Perchance to Dream

Running as fast as you can, you quickly turn for a fleeting glimpse of an unknown pursuant. Your heart and legs pump furiously, but you can’t run fast enough—your limbs feel heavy and move infuriatingly slowly, as if you’re slogging through water. As the entity behind you (is it a person? an animal?) continues to gain ground, you see a wall looming ahead and realize that you’re going to be caught. With nowhere further to run, you wait as the mysterious creature approaches. But just before the frightening hands (or claws?) draw you close, your eyes flare open, and you discover that it was all just a dream.

For the vast majority of human beings—and at least a few nonhuman species, too—the slumber hours are enlivened by a unique internal experience known as dreaming. Written records, oral traditions, and even ancient petroglyphs reveal that dreams have fascinated people at least since the appearance of the earliest historical records, yet humankind continues to know little about them. Recently, researchers from a variety of backgrounds, including theology, psychology, and neuroscience, have stepped forward to shed light on what exactly our brains are doing—and why—during the dark hours of the night.

Sleep to forget, sleep to remember

When you awaken after a night’s rest, the preceding hours may seem to have passed uneventfully, but in reality your brain cells were busily firing away, engaging in tasks that scientists are only beginning to understand. In the mid-20th century, researchers discovered that sleep, neurologically speaking, is not a homogenous state—rather than remaining static throughout the night, the brain cycles through two major phases of activity known as Rapid Eye Movement (REM) and non-REM sleep. Sleep scientists can distinguish between the four substages of non-REM sleep and REM sleep by examining electroencephalograms (EEG) and electromyograms (EMG), which depict the electrical activity patterns, or waves, within the brain and muscles, respectively.

A sleep study, also known as polysomnography (PSG), is conducted to determine various physiological changes that occur during sleep. Researches employ an assortment of instruments to measure these changes, including electroencepholograms (EEG), which record electrical activity patterns in the brain. Credit: Timothy Snyder

A sleep study, also known as polysomnography (PSG), is conducted to determine various physiological changes that occur during sleep. Researches employ an assortment of instruments to measure these changes, including electroencepholograms (EEG), which record electrical activity patterns in the brain.
Credit: Timothy Snyder

Much about the functional significance of the various sleep stages remains a mystery, but scientists are steadily chipping away at the unknowns. REM sleep is particularly fascinating because its electrical and neurochemical characteristics are quite different from those of non-REM sleep. Unlike the other stages, REM sleep produces an EEG dominated by high frequency, short amplitude waves—a pattern that shows striking resemblance to the EEG produced during waking. Similarities in brain waves aside, the neural processes of REM sleep and wakefulness are distinct. “REM sleep is different from non-REM sleep and it’s different from waking,” says Els van der Helm, a psychology graduate student in Associate Professor Matthew Walker’s lab. “If you just look at the EEG, it looks exactly like waking, except that your EMGs show that your muscular tone is completely damped, to the point that you’re basically paralyzed.” She also points out that during REM sleep, unique patterns of neural activation arise. While the emotional, visual, and memory areas become highly active, the prefrontal cortex, which provides inhibition to other regions of the brain and helps you assess the realism of your experiences, lies quiet. “What you’re left with is a really visual and highly emotional brain during REM sleep,” van der Helm says.

Van der Helm aims to illuminate our understanding of the brain’s special functions during REM sleep by focusing on the processing of emotional memories. Earlier findings that implicated REM sleep as a major player in the processing of emotions led van der Helm and Walker to develop the “sleep to forget and sleep to remember” hypothesis, which proposes that the neural activities occurring during REM sleep modify our emotional memories by stripping them down to their bare essentials. “When you think back on something emotional, you may be good at remembering the details and how you felt and what happened, but you’re not re-experiencing the emotionality of it,” says van der Helm. “You don’t get sweaty hands again when you think about that presentation you gave, and your heart doesn’t start racing again. So it seems as if the memory is contained really well, but the tone has been stripped away.”

