Conflicts with Extraretinal and Monocular Cues Cause the Small Range of the Induced Effect

A vertical magnifier before one eye causes the induced effect: an apparent rotation of frontal surfaces toward that eye. The rotation required to restore apparent frontoparallelism grows linearly up to ∼4% magnification, but plateaus at 8%. We examined the cause of the plateau.

Horizontal disparities (quantified by horizontal size ratios, HSRs) are ambiguous indicators of surface slant. Various retinal and nonretinal signals can allow veridical slant estimation from HSR, sensed eye position, vertical disparities (vertical size ratios, VSRs), and monocular cues. Vertical or horizontal magnification of one eye's image alters the natural relationships among HSR, VSR, eye position, and monocular cues. We argue that the induced-effect plateau is caused by conflicts between these means of estimating slant. A plateau is not observed in the geometric effect because some of the conflicts do not occur with horizontal magnification. Two experiments were designed to test this hypothesis. When strong monocular cues were present, plateaux occurred at ∼8% magnification in the induced, but not the geometric effect. When monocular slant cues were made useless, induced-effect plateaux were abolished. Even with strong monocular cues present, plateaux in the induced effect were eliminated when eye position was consistent with the vertical magnification in the retinal images. The smaller range of the induced effect can only be understood from consideration of all the signals involved in slant estimation.