Behave: The Biology of Humans at Our Best and Worst

by Robert M. Sapolsky

  —

  Probably the least controversial and most plausible is that mirror neurons mediate motor learning by observation.33 Downsides of this theory, though, are that (a) mirror neurons do their thing in species with minimal learning by mimicry; (b) the amount of mirror neuron activity is unrelated to the efficacy with which observational learning of movements occurs; (c) to the extent that mirror neurons are needed for types of observational learning, it’s a pretty low-level contribution in humans—after all, while we do learn how to carry out certain motoric acts by observation, far more interesting is our learning of context by observation—when to carry out that behavior (for example, observational learning may teach a subordinate primate the motoric features of kowtowing, but far more demanding and important is learning whom to kowtow to).

  Related to that is the idea of mirror neurons aiding learning from another person’s experience.34 If you observe someone biting into food and they grimace at its taste, having mirror neurons at the intersection of observing that expression and experiencing it yourself will certainly make more vivid your understanding that you should probably avoid that food. This is an idea advocated by Gregory Hickok of the University of California at Irvine, who, as we’ll see, is a hard-nosed critic of mirror neuron flights of fancy.

  This harks back to chapter 2 and Antonio Damasio’s influential somatic marker hypothesis, the idea that when we are choosing among difficult options, the frontal cortex runs as-if experiments, canvassing your mind’s and body’s responses to doing X or Y—a thought experiment combined with a (gut) feeling experiment. Mirror neurons, with their putative attunement to how things worked out for observed individuals, would certainly weigh into this process.

  Thus mirror neurons might be useful for learning the meaning of a movement, how to carry it out more effectively, and the consequences for someone else who did it. Nonetheless, such neuronal activity is neither necessary nor sufficient for observational learning, especially of the most interesting, abstract human kinds.

  Then there’s the next, more controversial realm, namely the idea that mirror neurons help you understand what someone else is thinking. This can range from understanding what behavior they are doing to understanding why they are doing it to grasping their larger motivations, all the way to peering into their souls with your mirror neurons. You can see why this has spawned debates.

  In this view mirror neurons aid Theory of Mind, mind reading, and perspective taking, suggesting that part of how we understand someone else’s world is by simulating (in our minds, in our PMC, in our mirror neurons) their actions.35 This orients a mirror neuron’s world in a very different way from the previous section, where mirroring is to improve your own motor performance and the most pertinent neuroanatomy about mirror neurons in the PMC is their talking to motor neurons that command muscles. In contrast, mirror neurons being concerned with understanding someone else’s actions should be talking to Theory of Mind–related brain regions, for which there is evidence.

  There was also the suggestion that mirror neuron–mediated perspective taking is particularly concerned with social interactions. Rizzolatti, for example, showed that mirror neuron activity was greater when the observed individual was closer.36 But importantly, this isn’t just literal distance but something resembling “social” distance; as evidence, mirror neuron activity would decrease if there was a transparent barrier between the observer and observed. In Gallese’s words, “this shows the relevance of mirror neurons when mapping the potentialities for competition or cooperation between agent and observer.”

  The notion that mirror neurons aid us in understanding someone else’s actions, leading to our understanding someone else, period, has been heavily criticized on two grounds, most notably by Hickok. First is the issue of causality—while some studies show that mirror neuron activity correlates with attempts at understanding someone else’s perspective, there is minimal evidence that such activity causes the understanding. The second criticism concerns something obvious: we can understand the intent behind someone else’s actions even if we can’t remotely perform them ourselves. This would apply to actions of the observed individual ranging from pole-vaulting eighteen feet to explaining special relativity.

  Supporters of this role for mirror neurons admit this but argue that they provide an extra level of understanding. Gallese writes, “I submit that it is only through the activation of Mirror Neurons that we can grasp the meaning of other’s behavior from within”37 (my emphasis). This is not my area of research, and I’m not trying to be snarky, but it seems like he’s saying that there’s understanding and then there’s super-duper understanding, and the latter requires mirror neurons.

  —

  These mirror neuron speculations have been extended to focus on autism, a disorder in which there are profound impairments in understanding other people’s actions and intentions.38 According to the “broken mirror” hypothesis of mirror neuron pioneer Marco Iacoboni of UCLA, mirror neuron dysfunction underlies those aspects of autism. This has been examined by scads of researchers, with findings varying depending on the paradigm; most meta-analyses conclude that there is nothing flagrantly wrong with the formal features of mirror neuron function in autistic individuals.

  Thus, while mirror neurons’ activity correlates with attempts to understand other people’s actions, their involvement seems neither necessary nor sufficient and is most pertinent to low-level, concrete aspects of such understanding. As for mirror neurons being the portal for peering into someone’s soul and attaining super-duper understanding from within, I think things are best summarized by the title of Hickok’s well-received 2014 book The Myth of Mirror Neurons.39

  Which leads to the Wild West of mirror neuron–ology, with speculations that mirror neurons are essential to language, aesthetics, consciousness.40 Most of all, within two seconds of people first hearing about mirror neurons, they started writing reviews where the last paragraph would say something like “Wow, mirror neurons! How cool is that? This opens up all sorts of interesting avenues. Maybe they even explain . . . EMPATHY!”

