In the Land of Invented Languages - Arika Okrent [57]
But he would then start learning a lot of characters that didn't look anything like what they meant, and a lot of compound characters that had no nice poetic explanation. So he would just have to memorize them, and the more characters he would learn, the harder they would be to remember. And so it makes sense that after a year of learning, he gave up.
But if he had been learning to speak Chinese as well as write it—if he hadn't been so impressed that he could read out the characters in his own language—perhaps he would have gotten further. Pronouncing the characters in Chinese, rather than in his own language, would help him to see why the character for “clamp,” for example, is formed like this:
It takes this form not because it has some conceptual thing to do with horses but because it is pronounced mà—just as the word “horse” is (but with a different tone). The tree part of the character provides a vague semantic clue that is open to interpretation (clamps are used on wood?), but the horse part is a much more reliable pointer to “clamp” because it doesn't take you on some roundabout journey of connotation to a concept. Instead, it sets you down on a nice straight path and gives you a little shove toward a word.
Unfortunately, the sound aspect of Chinese characters is not always so readily apparent. Thousands of years of language change coupled with a conservative writing tradition will do that. Look at English, after only a few hundred years of change, holding on to forms like “light” and “knee,” when the pronunciations that gave rise to those spellings are no longer used. The situation in Chinese writing is much worse.
Still, most characters, more than 90 percent, give you some clue about the pronunciation of the word. You can't depend on those clues entirely, but it makes the task of learning and remembering thousands of characters a little bit easier. Chinese writing doesn't represent spoken language in the way that alphabetic writing does, but it still represents spoken language—just in a much more complicated way.
But what of the observation, marveled at since Westerners began reporting from the Far East in the sixteenth century, that character writing is understood throughout Asia? How can it be that people who speak completely different, non–mutually intelligible languages understand each other in character writing? The truth is, they don't. At least not in the way you would imagine from the ever popular “characters transcend language and go straight to concepts” account.
The Chinese writing system is based on Mandarin Chinese. Other languages spoken in China, like Cantonese, are different but historically related—about as similar as French and Italian are. So what happens when a Cantonese speaker picks up a Mandarin newspaper? Does he just read it off into his own language? No. Essentially, he reads it in Mandarin. In order to become literate, he has had to learn the Mandarin way of marking grammatical distinctions and the Mandarin way of putting sentences together. He may not have learned the Mandarin way of pronouncing every word, but many of the Cantonese pronunciations are similar (as are the French jour and the Italian giorno), so the sound clues in the characters are sometimes helpful. However, they are much less helpful, so he has had to do a lot more brute memorization. This is why it has taken him a couple of years longer than a Mandarin speaker to become literate.
As for a Japanese speaker, he does not understand the Mandarin newspaper at all. His spoken language is about as similar to Mandarin as Hungarian is to English. However, for historical reasons, Japanese is partially written with Chinese characters (along with other characters that stand for sounds). So when a Japanese speaker sees a Mandarin newspaper, he may indeed be able to recognize a number of the characters, but that doesn't mean he will be able to form anything more than a fuzzy guess at what it all means. The