The Dragons of Eden: Speculations on the Evolution of Human Intelligence

Carl Sagan wrote this work of popular science as an expansion of the first Jacob Bronowski Memorial Lecture, delivered in 1975 at the University of Toronto. Bronowski, whose interdisciplinary approach to human learning provides a model for the book, is honored throughout. Sagan draws on neuroscience, evolutionary biology, paleontology, computer science, and myth to construct what he freely acknowledges is a speculative synthesis, crossing disciplinary boundaries to understand how human intelligence evolved, what its future might hold, and what it might reveal about the possibility of intelligence elsewhere in the universe.

Sagan begins by establishing a key distinction between genetic information, encoded in DNA and changing over immense time scales, and extragenetic information, which organisms acquire during their lifetimes through learning. Mammals, and humans in particular, depend far more on learned behavior than on prewired genetic instructions. Humans have also invented what Sagan calls extrasomatic knowledge: information stored outside the body, of which writing is the most important example. Because evolutionary change is far too slow to address the rapid transformations humanity now faces, Sagan argues that these learning systems are the only viable means of adaptation. He states a fundamental premise: The workings of the brain are entirely a consequence of its anatomy and physiology. He rejects mind-body dualism and commits to a materialist account of intelligence.

To convey the immensity of cosmic time, Sagan proposes a thought experiment he calls the Cosmic Calendar, compressing the roughly 15-billion-year history of the universe since the Big Bang into a single calendar year. On this scale, the Big Bang occurs on January 1 and life originates by approximately September 25. December chronicles the rapid proliferation of complex life: first worms on December 16, first vertebrates on December 19, first mammals on December 26. Humans appear at approximately 10:30 P.M. on New Year's Eve, and all of recorded history occupies the last 10 seconds of December 31. The exercise frames the book's central question: Given how recently intelligence appeared, what biological processes produced it, and what does its future hold?

Sagan then examines the information content of both genes and brains. He calculates that the DNA in a single human chromosome contains roughly 20 billion bits of information, equivalent to about 4,000 printed volumes. He argues that there is a practical upper limit to how much genetic information an organism can carry, because larger genomes produce unacceptably high mutation rates. Complex organisms must therefore store additional information in their brains. He estimates the information content of the human brain at roughly 10 to 100 trillion bits and notes that the number of possible brain states is so large that no two humans can ever be truly alike. Sagan identifies a critical crossover point during the Carboniferous Period, several hundred million years ago, when an early reptile became the first organism with more information in its brain than in its genes. After this threshold, brain evolution increasingly dominated genetic evolution.

The book's central structural idea is the triune brain model, MacLean's term for the three-layered evolutionary architecture of the human forebrain, developed by Paul MacLean, chief of the Laboratory of Brain Evolution and Behavior at the National Institute of Mental Health. MacLean proposes that the human forebrain contains three semi-independent layers, each corresponding to a major evolutionary step. The oldest is the R-complex, or reptilian complex, shared with reptiles and other mammals, which MacLean associates with aggressive behavior, territoriality, ritual, and social hierarchies. Surrounding it is the limbic system, shared with other mammals, which generates strong emotions and contains structures critical for aggression, fear, and memory formation. Sagan argues that love and sustained parental care are essentially mammalian inventions linked to the limbic system. The most recent layer is the neocortex, responsible for characteristically human cognitive functions: language, abstract reasoning, anticipation of the future, and spatial perception. Sagan divides the neocortex into four lobes, each associated with different capacities. Frontal lobe damage impairs a patient's ability to anticipate future events. Temporal lobe lesions can impair speech or facial recognition. Parietal lobe damage causes the greatest overall decline in intelligence, while the occipital lobes process vision. Sagan compares this layered architecture to Plato's metaphor in the Phaedrus of a charioteer barely controlling two unruly horses: The neocortex struggles to maintain control while the R-complex and limbic system pull in their own directions.

