The Mind's Eye

by Oliver Sacks

  So whether we are examining patients with alexia, patients with lexical hallucinations, or normal subjects reading, in any language, we are forced to the same conclusion: that there exists, in every literate human being, an area in the dominant hemisphere—the language hemisphere—a neuronal system potentially available for the recognition of letters and words (and perhaps other forms of visual notation—mathematical or musical, for example).

  This raises a deep problem: Why should all human beings have this built-in facility for reading, when writing is a relatively recent cultural invention?

  Communication by the spoken word—and, therefore, its neural basis—has every mark of having evolved through the gradual processes of natural selection. The changing anatomy of the brain in prehistoric man has been worked out in some detail from endocranial casts and other fossil evidence, as have changes in the vocal tract. It is clear that the beginnings of speech go back hundreds of thousands of years. But this cannot be maintained in regard to reading, for writing emerged little more than five thousand years ago—far too recently to have occurred through evolution by natural selection. Though the visual word form area of the human brain appears so exquisitely tuned to the act of reading, it could not have evolved specifically for this purpose.

  We might call this the Wallace problem, for Alfred Russel Wallace (who discovered natural selection independently of Darwin) became intensely concerned with the paradox of the human brain’s many potential abilities—lexical, mathematical, and so on—abilities that would be of little use in a primitive or prehistoric society. While natural selection could explain the appearance of immediately useful abilities, he felt, it could not explain the existence of potential powers that might become manifest only with the development of an advanced culture hundreds of thousands of years in the future.

  Unable to attribute these human potentials to any natural process, Wallace found himself constrained to invoke the supernatural: God, he believed, must have implanted them in the human psyche. There could hardly, from Wallace’s perspective, be a better example of a divine gift—a unique new power, biding its time, in posse, waiting for the rise of a sufficiently advanced culture.12

  Darwin, understandably, was horrified by this idea and wrote to Wallace, “I hope you have not murdered too completely your own and my child.” Darwin, for his part, had a much more open view of the process of natural selection and adaptation, foreseeing that biological structures might find uses very different from those for which they had originally evolved. (Stephen Jay Gould and Elisabeth Vrba called this sort of redeployment an “exaptation” rather than a direct adaptation.)13

  How, then, did the visual word form area of the human brain arise? Does it exist in the brains of illiterate people? Does it have a precursor in the brains of other primates?

  We are all faced with a world of sights and sounds and other stimuli, and our survival depends on making a rapid and accurate appraisal of these. Making sense of the world around us must be based on some sort of system, some swift and sure way of parsing the environment. Although seeing objects, defining them visually, seems to be instantaneous and innate, it represents a great perceptual achievement, one that requires a whole hierarchy of functions. We do not see objects as such; we see shapes, surfaces, contours, and boundaries, presenting themselves in different illumination or contexts, changing perspective with their movement or ours. From this complex, shifting visual chaos, we have to extract invariants that allow us to infer or hypothesize object-hood. It would be uneconomical to suppose that there are individual representations or engrams for each of the billions of objects around us. The power of combination must be called on; one needs a finite set or vocabulary of shapes that can be combined in an infinite number of ways, much as the twenty-six letters of the alphabet can be assembled (within certain rules and constraints) into as many words or sentences as a language ever needs.

  There may be some objects that are recognized at birth, or soon after, like faces. But beyond this, the world of objects must be learned through experience and activity: looking, touching, handling, correlating the feel of objects with their appearance. Visual object recognition depends on the millions of neurons in the inferotemporal cortex, and neuronal function here is very plastic, open and highly responsive to experience and training, to education. Inferotemporal neurons evolved for general visual recognition, but they may be recruited for other purposes—most notably reading.

  Such a redeployment of neurons is facilitated by the fact that all (natural) writing systems seem to share certain topological features with the environment, features which our brains evolved to decode. Mark Changizi, Shinsuke Shimojo, and their colleagues at Caltech examined more than a hundred ancient and modern writing systems, including alphabetic systems and Chinese ideograms, from a computational point of view. They have shown that all of them, while geometrically very different, share certain basic topological similarities. (This visual signature is not evident in artificial writing systems, such as shorthand, which are designed to emphasize speed more than visual recognition.) Changizi et al. have found similar topological invariants in a range of natural settings, and this has led them to hypothesize that the shapes of letters “have been selected to resemble the conglomerations of contours found in natural scenes, thereby tapping into our already-existing object recognition mechanisms.”

  Writing, a cultural tool, has evolved to make use of the inferotemporal neurons’ preference for certain shapes. “Letter shape,” Dehaene writes, “is not an arbitrary cultural choice. The brain constrains the design of an efficient writing system so severely that there is little room for cultural relativism. Our primate brain only accepts a limited set of written shapes.”14

  This is an elegant solution to the “Wallace problem”—indeed, it shows that there is no problem. The origin of writing and reading cannot be understood as a direct evolutionary adaptation. It is dependent on the plasticity of the brain, and the fact that even within the small span of a human lifetime, experience—experiential selection—is as powerful an agent of change as natural selection. Natural selection, for Darwin, did not forbid cultural and individual developments on a timescale hundreds of thousands of times faster than evolutionary development—on the contrary, it prepared the ground for them. We are literate not by virtue of a divine intervention, but through a cultural invention and a cultural selection that makes a brilliant and creative new use of a preexisting neural proclivity.

  While the visual word form area is crucial in the recognition of words and letters, many other areas of the brain are involved in “higher” levels of reading. This enabled Howard, for instance, to infer words from their context. Even now, nine years after his stroke, he is unable to recognize many simple words at a glance—but his writer’s imagination does not just depend on reading.

  While he was still in the rehab hospital, one of his therapists suggested that he keep a “memory book” to remind himself of appointments and to record his thoughts. As a lifelong keeper of journals, Howard was delighted by this idea. His new memory book proved to be an invaluable aid not only in stabilizing his still erratic memory but in reinforcing his identity as a writer:

  I knew I could no longer rely on the “sticking plaster” of memory. I could forget a word in the second part of what I was saying, even though I had already used the word a moment earlier.… I learned to write things down in the “memory book” [the moment I thought of them].… The memory book gave a lift to my sense of being in the driver’s seat of my life. [It] became my constant companion: part diary, part appointment book, part commonplace book. Hospitals, to a degree … breed a passive spirit; the memory book returned a piece of myself to me.

  Keeping the memory book invited him, forced him, to write every day—not only at the level of forming legible words and sentences but at a much deeper creative level. His journal of hospital life, with its various routines and characters, began to stir his writer’s imagination.

  Occasionally, wit
h unusual words or proper names, Howard might be unsure of their spelling—he could not “see” them in his mind’s eye, imagine them, any more than he could perceive them when they were printed before him. Lacking this internal imagery, he had to employ other strategies for spelling. The simplest of these, he found, was to write a word in the air with his finger, letting a motor act take the place of a sensory one.

  The great French neurologist Jean-Martin Charcot, in an 1883 lecture on a case of word blindness, describes a patient who, like Howard, has alexia sine agraphia. Charcot writes down the name of the hospital (which the patient himself has written earlier) and asks him to read it: “[The patient] is unable to do so at first; but he makes further efforts to do it and while he is accomplishing the task we notice that he traces, with the end of his right index finger, one of the letters which constitute the word, and with much trouble he says ‘La Salpêtrière.’ ” When Charcot gives him the name of a street to read, the patient “traces with his finger in space the letters which compose the word, and after a moment or two says, ‘It is the Rue d’Aboukir, the address of my friend.’ ”

  Charcot’s patient improved rapidly in “reading” by tracing letters in the air, and within three weeks, his reading speed had increased nearly sixfold. He said, “I can read printing less well than writing, because in writing it is easier for me to mentally reproduce the letter with my right hand, whereas it is more difficult to reproduce the printed characters.” (“When reading printed matter,” Charcot noted, “it is convenient for him to have a pen in his hand.”) Concluding his lecture, Charcot emphasized, “Briefly put, one can say of him that he reads only in the act of writing.”

  Increasingly and often unconsciously, then, Howard started to move his hands as he read, tracing the outlines of words and sentences still unintelligible to his eyes. And most remarkably, his tongue, too, began to move as he read, tracing the shapes of letters on his teeth or the roof of his mouth. This enabled him to read considerably faster (though it still might take him a month or more to read a book he could previously have read in an evening). Thus, by an extraordinary, metamodal, sensory-motor alchemy, Howard was replacing reading by a sort of writing. He was, in effect, reading with his tongue.15

  More than three months after his stroke, Howard returned from rehab to a home he did not entirely recognize:

  The house looked strange and familiar at the same time.… It was as though a movie set had been assembled from sketches of the real house and its rooms. Most peculiar was my office. I looked at my computer with a strange feeling. My whole office, where I had written several of my books, resembled a diorama in a museum.… On scribbled stick-on notes, my own handwriting looked strange, unfamiliar.

  Would he ever be able to use this alien computer—once the main tool of his trade—again? With his son’s help, and to his own surprise, he started to test out his old computer skills and soon felt them coming back. But writing something creative was another matter. And reading, even reading his own erratic handwriting, was still agonizingly slow and difficult. Furthermore, as he later wrote,

  I had been out of the world for months. I could no longer keep things straight in my head. What business did I have imagining that I might go back to my old desk and begin again? I was clearly unfit for fiction. I turned off the computer and took a long walk.

  Nonetheless, Howard had been, in a sense, staying in practice, writing every day, if only in his memory book. At first, he wrote,

  I had no thoughts of writing a book. That was not only well beyond my abilities, it was also beyond my imagination. But without my knowing it, another part of my brain was beginning to plot out a story. Images began popping into my head. Plots and plot twists began haunting my imagination. While I [had been] lying in my hospital bed … I was hard at work inventing story and characters and situations for the book I still didn’t know I was writing.

