--- In
cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:
>
>
> Fair enough.
>
> PIE *d, Etr t(H)
> *sed to sit, satH to sit
> *den day, Tin diety associated with the day
> *deiw god, heaven , Tivr moon
> *dem to build, t(H)am to build
>
> ***
> these three are very doubtful:
Then the above 4 are acceptable?
> *deH3 to give, tur to give
> *deH2 to divide, itu to divide
> *demH to tame, tamna horse
> ***
What's doubtful about them?
> PIE eu/u, Etr u
> *H1neun nine, nurpH nine
> *teuteH2 people, tut(H)i community, state
> *dHeu to pass away, lup to have lived, to die
> *bHeu to grow, to be, amu to be
> *yeu youth, hus child
>
> ***
> here, only <tut(H)i> has a chance of passing the Limburger test
('smelling
> less bad than ...').
My reconstructions for each of the above are:
*?nu(G, m, n) nine The reconstruction in PIE is complicated. It's
usually reconstructed with a final *n but some have argued for a *m. t
might have been variable. The Greek from looks compatible with a final
*H2. IMO final *H3 merged with final *H2, which leaves open the
possibility of a final *H3. As we see in the root for "to give"
non-initial PIE *H3 corresponds with Etr r. The pH in Etr is
analogical from sempH.
*tuti people, community
*dlu to die The origin of the Etr p is unclear, however the dH~l
correspondence is also likely found in the PIE ablative *-d (< *dH#)
Etr genitive -la.
*amb&w to be, to grow (this one was used in error since the eu~u
correspondance has a different origin) Loss of initial /a/ is regular
in PIE as is the loss of /&/ in Etr. The *mb is a prenasalized stop.
*yus' child The loss of the ejective spirant is regular in PIE as is
its merger with s in Etr (4 other cognates exist). I'll grant that
this is the only example of *y~h. I've postulated this as coming from
PIT *y, but there is another possibility. It may come from PIT *xW (>
h in Etr) with an irregular PIE development of *xW > *xJ by
dissimilation > *y.
> ***
>
> GBS would have approved as, for him, <ghoti> = <fish>.
Have you been enslaved by Greenberg and Ruhlen's look-alike methodology?
> ***
>
> But, let us say that PIE *du (zero-grade) = Etruscan <t(h)u>
formally, which
> I do not belive.
>
> What evidence do you have that *du ever meant 'one' or anything like it?
Because PIE *dwo: is a dual formation. A dual of what? If we undualify
two we get one. Two ones are two. 1+1=2, does it not?