The Age of Empathy

by Frans de Waal

  We identify with a high-wire artist to the point that we participate in every step he takes.

  I prefer the term Einfühlung since it conveys the movement of one individual projecting him- or herself into another. Lipps was the first to recognize the special channel we have to others. We can’t feel anything that happens outside ourselves, but by unconsciously merging self and other, the other’s experiences echo within us. We feel them as if they’re our own. Such identification, argued Lipps, cannot be reduced to any other capacities, such as learning, association, or reasoning. Empathy offers direct access to “the foreign self.”

  How strange that we need to go back one century to learn about the nature of empathy in the writings of a long-forgotten psychologist. Lipps offered a bottom-up account, that is, one that starts from the basics, rather than the top-down explanations often favored by psychologists and philosophers. The latter tend to view empathy as a cognitive affair based on our estimation of how others might feel given how we would feel under similar circumstances. But can this explain the immediacy of our reactions? Imagine we’re watching the fall of a circus acrobat and are capable only of empathy based on the recall of previous experiences. My guess is that we wouldn’t react until the moment the acrobat lies in a pool of blood on the ground. But of course this is not what happens. The audience’s reaction is absolutely instantaneous: Hundreds of spectators utter “ooh” and “aah” at the very instant that the acrobat’s foot slips. Acrobats often perform such slips on purpose, without any intention of falling, precisely because they know how much their audience is with them every step of the way. I sometimes wonder where Cirque du Soleil would be without this instant connection.

  Science is coming around to Lipps’s position, but this wasn’t the case yet when Swedish psychologist Ulf Dimberg began publishing on involuntary empathy in the early 1990s. He ran into stiff resistance from proponents of the more cognitive view. Dimberg demonstrated that we don’t decide to be empathic—we simply are. Having pasted small electrodes onto his subjects’ faces so as to register the tiniest muscle movements, he presented them with pictures of angry and happy faces on a computer screen. Humans frown in reaction to angry faces and pull up the corners of their mouths in reaction to happy ones. This by itself was not his most critical finding, however, because such mimicry could be deliberate. The revolutionary part was that he got the same reaction if the pictures flashed on the screen too briefly for conscious perception. Asked what they had seen after such a subliminal presentation, subjects knew nothing about happy or sad faces but had still mimicked them.

  Expressions on a screen not only make our face muscles twitch, they also induce emotions. Those who had been exposed to happy faces reported feeling better than those who had been exposed to angry ones, even though neither group had any idea of what they had seen. This means that we’re dealing with true empathy, albeit a rather primitive kind known as emotional contagion.

  Lipps called empathy an “instinct,” meaning that we’re born with it. He didn’t speculate about its evolution, but it’s now believed that empathy goes back far in evolutionary time, much further than our species. It probably started with the birth of parental care. During 200 million years of mammalian evolution, females sensitive to their offspring outreproduced those who were cold and distant. When pups, cubs, calves, or babies are cold, hungry, or in danger, their mother needs to react instantaneously. There must have been incredible selection pressure on this sensitivity: Females who failed to respond never propagated their genes.

  A good illustration is a female chimpanzee, named Krom, whom I knew at a zoo. Krom was fond of infants, and cared well enough for them so long as she could see them. Being deaf, however, she failed to respond to the soft yelps and whimpers of tiny infants in trouble, such as when they can’t reach the nipple, are in danger of losing their grip on Mom’s hair, or feel squeezed. I once saw Krom sit down on her infant and fail to get up when it burst out screaming. She reacted only upon noticing the worried reaction of other females. We ended up having another female adopt and raise this infant. Krom’s case taught me how critically important it is for a mammalian female to be in tune with her offspring’s every need.

  Having descended from a long line of mothers who nursed, fed, cleaned, carried, comforted, and defended their young, we should not be surprised by gender differences in human empathy. They appear well before socialization: The first sign of emotional contagion—one baby crying when it hears another baby cry—is already more typical of baby girls than baby boys. Later on we see more gender differences. Two-year-old girls witnessing others in distress treat them with more concern than do boys of the same age. And in adulthood, women report stronger empathic reactions than men, which is one reason why women have been attributed a “tending instinct.”

  None of this denies male empathy. Indeed, gender differences usually follow a pattern of overlapping bell curves: Men and women differ on average, but quite a few men are more empathic than the average woman, and quite a few women are less empathic than the average man. With age, the empathy levels of men and women seem to converge. Some investigators even doubt that in adulthood there’s much difference left.

  Nevertheless, to seek the origin of empathy in parental care seems logical, which is why Paul MacLean drew attention to the calls of young mammals that are lost and want their mother back, known as “separation calls.” The pioneering American neuroscientist, who in the 1950s first described the “limbic system,” was interested in the origin of parental care. Young mammals call when lost or frightened, and their mother responds by picking them up. She’s in a great hurry to take care of the problem, and if she’s big and strong, you just don’t want to get in her way (which is another human-versus-bear story). The evolution of attachment came with something the planet had never seen before: a feeling brain. The limbic system was added to the brain, allowing emotions, such as affection and pleasure. This paved the way for family life, friendships, and other caring relationships.

  The central importance of social bonding is hard to deny. We have a tendency to describe the human condition in lofty terms, such as a quest for freedom or striving for a virtuous life, but the life sciences hold a more mundane view: It’s all about security, social companionships, and a full belly. There is obvious tension between both views, which recalls that famous dinner conversation between a Russian literary critic and the writer Ivan Turgenev: “We haven’t yet solved the problem of God,” the critic yelled, “and you want to eat!”

  Our nobler strivings come into play only once the baser ones have been fulfilled. If attachment and empathy are as fundamental as proposed, we had better pay close attention to them in any discussion of human nature. There is also no reason to expect these capacities only in humans. They should manifest themselves in any warm-blooded creature with hair, nipples, and sweat glands, which is part of what defines a mammal.

  This obviously includes those pesky little rodents.

  Commiserating Mice

  I don’t particularly enjoy telling this story, since it betrays prejudice, but I think anyone can understand why the Dutch, in the aftermath of World War II, were less than enamored with their neighbors to the east. As an undergraduate at the University of Nijmegen, I was taught by several German professors, who spoke Dutch with a thick accent. One of them was a grumpy old man, who was said to have been a concentration camp guard. Obviously, this couldn’t be true, since he now would have been in jail, or worse, but this is what the rumor mill said.

  To make matters worse, this professor manually killed the mice needed for our anatomy practicum. He didn’t believe in death by ether, and would simply take a box with live mice and stand with his back turned to us. A few minutes later, a pile of dead mice with cracked necks lay on the counter.

  In his defense, I must say that “cervical dislocation,” as the practice is known, is probably quicker and more humane than other forms of euthanasia. But you can imagine that we found this professor a bi
t scary. And that’s just us. How did the mice look at this procedure? The first mouse from the box didn’t know what was coming, but what about the last one? Can rodents detect one another’s pain? Do they feel one another’s pain?

  Before going any further, I must warn that reading up on the science of animal empathy can be a challenge for animal lovers. To see how animals react to the pain of others, investigators have often produced the pain themselves. I don’t necessarily approve of these practices, and don’t apply them myself, but it would be foolish to ignore the discoveries they’ve produced. The good news is that most of this research was carried out decades ago, and is unlikely to be repeated today.

  In 1959, American psychologist Russell Church published a scientific paper under the provocative title “Emotional Reactions of Rats to the Pain of Others.” Church trained rats to obtain food by pressing a lever, and found that if one rat noticed that lever-pressing shocked a neighboring rat, it would stop. This is remarkable. Why shouldn’t the rat simply continue to get food and ignore its companion dancing in pain on an electric grid? Did these rats stop pressing because they were distracted, worried about their companion, or fearful for themselves?

