The Passive Construction in English

Abstract

1. It has been recognized for a long time that English passives involve some sort of transformation, i.e. that they exhibit a regular structural relation to the corresponding active forms. Our problem is to find the most adequate formal device to represent this relation within the framework of current syntactic theory.l The earliest version of passive transformation was given by Chomsky in essentially the following form: (1) NP, Aux, Vt, NP', X =X NP', Aux + be + En, Vt, by + NP, X. Although this seems intuitively correct, it presents several difficulties when its full implications are examined in the light of the present theory of generative grammar. In particular, the derived phrase structure of the transform is not very clear, since no definite phrase structure can be assigned to the agentive byphrase, which is generated ex nihilo (Lees 1960:31). Katz and Postal (1964:72, 148-9) avoided this difficulty by deriving the passive from an underlying structure with a passive marker by + Passive, which they consider to be an expansion of the manner adverbial. The syntactic motivation for this treatment comes from the fact that the verbs that can undergo passivization are restricted to those that take manner adverbials freely. By making use of this passive marker, it is possible to dispense with subcategorial restrictions on the verb in the structure index of the passive transformation, since it can be non-transitive as well as purely transitive, provided it co-occurs with this marker (Katz and Postal, 149; Chomsky 1965:104-5). Thus the recent formulation by Katz-Postal-Chomsky would look like this: