On 2006-09-16 21:42, tgpedersen wrote:
> All I have accessible now is Gyldendals Dansk Etymologisk Ordbog
> which only has 'glda. ofn'. I'll check it up, when I get the
> opportunity. BTW my shibboleth-explanation can't account for Got.
> aúhns, which I had forgotten about.
It's been suggested that *x > *f between a back vowel and a nasal should
be regarded as a regular Northwest Germanic innovation. I believe Voyles
gives a couple of examples, but unfortunately the only one I remember
is... *uxna- > *ufna- (I'd have to check up the other(s)). If there's
any truth in it, though, the reconstruction doesn't have to involve a
labiovelar or a pre-Germanic variation of *kW ~ *p. PGmc. *uxna-/*uGna-
is all we need (*ufna- being a regular post-PGmc. dialectal
development), and something like *h2aukh2- (without an ambarassing
labiovelar after a u-diphthong) can underlie the Germanic words as well
as Lat. aula/olla, auxilla, and Ved. ukHá:, ukHá-.
Piotr