Abstract
Input from the left visual field (LVF) is primarily processed in the right hemisphere, while input from the right visual field (RVF) is processed in the left hemisphere. However, the two hemispheres are not functionally redundant: for example, face processing is lateralized to the right hemisphere. How can the contralateral organization of the visual system be reconciled with the asymmetric distribution of visual functions? Here we examine how high-level visual information is shared between hemispheres. Specifically, using divided visual field presentation and fMRI adaptation, we investigate the hemispheric transfer of facial identity information. While fixating centrally (as verified by eye-tracking), participants were presented with two sequential faces in the periphery. The first face was presented to either the LVF or RVF, followed by a second face to either the LVF or RVF. The second face always differed in viewpoint, depicting either the same identity or a different identity. Combining across left and right fusiform face area (FFA), we found greater identity adaptation (different – same identities) when the first face was presented in LVF vs. RVF. This pattern was apparent not only in right FFA, but also in left FFA. Indeed, left FFA did not show identity adaptation even when a face was repeated in the same preferred RVF location. Together, these findings indicate that identity information is transferred from right to left FFA, but not the other direction. Moreover, selective identity adaptation in left FFA for faces that were initially presented in LVF implies that left FFA can represent, but may not be able to compute, facial identity. In sum, hemispheric transfer from lateralized processes may help reconcile the contralateral organization and hemispheric asymmetry of the visual system. Moreover, such transfer may help the visual system recognize objects that move between visual fields.