The Dragons of Eden

by Carl Sagan

  We are now in a position to compare the gradual increase through evolutionary time of both the amount of information contained in the genetic material and the amount of information contained in the brains of organisms. The two curves cross (this page) at a time corresponding to a few hundred million years ago and at an information content corresponding to a few billion bits. Somewhere in the steaming jungles of the Carboniferous Period there emerged an organism that for the first time in the history of the world had more information in its brains than in its genes. It was an early reptile which, were we to come upon it in these sophisticated times, we would probably not describe as exceptionally intelligent. But its brain was a symbolic turning point in the history of life. The two subsequent bursts of brain evolution, accompanying the emergence of mammals and the advent of manlike primates, were still more important advances in the evolution of intelligence. Much of the history of life since the Carboniferous Period can be described as the gradual (and certainly incomplete) dominance of brains over genes.

  * To some extent the mutation rate is itself controlled by natural selection, as in our example of a “molecular scissors.” But there is likely to be an irreducible minimum mutation rate (1) in order to produce enough genetic experiments for natural selection to operate on, and (2) as an equilibrium between mutations produced, say, by cosmic rays and the most efficient possible cellular repair mechanisms.

  * Incidentally, as a test of the influence of animated cartoons on American life, try rereading this paragraph with the word “rat” replaced everywhere by “mouse,” and see if your sympathy for the surgically invaded and misunderstood beast suddenly increases.

  * By the criterion of brain mass to body mass, sharks are the smartest of the fishes, which is consistent with their ecological niche—predators have to be brighter than plankton browsers. Both in their increasing ratio of brain to body mass and in the development of coordinating centers in the three principal components of their brains, sharks have evolved in a manner curiously parallel to the evolution of higher vertebrates on the land.

  * Horizon to horizon comprises an angle of 180 degrees in a flat place. The moon is 0.5 degrees in diameter. I know I can see detail on it, perhaps twelve picture elements across. Thus my eye can resolve about 0.5/12 = 0.04 degrees. Anything smaller than this is too small for me to see. The instantaneous field of view in my mind’s eye, as well as in my real eye, seems to be something like 2 degrees on a side. Thus the little square picture I can see at any given moment contains about (2/0.04)2 = 2,500 picture elements, corresponding to the wirephoto dots. To characterize all possible shades of gray and colors of such dots requires about 20 bits per picture element. Thus a description of my little picture requires 2,500 × 20 or about 50,000 bits. But the act of scanning the picture takes about 10 seconds, and thus my sensory data processing rate is probably not much larger than 50,000/10 = 5,000 bits per second. For comparison, the Viking lander cameras, which also have a 0.04 degree resolution, have only six bits per picture element to characterize brightness, and can transmit these directly to Earth by radio at 500 bits per second. The neurons of the brain generate about 25 watts of power, barely enough to turn on a small incandescent light. The Viking lander transmits radio messages and performs all its other functions with a total power of about 50 watts.

  3

  THE

  BRAIN AND THE

  CHARIOT

  When shall we three meet again …?

  WM. SHAKESPEARE

  Macbeth

  HE BRAIN of a fish isn’t much. A fish has a notochord or spinal cord, which it shares with even humbler invertebrates. A primitive fish also has a little swelling at the front end of the spinal cord, which is its brain. In higher fish the swelling is further developed but still weighs no more than a gram or two. That swelling corresponds in higher animals to the hindbrain or brainstem and the midbrain. The brain of modern fish are chiefly midbrain, with a tiny forebrain; in modern amphibians and reptiles, it is the other way around (see figure on this page). And yet fossil endocasts of the earliest known vertebrates show that the principal divisions of the modern brain (hindbrain, midbrain and forebrain, for example) were already established. Five hundred million years ago, swimming in the primeval seas, there were fishy creatures called ostracoderms and placoderms, whose brains had recognizably the same major divisions as ours. But the relative size and importance of these components, and even their early functions, were certainly very different from today. One of the most engaging views of the subsequent evolution of the brain is a story of the successive accretion and specialization of three further layers surmounting the spinal cord, hindbrain and midbrain. After each evolutionary step, the older portions of the brain still exist and must still be accommodated. But a new layer with new functions has been added.

  The principal contemporary exponent of this view is Paul MacLean, chief of the Laboratory of Brain Evolution and Behavior of the National Institute of Mental Health. One hallmark of MacLean’s work is that it encompasses many different animals, ranging from lizards to squirrel monkeys. Another is that he and his colleagues have studied carefully the social and other behavior of these animals to improve their prospects of discovering what part of the brain controls what sort of behavior.

  Squirrel monkeys with “gothic” facial markings have a kind of ritual or display which they perform when greeting one another. The males bare their teeth, rattle the bars of their cage, utter a high-pitched squeak, which is possibly terrifying to squirrel monkeys, and lift their legs to exhibit an erect penis. While such behavior would border on impoliteness at many contemporary human social gatherings, it is a fairly elaborate act and serves to maintain dominance hierarchies in squirrel-monkey communities.

