Against Empathy: The Case for Rational Compassion

by Paul Bloom

Our own acts that upset others are innocent or forced; the acts that others do to upset us are crazy or cruel.

Pinker refers to this kind of thing often too, and how it can lead to an endless escalation of violence in an 'eye for an eye' world.

But when it comes to serious acts, it’s almost always the case that the ramifications are worse for the victims than for the perpetrators.

If you want to think about evil, real evil, a better way to proceed is this: Don’t think about what other people have done to you; think instead about your own actions that hurt others, that made others want you to apologize and make amends. Don’t think about other nations’ atrocities toward your country and its allies; think instead about the actions of your country that other people rage against. Your response might be: Well, none of that is evil. Sure, I did some things that I regret or that others blame me for. And yes, my country might have done ugly things to others. But these were hard choices, tough calls, or perhaps honest mistakes, never the consequence of some sort of pure malice. Precisely. This is how people typically think of their past evil acts.

A good rule of thumb

Indeed, some argue that the myth of pure evil gets things backward. That is, it’s not that certain cruel actions are committed because the perpetrators are self-consciously and deliberatively evil. Rather it is because they think they are doing good. They are fueled by a strong moral sense.

One’s perspective matters a lot in these cases. After the attacks on the twin towers on September 11, 2001, some Palestinians celebrated in the streets, a reaction that many Westerners took to reflect moral depravity. But when Americans celebrated after the killing of Osama bin Laden in 2011, or when Israelis hooted and cheered as bombs dropped over Gaza in 2014, the celebrants didn’t think they were doing anything shameful at all.

We’ve seen clear cases where violence and cruelty are motivated by moral views. But often this is not the case.

Not everyone is willing to make others suffer in order to achieve what he or she wants. Perhaps empathy provides the brakes. Greed makes us want to knock someone down and take their money; empathy holds us back. Anger makes us want to respond to an insult by punching someone in the face; empathy restrains us.

I concede that empathy can serve as the brakes in certain cases. But I will argue here that it’s just as often the gas—empathy can be what motivates conflict in the first place. When some people think about empathy, they think about kindness. I think about war.

Much of this makes sense. If we were to have empathy for our enemies, it would block us from hurting them. Unfortunately, though, this isn’t how empathy works. Consider what happens when a country is about to go to war. Do leaders gain support by making rational arguments with statistical assessments of costs and benefits? Is the decision driven by the sort of “unempathic cost-benefit calculation” that Baron-Cohen complains about? Does this cold-blooded calculation explain the psychology of those who supported either side of the conflict in Gaza—or the American invasion of Iraq? Not so much. What is more typical is that people feel deeply about the crimes done in the past toward their families or compatriots or allies. Consider how the Israeli reaction to the news of three murdered Israeli teenagers spurred the attacks on Gaza, or how Hamas and other organizations used murdered Palestinians to generate support for attacks against Israel. If you ask a proponent on either side why they are killing their enemies’ children, they don’t spout the sort of bureaucratic number crunching that Baron-Cohen worries about. They more often talk about the harm that’s been done to those they love.

This is an excellent point, and failure to understand this is what leads to so many infuriating takes lamenting the modern coldness of war. I would prefer people to be cold about war than emotionally invested in it! The idea that a calculated drone strike from afar is somehow worse than chopping an enemy down, feeling the blood spray on your face, is utterly absurd and the worst of all the romantic impulses.

Asking people to feel as much empathy for an enemy as for their own child is like asking them to feel as much hunger for a dog turd as for an apple—it’s logically possible, but it doesn’t reflect the normal functioning of the human mind.

Also, in this case and so many others, empathy is not sufficient to guide moral action. In the end, individuals who wish to do good have to be consequentialists at least to some degree, doing the sort of cost-benefit calculation that Baron-Cohen derides. Suppose that prior military action could have stopped Hitler from killing millions in concentration camps. I believe it would have been morally right to engage in such action even thought it surely would have led to the death of innocent people. If Baron-Cohen agrees with me here, then he too recognizes the limits of empathy and the value of cost-benefit calculation.

