by Sam Kean
Unfortunately, many Civil War commanders—steeped in antiquated tactics and drenched in the romance of Napoleonic charges—never adjusted to the new reality. Most notoriously, on the day Mitchell arrived at Gettysburg, some 12,500 Confederate soldiers stormed a stone fence held by the Union. Pickett’s Charge. Among other troops, soldiers with piles of minnie balls were waiting, and they pulped the guts and pulverized the bones of the chargers up and down the line.
An injured soldier might languish for days before a stretcher team or ambulance wagon lugged him to a clinic. There, he might wait hours more until a surgeon in a bloody apron appeared, a knife between his teeth. The surgeon would probe the wound with fingers still crimson from the last patient, and if he decided to amputate, one assistant knocked the patient out with chloroform or ether, another put the limb into a headlock, and a third got ready to clamp the arteries. Four minutes later, the limb fell. The surgeon hollered “Next!” and walked on. This work might continue all day—one Kentucky surgeon remembered his fingernails getting soft from absorbing so much blood—and fresh graves ringed every hospital.* Walt Whitman recalled the crude tombstones, mere “barrel-staves or broken boards stuck in the dirt.”
After Gettysburg, Mitchell returned to Philadelphia to deal with the deluge of casualties. And although he continued his private practice (military work paid just $80 per month), he spent the better part of most days at Turner’s Lane, arriving at 7 a.m. for an hour of rounds, then returning around 3 p.m. and often staying until midnight. He spent hours writing up case reports as well—an illuminating experience. His early research training had emphasized rigor and data, but Mitchell found he couldn’t capture these cases with numbers and charts alone. Only narrative accounts could get at what injured soldiers really felt. The narratives affected him so profoundly, in fact, that in later years he began writing novels about his experiences, and drew on these case reports for inspiration.
Mitchell did his best and most original research on phantom limbs. Before his time, relatively few people admitted to them, since they risked being pegged as loopy. A more sympathetic Mitchell determined that 95 percent of his amputees experienced ghost limbs. Interestingly, though, the distribution of phantoms wasn’t equal: patients felt upper-body phantoms more vividly than lower-body phantoms, and felt phantoms in the hands, fingers, and toes more acutely than phantoms in the legs or shoulders. And while most men’s phantoms were paralyzed—frozen into one position—some soldiers could still “move” their phantoms voluntarily. One man would raise his phantom arm instinctively, to grab for his hat, whenever a gust kicked up. Another man missing a leg kept waking up at midnight to use the privy; groggy, he’d swing the phantom leg onto the floor and tumble.
Mitchell also probed phantom pain. Cramps or sciatica might race up and down the phantom, in waves lasting a few minutes. Less acute, but possibly more maddening, people’s phantom fingers or feet would start itching—itches impossible to scratch. Stress often exacerbated the discomfort, as did yawning, coughing, and urinating. Perhaps most important, Mitchell determined that if a soldier had felt a specific pain right before his amputation—like fingernails digging into his palm, a common result of muscle spasms—that same pain often got “stamped” into his nerves, and would persist for years afterward in the phantom.
To explain where phantoms came from, Mitchell suggested a few interrelated theories. His patients’ stumps often had raised growths on them where the underlying nerves had been severed. These “buttons” proved quite sensitive to the touch; they prevented many men from wearing prosthetics. Mitchell deduced from this touchiness that the nerves beneath must still be active—and still pinging the brain. As a result, part of the brain didn’t “know” the limb had gone AWOL. As further proof, Mitchell cited a case where he’d actually resurrected a patient’s phantom. This man had stopped feeling his phantom arm years before (as sometimes happened), but when Mitchell applied an electric current to the stump buttons, the man felt his former wrist and fingers suddenly materialize at the end of his stump—exactly as George Dedlow had at the séance. “Oh, the hand, the hand!” the man hollered. This indicated that the brain did indeed take cues from the stump.
Mitchell also implicated the brain itself in phantom limbs, a crucial development. Many a veteran, despite losing his dominant hand decades before, kept eating meals and writing letters with that hand in his dreams. Unlike stump irritation, this was a purely mental phenomenon and therefore must have its origins within the brain. Even more arresting, Mitchell discovered that some people who’d lost a hand or leg in infancy, and therefore had no memory of it, nevertheless experienced phantoms. Mitchell concluded from these cases that the brain must contain a permanent mental representation of the full body—a four-limbed “scaffold” stubbornly resistant to amputation. The brain’s private metaphysics, then, trumped physical reality.
Later work by other scientists confirmed and built upon Mitchell’s insights. For instance, Mitchell focused on how preamputation pain or paralysis can carry over into the phantom, but later scientists found that less pernicious sensations can be stamped onto the ghost as well. Some amputees feel phantom wedding bands and Rolexes, and people whose arthritic knees or knuckles allowed them to sense impending thunderstorms can often pull off the same trick with their phantoms. Moreover, neuroscientists have confirmed Mitchell’s guess that the brain contains a hardwired scaffold of the full body, since children born without arms or legs sometimes still feel phantoms. One girl born without forearms did arithmetic in school on her phantom fingers.
