Processing number and gender agreement in L2 Spanish: An ERP investigation of advanced learners

Alemán Bañón, J. , Fiorentino, R. & Gabriele, A.

University of Kansas

Recent proposals have attempted to elucidate how L1 properties affect development and ultimate attainment in L2 acquisition. The Failed Functional Features Hypothesis (FFFH) (Hawkins & Chan, 1997) proposes that learners can only acquire those features of the L2 that are present in the L1. In contrast, Lardiere (2009) argues that L1/L2 differences in the morphological realization of a given feature will also constitute a stumbling block for learners, regardless of the inventory of features in the L1/L2. Finally, the Full Access/Full Transfer Hypothesis (FAFT) (Schwartz & Sprouse, 1996) proposes that advanced learners can overcome transfer and acquire all features of the L2. The present ERP study examines these proposals by examining how native speakers of English, a language with number but not gender, process number and gender agreement violations in Spanish sentences. Violations were realized between nouns and adjectives (1. edificio muy *seguros ‘building-SG very safe-PL’) where English does not realize agreement, and between demonstratives and nouns (2. estos *apartamento ‘this-SG apartment-PL’), an environment where English requires number agreement. FFFH predicts differences between number and gender overall. Lardiere predicts differences in number agreement between conditions (1) and (2) above. Finally, FA/FT predicts no difference between number and gender at advanced proficiency levels. Preliminary results for L1 Spanish (n=9) show that number and gender violations in both conditions yielded equal P600s, an ERP response assumed to reflect the repair of morphosyntactic anomalies, including agreement mismatches (Osterhout & Mobley, 1995). Preliminary results for the advanced L2 group (n=7) show that number and gender violations yielded equal P600s in the noun/adjective combination, but not in the demonstrative/noun condition, where only gender violations elicited a P600. This suggests that L2 learners are sensitive to agreement violations involving features not present in the L1 and with a different morphological realization.