Sleep learning is a recurring theme in fiction: it occurs as an educational method in Brave New World by Aldous Huxley or as a means of learning a foreign language in the cartoon Dexter. The idea that we can learn in our sleep is fascinating, but whether it’s fantasy or reality has long been a mystery. Thanks to neuroimaging, we know that the brain is far from being idle during sleep and continues to respond to information from the outside world. Up to remembering them and remembering when you wake up?
We’ve known for about a decade that the brain is actually capable of assimilating new information during sleep. It was first singled out for associations between sounds or smells. For example, people who want to quit smoking cut their consumption by 35% by imagining the smell of tobacco during sleep as the smell of rotten fish.
We then tried to see if an idle brain could perform more complex learning, such as learning a foreign language. Together with Sid Kuider of PSL-ENS, and in collaboration with Maxime Elbaz and Damien Léger of APHP/Hôtel-Dieu, we developed a protocol to study the meaning of Japanese words during sleep.
Learning Japanese in your sleep
Japanese is a language with a relatively simple structure, consisting mainly of a sequence of consonants and vowels, for example, the word neko means “cat”. Unlike other Asian languages, it is relatively devoid of accents and has a sound register similar to French. However, the meaning of the words is very far from French. This combination of sounds that are easily recognizable to our ears, but a priori meaningless to us, makes it an ideal language to explore this experience.
Therefore, we recruited 22 healthy young people who did not know Japanese or near-Asian languages. We first presented them with pairs of sounds and images when they are awake, like here in the illustration a dog making a barking sound. As we fell asleep, we played sounds and Japanese translations together, such as the word “inu” (dog) with a barking sound. When they woke up in the morning, we asked them to choose from two images the one that corresponded to the word in Japanese, for example, the word “inu” with the image of a dog and another word presented during sleep, here a bell.
In our study, we reproduced Japanese words associated with sleep sounds, such as a dog barking at a word.
We then noticed that people were accurate in identifying which picture corresponded to the Japanese word. We also asked them if their answers were given randomly or with a certain degree of certainty. Confidence was low and at the same level when the answers were correct and when they were incorrect. This fact suggests that sleep learning is implicit, that is, people do not know what information was received during sleep.
The role of slow waves
The most interesting was the discovery of what happens during sleep. Using electroencephalography, a technique that records the electrical activity of the surface layer of the brain, we were able to predict during learning which words would be remembered upon awakening. Indeed, remembered words generated slower waves than forgotten words. Our results, together with those of a recent publication showing that slow waves predict learning about the relative size of objects, support their important role in sleep learning.
Thus, the sleeping brain can memorize new words and associate meaning with them. This learning can even be observed in the brain rhythms of sleep, which tells us more about their functions. But is this training very useful and accessible to everyone? It is not yet known whether this learning persists over time and whether it depends on individual differences in memory capacity.[Plus de 80 000 lecteurs font confiance à la newsletter de The Conversation pour mieux comprendre les grands enjeux du monde. Abonnez-vous aujourd’hui]
In any case, we know that learning to fall asleep is much less effective than learning to wake up. We performed the same protocol while awake with 10 times fewer repetitions than during sleep. Learning was found to be 5 times more effective while awake than while asleep, and with greater confidence in learned words compared to forgotten words. Thus, learning in a dream, weak and unconscious, is different from fast and conscious learning in the waking state. Although it is possible to learn in sleep, it therefore seems more appropriate to view wakefulness and sleep as complementary, since sleep is optimal for reinforcing information received during wakefulness.
Our current research is focused on the role of the body, and in particular the heart, in the process of remembering sleep. By studying the interaction between the body and the brain during the formation and consolidation of emotional memories, we seek to better understand how the body interferes with sleep learning mechanisms.