Reconstruction talk:Proto-Japonic/onansi

Arisaka's law
Is this reconstruction valid? I see the Okinawan form as deriving from a borrowing from Japanese, given the lack of other Ryukyuan cognates. Kwékwlos (talk) 23:10, 22 February 2023 (UTC)


 * I am curious about the decision to reconstruct this as *ənansi, for a couple reasons.
 * Arisaka's Law states that we should never see ⟨o₂⟩ in the same morpheme as ⟨o₁⟩, ⟨a⟩, or ⟨u⟩.
 * So far as I'm aware, doesn't distinguish between vowel-only ⟨o₁⟩ and ⟨o₂⟩.
 * Consequently, I see good reason not to reconstruct initial ə-, and reason to reconstruct initial o- instead.
 * Separately, the KDJ entry here at Kotobank notes that this was also realized as onasi -- but then again I can only find onazi in ONCOJ.
 * There is also a separate form oyazi. The notes in the oyazi entry in the KDJ (linked above) suggest that oyazi may have been from  +, while onazi was possibly from  + .  This is interesting, and the semantics make some sense.  This idea could merit further exploration.  ‑‑ Eiríkr Útlendi │Tala við mig 00:28, 23 February 2023 (UTC)