The difficulty with publishing a nonfiction book is that you need a novel angle. It would be tough to get a contract to write something that simply expresses some basic and obvious truth, like Hooray for Love or Democracy: Isn’t It Nifty? Instead, you need the “counterintuitive” perspective, the contrarian #SlatePitch that demonstrates why that thing you thought was good was actually bad, or vice versa. Thus we get books like Against Democracy, Against Love, and even Against Everything. The market pressure for constant novelty, which operates similarly in academia and in publishing, creates a dangerous incentive toward trying to say things that are eye-catching rather than things that are true.

Paul Bloom isn’t, despite what the title of Against Empathy might imply, against empathy. In fact, towards the end of the book, he even says that he “worr[ies] that I have given the impression that I’m against empathy.” The title is meant more for the purpose of making people angry and selling books than for clearly communicating the book’s central point. Instead, Against Empathy is simply a book about why moral decisions should be based on reason rather than emotion. This is a far less controversial position than “empathy is bad,” but “we should try to be reasonable” is not a thesis that will land you on the bestsellers list, so Against Empathy it is. (Plus, as Bloom recounts, this allowed him to have tremendous fun by being a dick at parties, telling people he was working on a book about empathy, and then following up with “I’m against it” and watching them gasp.)

Bloom has said he is “not against empathy in general, just empathy as a moral guide.” This is not, of course, the same as being against empathy. But even here, Bloom is misleading: the thing he calls “empathy” throughout the book has little to do with our ordinary use of the term. Bloom says “empathy” is not the same thing as “compassion,” even though many people understand the word that way. In fact, Bloom’s definition of empathy is downright peculiar: he makes clear that what he is condemning is “emotional empathy,” by which he means “feeling what other people are feeling.” He distinguishes this from “cognitive empathy,” which he defines as trying to understand other people’s perspectives.

To those of us, like myself, who argue that people should generally have “more empathy” for one another, Bloom’s version of “empathy” is unrecognizable. I have always understood empathy to mean “trying to imagine what it is like to be other people” so that we can compassionately understand where they are coming from. But Bloom defines it as literally feeling other people’s emotions, suffering when they suffer, being distressed when they are distressed, etc. It’s possible to frame these two definitions so that they seem similar: both can be described using the expression “put yourself in another person’s shoes.” But the version in which we literally feel as if we are other people takes us far afield from the term’s everyday usage.

Because Bloom defines empathy as “experiencing what other people experience” rather than “imagining other people’s experiences for the purposes of better understanding and caring about them,” he is able to offer absurd caricatures of the pro-empathy position. He suggests that an “empathetic doctor” would be a bad doctor, because an empathetic doctor would be in pain while their patients were in pain, and this would inhibit their ability to offer good treatment. But does anyone who advocates having empathetic doctors believe they should literally feel as if they have whatever ailment the patient has? The point is not that you should literally experience what another person does (partly because, in the absence of swapping bodies, it’s not actually possible to experience someone else’s experiences), but that you may be better at caring for someone if you have gone through the exercise of imagining what it might be like to be them. Similarly, Bloom points out that we can “be concerned about starving people without having a vicarious experience of starving,” and that if we comfort a child who is afraid of a thunderstorm or a barking dog, we don’t literally need to be afraid of thunderstorms and barking dogs. For Bloom, the fact that we are not experiencing starvation or fear when we think about people who are hungry or afraid means that “there’s no empathy there.” That’s only true, though, if we adopt his bizarre definition of empathy. Of course, it’s true that if I comfort my child during a thunderstorm, I don’t need to be afraid of thunderstorms myself, and if I worry about starvation, I don’t need to feel hungry. But in order to understand why I should care about fear and hunger, it might help if I thought about what it feels like for a person who is experiencing those things. Pro-empathy people like myself do not advocate actually trying to become afraid of thunderstorms in order to understand how small children feel about them, but rather spending time remembering what it is like to be a small, scared child, in order to appreciate why thunderstorms might frighten them.

