3
Brain Crashes
Men in a war
If they’ve lost a limb
Still feel that limb
As they did before
He lay on a cot
He was drenched in a sweat
He was mute and staring
But feeling the thing
He had not
—Suzanne Vega
In her song “Men in a War,” Suzanne Vega captures the paradox of phantom limb syndrome: people who have had a limb amputated often have the vivid feeling that their limb is still in place. Amputees sometimes feel their lost arm or leg as being frozen in a fixed position, and, because these sensations are so authentic, may take the position of their nonexistent limb into account as they move. A man who felt that his amputated arm was permanently extended from his side would go through a doorway sideways so as not to bump his nonexistent arm; another would attempt to walk with his nonexistent leg.1 In many cases phantom sensations are not experienced merely as the presence of the lost limb, but as an all-too-real source of pain. Phantom pain is as genuine and debilitating as any other form of pain. Tragically, however, there is little hope of soothing it at the perceived point of origin.
The unbroken succession of wars throughout human history has ensured that human beings have been losing limbs for quite some time. Consequently, there are a number of historical accounts of people living with the ghosts of their former limbs. Yet it was only in the later half of the twentieth century that the medical establishment started to accept phantom limb syndrome as a neurological disorder. One can hardly blame physicians and laymen alike for assuming that accounts of phantom limbs reflected some sort of hysteria or longing for the missing limb. What could have been more counterintuitive than the concept of feeling something that no longer exists? As the term “phantom” itself suggests, the condition would seem to beg for a metaphysical explanation. Indeed, phantom limb syndrome has led to my favorite argument in support of the concept of a soul. The eighteenth-century British Admiral Lord Nelson who lost his right arm in battle experienced vivid phantom sensations, which he took as proof of the existence of a soul. His surprisingly cogent reasoning was that if the specter of his arm could persist after it is lost, so could the person.2
As counterintuitive as phantom limb syndrome may be, there is perhaps an even more peculiar syndrome that also relates to body awareness. After certain types of cortical trauma (often a stroke) people may fail to acknowledge part of their body as being their own. The limb itself is fully functional: the muscles and the nerves connecting the arm or leg to the spinal cord are intact. This is a rare and generally transient form of body neglect that has been referred to as somatoparaphrenia.3 If a doctor touches the affected arm of a patient with this syndrome, she will not report any conscious sensation of being touched; nevertheless, she may reflexively move her arm in response to a painful stimulus. When asked about that object resting on the table she will report that it is an arm, not her arm; when questioned as to whose arm it is, she may report not knowing or even claim that it belongs to someone else. In one case a patient who believed that her left hand was that of the doctor’s, commented: “That’s my ring. You have my ring, doctor.” In his book The Man Who Mistook His Wife for a Hat, Oliver Sacks tells of a patient who was in the hospital after having suffered a stroke. While in the hospital, the patient had fallen off his bed. He later explained that when he awoke, he had found a disembodied leg in bed with him and assumed it was some sort of prank. So he, understandably, shoved the leg off the bed. In the processes he also ended up on the floor—the leg was his own.4
Phantom limb syndrome and somatoparaphrenia are in a sense mirror images of each other. In one, people perceive a nonexistent limb, and in the other people deny the presence of a perfectly good limb, physically speaking. Together these conditions call for a deeper understanding about the nature of the mind, and about what it actually means to feel one’s own body.
THE ILLUSION OF BODY
We all know the names of famous painters who dazzle our sense of vision. We revere the musicians who seduce our sense of hearing. You can probably name a famous cook and, if not a famous perfumer, at least a famous perfume. This covers four of the five senses, leaving us with the sense of touch. There is no Picasso, Mozart, or even Thomas Keller or Ernest Beaux (creator of Chanel No. 5) of touch. There are many reasons for this, including that we cannot store touch on a DVD, in an mp3 file, in the fridge, or in a bottle. Tactile stimuli cannot be experienced from afar. Touch is the most intimate and personal of our senses. Although mostly neglected in the art world, touch has an impressive palette of sensations to sample from: painful, cold, warm, tickly, smooth, rough, and itchy and scratchy. Touch provides an emotional range that our other senses can only aspire to: from the nonnegotiable pain of a stubbed toe to the sensual bliss of sex.
