Skip to content

Steven Pinker on torture

I’ve recently been thinking about that expression, “A liberal is a conservative who’s been arrested.”

Linguist and public intellectual Steven Pinker got into some trouble recently when it turned out that he’d been offering expert advice to the legal team of now-disgraced financier Jeffrey Epstein.

I would not condemn Pinker for this. After all, everybody deserves a fair trial. Also I agree with the statement here that it’s not fair to tar Pinker with guilt by association just because he met Epstein a few times and had friends who were friends with Epstein. We’re all of us two or three links from some pretty bad stuff. I have friends who’ve done things they shouldn’t have, and I’d be surprised if I didn’t have friends of friends who’ve committed some horrible crimes. Social networks are large. (We’ve estimated that the average American knows 750 people. If each of them knows roughly that many people, then, well, do the math.)

Pinker has come up before on this blog, notably in this 2007 post regarding a newspaper article he wrote, “In defense of dangerous ideas: In every age, taboo questions raise our blood pressure and threaten moral panic. But we cannot be afraid to answer them.” That article contained a long list of ideas labeled as dangerous. At the time, I questioned how dangerous some of those ideas really were, and Pinker replied to my questions. You can go read all that and make your own call on how the discussion holds up, twelve years later.

But the “In defense of dangerous ideas” idea of Pinker that I want to focus on here is this one:

Would damage from terrorism be reduced if the police could torture suspects in special circumstances?

Recall Pinker’s statement:

In every age, taboo questions raise our blood pressure and threaten moral panic. But we cannot be afraid to answer them.

One thing we do know, though, is that the U.S. government supported torture of terrorist suspects back in 2007 when that article came out, and torture is supported by the current U.S. president and also Pinker’s friend at Harvard Law School. So I guess it’s not such a taboo question after all.

I googled *steven pinker torture* to see if he’d written anything on the topic since then, and I found this from 2012 or so:

Question: You say that cruel punishments and slavery have been abolished. But torture was practiced by the United States during the Bush administration, and human trafficking still takes place in many countries.

Response by Pinker: There is an enormous difference between a clandestine, illegal, and universally decried practice in a few parts of the world and an open, institutionalized, and universally approved practice everywhere in the world. Human trafficking, as terrible as it is, cannot be compared to the African slave trade (see pp. 157–188), nor can the recent harsh interrogation of terrorist suspects to extract information, as indefensible as it was, be compared to millennia of sadistic torture all over the world for punishment and entertainment (see pp. 130-132 and 144–149). In understanding the history of violence, one has to make distinctions among levels of horror.

This seems to contradict Pinker’s “In defense of dangerous ideas” position. In 2007 he was defending the idea that torture would save lives—not that he necessarily agreed with the point, but he thought it worth discussing. But a few years later he was arguing that modern torture is “indefensible” and not such a big deal because it is “clandestine, illegal, and universally decried” and only occurring “in a few parts of the world” (unfortunately, the U.S. is one of those few parts of the world, also I don’t think it’s at all accurate to describe it as universally decried, but no need to get into that here as the point is pretty obvious and so I guess Pinker was just doing some rhetorical overreach when he said that).

So what’s the take on torture? Is it indefensible and universally decried, or is it an idea worth discussing, supported by a large chunk of our national political leadership?

One argument that’s sometimes been made in favor of torturing terrorist suspects is that terrorism is a new danger, unique in modern times. But I don’t think Pinker would make that argument, as he’s on record as saying that we live in “an unusually peaceful time” and that terrorism is a hazard that “most years kills fewer people than bee stings and lightning strikes.”

It kinda makes you wonder if he’d support police torture of beekeepers, or manufacturers of faulty lightning rods. Only in special circumstances, I’m sure.

Just to be clear: I agree with Pinker that less torture is better than more torture, and I’m not equating current U.S. military practices with the Spanish inquisition.

OK, so what’s the connection to Jeffrey Epstein?

Two things.

First, torture may well be “indefensible” in any case, but I think we can all agree that it’s particularly horrible if, when instituted as part of an anti-terror program, it ends up administered to someone who isn’t actually a terrorist. From news reports, it seems that happens sometimes. In Pinker’s 2007 “In defense of dangerous ideas” article, there’s the suggestion that torture of terrorism suspects could be ok—at least, worth discussing—without so much concern that terrorist suspects might be guilty of nothing more than association. The problems with guilt-by-association could well be clearer now to Pinker.

The second connection is that Pinker’s “In defense of dangerous ideas” article also appeared as the preface to a book called “What’s Your Dangerous Idea?”, the 2007 edition of a set of collections of short essays published by the Edge Foundation. This 2007 volume did not include any contributions from Jeffrey Epstein, as he was consumed with legal troubles at the time, but the now-famous financier returned the next year as one of “the intellectual elite, the brains the rest of us rely on to make sense of the universe and answer the big questions” to contribute to the volume, “What have you changed your mind about? Why?”

My point in making this connection is not to tar Pinker with guilt by association (or, for that matter, to imply that Epstein ever supported the idea of torture). My point is that your perspective on legal actions will appear different, depending on which end of the telescope you’re looking into. If you’re talking about an unnamed terrorist suspect, maybe you think that torture is a bold and dangerous taboo idea worth defending; if the person accused is a friend of a friend, they get legal advice.

