Japanese has 5 vowels (a, i, u, e, o), which are pronounced as in Italian. Each vowel can be either short or long, and the difference is very often significant. Since English has no such distinction, many English-speakers have problems with pronunciation.

From Everything2's point of view, however, a more significant problem is writing these vowels with Roman letters. The problem is not helped by the fact that the de facto standard, the Hepburn system of romanization, recommends marking them with macrons, which are either a pain to work with or outright impossible to typeset with most systems (including HTML). There are no less than five alternative approaches:

  • Ignore (omit): Ignore the distinction entirely and always write single vowels, as in. Tokyo. The upside is that many Japanese words have already entered the English language bastardized in this way (eg. tofu); the downside is that the lost distinction is often critically important, as in isshobin (short O = a flight together, long O = a 1.8L bottle of sake). However, for node titles in E2, my recommendation is to ignore long vowels since this allows the nodes to be easily found.

  • Phonetic (elongate): Whenever a long vowel is pronounced, double the vowel. The capital of Japan would thus be rendered Tookyoo. The upside is that the system is logical and easy to learn; the downside is that phonetically spelled Japanese cannot be accurately retranscribed into Japanese. For example, both "throne" and "many" would become ooi with this system.

  • Transcribed hiragana (by syllable): Copy the hiragana down as written in Japanese, so our example would now be Toukyou. The long O is thus ou and the long E ei. The upside is that the written difference between oui (throne) and ooi (many) would be maintained; the downside is that pronunciation is difficult unless familiar with the odd conventions of the kana system and mora syllables (eg. why sumou is read "sumoo" but omou is read "omou"). This is the method usually used by word processors, in scholarly circles and when the audience is intimately familiar with Japanese.

  • Transcribed katakana (with dashes): Katakana uses dashes instead of additional vowels to indicate vowel doubling, as in To-kyo-. While technically a smidgen more accurate for katakana words, this way lies typographic madness since by convention dashes do not usually carry phonetic information.

  • Substitute (with accents): Follow Hepburn's recommendation, but substitute circumflexes or umlauts for macrons: Tôkyô. This is my recommendation for writeup content on E2, if not throughout the writeup, then at least the first time the word occurs in the content; but this should under no circumstances be used in titles, as they would make searching impossible.

  • Hepburn (macrons): A new option thanks to the magic of Unicode, since ō and ū can now be entered as ō and ū respectively. While more formally correct than substituting, these have a higher danger of not working correctly and are a pain to enter, so for time being I am holding off; many bakufu members, however, are already using these.

So the summary of my recommendations:

  • Node titles: ignore long vowels
  • Writeup content: use circumflexes or macrons

That pretty much covers vowels. Someday I'll have to write the treatises on the differences between Hepburn and kunreishiki romanization plus the problems posed by the syllabic n in Japanese and the multiple kana for tsu and ji in Japanese...

Ijoo desu.

Log in or register to write something here or to contact authors.