You find yourself at a banquet table. You feel disaffected because the people surrounding you are speaking a language you do not understand. Suddenly, you feel pressure on your foot—beneath the table, someone’s foot is on your own. You glance up. Your eyes meet those of the attractive person sitting opposite you. Intuitively, you sense the word that you must now say to captivate this person. You say it: “Phlegm.” The person stands, and suddenly the other people are gone. As is the table. As are your clothes. You fling yourselves at each other in passion. It is wondrous. The two of you are far up in the air, the sensuality of the experience heightened by the clouds brushing past you. Yet suddenly you begin to sob in shame, because you have been observed by your four deceased grandparents, who disapprove. You realize that the severe-looking man in the black frock coat comforting your maternal grandmother is William Seward, and with great clarity and an inexplicable sense of nostalgia, you recite, “William Henry Seward. U.S. secretary of state in the Andrew Johnson administration.”
You know, one of those dreams.
Just as the kidney is a kidney-shaped organ, dreams are dreamlike. But why should that be? In real life you wouldn’t wind up floating amid the clouds with someone seconds after the touch of a foot. Instead, at a key moment, you’d decide that he or she was actually kind of neurotic or note the bit of spinach stuck in his or her teeth or suddenly remember you’d forgotten to turn off the lights of your car. Dreams, by contrast, are characterized not only by rapid transitions but by a heightened sense of emotionality. Then there’s the disinhibition: not only do you do things that you couldn’t bring yourself to do in real life, but with two seconds of sensible reflection, you wouldn’t remotely want to do them.
There has never been a shortage of theories as to the utility of dreams being dreamlike. Maybe dreaming is the channel by which the gods speak to mortals. Or maybe this is a good way to work out how you really feel about your mother without all that repression getting in the way. Maybe it’s a way to get your brain to work in an unconventional, orthogonal manner to solve that pesky math problem you went to sleep thinking about. Or maybe this is how you keep underutilized neural pathways active (this theory has floated around for a while—if you spend all day long exercising those rational, sensible pathways in your brain, you need dreams to give an aerobic workout to those gibberish neurons, lest they shrivel up from disuse). Or maybe it’s so you can have a sex dream about some unlikely person at work and then act all weird and knowing around them the next day at the water cooler. Or maybe the dreaming evolved so that the surrealists and dadaists could make a living.
How does your brain bring about this state of disinhibited imagery? Until recently, scientists understood little about the actual nuts and bolts of dreaming. But we’ve known for some time that sleep has a structure, an architecture, if you will, with rhythmic cycles during the night of deep, “slow-wave” sleep, interspersed with the REM (rapid eye movement) sleep, most associated with dreaming. And the levels of activity in the brain are not uniform throughout the stages of sleep. Techniques that indicate the overall levels of electrical excitability and activity in the brain have uncovered something pretty intuitively obvious: during deep, slow-wave sleep, the average level of brain activity goes way down. This fits well with studies suggesting that the main purpose of slow-wave sleep is to allow for the replenishing of energy stores in the brain—the proverbial recharging of the batteries. But something very different happens during the onset of dreaming during REM sleep—a big increase in electrical activity. And this has a certain intuitive logic to it as well.
Advances in brain-imaging technology now allow scientists to study activity and metabolism in the small subregions of the brain, rather than just in the brain as a whole. In a series of studies, Allen Braun and colleagues at the National Institutes of Health have examined the neuroanatomy of metabolism during sleep. I think they may have uncovered the explanation for why dreams are so dreamlike.
The researchers utilized positron-emission tomography (PET) to measure the various rates of blood flow throughout the brain. One of the remarkably adaptive features of the brain is that blood flow in a particular region will increase when that area increases its level of activity. In other words, there is a coupling between demand for energy and the supply of it. Thus, the extent of blood flow in an area of the brain can be used as an indirect index of the activity there. That is why PET scans, which show blood flow, are so helpful in this type of research.
Braun and crew got some volunteers who allowed themselves to be sleep-deprived for an ungodly twenty-four to fifty-three hours. Each bleary volunteer was then rolled into a PET scanner and forced to stay awake even longer, while a baseline PET scan was made. Then, snug as a bug inside the scanner, each subject was finally allowed to sleep, with the scanning continuing.
