Abstract
Two different “orientations” present in a linearly-oriented array of parallel, equilength lines were separated and independently manipulated: The common orientation of the individual lines defines a “Local Orientation”; the orientation of the implicit straight line through the centers of the parallel lines defines a “Global Orientation”. The present experiments measure the relative contributions of Global and Local Orientations in inducing changes in perception of two spatial dimensions: (1) The dimension of elevation as measured by the subject's setting of a small circular visual target that appears to lie at eye level (VPEL); (2) orientation in the frontal plane as measured by the subject's setting of a short line to appear vertical (VPV). Subjects monocularly viewed in darkness two simultaneously-presented inducing arrays (68°-long), centered at ±25° horizontal eccentricity, of 25 parallel equilength lines that were either 1°-long or 3°-long. To obtain large influences, the two inducing stimuli were bilaterally-symmetrically-oriented for VPEL trials and were parallel for VPV trials (Spatial Vision, 1994). In the first experiment the roll-tilt of the Global Orientation of the arrays was varied over a ±10° range around true vertical with Local Orientation fixed at either horizontal or vertical, or was identical to the Global Orientation. In the second experiment, Global Orientation was fixed at vertical and Local Orientation varied over a ±10° range around true vertical. Both VPEL and VPV changed systematically with Global Orientation; Local Orientation had at most a small influence. This is not consistent with a model in which the orientation responsible for induction is set by orientation-selective neural units in V1 since those units select for Local line Orientation and not Global Orientation. We suggest that the retinotopic location information in V1 is passed on to higher level neural units for construction of Global Orientation.
Support: NIH grant EY10534.