The Dragons of Eden

by Carl Sagan

  When we examine an object to our right, both eyes are viewing what is called the right visual field; and to our left, the left visual field. But because of the way the optic nerves are connected, the right visual field is processed in the left hemisphere and the left visual field in the right hemisphere. Likewise, sounds from the right ear are processed primarily in the left hemisphere of the brain and vice versa, although there is some audio processing on the same side—for example, sounds from the left ear in the left hemisphere. No such crossing of function occurs in the more primitive sense of smell, and an odor detected by the left nostril only is processed exclusively in the left hemisphere. But information sent between the brain and the limbs is crossed. Objects felt by the left hand are perceived primarily in the right hemisphere, and instructions to the right hand to write a sentence are processed in the left hemisphere. (See the figure on this page.) In 90 percent of human subjects, the centers for speech are in the left hemisphere.

  A schematic representation, after Sperry, of the mapping of the outside world onto the two hemispheres of the neocortex. The right and left visual fields are projected, respectively, onto the left and right occipital lobes. Control of the right and left sides of the body is similarly crossed, as is, mainly, hearing. Smells are projected onto the hemispheres on the same side as the nostril doing the smelling.

  Sperry and his collaborators have performed an elegant series of experiments in which separate stimuli are presented to the left and right hemispheres of split-brain patients. In a typical experiment, the word hat-band is flashed on a screen—but hat is in the left visual field and band in the right visual field. The patient reports that he saw the word band, and it is clear that, at least in terms of his ability to communicate verbally, he has no idea that the right hemisphere received a visual impression of the word hat. When asked what kind of band it was, the patient might guess: outlaw band, rubber band, jazz band. But when, in comparable experiments, the patient is asked to write what he saw, but with his left hand inside a box, he scrawls the word hat. He knows from the motion of his hand that he has written something, but because he cannot see it, there is no way for the information to arrive in the left hemisphere which controls verbal ability. Bewilderingly, he can write, but cannot utter, the answer.

  Many other experiments exhibit similar results. In one, the patient is able to feel three-dimensional plastic letters which are out of view with his left hand. The available letters can spell only one correct English word, such as love or cup, which the patient is able to work out: the right hemisphere has a weak verbal ability, roughly comparable to that in dreams. But after correctly spelling the word, the patient is unable to give any verbal indication of what word he has spelled. It seems clear that in split-brain patients, each hemisphere has scarcely the faintest idea what the other hemisphere has learned.

  The geometrical incompetence of the left hemisphere is impressive; it is depicted by the illustration on the opposite page: A right-handed split-brain patient was able to copy simple representations of three-dimensional figures accurately only with his (inexperienced) left hand. The right hemisphere’s superiority at geometry seems restricted to manipulative tasks; this dominance does not hold for other sorts of geometrical functions that do not require hand-eye-brain coordination. These manipulative geometrical activities seem to be localized in the right hemisphere’s parietal lobe, in a place that, in the left hemisphere, is devoted to language. M. S. Gazzaniga of the State University of New York at Stony Brook suggests that this hemispheric specialization occurs because language is developed in the left hemisphere before the child acquires substantial competence in manipulative skills and geometrical visualization. According to this view, the specialization of the right hemisphere for geometrical competence is a specialization by default—the left hemisphere’s competence has been redirected toward language.

  The subject reads and verbally reports only the word flashed to his right visual field. No association is made, even unconsciously, of the words in left and right visual fields. After Sperry.

  A split-brain patient presented with a word in his left visual field correctly writes (and in script rather than capital letters) the word with the hand out of view. But when the subject is asked what his left hand wrote, he gives a totally incorrect response (“cup”). After Nebes and Sperry.

  Shortly after one of Sperry’s most convincing experiments had been completed, he gave a party, so the story goes, to which a famous theoretical physicist with an intact corpus callosum was invited. The physicist, known for his lively sense of humor, sat quietly through the party, listening with interest to Sperry’s description of the split-brain findings. The evening passed, the guests trickled away, and Sperry found himself at the door bidding goodbye to the last of them. The physicist extended his right hand, shook Sperry’s and told him what a fascinating evening he had had. Then, with a little two-step, he changed the positions of his right and left feet, extended his left hand, and said in a strangled, high-pitched voice, “And I want you to know I had a terrific time too.”

  Relative incompetence of the left hemisphere in copying geometrical figures. After Gazzaniga.

  When communication between the two cerebral hemispheres is impaired, the patient often finds his own behavior inexplicable, and it is clear that even in “good speaking” the speaker may not know “the truth of the matter.” (Compare with the remark on this page, from the Phaedrus.) The relative independence of the two hemispheres is apparent in everyday life. We have already mentioned the difficulty of describing verbally the complex perceptions of the right hemisphere. Many elaborate physical tasks, including athletics, seem to have relatively little left-hemisphere involvement. A well-known “ploy” in tennis, for example, is to ask your opponent exactly where on the racket he places his thumb. It often happens that left-hemisphere attention to this question will, at least for a brief period, destroy his game. A great deal of musical ability is a right-hemisphere function. It is a commonplace that we may memorize a song or a piece of music without having the least ability to write it down in musical notation. In piano, we might describe this by saying that our fingers (but not we) have memorized the piece.

