Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Upper theta-band power indexes noun-phrase interference during sentence comprehension
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster D101 in Poster Session D, Wednesday, October 25, 4:45 - 6:30 pm CEST, Espace Vieux-Port
Shannon McKnight1, Phillip Gilley2, Albert Kim2; 1Fort Lewis College, 2University of Colorado Boulder
Prior studies have assumed that activity in a frequency band reflects a single cognitive mechanism, unless separated by temporal or spatial dynamics. However, here we find two functionally and spectrally distinct responses within theta band (3-7Hz) in two large scale studies examining the neural oscillatory dynamics of sentence processing. Participants silently read sentences in rapid-serial visual presentation (SOA = 550; ISI = 150) while EEG was recorded and responded to simple Yes/No comprehension questions after one-third of all sentences. EEG data was epoched at the sentence level for pre-processing, including removing low-frequency (0.001Hz) drift, eye-blink, and high amplitude artifacts for each participant. Power in the theta band was derived by squaring the absolute value of complex wavelet coefficients (6-cycle Morlet wavelets 64 frequencies, octave spaced 2-8Hz). For each participant, lower (3-4.5Hz) and upper (4.5-7Hz) theta power was calculated by finding the local maxima within each range and averaging power around the peak within a mid-frontal group of electrodes. Independently for each theta range, repeated paired t-tests were conducted across time with false-discovery rate correction for multiple comparisons to evaluate differences between conditions. In Study 1 (N = 195) participants read sentences containing center-embedded relative clauses, semantic and syntactic anomalies, and well-formed control sentences. First, in response to object-relative (difficult) compared to subject-relative (easy) sentences we find an increase in upper theta activity at the main verb. The presence of two subject noun phrases (NP) in the object relative clause sentence creates interference when the comprehender encounters, and must assign a subject to, the main verb, reflected in this upper theta band increase. Secondly, in response to semantic anomalies compared to controls we saw an increase in upper theta band power. This effect was not present for syntactic anomalies; however a lower theta band increase was found here, indicating a specificity for the upper theta band reflecting the interference between a well-formed predicted NP and the anomalous noun. In Study 2 (N=192), participants read sentences containing prepositional phrases that either did or did not match the subject NP in number. There was an increase in upper theta band power for prepositional phrases that matched the subject in number, and thus interfered with the subject NP, at the main verb phrase compared to sentences that contained a mismatching prepositional phrase. Furthermore, when examining numerically anomalous verb phrases, we saw an increase in lower theta band power, mirroring the syntactic anomaly effect found above. In sum, two distinct peaks of theta band activity arose consistently across two groups of participants completing different comprehension tasks indicating that there is likely not only one cognitive process operating in the range. The current results demonstrate that upper theta band activity reflects the processing demand present when NP interference occurs during comprehension, a finding consistent with work generally linking theta band activity to cognitive control processes. Further work can determine the functional significance of activity in the lower theta band, however it appears to be linked to syntactic error processing, not interference between semantically distinct noun phrases.
Topic Areas: Control, Selection, and Executive Processes, Syntax and Combinatorial Semantics