THE NATIVE AND NONNATIVE PROCESSING OF NUMBER AND GENDER AGREEMENT IN SPANISH: AN ERP INVESTIGATION

Abstract

The present study utilizes EEG to examine the processing of number and gender agreement in Spanish by native speakers and adult English-speaking learners. With respect to native processing, the study focuses on how different agreement features (number, gender) are retrieved for the purposes of agreement and on how structural distance (number of intervening phrases) impacts agreement resolution. With respect to nonnative processing, the study investigates the extent to which second language (L2) morphosyntactic processing is impacted by the properties of the learners' first language (L1), focusing on whether novel features (gender) and novel instantiations of a shared feature (number on adjectives) can be processed in a native like manner. An additional question examined in the study is whether L2 morphosyntactic processing is impacted by structural distance. Agreement was examined between nouns and adjectives within the same phrase (edificio muy seguro "building-MASC-SG very safe-MASC-SG"), between nouns and adjectives across a verb phrase (cuento es anónimo "story-MASC-SG is anonymous-MASC-SG"), and between demonstratives and nouns (este apartamento "this-MASC-SG apartment-MASC-SG"), which is a syntactic context where both English and Spanish instantiate number agreement. Both native speakers (n=24) and advanced English speaking learners of Spanish (n=25) elicited a P600 for number and gender violations overall, which was not preceded by a Left Anterior Negativity. For native speakers, effects were equally robust for number and gender, suggesting that both features are processed similarly at the brain level. For learners, effects were more positive for number than gender, suggesting a quantitative advantage for the feature present in the L1. These results are in line with the Full Transfer/Full Access Hypothesis (Schwartz & Sprouse, 1996), which predicts that adult L2 learners can show native like processing for novel features. Furthermore, while both groups showed sensitivity to across-phrase violations, they were both affected by the distance manipulation, as suggested by the fact that within-phrase agreement yielded more positive waveforms than across-phrase agreement overall. These results suggest that L2 learners can establish syntactic dependencies outside of local domains (contra the Shallow Structure Hypothesis, Clahsen & Felser, 2006) and that both native and nonnative processing are modulated by structural distance similarly.