Abstract
In the real world, do we rely on our knowledge of object size (i.e., familiar size) for space perception or do we compute object dimensions from a combination of the retinal image and oculomotor cues? Under restricted viewing conditions (i.e., a monocular pinhole to minimize oculomotor cues), when an object’s retinal angle is the only cue to size and distance, the visual system relies on familiar size. In this case, when presented with objects, for which the actual size is inconsistent with the familiar size (e.g., unusually large chairs), perception is inaccurate (e.g., Ittelson, 1951). In contrast, under unrestricted binocular viewing in natural environments, familiar size may have little or no effect on perceived size (e.g., Predebon, Wenderoth, & Curthoys, 1974). We examined size and distance perception while manipulating not just the physical size and distance of objects from the viewer, but also the congruency of objects with their familiar sizes and the availability of oculomotor cues. We presented real Rubik’s cubes and dice, each either in a size congruent with expectations (5.7-cm Rubik’s cube and 1.6-cm die) or the reverse or incongruent size (5.7-cm die and 1.6-cm Rubik’s cube), at two distances (25 cm and 91 cm). Participants viewed one object at a time in a dark tunnel (to eliminate pictorial cues), either monocularly through a 1-mm pinhole (to eliminate oculomotor cues) or binocularly (with full oculomotor cues). Participants indicated the perceived size and distance of an object by moving their fingers apart (manual estimation). Regardless of the presence or absence of oculomotor cues, familiar size affected both size and distance perception: Rubik’s cubes were perceived as larger and farther than dice, even when objects had identical dimensions. In sum, familiar size is a potent visual cue that affects object perception even during binocular viewing.