Sagan traces the evolution of human ancestors through the fossil record. He describes the gracile Australopithecines, bipedal creatures from approximately 5 to 6 million years ago with brain volumes of 430 to 600 cubic centimeters and associated stone tool industries. He contrasts them with the robust Australopithecines, who had larger bodies and powerful teeth but no associated tools. Homo habilis, the first true human, appeared approximately 3.7 million years ago with brain volumes of 500 to 800 cc, fully bipedal posture, and sophisticated tools. Homo erectus followed, associated with paleoanthropologist Richard Leakey's 1976 discovery of a nearly complete skull 1.5 million years old and with the domestication of fire. Sagan connects the rapid growth of the human skull with the evolutionary reshaping of the pelvis and argues that human childbirth is uniquely painful because skull evolution has been spectacularly fast. He reads the Book of Genesis as a metaphor for these developments: The acquisition of knowledge of good and evil corresponds to the growth of the neocortex, the pain of childbirth reflects cranial expansion, and the exile from Eden parallels the transition from integration with nature to civilization and agriculture.

A central chapter examines whether abstract thought is uniquely human. Sagan describes the breakthrough of Beatrice and Robert Gardner, two psychologists at the University of Nevada, who recognized that chimpanzees' vocal anatomy is unsuited for speech and instead taught a chimpanzee named Washoe American Sign Language (Ameslan). Chimps showed remarkable creativity: Washoe invented "waterbird" upon first seeing a duck; Lucy, another chimp, described a watermelon as "candy drink" and a radish as "cry hurt food"; and Lana learned a computer language called Yerkish and once told her trainer to leave the room when he interfered with her typing. Sagan raises the ethical implications of these findings, asking why intelligent beings who have committed no crimes are kept in cages. He speculates that humans may have systematically exterminated nonhuman primates who displayed signs of intelligence, and that teaching language to chimps represents a belated attempt to make amends.

Sagan devotes a chapter to sleep and dreams, proposing the "vampiric hypothesis"—his term for the idea that sleep evolved as a survival strategy for early mammals hiding from predatory reptiles. Early mammals evolved during an epoch dominated by reptiles. Because reptiles are cold-blooded and largely immobilized at night, early mammals may have been active nocturnally and slept during the day to avoid predatory dinosaurs. Sleep, in this view, evolved not as a restorative process but as a survival mechanism. Sagan proposes that dreams represent the nighttime unleashing of R-complex impulses normally suppressed by the neocortex during waking hours, citing the correlation of dream content with aggressive, sexual, and hierarchical material as supporting evidence.

The book's longest analytical chapter examines the lateralization of the human brain. Sagan describes Roger Sperry's split-brain experiments at the California Institute of Technology, in which the corpus callosum, the bundle of nerve fibers connecting the two cerebral hemispheres, was severed to control severe epilepsy. These experiments revealed that the left hemisphere controls speech, reading, writing, and arithmetic, while the right hemisphere excels at three-dimensional vision, pattern recognition, and musical ability. Sagan labels these two modes "rational" and "intuitive," and argues that all significant creative achievements require the collaboration of both hemispheres. Human culture, he contends, is metaphorically a function of the corpus callosum.

In his final chapters, Sagan considers the future of the brain. He discusses treatments for mental illness, the potential for cognitive prosthetic devices, and the rapid development of electronic computers as extensions of human intelligence. Addressing the objection that computers preempt learning, he invokes Plato's Phaedrus, in which the god Thoth's invention of writing was criticized for creating forgetfulness, yet writing proved an immense benefit. Sagan predicts that the next major advance will be a partnership between humans and intelligent machines. Returning to the search for extraterrestrial intelligence, he argues that natural selection everywhere favors intelligence and that interstellar communication should be possible despite radical biological differences. He warns against the popularity of pseudoscientific doctrines and the retreat from rational thinking, echoing Bronowski's declaration that knowledge is humanity's destiny. The coordinated functioning of both cerebral hemispheres, Sagan concludes, is the tool nature has provided for survival.