Indeed, sometimes the right thing to do involves allowing one’s own citizens to die. In World War II, the British military had cracked the Enigma code and had advance notice of impending German attacks on Coventry. But if they prepared for the attacks, the Germans would have known the code was cracked. So Churchill’s government made the hard choice to let innocent people die in order to retain a military advantage, giving them a better chance of winning the war and saving a greater number of innocent lives.

Excellent point, and a great example of how empathy would have been worse than rational compassion (though it is arguable - and in my view more likely - that it was strategic rationality rather than rational compassion)

When scholars think about atrocities, such as the lynchings of blacks in the American South or the Holocaust in Europe, they typically think of hatred and racial ideology and dehumanization, and they are right to do so. But empathy also plays a role. Not empathy for those who are lynched or put into the gas chambers, of course, but empathy that is sparked by stories told about innocent victims of these hated groups, about white women raped by black men or German children preyed upon by Jewish pedophiles.

Bit of a reach maybe but this is not unfair - particularly when one takes into consideration Bloom's previous (well-made) points on how empathy is inherently and unavoidably biased.

Or think about contemporary anti-immigrant rhetoric. When Donald Trump campaigned in 2015, he liked to talk about Kate—he didn’t use her full name, Kate Steinle, just Kate. She was murdered in San Francisco by an undocumented immigrant, and Trump wanted to make her real to his audience, to make vivid his talk of Mexican killers. Similarly, Ann Coulter’s recent book, Adios, America, is rich with detailed descriptions of immigrant crimes, particularly rape and child rape, with chapter titles like “Why Do Hispanic Valedictorians Make the News, But Child Rapists Don’t?” and headings like “Lost a friend to drugs? Thank a Mexican.” Trump and Coulter use these stories to stoke our feelings for innocent victims and motivate support for policies against the immigrants who are said to prey upon these innocents.

I am naturally very sympathetic to this reasoning! But it does seem like a good point in how empathy can override reason and lead to immorality.

I’m not a pacifist. I believe that the suffering of innocents can sometimes warrant military intervention, as, again, in the decision by the United States to enter World War II. But empathy tilts the scale too much in favor of violent action. It directs us to think about the benefits of war—avenging those who have suffered, rescuing those who are at further risk. In contrast, the costs of war are abstract and statistical, and a lot of these costs fall upon those we don’t care about and hence don’t empathize with. Once the war is under way, one can try to elicit empathy for those who have suffered, particularly those on one’s own side, because now the costs have become tangible and specific. But by then, it’s often too late.

Yes, this is good - empathy will cut both ways during a war (see the retrospective justification of torture on the basis of injuries to American soldiers) and in the lead up, will naturally side with selling scores, as the other side hadn't yet suffered.

They first asked people to describe a time in the past year when someone they were close to was mistreated, either physically or psychologically. They asked their subjects how attached they felt to this victim and then asked them whether they aggressively confronted the person who caused this mistreatment. As predicted, the more warmly they felt toward the victim, the more aggressive they said they were, consistent with a connection between empathy and violence.

Subjects were told about a math competition for a twenty-dollar prize between two students, described as strangers, who were currently in another room of the laboratory. They then read an essay purportedly written by one of the students, which described her financial problems—she needed to replace a car and pay for class registration. The subjects were then told that they were involved in an experiment that explored the effect of pain on performance, and to make everything random they would get to choose how much pain to administer—by choosing a dosage of hot sauce—to the student the financially needy student was competing with. The trick here concerned how the essay purportedly written by the student ended. As in the Batson studies we talked about earlier, some of the subjects read a passage designed to elicit empathy (“I’ve never been this low on funds and it really scares me”), while others did not (“I’ve never been this low on funds, but it doesn’t really bother me”). As predicted, greater amounts of hot sauce were assigned to the competitor when the person was described as distressed. Keep in mind that this competitor didn’t do anything wrong; he or she had nothing to do with the student’s anxiety about money. Interestingly, the studies by Buffone and Poulin also found that there was a greater connection between empathy and aggression in those subjects who had genes that made them more sensitive to vasopressin and oxytocin, hormones that are implicated in compassion, ...

