Reconstruction talk:Proto-Indo-European/perḱ-

pulcher
Thank you for the help here. But note firstly that an o-grade stem form for pulcher is both unsourced (or at least, I can't seem to find it), and unneeded, as *e > o is regular before l pinguis (and later olC > ulC) cf. *welt > volt > vult. Secondly I'd prefer to not lsit any PIt form as the assumed dissimilation could have occured in any moment ranging from PIE to right before this rounding (which is post-PIt, Pokorny's form  is hence early Latin, and not PIt), so by giving a PIt form with *r we're assuming retention, for which we have no evidence; by giving a form with *l we're dating a possibly OL dissimilation back to PIt, for which there's no evidence as well; by giving both we make both mistakes at the same time. What do you think? I've left both for now, with a question mark. (Thirdly, accent is inreconstructible and unsourced.) Catonif (talk) 10:02, 28 May 2023 (UTC)
 * Pokorny reads, but Mallory gives later a dissimilation of OLat from , from earlier . I am aware of the Latin *el > *ol, but if we were reverse engineering from PIE, the only two available reconstructions are , a zero-grade r-stem adjective, or , some kind of denominal adjective from an original r/n-stem noun. Either way, an e-grade is unlikely, and  actually makes for a much better reconstruction. --  20:38, 28 May 2023 (UTC)
 * Oh I see, thank you, I didn't check Mallory. I trust your PIE morphological/accentological analysis, which favours a zero- or o-grade, but, newbie question, given that we don't know how late and "impure" the proposed formation is, is an e-grade really so unlikely? As far as I see the full grade made its way in the Greek descendant as well.
 * Analysing the situation: the PIE form is already wobbly (all three grades are formally conceivable by its descendant), and adding to that a dissimilation that we can't date, during the PIt period, if we follow this derivation, the word can be *porkros, *polkros, *perkros or *pelkros... can we just place  here? No source seems to be attempting to reconstruct a specifically Proto-Italic form either, as it wouldn't have much importance in any case, so we wouldn't be deleting any sourced reconstruction. Catonif (talk) 16:14, 29 May 2023 (UTC)
 * What's not being overtly accounted for on the article is an underlying mobile n-stem. That's why the -n- forms are irregular, thus from, and presumably the full-grade Greek as well.
 * Starting from a PIE, an Italic reconstruction is indisputable, and the dissimilation would have occured after this form. --  19:29, 29 May 2023 (UTC)
 * I'll trust the first point, but how are you dating the dissimilation. Catonif (talk) 14:34, 30 May 2023 (UTC)
 * I'm not, other than it didn't happen in PIE, which it wouldn't've. -- 18:32, 30 May 2023 (UTC)