Reconstruction talk:Proto-Celtic/kʷennom

Proto-Celtic *kʷennom

know it's been almost five years and an entire global pandemic, but do you remember where you got this suggested etymology of Proto-Celtic *kʷennom from Proto-Indo-European *kpwen-nó-m < *káp-wl̥ ~ *kap-wén-s? —Mahāgaja · talk 20:44, 27 August 2024 (UTC)

@Mahagaja: I do not, and a first search in my sources came up with nothing. --{{victar|talk}} 06:49, 28 August 2024 (UTC)

That's why I always write "my original etymology" in an edit summary when there is no source. It's not because of vanity. It's for keeping track. Vahag (talk) 07:25, 28 August 2024 (UTC)

Do you mind if I remove it then? It just seems a little too adventurous to be there without a source. —Mahāgaja · talk 07:45, 28 August 2024 (UTC)

Too adventurous? --{{victar|talk}} 07:56, 28 August 2024 (UTC)

Well yes, or too uncertain. All the suggestions of an l/n stem on RC:Proto-Indo-European/kap-#Etymology 2 are peppered with question marks, so it's far from certain that anything like *káp-wl̥ ~ *kap-wén-s existed at all, let alone that it formed a *kpwen-nó-m from which *kʷennom is derived. If someone has stated this hypothesis in print with peer review, great, we can mention it. But without a source, it just seems far-fetched. —Mahāgaja · talk 08:11, 28 August 2024 (UTC)

@Mahagaja, I'm confused, since when do we not allow personal research? --{{victar|talk}} 07:28, 13 September 2024 (UTC)

We do, but within reason. If an etymology seems plausible enough that it raises no eyebrows, it probably doesn't need to be sourced, but this one raised both Mellohi!'s and my eyebrows. It requires some pretty substantial justification to be believable, ideally in print by someone with a good reputation in Celtic/Indo-European historical linguistics. —Mahāgaja · talk 16:00, 14 September 2024 (UTC)

Pretty much. It's one thing to do original research to:

• Pick the best reconstruction to adopt whenever other scholars give conflicting theories
• Interpret/extend a morphological process or sound law that a scholar illustrated with one example to another word that scholar did not talk about
• Interpreting the context of what a scholar said
• Correct a blatant mistake or implausible reconstruction by a scholar; EDPC has tons of outright errors in it

... etc. within reason of course. It's a different thing entirely to come up with an implausible/morphologically tortured etymology of your own, especially in a situation where scholars have outright said an etymology can't be found. — Ceso femmuin mbolgaig mbung, mellohi! (投稿) 17:26, 14 September 2024 (UTC)

tl;dr the more extraordinary the claim, the more extraordinary the evidence needed. — Ceso femmuin mbolgaig mbung, mellohi! (投稿) 17:34, 14 September 2024 (UTC)

@Mahagaja: If we start with what we do know, P(W)G *habulō points to pseudo-PIE *kap-u-l-ō. PG terms ending in *-uLa(n)- quite often arise from original PIE *-wL/n-stems, compare *kelurō (“throat”), from *gʷélh₁-wr̥ ~ *gʷl̥h₁-wén-s, and *feturaz, from *péd-wr̥ ~ *pd-wén-s, a development found across IE, compare PBL *súlˀnis, from *séh₂ul-n-, from *sóh₂wl̥. The evidence for PIE *kap-wl̥ ~ *k(a)p-wén-s is quite strong, in my opinion.

In light of that, I see *káp-wl̥ ~ *kp-wén-s > *kpwén-ō ~ *kpun-nés > PC *kʷenū ~ *kʷennos > *kʷennom not out of the realm of possiblity, and at the very least, worth mentioning. --{{victar|talk}} 19:26, 16 September 2024 (UTC)

It's not utterly impossible, but it's nowhere near obvious enough for us to be mentioning it as original research. If it's your own idea, write up an article and submit it to a Celtic or Indo-European linguistics journal, and once it's published, we can mention it then. —Mahāgaja · talk 21:16, 16 September 2024 (UTC)

No need to be snide. Which part in particular do you take issue with? -- {{victar|talk}} 05:11, 17 September 2024 (UTC)

Well, are Proto-Indo-Iranian *kapā́las and Proto-West Germanic *habulō the only things that point to an l-suffix? How sure can we be that a form with an l-suffix was present in the protolanguage rather than being innovated in those two languages? And is there any evidence at all for a zero grade *kp-? Looking through Reconstruction:Proto-Indo-European/kap-#Etymology 2, I get the distinct impression that what appears whenever a zero grade is expected is actually *kap-. (Maybe this means the full grade is actually *kh₂ep- and the zero grade *kh₂p-, but if so, we'd hope to see Sanskrit forms in khap- and/or kip-.) But again, I'm not saying this reconstruction is totally out of the question, just that it's not so obvious that we can really mention it if it's never been published anywhere. —Mahāgaja · talk 10:51, 17 September 2024 (UTC)

The existence of *l in Proto-Indo-Iranian is still hotly debated, but evidence from more recent sources points otherwise. Kwékwlos (talk) 15:48, 18 September 2024 (UTC)

@Mahagaja: There are three PIE suffixes that I know of that contain *-l-: *-lós, *-l̥, and *-wl̥. Of the three, *kap-u-l-ō is most easily explained by starting with a *-wl̥ suffix, which I find the "near obvious" etymology.

A heavy onset cluster like *kp- would likely have either triggered an epenthetic vowel, causing it to appear as a full grade, or been simplified, like seen in *sɸenyos and *dū. Supposing *kpw- may have simplified to PC *kʷ- in this case is no large leap of faith. I also wouldn't be surprised if it went through an intermediary metathesis stage, *kpw- > *pkw- > *pkʷ- > *(ɸ)kʷ-. --{{victar|talk}} 07:56, 19 September 2024 (UTC)

Where actually is there an l/n-alternation reconstructed, apart from *sóh₂wl̥? If that's all there is, this doesn't exactly point to a productive suffix. Exarchus (talk) 13:11, 19 September 2024 (UTC)

The suffixes *-l̥, and *-wl̥ and generally considered variants of *-r̥, and *-wr̥, respectively, and terms ending in *-lós are likely secondary constructions derived from *-l̥. --{{victar|talk}} 20:15, 19 September 2024 (UTC)

"The suffixes *-l̥, and *-wl̥ and generally considered variants of *-r̥, and *-wr̥" Source? Exarchus (talk) 20:26, 19 September 2024 (UTC)

To clarify: insofar as a *(w)l̥ suffix exists, it is obviously a variant of *-(w)r̥. But it apparently only occurs in *sóh₂wl̥ (whether that has a *-l̥ or a *-wl̥ suffix is already not clear), so the logical thing to conclude is that it was a non-productive suffix in PIE. Exarchus (talk) 07:33, 20 September 2024 (UTC)

{{R:ine:HCHIEL|2112}}, {{R:ine:Beekes 2011|page=206}}, etc.

Variants in *l are indeed rare, but not as rare as *sóh₂wl̥ being the only occurrence. PIE *h₃érō is better reconstructed as *h₃ór-ōl ~ *h₃ér-n̥-s (“bird, eagle”); PG *aflaz (“hearth”), Hittite 𒄩𒀊𒁉𒈾 (/⁠ḫappenaš⁠/, “baking kiln, fire-pit”)}, is likely from *h₂óp-l̥, ~ *h₂ép-n̥-s; PG *himilaz, *himinaz, is likely from *kém-l̥ ~ *km-én-s, etc. --{{victar|talk}} 08:12, 20 September 2024 (UTC)