by Carl Sagan
A reconstruction of a Cretaceous landscape in a swampy region in Western Canada. Most of the dinosaurs shown are bipedal and herbivorous. So far as we know, all these forms became extinct shortly thereafter.
If this sequence of events is correct, the major course of biological evolution on the Earth in the last sixty-five million years, and indeed the very existence of human beings, can be traced to the death of a distant sun. Perhaps other planets circled that star; perhaps one of those planets enjoyed a thriving biology tortuously evolved over billions of years. The supernova explosion would surely have extinguished all life on that planet and probably even driven its atmosphere into space. Do we owe our existence to a mighty stellar catastrophe that elsewhere destroyed biospheres and worlds?
After the extinction of the dinosaurs, mammals moved into daytime ecological niches. The primate fear of the dark must be a comparatively recent development. Washburn has reported that infant baboons and other young primates appear to be born with only three inborn fears—of falling, snakes, and the dark-corresponding respectively to the dangers posed by Newtonian gravitation to tree-dwellers, by our ancient enemies the reptiles, and by mammalian nocturnal predators, which must have been particularly terrifying for the visually oriented primates.
If the vampiric hypothesis is true—and it is at best a likely hypothesis—the function of sleep is built deeply into the mammalian brain; from earliest mammalian times, sleep played a fundamental role in survival. Since for primitive mammals sleepless nights would have been more dangerous for the survival of the taxon than sexless nights, sleep should be a more powerful drive than sex—which, at least in most of us, it seems to be. But eventually mammals evolved to a point where sleep could be modified by changed circumstances. With the extinction of the dinosaurs, daylight suddenly became a benevolent environment for the mammals. Daytime immobilization was no longer compulsory, and a wide variety of sleep patterns slowly developed, including the contemporary correlation of mammalian predators with extensive dreaming and mammalian prey with a more watchful dreamless sleep. Perhaps those people who can do with only a few hours’ sleep a night are the harbingers of a new human adaptation that will take full advantage of the twenty-four hours of the day. I, for one, freely confess envy for such an adaptation.
These conjectures on the origins of the mammals constitute a kind of scientific myth: they may have some germ of truth in them, but they are unlikely to be the whole story. That scientific myths make contact with more ancient myths may or may not be a coincidence. It is entirely possible that we are able to invent scientific myths only because we have previously been exposed to the other sort. Nevertheless, I cannot resist connecting this account of the origin of mammals with another curious aspect of the Genesis myth of the exile from Eden. Because it is a reptile, of course, that offers the fruit of the knowledge of good and evil—abstract and moral neocortical functions—to Adam and Eve.
There are today a few remaining large reptiles on Earth, the most striking of which is the Komodo dragon of Indonesia: cold-blooded, not very bright, but a predator exhibiting a chilling fixity of purpose. With immense patience, it will stalk a sleeping deer or boar, then suddenly slash a hind leg and hang on until the prey bleeds to death. Prey is tracked by scent, and a hunting dragon lumbers and sashays, head down, its forked tongue flicking over the ground for chemical traces. The largest adults weigh about 135 kilograms (300 pounds), are three meters (about 10 feet) long and live perhaps to be centenarians. To protect its eggs, the dragon digs trenches from two to as much as nine meters (almost 30 feet) deep—probably a defense against egg-eating mammals (and themselves: Adults are known occasionally to stalk a nest-hole, waiting for the newly hatched young to emerge and provide a little delicacy for lunch). As another clear adaptation to predators, the dragon hatchlings live in trees.
The remarkable elaboration of these adaptations shows clearly that dragons are in trouble on the planet Earth. The Komodo dragon lives in the wild only in the Lesser Sunda Islands.* There are only about 2,000 of them left. The obscurity of their locale immediately suggests that dragons are near extinction because of mammalian, chiefly human, predation, a conclusion borne out by their history over the last two centuries. All dragons with less extreme adaptations or less remote habitats are dead. I even wonder whether the systematic separation of brain mass for a given body mass between mammals and reptiles (see chart on this page) might not be the result of a systematic extinction of bright dragons by mammalian predators. In any case, it is very likely that the population of large reptiles has been declining steadily since the end of the Mesozoic Age, and that there were many more of them even one or two thousand years ago than there are today.
