In 1996 neuroscientist Giacomo Rizzolatti and his co-workers at the University of Parma in Italy published some remarkable findings. They had run an experiment to record electrical activity from neurons specialized for hand movement in two pigtail macaques. As anticipated, these neurons fired when the animals reached for peanuts placed in front of them. What was entirely unexpected, however, was that these same neurons fired when a scientist in the lab reached for the nuts instead. The monkey remained stationary. Nevertheless, watching the scientist move had activated motor areas in the macaques brain, just as if the animal had carried out the action itself.

Using functional magnetic resonance imaging (fMRI), Rizzolatti and his colleagues soon documented the same phenomenon in humans and dubbed the responsible nerve cells "mirror neurons" [see "A Revealing Reflection," by David Dobbs; Scientific American Mind, April/May 2006]. These cells look like any other neuron but boast a surprising double function: they become active during any type of directed behavior--chewing food, throwing a ball, performing a dance--whether we do it ourselves or simply watch someone else do it. Indeed, our conscious brain generates an inner simulation of sorts when we follow the actions of another person. Mirror neurons are presumed to be abundant in brain regions responsible for planning and initiating actions, including the primary motor cortex, the premotor cortex and supplementary motor areas.

Since Rizzolatti's discovery, other scientists have revealed that mirror neurons reflect not only the actions of other people but their intentions and emotions as well. The discovery is offering scientists new insight into, among other things, human empathy, language evolution and theories of mind. In addition, mirror neurons may help explain certain neurological conditions. For example, some evidence suggests that autistic children may suffer from mirror neuron deficiencies, leaving them unable to intuit others' emotional states. Our own work indicates that the mirror system can be enlisted to expedite the rehabilitation of hemorrhagic stroke patients.

Monkey See, Monkey Heal
In 2001 another research team at Parma, led by one of us (Buccino), used fMRI to track brain activity in people watching video sequences showing mouth, hand or foot movements. As it turned out, when the subjects watched a mouth move, the part of their brain responsible for controlling their mouth lit up. Likewise, observing hand and foot images engaged the corresponding brain regions. These responses remained below the action threshold--the subjects did not actually move--but they matched the brain responses to video exactly. Given these findings, we speculated that patients who had suffered a cerebral hemorrhage might regain lost movements more readily if, as part of their therapy, they watched others coordinate these actions.

During physical therapy, brain regions near the site of damage do often take on lost functions, but it is a gradual process. It seemed logical that this transfer might happen faster if the neurons in question could rehearse their new role. To test the idea, we recruited stroke patients participating in a 40-day rehabilitation program at the University Medical Center of Schleswig-Holstein in Luebeck, Germany. We asked our subjects to watch a six-minute film showing a series of movements--stretching arms, opening hands, grasping apples and so forth [see box on opposite page]. We then asked them to try to imitate the actions. We found that indeed these patients improved their motor abilities considerably faster than those in the control group, who did not watch the videos.

In a recent follow-up study, we documented the same gains among 22 stroke patients who had difficulty using their arms and hands. Physical rehabilitation progressed more rapidly when they watched short films of people demonstrating everyday hand movements before and after each therapy session. With the aid of fMRI, we were able to show that as this motor improvement occurred the areas of the cortex involved became increasingly active. The video observations apparently strengthened those brain areas responsible for planning movements. The inner simulation made it easier for the subjects to carry out the real motions.

Armchair Athletes?
In fact, our mirror neurons react to many movements, but research suggests that their activity level depends on how familiar we are with the actions we witness. Learning to coordinate completely new actions--playing a new sport, for example--demands far more conscious control than is required by a routine task. Buccino and his colleagues showed subjects various video clips depicting a person, an ape or a dog eating or communicating: the person moved his mouth as if speaking, the ape pursed its lips and the dog barked. Of interest, chewing motions universally activated mirror neurons in the subjects. When it came to communication, however, this activation occurred only when the moving lips belonged to a fellow human.

It may well be that our mirror neurons react only to actions that are part of our own motor repertoire. So, too, observers apparently need to understand another's intentions to activate their premotor centers. Neuroscientist Marco Iacoboni of the University of California, Los Angeles, performed an interesting experiment, asking subjects to watch short films of people using identical motions to different ends. In one sequence, a person picked up a cup and drank from it. In another, he picked it up and washed it. Yet another clip showed the motion itself, without any purpose, and another showed simply a set of plates and cups, without any movement. Iacoboni found that neither the motor action nor the environment alone activated the mirror neurons as strongly as the combination of the two did. Motor activity lacking an obvious purpose may be less effective in helping someone relearn a particular movement.

Further research may reveal that mirror neurons lie at the root of many fundamental human traits--from the way we learn to the way we develop distinct cultures. In the meantime, stroke patients stand to gain enormous benefit from harnessing this remarkable system.