Ernest Hartmann, a professor of psychiatry at Tufts University School of Medicine and the director of the Sleep Disorders Center at Newton Wellesley Hospital in Boston, Mass., explains.
The questions, "Why do we dream?" or "What is the function of dreaming?" are easy to ask but very difficult to answer. The most honest answer is that we do not yet know the function or functions of dreaming. This ignorance should not be surprising because despite many theories we still do not fully understand the purpose of sleep, nor do we know the functions of REM (rapid eye movement) sleep, which is when most dreaming occurs. And these two biological states are much easier to study scientifically than the somewhat elusive phenomenon of dreaming.
Some scientists take the position that dreaming probably has no function. They feel that sleep, and within it REM sleep, have biological functions (though these are not totally established) and that dreaming is simply an epiphenomenon that is the mental activity that occurs during REM sleep. I do not believe this is the most fruitful approach to the study of dreaming. Would we be satisfied with the view that thinking has no function and is simply an epiphenomenon--the kind of mental activity that occurs when the brain is in the waking state?
Therefore I will try to explain a current view of dreaming and its possible functions, developed by myself and many collaborators, which we call the Contemporary Theory of Dreaming. The basic idea is as follows: activation patterns are shifting and connections are being made and unmade constantly in our brains, forming the physical basis for our minds. There is a whole continuum in the making of connections that we subsequently experience as mental functioning. At one end of the continuum is focused waking activity, such as when we are doing an arithmetic problem or chasing down a fly ball in the outfield. Here our mental functioning is focused, linear and well-bounded. When we move from focused waking to looser waking thought--reverie, daydreaming and finally dreaming--mental activity becomes less focused, looser, more global and more imagistic. Dreaming is the far end of this continuum: the state in which we make connections most loosely.
Some consider this loose making of connections to be a random process, in which case dreams would be basically meaningless. The Contemporary Theory of Dreaming holds that the process is not random, however, and that it is instead guided by the emotions of the dreamer. When one clear-cut emotion is present, dreams are often very simple. Thus people who experience trauma--such as an escape from a burning building, an attack or a rape--often have a dream something like, "I was on the beach and was swept away by a tidal wave." This case is paradigmatic. It is obvious that the dreamer is not dreaming about the actual traumatic event, but is instead picturing the emotion, "I am terrified. I am overwhelmed." When the emotional state is less clear, or when there are several emotions or concerns at once, the dream becomes more complicated. We have statistics showing that such intense dreams are indeed more frequent and more intense after trauma. In fact, the intensity of the central dream imagery, which can be rated reliably, appears to be a measure of the emotional arousal of the dreamer.
Therefore, overall the contemporary theory considers dreaming to be a broad making of connections guided by emotion. But is this simply something that occurs in the brain or does it have a purpose as well? Function is always very hard to prove, but the contemporary theory suggests a function based on studies of a great many people after traumatic or stressful new events. Someone who has just escaped from a fire may dream about the actual fire a few times, then may dream about being swept away by a tidal wave. Then over the next weeks the dreams gradually connect the fire and tidal wave image with other traumatic or difficult experiences the person may have had in the past. The dreams then gradually return to their more ordinary state. The dream appears to be somehow "connecting up" or "weaving in" the new material in the mind, which suggests a possible function. In the immediate sense, making these connections and tying things down diminishes the emotional disturbance or arousal. In the longer term, the traumatic material is connected with other parts of the memory systems so that it is no longer so unique or extreme--the idea being that the next time something similar or vaguely similar occurs, the connections will already be present and the event will not be quite so traumatic. This sort of function may have been more important to our ancestors, who probably experienced trauma more frequently and constantly than we (at least those of us living in the industrialized world) do at present.
Thus we consider a possible (though certainly not proven) function of a dream to be weaving new material into the memory system in a way that both reduces emotional arousal and is adaptive in helping us cope with further trauma or stressful events.
Answer originally posted July 14, 2003.