THE ABILITY TO MAKE SOMEONE FEEL AGONY in a limb that isn’t there is one of pain’s most impressive sleights of hand. And phantom pain is not a rarity or a fluke: Among people who have had a limb amputated, almost 70 percent will end up suffering some form of phantom pain. Nobody knows for sure what causes it. It may take weeks or years to develop—and it can last for a lifetime. Occasionally, over time, the pain may recede, although no one knows why. And although surgeons may snip away at nerves or stumps in an effort to alleviate it, phantom pain remains notoriously difficult to treat.
People report specific kinds of pain—burning, stabbing, or aching. They feel the pressure of a nonexistent ring on a missing finger, or fingernails digging into the palm of a phantom hand. For all its ghostliness, phantom pain is explicit about where it hurts and how it feels—warm or wet, gritty or metallic. The sensation is said to be almost holographic. If ever there was evidence for pain being “all in the head,” phantom pains provide it. They also demonstrate that pain is sometimes a complex perceptual illusion involving vision, body image, and sense memory.
Surgeons can cut the nerves that lead from the limb to the spinal cord (a rhizotomy). But cutting nerves is not considered a very effective way to treat phantom pain, which is too intimately connected with the dance of perception itself to be plucked out like a splinter.
The Civil War physician Silas Weir Mitchell came up with the name phantom limb pain. In those days, lopping off limbs was the most efficient way to prevent infection or gangrene. This led to a bumper crop of amputees, who began to report bizarre sensations or excruciating pain in their missing limbs, a phenomenon that Mitchell wrote about in detail. But the first mention of phantom limb pain goes back to the sixteenth century in descriptions by the French surgeon Ambroise Paré. And Lord Nelson also experienced it. In 1797, when he lost an arm in battle and discovered that he still felt the presence of his fingers, Nelson declared that this was proof that the soul really does exist. If his arm could persist when its physical attributes were gone, surely our spirits can survive the demise of the body.
Sometimes phantom limbs have an active life as well. They will “gesture” or try to carry on with their previous lives—absent arms that rise to swing at tennis balls, missing hands that reach for a cup. One man said his phantom hand still reached for the check in restaurants (“Oh, my phantom will pay—trust me”). For equally mysterious reasons, phantom limbs can also be paralyzed or become stuck in awkward positions. Amputees with frozen phantom limbs will automatically make room for them when they walk through doorways and protect them in tight situations.
For years, phantom limbs were treated almost as part of medical folklore—exotic case studies that made for lively reading in medical monographs. But for people who suffer from phantom pain, it is as real as having a broken neck in a halo cast.
As I rummaged through the literature, I was struck by how the successive historical explanations for this weird phenomenon—from the soul to the current concept of neurogenesis, or the brain’s ability to grow new cells—also reflect the way we’ve conceived of pain in general.
Phantom limbs were initially regarded as something that couldn’t be touched scientifically—they were the body’s equivalent of an alien abduction. In fact, Mitchell was so unsure of his speculations about them that he published his first article on the subject under a pseudonym in a popular magazine instead of a scientific journal.
This also reflected the nineteenth-century attitude that pain wasn’t legitimate unless it could be pointed to, probed, and measured; otherwise, it was “hysteria,” “neurasthenia,” or simply madness.
The next phantom pain theory, which lingers on, is that it arises from inflamed or irritated nerve endings, called neuromas, in the stump. Some doctors have tried to treat phantom pain by surgically removing these neuromas or even by doing an additional amputation. Occasionally this helps, but each new amputation can also breed a new phantom, raising the specter of a hall-of-mirrors effect and an infinite number of phantoms-within-phantoms. Finally, in the middle of last century, the attention shifted to the brain.
In the 1940s and ’50s, the great Canadian neurosurgeon Wilder Penfield began to map the brain in an alarmingly straightforward way. Since the brain itself has no pain receptors, he was able to do surgery on it using local anesthetic. Patients remained conscious. While the brain was exposed, he stimulated particular parts of it with an electrode. Then he asked the patients on the operating table what they felt as he poked about.
What they reported was remarkable—a kind of jump-cut movie that involved feelings, images, and vivid memories. Penfield was like a puppeteer, pulling the strings attached to different parts of their lives. One of the things he discovered was a narrow strip that runs down both sides of the brain and contains a representation of different parts of the body. If he stimulated an area at the top of the brain, the patient felt sensations in his genitals. If he moved the electrode down a bit, the patient then reported a tingling in his feet. What Penfield learned was that different parts of the body—the tongue, the lips, the hands, the genitals, and the face in particular—have a disproportionately large area of representation in the brain. It was an early version of “brain maps,” which now number in the hundreds, and it gave rise to that odd little medical-school character known as the Penfield homunculus.
