Comparing writing systems

The 'mixed' Japanese writing system

Japan has an ingenious solution to representing language in writing. Very high rates of literacy are achieved with its mix of writing systems.
Is it because of them or despite them?

A single page in Japanese can contain text in up to six very different writing systems -

kanji Chinese characters,
hiragana syllabic symbols,
katakana symbols for foreign words,
furigana keys to pronunciation of kanji,
arabic numerals
roman letters.

Perhaps English might require some similar type of solution - although in some respects the Japanese seem to have produced difficulties in their own orthography to parallel or even outrank standard English spelling.

Japanese literacy rates have been claimed to be over 99%. Japan's high proportion of tertiary graduates and its performance as a modern technological society give this claim some credibility - although it is not clear how much credit may be due to the orthography, rather than the combination of high motivation and high value on education, diligence, intelligence, opportunities to learn, intensive teaching methods, and over-optimistic reporting.
The Victorian ethic about English spelling saw an admirable preparation for a hardworking industrial society because it required exceptional diligence to learn. Does this apply to Japanese training in the 20th century? After working exceptionally hard to learn to read, can they now work exceptionally hard at anything? With a difficult orthography but a literate populace, Japan contrasts with Brazil (q.v) to show the importance of social factors.

The Japanese language has just over fifty basic syllables, only five vowels and nine basic consonants, and an inflected morphology. This simplicity presents both opportunities and problems for a writing system.

The writing system has a long history, like English, that has resulted in mixtures of principles and discrepancies. While the English adopted a roman alphabet that was relatively unsuited to their language, the Japanese adopted their kanzi logographs from Chinese, which is a very different language from their own. Unlike English, however, there have been some recent reforms (Chapter 3).

The unique blend of scripts in Japanese offers fascinating comparisons for research. Spacing in text is between clauses and sentences, not between words, but visual recognition of words and parts of speech is clarified by the type of script in which they appear, and the mixture adds to their distinctiveness on a page.

Hiragana and katakana are each made up of 48 simple characters, mora, rather like syllables in representing speech units, and this number increases to 71 with diacritics
The problem of many words that sound the same is met in the writing system by using:


Words with content and logical or fundamental meanings


Grammar and some native Japanese words

aquare katakana

Scientific, technical and other foreign words


Small kana spellings that can complement kanji characters, as a key to their pronunciation

roman alphabet romaji

New words, including foreign introductions

arabic numerals

For numbers


Kanji is the core of the writing system for adults. Kanji have more disadvantages than the Chinese hanzi characters they are derived from, which generally have a stable pronunciation in each language that uses them. The pronunciation of kanji can change according to context, combinations can be different from what the parts would suggest, and small phonetic furigana must sometimes be added as clues.

Kanji can also be read in two ways: Chinese loan words are given an on- reading, pronounced rather like Chinese without the tones, and therefore with more homophones, while native words are given a kun- reading. Chao (1968) gives the English analogy of reading e.g. as exempli gratia and etc. as etsetra as similar to on readings, whereas for example, and etc are similar to kun readings. Many characters have from two to four or even more on and kun readings. Compound kanji are 'the real killers' for learners after the first dozen or so, because of uncertainty about the readings to give to them.

Kana characters (hiragana and katakana) are few, simple, easy to learn and to read, and quick to write. The syllable-like mora are relatively easy to segment from speech without requiring sound-blending .

'It only needs the explanation, "This syllable is pronounced like this" and repeat for each of the 50 basic syllables, and the child can read' (Hofman, 1989).

Hofman described how his two-year old son could sound out words one syllable at a time, then repronounce them to hear the word. In contrast, English-speaking children who know their 26-letter ABC and associated 'sounds' still cannot read words with them until they learn the more difficult skills of segmenting phonemes from speech, blending them, and then compensating for complex and irregular sound-letter changes - or else by attempting

The design of romaji is interesting. A restricted set of letters represents a larger number of sounds, as in English, but unlike English, ease of use is preferred to formal accuracy when necessary. Romaji has only recently been mixed with the other scripts, now that writing left to right is acceptable, because romaji cannot be written top down in the traditional way like the other scripts. Now it appears in the popular press and in information for tourists, and most Japanese can spell out their own names and sound out romaji words.

'Japanese is the world's most complex writing system... It seems to make English spelling problems seem small . . . Why do they put up with it, and how?' (Hofman, 1988).