The distinction can be illustrated by thinking about how literature works. When people say that reading novels allows us to empathize with different perspectives, they do not mean empathy in Bloom’s sense of actually feeling the person’s feelings. When I am reading about the protagonists in The Stranger or Crime and Punishment, I do not literally feel as if I have murdered someone. But I do get a better understanding of what it might feel like to be someone like Raskolnikov or Mersault. Empathy, in the sense in which people advocate it, is not a plea for people to weep when they see others weeping, but to better appreciate what it is like to weep. It is not, as Bloom says, that when somebody else’s child is killed I literally feel as if my own child has been killed. It is that I do my best to imagine what the experience is like for the person experiencing it, so that I am able to deal sensitively and humanely with the person who is going through it.

Even though I think “empathy as an exercise in imagining other people’s perspectives” is both a workable and natural definition of the term, Bloom firmly rejects it. He sticks to his belief that an empathetic psychotherapist would be a psychotherapist who actually has depression, rather than a psychotherapist who has previously had depression or who seems to appreciate what it is like to be depressed. The latter, he says, has “nothing to do with” empathy, even though it’s the only version of the idea of the “empathetic psychotherapist” that even makes any sense. Bloom says that when we talk about empathy as simply “understanding other minds,” we are talking about something called “cognitive empathy,” which is different from the “emotional empathy” he is against. “Cognitive” empathy, he says, is morally neutral: it’s what psychopaths, con men, and seducers do, since they “understand” the workings of other people’s minds very well indeed. But since people who advocate that kind of empathy also advocate being a compassionate person, it’s hard to see how this matters.

Frankly, I think that because Paul Bloom defines empathy in such an idiosyncratic way, Against Empathy is the classic case of arguing against a “straw man” position that few people actually hold, and avoiding the much stronger case that people actually do make. Bloom says he “hates” terminological disputes, and that we should examine the concept being described rather than the word used to describe it, but this is like writing a book called Against Science and then revealing in the introduction that you are actually only against giving scientists dictatorial powers to run all human affairs, a position held by very few people. It is very easy to make a persuasive argument that doctors shouldn’t be crying all the time, but the only way such an argument would be useful is if anyone actually believed doctors should be crying all the time. Furthermore, since I think that the whole reason Bloom has to adopt a useless definition of empathy is because it’s the only way the Against Empathy framing can even work, this book is an intellectually dishonest exercise from the start, born from a desire to provoke rather than to reach clarity.

But for a moment, let’s set aside the fact that the entire premise of the book is both sneaky and useless, and examine the substance of Bloom’s anti-“empathy” argument. It runs roughly as follows: “empathy,” defined as feeling other people’s emotions, is neither useful or necessary in grounding human morality. It’s not necessary, because plenty of people do good things without ever feeling the pain of the people they are helping. And it’s not useful, because it creates dangerous biases that can result in worse moral decisions. When we hear about the victim of a crime, and we empathize with their suffering, we may rush to support punitive criminal justice policies that have deleterious social consequences. Or when we see a starving child on a charity’s leaflet, we may rush to donate, even if this particular charity isn’t the most effective at helping children. The more we know about someone who is suffering, the more we are likely to want them to receive special treatment, even at the expense of those who are more deserving but whose names we do not know. Empathy, Bloom says, is selective: it causes us to sympathize with particular victims and make rash choices on the basis of our feelings about those victims, rather than stepping back and making a cool-headed decision as to what the most moral course of action actually is. Bloom says that a concern with the well-being of our fellow humans is best acted on through dispassionate cost-benefit analysis. All of this is essentially just a restatement of classic utilitarian ethics, and indeed, Bloom closes by praising Peter Singer and the Effective Altruism movement.

Bloom makes plenty of accurate points here, though again, they are a critique of “allowing your gut feelings to determine your morals” rather than “empathy.” It’s absolutely true, as Bloom points out, that people can be whipped up into a warlike frenzy by tales about individual victims, and that compelling narratives about one or two people can blind us to larger, less tangible harms. Bloom gives the example of climate change, saying that because its harms are diffuse, “empathy favors doing nothing.” But that’s false. Actually, if we truly appreciated what life will be like for the inhabitants of a boiling planet, if the power of empathy could finally let us see them as people just like ourselves rather than a mere abstract potential, we might finally appreciate just why the issue is so urgent.