Students encounter this troll-like figure when they first learn about the somatosensory sites in the brain that correspond to parts of the body. Since the hands and the mouth have the most complex nerve endings for touch, warmth, and pain, they have accordingly large areas in the brain map. But the torso doesn’t need to discriminate so finely in its sensations, so it has a smaller area of representation. The resulting Mr. Homunculus (looking a bit like some licensed action figure) has gigantic lips, tiny little arms, and big steam-shovel hands.
From the point of view of phantom limb pain, what was revealing is that the map for the foot is right next to the map for the genitals, and the hand area is adjacent to the area representing the face. This may help explain why sensation in a missing thumb can be triggered by light stroking on the lips; according to some scientists, this is a result of the sensory input migrating from one area of the brain to its neighbor.
The problem of phantom limb pain gets right to the heart of our misunderstanding of how pain works. In the past, we assumed that pain originated in the injured body part. Cut it off, or cut the nerves, and the pain will end. Also, any pain that couldn’t be located in the body was a trick of the mind, a delusion, and therefore untreatable. But some researchers in the area of phantom pain—John Dostroevsky and Joel Katz at the University of Toronto, Ronald Melzack at McGill University, and neurologist V. S. Ramachandran at the Center for Brain and Cognition at the University of California in San Diego—have carried out studies that illustrate what Melzack has always argued: Pain is in the brain. “You don’t need a body to feel a body,” Melzack wrote. “The brain itself can generate every quality of experience which is normally triggered by sensory input.”
Melzack agrees that the brain has a map of the body, but he argues that the map is innate and genetic. Otherwise, how could one account for phantom pain in people born without a limb, with no possible “body knowledge” of pain in the missing part? Melzack has studied this phenomenon in 125 people, most of them teenagers, who had either lost a limb before the age of six or were born without a limb. About a quarter of the people missing a limb since birth reported feeling, sometimes vividly, the arm or leg that they had lost. “One woman’s phantom arm reached out to prevent a cupboard door from slamming shut and to catch a falling egg,” Melzack wrote in a 1998 article in Discover magazine.
These findings persuaded him that “the body we perceive is in large part built into our brain—it’s not entirely learned.” He thinks that a network of neurons forms in the embryonic brain to link up the somatosensory thalamus and cortex (regions that enable us to sense the location of our limbs) and the limbic system, which is involved in feelings of pain and pleasure and is the area that helps us learn from our experiences. These connections prepare the brain to respond to body parts that may not even form.
“The brain needs to have information about what is going to happen to it,” says Melzack. “The brain is not just born into the world willy-nilly, waiting for anything to come bombarding in from the senses. It anticipates that it will be getting information from a body that has limbs and other organs, that there will be a mother and two breasts and sources of food. And even if we are missing a part of the body, the brain is still able to generate the perception of that part.”
(This observation makes you wonder about conditions that involve persistent distortions of body image, such as anorexia nervosa. Is our body image laid down long before we look into a mirror?)
Melzack called this convergence of connections in the brain the neuromatrix, and attributed to it the totally convincing, detailed, baffling sense of pain in a limb that is no longer there. This is one theory. In truth, nobody knows why or how phantom pain arises. But nothing could illustrate the primacy of the brain in our experience of pain more dramatically than this utterly real, utterly “invented” pain.
The neuromatrix argument is compelling, however, because it takes pain out of the mechanical, tissue-damage, signal-transmission domain and makes emotions an integral part of pain “wiring.” It suggests that cultural attitudes toward pain—the meanings we attribute to pain—can play a profound physiological role in how we feel it. On the one hand, the neuromatrix model may look like another cuts-of-meat, reductionist argument; on the other, it offers a metaphor for what geneticists are discovering about the perception of pain as an inheritable, individual expression. (Melzack writes about each person’s “neurosignature.”) The concept of a neuromatrix suggests that pain is not an invasive, alien force or a learned response but part of the map of who we are. Pain should not surprise us (a point of view the Buddhists have been teaching for some time).
The neurologist V. S. Ramachandran takes a more … neurological view of things, as one might expect. In Phantoms in the Brain, his entertaining 1998 book, he describes a number of off-the-wall experiments with people who suffer from phantom pain, where he manages to visually trick the brain into revising its mistaken body image. While his experiments don’t offer any practical treatment for phantom pain, they do illustrate just how integrated the experience of pain is with other senses, especially seeing and hearing. (Elaine Scarry reports in her book The Body in Pain that forcing their victims to look at the instruments of torture is one of the strategies that torturers use to increase the intensity of the pain.)
Ramachandran’s book addresses itself first to the old conundrum of how the brain gives rise to the mind. We know the topography of the brain, but how does the activity of 600 billion neurons define “self”? By looking at a number of odd neurological syndromes, he not only comes up with enticing theories about the “how” of perception, but, in the course of his experiments, finds himself able—temporarily at least—to manipulate the pain intensity and posture of phantom limbs. These involve smoke-and-mirror strategies that Ramachandran reports with boyish enthusiasm, like someone demonstrating his prizewinning science-fair project.
