Pritha Chandra & Roberta D’Alessandro – talk

On Monday, April 25, Pritha Chandra (IIT Delhi) and Roberta D’Alessandro will be giving a talk at the Morpho-syn Syndicate (Penn State U) with the title Ergativity, phi-agreement and alignment. When things start going Accusative

If you wish to attend please let Roberta know and she’ll send you the link.

ABSTRACT

Whether structural case is parasitic on agreement or vice-versa has been a hot debate in recent years, started by Bobalijk’s (2008) article, according to which agreement is based on morphological case (m-case). On the assumption that m-case marking takes place postsyntactically (Marantz 1991), Bobalijk argues that agreement in languages like Hindi takes place after m-case has been assigned, with the highest accessible (i.e. morphologically unmarked) argument. The conclusion is that PF is the locus of agreement. A crucial piece of evidence in favor of this one-way dependence is presented in Bobalijk (2017) who investigates those Western Indo-Aryan (WIA) languages where ergative attrition appears to happen independently of agreement change. This view has been contrasted by Preminger (2014) among others; finally, some approaches take the stance that agreement takes place partially in syntax (Benmamoun, Bhatia&Polinsky 2019, Arregi&Nevins 2012 a.o).

The proposal This talk focuses on alignment shift, i.e. on the shift from ergative alignment to accusative alignment. Ergativity is often investigated from a synchronic viewpoint; we take a different route here, by trying to identify the locus of alignment change through focusing on a number of Western Indo-Aryan languages (Hindi, Marwari, Gujarati, Kutchi-Gujarati, Haryanavi, Punjabi, among others), examining phi-agreement patterns and their correlation with case and alignment. Through a microvariational and micro-diachronic approach, we identify the starting point of change in agreement with a DOM-marked object and the appearance of T as a separate head. This is where change starts to happen, and this, we claim, is where ergative alignment starts weakening. We also show that DOM-marked objects are accusative and not dative (contra Bobalijk 2017). This helps us argue against the Dependent Case Theory.