I’ve come up with similar findings in a series of studies done in collaboration with Yale graduate student Nick Stagnaro. We tell our subjects stories about terrible events, about journalists kidnapped in the Middle East, about child abuse in the United States. And then we ask them how best to respond to those responsible for the suffering. In the Middle East case, we give a continuum of political options, from doing nothing, to engaging in public criticism, all the way up to a military ground invasion. For the domestic version, we ask about increased penalties for the abuser, from raising their bail to making them eligible for the death penalty. Then we give people Baron-Cohen’s empathy scale. This has its problems, as discussed earlier, but it should give some rough approximation of how empathic people are. Just as with the genetic study of Buffone and Poulin, we found that the more empathic people are, the more they want a harsher punishment.

As we’ve seen, this isn’t always the case. Often people who commit terrible acts are empathic and caring in other parts of their lives. One manifestation of this, often pointed out by those who want to mock vegetarians, was the concern that many Nazis had for nonhuman animals. Hitler famously loved dogs and hated hunting, but this was nothing compared to Hermann Göring, who imposed rules restricting hunting, the shoeing of horses, and the boiling of lobsters and crabs—and mandated that those who violated these rules be sent to concentration camps! (This was the punishment that he imposed on a fisherman for cutting up a live frog for bait.) Or take Joseph Goebbels, who said, “The only real friend one has in the end is the dog …. The more I get to know the human species, the more I care for my Benno.”

But then again, some Nazis really did seem to revel in their cruelty, and some of the atrocities done at the time of the Holocaust were done with enthusiasm and relish. I said earlier that sadists are rare, but if they do exist, they were probably overrepresented among, say, concentration camp guards. Certain individuals seem to be drawn to violent conflicts, showing up not because of ideological, religious, or political commitments, but because they enjoy torturing, raping, and killing people.

Dr. Josef Menengle comes to mind here

So is lack of empathy the core deficit that underlies psychopathy, that makes psychopaths psychopaths? There are reasons to doubt it. For one thing, as Jesse Prinz points out, it’s not that psychopaths suffer from a specific empathy deficit. Rather, they might suffer from a blunting of just about all the emotions.

For Prinz, this raises the question of whether the nastiness of psychopaths has anything special to do with empathy, as opposed to arising from, or being associated with, an overall limited emotional life.

So here’s what we can say about psychopaths and empathy: They do tend to be low in empathy. But there is no evidence that this lack of empathy is responsible for their bad behavior.

A limited but fair statement

One decisive test of the low-empathy-makes-bad-people theory would be to study a group of people with low empathy but without the other problems associated with psychopathy.

Are they monsters? They are not. Baron-Cohen points out that they show no propensity for exploitation and violence. Indeed, they often have strong moral codes. They are more often the victims of cruelty than its perpetrators.

Well, this would seem to provide a weak reason to think an empathy deficit is not the issue with psychopaths - though as before, if empathy merely 'activates' pre-existing kindness, maybe the lack of it will prevent inhibition of pre-existing wickedness?

No discussion of cruelty and violence would be complete without considering dehumanization—thinking about and treating other people as if they are less than fully human. This is the cause of much of the cruelty in the world.

A search of racist websites can easily find contemporary examples of this, of blacks, Jews, Muslims, and other members of despised groups being talked about as if they were nonhuman animals, lacking deep feelings and higher intellectual powers. In laboratory studies, researchers have found that people are prone to think of members of unfamiliar or opposing groups as lacking emotions that are seen as uniquely human, such as envy and regret. We can see them as akin to savages or, at best, as children.

Dehumanization is indefensible. It’s obviously mistaken to think about blacks or Jews or women as lacking critical human traits like agency and self-determination and rich emotional lives, and it is a mistake that can have terrible consequences, motivating and excusing indifference and cruelty. For some people, this is why empathy is so important. Empathy blocks dehumanization and allows us to see people as they really are. If so, this would be a powerful case in its favor.

I think that empathy is not needed to treat people as people; it is not an essential aspect of avoiding dehumanization.

However, it could be a powerful adjunct factor, thus making dehumanization easier in its absence, though not impossible. This would not be a much weaker case for empathy compared to a situation where it was essential.

Sure, the Palestinians are literally described as dogs. But this taunting would be odd behavior if the Israelis actually did think of them as dogs because, really, what would be the point?

To reinforce the sense in the face of evidence to the contrary? This makes total sense to me and doesn't hit me as strange at all.

