The project studies displacement asymmetries from a cross-linguistic perspective. Subject displacement is more restricted than object displacement in that the former requires special morphological devices in many languages. The goal is to find out what underlies this constraint on displacement in the mental grammar and which features need to co-occur in a language to trigger the asymmetry. To this end, we focus on the comparatively little studied phenomenon of agreement-suppression (AS). Cross-linguistic patterns of variability and stability in AS, identified in a questionnaire study, enable us to determine (i) whether AS is in fact another instance of a general restriction against subject extraction and (ii) what causes this restriction.