Does initial fixation location rather than face inversion modulate the response in the fusiform face area?

Upside-down presentation (i.e. inversion) disproportionately reduces the recognition of faces compared to other objects. On a neural level, the response to inverted faces also differs from the response to upright faces. Functional MRI studies show that the response to inverted faces is lower than the response upright faces in the fusiform face area (FFA). What is often not taken into account is on which feature (e.g., eye or mouth) the subjects initially fixate. This could be of relevance since initial fixation location might be confounded by face inversion. Behavioral studies have shown that the inversion effect is significantly decreased when subjects’ first fixation is on the eyes, versus when it is on the nose or mouth. Evidence from an EEG study, using a variable viewing paradigm, shows that the N170, a face-sensitive ERP component, is modulated by the feature that is first fixated. Here we present an fMRI study replicating the same paradigm, which orthogonally varied face inversion and initial fixation location. We found a higher response to eye-fixated faces compared to mouth-fixated faces in the FFA. However, we did not find an effect of face inversion in this region. Based on this we wonder if the inversion effect as previously found in the FFA is actually an effect of initial fixation feature. We suggest that future studies control for initial fixation position, to properly disentangle the effect of attended feature and inversion.