Given the unique neurochemical milieu present in the brain during REM sleep, van der Helm and Walker proposed that the modification of emotions may occur during this stage. The tendency of emotional memories to persist more strongly than neutral memories appears to depend upon activation of the adrenergic system, which relies on cell-to-cell communication mediated in part by epinephrine (also known as adrenaline; this same system gives us that extra boost during the so-called fight-or-flight response). In other words, emotional memories stick around better than their neutral competitors because of an extra kick provided by the adrenergic system at the time of memory formation. But during REM sleep, when these memories are brought forth for processing, the adrenergic system is suppressed. Van der Helm and Walker proposed that this period of suppression allows the brain to shave away components that are unnecessary to remember, leaving behind only the most essential elements of the memory.

Amongst the sleep stages, REM sleep is unique in terms of its neurochemical and electrophysiological characteristics; muscle tone is extremely low (in fact, you’re effectively paralyzed during REM sleep) and, as shown below, the brain waves of REM sleep more closely resemble those of waking than those of non-REM sleep. In particular, REM sleep brain waves are dramatically different from the delta waves of slow wave sleep.

Amongst the sleep stages, REM sleep is unique in terms of its neurochemical and electrophysiological characteristics; muscle tone is extremely low (in fact, you’re effectively paralyzed during REM sleep) and, as shown below, the brain waves of REM sleep more closely resemble those of waking than those of non-REM sleep. In particular, REM sleep brain waves are dramatically different from the delta waves of slow wave sleep (below).
Credit: Timothy Snyder

Credit: Timothy Snyder

Credit: Timothy Snyder


Who needs sleep anyway?

In today’s fast-paced, busy-body world, speed is the secret to success. Do you want to scramble up the corporate ladder like a mountain goat? Make your kids shine on their college applications? Impress your academic advisor and future employers? The winning approach can be summed up in four words: do more, finish faster. But with a measly 24 hours in a single day, how do we accomplish this? The simplest solution: sleep less.

While cutting back on sleep in the short term may get that paper finished on time, in the long term, it may actually cause harm. Catching a 15 minute nap can’t replace a full night’s rest, since sleep is a dynamic process consisting of various stages, each of which correlates with characteristic brain waves and certain physiological changes. In stage 1 of non-REM sleep, the brain transitions out of wakefulness; during stages 2-4 of non-REM sleep, many physiological factors, including muscle tension and the rate of respiration, gradually decrease; and in stage 4, large amplitude, slow waves appear, marking the synchronized firing of cortical neurons (which carry out higher level information processing). A full cycle typically lasts for about 90 minutes, reaching completion when the brain transitions from stage 4 back to stage 2 or 3 and then moves into REM sleep. During an average night, humans progress through five or six sleep cycles, and these cycles evolve as the hours pass—initially, we predominately engage in slow wave sleep, but as the night wears on, we begin to devote increasing amounts of time to stage 2 and REM sleep.

Given its involvement in the processing of memories and emotions, REM sleep may seem like the golden child of the sleep stages, but recent work in Professor Matthew Walker’s lab suggests that non-REM sleep also affects our waking lives in important ways. In a study published in the March issue of Current Biology, the Walker lab discusses a link between learning ability and non-REM brain wave oscillations. Specifically, they found a significant correlation between sleep spindles—short, rapid bursts of electrical activity in the brain that occur during stage 2—and learning ability in participants who were tested after a 90 minute nap. Sleep spindles are associated with activity in the hippocampus, a key memory-processing area in the brain, and may help clear space for new memories by facilitating the transfer of older memories from the hippocampus to the prefrontal cortex. Cutting back on sleep may hinder this process, since spindles occur most commonly during the latter half of the night, when stage 2 non-REM sleep becomes more prevalent.

Want to ace that test? Make sure you get your daily dose of spindles.


