A voice says: “Close your left hand. Don’t ask yourself whether you’re asleep. Think about trees.”

I’m lying in bed. A sleep mask covers my eyes. A tangle of wires covers my left hand. At the tip of my ring finger, a sensor measures my heart rate. A flexible length of plastic embedded with circuits stretches from my palm to the top of my middle finger. This will record the hypnic jerks and spastic opening-hand motions that signal my entry into hypnagogia, the first stage of sleep, where thoughts slip free of conscious control.

There’s a laptop on the bedside table; the screen shows fluctuating green and red lines. Adam Haar Horowitz, who is running the experiment, speaks to me over Zoom, monitoring my somatic information. The device I wear is called a Dormio. It was developed by Adam and a team of professors and researchers at the MIT Media Lab to facilitate “dream incubation,” the shaping of dreams according to words or images chosen by the dreamer. I’m wearing a prototype. Adam envisions a time when the components of the Dormio will be widely available; anyone will be able to download its blueprint and, with a few cheap premade circuits, construct her own dream incubator.

The way it works is simple. The device connects to a website where you can record a voice message to yourself—“think about trees”—that will play as you begin to fall asleep. Dormio detects when you enter hypnagogia, waits a short period, then awakens you and prompts you to describe what you’re experiencing, and sends the recording to your hard drive. You can also alter the parameters of awakening, which enables you to enter deeper or shallower levels of sleep. This first time, Adam is manning the controls himself; his is the voice reminding me to think about trees.

I settle into bed. The Dormio feels light on my hand, and I soon forget about it. My eyes are closed under the mask.

“Hold the image of a tree in your mind.”

I’m not going to be able to do this, I think. A spark of nervous energy runs through my legs.

“Relax. Don’t ask yourself whether you’re asleep. Think about trees.”

Okay, I think. Trees. I don’t want to disappoint Adam. He’s been so nice. I’m not feeling sleepy at all. I briefly consider lying. I could tell him I’m seeing people turning into trees, I’m seeing tree people. What are they called? Ents.

Stop it, I think. He’s got me strapped to this device. He can see my heart rate and everything. He’d see I was lying instantly. Shut up and think about trees.

Okay. Focus. Just trees in general? I can’t think about trees in general. What was that Wordsworth line about trees? “Of many, one.” One tree. The magnolia tree in my backyard. I picture it. Pendulous pink blossoms, pale bark, spreading branches . . .

I shift a little on the mattress. The nervous electricity in my legs seems to be fading. I wonder if maybe a magnolia isn’t the best tree to focus on—it’s a flowering tree, after all, a kind of hybrid between tree and flower. But then I remember Adam telling me to just watch where my thoughts go, and I think, Adam’s a really laid-back guy, there’s no way he’d object to a magnolia tree. I briefly consider asking him—he can hear me, he’s on the screen on my bedside table, if I took off these ridiculous eyeshades I’d see him—but I decide not to.

Because something odd is happening to the magnolia tree. It’s the light. There’s no light. I mean there’s no sky. The tree is there, but there’s no sky. Yet I can see the tree. If I can see it, I think, there must be light somewhere . . .

I focus on the tree. There aren’t any blossoms anymore. The branches . . . The branches extend as I watch them, the grooves of the bark deepen, glowing a little. Inside, I think. The light is coming from inside the tree. Above, where the sky should be, there’s nothing.

“Michael,” says Adam. “What are you thinking about?”

The sound of his voice makes me jump. My eyelids open to the blackness of the sleep mask. I remember I’m on my bed. I remember about Adam.

“Um,” I say.

I recall what I’d just been seeing, what I’d seen after I stopped thinking, what I’d seen when thought had turned wholly into image.

“Some buildings, lit up as teeth. And then . . . it’s hollow in there. I’m moving inside.”

This was a crude sketch of the images I’d experienced. I remember the long thick line of a structure under a purplish absence of sky. Periodically the structure would light up from within, showing bones, like white teeth, with gaps. When they lit up, I began to move—moving into the central gap. Then I was inside, a long spiraling hollow opening above me.

