In Dreams Begin Technologies
The exotic contraptions and magical processes that fill our dreams may seem an unlikely source for useful technology. But practical devices, from the first armored warship to a computer with laser circuits, have originated in dreams.
Inventors throughout history have credited dreaming with pointing the way to new devices and formulas. I interviewed modern scientists and engineers for my book The Committee of Sleep: How Artists, Scientists, and Athletes Use Dreams for Creative Problem-Solving—And How You Can Too , and I found that the same phenomena kept recurring. There are patterns to when creative dreams take place, to what kinds of problems are solved by them, and to what solutions arrive during typical REM (rapid eye movement) dreams versus half-asleep hypnagogic states. Technological design involves formal study—and much straightforward hard work—but truly notable inventions also include inspiration, the birth of some novel idea. Inspiration cannot be summoned at will; this is where dreams typically come in.
Any break from focused problem-solving may allow a misleading assumption to dissipate. But the sleeping mind abandons conventional logic with unusual thoroughness in pursuit of novel approaches. Most of what we call dreams occur in REM sleep, which happens for roughly 20 minutes every hour and a half throughout the night. In REM sleep, our visual cortex is as active as when we are awake. But our frontal lobes, which control many of our highest brain functions, including inhibitions, are sluggish, and a very different mix of neurotransmitters circulates in the brain. A century ago, without knowing the physiology of this state, Freud described REM dreaming as “primary process” thinking, meaning it was highly visual, intuitive, and emotional. We’ve discarded other Freudian notions, such as that dreams always represent erotic wishes, but the characterization of dreaming as “primary process” has held up well. Dreaming is simply the mind continuing to think about problems in a very different biological and psychological state. No one would suggest that all waking thought achieves profound results. Neither does dreaming. The dream’s power lies in the fact that it is so different a mode. It can supplement and enrich what we’ve already done awake.
Most creative dreams occur when a thoroughly prepared, highly motivated person is stuck on a problem, and REM-sleep dreams usually address problems that we can represent visually. Perhaps the oldest recorded dreamed invention is the Korean kobukson , an armored “turtle boat” whose designer, Adm. Yi Sun-shin, commanded Korea’s navy at the end of the sixteenth century. When the Japanese mounted the 1592–98 Hideyoshi invasion and sank many Korean vessels, Yi despaired of repelling the enemy until he dreamed one night of a monstrous sea turtle. It spewed fire from a dragon’s mouth and had a shell that no sword could pierce. Yi awoke and ordered his men to construct a vessel shaped like a turtle, built of thick wood blocks, and covered with iron armor. The prow bore a dragon’s head through which a cannon fired and sulfurous smoke billowed just as in Yi’s dream. A small fleet of kobukson repelled the Japanese and changed naval history.
Three and a half centuries later, nations had to repel air attacks as well. In the first years of World War II, as Nazi bombers ravaged Holland and Belgium, David B. Parkinson, a young engineer at Bell Labs, worried about the news from Europe. In 1940 Parkinson developed an “automatic level recorder” to chart data from telephone transmissions. The recorder used a specially shaped, wire-wound potentiometer that he had invented. His work was nonmilitary, but one night he dreamed he was with Allied anti-aircraft gunners shooting down German planes, and they invited him to inspect their weapon. He was amazed to see it controlled by his potentiometer. Once awake, he realized that “if the potentiometer could control the high-speed motion of a recording pen with great accuracy, why couldn’t a suitably engineered device do the same thing for an anti-aircraft gun!” He and a colleague, Clarence A. Lovell, incorporated the potentiometer in a pioneering electronic analogue computer. It was much easier to use and more accurate than the complicated, slow-acting mechanical “predictors” then employed to get a fix on aerial targets. The new anti-aircraft gun director was adopted by the Army and dubbed the M-9. In a single week during August 1944 the Germans launched 91 V-1 missiles toward London from the Antwerp area, and guns controlled by the M-9 shot down 89 of them.
Dreams have also improved on devices already in use. During the Gulf War, soldiers used bulletproof vests made of the fiber Kevlar. In response to the sudden call for more Kevlar, DuPont started turning out huge quantities around the clock. Then a crucial machine broke down, costing the company $700 a minute and threatening the lives of thousands of soldiers. Floyd Ragsdale was one of many engineers DuPont assigned to fix the machine. None of them could even locate the problem on the first day, but that night, Ragsdale fell asleep and dreamed he was part of the piece of equipment. He saw springs and hoses and water spraying everywhere. He woke up and wrote down simply “hoses, springs.” Later that night he puzzled over the notation and the dream until he realized that the hoses in the real machine must be collapsing, blocking the flow of water and shutting everything down. What would prevent this? Coiled springs inside them.
