Getting It Right

Voices of Experience

Learning Tools

Japanese Study Past and Present


School & Courses

Proficiency Tests

Language in the News


Total Quality Japanese: One-stop Shopping on the Net forall your Japanese-Language Needs

The Romaji (Roomaji) Conundrum

By Andrew Horvat

 (Taken from Andrew's new book, Japanese Beyond Words: How to Walk and Talk Like a Native Speaker.)

The official had hardly finished giving his e-mail address over the phone when he launched into a lengthy apology. "I know it is very confusing but if you don't spell my name incorrectly you won't be able to reach me. Please don't blame me. The computer engineers came up with the e-mail system on their own."

Had the official's name been rendered correctly in the romanization system in most common use today, the modified Hepburn, it would have been Jun'ichi. Using the Japanese government's slightly different system, the same first name would be spelled Zyun'iti. As the ministry's computer wizards mixed up the two systems and placed an unnecessary y after the J, the name became part of a hodgepodge of Latin letters used with little regard for any accepted system of romanization: Jyunichi.

Until a few years ago, how Japanese wrote their language in Latin letters was a matter of concern to no more than a handful of educators, most of whom were critical of the Hepburn system because, by failing to take into account the actual forms and sounds (the morphophonology) of the Japanese language, the system added to the difficulties of teaching Japanese to non-natives. These days, however, the party most concerned with Japan's roomaji (romanization) chaos is the International Standards Organization, which has put Japan on notice to come up with a single, rational, unified system.

Among the ISO's worries is the difficulty Internet users are certain to have locating Jun'ichi if sloppy computer engineers spell his name Jyunichi. After all, how can Jyunichi be found at his desk via e-mail when there are no fewer than three inconsistencies in his name. First is the non-existent jy, then there is the absence of a marker "'" after the n, which turns an ordinary n into a syllabic nasal (n'). As explained earlier, this difference in Japanese is crucial. Kan'i (with syllabic nasal) means "summary" (as in kan'isaibansho or summary court); plain kani means "crab." Finally there is the problem of the last syllable in Jun'ichi. Should it be chi as in Hepburn or ti as in the government's own Kunrei (Kunreeshiki) system of romanization?

If precedent is any guide, neither the ISO nor anyone else should hold their breaths. Anarchy is likely to reign for a long time to come. Instead, it might be an idea for those with an abiding interest in Japan to learn that roomaji (the use of Latin letters to write Japanese) can be spelled as romaji, rômaji, roomaji, roumaji, roomazi, or roomadi.

The second example, rômaji, is representative of the Hepburn system named after James C. Hepburn, the Philadelphia medical missionary who arrived in Japan in 1859 and compiled the first modern Japanese-English dictionary about a decade later. As all pioneering enterprises, the first version was bold and original but missed a few key points including the existence of palatals (for example kya, kyu, kyo) in Japanese. This problem has been immortalized in the rendering of Tokyo as Tokio in the name of the Tokio Fire and Marine Insurance Company. (As mentioned in chapter 3, putting the macron "¯" on top of the vowel o is a matter of choice.)

The zi ending of roomazi comes from the Kunreeshiki system promulgated in the 1930s through a cabinet order, or kunree. The di ending is typical of Nipponshiki romanization, an earlier version of Kunree that, though hardly remembered, remains with us in the name of the upmarket grocery chain Meidiya, which generations of foreign residents have pronounced with a hard d instead of as "Meijiya." Then there is the Japanese precision measuring device maker Mitutoyo, which also uses Nipponshiki but which when listed in the English-language Tokyo Yellow Pages ends up as Mitsutoyo because most publications written for foreigners require names to be printed in Hepburn.

Perhaps the biggest reason for roomaji chaos is that no one single system satisfies all users. While Hepburn makes it easy for non-natives acquainted with the values English assigns to Latin letters to learn how to approximate the sounds of Japanese, it has been strongly denounced by serious educators for distorting the actual forms and sounds of the language.

For example, in Kunreeshiki the various forms of the verb to wait, matu (in Hepburn matsu), can be listed as follows:

matu informal (dictionary form)
matanai informal negative
matimasu deferential, polite
mate informal imperative (Wait!)
matoo informal invitational (Let's wait)

Advocates of Kunreeshiki romanization argue that only in Kunreeshiki is it apparent that the root of the verb is mat. Students taught using the Hepburn system must be told that the tsu of matsu is dropped and replaced by a chi, ta, te, or too. Such explanations are both clumsy and inaccurate in that they fail to show that, far from being an exception, matu follows the same rule as other u verbs such as kaku (to write), the root of which is kak, or sasu (to point), the root of which is sas. Incidentally, in Hepburn the polite form of sasu is sashimasu, and this too makes a mess of the simple rule that can be achieved in Kunreeshiki.

