Abstract
Twelve normal controls, twelve left-brain-damaged patients, and thirty-six right-brain-damaged patients with or without tactile extinction or tactile neglect were asked to report light touches delivered to the left or the right hand or simultaneously to both hands. The hands could be in anatomic position or one hand could cross over the other. Moreover, the two hands could be in the left or the right hemispace or across the corporeal midline. Controls and nontactile-extinction groups performed better when the hands were in anatomical than in crossed position. By contrast, patients with tactile extinction detected contralesional stimuli with higher accuracy in crossed than in anatomical position. This result suggests that, in these patients, impairments in detecting contralesional stimuli can be due not only to sensory but also to spatial factors contingent upon the position of the hands. There was no interaction between the effect of crossing the hands and the hemispace where the crossing took place. This suggests that coding the position of a hand as left or right does not necessarily occur in relation to the bodily midline, but it may arise from the computation of the position of the other hand.