Of Trees, nodes, and minimal paths (was Re: Urheimat in

[Hans Holm]

It's indeed crucial: What really are 'certain shared innovations'? They
are never 'certain'. Even with perfect sound correspondences these may be
just chance. They of cource then would have a high probability of not
being just chance.

>the UPenn tree being discussed was entirely built on (in form at least)
>a narrowed sample of 'data'

.. I again must insist:
 We do not have the data for the UPenn tree;
 and therefore we cannot discuss anything!

Regards
Hans J. Holm
D-30629 Hannover