  The explanation offered by Church was typical of a time when conditioning was thought to underlie all behavior. He argued that a rat fears for its own well-being when it sees a companion in distress. But does an untrained rat have any reason to associate the squeals of others with pain to itself? The animals in the experiment had grown up in the laboratory with controlled temperature and light, ample food, and no predators. They had never encountered a situation like this before. It seems more likely that the sight, sound, or smell of another rat in pain arouses an innate emotional response. One rat’s distress may simply distress another.

  This study inspired a brief flurry of experiments on animal “empathy,” “sympathy,” and “altruism”—always put between quotation marks so as to avoid the wrath of behaviorists, who didn’t believe in such concepts. This work was subsequently ignored, due partly to the taboo on animal emotions, and partly to the traditional emphasis on the nasty side of nature. As a result, animal studies are now seriously lagging behind what we know about human empathy. This may be changing, though, thanks to a new study by Canadian scientists, titled “Social Modulation of Pain as Evidence for Empathy in Mice.” This time, the word empathy is free of quotation marks, reflecting the growing consensus that emotional linkage between individuals has the same biological basis in humans and other animals.

  The news came too late for my old anatomy professor, but Jeffrey Mogil, head of the pain lab at McGill University, almost felt as if his mice were talking to one another about their pain. He was puzzled time and again by the fact that when he was testing mice from the same cage, the order in which they were used seemed to affect their response. The last mouse showed more signs of pain than the first. One possibility is that the last mouse was sensitized by having seen others in pain. Mogil compared it to sitting in a dentist’s lobby and seeing other patients coming out of the room after an obviously unpleasant experience. One can’t help becoming primed for pain.

  Pairs of mice were put through a pain test. Each mouse was placed in a transparent glass tube such that it could see the other. Either one or both mice were injected with diluted acetic acid, known to cause—in the words of the investigators—a mild stomachache. Mice respond to this treatment with stretching movements, suggesting discomfort. The basic finding was that a mouse would show more stretching with an injected partner, who was stretching, too, as opposed to a control partner. Since this applied only to mice that were cage mates, not to strangers, it couldn’t be due to a simple negative association, because then the reaction should have been the same regardless of whether they knew each other. Further experiments explored which sense was involved by comparing anosmic mice (which lack olfaction), deaf mice, and mice that were prevented from seeing each other. Vision turned out to be critical: The reaction occurred only between mice that could see each other.

  The mice showed pain contagion. That is, the sight of another in pain intensified their own pain response. Interestingly, in the presence of a stranger in pain, sensitivity went down: The mice became strikingly passive. This counterempathic reaction, however, was restricted to males, which are also potentially the most hostile to one another. Were they less than empathetic with their rivals?

  This gender effect reminds me of how humans empathize with another’s distress. Seeing the pain of a person we have just cooperated with activates pain-related areas in our own brains. This applies to both men and women. Yet in some studies the same procedure has been followed with partners who had been instructed to act unfairly in a game with the subject before the latter went into the brain scanner. Having been duped by someone, we show the opposite of empathy: At our seeing his pain, the brain’s pleasure centers light up. We’re getting a kick out of their misery! Such Schadenfreude occurs only in men, however, because women remain empathic. This may seem a typically human reaction, yet the underlying theme (male lack of empathy for potential rivals) resembles the mouse findings, and might well be a mammalian universal.

  Finally, the investigators exposed pairs of mice to different sources of pain. One was the same acetic acid as before, but the second was a radiant heat source that might burn them if they came too close. Mice observing a cage mate in pain from the acid withdrew more quickly from the heat source, thus indicating heightened sensitivity to a completely different pain stimulus that required a different reaction. This precludes motor mimicry as an explanation: The mice seemed sensitized to pain in general. Any pain.

  This study goes a long way toward reviving the tentative conclusions of the 1960s, showing that even with larger numbers of subjects and more rigorous methods, we get the same result: intensification of one’s own experience based on the perceived reaction of others. This is close enough to “empathy” to call it that.