This perceptual breadth comes courtesy of the brain’s somatosensory system, which is not only responsible for the sense of touch per se, but for the perception of the body itself. When the neurologist taps my knee, I do not feel only the rubber hammer, I feel that it hit my knee! It’s my knee and no one else’s, and it is my left not my right knee. The somatosensory system does not simply generate a report of whether the hammer was soft or hard, cold or warm, but localizes these sensations to the specific part of the body that was in contact with the object that triggered them. At the periphery, the somatosensory system includes sophisticated touch detectors distributed throughout the body. Mechanoreceptors embedded in our skin can detect tiny deformations of our skin. Unlike the touchpad of a laptop, the brain has multiple types of detectors at every location; some respond to a light touch, others detect vibrations or limb position, and others respond to temperature or painful stimuli.
If you dig your fingernails into the palm of your hand, pain receptors are activated and you feel the pain in your palm. Someone with phantom pain may live with this same feeling even though he has no fingers and no palm. Is his pain an illusion?
Phantom sensations reveal something fundamental about body awareness: it is not that phantom limbs are an illusion, rather it is the feeling of our actual limbs that is an illusion. When you stub your toe, pain receptors send signals from your toe to neurons in your brain that ultimately induce the sensation of pain. But you do not feel the pain as occurring in your head! Like a film projector that casts images on a distant screen, your brain projects the pain back to your toe. This projection is perhaps the most astonishing illusion of all: while our body is of course real, the fact that we feel it as residing outside the confines of our cranium is an illusion.
A ventriloquist can create the compelling illusion that it is a dummy insulting everyone in the room. This is achieved by manipulating his voice and providing misleading visual cues: minimizing his own lip movements while exaggerating those of the dummy. If an absentminded ventriloquist proceeded with his routine but had forgotten his most important prop, the act would be rather transparent and unamusing. Like a ventriloquist without a dummy, in absence of a limb the illusory nature of our body ownership is rapidly revealed. Phantom limbs are simply the normal body illusion gone awry because a sensation is projected to a limb that no longer exists.
To better understand how this could come to be it will be helpful to examine the flow of information from periphery to brain in more detail. When someone gently touches your finger with a Q-tip, receptors in the fingertip generate action potentials (the bioelectric “waves” that carry the output signals of neurons) that travel along the axons of sensory neurons toward the spinal cord. These axons synapse on neurons in the spinal cord, which in turn shuttle information to specific parts of the cortex. In the same manner that there are areas in the brain devoted to processing visual or auditory stimuli, parts of the cortex are dedicated to processing the information arriving from the sensory receptors distributed throughout the body. The first of such areas is sensibly named the primary somatosensory cortex. By recording the electrical a
ctivity of the somatosensory cortex of animals while touching different parts of the body, neuroscientists in the late 1930s discovered a map of the body laid out in the brain. Around the same time the Canadian neurosurgeon Wilder Penfield came to the same conclusion while performing surgeries in patients with epilepsy. Since the brain itself does not have sensory receptors, these surgeries could be performed in awake patients under local anesthesia. This allowed Penfield to take the opposite approach of those scientists conducting experiments in animals; rather than record the electrical activity of neurons in response to touch, he stimulated them electrically and asked the patients what they felt. Answers included: “My mouth feels numb” or that the patient felt a jerky sensation in his left leg. From Penfield’s experiments we now know that if one were to draw on the cortex the body part that each zone of the somatosensory cortex represented, one would end up with a little man referred to as the somatosensory homunculus (Figure 3.1). The drawing of the man, however, would be severely distorted: some body parts, such as the fingers, would be disproportionately large. In other words, there is proportionally more cortical space allocated to the fingers in comparison to larger parts of the body, such as the thighs. Although the map is distorted, the neighboring areas of the body are represented in neighboring areas of the cortex, which is to say, the map is topographic. Later studies revealed that there is actually not a single map, but multiple maps of the body side by side; each specialized for different modalities of touch, such as feeling a light touch or vibrations.5
Figure 3.1 Somatosensory cortex: A map of the human body is laid out on the surface of the somatosensory cortex. The map is said to be topographic because adjacent areas of the cortex represent adjacent surfaces of the body. Note that large areas of the cortex can be “allocated” to relatively small parts of the body, such as the fingers (Bear et al., 2007; modified with permission from Wolters Kluwer.)
