The Horizontal-vertical illusion in haptic and visual space

Abstract

The vertical line of an inverted-T or L figure is judged visually to be longer than the horizontal line of equal length, an effect commonly called the horizontal-vertical illusion. This illusion has also been observed in touch perception, but the precise relationship between the haptic (tactile-kinaesthetic) and visual illusions is not clear. The present study is concerned with a clarification of this issue. In the first series of experiments, constant errors in the haptic judgment of length in the horizontal plane were found to relate consistently to the time and velocity of limb movement. Radial movements, which are executed at a slower speed and for a longer time, are judged longer than tangential movements of equal length. The data were considered in relation to certain physiological and kinematic properties of the actively moving limb. Taken together with additional information on judgments of movement duration, the results suggest that the haptic illusion of length is governed by the perception of the difference in movement times. In this way, it is specific to properties of the moving limb and is thus quite unrelated to the analogous illusion in vision. In the second series of experiments, the developmental function of the haptic L illusion was found to resemble that obtained for the visual L figure. It was concluded that common developmental processes give rise to the similarity of the two illusions across age levels, and that such data could best be explained in terms of Piaget's account of perceptual development. The final series of experiments were concerned with the illusion with the T figure. In line with observations on the visual figure, the haptic form of the illusion was found to diminish and consequently disappear, when the dividing line was shifted from the midpoint of the divided line, to form an L or reversed-L, in the extreme. When the visual figure was exposed in a piecemeal fashion, an illusion was obtained only when the subject himself directed the exposure of the partial views, but not when the exposure of these was solely controlled by the experimenter. The haptic illusion was also found to be greater when the subject actively scanned the figure with the moving limb, compared with the condition where his limb movements were rendered passive. Taken in conjunction with the further finding that the illusion also occurs under line-drawing and walking conditions, it was concluded that the same central mechanism, possibly involving selective attention, governs the illusion in visual, haptic and locomotive space.