Wednesday 07 May 2003

Hannas revisited

Mark Griffith from pushed back against my post about William C. Hannas, the “Master Linguist” in such a pleasant and well-informed way that his comment deserves reproducing:

Mind you (trying to sound most demurring and non-confrontational here!) isn’t it the case that Japanese newspapers and books are crammed full of hiragana and katakana alongside the imported (and very occasionally home-grown) Chinese kanji?

The last time I did a rough count, something approaching sixty per cent of the ink marks on a Japanese newspaper page were from one of the two syllabaries. And Japanese dictionaries spell kanji for users using hiragana (most of which users I assume are above kindergarten age). Of course most kanji are multi-syllable, but a small number are also one-syllable in their spoken form.

Indeed Japanese text is crammed full of (monosyllabic) hiragana and katakana alongside the Chinese kanji characters, as can be seen in this introduction to the career of the actress Hara Setsuko, star of films by Ozu, Kurosawa, and Naruse (the hiragana appear in blue, katakana in red, kanji in black).

Introduction to Hara Setsuko's career in Japanese

And Mark’s estimate that syllabic hiragana and katakana characters outnumber the kanji is also correct, as Jack Halpern makes clear in the introduction to his New Japanese-English Character Dictionary:

A running Japanese text consists of a mixture of kanji and kana, with the latter normally outnumbering the former.

In the example, hiragana are used for particles such as wa, ga, nado, no, ni, mo as well as for verb endings such as -shita, -rete, and -shite imasu. Katakana are used for loan words such as terebi (television).

However, it’s important to note that, since Japanese text is written without spaces, the process of reading involves skipping from one set of kanji characters to another. As Halpern explains:

Hiragana characters server as natural borderlines that help the reader segment the text into meaningful units. For this reason, a Japanese text is easier to read than a running Chinese text, which consists of Chinese characters only.

It’s also important to stress that much of the “meaning” of the text comes from the Chinese characters (in the same way that one could get the gist of an English text even if the prepositions, pronouns, and verb endings were missing).

And while it is true that “Japanese dictionaries spell kanji for users using hiragana”, that only goes to prove that Japanese is not fundamentally a phonetic language, since the hiragana are provided to give the pronunciation, not the meaning—a Japanese reader can look at a forgotten or unfamiliar kanji and be able to figure out its meaning, without necessarily knowing its pronunciation. To be fair, the form of the kanji often suggests possible pronunciations, but that’s very different from the phonetic (“each character corresponds to a syllable of sound”) definition of Japanese that Mr Hannas suggests.

So isn’t it a bit steep to say it’s utter nonsense to claim Japanese uses characters representing syllables, and that only kindergarten children spell with kana? Sounds to me a rather reasonable simplification to introduce new readers to a tricky language with a fascinating hybrid script.

Not really, since only kindergarten children spell exclusively with kana whereas Hannas quite specifically (and wrongly) states that “the writing systems of East Asia, including Chinese, Japanese and Korean, are ‘syllabaries,’ in which each character corresponds to a syllable of sound.” That is only true for the writing system used by kindergarten students. As soon as they start elementary school, Japanese children are rapidly introduced to multisyllabic Chinese characters.

Is it Hannas’ main argument you don’t like? The bit about precision overriding innovation? That’s obviously a big claim of his, but the syllable component of Japanese seems slightly more than utter nonsense to me.

No, I don’t necessarily disagree with Hannas’ contention that East Asia has failed to make significant scientific and technological breakthroughs compared to Western nations (although, to be honest, there is definite disagreement about whether or not this is actually true). But I think he’s absolutely wrong to blame the writing systems of China, Japan and Korea for that. [Does he also blame the Thai and Vietnamese writing systems for those countries having failed to make “significant scientific and technological breakthroughs”?)

To the contrary, I strongly believe that one of the reasons for Japan’s rapid and successful industrialization after the Meiji Restoration in the mid-nineteenth century is that their uniquely flexible writing system—coupled with a historical willingness to accept ideas from abroad—allowed the Japanese to easily import, comprehend, and put to use an astonishing range of Western cultural, political, aesthetic, and technological ideas.