Currently, van der Helm is testing this hypothesis by placing study participants into an fMRI (or functional magnetic resonance imaging) machine and exposing them to photographs with varying valence (positive, negative, or neutral) and arousal (or intensity) values. “These pictures come from a big database used by researchers all over the world and are standardized, so they have specific scores on arousal and valence,” she explains. As participants assign scores to the pictures, the fMRI apparatus measures changes in blood flow, which indicates locations of neural activity in the brain. To assess the impact that sleep has on the intensity of emotional memories, van der Helm allows one group of subjects to get a full night’s rest in the lab after their first scoring session, and then asks them to score the same set of photographs again the following morning. To account for the effects of second exposure (and to determine if the modification of emotions requires an intervening period of sleep), a second group undergoes both sessions on the same day—first in the morning and the second 12 hours later.

Van der Helm expected that sleep would decrease the intensity of an emotional memory, which is exactly what her initial results suggest. First, fMRI scans reveal that as subjects view increasingly emotional pictures, the amygdala (a region of the brain that plays a key role in the processing of emotions) also increases in activity; interestingly, this effect appears dampened in the sleep group and augmented in the awake group. Second, a night of sleep modifies the scores that participants assign to photographs of high-level, but not mid-level, intensity. “It seems as if sleep decreases emotional reactivity for the most extreme pictures, but not for the pictures that were mediocre in their emotional intensity,” says van der Helm.

The well of creativity

Studies like van der Helm’s strongly refute old perceptions of the sleeping brain as a dormant mass of resting neurons. Today, we have an image of a brain that busily shuffles through a variety of nightly tasks, many of which may significantly affect the way we function while we’re awake. While van der Helm’s research suggests a role for sleep (and more specifically, REM sleep) in the processing of emotional memories, work by psychology graduate student Jared Saletin, also in the Walker lab, provides evidence that the neural processes of sleep may contribute to nothing less than our creativity. Saletin’s research brings to light the relationship between sleep and relational memory, or the ability to connect separate memories in novel ways. “Let’s say you know how to get from Sacramento to Berkeley, and you know how to get from Berkeley to Los Angeles, but you’ve never actually been told how to get from Sacramento to LA,” says Saletin. “You still know how to do that because you know how to combine the steps along the way. This is something that kids learn at an early age.”

Saletin examines relational memory using transitive inference tests, which evaluate a person’s ability to combine a set of learned inferences in new ways to solve unfamiliar problems. Subjects learn a series of premises—for instance, “choose A over B,” “choose B over C,” and so on (in his study, Saletin uses fractal images, rather than letters of the alphabet, to avoid biasing his participants). As in van der Helm’s study, participants are separated into two primary groups: in the first group, participants are taught the premises in the evening, allowed to sleep overnight, and then presented with transitive inference tests the following morning; in the second group, participants learn the premises in the morning and then undergo the test 12 hours later, without an intervening period of sleep.

“I dream my painting, and then I paint my dream.” - Vincent van Gogh Using transitive inference tests to evaluate the effects of sleep on the ability to link memories in novel ways, psychology graduate student Jared Saletin has discovered that the construction of new ideas, particularly those of higher complexity, may occur while we sleep. In short, sleep may facilitate our creativity. Credit: Vincent Van Gogh

“I dream my painting, and then I paint my dream.” – Vincent van GoghUsing transitive inference tests to evaluate the effects of sleep on the ability to link memories in novel ways, psychology graduate student Jared Saletin has discovered that the construction of new ideas, particularly those of higher complexity, may occur while we sleep. In short, sleep may facilitate our creativity.
Credit: Vincent Van Gogh

The test presents subjects with questions that require them to link the original premises in unfamiliar ways. The difficulty of the question depends upon the number of independent premises that must be combined in order to arrive at the correct answer. In some instances, participants must derive what are called “1-degree inferences”—for example, if asked to choose between A or C, a subject would have to combine the premises “choose A over B” and “choose B over C” to produce the right conclusion. More challenging are the 2-degree inferences, in which participants have to leap across two levels to figure out which option to choose; for example, to know that E is superior to B, subjects would have to link the premises “choose B over C,” “choose C over D,” and “choose D over E.”

