The Dragons of Eden

by Carl Sagan

  In the same way, the argument about the “potential” to be human seems to me particularly weak. Any human egg or sperm under appropriate circumstances has the potential to become a human being. Yet male masturbation and noctural emissions are generally considered natural acts and not cause for murder indictments. In a single ejaculation there are enough spermatozoa for the generation of hundreds of millions of human beings. In addition, it is possible that in the not-too-distant future we may be able to clone a whole human being from a single cell taken from essentially anywhere in the donor’s body. If so, any cell in my body has the potential to become a human being if properly preserved until the time of a practical cloning technology. Am I committing mass murder if I prick my finger and lose a drop of blood?

  The issues are clearly complex. The solution, equally clearly, must involve a compromise among a number of cherished but conflicting values. The key practical question is to determine when a fetus becomes human. This in turn rests on what we mean by human. Surely not having a human shape, because an artifact of organic materials that resembled a human being but was constructed for the purpose would certainly not be considered human. Likewise, an extraterrestrial intelligent being who did not resemble human beings but who had ethical, intellectual and artistic accomplishments exceeding our own should certainly fall within our prohibitions against murder. It is not what we look like that specifies our humanity, but what we are. The reason we prohibit the killing of human beings must be because of some quality human beings possess, a quality we especially prize, that few or no other organisms on Earth enjoy. It cannot be the ability to feel pain or deep emotions, because that surely extends to many of the animals we gratuitously slaughter.

  This essential human quality, I believe, can only be our intelligence. If so, the particular sanctity of human life can be identified with the development and functioning of the neocortex. We cannot require its full development, because that does not occur until many years after birth. But perhaps we might set the transition to humanity at the time when neocortical activity begins, as determined by electroencephalography of the fetus. Some insights on when the brain develops a distinctly human character emerge from the simplest embryological observations (see the figure on this page). Very little work has been done in this field to date, and it seems to me that such investigations could play a major role in achieving an acceptable compromise in the abortion debate. Undoubtedly there would be a variation from fetus to fetus as to the time of initiation on the first neocortical EEG signals, and a legal definition of the beginning of characteristically human life should be biased conservatively—that is, toward the youngest fetus that exhibits such activity. Perhaps the transition would fall toward the end of the first trimester or near the beginning of the second trimester of pregnancy. (Here we are talking about what, in rational society, should be prohibited by law: anyone who feels that abortion of a younger fetus might be murder should be under no legal obligation to perform or accept such an abortion.)

  Embryonic development of the human brain. Shown are A after three weeks of gestation; B after seven weeks; C after four months; and D in a newborn infant. The brains in A and B have strong resemblances to the brains of fish and amphibians.

  But a consistent application of these ideas must avoid human chauvinism. If there are other organisms that share the intelligence of a somewhat backward but fully developed human being, they at least should be offered the same protection against murder that we are willing to extend to human beings late in their uterine existence. Since the evidence for intelligence in dolphins, whales and apes is now at least moderately compelling, any consistent moral posture on abortion should, I would think, include firm strictures against at least the gratuitous slaughter of these animals. But the ultimate key to the solution of the abortion debate would seem to be the investigation of prepartum neocortical activity.

  And what of the future evolution of the human brain? There is a wide and growing body of evidence that many forms of mental illness are the result of chemical or wiring malfunctions in the brain. Since many mental diseases have the same symptoms, they may arise from the same malfunctions and should be accessible to the same cures.

  The pioneering nineteenth-century British neurologist Hughlings Jackson remarked, “Find out about dreams and you will find out about insanity.” Severely dream deprived subjects often begin hallucinating in daytime. Schizophrenia is often accompanied by night-time sleep impairment, but whether as a cause or an effect is uncertain. One of the most striking aspects of schizophrenia is how unhappy and despairing its sufferers generally are. Might schizophrenia be what happens when the dragons are no longer safely chained at night; when they break the left-hemisphere shackles and burst forth in daylight? Other diseases perhaps result from an impairment of right-hemisphere function: Obsessive-compulsives, for example, are very rarely found to make intuitive leaps.

  In the middle 1960s Lester Grinspoon and his colleagues at Harvard Medical School performed a set of controlled experiments on the relative value of various therapeutic techniques for treating schizophrenia. They are psychiatrists, and if they had any bias it was toward the use of verbal rather than pharmacological techniques. But they found to their surprise that the recently developed tranquilizer, thioridazine (one of a group of approximately equally effective antipsychotic drugs known as phenothiazines), was far more effective in controlling if not curing the disease; in fact, they found that thioridazine alone was at least as effective—in the judgment of the patients, their relatives, and the psychiatrists—as thioridazine plus psychotherapy. The integrity of the experimenters in the face of this unexpected finding is breathtaking. (It is difficult to imagine any experiment that would convince leading practitioners of many political or religious philosophies of the superiority of a competing doctrine.)

