Abstract
When an object's location is defined by redundant sensory information, the brain combines these sensory inputs to form a coherent estimate of the object's location in such a way that the more reliable sensory cue is assigned a greater weight (Ernst and Banks, 2002). In the present study we asked if the motor system can learn to change the way in which it integrates sensory cues. Participants reached to visual (V), proprioceptive (P; their left index finger) or visual + proprioceptive (VP; their seen left index finger) targets. Inaccurate endpoint visual feedback was provided on the V and P reaches, such that on V reaches the seen horizontal error was greater than (three times) the actual horizontal error achieved and on P reaches the seen horizontal error was smaller than (one third) the actual horizontal error achieved. No feedback was provided on VP reaches, which were completed before and after reaching to V and P targets. To determine the weight assigned to V and P cues when reaching to VP targets, we compared reach endpoints on VP reaches with reach endpoints achieved on V and P reaches. Results indicate that after experiencing misaligned endpoint error feedback, participants adjusted their reaches to VP targets such that their endpoints resembled those achieved on P reaches. These results suggest that the brain can change how it weights sensory information regarding target location, relying more on the sensory modality which has been experienced as most accurate.Acknowledgments: Research support: Natural Sciences and Engineering Research Council (EKC)