--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...>
> 'Old European' is not a descendant, but a sister to PIE.
>
> So, like this:
> pre-PIE: *kän-/*kan-/*kn.-
> 'Old European': *kan-/*kn.- (merger of /ä/ and /a/
> its sister, PIE: *ken-/*kon-/*kn.-
So you are making "Old European" a close relative of PIE? Well, that
is a novel idea, I suppose. Is that really possible? Is Old European
not being drawn upon when vocabulary of a completely alien stock is
to be established? But I admit that, on its own premises, your
scenario is completely logical. On the other hand, one must be
allowed to ask the question: Are there other reasons to assume a
close relationship between OEur. and PIE than the prospect of
getting rid of a-roots and their ablaut in IE? That strikes me as a
particularly unsatisfactory basis for a genetic classification.
Jens