Friday, November 18, 2005

Beijing Duck not Constantinople

I vaguely remember when the capital of China changed from Peking to Beijing overnight in 1979. To be more precise, the actual capital city didn't change, nor did the chinese characters for its name (北京, meaning "northern capital"). Rather, the government changed the official romanization system (the way of writing Chinese in the roman alphabet) to Pinyin, which represents the name as Beijing. The name of the capital's eponymous culinary masterpiece doesn't reflect the changes, or else today afficianados of Chinese food would be dining on Beijing Duck.

Sometimes geographical names change because of politics. Sometimes because of linguistics. (And many people would argue that the root of linguistic changes are still political.)

The APEC summit is being held this week in Pusan, Korea. But, I was surprised to read in the Toronto Globe in Mail that Paul Martin is at the APEC summit in Busan.


I did some digging and discovered that while the city I've alwasy knows as "Pusan" is often referred to as Pusan, is is now officially named Busan. How did I not know this? I have travelled and worked in Korea and I've been to Pusan several times. Most notoriously, I've even taken the Kanpu Ferry overnight from Shimonoseki to Pusan, possibly the world's largest floating ashtray. (Everyone was chain-smoking, but in September it was too cold to go out on the deck at night. A cold and oxygen-less experience.) Everyone in Korea called the city "Pusan" when I was there. How, and when, did it change to Busan?

The culprit seems once again to be different romanization systems.

The McCune-Reischauer system transliterates the city's name, 부산, as "Pusan." McCune-Reischauer was created in 1937, and it remains one of the two most frequently used romanization systems. Technically speaking, McCune-Reischauer does not transliterate Hangul, the Korean writing system; rather, it represents Hangul phonetically.

This may sound like the splitting of linguistic hairs, or at the very least sound confusing, but students of languages that don't use the roman alphabet will understand. I'll illustrate with an example from Japanese, where I'm on more familiar ground:
You may have seen adzuki beans in the canned goods aisle of your grocery store (or at least your health food store). The name of these beans in Japanese is written with three "letters," (or more properly "kana"): あ (A) + づ (Dzu) + き (Ki). Romanizing the first and last kana is easy; there's not much room for error, at least if you are romanizing for English-speakers (Portuguese-speakers, for example, use a very different romanization system). There's a one-to-one relationship. However the relationship between づ and the roman alphabet is a one-to-many relationship. When romanizing づ, you have to make a choice between transliterating the way the writing symbols equate with the roman alphabet, or representing the pronounciation accurately. The two don't always line up.

"Dzu" is an accurate representation of how Japanese people pronounce づ. Going back to our neglected can of beans, they are most likely to be labelled "Adzuki," and you are correct to pronounce the word just the way it looks.

づ is a combination of つ, which is pronounced "tsu" (as in tsunami) and can be transliterated as tsu or less commonly as tu (the pronounciation doesn't change in either case), and a "dakuten" or "ten ten," the two little strokes which indicate that a consonant is voiced.

"T" is an unvoiced consonant. "D" is a voiced consonant. Hold your hand to your throat and say "tip" and "dip." Feel the buzz in your throat when you say the D in dip? Your vocal chords are active when you produce a voiced consonant. You don't feel the buzz when you make a "t" sound because your vocal chords are passive.

Therefore, when we transliterate づ, we need to get around the fact that there is no tenten in the roman alphabet. The roman alphabet assigns different letters to represent voiced and unvoiced consonants. So, starting with the transliteration of つ, which is tsu or tu, we then change out those unvoiced consonants with their voiced equivalents, arriving at dzu ("z" is the voiced partner of unvoiced "s") and du.

While づ can be correctly transliterated as dzu or du, it is always and only pronounced dzu. No Japanese person would ever say "aDUki." On the other hand, some western food companies label their beans as aDUki. This is a technically accurate choice, but it will mislead non-Japanese speakers as to the pronounciation. And yes, when you work as a translator, these are the issues you wrangle over.

Why is this a big deal? Look at the Japanese sound "ji." It can be written two ways in Japanese, じ or ぢ. If my starting point is a romanized word, and I want to work back to a Japanese script, when I look at "ji" I don't know if I need to write it with じ or ぢ. Your choices for transliterating じ for English speakers are limited, but ぢ can be transliterated as "ji", more accurately as "dji," or rarely and less helpfully as "dzi." However, if I see a "dji" or "dzi," I know that in Japanese it definitely came from ぢ. So, the big deal comes when a technically accurate transliteration disconnected from the actual pronounciation takes on a life of its own in other languages. The classic example is the famous Japanese movie monster ごぢら: the name is pronounced Go-Ji-Ra, but you are more likely to know him by his misleading transliteration, Go-Dzi-La. (Keep in mind that Japanese does not make a distinction between, as they say, "the L in Rome and the R in London;" the sounds can be romanized using either letter.) Yes, ごぢら is how you write Godzilla in Japanese. We are all victims of crummy transliteration on that one.
Back to the the McCune-Reischauer system: if McCune-Reischauer had been in charge of bringing Japanese movies to the West, we'd know that monster as Gojira today, but we wouldn't be able to know with accuracy how to write his name in Japanese. Romanizing Korean hits similar snags to the ones we've examined in Japanese, because the pronounciation of a consonant can change depending on its position in a word.

McCune-Reischauer is widely used outside of Korea, and it was used as the official romanization system in South Korea from 1984 to 2000. It was the system of the day when I visited and worked in Korea, which is why I know the city in question as Pusan, not Busan.

In 2000, South Korean authorities replaced the McCune-Reischauer system with the Revised Romanization of Korean as the official Korean language romanization system. (A variant of McCune-Reischauer is still the official system in North Korea.) The new system makes it much easier to work backwards from the roman alphabet to Korean script, but, is less accurate in indicating how a word is pronounced.

If this all sounds like Istanbul not Constantinople, keep in mind that it is really more like Beijing Duck and Godzilla. Pusan is still Pusan, phonetically, but as of 2000 is officially romanized as Busan.

Reading international news coverage of the APEC summit, I found that Pusan and Busan both show up, and both are even used by English-language media in Korea. I am suprised at the variation: surely the professional, courteous, and diplomatic thing to do is to follow the romanization standards set by the government in question. Just as we all learned to say Beijing instead of Peking, we need to learn to write Pusan as Busan.

Or else Godzilla and the Beijing Duck may come after us.

Links to this post:

Create a Link

<< Home