“Okay,” says Adam. “Very good. Now just relax. Think about trees.”

I think about what I’d seen, about my hypnagogic encounter. Was that a tree? Was the dream incubation successful?

The sky, I think. The sky was inside. Inside of the tree. My thinking is heavy, slow. My eyelids flicker under the sleep mask. But the thought is persistent. What I saw was the inside of the tree. The tree was borrowing my vision and watching itself. That’s why it didn’t look the way a tree looks from the outside. That’s why it didn’t look the way a tree looks to humans.

“Pretty cool,” says my father over dinner. “What’s it for?”

It’s a good question. The Dormio doesn’t exactly allow you to control your dreams—at least not the deep, late-sleep-phase dreams of flowing events, nameless anxieties, and sudden revelations. The Dormio enables a limited shaping of the images that appear during sleep’s first stage. Yet this is enough to give bite to the question, to render it slightly less abstract. Why would I want to shape my dreams? What kinds of things can you do with dreams?

These aren’t, of course, new questions. Perhaps every era has its characteristic answer to the question of what dreams are for. Our own time might be defined by its characteristically bifurcated response. On the one hand, there’s scientific research into the question of what dreams do for our brains. And on the other, there’s a wild arc of actual and possible answers to the question: What can my dreams do for me? This gap between what science can tell us and what we want to know is perhaps the central spiritual feature of our situation. The subject of dreams provides an instructive case.

For a long time, scientists resisted studying dreams because of the inherent subjectivity of the exercise: the only way to tell whether or what someone is dreaming is to ask them. There’s no truly objective way to study dreams, no way around the messy, complex engagement with another person’s words—a person who, unlike an MRI scan or a slice of neural tissue, can lie, misremember, embellish, tell you what she thinks you want to hear.

When Adam first expressed his academic interest in dreams, his neuroscience professor suggested he’d been taking too many magic mushrooms. But despite the faintly disreputable aura that clings to dream research, the phenomenon is inescapable for scientists seeking to understand the mind. As we learn more about dreams, the various reasons for their dismissal as objects worth studying—that they’re trivial or epiphenomenal, just the mind’s attempt to make sense of the sleeping brain’s random neural firing—grow less convincing. Research has found that depriving someone of REM sleep—the sleep stage in which most dreams occur—might account for the health problems associated with sleep loss. Clearly dreams do something for us. If not, why would evolution have endowed us with the capacity?

Despite some intriguing speculation, scientists haven’t yet come up with a clear, satisfying answer to the question of why we dream. Part of the reason is doubtless because, as any time spent studying neuroscience will show you, our knowledge of the brain is in its infancy. And part of it is due to the special limitations of dream research. Animal studies—sometimes referred to as the gold standard of neuroscientific research (think of the things one can do to rat brains that one can’t do to humans)—are of no help here. Like many pet owners, I believe that my dog dreams. But when I see her lying on the couch, muttering and growling with her eyes moving behind closed lids, I can’t wake her and ask her what she saw. When I spoke about the state of the field with the dream researcher Erin Wamsley, she described a kind of disappointment, a sense that the breakthrough insights into the nature of dreaming that seemed imminent a decade or two ago haven’t materialized.

Over the past few years, this perceived impasse has led to the emergence of what a recent special issue of the journal Consciousness and Cognition dubbed “dream engineering.” To adapt Marx’s maxim, if hitherto the scientists have attempted to understand dreams, the engineers now seek to change them. In fact, the engineers argue that we can’t deepen our understanding of dreams unless we can change them. “The logic is simple,” the issue’s editors wrote. “If we cannot control dream content, then we cannot do controlled experimentation on dream content.” The efforts to manipulate dreams that have gotten the most media attention have to do with lucid dreams, in which the dreamer is aware she’s dreaming and can communicate to observers through eye movement. Some have claimed that electrical stimulation of brain waves can induce lucid dreams. Yet a 2020 study failed to replicate these claims, “raising caution against the flurry of commercial devices advertising electrical stimulation for lucid dream induction.”

