You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster E47, Friday, November 10, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Oscillatory dynamics identify unique neural processes beyond event-related responses during auditory sentence comprehension

Julie Schneider1, Mandy Maguire1;1Callier Center for Communication Disorders at the University of Texas at Dallas

Introduction: Event related potentials (ERPs) and time frequency analysis of the EEG can identify the temporally distinct coordination of groups of neurons across brain regions during sentence processing (Lam et al., 2016). Although there are strong arguments that ERP components are driven by the same changes in neural oscillations during similar processes (Hagoort et al., 2004; Davidson & Indefrey, 2007; Bastiaansen et. al., 2002) others argue that the lack of clear associations between the two suggests oscillatory dynamics are more than just time-frequency representations of ERP components (Wang, Zhu & Bastiaansen, 2012; Bastiaansen & Hagoort, 2015), making it unclear how the two are related. The current study seeks to examine the neural activity underlying auditory sentence processing of both semantic and syntactic errors, to clarify if ERP and time frequency analyses identify the same or unique neural responses. Methods: Thirty-nine right-handed, monolingual, English-speaking adults, ages 18-32 years (11 male; Mage=22.1, SD=3.8) completed an auditory semantic and grammatical judgement task. All sentence types included a ‘target agent-action pairing’ in which an inanimate agent (noun) was paired with a modal verb and action verb. In the correct sentences, the target agent-action pairing was both possible and grammatically correct. The grammatically incorrect sentences included either the intrusion or omission of a present participle (-ing) form of the verb (i.e. will be bake, will baking). The semantic violation introduced an unsuitable pairing of actions with agents (i.e. hose-bake). Analysis: Data were epoched from -100 to 1500 msec around the target verb and a morlet wavelet analysis was applied. Single trials were averaged together to obtain a stable waveform ERP for each condition and each electrode for every subject. For the time frequency analysis, epochs were averaged across trials, within each condition, and the mean baseline power at each electrode and frequency was subtracted (Delorme & Makeig, 2004). We conducted a monte-carlo cluster correction analysis in the theta (4-8 Hz) and lower beta (13-20 Hz) bands to determine significance. The mean amplitude for each condition was computed and a Pearson correlation coefficient was computed to determine if a relationship existed. Results: As expected the semantic judgment task, t-tests confirmed a larger N400 amplitude (t(38)= 5.25, p< 0.001) and greater increase in theta power (t(38)= -2.85, p< 0.01) for semantic errors compared to correct sentences. In the syntactic judgment task, P600 amplitudes were larger (F(2,114)=2.935, p=.05) and beta power decreased (F(2,114)=8.56, p<0.001) for both grammatical error types compared to syntactically correct sentences. Importantly, the correlation identified no significant relationship between the ERPs and neural oscillatory dynamics. Discussion: Increases in theta and the N400 effect are both related to semantic processing, while decreases in beta and the P600 effect are linked to syntactic processing; however, no relationship between the event-related and time frequency responses existed. Based on these findings we conclude that event-related and time frequency responses may be gauging similar language processes, but oscillatory dynamics add to research by uncovering unique, distinct neural changes.

Topic Area: Methods

Back to Poster Schedule