Reconstruction talk:Proto-Celtic/ruskos

Page move

@Victar: since both Brythonic and Goidelic point to a form with long ū, why did you move this to a form with a short u? —Aɴɢʀ (talk) 11:26, 16 October 2016 (UTC)

Thanks for the interest. Well, Matasovic says that it might be long or short and the DIL actually thinks that the Irish form comes from Brythonic, which might explain the long u. The Welsh form looks to be from some derivative like *rūsklos, which Matasovic cites, but I have no clue how that plays in. An earlier *rukskos makes the most sense to me if you're trying to derive it from Proto-Indo-European *h₃rewk-, *h₃ruk- (“to dig up”) (Mallory 2006). --Victar (talk) 15:06, 16 October 2016 (UTC)

Matasovic says if it's a long ū, it might be actually from Proto-Indo-European *(H)rewH- (“to tear out, pluck”), which has nearly the same meaning, which would certainly explain it. If you think that's a better way to go. I'd be happy to change it back to that. I guess I'm just also thrown off by the Breton and Cornish forms that seem to point to a short u. --Victar (talk) 15:19, 16 October 2016 (UTC)

I just noticed the Breton words which I think suggest a short u (though I don't know enough about Breton historical phonology to be sure). At any rate, the entries' etymology sections still point to the version with the long ū, so I've restored the redirect. Maybe we should list the long ū variant as an alternative form, since it looks like Welsh and Irish come from *rūskos but Breton (apparently) comes from *ruskos. —Aɴɢʀ (talk) 15:30, 16 October 2016 (UTC)

I would love to know why the DIL calls it a British loan-word. Is there some additional information we're missing? --Victar (talk) 15:56, 16 October 2016 (UTC)

Also, do you know the reason for the -l- in the Welsh form? I see that also in Welsh trwsgl (“rash, crude, rude, coarse, clumsy”) from Proto-Celtic *trusskos and Welsh tresglen, Old Breton trascl, Cornish traskel (“thrush”) from Proto-Celtic *trezdkos. --Victar (talk) 16:05, 16 October 2016 (UTC)

DIL says it's a Brythonic loanword because Pedersen says it's one, but Pedersen doesn't explain why he thinks it's one. He just lists the Brythonic words and then writes "ir. [entlehnt] rúsc". The section where he mentions it (vol. 1, p. 332) addresses your second question: he claims that it's a kind of hypercorrection resulting from the loss of syllabic l in pronunciations like /ˈpɛrɨ̞ɡ/ for perygl and /ˈhɨ̞ðɨ̞ɡ/ for huddygl (“soot”) where the l is etymological. —Aɴɢʀ (talk) 16:25, 16 October 2016 (UTC)

Thanks! That's really great info. Here I was trying to reconstruct a *rūzdkos or something from Proto-Indo-European *(H)ruH-sdʰ-. =) --Victar (talk) 16:41, 16 October 2016 (UTC)