  Sure, why not? Feeling someone’s pain is like mirroring their experience, feeling as if you are them. Tailor made, an irresistible idea. And in the decades since mirror neurons’ discovery, the “maybe they even explain empathy” reviews have continued. Gallese, for example, nearly twenty years into the mirror neuron era, speculates: “I proposed the mirroring could be a basic functional principle of our brain and that our capacity to empathize with others might be mediated by embodied stimulation mechanisms [i.e., mirroring].” Iacoboni, at the same time, writes, “Mirror neurons are likely cellular candidates for the core layer of empathy.” There have been some supportive hints—for example, people who self-report being particularly empathic show stronger mirror neuron–esque responses to matching movements. But for skeptics everything else is mere speculation.41

  That’s disappointing. But worse is people skipping over the “maybe” and concluding that mirror neurons have been proven to mediate empathy. Iacoboni, for example, mistakes correlation for causality: “Other studies, however, show that [PMC] activity correlates with empathy even when subjects watch grasping actions without overt emotional content. Thus, the mirror neuron activity is a prerequisite for experiencing empathy (my emphasis).”42

  A flagrant example of this is the neuroscientist Vilayanur Ramachandran of UC San Diego, one of the most flamboyantly creative people in the business, doing fascinating research on phantom limbs, synesthesia, and out-of-body experiences. He’s brilliant but has gotten a bit giddy with mirror neurons. A sampling: “We know that my mirror neurons can literally feel your pain.” He’s called them “the driving force behind the great leap forward” into human behavioral modernity sixty thousand years ago and famously said, “Mirror neurons will do for psychology what DNA did for biology.” I’m not trying to harp on Ramachandran, but how can you resist someone brilliant handing
out sound bites like calling mirror neurons “Gandhi neurons”? And this wasn’t just in the first heady days of mirror neurons in the early 1990s. Two decades later he stated, “I don’t think [the importance of mirror neurons for empathy is] being exaggerated. I think they’re being played down, actually.”43

  Ramachandran is certainly not alone. British philosopher Anthony Grayling has gone for the empathy link big time, writing, “We have a great gift for empathy. This is a biologically evolved capacity, as shown by the function of ‘mirror neurons.’” In a 2007 New York Times article about one man’s heroic actions to save another, those cells featured again: “People have ‘mirror neurons,’ which make them feel what someone else is experiencing” (emphasis added). And of course there was my daughter’s six-year-old classmate who, upon the class being complimented by their teacher for caring about the planet and cleaning up after their Earth Day cupcake celebration, shouted out, “It’s because our neurons have mirrors.”44

  I’d like to think that I’m being a maverick here, ahead of the crowd in terms of crucial thinking, but in recent years most in the field have charged overhype. Psychologist Gary Marcus of NYU calls mirror neurons “the most oversold idea in psychology,” philosopher and neuroscientist Patricia Churchland of UCSD calls them the “darling of the don’t-look-too-closely crew,” and Harvard’s Stephen Pinker concludes, “Mirror neurons do not, in fact, explain language, empathy, society, and world peace.”45 They simply haven’t been shown to have much to do with this chapter’s concerns.

  THE CORE ISSUE: ACTUALLY DOING SOMETHING

  The previous chapter considered the world of difference between highfalutin moral reasoning and whether, at a crucial juncture, someone actually does the right thing. As we saw, there is something consistent about that latter type of person: “What were you thinking when you leaped into that river to save the child?” “I wasn’t; before I knew it, I’d jumped in.” An act of implicit automaticity, the product of a childhood in which doing the right thing was ingrained as an automatic, moral imperative, light-years away from the frontal cortex calculating costs and benefits.

  We face a similar situation here, one that is the core of this chapter. Sympathy versus empathy, “for” versus “as if,” affect versus cognition, what we do versus what other species do—does any of this actually predict who does something compassionate to lessen someone’s pains? Similarly, does any of this predict whether the person acting compassionately acts effectively, and how much it’s an act of self-interest? As we’ll see, there is a yawning gap between being in an empathic state and acting effectively in a way that is truly selfless.

  Doing Something

  It is far from guaranteed that an empathic state leads to a compassionate act. One reason for this is captured superbly by the essayist Leslie Jamison:

  [Empathy] can also offer a dangerous sense of completion: that something has been done because something has been felt. It is tempting to think that feeling someone’s pain is necessarily virtuous in its own right. The peril of empathy isn’t simply that it can make us feel bad, but that it can make us feel good, which can in turn encourage us to think of empathy as an end in itself rather than part of a process, a catalyst.46

  In such a situation, saying “I feel your pain,” becomes a New Age equivalent of the unhelpful bureaucrat saying, “Look, I sympathize with your situation, but . . .” The former is so detached from action that it doesn’t even require the “but” as a bridge to the “there’s nothing I can/will do.” Having your pain validated is swell; having it alleviated is better.