In the morning, Ragsdale sought out his supervisor and told him the hoses must be closing and needed springs. Even without knowing the provenance of the insight, the supervisor was doubtful; no one had suggested that the hoses might be the problem. But later that day, in a try-anything spirit, he pulled off a hose and found that it had indeed collapsed, which couldn’t be detected from the outside. The supervisor allowed Ragsdale to fit the hoses with springs, and by the end of the day, the machine was up and running. That one dream saved DuPont at least three million dollars and America many lives.
In the early 1780s, William Watts, a British plumber from Bristol, invented the modern method of making buckshot after he dreamed of being showered with spherical raindrops. Buckshot had always been laboriously cast in molds or dropped a short distance into a vat of water, which yielded teardrop-shaped globules. Watts realized that if he dropped the molten lead from a great height, it would cool into small, neat spheres before hitting the water, which would cushion its fall.
Domestic devices have also been dreamed up. When Elias Howe was designing his first sewing machine, in 1844, he couldn’t figure out how to make it hold a needle. His needles had their holes at the end opposite the point, and every possible way of securing them obstructed the needles’ passage through fabric. One night, according to legend, he dreamed he was captured by savages who threatened to kill him with spears if he didn’t finish the machine right away. He noticed eye-shaped holes near the heads of the warriors’ spears and realized that this was the proper design for his needles. He awakened and arose at once to whittle a model. He was not the first to use an eye-pointed needle, but his innovative use of the idea led to a patent that eventually became quite lucrative.
The dreams of Howe, Yi, Watts, and Ragsdale contained fantastic elements along with practical solutions; Parkinson’s was very literal. But all involved visual representations of problems, as is typical of full-fledged dreams from REM sleep. Why are REM dreams so visual? Why does the biochemical state of that segment of sleep favor visual imagery and keep visual areas active? The evolutionary anthropologist Donald Symons, of the University of California at Santa Barbara, suggests what he calls the “vigilance hypothesis.” Sleepers need to monitor their environments—to smell smoke, hear intruders, and feel pain. It would be dangerous to hallucinate vividly in most sensory modes. We’d either constantly awaken in unnecessary panic or block out real warnings.
Eyes closed in the dark, we don’t need to monitor our visual environment, so we’re free to fantasize with images, Symons says. Moreover, the attention required to monitor these nonvisual warning signals is a fraction of what we need to process waking input. Dreams are therefore free to play out all sorts of potential solutions that wouldn’t have reached consciousness during waking—as long as the dreams do so visually.
Two inventors recently had whole series of problem-solving dreams about lasers, a technology that is inherently thought about visually. In the early 1990s, Alan Huang, then head of optical computing at Bell Laboratories, was working to create a computer using optical laser circuits in place of electrical ones but couldn’t produce a model that worked even as well as conventional circuitry. He began to have a recurring dream in which two lines of sorcerers’ apprentices carried pails of data along two trails. The lines would at some point cross and block each other’s progress. One night he had the same dream except that when the lines of apprentices converged, they passed through each other. He awoke realizing that light had this quality. Two laser circuits could pass through each other, while electrical signals could not. He could design laser computers that exploited this difference.
A few people I interviewed for The Committee of Sleep went so far as to say that any time they were stuck on a problem, they would dream a solution. The Harvard physicist Paul Horowitz was one of them. Horowitz’s great passion is SETI, the Search for Extraterrestrial Intelligence, and he designs radio telescopes; radio signals travel great distances in a focused form more easily than light. But by the fall of 1998, laser technology had developed far enough that Horowitz decided to build controls for a giant optical telescope to search for pulsed light signals.
“There were literally an infinite number of ways we could put it together,” he recalls. “It’s not one big problem but lots of little ones. You figure out how to arrange one group of lenses, then you design a circuit to manipulate that. There were many stages where we would get blocked on how to do something. Often I’d have a dream. These dreams have a narrator who’s sketching the problem verbally. Then the voice is giving the solution. I’m also seeing the solution. I watch a man working on the mechanical device, arranging a lens for the optics or building a circuit, whatever it is that I’m stuck on at that point. The dream always takes it a slightly different way from what we’ve been trying awake.”
He keeps a pencil and paper next to his bed to record these dreams. “They’ll evaporate in a second if I don’t get them down. I take my notes into work and tell my team, ’I’ve dreamed up a solution—literally.’ They’ve gotten quite used to this.” He says that while working on the laser telescope, he had at least three dreams about optics assembly and at least two about circuits, always about specific problems that could be visualized. He’s had similar dream series during radio-telescope projects. As is typical with REM dreams, they always occur in the middle of the night, after several hours of sleep.
