The Mind's Eye

by Oliver Sacks

  Bach-y-Rita wondered if one might connect the output of a video camera, point by point, to the skin, allowing a blind subject to form a “touch picture” of his environment. This might work, he thought, because tactile information is organized topographically in the brain, and topographic accuracy is essential for forming a quasi-visual picture. Eventually, he began using tiny grids of a hundred or so electrodes on that most sensitive part of the body, the tongue. (The tongue has the highest density of sensory receptors in the body, and it also occupies the greatest amount of space, proportionally, in the sensory cortex. This makes it uniquely suitable for sensory substitution.) With this device, the size of a postage stamp, his subjects could form a crude but nevertheless useful “picture” on the tongue itself.

  Over the years, the sophistication of such devices has increased greatly, and prototypes now have four to six times the resolution of Bach-y-Rita’s early version. Bulky camera cables have been replaced by spectacles containing miniature cameras, allowing subjects to direct the cameras by a more natural head movement. With this, blind subjects are able to walk across a room that is not too cluttered, or to catch a ball rolled towards them.

  Does this mean that they are now “seeing”? Certainly, they are showing what behaviorists would call “visual behavior.” Bach-y-Rita spoke of how his subjects “learn[ed] to make perceptual judgements using visual means of interpretation, such as perspective, parallax, looming and zooming, and depth estimates.” Many of these people felt as if they were seeing once again, and functional MRIs showed strong activations of visual areas in their brains while they were “seeing” with the camera. (“Seeing” occurred particularly when the subjects were able to move the camera voluntarily, pointing it here or there, looking with it. Looking was crucial, for there is no perception without action, no seeing without looking.)

  To restore sight to someone who once had it, whether by surgical means or by a sensory-substitution device, is one thing, for such a person would have an intact visual cortex and a lifetime of visual memories. But to give sight to someone who has never seen, never experienced light or sight, would seem to be impossible, in view of what we know about the brain’s critical periods and the necessity of at least some visual experience in the first two years of life to stimulate the development of the visual cortex. (Recent work from Pawan Sinha and others, however, suggests that the critical period may not be as critical as previously accepted.)16 Tongue vision has been tried with congenitally blind people, too, and with some success. One young musician, born blind, said she “saw” the conductor’s gestures for the first time in her life.17 Although the visual cortex in congenitally blind people is reduced in volume by more than 25 percent, it can still, apparently, be activated by sensory substitution, and this has been confirmed, in several cases, by fMRIs.18

  There is increasing evidence for the extraordinarily rich interconnectedness and interactions of the sensory areas of the brain, and the difficulty, therefore, of saying that anything is purely visual or purely auditory, or purely anything. The world of the blind can be especially rich in such in-between states—the intersensory, the metamodal—states for which we have no common language.19

  On Blindness is an exchange of letters between the blind philosopher Martin Milligan and a sighted philosopher, Bryan Magee. While his own nonvisual world seems coherent and complete to him, Milligan realizes that sighted people have access to a sense, a mode of knowledge, denied him. But congenitally blind people, he insists, can (and usually do) have rich and varied perceptual experiences, mediated by language and by imagery of a nonvisual sort. Thus they may have a “mind’s ear” or a “mind’s nose.” But do they have a mind’s eye?

  Here Milligan and Magee cannot reach agreement. Magee insists that Milligan, a blind man, cannot have any real knowledge of the visual world. Milligan disagrees and maintains that even though language only describes people and events, it can sometimes stand in for direct experience or acquaintance.

  Congenitally blind children, it has often been noted, tend to have superior memories and be precocious verbally. They may develop such fluency in the verbal description of faces and places as to leave others (and perhaps themselves) uncertain as to whether they are actually blind. Helen Keller’s writing, to give a famous example, startles one with its brilliantly visual quality.

  I loved reading Prescott’s Conquest of Mexico and Conquest of Peru as a boy, and felt that I “saw” these lands through his intensely visual, almost hallucinogenic descriptions. I was amazed to discover, years later, that Prescott had not only never visited Mexico or Peru; he had been virtually blind since the age of eighteen. Did he, like Torey, compensate for his blindness by developing huge powers of visual imagery, or were his brilliant visual descriptions simulated, in a way, made possible by the evocative and pictorial powers of language? To what extent can description, picturing in words, provide a substitute for actual seeing or for the visual, pictorial imagination?

  After becoming blind in her forties, Arlene Gordon found language and description increasingly important; it stimulated her powers of visual imagery as never before and, in a sense, enabled her to see. “I love traveling,” she told me. “I saw Venice when I was there.” She explained how her traveling companions would describe places, and she would then construct a visual image from these details, her reading, and her own visual memories. “Sighted people enjoy traveling with me,” she said. “I ask them questions, then they look and see things they wouldn’t otherwise. Too often people with sight don’t see anything! It’s a reciprocal process—we enrich each other’s worlds.”

  There is a paradox here—a delicious one—which I cannot resolve: if there is indeed a fundamental difference between experience and description, between direct and mediated knowledge of the world, how is it that language can be so powerful? Language, that most human invention, can enable what, in principle, should not be possible. It can allow all of us, even the congenitally blind, to see with another person’s eyes.