Doctors have also catalogued phantoms in brave new places. Dental extractions can produce phantom teeth. Hysterectomies can produce phantom menstrual cramps and labor pains. After colorectal procedures, people might feel phantom hemorrhoids and bowel movements and rumbling phantom flatulence. There are also phantom penises, complete with phantom erections. Most phantom penises arise after penile cancer or accidents with shrapnel that most of us prefer not to think about. But unlike phantom limbs—which are often frozen into claws, and excruciating—most men find a phantom penis pleasurable. And they’re so realistic that even decades after the penis is shorn, some men still walk a little funny when they get aroused. Heck, some men’s phantom penises lead to real orgasms. All this showed that quite a few sensations and emotions in the brain can be tied up with phantoms.* The work furthermore helped shift the focus of phantom limbs away from the stump and toward the brain itself.
Although Mitchell made phantom limbs an object of legitimate scientific study, this knowledge didn’t readily translate into treatments. For most of the twentieth century, in fact, no different than in Mitchell’s day, doctors merely fit amputees with prosthetics and, if the phantom pain got bad, plied them with opiates. But in the 1990s phantom research went through a renaissance, as neuroscientists realized that it provides a unique glimpse into the brain’s movement centers and especially into brain plasticity.
The brain’s primary movement center is the motor cortex, a strip of gray matter that starts near your ears and runs to the top of your head. It sends out the commands that spur the spinal cord to move your muscles. On its own, however, the motor cortex can produce only crude movements, like kicks and lunges. Think of a bucking bronco—powerful, but lacking grace. Synchronized movements actually arise from two adjacent regions, the premotor cortex and supplementary motor area. In essence, these two regions coordinate simple movements into something more balletic. To change analogies, they play the motor cortex like a piano, pressing different areas in quick succession to produce complex chords and arpeggios of movement—walking, for instance, requires different muscle groups to contract with a precise amount of force at different moments. Toddlers stumble so much in part because their brains still hit false notes.
To execute a complicated movement, the motor areas also need feedback from the muscles at each stage, to ensure that their commands have been carried out properly. Much of this feedback is provided by the somatosensory
cortex, the brain’s tactile center. You can think about the somatosensory cortex as the motor cortex’s twin. Like the motor cortex, it’s a thin, vertical strip; they in fact lie right next to each other in the brain, like parallel pieces of bacon. Both strips are also organized the same way, body part by body part; that is, each strip has a hand region, a leg region, a lips region, and so on. In effect, then, the motor cortex and somatosensory cortex each contain a “body map,” with each body part having its own territory.
In some ways this body map is straightforward; in other ways it’s not. For example, just like on your body, the map’s hand region lies right next to the arm region, which lies right next to the shoulder region, and so on. But in other spots, the topography is scrambled. In particular, the hand territory also borders the face territory, even though the hand itself doesn’t border the face. Just as randomly, your foot territory nestles against the crotchal region.
The brain’s body maps also contain another counterintuitive feature. Despite what you might think, big body parts don’t need big patches of gray matter to function. Legs, for instance, although powerful, don’t require complicated instructions to jump or kick, and they’re not very sensitive to touch, either. As a result, these big burly parts get by with minuscule, Luxemburg-sized territories on the touch and movement maps. The lips, tongue, and fingers, meanwhile, engage in intricate movements such as speaking and handling tools, and therefore need Siberia-sized tracts of neurons. Some body parts, in other words, are magnified on the maps. (This explains why amputee soldiers felt missing fingers more than missing hands, and missing hands more than missing arms: our brains pay more attention to fine-motor structures.)
With all that in mind, consider what happens when a hand is amputated. First, a huge territory on the brain map goes black. It would be like watching the United States from space at night, with all its patches of sprawling suburban illumination, and seeing the power grid in Chicago fail. The key point, however, is that this spot doesn’t stay blank. Because the brain is plastic, adjacent areas can colonize the hand region and use its neurons for their own ends. With a missing hand, it’s usually the resource-hungry face territory that encroaches.
This encroachment happens quickly, sometimes within days, and happens over long neural distances, up to an inch. For these reasons scientists suspect that colonization cannot simply involve new neuron tendrils sprouting up and “invading” empty territory. Instead, the colonizer probably fires up preexisting circuits that were lying dormant. Again, the brain has bazillions of neural circuits running every which direction, and some of these tracts happen to start in the face territory and spill over into the adjacent hand region. Most of the chatter on these circuits is irrelevant to the hand, so the hand region mutes them. But when the hand area falls quiet, it loses the ability to resist. The nearby cheek and lip areas suddenly face no opposition and can take over.
Still, as every colonial power in history has learned, occupying a territory is different from assimilating it. Too many “hand circuits” exist to reprogram them all, and the hand territory always retains a vestige of its identity. As a result, the new face circuits and the old hand circuits overlap and intermingle, and both can end up firing simultaneously.
