On the general acceptance of Austric

[David Stampe]

----------------------------Original message----------------------------
Summary: Austroasiatic yes, Austric not yet.
 
No one I know doubts the Munda--Mon-Khmer relationship (Austroasiatic).
A sizable percentage of their respective proto vocabularies is cognate,
and Donegan and I have showed how the polysynthetic structure of Munda
evolved from (and preserved much evidence of) the analytic structure of
proto-Austroasiatic (which remains in Mon-Khmer) [1].  In the century
since Schmidt published evidence for Aa, new data on Munda and MK have
considerably strengthened the case.
 
This is NOT so for Schmidt's hypothesis that Austroasiatic and Austro-
nesian are related (the Austric hypothesis): as we have learned more
about the languages and their histories, resemblances have had to be
discarded, and although new ones have appeared, after a century the
strength of the case for Austric has not increased.  We now have far
better documentation of South Munda, which, despite its distance, has
more cognates with Mon-Khmer than North Munda does.  But resemblances
between Aa and An cluster in South MK languages, which have had close
contacts with An languages since proto-Aa times.  According to the
Austric hypothesis, Munda and MK are equidistant from An.  But if only
MK resembles An, that is a problem for the Austric hypothesis.
 
For example, it once seemed impressive that Munda and Nicobarese, the
extreme ends of Aa, have suffixes, like An.  But if the suffixes are
not even similar between Munda and Nicobarese, much less between Munda
and An, they are not evidence for Austric.
 
Even the one resemblance that everyone agrees is most striking -- that
both Munda and An infix VC- prefixes in C-initial roots -- probably has
a functional basis: it keeps the prefix from forming a closed syllable
which could attract the accent away from the root [1]; Sora (S. Munda)
uses such pre/infixes as @n- 'nominalizer' or @r- 'instrumentalizer,
locativizer' productively in deverbal derivation, but VC- prefixes are
NOT infixed where they represent inflectional categories that naturally
take the accent, e.g. @r- 'reciprocal' or @d- 'negative' (which takes
the form @dn- before V-initial verbs, as if to insure that it does take
the accent).  Although a functional basis does not mean that infixation
could not have had a common origin in Aa and An, it weakens a case that
from the beginning has rested on similar morphological processes rather
than on cognate morphemes.
 
David <stampe at hawaii.edu>
 
 
[1] Patricia Donegan and David Stampe.  1983.  Rhythm and the holistic
organization of language structure.  In: John F. Richardson, Mitchell
Marks, and Amy Chukerman, eds.  Chicago: Chicago Linguistic Society.
337-351.