Color-Binding Errors During Rivalrous Suppression of Form

ABSTRACT— How does a physical stimulus determine a conscious percept? Binocular rivalry provides useful insights into this question because constant physical stimulation during rivalry causes different visual experiences. For example, presentation of vertical stripes to one eye and horizontal stripes to the other eye results in a percept that alternates between horizontal and vertical stripes. Presentation of a different color to each eye (color rivalry) produces alternating percepts of the two colors or, in some cases, a color mixture. The experiments reported here reveal a novel and instructive resolution of rivalry for stimuli that differ in both form and color: perceptual alternation between the rivalrous forms (e.g., horizontal or vertical stripes), with both eyes' colors seen simultaneously in separate parts of the currently perceived form. Thus, the colors presented to the two eyes (a) maintain their distinct neural representations despite resolution of form rivalry and (b) can bind separately to distinct parts of the perceived form.     

Orientation-specific aftereffects and illusions in the perception of brightness

Orientation-specific brightness aftereffects were found when vertical and horizontal gratings of the same space-average luminance were viewed following alternate exposure to vertical and horizontal gratings that differed in space-average luminance. The vertical test grating appeared bright following exposure to a dim vertical grating, and dim after a bright vertical grating had been viewed. This aftereffect did not occur when the adaptation gratings had been seen by one eye and the test gratings by the other eye. An orientation-specific illusion in the perception of brightness was also found, with the white sectors of a vertical grating appearing brighter against a background of horizontal lines than they did against a background of vertical lines. Both distortions imply that there are detectors in the human visual system that are conjointly tuned to luminance and contour orientation.     

How configurations of binocular disparity determine whether stereoscopic slant or stereoscopic occlusion is seen

A partially occluded contour and a slanted contour may generate identical binocular horizontal disparities. We investigated conditions promoting an occlusion resolution indicated by an illusory contour in depth along the aligned ends of horizontally disparate line sets. For a set of identical oblique lines with a constant width added to one eye's view, strength, depth, and stability of the illusory contour were poor, whereas for oblique lines of alternating orientations the illusory contours were strong, indicating a reliance on vertical size disparities rather than vertical positional disparities in generating perceived occlusion. For horizontal lines, occlusion was seen when the lines were of different lengths and absolute width disparity was invariant across the set. In all line configurations, when the additional length was on the wrong eye to be attributed to differential occlusion, lines appeared slanted consistent with their individual horizontal disparities. This rules out monocular illusory contours as the determining factor.     

Why do pitched horizontal lines have such a small effect on visually perceived eye level?

In two experiments, visually perceived eye level (VPEL) was measured while subjects viewed two-dimensional displays that were either upright or pitched 20 degrees top-toward or 20 degrees top-away from them. In Experiment 1, it was demonstrated that binocular exposure to a pair of pitched vertical lines or to a pitched random dot pattern caused a substantial upward VPEL shift for the top-toward pitched array and a similarly large downward shift for the top-away array. On the other hand, the same pitches of a pair of horizontal lines (viewed binocularly or monocularly) produced much smaller VPEL shifts. Because the perceived pitch of the pitched horizontal line display was nearly the same as the perceived pitch of the pitched vertical line and dot array, the relatively small influence of pitched horizontal lines on VPEL cannot be attributed simply to an underestimation of their pitch. In Experiment 2, the effects of pitched vertical lines, dots, and horizontal lines on VPEL were again measured, together with their effects on resting gaze direction (in the vertical dimension). As in Experiment 1, vertical lines and dots caused much larger VPEL shifts than did horizontal lines. The effects of the displays on resting gaze direction were highly similar to their effects on VPEL. These results are consistent with the hypothesis that VPEL shifts caused by pitched visual arrays are due to the direct influence of these arrays on the oculomotor system and are not mediated by perceived pitch.     

Representational bias does not affect bisection of lines with a pictorially or semantically defined top by patients with left hemispatial neglect

Normal subjects may have both representational and visual-based components determining bias in bisection of horizontal, vertical, and radial lines. The influence of these components is less clear in patients with neglect. We asked 25 patients with right hemisphere stroke and clinical features of neglect to bisect lines oriented horizontally, vertically, and radially above and below eye level. Objects including human silhouette figures, arrowheads, and the words 'TOP' and 'BOTTOM' were placed at either end of each line. These figures were presented either upright or upside down in some orientations, and presented rightward and leftward in other orientations, to pictorially or semantically define a "top" to each line independent of the actual top of the visual field. Patients demonstrated a rightward bias on all horizontal line bisections, with similar bias and greater magnitude than normal subjects. Patients also demonstrated visual-based biases on some of the vertical, radial down, and radial up lines presented. However, patients did not demonstrate a significant representational bias with any of the visual cues presented on any of the line orientations. Patients with acute right hemisphere stroke demonstrate a rightward bias when asked to perform line bisection in the horizontal line orientation, as well as an upward bias in vertical line bisection. The lack of representational bias in patients with neglect may be due to a greater degree of visual-based neglect as compared to representational neglect, or it may be due to an absence of representational bias in patients with right hemisphere stroke.     

