Reconstruction talk:Proto-Indo-European/ḱerh₂-

horn

The lemma horn says it comes from this and so does https://www.ahdictionary.com/word/indoeurop.html, but it's not listed in the derivatives here. --Espoo (talk) 12:39, 20 February 2018 (UTC)

It's from Germanic *hurną, which is given under *ḱr̥h₂nós (and which probably could use a separate PIE noun entry). --Tropylium (talk) 08:59, 22 February 2018 (UTC)

In view of the usual development of *r̥ to *ri in Proto-Celtic (cf. Schumacher 2004: 126; adding a laryngeal to the reconstruction does not help, it would not yield a different reflex), Proto-Celtic *karno- seems to go back to a form with an intervening fricative like *karsno- < PIE *ḱr̥(h₂)s-no-. This is supported by Proto-Germanic *hurznutō "hornet", originally "the one with horns", which implies that *hurna- "horn" was really *hurzna- originally, going back to the same preform. Latin cornum, too, can go back to *ḱr̥(h₂)sno-. In Slavic, the fricative would have been lost regularly, and in (East) Baltic the loss may be regular too, if the Baltic word is related at all (which is not certain; it can also be a loan from Slavic, and maybe st- is a reflex of Proto-Slavic *s- here, originally *ts-), and even the appurtenance of the Slavic word is unclear (maybe the word is an old Balto-Slavic word, but has a different origin), see Latvian stirna, so the Balto-Slavic evidence is inconclusive, and not decisive. The Sanskrit term is a different formation – not deriveable from *ḱr̥(h₂)no-, and certainly not from *ḱr̥h₂-n-, since there is no trace of a laryngeal –, so it cannot disprove the sibilant. (It frankly looks like a compound *ḱr̥-ng(ʷ)-o-, but the second element remains obscure.) Only the Germanic–Celtic–Italic equation is completely secure. Since formations like *ḱr̥(h₂)-s-n- are well-paralleled, while *ḱr̥(h₂)no- would be singular, the reconstruction with the sibilant, *ḱr̥(h₂)-s-no-, is to be preferred, especially on the strength of the Celtic evidence. --Florian Blaschke (talk) 05:02, 26 February 2018 (UTC)

On second thought, PIE *ḱr̥h₂-no- would yield *krāno- in both Italic and Celtic anyway (*ḱr̥h₂sno- would yield *krāsno- in Italic and *krasno- in Celtic), so our reconstruction with laryngeal cannot be correct in the first place. It's surprising that this problem has apparently not been recognised in the literature, nor the problem with the missing laryngeal reflex in Sanskrit. --Florian Blaschke (talk) 05:25, 26 February 2018 (UTC)

Hm! In fact, the laryngeal deletion rule CH.CC mentioned here would only work (at least in Italic and Celtic) if we add the -s-, as I originally suggested! --Florian Blaschke (talk) 05:37, 26 February 2018 (UTC)

This root probably should be "de-emended" into *ḱer- 'horn', with *ḱer-h₂- 'head' as merely one extension among several. See e.g. Gąsiorowski 2017. --Tropylium (talk) 14:06, 27 February 2018 (UTC)

@Tropylium If Proto-Celtic *karno- goes back to a form without laryngeal in the first place, a reconstruction with a plain suffix *-no- might or might not work. It depends on the outcome of the combination *-r̥n- in Celtic, which is uncertain. What's clear is that the PIE syllabic liquid *r̥ yields PC *ri before stop and *m, but *ar before fricatives; what happens before *n I can't say. Despite what I said above, I now think it's also uncertain what happens in *Tr̥Hs (I'm not sure what made me so certain above). However, Proto-Celtic *grānom decisively argues against the reconstruction *ḱr̥Hno- posed in the article. I'm still puzzled why this problem has not been recognised. In any case, that reconstruction is very probably wrong; and it still seems to me that a reconstruction with a suffix -sno- is preferrable, especially in view of formations in *ḱ(e)r(h₂)-s- and *ḱ(e)r(h₂)-s-n- that already exist. --Florian Blaschke (talk) 10:09, 17 April 2021 (UTC)

Also, you're right, already Nussbaum in his 1986 paper argued that the laryngeal is not part of the stem, and Kloekhorst (2008) agrees. --Florian Blaschke (talk) 11:43, 17 April 2021 (UTC)

*ḱr̥h₂-nó-s

Proto-Balto-Slavic *śírˀnāˀ comes from Proto-Indo-European *ḱr̥h₂-néh₂ and not directly from *ḱr̥h₂-nó-s Ентусиастъ (talk) 06:21, 9 May 2021 (UTC)