Why is the gemination mark / sokuon in Japanese written as a small tsu?

7 comments

1. I always assumed that it was just because some words were spelled with an actual tsu/tu sound that eventually evolved to become “swallowed” by the next consonant, and from there the small tsu was generalized to spell gemination in general.

   Like if you try to pronounce “いつて” fast, I think you’ll naturally trend to end up saying something close to “いって” because it’s just easier to pronounce.

   That’s a pure guess though, I too would be interested to learn about the evolution of this spelling convention. I think small kana in general is a fairly recent development, no?

2. ok first read this:

   https://japanese.stackexchange.com/questions/751/how-did-little-tsu-become-a-lengthener

   then go here and search for “Sokuon”:

   https://en.wikipedia.org/wiki/Historical_kana_orthography

3. which other kana would you prefer?

4. So the relevant geminates in Middle Japanese came about via a few different processes and in morpheme-medial position was subject to allophonic variation between a nasalized and unnazalised consonant, depending on the following consonant. These additional consonants were written in various ways, from not being written at all, to kana corresponding to the same place of articulation, to 牟 and its katakana deriviative ム, and eventually to つ. This seems to have been in part because つ was used to represent the pronunciation of /t/ with no following vowel in Sino-Japanese loanwords (final /t/ with no vowel not being something that occurred in Japanese natively in the period; eventually these words would acquire a following vowel.) Generally there was a lot of variation preceding that and I am not sure that there’s really any solid reason other than it being convenient and つ already being used for a weird sound.

   Unlike morpheme-medial position, nasality was distinctive morpheme-finally. This is how we get differences like 死んで vs 知って; in other words, the nasality here was a property of the additional consonant and not something that depended on the following consonant (like it did in morpheme-medial position). Both of our sounds here, the morpheme-final nasalised and unnasalised sounds and the morpheme-medial sound (that didn’t care about nasality) were also written in all of the ways given above; eventually, as you know, つ prevailed for the morpheme-medial position and for the morpheme-final position, while ん and ン came to be used for the morpheme-final nasalized sound (where it would make sense to distinguish the two sounds, because in this position, they were distinctive, unlike medial position.) If you want to look into the changes that gave some of those moraic consonants, look up ‘onbin sound changes’ or something like this.

   As for when it first started to be written in a smaller size, I’m not really sure, but I think it is quite recently.

5. In the ainu katakana orthography, small symbols like ク represent syllable-final consonants (like, アク with a small ku would be “ak”). So you could take it as a similar convention, they’re writing -t, which gets assimilated to the next consonant.

   This sort of makes sense at least to me because the gemination in japanese is slightly glottalized, which reminds me of English -t. But ymmv. And that’s probably not *why*.

6. Sokuon what ?

7. JFZジョージ先生のビデオから来たんだけど爆笑