Abstract
An accumulating body of evidence has suggested that each of sleep and reward is beneficial to visual perceptual learning (VPL). First, we found that sleep not only enhances performance of presleep VPL but also stabilizes presleep VPL to be resilient to retrograde interference by postsleep learning (Sasaki and Watanabe, 2019). Second, we found that reward interacts with post-training sleep to enhance visual plasticity and leads to a greater performance gains of presleep VPL (Berard et al 2015), suggesting that interactions of reward and sleep are generally beneficial for presleep VPL. If so, interactions of reward and sleep should reduce “catastrophic forgetting”, that is, the retrograde interference by postsleep learning with presleep learning. To test the hypothesis, we trained participants on two blocks of a texture discrimination task (TDT) whose background orientations were orthogonal to each other with a 100-min nap between these two blocks. When the two blocks of the TDT with orthogonal background orientations were trained sequentially, learning of the first and second TDTs interfere with each other (Yotsumoto et al, 2009), unless the first learning of TDT was stabilized by sleep (Tamaki et al, 2019). In the reward condition, participants were water-deprived for 5 hours before the experiment and given water drops as a reward for each correct response during the first training before sleep. In the no-reward condition, participants were neither water-deprived nor given water during training. No reward was provided during the second training after sleep in either condition. Contrary to our prediction, retrograde interference was observed in the reward condition, but not in the no-reward condition. These results indicate that interactions of reward and sleep are not always beneficial to presleep learning. Perhaps reward interacts with sleep to enhance visual plasticity, which renders presleep VPL so plastic as to be vulnerable to retrograde interference.