The role of cognitive load in automatic integration of emotional information from face and body

Abstract

We perceive emotions daily through facial expressions, often accompanied by a body posture that provide additional emotional context. Congruent facial and bodily expressions (conveying the same emotion) enhance emotional recognition compared to incongruent ones, suggesting interaction between these channels. Although behavioral evidence suggests that this integration occur automatic, its underlying neural mechanisms remains unclear. This study investigated the automaticity of facial and bodily expressions integration by manipulating cognitive load. Twenty-eight participants completed an emotion recognition task with congruent or incongruent facial and bodily expressions while performing a memory task under low or high cognitive load. EEG recordings captured brain activity, and emotion recognition accuracy and reaction times were measured. Results revealed that congruent expressions improved recognition, with bodily expressions exerting a stronger influence on facial expression recognition than vice versa. Early neural responses (P100, N100, P250, N250) were stronger during facial expression focus, while later responses reflected attention to body expressions. Bayesian analyses provided strong evidence for the absence of significant interaction between congruence and cognitive load, supporting the automaticity of integration. These findings suggest that emotional expressions are integrated automatically, independent of cognitive resources, and emphasize the differential influence of bodily expressions over facial expressions in shaping emotional perception.