In fact, Bloom’s advocacy of rationality over empathy misses something crucial: empathy offers data to aid rational decision-making. When I am trying to understand how other people feel, either just by using my imagination or literally trying to feel that thing, I am learning something about the world. If you want to make “cost-benefit” calculuses about the lives of other people, it helps to understand what their pains and pleasures are actually like. Sam Harris, in supporting Bloom’s argument, suggested that if we considered the pain of mothers who lose their children when we set laws designed to prevent traffic accidents, we would make everyone drive 15 miles an hour, because the pain is so great that any inconvenience would be justified in preventing it. But what Harris is advocating is excluding a certain kind of data (the feeling of a mother’s pain) from our cost-benefit calculus, because we don’t like the effects it has on the calculus. But that’s not rational; if understanding a mother’s pain made us feel that we needed to take greater precautions against allowing people’s sons to die, that would be because we had a better understanding of the world having more fully appreciated people’s subjective experiences. Bloom and Harris are right that emotion shouldn’t be a trump card that overrides rational thinking, and “a mother’s pain” is just one consideration to enter a cost-benefit calculus. But, fundamentally, empathy produces data on what it is like to be other people, and without it we will make worse decisions that simply impose our own conception of a person’s good upon them rather than understanding their good from their perspective.

People like Bloom and Harris are often oblivious to the pitfalls of cost-benefit analyses, which can give them the dangerous belief that their subjective ideas of the social good are somehow objective and mathematically correct. Bloom praises economists for their cool-headed refusal to put emotion above facts, and their willingness to support policies like free trade, even though these policies hurt some people, because they serve the greater social good. Bloom’s praise of economists is telling: it’s precisely the problem with economists that because they see things in the statistical aggregate, they fail to appreciate what the actual individual experiences of economic policies can be. They think it is “irrational” not to support free trade, without appreciating what it is actually like to live in a town whose industry been hollowed out by globalization and technological change. Their inability to empathize prevents them from realizing why many individuals have perfectly rational reasons for despising economists.

Pure cost-benefit analyses, untempered by empathy, can be just as dangerous as emotion-based policymaking. The classic example is the dropping of the atom bombs on Hiroshima and Nagasaki. From a cost-benefit perspective, the bombings were “rational”; they ended the war. (This is highly contested but let’s grant it for the purposes of argument.) But from a humane perspective, they were madness, an unspeakable form of cruelty. Now, it’s often argued that this shows precisely why you need to use “rational” rather than “emotional” decision-making: the rational person doesn’t allow feelings to keep them from making the tough but correct choice. But it’s also true that when one gets rid of feelings, one fails to appreciate just how horrific the ultimate consequences of the decision are. A person who sees the atom bomb in purely statistical terms doesn’t actually understand what it means for the people on whom it is being dropped. As a result, such a person won’t have the necessary moral horror needed to make them urgently want to ensure that no such bomb is ever dropped again. Empathy helps us understand what matters, even if our decisions should ultimately be grounded in reason. We can see in Bloom’s book what happens when one willfully abandons this: in one chapter, Bloom rejects the idea that “violence is wrong,” saying that violence will always be with us and is “right” in some circumstances. But those who truly understand the reality of violence will have a hard time ever saying that violence is “right.” It may be “necessary,” but it is always a necessary evil, and empathy helps us appreciate why violence is inherently evil even when it is morally compelled. The pure embrace of cost-benefit analysis creates moral blind spots. (We can also see this in Bloom’s praise of Peter Singer. Singer is a classic example of someone whose commitment to “rational” morality over empathy has actually made him a morally oblivious person; even though he strongly advocates helping alleviate suffering, he constantly makes appalling and insensitive remarks about disabled people, then wonders why they are upset, something that would cease to be mysterious to him if he learned to exercise a modicum of empathy.)

