changeable, like 少な-かった sukuna-ka-tta 'was few/little'.

4. Shape, sound and meaning
 The most significant characteristic of Kanji is that it can be analysed as having three basic features: shape, sound and meaning. The Kanji 時 toki/ji 'time', for instance, has a certain shape as it looks, and a sound which pronounced as /toki/ or /ji/, and also indicates a meaning 'time'. The Western alphabet, on the other hand, consists of letters that have only shape and sound. These letters can basically represent a meaning when they are combined with other letters to form a word, although some words consist of only one letter. The letter 't', for instance, has its own shape and a sound /ti/, although does not have a meaning by itself. It could have a meaning when (it is delete inside brackets) combined with other letters such as 'i', 'm' and 'e', for example, to give the word 'time'. The Western alphabet, thus, is called phonogram, which represents some aspects of pronunciation.

 Kanji is not simply considered to be either the same kind of script as an alphabet, or the same as a word. It is a script, which has three basic features connected to each other and called ideogram or logogram.

5. On and Kun readings
 Another unique characteristic of Kanji is it has two types of pronunciation, that is, Kun-reading and On-reading. Kun-reading is originally from the Native-Japanese vocabulary, which usually represents the meaning of the Kanji. In earlier days, Japanese people combined a certain Chinese vocabulary, that is, Kanji to the pronunciation of their own vocabulary. On the other hand, the On-reading, which will be represented by capital letters here, is originally from the Chinese pronunciation. The pronunciation of Chinese words has changed in their long history since Kanji was invented, thus, several On-readings possibly exist for one Kanji, according to the era in which the Kanji was imported into Japanese language. For instance, the Kanji 川 'river' can be read either kawa or SEN, while the Kanji 頭 'head' can be read atama, ZU, TOO, ZYUU. Other Kanji however have only one type of reading. In the inventory of Jooyoo Kanji (Kanji with common use), 1,900 Kun-readings and 2,187 On-readings (4,087 in total) appear. On average, one Kanji has 0.98 Kun-reading and 1.12 On-reading. That is, it has 2.1 ways of being read(6). Some Kanji might have some more readings which do not appear in Jooyoo Kanji inventory.

 On-readings show highly homonymous features. The On-reading KAI can be 海, 会, 開 and so on. If homonyms, which share the same pronunciation, were counted as one, there would be only 302 ways of reading compare to the actual number of 2,187 readings. In fact, there are a lot of homonyms in Japanese. The Kanji which have the pronunciation KAI, for instance, amount to 178 for On-reading and 3 for Kun-reading.

6. Kanji in Nihongo Noryoku Shiken
Nihongo Noryoku Shiken (Japanese Language Proficiency Test) presupposes examinees' Kanji competence as shown in Table 1. However, the actual numbers of Kanji used in textbooks which are on the market are varied shown in Table 2.(7).

Table 1. Presupposed Kanji competence for Japanese Language Proficiency Test
Grade Level Number of Kanji Time of learning (hrs)
1 Advanced 2,000 900
2 Intermediate 1,000 600
3 Basic 300 300
4 Beginner 100 150


Table 2. The number of Kanji used in textbooks
Level Number of Kanji
Advanced 370 - 1,400
Intermediate 80 - 1,428
Basic 51 - 700


7. Types of radicals
 The idea of radical is based on the old Chinese dictionary which was issued in 1716. It is used as a key to group Kanji.
 Radicals are grouped into seven types roughly. They are:
  • Left one (Hen)
  • Right one (Tsukuri)
  • Top one (Kammuri)
  • Bottom one (Ashi)
  • Top-left one (Tare)
  • Surrounding one (Kamae)
  • Left-bottom one (Nyoo)
8. Four types of Kanji classification
 Pictographic (Shookee) characters, which are evolved from pictorial representations, show highly pictographic features. "Each pictogram was an image of the object or objects (and, in some cases, concepts) it represented, and, as far as we know, offered no clues to pronunciation" (O'Grady et al 1997:594). 木 ki/MOKU, 目 me/GAN, and 川 kawa/SEN, for instance, are formed from a picture of a tree, eye and stream respectively.

