Social: Why Our Brains Are Wired to Connect (20 page)

Figure 6.1 The Mirror System in Macaques (left) and Humans (right)

Although this first imaging study suggested a role for the mirror system in imitation, two additional kinds of evidence are necessary to make the case.
Iacoboni’s next step was to test whether imitation was affected when the mirror system was temporarily impaired.
His group used
transcranial magnetic stimulation
(TMS), a technique that directs an electromagnetic field at a particular spot in the cortex and
temporarily “frazzles” the neurons in that area such that the region is essentially taken offline
.
It sounds like a scary technique, but when done properly with healthy individuals, it is a safe and temporary procedure.
In this study, individuals were asked to imitate sequences of button presses as TMS was applied.
When TMS was applied to a mirror system region, participants made more errors in their attempts to imitate the other person.
But when TMS was focused on a nonmirror region, there was no increase in error rates, suggesting that the mirror system plays a causal key role in imitation.

These studies demonstrated that the mirror system is involved in rudimentary forms of imitation when the to-be-imitated behavior is not novel.
Adults are already experts at tapping their fingers before they arrive for these studies.
To test whether or not mirror neurons support imitation-based
learning
(that is, spreading new ways of doing things), evidence was needed showing that the mirror system is involved in acquiring new behaviors through imitation.
Rizzolatti’s group examined the neural systems involved in non-musicians
imitating the fingering required to make a set of guitar chords that they were shown
.
As predicted, the mirror system was involved during the act of imitating previously unknown complex hand movements.
There is no doubt that other cognitive capacities play important roles in various kinds of imitation, but it does seem reasonably safe to say that when it comes to its role in imitation, the mirror system appears to live up to the hype.

Mindreading Mirrors?

A second major claim made by mirror neuron researchers is that the mirror system is responsible for understanding the minds of others.
This is the claim that interests us the most as we try to understand the social mind.
Just as President Clinton’s impeachment fate depended on a verbal dance around the precise meaning of words like
sex
and
is
, whether or not we end up believing that mirror neurons help with mindreading and interpreting the intentions of others will depend on what we mean by words like
mindreading, goals
, and
intentions
.
To understand the relation of the mirror system to mindreading, we have to retrace our steps a bit to a philosophical debate over how we know the minds of others.

In the early 1980s, developmental psychologists were all abuzz with the notion of
Theory of Mind
—that we have a theory that other people have minds, and with this theory we can logically infer their thoughts, beliefs, and desires in countless situations.
Various philosophers like Daniel Dennett and Stephen Stich, one of my undergraduate mentors, were very supportive of this account of how we make predictions about the behaviors of others.
For a few years, Theory of Mind was the only game in town.
But in 1986, philosopher Robert Gordon suggested an alternative for how we understand the minds of others.

Gordon’s key insight was that there are multiple possible ways by which we can predict another person’s intentions in a given situation.
One route was the one associated with Theory of Mind.
Given our theory of how minds in general operate, we could use propositional
if-then
statements to logically figure out what a person’s intentions might be.
For instance, if we know that someone has not eaten in eight hours, then we can infer he is probably hungry, and if we can infer that he is hungry, then we can also infer that he currently possesses the intention of finding something to eat.

The second route involves imagining what it would be like to be in that situation ourselves and to use our own natural reactions to this simulated experience as a guide to how another person is likely to think, feel, and act.
If I want to understand what someone is experiencing after getting dumped by a romantic partner via text message, I can try to mentally re-create the scene, imagining myself as the protagonist.
What reactions do I see myself having as this scene unfolds in my mind?
They may help me understand how someone else might react.

Often these two routes lead us to reach the same conclusion but through different processes.
In the one case, I am thinking logically
about
the situation and how anyone would likely respond to it.
In the second case, I imagine myself
in
the situation in order to find out what my own reaction would be.
In the first case, my accuracy depends on the quality of my logic, and how similar the individual’s mind is to the typical mind I am reasoning about.
In the second case, my accuracy depends on the quality of my re-creation of the scene and how similar the target’s mind is to my own.
Gordon’s account of the second route
has been referred to as
Simulation theory
; and there is little question that at least in some cases we project from our own experience to that of others.

The key question here is whether mirror neurons have anything to do with either of these accounts.
Vittorio Gallese, one of the original discoverers of mirror neurons, has argued that
mirror neurons are the neural implementation of Simulation theory
.
Moreover, he has argued that this is
the
way we come to know the minds of others in normal circumstances, suggesting that
“the fundamental mechanism that allows us a direct experiential grasp
of the mind of others is not conceptual reasoning but direct simulation of the observed events through the mirror mechanism.”

When I think of the kind of mental simulation that Gordon and others have generally described, it sounds like hard work.
They suggest mental simulations are analogous to building wind tunnels to test airplane wings in simulated flying conditions or building complex
computer simulations with countless variables to do the same digitally.
When they discuss social simulations, they frequently talk about mentally constructing all the relevant aspects of a situation before running themselves through the simulation to see how they would react.
That sounds like a lot of work.
But if as Gallese suggests, the mere perception of another person allows us to intuitively and automatically understand their experience, that could make the mirror system a much more plausible implementation of Simulation theory.

