Reflexive Possessive in Mongolian and the Morphosyntax of Binding and Agreement

Abstract: 

In many languages, binding influences not only interpretation but also surface morphology. Special phi-agreement patterns often arise only when there is a binding relationship (e.g., Kratzer 2009, Paparounas & Akkuş 2023). Additionally, bound anaphors may fail to control verbal agreement, or trigger unique agreement patterns (Anaphor Agreement Effect, e.g., Rizzi 1990, Burzio 1996, Woolford 1999, Tucker 2012, Murugesan 2019). Understanding the nature of these phenomena requires examining the interplay between morphological, syntactic, and semantic factors in anaphoric systems across typologically different languages.

In Khalkha Mongolian, reflexive possessive agreement morphology appears in three environments: (bound) pronouns, possessive DPs, and nominalized clauses. This morphology has two key functions: marking local binding relationships between nominal arguments, and serving as a reference-tracking device in embedded clauses to indicate coreference between embedded and matrix subjects (Guntsetseg 2011; Janhunen 2012). The present analysis suggests that these two functions stem from a unified grammatical source—binding.

Specifically, these patterns are explained through a general binding mechanism combined with the language's independent morphosyntactic properties. Mongolian reflexives structurally resemble possessive DPs, where binding of reflexives is analyzed as binding of the "possessor" inside the reflexive. The same binding mechanism extends to nominalized clauses, where reflexive morphology results from agreement between a nominalizing head D and a locally-bound embedded subject. Finally, I discuss the implications of the current case study for two interrelated theoretical domains: the nature of binding and reflexivity, and the effect of binding on morphosyntax.