Hacker News Reader

Comment by rjtobin | Hacker News Reader

About 4 years ago I went to a talk given by the author of the book, the subject was along the same lines as the book: how torture is ineffective because it damages the subject's ability to recall information. Afterwards someone asked a question that a lot of us had been thinking: that in a lot of situations where torture might be used, the information being asked for is very well known to the subject, so unless serious brain damage occurred, the subject wasn't going to forget after torture. Like, "who is your cell leader?" or "where are the guns hidden?". He seemed confused by the question, even after a few other audience members tried to rephrase it. Got the impression he was being intentionally obtuse. I guess it deflated the gist of his work, which more accurately should be something like "torture ineffective when the concealed information is hard to remember".

I haven't read the book, but I hope in the meantime he has worked a little to address this question. Seemed like a well meaning guy though, and maybe my memory of the talk has been eroded by the torturous years of grad school in between

View on HN

I would argue your audience is failing to understand what torture is. Most people are deathly afraid of any pain. But they're also not agents/operatives/soldiers for their government. They're not at war with an enemy they think worth fighting and killing.

So they don't have a particular strong desire to conceal information from anyone who threatens them. This means, in an actual scenario of torture interrogation, you're pretty likely to initially withhold whatever you can. But by the time you reach an actual breaking point...well it's called that for a reason. You are in fact very unlikely to remember much of anything because of pain, shock etc. You're pretty likely to say whatever the torturer is asking to hear.

Waterboarding is torture because once water gets into your oesophagus your rational mind shuts down completely - this post (http://boards.straightdope.com/sdmb/showthread.php?t=448717) pretty much summarizes it and the experience has been confirmed by everyone else who has had it done to them.

That's the point about torture being ineffective. By the time it's effective, you're dealing with someone who is pretty much going to confabulate whatever you want to hear.

> I would argue your audience is failing to understand what torture is. Most people are deathly afraid of any pain. But they're also not agents/operatives/soldiers for their government. They're not at war with an enemy they think worth fighting and killing.

It almost sounds as if you are trying to define away torture for a fringe case where it's guaranteed to be useless. This is a legalistic approach not anymore convincing than attempts to define waterboarding as "not torture".

Torture is used in the armed conflicts depressingly regularly: the corresponding clauses in Geneva Conventions for treatment of POW were added not for purely theoretic reasons. Torture is used by criminals around the world on daily basis. The reason they haven't given up after millenia of practice is that it tends to work in their unsophisticated, easily verifiable cases: getting the combination to a safe or finding where a merchant keeps their money stashed.

Anyway, torture should not be used because it's simply an abhorrent thing to do to a human being. Trying to frame it in terms of performance is damaging to the moral aspect of it, as you descend to the same inhuman cost/benefit mindset as torture proponent.

AFAIK, torture does not work for "getting the combination to a safe or finding where a merchant keeps their money stashed". It is still used because it is a very effective way to frighten the opposition. Torture is often shown on TV, where the good guys force the bad ones to tell the truth. It is completely unrealistic.

> It is completely unrealistic.

Back in the 1990s in former USSR bugrlary-torture-murders were daily occurrences. It sure as hell worked, for purposes far more simple than "frightening the opposition". Stuff much too savage for TV shows typically, like soldering iron up one's anus.

Such barbaric methods makes almost no difference for the victims who will probably die. The difference is the effect on population. If the purpose was to get information, I think there are more effective ways: chemistry or hypnosis for example.

The purpose of information, however, is money. Which adds more data to this: if you're in early 1990s former USSR and you want to steal a great deal of money, there were far less brutal ways to do so... Here's how one fellow made $10bn (no home invasions and soldering irons required...): https://en.wikipedia.org/wiki/MMM_(Ponzi_scheme_company)

Ah, "racketeer" (written as "PAKETEP" in Cyrillics, iirc -- a borrowed word from English, but used to mean an individual extorting someone vs. organized extortion of a business) with a soldering iron or a clothes iron... the reason my parents for the longest time told me 1) never tell others we have a home computer 2) the model of our home computer was a "Sinclair" (it was actually a 386, but they were afraid I'd blab about having a computer in the first place... of course I blabbed about having a Sinclair, to someone who first volunteered information about having an actual Sinclair, leaving him confused as to why our Sinclair ran MS DOS and why I programmed it in Turbo Pascal...).