Saletin discovered that a full night’s sleep selectively improves a person’s ability to make the most difficult connections. “What happens when I give you B and D, which you’ve never learned before?” he asks. “After 12 hours, whether you’ve been awake or asleep, you get better at what we call inferential judgment, picking B over D, as if you’re going through C. But after sleep, you’re much better at the more distant pair of B and E, which requires two levels of relation jumping.”

Based on these findings, Saletin suggests that sleep modifies our memories in more dynamic ways than previously suspected. “Traditionally, people have talked about memory in three stages: you learn it, you store it, and you recall it,” he says. “But you actually do a lot more than that—you transform it over time and you integrate it. The transitive inference study leads us to suggest that sleep helps you build an infrastructure to connect pieces of information that you’ve never been explicitly told go together. This may be related to creativity, the emergence of an idea from parts that you’ve never put together before.”


A stroke of genius while you sleep

In 1865, German chemist Friedrich August Kekulé was trying to determine the structure of benzene, a recently discovered molecule known to contain six carbon and six hydrogen atoms. At the time, the geometry of the molecule baffled scientists, who couldn’t figure out how to arrange all 12 of the atoms such that each carbon atom possessed four bonds and each hydrogen atom possessed one. According to organic chemistry lore, a portion of the answer came to Kekulé while he dozed by a fire. After dreaming about a snake that circled about until it bit its own tail, he awoke and realized the solution—that benzene is a closed, hexagonal molecule with a carbon atom at each of its corners.

Other stories suggesting the creative force of dreams and their purported roles in historical events exist: Mary Shelley claimed that the monster in her book Frankenstein originated in a dream; Otto Loewi, a German pharmacologist, stated in a lecture that the idea for his experiment on frog hearts (which demonstrated the chemical transmission of nerve impulses and eventually made him a Nobel laureate) came to him in his sleep; and the melody for the Beatles song “Yesterday” apparently emerged from one of Paul McCartney’s dreams.

With the “provocation hypothesis,” dream researchers have proposed that on rare occasions, people may experience intense, highly vivid and memorable dreams that provide opportunities for insight and the birth of new ideas. Visiting scholar Kelly Bulkeley believes that such dreams exist, but he cautions that it’s often difficult to validate the high profile reports of dream-based revelations. “It’s always interesting to hear such stories, but it’s a little dicey to rely too much on them. It’s better to have a broader base of evidence to work with.” In any case, keeping a dream journal might not be such a bad idea—especially if it leads to a Nobel Prize, emblazons your name on the walls of literary history, or produces one of the most beloved songs of all time.


So it was just a dream?

Aside from processing memories, the brain also happens to produce the most vivid and coherent dreams during REM sleep, as revealed by experiments in which subjects are woken up during various sleep stages. The coexistence of memories, emotions, and dreams during REM sleep has nourished speculation that dreams may have some unknown function, but very little is known about the biological basis of dreams, aside from the fact that most people appear to have them.

Although science remains unsure about them, humanity’s fascination with dreams has proven powerful enough to bring forth a number of theories. Perhaps the most famous originated with Sigmund Freud, an early 20th century German psychiatrist who pioneered a psychological approach to dreams and viewed them as the “royal road to the unconscious,” a means of accessing our innermost desires and neuroses. Today, dream theories like Freud’s are largely ignored by cognitive neuroscientists, lingering at best as dusty, outdated ideas, and at worst as residents in the halls of pseudoscience.

In their place, physiological conceptions of the brain have taken hold. Rather than developing abstract conceptions of the mind, scientists now investigate the processes of cognition by searching for concrete links between anatomy and function and delving into the brain at the levels of tissue, neuron, molecule, and gene. But, according to Eleanor Rosch, a professor in the Department of Psychology, this approach doesn’t paint a complete picture of the human mental experience.

Although her research career has not focused on dreams, she has maintained a long-standing interest in the topic and even taught a course on the psychology of dreams for several years. Her fascination with the subject grew out of her research into the psychology of Eastern religions, particularly Tibetan Buddhism, in which dreams play an important role. According to Rosch, psychology’s neurophysiological models are difficult to apply to many of the mental processes that people utilize and experience every day. “Psychology is becoming more and more divorced from the way that people’s minds function in daily life,” she says. “Psychology has become very focused on the brain sciences, while anything that’s difficult to approach from that perspective has been sidelined, especially in mainstream university psychology.”