  MacLean has found that a lesion in one small part of a squirrel monkey’s brain will prevent this display while leaving a wide range of other behavior intact, including sexual and combative behavior. The part that is involved is in the oldest part of the forebrain, a part that humans as well as other primates share with out mammalian and reptilian ancestors. In non-primate mammals and in reptiles, comparable ritualized behavior seems to be controlled in the same part of the brain, and lesions in this reptilian component can impair other automatic types of behavior besides ritual—for example, walking or running.

  The connection between sexual display and position in a dominance hierarchy can be found frequently among the primates. Among Japanese macaques, social class is maintained and reinforced by daily mounting: males of lower caste adopt the characteristic submissive sexual posture of the female in oestrus and are briefly and ceremonially mounted by higher-caste males. These mountings are both common and perfunctory. They seem to have little sexual content but rather serve as easily understood symbols of who is who in a complex society.

  Schematic diagrams comparing the brain of a fish, an amphibian, a reptile, a bird, and a mammal. The cerebellum and medulla oblongata are parts of the hindbrain.

  In one study of the behavior of the squirrel monkey, Caspar, the dominant animal in the colony and by far the most active displayer, was never seen to copulate, although he accounted for two-thirds of the genital display in the colony—most of it directed toward other adult males. The fact that Caspar was highly motivated to establish dominance but insignificantly motivated toward sex suggests that while these two functions may involve identical organ systems, they are quite separate. The scientists studying this colony concluded: “Genital display is therefore considered the most effective social signal with respect to group hierarchy. It is ritualized and seems to acquire the meaning, ‘I am the master.’ It is most probably derived from sexual activity, but it is used for social communication and separated from reproductive activity. In other words, genital display is a ritual derived from sexual behavior but serving social and not reproductive purposes.”

  In a television interview in 1976, a professional football player was asked by the talk-show host if it was embarrassing for football players to be t
ogether in the locker room with no clothes on. His immediate response: “We strut! No embarrassment at all. It’s as if we’re saying to each other, ‘Let’s see what you got, man!’—except for a few, like the specialty team members and the water boy.”

  The behavioral as well as neuroanatomical connections between sex, aggression and dominance are borne out in a variety of studies. The mating rituals of great cats and many other animals are barely distinguishable, in their early stages, from fighting. It is a commonplace that domestic cats sometimes purr loudly and perversely while their claws are slowly raking over upholstery or lightly clad human skin. The use of sex to establish and maintain dominance is sometimes evident in human heterosexual and homosexual practices (although it is not, of course, the only element in such practices), as well as in many “obscene” utterances. Consider the peculiar circumstance that the most common two-word verbal aggression in English, and in many other languages, refers to an act of surpassing physical pleasure; the English form probably comes from a Germanic and Middle Dutch verb fokken, meaning “to strike.” This otherwise puzzling usage can be understood as a verbal equivalent of macaque symbolic language, with the initial word “I” unstated but understood by both parties. It and many similar expressions seem to be human ceremonial mountings. As we will see, such behavior probably extends much farther back than the monkeys, back through hundreds of millions of years of geological time.

  From experiments such as those with squirrel monkeys, MacLean has developed a captivating model of brain structure and evolution that he calls the triune brain. “We are obliged,” he says, “to look at ourselves and the world through the eyes of three quite different mentalities,” two of which lack the power of speech. The human brain, MacLean holds, “amounts to three interconnected biological computers,” each with “its own special intelligence, its own subjectivity, its own sense of time and space, its own memory, motor, and other functions.” Each brain corresponds to a separate major evolutionary step. The three brains are said to be distinguished neuroanatomically and functionally, and contain strikingly different distributions of the neurochemicals dopamine and Cholinesterase.

  At the most ancient part of the human brain lies the spinal cord; the medulla and pons, which comprise the hindbrain; and the midbrain. This combination of spinal cord, hindbrain and midbrain MacLean calls the neural chassis. It contains the basic neural machinery for reproduction and self-preservation, including regulation of the heart, blood circulation and respiration. In a fish or an amphibian it is almost all the brain there is. But a reptile or higher animal deprived of its forebrain is, according to MacLean, “as motionless and aimless as an idling vehicle without a driver.”

  Indeed, grand mal epilepsy can, I think, be described as a disease in which the cognitive drivers are all turned off because of a kind of electrical storm in the brain, and the victim is left momentarily with nothing operative but his neural chassis. This is a profound impairment, temporarily regressing the victim back several hundreds of millions of years. The ancient Greeks, whose name for the disease we still use, recognized its profound character and called it the disease inflicted by the gods.

