Revisiting the evidence for objects in English

The present paper challenges the long-standing distinction between direct and indirect objects and the status of the latter as a central grammatical category. My points of departure are the frameworks developed by Quirk et al. (1985), Aarts and Aarts (1982) and Huddleston and Pullum (2002 and 2005) where the separation of syntax and semantics is not strict enough to yield a syntactically appropriate notion of objects in English. A new theory of verb complementation referred to as “Passivisable Object Theory” (“PO Theory”) is proposed which stipulates that objects must be passivisable by definition. It is also shown that passivisation itself neither introduces nor eliminates prepositions. Thus both 
(a) She gave Mary a book and (b) She gave a book to Mary allow for only one passivisation each and consequently nothing but the first NP is classed as object in either structure. A book in (a) will be reclassified as a “complement extension” or CE (a new type of complement) and the PP in (b) as a “predicator complement” or PC (category adapted from Aarts and Aarts). This has far-reaching consequences for the entire system of transitive verbs and will – against Mukherjee (2005) – marginalise ditransitive structures and show that syntax is more independent of semantics than many grammarians would acknowledge. My claims are supported by evidence from BNC Baby and contrasted with examples from Victorian novels.