Wallach was not a theorist, and he did not organize his research around an overarching theoretical system. He described his style of working as "pursuing a problem as long as the work yielded worthwhile results, and then shelving it until a new idea promised progress." His studies span a broad range of psychological topics, including the following:
Movement of lines behind apertures Wallach's doctoral dissertation examined perception of lines moving behind an opening in a masking surface – a phenomenon known as the
aperture problem. If a slanted line moves behind such an aperture, the physical stimulus presented to the eyes will not distinguish whether the movement is horizontal, vertical, or at some other angle. Wallach found that the motion the observer perceives is determined by the shape of the aperture. If the aperture is circular, the line (or lines) will appear to move in a direction perpendicular to their orientation. But if the aperture is rectangular, the lines will seem to move in a direction parallel to the long axis of the aperture. An example of this phenomenon is the familiar
Barberpole illusion. Wallach explained this finding by asserting that the perceptual system tends to preserve the individual identity of the line segments defined by the end points created by the aperture, and that this mode of movement best preserves that identity. Because the original paper was in German, this research was not well known to English-speaking psychologists for several decades. In 1976, Wallach published an English summary of his dissertation experiments,. Köhler and Wallach presented a series of experiments on figural after-effects. If, for example, an observer stares for about a minute at a fixation point in the center of a visual field that is white except for a large black rectangle on the left side, and then (with the rectangle removed) looks at the center of an array of four evenly-spaced squares, symmetrically arranged around the fixation point, the two squares on the left side will appear farther apart than the ones on the right. Many similar observations are discussed in the Köhler and Wallach paper. Köhler considered that this phenomenon supported his theory of
psychophysical isomorphism – that the perception of forms is mediated by electrical fields on the cortex of the brain, fields which he thought were isomorphic to the stimulus but which could be distorted through a process of satiation. Wallach explored the ability of humans to locate sounds in the median plane – that is, to determine whether a sound comes from a source at the same elevation as the ears or from a source that is higher or lower, or even in back of the head. Binaural sound cues, including the phasing or time of the sound's arrival at each ear and the sound's relative intensity at the two ears (known respectively as ITD and ILD) enable a listener to determine a sound's lateral location (whether it is on the left, right, or straight ahead). But two sounds at different elevations can present identical ITD and ILD information to the ears, and so binaural cues to a stationary ear do not suffice to identify a sound's location in the median plane. Their experiments demonstrated that when a localizable sound reaches the ears and is immediately followed by an identical sound coming from a different direction, the listener perceives a single sound at the location of the first-arriving stimulus. The delay between the first-arriving and the second-arriving sound can be in the range of 1 to 5 ms for clicks, and as much as 40 ms for complex sounds such as speech and music. At delays above these thresholds, the second sound is heard as an echo. This phenomenon illustrates how the auditory system suppresses local reverberations to enhance the intelligibility of perceived sounds and it is a critical factor in acoustical engineering and design of sound reinforcement systems. Wallach et al. also noted that the precedence effect plays an important part in perception of stereophonic sound. Wallach explored the stimulus conditions for the perception of neutral colors – that is, colors that vary in lightness but have no hue, thus ranging from white to gray to black. Wallach projected round patches of light (“disks”) of various brightnesses on a white screen in a dark room and found that, when presented alone, the disks always appeared to be luminous – i.e. they seemed to be emitting light, just as the moon appears when it is high in a dark sky. However, when a surrounding ring of a different brightness was added to such a projected disk, the disk ceased to appear luminous and looked like a patch of smooth paper whose color depended on the relative brightnesses of the central disk and the surrounding ring .If the surround was less bright than the disk in the center, the disk appeared white. If the surround was brighter than the center, the central disk appeared to be a shade of gray. The shade of gray depended on the brightness ratio of the center to the surround, regardless of the absolute luminance levels of the two elements in the display. Thus, for example, a disk with a physical luminance of 50 millilamberts (mL) surrounded by a ring of 200 mL would seem to be the same shade of gray as a disk of 500 mL surrounded by a ring of 2000 mL. Wallach proposed that this "ratio principle" could explain the phenomenon of lightness constancy – the fact that an object's apparent lightness remains constant despite large variations in illumination. In subsequent years, a large body of literature has explored the adequacy and limitations of the ratio principle. The ratio principle does not hold if the luminance ratio is extremely high; or if the two interacting luminances are not adjacent. Furthermore, Wallach's highly simplified experimental setup does not deal with three-dimensional spatial arrangements nor with complex visual fields that include many interacting luminances. Rather than providing a complete solution to the problem of lightness constancy, Wallach's 1948 paper served to "set the stage for computational models of lightness perception". This phenomenon illustrates how the visual system processes displays of dynamically-changing elements so that we perceive a world of rigid objects arranged in space. If a stationary three-dimensional figure (for example, a wireform cube) is illuminated from behind so that its shadow falls on a translucent screen, an observer in front of the screen will see a two-dimensional pattern of lines. But if the same object is rotated, the observer will (accurately) see it as a turning three-dimensional cube, even though only two-dimensional information is presented. This is the
kinetic depth effect (KDE), a potent
depth cue. It occurs spontaneously, it can be seen with monocular vision, it occurs with solid figures as well as wireforms, and the figures need not be regular geometric objects nor need they have familiar shapes. Wallach and O’Connell found only two essential conditions for obtaining the effect. The object must be composed of straight lines with definite endpoints or corners, and the projected shadows of those lines must change in both length and orientation as the object rotates (otherwise a flat, deforming figure is seen.) Others sought to build theoretical models of the essential conditions for dynamically representing rigid three-dimensional objects using only two dimensions, leading to development of a new field of study:
structure from motion, a part of the domain of
cognitive science. Practical applications have included representing the third dimension in computer displays, palmtop devices, and airport security scanners.