Visually perceived eye level: changes induced by a pitched-from-vertical 2-line visual field.

The physical elevation corresponding to visually perceived eye level (VPEL) changes linearly with the pitch of a visual field. Deviations from true eye level average more than 0.5 times the angle of pitch over a 65 degrees pitch range. A visual field consisting of 2 dim, isolated vertical lines in darkness is more than 4/5 as effective as that of a complexly structured visual field; 2 horizontal lines have a small and inconsistent effect. Differences in influence on VPEL between pitched-from-vertical and horizontal lines were predicted from an analysis that extracted differences in retinal perspective resulting from changes in pitch. The Great Circle Model (GCM), based on a spherical approximation to the erect, stationary eye, predicts the present results and results of 8 other sets of experiments. The model treats the influence of a single line on VPEL as systematically related to the elevation of the intersection between the great circle containing the image of the line and the central vertical retinal meridian; generalized GCM combines visual inputs with inputs from the body-referenced mechanism and maps onto the central nervous system.     

Visual and haptic perception of the horizontal-vertical illusion

The horizontal-vertical illusion consists of two lines of the same length (one horizontal and the other vertical) at a 90 degree angle from one another forming either an inverted-T or an L-shape. The illusion occurs when the length of a vertical line is perceived as longer than the horizontal line even though they are the same physical length. The illusion has been shown both visually and haptically. The present purpose was to assess differences between the visual or haptic perception of the illusions and also whether differences occur between the inverted-T and the L-shape illusions. The current study showed a greater effect in the haptic perception of the horizontal-vertical illusion than in visual perception. There is also greater illusory susceptibility of the inverted-T than the L-shape.     

Why do children confuse mirror-image obliques?

Thirty-five children aged from 4.17 to 6.58 years were given a delayed-matching task in which they had to choose on each trial which of two lines was the same as a previously displayed standard line. Their choices were no better than random when the lines differed in degree of slope but not in left-right orientation and were only marginally more accurate when the lines were left-right mirror images. Performance improved significantly when the lines differed both in degree of slope from the vertical and in left-right orientation and improved still further if at least one of the lines was horizontal or vertical. The results suggest that young children have extreme difficulty encoding in memory either the degree of slope or the left-right orientation of an oblique line.     

Test of the effect of attention on judged length of a line

Studies to assess experimentally whether attention affects the judged length of a line have produced discordant results. This paper reports the results of a test designed to avoid factors that were not controlled in previous studies. Stimuli were either two vertical lines or two horizontal lines of equal physical length presented briefly on opposite sides of a fixation cross. Subjects were asked to direct their attention to one line when the arm of the cross pointing to the line changed in luminance. This arm was used either as a precue or as a postcue. Subjects judged lines to be longer when a precue preceded the line than when a postcue followed the line.     

A spatial illusion from motion rivalry

A new dynamic visual illusion is reported: contrast reversal of a horizontal and vertical plaid pattern (produced by adding two orthogonal sinusoidal gratings) causes the pattern to appear as an array of lustrous diamonds, cut by sharp lines into a diagonal lattice structure. On the basis of computer simulations it is suggested that the illusion results from rivalrous interaction of motion detectors tuned to opposing directions of motion.     

The apparent length of a line as a function of its inclination

In the first of two experiments reported here, subjects adjusted the length of a variable line until it appeared to be as long as a standard line. There were two sizes of standard line, 3 and 6 inches, and each was shown vertically for some trials and horizontally for others. The variable line was presented in each of the 10° positions from 0° (horizontal), through 90° (vertical), to 170°. The principal results of the first experiment are:

(1) Vertical lines look longer than horizontal lines of the same length, but lines tilted 20°-30° to the left of vertical look longer than lines in any other orientation. The results are asymmetrical, because lines tilted to the right of vertical do not look as long as those tilted to the left of vertical.

(2) The variability of the settings increases as the angle increases between the standard and variable lines.

(3) When they are expressed in percentage terms, the data obtained with the 3-and 6-inch standards are virtually identical, i.e. the data for the 3-inch standard can be made to match those for the 6-inch standard simply by doubling the former.

