Skip to main content

Verified by Psychology Today

Theater of the Mind

Dreams have been described as dress rehearsals for real life, opportunities to gratify wishes, and a form of nocturnal therapy. A new theory aims to make sense of it all.

Erik Madigan Heck/Trunk Archive, used with permission
Erik Madigan Heck/Trunk Archive, used with permission

Excerpted from When Brains Dream: Exploring the Science and Mystery of Sleep by Antonio Zadra and Robert Stickgold. Copyright (c) 2021 by Antonio Zadra and Robert Stickgold. Used with permission of the publisher, W. W. Norton & Company, Inc. All rights reserved. This selection may not be reproduced, stored in a retrieval system, or transmitted in any form by any means without the prior written permission of the publisher.

I’m walking along a beach. I know my parents are a short distance behind me. I look out at the ocean and see a giant, pink capital letter A rising out of the water. It says, “I am the letter A; follow me!” Its voice is deep and powerful, like the voice of God in some movies. I enter the water and try to swim toward the giant letter, but it keeps receding and the waves keep getting larger. I wake up.

This is how a young woman who took part in one of our dream studies described her recurrent dream. Many dreams feature such unimaginably wide-​ranging—​and often inimitably bizarre—​content that they simply defy classification. And yet, sometimes dream features come together to create a dream with thematic content that is experienced by a large proportion of the population and that has been described across time and cultures. If you’ve ever dreamed of falling, of being inappropriately dressed, or of being unprepared for an exam, then you’ve experienced one of these common types of dream.

What function do these dreams serve? And why do we have to experience the dream for this function to be fulfilled? As we’ll soon see, we’ve got some pretty good ideas about what the most likely answers are. Countless theories of why we dream have been proposed, from Sigmund Freud’s assertion that dreams allow the expression of repressed wishes to the notion that dreams are essentially random and inscrutable. We’ll propose a new model, which we call NEXTUP—​network exploration to understand possibilities. It describes dreaming as a special form of memory processing: Dreams allow for the discovery and strengthening of unexplored associations between previously formed memories.

Dream content isn’t simply random. But dreams rarely depict concerns from our waking lives directly or offer concrete solutions to them. Instead, they identify and strengthen associations that in some way embody our concerns, associations that the brain calculates may be of use in resolving these or similar concerns—either now or in the future.

We will go deeper into the why of dreaming, but first, let’s take a closer look at the what. What do people most typically dream about?

Rodney Smith/Trunk Archive, used with permission
Rodney Smith/Trunk Archive, used with permission

The Stuff of Dreams

Our garden-variety, everyday dreams are as different from one another as are the events in our waking life, but certain patterns and preferences are seen in what we dream.

The vast majority of adults—at least 85 to 90 percent—will tell you that they dream. But we recall only a fraction of what we dream about. People can dream in all sleep stages; so we’re immersed in various forms of dreaming for at least two-​thirds of the night, ​and some researchers would even say we dream all night long. If you’re one of those lucky souls who falls asleep quickly and sleeps soundly through the night, it’s unlikely that you recall even 5 percent—​20 minutes—​of those dreams, and most commonly it’s the dream you were having right before waking up.

Virtually all dreams have a narrative structure and contain the dreamer as an active participant. They are typically experienced from an embodied, first-​person perspective. Yet we’re rarely alone in our dreams. Most dreams contain at least two other characters. About half of our dream characters are familiar to us—​relatives, friends, colleagues, or acquaintances​—while the other half are unknown, including strangers and people identified solely by their occupational role, such as policemen, doctors, or teachers.

When we look at the gender of our dream characters, a peculiar finding emerges: Women’s dreams contain an equal proportion of male and female characters, but men’s dreams contain twice as many male characters as female. Why this difference exists is still under debate, but it has been documented in numerous studies, across cultures, and even in the dreams of young girls and boys. Furthermore, physical aggression is significantly more common in men’s dreams than in women’s—​and women, in turn, are more likely than men to be victims of aggression in their dreams, mirroring differences in actual experiences seen in most cultures.

At a more global level, the dreamer—​or another character—​is usually faced with some kind of problem. These can range from relatively minor difficulties—​planning a course of action, trying to make sense of a situation, or finding a lost object—​to such serious physical or psychological dangers as being lost, falling ill, facing interpersonal conflicts, or dealing with environmental hazards. Misfortunes—​mishaps the character cannot avoid—​are seen in about a third of all dreams. That rate is, sadly, seven times higher than the rate of good fortunes. Nevertheless, we succeed in handling difficulties in dreams as often as we fail.

