A study of the Samoan "-cia" suffix is presented. It argues that, contrary to prevailing theory, Samoan does have an active/passive contrast but that it is indicated by a difference in word order rather than by verbal morphology. It is shown, however, that "-cia" is similar to a passive suffix in that passive involves the choice of a patient as subject, while "-cia" indicates the choice of a patient as absolutive. It is also proposed that "-cia" can not always be analyzed as a transitive suffix since, besides deriving transitive verbs, it also derives intransitive verbs. A brief historical explanation of why "-cia" appears in active as well as passive clauses is offered, based on passive-ergative reanalysis. (MSE)