Figure 1.
In each experiment, subjects were instructed to fixate a small red square in the center of an adapter stimulus, which consisted of a random dot kinematogram (RDK) shown within a stationary circular aperture for a random time period lasting between 2 and 2.5 s. The adapter stimulus' dots moved either left- or rightwards at 3°/s. The test stimulus was flashed briefly and was either static or moving slowly left- or rightwards at 0.6 or 1.2°/s. Subjects reported moving direction of the test stimulus by a keypress upon appearance of a small green square. A Baseline and Storage of the MAE: Both fixation target and adapter stimulus were centered on the screen. After a delay of 50 or 500 ms the test stimulus was shown centrally as well. B Retinal Specificity or Phantom MAE: The adapter stimulus was shown 14° left or right (dashed circle) from the fixation target while subjects fixated in the center, where the test stimulus was shown with a delay of 50 ms. C Pre-saccadic Remapping of the MAE: The fixation target, which was shown 7° left or right from the center, was presented for 300 ms before adaptation started. Subjects were instructed to make a saccade upon appearance of a red square at the beginning of the delay period (third panel). The saccade target was always located on the opposite side of the screen relative to the adapter stimulus. Conditions for rightward saccades are depicted. Location and timing of the test stimulus were controlled separately. It appeared either 50 or 500 ms after offset of the adapter stimulus, and it was shown either centered around the position of the original fixation target (dashed circle) or centered around the position of the saccade target. The illustrations are drawn to scale (width: 42.5° height: 32°) and fixation and saccade targets were red. Dashed circles and arrows were not part of the display.
Figure 2.
Psychometric Functions of a Representative Subject (S.F.) for all three Experiments.
In each panel, the percentage of the subject's rightward choices is plotted against the test stimulus' velocity. The diameter of the data points reflects the number of measurements in each condition. In principle, we measured 30 trials in each condition. But note that in B and C data were collapsed from mirror-inverted conditions, yielding 60 measurements in each condition. A The baseline and storage experiment is shown with delays of 50 and 500 ms. Red and green data points represent responses following left- and rightward adaptation, respectively. Logistic functions and error bars from bootstrap runs are colored accordingly (see Experimental procedures for more details). MAE size estimates were obtained from the difference between the percentage of rightward responses, following left- and rightward adaptation, upon presentation of static test stimuli, marked by the intercept of the logistic functions with the y-axis (left panel). In the left panel, the test stimulus was shown 50 ms after presentation of the adapter stimulus (baseline condition). In the right panel, the test stimulus was shown with a delay of 500 ms (storage condition). B Retinal Specificity or Phantom MAE: Adaptation was either in the left or right periphery, whereas the test stimulus was shown centrally after a delay of 50 ms. Data from both adaptation loci did not differ significantly and were pooled for clarity. C Pre-saccadic Remapping of the MAE: Data from rightward and leftward saccade trials representing mirror-inverted conditions were collapsed for clarity. Only the four principle conditions are shown, in which adapter and test stimulus were either on the same or opposite sides, and the delay of the test stimulus was either 50 or 500 ms. The first two panels represent remapping conditions (delay 50 ms), where the test stimulus was either shown at the fixation target (leftmost panel) or at the saccade target (middle left panel). The last two panels depict post-saccadic control conditions (delay 500 ms), where the test stimulus was shown either at the original fixation target (middle right panel) or at the already fixated saccade target (rightmost panel). Labeling of B and C as explained in A.
Figure 3.
Size of the Motion Aftereffect in All Three Experiments.
Bars show means and error bars represent SEM across subjects (N = 7). A MAE size of both the baseline and storage experiment are shown with delays 50 and 500 ms, respectively. B Size of the Phantom MAE: Data from both adapter locations (right and left periphery) were collapsed. C MAE size in the pre-saccadic remapping experiment: Data from rightward and leftward saccade trials representing mirror-inverted conditions were collapsed. Bars are shown in the same order as psychometric functions in Figure 2C. In the first remapping condition (leftmost bar) the test stimulus was shown at the fixation target shortly before the saccade, i.e. in the classical RFs of presumably motion adapted neurons. In the second remapping condition (second leftmost bar) the test stimulus was shown at the saccade target but shortly before the saccade, i.e. following the remapping hypothesis in the future RFs of presumably motion adapted neurons. In the first control condition (second rightmost bar) the test stimulus was shown after the saccade at an unadapted peripheral site. In the second control condition (rightmost bar) the test stimulus was shown after the saccade but at the adapted central retinal site.
Figure 4.
Saccade Latencies in the Remapping Experiment.
Number of saccades in each subject (scale on left y-axis) and mean of all subjects (scale on right y-axis) are plotted against saccade latency. Total saccade count is shown as ochre bar histogram. For clarity, histograms of single subjects are depicted by colored stair functions. All histograms are made up of 35 bins, equivalent to a bin size of approximately 11 ms. The horizontal red and black bars depict times of saccade target and test stimulus presentation. The two black vertical lines enclose the interval of saccade latencies chosen for analysis. A Saccade latencies for trials with a delay period of 50 ms (remapping conditions). B Saccade latencies for trials with a delay period of 500 ms (control conditions).