(4) There are enormous differences among subjects in the patterns of settings made at the various angles. A few subjects apparently experienced no illusory effects since they adjusted the variable line to about the same physical length irrespective of its orientation. Other subjects showed exaggerated overestimations of the variable line for vertical and near-vertical positions.

In the first experiment, the variable line was always to the left of the standard, and it was natural to assume that this position effect had somehow produced the asymmetry noted in paragraph 1 above. This hypothesis was tested in the second experiment which alternatively showed the variable line above, below, to the right of, and to the left of the standard line. The results of this experiment generally confirm the data of the first experiment in showing that lines tilted 20°-30° to the left of vertical look longer than lines tilted to any other position. In addition, the second experiment shows that this asymmetry in the results is not a function of the relative positions of the variable and standard lines. In general, however, overestimations of length are smaller when the two lines are one above the other, greater when the two lines are side by side.     

The horizontal-vertical illusion and knowledge of results

The Horizontal-Vertical (HV) Illusion was examined in two studies in which subjects adjusted the vertical line in L-shaped and inverted-T figures or produced lines in the vertical and horizontal planes. On the adjustment tasks, vertical lines were made significantly shorter than horizontal comparison lines, especially for the inverted-T figure. On the production tasks, lines drawn in the vertical plane were significantly shorter than lines drawn in the horizontal plane. The adjusted and created lines of subjects receiving intertrial feedback on illusion magnitude were significantly more accurate and less variable than the estimations of control subjects. Performance on either task or figure type did not differ as a function of sex of subject. The present results show that the HV illusion exists in the absence of line bisection or a comparison line and results from the overestimation of vertical lines. These findings further clarify the relative contributions of the structural and strategy mechanisms in the formation of the Horizontal-Vertical Illusion.     

Aftereffects in binocular rivalry

Five experiments are reported in which the aftereffect paradigm was applied to binocular rivalry. In the first three experiments rivalry was between a vertical grating presented to the left eye and a horizontal grating presented to the right eye. In the fourth experiment the rivalry stimuli consisted of a rotating sectored disc presented to the left eye and a static concentric circular pattern presented to the right. In experiment 5 rivalry was between static radiating and circular patterns. The predominance durations were systematically influenced by direct (same eye) and indirect (interocular) adaptation in a manner similar to that seen for spatial aftereffects. Binocular adaptation produced an aftereffect that was significantly smaller than the direct aftereffect, but not significantly different from the indirect one. A model is developed to account for the results; it involves two levels of binocular interaction in addition to monocular channels. It is suggested that the site of spatial aftereffects is the same as that for binocular rivalry, rather than sequentially prior.     

Integration of local features as a function of global goodness and spacing

In two experiments, the accuracy with which subjects detected a conjunction of features was examined as a function of the spacing between items and the goodness of the axis along which they were located. In each array, two items were arranged along a vertical, a horizontal, or a diagonal axis. Based on the well-established oblique effect, the vertical and horizontal axes were considered to be good global patterns and the diagonals were considered to be poor. In Experiment 1, the two items in an array could be two horizontal lines, two vertical lines, a vertical and a horizontal line, or a plus sign with one of the single lines. In Experiment 2, a positive- and a negative-diagonal line were used as the individual features, and an "X" was used as the conjunction. The results from Experiment 1 indicated that global goodness influenced only the rate of illusory conjunctions, and not of feature errors. Illusory conjunctions of vertical and horizontal line segments were more likely to occur in vertical and horizontal arrangements. The results from Experiment 2 revealed a reversal of the effect of global goodness on the rate of illusory conjunctions: Illusory conjunctions of negative- and positive-diagonal line segments were more likely to occur in diagonal arrangements. The results of both experiments taken together showed the existence of an important and new factor that influences the likelihood that features of shape will be conjoined: the ease with which line segments conjoin when they are translated along their extent toward each other. In both experiments, greater spacing between items produced more feature-identification errors and fewer feature-integration errors than did less spacing.     