I’m sure that most of us, myself included, behave this way to some extent. When it comes to intellectual debate, I can be as hard on my friends, and on myself (or here), as I am on acquaintances or people I don’t know at all. But when it comes to people actually getting hurt, then, sure, it’s a lot easier to summon empathy for people who are close to me, and it’s a lot easier to apply some principle of loyalty to friends of friends. Clashes between loyalty and other principles: that can be the stuff of tragedy.

So, again, the point of the this post is not to trap Pinker in some sort of Gotcha. It’s fine with me that his position on terrorism and torture changed between 2007 and 2012. In 2007, the World Trade Center attacks were still fresh in everyone’s mind, but by 2012, Pinker had been spending years reflecting on the global decline of violence so he was less willing to consider terrorism as a police tactic worth defending. Fair enough. If you’re an Ivy League pundit and you want to support police torture as an idea worth defending, you have to think that the benefits are greater than the damage—while taking account of the fact that lots of the damage would be happening to people you have no connection to, people who might not be clients of your law-professor friends.

And then there’s that line, “A liberal is a conservative who’s been arrested.” Due process and not tarring people based on guilt by association: that’s a principle that applies to associates of sex traffickers and also associates of terrorists.

Studying incoherence

The above might come off as anti-Pinker, but that’s not where I’m going here. So let me explain.

Coherence of beliefs and attitudes is an ideal or norm that should not be achievable in practice. We shouldn’t expect complete coherence among someone’s beliefs, for the same reason we shouldn’t expect complete coherence in the actions of a complex organization such as a corporation or a government.

To put it another way: Consider a house that’s been lived in for a few decades, long enough that it’s had various weather events, changes in occupancy, and other things that have required repainting of various rooms at various times. The colors of the different rooms won’t be quite in synch. Similarly with an economy: different products are coming at different times, prices keep changing, there never will be complete coherence. The point of noting an incoherence in someone’s views is not a Snap! You’re a hypocrite!, but rather: That’s interesting; let’s juxtapose these views and see if we can understand what’s going on.

We learn through anomalies. That’s what posterior predictive checking is all about, and that’s what stories are all about. The anomalies of Pinker—a prominent libertarian who’s open to the idea of police torture, a supporter of due process for sex trafficking suspects but not for terrorism suspects, etc.—are interesting in giving a sense of the contradictions in certain contrarian political perspectives.

P.S. Speaking of the Spanish inquisition, my googling turned up this Pinker quote from 2019:

Changing sensibilities: In 1988, I enjoyed A Fish Called Wanda. 30 years later, I found it cringeworthy: We’re supposed to laugh at stuttering, torture, cruelty to animals, the anguish of those who care, & a women using her sexuality strategically. Not un-PC, just un-funny.

I’m surprised that the torture bothered him so much. What ever happened to defending dangerous ideas??? Maybe the problem is that the torture in the movie wasn’t being done by the police, and it wasn’t being done on a suspected terrorist.

I thought A Fish Called Wanda was hilarious in 1988 and I happened to see it again last year, and I found it hilarious the second time as well. Not cringeworthy at all. My favorite scene was when Kevin Kline flipped the gun to himself as he was walking through the security gate at the airport. And that was just one of so many memorable scenes. I’d call it a modern classic, except that I guess 1988 isn’t really “modern” anymore.

I wonder what Pinker thinks about The Bad News Bears? I loved that one too!

P.P.S. I also came across this thoughtful and approximately 60% positive review from sociologist Claude Fischer of Pinker’s book, “The Better Angels of Our Nature.” I highly recommend Fischer’s review. Indeed, I think you’re better off reading Fischer’s review than my post above.


  1. Saying that a thing is worth discussing, and shouldn’t be dismissed out of hand is not the same as saying it’s a good idea. In particular one reason to discuss and defend something is so that it can be seen as indefensible.

  2. Randy says:

    The quote should be “A liberal is a conservative who’s been arrested” – yes? Not “A liberal who’s a conservative who’s been arrested.”

  3. Anonymous says:

    “Would damage from terrorism be reduced if the police could torture suspects in special circumstances?”


      • Phil says:

        For sure there are possible worlds in which the answer is ‘yes’ and others in which the answer is ‘no’, and I don’t know how we’d know which world we are in. Who defines the ‘special circumstances’, and who decides whether the actual circumstances meet the definition?

        There are certainly situations in which I would want someone to be tortured to reduce the damage from terrorism, and some in which I’d wield the pliers myself if I had to (imagine the too-often-filmed situation of an atomic bomb that will go off in a major city if it can’t be quickly located).

        But I am not aware of any real situation in which the police could have reduced the damage from terrorism if only they had applied some torture, and even if such a situation existed and torture had been applied, it’s possible that if there were a policy that allowed such actions that policy would be a recruiting tool to recruit more terrorists, thus leading to more rather than less terrorism.

