The (in)compatibility of Anaphora and Agreement

Abstract

In this paper, I argue that Rizzi’s anaphor agreement e?ect is in fact not universal as there are languages that show violation to this e?ect. To the question of why some languages follow anaphor agreement e?ect and why certain other languages violate it, I demonstrate that this is an independent consequence of whether in a given language the functional head that carries the agreement probe merges ?rst in the structure or the subject DP that that serves as antecedent to the anaphor merges ?rst in the structure. In the former case, the order is Agree > Binding, where the anaphor do not have any ? features to control the agreement resulting in anaphor agreement e?ect and in the later case, the order is Binding > Agree, where the anaphor will have acquired the required ? features to control the agreement resulting in violation of anaphor agreement e?ect.