Abstract

This article addresses the issue of why some languages permit the interpretation of what are basically simple‐action transitive verbs (e.g. beat) as causative change‐of‐state verbs in the context of a ‘resultative’ adjective (e.g. beat the metal flat), while others do not. This crosslinguistic asymmetry has not received an entirely satisfactory explanation, despite the fact that resultative predication has been widely discussed from a variety of perspectives. We note that languages show a systematic correlation between (1) the presence of a grammaticalized opposition between perfective and imperfective aspect lexically/morphologically encoded in verb forms and (2) the absence of secondary syntactic predication encoding result states in combination with verbs otherwise denoting simple actions. Greek and English are taken as representative of the two ‘types’ of language. Building on the insights of Talmy (1985), Tenny (1987, 1994) and Snyder (1995), among others, an explanation is proposed for the absence of such secondary predication in Greek and similar languages based on the restrictive effects of grammatical aspect marking on the contextual (re)interpretation of a verb's inherent aspectual character (Aktionsart)—effects which render impossible the conversion by syntactic means of VPs denoting atelic activities into VPs denoting accomplishments. We also show that the ‘failure’ of some languages to allow the unaccusativization of agentive verbs of motion (e.g. unergative walked in the park v. unaccusative walked into the park) is closely related to this issue, though the outcome is also subject to a given language having the resources to mark ‘goals’ unambiguously. An important conclusion of the analysis is that the ‘deficit’ of languages like Greek in not encoding such resultative predication syntactically is not syntactic in character, but located at the interface of inflectional morphology and lexical semantics