When Kunreeshiki was adopted in the late 1930s, the then powerful Japanese central government did much to try to further its use. Unfortunately, by forcing non-native speakers of Japanese with no intentions of learning the language to abide by a system intended for those who have some command of Japanese, the government gave the impression of intolerant language management that would have dire consequences later on.

To give an idea of just how pervasive Kunreeshiki became in the short period of strong central rule from 1937 to 1945, even pamphlets for tourists, such as the one I have on Japanese proverbs published in 1940 by the Japanese Government Railways, provides foreigners with a list of the new orthography next to the old Hepburn equivalents in brackets. The list is as follows: si (shi), ti (chi), tu (tsu), hu (fu), zi (ji), sya (sha), syu (shu), syo (sho), tya (cha), tyu (chu), tyo (cho), zya (ja), zyu (ju), and zyo (jo).

The Government Railways politely informs readers that Mt. Fuji is now Mt. Huzi, Juujutsu is Zyuuzyutu, and the Chion-in Temple, Tion-in. It was the Kunree system's misfortune to become associated with Japanese militarism. After Japan's defeat, scholars and educators attempted to introduce romanization both as a teaching device and as a possible simple substitute for kanji, but U.S. administrators during the Occupation were wary of Kunree's nationalistic origin. This was in spite of the fact that the Japanese-language textbooks of American linguist Eleanor Jorden, which were used throughout the 1960s in courses taught to U.S. diplomats, relied on a modified version of the Kunree system.

The existence of conflicting systems results in confusion. But this minor state of chaos is nothing compared to the full blown anarchy to be found on the Internet where the boundaries between Hepburn and Kunree have been demolished in favor of an "anything goes" philosophy made possible thanks to the ability of computers to respond with one correct symbol to both correct and incorrect inputs.

Thus, the syllables ju, zyu, and jyu will all bring onto the screen the same combination of two kana syllabic symbols, which can be converted to several kanji characters. Manufacturers of Japanese word processors, needing certain extra features to popularize the use of their products, decided to make it possible even for the most inept consumers to be able to write in Japanese using Latin letters. They achieved this by making their products respond correctly to the most outlandish mistakes in roomaji. Thus, tsi, ti, ci, and chi will all yield the same kana symbol even though only ti and chi conform to known systems.

Another reason for the "anything goes" philosophy of romanization is that&emdash;strange as it may seem&emdash;many Japanese, perhaps even the majority, use the Roman alphabet (and not kana syllables) to access Japanese-language word processors. The reason for this is simple. The few Japanese familiar with the QWERTY keyboard prior to the introduction of Japanese-language word processing had learned typing as part of their English language secretarial education. The kana keyboard was used on a daily basis by typesetters at Japanese newspapers, secretaries in law offices, and a few thousand employees of the National Railways, whose computers were kana based.

To make matters worse, romanization was not taught very thoroughly in Japanese schools. Older Japanese were familiar with the Kunreeshiki, but younger Japanese who spoke English felt more at home with Hepburn, which uses English-style rules to write Japanese sounds. As for the vast majority of Japanese, both systems posed challenges. Making matters worse yet, the QWERTY keyboard did not come with diacriticals for lengthening Japanese vowels.

Literature on Japan for foreign readers done in Hepburn normally dispenses with any kind of macrons. This is because in the case of popular books neither authors nor readers are familiar with the critical role of vowel lengths. But for e-mail addresses and for home page URLs, being able to differentiate between Marubatsu Kinko K.K. (a maker of strongboxes) and Marubatsu Kinkoo K.K. (a gold mine) might make a difference in business outcomes. But what if the gold mine's owners, unfamiliar with the complicated rules for lengthening Japanese vowels in differing roomaji systems, follow the example of native speakers and lengthen the o with a u using the rules of kana (the Japanese phonetic syllabaries)? The absence of macrons on QWERTY keyboards has inspired just such a recent change in orthography.

It is just these kinds of confusing situations that the ISO wishes to avoid. To make matters interesting, Japanese officials charged with proposing a standard are reported to be leaning toward Kunree, and not the more popular Hepburn, as the future standard. So, say good-bye to Mitsui and Mitsubishi and welcome Mitui and Mitubisi, but if you find the change too tough to take, just go for a hike in the woods of Titibu.


(c) 2000 Andrew Horvat. Reproduced from JAPANESE BEYOND WORDS: HOW TO WALK AND TALK LIKE A NATIVE SPEAKER by Andrew Horvat with permission of Stone Bridge Press, Berkeley, California, http://www.stonebridge.com, 1-800-947-7271.

Click on image to
return to the index page

Getting It Right

Voices of Experience

Learning Tools

Japanese Study Past & Present


Schools & Courses

Proficiency Tests

Language in the News




back to top