As the subjects slid into slow-wave sleep, the blood-flow changes observed made a lot of sense. Parts of the brain associated with arousal, known as the reticular activating system, shut down; ditto for brain regions involved in controlling muscle movement. Interestingly, regions involved in the consolidation and retrieval of memories did not have much of a decrease in blood flow, and hence metabolism. However, the pathways that brought information to and from those regions shut down dramatically, isolating them metabolically. The parts of the brain that first respond to sensory information had somewhat of a metabolic shutdown, but the more dramatic changes were in downstream brain areas that integrate and associate those bytes of sensory information, that give them meaning. The result: metabolically quiescent, sleeping brains.
While the scientists at the scanner’s console bided their time, eventually the sleeping subjects transitioned into REM sleep. And then the picture changed. Metabolic rates leapt upward throughout regions of the brain. Cortical and subcortical regions that regulate muscle movement, brain-stem regions that control breathing and heart rate, all showed increases. A part of the brain called the limbic system, which is involved in emotion, showed an increase as well. The same for areas involved in memory and sensory processing, especially those involved in vision and hearing.
Meanwhile, something particularly subtle went on in the visual processing regions. The primary visual cortical region did not show much of an increase in metabolism, whereas there was a big jump in the downstream regions that integrate simple visual information. The primary visual cortical region is involved in the first steps of processing sight—the changing of patterns of pixels of light and dark into things like lines or curves. In contrast, the downstream areas are the integrators that turn those lines and curves into the perception of objects, faces, scenes. Normally, an increase in activity in the downstream areas cannot occur without an increase in those primary processing areas. In other words, when you’re wide awake, you can’t get from the eyes to complex pictures without going first through the initial level of analysis. But REM is a special case, where you’re not using the eyes. Instead, you’re starting with the downstream integration of visual patterns. This, Braun and colleagues speculated convincingly, is what makes up the imagery of dreams.
So there are increases in metabolism during REM sleep in numerous parts of the brain. In some regions, metabolic rates even wind up being considerably higher than during wakefulness. Now it’s time for the exception that I think is the punch line about the dreaminess of dreams, in a region of the brain called the prefrontal cortex. Outside the prefrontal cortex, all of the brain regions most closely associated with the limbic system showed an increase in metabolism with the onset of REM sleep. But in the prefrontal cortex, only one of the four subregions increased. The rest of that area stayed on the floor of metabolic inactivity that it had sunk to during slow-wave sleep. This is interesting, given the functions of the prefrontal cortex. The human brain, when compared with your off-the-rack mammalian brain, has many unique features. Its sensory inputs and motoric outputs are uniquely fine-tuned to make it possible to whip off an arpeggio on a piano. The limbic system allows for something virtually unprecedented among mammals: sexual receptivity among females throughout the reproductive cycle, rather than merely at the time of ovulation. The vast cortex creates symphonies and calculus and philosophy, while the atypically numerous interconnections between the cortex and the limbic system allow for that dreadful human attribute, the ability to think oneself into a depression.
Yet in many ways, the most unique feature of the human brain is the extent of the development of, the power of, that prefrontal cortex—the region staying metabolically inhibited during REM sleep. The prefrontal cortex plays a central role in self-discipline, in gratification postponement, in putting a rein on one’s impulses. On the facetious level, this is the part of the brain that keeps you from belching loudly in the middle of a wedding ceremony. On a more profound level, it keeps the angry thought from being allowed to become the hurtful word, the violent fantasy from becoming the unspeakable act.
Not surprisingly, other species don’t have a whole lot of prefrontal function. Nor do young kids; the prefrontal cortex is basically the last part of the brain to fully mature, not coming completely online for decades. Violent sociopaths appear to have insufficient metabolic activity in the prefrontal region. And damage to the prefrontal cortex, such as after certain types of strokes, causes a disinhibited, “frontal” personality. The person may become apathetic or childishly silly, hypersexual or bellicose as hell, scatological or blasphemous.
