Opposed Optimal Strategies of Weighting Somatosensory Inputs For

32 Behavioral studies have suggested that the brain uses a visual estimate of the hand to 33 plan reaching movements towards visual targets and somatosensory inputs in the case of 34 somatosensory targets. However, neural correlates for distinct coding of the hand 35 according to the sensory modality of the target have not yet been identified. Here we 36 tested the twofold hypothesis that the somatosensory input from the reaching hand is 37 respectively facilitated and inhibited when planning movements towards somatosensory 38 (unseen fingers) or visual targets. The weight of the somatosensory inputs was assessed 39 by measuring the amplitude of the somatosensory-evoked potential (SEP) resulting from 40 vibration of the reaching finger during movement planning. The target sensory modality 41 had no significant effect on the SEPs amplitude. However, Spearman’s analyses showed 42 significant correlations between the SEPs and reaching errors. When planning 43 movements towards proprioceptive targets without visual feedback of the reaching hand, 44 participants showing the greater SEPs were those who produced the smaller directional 45 errors. Inversely, participants showing the smaller SEPs when planning movements 46 towards visual targets with visual feedback of the reaching hand were those who 47 produced the smaller directional errors. No significant correlation was found between the 48 SEPs and radial or amplitude errors. Our results indicate that the sensory strategy for 49 planning movements is highly flexible between individuals and also for a given sensory 50 context. Most importantly, they provide neural bases for the suggestion that optimization 51 of movement planning requires the target and the reaching hand to be both represented 52 in the same sensory modality. 53 54