The types and numbers noun phrases verb takes

B83

“Listen my story”: ERP sensitivity to argument structure violations in L2 Chia-Hsuan Liao & Ellen Lau (University of Maryland)
cliao@umd.edu
Event concepts of common verbs (e.g., eat, sleep) are likely to be the same for speakers of different languages, but the argument structure of a verb (e.g., the types and numbers of noun phrases a verb takes), is linguistic knowledge, and can vary from language to language. For example, in English, “listen” is intransitive and requires a preposition to introduce a second argument, whereas in Mandarin, “listen” can take a second argument without any additional functional material. How do different argument structures interact with each other in bilinguals during online sentence processing? We hypothesized that native language argument structure knowledge is particularly difficult for L2 speakers to override in online processing. Therefore, in our experiment, we constructed sentences with verbs of mismatched argument structures in L1 and L2. The prediction was English native speakers can easily reject argument structure violation sentences such as My sister listened the music,while Mandarin L2 speakers of English may often erroneously accept such sentences because they are unable to immediately access the argument structure of their L2 online. Both behavioral and ERP results supported the hypothesis: While L1 speakers showed a prominent P600 effect to argument structure violations, the L2 group showed no ERP differences for the violations, and poorer behavioral sensitivity, even though they showed behavioral and ERP sensitivity to gross phrase structure violations in a control comparison. Our finding suggests argument-verb computation is likely subject to L1 transfer, and has implications for existing L2 processing theory.

Table 1: Example stimulus for each condition

L1 speakers!

Figure 2:Grand average ERPs of the Phrase structure conditions from the preposition to the following word at Pz and the topographic distribution in the 300-500 and 600-900 ms intervals at the preposition. References: [1] Neville et al., (1991) Journal of Cognitive Neuroscience, 3(2), 151-165 [2] Clahsen & Felser (2006), Applied Psycholinguistics, 27,3-42