Scientists use “dream engineering” to boost creative problem-solving during REM sleep

A recent study has found that dreaming about a specific problem can help people find creative solutions the next day. By playing specialized sounds while participants slept, scientists were able to guide dream content toward unsolved puzzles and improve subsequent problem-solving success. These findings were published in the journal Neuroscience of Consciousness.

Scientists have observed that taking a break from a difficult problem tends to help people find a solution. When a person steps away, their fixation on incorrect approaches begins to fade. At the same time, the brain can form new, unexpected connections between the problem and existing knowledge.

Sleep appears to be an ideal state for this kind of creative restructuring. Historical anecdotes and survey studies suggest that dreams act as a source of creative insight. This seems particularly true for dreams occurring during rapid eye movement (REM) sleep, the phase of the sleep cycle most closely associated with vivid dreaming.

Despite these historical accounts, finding direct proof that dreaming causes creative breakthroughs has proven difficult. Previous studies largely relied on observational data. This means scientists could only show a correlation between recalling a dream and solving a problem later.

This correlation leaves open the possibility that other factors caused both the dream and the eventual solution. For instance, a person with a high level of motivation might naturally dream about a problem and also work harder to solve it the next day. To establish a clearer link, researchers needed a way to control what people dreamed about without waking them up.

“We wanted to test the theory that dreams in REM sleep can promote creative problem solving. The link between REM-sleep dreams and problem solving was supported mostly by anecdotal evidence, and we sought to use new methods in dream engineering to test this theory directly,” said study author Karen Konkoly, a postdoctoral researcher at the University of Cambridge, who conducted the study while at Northwestern University.

The researchers used a technique called targeted memory reactivation. This method involves pairing a learning task with a sensory cue, like a sound, while a person is awake. Later, playing that same cue while the person sleeps helps strengthen the memory of the task.

The researchers applied this technique to 20 adults, focusing primarily on individuals who frequently experience lucid dreams. Lucid dreaming is a state where a sleeping person becomes aware that they are dreaming and can sometimes control the dream’s narrative.

During an evening laboratory session, participants attempted to solve various brain-teaser puzzles. These included matchstick, spatial, and word puzzles designed to require creative thinking rather than a simple step-by-step approach. Before starting each puzzle, participants listened to a unique 15-second musical or environmental soundtrack to build an association. Participants were given three minutes to work on each puzzle while the specific soundtrack played in the background.

Experimenters gave participants real-time feedback on any proposed solutions during this evening session. The participants continued this process until they had accumulated exactly four unsolved puzzles for the night. Afterward, the scientists attached sensors to the participants’ heads and faces to monitor their brainwaves, eye movements, and muscle activity.

The participants then went to sleep in the laboratory for the first half of the night. At 4 a.m., researchers briefly woke the participants to administer a short training session designed to promote lucid dreaming as they returned to sleep.

Once the monitoring equipment indicated that a participant had entered rapid eye movement sleep, the researchers quietly played the soundtracks associated with two of that person’s four unsolved puzzles.The sounds were kept at a low volume to avoid waking the sleeping individuals.

Before going to sleep, the participants had been instructed to listen for these sounds in their dreams. If they realized they were dreaming and heard a puzzle’s soundtrack, they were told to signal the researchers in the real world. They did this by performing a specific pattern of rapid in-and-out sniffs. This sniffing movement can be detected by the recording equipment even while the body remains paralyzed in sleep. The researchers awakened the participants after these targeted sleep phases to ask for detailed dream reports.

“This is a relatively new method of communication that we’ve been using a lot at Northwestern, and we find it to be an excellent way to communicate with sleepers,” Konkoly told PsyPost. “It was also fascinating to see how dreamers can respond to cues during sleep, letting us know which problems they are working on, but then forget the details by the time they wake up. The real-time communication feels like we are able to rescue information about dreams that the dreamers themselves have forgotten.”

The next morning, the participants were given another chance to solve the puzzles from the previous night. The scientists found that the sound cues successfully influenced dream content. Cued puzzles appeared in dreams significantly more often than puzzles that were not paired with overnight sounds.

In total, 15 of the 20 participants reported at least one dream that conceptually included elements of an unsolved puzzle. Puzzles that made their way into these dreams were much more likely to be solved the next day. On average, participants solved 42 percent of the puzzles they dreamed about conceptually.

They only solved 17 percent of the puzzles that did not appear in their dreams. Notably, simply hearing the sound in the dream without incorporating the puzzle’s concept did not improve solving rates. The sound cues did not automatically increase problem-solving success for everyone.

“‘Dream engineering’ allowed us to remind people of specific unsolved problems while they were sleeping, and people were more likely to solve a problem if they dreamed about it the night before,” Konkoly said. “This study provides preliminary causal evidence that dreaming can promote creative problem solving.”

The technique worked exceptionally well for a subgroup of 12 participants whose dreams were heavily influenced by the sounds. For these specific individuals, the sound cues doubled their problem-solving success rate on the cued puzzles compared to the uncued ones. Conscious awareness during the dream was not strictly necessary for success.

Participants actually solved a higher percentage of puzzles after having regular, non-lucid dreams about them. Some individuals even responded to the sound cues with the sniffing signal in their sleep but could not remember the dream upon waking. Even without remembering the dream, these participants still went on to solve the puzzle at a high rate.

“I was most surprised that the cues we presented during sleep successfully influenced non-lucid dreams,” Konkoly said. “Even without realizing they were dreaming, some dreamers seemed to follow our instructions, for instance asking other dream characters for help solving the puzzles whose sounds we were presenting.”

“Many problems in the world today require creative solutions. By learning more about how our brains are able to think creatively, think anew and generate creative new ideas, we could be closer to solving the problems we want to solve, and sleep engineering could help,” said senior author Ken Paller, the James Padilla Professor of Psychology and director of the cognitive neuroscience program in the Weinberg College of Arts and Sciences at Northwestern.

But the sound cues did not produce a universal problem-solving boost across the entire group. Because the overnight sounds only improved puzzle-solving for the participants who successfully incorporated the cues into their dreams, the researchers could not definitively prove a direct chain of events for all participants.

The study also relied on a relatively small sample size. Participants tackled only four unsolved puzzles per night, which limits the statistical power of the findings. Future experiments with more participants and a larger variety of tasks would help confirm these effects.

It also remains difficult to separate the exact moment of the dream from the unconscious mental processing that happens afterward. A person might dream about a puzzle and then continue to process the information unconsciously before waking up. Determining exactly when the creative insight occurs is still a challenge.

The researchers also noted that their study focused on convergent creativity, which involves finding a single correct answer to a specific problem. Future research could explore whether dreaming also boosts divergent creativity, which involves generating many novel ideas or alternative uses for everyday objects. The scientists also plan to investigate why certain people incorporate sounds into their dreams while others block them out.

“I’m curious to use methods like these to test other hypothesized functions of dreaming,” Konkoly said. “In addition, I was personally fascinated to spend so many nights presenting cues to participants, and seeing how some got incorporated into dreams vertically, some in a subtle or distorted way, and some ignored all together. How could the same stimulus, in the same state of consciousness, be perceived in such different ways?”

The study, “Creative problem-solving after experimentally provoking dreams of unsolved puzzles during REM sleep,” was authored by Karen R Konkoly, Daniel J Morris, Kaitlyn Hurka, Alysiana M Martinez, Kristin E G Sanders, and Ken A Paller.