The data represent induced power at the bridge between a matrix clause and an embedded clause represented by the words dit 'said' and que 'that' in French linked to four conditions as in (1a - d).
(1a) Quel reportage la concernant est-ce que Laetitia a dit que Jérémie avait regardé avec appréhension?
(1b) Quel reportage à propos d’elle est-ce que Laetitia a dit que Jérémie avait regardé avec appréhension?
(1c) Quel reportage la concernant est-ce que Jérémie a dit que Laetitia avait regardé avec appréhension?
(1d) Quel reportage à propos d’elle est-ce que Jérémie a dit que Laetitia avait regardé avec appréhension?
‘Which report regarding/about her did Laetitia/Jérémie say that Jérémie/Laetitia had watched with apprehension?’
These four conditions manipulated the contains of interrogative expressions. (1a/condition A) involves a modifier with an early match antecedent. (1b/condition B) involves a complement with an early match antecedent. (1c/condition C) involves a modifier with a late match antecedent. (1d/condition D) involves a complement with an late match antecedent. Our the time-frequency analysis was performed with the FieldTrip toolbox as eight datasets for four conditions (1a - 1d) and two groups (NSs and NNSs). The time window of interest in each epoched trial included the two critical bridge words dit que ‘said that’, and so covered 1.85 seconds, plus a 700ms baseline. A 50ms interval was left between the baseline and the critical window to prevent spectral leakage in plotting and analyses. We convolved a family of Morlet wavelets of 7 cycles with the selected time window of each EEG trial, which yielded the time-frequency information of the induced neural activity. After the induced power was obtained for each condition of each subject, a log transformation was conducted across channels and frequencies for each subject to standardize the power unit as dB. The length of the wavelets was set as 3 standard deviations of the Gaussian kernel. At 1Hz, the wavelet duration was 2.228 seconds. The spectral bandwidth was 0.286Hz. At 60Hz, the wavelet duration was 0.037 seconds. The spectral bandwidth was 17.143Hz. This procedure only allowed us to analyze the data starting at 5Hz (wavelet duration = 0.446 seconds; spectral bandwidth = 1.429Hz).
The Timing versus Resource Problem in Nonnative Sentence Processing: Evidence from a Time-Frequency Analysis of Anaphora Resolution in Successive wh-Movement in Native and Nonnative Speakers of French
