Perceptual displacement of a test mark toward the larger of two visual objects.

Abstract

Contour-repulsion concepts of satiation, fatigue, or inhibition explain some illusions, but not the fact, shown in Exp. I, that a hashmark between two squares of unequal size is perceived closer to the larger square. The aforementioned illusion could be explained as a result of distance perspective, for the larger square might be perceived as closer. This perspective theory cannot explain the fact that a vertical hashmark, placed between two vertical lines of unequal length, appears closer to the long line. Both results agree in quantitative detail with an adaptation-level model saying that the length of a test extent is judged relative to other nearby extents.

Contour repulsion vs. adaptation-level interpretations of the Baldwin illusion

“…Preliminary analysis of results indicated that order effects (Series A before B or vice versa) and differences in ratios of sizes of boxes had no statistically Significant effect on judgments, though in Series A the illusion tended to be stronger with larger ratios of size of box, as found by Merryman and Restle (1970). Data were thus combined so that, for each subject over all 40 judgments in Series A and over all 24 judgments in Series B, a mean R2 -Rl score was obtained as a negative, positive, or zero score.…”

“…Following the pattern of data analysis used by Merryman and Restle (1970), the displays in each condition may be paired such that within each pair the hash mark is located in the same place, the boxes are of the same ratio of sizes, but only the position of the larger box (left or right of the smaller) is different. The response to the first display of a pair, with the larger box on the left, was called Rl and the response to the second display of that pair, with the larger box on the right, was called R2.…”

“…Thirty-two stimulus cards carrying different versions of the Baldwin figure were prepared in two series: 20 cards in Series A and 12 cards in Series B. For Series A, the horizontal line of the Baldwin figure was drawn to give the same visual angle as in the experiment by Merryman and Restle (1970), that is, 2 degrees 30 minutes, which was calculated to be 1.79 cm for the distance between the large and small boxes in the figure. For these end boxes, four large-to-small pairings of the boxes were selected giving size ratios corresponding to 8: 6, 6: 4, 8: 4, and 8: 2.…”

“…The experiment by Merryman and Restle (1970) does not, therefore, provide a critical test between the predictions of genuinely competing theories, as there are no grounds for expecting any outcome of the experiment other than that predicted from adaptation-level theory. In order to investigate the possible effects of contour repulSion within the Baldwin figure in situations in which contour repulsion should have an effect on judgment, the present experiment explores the possibility that the Baldwin illusion may reverse as contour separation between the end boxes is reduced.…”

“…As the stimulus for their investigation, Merryman and Restle (1970) used the Baldwin illusion shown in Figure 1. In this figure, the hash mark (h) divides a horizontal line into two equal segments, Xl and X2 • According to adaptation-level theory, both the larger box (Bd and the smaller box (B2 ) should influence the apparent lengths of both Xl and X2 • This theory would predict that Xl, being affected more by the adjacent larger box (B l ), should appear shorter than X2 , which is affected more by the smaller box (B2 ).…”