Talk:あるいは

@Eirikr What should we call kana spellings held to be historical spellings before WWII but later revealed to be not, such as つくゑ, もちひる, and あるひは? --Dine2016 (talk) 12:33, 5 August 2019 (UTC)

@Dine2016, I like your labeling as non-etymological historical spelling. That seems to be optimally brief while still clearly explaining things.

Separately, I'm curious about the romanization. While this is etymologically parseable as aru i wa, I suspect that this is realized more as a single unit -- unlike それ, それで, or other constructs, あるいは can only take that は, no も or any other particle, suggesting that this can be viewed as a unit, and romanized thus as aruiwa without spaces. ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:10, 5 August 2019 (UTC)

(Your ping didn't reach me. Only by adding a new signature instead of modifying an existing signature can any ping take effect.)

@Eirikr: Thanks. Well, I was wrong in the post above: the Japanese was aware of the etymology before WWII, as shown by the 大辞典（平凡社）:

アルイワ　或は　副　あるひはとも書く。接續助詞の或はより轉ず。恐らくは。若しかすると。…
アルイワ　或は　接　ありの連體形に助詞いのついたものより生ぜる語。このいは名詞・連體形の下につけてその主語たるを示す辭。…あるは・或るものは・ある時はの意。…後轉じて接續詞となり、異なる二つの事を竝べて云ふ語となる。若しくは。または。但し古來、あるひはとも書かれてゐる。…

Also, the 大日本国語辞典 (also published before WWII, and predecessor of Shōgakukan's KDJ) states the following:

もちふ(ウ)　用　（他動上二）　〔古き用例によれば、もちゐると上一段に活用せし方多くして正しきやうなれど、今は慣用に從ふ〕

Given that the Japanese did know the etymology before WWII, もちひる and あるひは may just have been "common variants in historical kana orthography".

By the way, according to the 大辞典（平凡社） entry above, the い (i) is a topic marker instead of a nominalizing suffix. --Dine2016 (talk) 08:57, 6 August 2019 (UTC)

(Re: the ping, I removed my own sig entirely and re-added ~~~~, but I guess that doesn't work? Ah, well...)

@Dine2016: There are various examples throughout historical texts of "incorrect" kana spellings, especially as we get closer to the modern era and more and more sounds had collapsed, resulting in homophonous spelling possibilities. "Variants" seems fair enough as a simple descriptor.

There are broad sets of verbs that developed "y" conjugation stems from such a shift in (around) the Muromachi period, of which 用いる was one: root form moti(w)u gave rise via normal processes to mochiwiru, then as h and w converged, we see the もちひる kana spelling, and as w softened and the y became more pronounced in front of bare vowels in certain environments (c.f. w:ja:え, w:ja:い), we find back-formation mochiyu and its conjugated forms such as mochiyuru and mochiyure. もちゆ is treated now as a separate (archaic / obsolete) conjugation pattern, while もちふ and もちひる are treated as historical spellings (from what I've seen so far, anyway).

Regarding い (i), I've never understood it to be a suffix per se: I've read it described as an emphatic nominative particle, albeit already rare in Old Japanese and restricted in use to 漢文訓読 contexts, largely disappearing over the course of the Heian period. Relevant bit from Daijirin:

い

〔上代語。平安時代には、限られた経典の訓読にのみ用いられ、院政時代以降は消滅した〕

㊀ （ 格助 ）

名詞・または名詞的な語に付く。主格を強めて示す。「いは」「いし」などの形が多い。 「紀伊きの関守－留とどめてむかも／万葉集 545」

The entries from Daijisen and KDJ both class this as a 副助詞, but explain that 「名詞、名詞に準じる語に付く。上接の語を特に示したり、語調を強めたりする。」 -- which basically sounds like an emphatic nominative to me.

This particle appears to have fused in certain specific terms, perhaps arising from a prehistoric noun class that required this particle for standalone usage -- words like reconstructed proto-Japanese ma (“eye”), found in OJP as compounding form ma or standalone me₂, or kamu (“god, deity, spirit”), arising in OJP as compounding form kamu or standalone kami₂, etc. (I'm intrigued by this noun-class possibility; from what I've found so far, nouns that exhibit this kind of compound / standalone vowel shift roughly correspond to (some of) the "inalienable" noun class found in Polynesian languages like Māori and Hawaiian... but I digress; word-nerd fun. :) )

Cheers, ‑‑ Eiríkr Útlendi │^{Tala við mig} 16:37, 6 August 2019 (UTC)