Scientific discoveries depending on observations usually call for many years of patient work. For example, William Herschel and his sister Caroline spent endless nights at his telescope, carefully counting stars in different parts of the sky in order to build up today's picture of the structure of our own galaxy. Gregor Mendel worked for years in the garden of his monastery at Brunn, breeding successive generations of peas and teasing out from their reproductive patterns the basic laws of genetics. Charles Darwin drew on five years of world travel on H.M.S. Beagle in order to formulate his ideas on evolution and natural selection.

So when you come across a basic discovery that resulted from a few weeks of effort, and learn that the work in question was done as recently as the 1950s, it is a shock. It is still more of a shock when you realize that the same finding could have been made in 1852, or 1752, or in fact at any time after humans learned the use of fire. The discovery was of the existence of a particular pattern of sleep known as Rapid Eye Movement, or REM, and it was found more or less by accident in the laboratory of Nathaniel Kleitman in 1952.

One reason that this discovery took so long was that no one had spent much time simply watching people sleep. Most people assumed that it was a most unproductive use of time, because the brain was believed to be inactive while you slept. Kleitman's graduate student, Eugene Aserinsky, was probably less than thrilled when he was assigned the task of watching people sleep, especially since it's a job you have to do at night, when you are sleepy yourself. To make things easier, Kleitman and Aserinsky found a way to automate the task. When your eyeball moves, it creates electrical signals; so if you put electrodes on the skin close to a person's eye, you can make a continuous record of eye movements.

Aserinsky hooked his volunteers up to his recording apparatus. To his astonishment, all his sleeping subjects displayed the same pattern of behavior. The time when they were asleep fell into distinct phases. When someone first went to sleep, there were normally only slow movements or no movements of the eye. Today this is referred to as SWS, or slow wave sleep. But after a while, the eyes would flicker rapidly from side to side. This would occur for maybe a total of two hours, in a subject who slept for eight. The unanswered question was, why? What was going on when someone moved from SWS into a REM sleep phase?

One reason that question is difficult to answer is that we don't really know why people (and animals) have a need for sleep. A couple of years ago I wrote a column on sleep debt and sleep rhythms, but I avoided the most basic question: Why do we sleep at all? We know that it is necessary, because humans systematically deprived of sleep first become insane, then die. But why do we need it?

Over the years, a variety of answers have been offered. The most obvious one, that we sleep now because for billions of years there was nothing else to do when it became dark, is hardly satisfactory. Many predators have evolved to be more active at night. Humans would be a good deal less vulnerable if we could remain awake all the time.

Another answer is that sleep is a holdover from the time when we were cold-blooded creatures. As the thermometer drops, an animal that cannot regulate its own body temperature has no choice but to become more lethargic. This also is not much of a reason. You would think that as we became warm-blooded, our need to sleep would gradually have diminished and finally disappeared. Sleeplessness would be an added advantage to maintaining a constant body temperature.

There have been plenty of other theories as to why sleep is essential: it is a method of energy conservation (the active brain eats up large amounts of energy); it is a way of performing thermal regulation within the brain; it forms a necessary period of brain "detoxification," and, popular since the widespread use of computers, it is a period when the brain organizes, sorts, and stores the information that it received while the owner was awake.

These theories are not mutually exclusive. Our need for sleep could be a combination of several of them. But what, then, is the purpose of REM sleep? Almost from the day of discovery, it was suggested that this form of sleep might be related to dreaming. That is believed to be true to this day, and much current research involves sleep, learning, dreams, and the processing of memories.

No matter how this turns out, it leaves me with unanswered questions. We have learned that animals not only sleep, they also have a REM phase in that sleep. In some of them, that REM phase far exceeds what we find in humans. To take one extreme example, out of an estimated fourteen hours of sleep a day, the duck-billed platypus spends eight hours in REM mode. The platypus is a very primitive mammal, one of only two species that still lay eggs. Its life is rarely enriched with a variety of strange events. What new information can it possibly be processing for so long, and why would it spend so much time dreaming?

At the other extreme, we have the dolphin, which spends less than twelve minutes a day in REM sleep mode. The dolphin's sleep is odd in other ways. Like any mammal that lives in the sea, it cannot relegate its breathing, as humans do, to the functions of the autonomic nervous system. Dolphins are obliged to retain conscious control over their breathing. They do this by having one hemisphere of the brain go to sleep at a time. The left side will sleep for a couple of hours, then the right one, until the total reaches about ten hours a day. The extreme measures taken by the dolphin to find some way of achieving sleep offers powerful proof that sleep, no matter how inconvenient, is absolutely necessary not only for humans, but for animals.

Sleep remains a mystery, and if I were going into research tomorrow that's where I would head. Our degree of ignorance remains so large, there are surely great discoveries to be made.

"Borderlands of Science" is syndicated by: