Perceptual and Motor Shills, 1977,44. 755-765. @ Perceptual and Motor Skills 1977
COMPONENTS OF ASYMMETRICAL VISUAL ENCODING OF GEOMETRICALLY T R A N S F O R M E D SCRIPTS1 AMOS S. COHEN
Swiss Federal Institute of Technology, Zurich Summdsy.-Austrian and Israeli subjects were presented tachistoscopically with geometrically transformed numbers, nonsense syllables in Latin letters and Hebrew words (only the Israeli). Opposite reading and writing habits and different hand dominance were used as experimental variables. Criteria for evaluating the data were scanning pattern (voluntary preferred reading direction) and reading performance. A difference in reading pattern was found between Austrian right- and left-handed subjects but no difference in their reading performance. Also between Austrian and Israeli right-handed subjects a difference in the reading pattern appeared as well as in reading performance. Furthermore, the scanning pattern used by Austrian right-handed subjects in reading geometrically transformed numbers was different from their pattern in reading syllables. Discinguished were two sources of asymmetries, an intrinsic in visual mechanisms and a structural one in the stimulus itself.
The investigation of asymmetry in visual perception was challenged through the controversy between Lashley ( 1942) and Hebb ( 1949), which relates mainly to the role of learning processes in vision. While Lashley suggested rhac equal perceptions occur due to the excitation of similar patterns of unspecific cells, Hebb proposed that their topographical localization within the visual cortex plays a major role, as those cells excited previously most ofren in a similar manner should produce a more accurate perception.. Dealing with this disputation Mishkin and Forgays (1952) proved that learning influences visual perception, i.e., causes a specific asymmetry. They observed that unilateral tachistoscopically presented English words were better recognized when they appeared on the right than on the left visual half-field contrary to Yiddish words (written from right to left with Hebrew letters), which were performed better on the left visual half-field. This objection led the authors to conclude that a particular perception depends on previous learning. The lateral superiority depends also on the language which bilingual subjects first learned. Those subjects who learned English first recognize more Yiddish words right of fixation, contrary to those subjects whose prior language was Yiddish (Orbach, 1953). The superiority of right visual half-field 'Presented in part at the XXIst International Congress of Psychology, Paris, France, 18-25 July, 1976. Thankful acknowledgement is given Professor Ivo Kohler whose personal communication I greatly appreciated. This study was supported by a Scholarship of the Austrian Ministerg for Science and Research and aided by the Swiss Federal Institute of Technology, Zurich. Requests for reprints should be sent to Amos S. Cohen, 57, Smilanski St., 42434 Natania, Israel.
756
A. S. COHEN
for perceiving verbal stimuli written in Latin script and by using unilateral presentation was proven in further studies (McKeever & Gill, 1972; Hines & Satz, 1974; Mackavey, et al., 1975) but not when using bilateral expositions (Mackavey, 1974). Also, superiority of the right visual half-field depends on specific typeface used (Bryden & Allard, 1976). The advantage of the left visual half-field was pointed out for non-verbal stimuli like for figures, dots, faces, etc. (Kimura, 1959; Bryden, 1960; Bryden & Rainey, 1963; Brandshaw, Gates, & Patterson, 1976). The observed differences between verbal and non-verbal tasks, termed analytical vs holistic, were interpreted in terms of dual functional asymmetry of the brain (Kimura, 1966). The dominant brain hemisphere should process the analytical tasks better whereas the non-dominant the holistic ones, either in vision or audition (Kimura, 1961; Gordon, 1970; Bever & Chiarello, 1974; Bever, 1975). Therefore, superiority of a particular visual half-field depends on the nature of the task as well as on the subjects' relative brain dominance. Superiority of a particular visual half-field is determined by [he mode of presenting stimuli also. When verbal stimuli are presented bilaterally an advantage for the left visual half-field results, as Heron (1957) pointed out. He emphasizes that two types of processes are involved in perceiving stimuli presented, a prior "attentional set" and "post-exposure process," meaning trace scanning mechanisms. Accordingly perception depends also on-through learning affected-scanning processes as well. Carmon and Nachshon (1973) found a more effective binocular fusion at a certain retinal locus on the right visual halffield by English native readers whereas the left visual half-field was superior for Hebrew