Dissociations between slant-contrast and reversed slant-contrast

A vertical test probe is misperceived as slanted in the opposite direction to an inducer when disparity specifies the inducer slant while monocular cues specify a frontoparallel surface (slant-contrast). In reversed cue conditions with vertical axis slant the test probe is misperceived as slanted in the same direction as the inducer (reversed slant-contrast). We found reliable slant-contrast and reversed slant-contrast with inducers having horizontal-axis slant. The reversed slant-contrast was not influenced when the inducer and probe were separated in the frontal plane or in disparity depth whereas slant contrast was degraded, especially in the latter condition. Slant contrast was most pronounced when the inducer was slanted like a ceiling compared to like a ground. No such difference was found for the reversed slant-contrast. When the cue conflict was minimized slant-contrast was reduced, but only with inducers having ground-like slant. Implications for an existing model explaining the slant effects are discussed.