by JohnMark Taylor
figures by Youngeun Kaitlyn Choi

What about the human brain allows a person to perform such feats as learning guitar through imitation, empathizing with anothers’s pain, or intuiting where a fencer will strike next? Nearly twenty-five years ago, scientists discovered a special kind of cell called a mirror neuron that many both in science and the popular press came to believe might enable social skills like these, skills that underlie much of what makes us uniquely human. However, after a quarter century, dozens of experiments, and reams of popular articles, the true significance of these cells has become increasingly controversial. What have mirror neurons really told us so far about the human mind, and what remains to be learned from them?

What are mirror neurons?

The story of mirror neurons began simply enough. In 1992, a team of neuroscientists led by Giacomo Rizzolatti inserted tiny electrodes into the brains of macaque monkeys, hoping to better understand how the brain orchestrates the delicate interplay of muscles involved in moving the hand. Using these electrodes, the researchers monitored the activity of neurons, the cells that constitute the smallest processing units of the brain, analogous to microchips in computers. A neuron can be specialized to perform any of a dizzying number of functions, from perceiving a face to regulating sleeping and waking.

Rizzolatti’s team, examining neurons in a part of the macaque brain involved in controlling the muscles of the hand, expected to find neurons that fired specifically when the monkey performed particular actions, such as reaching for or grabbing something. They indeed found neurons that fired when the monkey performed these actions, but it turned out that this was only half the story. One day, when the experimenters ate lunch in the same room as the monkeys, they observed something entirely unexpected: some of these neurons also fired when the monkey observed an experimenter performing the same action (in this case, bringing food to one’s mouth). In short: these neurons fired both when monkey see, and when monkey do.

Figure 1: Mirror neurons in action. A mirror neuron fires an electrical pulse, or action potential, when the monkey either observes or executes a specific action. In this case, the mirror neuron responds to grasping actions. The graph at the bottom shows what the action potentials (each depicted as a hump) would look like when measured with an electrode, as used by the researchers.
Figure 1: Mirror neurons in action. A mirror neuron fires an electrical pulse, or action potential, when the monkey either observes or executes a specific action. In this case, the mirror neuron responds to grasping actions. The graph at the bottom shows what the action potentials (each depicted as a hump) would look like when measured with an electrode, as used by the researchers.

Mirror neurons’ great potential

For nearly a decade, these neurons, termed “mirror neurons,” remained relatively unknown to the public. However, their reputation began to change in 2000, when the famous neuroscientist and popularizer of science V.S. Ramachandran wrote an edge.org article speculating that “mirror neurons would do for psychology what DNA did for biology: they will provide a unifying framework and help explain a host of mental abilities that have hitherto remained mysterious and inaccessible to experiments.”

In a series of elegant, compelling proposals, Ramachandran theorized that mirror neurons might help explain a wide variety of human social abilities. For example, how, biologically, do people imitate the actions of others, an ability that in part enables the spread of culture? Ramachandran proposed that mirror neurons translate an observed action into a series of commands for the muscles to execute. How do people understand the intentions behind another’s actions? Mirror neurons may run a sort of virtual reality simulation of what it would be like for oneself to perform that action. Why are autistic individuals impaired when it comes to understanding the thoughts of others? Perhaps they have deficient mirror neurons (an idea that came to be called the “broken mirror” hypothesis). Within a year, the use of the phrase “mirror neurons” more than doubled, and over the next decade, mirror neurons captured the public imagination, being touted as able to offer insight into everything from empathizing with therapy clients to international diplomacy, how children learn music, and how people appreciate art. Not bad for a finding that was initially rejected from the top science journal, Nature, for “lacking public interest.”

As interest in mirror neurons exploded among the public, scientists remained divided regarding their significance. Some scientists, such as Rizzolatti and Ramachandran, were optimistic that mirror neurons would prove crucial for many of humans’ social abilities, while others thought that their importance was overblown. For some time, skeptics had one particularly effective arrow in their quiver: despite claims that mirror neurons might underlie much of what makes humans unique (such as language and culture), until 2008, they had never once been decisively identified in humans. Even as of 2016, only one study, using electrodes implanted into the brains of epilepsy patients, has successfully identified human neurons with properties similar to those found in the macaques.

