In many languages, subjects are much more difficult to extract from embedded clauses than objects. In English, subject extraction from that-clauses appears to be prohibited:
Interestingly, German and Dutch do not appear to have such a constraint and subject extraction from embedded complementizer clauses is relatively unproblematic. In my current research I investigate whether there really is no subject/object asymmetry in German and Dutch and if so, why this is the case. The ultimate goal is to find out which constraints are at work when long-distance subject dependencies are created.
I also work on intervention effects, specifically negative intervention and case-intervention. Intervention effects are associated with degraded acceptability and/or an increased processing load. They arise when an intervenor ‘intervenes’ within a dependency (such as a dependency created by movement). My current work includes research on negative intervention and case-intervention.
In this joint project with Andreas Hiemstra (University of Oldenburg & University of Groningen), we investigate the acceptability and processing of different types of long-distance wh-questions in Dutch by L3 learners who either have German as the L1 and English as the L2 or English as the L1 and German as the L2. Long-distance questions involve movement of a question word from a subordinate clause into a higher clause and are highly complex structures, which are not explicitly taught in language courses. They show a lot of variability in their acceptability and the ways in which they are realized across languages. In our study, we aim to find out how L3 learners of Dutch differ from native speakers in how they process, judge and produce different types of long-distance questions. In particular, we are interested to see whether deviations from non-native behaviour can be explained by transfer from the L1, transfer of the L2, access to UG principles or processing factors, or a combination of several of these factors.
In this project, we investigate the acceptability, processing and production of superlatives and 2-verb clusters in Dutch by L3 learners who either have German as the L1 and English as the L2 or English as the L1 and German as the L2. Dutch has two optional variants for superlatives and 2-verb clusters, whereas German and English allow only one opposing option: natuurlijkste (GER)/meest natuurlijke (ENG) and gedanst heeft (GER)/heeft gedanst (ENG). In Dutch, both variants of the superlative and 2-verb clusters are grammatical. In our study, we aim to find out how L3 learners of Dutch differ from native speakers in how they process, judge and produce different types of superlatives and 2-verb clusters . In particular, we are interested to see whether deviations from non-native behaviour can be explained by transfer from the L1, transfer from the L2 (Bardel & Falk, 2007), (psycho-)typological factors (Rothman, 2011), foreign language proficiency, or a combination of several of these factors.
Principal investigator: Andreas Hiemstra (University of Oldenburg & University of Groningen). Collaborators: Esther Ruigendijk (University of Oldenburg), Marije Michel & Greg Poarch (both University of Groningen).