Poster D30, Friday, August 17, 4:45 – 6:30 pm, Room 2000AB
Beta-band ERP activity during cyclic wh-movement in French as a physiological index of non-nativeness
Laurent Dekydtspotter1, Kate A. Miller2, Mike Iverson1, Yanyu Xiong1, Swanson Kyle1, Gilbert Charlene1;1Indiana University Bloomington, 2Indiana University Purdue University Indianapolis
Non-native processing may involve procedural delay (Boxell & Felser, 2017). We examined EEG activity in native (NSs; n=24) and advanced L1-English non-native (NNSs; n=22) speakers of French as information accompanying displaced wh-expressions (1a-d) passes through bridge dit que ‘said that’. ERPs confirmed delayed distinctions in NNSs. Time-frequency analysis revealed signatures of non-nativeness in beta-band ERP activity. Information passing between clauses in wh-dependencies (1a-d) exemplifies the specificity of recursion in language. N-complements (à propos de lui; 1b,d) are re-represented in recursion, but not NP-modifiers (le concernant; 1a,c) (Chomsky, 2005; Lebeaux, 1988). When matching antecedents are available, re-represented pronouns can be syntactically bound (1b) or discursively coreferential (1a). Without matching antecedents (1c,d), these processes are thwarted. (1.) a. Quelle décision le concernant est-ce que Paul a dit que Lydie avait rejetée sans hésitation? b. Quelle décision à propos de lui est-ce que Paul a dit que Lydie avait rejetée sans hésitation? c. Quelle décision le concernant est-ce que Lydie a dit que Paul avait rejetée sans hésitation? d. Quelle décision à propos de lui est-ce que Lydie a dit que Paul avait rejetée sans hésitation? ‘Which decision regarding/about him did Paul/Lydie say that Lydie/Paul had rejected without hesitation?’ Participants completed a RSVP task (300ms/word, 250ms/ISI) with 25 quadruples crossing Structure (N-complement/NP-modifier) and Antecedent Gender (Match/Mismatch) (1a-d). EEG was recorded via a 64-electrode EGI system with 50 kΩ maximal impedance, a Net Amps 300 amplifier, and 1000Hz sampling rate. Data were preprocessed with a .05-100.5Hz band-pass filter and cleaned of artefacts via epoch/channel rejections and Independent Component Analysis. 87% of NS and 86% of NNS trials were retained. Average amplitudes, with 50ms-baseline into critical words dit ‘said’ and que ‘that’ (Phillips et al., 2005), were compared over four regions (Fiebach et al., 2002). Mixed-effect models for 250-550ms after critical words’ onsets revealed modifier-complement differences at dit in NSs (p=.02) but at que in NNSs (p=.027). Topographical differences accompanied delayed effects. Power analyses between 4-40Hz revealed a main group effect at 18-21Hz 250-400ms into dit in two clusters, with greater power for NNSs than NSs (p=.005). Main-effect modifier-complement differences arose during dit at 34-40Hz with greater modifier-complement differences in mismatch than match (p=.003) and during que at 4-5Hz with increased modifier-complement power-differences in match than mismatch (p=.035). Group-conditions interactions arose over anterior clusters for 30ms during dit at 36-40Hz (p=.038) and 50ms during que at 21-24Hz (p=.013). Analysis of average power over relevant clusters revealed the interaction at 36-40Hz resulted from greater modifier-complement power difference in match for NSs (p=.001) and in mismatch for NNSs (p=.002). The interaction at 21-24Hz resulted from a modifier-complement difference in mismatch in NNSs (p=.0005) not found in NSs (p=.077). Power-differences in gamma- and theta-band rhythms at the bridge reflected reference-related grammatical procedures linked to displaced wh-expressions. Power-differences in beta-band rhythms distinguished NNSs from NSs. This beta-band activity plausibly reflects second-language brain processing focused on maintaining weakly activated representations (Dekydtspotter & Miller, 2013; Miller, 2014, 2015), which could reduce the overall speed of processing.
Topic Area: Grammar: Syntax