Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neural Signatures of Filler-Gap Dependency Processing in L1 and L2 English Speakers
Poster Session B, Friday, September 12, 4:30 - 6:00 pm, Field House
Shuzhen Wang1, Laura Sabourin1; 1University of Ottawa
Language processing is a complex task requiring our brain to rapidly integrate various linguistic information from unfolding words and quickly build comprehension. While native speakers (L1) are widely assumed to engage in both incremental and predictive processing in real time, debates persist regarding how second language speakers (L2) process sentences in real time and whether their processing mechanisms align with or differ from those of native speakers. The present event-related potential (ERP) study thus aimed to examine whether late L2 speakers recruit the same neural mechanisms as L1 speakers when resolving the English filler-gap dependency – a long-distance relationship between a fronted wh-word (filler) and its canonical syntactic position (gap). We compared online wh-dependency resolution in English between 11 native English speakers and 20 late Mandarin-English learners, divided by a cloze test median into High versus Low proficiency groups. Participants read 24 globally grammatical sentences with manipulations of filler plausibility (plausible vs. implausible) and gap licenser types (transitive verb vs. preposition). For example, the filler could be either a plausible or implausible match for the first verb “wrote,” as in “Lucy identifies which article/woman the author wrote a column to describe _ in the newspaper.” The gap licenser could be either a transitive verb or a preposition, as in “Lucy identifies which article the author wrote a column to describe/about _ in the newspaper.” Sentences were presented word by word for 500 ms with a 400 ms interval using the Rapid Serial Visual Presentation (RSVP) paradigm, while participants’ EEG was recorded. Mean amplitudes were analyzed at three regions of interest: the first verb (wrote), the filled-gap noun phrase (column), and the gap licenser (describe vs. about). At the first verb, the L1 group showed a robust N400 effect for the implausible fillers compared to plausible fillers in the 300-500 ms time window. L2 speakers with high proficiency showed a delayed N400 in the 500-600 ms time window. At the filled-gap noun phrase, high proficiency learners displayed an N400 when an expected gap appeared to be filled, indicating an expectation violation, whereas low proficiency learners showed a frontal P600-like positivity consistent with syntactic reanalysis. Native speakers showed no significant effect, likely because reanalysis had already been completed at the determiner. At the gap licenser, all three groups – L1, L2-High, and L2-Low – showed more negative-going waveforms for prepositions than transitive verbs. This pattern was consistent across time windows: an N280 (250–320 ms) typically associated with function words like prepositions, an anterior negativity (300–500 ms) reflecting increased working memory demands, and a late negativity (500–900 ms) indicating continued working memory processing. Crucially, the absence of Group × Condition interactions across all critical regions challenged theories suggesting fundamental differences between L1 and L2 processing. Overall, the findings of the current ERP study provide evidence for qualitatively similar L1 and L2 neural mechanisms during the real-time processing of wh-dependencies in English. Moreover, high-proficiency L2 speakers showed delayed sensitivity to filler-verb plausibility compared to native speakers but more advanced predictive processing than low-proficiency speakers.
Topic Areas: Language Development/Acquisition, Syntax and Combinatorial Semantics