Braun and colleagues found that during REM sleep, much of the prefrontal cortex was off-line, unable to carry out its waking task of censoring material, while the complex sensory-processing parts of the brain concerned with emotion and memories were highly active.
So bring on those dreams, now free to be filled with disinhibited actions and labile emotions. You breathe underwater, fly in the air, communicate telepathically; you announce your love to strangers, invent languages, rule kingdoms; you even star in a Busby Berkeley musical.
Mind you, even if it turns out that the inhibition of prefrontal metabolism during REM sleep explains the disinhibition of dream content, it still doesn’t tell us anything about why some people’s brains would want to spend some REM time in a Busby Berkeley musical. The specific content of dreams remains a mystery. Moreover, if true, this speculation would constitute one of the classic features of science—in explaining something, you’ve merely redefined the unknown. If the answer to the question “Why is dream content so disinhibited?” turns out to be “Because prefrontal cortical regions are atypically inactive during REM sleep,” the new question obviously becomes “Then why are prefrontal cortical regions atypically inactive?”
Just as with anything else that can be studied and measured in living systems, the level of activity of the prefrontal cortex varies considerably in different individuals. As noted, there seem to be decreased metabolic rates in prefrontal regions in sociopaths. At the other end of the spectrum, Richard Davidson and colleagues at the University of Wisconsin have observed elevated prefrontal metabolic rates in people with “repressive personalities.” These are highly controlled folks with superegos going full throttle, working overtime to keep their psychic sphincters good and tight. They dislike novelty, prefer structure and predictability, are poor at expressing emotions or at reading the nuances of emotions in other people. These are the folks who can tell you what they’re having for dinner two weeks from Thursday.
This leads me to an idea that seems to flow naturally from the findings of Braun and his colleagues. The data regarding the sociopath/repressive continuum come from studies of awake individuals. Most certainly, there will also be considerable variability among people in prefrontal cortex functioning during REM sleep. While prefrontal metabolism may generally remain on the floor with the transition into REM sleep, there’ll be exceptions. So I suspect it’s likely that the more prefrontal metabolism remains suppressed during REM, the more vivid and disinhibited dream content will be. Better yet would be some comparative studies of prefrontal metabolism during wake and sleep. Do people who have the most active prefrontal cortices when awake have the least active when asleep? This would certainly fit the old hydraulic models of psychoanalysis, where if you repress something important during the day, it’ll come oozing out during dreams.
I’ve occasionally heard med students come up with a witticism to express their typical disdain for classes in psychiatry: “What classes are you taking this semester?” “Oh, pathology, microbio, pharmacology, and this required seminar in laser psychotherapy.” The last is meant to be an eccentric oxymoron. Laser-something-or-other equals high tech, as opposed to psychotherapy, as the pejoratively low-tech art of talk therapy. Thus, the student is saying, “They’re forcing us to take some class with these shrinks who are trying to dress up their stuff as modern science.” Wouldn’t it be ironic if some reductive support for the seemingly antiquated Freudian concept of repression were to emerge from the bowels of a gazillion-dollar brain scanner?
NOTES AND FURTHER READING
For a nontechnical introduction to the neurobiology of sleep, see chapter 11 in Sapolsky R, Why Zebras Don’t Get Ulcers: A Guide to Stress, Stress-Related Diseases, and Coping, 3rd ed. (New York: Henry Holt, 2004).
Braun’s work is reported in Braun A, Balkin T, Wesensten N, Gwadry F, Carson R, Varga M, Baldwin P, Belenky G, and Herscovitch P, “Dissociated patterns of activity in visual cortices and their projections during human rapid eye movement sleep,” Science 279 (1998): 91.
The late maturation of the frontal cortex is documented in Paus T, Zijdenbos A, Worsley K, Collins D, Blumenthal J, Giedd J, Rapoport J, and Evans A, “Structural maturation of neural pathways in children and adolescents: in vivo study,” Science 283 (1999): 1908. The function of the frontal cortex also figures in essay nine, “The Pleasure (and Pain) of ‘Maybe’.”