Some women, Ramachandran notes, have reported phantom breasts after radical mastectomy, complete with sensation in phantom nipples. Men who have had the misfortune of losing their penis have reported phantom erections. Ramachandran tells a story of a patient who confided to him that ever since his leg had been amputated from below the knee, he was experiencing strange new sensations whenever he had sex. His orgasm had spread, as it were, to his phantom foot—which, as he pointed out, offered more terrain for sensation. He seemed to enjoy his new, enlarged pedal orgasms, but he wanted to know if they were within the realm of normal phantom behavior. (They were.)
Ramachandran did a simple experiment on a man who had lost his arm and hand below the elbow. Using a Q-tip, he stroked various parts of the man’s body. When he touched his cheek and other areas on his face, the man felt the sensation both in his face and in his phantom hand. He could travel from baby finger to index, by slowly moving the Q-tip from one facial area to another. Soon Ramachandran had mapped out the whole hand—on the man’s face. The only explanation for this weird correspondence between missing hand and sensations in the face, Ramachandran felt, was the fact that the brain maps for the hand and the face were side by side. (“More Descartes thinking,” Melzack humphed when I asked him later about Ramachandran’s theories.)
When amputation interrupted signals from the hand, Ramachandran believes, sensory fibers belonging to the face began to invade the “empty” hand area and become active there. Neural colonialism, if you like, or squatters’ rights. As he writes, “This finding flatly contradicts one of the most widely accepted dogmas in neurology—the fixed nature of connections in the adult human brain.” Contrary to what is taught in textbooks, he continues, “new, highly precise and functionally effective pathways can emerge in the adult brain as early as four weeks after injury.”
The brain, formerly thought of as fixed, is capable of change and remapping. This idea of neuroplasticity has emerged only in the last twenty years, and it has far-reaching implications for pain patients and for people with nerve damage from strokes or injury. It means not only that there is a possibility of nerve cells regenerating or invading damaged areas—“like ivy growing over a brick wall,” as one doctor put it—but that there may be a built-in redundancy to the brain: When one area conks out, another can kick in.
But how does phantom pain arise? That was still a mystery. The brain cells might make new connections. Or maybe the “volume control mechanisms” involving pain go awry as a result of the remapping. Ramachandran likens it to a wa-wa pedal in the brain that creates a jumbled, echo effect that is eventually perceived as pain. The fact is, he writes, despite our progress in locating what happens where in the brain, “we really don’t know how the brain translates patterns of nerve activity into conscious experience, be it pain, pleasure or color.” In matters of the brain, where is less important than how.
Ramachandran found that although he couldn’t cure phantom pain, he could sometimes “conduct” the missing limbs. A patient with cancer came to see him with a vivid phantom limb that was doing very strange things. It would sometimes go into a clenching spasm, and nothing could get his phantom hand to open. He even felt phantom fingernails digging into his palm, and it was excruciating.
The neurologist did an experiment using a simple device that involved a vertical mirror in a box with two holes in the front. He had the man put his good arm inside the box on the right side of the mirror and “insert” his phantom on the left. Then he asked his volunteer to use the mirror to superimpose the image of his good right arm on the missing limb. This gave the man the visual impression that his left arm had been restored. Then Ramachandran asked him to clench and unclench his good hand. This, of course, resulted in the image of both hands unclenching. As a result of this visual feedback, the man felt his phantom hand opening up. Not only that, but the pain went away. It stayed away until another spasm arrived, which responded again to the mirror-box exercise. The mirror-box not only gave the image of his hand back to the man, but also erased the pain. Alas, the results faded after a while.
Is body image a mere hallucination then? And is this why one-hundred-pound teenage girls can look in the mirror and still say, “I’m so fat”? Since vision accounts for thirty brain maps itself, and even plays a role in pain perception, the idea of the brain as a collection of fixed “modules” begins to look hopelessly old-fashioned. Ramachandran concluded that “pain is an opinion on the organism’s state of health rather than a mere reflexive response to an injury…. There is so much interaction between different brain centers, like those concerned with vision and touch, that even the mere visual appearance of an opening fist can actually feed all the way back into the patient’s motor and touch pathways, allowing him to feel the fist opening, thereby killing an illusory pain in a nonexistent hand.”
What we are learning about neuroplasticity suggests that the brain is both site-specific, with certain functions organized in certain areas, and dynamic, with the ability to change and reorganize itself. Far-flung areas of the brain are linked up in such “simple” activities as seeing. And, as Ramachandran’s mirror-box proves, even vision can affect the pain we feel. Seeing is not just believing—seeing is feeling, too.