Richard Wordingham wrote:

> ---
>
> Sorry, I didn't read carefully enough. I was skimming because I hope
> to look throught these lists in more detail later.
>
> Germanic (as a group or proto-language) has next to nothing to do
> with the relationship of English nephew and Latin nepos, nepot-. The
> connecting route is English nephew < (Old) French neveu < Latin
> nepotem. (I have missed out some intermediate stages because I don't
> have my reference books to hand.) The intervocal Latin /t/ was, as
> usual, ultimately dropped in the development of the Old French form.
> English then borrowed the Old French word, the English spelling was
> modified to better match the Latin original, and in many people's
> speech the pronunciation has followed. The English-French-Latin
> correspondences in this word are n-n-n, e-e-e, ph-v-p, ew-eu-o, 0-0-
> t, 0-0-em. No w-t required! Kondrak looked at algorithms which can
> achieve this level of matching in the thesis to which you posted a
> reference at this group, and notes the correct matching in the
> relationship of English wolf and Latin lupus (i.e. l-l and f-p).
>
> The nepot- stem is PIE, which I think is the point you wanted to
> make. I can't say I like the suggested Nostratic *k > PIE *p
> correspondence, though I can dimly see how it might just be a
> conditioned change spread by analogy. There are some German forms in
> Pokorny one could be tempted to reach down to, but that would be
> demonstrably wrong.
>
> Richard.

Ok, I see what you mean now. When comparing forms without looking into
history
one can easily make mistakes. The thing is that I think the errors are
symmetrically
distributed (e.g. like the Gaussian) so that given enough data they will
cancel out.
Of course, one can also make mistake tracing words across time. The more
that is
done the more chances of errors creeping in.

In general I think for me the best thing is to go back to the originals,
Hittite,
Akkadian, Sumerian, Sanskrit, etc. After all, I do believe in something
like
Nostratic and that I think is the best hope of recovering anything.

>

--
Mark Hubey
hubeyh@...
http://www.csam.montclair.edu/~hubey