By contrast, the Dormio device, the efficacy of which I’d experienced myself, represents the most fully realized piece of technology to have emerged from the nascent dream engineering movement so far. Adam invented it in collaboration with a team of other students, as part of his 2019 master’s thesis at MIT, which synthesized his interests in dreams, creativity, and technology. After graduating, he stayed on to probe the device’s potential with a team at the Media Lab, an institution loosely affiliated with the university’s architecture school, where scientists, engineers, and artists collaborate on projects untethered to traditional disciplines.

In the days following my tree experience, I continued using the Dormio. A couple of times, the word I selected was “quake,” the title of the classic 1996 computer game. I wanted to see what the chunky geometry and minor-tone palette of late-Nineties computer graphics would do inside me. What surprised me when I awoke was not that the tunnels and corridors of the game unfurled through a purple absence of sky at my wish, but that the color red predominated. That color is absent from my conscious memories of the game. But when I awoke, what I talked about was the red, the most vivid red, as if there’d been some kind of veil thrown over the reds I’d encountered in my waking life—as if in the dream I was standing very, very close to the color red. Maybe inside it.

After a few days of this, I found I was developing a sense, even a taste, for hypnagogia. Now, when I go to sleep, I watch for it. It’s the moment when thoughts take on a life of their own. Or more accurately, when they resume the life they’ve always led when I’m not there.

There is nothing new about the desire to shape our dreams. In his 1995 book Private Myths, the psychiatrist Anthony Stevens describes theories of dreams throughout history as belonging to three major categories. Some cultures believe that dreams are messages from supernatural agents. Some believe that dreams are “actual experiences due to the external wanderings of the soul during sleep.” And some believe that dreams are the natural result of normal mental processes.

It is possible, of course, for a culture to hold several of these theories at once. The Bible, as the scholar Richard Walsh notes, contains some dreams that serve as messages from God and others that represent mystical experiences, such as Jacob’s dream of the ladder. The naturalistic theory of dreams, meanwhile, predominated in ancient Greece in the writings of philosophers like Heraclitus and Aristotle, and in much of the West since the nineteenth century.

If one believes that dreams are divine messages, as the ancient Egyptians did, then dream incubation is a way of encouraging such messages. This practice was eventually incorporated into institutional religion. Worshippers of the Egyptian god of healing used light, scents, and sounds to make the dreamer more receptive to divine messages.

Historical efforts to shape dreams were typically based on faith, on a knowledge of the supernatural world as revealed by religion. Dream incubation was rarely an individual enterprise, but was embedded in a religious culture. The prospect of simple, widely available devices for incubating dreams raises the prospect of a new dream culture, one based on naturalistic premises and embedded in a scientific worldview. Adam, for his part, speaks of his desire to create a “community” of dream incubation. Such a community or culture would represent the opening of a new route between dreams and public life.

What do we need to build a dream incubation culture? What beliefs must we have? How do we get from the dry, incomplete scientific picture of dreams to thousands, or millions, or billions of people hooked up to Dormios? And for what? Do we need to start with a kind of faith in what dreams can do? Or can we discover something like faith through the practice of dream incubation itself?

Adam told me an anecdote that expressed the challenges of a contemporary dream incubation culture. The Coors Brewing Company had gotten wind of his device and contacted him with a proposal. They’d hoped to provide a number of people with Dormios and incubate their dreams with the words “Coors beer,” the results of which would then be used in a new advertising campaign.

Adam, horrified, turned them down instantly. But the incident suggests the dystopian possibilities of dream incubation in our commercial culture. It would be easy to fold devices like the Dormio into the ever-proliferating technical means by which advertisers seek to shape our desires. Adam fears that the ubiquity of smart speakers in bedrooms may already represent another way for companies—or political actors—to infiltrate our subconscious.

If we put aside, for the moment, our fears that dream engineering might be used to control us, what other uses might we find for dreams? Adam’s suggestion, expressed on his website as well as in his grant applications, is that dream engineering could foster creativity.

