Abstract
Perception of unfamiliar faces is right-lateralised. However, familiar faces have been reported to be processed differently: familiar faces activate the left hemisphere more strongly than unfamiliar faces. This difference has been proposed to reflect higher processing of verbal information associated with familiar faces, e.g. the name. However, it remains unclear whether this increased left-hemispheric processing of familiar faces has an effect on how familiar faces are perceived. Therefore our study investigated the effect familiarity has on the right-hemispheric bias in face perception. To address this question, chimeric faces of familiar and unfamiliar faces were used. Familiar faces depicted pictures of celebrities participants could name. Since pictures of celebrities might differ from unfamiliar faces, e.g. be more symmetrical, we used pictures of celebrities unknown to the participants as our unknown faces. Next, chimeric faces were created by splitting a face image into a left and right half, and then mirroring each. This resulted in two new images. One image contained solely information of the left image half of the original (left chimeric) processed by the right hemisphere. Whilst the other image contained solely information of the right image half of the original (right chimeric) processed by the left hemisphere. To measure lateralisation of face perception, participants were asked to choose which chimeric face looked more like the original face. A preference for the left chimeric would suggest a right-hemispheric bias. In line with previous literature, participants preferred the left chimeric when unfamiliar faces were used supporting the known right-hemispheric bias for the perception of unfamiliar faces. However, familiar faces did not elicit a bias in perception: no preference for either chimeric face emerged. Thus, familiarity seems to change how faces are perceived. Familiar faces seem to be perceived more bilaterally, potentially due to an increased activation of the left hemisphere.
Meeting abstract presented at VSS 2017