I have an alternative (language-based) theory for why Japanese science and technology may be less innovative than their Western counterparts, but that will have to be the subject of another post.


Trevor Hill at Glome offers an absolutely first rate rejoinder to Hannas’ nonsensical assertions about East Asian languages, including a persuasive argument that Chinese characters actually facilitate abstract thinking. (I’d actually been planning to write a post along the same lines, based on some ideas in the Halpern essay I cited above, but Trevor has done such a great job that I may not bother.)

Permalink | Technorati


Excellent points. One minor quibble: when you say "Japanese is not fundamentally a phonetic language," you mean "kanji are not fundamentally phonetic," right?

Posted by language hat on 7 May 2003 (Comment Permalink)

Aha - I shall have to back away from this most civilised exchange about where exactly Japanese speakers locate the meaningfulness, as it were (in the kana or the kanji), though I assume you'd agree that the attractively ball-like 'no' kana, dotted in soft blue through the text you cite does often encode 'belongs to' for Japanese readers?

A bit like claiming (I nervously suggest) that the -nak or -nek on the end of Hungarian nouns doesn't actually mean anything in itself, even if it tells readers/listeners that something belongs to that noun?

We're going right to the edge of my sad crumbs of knowledge of Far Eastern languages here, _and_ I haven't read a word of Mr Hannas' text, so really must defer.

I can also buy the idea they have a uniquely flexible writing system with lots of clever features. Just having two syllabaries (alphabets?) gives them a fantastic way to shift voice in a text, for example. My hunch (and perhaps Hannas also only works from hunches) is that all the exquisitely squiggly Oriental scripts, Thai, Cambodian, Arabic etc included do have a slight double effect (a) emphasising fidelity and precision and (b) cramping innovation a bit, but there could be so many positive effects bundled into (a) we should all celebrate and enjoy their survival.

If anything Arabic might militate against the Hannas thesis, since Far Eastern societies like Japan, Korea, China have in their recent catch-up (ignoring their earlier glory days for a moment) did a lot better than Arabic-speaking countries. Yet, from my position of hopeful ignorance, Arabic strikes me as a more flexible writing system than Japanese - at least in the Hannas sense of allowing a bigger range of accepted calligraphic styles. There's some gorgeous Far-Eastern calligraphy, but Arab calligraphers could produce elaborate geometric grids, pictures of camels etc, all woven out of fairly readable script. Yet those cultures are more troubled by inflexibility the last couple of centuries, so it's clearly about a bit more than writing.

Should I read Hannas, or is the poor man already in the outer darkness?

Posted by mark on 7 May 2003 (Comment Permalink)

Regarding Arabic...

It's a very good counterexample, considering what a failure modernization has been in most Arab countries, coupled with the fact that Arabic is an alphabet not too different from western ones. Clearly, the problems in the Arab world are not as much linguistic as cultural, although this brings me to a couple of interesting points.

Partly because of the many cultural problems Arabs have in accepting western thought, the literary world of Arabic is horribly barren. I read recently in a journal on MidEast affairs that last year only about 100 books were translated into Arabic from foreign languages.

Can you imagine that! They contrasted that figure with Spain, where over 100,000 books were translated into Spanish the same year... No wonder they have a serious dearth of new ideas and a problem with modernization and education...

BTW, the Arabic script is an example of a so-called "defective" script. This means linguistically that it fails to represent enough information to completely pronounce a word. Arabic does not write short vowels (most of the vowels are short), so basically you have texts full of only consonants, forcing the reader to infer the vowels much of the time.

This isn't a problem really though, since semitic languages carry most of their semantic information in the consonants and few long vowels...

Posted by Trevor Hill on 9 May 2003 (Comment Permalink)

This discussion is now closed. My thanks to everyone who contributed.

© Copyright 2007 Jonathon Delacour