Viewpoint (In)dependence of Action Representations: An MVPA Study

摘要

The discovery of mirror neurons—neurons that code specificnactions both when executed and observed—in area F5 ofnthe macaque provides a potential neural mechanism underlyingnaction understanding. To date, neuroimaging evidence fornsimilar coding of specific actions across the visual and motornmodalities in human ventral premotor cortex (PMv)—the putativenhomologue of macaque F5—is limited to the case of actionsnobserved from a first-person perspective. However, itnis the third-person perspective that figures centrally in ournunderstanding of the actions and intentions of others. Tonaddress this gap in the literature, we scanned participantsnwith fMRI while they viewed two actions from either a firstornthird-person perspective during some trials and executednthe same actions during other trials. Using multivoxel patternnanalysis, we found action-specific cross-modal visual–motornrepresentations in PMv for the first-person but not for thenthird-person perspective. Additional analyses showed no evidencenfor spatial or attentional differences across the twonperspective conditions. In contrast, more posterior areas innthe parietal and occipitotemporal cortex did show cross-modalncoding regardless of perspective. These findings point to anstronger role for these latter regions, relative to PMv, in supportingnthe understanding of othersʼ actions with reference tononeʼs own actions