Abstract
To explain how our visual system encodes multiple objects simultaneously, based on fMRI results from visual short-term memory (VSTM) studies, the neural object file theory argues that the inferior intra-parietal sulcus (IPS) first individuates objects based on their spatial/temporal properties and then the superior IPS and higher visual areas encode their detailed feature information (Xu & Chun, 2009, TICS). Presently it is unknown whether IPS regions from both hemispheres contribute equally or whether one hemisphere plays a more important role in such visual processing. By subtracting ipsilateral signals from contralateral ones, previous event-related potential (ERP) studies have reported contralateral delay activities (CDA) that share certain similarity with those reported in fMRI studies. This suggests that some of the parietal activities seen in fMRI studies may have a contralateral bias. Meanwhile, neuropsychological studies on neglect and fMRI studies on visual attention have all reported a more bilateral representation in the right parietal cortex and a more contralateral representation in the left parietal cortex. Here we asked observers to encode two or four object shapes either unilaterally or bilaterally. Regardless of the encoding load, both the inferior and the superior IPS responded strongly to both ipsilateral and contralateral shape presentations. Nevertheless, both regions also showed a small but significant contralateral bias, with the bias being greater in the inferior than in the superior IPS. This is consistent with inferior IPS' involvement in object individuation which is location-based and superior IPS' involvement in object identification which is more feature-based. Interestingly, we did not find any laterality differences between the two hemispheres. Taken together, these results suggest that (1) CDA results from previous ERP studies may largely originate from the inferior IPS region, and (2) top-down visual attention induced parietal laterality effects reported previously may be dissociable from those seen during object individuation and identification.
This research was supported by NSF grant 0855112 to Y.X.