We discuss the recent proposal by Koontz-Garboden (2009) (cf. also Chierchia 2004) that reflexively marked anticausative verbs (in Romance languages and beyond) are semantically reflexive. This proposal predicts that a sentence headed by a lexical causative verb should not entail the sentence headed by the reflexively marked anticausative counterpart. We uncover problems with the main argument for this claim and add further tests which show that a causative sentence does, in fact, entail its anticausative counterpart, whether reflexively marked or not. Our findings support standard semantics of the causative alternation according to which anticausatives, whether reflexively marked or not, denote inchoative one-place predicates. They also reconfirm that the relevant reflexive morphology is syncretic and does not necessarily derive reflexive semantics.