Face or building superiority in peripheral vision reversed by task requirements

Peripheral vision has been the topic of few studies compared with central vision. Nevertheless, given that visual information covers all the visual field and that relevant information can originate from highly eccentric positions, the understanding of peripheral vision abilities for object perception seems essential. The poorer resolution of peripheral vision would first suggest that objects requiring large-scale feature integration such as buildings would be better processed than objects requiring finer analysis such as faces. Nevertheless, task requirements also determine the information (coarse or fine) necessary for a given object to be processed. We therefore investigated how task and eccentricity modulate object processing in peripheral vision. Three experiments were carried out requiring finer or coarser information processing of faces and buildings presented in central and peripheral vision. Our results showed that buildings were better judged as identical or familiar in periphery whilst faces were better categorised. We conclude that this superiority for a given stimulus in peripheral vision results (a) from the available information, which depends on the decrease of resolution with eccentricity, and (b) from the useful information, which depends on both the task and the semantic category.