Central facial emotion accelerates automatic processing of multiple facial expressions: An EEG study

Abstract

Automatic processing of multiple facial expressions (MFE) is essential for adaptive social interactions, enabling the rapid detection of emotional cues in complex social environments. Behavioral studies have shown that, during explicit judgments of overall emotion, faces in the central visual field contribute more strongly than peripheral faces. However, it remains unclear whether centrally presented emotional faces make a distinct contribution to automatic MFE processing compared with non-emotional central faces and, if so, whether their influence is primarily reflected in the speed or strength of neural responses. We conducted three EEG experiments using visual mismatch negativity (vMMN), a well-validated neural marker of automatic processing. Forty-eight college students participated (24 in Experiments 1/2; 24 in Experiment 3). A forward–reverse oddball paradigm was used to control for physical stimulus confounds. In Experiment 1, participants performed a central gender-detection task on MFE stimuli (one central face + four peripherals); central faces were either emotional (happy/surprised, congruent with peripherals) or neutral. We compared vMMN latency and amplitude between the central emotional and neutral conditions to examine whether central emotional faces make a distinct contribution to automatic MFE processing. Experiment 2 presented isolated happy or surprised faces at fixation and served as a single-face benchmark for comparison with Experiment 1. Experiment 3 replaced central neutral faces with scrambled faces and used a face-irrelevant color-detection task to test whether the distinct contribution persisted. Experiment 1 revealed an acceleration effect of central emotion: central emotional faces elicited significantly earlier vMMN latency than neutral ones, with no significant difference in amplitude. Experiment 2 showed shorter vMMN latency for MFE than for single emotional faces. Experiment 3 replicated the acceleration effect of central emotion during a color-detection task. A consistent negativity bias emerged: surprised (threat-related) faces elicited earlier vMMN than happy faces. These findings suggest an acceleration effect of central emotion in automatic MFE processing: central facial emotion accelerates, rather than enhances, automatic MFE processing. These results advance our understanding of MFE ensemble perception and provide neural evidence for its theoretical development.