Figural after-effects, retinal size, and apparent size

The results of earlier experiments on the question of whether figural after-effects are affected by apparent as opposed to retinal size are shown to be inconclusive. A new hypothesis is proposed namely that both factors may be responsible for producing figural aftereffects, and four experiments have been made to test it. Situations were used in which the apparent sizes of the figures were determined by the size-constancy effect. It was found that where retinal sizes of test and inspection figures are the same and apparent sizes are different, figural after-effects in the direction which would be predicted on the basis of apparent sizes are obtained. It was further shown that where retinal and apparent sizes are in conflict, whether a figural after-effect is seen or not, and the direction of the figural after-effect, depends upon the balance between these two factors.     

Apparent distance vs. apparent size as determiners of figural after-effects

Abstract.— Sutherland's (1954) much discussed experimental design was analysed in a new fashion. When the original experiment was repeated as such, the expected FAE was found. However, when the same IFs and TFs were presented in a vertical position, no significant FAEs were found. The results lend support to the interpretation presented earlier by Kolehmainen (1968): The FAE found by Sutherland was apparently due to an uncontrolled binocular disparity resulting in a different localization of stimulus figures in respect to the fixation point.     

Retinal size as a determiner of the direction of size distortion in figural after-effects

When dealing with size distortions, the direction of the distortion is determined by the retinal size of the T-figure with respect to the retinal size of the I-figure. When the T-figure is retinally smaller than the I-figure it is distorted toward appearing even smaller; when it is retinally larger than the I-figure it is distorted toward appearing even larger; and when the two are retinally equal, there is no distortion. If the T-figures are too much larger or smaller than the I-figure, the distortion ceases. The traditional distance paradox is observed where the distance is defined retinally. The apparent size of the figures is irrelevant to a prediction of what sort of distortion takes place. Sutherland's hypothesis that apparent size determines the direction of distortion when retinal sizes are equated is not supported.