Abstract
When perceiving the emotion of ambiguous faces, individuals tend to judge them as negative – the ‘negativity bias’ of emotion perception. The negativity bias has also been found in peripheral vision and with groups of faces (ensemble perception). Yet little is known about the role of interactions between neighboring faces in biased emotion perception. In the current study, we presented single faces or two identical faces, with either happy or neutral expressions. The single face was presented at 5.5° eccentricity and the two faces at 5.5° and 6.5°. Participants indicated whether the facial expression was happy or neutral. The discriminability (d') was comparable in the single-face (2.21+/-0.24) and the two-face condition (2.03+/-0.21). In the single-face condition, participants tended to report happy facial expressions as ‘neutral’ (criterion: -0.76+/-0.07), revealing a strong negativity bias. In the two-face condition, there was only a slight negativity bias (-0.23+/-0.16), clearly smaller than in the single-face condition. These results suggest that a second, identical face did neither help nor interfere with facial expression discrimination, but made observers less biased to respond ‘neutral’ when happy faces were presented. In a control experiment, we varied the location of the single face, presenting it at the same location as the inner or outer face, or in the center of the two face locations. The result showed similar discriminability at all single-face locations and in the two-face condition. In the single-face conditions, there was again a clear negativity bias (-0.42+/- 0.18). There was no bias in the two-face condition (0.06+/-0.12). Interestingly, most observers reported that the two identical faces appeared to be different. Our results suggest that presenting two identical faces instead of a single face abolishes the negativity bias of emotion perception. We discuss our results in light of crowding, ensemble perception, and the processing of redundant signals.