Centred egocentric, decentred egocentric, and allocentric spatial representations in the peripersonal space of congenital total blindness.

The distinction between different spatial representations in the peripersonal space was examined in two experiments by requiring sighted blindfolded and blind participants to remember the locations of objects haptically explored. In experiment 1, object relocation took place from either the same position as learning-with the same (centred egocentric condition) or 90 degrees-rotated (rotated egocentric condition) object array-or from a position different from the learning position (allocentric condition). Results revealed that, in both sighted and blind people, distance errors were higher in the allocentric and rotated conditions than in the centred egocentric condition, and that blind participants made more distance errors than sighted subjects only in the allocentric condition. Experiment 2 repeated rotated egocentric and allocentric conditions, while the centred egocentric condition was replaced by a decentred egocentric condition in which object relocation took place from the same position as learning (egocentric) but started from a decentred point. The decentred egocentric condition was found to remain significantly different from the rotated condition, but not from the allocentric condition. Moreover, blind participants performed less well in the allocentric condition, but were specifically impaired. Overall, our results confirm that different types of spatial constraints and representations, including the decentred egocentric one, can be distinguished in the peripersonal space and that blind people are as efficient as sighted in the egocentric and rotated conditions, but they encounter difficulties in recalling locations also in the peripersonal space, especially when an allocentric condition is required.