More Evidence That Dreams Reflect Learning During Sleep

A recent study shows dreaming of a picture task correlates with better memory.

Posted May 24, 2018

There is a general consensus that sleep is important for learning. It is thought that memory traces are reactivated in the brain and then actively strengthened during sleep. Several researchers have proposed that in dreams, we may be witnessing this memory processing as if it were a real experience, our own virtual reality. This has been supported to some extent by findings that dreams are often made up of snippets of recent experience and of autobiographical memories. Also, certain qualities of dreams, such as emotional intensity, are associated with sleep neurophysiology (for an example, see this post on dreaming and NREM sleep spindles).

Nevertheless, there is relatively little experimental evidence that tasks learned prior to sleep are incorporated into dreams. I previously described one experiment in which an alpine racing video game was incorporated into sleep onset dreams, and this later correlated with performance, suggesting that dreaming of the task was associated with memory strengthening during sleep.

Recently, researchers at the Zurich Center for Interdisciplinary Sleep Research conducted an overnight study to assess whether a word-image association task learned prior to sleep would be incorporated into dreams, and whether this would then relate to performance the following morning. They were also interested in whether repeated awakenings for dream recall across the night would impact memory consolidation. This is one potential problem in dream research, as it is necessary to repeatedly awaken participants in order to collect their dreams, this could interfere with the memory processing that occurs during sleep.

The researchers recruited twenty-two participants between the ages of 19-35 for the study. On the experimental nights, participants came to the sleep laboratory and were hooked up to polysomnography, including electrodes on the scalp to measure brain activity, along with electrodes around the eyes to measure ocular movements that occur during REM sleep, and cardiac and muscle tension electrodes. Participants then completed a word-picture association task, which required them to first learn associations between words and pictures, and they were then tested on their memory for these associations later.

During the experimental night, participants were awakened 3-6 times via an intercom and asked “What went through your mind before you woke up?”. These awakenings occurred in both NREM and in REM sleep, and every participant managed to recall at least 1 dream. In the morning participants completed a recall task again for the word-picture associations.

This condition with multiple awakenings was compared to a condition where participants were not awakened at all throughout the night, in order to see whether repeated awakening interferes with memory consolidation.

To assess whether the dreams were related to the pre-sleep task, two raters looked for any correspondence between the categories of pictures presented prior to sleep and the content of the dream. To elaborate, the task presented prior to sleep used neutral and positive pictures from 3 categories; on one night, the pictures contained children, sports, and animals; on the other night, images contained water, transportation, or food. The raters examined the dreams reported by participants to look for any content related to these categories. From this, the experimenters could determine to what extent participants dreamt about the categories which were presented just prior to sleep (as compared to the categories which were not presented, and which may appear in dreams anyway just by chance).

Following analyses, it was found that repeated awakening reduced sleep quantity, and increased amount of time spent awake during the night. Despite this interference with sleep quantity, there was no difference in memory performance following the night with multiple awakenings, as compared to the night without awakenings. This is perhaps surprising, but also reassuring, and suggests that even with repeated awakenings, the process of memory consolidation continues unimpeded.

Overall, there were 106 dream reports (of 121 awakenings) for the 22 participants in the condition with multiple awakenings. The authors found that the dreams from this night incorporated the picture categories that had been presented prior to sleep, moreso than the categories that had not been presented. This was not found for dreams that were reported only in the morning following a night of uninterrupted sleep. From this finding, the authors suggest that dream researchers should use multiple awakenings across the night if they hope to observe task-related content in dreams.

Finally, the authors found a positive correlation between the amount of task-related content in NREM dreams, and performance following sleep. This supports the theory that dreaming of a task corresponds with learning, and that dreaming may reflect memory consolidation that is ongoing during sleep. However, this correlation was not found for REM dreams, which may be due to the task relying more on NREM sleep for consolidation.

Overall, this study joins several others in the dream research field that are using multiple awakenings across the night in order to better understand the quality and quantity of dream recall as it changes across sleep stages and across the night. That repeated awakenings do not interfere with memory consolidation bodes the question of whether memory processing continues at some level despite awakening, or whether it can pick up where it left off at each sleep onset to some extent.

Finally, that higher incorporation of task stimuli in NREM dreams correlated with memory also provides solid evidence supporting the theory that dreaming reflects memory consolidation, and points to the usefulness of collecting dreams to better understand memory processing during sleep.


Schoch, S., Cordi, M. J., Schredl, M., & Rasch, B. (2018). The effect of dream report collection and dream incorporation on memory consolidation during sleep. bioRxiv, 323667.