[Note: I am late to the discussion on Paul Bloom’s “Against Empathy,” which appeared in the Boston Review earlier this year. But as far as I can tell, though Bloom’s essay received plenty of attention, it has not gotten nearly enough of the outright ridicule that it deserves.]
Fuck Paul Bloom. Fuck this pointlessly contrarian, Slatepitchy, technocratic, inhuman BS, and fuck Yale professors who live perfect lives and have happy families and then say we should feel others’ pain less because it might hurt us too much.
Such was my initial reaction to Paul Bloom’s “Against Empathy.” But then I began to empathize with him. Hang on a minute, he’s a human being, just like me. He’s a smart guy, he’s trying hard, and he’s fallible. Let’s give him the benefit of the doubt and engage his arguments. Thanks to a dash of empathy, then, I was able to calm down and think through where exactly the train of his thought plummets off the tracks and into the canyon. I have about nine carefully-considered reasons for despising him.
First, Bloom is lying, or at least guilty of massive intellectual dishonesty. He opens the piece with an anecdote. He says he goes up to people at parties and tells them he is writing a book about empathy. Then he pauses and says “I’m against it.” He does this because he enjoys horrifying them (though, like a dick, he pretends to be surprised when they react badly). But Bloom is misleading these people. He is not against empathy. As he concedes in his reply to critics, “I am not against empathy in general, just empathy as a moral guide.” This is not the same thing as being against empathy. But this second statement, too, is a lie; Bloom is not even against empathy as a moral guide. What he is in fact against is a particular kind of empathy as a moral guide, what he calls “emotional empathy,” which he distinguishes from “cognitive empathy.” (Actually, he even narrows this further; by “moral guide,” he actually means “not by itself a good basis for public policy,” which is also completely different.) The whole framing of his argument as being “against empathy,” then, is purely so that Bloom can make his point seem more significant than it is. (A cynic might say it is to drum up early publicity for the book he mentions in his first sentence of his essay. But I am not and have never been a cynic.)
I think I may be being too fair to Bloom here, however, by assuming that he has a coherent position and is simply misleading people about it in order to be provocative. In fact, his use of the word “empathy” is so vague and contradictory that it is impossible to actually know what he is arguing. Bloom likes to pretend he has found a neat distinction between the “good” cognitive empathy and the “bad” emotional empathy. But I defy any reader to come out of the essay actually knowing what Bloom is referring to, how one can actually, practicably follow his proposal to retain “compassion” and ditch “empathy.” Yes, he offers a definition. He says emotional empathy is “feeling the pain of others,” and that this is the “common sense of the word.” But then his examples don’t follow the “common sense” of the word at all: he suggests empathy would mean feeling all of the sorrow of a friend whose child has drowned. Yet nobody thinks of empathy this way, as being some kind of one-to-one correlation between one’s feelings and the feelings of others where every time I learn that someone’s child has been killed I literally feel as if my own child has been killed. (He has the brazenness to say that those who defend empathy are “confused” by the term’s meaning, when he himself has asked us to use both the “common sense” and his totally unused, extreme sense.)
It’s a tribute to Bloom’s expertise in fudging and circumlocution that he says of his interlocutors that “some of them found my argument persuasive—one even called it obvious—while others responded with abject horror.” Such a diversity of reactions often occurs when one has written so unclearly as to leave open the question of whether one has even said anything at all. What Paul Bloom is doing is the old trick of provoking his audience with a bold statement, then spending the rest of the text hedging and qualifying so completely that when one reacts viscerally to the thesis statement, he can reply “Aha, you’ve totally misinterpreted me, how terribly uncareful a reader you are.” I call this “Žižek’s Gambit.”
So Bloom’s argument could be either so banal as to be worthless, or an absolute rejection of a fundamental human belief. We don’t know, and this makes it hard to actually engage with his position. But let’s assume that he is arguing something moderately significant, and imagine that some of the familiar recurring themes in his text hold together as a coherent assertion, namely something about how feeling others’ emotional pain does not alone constitute a sound basis for social policy. Does it hold up?
First, most of Bloom’s arguments do not prove what he thinks they prove. He says many times that empathy is biased towards those we feel close to. But this isn’t actually a point against empathy, because it could just as easily be used to argue for more empathy (toward those who we are biased against), or simply as a reminder that we should check our biases before making decisions based on empathy. One might conclude that the bias cannot be corrected for, but there would have to be some evidence for this, of which Bloom provides none.
