Reconstruction talk:Proto-Japonic/u

Reconstruction?

This strikes me as extremely unlikely. The contortions required to go from /airu/ to /u/ leave me most unconvinced. The root of ある (aru) is /ar/, and that final /r/ is an integral part of this verb stem. Any derivation or shift from the root /ar/ will include this /r/. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:07, 3 June 2019 (UTC)

@Kwékwlos, thank you for the go at improving things.

A derivation from -ay is still problematic, however. The underlying root form of this verb is /u/. I see no clear path phonologically for -ay to result in /u/. ‑‑ Eiríkr Útlendi │^{Tala við mig} 22:22, 12 February 2020 (UTC)

Perhaps this was a post-Japonic innovation, in which the PJ potential suffix was separated and then taken as a verb meaning "get, attain". In any case, if it was an actual verb, reconstructing *u would be the go-to case here, but where is the stem of the verb? (e.g. *kuru ~ *kuraCi (“to do for someone”) with a stem *kur-, yielding くれる (kureru)). Kwékwlos (talk) 22:27, 12 February 2020 (UTC)

@Kwékwlos, the terminal form of the verb would have been just /u/. This was conjugated as a 下二段活用 (shimo nidan katsuyō) verb. See w:Classical_Japanese#Verbs_(動詞_Dōshi) for details on this conjugation pattern. We have ample historical examples of this verb displaying the root /u/ as far back as at least the Man'yōshū, completed in 759.

In short, even had the potential suffix -ay been treated as a standalone verb stem, there is no phonologically sensible way to derive u from ay. No such derivation appears to be possible. Going the other way, we potentially have both historical and reconstructed examples in the ways that verb conjugations have evolved, were u becomes e. But as you note, that's more a case of suffixing and fusion, and we'd still need to have the verb core to start from. ‑‑ Eiríkr Útlendi │^{Tala við mig} 20:18, 13 February 2020 (UTC)

@Kwéklos, revisiting this, I realize that I'm not sure what you mean by "the potential suffix *-ay. Are you referring to the -e that appears in modern transitive godan verbs, such as 飲む (nomu, “to drink”) → 飲める (nomeru, “to be able to drink; to be drinkable”)?

If so, there's a paper from 2016 by 三宅俊浩 (Miyake Toshihiro) that lays out a compelling case that this -e is not from this same verb 得る (eru), but instead arose through a shift from 四段活用 (yodan katsuyō, “quadrigrade conjugation”) to 下二段活用 (shimo nidan katsuyō, “lower bigrade conjugation”) as a means of using transitive yodan verbs in an ergative fashion to talk about how the objects can be [VERB]-ed. Similar to English expressions like "this car drives well" or "this cake bakes quickly", using usually-transitive verbs in an ergative fashion with the usually-objects as subjects instead, to talk about a quality of the usually-objects in relation to the action of the verb. This derivation is also why the grammar of potential verbs is often weird for English speakers learning Japanese -- in English, potential is talked about as an ability of the agent, whereas in Japanese, it's instead a quality of the patient. 私は日本語が話せます (Watashi wa Nihongo ga hanasemasu) in Japanese conceptual terms is not so much "I can speak Japanese" and instead "for me, Japanese is speakable". See also this post I wrote over on the Japanese Stack Exchange about this specific class of potential verbs. ‑‑ Eiríkr Útlendi │^{Tala við mig} 17:21, 8 December 2022 (UTC)

@Eirikr continuing this, what would be the best reconstruction instead of mistaken *airu? Kwékwlos (talk) 23:44, 27 February 2023 (UTC)

@Kwékwlos: I suspect just /u/. The Ryukyuan reflexes of Japanese vowel-stem verbs and irregular verb 来る (kuru, “to come”) seem to all be reflections of an ancient compound formation, whereby we use the continuative / gerundive / infinitive / -masu stem + some other element (possibly wiru?) that reduces to /uN/ or just /N/. For example, modern Japanese 来る (kuru) had Old Japanese terminal form /ku/, and this reflects in Okinawan as /t͡ɕu:N/ -- apparently from continuative stem ki- + this other element. Modern Japanese 知らせる (shiraseru, “to let know, to inform”) from older /sirasu/ reflects as sirasjuN. Poking around JLect, I cannot find any reflections of Japanese 得る (eru), Old Japanese 得 (u), so I cannot confirm the Ryukyuan descendants we currently list. That said, I do see some places where Japanese verb stems ending in /-e/ reflect as /-iN/, such as keru "to kick" appearing as Okinawan /kiːN/. ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:24, 28 February 2023 (UTC)