The impact of scene inversion on early scene-selective activity

Category-selectivity is a ubiquitous property of high-level visual cortex manifested in distinct cortical responses to faces, objects, and scenes. These signatures emerge early during visual processing, with each category sensitive to specific types of visual information at different time points. However, it is still not clear what information is extracted during early scene-selective processing, as scenes are rich, complex, and multidimensional stimuli. Here, we tested the hypothesis that early scene-selective activity involves global processing by examining the impact that scene inversion has on visually-evoked Event-Related Potentials (ERPs), particularly the scene-selective P2 component (peaking around 220 ms post-stimulus onset). We recorded ERPs from participants as they were viewing images of faces, objects, and scenes. Images were presented in both upright and inverted orientations and spanned a wide range of category dimensions and properties, preventing the possibility of salient image properties influencing the results. Replicating previous studies, P2 was found to be the first scene-selective component, with higher amplitude in response to scenes compared to faces and objects. Inversion had a clear effect on P2, with a decrease in its amplitude to inverted scenes. No scene inversion effects were observed on P1 and N1, the earlier visually-evoked components. Conversely, face inversion impacted all three components, while inverting objects had no effect on any of the three components. The current findings support the notion that P2 indexes scene-selective processing and suggest that global scene information is extracted at the P2 time window.