How bureaucrats spell logic in Romanized Japanese

by Colin P.A. Jones

Tomorrow I will go to Sinzyuku to meet my old friends Mr. Tutida and Ms. Oisi. We will get some susi and then end up in Kabuki-tyo, drinking syoutyū until the syūden.

What’s that you say? That is not how those Japanese words are pronounced or written?

It just so happens that you are mistaken. I am using the Official Japanese government method of Romanizing Japanese. Cabinet Directive No. 1 of Dec. 9, 1954, mandates the way Japanese is to be rendered in Roman letters. Never mind that it often results in Japanese being rendered an unpronounceable mess to foreigners who have not taken the time to understand the system (i.e., a great many of them).

Japanese is actually not very difficult to pronounce, having fewer consonant and vowel sounds than many other languages. The government has added an unnecessary layer of complexity by establishing rules on Romanization that value consistency over pronunciation despite the Japanese language itself lacking the consistency that the rules seek to impose. To be specific, kana (phonetic Japanese alphabet symbols) for the sounds ka, ki, ku, ke and ko consist of the common consonant “k” attached to each of the different vowel sounds. Kana for sa, shi, su, se and so attach the same set of vowels to the consonant “s.” But when we come to “i,” the pronunciation becomes an irregular “shi” (there is no “si” sound in Japanese, which is why native speakers of the language have to be careful with English words such as “sit”).

Similar irregularity occurs with “t” sounds. There’s no problem with the kana for ta, te and to ? but instead of ti, we have chi, and tu becomes tsu. The same goes for variations of the “s” and “t” sounds such as za, ji, zu, ze, zo, cha, chu and cho. The official government Romanization ignores these distinctions, demanding consistency (sa, si, su, se, so; ta, ti, tu, te, to, etc.) regardless of actual pronunciation. By this method, the number seven in Japanese becomes siti (instead of shichi), and green tea becomes otya (not ocha).

This is a system that only a bureaucrat (or perhaps a linguistic scholar) could appreciate, since it values logic over usefulness — the same kind of logic that underlies the once-prevalent Wade-Giles system of Romanizing Mandarin Chinese, thanks to which most people in the world mispronounce words like “kung fu” (which is more properly pronounced gong fu). It is the sort of logic that made it necessary for me to get a Canadian passport for my baby daughter before I could get her a Japanese passport. Without official proof of the “correct” spelling of her family name (i.e., mine), the passport would have been issued in the name of a child with a last name, in Roman letters, of “Zyonzu.”

Then again, maybe it is presumptuous of me to think the main purpose of a system for Romanizing Japanese is to help nonnative speakers to understand and pronounce the language.

I once discussed this matter with a big group of Japanese people, none of whom seemed to think that the system was intended for anyone other than, well, Japanese people who already know how the language is pronounced. As it was explained to me, Japanese was Romanized in a number of different, inconsistent ways before World War II. So the government stepped in, solving the problem by mandating a single, unified system of Romanization.

Except that they didn’t. As anyone who rides the trains or looks at a map of Tokyo will know, the station and the town are written “Shinjuku.” Most other well-known place names are also written the old-fashioned, nonofficial way. In fact, the Cabinet directive specifies that if use of the “new” Romanization is not possible, because of “custom or other reasons,” an alternative system can be used — one that results in words (primarily place names) being rendered in the ways most of us find familiar. (Amen to that: Who wants to spend time in a place named “Kitizyozi”?)

This system gives us Tsuchida, Oishi, sushi, Kabukicho, shōchū and shūden. It is also a system that renders Japanese in a form closer to how it might be pronounced (by a speaker of English, at least). So the government-approved system both encourages mispronunciation and is inconsistent in application.

“Shinjuku” may prevail in the real world, but “Sinzyuku” still generates thousands of results on Google. Isn’t it time the government consigned this antiquated system to the dustbin of history before the confusion spreads any further onto the Web?