Excluding SVO in Ergative Languages: A New View on Mahajan's Generalisation
Antje Lahne
November 2008
 

This paper offers a new derivation of Mahajan's Generalisation. The basic idea of the new analysis is that in ergative languages, the external and the internal argument always end up as specifiers of the same head. Consequently, there is no position between them that the verb could possibly fill. In accusative languages, on the other hand, the external and the internal argument end up as specifiers of different heads, so that there is at least one head between them that the verb can fill. The reason for this structural difference is that in ergative languages, in contrast to accusative languages, case assignment dependencies are nested. This in turn is due to the different order of application of Merge and Value operations in the two classes of languages.
Format: [ pdf ]
Reference: lingbuzz/000810
(please use that when you cite this article)
Published in: Manuscript, University of Leipzig
keywords: mahajan's generalisation, sov, svo, ergativity, syntax
Downloaded:627 times

 

[ edit this article | back to article list ]