by Robert Sapolsky
You find yourself at a banquet table. You
feel disaffected because the people surrounding you are speaking a language
you do not understand. Suddenly, beneath the table, you feel someone's foot
on top of your own. You glance up. Your eyes meet those of an attractive
person and you sense there is one word that you now must say: "Phlegm." The
person stands, and suddenly everyone else at the banquet is gone. As is the
table. As are your clothes.
You fling yourselves at each other in passion. You rise up in the air, the
sensuality of the experience heightened by clouds brushing past. Yet you
begin to sob in shame because you have been observed by your four deceased
grandparents, who disapprove. You suddenly 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're dreaming.
To state a truism, just as the kidney could accurately be described as a
kidney-shaped organ, dreams are dreamlike. 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 such a moment you might remember you forgot 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 and
irrationality. In dreams, you do things that with two seconds of sensible
reflection you couldn't bring yourself to do in real life.
There has never been a shortage of theories about the utility of dreams
being dreamlike. Maybe dreaming is the channel through which the gods choose
to speak to mortals. Maybe it's the means to work out how you really feel
about your mother without all that repression stuff 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.
Maybe it's how you keep underutilized neural pathways active by giving a
workout to neurons that don't get exercised during the day. Or maybe the
whole thing 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 one thing we've known for some time is that there is a
structure--an architecture, if you will--to sleep, with rhythmic cycles of
deep, slow-wave sleep interspersed with the rapid-eye-movement (REM) 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 activity in the brain have uncovered something
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--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 small subregions of the brain rather than just the brain
as a whole. In a pioneering series of studies, Allen Braun and his
colleagues at the National Institutes of Health have taken a close look at
the neuroanatomy of metabolism during sleep. In the process I think they may
have uncovered the explanation for why dreams are so dreamlike.
The researchers utilized positron emission tomography, or PET scans, to
measure the various rates of blood flow throughout the brain. One of the
brain's remarkably adaptive features 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 a particular area of the brain can be
used as an indirect index of the level of activity there. That is why PET
scans, which easily 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 24 to 53 hours. Each bleary volunteer was
eventually rolled into a PET machine and forced to stay awake while a
baseline scan was made. Then, snug as a bug inside, each study subject was
finally allowed to sleep while the scanning continued.
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 regulating 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 bring information to and from those
regions shut down dramatically, thus isolating them metabolically. The parts
of the brain that first respond to sensory information had something of a
metabolic shutdown, but the more dramatic changes were in downstream brain
areas that integrate and associate those bytes of sensory information and
give them meaning.
The result: metabolically quiescent, sleeping brains.
While the scientists at the scanner's console bided their time,
the sleeping subjects transitioned into REM sleep. And then the picture
changed. Metabolic rates lept upward throughout subregions of the brain.
Cortical and subcortical regions that regulate muscle movement and
brain-stem regions that control breathing and heart rate all showed
increases. In a part of the brain known as the limbic system, which is
involved in emotion, there was an increase as well. The same was true for
areas having to do with memory and sensory processing, especially those
connected to vision and hearing.
Meanwhile, something subtle went on in the visual processing regions. The
primary visual cortical region did not show much of an increase in
metabolism, but 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, and scenes. Normally, an increase in
activity in the downstream areas cannot occur without an increase in the
primary areas. In other words, when you're wide awake, you can't get your
eyes to see complex pictures without first going through an initial level of
analysis. But REM sleep is a special case--you're not using the eyes.
Instead, you're starting with the downstream integration of visual patterns.
This, Braun and his colleagues have 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 when someone is awake. But researchers have also found an
exception that I think may be the answer to why dreams are dreamlike, 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. Within
the prefrontal cortex, however, only one of the four subregions increased in
activity. The rest of those areas stayed on the floor of metabolic
inactivity they had sunk to during the period of slow-wave sleep.
This is intriguing, given the functions of the prefrontal cortex. The human
brain has many unique features when compared with an off-the-rack mammalian
brain. Its sensory inputs and motor outputs are uniquely fine-tuned to make
it possible to whip out an arpeggio on a piano. The limbic system allows for
something virtually unprecedented among mammals: sexual receptivity among
females throughout the reproductive cycle, not only 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 a particularly dreadful human attribute--the ability
to think oneself into a depression.
Yet in many ways, the most remarkable feature of the human brain is the
extent of the development and the power of that prefrontal cortex, the
region that stays metabolically inactive during REM sleep. The prefrontal
cortex is the brain region that 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 or an important business meeting.
On the 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 that created
by strokes, causes a disinhibited, frontal personality. The person may
become apathetic or childishly silly, hypersexual or bellicose as hell,
scatological or blasphemous.
Braun and his 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 there were high rates of activity in the complex
sensory processing parts of the brain concerned with emotion and memories.
So bring on those dreams, now free to be filled with uninhibited actions and
labile emotions.
You can breathe underwater, fly in the air, communicate telepathically; you
can announce your love to strangers, invent languages, rule kingdoms; you
can even star in a Busby Berkeley musical.
Mind you, even if it turns out that the lack of metabolic activity in the
prefrontal cortex during REM sleep explains the disinhibition of dream
content, it still doesn't tell us anything about why anyone's brain would
spend time staging that particular 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, we've merely
managed to redefine the unknown. Suppose 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, there is considerable variability in the level of activity of the
prefrontal cortex in different individuals. At one end of the spectrum,
there seem to be decreased metabolic rates in prefrontal regions in
sociopaths. At the other end of the spectrum, Richard Davidson, at the
University of Wisconsin at Madison, and colleagues have observed elevated
prefrontal metabolic rates in people with so-called 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 as to how the prefrontal
cortex functions during REM sleep. While prefrontal metabolism may remain on
the floor with the transition into REM sleep on average, there will 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, perhaps in a subject-specific manner. Most revealing would be some
comparative studies of prefrontal metabolism during waking and sleep- ing
periods. Do peo-ple 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, which postulate that if you repress
something important during the day, it will most likely come oozing out
during dreams.
At Stanford, where I direct a neuroscience lab, I've occasionally heard
medical students come up with a witticism to express their disdain for
classes in psychiatry. Question: "What classes are you taking this
semester?" Answer: "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, 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 that seemingly antiquated Freudian
concept of repression were to emerge from the bowels of a gazillion-dollar
scanning machine?