An updated perspective

Accordingly, over the past ten years, the pendulum of scientific opinion has begun to swing towards the skeptics. Many of the more prominent theories regarding the function of mirror neurons have not survived scrutiny. First, it was seen as increasingly implausible that mirror neurons alone could explain the human capacity for imitation; adult macaques, it became increasingly clear, did not engage in mimicry despite having mirror neurons, and so mirror neurons could not explain this ability by themselves. Second, the theory that the ability to mentally simulate others’ actions (putatively enabled by mirror neurons) is necessary to understand others’ actions has become increasingly shaky. For instance, some patients with brain damage that prevent them from performing certain actions (such as brushing one’s teeth) are nonetheless able to understand the meaning of these actions when performed by others.

Finally, the theory that mirror neuron defects might underlie autism—the “broken mirror” hypothesis—has proven most dubious of all. An exhaustive recent review of 25 different studies presents a wide array of behavioral and neurological evidence that deficient mirror neurons probably do not lie at the core of autism. For example, while the broken mirror hypothesis predicts that autistic individuals should show severe impairments in understanding and imitating actions such as reaching, several studies have found no such impairments. Moreover, while many studies have reported differences between the brains of autistic and non-autistic individuals, these differences do not appear to lie in parts of the brain thought to contain mirror neurons.

Using magnetic resonance imaging (MRI), several studies have examined the cortical thickness (the size of the sheet of neurons covering the brain) of various brain areas, and have found only sparse evidence that structural differences in mirror neuron regions might be involved in autism. Rather, structural differences between autistic and non-autistic individuals appear to extend widely throughout the brain, and differences in mirror neuron regions do not appear to show reproducible patterns between subjects. Additionally, these mirror neuron regions appear to show similar activity in autistic and non-autistic individuals when they view or perform various actions, suggesting that the neural basis of autism probably lies elsewhere.

Counterarguments like these have pushed mirror neuron proponents to fine-tune their claims. For example, Rizzolatti, the original discoverer of mirror neurons in macaques, now suggests that mirror neurons might only be required for understanding the actions of others from a first-person perspective. He explains that this internalization of behaviors we see might provide us a deeper level of understanding about another person’s goals, but cedes that mirror neuron activity might only constitute one among several ways of comprehending others’ behavior.

Mirror neurons have begun to assume a humbler identity than was initially theorized, but it is important to remember that despite recent criticism, their activity may still play an important role in many behaviors. For instance, even Gregory Hickok, perhaps the most prominent critic of the hype surrounding mirror neurons, accepts that they probably play a role in enabling imitation, given that there must be some mechanism in the brain that converts an observed action to a series of muscle commands. Much research remains to be done; for instance, there has yet to be a study that specifically disables mirror neurons (an experiment that recent technological advances may make possible in monkeys), which would help to elucidate what exact behaviors rely on these neurons. Now that the hype around mirror neurons has begun to dissipate, it will be interesting to see what role remains for these curious cells.

JohnMark Taylor is a PhD student in the Harvard University Psychology Department.

For more information:

Read the original study in which mirror neurons were discovered here. For a detailed, critical look at theories regarding mirror neurons, see this recent book by psychologist Gregory Hickok. Rizzolatti, the original discoverer of mirror neurons, writes a response to Hickock’s criticisms here, to which Hickok replies here.

8 thoughts on “Mirror Neurons After a Quarter Century: New light, new cracks

  1. Posted on behalf of Paolo B. Pascolo, who asked that these remarks be posted as a comment:

    I must say in advance that I’m rather skeptic about the core of the interpretations given by Rizzolatti of his own measurements. On my opinion, the mirror effects or artefacts could be generated by common neurons in their normal operations inside our complex brains.
    I have to say that a behavior correlation doesn’t always mean behavior causation.
    On one side the physiology of the so-called neuron mirror is still undistinguished from that one of the “common” neurons; on the other one many of the abstract concepts arbitrarily forced to be tied to the NMS, such as empathy, autism, could require a totally different, explanation.

    1. Hi JohnMark
      I totally agree with you when you say that “a behavior correlation doesn’t always mean bahavior causation”. And in my opinion most of the research in mirror neurons lies in a great misconception that rests in the consideration that the neural system works representing the environment. Here I stand with the Santiago School and understand that the neural system works in operational closure. Therefore, an advance in the studies concerning these cells requires a shift in the starting point of view of the research. Until then, nothing important will be achieved for a real explanation of the functioning of the system and its behavioral correlation.

  2. When I’m in a bar, I often observe other patrons that will lift their glass right after the moment I lift mine. Perhaps this is just a sign of anxious behavior…a feeling of insecurity I presume.

