The passive construction as a strategy for an agent-patient construction: Some comments on terminology

Abstract: This chapter addresses longstanding terminological disagreement sur- rounding the notions of passive and voice and argues for a principled distinction between functionally defined constructions and formally defined coding strat- egies. It is proposed that grammatical terms such as passive are most usefully employed in line with both traditional and much current usage, namely to denote formal strategies. On this view, a passive construction is defined as an overtly verb-marked alternant in a valency alternation, in which the agent of a basic transitive construction is downgraded and the patient is promoted to subject posi- tion. Such a definition excludes unmarked patient-initial constructions, lability alternations, and purely discourse-driven word-order inversions, all of which have at times been classified as passives in recent functional approaches. This terminological precision is argued to reduce conceptual ambiguity and to permit the well-attested association between passive constructions and patient topicality to be treated as an empirical generalization rather than a definitional criterion. Finally, it is emphasized that a narrow definition of the passive as a formal strat- egy does not entail a narrow conception of voice phenomena. On the contrary, a clear separation of form and function enables systematic comparison between passive constructions and a broader set of functionally related structures across languages, while providing a shared terminological foundation for the contribu- tions to this volume.