Richard Wordingham:
> --- In cybalist@yahoogroups.com, "Joao S. Lopes" <josimo70@...> wrote:
>
>> If this scenario is true, Satemization appeared in different lines:
>> Balto-Slavic, Albanian, Armenian and Indo-Iranian. Is it convergent
>> evolution, or consequence of a common isogloss?
>
> Horizontal transfer seems to be the best explanation.
>
> As far as I can work out from Atkinson's expansion clip, Balto-Slavic
> and Indo-Iranian are almost at opposite extremes of the IE range,
> rather than neighbours. This looks like a mistake. Perhaps they can
> rescue it by making movement across the steppes easier.
This borders on something that's been bothering me.
From what I gather, their model implies (or produces the result) that
the branches pretty much took the shortest route to where they are
today. E.g., the maximum density Urheimat of IIr is in Eastern Turkey,
and that of Greek in the Aegaean, both in blatant contradiction to what
we think we know from archaeology and ancient myths.
They also say their model is robust to changes in variables, also
differences in ease of movement, and that maximum probability density
for point of departure is in Anatolia anyway. But I don't understand how
that can be. Like a lightning overwhelmingly follows the best conductor,
so will the most probable path of migration follow the fastest trail. I
simply can't understand how ease of movement along the steppe can fail
to funnel migration.
But maybe it does, and it's only the Urheimat that's virtually
invariant. If so, that could be an artefact of the early date for the
splitting off of the Anatolian branch, making the paths swirl aimlessly
around in the Caucasus for Millennia before hitting the steppe.
Anyway, I'd like to see results for paths, not only point of departure.
--
Trond Engen