Varanus komodoensis, the Komodo dragon, Komodo Island, Indonesia. Courtesy of The American Museum of Natural History
St. George slaying the Dragon, by Donatello from the Chiesa di Or San Michele, Florence.
PHOTO ALINARI
The pervasiveness of dragon myths in the folk legends of many cultures is probably no accident.†
The implacable mutual hostility between man and dragon, as exemplified in the myth of St. George, is strongest in the West. (In Chapter 3 of the Book of Genesis, God ordains an eternal enmity between reptiles and humans.) But it is not a Western anomaly. It is a worldwide phenomenon. Is it only an accident that the common human sounds commanding silence or attracting attention seem strangely imitative of the hissing of reptiles? Is it possible that dragons posed a problem for our protohuman ancestors of a few million years ago, and that the terror they evoked and the deaths they caused helped bring about the evolution of human intelligence? Or does the metaphor of the serpent refer to the use of the aggressive and ritualistic reptilian component of our brain in the further evolution of the neocortex? With one exception, the Genesis account of the temptation by a reptile in Eden is the only instance in the Bible of humans understanding the language of animals. When we feared the dragons, were we fearing a part of ourselves? One way or another, there were dragons in Eden.
The Temptation by a man-serpent and the expulsion from Eden. Michelangelo, from the ceiling of the Sistine Chapel.
SCALA/EPA
The most recent dinosaur fossil is dated at about sixty million years ago. The family of man (but not the genus Homo) is some tens of millions of years old. Could there have been manlike creatures who actually encountered Tyrannosaurus rex? Could there have been dinosaurs that escaped the extinctions in the late Cretaceous Period? Could the pervasive dreams and common fears of “monsters,” which children develop shortly after they are able to talk, be evolutionary vestiges of quite adaptive—baboonlike—responses to dragons and owls?*
What functions do dreams serve today? One view, published in a reputable scientific paper, holds that the function of dreams is to wake us up a little, every now and then, to see if anyone is about to eat us. But dreams occupy such a relatively small part of normal sleep that this explanation does not seem very compelling. Moreover, as we have seen, the evidence points just the other way: today it is the mammalian predators, not the mammalian prey, who characteristically have dream-filled sleep. Much more plausible is the computer-based explanation that dreams are a spillover from the unconscious processing of the day’s experience, from the brain’s decision on how much of the daily events temporarily stored in a kind of buffer to emplace in long-term memory. The events of yesterday frequently run through my dreams; the events of two days ago, much more rarely. However, the buffer-dumping model seems unlikely to be the whole story, because it does not explain the disguises that are so characteristic of the symbolic language of dreams, a point first stressed by Freud. It also does not explain the powerful affect or emotions of dreams; I believe there are many people who have been far more thoroughly frightened by their dreams than by anything they have ever experienced while awake.
The buffer-dumping and memory-storage functions of dreams have some interesting social implications. The American psychiatrist Ernest Hart
mann of Tufts University has provided anecdotal but reasonably persuasive evidence that people who are engaged in intellectual activities during the day, especially unfamiliar intellectual activities, require more sleep at night, while, by and large, those engaged in mainly repetitive and intellectually unchallenging tasks are able to do with much less sleep. However, in part for reasons of organizational convenience, modern societies are structured as if all humans had the same sleep requirements; and in many parts of the world there is a satisfying sense of moral rectitude in rising early. The amount of sleep required for buffer dumping would then depend on how much we have both thought and experienced since the last sleep period. (There is no evidence that the causality runs backwards: people drugged with phenobarbital are not reported, during interstitial waking periods, to perform unusual intellectual accomplishments.) In this respect it would be interesting to examine individuals with very low sleep needs to determine whether the fraction of sleep time they spend dreaming is larger than it is for those with normal sleep requirements, and to determine whether their amount of sleep and dream time increases with the quality and quantity of their learning experiences while awake.
