I noticed that in songs, the vowel sounds of the morae that come before geminations are sometimes repeated.

For example, the first lyric line of “マジカルちょーだいっ” is sung as しらんぷりをしたあって where the line is actually しらんぷりをしたって.

Another example: in the second lyric line of “片道きゃっちぼーる”, the final あった sounds like ああった.

Also, at the beginning of the song “Gem Stone”, the repeated もってる is pronounced as もおってる.

My questions are:

  1. Is this way of pronouncing gemination limited to songs, or are there instances where it's used elsewhere?

  2. It seems to be limited to songs with children or children-like voices, so is it by chance the way Japanese children initially pronounce gemination?

  • I perceive the lengthening in the first two of your examples, but not in the third one.
    – user458
    Oct 23, 2011 at 15:01
  • 1
    @sawa In the third one, the lengthening is in the repeated phrase. It's the second もってる in the line: だれもがもってる、もってる〜
    – Lukman
    Oct 24, 2011 at 0:47

3 Answers 3


It is common in songs, and it is not specific to children’s songs.

In the first case, the pitch of the lyric line is probably something like:

し(G) ら(G) ん(G) ぷ(G) り(G) を(F#) し(G) た(E) っ(F#) て(D)

but if you try to sing this as it is, there is a problem: gemination is not a sound but just a pause, and you cannot sing it with any pitch. Therefore, the vowel preceding the gemination is prolonged to fill the first part of the mora which should filled by the gemination:

し(G) ら(G) ん(G) ぷ(G) り(G) を(F#) し(G) た(E) ーっ(F#) て(D)

I did not check the other two videos, but I guess that they arose for the same reason.

  • Thanks for the extra explanation on the reason for the lengthening. It totally makes sense :)
    – Lukman
    Oct 25, 2011 at 7:43

Not hugely confident in this answer, but I'll try.

The gemination is supposed to be accomplished by a glottal stop in speech, and singing with a glottal stop is awkward at best and would sound strange even done properly. I imagine that the vowel lengthening is done to fill in a mora for rhythm/time purposes, and to indicate the omission. (That is, I know exactly what you're talking about, and in the examples I've heard, I don't even hear the in the sung lyric; it's entirely replaced by the vowel extension.)

It seems to me exactly comparable to the phenomenon in English songs of minimizing or omitting a final sibilant s sound (especially at the end of a musical phrase), which indeed is often combined with an extension of the preceding vowel.

Arguably it's different in Japanese because such vowel lengthening has much more potential to cause confusion with other words (as if there weren't enough homonyms in the language to begin with!) However, I think a certain amount of leeway has to be given here. After all, if I call after 雪【ゆき】ちゃん with a hearty 「ゆう~き!」, nobody would think her name had suddenly changed to 勇気【ゆうき】, right? (I think I got the correct kanji...)

  • 1
    I am not sure what you mean by “singing with a glottal stop is awkward at best.” Even when the preceding vowel is prolonged, the glottal stop is usually not omitted. Oct 26, 2011 at 12:17
  • Well, I didn't hear a glottal stop in the examples I'm recalling, and it certainly seems awkward to me when I attempt it. Oct 26, 2011 at 13:19
  • 2
    I had been confusing glottal stop with something else. I do not think that gemination of the consonant /t/ (as in the question) is pronounced as a glottal stop, regardless of whether it is in a song or not. Oct 26, 2011 at 15:57
  • 1
    In other words, your statement “The gemination is supposed to be accomplished by a glottal stop in speech” is false, at least in the case of gemination of consonant /t/. Oct 26, 2011 at 20:34
  • 1
    Gemination in Japanese generally doesn't involve a glottal stop, so I think it's weird to indicate it with /ʔ̩/ or to refer to it as a "glottal stop". But some Japanese phonologists do consider it a phoneme, which they label /Q/. For example, the prefix ぶっ (as in ぶっ飛ばす) could be written buQ-.
    – user1478
    Jun 13, 2014 at 2:41

My first impression is that the only purpose of the extra mora is to create an additional mora for rhythmic reasons. Vowel lengthening does occur for expressive reasons, but I don't think any connection can be drawn to gemination. From a phonological perspective, I see no reason a vowel would lengthen before a geminate. I don't recall ever seeing such a process in any language.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .