There is general consensus that performance on a number of cognitive tasks deteriorates following total sleep deprivation. At times, however, subjects manage to maintain performance. This may be because of an ability to switch cognitive strategies including the exertion of compensatory effort. The present study examines the effects of total sleep deprivation on a semantic word priming task. Word priming is unique because it can be carried out using different strategies involving either automatic, effortless or controlled, effortful processing. Twelve subjects were presented with word pairs, a prime and a target, that were either highly semantically associated (cat…dog), weakly associated (cow…barn) or unassociated (apple…road). In order to increase the probability of the use of controlled processing following normal sleep, the subject’s task was to determine if the target word was semantically related to the prime. Furthermore, the time between the offset of the prime and the onset of the target was relatively long, permitting the use of an effortful, expectancy-predictive strategy. Event-related potentials (ERPs) were recorded from 64 electrode sites. After normal sleep, RTs were faster and accuracy higher to highly associated targets; this performance advantage was also maintained following sleep deprivation. A large negative deflection, the N400, was larger to weakly associated and unassociated targets in both sleep-deprived and normal conditions. The overall N400 was however larger in the normal sleep condition. Moreover, a long-lasting negative slow wave developed between the offset of the prime and the onset of the target. These physiological measures are consistent with the use of an effortful, predictive strategy following normal sleep but an automatic, effortless strategy following total sleep deprivation. A picture priming task was also run. This task benefits less from the use of a predictive strategy. Accordingly, in this task, ERPs following the target did not differ as a function of the amount of sleep.
Read full abstract