Reconstruction talk:Proto-Japonic/koi

Derivation, derivatives

This is another highly problematic entry.

• The compounding form in recorded Old Japanese was probably originally ku, with compounding form ko treated in Japanese sources as a shift from earlier ku. See the KDJ entry here in Kotobank, and the Daijisen entry at the top of this page, for instance. Daijirin likewise states that kogane is a shift from kugane.
• Old Japanese kugane could not derive from the hypothesized *konkanay. I believe this *konkanay is a mistake in reconstruction, and never existed.
• The vowel shift from u to i is also in evidence in OJP kutu ("mouth", compounding form) and kuti (standalone form), so we don't need to posit any vowel value /o/ to allow for the shift to i.

I'll remove *konkanay for now. Longer term, I think we should consider moving this entry, unless we can clearly show that Japanese sources are incorrect about kugane. ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:05, 26 March 2021 (UTC)

I've read through Pellard's paper "Ryukyuan perspectives on the proto-Japonic vowel system", accessible online at https://hal.archives-ouvertes.fr/hal-01289288/file/Pellard_2013_Ryukyuan_perspectives_on_the_proto-Japonic_vowel_system.pdf. It would be good to include references for Proto entry reconstructions.

Pellard lays out a good argument for Proto-Japonic *koi. However, he doesn't go into the core issue we face with the standalone vs. compounding forms, where standalones appear to have that final -i that is missing from compounds. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:55, 30 March 2021 (UTC)