Perception

Subjects listened to descriptions of familiar objects and the spatial relations between them and attempted to visualize these objects as a composite scene. The time taken to generate a composite visual image, measured from the end of the verbal description, increased dramatically and linearly with the number of objects in the scene. As this relationship was independent of the rate of presentation of the verbal description, over a wide range, it could not be due to the visualization process cumulatively lagging behind the verbal description. Our evidence indicates that this relationship is due to some process, necessary to the subjective visualization of a composite array as a whole, which commences only after the verbal description is complete.

Three studies are reported with children aged to years and also with educationally subnormal children, in which it was found that up to about 7 years children's selection of one of a pair of line stimuli as ‘falling over’ is affected by the nature of the visual surround contours provided. Both the form of the stimulus surround and the presence or absence of a baseline tilted 45° relative to the horizontal were found to affect the children's judgements. The plane (horizontal or vertical) in which the stimuli were presented had no effect on the results.

It is argued that while the results show the influence of the visual surround on children's comprehension of ‘falling over’, this may not be wholly explicable in terms of surround contour matching, as conventionally correct judgements were obtained in the absence of all straight line contours in the immediate surround.

Piaget and Inhelder showed that children do not realize that the surface of a fluid remains horizontal in a tilted vessel. Several studies have since shown that many adults do not have an adequate concept of the water-level principle. However, in all these studies, drawings of vessels, or other abstract displays, were used. The present experiment is an investigation of whether adults who do not know the water-level principle are able to recognize the correct orientation of a fluid surface in realistic three-dimensional scenes and in cinematographic sequences. It was found that all subjects who could state the principle clearly, could precisely and accurately recognize the correct fluid level. More than half the subjects did not know the principle and all these subjects showed evidence in their judgments of only the crudest perceptual schema.

Spatial-frequency masking was studied with briefly pulsed (25 ms) vertical gratings. The mask was a noise grating, and the test pattern was a sinusoidal grating. A low-frequency band of noise masked a low- but not high-spatial-frequency test grating when the patterns were presented simultaneously. A high-frequency band of noise did not mask a low-frequency test grating when the patterns were presented simultaneously or when the mask was presented after the test pattern (backward masking). Masking was, however, observed when the mask or test pattern was of sufficiently high contrast so that the stimuli had nonlinear distortion and thus produced DC shifts of the field luminance.

The encoding of simple features such as lines and angles leads naturally to size and position invariance when these features are analyzed symbolically in terms of structure and connectedness. No such analysis is possible for a spatial frequency encoding, however, as all a priori indications of pattern unity are lost. An additional transformation of spatial frequency information is therefore demonstrated that automatically derives a size invariant representation of form. The feasibility of the transform scheme in the visual system is outlined with a hypothetical processing sequence. It is found that the organization of cells in the striate cortex shows a marked similarity to an intermediate stage of the proposed sequence.

Adult subjects were instructed to select either the wider physical or the wider perspective shape of two rotated rectangles. With equal differences between the physical and perspective widths, responses were governed by instructions and by the magnitude of the differences between widths of the forms. However, with unequal differences between physical and perspective widths of the two rectangles, responses were governed primarily by the magnitude of the differences. Regardless of instructions, responses were to the dimension with the greater difference between widths. The results suggest that performance in the shape constancy experiment is a function of relative dimensional discriminability, as defined by instructions, and relative differences of widths between forms presented for comparison.

Illusory contour figures can arise in figures that do not require completion, and hence no appearance of overlap.

Frisby and Clatworthy have suggested that subjective contours depend on special contrast effects acting at the tips of lines and at right angles to the lines. Their suggestion depends in part on the absence of such contours in dot figures. However, if the dots are grouped regularly and make a ‘good figure’, strong subjective contours appear, including a pattern comparable to Kanizsa's triangle. The kind of contrast explanation required for these figures would be one in which individual dots create small contrast effects which are only manifest in perception when the effects are grouped together, by Gestalt form indicators, overlap cues, or the like.

The superiority of binocular vision over monocular vision has been compared for the detection of stationary sinusoidal grating patterns, and for the detection of the apparent movement induced by rapidly phase-reversing such gratings. The thresholds for binocular and monocular pattern perception were in the ratio 1: 2 ^1/2, as found by previous workers. For apparent movement, however, binocular thresholds were lower than monocular thresholds by a factor of 1 ·9; for every subject tested (n = 20) the ratio for movement detection was larger than the ratio for pattern detection. The effects of combining inputs from the two eyes cannot be explained solely by linear summation models, but may in some circumstances depend on the nonlinearities of certain types of nerve cell.

Interocular transfer of the adaptation aftereffect was measured and correlated with residual binocular function for two strabismic amblyopes who represent different types of neural loss. The results indicate that for these two subjects there is paradoxical correlation between efficacy of interocular transfer and residual binocular function.

It is hypothesized that a negative correlation exists between the readiness with which two visual stimuli display rivalry and the magnitude of the inhibition effects between the corresponding neural channels. With binocular rivalry being more readily observed than monocular rivalry, it is predicted that lateral inhibition between neural channels selectively sensitive to such fundamental parameters as orientation, is primarily confined to those monocularly driven channels deriving their sensory input from the same eye. In confirmation, it is shown that the visual tilt illusion, thought to reflect lateral inhibition between orientation-sensitive channels, is very much reduced under dichoptic viewing conditions. Moreover, it is shown that those subjects displaying the greatest interocular transfer of the illusion least readily experience binocular rivalry.

Magnitude estimates were obtained for the apparent lengths of (i) seven vertical lines of different lengths at eleven positions along the horizontal meridian and (ii) the same seven lines presented horizontally at ten different angles of elevation. In experiments 1 and 2 (experienced and naive subjects, respectively) the seven vertical lines were permutated randomly in the eleven azimuth positions within a session. In experiment 3, the seven vertical lines were all presented first at a single azimuth position before being presented at another azimuth position. In experiment 4, the same seven line lengths (presented in horizontal orientation) were permutated randomly within the ten angles of elevation. The results showed that the apparent length of a line decreases as the line was moved away from the midline position into the periphery. Power functions described the growth of line lengths at each azimuth position and each angle of elevation. The exponents of the power functions in experiments 1,2, and 4 were smallest at the midline position and became larger for the line lengths presented more peripherally. For lines presented vertically the exponents tended to decrease again at the more extreme azimuth positions. Equal-length contours were derived from the power functions of experiments 1, 2, and 4.

A new optical—geometrical illusion is described. The parallelism of short rows of dots is affected by some unknown factor, so that the rows appear as pivoting on their middle point. Some explanations of the illusion are considered, but with no success.

Tilt aftereffects induced by line gratings are maximal when the test and inducing stimuli have an angular separation of 10°–15°. Similar effects induced by prismatically tilted real-object scenes have been found to increase monotonically with scene tilt. The difference between these two angular functions of aftereffect (that induced by gratings and that induced by real objects) has been attributed to the ‘meaning’ inherent in the real-object scenes.

The preliminary experiments described here suggest that tilt aftereffects and illusions induced by projected slides of tilted real-object scenes have angular functions similar to that induced by a line grating. Hence, the monotonically increasing angular function obtained in the prism studies is not necessarily determined by the use of real-object scenes.