The illusory distortion of space in spatial neglect

The hypothesis of space representation in neglect postulates an anisometric representation along the horizontal dimension that takes the form of a progressive expansion of the contralesional side and a compression of the ipsilesional one, following a logarithmic scale. This spatial distortion in neglect can be compared to the effects of the Oppel-Kundt illusion, that consists of the perception of a filled space as larger than an empty space of equal size. In experiment 1 we used this illusion in two different tasks: reaction times and size adjustment. Neglect patients judged as larger the stimuli presented in a compressed portion of space than those presented in an empty hemispace. According to Gregory’s theory of size constancy scaling, the stimuli presented in a filled space are perceived as larger because they are perceived to be farther from the observer. The O-K illusion, thus, could induce an inappropriate size-constancy scaling that reproduces the misrepresentation of space in neglect patients. In the second Experiment we tested Gregory’s theory in neglect patients by using depth cues: the Ponzo illusion. This illusion consists of the fact that the perceived size of an object of constant size is directly proportional to its apparent distance. We expected that to be perceived as of the same size a further stimulus needs to be shorter than a nearer one. In addition, we expected that neglect patients will be affected by their distortion of the horizontal extent of the stimuli along the left-right dimension. In the first session of the second Experiment, subjects were requested to modify the upper target stimulus to make it appear of the same size of the standard stimulus. When neglect patients compared an upper stimulus with a lower one placed at the left side, they shrunk more the first one because they perceived the reference left stimulus as smaller than its real size. In the second session we have changed the position of standard stimuli: they were placed upward, and the left or right position were less lateralized along the horizontal space. In this case there are no differences of side presentation of stimuli among the different groups. According to the anisometric hypothesis of space, indeed, if it is true that neglect signs follow a gradient along the horizontal dimension, with more impairment for the left side and progressively less impairment for right positions, less lateralised stimuli should produce less impairment. In the third Experiment we used an indirect method to test the anisometric hypothesis of space: we analyzed the burden of the distortion of horizontal dimension in neglect patients in a task of vertical dimension judgment. Based on the results obtained we can state that the misjudgment of the vertical extent found with neglect patients is due to a misrepresentation of the horizontal dimension (irrelevant to the task), thus in line with the space anisometry hypothesis. When neglect patients had to compare two identical stimuli (tall and large the same), they judged more often the leftmost as being taller than the right one because it appears to them thinner. In a second session we presented the same stimuli as for session one but reducing their horizontal size. The results are in agreement with the crossover effect: the left stimulus is perceived by neglect patients as larger than the right one, and consequently shorter.