banniere4.jpg

Conférence linguistique (25 nov.)


Le 25 novembre, vous êtes cordialement invités à apporter votre lunch au DS-3470 de 12h45 à 13h45. Laura Grestenberger de l'Université Concordia nous présentera « Participles and voice mismatches: deponency in non-finite contexts ». Bienvenue à tous! 

Voici le résume de sa conférence: Deponents are verbs that exhibit a synchronic "voice mismatch" in the languages in which they occur: they are syntactically active, but have non-active morphology. Previous work has focused on this phenomenon in Modern Greek and Latin, and it has been suggested that deponency depends on finiteness (e.g., Papangeli and Lavidas 2009, Pesetsky 2009). In this talk I will use data from several other languages in which deponency occurs to show that the decisive factor in triggering the mismatch is not finiteness, but the presence vs. absence of vP. Non-finite forms (especially participles) provide evidence for this analysis. Building on Anagnostopoulou 2003 and subsequent work (e.g., Alexiadou and Anagnostopoulou 2008), I argue that the properties of participles are determined by different levels of incorporation of functional structure: in target state participles, the nominalizer attaches directly to the root and does not include the functional projection vP (or Voice) while in result state participles, it does include vP. I propose an analysis in which the mismatch is triggered if vP introduces a non-canonical agent argument. We therefore expect voice mismatches to be preserved in nominalizations that include vP, but disappear in nominalizations that do not include vP. The syntax of deponent participles in Hittite, Vedic Sanskrit, Ancient and Modern Greek and Latin confirms this prediction and makes it possible to pinpoint exactly where and why voice mismatches occur.