25-11-03 13:11, tgpedersen wrote:
>> It's pure speculation,
>> of course, but I prefer this kind of speculation to proposing
>> Proto-World *kuan- a la Merritt Ruhlen.
>>
>
> But what's to become of Ruhlen's other "dog" words if you orphan
> them? Have you even considered that? Tsk, tsk. ;-)
We've been through that. The method consists in selecting a number of
roughly similar words from random languages (sometimes incorrectly
quoted -- see the discussion of the Old Chinese item some time ago),
with referents ranging from 'dog' to 'hyaena', without even attempting
to establish systematic correspondences or to analyse the individual
etyma. Mysteriously, the impressionistic "reconstruction" resembles the
PIE word more than anything else (by the same token, Ruhlen's
Proto-World *akwa means ... [answers on postcards]). I disagree with
Ruhlen, because by comparing English dog 'dog' with Mbabaram dOg 'dog'
and Proto-Kartwelian dz^aG 'dog' I have already established beyond
reasonable doubt that the Proto-World word for 'dog' was *dog ... well,
perhaps *dag, because some details can be expected to change as time
goes by.
Piotr