Reconstruction talk:Proto-Indo-European/h₂éǵr̥yos

@CodeCat: I'm...skeptical of this. If it were *h₂ǵr̥yós with Sanskrit *जृय (jṛyá), that would seem more plausible, as AG seems to have moved the stress in the case of most -ιος adjectives, but the lack of syllabic r in Sanskrit with the oxytone makes me think that these are both secondary. —JohnC5 23:16, 9 May 2016 (UTC)

But what part are you skeptical about? —CodeCat 23:18, 9 May 2016 (UTC)

I guess I'm skeptical that this form is reconstructable. *h₂éǵr̥yos should give Sanskrit *अजृय (ájṛya), but I suppose the analogy with other -yá adjectives could have moved the accent and analogy with the word अज्र (ajra) could have desyllabified the r. For AG, I would expect something like *ágrəyos > ἀγρός but again analogy could have reformed this to ἄγριος (ágrios). That, however, doesn't explain the irregular PIE adjective h₂éǵr̥yos as opposed to *h₂ǵr̥yós.

If you reconstructed *h₂ǵr̥yós, analogy could get you both to the AG and Sanskrit forms, but again, that is a lot of appeal to analogy. I feel that too much work has to be done to either candidate PIE form in order to produce the Sanskrit or AG, particularly when both sould easily be made by secondary -yá or -ιος respectively. This just doesn't seem like a grade a reconstruction to me. On the other hand, Old Armenian արտի (arti) could perhaps be a reflex, which would add a little credence. —JohnC5 23:43, 9 May 2016 (UTC)

See the usage note on the *-yós page. The lack of accent is regular for derivations of thematic nouns. This isn't the only one that Ringe gives. —CodeCat 23:58, 9 May 2016 (UTC)

That helps, and I'm frustrated at myself for missing that usage note. Should we have an alternative form for *-yos then? It still means that a fair amount of reanalysis had to take place, but it makes me think that adding Armenian would be appropriate. Also, thanks for the discussion. —JohnC5 00:12, 10 May 2016 (UTC)

We can have *-yos, but it's not an alternative form so much as an allomorph. What Ringe doesn't tell is what happens when the suffix is added to a noun with a stressed thematic vowel. Since the vowel is dropped, where does the stress go? I would guess it moves onto the suffix, but who knows. —CodeCat 00:27, 10 May 2016 (UTC)

Yep, allophonemorph (I need to go to sleep). The only example I can must find of a stressed thematic vowel that that seems to take this suffix is *yugóm with Sanskrit युज्य (yújya) (maybe युग्य (yugya)) and AG ζῠ́γιος (zúgios). —JohnC5 03:21, 10 May 2016 (UTC)