        My feeling on torture to reduce terrorism is: it should be illegal, and if a police officer (or anyone else) does it they should be prosecuted for it. If the torture was really justified, to the extent that ‘society’ is willing to endorse it, then the jury and the judge will find a way to either prevent punishment or make it minimal. But I don’t want the police or anyone else getting comfortable with the idea that that can be counted on. The person doing the torturing has to be willing to pay a huge penalty. I like to think that if there really truly were an atomic bomb set to go off in a major city, and I could stop it by torturing someone but I would spend the rest of my life in prison, I would go ahead and do what I had to do. I’d rather spend the rest of my life in prison than allow hundreds of thousands of innocent people to be killed, if that’s the choice.

        In short (too late!) I think the answer to “Would damage from terrorism be reduced if the police could torture suspects in special circumstances?” is “Yes”, but I don’t think that means we should define those circumstances and make torture part of accepted policy even if those circumstances occur.

        At any rate I have no problem at all with the idea that it’s OK to discuss this concept. I don’t actually think it’s a worthwhile discussion but I don’t think it’s immoral to even consider the question.

        • Anonymous says:

          It’s likely I have far more direct experience with this topic than anyone else commenting here. Really, the answer’s “no”.

          -I don’t trust an academic like Pinker, or any of these other armchair warriors, to correctly estimate the value of things “willingness of the enemy to surrender to US troops”. The experiences in the last months of the WWII in Europe between Germans fighting the Russians, and Germans fighting the Anglo-Saxons is illustrative, but there are innumerable more recent examples which illustrate the effect well.

          -Then there’s the matter of “would I want to serve with people who enjoy torturing enough to do it as a job?” No, no I wouldn’t. I really wouldn’t.

          -If you were that eager to get false confessions, you don’t need torture. The police have found it’s not that difficult to coerce false confessions without it.

          -I think you all are underestimating how good regular old interrogations (combined with other sources of intel) actually is.

    • anonima says:

      “Would damage from terrorism be reduced if the police could torture suspects in special circumstances?”


      It´s called “State terrorism” or “human rights violation”

  4. Dogen says:

    I, too, loved A Fish Called Wanda when it came out. And I saw it again recently! I didn’t find it cringeworthy—worse! I found it boring.

    And my biggest criticism of Pinker is that he’s super boring. The difference being that I’ve always found him super soporific—in that regard he’s incredibly coherent! He’s such an “oh, so sincere” blowhard droning on and on in the intellectual shallows. A perfect illustration of TED-talk profundity. (I have no idea if his academic research is any good, I assume it must be, but his public persona is truly cringeworthy.)

    I think a great strength of the movie in the 1980s was in having a major female character with her own, unapologetic (mostly), agency. But she was still a caricature (as were the males).

    I think the portrayal of women in mainstream movies has enormously improved, even though there’s still a long way to go. So a lot of what made the movie fresh and interesting to me in 198x is now seriously dated. There are still funny moments, its a great cast after all, but they weren’t enough for me in 2019.

  5. RAD says:

    I think your argument turns on how you interpret Pinker’s use of the word “indefensible”. If it is synonymous with the word “taboo” and he intended this meaning then I think you are correct. A more generous interpretation, given the context of the quote, is that indefensible is shorthand for “a losing argument”.

  6. paul alper says:

    Torture advocates in the U.S. use linguistic acrobatics. Torture is reasonable response to conflict and at the same time it is only enhanced interrogation, a euphemistic extension of commonly accepted practice. Here is a waterboarding exercise:

    Here is another enhanced interrogation exercise:


  7. Jonathan (another one) says:

    Not completely off-topic, but I highly recommend The Report, a not-boring movie about investigating the roots of post-9/11 torture. There are plenty of villains in the piece, but prime among them are Jim Mitchell and Bruce Jessen, who style themselves as social scientists.

  8. Koray says:

    I think you’re misinterpreting Pinker’s response to the question “You say that cruel punishments and slavery have been abolished. But torture was practiced by the United States…”. All he’s saying is while what the US did torture, it was at such a small scale compared to the “millennia of widespread sadism” that we can _say_ that “cruel punishments have been (virtually) abolished”.

    With regards to his openness to discussion of torture under exceptional circumstances, it’s not a new philosophy puzzle. The current (expert) consensus is that torture doesn’t actually work, so we don’t have to answer it right now. But, if something like a hypothetical truth drug existed, which made its recipient answer every question and answer them truthfully, would we administer this drug against their consent? This question should not be dismissed because regardless of our current answer, future generations will continue to re-raise it. We should at least let history know why we answered it our way.

    • Andrew says:


      To me the interesting question here is not about torture but about perspective. When it comes to himself, Dershowitz, and Epstein, Pinker feels strongly opposed to the concept of guilt by association. But when police or the military are torturing terrorism suspects . . . lots of this is guilt by association, right? The army picks people up from a war zone and then tortures them into providing information. I’m thinking that Pinker may be less supportive of this principle now that he can draw the analogy to the accusations being leveled at him. All Pinker did was provide Epstein with some legal advice and now people keep associating him with Epstein. Nobody’s being tortured here, but there’s a general issue of what’s ok to do if someone has an association with a serious criminal.

