Learning during Sleep: An Indirect Test of the Erasure-Theory of Dreaming
In this study the hypothesis, put forward elsewhere, that dreams are functional through the erasure of ‘incidental’ and weakly represented information was indirectly explored. 12 subjects were presented paired-associate word lists during each Stage-2 period of their sleep. According to the erasure-hypothesis these associations are destroyed during the subsequent dream if their representation is weak. Two effects might be expected. Firstly, associations which are formed during the last Stage-2 period (which is not followed by a dream-stage) will not be destroyed. Secondly, a stimulus frequency threshold-effect could be expected. Associations which are repeatedly presented (more often than a certain critical number) might become strong enough to withstand the ‘erasure.’ In the present study no indication was found for the latter expectation but a significant effect was found for those associations presented during the last Stage-2 period. Furthermore, there was suggestive evidence that sleep-rehearsal (of previously learned associations) yielded long-term effects.