The Acquisition of Correlates of Unaccusativity in L2 French

The Unaccusative Hypothesis (Perlmutter 1978) maintains that there are two types of intransitive verbs: unergative, where the grammatical subject is the Agent and the verb describes an activity (laugh, run) and unaccusative, where the grammatical subject is the Theme and the verb describes an achievement, accomplishment or state (open, arrive, exist). Levin and Rappaport Hovav (1995) argued that unaccusatives are not a monolithic class but rather can be divided into distinct sub-classes based on syntactic and semantic criteria: alternating change-of-state verbs such as break (henceforth “CS verbs”), verbs of inherently directed motion such as arrive (henceforth “IDM verbs”), verbs of appearance, disappearance and existence such as appear and die, and verbs of occurrence such as happen. While there is much overlap cross-linguistically, languages differ as to what aspect of meaning—telicity for unaccusatives or agentivity for unergatives—determines class membership (Dowty, 1991; Levin and Rappaport Hovav, 1995; Van Valin, 1990 for example). Research in the L2 acquisition of unaccusativity in English has shown that learners with different L1s and at different levels of proficiency produce and accept inappropriate passive morphology with unaccusative verbs, although their production and/or judgements may become native-like. However, some learners fossilize before having attained native-like performance (Han, 1998). Sorace (1993a) showed that near-native Francophone and Anglophone learners of Italian fell short of native-like performance, but that they were sensitive to an unaccusative hierarchy. Subjects’ performance was more accurate with verbs of change of location such as andare (‘go’) and venire (‘come’), which are high on the unaccusativity hierarchy, and less consistent with alternating change-of-state verbs, which are at the lower end of the hierarchy. Oshita (2001), based on previous research on the L2 acquisition of unaccusativity in English, Italian and Chinese put forth the Unaccusative Trap Hypothesis (UTH). Under the UTH, beginning learners do not distinguish unergative and unaccusative verbs, both of which are subject to the same “single-argument linking rule”. Based on the input, intermediate learners begin to distinguish the two classes and may eventually sort them out, though some remain in the “unaccusative trap”, that is they fossilize. Montrul (2005) tested Sorace’s (2000) Auxiliary Selection Hierarchy (ASH) and the UTH in the L2 acquisition of Spanish using a judgement task including diagnostics of