Reconstruction talk:Proto-Japonic/wu

Lemma form

@Eirikr Should this be *wi instead? I found out that -u is probably a Japanese innovation; PR uses (infinitive) + *wori. Kwékwlos (talk) 17:47, 9 July 2019 (UTC)

• @Kwékwlos, no idea what "PR" is, although I'm pretty sure it's not Puerto Rico.

FWIW, wori is the continuative or stem form of verb woru, modern 居る (oru). This is probably an ablaut derivation from this same wu root.

It seems that the root wu does appear in historical texts, albeit rarely, such as in the Nihon Shoki: 「急居、此云菟岐于」 → 「急居、此これ菟つ岐き于うと云いう」. See also the Nihon Shoki text and explication here (about halfway down, search the page for 急居), and the KDJ entry for 急居 here. In this compound, the 居 kanji is given the reading う, and that is analyzed by Japanese lexicographers as the terminal form う of what became the 上一段 verb ゐる. And since ゐる begins with w, the terminal must also, thus pointing towards the existence of wu, even though historical (i.e. written-down) Japanese makes no u ↔ wu distinction.

So in answer to your question, "Should this be *wi instead?" I'd say "no", this belongs at *wu. Cheers, ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:56, 9 July 2019 (UTC)

@Eirikr You know that PR = Proto-Ryukyuan. As per talk page, I don't really think /wu/ would be reocnstructed here as there is no Man'yōgana character for this sequence. Kwékwlos (talk) 21:44, 2 March 2020 (UTC)

@Kwékwlos: Hmm. "You know that PR = Proto-Ryukyuan." Do I? Now, in this specific context, I do. I did not when I posted that previously. I've noticed various places in your posts where you seem to assume that your audience is much savvier to your mental context than we sometimes are. Avoiding ambiguity can make for more productive conversations.

Lack of man'yōgana is germane to Old Japanese. It says nothing necessarily definitive about earlier stages of the language. The existence of inflected forms with morae that begin with /w-/ or /y-/ would seem to suggest the existence of logical phonemes /wu/ and /yi/, even if such forms do not have any substantive presence in the historical language (that is, language for which we have texts). Monolingual Japanese sources describe the existence of such logical roots, such as the Kotobank entries for 植える here and here. The modern reading うえる is described as deriving from older うゑる, from うう, itself listed as ワ行二段.

Moreover, analysis of man'yōgana usage patterns is ongoing. The current state of the art might not paint a complete picture. For instance, the man'yōgana used for the specific root う for ゐる appears in the Nihon Shoki as 于 (entry), and the root う for うゑる appears as 宇 (entry). Both of these are reconstructed consistently for Middle Chinese as starting with the voiced glottal fricative /ɦ/, with some instead using the voiced velar fricative /ɣ/. Both of these are back-of-the-throat sounds, which could prompt rounding into a /w/ sound in some speakers. Both reconstructions also include a possible /ɦu-/ initial. While inconclusive, this does suggest the possibility that there may have been a /wu/ sound, or something close to it, at the very early edge of historical Japanese.

Beyond the concrete textual evidence, if we state flatly that there were no wu, then where would the w- have come from in other inflected forms? Japanese is not prone to random excrescences. ‑‑ Eiríkr Útlendi │^{Tala við mig} 22:55, 2 March 2020 (UTC)