  Recent research shows that endorphins, small protein molecules which occur naturally in the brains of rats and other mammals, can induce in these animals marked muscular rigidity and stupor reminiscent of schizophrenic catatonia. The molecular or neurological cause of schizophrenia—which was once responsible for one out of ten hospital-bed occupancies in the United States—is still unknown; but it is not implausible that someday we will discover precisely what locale or set of neurochemicals in the brain determines this malfunction.

  A curious question in medical ethics emerges from the experiments of Grinspoon et al. The tranquilizers are now so effective in treating schizophrenia that it is widely considered unethical to withhold them from a patient. The implication is that the experiments showing tranquilizers to be effective cannot be repeated. It is thought to be an unnecessary cruelty to deny the patient the most successful treatment for his condition. Consequently, there can no longer be a control group of schizophrenics that is not given tranquilizers. If critical experiments in the chemotherapy of brain malfunction can be performed only once, they must be performed the first time very well indeed.

  An even more striking example of such chemotherapy is the use of lithium carbonate in the treatment of manic depressives. The ingestion of carefully controlled doses of lithium, the lightest and simplest metal, produces startling improvements—again as reported from the patients’ perspective and from the perspective of others—in this agonizing disease. Why so simple a therapy is so strikingly effective is unknown, but it most likely relates to the enzyme chemistry of the brain.

  A very strange mental illness is Gilles de la Tourette’s disease (named, as always, after the physician who first drew attention to it, not after the most celebrated sufferer of the malady). One of the many motor and speech disorders that are among the symptoms of this disease is a remarkable compulsion to utter—in whatever language the patient is most fluent—an uninterrupted stream of obscenities and profanities. Physicians describe the identification of this disease as “corridor diagnosis”: The patient can, with great difficulty, control his compulsion for the length of a short medical visit; as soon as the physician leaves the room for t
he corridor, the scatologies overflow like the flood from a burst dam. There is a place in the brain that makes “dirty” words (and apes may have it).

  There are very few words that the right hemisphere can deal with competently—not much more than hello, goodbye, and … a few choice obscenities. Perhaps Tourette’s disease affects the left hemisphere only. The British anthropologist Bernard Campbell of Cambridge University suggests that the limbic system is rather well integrated with the right cerebral hemisphere, which, as we have seen, deals much better with emotions than the left hemisphere does. Whatever else they involve, obscenities carry with them strong emotions. Yet Gilles de la Tourette’s disease, complex as it is, seems to be a specific deficiency in a neuronal transmitter chemical, and appears to be alleviated by carefully controlled doses of haloperidol.

  Recent evidence indicates that such limbic hormones as ACTH and vasopressin can greatly improve the ability of animals to retain and recall memories. These and similar examples suggest, if not the ultimate perfectability of the brain, at least prospects for its substantial improvement—perhaps through altering the abundance or controlling the production of small brain proteins. Such examples also greatly relieve the burden of guilt commonly experienced by sufferers from a mental disease, a burden rarely felt in victims of, say measles.

  The remarkable fissurization, convolutions and cortical folding of the brain, as well as the fact that the brain fits so snugly into the skull, are clear indications that packing more brain into the present braincase is going to be difficult. Larger brains with larger skulls could not develop until very recently because of limits on the size of the pelvis and the birth canal. But the advent of Caesarean section—performed rarely two thousand years ago but much more commonly today—does permit larger brain volumes. Another possibility is a medical technology sufficiently advanced to permit full-term development of the fetus outside of the uterus. However, the rate of evolutionary change is so slow that none of the problems facing us today is likely to be overcome by significantly larger neocortices and consequent superior intelligences. Before such a time, but not in the immediate future, it may be possible, by brain surgery, to improve those components of the brain we consider worth improving and to inhibit further those components that may be responsible for some of the perils and contradictions facing mankind. But the complexity and redundancy of brain function make such a course of action impractical for the near future, even if it were socially desirable. We may be able to engineer genes before we are able to engineer brains.

  It is sometimes suggested that such experiments may provide unscrupulous governments—and there are many of them—with tools to control their citizenry still further. For example, we can imagine a government that implants hundreds of tiny electrodes in the “pleasure” and “pain” centers of the brains of newborn children, electrodes capable of remote radio stimulation—perhaps at frequencies or with access codes known only to the government. When the child grows up, the government might stimulate his pleasure centers if he has performed, in work quota and ideology, an acceptable day’s work; otherwise it might stimulate his pain centers. This is a nightmarish vision, but I do not think it is an argument against experiments on electrical stimulation of the brain. It is, rather, an argument against letting the government control the hospitals. Any people that will permit its government to implant such electrodes has already lost the battle and may well deserve what it gets. As in all such technological nightmares, the principal task is to foresee what is possible; to educate the public in its use and misuse; and to prevent its organizational, bureaucratic and governmental abuse.

