Our daily experience shows that the CNS is a highly efficient machine to predict the effect of actions into the future; are we so efficient also in reconstructing the past of an action? Previous studies demonstrated we are more effective in extrapolating the final position of a stimulus moving according to biological kinematic laws. Here we address the complementary question: are we more effective in extrapolating the starting position (SP) of a motion following a biological velocity profile? We presented a dot moving upward and corresponding to vertical arm movements that were masked in the first part of the trajectory. The stimulus could either move according to biological or non-biological kinematic laws of motion. Results show a better efficacy in reconstructing the SP of a natural motion: participants demonstrate to reconstruct coherently only the SP of the biological condition. When the motion violates the biological kinematic law, responses are scattered and show a tendency toward larger errors. Instead, in a control experiment where the full motions were displayed, no-difference between biological and non-biological motions is found. Results are discussed in light of potential mechanisms involved in visual inference. We propose that as soon as the target appears the cortical motor area would generate an internal representation of reaching movement. When the visual input and the stored kinematic template match, the SP is traced back on the basis of this memory template, making more effective the SP reconstruction.
Keywords: internal models; kinematics; motion inference; visual perception.