native readers. In terms of learned scanning prosesses Harcum and Dyer ( 1962) explain che more accurate perception of a bilaterally presented array of binary patterns on the left side by American subjects and the contrary results when these stimuli were presented to Israeli subjects (Harcum & Friedman, 1963). This lateral advantage can be influenced by appropriate instructing of subjects (Harcum, et al., 1963). Furthermore, Cohen ( 1976) has shown that Austrians encode an array of patterns without any directional clues from left to right whereas Israeli subjects encode most often from right to left as they would read in their native languages. Generally perceptual asymmetry results from the interaction of the subject's characteristics and those of stimuli, which might be termed respectively as subject's intrinsic ajymmetry and as strivctvral asymmetry of stimuli. Perceptual asymmetry in horizontal direction reflects, therefore, the processing accuracy of defined scimuli wichin a particular part of visual cortex as well as encoding mechanisms. While variables such as the relative brain dominance, learning, or maturation belong to a subject's intrinsic asymmetry, the nature of stimuli, their spatial arrangement, and the required task relate to structural asymmetry. There-
VISUAL ENCODING OF TRANSFORMED SCRIPT
757
fore scanning mechanisms must be considered a n integral part of perceptual asymmetry. In this study the role of learning in the encoding of verbal stimuli was investigated by comparing data o f Austrian and Israeli subjects because the direction of their reading and writing habits is of interest. Secondly, t o investigate the role of t h e relative brain dominance data of right- and left-handed subjects with equal reading and writing habits must be compared. In this case handedness is important, assuming a relationship between a dominant right hand and the contralateral hemisphere. A third investigational goal was to study the influence of structural asymmetry on encoding process. Therefore the perception of Latin script and that of numbers has to be compared. A further objective of the present experiments was to study wherher these variables influence perceptual performance.
EXP. 1: READINGGEOMETRICALLY TRANSFORMED NUMBERS Method Subjects.-The three following samples of subjects participated in this experiment: 30 right-handed Austrians, 10 left-handed Austrians and 30 right-handed Israeli subjects. All subjects were students, older than 18 yr., who had normal or corrected vision. Stimuli.-In each presentation 12 numbers were shown, ordered in four items, rwo items above and two items below. Every item consisted of three random numbers, with the exception of the number zero, which never appeared at the beginning or at the end of an item. The items were shown in one of the four following geometrical transformations: normal orientation of the script; inverted form, a rotation of 180" around the x axis; mirror-image, a rotation of 180" around the y axis; and rotated form, a rotation of 180" around the z axis of the plane (see Fig. 1). normal
N
inverted
I
mirrored
M
rotated
R
FIG.1 Examples of geometrically transformed numbers used in Exp. 1 Duration of presentation and its location.-The duration of every presentation was 1 sec. During the exposure there was enough rime to carry out one or more eye movements, therefore, no fixation point was used. Each subject was only asked to glance at the screen. Nevertheless, the presentations occurred on the right, the left, or in the middle of the screen but these locations do not necessarily correspond to visual input to a specific brain hemisphere. The subject was sitting with his head stabilized in the median direction and the center of the screen at a distance of 110 cm. Each presented number was 7 mm high and appeared white on black with a maximum lateralicy of 47.5 mm from the screen's middle to the right or to the left. Number of presentationr.-Eight presentations were given for each geometrical transformation and visual location in random sequence, which total 96 exposirions.
758
A. S. COHEN
The ta~k.-The task was to reproduce as many numbers separdely as possible. He was told that there are some transformations which can be read more easily from right to left and others in opposite direction, but no specific reading direction was suggested. Also, the subject was told that he could use any direction he chose. Once a reading direction had been chosen, the subject was not allowed to change it until he finished with that item. Afterwards, he could do so if he wished. Scores.-There were two scores used, scanning pattern, which is the reading direction from left to right vs a well defined sequence of reproducing the three numbers of the first item and reading performance, which is the number of reproduced items in each exposure.
