In the Hollywood blockbuster “Inception” (2010), a dedicated team of “dream extractors” is hired to alter a CEO’s decision-making by manipulating his dreams. In the movie, this feat involves a private jet and liters of sedative gas — but a new study suggests they could have achieved a similar effect with only some steel-drum jingles and a comfy bed in a research lab.
The new work shows that audio cues played to sleeping volunteers during rapid-eye-movement (REM) sleep, the stage of sleep when most dreaming occurs, can manipulate dream content.
‘Sleep on it’
If you are stuck on a problem, people often advise you to “sleep on it.” And there’s some scientific evidence to back this up, said study co-author Ken Paller, a cognitive neuroscientist at Northwestern University. For example, in one 2012 study, volunteers asked to solve association-based problems performed better after sleeping than did another group that stayed awake.
But how sleep might achieve this was unclear.
“The motivation for this study was to see if dreaming has something to do with the benefits of sleep we get for problem solving,” Paller told Live Science.
Paller’s team recruited 20 participants who reported having a history of or interest in lucid dreaming, a dream state in which the sleeper becomes aware that they are dreaming and can manipulate their dream to some degree.
Before taking a snooze in the research lab, these participants were tasked with solving puzzles that tested creative cognition within a certain time limit. These included tasks in which volunteers had to alter matchstick diagrams to make certain shapes by moving a limited number of sticks.
As the volunteers considered each puzzle, a short soundtrack played; the tune was unique to each conundrum. These themes included guitar riffs, whistling tunes and steel-drum songs. The puzzles were difficult enough that each participant was left with several unsolved puzzles by the time the testing concluded.
Lead study author Karen Konkoly, who worked on the project while studying dreams in Paller’s lab, also taught the volunteers specific eye movements, with the idea that, if the participants were experiencing lucid dreaming, they could attempt to communicate that to the researchers by moving their eyes.
Then, the researchers fitted the participants’ scalps with electrodes to measure their brain activity and eye movements while they slept. The participants were allowed to watch “Inception” or “Waking Life” (2001), another film about lucid dreaming, while the electrodes were applied.
Hours later, as the volunteers entered REM sleep, the research team, led by Konkoly, began playing soundtracks associated with puzzles they had failed to solve. Immediately afterward, they woke up the participants to record any dreams they’d had in diaries. The participants recorded their dreams over the next two weeks and spent one more night in the lab solving puzzles.
Three-quarters of the volunteers reported having dreams that related to the unsolved puzzles, and the data suggested they were more likely to dream about the puzzles the researchers had cued up with audio. Six dreamers, upon hearing the puzzle soundtracks, signaled to Konkoly they were lucid by moving their eyes or changing their breathing in preset patterns.
The next day, all of the volunteers attempted the puzzles again. The results were mixed.
If certain unsolved puzzles had appeared in the volunteers’ dreams, the volunteers were more likely to solve those puzzles the next day, compared with puzzles they had not dreamed about. The volunteers solved 42% of the puzzles they dreamed about and only 17% of those that didn’t appear in their dreams.
Does lucid dreaming help or hinder?
This finding doesn’t definitively prove that dreams help us solve puzzles, though. It’s possible that the volunteers simply dreamed about the puzzles they were most curious about and were more likely to solve at baseline.
To the authors’ surprise, the volunteers whose eye movements suggested they had lucidly dreamed were less likely to solve the puzzles than those who’d had non-lucid dreams about the puzzles. Paller said the study’s small sample size may have produced this effect.
“I think we didn’t have enough lucid dreams to really be sure about that,” he said.
Emma Peters, a dream engineer at the University of Bern in Switzerland who was not involved in the study, said a major talking point in the field had been whether lucid dreaming might actually impair creative thinking, compared to non-lucid dreaming.
“The idea is, you can do creative problem-solving in dreams because your dreams are so bizarre,” she said, “and they make associations that you would normally not do if you were consciously there.”
For Paller, interpretations of dream research face another important limitation: the other parts of the sleep cycle in which dreams don’t occur as often. At present, it’s impossible to rule out the possibility that brain activity in these stages could be a driving force in creative thinking; the downstream results of that thinking may then emerge in remembered dreams.
But the field is gradually building a picture of what goes on in the sleeping brain. For Paller, these unresolved mysteries are what keeps dream science exciting.
“I think science is fun when there’s still things you need to understand,” he said, “and you haven’t got there.”
Konkoly, K. R., Morris, D. J., Hurka, K., Martinez, A. M., Sanders, K. E., & Paller, K. A. (2026). Creative problem-solving after experimentally provoking dreams of unsolved puzzles during REM sleep. Neuroscience of Consciousness, 2026(1). https://doi.org/10.1093/nc/niaf067
