Much psycho- and neurolinguistic research has been dedicated to filler-gap dependencies, identifying the cognitive mechanisms and brain structures underlying the processing of these structures. However, this research tended to focus on languages with one shared property, namely the lack of grammaticized resumptive pronouns. The objective of the proposed project is to expand the empirical scope of the explorations of unbounded dependencies by investigating the processing of these constructions in Hebrew, a language in which gaps and resumptive pronouns alternate (relatively) freely. The project will focus on two interrelated research questions:1. Are differences between the grammars of languages reflected in processing? Specifically, are the same processing strategies employed in languages with and without resumptive pronouns? For example, it has been shown that in English, once the syntactic processor identifies a filler, it engages in active search for a possible gap. Given that in Hebrew some fillers are not associated with gaps but rather with resumptives, does the parser in this language actively construe gaps like its English counterpart, or does it adopt a 'wait and see' strategy?2. Are resumptive pronouns easier to process than gaps, as is sometimes assumed in the literature, or are they in fact harder to process, since their processing invariably includes a misguided extra step, namely the (unwarranted) construal of a gap? Or does their processing involve some aspects which are easier (e.g. accessing their antecedent), and some which are harder?These issues will be addressed using techniques and paradigms proven successful in the study of online sentence processing, e.g. cross-modal lexical priming and self-paced reading, as well as event-related potentials, utilizing specific effects as indexes of known processing mechanisms, and functional magnetic resonance imaging, assessing the processing load associated with resumptives compared to gaps.