The Commissar wrote:Caesius wrote:To be clear, all the "u" does is lengthen the "o" sound; normally lengthened vowels would be represented with a macron i.e. Katawa Shōjo. The other vowels don't have a similar way of lengthening them by adding a letter, so it's kind of dumb to single out "o" for lengthening, but at least it's relatively unambiguous (or so it should be). You can't just double vowels like you would in Finnish either, because doubled vowels in Japanese have a glottal stop between them, changing the meaning - e.g. the difference between "shi" and "shii" is that the former sounds like English "she" while the latter sounds like English "she-e" (the letter "e"), but faster and with a much more subdued glottal stop.
At least I think that's how it is; I'm no expert in Japanese, like I said I just browse Wikipedia.
Not exactly. They are two vowel lengths, Shii would be pronounced like a longer she, but with the way they Japanese language works, they are considered two...I forget what they're called, not syllables, but something else. Anyway, double vowels usually use two kana, but are pronounced together.
And I gathered this from Wikipedia as well.
The word you're looking for is mora; plurals moras or morae. They factor into most languages. The idea is that a syllable is composed of two things: a optional onset, which may be simple, like /s/ in "sit" or complex, like /st/ in "sting" or /spl/ "splash", and a rime (or rhyme) comprised of one or more morae, which serve roughly as units of timing. There will be at least one mora for the nucleus of a syllable (usually a vowel). It is important to understand the morae are not true units of time; they are an abstract units we use in understanding syllable "length".
Codas can optionally end the syllable, like simple /t/ in "sit" or complex /ts/ in "sits". In English, the elements of the coda each get their own mora if the syllable is stressed, as in "met" (/mɛt/), which has two morae:
The sigma means syllable, the mu means mora, and in actual phonological work the onset is directly attached to the syllable instead of getting a label, but I wanted to point it out here.
In some languages, this is not the case. Instead, elements of the coda belong to the same mora as the nucleus, so "met" would have one. In unstressed syllables in English, the moraic-ness of the coda is disputed. Here is mono-moraic "met":
In languages with a distinction between long and short vowels, as in there is a difference in the actual length they are said, the actual distinction is whether the nucleus is carried by one or two mora. Here is "shoujo" (/ʃo:dʒo/), with "ou" as long "o". Please note that /dʒ/ is realized as one phoneme (i.e. sound), not two:
In contrast to the first "o" in "shoujo", the second is short and only has one mora attached to it. In Japanese, morae are rattled off at approximately the same speed and loudness; it is a mora-timed language. This is in contrast to English, where syllables (not mora) are louder and longer when stressed. In Japanese, mora are far mora important than syllables phonologically. For example, haiku are supposed to be 5-7-5 mora, not syllables.
Caesius, I think what you might be hearing in "shii" and the like may be Japanese pitch accent. Japanese doesn't have stress like English, but instead about 1/5 of the words feature pitch accent which applies not to syllables, but to mora. An accented mora rises in pitch, then drops suddenly for the following mora, which I imagine sounds very startling.