Disparity minimisation, cyclovergence, and the validity of nonius lines as a technique for measuring torsional alignment

Frisby et al (1993 Perception 22 Supplement, 115) proposed that the visual system might make cyclovergent eye movements in order to minimise the overall pattern of both vertical and horizontal disparities when an observer views an inclined stereoscopic surface. Their measurements of cyclovergence, which used vertically oriented nonius lines, were found to be consistent with that proposal. In our experiment 1, we measured torsional eye movements objectively, using scleral coils, and found no evidence of a cyclovergent response to either a real inclined surface or to a simulated inclined surface in which the two stereoscopic images were related by a horizontal shear transformation. These results are inconsistent with the disparity minimisation hypothesis. In order to account for the discrepant findings of the two studies, we propose that vertically oriented nonius lines may not be a valid method for assessing cyclovergence because the lines can be seen as lying 'within' the inclined surface. In experiment 2, we tested the predictions of the cyclovergence hypothesis of Frisby et al against our own 'within surface' explanation, using both horizontally and vertically oriented nonius lines and dichoptic images related by either a horizontal or a vertical shear. If cyclovergence were the cause of the misalignment, both horizontal and vertical nonius lines should appear misaligned to the same extent. This was not found to be the case. We conclude that vertical nonius lines may not be a valid technique for measuring cyclovergence when the lines are seen against a background of an inclined surface.     

The visuomotor system resists the horizontal-vertical illusion

The effect of the horizontal-vertical illusion on the visual and visuomotor systems was investigated. Participants (N = 8) viewed horizontal and vertical lines in an inverted-T stimulus and judged whether the two line segments were the same or different lengths. Participants also reached out and grasped either the vertical or the horizontal line segment of the stimulus. Perceptually, participants succumbed to the illusion; that is, they judged Ts of equal horizontal and vertical line lengths to be different and Ts of unequal line lengths to be the same. When reaching toward the same stimuli, however, the size of their grip aperture was scaled appropriately for the various line lengths. Thus, whereas the perceptual system succumbed to the illusion, the visuomotor system did not. Those results support a model proposed by M. A. Goodale and A. D. Milner (1992), who posited separate cortical pathways for visual perception and visually guided action.     

The Visuomotor System Resists the Horizontal-Vertical Illusion

The effect of the horizontal-vertical illusion on the visual and visuomotor systems was investigated. Participants (N = 8) viewed horizontal and vertical lines in an inverted-T stimulus and judged whether the two line segments were the same or different lengths. Participants also reached out and grasped either the vertical or the horizontal line segment of the stimulus. Perceptually, participants succumbed to the illusion; that is, they judged Ts of equal horizontal and vertical line lengths to be different and Ts of unequal line lengths to be the same. When reaching toward the same stimuli, however, the size of their grip aperture was scaled appropriately for the various line lengths. Thus, whereas the perceptual system succumbed to the illusion, the visuomotor system did not. Those results support a model proposed by M. A. Goodale and A. D. Milner (1992), who posited separate cortical pathways for visual perception and visually guided action.     

Voluntary visual attention and phenomenal line length

In 1956, Fraisse, et al. reported subjects judged that lines were longer when voluntary attention was focused on the lines than when attention was distracted from the lines. In the many attempts to repeat these results, none has ascertained whether attention on reported line length was a phenomenal effect. In the present study, 46 subjects were shown as stimuli pairs of horizontal or vertical briefly flashed lines with a fixation cross placed equidistant between the lines but far from each one. A change in color of one arm of the cross was used as a cue to focus subjects' voluntary attention on one line. Analysis showed attention increased the judged length of attended lines. Since this effect of attention also occurred when subjects were absolutely certain they saw the stimulus lines differed in length, this effect indicates that attention increased the phenomenal length of the attended lines. This lengthening was quite small: it involved a maximum mean increase of about .15 in the probability of the comparative response that the attended line was longer. This effect occurred in the horizontal dimension and was almost absent in the vertical dimension. In agreement with data indicating that flashed lines expand phenomenally by activating motion detectors and that focused attention makes neural motion responses increase in amplitude, the present results suggest that focused attention makes attended lines look longer because it makes these lines expand phenomenally more rapidly.     

Attentional scanning and space errors

Using the method of paired comparisons, pairs of simultaneous horizontal or vertical lines, with one line above and one below or one on the left and one on the right of a fixation point, respectively, were presented tachistoscopically for length comparison. Space errors were found to have a pattern similar to that of time errors. The tendency to guess the comparative response from the absolute magnitude of stimuli is proposed as a basis for time and space errors. Manipulation of attentional scanning, which implies a more frequent usage of this guessing strategy for one of the two lines in a pair, was shown to affect space errors.     

Attentional scanning and space errors.

Using the method of paired comparisons, pairs of simultaneous horizontal or vertical lines, with one line above and one below or one on the left and one on the right of a fixation point, respectively, were presented tachistoscopically for length comparison. Space errors were found to have a pattern similar to that of time errors. The tendency to guess the comparative response from the absolute magnitude of stimuli is proposed as a basis for time and space errors. Manipulation of attentional scanning, which implies a more frequent usage of this guessing strategy for one of the two lines in a pair, was shown to affect space errors.