At present I'm investigating the roots proposed for the Spey in Scotland. It's attested as 'Spe' in the twelfth century, by which time the area had been Gaelic-speaking for some centuries.
A link with *skwijat- 'thorn' has been proposed but faces many challenges on both phonetic and semantic grounds. This might be less problematic were it cognate with the suffixless form apparently attested in OIr 'sce', but there are perhaps too many objections, with this rather
challenging river-name.
The 'p' more or less rules out a coining in Goidelic which had long lost this phoneme, hence one assumes that the name must be earlier either Brittonic/Pictish, 'Old European' or perhaps even non-Indo-European.
Being able to argue for **spue or ** spwe would tie in with another piece of evidence. The loss of /w/ could be ascribed to mediation by Goidelic which had lost its Celtic approximants.
However I can't find any convincing way of explaining the final vowels, and I'm not entirely satisfied with the initial cluster, and I seem to be down to choosing between the solution with fewest objections.
Any suggestions for suitable suffixes?
From: Bhrihskwobhloukstroy <bhrihstlobhrouzghdhroy@...>
To: cybalist@yahoogroups.com
Sent: Friday, January 11, 2013 10:58 PM
Subject: Re: [tied] PIE suffixes
Which one? (River-name)
2013/1/11, gwalstawd
gwalstawd@...>:
> I'm working with a rather problematic Celtic river-name.
>
> PIE *skw- would give sp- in P-Celtic.
>
> Would it be reasonable to propose that a zero grade of *sekw- (1 or 2) could
> also give sp-?
>
> Are there any PIE suffixes or extensions which could account for the vowels
> in a hypothetical *spue(:)?
>
> I can't find anything in the standard works or by googling.
>
> Thanks
>
> Guto
>
>
>
>