View article

The present study tested the hypothesis that facilitated processing of threatening faces depends on working memory load. Participants judged the gender of angry versus happy faces while event-related brain potentials were recorded. Working memory load was manipulated within subjects by the mental rehearsal of one-versus eight-digit numbers. Behavioral results showed that the relative slow-down to angry compared to happy faces in the gender-naming task (ie, the negativity bias) was eliminated under high working memory load. Under low (but not high) load, N2 amplitudes were smaller to angry compared to happy faces. Moreover, high load reduced LPP amplitude and eliminated the enhanced LPP to angry compared to happy faces that were present under low load. These results suggest that working memory load improves attentional control, and reduces sustained attention for distracting negative …