In Aldous Huxley’s Brave New World, a boy memorizes every word of a conference in English, even though he doesn’t know the language… While he’s sleeping near a radio station that broadcasts the conference. When the child wakes up, he or she is able to recite it in its entirety. Based on this discovery, the totalitarian authorities of the dystopic world of Huxley adapted the method to shape the unconscious of all citizens.

Everyone knows that we learn better when we are well-rested. On the other hand, most of us think a priori that it is impossible to learn in our sleep. Yet new scientific discoveries suggest that the facts are not so simple: an essential part of learning happens well during sleep.

Indeed, newly formed memories resurface during the night, and this re-reading helps to “consolidate” them, or even to memorize some of them permanently.

And if the scenario imagined by Huxley in 1932 remains a fiction today, experiments indicate that it is possible to manipulate the memories of a person immersed in a deep sleep, laying the foundations for a new science of learning during sleep.


For such learning to work, the focus is on exploring how a person can absorb information at a time when their consciousness seems to be at rest.

By the time Huxley wrote his novel, serious research was already beginning on the possibility of intervening during sleep. In 1927, New Yorker Alois Saliger invented an “automatic suggestion machine with a delayed start”, marketed under the name PsychoPhone, capable of re-transmitting a recorded message during the night. Then, in the 1930s and 1940s, various studies provided some alleged examples of learning during sleep…

One of them, published in 1942, described an experiment that Lawrence LeShan (then at William and Mary College in the United States) conducted in a summer camp where many boys bit their nails.

In a room where twenty boys slept, LeShan installed a portable phonograph that repeated the sentence “My nails taste terribly bitter” over and over again. And this, 300 times a night, 150 minutes after falling asleep, for 54 consecutive nights (even during the last two weeks when, after the phonograph broke down, LeShan himself pronounced the sentence).

Eight of the twenty boys stopped biting their nails, while none of the twenty others who slept without being exposed to the recording had changed their behavior.

However, this result remained suspicious, as the researcher had not used any physiological monitoring device to verify that his subjects were asleep when the phonograph was in use.


The field of research underwent a dramatic turnaround in 1956 when RAND Corporation scientists used electroencephalography (a non-invasive brain imaging technique that records brain waves related to different sleep phases) to record the brain activity of participants to which 96 questions and answers were read during their sleep. For example: “In which type of store did Ulysses Grant work before the war? “and the answer: “A hardware store. »

The next day, participants could only remember the answers when they had been read to them at times when the electroencephalogram showed signs of waking up. In other words, participants seemed to have only learned when they were not really sleeping… A hard blow to this field of research, which instilled doubt in the minds of scientists about the reality of learning while sleeping.

For fifty years, many researchers have therefore avoided the subject. Nevertheless, others have focused more generally on the possibility that sleep may have a beneficial influence on memory.

A typical study design was to determine whether night-time sleep deprivation disrupted the memorization of information learned the day before. Another tested whether memory was better after a nap or after a similarly long period of wakefulness. But here again, various factors are likely to interfere with such studies. For example, stress related to sleep deprivation often affects cognitive functions, which then reduce memory performance.

While recognizing the difficulties, cognitive neuroscience researchers have taken up the challenge and have accumulated evidence in support of the idea that sleep is a way to revive memories acquired the day before. The link between sleep and memory then resurfaced.


At first, many researchers focused on so-called periods of Rapid Eye Movement sleep (REM), better known as REM sleep, where dreams are the most frequent and striking. Why? Why? Because we thought that the processing of memories during the night was linked to dreams. But few data have confirmed it.

In 1983, two renowned British biologists, Graeme Mitchison and Francis Crick, even speculated that REM sleep was used to forget. In the same context, in the early 2000s, Giulio Tononi and Chiara Cirelli, both at the University of Wisconsin at Madison, proposed that sleep was a time when connections between brain cells were weakening, giving way to new information acquired the next day.


Gradually, scientists became interested in another period of sleep without fast eye movements and dreams: Deep Slow Sleep.

In 2007, Björn Rasch, then at the University of Lübeck, Germany, and his colleagues organized an experiment where participants had to memorize the location of a set of objects while smelling a rose. Later, in their beds in the laboratory, subjects were again exposed to the same smell at different stages of sleep, confirmed by electroencephalography.

The smell activates the hippocampus, an essential brain region to learn how to find one’s way around and to move around in one’s environment, and to store new knowledge that is acquired. Upon awakening, participants then remembered the locations more precisely, but only if the smell had been presented to them during Deep Slow Sleep, not during REM sleep.


In 2009, the laboratory extended this method by using sounds rather than smells. They found that the sounds played during deep slow sleep improve the memorization of the location of unique objects (instead of a whole set of objects, as in the study with the smell of rose).

In this method, called Targeted Memory Reactivation (TMR), volunteers first memorize the locations of fifty objects. For example, they learn how to place a cat in a designated place on a computer screen and a teapot in another. At the same time, they hear a sound corresponding to each object – a mewing for the cat, a whistling for the teapot… After this learning phase, the participants take a nap in a comfortable place in the laboratory.

Using electrodes placed on their heads, researchers monitor their brain activity to ensure that everyone is sleeping well. In fact, they are targeting a particularly important time for memory reactivation, when the activity of neural networks in the cerebral cortex, the upper layer of the brain, is synchronized.


As soon as the team of researchers detects a Deep Slow Sleep phase, they play a sound; whistling or other sounds associated with objects in the learning phase.

The sounds are played softly, no louder than the background noise, to avoid waking the participants. But when they wake up, they remember the positions of the objects better than when there is no sound.

Thus, whatever sensory signal was used – sounds or smells – it clearly triggered the reactivation of spatial memory and attenuated the mechanisms of forgetting.

At first, this memory reactivation technique was very controversial. One reason was that sleep specialists generally thought that the sensory circuits in the cortex were largely deactivated when sleeping, with the exception of those of the sense of smell.

Nevertheless, researchers continued to follow their intuition. And in fact, other subsequent studies have shown the same benefits on memory. Functional magnetic resonance imaging has even highlighted the brain regions involved in memory reactivation, and electroencephalography has revealed the importance of specific brain wave oscillations in memory.

In particular, two studies published in 2018, one by Scott Cairney of York University in England and his colleagues, the other by James Antony of Princeton University in the United States, and his colleagues associated a particular oscillation in the brain waves, called a “sleep spindle“, with the same memory reactivation benefits shown above.


In addition to spatial memory, the memory reactivation technique has made it possible to consolidate memory in other contexts, such as mastering a piano melody and memorizing new vocabulary or grammatical rules.

It is also capable of reducing reflexes conditioned by fear.
For example, in 2013, researchers at Northwestern University in the United States showed participants neutral faces while exposing them a neutral smell; but some faces were associated with a slight electric shock.

Then, during sleep, they exposed the smell from seeing a “frightening” face (associated with the electric shock): this simple reactivation was sufficient to reduce, the next time they woke up, their emotional reaction to that face.

Ongoing studies are examining other types of memorization, such as associating names with faces when first meeting new people.

As it evolves, this technique should be tested in the treatment of various disorders, for example to reduce addiction or accelerate recovery from disease.

Researchers, such as the neurologist Marc Slutzky of Northwestern University in the United States, are currently testing a new rehabilitation procedure to recover motor skills in the arms of stroke patients.

Sounds are part of the therapy and are replayed during their sleep in an attempt to accelerate the relearning of lost movements. The prospects seem promising, as SCIENTISTS have already shown that this method can change similar forms of motor learning in healthy people.


As shown above, It is therefore possible to consolidate memories. Will it also be possible to load new information into the brain of a sleeping person – at the risk of reviving the ethical spectrum of mind control invoked by Huxley?

Anat Arzi, at Cambridge University, and her colleagues have managed to create relatively simple memories using smells. In an experiment published in 2014, researchers reduced the desire for tobacco among smokers who wanted to quit smoking.

Once asleep, participants were exposed to two smells: cigarette smoke and rotten fish. A week later, those who had smelled the sequence of two smells consumed 30% less cigarettes. Their brains had associated tobacco with the repellent smell of fish.

Acquiring more complex memories could also become a reality. In 2015, Karim Benchenane of the Graduate School of Industrial Physics and Chemistry in Paris, and his colleagues, showed how to create a false memory in mice.

The team knew that when a mouse explores a new environment, a specific set of neurons, called place-cells, are activated as the animal passes through specific parts of a pen. These same neurons light up again during sleep when the memory of the journey seems to be replayed.

The researchers stimulated the mouse reward circuit (brain areas associated with pleasure and motivation) at the very moment when place-cells were spontaneously active during their sleep.

Surprisingly, the mice then spent more time at the locations corresponding to the place-cells associated with the stimulus, moving there immediately after waking up.

However, further experiments are needed to determine whether false memories were implanted in mice during sleep or whether the rodents were actually guided by a conditioning process.


Recent work, also on place-cells, has clarified the mechanisms of memory enhancement during sleep. When a rodent moves, its place-cells are activated sequentially, “registering” the movement in the hippocampus. At the same time, this brain region emits much faster, “entangled” sequences.

So, which of these sequences allow reactivation during sleep when it is time to consolidate the memory of the movement?

To answer this question, Michaël Zugaro and his colleagues (from the Interdisciplinary Center for Research in Biology in Paris) placed rats… in the wagon of an electric train equipped with a conveyor belt.

When the train is running, but the conveyor belt is inactive (rodent movement is passive), only the slow sequences are activated. This is not the case when the carpet is running: both types of sequences are activated.

By observing the hippocampus of rodents during their sleep, biologists found the activation sequences corresponding to the movement only in the case where the conveyor belt had been activated. They deduced from this that the entangled sequences are essential for the formation of memories.


Learning new knowledge vs. Consolidating old one

Reactivating learning in a targeted way during sleep with sounds that have previously been associated with it, as Delphine Oudiette and Ken Paller did in their work, is one thing. Hypnopedics (i.e. teaching through sleep) is another.