The idea seems compelling. The Dormio is, after all, a technological version of the method creators like Thomas Edison and Salvador Dalí used in their work. Dalí, for example, would hold a key between his fingers; once he started to drift off, his hand would open, the key would drop, and the noise would awaken him, the images of hypnagogia fresh in his mind and ready for harvesting. There are numerous anecdotes of scientists, inventors, writers, and artists solving problems in their dreams. The dreaming mind’s alertness to unexpected, concealed connections suggests parallels between the dream state and the state of creative inspiration often described by writers and artists.

My Dormio-assisted adventures in hypnagogia showed me how unbelievably, compulsively, naturally, and irresistibly creative the mind becomes once it slips loose of conscious control. At times it felt as if my awareness was coming apart under the pressure of creative energy, like a thin cotton shirt under a fire hose. One afternoon, I choose “basketball” as my incubating word. I close my eyes and picture a basketball. Soon the image starts to spiral out from its shape—the ball’s pattern of tiny leather bumps and the indented black lines that swim over its surface become directions for the evolution of an entire space, a world crisscrossed by unspooling lines, overlaid with orange squares, a world through which I’m moving at light speed. When, after six or ten minutes, I am awakened, I’ve lived through several phases of this constantly evolving basketball world.

I’ve long been haunted by the Buddhist idea that we die and are reborn dozens of times per second. The ancient maxim intimates that beneath the scale of consciousness and conventional experience lies a ceaseless torrent of change, a swarm of chasms and metamorphoses. In one sense, it’s a kind of therapy for the fear of death. If you are afraid of dying, think of how much you have changed since you were six—since you were a fetus—since before that—can you imagine a greater change? My hypnagogic experiences suggest that similar changes are constantly taking place beneath the illusory continuity of ordinary waking consciousness.

The idea that we are subject to an incessant stream of transformations would seem to disintegrate the idea of death. Hypnagogia turns this abstract, mystical idea into experience, supplies it with a kind of empirical evidence, albeit one accessible only from the first person, evidence as individual as one’s own death. Just below the surface of wakeful awareness, just a minute or two under it, everything is change. The mandalas hypnagogia relentlessly sketches under the eyelids don’t so much provide a respite from the distractions of ordinary consciousness as a shocking realization that the sluggish movements of regular experience might be composed of millions of shining, vibrating filaments.

And this is sleep? I think, jettisoned into wakefulness by the sound of my iPad. This is rest? This is unconsciousness?

I experience a vague horror. I feel, very slightly and for the first time, how one might long for nothingness. I feel a new sympathy for the expressions I’ve sometimes encountered among the old, the ill, the insane, or the ancients—expressions of the desire for everything to stop. When I finally quit using the Dormio, it’s because I’ve grown a little afraid of what it reveals.

This helpless creativity of my mind, this incessant hypnagogic generation of forms and worlds, isn’t like the productions of human artists. Looking at surrealist images by Dalí or even Max Ernst while my vision is still saturated by the shapes and colors of hypnagogia, I’m most taken by the way, in the pictures, creativity has stopped. It’s suspended, pinned like a butterfly between the wooden edges of the frame. But the images of hypnagogia never stop; the creativity of the dreaming mind is a transformative force defined by the fact that it can’t be distilled into intelligible sentences, paintable images, tolerable music.

And when it escapes hypnagogia, this creative energy dives deeper into the sleeping mind, into late-stage dreams, where it takes a viral form, a form that you can bring back up into consciousness in deceptive, melting images and phrases, all marked by a trademark blur, a kind of birthmark, an emblem of its continuing participation in the ceaseless transformations of dreams. And these fragments, when examined closely, defamiliarize ordinary consensus reality in a manner different from that of hypnagogia.

For example, the night after my basketball experiment, I dreamed that I was a basketball player. In this late-sleep dream, I find the restless transformations of hypnagogic basketball have moderated. There’s still a fluidity to the outlines, a sense that I’m always standing a little too close to things to see them fully. But now my dreaming mind imitates a world stable enough for me to move around in the form of a human.