  There is already a range of psychotropic and mood-altering drugs which are, to varying degrees, dangerous or benign (ethyl alcohol is the most widely used and one of the most dangerous), and which appear to act on specific areas of the R-complex, limbic system and neocortex. If present trends continue, even without the encouragement of governments people will pursue the home-laboratory synthesis of and self-experimentation with such drugs—an activity that represents a small further step in our knowledge of the brain, its disorders and untapped potentials.

  There is reason to think that many alkaloids and other drugs which affect behavior work by being chemically similar to natural small brain proteins, of which the endorphins are one example. Many of these small proteins act on the limbic system and are concerned with our emotional states. It is now possible to manufacture small proteins made of any specified sequence of amino acids. Thus, the time may soon come when a great variety of molecules will be synthesized capable of inducing human emotional states, including extremely rare ones. For example, there is some evidence that atropine—one of the chief active ingredients in hemlock, foxglove, deadly nightshade, and jimson weed—induces the illusion of flying; and indeed such plants seem to have been the principal constituents of unguents self-administered to the genital mucosa by witches in the Middle Ages—who, rather than actually flying as they boasted, were in fact atropine-tripping. But a vivid hallucination of flying is an extremely specific sensation to be conveyed by a relatively simple molecule. Perhaps there are a range of small proteins which will be synthesized and which will produce emotional states of a sort never before experienced by human beings. This is one of many potential near-term developments in brain chemistry which hold great promise both for good and for evil, depending on the wisdom of those who conduct, control and apply this research.

  When I leave my office and get into my car, I find that, unless I make a specific effort of will, I will drive myself home. When I leave home and get into my car, unless I make a similar conscious effort, there is a part of my brain that arranges events so that I end up at my office. If I change my home or my office, after a short period of learning, the new locales supplant the old ones, and whatever brain mechanism controls such behavior has readily adapted to the new coordinates. This is very much like self-programming a part of the brain that works like a digital computer. The comparison is even more striking when we realize that epileptics, suffering from a psychomotor seizure, often go through an exactly comparable set of activities, the only difference being perhaps that they run a few more red lights than I usually do, but have no conscious memory of having performed these actions once the seizure has subsided. Such automatism is a typical symptom of temporal-lobe epilepsy; it also characterizes my first half-hour after awakening. Certainly not all of the brain works like a simple digital computer; the part that does the reprogramming, for example, is rather different. But there are enough similarities to suggest that a compatible working arrangement between electronic computers and at least some components of the brain—in an intimate neurophysiological association—can be constructively organized.

  The Spanish neurophysiologist José Delgado has devised working feedback loops between electrodes implanted in the brains of chimpanzees and remote electronic computers. Communication between brain and computer is accomplished through a radio link. Miniaturization of electronic computers has now reached the stage where such feedback loops can be “hardwired” and do not require a radio link with a remote computer terminal. For example, it is entirely possible to devise a self-contained feedback loop in which the signs of an on-coming epileptic seizure are recognized and appropriate brain centers are automatically stimulated to forestall or ameliorate the attack. We are not yet at the stage where this is a reliable procedure, but the time when it will be does not seem very far off.

  Perhaps some day it will be possible to add a variety of cognitive and intellectual prosthetic devices to the brain—a kind of eyeglasses for the mind. This would be in the spirit of the past accretionary evolution of the brain and is probably far more feasible than attempting to restructure the existing brain. Perhaps one day we will have surgically implanted in our brains small replaceable computer modules or radio terminals which will provide us with a rapid and fluent knowledge of Basque, Urdu, Amharic, Ainu, Albanian, Nu, Hopi, !Kung, or delphinese; or numerical values of the incomplete gamma
function and the Tschebysheff polynomials; or the natural history of animal spoor; or all legal precedents for the ownership of floating islands; or radio telepathy connecting several human beings, at least temporarily, in a form of symbiotic association previously unknown to our species.

  But the real extensions of our brains, particularly for the uniquely human aspects of the neocortex, are already in the course of being accomplished. Some of them are so old we have forgotten that they have occurred. Rich and unrepressive learning environments for children represent a remarkably promising and successful educational tool. Written language is a notable invention that is essentially a simple machine for the storage and retrieval of quite complex information. The amount of information stored in a large library far exceeds the amount of information in either the human genome or the human brain. The information is certainly not stored as efficiently as it is in biological systems, but it is still serviceably compact, and the development of microfilm, microfiche and the like has greatly improved the extrasomatic information storage capabilities of mankind. The number of bits of information contained in human libraries, works of art, and other cultural institutions would provide a point in the chart on this page many factors of 10 beyond the right-hand edge of that figure.

  Writing is, however, very simple machinery. Much more sophisticated and far more promising extrasomatic information storage and processing lies in the electronic computer. To give some idea of the level of development of such computers and their programs, they can today play perfect tic-tac-toe, world-class checkers, expert chess and practice passable psychotherapy. A computer system now exists which will transcribe a live performance of a classical or other melody into conventional musical notation. The amount of information storage and the speed of information processing by computers are, of course, prodigious.