  MacLean has distinguished three sorts of drivers of the neural chassis. The most ancient of them surrounds the midbrain (and is made up mostly of what neuroanatomists call the olfactostriatum, the corpus striatum, and the globus pallidus). We share it with the other mammals and the reptiles. It probably evolved several hundred million years ago. MacLean calls it the reptilian or R-complex. Surrounding the R-complex is the limbic system, so called because it borders on the underlying brain. (Our arms and legs are called limbs because they are peripheral to the rest of the body.) We share the limbic system with the other mammals but not, in its full elaboration, with the reptiles. It probably evolved more than one hundred and fifty million years ago. Finally, surmounting the rest of the brain, and clearly the most recent evolutionary accretion, is the neocortex. Like the higher mammals and the other primates, humans have a relatively massive neocortex. It becomes progressively more developed in the more advanced mammals. The most elaborately developed neocortex is ours (and the dolphins’ and whales’). It probably evolved several tens of millions of years ago, but its development accelerated greatly a few million years ago when humans emerged. A schematic representation of this picture of the human brain is shown opposite, and a comparison of the limbic system with the neocortex in three contemporary mammals is shown above. The concept of the triune brain is in remarkable accord with the conclusions, drawn independently from studies of brain to body mass ratios in the previous chapter, that the emergence of mammals and of primates (especially humans) was accompanied by major bursts in brain evolution.

  It is very difficult to evolve by altering the deep fabric of life; any change there is likely to be lethal. But fundamental change can be accomplished by the addition of new systems on top of old ones. This is reminiscent of a doctrine which was called recapitulation by Ernst Haeckel, a nineteenth-century German anatomist, and which has gone through various cycles of scholarly acceptance and rejection. Haeckel held that in its embryological development, an animal tends to repeat or recapitulate the sequence that its ancestors followed during their evolution. And indeed in human intrauterine development we run through stages very much like fish, reptiles and nonprimate mammals before we become recognizably human. The fish stage even has gill slits, which are absolutely useless for the embryo who is nourished via the umbilical cord, but a necessity for human embryology: since gills were vital to our ancestors, we run through a gill stage in becoming human. The brain of a human fetus also develops from the inside out, and, roughly speaking, runs through the sequence: neural chassis, R-complex, limbic system and neocortex (see the figure on the embryology of the human brain on this page).

  A highly schematic representation of the reptilian complex, limbic system, and neocortex in the human brain, after MacLean.

  The reason for recapitulation may be understood as follows: Natural selection operates only on individuals, not on species and not very much on eggs or fetuses. Thus the latest evolutionary change appears postpartum. The fetus may have characteristics, like the gill slits in mammals, that are entirely maladaptive after birth, but as long as they cause no serious problems for the fetus and are lost before birth, they can be retained. Our gill slits are vestiges not of ancient fish but of ancient fish embryos. Many new organ systems develop not by the addition and preservation but by the modification of older systems, as, for example, the modification of fins to legs, and legs to flippers or wings; or feet to hands to feet; or sebaceous glands to mammary glands; or gill arches to ear bones; or shark scales to shark teeth. Thus evolution by addition and the functional preservation of the preexisting structure must occur for one of two reasons—either the old function is required as well as the new one, or there is no way of bypassing the old system that is consistent with survival.

  Schematic views from the top and from the side of the rabbit, cat, and monkey brains. The dark stippled area is the limbic system, seen most easily in the side views. The white furrowed regions represent the neocortex, visible most readily in the top views.

  A photograph taken with an electron microscope of a small plant called a red alga. Its scientific name is Porphyridium cruentum. The chloroplast, this organism’s photosynthetic factory, almost fills the entire cell. The photograph is magnified 23,000 times and was taken by Dr. Elizabeth Gantt of the Smithsonian Institution’s Radiation Biology Laboratory.

  There are many other examples in nature of this sort of evolutionary development. To take an almost random case, consider why plants are green. Green-plant photosynthesis utilizes light in the red and the violet parts of the solar spectrum to break down water, build up carbohydrates and do other planty things. But the sun gives off more light in the yellow and the green part of the spectrum than in the red or violet. Plants with chlorophyll as their only photosynthetic pigment are rejecting light where it is most plentiful. Many plants
seem belatedly to have “noticed” this and have made appropriate adaptations. Other pigments, which reflect red light and absorb yellow and green light, such as carotenoids and phycobilins, have evolved. Well and good. But have those plants with new photo-synthetic pigments abandoned chlorophyll? They have not. The figure on this page shows the photosynthetic factory of a red alga. The striations contain the chlorophyll, and the little spheres nestling against these striations contain the phycobilins, which make a red alga red. Conservatively, these plants pass along the energy they acquire from green and yellow sunlight to the chlorophyll pigment that, even though it has not absorbed the light, is still instrumental in bridging the gap between light and chemistry in all plant photosynthesis. Nature could not rip out the chlorophyll and replace it with better pigments; the chlorophyll is woven too deeply into the fabric of life. Plants with accessory pigments are surely different. They are more efficient. But there, still working, although with diminished responsibilities, at the core of the photosynthetic process is chlorophyll. The evolution of the brain has, I think, proceeded analogously. The deep and ancient parts are functioning still.

  1 THE R-COMPLEX

  If the preceding view is correct, we should expect the R-complex in the human brain to be in some sense performing dinosaur functions still; and the limbic cortex to be thinking the thoughts of pumas and ground sloths. Without a doubt, each new step in brain evolution is accompanied by changes in the physiology of the preexisting components of the brain. The evolution of the R-complex must have seen changes in the midbrain, and so ort. What is more, we know that the control of many functions is shared in different components of the brain. But at the same time it would be astonishing if the brain components beneath the neocortex were not to a significant extent still performing as they did in our remote ancestors.