IT’S NOT THE MOST comfortable way to raise funds for a major American charity. In 1959, New York disk jockey Peter Tripp decided that he would stay awake for 200 straight hours. He got into a glass booth in the most visible place possible in New York—Times Square—and rigged up the radio so that he could broadcast his show. He even allowed scientists (and, wisely, medical professionals) to observe and measure his behavior as he descended into sleeplessness. One of those scientists was famed sleep researcher William Dement. For the first 72 hours, everything seemed fine with Tripp. He gave his normal three-hour show with humor and professional aplomb. Then things changed. Tripp became rude and offensive to the people around him. Hallucinations set in. The researchers testing his cognitive skills halfway through found he could no longer complete certain mental skill tests. At the 120-hour mark—five days in—Tripp showed real signs of mental impairment, which would only worsen with time. Dement described Tripp’s behavior toward the end of the adventure: “The disk jockey developed an acute paranoid psychosis during the nighttime hours, accompanied at times by auditory hallucination. He believed that unknown adversaries were attempting to slip drugs into his food and beverages in order to put him to sleep.” At the 200-hour mark—more than eight days—Tripp was done. Presumably, he went to bed and stayed there for a long time.

      Some unfortunate souls don’t have the luxury of experimenting with sleep deprivation. They become suddenly and permanently incapable of ever going to sleep again. Only about 20 families in the world suffer from Fatal Familial Insomnia, making it one of the rarest human genetic disorders that exists. That rarity is a blessing, because the disease follows a course straight through mental-health hell. In middle to late adulthood, the person begins to experience fevers, tremors, and profuse sweating. As the insomnia becomes permanent, these symptoms are accompanied by increasingly uncontrollable muscular jerks and tics. The person soon experiences crushing feelings of depression and anxiety. He or she becomes psychotic. Finally, mercifully, the patient slips into a coma and dies.

      So we know bad things happen when we don’t sleep. The puzzle is that, from an evolutionary standpoint, bad things also could happen when we do sleep. Because the body goes into a human version of micro-hibernation, sleep makes us exquisitely vulnerable to predators. Indeed, deliberately going off to dreamland unprotected in the middle of a bunch of hostile hunters (such as leopards, our evolutionary roommates in eastern Africa) seems like a plan dreamed up by our worst enemies. There must be something terribly important we need to accomplish during sleep if we are willing to take such risks in order to get it. Exactly what is it that is so darned important?

      To begin to understand why we spend a walloping one-third of our time on this planet sleeping, let’s peer in on what the brain is doing while we sleep.

      You call this rest?

      If you ever get a chance to listen in on someone’s brain while its owner is slumbering, you’ll have to get over your disbelief. The brain does not appear to be asleep at all. Rather, it is almost unbelievably active during “rest,” with legions of neurons crackling electrical commands to one another in constantly shifting, extremely active patterns. In fact, the only time you can observe a real resting period for the brain—where the amount of energy consumed is less than during a similar awake period—is during the phase called non-REM sleep. But that takes up only about 20 percent of the total sleep cycle. This is why researchers early on began to disabuse themselves of the notion that the reason we rest is so that we can rest. When we are asleep, the brain is not resting at all. Even so, most people report that sleep is powerfully restorative, and they point to the fact that if they don’t get enough sleep, they don’t think as well the next day. That is measurably true, as we shall see shortly. And so we find ourselves in a quandary: Given the amount of energy the brain is using, it seems impossible that you could receive anything approaching mental rest and restoration during sleep.

      Two scientists made substantial early contributions to our understanding of what the brain is doing while we sleep. Dement, who studied sleepless Peter Tripp, is a white-haired man with a broad smile who at this writing is in his late 80s. He says pithy things about our slumbering habits, such as “Dreaming permits each and every one of us to be quietly and safely insane every night of our lives.” Dement’s mentor, a gifted researcher named Nathaniel Kleitman, gave him many of his initial insights. If Dement can be considered the father of sleep research, Kleitman certainly could qualify as its grandfather. An intense Russian man with bushy eyebrows, Kleitman may be best noted for his willingness to experiment not only on himself but also on his children. When it appeared that a colleague of his had discovered rapid eye movement (REM) sleep, Kleitman promptly volunteered his daughter for experimentation, and she just as promptly confirmed the finding. He also persuaded a colleague to live with him underground to see what would happen to their sleep cycles without the influence of light and social cues. Here are some of the things Dement and Kleitman discovered about sleep.

      Sleep is a battle

      Like soldiers on a battlefield, we have two powerful and opposing drives locked in vicious, biological combat. The armies, each made of legions of brain cells and biochemicals, have very different agendas. Though localized in the head, the theater of operations for these armies engulfs every corner of the body. The war they are waging has some interesting rules. First, these forces are engaged not just during the night, while we sleep, but also during the day, while we are awake. Second, they are doomed to a combat schedule in which each army sequentially wins one battle, then promptly loses the next battle, then quickly wins the next and so on, cycling through this win/loss column every day and every night. Third, neither army ever claims final victory. This incessant engagement is referred to as the “opponent process” model. It results in the waking and sleeping modes all humans cycle through every day (and night) of our lives.

      One army is composed of neurons, hormones, and various other chemicals that do everything in their power to keep you awake. This army is called the circadian arousal system (often simply called “process C”). If this army had its way, you would stay up all the time. It is opposed by an equally powerful army, also made of brain cells, hormones, and various chemicals. These combatants do everything in their power to put you to sleep. They are termed the homeostatic sleep drive (“process S”). If this army had its way, you would go to sleep and never wake up. These drives define for us both the amount of sleep we need and the amount of sleep we get. Stated formally, process S maintains the duration and intensity of sleep, while process C determines the tendency and timing of the need to go to sleep.

      It is a paradoxical war. The longer one army controls the field, for example, the more likely it is to lose the battle. It’s almost as if each army becomes exhausted from having its way and eventually waves a temporary white flag. Indeed, the longer you are awake (the victorious process C doing victory laps around your head), the greater the probability becomes that the circadian arousal system will cede the field to its opponent. You then go to sleep. For most people, this act of capitulation comes after about 16 hours of active consciousness. This will occur, Kleitman found, even if you are living in a cave.

      Conversely, the longer you are asleep (the triumphant process S now doing the heady victory laps), the greater the probability becomes that the homeostatic sleep drive will similarly cede the field to its opponent, which is, of course, the drive to keep you awake. The result of this surrender is that you wake up. For most people, the length of time prior to capitulation is about half of its opponent’s, about eight hours of blissful sleep. And this also will occur even if you are living in a cave.

      Such dynamic tension is a normal—even critical—part of our daily lives. In fact, the circadian arousal system and the homeostatic sleep drive are locked in a cycle of victory and surrender so predictable, you can graph it.

      In one of Kleitman’s most interesting experiments, he and a colleague spent an entire month living 1,300 feet underground in Mammoth Cave in Kentucky. Free of sunlight and daily schedules, Kleitman could find out whether the routines of wakefulness and sleep cycled themselves automatically through the human body. His experiment provided the first real hint that such an automatic device did exist in our bodies. Indeed, we now know that the body possesses a series of internal clocks, all controlled