by Oliver Sacks
Above all, the recognition of faces depends not only on the ability to parse the visual aspects of a face—its particular features and their overall configuration—and compare it to others, but the ability to summon the memories, experiences, and feelings associated with that face. The recognition of specific places or faces, as Pallis emphasized, goes with a particular feeling, a sense of association and meaning. While purely visual recognition of faces is mediated by the fusiform face area and its connections, emotional familiarity is mediated at a higher, multimodal level, where there are intimate connections with the hippocampi and amygdala, areas dedicated to memory and emotion. Thus A.H., after his stroke, lost not only his ability to identify faces but this sense of familiarity; every face and place appeared new to him and continued to do so even if seen again and again.
Recognition is based on knowledge; familiarity is based on feeling; but neither entails the other. The two have different neural bases and can be dissociated; thus, although both are lost in tandem with prosopagnosia, one can have familiarity without recognition or recognition without familiarity in other conditions. The former occurs in déjà vu and also in the “hyperfamiliarity” for faces described by Devinsky. Here a patient may find that everyone on the bus or on the street looks “familiar”—he may go up to them and address them as old friends, even while realizing that he cannot possibly know them all. My father was always very sociable and could recognize hundreds or even thousands of people, but his feeling of “knowing” people became exaggerated, perhaps pathological, as he moved into his nineties. He often attended concerts at the Wigmore Hall in London, and there, during the intermissions, he would accost everyone in sight, saying, “Don’t I know you?”
The opposite occurs in patients with Capgras syndrome, for whom people’s faces, though recognized, no longer generate a sense of emotional familiarity. Since a husband or wife or child does not convey that special warm feeling of familiarity, the Capgras patient will argue, they cannot be the real thing—they must be clever impostors, counterfeits. People with prosopagnosia have insight; they realize that their problems with recognition come from their own brains. People with Capgras syndrome, in contrast, remain immovable in their conviction that they are perfectly normal and it is the other person who is profoundly, even uncannily wrong.
People with acquired prosopagnosia, like A.H. or Dr. P., are relatively rare—most neurologists are likely to encounter such a patient once or twice in their career, if at all. Congenital prosopagnosia (or, as it is sometimes called, “developmental” prosopagnosia), such as I have, is much commoner, yet remains completely unrecognized by most neurologists. Heather Sellers, a lifelong prosopagnosic, wrote about this in a 2007 autobiographical essay: “I couldn’t recognize my husband’s children.… I hugged the wrong man in the grocery, thinking it was [my husband].… My colleagues remained unidentifiable after a decade.… I kept introducing myself to neighbors.” When she consulted two separate neurologists for her problem, they both said that they had never seen it before, and it was “very rare.”12
One eminent neurologist who has written about visual agnosia confessed to me that he had not even heard of congenital prosopagnosia until very recently. This, however, is not entirely surprising, for people with congenital prosopagnosia do not generally consult neurologists about their “problem,” any more than someone with lifelong colorblindness would complain about it to an eye doctor. It is just the way they are.
But Ken Nakayama at Harvard, who investigates visual perception, has long suspected that prosopagnosia is relatively common but underreported. In 1999, he and his colleague Brad Duchaine, at University College London, began using the internet to seek subjects with face-blindness, and they received a startling response. They are now investigating several thousand people with lifelong prosopagnosia ranging from mild to cripplingly severe.13
While people with lifelong prosopagnosia do not have gross lesions in the brain, a recent study by Lucia Garrido and her colleagues showed that they do have subtle but distinct changes in the brain’s face-recognition areas. The condition also tends to be familial: Duchaine, Nakayama, and their colleagues have described one family in which ten members—both parents and seven of their eight children (the eighth could not be tested), as well as a maternal uncle—have it. Clearly there are strong genetic determinants at work here.
Nakayama and Duchaine have explored the neural basis of face and place recognition, generating new knowledge and insights at every level from the genetic to the cortical. They have also studied the psychological effects and social consequences of developmental prosopagnosia and topographical agnosia—the special problems these conditions can create for an individual in a complex social and urban culture.
The range seems to extend in a positive direction, too. Russell, Duchaine, and Nakayama have described “super-recognizers,” people with extraordinarily good face-recognition abilities, including some who seem to have indelible memories of virtually every face they have ever seen. Alexandra Lynch, one of my correspondents, described her own uncanny ability to recognize people:
It happened again yesterday. I was on my way down into the subway in Soho when I identified someone fifteen feet ahead of me (back turned, talking intimately with his friend) as a man I knew, or had seen before. In this case, it was Mac, who used to be a family friend’s art dealer. I had last seen him (briefly) two years earlier, at an opening in midtown. I’m not sure I’ve ever spoken with him beyond an introduction a good ten years ago.
This is an integral part of my life—I catch a passing glimpse of someone and, with no real effort, flash, place the face—yes, that’s the girl who served us wine at an East Village bar last year (again, in a totally different neighborhood, and at night not during the day). It is true that I’m a big fan of people, of humanity and diversity … but to my knowledge I make no effort to record the physical traits of ice cream servers, shoe salesmen and friends of friends of friends. Even a slim wedge of face, or the way someone walks two blocks away at dusk, can trigger my mind to zero in on a match.
The super-recognizers, Russell et al. write, “are about as good as many [lifelong] prosopagnosics are bad”—that is, they are about two or three standard deviations above average, while the most severe prosopagnosics have face-recognizing abilities two or three standard deviations below average. Thus the difference between the best face recognizers and the worst among us is comparable to that between people with an IQ of 150 and an IQ of 50, with others filling every level in between. As with any bell curve, the vast majority of people are somewhere in the middle.
Severe congenital prosopagnosia is estimated to affect at least 2 percent of the population—six million people in the United States alone. (A much higher percentage, perhaps 10 percent, are markedly below average in face identification but not cripplingly face-blind.) For these people, who have difficulty recognizing their husbands, wives, children, teachers, and colleagues, there is still no official recognition or public understanding.
This is in marked contrast to the situation with another neurological minority, the 5 to 10 percent of the population with dyslexia. Teachers and others are more and more aware of the special difficulties and often special gifts which dyslexic children may have, and are starting to provide educational strategies and resources for them.
But for now, people with varying degrees of face-blindness must rely on their own ingenuity and strategies, starting with educating others about their unusual, but not rare, condition. Increasingly, prosopagnosia is the subject of books, websites, and support groups, where people with face-blindness or topographical agnosia are able to share experiences and, no less important, strategies for recognizing faces and places when the usual “automatic” mechanisms are compromised.
Ken Nakayama, who is doing so much to further the scientific understanding of prosopagnosia, also has a personal acquaintance with the subject, and posts this notice in his office and on his website:
Recent eye probl
ems and mild prosopagnosia have made it harder for me to recognize people I should know. Please help by giving your name if we meet. Many thanks.
1. This is an exaggeration—I had no trouble recognizing my parents or my brothers, though I was less adept with my huge extended family and completely lost, sometimes, when I saw photographs of them. I had dozens of aunts and uncles, and when I published my memoir Uncle Tungsten, I selected for the hardcover edition a photograph of another uncle, whom I mistakenly identified as Uncle Tungsten. This upset and bewildered his family, who said, “How could you make such a mistake? They look nothing like one another.” (I corrected the error in the paperback edition.)
2. Our other two brothers seemed to have normal powers of facial recognition. My father, a general practitioner, was immensely gregarious and knew hundreds of people, not to mention the thousands of patients in his practice. My mother, in contrast, was almost pathologically shy. She had a small circle of intimates—family and colleagues—and was very ill at ease in large gatherings. I cannot help wondering, in retrospect, if some of her “shyness” was due to a mild prosopagnosia.
3. A most remarkable and creative reaction to face-blindness—the word “compensation” seems inadequate—is that of the artist Chuck Close, who is famous for his gigantic portraits of faces. Close himself has severe lifelong prosopagnosia. But this, he believes, played a crucial role in driving his unique artistic vision. He says, “I don’t know who anyone is and have essentially no memory at all for people in real space, but when I flatten them out in a photograph, I can commit that image to memory in a way; I have almost a kind of photographic memory for flat stuff.”
4. It is similar with milder degrees of colorblindness or stereo blindness. People may be unaware of these “deficits,” considering themselves normal, until the deficit is revealed through, for example, a routine eye examination or driver’s license test.
5. Once, as I was being interviewed on the radio about The Man Who Mistook His Wife for a Hat, a listener phoned in and said, “I can’t recognize my wife, either.” (This, he added, was because he had developed a brain tumor.) I arranged to see Lester C. and find out more about his experiences.
While Lester had found various strategies for recognizing people, he told me, he was distressed by his inability to appreciate the beauty of faces. He had had “a great eye for the girls,” he said, before the tumor. Now he had to judge beauty indirectly, by taking seven criteria (color of eyes, shape of nose, symmetry, etc.) and rating each of these on a scale of one to ten. This way he could construct a “mental histogram,” as he put it, for beauty. But he soon found that such histograms did not work and were sometimes ludicrously at odds with a direct or intuitive judgment of beauty such as he had once had.
Most people with prosopagnosia remain sensitive to facial expressions, seeing at a glance whether someone looks happy or sad, friendly or hostile, even though the faces themselves may be unidentifiable. The reverse also occurs: Antonio Damasio has described how people with damage to the amygdala (a part of the brain crucial to the perception and feeling of emotion) may have difficulty “reading” faces, judging their emotional expressions, even though they recognize faces normally. This may also be the case with some autistic people. Temple Grandin, who has Asperger’s syndrome, says, “I can recognize major expressions on a person’s face, but I do not pick up subtle cues. I did not know that people had little eye signals until I read about them in Simon Baron-Cohen’s book Mindblindness when I was fifty.” (Though Temple is a “visual thinker” and can easily visualize complicated engineering problems, she seems to be no better or worse than average at recognizing faces.)
