Ergativity in Austronesian languages

Ileana M. Paul &Lisa Travis

University of Western Ontario

             &

Lisa Travis

McGill University

Within the Austronesian family, many languages are classified as ergative (e.g. Samoan) or as having some ergative properties (e.g. Tagalog). For one particular language to exhibit some but not all of the characteristics of ergativity is problematic for an ergativity macroparameter. In this paper, we look at a sample of Austronesian languages to assess the range of ergative properties within this family.

            We begin with Malagasy – a language that appears in some respects to be ergative (similar to Tagalog, as discussed by Schachter 1976 and Maclachlan 1996). The most compelling evidence in favour of an ergative analysis comes from the prominence of the agent in Theme Topic (TT) constructions. The actor (or ergative NP) can both bind reflexives and control null arguments, unexpected behaviour for a “demoted” agent. (1) is an example of Actor Topic (AT) construction (considered active in the nom/acc analysis and anti-passive in the ergative analysis). (2) and (3) illustrate binding and control respectively in a Theme Topic (TT) form (considered passive in the nom/acc account and transitive in an ergative account).

1.         Manaja an’i i  Bakoly Rabe.               AT: transtitive (nom/acc) or antipassive (erg)

            at.respect i Bakoly Rabe

            ‘Rabe respects Bakoly.’

2.         Hajain-dRabe ny tenany.                         TT: passive (nom/acc) or transitive (erg)

            tt.respect-Rabe det self.his

            ‘Rabe respects himself.’

3.         Kasain-dRabe sasana ny zaza.

            tt.intend-Rabe tt.wash det child

            ‘Rabe intends to wash the child.’

            Other characteristics of Malagasy are similar to ergative languages. The Actor is the imperative addressee and is marked with the same case (genitive/ergative) as possessors. Moreover, extraction famously targets the absolutive argument, as is common among ergative languages. Finally, the lack of derived objects in Malagasy can be easily explained from an ergative view: one wouldn’t expect movement to a demoted theme in an anti-passive construction.

            There are, however, several problems with the ergative analysis. The first is the anti-passive (Actor Topic). The verbal morphology in the anti-passive occurs in other verb forms, such as regular intransitives and applicatives. Second, the theme is clearly not demoted, as it is not optional and can control both secondary predicates and PRO. Moreover, the agent in the apparent active (Theme Topic) is always omissible, unexpected behavior for a transitive subject. Finally, Theme Topic constructions fix weak crossover violations, unlike what occurs in other ergative languages (Bobaljik 1993).

            We compare Malagasy with a range of Austronesian languages and show that the “mixed ergative” nature of Malagasy is not unique. On the other hand, not all languages share the same combination of ergative properties. This lack of cohesion is problematic for theories of ergativity that attempt to tie ergative properties to a particular syntactic structure. Any analysis of ergativity, however, must take into account the range of properties exhibited by Austronesian. In other words, any analysis for which the syntactic properties of ergativity fall out from a single macroparameter must explain why Austronesian languages have such “mixed” properties.

References

Bobaljik, Jonathan. 1993. Ergativity and ergative unergatives. In Papers on Case and Agreement II: MIT working Papers in Linguistics, ed. by Colin Phillips, 45-88.

Maclachlan, Anna. 1996. Aspects of Ergativity in Tagalog. PhD, McGill University.

Schachter, Paul. 1976. The Subject in Philippine Languages: Topic, Actor, Actor-Topic, or None of the Above? In Subject and Topic, ed. by Charles N. Li, 491-518. New York: Academic Press, Inc.