Some observations on the Poggendorff and Müller-Lyer tactual illusions

Abstract

Two experiments were conducted in which illusory and nonillusory figures were presented in the right and left tactual fields. Experiment 1 used 24 men (aged 18-24 years) and the Poggendorff'(P)illusion and an alignment (AI) task; Experiment 2 used another 24 men and the MullerLyer (ML) illusion and a line-match (Lm) task. Standard and mirror-image presentations of the figure were used in each case. Nested designs of analysis of variance and analysis of covariance were adopted for comparing left-and right-hand performance under each experimental condition. Covariance analysis on distortion scores of P and ML was performed separately with discrepancy measures on Al and Lm tasks as covariates. An investigation was also made into whether the direction of tactually perceived distortions was the same as or opposite to that reported for the visual illusions. The results show large P and ML tactual distortions. For ML, the tactual effects found were similar to those predictable from visual illusions, whereas for P, a high value ofinversion emerged. Differencesbetween hands were significant for P under the standard presentation, exploration with the right hand inducing a higher value of inversion. The latter results are discussed with reference to differences between the functions of human hemispheres which some authors have suggested for perceptual distortions.Susceptibility to visual illusions has been only marginally investigated in the context of functional hemispheric lateralization studies. As is well known, human hemisphere functioning is described in terms of contrasting modes of cognition, information processing, and coding (e.g., Bradshaw & Nettleton, 1981). According to the most current formulation, the left hemisphere may use analytic, serial, or sequential strategies, while the right uses global or parallel strategies (Simion & Bagnara, 1982); other studies point out the differences in coding information (Posner, 1978;Umilta, 1982). Friedman and Campbell-Polson (1981) proposed a model of the hemispheres as two independent and limited pools of resources, or subprocessors (Allen, 1983), which could be used in different ways. Whatever mechanism underlies the experimental data, however, there is general agreement that the left hemisphere is verbal, while the right is visuospatial. Basso, Bisiach, and Faglioni (1974) proposed that if illusions were due to perceptual distortion depending on structural features of a particular stimulus, right-hemisphere damage, by impairing the holistic attitude, should reduce susceptibility to the illusion, and left-hemisphere damage should enhance it. Basso et al. therefore studied performance in a linear-lengthdiscriminationtask and the extent of the Miiller-Lyer(ML) visual illusion in normal and unilateral-brain-damaged patients.Although linear-length discrimination was found to be impaired in right-brain-damaged patients with a visual field defect, Basso et al. found that, contrary to the hypothesis, left-brain-damaged patients showed reduced susceptibility to the ML. A seco...

Lack of evidence for a tactual Poggendorff illusion

“…Scrutiny of the Lucca et al (1986) paper reveals that the data they present are not really informative regarding the magnitude or the direction of the tactual Poggendorff illusion. First, to obtain their overall mean illusions, they averaged the unsigned setting errors, as did Witkin and Asch (1948) many years ago in their experiments on the rod-and-frame illusion.…”

“…Second, Lucca et al (1986) referred to previous work, some of which was their own, which had revealed "inversions," settings in the direction opposite the usual visual effects. Following a procedure adopted by Pasnak and Ahr (1970), Lucca et al (1986) then calculated the mean of all the individual settings that were in the same direction as the visual effect (here called positive effects) and, separately, the mean of all the negative settings.…”

“…Further details of the Lucca et al apparatus and method need not be considered here, since our first experiment is an attempt to replicate theirs. Lucca et al (1986) claimed to have obtained large illusions (of the order of 4 mm) in the full Poggendorff figure with parallels (P2 and PI in Figure I) but small effects (of the order of 1 mm) in the parallel-less ("simple alignment") figures (ALl and AL2 in Figure 1). More importantly, they claimed to have found that the majority This research was supported by an Australian Research Council Grant A78831871 to P. Wenderoth.…”

“…Recently, Lucca, Dellantonio, and Riggio (1986) had blindfolded subjects use two fingers actively to explore Poggendorff-type figures constructed from raised aluminum strips. Both right and left hands were used to explore four figures, such as those represented in Figure 1.…”

“…Following a procedure adopted by Pasnak and Ahr (1970), Lucca et al (1986) then calculated the mean of all the individual settings that were in the same direction as the visual effect (here called positive effects) and, separately, the mean of all the negative settings. They reported mean positive effects near zero but mean "inversions" of 3-4 mm.…”