Reading, writing, and national power

From time to time I've touched on why I'm skeptical about the "inexorable rise of China" meme which dominates a lot of American thinking about the future -- the issue of "zombie" state-owned enterprises which produce little value but are propped up at huge expense to absorb what would otherwise be dangerous numbers of unemployed workers, the likelihood that official figures on economic growth are exaggerated, and the stultifying effect of a totalitarian state upon the open society and free flow of information which are essential to real modernity.  But there's another problem which, while it superficially seems trivial, I believe will be a major factor holding China back.  It's the writing system used by the Chinese language.

The world's two most widely-used writing systems, the Roman and Arabic scripts, look very different but are both alphabets -- systems in which symbols more-or-less represent individual sounds, so that a couple dozen letters and (in some cases) a few diacritical marks suffice to write any language.  The same is true of other alphabets such as the Cyrillic, Greek, Devanagari, etc.  Some languages like English and French have spelling systems which deviate substantially from an exact fit to pronunciation, but the alphabetic principle still holds.

The Chinese character system is fundamentally different.  In principle each symbol represents a morpheme, a spoken unit of meaning.  A morpheme is not necessarily a word.  The English word "teacher" consists of two morphemes, "teach" and "er", each of which has an identifiable meaning even though the latter cannot stand alone as a word in its own right; Chinese has numerous compound words formed from two or more morphemes in the same way.  In Chinese, almost all morphemes are single syllables, and syllables are highly distinct units of speech.  A writing system where symbols represent morphemes rather than sounds "fits" Chinese well.

The problem is that in any language, the number of morphemes is vastly larger than the number of individual sounds.  An alphabet typically has two or three dozen letters; even allowing for complications like capital vs lower-case letters or the joined vs unjoined forms of Arabic letters, the total number of symbols to be learned is well under a hundred.  In Chinese, one must know about three thousand characters for basic literacy, and the ability to read sophisticated texts requires six thousand.  The demand on the student's memory and learning capacity is vastly greater.

A few Chinese characters are recognizably pictograms of the things they mean, which makes them easy to remember, but most are not.  Look at any ordinary page of Chinese writing and try to guess the meanings of characters from their shape.  You won't have much luck.

Most Chinese characters are combinations of simpler elements in which one part tells you something about the pronunciation while the other gives a hint at the meaning.  For example, the words for "sheep" and "ocean" are pronounced alike, and the character for "ocean" incorporates the character for "sheep" plus an added element which means "water":

This is surely helpful to native speakers, but not as much as you might think.  The writing system was standardized more than two thousand years ago, and Chinese (like all languages) has changed enormously over that time, so much so that several regional "dialects" are no longer mutually intelligible and really qualify as separate languages.  Many morphemes which were pronounced similarly back then are no longer so similar, and many words have changed in meaning.  But it's the sheer number of symbols to memorize which is the primary problem.

China's regime claims a national literacy rate of 96%.  From what I've read, you'd be hard pressed to find anyone who has spent a lot of time in China (especially rural China, where most of the population lives) who believes this.  Functional illiteracy or semi-literacy is still very common among adults.  Part of the problem is that literacy is officially defined as knowledge of 1,500 (or in some areas 950) characters, which isn't enough for real basic literacy.  The bigger issue is what is called "character amnesia".

Literacy rates obtained by testing people as they leave school can be impressive, but any extremely complex body of knowledge will tend to fade from the memory if it isn't intensively used.  In pre-modern China, as in all pre-modern societies, full literacy was confined to an educated minority of bureaucrats, intellectuals, authors, and so on.  Such people spent much of their time reading and writing, which enabled them to sustain mastery of thousands of characters.  The common people, if they were literate at all, knew a few hundred characters relevant to however they earned their living, which is a far cry from literacy in the modern sense.

An alphabetic writing system is simple enough that the great mass of people who don't spend much time reading and writing can retain their knowledge of it for a lifetime.  A system requiring 3,000 symbols for even basic literacy is a very different matter.

The standard objection normally raised at this point is that Japan also uses Chinese characters for writing, and is among the most literate nations on Earth.  In fact, Japanese is written with a mix of Chinese characters (called kanji in Japanese) and syllabic symbols called kana.  Each kana system (there are two, hiragana and katakana) contains just 46 symbols -- not much more than an alphabet -- and any Japanese word can be written with just those symbols, even though standard writing uses Chinese characters for many words.  Even so, full literacy in Japanese requires knowledge of about 2,000 characters as well as the kana systems.

Most Japanese adults can probably read most of those 2,000 characters, but can write far fewer as the years pass after leaving school.  This "man-on-the-street" test shows a few examples:

An English-speaker with any education at all would have no trouble spelling basic words like "bribe" or "battle", but you can see for yourself the complexity of the characters that have to be remembered for the Japanese equivalents.  Even those who get the "spelling" right often struggle a bit.  It's much easier, when writing, to just use kana when one is uncertain about a character.  And Japan is a much more "bookish" culture even than the West, never mind China.  It's highly unlikely that character retention among the broad mass of China's population is even this good, or even anywhere near this good.

Alphabetic writing has other advantages we rarely think of, such as the ability to misspell words.  Yes, that's an advantage.  An American of limited education who can't remember the correct spelling of "similar" or "embarrassing" can write "similer" or "embarasing" and be understood, which is better than nothing -- just as a Japanese who can't remember the right characters for what he wants to write can resort to kana.  A Chinese in the same position is simply stuck.  Every alphabet also has an "alphabetical order" which makes it easy to organize and look up information; there is a standard way of ordering Chinese characters, based on the number of strokes used to write particular parts of them, but it's much more complex and difficult to use than alphabetical order.

During Mao's rule, China did make one reform by simplifying many of the characters.  But this did not address the real problem of the sheer number of characters, and the simplified characters are also noticeably less visually distinct from each other in many cases.  It's unlikely that the reform has made real literacy significantly easier.

Even in school, the system makes learning to read and write take years longer than in a country that uses an alphabet -- years which are thus not available for study of other subjects.

Might China someday switch from characters to the pīnyīn Romanization system, which was developed in China and does an excellent job of representing the sounds of standard Mandarin Chinese?  That too seems very unlikely.  A country which changes its writing system faces the question of what to do with the existing body of books and other documents and records printed in the old system.  It can either (a) reprint everything in the new system, a massive and expensive task; (b) teach each new generation both systems, which would negate much of the benefit of the switch; or (c) do neither, meaning that future generations will be largely cut off from the records and literature of the past.  And the traditional attachment to the characters as an integral part of Chinese culture is simply too strong.  If you doubt the power of cultural inertia in such matters, consider that the US still has no plans to switch from our chaotic old jumble of weights and measures to the simpler and easier metric system -- a far less drastic move than changing the writing of the language.

The Chinese writing system served its purpose well in the days when full literacy was confined to an educated minority; indeed, during much of history, Chinese civilization was among the world's most advanced (the Chinese text in the image at the top of this post is from the Guǎngyùn, a printed dictionary published in China in the year 1011, fifty-five years before the Norman conquest of England and four centuries before Gutenberg).  But true modernity requires full literacy among most of the adult population.  In the real world, this makes the Chinese writing system a massive handicap.