Rosch contends that because dreams are difficult to fit into current conceptions of the brain, they have largely been overlooked by neuroscientists. “Dreams are hard to fit into your prototype of what a scientific experiment is,” she states. “If you work with a dream and you analyze it according to two different dream theories, you will come to very different conclusions that are difficult to compare, which makes testing the theories against one another hard.” Rosch believes that the difficulty of developing mutually exclusive dream theories that are testable using conventional neuroscience techniques has made dreaming a taboo topic of research for modern day scientists. “This may be part of the allergy to actually studying anything about people’s inner experiences,” she says.

Dreams are also difficult to study because of the manner in which scientists must collect information about them. “You have to be awake to talk about your dream, and at that point, your brain is completely different in terms of neurochemicals and neurophysiology,” says van der Helm. “So basically you’re asking subjects to travel back in time, to a state where they were unconscious, and describe what happened. We have no idea what’s happening during that transition from your dreaming brain to your waking brain. All you can rely on is the waking brain to recall it, and we already know that people differ dramatically in their ability to recall dreams.”

In spite of the challenges inherent in dream research (or perhaps because of them), one hypothesis regarding the origin of dreams has become prevalent within the brain sciences—that dreams serve no purpose and actually originate from chaotic bits of images, sensations, emotions, or memories brought on by random stimulation of the relevant circuits as the brain goes about its nightly business. These jumbled, nonsensical fragments are then woven into a coherent narrative by the waking brain as it comes online again. “You can imagine that dreams are basically just an epiphenomenon—you’re reactivating certain networks in the brain, and this is activating your visual and emotional areas,” says van der Helm. “So you have these emotional feelings, and you see things, but perhaps the dream itself doesn’t serve any purpose. Our lab doesn’t necessarily think that dreams lack function, but this specific view is hard to disprove.”

This image depicts the Hindu god Vishnu as he sleeps on Ananta, the cosmic serpent, which floats on the cosmic ocean beyond space and time. In Hinduism, Vishnu is the divine dreamer and the world as we know it is his dream. Credit: Victoria and Albert Museum, London

This image depicts the Hindu god Vishnu as he sleeps on Ananta, the cosmic serpent, which floats on the cosmic ocean beyond space and time. In Hinduism, Vishnu is the divine dreamer and the world as we know it is his dream.
Credit: Victoria and Albert Museum, London

Pointless? Maybe not…

At the Graduate Theological Union, a partnership of seminaries and graduate schools that focuses on interreligious collaboration and offers two affiliate PhD programs with UC Berkeley, visiting scholar Dr. Kelly Bulkeley spends his waking hours investigating the contents of dream journals, in which people record the details of their dreams immediately after emerging from sleep. Bulkeley, who suffered from recurrent nightmares as a child, began his academic journey as a student of psychology, but quickly discovered that he needed an interdisciplinary approach to understand the common elements of the overall human dreaming experience.

After studying philosophy and religion as an undergraduate at Stanford University, Bulkeley went on to divinity school, where he completed programs in psychology and religion. In particular, his exploration of world religions provided him with a wide perspective on the significance and nature of dreams. “The best historical records of dreams are often in religious texts or records of religious practices,” he says. “We have only had psychology for the past couple hundred years. If we want to learn about dreams in the broader perspective of human history, you really have to have that bigger historical picture. The only way to get that was to understand world religions.”

After sifting through both ancient and modern records of dreams, Bulkeley began to realize some common trends that transcended culture, ethnicity, and time. Such “psychological universals” suggested to Bulkeley that dreams are more than pointless offshoots of neurological babble. To investigate this hypothesis, Bulkeley needed to analyze the content of numerous dream journals produced by a large number of people. Unfortunately, the traditional approach to cataloguing dream content involves a tedious scoring process in which human readers pore over dream journals, searching for words that connote certain emotions, concerns, experiences, or activities. Although it has produced intriguing results, the human scoring system is highly subjective (the scoring of a particular word could vary between readers) and labor intensive, which has prohibited large-scale analyses of numerous dream reports.

