Lec 7: Mirror Neurons

How is the brain originally thought to be organized, and why were mirror neurons such a surprise?

• Neurologist Alexander Luria proposed the brain could be divide into zones responsible for separate, independent jobs (ex: PFC is for thinking, back of brain is for perception, in between is a strip of cortex devoted to action)

• This states that perception and action are fundamentally different

• Hence, why mirror neurons were a stunning discovery since they’re located in the motor-related cortex part of the brain which is supposed to be for doing, not for seeing. Yet, they fire when you watch someone perform an action (ex: a neuron fires when you make your hand move to reach for something BUT it also fires when you see someone else do the same thing).

• Brings up the possibility that if neurons can bridge seeing and doing, they might also bridge self and other

primary motor cortex (M1)

• issues direct motor commands that drive specific MOVEMENTS via muscle contractions, w/a somatotopic map (the motor homunculus) where each region controls a different part of the body

• where the motor homunculus is located —> sends signals down spinal cord with instructions for every movement of your body

premotor cortex (PMC)

• selects, prepares, and sequences ACTIONS (especially externally guided or visually cued movements) by shaping the motor plans that M1 executes

• provides intention-based oversight of M1

• cares about ACTIONS (movement in service of goal) so it plans them and turns them into the set of coordinated movements that M1 carries out

• mirror neurons found here in PMC which is why they turn out to be about goals and not just muscle movements

How were motor neurons discovered?

• by neuroscientist Giacomo Rizzolatti

• He would conduct typical studies that didn’t recieve much attention. One of them was placing electrodes into premotor cortex in macaque monkey to see how those neurons respond when the monkey grabs an object 2 diff ways.

• When doing a study like this, the monkey didn’t pick up the object at all so someone went over to pick it up and restart the trial. However, when the money saw the object being picked up, the same neurons fired in the monkey that would fire if the monkey itself picked up the object. The reset wan’t originally part of the study.

• b/c these neurons were in the motor cortex, firing during performance was expected but firing during observation was surprising that opened more research and debate

What are the 3 regions of the mirror system?

• ventral premotor cortex —> controls complex hand-to-object interactions (such as grasping and shaping the hand), coordinates mouth and face movements, and houses part of the mirror neuron system for action observation and understanding

• rostral inferior parietal lobule —> integrates sensory, motor, and cognitive data. It handles attentional reorienting, tool use and motor intentions, and peripersonal space awareness

• anterior intraparietal sulcus —> coordinates hand grasping, fine finger movement,. It processes object physical features (like shape and size) to shape hand grips, controls real-time action adjustments, and represents action goals for both self and others

Do mirror neurons respond to what an action looks like, or to its goal?

• Study conducted by Umilta + colleagues

• What they did + findings: In one condition, the monkey sees block on table and watches a hand reach out and grasp it. When it becomes clear hand is going for block, mirror neurons respond strongly. In key condition, the monkey sees block, but as hand starts moving, a divider drops down and hides the block. Monkey sees hand go behind divider but never sees the grasp. Mirror neurons respond just as strongly. Then they did two more versions: same hand movements, w/ and w/o divider, but with no block to grab. Here the mirror neurons didn’t respond.

• what it means: what matters is not seeing the action completed, but UNDERSTANDING the goal of what’s happening. The most telling comparison is between the two dividers w/diff conditions. When the divider is down, they’re visually identical, yet only the one with the block actually produces mirror activity. This means, mirror neurons don’t understand pretend. Actions always involve goals, and mirror neurons code the goal.

Mirror neurons

• neuron active BOTH when performing a particular action & when observing that action being performed by another being (SAME neuron must be active for both, neighboring neurons don’t count)

• these neurons represent the GOAL of an action, not just mere movements, & not just what’s visible

• Mirror neurons seen in macaques; whether humans have them is still a debate

Why is it hard to show that humans have mirror neurons?

• not many humans would volunteer to have a patch of skull removed and place electrodes directly on individual neurons like they did with monkeys

• so, we rely on fMRI for human studies which have shown similar activity to mirror system of the monkeys.

• This evidence is suggestive but doesn’t settle the question because of resolution. fMRI images are built from voxels and each one contains about 600,000 neurons. Each voxel gets a single color assigned. so when a voxel lights up for both performance & observation, its hard to tell whether singular neurons were going BOTH or if some neurons WITHIN the voxel were active during performance while others were only active during observation

• Kilner study found solution to this problem

What is repetition suppression?

• def: repeating a stimulus produces a reduced response in brain regions sensitive to it

• ex: fMRI compares two kinds of things like a house & face. FFA responds more to the face but when you show the face again, it responds less strongly.

• this is due to efficiency. The 1st time, a voxel recruits a broad population of neurons, some more strong/useful than others. The 2nd time, the poorly tuned neurons drop out, leaving the strong ones to fire only, dropping voxel’s overall signal. Processing hasn’t gotten worse, representation has gotten sharper.

• repetition suppression works for more than just vision —> works with feeling & action (ex: watching an action twice)

How did the Kilner study get around the voxel problem, and what did it find?

• his reasoning before study: if neurons in human brain handle both perception and action, then it should treat watching the action and performing it as the same thing happening twice. Therefore, suppression repetition would occur just as it does for being shown an image twice. If there’s no repetition suppression, then it’s evidence that there’s no mirror neurons because it handles performance and observation as two different things.

• Study: participants lay in scanner with 2 levers. One lever pushed toward side or pulled out and the other pushed away or pulled toward them. Each trial was a video of a hand performing one of these actions. Some trials, you had to repeat the action you just watched while for other trials you did a different action than the one watched.

• Findings: only repetition trials showed repitition suppression. Suggesting among 600,000 neurons in voxels, some of the same neurons handle both performance and observing. Same three regions of the monkey showed up.