 Indicational (Shiji) characters are similar to Shookee characters, although they are more symbolic to represent more abstract notions such as number and position. Examples of Shiji characters are 一 ichi 'one', 上 ue/JOO 'up, above', 末 sue/MATSU 'end' and so on. Shookee and Shiji might be grouped into one category and called Basic-Form, as they can not be divided into smaller structural elements.

 Semantic (Kaii) characters, on the other hand, are multi-element characters, which are formed from some Basic-Form Kanji, based on meanings. 明 aka(-rui)/MEI 'light', for instance, is formed from 日 hi 'sun' and 月 tsuki 'moon', and amounts to the meaning of 'light'. Three elements of 木 ki 'tree' combine into 森 mori and have a meaning of 'forest'.

 The forth category, phonetic (Keesee) characters are also formed from more than two elements. One character consists of two parts, both of which can often be Basic-Form Kanji themselves, and each of which is called phonetic determinative and radical (or key) respectively. Phonetic determinative provides information about the pronunciation of the Kanji. Radical is a semantic component which provides information about the meaning of the Kanji. 交 maji(-waru)/KOO 'cross', for instance, is a Basic-Form Kanji, and can be combined with radicals such as 木 ki/MOKU 'tree', 力 chikara/RYOKU 'power' and 糸 ito/SHI 'string', each of which forms 校 KOO 'school', 効 KOO 'effect' and 絞 KOO ' squeeze' (Takebe 1989:44). It is said that these characters are in the same 'Kanji family', as they share the same part, the phonetic determinative.

 However, the phonetic determinative is not perfect, as it can not always provide complete information about the pronunciation. The determinative 尺 SHAKU (an unit for length/to connect something successively, in original meaning), for instance, forms 釈 SHAKU 'explain', 訳 YAKU 'meaning of words', 沢 TAKU 'stream', 択 TAKU 'to choose' (Takebe 1989:46). Among these examples, the original sound of the determinative remains only in 釈 SHAKU. Thus, the determinative of a Keesee character does not always represent the predictable pronunciation. Kaiser (1997:37-40), however, claims that they could be in the same group, in which each Kanji is related to each other by means of rhyme. The pronunciation shaku, yaku and taku, in fact, rhyme. Although, there are still some exceptions such as 駅 EKI 'station', which does not rhyme with other Kanji of the same group.

 Examples of each category are shown in Table 3.

Table 3. Four categories of Kanji Structure
      Origin
Pictographic
(Shookee)
yama 'mountain'
kawa 'river'
hi 'sun'
tsuki 'moon'
ki 'tree'
hi 'fire'
me 'eye'
sakana 'fish'
Indicational
(Shiji)
ue 'above'
shita 'below'
HON 'book'
sue 'end'
Semantic
(Kaii)
aka(-rui) 'light' 日 'sun' + 月 'moon'
shin(-jiru) 'believe' 人 'person' + 言 'say'
na(-ku) '(birds) sing' 口 'mouth' + 鳥 'bird'
honoo 'flame' 火 'fire' + 火
Phonetic
(Keesee)
DOO 'copper' 金 'metal' + 同 DOO 'same'
DOO 'body' 月 'body' + 同 DOO 'same'
(All examples are in Ito et al 1982)

Reference:
  1. Mizutani, Shizuo et al (1987) p.31, Moji・Hyooki to Gokoosee [Scripts and Word structure], Tokyo: Asakurashoten
  2. Mizutani et al (1987) p.111
  3. Mizutani et al (1987) p.109
  4. The National Language Research Institute (1962) in Yoshimura, Yumiko (1989) p.221, 'Kanji no Shidoo [Teaching Kanji]' in H. Teramura ed., Kooza Nihongo to Nihongo Kyooiku 13 [Studies in Japanese and Teaching Japanese 13], Tokyo: Meejishoin
  5. Hayashi (1982) in Yoshimura (1998) p.221
  6. Mizutani et al (1987) p.22
  7. Yoshimura (1989) pp.221-223