Reconstruction talk:Proto-Indo-European/peyḱ-

Balto-Slavic
The Balto-Slavic languages have sibilants that reflect *ḱ rather than *k. Should this be instead? 19:58, 30 September 2013 (UTC)
 * That's just what I was about to ask. Unfortunately I don't know enough about Iranian to know whether Avestan -xš- (which generally corresponds to Sanskrit -kṣ-) can come from a cluster involving *ḱ. —Aɴɢʀ (talk) 21:17, 30 September 2013 (UTC)
 * It looks like the effect of the RUKI law on the sequence -ks-, but I don't know what would happen to a hypothetical -ḱs-. Maybe the RUKI-ization would cause the palatovelar to depalatalize? 21:19, 30 September 2013 (UTC)
 * If our etymology of texo is correct, then -ḱs- becomes -kṣ- in Sanskrit but plain -š- (not -xš-) in Avestan. —Aɴɢʀ (talk) 21:42, 30 September 2013 (UTC)
 * The source confirms that. It says that PIE *ḱs emerges in Proto-Indo-Iranian as *ćs, with the regular change *ḱ > *ć as in other environments. This *ćs then surfaces in Iranian as *š, and in Indo-Aryan (Sanskrit) as *kṣ. The cluster *ks emerges as *kš, which becomes also *kṣ in Indo-Aryan and *xš in Iranian. The "thorn cluster" combinations *tk and *tḱ metathesize into *ks and *ḱs and develop identically from there onwards. So the only two sources for Avestan *xš are PIE *ks and *tk, neither of which seem to match Balto-Slavic. In Balto-Slavic, *ks gives *kš (found in the name Aukštaitian if I'm not mistaken) and this then develops into *x in Slavic (seen in the s-aorists of velar-stem verbs). The combination *ys becomes *yš by RUKI, but I don't think *yk evolves in any special way. *ḱ on the other hand gives Balto-Slavic *ś, which becomes š in Lithuanian and s in Latvian and Slavic, which is the pattern we see here.  21:47, 30 September 2013 (UTC)
 * So Avestan seems to come from peyk- while B-S seems to come from peyḱ-. There are some cases where it's the other way round, where B-S seems to have a plain velar while I-I seems to have a palatal (e.g. the goose word). Looks like some inner-PIE dialectal variation, unsurprising especially if the palatals and the plain velars were originally allophones of the same phoneme whose distinction only later became phonologized, as has often been suggested. —Aɴɢʀ (talk) 21:59, 30 September 2013 (UTC)
 * Avestan seems to come from *peyk-st- or something similar. Some people have formulated rules for when palatovelars appear in Balto-Slavic as sibilants and when as velars. I think it's related to adjacent vowels but I don't remember how exactly or which sources it is from. 22:08, 30 September 2013 (UTC)
 * I see the page has been moved, and the Sanskrit descendants match the new name. But how is the Avestan form accounted for? 11:43, 1 October 2013 (UTC)


 * has created many similar PIE appendices from Pokorny's IED, in what is today obsolete notation, and some wrong descendants which now have better etymologies. --Ivan Štambuk (talk) 19:23, 1 October 2013 (UTC)
 * Still, how did Avestan fra-pixšta- come from this? It shouldn't be phonologically possible. 19:26, 1 October 2013 (UTC)
 * See here, §86, esp. the end. --Ivan Štambuk (talk) 22:37, 1 October 2013 (UTC)
 * Ok, thank you. You might want to add that as a reference for the Avestan descendant? 23:04, 1 October 2013 (UTC)

Mistake
Proto-Indo-Iranian *pinášti comes from PIE *pi-ne-s-ti < *peis- (Lat. pīnsō). Not from *pi-ne-ḱ-ti < *peiḱ-. Evlampij (talk) 03:35, 9 September 2023 (UTC)