Examining moral matters carefully and rationally, and treating human beings equally rather than singling out particular emotionally touching cases, is manifestly a good thing. But nobody needs to be against empathy in order to be in favor of that. Paul Bloom could have written a very good book on why we needed to empathize fairly with all human beings, and examine our biases. It could have been a beautiful manifesto for loving and understanding each other. But instead, he simply concluded that it’s “impossible” to empathize with more than a few people (unless I am failing to empathize correctly, I have never found this to be the case), so we shouldn’t try to do it. And instead of exhorting us to understand other minds, he wrote a book that will help people justify telling others what’s good for them without having to actually listen to them.

I have dwelled too long on Bloom’s terrible piece of work. He is not an intellectually serious person; if he was, he would have dealt with the strong pro-empathy case (the one that says it’s a useful exercise to compassionately imagine other lives) rather than the ludicrous one (the one that says every time a child dies in the world I should burst into tears). But it’s worth pointing out just where Bloom goes wrong, because the book’s reception has been positive (Jesse Singal of New York magazine said Bloom’s “great” case is “tough to crack” and “absolutely succeeds”) and he is a tenured professor at Yale, an institution whose affiliates are granted a level of public credibility completely out of proportion to the actual social worth of their ideas.

Against Empathy does, however, offer a useful opportunity to reaffirm what empathy is and why it matters. Empathy is probably best thought of as something you do regularly, rather than all the time. It means making sure that you frequently think about the lives of those who are different from yourself, especially those who are very different. It’s an exercise to help connect you with your fellow human beings. Our imaginations are poor things with limited capacities, but listening to strangers and reading memoirs and stories can assist us. The aim is to get a real sense of the diversity of human perspectives, and to remind ourselves that other people experience consciousness just like we do. Empathy isn’t a moral philosophy in and of itself, it’s a technique for acquiring a better sense of the world as it really is.

The reason why empathy is so important is that it’s so easy to lapse into solipsism, to forget just how human our fellow humans are. (And, by the way, how human other animals are. One reason I think people are unmoved by the industrial killing of animals, even though we all know in the abstract that causes extraordinary amounts of harm to conscious creatures, is that we have failed to sufficiently empathize with cows and pigs. The moment we understand what it is like to be them, it becomes far harder to turn away from their pain.) It’s so easy to see others as statistics, or not even to see them at all. But the more I realize that others truly are just slightly different versions of myself, that they have dreams, itches, and fears just like I do, that they have eyeballs, teeth, and an anus just like I do, that they must fumble their way through the bewildering process called life just as I must, the more I begin to feel a powerful and moving sense of community, one that I believe is essential for creating a peaceful and mutually supportive world. Bloom quotes George Orwell, who found himself unable to shoot a fascist during the Spanish Civil War because the man was trying to hold his trousers up, and “a man who is holding up his trousers isn’t a ‘Fascist’, he is visibly a fellow-creature, similar to yourself, and you don’t feel like shooting at him.” For the cost-benefit analyst, that only demonstrates the case against empathy: a soldier who empathizes with the enemy won’t be able to shoot, and lives will be lost. And that may be right. But if we’re ever to actually eliminate war, to create a world based on mutual understanding, it’s vital for everyone to realize that the planet is filled with nothing but fellow creatures, that we’re alljust holding our trousers up. Not only is empathy a “good thing,” but until we learn to empathize, we will never truly know who we are.

Books like Against Empathy make me viscerally angry. First, there’s the contrarianism; I believe that bad things should be criticized, but Bloom is just trying to wind people up, and a true intellectual would frame their argument in the most helpfulmanner rather than the most attention-getting manner. Second, though, this just seems to me like the last thing we need right now. In a time where nobody understands each other, where there is so much isolation, when there are so few attempts to break down communications barriers and genuinely connect with other people, do we really need someone to come along and encourage us to feel other people’s feelings less? Surely Bloom can’t really believe what he’s saying, that humankind suffers from a surfeit of empathy. It’s already so difficult to make oneself understood, so difficult to get people to appreciate that those different from themselves are real and conscious, that I can’t imagine anyone choosing to write this book of all the possible ones they could write. And I can’t help but feel a seething hatred for Paul Bloom, more concerned with trolling people at parties than with seriously helping humankind.

But then again, I can empathize with him. After all, it can be tough to sell a book these days.