Faces, vases, subjective contours, and the McCollough effect

Précis of Sensory Analysis

Sensory analysis is that initial, preconscious stage of perception at which primitive features (edges, temporal discontinuities, and periodicities) are picked out from the random fluctuations that characterize the physical stimulation of sensory receptors. Sensory analysis may be studied by means of signal-detection, psychometric-function, and threshold experiments, and Sensory Analysis presents a succinct, quasi-quantitative account of the phenomena revealed thereby. This account covers all five sensory modalities, emphasising the similarities between them.

A succinct account depends on identifying simple principles of wide generality, of which the most fundamental are that (a) sensory discriminations are differentially coupled to the physical stimuli and that (b) small stimuli are subject to a square-law transform which makes them less detectable than they would otherwise be. These two principles are established by comparisons between different configurations of two stimulus levels to be discriminated; they are realized within a simple physical-analogue model which affords certain low-level comparisons with neurophysiological observation. That physical-analogue model consists of a sequence of elementary operations on the stimulus constituting a stage of sensory processing. The concatenation of two or three stages in cascade accommodates an increased range of experimental phenomena, especially the detection of sinusoidal gratings.

This précis is organized in three parts: Part I surveys Sensory Analysis as economically as may be, beginning from the simplest, most fundamental ideas and working toward phenomena of increasing complexity. A rather shorter Part II reviews the most important alternative models addressed to some part or other of the phenomena surveyed. Finally, a very short Part III contributes some metatheoretic remarks on the function of a theory of sensory discrimination.

Color constancy and the functional significance of McCollough effects

A central problem in visual perception concerns how humans perceive stable and uniform object colors despite variable lighting conditions (i.e. color constancy). One solution is to 'discount' variations in lighting across object surfaces by encoding color contrasts, and utilize this information to 'fill in' properties of the entire object surface. Implicit in this solution is the caveat that the color contrasts defining object boundaries must be distinguished from the spurious color fringes that occur naturally along luminance-defined edges in the retinal image (i.e. optical chromatic aberration). In the present paper, we propose that the neural machinery underlying color constancy is complemented by an 'error-correction' procedure which compensates for chromatic aberration, and suggest that error-correction may be linked functionally to the experimentally induced illusory colored aftereffects known as McCollough effects (MEs). To test these proposals, we develop a neural network model which incorporates many of the receptive-field (RF) profiles of neurons in primate color vision. The model is composed of two parallel processing streams which encode complementary sets of stimulus features: one stream encodes color contrasts to facilitate filling-in and color constancy; the other stream selectively encodes (spurious) color fringes at luminance boundaries, and learns to inhibit the filling-in of these colors within the first stream. Computer simulations of the model illustrate how complementary color-spatial interactions between error-correction and filling-in operations (a) facilitate color constancy, (b) reveal functional links between color constancy and the ME, and (c) reconcile previously reported anomalies in the local (edge) and global (spreading) properties of the ME. We discuss the broader implications of these findings by considering the complementary functional roles performed by RFs mediating color-spatial interactions in the primate visual system.

Thalamocortical dynamics of the McCollough effect: boundary-surface alignment through perceptual learning

This article further develops the FACADE neural model of 3-D vision and figure-ground perception to quantitatively explain properties of the McCollough effect (ME). The model proposes that many ME data result from visual system mechanisms whose primary function is to adaptively align, through learning, boundary and surface representations that are positionally shifted due to the process of binocular fusion. For example, binocular boundary representations are shifted by binocular fusion relative to monocular surface representations, yet the boundaries must become positionally aligned with the surfaces to control binocular surface capture and filling-in. The model also includes perceptual reset mechanisms that use habituative transmitters in opponent processing circuits. Thus the model shows how ME data may arise from a combination of mechanisms that have a clear functional role in biological vision. Simulation results with a single set of parameters quantitatively fit data from 13 experiments that probe the nature of achromatic/chromatic and monocular/binocular interactions during induction of the ME. The model proposes how perceptual learning, opponent processing, and habituation at both monocular and binocular surface representations are involved, including early thalamocortical sites. In particular, it explains the anomalous ME utilizing these multiple processing sites. Alternative models of the ME are also summarized and compared with the present model.

