Think of what this means. Anytime you watch someone doing something, the neurons that your brain would use to do the same thing become active—as if you yourself were doing it. If you see a person being poked with a needle, your pain neurons fire away as though you were being poked. It is utterly fascinating, and it raises some interesting questions. What prevents you from blindly imitating every action you see? Or from literally feeling someone else’s pain?
In the case of motor mirror neurons, one answer is that there may be frontal inhibitory circuits that suppress the automatic mimicry when it is inappropriate. In a delicious paradox, this need to inhibit unwanted or impulsive actions may have been a major reason for the evolution of free will. Your left inferior parietal lobe constantly conjures up vivid images of multiple options for action that are available in any given context, and your frontal cortex suppresses all but one of them. Thus it has been suggested that “free won’t” may be a better term than free will. When these frontal inhibitory circuits are damaged, as in frontal lobe syndrome, the patient sometimes mimics gestures uncontrollably, a symptom called echopraxia. I would predict, too, that some of these patients might literally experience pain if you poke someone else, but to my knowledge this has never been looked for. Some degree of leakage from the mirror-neuron system can occur even in normal individuals. Charles Darwin pointed out that, even as adults, we feel ourselves unconsciously flexing our knee when watching an athlete getting ready to throw a javelin, and clench and unclench our jaws when we watch someone using a pair of scissors.1
Turning now to the sensory mirror neurons for touch and pain, why doesn’t their firing automatically make us feel everything we witness? It occurred to me that perhaps the null signal (“I am not being touched”) from skin and joint receptors in your own hand block the signals from your mirror neurons from reaching conscious awareness. The overlapping presence of the null signals and the mirror-neuron activity is interpreted by higher brain centers to mean, “Empathize, by all means, but don’t literally feel that other guy’s sensations.” Speaking in more general terms, it is the dynamic interplay of signals from frontal inhibitory circuits, mirror neurons (both frontal and parietal), and null signals from receptors that allow you to enjoy reciprocity with others while simultaneously preserving your individuality.
At first this explanation was an idle speculation on my part, but then I met a patient named Humphrey. Humphrey had lost his hand in the first Gulf War and now had a phantom hand. As is true in other patients, whenever he was touched on his face, he felt sensations in his missing hand. No surprises so far. But with ideas about mirror neurons brewing in my mind, I decided to try a new experiment. I simply had him watch another person—my student Julie—while I stroked and tapped her hand. Imagine our amazement when he exclaimed with considerable surprise that he could not merely see but actually feel the things being done to Julie’s hand on his phantom. I suggest this happens because his mirror neurons were being activated in the normal fashion but there was no longer a null signal from the hand to veto them. Humphrey’s mirror neuron activity was emerging fully into conscious experience. Imagine: The only thing separating your consciousnesses from another’s might be your skin! After seeing this phenomenon in Humphrey we tested three other patients and found the same effect, which we dubbed “acquired hyperempathy.” Amazingly, it turns out that some of these patients get relief from phantom limb pain by merely watching another person being massaged. This might prove useful clinically because, obviously, you can’t directly massage a phantom.
These surprising results raise another fascinating question. Instead of amputation, what if a patient’s brachial plexus (the nerves connecting the arm to the spinal cord) were to be anesthetized? Would the patient then experience touch sensations in his anesthetized hand when merely watching an accomplice being touched? The surprising answer is yes. This result has radical implications, for it suggests that no major structural reorganization in the brain is required for the hyperempathy effect; merely numbing the arm is adequate. (I did this experiment with my student Laura Case.) Once again, the picture that emerges is a much more dynamic view of brain connections than what you would be led to believe from the static picture implied by textbook diagrams. Sure enough, brains are made up of modules, but the modules are not fixed entities; they are constantly being updated through powerful interactions with each other, with the body, the environment, and indeed with other brains.
MANY NEW QUESTIONS have emerged since mirror neurons were discovered. First, are mirror-neuron functions present innately, or learned, or perhaps a little of both? Second, how are mirror neurons wired up, and how do they perform their functions? Third, why did they evolve (if they did)? Fourth, do they serve any purpose beyond the obvious one for which they were named? (I will argue that they do.)
I have already hinted at possible answers but let me expand. One skeptical view of mirror neurons is that they are just a result of associative learning, as when a dog salivates in anticipation of dinner when she hears her master’s key in the front door lock each evening. The argument is that every time a monkey moves his hand toward the peanut, not only does the “peanut grabbing” command neuron fire, but so does the visual neuron that is activated by the appearance of his own hand reaching for a peanut. Since neurons that “fire together wire together,” as the old mnemonic goes, eventually even the mere sight of a moving hand (its own or another monkey’s) triggers a response from the command neurons. But if this is the correct explanation, why do only a subset of the command neurons fire? Why aren’t all the command neurons for this action mirror neurons? Furthermore, the visual appearance of another person reaching toward a peanut is very different from your view of your own hand. So how does the mirror neuron apply the appropriate correction for vantage point? No simple straightforward associationist model can account for this. And finally, so what if learning plays a role in constructing mirror neurons? Even if it does, that doesn’t make them any less interesting or important for understanding brain function. The question of what mirror neurons are doing and how they work is quite independent of the question of whether they are wired up by genes or by the environment.
Highly relevant to this discussion is an important discovery made by Andrew Meltzoff, a cognitive psychologist at the University of Washington’s Institute for Learning and Brain Sciences in Seattle. He found that a newborn infant will often protrude its tongue when watching its mother do it. And when I say newborn I mean it—just a few hours old. The neural circuitry involved must be hardwired and not based on associative learning. The child’s smile echoing the mother’s smile appears a little later, but again it can’t be based on learning since the baby can’t see its own face. It has to be innate.
It has not been proven whether mirror neurons are responsible for these earliest imitative behaviors, but it’s a fair bet. The ability would depend on mapping the visual appearance of the mother’s protruding tongue or smile onto the child’s own motor maps, controlling a finely adjusted sequence of facial muscle twitches. As I noted in my BBC Radio Reith Lectures in 2003, entitled “The Emerging Mind,” this sort of translation between maps is precisely what mirror neurons are thought to do, and if this ability is innate, it is truly astonishing. I’ll call it the “sexy” version of the mirror-neuron function.