Topicality, accessibility, and causality in anaphora resolution: An eye-tracking study of null and overt pronouns in Italian

Abstract It has been argued that the distribution of pronominal forms in pro-drop languages is shaped by a complex interplay of syntactic, semantic, and pragmatic factors. Null pronouns are typically interpreted as co-referring with the most prominent antecedent, whereas overt pronouns often signal a topic shift, co-referring with less prominent antecedents. Although many frameworks have been proposed to explain the division of labor between null and overt pronouns, empirical studies on anaphora resolution in causative sentences remain scarce. In this study, 41 Italian native speakers completed a referential selection task involving subject-biased or object-biased complex causative sentences with either a null or an overt pronoun. Results revealed a strong reliance on discourse-level factors, with participants consistently selecting the pragmatically appropriate antecedent regardless of pronoun type. However, eye-tracking data showed higher processing costs in accommodating violations of the typical pronoun-antecedent combinations. This was especially true in subject-biased sentences with an overt pronoun but was also observed in object-biased sentences with a null pronoun. These findings suggest that, while pragmatic factors largely drive anaphoric resolution in causative sentences, pronoun type leads to different processing costs in contexts where the pragmatic antecedent does not align with the expected pronoun type.