Here’s how Gallese’s argument works.
The fact that your “reaching-for” neurons are active when you see someone else “reaching for” something literally means that neurons in your brain are matching the neural state of the person you are observing.
When you see a person you are looking at “reaching for” a cup, both your and her “reaching-for” neurons are active.
Gallese and others characterize this as
motor resonance
between you and the other person.
If you are experiencing the same motor state as another person, your brain is in essence simulating key aspects of the other person’s brain, allowing you to automatically understand the mental state of that person as it relates to her action or activity.
My brain is mirroring your brain
, and thus by simply knowing the state that I am in, I know your mind as well.
In other words, mirror neurons would seem to provide us with an almost magical mindreading device that operates automatically whether we are trying to understand the other person or not.

Cracks in the Mirror

But a growing chorus of critics suggests that the mirror neuron camp has not done enough to prove that the mirror system in humans is central to mindreading.
Still other critics think that enough research
has
been done and that the conclusion is clear: the mirror system
is not
central to mindreading.
These critics are doing
conceptual and empirical work to ensure that our ultimate understanding of the mirror system and its contributions is correct.
It’s scientific democracy in action.

Mirror neuron researchers have argued that one of the key qualities of these neurons, as they relate to mindreading, is that they are sensitive to the abstract meaning of other people’s actions.
Imagine you saw someone opening a peanut shell.
There’s the raw visual information associated with seeing that action.
There are also the sounds associated with it.
But whether we see or hear the action, our mindreading is focused on the meaning—someone wants to get inside the shell so the peanut can be eaten.
On the one hand, if a neuron responds to the sight of the action but not to the sound, or if it responds to the sound but not to the sight, then it is only mirroring something sensory (that is, at the level of the senses).
On the other hand, if a neuron is responding to the meaning of another’s action, it should not matter whether we see or hear the action.
In 2002, Rizzolatti’s team
found mirror neurons that fit the bill
.
A subset of mirror neurons responded to both the sound and the sight of actions, suggesting that the increased activity in these neurons could be responding to the meaning of the action, not just to its appearance or sound.

Greg Hickok, one of the most vocal critics of the mirror neuron camp,
highlighted an important limitation of the sight-sound study
.
The researchers had started by identifying neurons that showed the standard mirror neuron properties (that is, neurons that responded during the visual observation and performance of an action), and then they tested those same neurons to see which responded to the sound of the actions as well.
Of the neurons initially identified as mirror neurons, only 15 percent also showed the response to the sound of the actions.
This means that 85 percent of the mirror neurons responded only to the visual characteristics of an action.
As such, they could not be representing its meaning.
On the one hand, yes, some mirror neurons in macaque monkeys do seem to respond to the meaning of an action and not just to its appearance.
On the
other hand, the vast majority of mirror neurons don’t do this.
This 1-to-5 ratio is important because in an fMRI scan, we aren’t looking at individual neurons.
Rather, we see the summed effects across large populations of neurons.
When the results of fMRI studies are used to claim that the mirror system represents the meanings of actions, there is no way to be sure that it is the meaning-representing mirror neurons that are driving the effects.
This analysis does not mean that mirror neurons cannot support mindreading in humans, but it does suggest that it will be very difficult to assess this ability with fMRI.

A second finding that is used to defend the notion that mirror neurons represent the abstract meaning of an action, rather than just the sensory aspects, is
that the mirror neurons respond to actions that involve objects that cannot be seen
.
Rizzolatti’s group showed monkeys an object but then placed a dividing wall between the object and the monkey so that the object could no longer be seen.
The same mirror neurons that would respond when the monkey saw an experimenter reach for the visible object also turned on when the experimenter reached for the hidden object.
Rizzolatti argued that if the mirror neurons were responding only to the visual properties of the action, they would not respond to someone’s reaching for a hidden object.

But Hickok points out that this argument is flawed.
The monkeys could be using working memory to hold the image of the hidden object in mind.
Humans are certainly capable of seeing an object
and then continuing to visualize the object in their mind after it can no longer be seen.
Perhaps the monkeys are doing the same.
Mirror neurons may be responding not to the meaning of the action but rather to a visual representation held in our brains.

Cecilia Heyes, a psychologist at Oxford University, makes a very different argument against mindreading mirror neurons.
She suggests that the purpose of mirror neurons
cannot be motor resonance–induced understanding of others because these neurons are not intrinsically wired for mapping observed actions or
meaning from performed actions.
Instead, Heyes suggests, the reason the “reaching-for” neurons are activated both when I perform and when I observe a “reaching-for” action is past experience, rather than an intrinsic mirroring function.
Heyes believes mirror neurons are really just motor neurons that become conditioned over time to respond to the sight of our own actions as we perform them (the neurons then generalize to seeing others perform the same actions).
Given that sea slugs can be conditioned to learn that one event is associated with another event, conditioning does not imply any sort of meaningful understanding.
I have seen my own hand reach for my spoon thousands of times since infancy, so whether mirror neurons are specifically designed for motor resonance or merely conditioned to link the action and the sight of the action, both Heyes and the mirror neuron camp agree that when I see someone else use a spoon, my “spoon-reaching” mirror neurons will turn on.
Only in special cases would Heyes and the mirror neuron camp predict different outcomes.