Reconstruction talk:Proto-Germanic/dulgaz

@Leasnam Hey, from what source did you get the PIE etymon? Pronk-Tiethoff in The Germanic Loanwords in Proto-Slavic (2013) suggests that the Germanic term is not a cognate of but a loan from the Celtic, which would mean that the entire PIE construction is based upon the Celtic only (since the Slavic terms were loaned from Germanic, probably Gothic). Are there reflexes of the PIE in other branches? — Kleio (t · c) 15:50, 3 May 2017 (UTC)

Most likely I would have gotten it from Koebler, who draws on multiple other sources. The current Etymology being displayed, however, was added by User:Angr. Perhaps they might be able to shed more light Leasnam (talk) 16:27, 3 May 2017 (UTC)

I think I got it from {{R:cel:Matasovic 2009}}, but I'm not sure. I'll check when I get home if I remember. —Aɴɢʀ (talk) 16:32, 3 May 2017 (UTC)

If it's in Matasovic, please share your findings. That's one of the Leiden dictionary series my university library doesn't have, which frustrates me a bit. — Kleio (t · c) 17:13, 3 May 2017 (UTC)

It is extremely unlikely to be a loan from Celtic, because you'd have to explain how the regular Celtic zero-grade *dlig- became the regular Germanic zero-grade *dulg-. It would only make sense as a pre-Celtic loan to pre-Germanic, *dʰl̥gʰ-, but at that point there's really no hard evidence either way whether it's a loan or cognate, especially considering they don't even remotely share the same nominal formation (Celtic *dligetos, Germanic *dulgaz). Anglom (talk) 16:43, 3 May 2017 (UTC)

Good point. I'm also not sure why the Slavic word has to be a borrowing from Germanic rather than a cognate. —Aɴɢʀ (talk) 16:57, 3 May 2017 (UTC)

This claim too is all over the literature, though I'm not sure why either. In any case it's more similar to the Germanic word than the Celtic. So far I've not been able to borrow the monograph by Pronk-Tiethoff I mentioned in my original post above (only Google Books previews so far, which sucks) -- but when I do (soon, via interlibrary loans from another uni..) I hope it'll have some answers. — Kleio (t · c) 17:13, 3 May 2017 (UTC)

Lehmann 1986 (citing Scardigli 1973), Pronk-Tiethoff 2013, Köbler 2014 are the ones I've found so far that either claim a loan from Celtic or regard it to be plausible which is why I was curious about the etymology here; personally I don't know enough about early developments of Celtic especially to argue for or against what you just said. In the sources I mentioned, the lack of other Germanic cognates and the existence of Celtic-to-Germanic loanword traffic in the field of legal/military terms is brought as an argument for borrowing from Celtic. Furthermore, the lack of Indo-European cognates further afield would suggest an ultimate non-Indo-European origin. — Kleio (t · c) 17:13, 3 May 2017 (UTC)

I just checked Matasović, and he makes no claim at all about its being a loanword or not a loanword. Obviously we can't rule out the possibility of its being a loanword in some direction or other, or its being non-Indo-European but (1) there are plenty of words that are found in only three branches of Indo-European, and (2) the root dʰelgʰ-, of which this is the zero grade, just looks eminently like a PIE root. Usually when a reconstructed form is called non-IE, it's because it looks non-IE somehow, e.g. by having the vowel a or the consonant b, or by having too many vowels (hemp from *kanabi- is a good example where all three factors come together). But not here: you couldn't ask for a better formed PIE root! —Aɴɢʀ (talk) 18:33, 3 May 2017 (UTC)