Kate Manne makes a similar argument in her discussion of the aftermath of a police shooting in Ferguson, Missouri, where police officers screamed at protesters, “Bring it, you fucking animals, bring it!” For Manne, this can best be seen not as a failure to acknowledge the protesters’ humanity, but as “a slur and a battle cry,” as an “insult that depends, for its humiliating quality, on its targets’ distinctively human desire to be recognized as human beings.”

But again, if the purpose of a battle cry is to prepare for a fight, then reinforcing the inhumanity of your opponents makes total sense and fits with dehumanization quite neatly.

While much of what happened during acts of mass killing did reflect thinking of the Jews as less than human, some of the actions prior to this—the various humiliations and degradations of Jews in the Ukraine, for instance, and the delight that people took in this—reflect an appreciation of the humanity of those who were being tormented. If you don’t think of them as initially possessing dignity, where’s the pleasure in degrading them?

Again, this seems very strange. People can obtain pleasure from torturing animals, and directions if indignity can again be about reinforcing the idea that your victims are not human.

I’m framing this point as an alternative to Smith’s dehumanization analysis. But in response, Smith points out that this sort of degrading treatment, while not reflecting dehumanization, might reflect a wish to dehumanize, a desire to bring people down to the point where they are seen, and will see themselves, as less than human. Calling people “dogs” and “animals,” then, is more than just insult; it’s different from saying that someone is ugly and stupid. It’s an attempt to shift how these people are thought of.

Okay, good that my objections aren't novel to Bloom!

Is lack of empathy another force that supports dehumanization? I think not. There is a big difference between actively denying someone’s human traits—dehumanization—and not thinking about these human traits but instead focusing on other aspects of the person. The first is terrible; the second is not. To elaborate on this, consider some examples. A couple is lying in bed and the woman uses her partner’s stomach as a pillow. Or a man in a crowd moves behind someone to keep the sun out of his eyes. Or a host is having several people over for dinner and needs to figure how much food to order from Royal Palace and where to put the chairs around a too-small table. This can all be done without considering people’s thoughts and feelings, by literally thinking of people in the same way that one would think of objects. But none of this is immoral.

A weak response. I think this is an important chink in Bloom's armour.

Similarly, I’ve been arguing throughout this book that fair and moral and ultimately beneficial policies are best devised without empathy. We should decide just punishments based on a reasoned and fair analysis of what’s appropriate, not through empathic engagement with the pain of victims.

This is correct but tangential. Bloom has failed to adequately address this point in my view.

Owen Flanagan once described a meeting with the Dalai Lama in which he asked the leader of the Tibetan Buddhists a great question: If it would stop the Holocaust, would you kill Hitler? “The Dalai Lama turned to consult the high lamas who were normally seated behind him, like a lion’s pride. After a few minutes of whispered conversation in Tibetan with his team, the Dalai Lama turned back to our group and explained that one should kill Hitler (actually with some ceremonial fanfare, in the way, to mix cultural practices, a Samurai warrior might). It is stopping a bad, a very bad, karmic causal chain. So ‘Yes, kill him. But don’t be angry.’” The Dalai Lama is conceding that a rational and caring individual is going to have to engage in, or at least support, certain acts of violence, including murder. But he sees it as a necessary evil, a last resort. If there were some way to stop that very bad karmic chain without violence, that would be better. This is not the perspective of an angry person—anger feeds off the suffering of others; an angry person wants wrongdoers to suffer.

I actually really like this. If you are to kill Hitler, it should be a regretful act out of compassion for his victims, not out of revenge and anger.

So when it comes to my imaginary genetically engineered child, I would put in some anger, but not too much, and I would make sure to add plenty of intelligence, concern for others, and self-control. I would be wary of removing anger altogether, but I would ensure that it could be modified, shaped, directed, and overridden by rational deliberation; that, at most, it could be a reliable and useful servant—but never a master. That’s how we should think about empathy.

This is not a bad view. Empathy, but with limitations. It is weaker that the thesis initially presented by Bloom, but arguably this form of empathy would be different enough in important ways as to warrant a separate analysis.

One might also believe that some partiality makes sense in a personal context—if my child and a stranger were drowning and I could save just one, I’d save my child, and I don’t feel that this is the wrong choice. So the partiality of empathy and other psychological processes might be morally appropriate at least some of the time. These are concerns worth taking seriously, and I’ve tried to respond to them throughout this book.