Motion after-effects and word contingency

Stimulus-selectivity in phenomena such as the McCollough effect and other contingent after effects are controversial. Word specific McCollough effects have been reported (Allan et al., Percept Psychophys 1989;45:104-113) that suggest an associative model rather then a neural one. However, failures to replicate make this finding controversial (Humphrey et al., J Exp Psychol: Gen 123:86-90). We applied the same contingency to the motion after-effect. Moving words, words paired with sine wave gratings and words composed of sine wave gratings failed to generate text contingent after-effects in stimulus situations that normally evoke motion after-effects. Thus, there was little evidence that motion adaptation can be made textually contingent.

Orientation and color processing for partially occluded objects

The McCollough orientation-contingent color aftereffect could be equally well elicited by either a full test pattern of black and white stripes or a similar test pattern that was largely occluded by white surfaces, provided the latter stripes were made to appear as through continuing under the white surfaces--by means of stereo depth cues. The color aftereffect appeared concentrated around the edges of the stripes that protruded out from under the white surfaces; surfaces that themselves continued to appear a uniform white as shown by color matches. These results suggest that occluded, perceptually-continued edges can elicit the McCollough effect, which is generally thought to occur quite early in the visual pathway.

The McCollough effect: dissociating retinal from spatial coordinates

Three experiments were conducted to dissociate the perceived orientation of a stimulus from its orientation on the retina while inducing the McCollough effect. In the first experiment, the typical contingency between color and retinal orientation was eliminated by having subjects tilt their head 90 degrees for half of the induction trials while the stimuli remained the same. The only relation remaining was that between color and the perceived or spatial orientation, which led to only a small contingent aftereffect. In contrast, when the spatial contingency was eliminated in the second experiment, the aftereffect was as large as when both contingencies were present. Finally, a third experiment determined that part of the small spatial effect obtained in the first experiment could be traced to hidden higher order retinal contingencies. The study suggested that even under optimal conditions the McCollough effect is not concerned with real-world properties of objects or events. Implications for several classes of theories are discussed.

Orientation-contingent color aftereffects mediated by subjective transparent structures

We examined whether the orientation-contingent color aftereffect (the McCollough effect) could be mediated by subjective horizontal and vertical structure induced by the perception of transparency. In our experiments, red vertical bars and green horizontal bars were alternated as an adapting stimulus. After adaptation, subjects (n = 6) were asked to adjust the green and red saturation of a test pattern until they obtained a neutral gray. Horizontal and vertical stripes were combined in the test pattern in three different ways: (1) overlapping with a luminance combination that gave rise to a perception of transparent overlays of horizontal and vertical stripes (valid transparency condition), (2) overlapping with luminance combinations that did not induce a perception of transparency (invalid transparency condition) and that appeared more as a patchwork of checks, and (3) presented in adjacent, nonoverlapping areas. Our results showed that the McCollough effect was significantly greater in the valid transparency condition than in the invalid transparency conditions. The effect in the valid transparency condition was nevertheless less strong than was the effect seen with the standard test stimulus made up of nonoverlapping vertical and horizontal stripes. Our results suggest that the McCollough effect can be mediated by the subjective spatial organization (inner representation of vertical and horizontal stripes) that accompanies the perception of transparency in our stimulus.

Subjective contours 1900-1990: research trends and bibliography

A bibliography on subjective contours and a brief summary of trends in research on this problem are presented. The bibliography covers the years 1900-1990 and contains 445 entries, each briefly annotated with a code that indicates the general content and theoretical orientation of the item.