I have the body of a basketball player. I’m nearly seven feet tall, able to reach the basket with barely a hop. I can move at incredible speeds for a long time without losing my breath. But this body is new to me. I don’t really know how to use it, especially on the court. My shot is way off. So I hang out near the basket, dunk the ball—which I find to be easy. The people around me—my teammates, my coaches, and the people watching, the invisible pressure of millions of eyes, and, still worse, millions of dollars—are disappointed in me. Shocked even. Disgusted by the jerky, awkward way I move.

So I pretend I’ve been injured. I need help; I need to relearn how to move, how to score, how to make this incredible body work again. Weeks or maybe years pass in the dream—and now we see the other side of hypnagogia’s micromysticism, now we see the mysticism of the large scale, of weeks, months, and years that pass in minutes of dream time.

I wake directly from immersion in a network of basketball anxieties, basketball relationships, and basketball movements accrued over years of dream time, into the present tense of wakefulness, with the sun coming through the blinds, my wife stirring beside me, the chirp of my daughter’s voice through the baby monitor.

And the dream, through its memory, injects two viral thoughts into my day, two basic reflections on personal identity, two reflections derived not from arid logic, but from long years of dream basketball experience. First: I am not my body. (I can inhabit different bodies; I have inhabited different bodies.) Second: I am not my thoughts. (My thoughts—my mind—might be bigger than I am; my experience in the dream was of dwelling in a vast, peopled world that was also, somehow, my thoughts.)

Creativity is a soothing, positive, productive presence in advertisements and tech slogans. Direct insight into the relentless creativity of the dreaming brain, on the other hand, can be disturbing. Turning dream content into creative work is, as the dream researchers Robert Stickgold and Antonio Zadra show in their book When Brains Dream, rare. It seems to me that creation—the creation of stable, delimited objects—may even involve a kind of struggle against the mind’s ceaseless, devouring creativity. In any case, my encounter with dream creativity inspired at least as much fear as exhilaration. We might call the kinds of fears embodied by my basketball dreams the metaphysical fear of creativity. These fears tend to be private, personal, existential.

But there’s another dimension of the fear of the compulsive mental creativity revealed by dreams—the moral fear. And this fear poses, to my mind, perhaps the largest challenge to Adam’s compelling and utopian vision of a public dream culture. To get this fear in focus, we might reflect on public reactions to the images generated by the collective dreamer of the internet every minute of every day. Take as an example the utterly surreal, dreamlike, even nightmarish photograph of the Bidens and the Carters that appeared everywhere for about a week in early 2021. If you’ve somehow never encountered it, google “Bidens and Carters” and take a look.

I think this image is about as dreamlike an artifact as we have in public circulation. Almost as soon as it was released, practically every major news organization in the world attempted to explain away the surreal effect. The answer, a former White House photographer told the New York Times, could be found in Jimmy Carter’s huge shoes, “an indication that the foreground was distorted by a super-wide-angle lens.”

The photographer who shot the image refused to explain it. “It’s for people to figure out and think about,” he said. The spectacle of so many prominent outlets immediately offering a variety of detailed technical explanations—despite the speculative nature of the explanations invoked—indicates the scale of the discomfort aroused by the image, the intensity of the urge to make it go away, to reduce it to normal size, to rationalize it. Moreover, the explanations characterizing the first wave of the photo’s reception contained a logical flaw. They sought to explain the physical cause of the image’s distortion while presenting this as an explanation of the image’s disturbing psychic fascination.

As a sign that this initial, technical method of warding off the photo’s dreamlike power had failed, a second genre soon arose. This genre more logically, if no more persuasively, sought to explain the source of our fascination. And it did this by arguing that the photo’s visual idiosyncrasies serve to emphasize the presidential couples’ virtues. The Washington Post, for instance, paraphrased one of Carter’s biographers to the effect that his “slim build” might be “another factor” explaining his tiny stature in the image, going on to say that he stayed fit in his old age by skiing, fly-fishing, and long-distance running.

A BBC article highlighted a tweet that took a different tack, acknowledging and embracing the photo’s cartoonish grotesquerie while connecting its distortions to positive, progressive political feelings: “We voted for Biden because he’s a decent human being with sound policies but also because he and Jill are giants who will crush you if you make them angry.”

