Re: Family terms [was: Kluge's Law in Italic?]

From: Tavi
Message: 68562
Date: 2012-02-14

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:
>
> > Are you saying that 'brother', etc., all happened to have a
laryngeal
> > each but 'mother' didn't? Analyzing the common ending as *-ter- not
> > -xter- leads only to baseless folk etymology. Even if from babbling,
> > such a *ma- could have been old enough to undergo a>e, back to a
only by
> > the following x (even if not so old, it causes the lengthening seen
in
> > historical IE).
>
> I'm open to any explanation that makes sense od the peculiarities of
the
> PIE family terms. 'Father' can hardly have been originally segmented
as
> *p-h2ter- (even if it should have been resegmented in this way later
> on), and I find *ph2-ter- (a transparently formed agent noun) easier
to
> swallow than *ph2t-er-. Of course 'mother' can be *ma(:)-h2ter-, as
far
> as I'm concerned (with *-h2ter- taken from 'father'); all that I'm
> saying is that *h2 is unlikely to be part of the "baby talk" element
so
> comon in 'mother' words the wide world over.
>
That's right, because *h2 is actually a conventional symbol of "IE-ists
talk" and not a real sound. Incidentally, Mallory & Adams reconstruct
*pºh2té:r, with a "vocalic" consonant and a long /e:/.

I also haven't seen any explanation of why this *h2 gives /a/ in Latin
pater but /i/ in Iu:piter, Iuppiter like in Indic.