Dream themes are called typical when many people report having had them at least once. Indeed, people have been offering explanations for these dreams—​of being chased, of falling, or of losing one’s teeth—​for millennia. (According to a dream interpretation dictionary written by the Duke of Zhou in the 11th century BCE, dreams of your teeth falling out mean that your parents may suffer a misfortune.) So it’s somewhat surprising that the first major scientific study of typical dreams didn’t appear until 1958, when researchers investigated the prevalence of 34 typical dreams in Japanese and American students.

Although some cross-​cultural differences were noted—​Americans reported fewer dreams of fire and more of nudity—​the similarities were striking. In both groups, dreams of being attacked or pursued; of falling; of trying again and again to do something; of school, teachers, or studying; and of sexual experiences all figured—​with almost identical rank orderings—​among the six most frequently reported dream themes.

It was another 40 years before one of us (Tony Zadra) and fellow Canadian dream researcher Tore Nielsen followed up on this study. One key finding was how consistent students’ prevalence profiles of typical dreams were from one year to the next as well as across student populations from different regions of Canada. Another, even more striking result was how stable the typical dream profiles had remained over decades. For example, the four most frequently reported dream themes in the 1958 study were all among the top five reported by Canadian students in the late 1990s and early 2000s. Subsequent studies in Germany and Hong Kong have revealed remarkable similarities in the dream themes across all of these populations.

Anronio Zadra Ph.D and Robert Stickgold Ph.D
Anronio Zadra Ph.D and Robert Stickgold Ph.D

Several interesting observations can be made. First, no typical dream is reported by everyone; only four of the top 15 themes have a prevalence exceeding 70 percent. Themes of being chased or pursued, of sexual experiences, of school, teachers, and studying, and of falling are the most commonly reported dreams in both men and women—though the studies also show consistent gender differences. And, many of the typical dreams most frequently discussed in popular media aren’t all that common. For instance, fewer than 30 percent report ever having dreamed of being unable to find (or embarrassed about using) a toilet, or of their teeth falling out, or of finding money.

These numbers reflect lifetime prevalence, the percentage of people who have had such dreams at least once in their lives. They don’t tell us how often people had the dreams. When Tony evaluated the thematic content of 3,000 randomly selected dreams from 450 individuals, he found that only five typical dreams—​falling, flying, a person now dead being alive, being inappropriately dressed, and being unable to find or use a toilet—​occurred in more than 3 percent of the dream reports. Still, when you add up the frequencies of all these categories of typical dreams, there’s a better than 50-50 chance that one of the themes will pop up in your dreams tonight.

What Are Our Dreams For?

Whatever the function of dreams may be, it cannot depend on our remembering them once we awaken. As mentioned earlier, people recall only a small fraction of their dreams. Even those dreams that are remembered, unless written down, tend to be evanescent, their details quickly fading as we make our way through the day. It’s important to distinguish between the uses we choose to make of the dreams that we do remember—​for interpretation, personal growth, inspiration, or entertainment—​and the biological or adaptive function of all of our dreams, including those we do not remember.

Over millennia, hundreds of ideas have been put forth to explain the nature and function of dreams. Perhaps the most famous are the theories of Freud, who claimed that dreams allow the partial expression of repressed wishes (of a sexual or, at times, aggressive nature), and Carl Jung, who suggested that dreams play a vital role in the development of personality. Given that decades of research found little or no empirical support for Freud’s model of dreams, the overwhelming majority of sleep and dream scientists have long abandoned Freudian conceptualizations. While some Jungian ideas about dreams find an echo in contemporary clinical models of dreaming, these models are largely driven by new discoveries and ways of thinking about the nature of the human mind and brain function.

But what if dreams are instead nothing more than a meaningless by-​product of the sleeping brain? Next to the dream theories of Freud and Jung, the activation-​synthesis hypothesis proposed by Allan Hobson and Robert McCarley of Harvard Medical School is probably the most widely known theory of dreaming. In a pair of articles published in 1977, Hobson and McCarley presented a model of dreaming based on the neurobiology of REM sleep.

