Posts

  • QI — Tell us about something interesting!
  • Romanization
#1

Hey Eevee. I know you're busy, so you probably don't remember this, plus you probably can't look it up, cause old Veekun is dead, so I'll refresh your memory. I mentioned once on the old forums (I think it was when you posted the Kana for "HeartGold" and "SoulSilver" back when it was first announced) that I didn't much care for the Hepburn romanization of Japanese.

Well, I was playing around with Wikipedia this morning and just randomly looked up "romanization" in general, and found a link to something interesting. It's called "Nihon-shiki" or sometimes "Nippon-shiki" romanization. It was invented by a guy to be a sort of one-for-one transliteration, rather than phonetic transcription, of Japanese.

He meant for it to replace Japanese rather than for Westerners to use it, so there was no attempt to change the way certain romanized Kana characters sounded to make them easier for Westerners to pronounce. Basically, it's romanization for people who already know how to pronounce Japanese, and it sticks to a strict "two-letters for each Kana" rule.

Then there is "Kunrei-shiki", which came later and wasn't as strict. It incorporated more modern Japanese pronunciation, and was actually used as a basis for an ISO standard.

Anyway, I'll shut up now. I just thought you'd find it interesting to know that there was actually a romanization specifically for Japanese speakers. Look it up on Wikipedia if you want to know more.

#2

Yeah I can't bloody stand Hepburn. It goes to contortions to help native English speakers pronounce words correctly, only partially succeeds at that, and destroys the actual Japanese spelling in the process. Sort of useless to me. More than once I've run across a Hepburn word and been unable to find it in a dictionary, because some おう became ō, and that quickly became just o.

I'm familiar with nihon-shiki, but I don't much fancy it either; where Hepburn actively pursues native English pronunciation, nihon-shiki forsakes it entirely. I don't think that's entirely true to the source, either, and it always throws me for a loop when I see it—especially because I'm used to something like "si" actually being spelled with some wacky katakana combination like スィ.

So I try to stick with waapuro, which unfortunately gets far too little mention in Wikipedia's article on romanization. It's written how you'd type it in an IME, it's fairly consistent, and it has exceptions when the pronunciation is exceptional. The only ambiguity is with ぢ/づ because they're usually written ji/ju, but they're uncommon enough that it's rarely a problem, and in practice I think I write them as dzi/dzu or something.