Perhaps the most developed example of this genre came from the literary magazine n+1. The author argued that we’re drawn to the image because the shrunken, extremely elderly Carters represent the old people we’ve lost or feared losing to the pandemic. And we are drawn to the gigantic image of Biden because of our astonishment at the “scope and speed of his legislative proposals.” In effect, this second method of dealing with the photo represents a kind of communal dream interpretation. And it is driven by an effort to sanitize our collective unconscious, to make our fascination with this dreamlike image express virtues such as compassion and hope.

To me at least, the kinds of explanations offered by n+1, the BBC, and the Washington Post seem wildly, spectacularly wrong. But I think I can understand their motives. Behind these somewhat desperate efforts to convince readers that their intrigue derives from their progressive values lies the specter of a different account. When it first materialized, the image appeared on a conservative website as an almost uncanny realization of the Trump-era, QAnon-inflected idea of a “deep state” controlled by ancient and unnatural figures.

This seems to be part of the larger process by which transgressive, surreal images and writing, once associated with the liberation movements of the Sixties, have been forsworn by their progressive successors. The n+1 piece in particular embodies the role that literary creation—and artistic creation more widely—has come to occupy. Literature stands as the emblem of reason, as the superego defending against the subversive power of the nightmarish free associations that gave us conspiracies like Pizzagate—perhaps the most extraordinary public instance of the logic of dreams in modern history.

The values of creativity—the shock of violating convention, the wildness of uninhibited imagination—are no longer the values of the literary and artistic establishment. Art has taken on an increasingly didactic and moralistic tone. If it approaches the dreamlike, it comes to dispel and discipline the dream’s unruly power, to ward off contact with the sources of that power. Here we have a second sense in which artistic creation has come to depend on a struggle against creativity, and the wild, convention-demolishing, weak-association-hunting inventiveness of uncontrolled thought.

I don’t mean to suggest that our culture has given up on creativity. The arts and sciences still speak in terms of its rhetoric. The image of the transgressive creator who breaks through the crust of convention to discover new forms and new ways of life retains its emblematic power, however inhospitable to creation the institutions who use this emblem have become. The intensity of the rhetoric of innovation issuing from Silicon Valley, which heralds each new trinket as if it represented a breakthrough on the order of the combustion engine or the superconductor, testifies to the allure of creativity. We still believe in creativity; we are still creationists.

As I’ve become more aware of my own dreams and more sensitive to the uncanny transformations of hypnagogia, I’ve begun to think that it would be foolish to underestimate what a truly democratic distribution of Dormio-like technology might accomplish. We might use dream incubation and enhanced dream awareness not necessarily to experience more strange linkages or weak associations, but as a kind of therapy that could lessen our fear of the mind’s creativity.

We might attend to our dreams, not to become more creative, but to become more open to the creativity that is always happening in what the dream researcher Rosalind Cartwright has called our “twenty-four hour mind.” The more frequently we observe ourselves emerging intact and invigorated from regular hygienic baths in unconventional, nonsocial, self-dissolving thought, the more comfortable we might become with such modes. We might invent new methods of sharing our dreams, which could lead to new methods of sharing creative ideas with one another.

The ancient Egyptians, medieval Christians, and twentieth-century Freudians found in dreams what their faiths told them they’d find. But if we turn to the Dormio to help with creativity, we’ll be using our dreams to repair our faith, to heal the fears that haunt our belief in creativity, that repress our awareness of our natural changefulness.

Perhaps the dominant way individuals have used dreams in the modern era is as a means of achieving personal insight. Freud and Jung both regarded dreams as texts that can be interpreted to reveal different facets of the self, its conflicts and aspirations. Various techniques used to remember one’s dreams have proliferated, and by simply saying “I will remember my dreams” before going to sleep, Stickgold claims, one can enhance one’s capacity to recall dream content. While working on this essay, without especially trying, I found myself remembering a much larger share of my dreams than I usually do.

But dream engineering, the prospect of changing one’s dreams, doesn’t seem obviously compatible with the project of understanding the self by interpreting one’s dreams. Presumably, if the goal is to understand who I am through my dreams, then receptivity, not interference, should be the technology’s focus. To change the self by changing one’s dreams requires a different orientation.

