CZ:Romanization/Japanese
This page provides a forum to discuss how Japanese words can best be romanised on the Citizendium. It is not yet policy. Work in progress! Please feel free to add or edit sections here. Discussion of the issues should take place on the Talk page.
Major systems
There are two major ways of romanising the Japanese language: Hepburn and Kunreishiki. The former was created with the needs of Japanese language learners in mind, particularly those familiar with the spelling conventions of English. Kunreishiki, on the other hand, encodes the phonology of Japanese more accurately.
Example: /h/
The phoneme /h/ before /u/ is written as <f> in Hepburn and <h> in Kunreishiki. Because the usual pronunciation of /hu/ is [ɸɯ] in Japanese - with an initial bilabial fricative sound - Hepburn encourages the idea that it is functionally and phonetically equivalent to [f] in English, which is actually a labiodental fricative involving the top teeth rather than lip. Kunreishiki, however, maintains <h> because underlyingly the sound is really a /h/: in the sequence ha ([ha]), hi [hi], hu ([ɸɯ]), he ([he]) and ho ([ho]) it simply surfaces as a different sound. It is unlikely that hu is underlyingly /ɸ/ because that fricative only emerges in one context, i.e. before [ɯ].
Long vowels
These are represented in several ways: by placing a 'macron' (a bar) over the vowel letter; by doubling the vowel letter; by using another vowel letter; or by nothing at all.
Macron
Example: 東京 とうきょう Tōkyō; 大阪 おおさか Ōsaka
For:
- This is often seen on official signs in Japan.
- Wikipedia uses it.
Against:
- It may not be clear to lay readers what the macron signifies.
- It is fiddly to type.
- It eliminates the contrast between historically long vowels and diphthongs that remains encoded in the modern kana writing system: e.g. 大阪 'Osaka' is written Ōsaka, the same way as the vowels in Tōkyō; however, kana indicates that the vowels were once distinct: おおさか not *おうさか.
Double vowels
Example: 東京 とうきょう Tookyoo; 大阪 おおさか Oosaka
This is the way argued for by the journalist Andrew Horvat in his book Japanese Beyond Words: How to Walk and Talk Like a Native Speaker; it has also been advocated by others at various times.
For:
- It is slightly more clear to lay readers that it is supposed to represent a long vowel.
- It avoids typing fiddly diacritics; i.e. it is simple to use.
Against:
- It eliminates the contrast between historically long vowels and diphthongs that remains encoded in the modern kana writing system: e.g. 大阪 'Osaka' is written Oosaka, the same way as the vowels in Tookyoo; however, kana indicates that the vowels were once distinct: おおさか not *おうさか.
- It may mislead people into using English pronunciations: e.g. 和製英語 'Made-in-Japan English' would be written wasee Eego, possibly making readers parse the vowels as [i:] (as in English sheep) and thus producing something like 'wassy ego'.
- In slower speech, two different vowels may surface for many speakers, e.g. おう may be [o-u].
Other letters
Example: 英語 えいご Eigo
The vowels [e:] and [o:] in modern Japanese were originally two vowels, [ei] and [ou], and this is reflected in kana: えい and おう are much more common than ええ and おお. In some speech styles, [ei] and [ou] might still be heard. These could be kept in Citizendium romanization.
For:
- Maintains a link with kana, with a one-to-one correspondence between the kana and Roman symbols. Readers may see that え is e and い is i in えいご 'English'.
Against:
- May mislead readers into assuming Japanese has diphthongs, and that these are similar to English. Eigo might well be pronounced with [eɪ], not [e:].
- Causes confusion with words that really do contain a sequence of the two separate vowels, e.g. 小売 kouri ('retail', ko plus uri - not [ko:ɾi] but [koɯɾi]). This could be solved by using hyphens to indicate meaningful boundaries (ko-uri) but causes further confusion in itself.
No representation
Example: 東京 とうきょう Tokyo; 大阪 おおさか Osaka
Here there is no difference between long and short vowels; unwise as the difference is contrastive in Japanese. However, it is probably acceptable to do this for placenames such as 'Tokyo'; indeed, this is often standard practice across Japan.
Apostrophe
The apostrophe ' is sometimes used in romanizing Japanese to indicate that a preceding <n> is not part of the following syllable, e.g. 山陰 San'in (a region of western Japan). Here, <san> includes the so-called 'syllabic nasal' /n/ (actually not a syllable at all, but a 'mora'), and so does not continue into the next syllable or move into it entirely. In English, that [n] would naturally close the first syllable and begin the second (ambisyllabicity), causing potential confusion for listeners.
For:
- Accurately distinguishes moraic /n/ from non-moraic /n/ (which merely opens a syllable, e.g. に ni).
Against:
- Not at all obvious what it represents to lay readers, who may not need the pronunciation anyway. It may be misread as a glottal stop, the possessive apostrophe before s (Harry's), or its use in words like going versus goin'.
An alternative is to use a hyphen: San-in. This can be confused with the hyphen that links morphemes, but may be easier to interpret than the apostrophe.