The role of binocular viewing in a spacing illusion arising in a darkened surround.

A study is reported of the binocular-oculomotor hypothesis of the moon illusion. In a dark hall, a pair of light points was presented straight ahead horizontally, and another pair was presented at the same distance but 50 degrees upward. Twenty subjects compared the spacings of these two pairs. Half of the subjects viewed the stimuli first monocularly and then binocularly, and the other half viewed them in the reverse order. Eye position was also systematically varied, either level or elevated. A spacing illusion was consistently obtained during binocular viewing (with the upper spacing seen as smaller), but no illusion arose during monocular viewing unless it was preceded by binocular viewing. Furthermore, an enhancement of the illusion due to eye elevation was found only during binocular viewing. These findings replicate the report of Taylor and Boring (1942 American Journal of Psychology 55 189-201), in which the moon was used as the stimulus, and support the binocular-oculomotor hypothesis as a partial explanation for the moon illusion.