Abstract
“Deagentivisation” is the process by which an agent loses its status ‐ usually considered as semantico‐syntacti‐cally privileged in clause‐structure ‐ and is either (a) not expressed at all, or (b) expressed as a secondary constituent (e.g., as a PP, a Genitive NP, etc.). If the agent is directly expressed its role is salient, i.e., the action is considered as having been caused by the agent (ultimately or directly), in deagentivised constructions it is the action itself, its result, or any other of its participants (the patient 1, goal, theme, etc.) that is taken as the central element in the clause. Deagentivisation goes together with “detransitivisation”, as it changes a prototypical transitive construction into a (much) less prototypical one. It is a very frequent process which in a language like English is mainly, though by no means only, visible in passive constructions with or without a by‐phrase. Deagentivisation is a syntactic‐semantic process but its origin is pragmatic, as the speaker decides to “demote” the agent which, under normal circumstances, would be selected as the subject, i.e., as the first perspective in the organisation of the sentence (see Dik, 1989). This decision is taken through the weighting of several parameters of various natures, again ranging from syntax to pragmatics. This paper will attempt a (rather sketchy) preliminary presentation of the general process which would have to be subject to empirical validation.
Notes
This paper was partly financed by the Research project DGICYT PS‐94–0014. It also benefited from a stay at the University of California, Berkeley, with a Complutense Del Amo scholarship for 1996–1997.
Address correspondence to: E. Bernárdez, C. Filología Inglesa, (U.C.M.), Vicente Aleixandre, 21, 28700 San Sebastián de los Reyes, Madrid, Spain.