Often, peripheral explanations are the first to be proposed. In my opinion, at least, they are usually wrong and they die hard. Because such an explanation will generally be simpler than a central one, it is good science to begin an investigation by testing it. But it is bad science to cling to that explanation in the face of strong evidence to the contrary. Explanations in terms of eye movements have always been popular. Thus, in the domain of motion perception, such phenomena as apparent motion, induced movement, and the autokinetic effect (see Chapter 7) have all been held to be caused by eye movements at one time or another. Max Wertheimer, in his famous research on apparent movement that launched the Gestalt approach, found it necessary to perform a special experiment to demonstrate that such motion could not be explained by eye movements. He arranged for one object to appear to move leftward simultaneously with another object appearing to move rightward. The eyes cannot move in opposite directions at the same time. Eye movements have been invoked as the explanation of the origin of form perception, of depth perception based on retinal disparity, and of the perceived direction of all regions in the field, an example of which is the uprightness of vision despite the inverted orientation of the retinal image. As a contemporary example of this controversy, consider a phenomenon not yet mentioned. If a figure is moved behind a narrow, stationary slit in an opaque surface so that all but the segment behind the slit is occluded at any given moment, one still tends to perceive the entire figure. Helmholtz and others studied this effect in the last century, and Theodore Parks of the University of California at Davis rediscovered it in this century. It is now referred to as anorthoscopic perception—–an abnormal way of presenting something. How is such perception possible? Isn’t an extended, simultaneous retinal image of a figure necessary for the perception of its shape?