29_05Mirror neurons

Imitation is an advanced process, unique to man.

Prof. Giacomo Rizzolatti, University of Parma, interview

Mirror neurons activate when a monkey takes a certain action, but also when observing someone doing the same action. At first, honestly, we did not attach much importance to it, but the phenomemon was repeated too frequently and we began a series of checks and saw that it was a very real phenomenon. The revolutionary aspects was the opportunity to immediately connect a visual acquaintance with a motor knowledge.

The classic model of our brain is that we need to do complicated logical deductions. In fact, if you think about it, it is not so. If you see me take this pencil, you’re not going through a big reasoning process: the meaning of taking is understood instantly without a complicated process. Then, the execution is much more complicated than its understanding and its formal aspect. We have a sort of “matching system”, an overlay system between what is seen and what is done. The mirror system was the first demonstration that this neurophysiological  mechanism, which had been posed as a hypothesis, exists.

With the discovery of mirror neurons, learning by imitation seems to be a minor phenomenon…

This is another very important aspect. We have studied the mirror neurons on the monkey, and we thought that the monkey was able to imitate. Speaking with ethologists, we learned that monkeys know how to imitate some facial expressions, but they do not know how to imitate the taking. Meaning that imitation is an advanced process, proper to man, because man has a mirror system even for gestures devoid of apparent purpose. The monkey only includes a movement that has a clear goal: to take food, crush, break. When he sees a meaningless gesture, this gesture is lost in the visual system and does not reach the motor system: the monkey understands that something moves, but does not give it a motor meaning.

It is through the mirror neurons that imitation has become an important topic. In psychology, it was said that the imitator is stupid, and the creator a genius. But we must first imitate and then we can do something new. If you do not know how to imitate, there is no culture. Therefore I now say to teachers: “Let them imitate, teach them how to copy, this is how culture is transmitted”.


One theory says that we are only able to learn what is part of what we believe is possible. We reject the rest…

I honestly do not know. I could say that we learn by imitation whata is part or our motor  heritage, by assembling elementary motor acts. Motor acts that do  not correspond to what we do, do not make the mirror neurons resonate. But your question was more specific…

if we are able to learn only what somehow we already believe to be possible…. If we see something that is part of our motor heritage, we learn, and then we can improve it. In this regard, a great experience was made with dancers: three groups of volunteers, a group of capoeira masters, a classical dancers group and people who had never taken dance lessons were put before a screen on which were presented capoeira steps. Among the masters of this art, we saw greater activation in the mirror neuron system as the other two groups… Another experiment, more recent, shows that in a person who teaches physical exercices, mirror neurons increase always more. So there is a plasticity of the brain to learn. The understanding becomes more experiential, more interior, less cognitive – although cognitive word is somewhat ambiguous. In short, it becomes more phenomenological.

Can you tell us about the sensations and emotions mirror ? For the emotion we experienced with smells (those that work best are foul odors, like rotten eggs). When someone smells rotten eggs, a whole series of zones get active, which are not areas of the mirror system, but a high-level structure called the insula, located deep in the cortex. We felt our students with a rotten eggs smell, and the areas of disgust were activated, with strong emotions, urge to vomit and so on. We previously felt rotten eggs to actors that we had filmed and then we showed the faces of these actors to our students: the same areas are activated in the insula, whether the person felt it, or saw someone else smelling it…

Mirror neurons have been linked with empathy, as a capacity for animal survival. But empathy is also considered as one of the noblest manifestation of human consciousness…

Yes, we have a natural system of co-sympathy. On this subjet, I recommend the excellent work of Tania Singer: she put electrodes on a woman and gave her small shocks; after which she made her see a film in which her fiance suffered the same treatment. The same areas were activated when experiencing pain or when she saw the pain in the other. This means that when we see another suffer, it is as if we suffer ourselves.

So there really is a link between people. All the mirror system indicated that we are connected to others, it is a continual joint-participation, really the opposite of the idea of isolated monads who travel the world. This being said, if we read Adam Smith, he almost predicts what the mirror neurons are since his first book (Theory of moral feelings, 1759), he says more or less the following: “There is no human so selfish who, seeing another suffer, does not feel himself something in his heart.” It is not true that we are born selfish or individualists: we are in constant contact with others and I can’t do what I want, otherwise I disappear and the society dies.

