Perception

Any attempt to unravel the mechanism underlying the process of human face recognition must begin with experiments that explore human sensitivity to differences between a perceived image and an original memory trace. A set of three consecutive experiments are reported that were collectively designed to measure the relative importance of different facial features. The method involved the use of image-processing equipment to interchange cardinal features among frontally viewed target faces. Observers were required to indicate which of the original target faces most resembled the modified faces. The results clearly establish the dominant influence of the head outline as the major recognition feature. Next in importance is the eye/eyebrow combination, followed by the mouth, and then the nose. As a recognition feature in a frontally presented face, the nose is hardly noticed. The number of apparently random responses to some faces indicates that a surprisingly different face can sometimes arise from a fortuitous combination of the old features.

Thresholds for stereoscopic-depth perception increase with decreasing spatial frequency below 2.5 cycles deg⁻¹. Despite this variation of stereo threshold, suprathreshold stereoscopic-depth perception is independent of spatial frequency down to 0.5 cycle deg-¹. Below this frequency the perceived depth of crossed disparities is less than that stimulated by higher spatial frequencies which subtend the same disparities. We have investigated the effects of contrast fading upon this breakdown of stereo-depth invariance at low spatial frequencies.

Suprathreshold stereopsis was investigated with spatially filtered vertical bars (difference of Gaussian luminance distribution, or DOG functions) tuned narrowly over a broad range of spatial frequencies (0.15–9.6 cycles deg⁻¹). Disparity subtended by variable width DOGs whose physical contrast ranged from 10–100% was adjusted to match the perceived depth of a standard suprathreshold disparity (5 min visual angle) subtended by a thin black line. Greater amounts of crossed disparity were required to match broad than narrow DOGs to the apparent depth of the standard black line. The matched disparity was greater at low than at high contrast levels. When perceived contrast of all the DOGs was matched to standard contrasts ranging from 5–72%, disparity for depth matches became similar for narrow and broad DOGs. 200 ms pulsed presentations of DOGs with equal perceived contrast further reduced the disparity of low-contrast broad DOGs needed to match the standard depth.

A perceived-depth bias in the uncrossed direction at low spatial frequencies was noted in these experiments. This was most pronounced for low-contrast low-spatial-frequency targets, which actually needed crossed disparities to make a depth match to an uncrossed standard. This bias was investigated further by making depth matches to a zero-disparity standard (ie the apparent fronto-parallel plane). Broad DOGs, which are composed of low spatial frequencies, were perceived behind the fixation plane when they actually subtended zero disparity. The magnitude of this low-frequency depth bias increased as contrast was reduced. The distal depth bias was also perceived monocularly, however, it was always greater when viewed binocularly. This investigation indicates that contrast fading of low-spatial-frequency stimuli changes their perceived depth and enhances a depth bias in the uncrossed direction. The depth bias has both a monocular and a binocular component.

Adaptation and reaction-time techniques were used to examine the role of different spatial-frequency channels in the perception of local and global structure. Subjects were shown figures consisting of a large C composed of smaller Cs and asked to identify the orientation of either the global C or its local elements. Prior to performing the task subjects were adapted to different spatial frequencies and the effect on subsequent performance was assessed. Two main results were found. First, the adapting frequency that most affected the global task was often lower than that most affecting the local task, suggesting that high and low frequencies independently code the structure of an image. Second, reaction time to global figures was often faster than to local figures at all levels of detectability, again suggesting a role of low-frequency channels in global processing.

Patterns of vibrotactile stimulation were delivered to the index fingertips of naive blindfolded subjects. The attributions made by these subjects when they were allowed to experience transformations of vibrotactile stimulation correlated with self-movement were assessed. Although the subjects became aware of the relationship between self-movement and stimulation transformation, they never developed the hypothesis of distal attribution, ie the hypothesis that the ultimate cause of their vibrotactile experience was an encounter with an object in the environment. It is proposed that further investigations of the course of acquisition of distal attribution in the situation described may be instructive in the study of externalization in other modalities.

Cutaneous points on different fingers were brought into contact with each other to ascertain whether one stimulus was perceived to be single or double. The perceptual responses were mapped on the five fingertip pads. The results show that there is an extensive cutaneous area on each fingertip pad which elicits one-stimulus perceptions when a single stimulus is applied between each finger and the thumb. This area decreases when the stimulus is applied to adjacent fingers (2–3, 3–4, 4–5), and even more so when it is applied to nonadjacent fingers (2–4, 2–5, 3–5). In fact, in adjacent and nonadjacent fingers the cutaneous surface eliciting a doubling of the single stimulus (diplesthesia) is very extensive. The spatial arrangement of the cutaneous areas eliciting single perceptions appears to be invariant in the proximo-distal plane. Instead, the shift from one kind of perception to another occurs in the medio-lateral plane. It is suggested that this perceptual organization could reflect a neural organization.