Difficulty making social contact with others can also be a central problem in schizophrenia, and Yong-Wook Shin et al. have obtained preliminary results suggesting that schizophrenic people have difficulty not only in reading facial expressions but in face recognition, too.
6. Determined to provide some objective correlate, Gall went further, trying to measure and correlate personality and moral faculties of individuals with the shapes and bumps of their skulls, using a method he called “cranioscopy.” One of his students, Johann Spurzheim, went on to popularize this idea as “phrenology,” a pseudoscience that gained much attention in the early nineteenth century and influenced Lombroso’s theories of criminal physiognomy. Spurzheim and Lombroso’s work has long been discredited, but Gall’s idea of localization in the brain had a lasting impact.
7. In 1869, Hughlings Jackson debated this issue with Broca, insisting that “to locate the damage which destroys speech and to locate speech are two different things.” Jackson, it was generally thought, lost this debate, but he was not the only one with reservations. Freud, in his 1891 book On Aphasia, suggested that the use of language demanded many interconnected areas of the brain, and that Broca’s area was only one node in a vast cerebral network. The neurologist Henry Head, in his monumental 1926 treatise Aphasia and Kindred Disorders of Speech, inveighed against “the diagram-makers,” as he called the aphasiologists of the nineteenth century. Head argued, as Hughlings Jackson and Freud had, for a much more holistic view of speech.
8. Much that we now take for granted in neuroscience was very unclear when Gross began this work. Even in the late 1960s, it was widely believed that the visual cortex did not extend far beyond its main locus in the occipital lobes (as we now know it does). That the representation and recognition of specific categories of objects—faces, hands, etc.—might rely on individual neurons or clusters of neurons was considered improbable, even absurd; the idea was good-humoredly mocked by Jerome Lettvin in his famous comments about “grandmother cells.” Very little attention, therefore, was paid to Gross’s early findings, and it was not until the 1980s that they were confirmed and amplified by other researchers.
9. Different inferotemporal cells, they write, are “selective for different face parts and interactions between parts, and even the same cell can respond maximally to different combinations of face parts. Thus, there is no single blueprint for detecting the form of a face.… This diversity of feature tuning provides the brain with a rich vocabulary to describe faces and shows how a high-dimensional parameter space may be encoded even in a small region of [the inferotemporal cortex].”
10. Koch, Fried, and their colleagues have published many papers on their work; those most relevant here include Quian Quiroga et al., 2005 and 2009.
11. Yoichi Sugita points out, however, that this narrowing is easily reversible, at least in childhood, by experience.
12. Despite its unfamiliarity to modern physicians, congenital prosopagnosia entered the medical literature as early as 1844, when A. L. Wigan, an English doctor, described one of his patients:
A gentleman of middle age … lamented to me his utter inability to remember faces. He would converse with a person for an hour, but after an interval of a day could not recognise him again. Even friends, with whom he had been engaged in business transactions, he was unconscious of ever having seen. Being in an occupation in which it was essential to cultivate the good-will of the public, his life was made perfectly miserable by this unfortunate defect, and his time was passed in offending and apologizing. He was quite incapable of making a mental picture of anything, and it was not till he heard the voice, that he could recognise men with whom he had constant intercourse.… I endeavoured in vain to convince him that an acknowledgment of the defect would be the best means of removing the unfortunate effect it had produced in alienating friends. He was quite determined to conceal it, if possible, and it was impossible to convince him that it did not depend solely on the eyes.
13. Information is available at their website, www.faceblind.org.
Stereo Sue
WHEN GALEN, in the second century, and Leonardo, thirteen centuries later, observed that the images received by the two eyes were slightly different, neither of them appreciated the full significance of these differences. It was not until the early 1830s that Charles Wheatstone, a young phy
sicist, began to suspect that even though the brain somehow fused these images automatically and unconsciously, the disparities between the two retinal images were in fact crucial to the brain’s mysterious ability to generate a sensation of depth.
Wheatstone confirmed the truth of his conjecture by an experimental method as simple as it was brilliant. He made pairs of drawings of a solid object as seen from the slightly different perspectives of the two eyes and then designed an instrument that used mirrors to insure that each eye saw only its own drawing. He called the device a stereoscope, from the Greek for “solid vision.” If one looked into the stereoscope, the two flat drawings would fuse to produce a single three-dimensional drawing poised in space.
(One does not need a stereoscope to see stereo depth; it is relatively easy for most people to learn how to “free-fuse” such drawings, simply by diverging or converging the eyes. So it is strange that stereopsis was not discovered centuries before: Euclid or Archimedes could have drawn stereo diagrams in the sand, as David Hubel has remarked, and discovered stereopsis in the third century B.C. But they did not, as far as we know.)