I think at this stage, my take is that Paul Bloom has a good case against empathy in many contexts, particularly in considerations of morality, but his concessions in some places and weak responses in others make his overall thesis inadequately support and his arguments only support a more limited case than he thinks.

Then, because everyone loves a surprise ending, I’ll finish off by saying some nice things about empathy.

Heh, you already have!

David Eagleman makes this argument with a series of striking examples. He tells the story of how, in 2000, an otherwise normal Virginia man started to collect child pornography and make sexual overtures toward his prepubescent stepdaughter. He was sentenced to spend time in a rehabilitation center only to be expelled for making lewd advances toward staff members and patients. The next step was prison, but the night before he was to be incarcerated, severe headaches sent him to the hospital, where doctors discovered a large tumor in his brain. After they removed it, his sexual obsessions disappeared. Months later, his interest in child pornography returned, and a scan showed that the tumor had come back. Once again it was removed, and once again his obsessions disappeared.

I'd heard of this case before and it is truly fascinating. And it is indeed a difficult one for Cartesian dualists.

Eagleman would argue that this distinction is an illusion. Tumor Man is not a bizarre anomaly; he is just a case where the determined nature of behavior is particularly obvious.

Yes - it is merely that for 'normal' people, the neurological causes of our behaviour are less clear.

Blame the brain, not the person!

Ah, but the brain is the person (which I think Bloom basically agrees with)

Rather, I’m making the distinction in a different way. My suggestion is that cases like Tumor Man are special because they involve actions that are disengaged from the normal neural mechanisms of conscious deliberation. One way to see this is that when people in these states are brought back to normal—the tumor is removed, the drug wears off—they feel that their desires and actions were alien to them and fell outside the scope of their will. Accordingly, such individuals in their altered states are less responsive to carrots and sticks: Even the threat of imprisonment did not slow down Tumor Man, because the part of his psyche that motivated his sexual behavior was disengaged from the part of his psyche that computed the long-term consequences of his actions.

Yeah, this is fair.

To see this, imagine two computers. One behaves randomly and erratically; it doesn’t have a rational bone in its mechanical body. The other is a deliberating cost-benefit analyzer. Plainly, both are machines: no souls here. Yet they are as different as can be. The question that remains for the psychologist is: What kind of computer are we? Or better than that—since the answer here is plainly both—to what extent are we irrational things and to what extent are we reasoning things?

Still, even the most robust and impressive demonstrations of unconscious or irrational processes do not in the slightest preclude the existence of conscious and rational processes. To think otherwise would be like concluding that because salt adds flavor to food, nothing else does.

A while ago, John Macnamara pointed out that the discovery of these failures of reason reveal two very different things about our minds. Most obviously, they illustrate irrationality, how things go wrong, how we are limited. But they also illustrate how intelligent we are, how we can override our biases. After all, we know that they are mistakes!

Much of this book has been observing this dynamic. Just as one example among many, yes, we often favor those who are adorable more than those who are ugly. This is a fact about our minds worth knowing. But we can also recognize that this is the wrong way to make moral decisions. It’s this ability to critically assess our limitations—with regard to our social behavior, our reasoning, and our morality—that makes all sorts of things possible.

But when it comes to intelligence, there is a law of diminishing returns. The difference between an IQ of 120 and an IQ of 100 (average) is going to be more important than the difference between 140 and 120. And once you pass a certain minimum, other capacities might be more important than intelligence.

Indeed, high intelligence is not only related to success; it’s also related to good behavior. Highly intelligent people commit fewer violent crimes (holding other things, such as income, constant), and the difference in IQ between people in prison and those in the outside world is not a subtle one. There is also evidence that highly intelligent people are more cooperative, perhaps because intelligence allows one to appreciate the benefits of long-term coordination and to consider the perspectives of others.

Reason and rationality, then, are not sufficient for being a good and capable person. But my argument is that they are necessary, and on average, the more the better. It’s not just intelligence, however. I said that if you were curious about what sort of person a child would grow up to be, an intelligence test would be a great measure. But there’s something even better. Self-control can be seen as the purest embodiment of rationality in that it reflects the working of a brain system (embedded in the frontal lobe, the part of the brain that lies behind the forehead) that restrains our impulsive, irrational, or emotive desires.