Item

Abstract

Previous studies have demonstrated that visual-to-tactile sensory substitution devices can be used by blind humans to localize objects within a three-dimensional environment. Yet the brain mechanisms that enable touch to act as a "substitute" for vision remain unclear. It is obvious that
self-movement is an essential part of the process, but the actual role of movement in sensory substitution is not well-understood. In this study, blindfolded participants used a finger-mounted photodiode coupled with a vibrating motor, and judged the distance of a light source in a visual matching task. Participants never saw the light nor received feedback about their performance.
There was no spatial information in the tactile stimulus, ensuring that all spatial awareness was attained through movement. One group of participants was instructed to attend primarily to arm location (the proximal object) and a second group was told to attend only to the felt location of the
light (the distal object). In the Distal group, distance judgments improved significantly over the course of 2 one-hour sessions, with the accuracy of some subjects approaching that of subjects in a visual control condition. The Proximal group showed no improvement in mean accuracy. Importantly, the degree to which all subjects experienced the light as a solid object--rather than an imaginary one--was positively correlated with the accuracy of their distance judgments. Finally, improved performance transfered to new conditions, e.g. in which the photodiode was placed on the opposite hand, or the body was rotated by 90º. This demonstrates that performance is not specific to
particular muscles or joint angles. The results show that visual-to-tactile sensory substitution is a true analogue to everyday visual perception. It can therefore be used in future experiments to better
understand how movement contributes to our experience of a three-dimensional world. The success of the transfer phase supports an interpretation of the results in terms of ecological invariants, although sensorimotor and neural-plasticity accounts of sensory substitution are also considered.