In a nutshell, activation-​synthesis proposes that dreaming is triggered by the “largely random” firing of giant neurons in the pontine reticular formation (PRF) of the brainstem. The PRF plays a role in the regulation of REM sleep, and Hobson and McCarley proposed that the firing of these giant neurons stimulates the visual cortex while simultaneously initiating the rapid eye movements that characterize REM sleep. According to the hypothesis, the forebrain responds to this stimulation by attempting to construct a narrative that explains these visual sensations. The dream is, according to Hobson, the result of the forebrain “making the best of a bad job in producing even partially coherent dream imagery from the relatively noisy signals sent up to it from the brainstem.”

Somewhere between this Scylla of random, meaningless dreams and the Charybdis of dreams carrying messages from the gods or the unconscious, there slowly grew a sometimes inchoate movement proposing cognitive and emotional functions for dreaming.

One of the more straightforward ideas about why we dream is that dreaming helps us find solutions to our personal problems. Proponents of this idea often point to famous discoveries made in people’s dreams, including Elias Howe’s invention of the sewing machine, August Kekulé’s discovery of the ring structure of the benzene molecule, or Paul McCartney’s writing “Yesterday.” Great discoveries and breakthroughs aside, however, studies have repeatedly shown that dreams only rarely contain practical solutions to real-​life problems. This is not to say that dreams can’t help us grapple with impor­tant issues. People will sometimes make a decision, plot a course of action, or reconsider a previous plan based on a dream. But actual problem solving within dreams occurs too seldom for it to be the reason that dreaming evolved—and, in any case, it would only offer a function for those dreams we remember.

The Finnish philosopher and cognitive neuroscientist Antti Revonsuo proposed, in 2000, that dreaming evolved as a mechanism for simulating threatening events and rehearsing successful responses to these threats. Taken as a whole, empirical support for his threat simulation theory has been mixed. Although it is true that much dreaming contains threatening material, realistic life-​threatening events occur infrequently in dreams. Furthermore, only a small percentage of such dreams contain effective avoidance responses. In 2016, Revonsuo and his colleagues put forth an alternative theory, proposing that the function of dreaming is to simulate and thereby strengthen “the social skills, bonds, interactions, and networks that we engage in during our waking lives,” but it’s too early to tell how well this model will fare in comparison.

Solve Sundsbo/Art + Commerce, used with permission
Solve Sundsbo/Art + Commerce, used with permission

In the past decade, there has been an explosion in the number of neurobiological studies supporting the idea that sleep—​REM sleep in particular—​plays a key role in emotional processing. These studies have not explained what, if anything, dreaming contributes. That said, many clinical theories of dream function proposed over the past 40 years have focused on the idea that dreaming is involved in emotional regulation.

One model views dreaming as a kind of emotional damper, tasked with containing emotional surges or regulating the dreamer’s mood across successive periods of REM sleep. According to this theory, when dreams succeed in their overnight modulation of emotions, there is a pre-​ to post-​sleep improvement in the dreamer’s mood; or as co-author Bob Stickgold’s mother used to say, “Things will look better in the morning.” More recently, models of dysphoric dreaming—​bad dreams and trauma-​related nightmares—​propose that one function of dreaming is to reduce or extinguish fear-​based memories by allowing fear stimuli to be experienced in novel and emotionally varied circumstances.

Perhaps the best-​known modern clinical theory of dreaming is that of the late Ernest Hartmann, a former psychiatrist and professor at Tufts University. Hartmann proposed that dreaming is a form of “nighttime therapy” that helps weave emotional concerns as well as traumatic events into existing memory systems, all within the “safety” of sleep. Dreams, according to Hartmann, accomplish this function by creating connections between new and old memories that are broader and looser than connections made during wakefulness.

The Memory Function of Dreams

What if each of these theories was partly true? Could it be that dreams sometimes help us solve problems but at other times provide a unique environment in which to rehearse social interactions, learn to avoid threatening situations, or process emotions? We think the answer to this question may well be yes. But we also propose that a critical aspect of dreaming lies in the processing of memories.

When Erin Wamsley joined Bob’s lab in 2007, researchers knew that sleep supported memory evolution by processing new memories from the day before. But they didn’t know if or how dreaming contributed to this process.