      • Koray says:

        You wrote: “Nobody’s being tortured here, but there’s a general issue of what’s ok to do if someone has an association with a serious criminal.”

        I don’t see how guilt by association applies to torture. The accomplices or family members of terrorists are not tortured as punishment for a past crime; they’re tortured to extract information that can prevent a future crime. This is the (hypothetical) payoff for it. It is also not meant to discourage families of other terrorists from staying silent on overheard bombing plans in the future.

        In case of Pinker & Epstein, some people do want to punish Pinker for Epstein’s crimes in the past, and they also want to discourage everyone from offering legal advice to or socializing with people that may end up being criminals. This is indeed guilt (of past crimes) by association and also a deterrent.

        • Andrew says:


          The connection to guilt by association is that often the people who are being tortured are not accomplices of terrorists or even accomplices of suspected terrorists but just associates of terrorists, if that.

          We have no reason to think that Pinker was an accomplice of Epstein, but he was certainly associated with Epstein. Pinker was angry that people were saying bad things about him—treating him like an accomplice, when he was just an associate. Given this experience he’s had, I suspect that he might now have a better understanding of why people oppose the torture of associates of suspected terrorists.

          • Koray says:

            Yes, I already acknowledged that who’s being tortured is sometimes merely an associate (e.g. “family member”), who may or may not overheard something about a future plot. The torturers may only be hoping that these associates may know something. Note that if the torturers know for a fact that a certain associate cannot possibly know something, they will not torture that associate. So, this is not about guilt by association at all.

            A hypothetical example: terrorist groups Red and Green both hate the USA, but they also hate each other. Group Red is spying on Group Green and may have accidentally discovered their plans against the USA. The CIA knows that Group Red spies on Group Green, and the CIA was able to capture some members of Group Red. The CIA doesn’t know whether these Red members know anything about the plot, nor do they know whether the Red members would tell anything if they knew because they also hate the USA and may want to see the plot succeed.

            The CIA is not accusing the Red members of “guilt by association” because they definitely had nothing to do with the plot and they hate the Greens. Torture is entirely about obtaining information from a potentially “knowledgeable but unwilling” party. These parties may be accomplices, mere associates or even enemies of the actual perpetrators, and you don’t even have to be certain that they know or certain that they wouldn’t tell anything if they knew.

            • Andrew says:


              You write, “Torture is entirely about obtaining information.” I don’t think this is historically the case. But, even if torture is about obtaining information, I don’t think it would be right to torture Steven Pinker, just on the off chance that he might have information about Dershowitz or other people he is associated with who have associated with sex traffickers and other criminals. I’m guessing that Pinker feels the same way, and he might even have some more general feelings about torturing people whose only offense is to be associated with criminals. You and Pinker can debate this one between yourselves!

            • Curious says:


              It is absolutely about guilt by association according to your own description of the problem:

              “…if the torturers know for a fact that a certain associate cannot possibly know something, they will not torture that associate”.

              So, the associate is only of interest due to their connection to a particular person/group. Unless it can be known with certainty that they do not know something — a logical impossibility — they are assumed guilty of intentionally withholding information. It is then decided that any individual rights for due process and civil liberties can justifiably be ignored because of a *possibility* that some lives will be saved.

  9. Mikhail Shubin says:

    I visited the post from 2007, and oh

    “Were the events in the Bible fictitious — not just the miracles, but those involving kings and empires?”

    Is this such a radical thought that everybody afraid to discuss it? Seems to me like a regular question from Religious Studies 101 textbook. Bible was treated by researchers like a regular ancient literature for like ages.

    This made me think, is “indefensibility” of an idea depends on the context? Take two ideas: “Torture is fine sometimes” and “Hitler was right”. Both are indefensible. But I can see the first idea being discussed in some debating society or a philosopher publishing a paper on morality of torture in Political Philosophy Letters. But not the second.

  10. oh blessed 2007, one was able to write a post mentioning feminism and get less than 1000 comments.

  11. paul alper says:

    Keith Olbermann in 2009 issued a waterboarding challenge to Sean Hannity with the money going to charity. A different right-wing radio host, Erich “Mancow” Muller who likewise believed waterboarding is not torture, took the challenge:

    “Mancow admitted that it was “absolutely torture” and was “way worse” than he expected.”

    “Yesterday, MSNBC’s Keith Olbermann interviewed Mancow about his experience under the bucket. “I would have said anything to make it stop,” Mancow said, further confirming that torture does not produce reliable intelligence. “I don’t think drowning is harsh enough. … This is worse. This isn’t gulping for air. This is your brain is shut off.” Mancow said that despite the “horrific” event, Hannity called him afterwards to insist that waterboarding still isn’t torture.”

    “Mancow, who initially scoffed at the tactic, explained to Olbermann: “Look, I see the video…the sprinkling of the water, big deal. … I was laughing at it. I was willing to prove and ready to prove that this was a joke. And I was wrong.”

