The typology of anaphor agreement effect

Abstract

In this paper, I argue that Rizzi’s anaphor agreement effect is in fact not universal as there are languages that show violation to this effect. To the question of why some languages follow anaphor agreement effect 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 first in the structure or the subject DP that that serves as antecedent to the anaphor merges first in the structure. In the former case, the order is Agree > Binding, where the anaphor do not have any phi features to control the agreement resulting in anaphor agreement effect and in the later case, the order is Binding > Agree, where the anaphor will have acquired the required phi features to control the agreement resulting in violation of anaphor agreement effect.