From: Piotr Gasiorowski
Message: 10446
Date: 2001-10-19
--- In cybalist@..., george knysh <gknysh@...> wrote:
*****GK: But surely the proto-Hittites etc. should have
known [the horse] in their pre-Anatolian habitat.*****
Not so surely. If they came from the Middle Danube basin (as
I propose) and spent some time in the Balkans before
migrating to Asia Major, they wouldn't have had much chance
to see wild horses, and would have had to wait for someone
to show them domesticated ones. Even if originally living
close to the tarpan's range (if we speculate that *h1ek^wos
is really PIE), the Proto-Anatolians may have lost the word
as they moved south into the Balkans, where _Equus ferus_
did not occur in PIE times.
*****GK: This is where I have a problem. I take it that the
theory of the very early "Anatolian"
branch-off must be based on something more solid than
lexical matters. But would these other elements (morphology,
syntax, etc..) be sufficiently compelling evidence for such
a major split without the assumed lexical points? A
situation where "An." has a term for X missing from "non-An"
could easily be explained as a foreign borrowing in "An."
given the known cultural context. I realize that each case
should be analyzed separately but on balance the probability
seems higher that it is "An." which has suffered substantial
lexical losses from IE and gains from non-IE . Again,
because of the geography. BTW is there any information which
can be derived from the extant common vocabulary of "An" and
"non-An" which might be helpful in the matter of determining
the time line of the split? Or is this too tenuous or
suspect (like the "horse" word issue?)******
In terms of grammatical structure, the Anatolian branch is
_very_ different from the rest of IE -- enough to justify
the theory that the differences are at least partly due to
shared innovations in the non-Anatolian part of the family.
Even "stray sheep" like Armenian and Tocharian, in which a
non-IE substrate is equally evident and which were spoken in
relative isolation from the bulk of IE, don't diverge from
the "classic" IE type to a comparable degree.
Piotr