Monkey See, Monkey Do

The other day I was teaching my sister to play the drums. After she struggled for a bit, my dad told her about this video he watched recently where this novice tennis player was taught how to play tennis by simply watching a professional player play in front of her. So I played her the beats that she was learning and told her to watch. After around 10 minutes, she actually got the beat! I was extremely fascinated by this, and I wondered how that worked and after enough research I came across this article which spoke about “Mirror Neurons” and that made perfect sense.

A mirror neuron is a neuron that fires both when an animal acts and when the animal observes the same action performed by another. Thus, the neuron "mirrors" the behavior of the other, as though the observer were itself acting. Such neurons have been observed in humans, other primate species, and birds.

An observation done on this type of neuron in macaque monkeys showed how the neuron’s firing rate increases when the monkey views a person grasping a small morsel of food, such as a raisin, or one person grasping a raisin from the hand of another. The cell also responds when the monkey itself grasps the raisin. Is this cell simply interested in raisins? The answer is no, because the neuron does not respond to the sight of the raisin alone. Is this cell simply interested in grasping? Again the answer is no, because it does not respond to the sight of a grasping action in the absence of the raisin. Interestingly, these neurons do not respond when the monkey watches the raisin being grasped with pincers.

The fascinating part is that mirror neurons fire in the same way when we recreate that action ourselves. Apart from imitation, they are responsible for a myriad of other sophisticated human behavior and thought processes such as learning, empathy, and sleep paralysis, to name a few. Moreover, defects in this system have been linked to disorders like autism and schizophrenia.

Another interesting fact about mirror neurons is that they don’t just activate when the observer is watching someone perform a particular task, but also by the sound of the action being performed. Two papers (1, 2) have reported that neurons in F5 (a section of the brain- pictured below) responded to the sound of an action: so-called auditory mirror neurons. On average, 17% of F5 neurons have been reported to have auditory properties (12.7% and 21.3%, respectively, in the two papers).

Multiple papers have reported that mirror neurons get discharged during action observation and that their firing is further modulated by different factors: occlusion, the relative distance of observed action, reward value, and the viewpoint of the observed action. A study also showed that 19/37 mirror neurons were discharged even when the observed action was occluded or hidden from the observer, demonstrating that direct vision of the action was unnecessary to elicit mirror neuron discharge.

This brings us to an important question: What are the implications of auditory mirror neurons and mirror neurons? This can be answered by the “Theory Of Mind” hypothesis.

Theory of mind refers to our ability to infer another person's mental state (i.e., beliefs and desires) from experiences or their behaviour. Several competing models attempt to account for our theory of mind; the most notable concerning mirror neurons being simulation theory. Simulation theory suggests that the theory of mind is present because we subconsciously empathize with the person we're observing, from the auditory part of it to the possible surrounding environment they are in, accounting for relevant differences, where we imagine what we would desire and believe in that scenario. Mirror neurons have been interpreted as the mechanism by which we simulate others in order to better understand them. Therefore, their discovery has been taken by some as a validation of simulation theory. All of us most likely have picked up many habits and actions implicitly due to the mirror neuron systems from the people we interact with, which could also explain who you surround yourself with, matters.

Autism, on the other hand, is a neuropsychiatric abnormality characterised by two main defects in the mirror neuron system. First is the defect in the social-cognitive domain, which presents itself as cognitive loneliness, a lack of empathy and contact with the external world. The second is sensorimotor defects like temper tantrums, head banging, and some form of repetitive rituals. A phenomenon in autism, one which is particularly interesting, is the inability to comprehend abstract reasoning and metaphors, which in healthy individuals is subserved by the left supramarginal gyrus which is rich in mirror neurons. Mirror neuron abnormalities have also been called into question for several other autistic problems like language difficulties, self-identification, lack of imitation, and intention understanding. This is why people with autism require a completely different approach to be taught lessons and different techniques to adapt to life around them.

Now coming back to the example we started with, if the person watching the professional tennis player play tennis lay somewhere on the Autism spectrum or if that person had a cocaine addiction, that person would be unable to learn the techniques required to play that game just by watching, as that person would not possess a functional ‘Theory of Mind’ and in turn would have a deficit in their mirror neuron system.

In conclusion, numerous studies indicate that the mirror neuron system plays a major role in domains such as empathy, autism and other social abilities of humans. Mirror neurons do not activate when a person does any random action, but they do activate when the intention behind the action is understood. It also makes sense from an evolutionary perspective, if another monkey is going to gain an advantage by eating a banana, the observing monkey would want to prepare to do the same to not miss out on a resource of food. I would like to end this blog with a thought-experiment - if humans learn actions based on observation and get better at that particular action, what makes us any different from Artificial Intelligence in today’s world? And how real is the reality of us living in a simulation?