Erin designed a study in which participants explored a virtual maze, trying to learn its layout. Then she let them take a 90-​minute nap. After their nap, she asked them whether they remembered dreaming about the task and then tested them on the maze again. The results were astonishing. Participants who had no memory of dreaming about the task took, on average, one and a half minutes longer to find their way out of the maze after their naps, while those who reported that they had dreamed about it found their way out two and a half minutes faster than before. When Erin repeated the experiment, actually waking subjects to collect dream reports and identifying those whose dreams were related to the task, she found that the latter showed almost 10 times more improvement after their naps compared with the participants who reported no related dreams.

What exactly did they dream about? One participant reported: “I was thinking about the maze and kinda having people at checkpoints, I guess [there were no people or checkpoints in the actual maze], and then that led me to think about when I went on this trip a few years ago, and we went to see these bat caves, and they’re kind of like, maze-​like.” Another recalled “just hearing the music” that played in the background while they had been exploring the maze.

Dreams like these seemed unlikely to help participants enhance their memories of the maze’s layout. And yet they were reported by the very participants who showed the greatest improvement. The sleeping brain was both enhancing its memory of the maze layout and creating related dreams. So, these dreams must be serving some other function. But what?

The sleeping brain performs multiple forms of memory evolution. It stabilizes and strengthens some memories while extracting rules and gist from others, and it integrates new memories into older, pre-existing knowledge networks. Fortunately, the brain can probably carry out multiple forms of memory processing at the same time. For example, after Erin’s participants fell asleep, the hippocampus presumably replayed and strengthened its memories of the paths they followed in the maze. But this left the rest of their brains free to deal with other aspects of the evolution of these memories—such as how to file them.

This is not a trivial task. Our brains store immense amounts of information in an unbelievably complex collection of interlocking neural networks. Related memories are physically connected so that activation of any memory in the network will tend to activate others in that network. How the brain decides to file new information—​into exactly which networks it will link a new memory—​determines whether and when this new information will come to mind during subsequent wakefulness.

Perhaps some strategy you learned while exploring a cave will help you the next time you try the maze task, or conversely, maybe something you learned from the maze task will help you next time you’re down in a cave. Your brain suddenly realizes, hey, exploring mazes and caves is really the same thing. It’s a perfect example of the function of dreaming that we propose: ​the extraction of new knowledge from existing information through the discovery of unexpected associations.

Shaun Higson/Thailand-Bangkok/Alamy, used with permission
Shaun Higson/Thailand-Bangkok/Alamy, used with permission

Nocturnal Adventures

Our model, NEXTUP, specifically proposes that dreaming involves the discovery and strengthening of previously unexplored weak associations. Typically, the brain starts with some new memory, encoded that day—​maybe an important event, a discussion overheard at work, or something related to a personal concern—​and searches for other, weakly related memories. These can be from the same day, or they can be older memories from any time in the dreamer’s past. The brain then combines the memories into a dream narrative that explores associations the brain would not normally consider.

In the glare of day, the usefulness or “rightness” of these newly found associations might be incomprehensible. But that’s fine. We don’t need to understand why our brain chose these associations. We don’t even need to remember the dream. All the important work was done while we slept: Associations were discovered, explored, and evaluated while we dreamed, and if our brain calculated that some of them were indeed novel, creative, and potentially useful to us, then it strengthened the associations and filed them away for later use.

From our perspective, the bizarreness of many dreams is simply a predictable consequence of weak, and hence unexpected, associations being incorporated into the dream narrative.

In Bob’s first faculty position, at the University of Massachusetts Medical Center, he had the unpleasant task of helping to teach the so-​called dog lab. This long-​since-​abandoned lab was the medical students’ introduction to death, albeit in the guise of a laboratory study of cardiovascular function. When students arrived, they were confronted with anesthetized dogs and a lab manual describing how to insert catheters into veins, measure intravenous blood pressure, and more. Near the end of the lab, they would cut through the skin and muscles of the dog’s chest, use a buzz saw to go through the rib cage, and apply drugs directly to the pumping heart. The first night that Bob taught the lab, he had a dream:

I was in the dog lab again, and we had just cut open the dog’s chest. As I looked down, I suddenly realized that it wasn’t a dog; it was my 5-​year-​old daughter, Jessie. I stood there dumbfounded, not understanding how we could have made such a mistake. And as I watched, the edges of the incision drew back together and healed without a hint of a scar.