Michel Jouvet, a French neurologist at the University of Lyons, has found that dream sleep is triggered in the pons, which, while it resides in the hindbrain, is a late and essentially mammalian evolutionary development. On the other hand, Penfield has found that electrical stimulation deep into and below the temporal lobe in the neocortex and limbic complex can produce a waking state in epileptics very similar to that of dreams denuded of their symbolic and fantastic aspects. It can also induce the déjà vu experience. Much of dream affect, including fear, can also be induced by such electrical stimulation.
I once had a dream that will tantalize me forever. I dreamt I was idly thumbing through a thick history text. I could tell from the illustrations that the work was moving slowly, in the usual manner of such textbooks, through the centuries: classical times, Middle Ages, Renaissance and so on, gradually approaching the modern era. But then there was World War II with about two hundred pages left. With mounting excitement I worked my way more deeply into the work until I was sure that I had passed my own time. It was a history book that included the future—like turning the December 31 page of the Cosmic Calendar and finding a fully detailed January 1. Breathlessly I attempted literally to read the future. But it was impossible. I could make out individual words. I could even discern the serifs on the individual characters. But I could not put the letters together into words or words together into sentences. I was alexic.
Perhaps this is simply a metaphor of the unpredictability of the future. But my invariable dream experience is that I am unable to read. I can recognize, for example, a stop sign by its color and its octagonal shape, but I cannot read the word stop, although I know it is there. I have the impression of understanding the meaning of a page of type, but not by reading it word by word or sentence by sentence. I cannot reliably perform even simple arithmetic operations in the dream state. I make a variety of verbal confusions of no apparent symbolic significance, like mixing up Schumann and Schubert. I am a little aphasic and entirely alexic. Not everyone I know has the same cognitive impairment in the dream state, but people often have some impairment. (Incidentally, individuals who are blind from birth have auditory, not visual dreams.) The neocortex is by no means altogether turned off in the dream state, but it certainly seems to suffer important malfunctions.
The seeming fact that mammals and birds both dream while their common ancestor, the reptiles, do not is surely noteworthy. Major evolution beyond the reptiles has been accompanied by and perhaps requires dreams. The electrically distinctive sleep of birds is episodic and brief. If they dream, they dream for only about a second at a time. But birds are, in an evolutionary sense, much closer to reptiles than mammals are. If we knew only about mammals, the argument would be more shaky; but when both major taxonomic groups that have evolved from the reptiles find themselves compelled to dream, we must take the coincidence seriously. Why should an animal that has evolved from a reptile have to dream while other animals do not? Could it be because the reptilian brain is still present and functioning?
It is extremely rare in the dream state that we bring ourselves up short and say, “This is only a dream.” By and large we invest the dream content with reality. There are no rules of internal consistency that dreams are required to follow. The dream is a world of magic and ritual, passion and anger, but very rarely of skepticism and reason. In the metaphor of the triune brain, dreams are partially a function of the R-complex and the limbic cortex, but not of the rational part of the neocortex.
Experiments suggest that as the night wears on our dreams engage increasingly earlier material from our past, reaching back to childhood and infancy. At the same time the primary process and emotional content of the dream also increase. We are much more likely to dream of the passions of the cradle just before awakening than just after falling asleep. This looks very much as if the integration of the day’s experience into our memory, the forging of new neural links, is either an easier or a more urgent task. As the night wears on and this function is completed, the more affecting dreams, the more bizarre material, the fears and lusts and other powerful emotions of the dream material emerge. Late at night, when it is very still and the obligatory daily dreams have been dreamt, the gazelles and the dragons begin to stir.
One of the most significant tools in studying the dream state was developed by William Dement, a Stanford University psychiatrist, who is as sane as it is possible for a human being to be, but who bears an exceedingly interesting name for a man of his profession. The dream state is accompanied by rapid eye movements (REM), which can be detected by electrodes taped lightly over the eyelids in sleep, and by a particular brain wave pattern on the EEG. Dement has found that everyone dreams many times each night. On awakening, an individual in the midst of REM sleep will usually remember his dream. Even people who claim never to dream have been discovered by REM arid EEG criteria to dream as much as anyone else; and, when awakened at appropriate times, they admit with some surprise to having dreamt. The human brain is in a distinct physiological state while dreaming, and we dream rather often. While perhaps 20 percent of the subjects awakened during REM sleep do not recall their dreams, and some perhaps 10 percent’ of subjects awakened during non-REM sleep report dreams, we will, for convenience, identify REM and accompanying EEG patterns with the dream state.
