PIE Ablaut [was] Re: Gypsies again

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:

> ----- Original Message -----
> From: "Miguel Carrasquer" <mcv@...>

> > >The example you gave to prove your point: cakara/caka:ra has a long,
> > >checkered history of disputation. My best guess at present is

that the

> > >lengthened vowel was simply introduced to provide a means of
> > >differentiating
> > >1st and 3rd persons.
> >
> > And this happened only in open syllables?
>
> Patrick:
>
> But, as for the example you gave, I find it most unconvincing. Richard
> mentioned that caká:ra was also in use for a 1st p. form and he also

pointed

> out that 1st p. would be more likely to be marked than 3rd p. I

would say

> that constitutes a medium argument for regarding the <a:> as vRddhi,

and

> purely for the purpose of differentiation.

To differentiate what? On the analogy of the CVCC roots, one would
expect _cakara_ for 1s (as occurs) and 3s (as does not occur). You
seem to be arguing that _caka:ra_ was substituted for *cakara in the
3s, to distinguish it from the 1s, but that the 1s then came to copy
the 3s, so the attempt at distinction started to fail. Do I
understand you correctly?

Brugmann's law also operates on the vowel of the 1st person present
indicative active. While the pre-merger form of the 1s is not obvious
to me - *-omi or *-o:mi? - it explains the form of the thematic vowel
in 1du -a:vas and 1pl -a:mas. Note also the secondary thematic 1p
endings: -am, -a:va, -a:ma. It it not a case of blanket lengthening
in the first person.

> Also, I am not convinced that the lost laryngeal could effectively
> close the syllable of the 1st p.

Lost laryngeals can also make syllables long for the purposes of
scansion in the Vedas. However, I'm not sure what to make of the
presence of the 1s form _caka:ra_. Is it analogy from the CVCC roots,
where 1s is the same as the 3s, or is it a case of consonant plus
laryngeal not always causing PIE *o to have short reflex.

> Patrick:
>
> I believe that during the pre-PIE (my Pontic) stage, PN *ke had

become p-PIE

> *kYa so that it entered PIE as *k^A. I do not believe that this
> palatalization occurred as late as Proto-Anatolian; it was already

there in

> PIE. Satem and centum are just different responses to it.

PIE *k^ yields Sanskrit /s'/. Sanskrit /c/ derives from softened PIE
*k or *kW.

Richard.