Dating the final IE unity

[Hans Holm]

RC>very closely related" is extremely misleading. They could be considered
RC>closely related only by contrast to the highly diverse (lexically and
RC>typologically) other families of the north and west of Australia.
RC>Consider just the immediate neighbours of Dyirbal, as described by
RC>Dixon: Yidin (27% shared vocabulary), Mbabaram (18%), Warungu (47%),
RC>Wargamay (60%)

.. misleading are such percentages, if taken as representing proportional
genealogical relationship. Direct genealogical relationship means that two
daughter-languages directly stem from the same mother language.
Replacements, even severe and/or quite different ones, taking place after
that split, have no influence on the notice. Thus, two languages with a
smaller number of retained common lexemes may be related closer than
others with higher percentages. I shall try to make this clearer in the
near future.

(BTW, you signed with my Footer.)

Mit freundlichen Grüßen
Hans J. Holm, Meckauerweg 18, D-30629 Hannover
 Tel=FAX x49-511-9585714.