Abstract

Some patients with damaged striate cortex have blindsight-the ability to discriminate unseen stimuli in their clinically blind visual field defects when forced-choice procedures are used. Blindsight implies a sharp dissociation between visual performance and visual awareness, but signal detection theory indicates that it might be indistinguishable from the behavior of normal subjects near the lower limit of conscious vision, where the dissociations could arise trivially from using different response criteria during clinical and forced-choice tests. We tested the latter possibility with a hemianopic subject during yes-no and forced-choice detection of static and moving targets. His response criterion differed significantly between yes-no and forced-choice responding, and the difference was sufficient to produce a blindsight-like dissociation with bias-sensitive measures of performance. When measured independently of bias, his sensitivity to static targets was greater in the forced-choice than in the yes-no task (unlike normal control subjects), but his sensitivity to moving targets did not differ. Differences in response criterion could therefore account for dissociations between yes-no and forced-choice detection of motion, but not of static pattern. The results explain why patients with blindsight are apparently more often ''aware'' of moving stimuli than of static stimuli. However, they also imply that blindsight is unlike normal vision near threshold, and that pattern- and motion-detection in blindsight may depend on different sets of neural mechanisms during yes-no and forced-choice tests.