Reconstruction talk:Proto-Indo-European/ǵʰelh₃-

Comments
Why doesn't Latin word belong here? It has long been consensus among linguists that "gilvus" is another reflex of *ǵʰelh₃-wo-s, alongside regularly expected helvus. I could quote more modern sources, but I really see no point in it. And may I add that the "etymology" section on the very page cites it as a descendent? 187.23.90.44 03:59, 8 July 2016 (UTC)
 * Lewis & Short do not represent modern Indo-European theory by any stretch an cannot be counted as a source. The page itself refers to a dialectal variant, *ǵeylh₃- which is so different that I have trouble believing it exists, much less is related. De Vaan, the authority on Latin etymology, does not mention gilvus at all, even in the section on helvus, and he would if there were a reliable connection. What linguists support your claim? — JohnC5 04:45, 8 July 2016 (UTC)


 * There's no way that gilvus can be a regular development in Latin of *ǵʰelh₃-wo-s. In fact, given that (per Rix, cited in de Vaan) the /l/ would precede /w/ and therefore be realised as the velarised allophone in Italic, the expected result would likely be > *hulwos >  (compare  > ) – while  could be the result of a zero-grade variant *ǵʰl̥h₃-wo-s. The only sensible explanation is that gilvus is a loanword, possibly an unattested Celtic cognate of  – in fact, even Proto-Germanic itself can't be ruled out as a source, AFAICS. --Florian Blaschke (talk) 18:45, 13 November 2021 (UTC)