(A) A field where forces are linearly dependent on both limb position and velocity.
(B) A field where forces are linearly dependent on limb velocity but nonlinearly dependent on limb position. Gain-field encoding predicts that the field in (B) will be harder to learn than one in (A).
(C) Learning index of subjects (n = 6) for the paradigm in (A) and subjects (n = 5) for the paradigm in (B).
(D) Gain-field encoding predicts hypergeneralization. The figure shows movements and its associated force field during training and test sets.
(E) Performance of subjects (n = 4) for the paradigm in (D). Dark lines are errors in center movements and gray lines are errors in right movements. The shaded areas represent the SEM. Filled diamonds show the catch trials for the left movements during test set; filled squares show the catch trials for center movements.