Waking from the dream, Bob told his wife about it, and she suggested that the dog lab had clearly aroused his fears of mortality. And where were these fears greatest? For his child Jessie, of course. But Bob disagreed. That’s not what it felt like to him. To him, the dream seemed to ask the question, “If it’s okay to do this to a dog, why isn’t it okay to do it to Jessie?” Of course, both explanations were reasonable.

This is a classic example of what dreams like to do. At the time, Bob’s brain took an emotional event from his day and replayed it with an entirely improbable and bizarre rescripting. Obviously, this dream wasn’t designed to improve his ability to perform the surgery. Instead, while his brain dreamed, it searched through his memory networks for weak and potentially useful associations: The dog and Jessie were both small and helpless; he felt responsible for them both; he didn’t want either of them to die; he loved them both; or all of the above.

Finding multiple links to Jessie, his brain built her into the dream. But why? Not to answer a question and not to solve a problem. Rather, the brain asked “What if?” and watched its own emotional and cognitive response, observing how this response affected the dream narrative. The intensity of that response and the way it influenced the rest of the dream told the brain what it needed to know: This association, Jessie and the dog lab, was a valuable one. Something was uncovered about the fragility or sacredness of life that was important, something worth marking and strengthening and keeping available for the future. Once these connections were enhanced, the brain’s job was done. Whether Bob remembered the dream when he woke up didn’t really matter.

Of course, not all dreams are straightforward. It’s often a real stretch to identify any connections between our dream content (or at least what we can remember of it) and what happened that day or at any point in the past.

A few years back, Tony was giving a talk to a group of students when something about the feel of the auditorium made him realize that he was dreaming. He closed his eyes (in the dream) and imagined himself standing on a wide, sandy beach with the ocean surf crashing onto the shoreline. When he opened his eyes, he was indeed standing on a beach. But, strangely, there were also penguins—thousands of penguins—all over the place, just minding their own business, wobbling about, a bobbing mass of black and white. Tony woke up perplexed. Try as he might, he was unable to find a source for the birds that made any sense in the context of his current life.

Several days later, while riding in a friend’s car, a billboard caught his eye. It featured two beaches, side by side. The beach on the left had people enjoying the sand and wading into the water. But the one on the right was covered in penguins! Tony had driven that stretch of road recently, and in all likelihood, his brain had registered the image without his even noticing it. So, when he became lucid in his dream and tried to imagine an idyllic beach, his brain searched through his memory networks for weak and potentially useful associations. What it came up with was an unusual “What if?”

When brains dream, they can search for weak, novel associations anywhere, including in events people have taken little or no note of. This is one reason why linking our dreams to life events is often trickier than people imagine.

Geof Kern/Trunk Archive, used with permission
Geof Kern/Trunk Archive, used with permission

It’s also important to keep in mind that dream content sometimes dramatizes current concerns and other meaningful life events without displaying any concrete elements from them. For example, a sense of being overwhelmed may take the form of a tidal wave in a dream. A daytime fear of not meeting expectations at work may give rise to an exam dream. Showing an unexpected degree of resilience or creativity in the face of a problem may lead to a delightful dream of discovering a new room in your childhood home.

Whether in movies, books, or other forms of storytelling, humans use dramatization, with its figurative plots and metaphors, to describe the emotional events and concerns of our lives. What our brains do while dreaming is not that different from what they do when we go to the theater—​they imagine and explore possibilities embedded in a narrative with the hope of gaining new understanding about ourselves and the world.

Just as movies and plays typically involve themes of universal human significance, so our dreams are often centered on common themes—​themes of being chased or running late, being unprepared for an exam, soaring through the air, falling ill, dying, or discovering something wondrous. And much like our preferences for stereotypical movie genres, the kinds of stories the dreaming brain chooses to tell itself also depend on what thematic contents we as individuals most strongly relate to. Our dreams are both universal and unique, and they all help us to explore a multitude of worlds that might be.

Antonio Zadra, Ph.D., is a professor at the Université de Montréal and a researcher at the Center for Advanced Research in Sleep Medicine. Robert Stickgold, Ph.D., is a professor at Harvard Medical School and director of the Center for Sleep and Cognition.

Submit your response to this story to If you would like us to consider your letter for publication, please include your name, city, and state. Letters may be edited for length and clarity.

Pick up a copy of Psychology Today on newsstands now or subscribe to read the rest of the latest issue.