Yet strong movements to shift to simple kana or romaji have lapsed. Japanese can still maintain their popular belief that 'you cannot write Japanese without Chinese characters'. Ministry of Education attempts to reduce or replace kanji to lessen the burden on learners have been resisted by adults who believe they can read kanji better and faster, even though they learned hiragana first. The distinctive forms of the complex kanji used for important words stand out visually from the simple outlines of the hiragana background. The make-up of kanji characters helps readers to get the meaning of an unfamiliar word, similar to the way that English readers can use Latin and Greek roots. Chinese or Koreans can generally understand Japanese text fairly well using kanji and guesswork - which gives Japanese a certain degree of pan-Asian intelligibility . Finally, possibly the most weighty point, Japanese adults are accustomed to their mixed script.

As well, text made up only of hiragana has been found to require more time, eye-fixations, and eye-regressions to read than mixed Japanese script (Sakamoto & Makita, 1973) - although this finding may be due its relative unfamiliarity once readers are adult. Kana syllables cannot distinguish between the many words that sound the same, and each hiragana character takes about the same space as one kanji. The famous word shikaishikaishikai is easier to read when written as a single kanji character. Research to simplify kanji has been primarily to seek greater ease in writing, - since complex rather than less complex characters appear easier to read, because they contain more cues to discriminate characters.

It is possible too that kanji may be integrated into the conceptualisation of vocabulary for literate Japanese. For example, in translating English into their own language, Japanese may write a kanji in the air to obtain the spoken Japanese word from it. This is of interest for English reading theory, since it suggests that there can be a close relationship of phonology to an apparently completely 'visual' character.

Kana are read phonetically and kanji are read visually, with a dissociation between the processes involved, according to Morton & Sasanuma and popular Japanese belief. (This must be a little awkward in reading pages of mixed text, surely?) Nomura found that meaning was extracted faster from kanji than kana words, and thought that kana pronunciation was data-driven and that kanji pronunciation was conceptually-driven. Morton & Sasanuma (1982) also claimed that evidence supports the intuitive belief that kanji can give direct access to the meaning of words, but that kana always require translation into a phonological code when they are being read, and there is no development of automatic visual recognition of the kana symbols. One of the most intriguing ways of studying what differences there may really be in processing is in observing brain-damaged patients - following certain lesions some patients can still read kana but not kanji, while other insults to the brain leave the ability to read kanji but not kana.

Learning Japanese. Children are introduced to reading with hiragana, usually at home or kindergarten before they start school. Once they grasp the key clue that each symbol represents a segment of the spoken language, they are reputed to soon be able to read or write anything in hiragana - although some express doubts about how easy it really is for them. 'Most descriptions of written Japanese greatly oversimplify the problems facing the young reader.' (Stevenson et al. 1984.)

There is also strong cultural pressure to work hard. Heavy rote learning can intensify with the introduction of kanji - described to me by one Japanese: 'The teacher say, 'Today we will learn the word mountain. So we write out mountain 350 times, and then we know the word mountain.'

Near the end of the first grade Japanese children can be learning three forms of writing, with all 1850 kanji essential for literacy taught by late middle-school. Stevenson et al (1982) made a large-scale cross-language study of beginning readers in three cities in America, Taiwan and Japan. They found that Japanese Grade l children were better at sentence comprehension and oral text reading than the children in Taiwan and USA, although they were lagging by grade five, when kanji has a majro emphasis. These two skills are possibly the most crucial measures of 'real' reading, and show that the children were not just blindly decoding the sounds, which has been thought in English-language countries to be the dead-end result of intensive drill-learning.

For a time it seemed that the technological revolution would force Japan to move to an alphabetic script, but in spite of the enormous problems, Japanese are now able to use their ancient and complex writing system most ingeniously in operating the electronic communications technology in which they are leaders.

The Japanese writing system thus has disadvantages that result from its advantages, and other advantages that are linked to its apparent disadvantages, illustrating the problem of orthographic 'best fit' to meet competing demands. Some problems are intrinsic to the language, but others, as with English, are the effects of deviations and aberrations that have developed over time or have never been sorted out.

To foreigners, the Japanese writing system may appear even worse than English, both for learning and for using, as when two Japanese may be unable to give the same verbal translation of an ordinary notice, because the printed words do not carry precise meanings. This would appear to be direct 'reading for meaning' that even bypasses language. But it is at the risk of misinterpretations of that meaning. It may be due to a helpful 'initial learning script', and equally ideal pupil docility, that basic literacy is achievable by all, although some claim that there is much hidden illiteracy and a high proportion of Japanese cannot even read their newspapers (Hofman, 1989).

However, to master the system and its complexities may require a dedication to disciplined learning which has social as well as literate consequences - and if this should slacken, Japanese literacy may suffer at its upper levels.