As interest in dream engineering has grown, researchers have attempted to move beyond the textualist model of interpretation long dominant in the West. For the Onge of the Andaman Islands, writes the scholar Matthew Spellberg, “dreams are understood to be sites of action; not texts but places, not a coded language but a part of reality.” Drawing on the work of the anthropologist Vishvajit Pandya, he claims that the Onge transform individuals’ dream lives through a collective process of negotiation. Every night, a group gathers to describe the dream version of an island they visit together at night. This process involves subtle modifications so that “everyone’s dream accounts [are] gradually aligned with one another,” and the island they visit becomes, in a sense, a shared space.

Spellberg finds in this practice of dream sharing “a protocol for building a consensual, democratic dream-world, for negotiating the . . . imaginative world that governs so many behaviors and decisions in the physical one.” Perhaps a contemporary public dream culture in the United States, facilitated by devices such as the Dormio, might establish similar protocols, enabling us to transform and coordinate our dream worlds.

I wonder whether such a practice, in American society, might suffer from the kind of anxiety that marked the responses to the Biden-Carter photograph. It’s not hard to imagine our community engaging in efforts to sanitize and purify dreams of unorthodox or incorrect elements. There seems to me to be a value in keeping our dreams private and asocial, particularly in a world where social technologies burrow ever deeper into our conscious lives. But even if one wanted to replicate the practice of Onge dream sharing (which Spellberg notes has died out), the logistics of such an endeavor are hard to imagine in our fragmented society.

So what can dreams offer us? Reading through recent research, I’ve often thought of the Sex Pistols line “Don’t know what I want, but I know how to get it.” Dream engineers pursue different things—from personal insight to social change—by recording and transforming dreams. I’ve offered my own addition to this pool by suggesting that technologies like the Dormio can also allow us to bathe in the endless creativity of hypnagogia, thereby providing a kind of therapy for the fear of creativity that afflicts our culture.

I might offer another, entirely personal and doubtlessly idiosyncratic answer to my father’s reasonable question about dream engineering, which came to me while reading When Brains Dream. Stickgold and Zadra suggest that the dreaming mind categorizes our experiences, selecting which memories will become part of ourselves, which we will draw on to make sense of new situations, and which we will forget. The bizarre quality of most dreams may result from the strange association scheme the brain uses to store memories. Think of the weird juxtapositions caused by conscious categorization, the way alphabetical organization, for example, puts cookies next to cows. Or perhaps the bizarre quality might derive from the mind’s discovery of hidden affinities among entities that, to the waking mind, have little in common.

We don’t remember everything we experience. And memories, in the sense of the fully formed suites of images that come to mind when we think of a certain person, place, or time, betray subtle and not-so-subtle differences from actual past events. Scientists increasingly believe that, as the neuroscientist Roland Benoit puts it, “our memory is not made for the past, but for the future.” Our memories are constantly transforming as we encounter new challenges and new relationships. As we grow older, our sense of self changes, and our memories reshuffle to support this new self.

I was particularly struck by an experiment in which Stickgold attempted to demonstrate the dreaming brain’s method of deconstructing and reconstructing memories. In it, subjects were shown a group of letters and asked to press one key if the letters formed an English word, and another key if they didn’t. Stickgold showed the subjects a different word before showing them the target word. This word would either be related strongly to the target word (like “right” followed by “wrong”), weakly related (like “thief” followed by “wrong”), or unrelated (like “tree” followed by “wrong”).

Subjects tested in the middle of the day identified the target word more quickly if the word was strongly related to the target word. But when Stickgold woke subjects up in the middle of REM sleep, this effect disappeared; showing them a strongly related word no longer had any impact on how quickly they identified the target word. But showing them a weakly related word made the subjects identify the target word more quickly.

Stickgold argued that this study shows that, during REM sleep, the brain preferentially searches through its memories for weak associations—associations that aren’t already known, links between experiences, ideas, people, or images that aren’t obvious, subterranean connections between memories that may help us face a certain kind of challenge or solve a certain kind of problem. Associations that prove useful are then saved by the brain—to use the inescapable computing metaphor of our era—which may allow new memories, thoughts, and connections to creep into consciousness.

