Re: [pieml] Labiovelars versus Palatals + Labiovelar Approximant

From: tgpedersen
Message: 61190
Date: 2008-11-01

> > We're talking different things. Of course PIE -we- gives -u- in
> > zero grade morphologically. I'm saying that unstressed -we- or
> > -wo- -> -u- doesn't make much sense phonetically. I'd like to see
> > an example of it elsewhere.
> >
>
> Perhaps it's more like Spanish, where /u/ > /w/ when paired with a
> vowel (unless marked otherwise).

Even after accepting that, we haven't explained the 'pairing' thing -
where does that extra vowel comes from? The problem with trying to
explain a language with ablaut type phenomena from a stage which
hasn't is that once you reduce the wide bandwidth that the ablauting
vowel took up to one vowel, the vowel system of that earlier stage
becomes extremely sparse. Assuming that some earlier vowels were
sucked up into the ablaut system by analogy might make it less so.

> Or perhaps, it really is /k^wn-/ and /swp-no/

And what does that mean?


Torsten