3

A question about rōmaji: Is it more common to use ī or ii? For loanwords, it seems to be common to use ī, e.g.

コーヒー  kōhī
アンチーク antīku

For other words, using ii seems more common, e.g.

涼しい suzushii
大きい ōkii
小さい chiisai

Google translator shows suzushī, ōkī and chīsai, though. (I know, results from Google translator are usually poor, but still.)

Similar question: I believe 湖 is actually mizu+umi (right?), so the natural romanization would seem to be mizuumi rather than mizūmi, but I think the latter is more common?

  • I don't know if this is on-topic here, but here's something to consider: normal romaji where? (in Japan? In America? In France? in academic literature in a specific discipline?) – virmaior Aug 5 '18 at 6:32
  • Well, in Japan, or at least in the material that is usually taught to learners. Maybe people don't consistently use it, so that could be an answer, too. – digory doo Aug 5 '18 at 6:33
5

One basic rule (or perhaps "convention" might be a better word) for determining romanized spellings, is whether two identical vowel sounds -- such as i + i, or u + u, belong to the same morphophonemic unit. In more common parlance, whether the two sounds belong to the same root or stem.

For adjectives that end in two i sounds, the last i is usually the conjugating portion -- this is the i that changes to ku in the adverbial form, like すず[し]{●}[い]{●} as an adjective changes to すず[し]{●}[く]{●} as an adverb. Since the final i belongs to a separate functional piece of the word, romanization practices generally keep this separate -- so すずしい = suzushii with the two i's spelled separately.

Likewise, みずうみ "lake" derives from 水 + 海, and the two u sounds belong to different roots. So this is similarly spelled spearately -- みずうみ = mizuumi with the two u's spelled separately.

However, in おにいさん, the two i's are part of the same root. Some romanization systems spell this as onīsan as a result. Likewise with borrowed terms like コーヒー: the two i's are part of the same root, and these are thus generally spelled together as ī as a result.

Other romanization systems spell *all* double-i's separately, sometimes out of a stated concern that the macron (the long bar) over the ī is too hard to distinguish visually from the tittle (dot) over the regular i.

There are multiple systems for romanizing Japanese, as Virmaior points to in his comment. Some of the main ones are Kunrei-shiki in Japan, and Hepburn, which is commonly used in English-language materials about Japanese. More at the Wikipedia article on Romanization of Japanese.

  • 1
    So, 小さい would be chīsai, right? innovativelanguage.com teaches chiisai, though. (See "Japanese word of the day", 2017-09-03.) They do use the ī all right, e.g. in kōhī. – digory doo Aug 5 '18 at 6:54
  • Also, the same site has at 2018-07-17 the expression fukai mizūmi, and not mizuumi as you suggest. – digory doo Aug 5 '18 at 7:33
  • 3
    Inconsistencies in the site mentioned might easily be explained by different staff members (who use different conventions) inputting the word of the day or one individual forgetting which 'convention' they used the day before. Until there is more universal acceptance of one convention there are going to be inconsistencies. Memorizing (or just getting used to) the different conventions is likely your best bet. – BJCUAI Aug 5 '18 at 9:48
  • 1
    So, the answer should probably start with saying, "There is no universally accepted convention, but here is one". I haven't been able to verify that Utlendi's suggested convention is a widely used one. – digory doo Aug 5 '18 at 10:07
  • 1
    @sumelic: I think that most adjectives ending in しい have a pitch downstep between the し and the い. In that way, the two //i// vowel values are somewhat distinct. The vowel values themselves are identical, but the voice pitch goes from higher to lower. Likewise for いい "good", though in that case, the word sounds like a single //i// sound that goes from high pitch to low. Phonetically, this しい or いい is //iː//. Tonotactically, this is //íꜜì//. Similarly, the //u// vowel in みずうみ is two morae, with a downstep in between: phonetically //uː//, tonotactically //úꜜù//. – Eiríkr Útlendi Aug 6 '18 at 17:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.