Testing the predictions of the feature-assembly hypothesis: evidence from the L2 acquisition of Spanish aspect morphology

According to the Feature-Assembly Hypothesis (FAH) (Lardiere 2005, 2009, Choi and Lardiere 2006a) convergence in second language acquisition is determined by whether L2 speakers can effectively reassemble existing features into new (L2) configurations. We discuss the validity of this prediction for the L2 acquisition of Spanish imperfect, an area of attested difficulty, which requires native speakers of English to remap semantic concepts regarding the temporal status of events onto new morphological configurations. Data from 60 L1 English learners of Spanish and 15 native speakers who completed a context/sentence matching task show that only the meaning associated with Spanish imperfect which requires a new semantics-morphology mapping is problematic. We argue that a hypothesis, such as the FAH, which takes into account the conditions which determine the expression of aspect-related features, can adequately provide a fine-grained account of L2 variation in this grammatical domain