Laterality and (in)visibility in emotional face perception: Manipulations in spatial frequency content

Abstract

It is widely agreed that hemispheric asymmetries in emotional face perception exist. However, the mechanisms underlying this lateralization are not fully understood. In the present study, we tested whether (a) these asymmetries are driven by the low spatial frequency content of images depicting facial expressions, and (b) whether the effects differed depending on whether the emotional facial expressions were clearly visible or hidden (i.e., embedded in low spatial frequencies). The manipulation sheds light on the contribution of cortical and subcortical routes to emotional processing mechanisms. We prepared both unfiltered (broadband) and hybrid faces. Within the latter, different bands of spatial frequency content from images of 2 different expressions were combined (i.e., low frequencies from an emotional image combined with high frequencies from a neutral image). We presented these broadband and hybrid images using the free-viewing emotional chimeric faces task (ECFT) in which 2 images are presented above and below fixation and asked participants to report which of the 2 mirror reversed images appeared more emotional. As predicted, the results showed that only broadband expressions produced the well-known left visual field/right hemisphere (LVF/RH) bias across all basic emotions. For hybrid images, only happiness revealed a significant LVF/RH bias. These results suggest that low spatial frequency content of emotional facial expressions, which activates the magnocellular pathway in subcortical structures and bypassing cortical visual processing, is not generally sufficient to induce an LVF bias under free-viewing conditions where participants deny explicitly seeing the emotion, suggesting that the LVF bias in ECFT is primarily cortically mediated.