Richard Wordingham wrote:

> --- In Nostratica@yahoogroups.com, "H.M. Hubey" <hubeyh@...> wrote:
> > > IV. Sumerian r/ vs Turkic z/
> > Despite PA "to dig" being reconstructed as *karu/*karo, and despite
> PU
> > also being
> > constructed similary, there are still "well-known" Turkologists who
> > still try to pass of the
> > lark, that PT was sh-z. That would imply that *kar came from *kaz
> and
> > that words like
> > karim /garim (canal e.g. digged thing) should have come from *kazim
> > despite total
> > lack of evidence and being totally against laws of linguistics.
>
> Could you expand on that, please. You seem to be saying that z > r
> is inherently impossible, but it is assumed for Latin (final step in
> change of intervocalic PIE *s to /r/) and North and West Germanic
> (conditional PIE *s > Proto-Germanic *z by Werner's law, and then *z
> > /r/). In the case of North Germanic, this *z appears in runic
> inscriptions, where it is distinct from /s/ and /r/. (Gothic merged
> *z with *s.) As far as I can make out from the phonological sketch
> of Mak at the Rosetta Project, the change [z] > [r] is in progress in
> Mak.

I think r is more archaic. Secondly in some cases it seems like z came
from something like
rs or rsh. Third, I do not know enough about Germanic to know exactly
what is known
and what is inferred. All the books I read force you to spend your life
collecting journal
articles. Instead of pointing out why X is posited and showing the
evidence, they ignore
the evidence and instead decide to spend chapters on glottalics,
laryngeals etc and
the data gets lost in the shuffle. If dialects have r and z and one of
them takes over it does
not mean that there was really a change r>z or z>r. Specifically in this
case, I pointed out that
if Sumerian has r 5,000 years ago why would I believe that Turkic had z
3,000 years ago
and changed to r 500 years ago?

In general I do not believe much of what I read. I am still not sure
enough to insist on
what Nostratic looked like, or even PIE looked like, but from evidence
that I have from
Hittite, Turkic, Akkadian, and some of the modern languages words I
know, here is a great
possibility.

I think there is plenty of evidence for this:

*th > {l,s/sh,w} u {t} u { dh}
*dh > {r,z,y} u {d} u {ng}

or something like it. I like it for several reasons. One of them is that
it obeys some
basic principles. The other is that it obeys Occam's razor. Third is
that it is symmetric
and thus beautiful. Fourth is that here are some of the things it does;

There were consonant clusters of the type -nth-, -rth-, -lth-. Now we
know that
liquids tend to disappear. Ditto for nasals like n. With a sound change
like th> s or
th>sh, we can get -rth-> rs > z or with th>0 we can get rth>r. In
short I have an
Occam's razor description of many phenomena including some of the PIE
sound changes that "need" things like laryngeal vowel-colorings and the
anomolous
changes like r>z or z>r etc.

There is more but I think this is enough for now. I posted much material
on this
on other mailing lists. And finally, I have reconstructed Turkic in ways
that have
finally shown some sense and they can be found on other mailing lists
e.g. turkoloji, and
even more to the point, I have words in Turkic which still retain
consonants that were
lost in Akkadian. Nobody can claim that these were added by Turkic
speakers because
they could not handle vowels in initial or final positions. In my view,
that puts the final nails
on the coffin.

>
>
> Do Proto-Turkic reconstructions work with /r/ in place of /z/?

I think the z>r changes are the worst possible explanations for Turkic.


--
Mark Hubey
hubeyh@...
http://www.csam.montclair.edu/~hubey