Rerilltr Scanning pattern.-A 2 X 3 X 4 analysis of variance was carried out to compare data of the right-handed Austrian and Israeli subjects by scoring the reading direction from left to right only. Similar analysis of variance was carried out for data of right- and left-handed Austrian subjects, after having reduced the number of right-handed Austrian subjects at random to 10. In both comparisons there was a difference between the experimental groups ( b e t w e e n right-handed Austrian and Israeli subjects; = 87.77, p < 0.001 and berween right- and left-handed Austrian subjects; Fl,zls = 4.22, p 0.05). Also, there was a difference between geometrical transformations for righthanded Austrian and Israeli subjects (F3,GS6 = 89.28, p < 0.001) and for right= 29.57, p < 0.001). There was no and left-handed Austrian subjects (F3,zI0 significant difference among the three visual locations on the first or the second comparison. Furthermore, no significant interaction was observed between
COMPONENTS OF ASYMMETRICAL VISUAL ENCODING OF GEOMETRICALLY T R A N S F O R M E D SCRIPTS1 AMOS S. COHEN
Swiss Federal Institute of Technology, Zurich Summdsy.-Austrian and Israeli subjects were presented tachistoscopically with geometrically transformed numbers, nonsense syllables in Latin letters and Hebrew words (only the Israeli). Opposite reading and writing habits and different hand dominance were used as experimental variables. Criteria for evaluating the data were scanning pattern (voluntary preferred reading direction) and reading performance. A difference in reading pattern was found between Austrian right- and left-handed subjects but no difference in their reading performance. Also between Austrian and Israeli right-handed subjects a difference in the reading pattern appeared as well as in reading performance. Furthermore, the scanning pattern used by Austrian right-handed subjects in reading geometrically transformed numbers was different from their pattern in reading syllables. Discinguished were two sources of asymmetries, an intrinsic in visual mechanisms and a structural one in the stimulus itself.
The investigation of asymmetry in visual perception was challenged through the controversy between Lashley ( 1942) and Hebb ( 1949), which relates mainly to the role of learning processes in vision. While Lashley suggested rhac equal perceptions occur due to the excitation of similar patterns of unspecific cells, Hebb proposed that their topographical localization within the visual cortex plays a major role, as those cells excited previously most ofren in a similar manner should produce a more accurate perception.. Dealing with this disputation Mishkin and Forgays (1952) proved that learning influences visual perception, i.e., causes a specific asymmetry. They observed that unilateral tachistoscopically presented English words were better recognized when they appeared on the right than on the left visual half-field contrary to Yiddish words (written from right to left with Hebrew letters), which were performed better on the left visual half-field. This objection led the authors to conclude that a particular perception depends on previous learning. The lateral superiority depends also on the language which bilingual subjects first learned. Those subjects who learned English first recognize more Yiddish words right of fixation, contrary to those subjects whose prior language was Yiddish (Orbach, 1953). The superiority of right visual half-field 'Presented in part at the XXIst International Congress of Psychology, Paris, France, 18-25 July, 1976. Thankful acknowledgement is given Professor Ivo Kohler whose personal communication I greatly appreciated. This study was supported by a Scholarship of the Austrian Ministerg for Science and Research and aided by the Swiss Federal Institute of Technology, Zurich. Requests for reprints should be sent to Amos S. Cohen, 57, Smilanski St., 42434 Natania, Israel.
756
A. S. COHEN
for perceiving verbal stimuli written in Latin script and by using unilateral presentation was proven in further studies (McKeever & Gill, 1972; Hines & Satz, 1974; Mackavey, et al., 1975) but not when using bilateral expositions (Mackavey, 1974). Also, superiority of the right visual half-field depends on specific typeface used (Bryden & Allard, 1976). The advantage of the left visual half-field was pointed out for non-verbal stimuli like for figures, dots, faces, etc. (Kimura, 1959; Bryden, 1960; Bryden & Rainey, 1963; Brandshaw, Gates, & Patterson, 1976). The observed differences between verbal and non-verbal tasks, termed analytical vs holistic, were interpreted in terms of dual functional asymmetry of the brain (Kimura, 1966). The dominant brain hemisphere should process the analytical tasks better whereas the non-dominant the holistic ones, either in vision or audition (Kimura, 1961; Gordon, 1970; Bever & Chiarello, 1974; Bever, 1975). Therefore, superiority of a particular visual half-field depends on the nature of the task as well as on the subjects' relative brain dominance. Superiority of a particular visual half-field is determined by [he mode of presenting stimuli also. When verbal stimuli are presented bilaterally an advantage for the left visual half-field results, as Heron (1957) pointed out. He emphasizes that two types of processes are involved in perceiving stimuli presented, a prior "attentional set" and "post-exposure process," meaning trace scanning mechanisms. Accordingly perception depends also on-through learning affected-scanning processes as well. Carmon and Nachshon (1973) found a more effective binocular fusion at a certain retinal locus on the right visual halffield by English native readers whereas the left visual half-field was superior for Hebrew native readers. In terms of learned