Representations Of Visual Working Memory Guide The Detection Of Emotional Faces

Previous studies have documented that stimulus matching the contents of visual working memory (VWM) have an automatic advantage in the competition of attentional process. The present study aims to explore the electrophysiological correlates of influence from VWM on visual search of emotional faces. In experiment 1, Participants performed a search task to detect happy or angry faces among groups of neutral faces while simultaneously keeping in VWM a color cue shown at the beginning of each trial. A VWM test was required at the end of each trial to ensure that the cue had been maintained in VWM. Happy faces elicited a larger amplitude of N2pc component when VWM features matched the target face (valid condition) and a smaller amplitude when VWM features matched a distracter face (invalid condition), compared with the neutral condition (VWM features did not match any face in the search array). Additionally, angry faces elicited a greater N2pc amplitude in valid trials than in neutral and invalid trials. Also, the anger superiority effect in valid conditions was significantly smaller than that in neutral trials, which indicated that anger superiority effect was modified by the representations of VWM. But we did not find the anger superiority effect to be eliminated under the guidance of VWM, as indicated by the larger N2pc potential elicited by angry faces than happy faces in all three cue conditions.Can the results of the above experiment be unambiguously attributed to actively maintaining the cue in VWM? Perhaps mere exposure to the sample without memorising it would also produce the same effects. Specifically, priming an object’s representation might direct attention to that object compared with a novel one. Thus, priming and VWM effects may be confounded in Experiment 1. Therefore, we conducted Experiment 2 as a control. In this experiment, participants were required to only pay attention to the cue items, with no memorisation task enforced. The result of this experiment showed that the search efficiency for angry and happy target faces remained stable across the three perceptual cue conditions. Also, anger superiority effect was not modulated by different perceptual cue conditions. These results support the findings from Experiment 1, suggesting that the guidance effect of VWM on search for emotional faces cannot be explained by mere exposure to the cues.Together, the present study demonstrates that the content of VWM can guide the detection of angry or happy faces, as indicated by a modulation of N2pc component. However, the anger superiority effect is not eliminated under the guidance of VWM representations. These findings implied that the top-down guidance of attention from VWM is effective when processing complex items that convey social information (i.e., faces), but is not powerful enough to modify the anger superiority effect.