Hallo!

On Sun, 6 Dec 2009 21:37:18 +0100, I wrote:

> Hallo!
> 
> On Sun, 6 Dec 2009 17:03:21 +0000, David McCann wrote:
> 
> > One could explain PIE from something like
> > soft: u, ə, i > u, e, i
> > hard: o, a, e > o/a, o/a, e
> 
> No.
> 
> These alternations simply *do not occur* in Indo-European, at least
> not systematically.  One would expect i~e and u~o alternations in
> the same function as e~o ablaut, and that is not the case.  There
> are only few pairs of forms where these "alternations" occur, and
> there is no reason that those have anything to do with PIE ablaut.

An afterthought:

Any attempt to link PIE ablaut to whatever kind of vowel harmony system
is foiled by the behaviour of the alternants in PIE.  In a vowel harmony
system, all morphemes in a word form appear in the same grade (e.g., all
palatal, all "soft", whatever) - otherwise it wouldn't be vowel *harmony*.
In PIE ablaut, usually *one* morpheme is in e-grade, and all others in
o-grade or zero grade.  This does not look like a vowel harmony system
at all, but rather like something conditioned by some kind of accent.

... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html