ablative as ergative marker

[Scott DeLancey]

On Thu, 22 Oct 1998, aldai wrote:

> Am I right if I assume that unidirectional cross-linguistic paths of
> evolution are easier to find for tense-aspect-mood systems than for case
> systems?

I don't think so.  As far as I know there's no synoptic work yet to
parallel BP&P, but there certainly are robustly attested recurring
patterns.  Generally speaking case markers come from adpositions
(except for oddities like Celtic initial mutation this may well
be universal).  Adpositions originate in relator noun constructions
or serial verb constructions.  Dative markers originate either in
locatives (or allatives, in languages like English that maintain that
distinction) or in serialized 'give' verbs, which probably usually
go through an intermediate stage as benefactives:

	LOC > DAT
	'give' > BEN > DAT

It looks like accusatives mostly originate from datives, though
it would be dandy to have more empirical work on that available.

Ergatives seem to have a more diverse set of origins, but not
unlimited.  They may derive from genitives (probably through
a grammaticalized nominative construction in which the subject
is marked as a possessor, i.e. something like _John's shooting
the dog_ being reanalyzed as a finite construction), or from
instrumentals or (and?) ablatives, for fairly straightforward
semantic reasons (again, see Clark and Carpenter, or my 1981
paper in Language, or the work of John Anderson).  Does anybody
have suggestions of other paths?)

Anyway, none of this seems any more diverse than the set of
possibilities for, say, perfective or progressive constructions.

Scott DeLancey	
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA

delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html