Abstract
A number of studies have demonstrated that reward evokes task-irrelevant perceptual learning (Seitz et al, 2009). However, the underlying mechanisms are still unknown. There are at least 2 possible hypotheses for the mechanisms: a reinforcement hypothesis or an alertness hypothesis. A critical factor to dissociate the two hypotheses is the timing between the reward signal and the paired visual stimuli. If the reinforcement hypothesis is true, reward needs to follow the rule of contiguity. Thus, learning should occur only for a visual stimulus presented after the reward is given. On the other hand, according to the alertness hypothesis, reward enhances alertness to a stimulus presented temporally close to reward presentation, irrespective of whether it is presented before or after the stimulus. Thus, if the alertness hypothesis is the case, learning should occur for a visual stimulus presented both before and after the reward is given. To test which hypothesis is true, we manipulated the timing between an exposed visual stimulus and reward. Prior to each of 12 days' training sessions, subjects underwent 5 hours of fasting. In training, a sequence of two orientations was exposed in a random order to the trained eye, whereas dynamic color patches were presented to the other eye to perceptually erase the orientations in the other eye (CFS paradigm). One of the orientations was paired with the reward presented 400 ms prior to the stimulus onset, and the other orientation paired with reward delivered 400 ms after the stimulus onset. Subjects' sensitivity to the 2 orientations was tested by a detection task in both eyes before and after training. Subjects showed significant improvement in the trained eye for both reward timings. There was no improvement in the untrained eye. These results are consistent with the alertness hypothesis for task-irrelevant perceptual learning in association with reward.
Meeting abstract presented at VSS 2017