Bill McGregor, University of Aarhus. AJL 27/2
As in a number of ergative languages, the ergative case-marker -na∼-ma in Warrwa is occasionally found on the subject of intransitive clauses, indeed even on the subject of verbless clauses. I argue that the presence vs. absence of the ergative marker in this environment is not random free variation, but is motivated and highly constrained. The paper is concerned with identifying the motivations. It is proposed, based on an investigation of uses in a corpus of narrative and other texts, that two features are relevant: (a) semantic – the subject is highly agentive; and (b) referential – the identity of the subject is not predictable: it is unexpected. Use of the ergative on an intransitive subject thus highlights both the agentivity and the unexpectedness of the subject. I argue that, contrary to recent claims by some, Warrwa is not an active language: it is not the case that -na∼-ma groups together some intransitive and transitive subjects, while zero marking groups some intransitive subjects with transitive objects; these groupings are, I argue, purely formal and epiphenomenal. Finally, I situate optional marking of intransitive subjects in Warrwa in a wider theory of optional case marking.
This is a revised version of a paper presented to the Australian Linguistic Society Conference, University of Queensland, 7 July 2006. Thanks to the audience for useful questions, and to Jane Simpson and two anonymous referees for comments on an earlier draft, which was written during a two month stay in the Department of Linguistics, Research School of Pacific and Asian Studies, Australian National University, July–August 2006. I am grateful to Andy Pawley for making me welcome in his department, and providing infrastructure support for the duration of my stay. I am also grateful to Alec Coupe, Alice Gaby, Felicity Meakins, Carmel O’Shannessy, Edgar Suter, and Jean-Christophe Verstraete for making unpublished work available to me, and for discussions of optional ergative marking. The fieldwork on which this paper is based was funded by Australian Research Council Large Grants A58930745 and A59332055. The initial research was undertaken during tenure of a Research Fellowship at the Katholieke Universiteit Leuven, 1998; this was followed up by further research at the Max Planck Institute for Psycholinguistics in Nijmegen. Thanks are due to these organizations for their support. My greatest debt of gratitude, of course, goes to my Warrwa teachers, Maudie Lennard and †Freddy Marker.