The effect of edge orientation and movement direction on the apparent misalignment of collinear bars.

Four experiments on apparent misalignment of oblique collinear bars are reported. The data from the first three experiments showed that the misalignment was slight but significant when no direction other than that of the bars themselves was delineated and about double when the vertical was delineated by movements of the adjustable bar. When the vertical was delineated by both these movements and the ends of the bars (or by vertical parallel lines), the misalignment was more than six times greater. Conversely, it was reduced when the bar ends delineated the vertical and bar movements delineated the direction at right angles to the bars. The data from a fourth experiment showed that the inner pair of edges were closely involved in apparent misalignment and that the outer ends were not. The relationship between the misalignment effect with bars and the Poggendorff figure (oblique lines separated by parallels), the delineation of direction by edges and movement, and the implications of these data for an explanation are discussed.

The Poggendorff illusion and apparent interparallel extents.

An explanation of the Poggendorff misalignment effect in terms of apparent contraction of interparallel extent resulting from the Müller-Lyer illusion was tested in three experiments. Three of the eight stimulus figures had oblique transversals outside the parallels in the usual way, three had them inside, and two were controls consisting of the transversals only. Müller-Lyer forms were differently delineated between the parallels for the inside-transversal and outside-transversal figures, and were not delineated in the control figures. In the first experiment apparent misalignment occurred in four of the six parallel-line figures and in neither of the controls. In the second experiment oblique extent between the parallels was underestimated in six of the eight figures and right-angle extent was overestimated in all of them. The results of the third experiment showed that right-angle (horizontal) extent between the parallels without transversals is estimated without significant error. The data from the three experiments do not support the interparallel-extent explanation of apparent misalignment. Instead, the results are interpreted in terms of independent perceptual compromises, one involving alignment of the transversals and the other the distance between them.

Sine of an illusion.

In an informally observed sine-wave figure in which the vertical extent between contours was constant, apparent extent in the crest and trough (the 'turns') appeared greater than in the straight oblique sections of the figure. This observation was confirmed in two experiments in which the vertical extents were matched by two vertically arranged dots. It was found that in a turn the apparent extent was greater than the true extent, but in a straight section both extents were about equal. These outcomes were confirmed when the two sections were each separated from the figure and presented alone. The illusion is explained in terms of a perceptual compromise between the vertical extent and the greater overall dimensions of the section at the turn of the sine-wave figure and is thereby held to be the same in principle as the Müller-Lyer illusion.

The Bourdon illusion occurs with straight-, right-angle-, and parallel-edge figures.

The Bourdon illusion is the apparent bentness of the straight edge of a figure consisting of two elongated triangular components arranged apex to apex. In three experiments, the illusion was shown to occur in the opposite direction, with the components arranged base to base. It was also shown to occur with the component edges at right angles and parallel. With the edges at right angles, the illusion also occurred in one direction when the components were apex to apex and occurred in the opposite direction when they were base to base. Supplementary observations indicated that the illusion is stronger when the components are relatively small and widely separated and eliminated when one of the two edges is curved.

Perceived causality occurs with stroboscopic movement of one or both stimulus elements.

It was shown by magnitude estimation that the perception of causality first described by Michotte (1946/1963) also occurs consistently and strongly with stroboscopic, ie apparent or phi, movement. This is so when the 'causal' movement is stroboscopic and the 'caused' movement real, when these movements are reversed, and when both movements are stroboscopic. The effect is not due to prior experience with Michotte-type displays.