  3. Nice, short, powerful piece (unlike Rizzolati’s and Hickock’s responses. Thanks. Another area to look at is embodied simulation, as per Bergen’s Louder than Words. His work, and Schacter’s on memory have convinced me that the brain, doing “massive redeployment” has taken the same neural networks used to process sensory input (sensory & motor cortices), to interpret it in others, to remember it, and to use it with language. When you hear throw the ball, the same visual, somatosensory, and motor areas fire up as if your were doing it yourself. (That is why you shouldn’t talk much while driving.) Could you send me an email? ctskelly at gmail.

  4. Ah those mysterious mirror neurons……mainly because it seems to me that monkeys deafferented by rhizotomy (cervical and thoracic roots all cut) seem to adjust similar movements more economically than intact monkeys, though it takes them longer to master them. However, in a colony situation, deaff monkeys show a far superior acquisition of normal movements than when isolated post op. Furthermore, neonate deaffs (2nd trimester rhizotomy and reinsertion into womb for normal delivery) in isolation learn to perform reach and grasp behaviors much faster than unisolated neonate deaffs, often just as fast as intact infants. We managed to save the hands of deaff monkeys from self-mutilation for up to ten years using no restraint. We found the slower isolated vs. socialized post op recovery of movement in deaff adolescents to be the reverse of neonate deaff isolated vs socialized, the latter taking longer, very perplexing. What is most regrettable to me is that PETA’s successfully eliminating deaff research from neuroscience– using a self-mutilating phenomenon in deaffs that we had eliminated for emotional appeal to the media– has devastated a very productive area of research by exploiting hand biting phenomenon as a horror show, a phenomenon which, as I said, we found is no longer a problem once monkeys are functionally using the deaff hand in their home environment. I especially regret the absence of this preparation for use in MIRROR NEURON research. It would be very interesting to see a systematic study of whether a deaff monkey does better after watching an intact monkey attain reward with reach and grasp or any other complex forelimb movement, as movements into INTERPERSONAL space (such as self touching and feeding….and, indeed, hand biting!) recover much earlier than movements into extrapersonal space such as climbing or reaching for food through a hole in an opaque box at reach distance. Also, it was our general observation that if an animal never seems to acquire a successful movement, we must have left a rootlet intact. And indeed, in every monkey that recovered poorly or severely bit its hand, this was the case. Essentially, from my perspective, mirror neurons in deaff monkeys, as described, seem to suffer impediment of their function due to residual sensory input, as when a rootlet is left intact, until the monkey learns its meaning and uses it advantageously. Other studies we did showed that lesion of the lemniscal pathways severely impede precision grip, it being replaced by a crude whole-handed scooping movement only occasionally effective. But thumb-index apposition would not recover: thumb and index seeming incapable of coming together at the tips to seize a small food pellet placed in a device where this movement is required for successful seizure of the pellet. If however,deafferentation is superimposed, thumb index apposition returns. The reaching movement and the shoulder’s ability to hold the hand over a target– both is in no way affected by central lemniscal lesion– after deaff, suffer from severe dysmetria, though accurate ballistic movements with an unseen arm are quite accurate and smooth. The deaff’s problem seems to be an inability to perform anything other than a ballistic movement which, though accurate cannot hold the hand over the target. The adaptive behaviors of monkeys so as to stabilize the hand were nothing short of amazing. All in all, our impression was that what Rizzolatti et al describe as mirror neurons might be severely disrupted by uncontrolled or aberrant sensory input. However, over time, this is overcome. I propose that we still do not understand how somesthetic and proprioceptive inputs help and hinder purposeful movements and would suggest that the role of uncontroled afferent “noise” in the system, as after incomplete deafferentation, somehow impedes the internally formulated behavior much as vestibular aberrancy impedes Hippocampal function in spacial localization. It has been several decades since AJ Berman, TA Tran and I did this work, having only reported it at the Society of Neurosciences meetings. But Prof Ed Taub is still around and I would strongly suggest that all this work, in which he was a pioneer until hobbled by PETA, be rewritten and theoretically updated in light of the tremendous of work done on mirror neurons to date so that researchers in mirror systems might study the role of these neurons in reach-grasp behavior. It is tantalizing that monkeys with Areas 4 & 6 bilat lesion several years prior to deaff did not show a significant recovery difference from monkeys without cortical lesion. Rizzolatti’s F5 neurons thus seem to have been eliminated.

