Stereodepth from Afterimages
T. Bower,W. M. Goldsmith,J. Hochberg
DOI: https://doi.org/10.2466/PMS.1964.19.2.510
1964-10-01
Perceptual and Motor Skills
Abstract:W e cannot tell which eye receives a monocular stimulus (Pickersgill, 1961); however, if we can't detect which eye receives each of two views, binocular disparity must be an ambiguous depth cue. Disparity changes due to convergence changes are un-ambiguous, but these could not explain the correct depth judgments obtained by Ogle and Reiher (1962) with rwo flash tubes' binocular afterimages (whose disparity was fixed). However, their data are inconclusive, since their Ss may have identified each afterimage's eye-of-origin by its rivalry with the contralateral images of the comparison stimuli, which were freely viewed. In the present experiment, a glass stereogram was transilluminated by a flash (Heiland, 65 A ) , located 1 ft. behind the stereoscope. Each half-view bore three slits, 1 /32 in. wide and 3 in. tall, and a 1/32-in. luminous dot so placed next to the central slit in each half-view that, when the dots were fused, the combined view contained rwo outer lines (each line viewed by both eyes), 16' apart. The center lines then formed a double image with 0.12" disparity. Reversing the half-views changed the disparity from "center near" to "center far." In Condition 1, 1 0 naive Ss dark adapted for 5 min., adjusted the stereoscope in semidarkness, then fired the flash and reported immediately whether the centerline in the afterimage was near or far. There were 5 near and 5 far trials in randomized order, with an intertrial interval of 1 min. Condition 11 differed only in that the afterimages were sustained by a flickering (4-cps) white field for 1 5 sec, after each flash, following which depth was judged. The order of conditions was equated berween Ss, the intercondition interval being 45 min. In Cond. I (immediate response), Ss judged better than chance ( M correct = 8.4, p < 0.01 by t test) and better than in Cond. I1 ( p < 0.01). Judgments in Cond. I1 (protracted inspection) did not differ from chance ( M = 5.5). W e must conclude, as did Dove in 1841, that eye-movement-produced changes are unnecessary to stereopsis, and that some other sign must identify which eye receives each half-view in the very first moment of binocular input.