  And there’s a broader reason why an empathic state may not produce action, first raised in chapter 6 when considering those strange creatures, adolescents. In that discussion I emphasized a wonderful feature of so many adolescents, namely the frenzied feeling of the world’s pains, but noted how that intensity often leads to little more than frenzied self-absorption. If instead of imagining how someone else is feeling (an other-oriented perspective), you are imagining how it would feel if this were happening to you (a self-oriented perspective), “you” has just come to the forefront and the main point is that feeling someone’s pain feels painful.

  The biological substrates of this are clear. Look at someone in pain with the instruction to take a self-oriented perspective, and the amygdala, ACC, and insular cortex activate, along with reports of distress and anxiety. Do the same with an other-oriented perspective, and all are less likely. And the more extreme the former state, the more likely that someone’s focus will be to lessen their own distress, to metaphorically look the other way.47

  This can be predicted with remarkable ease. Expose subjects to evidence of someone else in pain. If their heart rate increases a lot (a peripheral indicator of anxious, amygdaloid arousal), they are unlikely to act prosocially in the situation. The prosocial ones are those whose heart rates decrease; they can hear the sound of someone else’s need instead of the distressed pounding in their own chests.*48

  Thus, if feeling your pain makes me feel awful, I’m likely to just look out for number one, rather than helping you. Likewise if you’ve got your own issues. We saw this earlier, with the demonstration that if you increase people’s cognitive load, they become less prosocial toward strangers. Similarly, when people are hungry, they are less charitable—hey, quit bellyaching about your problems; my belly is aching. Make people feel socially excluded and they become less generous and empathic. Stress has the same effect, working via glucocorticoids; Mogil’s group (with my involvement) recently showed that if you use a drug to block glucocorticoid secretion, both mice and humans become more empathic toward strangers. Thus, if you feel highly distressed, whether due to resonating with someone else’s problems or because of your own, tending to your own needs readily becomes the priority.49

  In other words, empathic states are most likely to produce compassionate acts when we manage a detached distance. This brings to mind the anecdote from many chapters ago about the Buddhist monk I encountered who said that, yes, sometimes he cuts short his cross-legged meditation because of his knees, but not because he feels them hurting—“I do it as an act of kindness to my knees.” And this is certainly in line with the Buddhist approach to compassion, which views it as a simple, detached, self-evident imperative rather than as requiring vicarious froth. You act compassionately toward one individual because of a globalized sense of wishing good things for the world.*

  A handful of fascinating studies of Buddhist monks have been carried out, both by Richard Davidson of the University of Wisconsin and Tania Singer of the Max Planck Institutes in Germany. Remarkably, given the science-versus-religion culture wars, such work was given its, er, blessing and facilitated by the Dalai Lama, who is famously intrigued by neuroscience and who has said that if his Dalai Lama gig hadn’t come up, he would have wanted to be a scientist or engineer. The most publicized work revolves around the neuroimaging of Matthieu Ricard, a French-born Buddhist monk (who is the Dalai Lama’s French translator and who just happens to have a PhD in molecular biology from the Pasteur Institute—this is one interesting guy).50

  When confronted with examples of human suffering and instructed to empathically feel the pain of those people, Ricard showed activation of the same circuitry as you’d see in most everyone else. And it was extremely aversive—“The empathic sharing very quickly became intolerable to me and I felt emotionally exhausted,” he explained. When instead he did his Buddhist thing, focusing on thoughts of compassion, a totally different picture of activation emerged—the amygdala was silent, and instead there was heavy activation of the mesolimbic dopamine system. He described it as “a warm positive state associated with a strong prosocial motivation.”

  In other studies volunteers underwent either empathy training (focusing on feeling the pain of someone in distress) or compassion training (focusing on a feeling of warmth and care toward that distressed person).51 The former would generate the typical neuroimaging p
rofiles, including heavy amygdala activation, and a negative, anxious state. Those with compassion training did not, showing heavy activation instead in the (cognitive) dlPFC, coupling of activation between the dlPFC and dopaminergic regions, more positive emotions, and a greater tendency toward prosociality.

  Okay, caveats. This is a tiny literature (i.e., not much larger than the study of Ricard). And all-star Buddhist monks apparently meditate eight hours a day, not a trivial path to take. The point is merely to emphasize this scenario of detachment. Which brings us to the next issue, which is whether compassionate acts fostered by empathy are necessarily useful.

  Doing Something Effectively

  In a provocatively titled 2014 article, “Against Empathy,” Paul Bloom explored the ways in which empathy can lead to compassionate acts that are far from ideal.

  There is the realm of what has been termed “pathological altruism,” the type associated with codependency.52 This is the scenario of someone so consumed with the vicarious pain of a loved one that they endure and facilitate his dysfunction rather than administering tough love. Then there’s the danger that the empathic pain is so intense that you can only come up with solutions that would work for you, rather than ones that might help the sufferer. And there is the problem of empathy impeding your doing what’s necessary—it’s not great if a parent is so vicariously distressed by their child being distressed that they forgo vaccinations. A large piece of the training of health-care professionals is teaching them to keep empathy at bay.* For example, the various behavioral and neurobiological responses to seeing someone poked with a needle do not occur in acupuncturists. As Jamison describes, when anxiously seeing a doctor about something worrisome, “I needed to look at him and see the opposite of my fear, not its echo.”