  He decided to write—if he could—a new novel, following his mother’s old advice:

  Write about what you know.… What I knew about now was my illness. I knew the hospital routines and the people around me. I could do a book that described what it was like to be out of things, flat on my back for a time with nurses and doctors ordering and reordering my days.

  He would reintroduce his alter ego, the detective Benny Cooperman, but it would be a Cooperman transformed: the great detective, waking in a hospital bed, finds himself not only alexic but amnesic as well. His powers of inference, however, are intact and enable him to stitch together disparate clues, to figure out how he landed in the hospital and what happened in the mysterious few days he can no longer remember.

  Howard moved into high gear, typing for hours each day on his computer. Within a few weeks, his imagination and creative flow enabled him to produce a first draft. The problem now was how to correct and revise the draft, given his problems with short-term memory and his inability to read in the normal way. He employed many devices using his word processor—indenting certain paragraphs, marking passages with different font sizes—and after he had done as much as he could by himself, he got his editor to read the entire book aloud to him, so that he could engrave its overall structure in his memory and reorganize it in his mind. This painstaking process took many months of hard labor, but his abilities to remember and revise mentally, like Lilian Kallir’s ability to arrange piano scores in her mind, steadily increased with practice.

  His new novel (which he called Memory Book) was published in 2005, and this was followed in fairly rapid succession by another Benny Cooperman novel and, in 2007, a memoir, The Man Who Forgot How to Read. Howard Engel is still alexic, but he has found a way to remain a man of letters. That he was able to do so is a testament to many things: the dedication and skill of his therapists in rehab, his own determination to read again, and the adaptability of the human brain.

  “The problems never went away,” Howard writes, “but I became cleverer at solving them.”

  1. This was published as a chapter in An Anthropologist on Mars.

  2. Lilian Kallir, too, had alexia sine agraphia, and continued to write letters to her friends around the world. But since her alexia for words developed slowly, over the course of years, she seemed to have insensibly accommodated to the fact that reading and writing could be dissociated in this way.

  3. The current term, “visual agnosia,” was introduced by Sigmund Freud the following year.

  4. Congenital “word blindness” (which we now call dyslexia) was recognized by neurologists in the 1880s, around the same time that Charcot, Déjerine, and others were describing acquired alexia. Children with severe difficulties in reading (sometimes in writing, reading music, or calculating, too) were often seen as retarded, despite clear evidence to the contrary. W. Pringle Morgan, writing in the British Medical Journal in 1896, detailed a careful study of an intelligent and articulate fourteen-year-old boy who had severe difficulties in reading and spelling:

  In writing his own name he made a mistake, putting “Precy” for “Percy,” and he did not notice the mistake until his attention was called to it more than once.… Words written or printed seem to convey no impression to his mind, and it is only after laboriously spelling them that he is able, by the sounds of the letters, to discover their import.… He can only recognize such simple ones as “and,” “the,” “of,” etc. Other words he never seems to remember, no matter how frequently he may have met them.… The schoolmaster who has taught him for some years says that he would be the smartest lad in the school if the instruction were entirely oral.

  It is now recognized that as much as five to ten percent of the population has dyslexia and that, whether by way of “compensation” or simply because of their different neurological makeup, many dyslexic people have exceptional talents in other areas. These and many other aspects of dyslexia are discussed in depth by Maryanne Wolf in Proust and the Squid: The Story and Science of the Reading Brain, and by Thomas G. West in In the Mind’s Eye.

  5. I am quoting here and
elsewhere from the translation provided by Israel Rosenfield in his excellent book The Invention of Memory.

  6. Israel Rosenfield also remarks that Oscar C.’s central problem was not just in recognizing letters but in perceiving their sequence, and that he had similar problems with numerical notation. Numbers, Rosenfield writes, “are always read the same way in every context. A 3 is three whether it appears in the phrase ‘3 apples’ or ‘a 3 percent discount.’ But … the meaning of a number in a multidigit numeral depends on where it is placed.” It is similar with musical notes, whose meaning depends on context and placement.

  Words, Rosenfield continues, are similar:

  Changing a single letter in a word can alter both its pronunciation and its meaning. Its significance depends on what precedes and what follows.… It is the failure to capture this overall organization—in which identical stimuli, letters, are constantly changed in significance—that is characteristic of patients with verbal blindness. They cannot organize the stimuli in a way that makes sense of the symbols.

  7. In the few days that he lived after his second stroke, Oscar C. had aphasia as well. He would say one word in place of another or make garbled sounds and had to rely on mime and gesture to communicate. His wife noticed (“with dread”) that he could no longer write. Israel Rosenfield, analyzing Déjerine’s case in The Invention of Memory, suggests that one may have alexia without agraphia—this is relatively common—but not agraphia without alexia. “Agraphia,” Rosenfield writes, “is always associated with an inability to read.” And yet extremely rare cases of isolated agraphia have been reported, and the debate is not yet resolved.