  1. Despite an initially overwhelming sense of despair on losing their sight, some people, like Hull, have found their full creative strength and identity on the other side of blindness. One thinks especially of John Milton, who started to lose his sight around the age of thirty (probably from glaucoma), but produced his greatest poetry after becoming completely blind a dozen years later. He meditated on blindness, how an inward sight may come in place of outward sight, in Paradise Lost, in Samson Agonistes, and—most directly—in letters to friends and in a very personal sonnet, “On His Blindness.” Jorge Luis Borges, another poet who became blind, wrote about the varied and paradoxical effects of his own blindness; he also wondered how it might have been for Homer, who, Borges imagined, lost the world of sight but gained a much deeper sense of time and, with this, a matchless epic power. (This is beautifully discussed by J. T. Fraser in his 1989 foreword for the Braille edition of Time, the Familiar Stranger.)

  2. In his book The Invention of Clouds, Richard Hamblyn recounts how Luke Howard, the nineteenth-century chemist who first classified clouds, corresponded with many other naturalists of the time, including John Gough, a mathematician blinded by smallpox at the age of two. Gough, Hamblyn writes, “was a noted botanist, having taught himself the entire Linnean system by touch. He was also a master of the fields of mathematics, zoology and scoteography—the art of writing in the dark.” (Hamblyn adds that Gough “might also have become an accomplished musician had his father, a stern Quaker … not stopped him playing on the godless violin that an itinerant fiddler had given him.”)

  3. Tenberken also has an intense synesthesia, which has persisted and been intensified, it seems, by her blindness:

  As far back as I can remember, numbers and words have instantly triggered colors in me.… The number 4, for example, [is] gold. Five is light green. Nine is vermilion.… Days of the week as well as months have their colors, too. I have them arranged in geometrical formations, in circular sectors, a little like a pie. When I need
to recall on which day a particular event happened, the first thing that pops up on my inner screen is the day’s color, then its position in the pie.

  4. Although I myself am a poor visualizer, if I shut my eyes, I can still “see” my hands moving on the piano keyboard when I play a piece that I know well. (This may happen even if I just play the piece in my mind.) I feel my hands moving at the same time, and I am not entirely sure that I can distinguish the “feeling” from the “seeing.” In this context, they seem inseparable, and one wants to use an intersensory term like “seeing-feeling.”

  The psychologist Jerome Bruner speaks of such imagery as “enactive”—an integral feature of a performance (real or imaginary)—in contrast to an “iconic” visualization, the visualization of something outside oneself. The brain mechanisms underlying these two sorts of imagery are quite different.

  5. Though I have almost no voluntary imagery, I am prone to involuntary imagery. I used to have this only as I was falling asleep, in migraine auras, with some drugs, or with fever. But now that my sight is impaired, I have it all the time.

  In the 1960s, during a period of experimenting with large doses of amphetamines, I experienced a different sort of vivid mental imagery. Amphetamines can produce striking perceptual changes and dramatic enhancements of visual imagery and memory (as I described in “The Dog Beneath the Skin,” a chapter in The Man Who Mistook His Wife for a Hat). For a period of two weeks or so, I found that I had only to look at an anatomical picture or specimen, and its image would remain vivid and stable in my mind for hours. I could mentally project the image onto a piece of paper—it was as clear and distinct as if projected by a camera lucida—and trace its outlines with a pencil. My drawings were not elegant, but they were, everyone agreed, quite detailed and accurate. But when the amphetamine-induced state faded, I could no longer visualize, no longer project images, no longer draw—nor have I been able to do so in the decades since. This was not like voluntary imagery—I did not summon images to my mind or construct them bit by bit. It was involuntary and automatic, more akin to eidetic or “photographic” memory, or to palinopsia, an exaggerated persistence of vision.

  6. The physicist John Tyndall referred to these in an 1870 lecture, a few years before Galton’s Inquiries: “In explaining scientific phenomena, we habitually form mental images of the ultra-sensible.… Without the exercise of this power our knowledge of nature would be a mere fabulation of co-existences and sequences.”

  7. I described Temple more fully in An Anthropologist on Mars, and she speaks about her visual thinking especially in her book Thinking in Pictures.

  8. Kosslyn’s latest book on the matter, The Case for Mental Imagery, details the history of this debate.

  9. Functional MRIs also showed that the two hemispheres of the brain behaved differently in regard to imagery, the left hemisphere concerned with generic, categorical images—e.g., “trees”—and the right hemisphere with specific images—e.g., “the maple in my front yard”—a specialization also present in visual perception. Thus prosopagnosia, an inability to recognize specific faces, is associated with damaged or defective visual function in the right hemisphere, though people with prosopagnosia have no problem with the category of faces in general, a left-hemisphere function.

  10. Mr. I.’s case is described in An Anthropologist on Mars.

  11. While it seems clear that perception and imagery share certain neural mechanisms at higher levels, this sharing is less evident in the primary visual cortex—hence the possibility of a dissociation such as occurs in Anton’s syndrome. In Anton’s syndrome, patients with occipital damage are cortically blind, but believe they are still sighted. They will move about without restraint or caution, and if they bump into a piece of furniture, they will ascribe this, perhaps, to the furniture being “out of place.”