  12. jim says:

    I don’t recall seeing A Fish Called Wanda. For me Movies don’t age well and watching any movie that’s more than a decade old is torture, and I would say almost anything to get out of it – so torture would in fact work because I’d be likely to say something, but it wouldn’t in fact work because there’s no telling whether what I said was reliable. But it would in fact work because, seeing how tortured I am by old movies, my torturer would know what I wouldn’t risk providing bad information; but it wouldn’t in fact work because I would know that my torturer would trust me, and that I could probably slide one by the dude.

    The only exception is The Princess Bride. “My Name is Inigo Montoya. You killed my father prepare to die!”

  13. Martha (Smith) says:

    Andrew said,
    “P.P.S. I also came across this thoughtful and approximately 60% positive review from sociologist Claude Fischer of Pinker’s book, “The Better Angels of Our Nature.” I highly recommend Fischer’s review. Indeed, I think you’re better off reading Fischer’s review than my post above.”

    Thanks for the link. Indeed very interesting.

  14. billo says:

    It seems to me that what you are really asking is how much of a utilitarian a person is. From a utilitarian perspective if one is given two bad choices, then the one that does the least harm is the “good” choice and is morally justifiable. Thus, if torturing one person will save a thousand lives, torture is the “good” choice. Oh sure, it’s regrettable and unfortunate, but basically “justifiable.” There are, essentially, no “Sophie’s choices.”

    From a deontologic perspective, the circumstances do not change the moral meaning of the act. Torturing one person may save a thousand lives, but it is still an evil act, and as such requires condemnation. However, declining to save a thousand lives is *also* evil, and is *also* worthy of condemnation. There is no “good” solution. Torture is torture. Obfuscating what it is by splitting hairs about definitions and using euphemisms like “enhanced interrogation” doesn’t change what it is. Torture is evil, regardless of how or why you do it. Anybody who tortures someone takes the moral hit.

    A classic example of this in the Old Testament/Tanakh is the story of Uzzah, who broke God’s law by touching the Ark of the Covenant. While transporting the Ark, the ox pulling the cart stumbled and Uzzah reached up to keep the Ark from falling. God struck him dead. Some argue that Uzzah reaching up showed a lack of faith — that God would not let the Ark fall, and Uzzah’s intervention was not necessary and showed that Uzzah didn’t believe in the power of God. Others argue that the motive of Uzzah doesn’t matter. It was wrong for Uzzah to touch the Ark *and* is would have been wrong to allow it to fall. That is, in fact, one interpretation of so-called “original sin.” The problem of humanity is not so much that we are inherently horrible as much as the fact that we live in a universe that constantly presents us with lose-lose moral choices, and thus there is no possible way to live a life without making a poor choice.

    Whether or not the consequences of an evil act mitigate legal culpability is tangential to the moral culpability.

    The problem is often, it seems to me, that a lot of people want to claim deontologic morality but have utilitarian/consequentialist ethics. Eventually the disconnect between the two cannot be plastered over.

    • jim says:

      I guess another utilitarian framing of the torture question is: does it provide reliable information? If not then the moral dilemma is irrelevant.

      If the victim of the torture has no information, then s/he is forced to lie. Since the torturer has no reliable information s/he has no way of detecting a lie. The information from this lie might be worse than useless, it might be dangerous – it might cost more lives than having no information. Then again, the victim might not be lying and the information might save many lives. But then again again, there’s no way to know.

      This I believe is why most experts think torture is at least useless, whether or not it’s morally wrong.

      • Martha (Smith) says:

        Good point.

      • billo says:

        The problem with the “torture provides no useful information” is that it is a very broad generalization, both about the application of torture and the purpose of it. Most torture is not for the purpose of extracting information in the first place. When used by the State on its citizens, it is most often used to establish and define the relationship between the State and the citizenry. This was particularly true in places like Argentina, Nazi Germany, modern China and North Korea, and many places in the Middle East today. Consider the Uyghurs and Tibetans in China. For an excellent description of this with respect to the the “Dirty War” in Argentina, look at “Torture and Eucharist” by William Cavanaugh. For a good discussion of one type of application — more psychological than physical — in China, take a look at Thought Reform and the Psychology of Totalism: A Study of “Brainwashing” in China In the Middle East, by Robert Lufkin. In the Middle East, torture is often used not to extract information but often for propaganda purposes, much like the ISIS execution videos. In the 1990s, the Chechens made a cottage industry of making videos of torture and execution of Russian prisoners. I’m a forensic pathologist and computer scientist. When I was in the military, my job was analysis of this kind of imagery for operational purposes.

        My experience with folk who claim to have first hand knowledge of interrogations under torture (I don’t have any) is mixed. I have heard some folk say it’s not particularly useful, and others who claim that it has been useful in very specific circumstances where knowledge about the victim and surrounding the information being extracted is firm and the information can be corroborated. The impression I get is that there is a nontrivial space between “low yield” and “no yield,” particularly with high value information and strict time constraints. But I’m an expert on the physical manifestations of torture in the victim, not on its usefulness.

        That doesn’t make torture any less evil. But to those I have known who are more operationally familiar with this kind of thing the claim that it is “never” useful isn’t held with the dogmatism I see in a lot of discussions in forums like this.