Intermanual tactile recognition of laterally inverted mirror shapes was studied, with special reference to the role of the thumb. Children were allowed to feel the shapes either with the whole hand, with only four fingers (excluding the thumb), or with only the index finger. Intramanual recognition was also studied after rotation of the hand from the palm down/up to the palm up/down orientation. The thumb was found not to be important for intermanual mirror reversals, and only of limited importance for intramanual reversals. There was no evidence that coding with reference to the hand is of importance for either inter- or intramanual reversals. The explanations for the two kinds of reversals are quite different.

Quantitative estimates of the spatial discriminative capacities of the visual and kinaesthetic systems in adults and children were obtained. Intersensory integration was investigated by including spatial discriminations based on congruent visual plus kinaesthetic reafference. The psychophysical method of adjustment was used with simultaneous comparisons of a fixed and a variable stimulus. The subject's task was to estimate when the variable stimulus (ellipse) was identical to the standard one (circle), under one of three modality conditions: vision, kinaesthesis, and vision plus kinaesthesis.

After a pilot study with adults, children (aged 8, 10, and 12 years) and adults were both tested. Subjects from each age group were randomly allocated to each of the three experimental conditions. Results show that the visual and kinaesthetic estimates of the 8- and 10-year-old subjects did not differ significantly, but the visual responses of the adults and 12 year olds were significantly more accurate than corresponding kinaesthetic estimates. Bisensory estimates were significantly more accurate than visual responses only for the 8- and 10-year-old age groups. Intramodal comparisons showed the kinaesthetic estimates of the 8, 10, and 12 year olds to be significantly more accurate than the corresponding adult performance. Adult visual estimates were significantly more accurate than those made by 8 year olds, but were not significantly different from the visual responses of 10 and 12 year olds. Estimates based on bisensory reafference did not differ from each other across the four age groups. It is concluded that modality adeptness and dominance are task dependent and empirically determined rather than being innate properties of sensory systems. The data indicate that intersensory differentiation rather than integration occurs with maturity.

Although the oblique effect has been conceptualized as a purely visual phenomenon, recent studies report its occurrence in a haptic matching task and present the hypothesis that differences in haptic orientational sensitivity might be responsible for the results. The possibility that procedural variables could be responsible was investigated. Specifically, the effect of prior knowledge of the stimulus orientation standards and of use of bilateral haptic exploration of standard and comparison orientations was examined. The results indicate that the reported oblique effect is reduced either when subjects are not informed which orientations will be tested, or when a unilateral matching procedure instead of a bilateral one is used. When both conditions are combined, the haptic oblique effect is eliminated. It is concluded that this particular manifestation of the oblique effect is not related to haptic sensitivity, but stems from (i) the use of well-established imagery as referent for a match (imagery for oblique stimulus orientations is inferior) and (ii) the inherently different scanning patterns required in bilateral exploration of obliques (percepts of standard and comparison obliques will be necessarily different).

The minimum principle states that a perceiver will see the simplest possible interpretation of a pattern. Some theorists of human perception take this principle as a core explanatory concept. Others, especially Rock and Hochberg, hold the view that a perceptual minimum principle is untenable. Rock presents a great number of demonstrations which, in his opinion, rule out the minimum principle. Hochberg states that ‘impossible’ figures especially present a difficulty for this principle. It is argued here that, in order to test the minimum principle, a method is needed to describe interpretations of patterns in such a way that they can be ordered according to simplicity. To achieve this, Leeuwenberg's coding system was used. The analyses reported here of the patterns which Rock produces as evidence against the principle show that, contrary to Rock's claim, the way these patterns are preferentially perceived provides strong support for the minimum principle. Next, it is demonstrated that interpreting certain patterns as ‘impossible’ figures is not incompatible with the principle. Finally, it is argued that a test of the minimum principle is necessarily conflated with two other hypotheses, one concerning the metric of simplicity and one concerning the task conception of the experimental subjects.

An investigation is reported of the ability of normal subjects and patients with right-hemisphere lesions to identify 3-D shadow images of common objects from different viewpoints. Object recognition thresholds were measured in terms of angle of rotation (through the horizontal or vertical axis) required for correct identification. Effects of axial rotation were very variable and no evidence was found of a typical recognition threshold function relating angle of view to object identification. Although the right-hemisphere-lesion group was consistently and significantly worse than the control group, no qualitative differences between the groups were observed. The findings are discussed in relation to Marr's theory that the geometry of a 3-D shape is derived from axial information, and it is argued that the data reported are more consistent with a distinctive-features model of object recognition.