In most cases, the only records of dreams from ancient times reside in religious texts. Visiting scholar Kelly Bulkeley uses his training in both religion and psychology to develop an interdisciplinary understanding of dreams. Drawn from the Bible, this scene depicts Jacob’s dream of a ladder ascending to heaven.  Credit: Kelly Bulkeley

In most cases, the only records of dreams from ancient times reside in religious texts. Visiting scholar Kelly Bulkeley uses his training in both religion and psychology to develop an interdisciplinary understanding of dreams. Drawn from the Bible, this scene depicts Jacob’s dream of a ladder ascending to heaven.
Credit: Kelly Bulkeley

To address this issue, Bulkeley used word search technology, a data-mining technique that has been successfully employed by literary scholars to rapidly examine large swaths of works. Using 40 different categories of word strings partially constructed using words catalogued by human scorers, Bulkeley was able to quickly evaluate the content of several dream journals. The word search system reports the percentage of dreams that contain at least one word from a particular category, each of which represents something general like an emotion or use of a particular sense, such as vision or hearing. Using the automated word search system, Bulkeley demonstrated that a single person with access to a computer could quickly and reliably replicate the results of the human scoring system. The additional advantage of the word search approach is that researchers can conduct a blind analysis—that is, they can evaluate the content of a journal without ever reading (and becoming biased by) the dream narratives.

By employing the word search method, Bulkeley has shown that a dream journal can be used to accurately predict many aspects of a subject’s waking life: religious convictions, the nature of personal relationships, jobs, hobbies, and more. When combined with the finding that most dreams contain relatively common scenarios—interactions with family and friends, walking and driving, going to work—rather than fantastical experiences like flying or falling from great heights, Bulkeley’s studies suggest that our dreams may be more than neurological nonsense. “The recurrent themes and content of dreams are accurate reflections of what’s important in the dreamer’s waking life,” Bulkeley says. “Dreams turn out to be much more mundane and normal than most people assume. Every now and then there’s something odd, but if you look at the broad patterns of what people actually dream about, it tends to be about people we know, places we usually go to, things we often do in the day.”

So do dreams mainly provide us with an opportunity to rehash the day-to-day occurrences in our lives? Possibly, but Bulkeley points out that dreams don’t mirror our waking hours in every respect. “We seem to do less reading and writing and computer work in our dreams compared to the proportion that many of us, particularly in the academic world, do in our waking lives,” he says. “Dreaming seems to have more of a bias towards social activities and less towards reading, writing, and arithmetic activities.” According to Bulkeley, it’s also plausible that dreaming, like play in young mammals, provides us with the chance to safely practice behaviors relevant to our survival. “We find recurrent patterns of fight or flight behavior, we find all sorts of sexual behavior, and we find all sorts of bonding behavior—kind of the basic stuff of human survival and reproduction,” he says.

Pushing the boundaries

While varying cultural and religious beliefs about the significance of dreams abound, the underlying roots of dreaming continue to mystify non-scientists and scientists alike, even as they bind us together under the banner of universal experience. Are dreams inherently functionless, constructed piecemeal from random sparks released by the brain as it conducts the important work of processing our memories and emotions while we sleep? Or do they serve a purpose, providing us with the means to safely test the boundaries of existence? While proponents of either view may dismiss the competing idea as erroneous, dream research may actually benefit from the fertility of thought that can accompany controversy, especially if the answer lies somewhere between the two extremes.

“There may always be a horizon of skepticism, and that’s fine—that’s where we have the debates, and that’s where we do the research,” Bulkeley says. “We’re pushing the edges, trying to figure out how far they extend and whether we do reach a point where maybe it is chaos, maybe it is random neural nonsense. I don’t think we should stick to an ultimate psychological determinism, where every element of a dream means something. Maybe there is some crazy stuff in there, but we’ll never know unless we look.”

Leave a Reply