Re: Spirantized puzzle

> > >
> > > > This is strange:
> > > >
> > > > Skt. Nsg pitá, Ds pitré (< *pitréi)
> > > >
> > > > Av. Nsg ptá:, Ds fðrái (< *pitréi)
>
> > > It's a result of the r:
>
> > You didn't get it. The puzzle was why Skt. looks like a
> > generalization from Avestan Nsg, and PGerm like a generalization
> > from Avestan Dsg.
>
> The Avestan paradigm is the result of Avestan rules.

This can be read either as a truism or as an unreasoned rejection of
what I proposed.

> > I'm arguing that PIE paradigms alternated like Avestan and that
> > various daughter language arrived at their paradigms by various
> > generalizations of those paradigm. If that is so, Grimm and Verner
> > was something that happened inside PIE, not in PGerm.
>
> Not all words have paradigms;

No, but the substitutions generated by generalizing paradigms spread
to words which do not have paradigms
http://www.angelfire.com/rant/tgpedersen/Shibbolethisation.html

> not all paradigms have forms that put all consonants next to other
> consonants.

Those nouns that have the ablaut vowel e/o/zero (<- PPIE /a/) do.
Those that have -i-(-ei-) and -u-(-eu-) will traditionally have -i-
and -u- in zero grade. If we instead posit their consonantal forms,
-y- and -w- we can treat them as semivowels, which would separate the
unvoiced initial set of sprirantized stops from the subsequent voiced
set. And if we posit zero grade for the ablaut vowel instead as some
'consonantal schwa' &, we get PIE *f&ðrí -> PGerm. *fáðri.


Torsten