  5. I am not a scientist. I am an MS patient with a very strange symptom. When I see a person subjected to pain in real life or even in a movie like when I saw a cowboy have his bleeding leg cauterized by a branding iron to stop the bleeding, I feel an extreme burning, stinging pain sometimes in the same exact area of the body as the person subjected to the pain. If I don’t see the action and only see the resulting injury, the pain usually stars in my feet and radiates up my legs. I don’t get nauseated. I don’t faint. I feel pain. The pain is sudden, unexpected, and severe. I have to look away and try to push the scene out of my mind.
    My MS doctor who has an MD and a PhD and conducts research in MS for a well-known university suggested I Google mirror neurons to see if I could find any discussion on this type phenomenon. He has no other MS patients with this symptom. Just wish I knew what causes this and hoping someone out there has an explanation.

  6. We are sort of realizing that individual neurons don’t really exist in the brain, but rather, information cycles through inumerable neurons innumerable times, each time modifying that incoming rather muddled signal. I could go into a most prolix discussion of what happens to it as it cycles, but surely we would all agree that it picks up or looses “rotational momentum” as it is bombarded by new input until finally the various inputs activate or deactivate other cycles, ultimately resulting in a thought, or extinction into just nothing , or the triggering of other internal circuits that eventually lead to thought, perception or action. I fear that the mental image I have of the brain is much like that of an alien in a flying saucer looking at the traffic on all NY/NJ roadways during rush-hour and wondering where any one car is going and when it will get there . Thus I could devote myself to studying where a particular blue 2010 Honda Sedan might be going and how that will affect tomorrow’s traffic at the same time and on that same segment of roadway, but I really doubt that it alone would be much help over the next decade of daily observation of its comings and goings, as it may be very relevant to the perceptions, goals, habits and state of mind of the driver, but certainly not to the history of America for the next 24 hrs. Perhaps Rizzolatti’s brilliant work , using unicellular recordings from region F5 Frontal Cortex may well have cued us with a very important clue as to the brain’s motor function, but, reading much of the “mirror neuron” literature I fear too many are too prone to report on the INTENT of all the traffic in that region on the basis of the meticulously documented activity of that very interesting blue Honda.

    To my mind, today’s science is not a mad search for truth, but a BALANCING SCALE on which scientists try to balance COMPLEXITY and NECESSITY. On the one hand, every experimenter knows full well that the phenomenon from which he extracts a principle must be much, much more COMPLEX than it seems on careful analysis, but then, on the other hand, there’s the NECESSITY of funding further research under the rule of PUBLISH OR PERISH. And so, too often, the brilliant skeptic becomes a Jesuit for the research funding Popes. Science thus doggedly hangs on to DOGMAS as solid rock bottom foundations on which it builds bridges of “knowledge” upon which SCIENCE MARCHES ON. But, as all scientists and students march forward in cadence, building castles in the sky, guided by by current DOGMA, eventually the bridge’s strength gives out, it collapses, dogma and all, and a new COMPLEXITY/NECESSITY bridge, based on new DOGMAS is built. Like worker ants locking together in self sacrifice, scientists sacrifice the diversity of erudite thoughts to focus on a new bridge of self-sacrificing ants which their Nobel Laureate Queens can cross. Can anyone be surprised then that it is estimated that 1/3 of all PUBLISHED literature is tainted with hanky-panky? Are the highs and lows mankind suffers thanks to the dogmatism of science necessary ? Or can we find our way to more reliable foundations, unshattered by recurrently decreed “uncertainty principles,” always a decade or so after the Eureka Dogma was so pompously established. For us, False Flags are cheap, mere PdFs we can call forth on or erase from our computers, but for the starting college students the “all you needed to know about….” textbooks in biology, chemistry, physics, mathematics– full of fantasies at $250 or so each– all becoming part of a life-long repaid usurious student loan– may well explain why so many Americans, in this “ERA OF SCIENTIFIC MARVELS,” are so “turned off” to science. SINCE FACT TODAY IS SO TENUOUS, WOULD IT NOT BE BETTER TO MAKE HIGHER EDUCATION *FREE* LEST OUR CHILDREN BE FOREVER PAYING FOR THEIR PARENTS’ ILLUSIVE MIRROR NEURON DELUSIONS?

  7. De Teodoru, very interesting comment.

    Mirror neurons are yet another fantasy homunculus linking mind to brain to environment. The hard problem of consciousness (and culture) remains inscrutable from a physicalist perspective. It may not always be so. But we have seen repeated postulates of Descartes’ pineal gland over centuries. We need a new paradigm.

Leave a Reply to DE Teodoru Cancel reply

Your email address will not be published. Required fields are marked *