Adaptation in perception of depth and distance Since a human's two eyes are approximately 6.5 cm apart, they see the world from different viewpoints: the image projected on the left retina is slightly different from the image projected on the right. This difference (known as
binocular disparity) is the fundamental cue underlying
stereoscopic depth perception . The importance of stereoscopic perception is familiar to anyone who has ever attempted to thread a needle with one eye closed; and when two slightly disparate photographs are viewed through a
stereoscope (a device that makes it easy to fuse the two images), the fused scene takes on a three-dimensional appearance. In 1963, Wallach, Moore and Davidson artificially increased disparity by having subjects look through a telestereoscope, a device which uses a mirror arrangement to simulate an increased distance between the eyes. When they viewed a wireform cube through the telestereoscope, subjects reported that the cube's depth appeared greater than its width and height, a result of the greater disparity generated by the telestereoscope. After this, the cube was made to rotate slowly while the subjects watched. This created a conflict between two depth cues: while the artificially-increased disparity was indicating that the cube's depth was greater than its other two dimensions, the kinetic depth effect (which is not affected by disparity) was presenting cues consistent with a normal rotating cube, of equal size on all sides. After watching the rotating cube in this cue-conflict situation for a 10-minute adaptation period, subjects were again showed the stationary cube (still through the telestereoscope) and asked to indicate its depth. They reported less apparent depth than before the adaptation period, indicating that conflict from the competing KDE cue had modified the way the visual system interpreted the stereoscopic depth cues. The altered perception of depth was temporary: it could easily be unlearned (by watching the cube rotate without the telestereoscope), and the effect dissipated spontaneously after a few minutes, even with if the subjects simply sat with eyes closed during that time. Subsequently, Wallach and Frey performed similar experiments creating a conflict among different cues that the visual system uses to compute the distance of an object from the observer. Two such cues are
accommodation (adjustments of the eye's lens to bring near or far objects into focus) and
convergence (the inward turning of the eyes necessary to fixate on near objects). These two cues together are called oculomotor cues. Other cues also play a part in distance perception; among these are perspective, texture gradients, and motor cues (when we reach out to touch an object, we acquire information about how far away it is.) Wallach and Frey constructed special goggles that artificially distorted oculomotor distance cues, such that the wearer would see objects with accommodation and convergence cues appropriate to distances closer than the objects’ actual distances. The subjects wore the glasses while physically manipulating a set of small wooden blocks set on a table, and thus perspective, texture and motor cues gave veridical information. After 15 minutes of adaptation, tests showed that subjects (now without the goggles) registered the distance of test objects as being farther than their objective distances. A different set of goggles, simulating the oculomotor cues for distances greater than veridical, yielded the opposite result. These findings – that exposure to cue-conflict situations modifies the way in which the visual system evaluates cues – represented a definite step away from the Gestalt tradition in which Wallach was trained. Gestalt psychologists preferred to explain perceptual phenomena through the characteristics of the stimulus complex taken as a whole, and through inborn, invariant functions of the perceptual system. They generally downplayed the role of experience and adaptation. devised a mechanical apparatus that enabled head movements to cause displacements of an image by any desired percentage of the extent of that head movement, and discovered that subjects could detect deviations of as little as 2% from the normal degree of displacement. This showed that a highly accurate compensating process corrects for the image displacement that normally accompanies a head movement, thus yielding an appearance of stability. Wallach called this process
constancy of visual direction (CVD), and he noted with interest that it could be easily modified through perceptual adaptation. To demonstrate this, Wallach & Kravitz set the apparatus so that during head movements the visual image moved by 150% of what would be the normal displacement, and had subjects turn their heads back and forth watching this altered displacement for 10 minutes. After this brief adaptation period, subjects were shown an objectively stationary target as they turned their heads. They reported that it no longer appeared motionless, but swung back and forth in the direction opposite the movements that had occurred during adaptation, In order to make the target appear stationary, the apparatus had to be set so that the target actually moved by about 14% in the same direction it had moved during the adaptation period. The CVD process that correlates head movements and image shifts had been modified by exposure to an abnormal stimulus condition. (The adaptation of the CVD process was temporary and dissipated after a few minutes.) As in the case of depth and distance perception, Wallach's finding that the constancy of visual direction adapts readily when stimulus conditions are altered represented a marked departure from the Gestalt tradition, which focused on innate and unmodifiable processes. In fact, Wallach came to regard adaptation as an analytical tool in itself. For example, Wallach & Bacon were able to demonstrate that two distinct processes are involved in the constancy of visual direction by showing that they adapt differently. In addition to the processes compensating for image displacements during head rotation, Wallach and various collaborators examined other sorts of compensations related to perceptual stability during bodily movement, including displacements caused by nodding and by eye movements, the changing orientation of objects as one walks past, optical expansion caused by moving forward, displacement in a dimension unrelated to the physical movement, and movement-correlated alterations in form perception. ==Teaching and impact on students==