>
> Analogy between "literate" American and "pre-literate" European
naming does
> not of course prove anything decisively, but it is evidence. And
this
> evidence does not favor universal, standard or official names for
rivers or
> seas (and perhaps other features) before writing. At best, this
kind of
> evidence suggests the earliest names preserved in writing would
seem to be
> coincidental and quite probably relatively recent.
>
You may be right there. The assumption that Germania was uniformly
Germanic-speaking and Gaul uniformly Celtic-speaking may be just
that, an assumption. Who knows which relic languages survived then,
to become extinct later? Some ancient writer, I forgot, says they
spoke 140 languages in the Caucasus. Mithridates is supposed to have
spoken 25 languages. Which are they? Can we identify them? Did any of
them influence today's languages? Is the Germanic 30% substrate
language one or several languages? Is there a megalithic component in
it (them)? "drink" is usually assigned to that substrate language.
However, Bomhard has a (approx, by memory) root *tr(k)- "drink" with
cognates in Kartvelian and Dravidia, but nothing similar for other
roots of Germanic substrate language.
> Regards,
> Steve Long
Torsten