Re: [tied] Scythians, Zoroastrians, etc.

----- Original Message -----
From: "george knysh" <gknysh@...>
To: <cybalist@yahoogroups.com>
Sent: Tuesday, February 26, 2002 7:34 PM
Subject: Re: [tied] Scythians, Zoroastrians, etc.


> *****GK: "Alizones" is certainly not easy to "crack".
> If direct affinities to Indic seem difficult because
> of the considerations enumerated above, then perhaps
> other solutions might be entertained.

Of course they might.

> In any case,
> "unlikely" or "highly unlikely" is not the same as
> "impossible" or "absolutely impossible".

Well, if I had said "absolutely impossible", you could
rightly object that nothing is ever absolutely impossible,
even alien abductions and faster-than-light travel. I would
bet my bottom zloty that Skt. s'ana- and (Graeco-)Scythian
<-zon-> cannot be the same, but I wouldn't be equally
adamant about my own etymological proposal. I'm sure
plausible alternatives can be offered; I just can't think of
any. BTW, the analysis of *-zana- <-zon-> (pre-/n/ rounding
is common in Iranian) as *g^onh1-o- (and so cognate to Gk.
gonos and Skt. jana-) gives us more etymologies, e.g.
<ama-zon-> with <ama-> = Avestan ama- 'strength, courage,
virtus militaris'.

> I have some
> doubts about the Iranic solution because it seems to
> stretch the "lambdacizing" explanation a bit too much
> (now on the first "r" instead of the second as in
> "parala-") and very early too.

No man knows for sure where and how long ago lambdacism
began to spread; we only know that it was already there in
Old (Alanic) Ossetic, and that it allows one to propose
quite plausible Scythian etymologies (including that of the
Paralatae). For all I know, *ry > *l may be an old process
in the Pontic steppe dialects. There is only one *r in
(hypothetical) *aryazana-, so the Ossetic-type lateral
dissimilation I posited in *pari-a(:)rya- > *palala- >
*parala- cannot apply.

> I noticed that Duridanov's Thracian plays
> around a little with its "z" and "s" (e.g. EZVAS for
> Indic AS'VAH; ZURA for SIRA ('current') PIZAS for the
> 's' in Greek and Lithuanian 'bog, meadow' etc..== So
> in spite of your explanations I can't quite rule out
> some Thrakoid expression having changed the "s'" to a
> "z" which then gave us the Greek. The diphtong issue
> would not be a major problem. Lots of examples of
> variety in Duridanov. Well, nothing certain here I
> guess.*****

The only acceptable (in fact, certain) etymology among the
above is <ezba-> 'horse' (also with variants like <esba->,
<espa-> frm *ek^wo-. Here, however, we have the Thracian
change of *w > b (perhaps pronounced [v]), plus voicing
assimilation in the sequence <sb> (or [sv]), with <s>
realised as [z]. In other contexts PIE *k^ becomes Thracian
<s>, not <z>.

Piotr