Penfield’s experiments beautifully demonstrated that sensations could be elicited by direct brain stimulation, even though all the normal input channels had been bypassed. For the most part, however, the sensations elicited in Penfield’s and subsequent research in humans were rather “fuzzy”—patients generally would not mistake the feeling produced by direct brain stimulation with someone actually touching their body. But in principle it should be possible to fool the brain if we knew precisely which neurons to stimulate. Whether direct brain stimulation can substitute for the real physical stimulation has been addressed in some clever Matrix-like experiments performed in monkeys. The Mexican neuroscientist Ranulfo Romo trained monkeys to perform a task in which they had to judge the frequency of a vibrating metal “probe” placed on the tip of their index finger.6 During each trial the monkey first received a “reference” stimulus, for example, the probe might vibrate at 20 cycles per second. Next the monkey received the “comparison” stimulus, in which the probe vibrated at a lower or higher frequency. The monkey was trained to press one of two buttons to indicate whether the second stimulus was lower or higher in frequency than the first. Monkeys performed this task well, routinely discriminating between stimuli vibrating at 20 and 30 cycles per second. Key to this experiment was the fact that the monkeys had electrodes implanted in their brains, precisely in the area of the primary somatosensory cortex devoted to processing information from the finger stimulated by the probe. These electrodes allowed the experimenters to artificially stimulate the same neurons that would normally be activated by the probe at the tip of the monkey’s finger.
Romo and his colleagues wondered if they could fool the monkey into doing the task “virtually”—what would happen if after training a monkey to compare the frequency of real-world physical stimuli, they used direct electrical stimulation of the somatosensory cortex? In these virtual trials the first event was again the metal probe at the finger; however, rather than apply a second stimulus at a different frequency to the finger of the animal, they applied a brief series of electrical pulses directly to the monkey’s brain through the implanted electrodes—completely bypassing the peripheral somatosensory system. Since the electrical stimulation could also be applied at specific frequencies, the experimenters were able to ask the monkey to compare the frequency of real and virtual stimuli. If the monkey did not feel this second stimulus at all, presumably he would not complete the task or he would guess. On the other hand, if the physical and direct-brain stimulation were in some way equivalent, he would continue performing the task with a high success rate. Amazingly the monkeys continued to perform the task, correctly comparing the frequency of the physical and virtual stimuli, just as they had with two physical stimuli. We, of course, do not know if the physical and virtual stimuli felt the same to the monkey, or if the monkey thought, “Whoa! I never felt something like that before.” Nevertheless, these experiments confirm that relatively primitive forms of direct brain stimulation can functionally substitute for real stimuli.
Knowing that the sensation of touch, or of the feeling of one’s arm, can be achieved solely by the activation of neurons in the brain allows us to understand how phantom sensations might arise. One of the first scientific hypotheses of the cause of phantom limb sensations was that they were a result of the regrowth of the severed nerves at the site of the amputation. This is a logical hypothesis since the distal ends of cut nerve fibers can indeed sprout into the remaining part of the limb, referred to as the stump. In this manner the nerves that used to innervate the hand could innervate the stump and send signals to the central nervous system, which would continue to interpret these signals as if the lost limbs were still present. This hypothesis was behind one of the early treatments for phantom pain, which was to surgically cut the nerves in the stump or as they enter the spinal chord. This procedure was beneficial in some cases, but generally did not provide a permanent cure for phantom pain.