  It’s obviously not the imaginative kind of empathy that makes us truly understand how someone else feels, even someone we don’t see, for example, when we read about the fate of a character in War and Peace. Yet it’s good to keep in mind that imagination is not what drives empathy. Imagining another’s situation can be a cold affair, not unlike the way we understand how an airplane flies. Empathy requires first of all emotional engagement. The mice show us how things may have gotten started. Seeing another’s emotions arouses our own emotions, and from there we go on constructing a more advanced understanding of the other’s situation.

  Bodily connections come first—understanding follows.

  Oscar the Cat

  Oscar the Cat stares at us from a photograph in the prestigious New England Journal of Medicine along with an admiring description by a fellow expert. The author relates how Oscar makes his daily rounds at a geriatric clinic in Providence, Rhode Island, for patients with Alzheimer’s, Parkinson’s, and other illnesses. The two-year-old cat carefully sniffs and observes each patient, strolling from room to room. When he decides that someone is about to die, he curls up beside them, purring and gently nuzzling them. He leaves the room only after the patient has taken his or her last breath.

  Oscar’s predictions have been so dependable that the hospital staff counts on them. If he enters a room and leaves again, they know the patient’s time isn’t up yet. But as soon as Oscar starts one of his vigils, a nurse will pick up the phone to call family members, who then hurry to the hospital to be present while their loved one passes away. The cat has predicted the deaths of more than twenty-five patients with greater accuracy than any human expert. The tribute to the tomcat states: “No one dies on the third floor unless Oscar pays a visit and stays awhile.”

  How does Oscar do it? Is it the smell, skin color, or a certain pattern of breathing of dying patients? With so much variation in what ails them, it seems a bit unlikely that all patients end up showing the same telltale signs, but it’s a possibility. Even more baffling is the question of what drives the cat.
He has sometimes been the only one to be with an expiring patient, and the staff interprets this as him giving succor. But is this really what motivates our feline hospice?

  I see two possible reasons for his behavior: It is either an attempt to comfort himself, if he’s upset by what he senses is happening to a person, or an attempt to comfort the patient. But both possibilities remain perplexing. The first is so because it’s unclear why Oscar would seek comfort with patients who have mostly become incapacitated: Wouldn’t he be better off getting petted by some of the many people who’d love to do so? The second possibility is even harder to believe: Belonging to a species of solitary hunters, why would Oscar be so much more generous than any other cat I’ve ever known? I’ve had many in my lifetime, and whereas most cats do like to snuggle, I don’t read much concern about our well-being in their behavior. To be perfectly cynical: I sometimes wonder why our cats love us so much more the colder it gets.

  I’m exaggerating, of course. Cats do give affection and can show strong emotional connectedness. Otherwise, why would they always want to be in the same room we are in? The whole reason people fill their homes with furry carnivores and not with, say, iguanas or turtles—which are easier to keep—is that mammals offer us something no reptile ever will: emotional responsiveness. Dogs and cats have no trouble reading our moods and we have no trouble reading theirs. This is immensely important to us. We feel so much more at ease, so much more attached to animals with this capacity. Even if Oscar wasn’t exactly acting out of concern, as I surmise, it would still be a mistake to dismiss his behavior as irrelevant to the issue of empathy.

  Every evolved capacity is assumed to have advantages. If emotional contagion was indeed the first step on the road toward full-blown empathy, the question is, how does it promote survival and reproduction? The usual answer is that empathy produces helping behavior, but this hardly works for emotional contagion, which by itself doesn’t do so. Take the typical reaction of a human toddler who hears another child cry. Her eyes fill with tears, upon which she runs to a parent to be picked up and comforted. In doing so, she in fact turns her back on the source of discomfort. Due to this lack of other-orientation, psychologists speak of “personal distress.” It is a self-centered response that doesn’t provide a good basis for altruism.