Purchase this article with an account.
Anna Schapiro, Timothy Rogers, Kenneth Norman, Lang Chen, Elizabeth McDevitt, Sara Mednick; The role of sleep in consolidating semantic knowledge. Journal of Vision 2013;13(9):666. doi: 10.1167/13.9.666.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Though sleep is thought to play an important role in many domains of learning and memory, the impact of sleep on the acquisition of new semantic information has remained relatively unexplored. We investigated how sleep affects learning the names, category memberships, and parts of 15 novel objects ("satellites") organized into three classes. Each satellite had five visual parts, four shared by other category members and one unique to the item, as well as a class name shared by other category members and a unique code name. To learn this structure, participants guessed, with feedback, the missing part or name of one satellite on each trial. After reaching a criterion of 66% correct, participants were tested (without feedback) immediately and again later, after having slept or not, on their memory for unique and shared properties of the items, and on generalization to novel category members. In Experiment 1, half of the participants began in the evening and were tested 12 hours later after a full night of sleep, and half began in the morning and were tested after 12 hours of wake. Memory for shared properties and generalization improved over both sleep and wake, but memory for idiosyncratic properties improved only with sleep, declining over wake. Over both sleep and wake, generalization improved more for category exemplars closer to the prototype. To assess the contributions of different sleep stages to these effects, participants in Experiment 2 took polysomnographically-recorded naps between tests that contained either non-rapid eye movement (REM) sleep only or non-REM and REM sleep. Preliminary results suggest that non-REM benefited unique properties, while REM (and not simply additional sleep) benefited shared properties. We developed a neural network model that accounts for these results in terms of stronger hippocampal influence on autonomous offline learning during NREM than during REM or wake.
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only