Re: [tied] Terminology (Re: Piotr-)

01-09-03 18:50, alex wrote:

> Nope , But you will recognise in Gree the "ipp" there. In Thracian we
> have -esb-, -asb-. I very doubt if one will find in Greek a name like
> *Leukappos and *Happarchos this one will think anymore at the
> "ippo" -relationship.

But Greek is Greek and Thracian is Thracian. Greek forms vary from
<hippo-> to <-ipp-> (not to mention dialectal <ikko->) for reasons that
can be explained. Thracian <a> may be due to vowel contraction in
compounds, e.g. *kaka-esba- > kaka:sba-.

>> Skt. as'va-; Av. aspa-; OLith. as^va ~ es^va 'mare', as^vis 'foal';
>> OPr. aswinan 'horse milk'. I wonder why you chose just these
>> languages. Reflexes of *h1ek^wos are found in many other groups
>
> I wonder how you wonder. They are the satem forms of the IE root. Why
> should I mention the centum forms here? Comparation should be done
> within the same group in this case, imho.

What was the point of your listing all those Satem forms? The people
potentially interested in this discussion know them better than you do.

>> If, however, the
>> name _is_ Dacian, the *g^ in *h2arg^- would probably have become *dz
>> (related to Alb. <dh>). This is possibly what lurks beneath ancient
>> Ordessos (for *ardzes^- < h2arg^-es-jo-), in which Romanian
>> substituted /dZ/ for /dz/, perhaps through distant assimilation to
>> /S/.
>
> Seems you are a bit wrong here. First of all, so far I know the
> generally aception is as follow:
> Alb. /dh/ is a result from PIE *d, *dh ( after /r/, Pekmezi, Gr. Alb.
> Spr. 32)
> The coresponding sound of Alb. /dh/ is not Rom. /dz/ or /z/ but simply
> /d/
> Alb-Rom: hurdhë - leurdã; shkardhë - zgardã

Partly inaccurate, and completely irrelevant. We are not talking about
_these_ <dh>'s.

> The Alb. /d/ should be the result of PIE *g^ and *g^h ( Pekmezi, Gr.
> Alb. Spr. 29)
> And for this example I shoued by myself the PIE *g^her- which gave in
> Alb. "dorë" and in Rom. "ghearã".

??? -- The PIE word was actually *g^Hesr. (check the archive), and
<ghearã> has nothing to do with it.

Thus, the Albanian form of

> '*h2arg^-es-jo-' should have been '*ardes-'. No /s/, no /z/ but simply
> "d".

/d/ _and_ /ð/ (<dh>), actually, and the latter is regular after /r/
(I've written about this before!). In a similar context we have Alb.
bardhë 'white' < *bHr.h2g^-. The substratal cognate in Romanian is
<barzã> -- you _do_ see the <z>, right? We normally find Rom /z/
(dialectal /dz/) in substratal words that have Albanian cognates with
/d/~/ð/ from *g^(H). I take the affricate *dz (partly surviving in
Romanian) to have been an intermediate stage between PIE *g^(H) and
Mod.Alb. /d/~/ð/.

Piotr