There is some evidence that dreaming is necessary. When people or other mammals are deprived of REM sleep (by awakening them as soon as the characteristic REM and EEG dream patterns emerge), the number of initiations of the dream state per night goes up, and, in severe cases, daytime hallucinations—that is, waking dreams—occur. I have mentioned that the REM and EEG patterns of dreams are brief in birds and absent in reptiles. Dreams seem to be primarily a mammalian function. What is more, dream sleep is most vigorously engaged in by human beings in the early postnatal period. Aristotle stated quite positively that infants do not dream at all. On the contrary, we find they may be dreaming most of the time. Full-term newborn babies spend more than half their sleep time in the REM dream state. In infants born a few weeks premature, the dream time is three-quarters or more of the total sleep time. Earlier in its intrauterine existence, the fetus may be dreaming all the time. (Indeed, newborn kittens are observed to spend all of their sleep time in the REM stage.) Recapitulation would then suggest that dreaming is an evolutionarily early and basic mammalian function.
There is another connection between infancy and dreams: both are followed by amnesia. When we emerge from either state, we have great difficulty remembering what we have experienced. In both cases, I would suggest, the left hemisphere of the neocortex, which is responsible for analytic recollection, has been functioning ineffectively. An alternative explanation is that in both dreams and early childhood we experience a kind of traumatic amnesia: The experiences are too painful to remember. Bu
t many dreams we forget are very pleasant, and it is difficult to believe that infancy is that unpleasant. Also some children seem capable of remembering extremely early experiences. Memories of events late in the first year of life are not extremely rare, and there are possible examples of even earlier recollections. At age three, my son Nicholas was asked for the earliest event he could recall and replied in a hushed tone while staring into middle distance, “It was red, and I was very cold.” He was born by Caesarean section. It is probably very unlikely, but I wonder whether this could just possibly be a true birth memory. At any rate, I think it is much more likely that childhood and dream amnesia arise from the fact that in those states our mental lives are determined almost entirely by the R-complex, the limbic system and the right cerebral hemisphere. In earliest childhood, the neocortex is underdeveloped; in amnesia, it is impaired.
There is a striking correlation of penile or clitoral erection with REM sleep, even when the manifest dream content has no overt sexual aspects whatever. In primates, such erections are connected with sex (of course!), aggression and the maintenance of social hierarchies. I think that when we dream there is a part of us engaged in activities rather like those of the squirrel monkeys I saw in Paul MacLean’s laboratory. The R-complex is functioning in the dreams of humans; the dragons can be heard, hissing and rasping, and the dinosaurs thunder still.
One excellent test of the merit of scientific ideas is their subsequent validation. A theory is put forward on fragmentary evidence, then an experiment is performed, the outcome of which the proposer of the theory could not know. If the experiment confirms the original idea, this is usually taken as strong support for the theory. Freud held that the great majority, perhaps all, of the “psychic energy” of our primary-process emotions and dream material is sexual in origin. The absolutely essential role of sexual interest in providing for the propagation of the species makes this idea neither as silly nor as depraved as it appeared to many of Freud’s Victorian contemporaries. Carl Gustav Jung, for example, held that Freud had severely overstated the primacy of sex in the affairs of the unconscious. But now, three-quarters of a century later, experiments in the laboratories of Dement and other psychologists appear to support Freud. It would, I think, require a very dedicated puritanism to deny some connection between penile or clitoral erection and sex. It seems to follow that sex and dreams are not casually or incidentally connected but rather have deep and fundamental ties—although dreams certainly partake of ritual, aggressive and hierarchical material as well. Particularly considering the state of sexual repression in late-nineteenth-century Viennese society, many of Freud’s insights seem hard-won and courageous as well as valid.