Tachistoscopic construction of four orientations of a square

Bulletin of the Psychonomic Society 1977, Vol. 9 (5),333·336 Tachistoscopic construction of four orientations of a square RICHARD M. JOHNSON University of Manitoba, Winnipeg, Manitoba, Canada and JOHN UHLARIK Kansas State University, Manhattan, Kansas 66506 Subjects were given repeated brief exposures to an outline of a square, a 45-deg rotation of a square (diamond), or one of two rotations of a square which were not symmetrical about the vertical axis. The stimulus forms were presented tachistoscopically for as many repetitions as were necessary for correct identification, and the nature of fragments reported prior to correct identification was examined. The obtained patterns of fragmentation were most consistent with an explanation based on the importance of the vertical axis of symmetry as opposed to explanations based on preferences for lines of certain orientations or the internal geometry of the forms. Rotation of some two-dimensional figures in the frontal plane tends to produce distinctly different phenomenal percepts. For example, Mach (1914/1959) observed that a square and a diamond give rise to the perception of distinctly different forms, even though their internal geometry is identicaL Attneave (1968) attributes this difference to competition between the axes of symmetry. In both cases, two of the axes of symmetry are bisectors oriented obliquely 45 deg from vertical, and the other two axes of symmetry are the vertical and horizontal bisectors. Attneave suggests that the perceptual system manifests a preference for analyzing symmetry about the vertical axis and, since the square and diamond have different distributions of characteristic features about their vertical axes, different phenomenal percepts result. Rock (1973) also attributes special importance to the vertical axis of symmetry in the analysis of form. The finding that subjects judge figures to be similar on the basis of symmetry about the vertical axis, rather than about the horizontal axis, has been used to provide support for this notion (Rock & Leaman, 1963). A tachistoscopic repetition effect which has been recently utilized to investigate form perception (Johnson & Uhlarik, 1974) may be useful in examining differences in the way orientation of figures affects perceived form. The repetition effect obtains when a visual stimulus is repeatedly presented for brief exposures at constant durations. Initially subjects are typically unaware of This research was supported by grants from the Bureau of General Research, Kansas State University. The authors would like to thank Paul Goldhorn and Ross Teske for their assistance in conducting this experiment. Requests for reprints should be sent to John Uhlarik, Department of Psychology, Kansas State University, Manhattan, Kansas 66506. any aspect of the stimulus. With well-spaced repeated presentations, parts of the stimulus are reported. Gradu· ally, these fragments become more complex. Eventually, the entire stimulus is correctly recognized on every exposure, even though the exposure duration is the same as for the initial exposures when nothing was reported. For example, the most frequent initial types of reports for presentation of a square were single vertical·line elements and parallel vertical lines. Gradually subjects reported more complex fragments, until the entire stimulus pattern was identified (Johnson & Uhlarik, 1974). In other words, it appeared as though the subjects were gradually constructing the forms over repeated exposures. To the extent that the fragmentation effects provide information concerning the manner in which forms are constructed from components or elements of the figure, one might expect different patterns of features to be reported as a function of the orientation. In the present study, a square, a 45-deg rotation of the square (diamond), or one of two rotations of the square which were not symmetrical about the vertical axis was repeatedly presented to subjects at initially subthreshold durations. If only the internal geometry of a form is important, one would expect no qualitative differences in the pattern of fragmentation for the four different orientations of the square. On the other hand, the distribution of features about the vertical axis of symmetry is different for the square and diamond. There· fore, the qualitative nature of the fragments reported for these two orientations of the same form should be different if the vertical axis of symmetry is the critical determinant of perceived form. The 22.5-deg rotations had no vertical axis of symmetry. In this case, one would expect fragments to be reported that are representative 333 334 JOHNSON AND UHLARIK of both of the obliquely orientated axes of symmetry since these axes of symmetry are competitive in nature, according to Attneave (1968). METHOD Subjects The subjects were 48 undergraduates enrolled in general psychology classes at Kansas State University. Subjects were required to have normal or corrected 20/20 vision. Stimuli and Design The stimulus patterns were line drawings of a square, a diamond, and a 22.S-deg rotation (with respect to the vertical' axis of symmetry) of the square. There were 16 subjects randomly assigned to each viewing condition. For half of the subjects in the 22.S-deg rotation condition the direction of rotation was clockwise (+22.S deg); for the remaining subjects rotation was counterclockwise (-22.S deg). Each stimulus was drawn in black ink on a white background, and subtended a visual angle of 2 deg along its largest dimension. The width of the lines sub tended 2 min of visual angle. The stimuli are shown in Figure 1. The stimuli were individually presented in one channel of an Iconix 6131 three-channel tachistoscope. The tachistoscopic field of view was rectangular, subtending a visual angle of S deg vertically and 8 deg horizontally. A homogeneous preexposure field containing a fixation point was presented in the second channel of the tachistoscope. The fixation point was centered in the preexposure field and coincided with the center of the stimulus pattern in the first channel. A homogeneous white postexposure field was presented in the third channel. The field luminance was 43 mL for each channel. Procedure The preexposure field was on at the beginning of each exposure. All subjects viewed the stimuli binocularly. The subjects were instructed to fixate on the dot in the tachistoscope and then press the button on a hand-held switch to trigger the stimulus presentation. When triggered, illumination of the preexposure field terminated and was immediately followed by illumination of the stimulus field. Immediately upon termination of the stimulus field, a postexposure field followed for SOO msec . The preexposure field then came back on until the subject initiated another exposure. The subject recorded his response and then initiated a new trial. Subjects were allowed to work ,, CD I I I I ! Figure 1. The four st (...truncated)