[PS-1.5] Morphological decomposition and second language proficiency

Gor, K. 1 & Jackson, S. 2

1 School of Languages, Literatures, and Cultures University of Maryland College Park USA
2 Center for Advanced Study of Language University of Maryland College Park USA

The processing of inflectional morphology reveals important differences between L2 learners and native speakers (NSs), as well as differences between L2 learners at different stages of proficiency/development. We report morphological priming effects in an auditory lexical decision experiment, with data from three groups of L2 learners of Russian (advanced, advanced high, and superior proficiency; N=57) and a group of Russian NSs (N=11). In matched-prime conditions, the primes were verbs in the 1st person singular nonpast tense, and the targets were infinitives of the same verbs. Russian regular, irregular, and semi-regular verbs were balanced across two frequency ranges, high and low (20 items per condition). Reaction time (RT) data were analyzed using linear mixed-models, and revealed a striking asymmetry in the role of priming with irregular verbs. For NSs, priming was observed for all verb types, and was significantly larger (p < 0.001) for irregular verbs. In contrast, for the advanced and advanced high L2 learners, priming was observed in every condition except for low-frequency irregular verbs. Only at the superior level of proficiency did significant priming emerge for low-frequency irregular verbs in L2 learners. The results do not support a categorical difference in the processing of regular and irregular verbal morphology by NSs (cf. Orsolini & Marslen-Wilson, 1999, for Italian verbs). Conversely, less proficient L2 learners did not take advantage of the mapping of irregular primes and targets, which indicates that they had problems with decomposition of irregular inflected verbs involving complex allomorphy. These results support general claims that L2 learners should have reduced or delayed access to morphological decomposition (e.g., Ullman 2005, Clahsen et al. 2010), but they provide a richer picture of the development of morphological processing, even to the point of demonstrating fairly deep processing of less transparent irregular forms at higher levels of L2 proficiency.