Solutions to more varied types of problems show up in dreams from the hypnagogic state, in the time when we are just falling asleep or awakening. This is the next closest state to wakefulness in terms of brain-wave patterns as well as in when they occur. When Freud characterized REM dreams as “primary process,” being visual, intuitive, and emotional, he described waking thought as “secondary process”: logical, linear, and focused. In hypnagogic, half-awake dreams, primary process and secondary process enjoy a rare coexistence and interaction. We can critically evaluate images that are still before our eyes. Elaborate abstract problems and mathematical concepts that must be represented algebraically rather than geometrically are likelier to be solved in hypnagogic dreams.
Stephen Bailey, who has worked at Chicago’s Fermilab, is a physicist by training but is also adept at computer programming. One evening, a statistician friend of his described a program he needed that would allocate computer memory for working with complex mathematical matrices. Bailey wasn’t planning to take on the problem, but as he fell asleep, he says, “the data were floating around in three-dimensional space. I was seeing abstract geometric shapes with numbers in them, floating, twisting, coming around to the arrangements in which they needed to be. At the same time, I knew what the algorithm had to be in order to achieve this. It would need to use a method called recursion, where you break a large problem into smaller problems repeatedly. All the details were in the dream.”
He awoke, went to his computer, and wrote out the program. It worked perfectly. This experience was not isolated, he says. “When I’m working on a project, I tend to dream about it. I’ll even dream in computer language: C++ is what I use to program.” These problem-solving dreams seem to be hypnagogic ones, and Bailey distinguishes them from his usual dreams: “Other times I have people, places, plot. These problem-solving dreams are much more focused.”
Some computer dreams contain only the outlines of solutions. Bill Markwick, a Canadian programmer, says that when he dreams, “the answer isn’t neatly worked out. Instead, it tends to be a concept or method of solution that I hadn’t thought of while awake. It’s as if some mysterious part of my brain handed up an algorithm and said, ‘Here, dolt. You do the figures. You’re the one with a calculator.’”
Stephen LaBerge, a dream researcher at Stanford University, has interviewed a computer scientist who has learned to use dreaming as reliably as has Paul Horowitz. His solutions occur in “lucid dreams,” a rare phenomenon in which one realizes one is dreaming and where again primary process and secondary process seem to coexist and overlap. This programmer’s dreams are always set in an old-fashioned parlor. The programmer is sitting with Albert Einstein, who in the dream world is an old friend. They discuss the program in question and write on a blackboard. When they’re done, Einstein goes off to bed. The programmer then concentrates and tells himself, “I want to remember this flow chart when I wake up.” When he does awaken, he reaches for his pad and writes out what invariably proves to be a usable program.
I’ve conducted research with college students who’ve tried to dream solutions to more modest problems—their math homework, or fitting bulky furniture into their new apartments. Half have had dreams about such problems, and a quarter have had dreams that they—and independent judges—felt solved the problems. Ordinary people whom I interviewed for my book also reported using dreams to solve thorny little problems, such as chess moves or how to perform a magic trick. These were problems within the range of their abilities, but barely, ones on which they’d been stuck while awake.
Cultures that emphasize and value dreams seem to produce even more examples of problem-solving ones. Anjali Hazarika, of India’s National Petroleum Management Programme, runs workshops on dreaming for oil personnel, a curriculum difficult to imagine in the United States. She emphasizes solving personal problems that are interfering with job performance, but some of her workshop participants have focused on objective problems too. She tells of a chemist who was developing enzymes to refine crude oil and dreamed he was next to a road when a big truck full of rotten cabbages rumbled by. When he resumed work on his project the following Monday, he suddenly realized that as cabbages decompose, bacteria break them down into exactly the kind of enzymes he was seeking. And cabbages are dirt cheap; rotten ones are simply thrown out. The chemist developed a technique for crude-oil refining using the exact material his dream had trucked in for him.
A petroleum engineer in another of Hazarika’s groups was working on improving a device used in pumping oil from the wells. An S -shaped tube coming out of the pump kept clogging, whereupon the machine would have to be stopped and cleaned. The engineer incubated a dream about this and saw the letter S changing into a U . He awoke and realized the pipe could have just one gentle curve and still get the oil to where it was needed. His new design did not clog.
In some dreams, a problem’s solution is unmistakable, as in seeing the computer directing the anti-aircraft guns or having the design of telescope controls acted out in front of you. But in others—like that of the sorcerer’s apprentices who could march through each other—they can be missed if the dream isn’t specifically examined for solutions. Our society has benefited greatly from the usefulness of dreams. But we might gain even more if we paid serious attention to this resource.