  15. A.G.McDowell says:

    As evidence for the (lack of) efficacy of torture I note the end of the web page at – written by the excellent SF Author David Drake, who was a prisoner interrogator in Vietnam.

    I also note that the possibility of torture is an excellent motivator for the target organization to practice “Need to know” and other security measures. If you have any other sources of intelligence, such as hacking or penetration (e.g. the last thing you want is for the target organization to be good at enforcing “Need to know”.

  16. Roger says:

    There are situations where torture does not work, but there are also situations where it does. Some examples were on the TV show 24 a few years ago. Say someone planted a bomb that is timed to explode in 30 minutes. The bomber is caught, and tortured to reveal how to deactivate the bomb.

    Opinions differ on what is torture. There are some accepted interrogation methods that you might consider torture.

    “the U.S. government supported torture of terrorist suspects” — This is overstated. I think that 2 or 3 suspects were waterboarded, and the CIA called it enhanced interrogation, not torture.

    • jim says:

      How do you know you caught the real bomber? How do you know the bomb is real? How do you know the person you caught knows how to deactivate the bomb? How do you know they wont lead you into a worse situation? How will torture help you know any of these things?

    • Michael Schwartz says:

      Well, if it was on a fictional TV show, then it must be true.

      I mean, there must be thousands of documented, real cases just like this (“ticking time-bomb” scenario) where torture was shown to work.

      Thanks, Roger, for clearing up this complicated topic with straightforward example from TV.

  17. Conor says:

    Not your point, but it’s an area I’ve spent a lot of time researching: in what way is the slave trade not comparable to human trafficking, Pinker? There’s a sometimes healthy debate around particulars that I doubt will ever be resolved, but the definition of trafficking as defined by the Palermo Protocols adopted by the UN is:

    “‘Trafficking in persons’ shall mean the recruitment, transportation, transfer, harbouring or receipt of persons, by means of the threat or use of force or other forms of coercion, of abduction, of fraud, of deception, of the abuse of power or of a position of vulnerability or of the giving or receiving of payments or benefits to achieve the consent of a person having control over another person, for the purpose of exploitation.”

    Seems like it ticks all of the boxes.

    • steven t johnson says:

      You can compare ravens to writing desks.

      Compare the estimated death tolls for the Middle Passage to the death tolls for human trafficking, if you must. But I think human trafficking is terrible even though the death toll is lower. Also, the children of trafficked humans are not property in any sense I understand. I think Pinker is shady but this isn’t one of his offenses to say a Florida massage parlor isn’t the Middle Passage.

      If you insist on the formalities, the provision in the Palermo Accords relating to abduction, fraud and deception, abuse of power or position of vulnerability don’t seem to technically apply to the Atlantic slave trade when it was legal, which was most of the time. The actual kidnapping of people for the slave trade seem to have been regarded as acts of war by other Africans, like the kingdom of Dahomey. Ending the slave trade was part of the colonial project. (It would be foolhardy to conclude colonization was therefore good, or even “good.”)

      The final word on torture as useful tool for interroation is the way it confirmed the existence of witches. (The estimated death toll in Europe from witch hunting is in the hundreds of thousands if I remember correctly.) It is alleged that the use of torture in Algeria enable the French to suppress the first phase of revolt. They would torture one or more prisoners for names and other information, then they would torture other prisoners separately to see if they got the same information. But of course, this was not torture justified by an exigent energency It was standard operating procedure, business as usual. Yet, the Algerians still won, didn’t they?

      The real purpose of torture is to terrorize people. Sometimes that works, sometimes it doesn’t.

  18. Dan F. says:

    “and I’m not equating current U.S. military practices with the Spanish inquisition.”

    I suspect this reflects an inaccurate picture of what the Spanish inquisition entailed.

    At the time of the inquisition, torture was standard in trials in Europe, and the inquisition stands out positively for its procedural guarantees with respect to torture and that it used torture less than was standard.

    Contemporary US practices stand out negatively for their evasion of procedural guarantees and that they use torture and other similar tactics more than is standard among comparably developed nations.

  19. steven t johnson says:

    This is in many ways a confusing post. Perhaps the most important takeaway is the endorsement of the review of Pinker’s Better Angels. Fischer managed to cite the An Lushan rebellion as an example of how world wars are not so bad in really long-term perspective. The problem of course is that the estimates for casualties seem to rely upon tax records. This naturally raises the likelihood that as many people as possible didn’t register for taxes when they fled somewhere else and got away with it because of, An Lushan rebellion. The assumption every last one of them died is highly doubtful. When there is great inaccuracy in a limited set of data points, it can play hob with calculations of trends. Pinker’s many graphs are the real carriers of the burden of argument. There might be issues in the graphical representation of data.

    Further, Fischer was rather cautious in talking about the presentation of the world wars as random glitches. If (as I gather Nassim Taleb argued) the world wars were the necessarily infrequent breakdowns of global order, due to improbable concatenations of events, it is not at all clear what “trend” even means. Given enough time, complex systems will break down (barring preventive repair, but can that be assumed? The trend is for world wars to get worse in times of violence. WWI, young men in armies died, decimating cohorts. In WWII, the generals conserved their soldiers but attacked the civilian populations. The trend seems to suggest WWII will be marked by outright extermination of whole populations, judging from the attempts to starve whole population. (This is not even regarded as a bad thing!)