Teaching consists in directly creating new associations during sleep, for example an unpleasant smell mixed with tobacco, which then influences behaviour once awake: smokers subjected to this association reduce their cigarette consumption, as shown by the Anat Arzi team at Cambridge University in 2014.

These two paths have each provided new information on the mechanisms for consolidating new learning during sleep.

It matters when exactly, while sleeping, to try to reinforce learning

Targeted Memory Reactivation (TMR) has thus contributed significantly to showing that neural activity outbursts called sleep zones play an important role in the consolidation of so-called declarative memory, in conjunction with other neurophysiological markers of slow wave sleep.

When, during slow sleep, a participant is presented with a sound related to prior learning, his brain activity increases in the 12-15 hertz frequency band (which corresponds to sleep zones), and he remembers the associated words better the next day.

However, scientists have shown that if a second auditory stimulation is presented quickly (less than 1.4 seconds) after the first one, then the activity related to the spindles is abolished and the benefit of the technique is lost.

These results suggest that the sleep spindle is an important element for the propagation of information and its transfer from the hippocampus to the cortex. They also show that the success of Targeted Memory Reactivation procedures depends on the parameters used to present stimuli during sleep.

When it’s too complicated…

sound patterns for learning during sleep
Our brain detects sound patterns in the awake state (activated brain regions, seen by magnetoencephalography, shown in upper left), but not during deep slow sleep (upper right), while in both states it perceives isolated sounds

Recent research has also shown that it is impossible to create too complex associations while sleeping. By monitoring the brain activity of participants during sleep, their brains do not detect the regularities of a sound flow, whereas they detect them when they wake up, even if the participant is not aware of it (see the figure above).

Our ability to learn during sleep is therefore limited to elementary associations. Probably because our sleeping brain needs to be protected from the environment to initiate the recovery processes that allow us to be efficient when we wake up.

The elaborate processing of information presented during sleep would be confused with the mechanisms underlying memory consolidation.

It could disrupt other important processes during sleep

Moreover, there is nothing to suggest that the consolidation processes linked to external stimuli depend on the same neural networks as those set up during spontaneous consolidation during sleep.

For example, it has been known for more than a decade that the concentration of a neurotransmitter, acetylcholine, decreases in the brain during slow sleep, and that without this decrease, the brain is not able to consolidate new learning in declarative memory. However, in January 2018, Jens Klinzing of Tubingen University and his colleagues showed that Targeted Memory Reactivation enhance participants’ memory even when acetylcholine concentrations are kept high during slow sleep.

This suggests that at least two distinct brain strategies lead to the consolidation of new memories, and that Targeted Memory Reactivation, hypnopedics and spontaneous consolidation during sleep should not be confused.

Finally, let us not forget that our sleep takes place in a protected environment. What effects would the systematic application of stimuli have on the other essential functions it performs, in particular our biological balance and the restoration of our capacity for vigilance?


A study conducted in 2012 by psychologist Carmen Westerberg is leading the way. Carmen Westerberg has performed tests in patients with memory disorders that often precede Alzheimer’s disease.

She found a correlation between poor sleep quality and a reduced ability to remember information after a one-night delay.

All this knowledge may lead to learning while sleeping programs to preserve memories, accelerate the acquisition of new knowledge and skills, and even change bad habits such as smoking.

Sleep is necessary not only to stay alert and recharge your batteries, but also to consolidate the memories acquired during waking up.

We still have a lot to learn about how memory works, including what mechanisms preserve memories. It is also essential to know more about the dangers of insufficient or inadequate sleep, whether due to life stress, various illnesses, or aging.


It has long been thought that, during sleep, the brain was turned off and unable to learn. But research over the past decades has shown that it remains active when you sleep, and that recent memories are reactivated.

By controlling this reactivation of memory, neuroscientists have improved the learning of new data during a specific phase of deep sleep. Their method would have medical applications, for example for the treatment of addictions.

The limits of the possibilities remain to be explored, but this research has already established that a normal component of learning continues at night when the brain is disconnected from the outside world.

In the longer term, scientists could also examine whether it is possible to take control of dreams. A new world would then open up, where therapies for nightmares would be explored, problems would be solved while sleeping, and even adventures of our choice would be dreamed of.

All these are new ways of transforming daily rest periods into productive actions, in a society that already offers many gadgets to record physical activity and genetic tests. A frightening prospect for some, but for others, a new opportunity to unlock the secrets of the mind.


Jim Miller

Jim spent his twenties trying tounderstand how our primitive minds get in the way of self-growth. He has a lot of interesting ideas about human psychology that he doesn't hesitate to share when opportunity presents itself.

Subscribe to our newsletter

Be the first to receive the latest articles and exclusive offers on our products directly in your inbox

newsletter signup form

*You can unsubscribe any time via the link provided in Newsletter.

Related Articles


Jim Miller

Jim spent his twenties trying tounderstand how our primitive minds get in the way of self-growth. He has a lot of interesting ideas about human psychology that he doesn't hesitate to share when opportunity presents itself.

Subscribe to our newsletter

Be the first to receive the latest articles and exclusive offers on our products directly in your inbox

newsletter signup form

*You can unsubscribe any time via the link provided in Newsletter.