  Such memorization can be quite complex. I recently had the pleasure of witnessing the rehearsal of a new piano concerto by a major symphony orchestra. In such rehearsals the conductor does not often start from the beginning and run through to the end. Rather, because of the expense of rehearsal time as well as the competence of the performers, he concentrates on the difficult passages. I was impressed that not only had the soloist memorized the entire piece, she was also able to begin at any requested place in the composition after only a brief glance at the designated measure in the score. This enviable skill is a mixed left and right hemisphere function. It is remarkably difficult to memorize a piece of music you have never heard so that you are able to intervene in any measure. In computer terminology, the pianist had random access as opposed to serial access to the composition.

  This is a good example of the cooperation between left and right hemispheres in many of the most difficult and highly valued human activities. It is vital not to overestimate the separation of functions on either side of the corpus callosum in a normal human being. The existence of so complex a cabling system as the corpus callosum must mean, it is important to stress again, that interaction of the hemispheres is a vital human function.

  In addition to the corpus callosum there is another neural cabling between the left and right hemispheres, which is called the anterior commissure. It is much smaller than the corpus callosum (see figure on this page), and exists, as the corpus callosum does not, in the brain of the fish. In human split-brain experiments in which the corpus callosum is cut, but not the anterior commissure, olfactory information is invariably transferred between the hemispheres. Occasional transfer of some visual and auditory information through the anterior commissure also seems to occur, but unpredictably from patient to patient. These findings are consist
ent with anatomy and evolution; the anterior commissure (and the hippocampal commissure; see the figure on this page) lies deeper than the corpus callosum and transfers information in the limbic cortex and perhaps in other more ancient components of the brain.

  Humans exhibit an interesting separation of musical and verbal skills. Patients with lesions of the right temporal lobe or right hemispherectomies are significantly impaired in musical but not in verbal ability—in particular in the recognition and recall of melodies. But their ability to read music is unimpaired. This seems perfectly consistent with the separation of functions described: the memorization and appreciation of music involves the recognition of auditory patterns and a holistic rather than analytic temperament. There is some evidence that poetry is partly a right-hemisphere function; in some cases the patient begins to write poetry for the first time in his life after a lesion in the left hemisphere has left him aphasic. But this would perhaps be, in Dryden’s words, “mere poetry.” Also, the right hemisphere is apparently unable to rhyme.

  The separation or lateralization of cortical function was discovered by experiments on brain-damaged individuals. It is, however, important to demonstrate that the conclusions apply to normal humans. Experiments carried out by Gazzaniga present brain-undamaged individuals with a word half in the left and half in the right visual fields, as in split-brain patients, and the reconstruction of the word is monitored. The results indicate that, in the normal brain, the right hemisphere does very little processing of language but instead transmits what it has observed across the corpus callosum to the left hemisphere, where the entire word is put together. Gazzaniga also found a split-brain patient whose right hemisphere was astonishingly competent in language skills: but this patient had experienced a brain pathology in the temporal-parietal region of the left hemisphere at an early age. We have already mentioned the ability of the brain to relocalize functions after injury in the first two years of life, but not thereafter.

  Robert Ornstein and David Galin of the Langley Porter Neuropsychiatric Institute in San Francisco claim that as normal people change from analytic to synthetic intellectual activities the EEG activity of the corresponding cerebral hemispheres varies in the predicted way: when a subject is performing mental arithmetic, for example, the right hemisphere exhibits the alpha rhythm characteristic of an “idling” cerebral hemisphere. If this result is confirmed, it would be quite an important finding.

  Omstein offers an interesting analogy to explain why, in the West at least, we have made so much contact with left-hemisphere functions and so little with right. He suggests that our awareness of right hemisphere function is a little like our ability to see stars in the daytime. The sun is so bright that the stars are invisible, despite the fact that they are just as present in our sky in the daytime as at night. When the sun sets, we are able to perceive the stars. In the same way, the brilliance of our most recent evolutionary accretion, the verbal abilities of the left hemisphere, obscures our awareness of the functions of the intuitive right hemisphere, which in our ancestors must have been the principal means of perceiving the world.*

  The left hemisphere processes information sequentially; the right hemisphere simultaneously, accessing several inputs at once. The left hemisphere works in series; the right in parallel. The left hemisphere is something like a digital computer; the right like an analog computer. Sperry suggested that the separation of function in the two hemispheres is the consequence of a “basic incompatibility.” Perhaps we are today able to sense directly the operations of the right hemisphere mainly when the left hemisphere has “set”—that is, in dreams.

  In the previous chapter, I proposed that a major aspect of the dream state might be the unleashing, at night, of R-complex processes that had been largely repressed by the neocortex during the day. But I mentioned that the important symbolic content of dreams showed significant neocortical involvement, although the frequently reported impairments in reading, writing, arithmetic and verbal recall suffered in dreams were striking.

