Neural Basis of the Interaction between Alerting and Executive Control

Filip KottikBartłomiej PanekIlona KotlewskaMikołaj CompaDariusz Asanowicz^*

• Institute of Psychology, Jagiellonian University, Kraków, Poland

Attentional alerting – evoked by an accessory stimulus such as a tone presented briefly before target onset – generally decreases response time (RT) but this decrease is smaller in trials with a conflict (induced e.g., by presenting flankers that are incongruent with the target stimulus). This somewhat paradoxical interaction is usually interpreted as an increased conflict cost, possibly indicating less efficient conflict resolution. The present study investigated the electrophysiological activity underlying the impact of alerting on response conflict processing. Human participants performed a modified version of the Eriksen flanker task while EEG was recorded. Alerting tone was presented either with a short stimulus-onset asynchrony (SOA 100 ms) or long (SOA 800 ms), or was not presented at all (no alerting condition). To examine how alerting modulates motor, visual, and central executive processing, we analyzed evoked (i.e., phase-locked) activity (event-related potentials or ERPs), induced (i.e., non-phase-locked) activity (local power modulations), and phase coherence-based functional connectivity. Time-frequency power and phase of the EEG signal were measured from EEG sources isolated with a method employing the generalized eigenvalue decomposition (GED). Behavioral results replicated the alerting-conflict interaction in RT (but not in error rates). In the EEG results, effects of alerting were observed as: (i) an increase in conflict-related midfrontal theta power, (ii) a decrease in midfrontal N2 amplitude, (iii) a decrease in LRP latency, and (iv) an increase in N2pc amplitude. Moreover, several alerting effects were present only in the SOA 800 condition, suggesting that they may be specific to endogenous alertness: (i) a suppression of the flanker effect on response-related lateralization of alpha/mu power, (ii) an increase in the flanker effect on LRP latency, and (iii) an increase in the flanker effect on the target-related contralateral suppression of visual alpha power. The findings suggest that alerting dynamically modulates both the emergence and resolution of response conflict through widespread changes within a neural network that can be characterized as a “selection-for-action” system. Alerting may serve as a key modulator of neural dynamics in this system.