Abstract
Numerous studies designed to test Milner and Goodale's (1992) two-stream model of visual processing have used visual illusions as stimuli. The general claim is that finding a dissociation between perception and action in responses to an illusion provides evidence for the operation of two independently operating streams.The present study used the Poggendorf illusion to investigate the perception-action dissociation (see also Melmoth, Tibber, Grant and Morgan, 2009). A novel stimulus configuration was used, consisting of a rectangular bar, a line that intersected one (long) side of the bar, and a column of numbers to the other side of the bar. In separate experiments, participants were required to project (mentally or by various hand movements) the intersecting line beyond the other side of the bar to the number column. Figures were presented on legal sized paper tacked up at eye level; when required, movements were made directly to the paper. Effects of the illusion were assessed based on projected lines' degree of misalignment from the original. Across experiments, the oft-reported dissociation between perception and action in responses to visual illusions was not found. The apparent effect of the Poggendorff illusion on actions was consistently as great - or greater - than its effect on perceptual judgments. This was also true for an altered version of the figure in which the rectangle was removed altogether and just the line and number column remained. The prevalence of the Poggendorf illusion's affect on actions does not support the two-stream model. Alternately, this suggests that the locus of this illusion's effect is early in visual processing before the two streams have diverged (see Dyde and Milner, 2002).