I asked Stickgold about this idea of weak association, whether it was the kind of connection that was there all along but buried, or whether the connection was discovered, perhaps even created, by the dream. He told me that what the mind investigates during dreams aren’t “the funny associations that you couldn’t find when you’re awake, but that you wouldn’t.” The dreaming mind is freer, less inhibited by a sense of logical or social or conventional patterns, alert to resonances that elude us when awake.

Perhaps it’s not particularly surprising to say that dreams put unexpected things together. But Stickgold’s experiments represent cunning ways to smuggle something every dreamer knows into the objective world of science. And once there, these insights connect with other painstakingly established facts—about sleep, memory formation, or dream-eroding maladies. What emerges from this mosaic is a still fuzzy sense of dreams as workshops in which the memories of waking life are refined and split into components that are then magnetized, attracting lost twins across social and temporal gulfs. These pieces are then reassembled into the memories that form the background, the perspective from which our conscious selves gaze out at the world, at you, at the mirror.

I was sitting in my backyard reading Stickgold’s research, and I looked up to see the red brick of my house suffused with the late summer sun. I immediately recalled a dream—a series of dreams—which featured redbrick surfaces in the same slant of light. I couldn’t recall anything about the dreams except for that image, and the fact that it was associated with a neighborhood I used to frequent in Baltimore nearly two decades ago.

It struck me that the vision before my eyes at that moment—the brick of my house in the sun, in 2021—had a thickness, an emotional density, that few other images have for me. I connected the felt density of this image with what I’d been reading, and concluded that my dreams about red brick in the sun were responsible for endowing the quotidian wall of a house I’d lived in for just a couple of years with the presence, the heaviness, the nearly hallucinatory intensity of childhood memories.

What if I could give places in my current life the same kind of density, the same intimacy, the same closeness, that my dreams had given the brick wall of my house?

I decided to conduct an experiment. Stickgold had told me that what had surprised him most in his research was how easy it was to influence people’s dreams. The Dormio, as I’d experienced, offers a technological method for influencing hypnagogia, the initial stage of dreaming. But Stickgold suggested you could also influence the deeper stage of dreaming, the stage that is less like a liquid mosaic of images and more like life—the stage where one encounters people, places, and things from one’s past.

I picked a place that has always seemed fundamentally sterile and alien to me—my university office, a space I primarily use for meetings, and where I’ve been unable to work. My office has a perceptual thinness to it—the surfaces seem empty. There’s nothing behind them, no memories, no significance.

So I took photos of my office using my iPad—the walls, the carpet, the desk. Every night for a week, I’d gaze at those images before going to sleep. Could I make my office a part of me? Could I intervene in the subconscious workshop where my significant memories were assembled, and align my internal and external environments?

The result of my little experiment was ambiguous. I still don’t feel comfortable working in my office, though I am able to recall its contents vividly now in a way I wasn’t able to previously. In fact, I must confess that if anything, the effect has been the opposite of what I intended. I’ve developed a conscious aversion to the metal bookcase in my office, and have gradually been removing the books and taking them home.

Perhaps I can’t finally control which memories form a part of me and which don’t. Perhaps all I could do was show my office to the submerged part of me that decides what I hold on to and what I expel. For some reason, it likes the red brick of Baltimore row houses in the late summer sun. Maybe nothing will induce it to accept the drywall and carpet of my office.

While the result may not have been what I expected, I like to think that my experiment in dream engineering represents a transaction between my conscious and unconscious mind. The few large-scale changes I’ve been able to make in my life have convinced me of the value of such transactions. Waking consciousness is very thin; the waking will is weak. To change oneself, one must go below consciousness to the place of dreams and habits. Perhaps my exercises in dream engineering, with my Dormio and iPad, are not fundamentally unlike the implements of Egyptian dream incubation after all. I place my word or image, my tree or office or basketball, on the altar of sleep, and a mind greater and vaster than my own contemplates it, taking it up or letting it fall.