scanning prosesses Harcum and Dyer ( 1962) explain che more accurate perception of a bilaterally presented array of binary patterns on the left side by American subjects and the contrary results when these stimuli were presented to Israeli subjects (Harcum & Friedman, 1963). This lateral advantage can be influenced by appropriate instructing of subjects (Harcum, et al., 1963). Furthermore, Cohen ( 1976) has shown that Austrians encode an array of patterns without any directional clues from left to right whereas Israeli subjects encode most often from right to left as they would read in their native languages. Generally perceptual asymmetry results from the interaction of the subject's characteristics and those of stimuli, which might be termed respectively as subject's intrinsic ajymmetry and as strivctvral asymmetry of stimuli. Perceptual asymmetry in horizontal direction reflects, therefore, the processing accuracy of defined scimuli wichin a particular part of visual cortex as well as encoding mechanisms. While variables such as the relative brain dominance, learning, or maturation belong to a subject's intrinsic asymmetry, the nature of stimuli, their spatial arrangement, and the required task relate to structural asymmetry. There-
VISUAL ENCODING OF TRANSFORMED SCRIPT
757
fore scanning mechanisms must be considered a n integral part of perceptual asymmetry. In this study the role of learning in the encoding of verbal stimuli was investigated by comparing data o f Austrian and Israeli subjects because the direction of their reading and writing habits is of interest. Secondly, t o investigate the role of t h e relative brain dominance data of right- and left-handed subjects with equal reading and writing habits must be compared. In this case handedness is important, assuming a relationship between a dominant right hand and the contralateral hemisphere. A third investigational goal was to study the influence of structural asymmetry on encoding process. Therefore the perception of Latin script and that of numbers has to be compared. A further objective of the present experiments was to study wherher these variables influence perceptual performance.
EXP. 1: READINGGEOMETRICALLY TRANSFORMED NUMBERS Method Subjects.-The three following samples of subjects participated in this experiment: 30 right-handed Austrians, 10 left-handed Austrians and 30 right-handed Israeli subjects. All subjects were students, older than 18 yr., who had normal or corrected vision. Stimuli.-In each presentation 12 numbers were shown, ordered in four items, rwo items above and two items below. Every item consisted of three random numbers, with the exception of the number zero, which never appeared at the beginning or at the end of an item. The items were shown in one of the four following geometrical transformations: normal orientation of the script; inverted form, a rotation of 180" around the x axis; mirror-image, a rotation of 180" around the y axis; and rotated form, a rotation of 180" around the z axis of the plane (see Fig. 1). normal
N
inverted
I
mirrored
M
rotated
R
FIG.1 Examples of geometrically transformed numbers used in Exp. 1 Duration of presentation and its location.-The duration of every presentation was 1 sec. During the exposure there was enough rime to carry out one or more eye movements, therefore, no fixation point was used. Each subject was only asked to glance at the screen. Nevertheless, the presentations occurred on the right, the left, or in the middle of the screen but these locations do not necessarily correspond to visual input to a specific brain hemisphere. The subject was sitting with his head stabilized in the median direction and the center of the screen at a distance of 110 cm. Each presented number was 7 mm high and appeared white on black with a maximum lateralicy of 47.5 mm from the screen's middle to the right or to the left. Number of presentationr.-Eight presentations were given for each geometrical transformation and visual location in random sequence, which total 96 exposirions.
758
A. S. COHEN
The ta~k.-The task was to reproduce as many numbers separdely as possible. He was told that there are some transformations which can be read more easily from right to left and others in opposite direction, but no specific reading direction was suggested. Also, the subject was told that he could use any direction he chose. Once a reading direction had been chosen, the subject was not allowed to change it until he finished with that item. Afterwards, he could do so if he wished. Scores.-There were two scores used, scanning pattern, which is the reading direction from left to right vs a well defined sequence of reproducing the three numbers of the first item and reading performance, which is the number of reproduced items in each exposure.
Rerilltr Scanning pattern.-A 2 X 3 X 4 analysis of variance was carried out to compare data of the right-handed Austrian and Israeli subjects by scoring the reading direction from left to right only. Similar analysis of variance was carried out for data of right- and left-handed Austrian subjects, after having reduced the number of right-handed Austrian subjects at random to 10. In both comparisons there was a difference between the experimental groups ( b e t w e e n right-handed Austrian and Israeli subjects; = 87.77, p < 0.001 and berween right- and left-handed Austrian subjects; Fl,zls = 4.22, p 0.05). Also, there was a difference between geometrical transformations for righthanded Austrian and Israeli subjects (F3,GS6 = 89.28, p < 0.001) and for right= 29.57, p < 0.001). There was no and left-handed Austrian subjects (F3,zI0 significant difference among the three visual locations on the first or the second comparison. Furthermore, no significant interaction was observed between