What does this all mean on a higher scale, the scale of perception? It means that for some amputees, touching or moving their faces will summon up sensations in their missing hands. If an amputee strokes his cheek, for instance, he might feel his missing thumb being brushed. If he whistles or chews gum, the index finger twitches. If he pops a chin zit, the pinkie feels the squeeze. Even those people who don’t consciously register the dual sensations will still have signals intermixing in the brain. The net result is that face sensations keep stoking the mental memory of the hand and keep stirring the phantom awake.
(Similarly, because the foot and genital territories border each other on the brain map, when the lower leg disappears, the genital spot can take over. Sure enough, some lower-limb amputees feel their phantom feet most insistently during sex. A few even report feeling orgasms quaking all the way down to their phantom tiptoes. And, like striking a bigger tuning fork, this expansion of the orgasmic territory gives them a proportionately greater pleasure.*)
Scientists gained another crucial insight into phantom limbs from a series of almost comically low-tech experiments conducted by a neurologist in Southern California named V. S. Ramachandran. Ramachandran had a patient named D.S. who’d lost his left arm after a motorcycle wreck and had experienced severe phantom cramping ever since. To treat him, Ramachandran took an open-topped cardboard box and mounted a mirror inside. The mirror divided the box’s interior into two parts, a left chamber and a right chamber. Ramachandran cut a hole in the box on either side of the mirror and had D.S. slide his right hand into the right chamber. (D.S. also imagined sliding his phantom left hand into the other chamber.) The crucial point is that the reflective surface of the mirror faced right. So when D.S. inserted his hand into the hole and glanced down, it looked like he had two intact hands again.
Ramachandran had D.S. close his eyes and start swinging his hands back and forth symmetrically, like someone conducting the philharmonic. At first nothing happened. The phantom stayed frozen, mute. D.S. then opened his eyes and repeated the motion while looking in the mirror. That’s when the orchestra burst into song. As his hands swayed back and forth, his phantom fingers unfurled for the first time in a decade. His cramps abated, his rigid wrists went loosey-goosey. “My god!” he yelled, and began jumping up and down. “My arm is plugged in again.”
Over the next few years many more amputees would share that same glee in Ramachandran’s office. The “mirror box” looked hokey, sure. But something about seeing a lost limb in motion unfroze the phantom in people’s minds. Again, we dedicate loads of brainpower to vision, and we implicitly trust sight over our other senses—seeing is believing. So when the eyes see a limb moving again, the brain believes it can.
Based on this and other insights, scientists like Ramachandran have now sketched out an explanation for why phantoms exist and why they often snarl with pain. Because the brain has a hardwired mental scaffold of the body, it expects to find four full limbs at all times—that’s its default setting, and that’s why even people born without limbs can experience phantoms. Moreover, the reality of phantoms is reinforced when the brain keeps receiving spurious signals, both from the inflamed stump and especially from any greedy brain territories that colonize an empty neural landscape. All this activity fools the brain into thinking that the hand or leg still exists. So the brain keeps sending motor signals down there, and armless men keep grabbing for their hats in gusty weather.
That explains the sensation. The paralysis and pain arise for different reasons. If the limb was paralyzed before the amputation, the phantom is usually paralyzed afterward, too. But even people who can “move” their phantoms at first often lose that ability later. Remember that the brain, after it sends off a movement command, looks for sensory feedback to confirm that movement took place. Arms that don’t exist obviously don’t provide this feedback. So, over time, most people’s brains conclude that the phantom is paralyzed.
Pain can get stamped into a phantom limb just like paralysis does, when preamputation aches and pains carry over. But motor commands can also exacerbate the pain. Because an AWOL limb can’t respond to motor commands, the brain—which hates being disobeyed—often ratchets them up: a failed command to squeeze the left hand gets transformed into squeeze hard, then squeeze harder, then squeeze like hell. This causes pain for two reasons. One, pain signals alert the body that something is wrong, and with this mismatch between motor commands and sensory feedback, there’s clearly something amiss here. Second, brutal commands like those were usually accompanied by pain in the past: your brain learned, for instance, that clenching your fist caused your fingernails to gouge your palm. Eventually the hand-clenching circuit and the pain circuit got wired together. As a result, whenever the brain tr
ies to rouse the phantom with a hard squeeze, pain sensors can’t help but fire.
The mirror box, however, slits the neural Gordian knot. It resolves the mismatch between the motor and sensory systems, and because the brain literally sees its commands being obeyed, it can stop sending out orders to squeeze and squeeze harder. In the sudden tranquility, the pain seeps away. To be sure, at first the relief lasts only a few hours before the phantom seizes up again. Not everyone finds relief from mirror therapy, either. But those who do, and who practice with the mirror box, can see profound improvement over time as their brain maps reorganize. In many cases the pain all but disappears. (You can think about this decoupling as the converse of neurons that fire together wire together. Here, neurons out of sync fail to link.) And in some cases the phantom itself vanishes. After Ramachandran’s first patient, D.S., drilled with the mirror box for several weeks, he felt his phantom left arm shrinking inch by inch, “telescoping” up into his shoulder. Finally, just a nub of sensation remained. Ramachandran called this the first successful amputation of a phantom limb.