In fact, for a psychologist, Bloom’s hard evidence is worryingly thin throughout. He repeatedly acknowledges the deficiencies in current research (“[m]any of these studies are poorly done,” “[t]o my knowledge, this has never been studied”), but then goes on to draw surprisingly solid conclusions about the negative effects of empathy. He cites studies to prove that empathy and morality are uncorrelated, but if those conducting the studies have as cloudy a notion of what empathy is as Bloom does, it is doubtful they even knew what they were testing. (The test in which people were given either “compassion training” or “empathy training” seems a particularly useless exercise; Bloom concludes that empathy causes burnout, but I too would expect this if empathy is taken to mean “attempting to suffer as much as possible.”) Much of his other evidence consists of stray experiences recounted by friends and relatives; someone he knew was grateful for a doctor they had who was very matter-of-fact, etc.
Bloom goes on extensively about how doctors who go to pieces in a crisis would not be much use. Of course, nobody whatsoever disagrees with this, but it does not prevent him from pressing the point. He is also insistent that “a doctor can deal perfectly well with a depressed person without herself feeling depressed, not even a little bit.” And it’s here that Paul Bloom reveals that he has never suffered from depression. First, nobody is suggesting that the doctor needs to be feeling depressed. But if Paul Bloom had talked to any depressed people, or had ever been depressed himself, he would know that a doctor who knew what it was like to be depressed makes a tremendous difference. My own experiences with psychiatric medicine have been uniformly made worse by doctors who had no idea what it is I was going through, who thought simple “compassion” was an adequate substitute for “being able to truly understand.” You don’t need to be depressed, but to have been depressed would be great. But because Bloom doesn’t believe in practicing empathy, he can’t understand why this is the case. (I would feel queasy here about arguing from personal experience, but if he can cite some old relative who once liked their tough-minded doctor, I don’t feel bad about going anecdotal.)
One of Bloom’s worst intellectual offenses is how completely he distorts the anticipated opinions of those he opposes. Throughout his piece, Bloom sets himself up as a tough-minded truth teller against a bunch of touchy-feely saps. He talks consistently as if the empathetic are weeping at all moments of their day. And he continually pretends that those who support empathy believe ridiculous things. “You don’t need to feel as if you are drowning to choose to rescue a drowning child,” he points out. No shit, but this is not what someone would reply with. Instead, they would say that one would be more motivated to stop something bad if one understood what it was like (e.g. John McCain’s torture experience informing his opposition to torture.) Of course, Bloom can disagree with even this position, but by arguing against the most absurd and caricatured version of the pro-empathy position, he reveals that he is uninterested in actually seeking truth.
The arguments, then are both weak and unfair. But I also believe they’re dangerous. Fortunately, I don’t have to strain to prove this point, because Sam Harris shows up in the replies to make my case by example. Harris’s reply goes as follows (this is a paraphrase, but I will vigorously dispute any suggestion that I am being unfair to Harris):
I agree completely with Paul Bloom. Empathy often gets in the way of sound moral decision-making. People forget that “dead babies are not an argument.” For instance, when Gaza is bombed, we think about the tears of Palestinian mothers whose children have been blown to smithereens. But this is a mistake. If we stopped allowing our feelings for their suffering to come in the way of our reasoning, we would see that Israel’s policies are fair and just. Too much empathy is a central problem in evaluating Israel’s actions. We know this, because 30,000 people die in traffic accidents every year. If empathy guided our traffic policy, we would all drive at 15 miles an hour.
One might think this a devastating satire of Bloom, showing the bloodthirsty consequences of abandoning our empathy. One might also hope Harris’s argument would ignite a qualm or two in Bloom, a sort of “Oh, God, what have I done” moment. But Bloom thanks Harris for his “thoughtful parallel,” and if there is a trace of the Swiftian in Harris, he has disguised it carefully beneath an entire career of uttering hateful filth.
First, as a minor point, we should note that Harris would only ever have made this argument about Palestinian children. Otherwise, why would he think the fact that Hamas is “genocidal” is important to consider, but Palestinian dead babies are not? If dead babies aren’t an argument, what’s the problem with genocide? The test of Harris’s sincerity is whether he will reverse the argument and apply it to Israelis or Americans.
