Temporal prediction of touch instant during observation of human and robot grasping.

The aim of the present work was to test the ability to predict the instant at which a grasping hand touches an object. Our hypothesis was that, because of the activation of the mirror-neuron system, the same predictive process necessary for action execution should be active during observation. Experimental evidence indicates, however, that not only observed actions but also observed objects automatically activate observer's motor repertoire. What happens, therefore, if the observed action is different from the one automatically evoked by the vision of the object? To answer this question we presented subjects with two different grasping actions: the one most suitable for the presented object and a less appropriate one. Subjects were required to detect the instant at which the demonstrator's hand touched the object. In a further condition, subjects were required to detect the outcome of an action performed by a robotic arm moving with constant kinematics. Results showed that while in the case of robot grasping subjects responded before the touch instant, in the case of human grasping the response followed the touch instant, but occurred much earlier than simple reaction times. This demonstrates that subjects were able to predict the outcome of the seen action. The predictive capability was specifically enhanced during observation of the "suitable" grasping. We interpret these results as an indication of the synergic contribution of both object-related (canonical) and action-related (mirror) neurons during observation of actions directed towards graspable objects.