    In short, I’m not quite sure anyone is safe in assuming Pinker is both objective and reliable outside his actual field of study, which I think is linguistics.

    In regard to the liberal mugged/conservative arrested dichotomy, they are equally irrelevant to the class issue of who is apt to be arrested and whose mugging is a serious issue.

    And on the meanness of comedy? There is a wrong cliche about how drama is conflict. Instead, comedy is conflict. The point about farce is that the conflict is treated as though it does not have real consequences. The pain of caricatures is a joke. Whether this means you disapprove of farce as immoral is up to you. Whether you wish to prescribe whose pain in farces is deemed acceptably funny requires you have the authority to prescribe, I should think.

  20. Michael L. says:

    You quote Pinker here:

    ‘But a few years later he was arguing that modern torture is “indefensible” and not such a big deal because it is “clandestine, illegal, and universally decried” and only occurring “in a few parts of the world”’

    But as I read his “clandestine, illegal, and universally decried…” point, I thought he was claiming that human trafficking is not the same as the African slave trade. I didn’t think he was saying anything about torture. He didn’t do that until he said, “nor can the recent harsh interrogation….”

  21. David Baer says:

    People who are against torture which is used in order to save the life or serious injury of another put their elitist views ahead of human life.

    They lack skin in the game, more than willing to sacrifice human lives for their principles.

    I happen to believe that human life should take priority over hurting someone with torture.

    Specifically, if you knew that your entire family would die, unless torture were used to intervene, would you be against torture?

    I am not trying to be provocative, I simply believe human life takes precedence over worrying about torture and I’m confused why others (unless they are some sort of elites) could possibly think otherwise. Please answer my question about your family at risk from torture.

    • Andrew says:


      I would not want anyone in my family to be tortured, even if it would help some inquisition dude maintain his quota of confessions for the month. But I guess elitist inquisition officials might feel differently; they might be more than willing to sacrifice human lives for their principles.

      • Phil says:

        Andrew, you’re being unfair to David here. He didn’t say anything about ‘maintaining a quota of confessions’, he is hypothesizing a situation in which torture would definitely save a human life or the serious injury of someone.

        David, your example seems to implicitly assume the event is somehow isolated, just a simple choice between (torture, save lives) and (no torture, people die). There may be some real examples of that, and it’s almost tautological that we wouldn’t know about them. All of the examples of institutionalized torture that I know of are parts of torture campaigns.

        Rather than your simplified view, I think what we face is more like a choice between (torture, save lives now, provide propaganda to recruit more terrorists so more people die in the future) and (no torture, people die now, terrorists have a victory but have no moral high ground, unclear whether more or fewer people die in the future). I very much doubt the waterboarding the U.S. did in the past nineteen years saved more lives than it cost.

        Many people have been tortured, most of them not ‘elite’, and abhor the practice and think we should have no part of it. You might consider that before calling rejection of torture ‘elitist’.

        • Andrew recently dismissed the use of game theory as having relatively little influence these days, however it’s interesting how fundamental the example you give was to early game theory thinking. iterated prisoner’s dilemma is a very different situation than single instance. when it comes to international policy everything is an iterated “game”. the consequence of doing the locally optimal thing can sometimes be the worst outcome when iterated

      • David Baer says:

        I believe that the potential to save a life or serious harm is far more important than worrying about torturing someone to save a life.

        There are some elites (typically elites) that put principles such as “no torture” above human life (as long as it is someone that isn’t *their* family member or loved one).

        I put preserving human life or serious injury above that of torture.

        I believe terrorists should be tortured if it can save lives or prevent serious harm and I don’t understand those who don’t make the saving of human life or bodily harm a priority.

        • Your principle could be fine, if there were any evidence that torture, on the whole, could save human lives in bulk.

          But while it’s possible that in some specific instance, some specific person being tortured could save some specific victims from being killed… The global reality is that this seems unlikely to scale.

          Suppose for example you recognize that some specific person has plans to do some kind of terrorist activities, so you use a drone to kill this person. Then 1000 marginally angry people become 1000 extremely angry people, and each of them plots to do the maximal damage possible to your country. Having killed this one person saved maybe 30 people from dying, but turned 1000 people into terrorists who on average each kill 3 people, resulting in 3000 extra deaths in total.

          This is not an implausible dynamic.

          So, there’s a difference between “this instance of torture will save these N people from a particular plot to kill them” and “torture/killings/etc as a policy will reduce the number of people killed by terrorism”

        • Phil says:

          David, when it comes to institutionalized torture I just don’t know where this ‘elites’ thing comes from. It’s the ‘elite’ who do the torturing; it’s the non-elite who get tortured.

  22. Renzo Alves says:

    Perhaps it would help to learn something about the history of torture during the so-called Spanish Inquisition before drawing conclusions from it.
    Recommended reading: “For the Glory of God,” (esp. pp. 256-261) by Rodney Stark (Princeton UP, 2003).

