Brian:
> Are you claiming that <mice> is analogical? It's completely
> regular: the PGmc nom.pl. is *mu:siz, yielding OE my:s by
> I-umlaut.
Two-to-one. Even Jens just said that [ab/um]laut is analogical.
So yeah. At any rate, the point is that it isn't IE ablaut
even though it looks like it and even though IE ablaut
was the basis for more recent processes like this I-umlaut.
This is what can be deceiving when working a language
back in time.
At any rate, it looks like you're more interested in
nitpicking offtopic details rather than follow the original
point now. Who cares?
= gLeN