Anomalous contours and illusion of angularity: phenomenal and theoretical comparisons

Many experimental comparisons between real and anomalous contours have proven the functional equivalence of the two conditions; however, there are some contradictory findings. One of these is obtained by analyzing the anomalous contours in the light of a new illusion, called the 'illusion of angularity'. A circle becomes a polygon when it covers the centre of a radial arrangement of black stripes, and a polygon changes its perceptual shape depending on its orientation with respect to the same radial arrangement. Phenomenally, it appears like a very pointed polygon, in which every side is concave or, alternatively, a shape that looks like a circle with angles added in the spaces between the radial stripes, or a polygonal shape in which every side is convex. The reciprocal anomalous counterparts of these conditions, obtained by removing the geometrical/polygonal contours, reveal different results. In the first case, one sees a perfect circle; in the second case, a polygon with blunted vertices, or a circular shape with angular protrusions; in the third case, a deformed circle. These results are inconsistent with some theoretical models proposed to explain the emergence of anomalous contours, namely, all the top-down models expressed in terms of cognitive constructions and perceptual hypotheses, or in terms of global figural organizations. Rather, these comparisons suggest a different interpretation for the two phenomena (the illusion of angularity and anomalous contours). This interpretation is based on dynamic interactions or on network computations that synthesize both real and anomalous contours.

Color aftereffect contingent on text

During adaptation, two different letter strings (each five or six letters) were presented to subjects alternately, one in green and the other in magenta. The extent to which these letter strings subsequently elicited a color aftereffect was assessed. In different experiments, the chromatic letter strings consisted of words and nonwords. The results indicated that letter strings that form English words can contingently elicit a color aftereffect. This was the case even when the words were anagrams. There was no evidence that nonword letter strings could contingently elicit such an aftereffect, even when the nonwords conformed to English orthography. The results are relevant to understanding other contingent color aftereffects (McCollough effects), illusory color noted by computer operators who work at monochrome (green or amber) displays, and the processing of text.

Perceptual organization and apparent brightness in subjective contour figures

According to a number of theories subjective contours arise from brightness contrast and/or assimilation. The apparent brightness gradients generated by these effects are assumed to give rise to the perception of contours delineating the gradients. A study is reported in which naive observers were shown a subjective contour display and asked to report what they saw. They were then asked to judge whether the center or the surround of the display appeared brighter. Subjects whose reports indicated that they had seen the subjective contour figure showed an overwhelming preference for the center of the display being brighter than the surround. However, subjects who did not see the subjective contour figure did not differ significantly in their selection of the center over the surround. This finding presents difficulties for any theory which derives subjective contours from the apparent brightness difference.

Evidence for color-contingent organizational aftereffects

An effort was made to produce negative organizational aftereffects in response to a perceptually ambiguous stimulus that can be seen as a cube in either of two orientations. 6 subjects were adapted to alternative disambiguated versions of the ambiguous cube, shown either in green or magenta light. When tested with the ambiguous figure shown in either green or magenta light, 5 subjects showed negative aftereffects while one subject showed a positive aftereffect. The aftereffects showed an average change in preferred organization of 25%. The results are interpreted in terms of both cognitive and hard-wired processes.

The effects of illumination level and retinal size on the depth stratification of subjective contour figures

The apparent stratification in depth of subjective contour figures over their backgrounds was investigated as a function of illumination level, figure size, and viewing distance. Magnitude estimation, with a real contour figure serving as the modulus, was used to measure the stratification in depth of a subjective contour figure over its background. Illumination level and retinal size both had significant effects on the depth stratification of the subjective contour figures. The greatest apparent depth differences were obtained for figures of small retinal size under low levels of illumination. These results paralleled previous findings for judgments of subjective contour strength. Consequently, both contour clarity and depth stratification of subjective contour figures are affected in similar ways by illumination level, figure size, and viewing distance. The implications of this response coupling are discussed in terms of current theories of subjective contours.