• one of the most convincing studies for mirror neurons

What big claims were made about mirror neurons?

• V.S. Ramachandran: predicted mirror neurons would do for psych what DNA did for biology. They would provide a unifying framework & explain various mental abilities that had remained mysterious & inaccessible experiments

• some papers claimed mirror neurons explained almost everything —> language, imitation, cultural transmission, autism, placebo, hypnosis

• focus claim: mirror neurons explain how we understand other minds

What is the simulation theory & how does it differ from ToM?

• simulation theory proposed by Robert Gordon —> to understand another mind, you don’t rely on theory. Instead you, imaginatively put yourself into the other person’s situation & let your own cognitive & emotional machinery run, & read off what they would think, feel, or decide from what comes out of your own simulation. Meaning, understanding others becomes closer to imaginative reenactment than to theorizing about them.

• Ex: you want to now what wind will do to an airplane wing. Approach 1: develop the physics of aerodynamics and work through formulas. Like ToM, you reason through various “If __, then __” rules. Approach 2: build a wind tunnel with a model of the wing and simply watch what happens. That is simulation.

• ToM works best when another person is typical

• Simulation depends on how accurate and complete it is & works best when the other person is like you because you’re using your own reactions as stand-in for theirs

What is the motor resonance account of understanding other minds?

• Vittorio Gallese turned simulation theory into motor resonance account

• argued the fundamental mechanism giving us direct experiential grasp of other minds is not conceptual reasoning, but direct simulation of observed events through the mirror mechanism

• ex: Jared’s mirror neurons are active bc he’s performing the goal-oriented action of reaching toward a light switch. Your mirror neurons become active from observing him. Since your brain is now in a state of resembling his, you directly experience his intention to turn of the light. The shared activation is your understanding of his action. so, mirror neurons build a direct, pre-reflective bridge between self and other.

• major point: speed. Mirror neurons are far faster than logical reasoning, letting us make sense of other minds in real time w/o stopping to think.

• this account faces serious problems though… 3 critiques

Critique 1: that responding is not the same as understanding?

• critics (Cecilia Hayes & Gregory Hickok) note that on the motor resonance view, you count as having understood an action simply if seeing it activates the neural mechanisms you’d use to produce it. That activation need not have any further consequence, such as letting you name the action or reason about it

• critics then argue that the use of “understanding” is misleading b/c it comes close to erasing the difference between responding & understanding. Mirror neurons activating at the right moment doesn’t mean they understand anything either

Critique 2: What does neuropsychology say about the motor resonance account?

• Study (Alfonso Caramazza’s group): tested patients who had strokes and had trouble using everyday objects. Researchers pantomimed the same object-related actions to see if patients could recognize them.

• findings: many patients who were bad at using the objects were fine at recognizing the actions they couldn’t perform.

• if you recognize an action you can’t perform, then the ability to perform it is not what makes recognition possible. That’s the direct problem for the motor resonance claim that understanding others’ actions depends on simulating them in your own motor system.

Critique 3: What’s Heye’s associative account? how did the countermirroring study test it?

• argues mirror neurons were not specially built by evolution for understanding other minds. Instead, they’re ordinary neurons wired by ordinary experience. The key is the Hebbian rule: neurons that fire together wire together

• ex: a child picking up a pencil for the first 10000 times. Almost every time, the child is both doing the action + watching it, activating visual and motor neurons at the same moment, over and over. Through Hebbian learning, those neurons become linked—not because evolution required it but because the 2 kinds of activity kept happening together. Later when you watch someone pick something up, the visual AND motor neurons activate b/c of that long history. That looks like mirroring.

• fMRI Study (Catmur & colleagues w/Heyes): The mirror condition had participants watch a hand flex & respond by flexing their hand, just like ordinary experience. The counter mirror condition had participants watch a foot flex & respond by flexing their hand.

• Findings: Evolution never built neurons linking a seen foot with a moving hand, so if those responses appear they must come from learning. They did, both conditions produced mirror-like activity in the classic regions.

• this showed mirror responses can come from learned associations

What’s the “fad cycle," and where do mirror neurons fit in it?

• Paul Rosin recognized this cycle

• fads def: short-lived, enthusiastically followed trends in diagnostic labels, therapies, or research ideas that quickly surge in popularity before fading when evidence fails to support them

• fads build enthusiasm partly by exaggerating what they can do and many branches have claimed to explain most of human behavior.

• a valuable idea that might realistically explain 20%, sells itself as explaining nearly all of it. The cycle is oversell followed by overkill. (aka hype cycle: inflated expectations —> a crash into a trough of disillusionment —> recovery to a level of real —> practical value that’s below original hype)

• For mirror neurons, the long list of everything they were supposed to explain was peak hype. Most of it didn’t survive, but something real survived

What do mirror neurons actually seem to be good for?

• imitation

• First ever mirror neuron fMRI study (Marco Iacoboni): participants watched a hand lift one finger & sometimes imitated it. mirror regions responded to both during observation & during imitation, the characteristic mirror pattern.

• Dance Dance Revolution (DDR) study (Emily Cross): participants watched 18 dance sequences in the scanner, then trained at home for 5 days. 6 sequences they practiced at home, 6 they only watched, 6 they did not train at all. Afterward, they watched all 18 again & took a surprise test dancing sequences from each group.

• DDR findings: performance was best for the practiced sequences and worst for the untrained ones. sequences they only watched landed in the middle. meaning, watching alone helped and the benefit from watching was related to the mirror system activity.

• this reverses the motor resonance story. motor resonance says seeing an action leads me to understand you. the imitation story says watching someone else prepares me to perform the action myself.