Role of peripheral vision in the representation of objects' spatial location

The role of peripheral vision in coding the spatial location of visual objects remains unclear. It has been suggested that it plays an indirect role only, providing information to guide saccadic eye movements from which efficient spatial coding is achieved (Yamamoto & Philbeck, 2012). Here we investigated whether peripheral vision might also play a more direct role through the allocentric coding of objects' location with respect to peripheral elements within real world visual scene. Subjects faced a wide-field screen and were asked to gaze and memorize the spatial location of a red dot presented on top of a real-world picture. During the learning phase, a portion of the picture surrounding the red dot was masked in order to remove central visual cues (mask diameter: 0°, 5°, 10° or 20°). After a delay with a grey screen during which subject made a saccade, the real-world picture was presented again and subjects had to point to the memorized location of the red dot on the picture using a laser. Crucially, the reappearing picture could be shifted 3° to the left or to the right, inducing a conflict between the spatial locations defined by visual information contained in the picture or by saccadic and other extra-visual information. Although the shift was unnoticed by the subjects, pointing responses were significantly impacted and biased toward the shift. The relative weight of visual information in estimated locations decreased gradually with increasing mask diameter: 100% with no mask, 60% with 5° mask, 40% with the 10° mask and 15% with the 20° mask. Thus, even with the largest masks, perifoveal and peripheral visual information are used for coding objects' spatial location. These results indicate that peripheral vision can play a direct role in building an internal representation of objects' location in the surrounding