Abstract
We are rarely exposed to faces in the absence of situational context. Previous ERP research using faces primed with contextual sentences suggests that valence and self-relevance modulate electrocortical responses to faces. However, the time-course of these effects is unclear and no studies have investigated whether they interact with another key sign of self-relevance: whether the face is looking at or away from the participant. We used emotional sentences to vary the context within which neutral faces were placed. These sentences referred to the face having a positive or negative opinion (valence manipulation) of the participant or of someone else (self-relevance manipulation). Participants read each sentence before viewing the face that the sentence referred to, and faces had either direct or averted gaze (gaze manipulation). Eye-tracking was used to ensure that participants read the sentences and enforced fixation to the face, while ERPs were recorded. In a preliminary sample of 15, mean amplitude analyses of 100ms time-windows from 150-750ms after face onset were performed on a subset of occipito-temporal electrodes to track the time-course of contextual modulation. Self-relevance elicited a larger amplitude response for other versus self-relevant context on the right hemisphere, consistently from 250 to 650ms. Gaze direction interacted with valence consistently from 250 to 750ms for faces in the other-relevant, but not self-relevant context. In the other-relevant positive context, direct gaze produced a smaller amplitude response than averted gaze. In contrast, direct gaze produced a larger amplitude response than averted gaze in the other-relevant negative context. Participants also rated faces in a self-relevant context as more arousing than faces in the other-relevant context, suggesting a potential arousal-based mechanism for self-relevance effects. These results indicate that the contextual cues of self-relevance and valence impact perception of neutral faces in a complex way depending on gaze direction.
Meeting abstract presented at VSS 2016