Every one knew how laborious the usual Method is of attaining to Arts and Sciences; whereas by his Contrivance, the most ignorant Person at a reasonable Charge, and with a little bodily Labour, may write Books in Philosophy, Poetry, Politicks, Law, Mathematicks and Theology, without the least Assistance from Genius or Study.

The image ?http://www.jaffebros.com/lee/gulliver/faulkner/p5.gif? cannot be displayed, because it contains errors.
Gulliver's Travels:
Voyage to Laputa





Laputan Logic

Atom Feed

Subscribe with Bloglines

Laputan Logic*
Fanciful. Preposterous. Absurd.
The Perils of Reading Chinese, Japanese and Korean Characters in English

Posted on Tuesday 26 November 2002

In an interesting essay, George Leonard discusses the problems of transliterating Asian languages into the English. In the process he discusses the more noteworthy features of each of these languages and their various writing systems.

In an ironic twist of translation, this essay has been so poorly OCR'ed on to the website that a lot of it ended up being unreadable. I've included a cleaned up version of the section on Chinese here.

One of the most respected English professors in the world once remarked confidently to an audience, that he had become more multicultural, and had recently been reading Qing Dynasty poetry. He pronounced it, Kwing. Presumably he thought the Qing Dynasty and the Ch'ing Dynasty were two different dynasties. They are not, and they are both pronounced, simply, Ching.

Why the bizarre spellings then? Spellings so unexpected they confuse even great scholars? Students will not get far if they think Qing and Ch'ing are different eras.

In this essay I address myself to the professor who will have to deal, as soon as books are assigned, with this pedagogical emergency. I will deal with Chinese, then with Japanese and Korean romanizations-- the three most populous groups of Asian Americans with transliterated old country languages.

The colonial heritage makes it unnecessary to devote similar attention to the Filipino or Vietnamese. Colonialism led the Philippines to use a romanization worked out by the Spanish; and led Vietnam to use a romanization (called quoc ngu) worked out by a French missionary, Alexander de Rhodes, in 1651, and enforced by the French when they closed the China-oriented civil service schools earlier in this century. Vietnam is the only country in Asia using romanized letters1. Thai's situation will be more understandable when we know Korea's.


Not only "Ch'ing" and "Qing," but "Zhou" and "Chou" are the same dynasty as well, and the latter two are both pronounced like the English name "Joe." As American students move from book to book during an introductory course, they become confused, as Zhou in one book changes to Chou in another, and all the philosophers, titles of books and names of artists shift under their eyes. If only for that reason, when studying Chinese topics, the professor must start the class by identifying which system of "romanization" he or she will use-- and insist the class follow it. He or she should also insist, when the reading is assigned, on the whole class using one translation only-- if only to avoid orthographic confusion.

The problem: there currently is (and perhaps can never be) a good "romanization" of Chinese-- a method of spelling out Chinese sounds into our alphabet. The older, still most widely used version, the Wade-Giles romanization, is so bad it's comic. Its defenders argue, at best, that it was conceived back when Asian studies was the province of specialists and experts, who had the time to learn Wade Giles's fine distinctions. Others, less charitable, actually have hinted that Wade and Giles were not unhappy about how convoluted their romanization was, for it meant that only professional Chinese scholars like themselves would ever pronounce the Chinese sounds correctly.
For instance, the most important virtue for Confucius is spelled, in Wade-Giles, "jen." American readers pronounce it, naturally, like the first syllable of Jennifer. How do you think that "j" is supposed to be pronounced? Were they thinking of "j" in German, pronounced like "y"? No. As in Spanish, like "h"? No luck. Wade Giles used a "j" as their symbol for the "r" sound at the beginning of words like "rip." The Confucian term is pronounced "ren."

Watch how the damage unfolds. Wade and Giles have used up the letter "j" to mean "r." How will they spell the name of the Taoist philosophic classic, which is pronounced Dao de Jing? They have used up the "j". So for the j sound they use... "ch".

Why? I've no idea. They have to spell Dao de Jing as "Tao te Ching." They also, you'll notice, chose for some reason to spell the "d" sound with a "t." Generations of Americans have called the Dao the "Tao" (and soybean cake, dofu, "tofu." )

To recap, having used up "j" on the "r" sound, they then-- for the "J" sound-- used up "ch." So be it. But they're only painting themselves further into the corner. Having used up "ch" how can they spell the dynasty whose name starts with the "ch" sound, the Ching dynasty? Their amazing solution: when they print "ch" with an apostrophe in front of it, just say it the way you normally would. So the Ching dynasty they write Ch'ing.

Everyone, naturally, sees that strange apostrophe, thinks the Chinese must pause for a second in the middle of the word; and everyone says "Ch... ing", two separate sounds. But all the apostrophe means is, "this "Ch" you shouldn't say as "j" but the way you usually say ch." If you must resort to an apostrophe, wouldn't it have made more sense to add the odd apostrophe when you're not supposed to say it the way you normally would? Why should the odd spelling symbolize the normal pronunciation?

Alas, the name of the country was China, and the language, Chinese. According to their logic, Wade and Giles should now ask us to spell those two words "Ch'ina" and "Ch'inese." At this point Wade and Giles threw in the towel and said, "But go on spelling China and Chinese the way they've always been spelled, even though it violates our system." Chaos.

The proof that Wade Giles has been a catastrophe is that many authors felt they had the right, even the duty, to make up a simpler system of their own. Then, in the 1950s, the Communists developed a new romanization called Hanyu Pinyin (or "pinyin" for short ) which avoids many of the Wade
Giles problems, and is increasingly used throughout the West.

Pinyin is better, and is becoming standard in newer books. Ren is, mercifully, ren, not "jen." But , as Molly Isham explains, there are still six symbols that Americans find difficult: Q like 'ch' in 'cheese'; X like 'sh' in 'sheep'; Z like 'ds' in 'beds'; C like 'ts' in 'cats'; Zh like 'dg' in 'edge'; E like the 'e' in 'the' when it appears before a consonant.

That is how the Ch'ing Dynasty became the Qing Dynasty and befuddled the famous professor. That is why Chou in an old book is Zhou in the newer ones, and it's still pronounced Joe. (Even closer than a J, would be the harsh grinding sound made at the start of the name George.)

All these romanizations, from Wade-Giles to Pinyin, only capture the way one of the eight main Chinese languages pronounce these words: putonghua, the language of the North, of Beijing, of the Emperor's city, and his ruling class-- the language we call "Mandarin." It is more accurate to say that "Chinese" is the name of a language family, like the "Romance" language family. Cantonese is as far from Mandarin as Spanish is from French. A 1955 PRC government conference estimated that 71.5% of Chinese spoke Mandarin, while the second most common fangyan, "regional speech", accounted for only 8.5% and behind that came Cantonese with only 5%. Pinyin does not represent the sounds of any language but Mandarin.

(Among American Chinese, however, Mandarin isn't the most common. Until recently, ninety five percent of American Chinese came from the six counties around the southern seaport of Canton, or Guangdong, and the language there was primarily Guangdonghua, which we call Cantonese. Toishanese was spoken a lot nearby, as well, and is in the United States as a consequence.)

But everyone in China has to study Mandarin in school, and all now learn pinyin as well as Chinese characters. Mao decided some central language, some lingua franca was needed, and decreed it would be Mandarin, the language of the Northern majority and of the capital.
To further complicate matters, Chinese dialects within the languages are further apart than dialects within English. It is only as far from Beijing to Shanghai as from San Francisco to Los Angeles, but the difference in their Mandarin is about as wide as the difference between the English spoken in California and in Scotland. I've watched Shanghaiese listen to a Beijingese, slowly repeat what they'd heard, mulling over the sounds, then suddenly exclaim "Ai-ya!" and understand.

Worst, all the Chinese languages (or "regionalects," to use a more precise term) are tonal languages, and the same sound, said four different ways, means four things. Two different tones, two different words: the difference between Oh? and Oh! Pinyin tries to cope with the tones by using a series of accent marks but American books often omit them as just too much to handle. That means, however, that you'll read of Confucius's love for li, courtesy/ritual, and a few pages later of his opponent's (the Law and Order School's) belief in li, force. Confucius's li was said like a skeptical "li?" and his opponents' li was said like an emphatic "li!" The two tones make two different words-- indeed, two opposing philsophies!

To avoid conflicting romanizations, the professor, if books are assigned, should insist the class all work from one translation. There's another reason. Chinese translations can legitimately differ from each other much more widely than two translations from, say, French into English, legitimately can.

Here are two translations from the Tao te Ching, chapter 6:

The valley spirit never dies;
It is the woman, primal mother.
Her gateway is the root of heaven and earth.

(Feng and English)

The Valley Spirit never dies.
It is named the Mysterious Female.
And the Doorway of the Mysterious Female
Is the base from which Heaven and Earth sprang.

(Arthur Waley)

Is the valley spirit "primal mother" or Mysterious Female? There's a difference! In the first translation, her gateway "is", present tense, right now, the root of heaven and earth. But in the second, the Tao seems to be talking about a cosmic event long ago, the origin of "Heaven and Earth," some sort of creation myth: "the Doorway of the Mysterious Female is the base from which Heaven and Earth sprang."

Yet neither translation is wrong. The Chinese text permits both of them, and indeed, many others. It has to do with the great differences between Chinese and English. Chinese has no plurals, no "a" or "the", often avoids verbs, and has only the present tense. You put time first in a sentence to specify when something happened. We sometimes do that too: "Yesterday I'm walking along the road and I see a girl walking toward me." The Chinese say that, in effect, to create a past tense. To create a future tense, they say something like, "Tomorrow, I am walking along a road and I see a girl walking toward me." So literary texts can justifiably be translated many different ways. The Chinese sentence to be translated literally says, "He/she/it says/said/will say to him/her/it...." and the translator has to pick the sense that makes the most sense.

Pick one translation; be aware from the start which romanization it uses; compare it with other translations if it seems odd. Chinese affords tremendous room for interpretation and many translators take advantage of it. Be particularly wary if the translation seems surprisingly trendy. Chinese gives plenty of room for a translator who's grinding some axe to turn ancient Chinese sexists into feminists, or pacifists, or whatever the soup du jour is. (Stephen Mitchell of Berkeley comes to mind.) Watch out for "orientalism," for hippie translations full of mystic magic mumbo jumbo, and, equally obnoxious, for yuppie translation that turn old generals Sun Tse into CEO's giving tips on how to run an office.

1 - That's not really true, in addition to Vietnam and the Philippines, Malaysia and Indonesia also use the roman alphabet - JH.