  In addition to the symbolic content of dreams, other aspects of dream imagery point to a neocortical presence in the dream process. For example, I have many times experienced dreams in which the dénouement or critical “plot surprise” was possible only because of clues—apparently unimportant—inserted much earlier into the dream content. The entire plot development of the dream must have been in my mind at the time the dream began. (Incidentally, the time taken for dream events has been shown by Dement to be approximately equal to the time the same events would have taken in real life.) While the content of many dreams seems haphazard, others are remarkably well structured; these dreams have a remarkable resemblance to drama.

  We now recognize the very attractive possibility that the left hemisphere of the neocortex is suppressed in the dream state, while the right hemisphere—which has an extensive familiarity with signs but only a haltting verbal literacy—is functioning well. It may be that the left hemisphere is not entirely turned off at night but instead is performing tasks that make it inaccessible to consciousness: it is busily engaged in data dumping from the short-term memory buffer, determining what should survive into long-term storage.

  There are occasional but reliably reported instances of difficult intellectual problems solved during sleep. Perhaps the most famous is the dream of the German chemist Friedrich Kekulé von Stradonitz. In 1865 the most pressing and puzzling problem in organic structural chemistry was the nature of the benzene molecule. The structure of several simple organic molecules had been deduced from their properties, and all were linear, the constituent atoms being attached to each other in a straight line. According to his own account, Kekulé was dozing on a horse-drawn tram when he had a kind of dream of dancing atoms in linear arrangements. Abruptly the tail of a chain of atoms attached itself to the head and formed a slowly rotating ring. On awakening and recalling this dream fragment, Kekulé realized instantly that the solution to the benzene problem was a hexagonal ring of carbon atoms rather than a straight chain. Observe, however, that this is quintessentially a pattern-recognition exercise and not an analytic activity. It is typical of almost all of the famous creative acts accomplished in the dream state: they are right-hemisphere and not left-hemisphere activities.

  The American psychoanalyst Erich Fromm has written: “Must we not expect that, when deprived of the outside world, we regress temporarily to a primitive animal-like unreasonable state of mind? Much can be said in favor of such an assumption, and the view that such a regression is the essential feature of the state of sleep, and thus of dream activity, has been held by many students of dreaming from Plato to Freud.” Fromm goes on to point out that we sometimes achieve in the dream state insights that have evaded us when awake. But I believe these insights always have an intuitive or pattern-recognition character. The “animal-like” aspect of the dream state can be understood as the activities of the R-complex and the limbic system, and the occasionally blazing intuitive insight as the activity of the right hemisphere of the neocortex. Both cases occur because in each the repressive functions of the left hemisphere are largely turned off. These right-hemisphere insights Fromm calls “the forgotten language”—and he plausibly argues that they are the common origin of dreams, fairy tales and myths.

  In dreams we are sometimes aware that a small portion of us is placidly watching; often, off in a corner of the dream, there is a kind of observer. It is this “watcher” part of our minds that occasionally—sometimes in the midst of a nightmare—will say to us, “This is only a dream.” It is the “watcher” who appreciates the dramatic unity of a finely structured dream plot. Most of the time, however, the “watcher” is entirely silent. In psychedelic drug experiences—for example, with marijuana or LSD—the presence of such a “watcher” is commonly reported. LSD experiences may be terrifying in the extreme, and several people have told me that the difference between sanity and insanity in the LSD experience rests entirely on the continued presence of the “watcher,�
�� a small, silent portion of the waking consciousness.

  In one marijuana experience, my informant became aware of the presence and, in a strange way, the in-appropriateness of this silent “watcher,” who responds with interest and occasional critical comment to the kaleidoscopic dream imagery of the marijuana experience but is not part of it. “Who are you?” my informant silently asked it. “Who wants to know?” it replied, making the experience very like a Sufi or Zen parable. But my informant’s question is a deep one. I would suggest the observer is a small part of the critical faculties of the left hemisphere, functioning much more in psychedelic than in dream experiences, but present to a degree in both. However, the ancient query, “Who is it who asks the question?” is still unanswered; perhaps it is another component of the left cerebral hemisphere.

  An asymmetry in the temporal lobes in left and right hemispheres of humans and of chimpanzees has been found, with one portion of the left lobe significantly more developed. Human infants are born with this asymmetry (which develops as early as the twenty-ninth week of gestation), thus suggesting a strong genetic predisposition to control speech in the left temporal lobe. (Nevertheless, children with lesions in the left temporal lobe are able, in their first year or two of life, to develop all speech functions in the comparable portion of the right hemisphere with no impairment. At a later age, this replacement is impossible.) Also, lateralization is found in the behavior of young children. They are better able to understand verbal material with the right ear and nonverbal material with the left, a regularity also found in adults. Similarly, infants spend more time on the average looking at objects on their right than at identical objects on their left, and require a louder noise in the left ear than in the right to elicit a response. While no clear asymmetry of these sorts has yet been found in the brains or behavior of apes, Dewson’s results (see this page) suggest that some lateralization may exist in the higher primates; there is no evidence for anatomical asymmetries in the temporal lobes of, say, rhesus monkeys. One would certainly guess that the linguistic abilities of chimpanzees are governed, as in humans, in the left temporal lobe.