Next, a question should be asked that applies equally to Harris and Bloom: What are you talking about when you refer to the harms of excess empathy? Where are they? The Palestinians were crushed entirely in the Israeli bombing campaign, not a lick of backlash seems to have halted Israel’s agenda in the region: how could it possibly be an example of policy being shaped by too much empathy? If Harris believes a central problem of the Israel-Palestine conflict is that Palestinian suffering is given excess weight, does he have a clue how many civilians died in the summer Gaza strikes? Some boon the empathy surplus was for Palestinians…
But perhaps the most scandalous aspect of the Harris position, beyond the hypocrisy, is that Harris and Bloom are wrong even on the rational consequentialist framework they both adopt. They both insist that we should make careful, emotion-free, data-driven decisions. But they don’t understand that empathy produces data. If you’re trying to make a rational decision based on the evidence, the feelings of others are critical evidence. Why would they not be? What kind of selective consequentialism thinks they should not enter into the cost-benefit calculus?
Harris assumes that if we factored what it’s like to lose a child in a traffic accident into decisions over speed-limits, we would all drive at 15 miles an hour. But how is that the case? Harris presumes that a speed limit change would inevitably be the consequence of empathetic decision-making. But this is not so. We might be empathetic but still believe there is a good to be had in having efficient transit. But unless we understand what a mother’s pain is like when her child is lost, we’re not going to be very good at assigning a value to it. We might lower the speed limit if we had this information, but if we did, it would be because we had a better understanding of the stakes.
Harris and Bloom are advocating selectively excluding certain kinds of data from a cost-benefit calculus, because they might lead to decisions that Harris and Bloom don’t like. If we value Palestinian children too highly, because of what their mothers’ pain is like, we might be excessively critical of Israeli policy. They are working backward from their favored outcomes, admitting only the data that support them. And they have the gall to suggest that this profoundly irrational move prioritizes rationality!
So of course dead babies are an argument. They are not necessarily the winning argument. They aren’t a trump card that destroys all other considerations. But they’re an argument, and a powerful one. And that’s where the whole “compassion instead of empathy” public policy framework falls apart completely. If you don’t know (or care) what it feels like to be another person, how on earth are you going to make policy calls that involve placing a value on that feeling? What is compassion if you’re deliberately blinding yourself to the actual internal experience of people, the very thing that constitutes what you’re supposed to be compassionate about? Can we coolly measure “units of suffering” without having any idea what they look like? (Bloom actually replies to a version of this argument, which he fairly characterizes as trying to describe redness to a blind person. But all he says is “I am not convinced.”)
Unfortunately, the Boston Review symposium commenters do a miserable job of exploding Bloom’s position. They range from quibbles about the cited studies to pompous theological irrelevancies, and nearly all of them treat Bloom’s article as interesting and reasonable. Of the twelve, Simon Baron-Cohen does the best job of slicing up Bloom, pointing out that the thumpingly obvious truth that reason and feeling are not mutually exclusive. Unfortunately, Baron-Cohen falls on his face about two paragraphs into his reply, by suggesting that Bloomian logic would justify a Holocaust. By forgetting that Bloom has played Žižek’s Gambit, Baron-Cohen allows Bloom to return in his reply and say “No, no, as you saw, I specifically insisted that my position includes ‘compassion,’ which is anathema to Nazism. What an uncareful reader you are…” Because Baron-Cohen did not succesfully anticipate this reply, he ends up losing the argument.
But Bloom’s position does lead to the Holocaust, or at least opens up the path to it. It does this because compassion without empathy is devoid of content. If one is not thinking about how the other feels, it is very simple to think one is being compassionate when one is in fact being monstrous. For example, one might consider it compassionate to require abortions for mothers pregnant with mentally disabled fetuses; it spares mothers the tremendous cost and suffering of raising a deeply ill child, and spares the child a life of pain. But half a second of empathy, actually imagining what it would be like to be such a mother, would reveal the monstrosity of this ostensibly-merciful program. “Compassion” for someone is meaningless unless one knows what one is being compassionate about, and delusions of logic can be the most illogical force of all.
What made me most upset with Bloom, though, is his overarching assumption that somehow there is too much empathy in the world. Out of the fact that in some studies, suffering others’ pain has been shown to cause one a hard time, Bloom concludes a society we have gone mad with empathy-fever. But where is this huge, devastating surfeit of empathy? Bloom never hints at its location.
Bloom compares empathy to anger. But what is empathy’s body count? How can an emotion with proven destructive power be compared with a benign-to-positive one? Bloom has an answer to this: nations can be roused to war through atrocity tales, which appeal to empathy. Certainly this is true, but does he have any reason to think that his “compassion” framework will not be equally biased? We can easily recall the Harris example, where one’s already-favored victims are just as easily given priority under a “cool, rationalistic” framework as under an empathetic, passionate one.