Coding space within but not between objects: evidence from Balint's syndrome.

The ability to make spatial judgements was examined in a patient demonstrating poor perception of multiple objects following bilateral parietal lesions, under conditions in which the presence of the stimuli to which judgements were made could be detected. The tasks required judgements of spatial length or the position of coloured parts of stimuli. We manipulated the degree to which two uprights in a display could be encoded into a single perceptual object using either stored knowledge or bottom-up cues based on 2D or 3D image relations. Performance was dependent on the presence of both bottom-up grouping and familiarity. However, connectedness in the image was not sufficient to benefit performance, when stimuli were separate objects in 3D space. This deficit in spatial judgements, arising following detection of the relevant stimulus elements, is attributed to an impairment in coding the spatial relations between separate perceptual objects. This deficit could be overcome if stimuli could be grouped in 3D, using bottom-up cues and top-down knowledge.