  Anton’s syndrome is sometimes attributed to the preservation of some visual imagery despite occipital damage, and to patients mistaking this imagery for perception. But there may be other, stranger mechanisms at work. The denial of blindness—or, more accurately, the inability to realize that one has lost one’s vision—is very like another “disconnection syndrome,” known as anosognosia. With anosognosia, following damage to the right parietal lobe, patients lose awareness of their left side, and of the left half of space, along with the awareness that anything is amiss. If one draws their attention to their left arm, they will say it is someone else’s—“the doctor’s arm,” or “my brother’s arm,” or even “an arm someone left here.” Such confabulations seem similar in a way to those of Anton’s syndrome, attempts to explain what, to the patient, is a bizarrely inexplicable situation.

  12. Einstein described this in regard to his own thinking:

  The psychical entities which seem to serve as elements in thought are certain signs and more or less clear images which can be “voluntarily” reproduced and combined.… [Some] are, in my case, of visual and some of muscular type. Conventional words or other signs have to be sought for laboriously only in a second stage.

  Darwin, on the other hand, seemed to describe a very abstract, almost computational process in his own thinking, when he wrote in his autobiography, “My mind seems to have become a kind of machine for grinding general laws out of large collections of facts.” (What Darwin omitted here was that he had a fantastic eye for form and detail, an enormous observational and depictive power, and it was these which provided the “facts.”)

  13. Dominic ffytche, who has investigated the neurobiology of conscious vision—imagery and hallucination as well as perception—feels that visual consciousness is a threshold phenomenon. Using fMRIs to study patients with visual hallucinations, he has shown that there may be evidence of unusual activity in a specific part of the visual system—for example, the fusiform face area—but this has to reach a certain intensity before it enters consciousness, before the subject actually “sees” faces.

  14. The heightened (and sometimes morbid) sensitivity of the visual cortex when deprived of its normal perceptual input may also predispose it to intrusive imagery. A significant proportion of those who go blind—10 to 20 percent, by most estimates—become prone to involuntary images, or outright hallucinations, of an intense and sometimes bizarre kind. Such hallucinations were originally described in the 1760s by the Swiss naturalist Charles Bonnet, and we now speak of hallucinations secondary to visual impairment as Charles Bonnet syndrome.

  Hull described something akin to this which occurred for a while after he lost the last of his sight:

  About a year after I was registered blind, I began to have such strong images of what people’s faces looked like that they were almost like hallucinations.… I would be sitting in a room with someone, my face pointed towards my companion, listening to him or her. Suddenly, such a vivid picture would flash before my mind that it was like looking at a television set. Ah, I would think, there he is, with his glasses and his little beard, his wavy hair and his blue, pinstriped suit, white collar and blue tie.… Now this image would fade and in its place another one would be projected. My companion was now fat and perspiring with receding hair. He had a red necktie and waistcoat, and a couple of his teeth were missing.

  15. Ben, who had retinoblastoma, had both eyes removed at the age of three, but then, tragically, died at sixteen from a recurrence of his cancer. Videos of Ben and his echolocation can be seen at the website www.benunderwood.com.

  16. See Ostrovsky et al., for example.

  17. Congenitally blind people, we might suppose, can have no visual imagery at all, since they have never had any visual experience. And yet they sometimes report having clear and recognizable visual elements in their dreams. Helder Bértolo and his colleagues in Lisbon, in an intriguing 2003 report, described how they compared congenitally blind subjects with normal sighted subjects and found “equivalent visual activity” (based on analysis of EEG alpha-wave attenuation) in the two groups while dreaming. The blind subjects were able, upon waking, to draw the visual co
mponents of their dreams, although they had a lower rate of dream recall. Bértolo et al. conclude, therefore, that “the congenitally blind have visual content in their dreams.”

  18. Would acquiring “sight” if one has never seen before be bewildering or enriching? For my patient Virgil, who was given sight, through surgery, after a lifetime of blindness, it was utterly incomprehensible at first, as I described in An Anthropologist on Mars. Thus although sensory-substitution technologies are exciting and promise a new freedom for blind people, we need to consider equally their impact on a life that has already been constructed without sight.

  19. In a recent letter to his colleague Simon Hayhoe, John Hull expanded on this:

  For example, when the thought of a car occurs to me, although my front-line images are of recently touching the warm bonnet of a car, or of the shape of the car as I feel for the door handle, there are also traces of the appearance of the whole car, from pictures of cars in books, or memories of cars coming and going. Sometimes, when I have to touch a modern car, I am surprised to find that this memory trace does not correspond to reality, and that cars are not the same shape they were twenty-five years ago.

  There is a second point. The fact that an item of knowledge is so much buried in the sense or senses that first received it, means for me that I am not always sure whether my image is visual or not. The trouble is that tactile images of the shape and feel of things also often seem to acquire a visual content, or one cannot tell if the three-dimensional memory shape is being mentally represented by a visual or a tactile image. So even after all these years, the brain can’t sort out where it is getting stuff from.