Absolutive movement in Polynesian: Syntactic ergativity and postverbal word order variation
Lauren Clemens
This talk details a unified account of ergative subject extraction restrictions (a manifestation of “syntactic ergativity”) and variable verb-initial word order (VSO ~ VOS) in the Polynesian languages Niuean and Tongan following Clemens and Tollan to appear. The account I advance is based on the finding that syntactic ergativity can arise from the locus of ABS case assignment (Campana 1992; Bittner & Hale 1996; Aldridge 2004; Coon et al. 2014). I demonstrate how extraction asymmetries in Tongan are the result of high ABS case assignment, whereas the absence of syntactic ergativity in Niuean is explained by ‘low ABS’ case assignment (Massam 2006; Legate 2008; see also Tollan 2019 and Clemens and Tollan 2020). Next, I compare data from coordination in Niuean and Tongan to support the connection between the locus of ABS and the presence of syntactic ergativity. While Niuean’s coordination follows a consistently accusative pattern, coordination in Tongan is only accusative when XPs smaller than TP are coordinated, suggesting that syntactic ergativity stems from the presence of a relatively higher functional head. Finally, I provide a novel account of variable post-verbal word order in Tongan (VSO ~ VOS) and the absence of VOS in Niuean based again on the different position of ABS in these two languages. Object A-movement in Tongan VOS (Otsuka 2005) is a reflex of ABS case assignment: the base position of the ABS object follows the subject, while the case position precedes it. The object is pronounced in either position according to pragmatic factors. Movement is covert in VSO, which can give rise to weak crossover effects. In Niuean, the ABS object is generated and pronounced in the low position. This project contributes to research connecting syntactic ergativity and the locus of ABS, as opposed to the properties of ERG (cf. Polinsky 2015; Deal 2016). Further, the connection of syntactic ergativity to word order supports the view that ERG extraction restrictions arise from movement of the ABS, rather than the ERG argument (cf. Assmann et al. 2015).