I read that through mirror neurons, we can access someone else’s intention,  its mental state that determined his actions and behaviors. Can you comment ? This is what we study here a lot now, because we believe that there are practical implications. If I take a cup by the handle, I want to drink; if I take it by the top, I want to move it, and so on. In a gesture, there are always two components, the how and the why. When we look how the other does it, we understand quickly what is happening and, in general, we deduce even automatically  why he takes it. Here in Parma, professor Fogassi  discovered that chains of neurons are activated in succession when we do an action. If I take the cup like this, neurons are activated to bring the cup to the mouth, to open the mouth and so on: my intention is manifested by a series of immediate activations, individual motor acts. When I will take the cup, I have already activated the arm movement towards the mouth, the opening of the mouth… there is activation in advance, everything is ready.

And when I see how you take the cup, I understand if you take it to offer it to me or to throw it to my face…

Yes, because the mechanism itself is “mirror”, so you understand my intention. This is a very simple mechanism, but that explains a very complex philosophical concept: intentionality. Children with autism who have a normal intelligence quotient (some autistic have no lack of intelligence, only relationship difficulties) fully understand the gesture, but in them, no activation occurs…. we tested it with gestures such as bringing to mouth. If a normal child sees you grab a chocolate while her system is active, and if you ask her, “Why ?”, she replies “To eat”. The autistic child too, but its system is not activated beacause he only understands rationally.

This is very interesting, because most of the work on autism are based on improving behavior (these children are often destructive, they cry, shout..) then it would be important for them to successfully reacquire the ability to understand others, not just to act because they are either punished or  rewarded. We do not yet know exactly how to do this, but we think of theatrical techniques or martial arts… The current idea is to successfully exploit what remains in them of awake mirror neurons, and make them work.

Mirror neurons were found first in the motor areas, then in emotional areas. What about the language ? The hypothesis is that the brain’s overall architecture is mirrored…

Here we must distinguish two aspects in which mirror neurons play a role: the evolution of language and phonology of modern language. We believe that the language derives from mirror neurons. It was born and developed as a language of signs to when it then associates sounds. The other aspect in which mirror neurons are certainly involved is phonology, we learn by imitation, and we showed that there is a mirror system.

By the time you hear the word “beer” even if you do nothing, the excitability of a part of the brain that codifies the “r” increases and the “r” sound resonates. With magnetic stimulation, we can show that when I listen, I have a motor copy of what you say. This is very useful for the child. Do you know that children, after a while, lose the ability to mimic other sounds and only remains those they have heard ? A Japanese child, for instance, does not distinguish one “l” of an “r” (he means a kind of “r” in both cases), because something was stuck in him. At first the children are completely plastic. Then, with learning, they eliminate what is not useful and what remains is what they know how to pronounce.

But faced with such an argument, is there something that captures the reasoning to a mirror level ?

The semantics very likely. In any case, many think so. When we hear a sentence like: “He kicks the ball”, it is not only an activation of language areas, motor areas are also active: some say that they are even indispendable for understanding the sentence. Some say that we understand the language with the motor system, but I do not know how to explain the syntax: syntax is very complicated…. What motor areas are activated when I say: “Freedom is beautiful ?” the extreme positions bother me, because it explains everyting, explains nothing… You have to stay within certain limits!

In the next perspective of creating computerized humanoids gifted with adaptive intelligence, would the use, in their construction plans, of artificial neural mirrors be possible, and with what consequences ?

One thing is to successfully teach a robot to act by imitation, another to respond by having to choose a particular action. Suppose there is in the robot mechanism I was mentionning to  understant the intentionality and that if he sees an X input, it excites a chain to bring the object to the mouth or another string to throw it away. If the patient extends a hand to an object, the robot takes it and brings it to him. If the patient pushes the object, the robot take it away. We can try to give intentionality to the robot, so depending on what the camera shows him, he reconstructs the possible motor programs and acts accordingly.

There are European projects on this subject, we ourselves were involved in a few, but they are two very different worlds, robotics and this one… In Genoa, the team of professor Sandini works incidentally on the insertion of these biological aspects in robots… But the most interesting thing, in my opinion is  the robot which includes intentions and then does what you do not manage to do, or you did not want to do…. For a sick person, this is very convenient, especially if he is paralyzed.