Blindsight is unlike normal conscious vision: evidence from an exclusion task

We explored whether information processed subconsciously in blindsight is qualitatively different from normal conscious processing. On each trial the blindsight patient GY was presented with a square-wave grating either in an upper or lower quadrant of his visual field and was asked to report the opposite of its location (e.g. to say 'Up' if it was in the lower quadrant). We found that while GY was able to follow these exclusion instructions in his normal field, he tended to erroneously respond with the real location when the grating appeared in his blind field. Remarkably, his error rate actually increased with increasing grating contrast in his blind field. The interpretation of these results does not rely on subjective reports and thus cannot be criticized on the grounds that subjective reports are unreliable. We conclude that blindsight is unlike normal conscious vision.