Today scientists agree that in many cases phantom sensations do not reflect an abnormal signal from the nerves that used to innervate the missing limb, but are caused by changes that take place within the brain. Specifically, as in the monkey experiment in which direct brain stimulation appeared to substitute for a real stimulus, neurons in the brain that would normally be activated by the arm continue to fire, driving the perception of a phantom limb.7 But a question remains: why would the neurons in the brain that are normally driven by the limb continue to be active even when the limb is long gone? The answer to this question provides important insights into how one of the most powerful features of the brain, its ability to adapt, can become a brain bug.
NEURONS ABHOR SILENCE
Like space on a computer chip, cortical real estate is an extremely valuable and limited resource. So how does the brain decide how much cortical area should be allocated to each part of the body? Does a square centimeter of skin in the small of your back deserve as much cortical computational power as the square centimeter of skin on the tip of your index finger?
One might guess that the amount of cortical area devoted to different body parts is genetically determined and, indeed, to some extent cortical allocation is hardwired. For instance, per square centimeter there are fewer sensory fibers innervating your back than your hand (your back is a low-resolution input device, while your fingertips have high input resolution). This is a function of our neural operating system as defined in the Introduction. But such a predestined strategy alone would be an overly rigid and ill-conceived evolutionary design. The elegant (and somewhat Darwinistic) solution to the cortical allocation problem is that different body parts have to fight it out: the most “important” parts of the body are awarded more cortical real estate.
If you close your eyes, and ask someone to touch a finger on your hand you can easily report which finger was touched. If this person is willing to repeat this experiment by touching one of your toes, you may find that you are not so sure which toe was touched, and may even get the answer wrong. This is in part because, in all likelihood, your brain devotes more of the somatosensory cortex to your fingers than t
o your toes. The amount of cortex allocated to each part of the body contributes to how precisely we can localize the point of contact, and how easily we can determine if we were touched by a pin, a pen, or someone’s finger. One can envision that a seamstress, surgeon, or a violinist, compared to a professor, lawyer, or a cymbalist, would benefit immensely from having a larger amount of somatosensory cortex allocated to processing information from the fingertips. Furthermore, if you were to try to learn Braille, it would be very convenient to perform an upgrade to the part of your somatosensory cortex devoted to your fingertips. It makes sense to be able to allocate different amounts of cortical surface to the fingertips on a case-by-case manner along the lifespan of an individual. It turns out that the brain can dynamically allocate cortical resources depending on computational need; that is, the cortical area representing different parts of the body can expand or contract as a result of experience.
For many decades, it was thought that the somatosensory maps observed in humans and other animals were rigid throughout adult life. But this view was overturned in the early eighties by a series of groundbreaking studies by the neuroscientist Michael Merzenich, at the University of California in San Francisco. Merzenich and his colleagues demonstrated that cortical maps were “plastic”—like sand dunes in the desert the cortex was constantly being resculpted.8 Merzenich first showed that after cutting one of the nerves innervating the hand, the somatosensory cortex of monkeys reorganized—the map changed. Neurons in cortical areas that originally responded to the affected hand initially became unresponsive, but over the course of weeks and months. The neurons that were deprived of their normal input as a result of the nerve transection “switched teams”—they became progressively more responsive to other parts of the body. More importantly, subsequent studies showed that when monkeys were trained to use a few of their fingers for tactile discrimination over the course of months, the areas of the somatosensory cortex that represented those fingers expanded. It is as if there was some manager in the brain that went around redistributing precious cortical space to the parts of the body that needed it the most.
Brain Buys Page 8