Dissociation between perception and visuomotor transformation during reproduction of remembered distances

1. In complete darkness subjects were presented with two visual stimuli whose distance was randomly varied. The subjects were required to reproduce the interstimulus remembered distance in two conditions. In one condition (reproduction by pointing) they pointed to a virtual position in space. In the other condition (visual reproduction) they used two other visual stimuli. One of them was fixed, and the other could be manually moved. Constant and variable errors were measured in the two conditions. 2. Constant error varied between the two conditions. In the pointing task subjects slightly overestimated the shorter distances, and underestimated the longer ones. During visual reproduction, they consistently overestimated all distances, and the error of overestimation tended to increase with distance. Statistical comparison between the errors in the two conditions was significant. Variable error increased with distance in both conditions, but did not show any significant difference between the two tasks. 3. The results of the present experiment support the hypothesis that perception and visuo-motor transformation are two separate processes in which the same object attributes are independently analysed. However, the finding that variable error did not change between the two tasks suggests that some stages are common to the two processes.