    • Andrew says:


      OK, I got to this part, from p. 259:

      The Inquisition in Aragon and Castile sustained charges of witchcraft and imposed severe punishments only when they confronted seemingly valid evidence of explicit satanic invocations—as in cases of third- and fourth-time offenders or those who defied the court and refused to express contrition.

      Look, I’m all for cultural relativism or whatever you want to call it, but . . . witchcraft isn’t real! There’s no such thing as a third- or fourth-time offender when there’s no actual offense. Also not clear why people should express contrition for something that’s not real. But, OK, I get it: they’re imposing severe punishments not for witchcraft but for dissent: unwillingness to tell a lie, just cos the authorities want you to.

      The book follows with a bit about how the Inquisition was “extremely reluctant, and eventually unwilling, to resort to torture,” while, “In contrast, many local, secular authorities in Spain . . . resorted to torture to prove their cases.” Fine. I can believe the locals were worse. It also seems to me that the church authorities weren’t helping any by creating a framework where there was such a thing as legal execution of witches, setting up a system where they could act as the moderates and their more extreme allies could do the burning.

      And this:

      Thus was witch-hunting minimized in Spain. The inquisitors refused to assume satanism unless driven to do so by substantial and uncoerced evidence, and even then they were extremely reluctant to execute, seeing it as their proper duty to bring the offender back into the good graces of the Church.

      Well, I guess if there was substantial evidence then it was ok, huh? And they were “reluctant to execute.” How moderate of them!

      Anyway, sure, I know that life was different in the past: they used to have the death penalty for nonviolent offenses. And, arguably, the prison sentences that are given today are more barbaric than the flogging that was done in the past.

      So I’m not saying the Spanish Inquisition was uniquely bad. The inquisitors may have been no worse than other power-mad despots of their era. Maybe what particularly bugged people about the Inquisition was their combination of brutality, legalism, and fine words. They weren’t murdering people who threatened their power; no, they were “reluctantly executing” people who, based on “substantial and uncoerced evidence,” were “refusing back into the good graces of the Church.” A bit Orwellian, I’d say (to use an anachronistic term).

      • Dogen says:

        Yes, not to mention that the Inquisition in general, and the Spanish Inquisition in particular, also tortured and burnt at the stake: Jews, Moors, Lutherans and other Protestants, bigamists, and many other “heretics”. The explicit goal was to terrorize people opposed to the Catholic orthodoxy, and it was apparently very successful.

        For example, the Spanish Inquisition pulled in something like 100,000 accused Jews (numbers rough, from my poor memory) and gosh! They ONLY BURNED ABOUT 2,500 of those horrible sub-humans. OBVIOUSLY THE SPANISH INQUISITION WASNT SOOOOO VERY BAD!!!!!

      • Terry says:

        Look, I’m all for cultural relativism or whatever you want to call it, but . . . witchcraft isn’t real! There’s no such thing as a third- or fourth-time offender when there’s no actual offense. Also not clear why people should express contrition for something that’s not real.

        If everyone involved believes witchcraft to be real, are you not faulting people for not being able to see hundreds of years into the future? Isn’t there a moral distinction between honest error and malicious dishonesty? Further, it is my understanding that some people actually believed themselves to be witches and actually practiced witchcraft at the time.

        The passages you cite show the Spanish Inquisition to be a significant step into a more progressive future even if it fell short of omniscient perfection. Is your heart so hard that you cannot acknowledge that? You can see the Spanish Inquisition as executing X innocent people or as saving Y innocent people, where Y seems to be >> X. It is possible to cheer Newton’s achievements even though he was completely wrong about alchemy.

        To be fair, I have the same tendency. I fell righteous outrage when I think of doctors bleeding patients to death, even though I know most doctors had the best intentions.

        Wikipedia has something on the numbers executed. The Spanish Inquisition looks to be a big step forward.

        García Cárcel estimates that the total number prosecuted by the Inquisition throughout its history was approximately 150,000; applying the percentages of executions that appeared in the trials of 1560–1700—about 2%—the approximate total would be about 3,000 put to death. Nevertheless, some authors consider that the toll may have been higher, keeping in mind the data provided by Dedieu and García Cárcel for the tribunals of Toledo and Valencia, respectively, and estimate between 3,000 and 5,000 were executed.[144] Other authors disagree and estimate a max death toll between 1% and 5%, (depending on the time span used) combining all the processes the inquisition carried, both religious and non-religious ones.[117][145] In either case, this is significantly lower than the number of people executed exclusively for witchcraft in other parts of Europe during about the same time span as the Spanish Inquisition (estimated at c. 40,000–60,000).

        Similarly, the people involved in the Salem witch trial acted with surprising humility and morality. Within about a year, they acknowledged their mistake and performed public contrition. For one judge it triggered a profound moral introspection that led him to free his slaves, so the Salem Witch Trial can be seen as the beginning of the abolition movement. That sort of thing doesn’t happen much today.

    • Michael Schwartz says:

      It is my understanding that torture was not the main weapon of The Spanish Inquision. Rather, their main weapons were fear, surprise, a ruthless efficiency, etc…

      I thought everyone knew this.

Leave a Reply