Quantitative rejection of hypotheses predicting learning-induced changes in the direction of the initial eye acceleration that depend only on the direction of the prelearning smooth eye movement and not on whether the stimulus consisted of one or two targets. A, Comparison of the actual changes in direction of responses to oblique motion of single targets with the shift predicted if learning altered only the horizontal component of pursuit. Eight points are plotted for each monkey: one point for each of four single-target stimuli and each of two learning directions. The four single-target stimuli moved obliquely 30 and 60° up and down relative to the learning direction. B, Comparison of the shifts in direction of responses to two-target stimuli (x-axis) with those for single-target stimuli (y-axis) that evoked responses in the same direction. Each data point shows the average for the upper and lower quadrants in the learning direction, so four data points are plotted for each monkey: one each for right-up, right-down, left-up, and left-down two-target motions. Different symbols show data from different animals: circles, monkey E;squares, monkey K; triangles, monkey I;diamonds, monkey N.