Re: Schwebeablaut (was Re: tolkien?)

Hallo!

On Sun, 14 Dec 2003 20:06:41 +0100,
Andreas Johansson <andjo@...> wrote:

> Quoting Jörg Rhiemeier <joerg_rhiemeier@...>:
>
> > [...]
> >
> > These are indeed quite common in PIE, common enough to have a terminus
> > technicus for this phenomenon: it is called Schwebeablaut.
> > The origin might have been a difference in accent position:
> >
> > *CáRaC > *CeRC
> > *CaRáC > *CReC
> >
> > I am planning to use Schwebeablaut in my IE-related conlang family
> > "Hesperic", though I don't know yet what exactly to do with it.
>
> What's the reason to reconstruct with a's rather than as CéreC and CeréC?
In pre-ablaut pre-PIE, there were only three vowels, namely *i, *u
and a third vowel which at that stage must have been *a, for typological
reasons.  Then ablaut came into play, under the terms of which
each of the three vowels had a strong and a weak grade:

    Strong    Weak

*a  *á > *e   *a > 0/*@ > 0/*o
*i  *í > *ei  *i
*u  *ú > *eu  *u

...resulting in the familiar ablaut patterns.

The o-grade, which originally occured where a deletion of *a
would have left behind an inadmissible consonant cluster, was later
reinterpreted as a separate grade from the zero grade.
Further cases of *o and of *a then resulted from recolouring of *e
under the influence of laryngeals: *h2e > *h2a, *h3e > *h3o.
There are a few cases of *a and *o that remain unexplained, though.

So I could as well have written the pre-ablaut forms with *e's,
but the use of *a is justified by typological reasons.

Greetings,

Jörg.