Filler-Gap Parsing In Wh-movement: An ERP Study Of Chinese Learners Of English

There have been a number of behavioral and neural studies concerned with Shallow Structure Hypothesis which claims that L2 speakers and L1 speakers are fundamentally different in parsing syntactically complex sentences, and L2 speakers mainly depend on semantic information. The results of previous studies are controversial. By using Event-Related Potentials (ERP) technique, our present study investigates whether Unbounded Filler-Gap dependency can be parsed in a same way between L1 and L2 and to what extent L2 are more inclined to semantic information.The fillers in unbounded filler-gap dependency sentences may have a Plausible Intermediate Gap (IG) or Implausible IG or None IG in this present study (eg. Who did the police know/claim/police’s claim about the women killed).In our present study,None IG situation elicited a much larger P600 than both Plausible IG and Implausible IG situation in L1 indicating there was a significant main effect of syntactic structure IG and L1 could successfully apply Intermediate Gap during processing to facilitate the Filler-Gap integration. While L2 had no sign of P600 in None IG situation which supposed to be the most difficult situation to parse,there was a significant main effect of semantic plausibility in L2. Implausible IG situation produced a Sustained Negativity from 300ms to 800ms in L2, indicating the semantic violation between filler and main verb had important impacts on L2.These results demonstrate Filler-Gap dependency and IG are processed differently in L1 and L2. L1s mainly rely on syntactic structure during the processing and are hardly affected by semantic violation, while L2s are more sensitive to semantic information and can not apply abstract syntactic structure during the processing.