Reconstruction talk:Proto-Indo-Iranian/twárštā

@AryamanA: Hi! Tvaṣṭar not tvaṣṭṛ! Tvaṣṭar is a "Vedic god-creator". Thank you.--Calak (talk) 14:38, 26 January 2018 (UTC)

@Calak: tvaṣṭar is the vocative case and the stem form, and tvaṣṭṛ is the regular lemma for r-stems in Sanskrit. It's the same word. —Aryaman^{A (मुझसे बात करें • योगदान)} 15:31, 26 January 2018 (UTC)

Accent

@AryamanA, Victar, it seems that Vedic is a *-tōr-stem, not a *-tḗr-stem as seen both by its accentuation and acc.sg. Tváṣṭāram as opposed to Tvaṣṭáram. I don't know what to say about the Avestan since -ōrə- is a very funny vocalism, but this reconstruction does not work for Sanskrit. —*i̯óh₁nC^[5] 04:54, 29 January 2018 (UTC)

@JohnC5: *-ā́ would be correct if it was *-r̥-, but it's probably not. The development of the Iranian could easily be PIr *twárštā > Proto-Avestan *θwərᵊštar > Avestan θβōrᵊštar. *twŕ̥štā would have yielded **θβə́hrštā though. --Victar (talk) 05:31, 29 January 2018 (UTC)

@AryamanA, Victar, JohnC5: *twarštā is incorrect, correct form is twr̥štā, please read this article first. Thank you.--Calak (talk) 10:00, 29 January 2018 (UTC)

@Calak, first off, there is no such thing as *twr̥štā -- it's either twr̥štā́, *twárštā or *twŕ̥štā, and I can tell the latter does not work for Avestan. Secondly, if you're creating entries from sources, you need to include those sources. You've been creating a lot of proto entries that just site {{R:pal:Mackenzie}}, which is rather pointless as a source as it gives no etymology for the MP words.

Reading the paper through, Lubotsky makes an incorrect supposition that *twárštā renders θβarᵊštar. In fact, PIr *-wa- > *-wə- is a predictable outcome, and -ōrᵊ- is an outcome of -ərᵊC-. Everything is on the level with the Avestan. I'll let AryamanA and JohnC5 comment on the outcome in Sanskrit though. --Victar (talk) 14:56, 29 January 2018 (UTC)

@Calak: Yeah, Lubotsky's reasoning is unnecessary in the case of the Sanskrit as the accent and suffix ablaut are far more elegantly explained by a *-tōr-stem. Then the loss of the stem-internal *r is simple dissimilation. His claim that a zero-grade is required for Celtic is refuted by Matasović, which puts Lubotsky's pleading that *twerḱ- had not full-grade forms on poor footing. Also, there are no comparanda to my knowledge for *R(ź)-tōr in PIE. Overall, this article constructs a far less compelling solution for problems that are solved more easily through preëxisting formations.

@Victar: I'd love to know more about PIr *-wa- > *-wə-. —*i̯óh₁nC^[5] 21:07, 29 January 2018 (UTC)

@JohnC5: Forgive me, I misspoke. I was thinking *-wa- > *-wā-, so *twárštā > (y_, w_) *θwārᵊštar > (_rC) θβōrᵊštar. --Victar (talk) 01:18, 30 January 2018 (UTC)

@Victar: Cool. Where'd you find out about this? —*i̯óh₁nC^[5] 01:33, 30 January 2018 (UTC)

@JohnC5: {{R:ae:Brill}} §7.4.3. and §7.13.3. --Victar (talk) 01:41, 30 January 2018 (UTC)

@Victar: Excellent! Thank you. —*i̯óh₁nC^[5] 02:27, 30 January 2018 (UTC)