LINGUIST List 9.924

Sat Jun 20 1998

Disc: State of Comparative Linguistics

Editor for this issue: Martin Jacobsen <martylinguistlist.org>

Directory

1. nacal26, Reconstructability

	I am not a member of any networks, preferring to spend my
retirement years in work on Nilo-Saharan and Omotic to chit-chat, but
I occasionally see something which has appeared on the network(s).

	I just saw something from Alexis Manaster-Ramer which refers,
I suppose, to my 1973 paper in "Language Sciences" on why one cannot
reconstruct "Proto-Human". The comparison of the argument therein to
Goedel or Turing is not apposite. The latter are dealing with
axiomatic systems having no necessary analogs in the physical
world. If there is any connection, it would be in showing that human
language has an underlying abstract code which is subject to Goedel's
results in some way- I tried this years ago but gave it up. I'm just
no longer mathematician enough to handle this sort of stuff (maybe
never was).

	A better comparison might be archeology: from the physical
assemblages, one can arive at limited conjectures about such things as
prehistoric shelters, diets, etc., much less so about ideology. The
names of individuals,even of leaders, their personality quirks, the
particular incidents of daily life, etc. These are not reconstructable
by archeological methods unless so-far-unknown written records or
other such recordings are discovered. So, too, we can make
speculations about the nature of prehistoric languages before the
comparative method (a probability-based method!) allows us to
reconstruct with even a low level of certainty. But these will be
general and vague: not specific morphemes.

	Unless some alien species contacts us with recordings they
made in ancient contacts, it is hard to see how we can compensate for
millenia of probabilistic change in a system whose basis includes a
high degree of arbitrariness.

Lionel, June 3