This paper investigates German light verb constructions (LVCs) with experiencer objects that follow the pattern ‘in N versetzen’ and their contrast to derivationally related simplex verbs (SVs), e.g. in Angst versetzen (lit. ‘transfer into fear’) vs. ängstigen ‘frighten’. Such pairs are often referred to as synonyms. We address the question whether there are systematic differences in usage and meaning between this family of object-experiencer LVCs and corresponding SVs by means of a corpus study on 1,983 tokens. The quantitative analysis indicates that there are differences between the LVC and SV with respect to the frequency of subject as well as object types. A mixed-effects model shows how the choice of SV vs. LVC is conditioned by the semantic type of both arguments: for instance, collective experiencers and animate non-experiencer arguments prefer the LVC. The lexical stem plays the role of a random effect. The quantitative picture gives rise to a semantic analysis that develops explanations for the correlations between argument properties and predicate shape. We ultimately classify the choice of the LVC over the corresponding SV as overt marking of causative semantics, motivated by the corpus data and corroborated by semantic tests.