multiple "ancestors"

[Ante Aikio]

I would like to make a couple of remarks on Steve Long's claim that a
language could have multiple genetic ancestors, perhaps from a slightly
different perspective than the objections already brought up by Larry
Trask and others.

It should be noted that the assumption of multiple genetic ancestors
actually invalidates the most central concepts of historical linguistics
(e.g. language family, descent, convergence, divergence, borrowing, family
tree, Indo-European, etc.). These terms are defined in the context of the
comparative method and the assumption of a single parent. They are quite
simply meaningless, if the single parent asumption is *not* accepted - in
which case one would either have to redifene the terms or not to use them
at all. Consequently, the acceptance of the concept of "multiple genetic
ancestors" would immediately invalidate *all* established results ever
obtained in the field of historical linguisics.

As for the assumption of a single genetic ancestor (to which there are
some rare exceptions, to which the standard tools of historical
linguistics, especially the comparative method, are not applicable), it
should be fairly obvious that it is the only reasonable starting point.
This assumption is based on direct observations of language acquisition
(we do not observe the development of mixed idiolects in cases where a
child receives input from more than one language) and stability of
language over time - it is extremely rare for a langauge to change so
rapidly that the continuity between it and its ancestor would clearly be
disrupted. (The stability is of course merely an effect of how language
acquisition works.)

Regards,
Ante Aikio