No abstract

An explanation of the Poggendorff misalignment illusion in terms of three basic components, the horizontal-vertical, longitudinal-transverse and obtuse angle effects, is proposed. It is argued that these effects amy either singly or jointly give rise to an apparent elongation- contraction of the space between the aligned elements with consequent change in apparent oblique direction. Experimental evidence for the involvement of the three components in various combinations and for modification of perceived direction between the transversals is presented. It is shown that the illusion in various forms of the Poggendorff figure can be accounted for by the involvement of one or more of the three components, each of which can be demonstrated independently of the illusion. It is suggested that other classical illusions, including the Müller-Lyer, might also be analysed in terms of basic components.

Miiller-Lyer first drew attention to differences between the apparent length of lines bounding angles of different sizes. This simple form of his illusion was investigated in Experiments 1 and 2 in which the angle between a horizontal 50-mm test line and an attached inducing lihe was varied between 0 and 180° in 150 steps. With acute angles the test line Was underestimated relative to a plain line and with obtuse angles overestimated. Maximum effects were found for angles of about 30 and 150° with the magnitude of overestimation exceeding that of underestimation. In Experiments 3 and 4 the contribution of the angle effect in more complex figures was investigated using figures consisting of one, two and four angles, the last being the conventional Miiller-Lyer figures. In acute angle figures (Experiment 3) the illusion appeared to be due to two components, that due to angle and that due to the longitudinal space enclosed between the obliques but in obtuse angle figures it appeared to derive from the additive effect of angles.The implications of these findings for an explanation of the Miiller-Lyer illusion are discussed.

Using a stationary target and moving field, both consisting of gratings of vertical light and dark bars, Over and Lovegrove (1973) reported that, with monoptic viewing, induced target movement is weaker when the light bars of the two components are different in color. This reduction did not occur for dichoptic viewing, for which the aftereffect was almost negligible. Six experiments are described. The effect of different colors was not confirmed, using a stationary point and moving frame or stationary and moving gratings. Reduced effects for different colors and greatly reduced effects for dichoptic viewing occurred only when there was a stationary boundary to the moving bars of the field grating, as in Over and Lovegrove's experiment. It is concluded that the effect studied by Over and Lovegrove is not the classical induced movement described by Duncker (1929Duncker ( /1938 but one due to periodic coincidence and noncoincidence of moving and stationary bars in grating patterns. This effect is absent when target and field bars are rendered more distinguishable by different colors.

Four experiments on induced movement and induced stationariness are described. Experiments 1 and 2 showed that mere enclosure of a stationary spot in a moving frame does not necessarily result in induced movement. Nor does enclosure of a moving spot in a stationary frame necessarily result in perceived real movement of the spot. Duncker's principles of enclosure is thus called into question. Two further experiments showed that both induced and perceived real movement of a spot are much more frequent when the frame is replaced by either two or more similar spots which enclose or flank the target spot. It can be concluded that the principle of enclosure obtains when the reference field consists of more than one element which move or remain stationary together. When such a field moves, it is the single, enclosed element which appears to move while the field itself appears stationary.It was shown earlier (Day, 1978) that when one of two small spots of light moves slowly in an otherwise dark field, either of the spots is perceived to move. Which of the two appears to move is a matter of chance. Occasionally it's the moving spot that appears to move, and occasionally it's the stationary spot. This state of affairs is changed by the addition to the display of a surrounding frame. When the two spots are enclosed in a stationary frame, perception is veridical; the moving spot is almost always seen to move and the stationary spot to remain stationary. However, when the frame moves in the same direction and at the same speed as the moving spot perception is nonveridical; the stationary element is usually seen to move in the opposite direction to the frame (induced movement) and the moving element to remain stationary (induced stationariness). These states of affairs obtain when the speed of the moving spot or frame does not exceed the subject-relative threshold, the threshold for movement of an object relative to the observer in an otherwise completely dark field (Kinchla, 1971).1 When the speed of the moving elements of the display progressively exceeds this threshold, veridical perception is increasingly common.These outcomes confirmed that for speeds and intervals below the subject-relative threshold, displacement of one spot relative to another in a dark field is ambiguous (Day, 1978;Kinchla, 1971;Wallach, 1959).