The general big fear was being burgled and tortured for money you were presumed to have had but actually didn't. I suspect this was the intent of the burglars: most any rational person would prefer to give up their money rather than being tortured with a soldering iron -- and these were heavy soldering irons used for repairing 220v home circuits (culture shock coming to the US: people didn't fix simple things like an outlet/switch failing to work, but I digress...) -- but the stories of those who didn't have money but were tortured and murdered anyway, made those who had the money (and might decide to volunteer only a portion of it) realize that the threats weren't kidding.

However, that's not the full story. I really do wonder how much the soldering iron added over simply barging into a home, demanding the money. I suspect there is also huge element of, as Orwell said in 1984, torture for the sake of torture itself (sadistic personalities amongst the criminals, rather than pure rational calculation as means to making money).

Re, TV: the soldering iron thing _did_ make it into some Russian/late Soviet movie I watched back in the 90s, not in a graphical scene, but the threat: the militisia agent -- it was stilled called that back then -- was refusing to take the reported threat seriously and was simply cracking jokes about it.... A graphical scene (albeit comedic) with a clothes iron did make into one the more famous movies of that era, ironically as something that happens in US (ah, projection!) https://en.wikipedia.org/wiki/Weather_Is_Good_on_Deribasovsk... )

That said, growing up in former USSR (Belarus, specifically), I was also surrounded by plenty of stories of partisans (anti-Nazi resistance fighters) saying nothing when questioned by the Gestapo, as well as tactics the partisans used: assume anyone can be captured or be a traitor, make any one piece of information useless by itself, and so forth. This couldn't be propaganda, as the French resistance stories I was exposed to in the US were also quite similar. I suspect resistance to giving useful information under torture when 1) one has a lot more at stake than money 2) the organization one is a part of has a choice about how to distribute the information.

In fact in France, there is a strange concept called "juge d'instruction". Basically a district attorney with the power of a judge. And this power includes sending someone to jail pending a trial.

Over time, these judges have been known to send people against who they only have a weak case to jail to make them talk. The state of French jails is appalling, many date from the XIX century, infested with pest, and all are overcrowded.

One could argue that this is a form of torture. Particularly for white collar crime.

If it's intentional infliction of suffering then it indeed can be argued that way.

This is some other thing though: police/prosecution using torture as coercion into admitting guilt. They don't really care whether it's true or not, their goal is just chalking up some bureaucratic milestone ("case closed").

> Trying to frame it in terms of performance is damaging to the moral aspect of it,

Well, I prefer performance-based explanation as to why not use torture. The problem with morals is that they change. And they tend to provide a lot of backdoors. Like, it's obviously bad to torture, but what if you could save X lives by torturing one person?

With a performance-based explanation it's easy: nothing because torture won't give you reliable information and you won't save the people that way; better start thinking about some other way.

But arguing from a performance perspective on torture is often dishonest, because people start from a conclusion they would like to reach, and then work backward in any way possible to make their claim intellectually respectable.

It's dishonest because high-quality research into torture is not really done on a scientifically known level (maybe governments have such knowledge), and it's nigh-impossible for any serious body of scientific evidence and insight to occur. The consequence of this is that people are free to speculate anything at all, insofar as they erect an intellectual barrier capable of thwarting an audience. Like that author did as they appeared deaf to audience questioning.

We make cost-benefit analyses all the time. With any military operation the collateral damage has to be weighed against the target. The fact that we freak out over torture yet kill innocent people in collateral damage is completely irrational. On the scale of harm, torture that leaves no physical harm is behind cutting off people's limbs, and that is far behind killing people. Now, torturing innocent people is worse than torturing (most likely) guilty people who (most likely) know something that can save lives. But for some reason mentioning torture reflexively shuts people's brain off yet killing innocent people in collateral damage is fine if subject to a cost-benefit analysis.

Using war to justify torture is like using murder to justify rape.

Are saying that collateral damage is not justified under any circumstance? Otherwise your witty response does nothing to adress the point that I made. We already do not condone murder nor rape.

War is a barbarism at industrial scale, which humanity, unfortunately, haven't managed to eradicate. There are however certain undertakings to reduce human suffering: bans of certain particularly inhumane or indiscriminating weapons, prohibitions of deliberate targeting of civilians, and yes, elimination of torture.

If you make an effort to reduce suffering it comes as a complete package: people quick to torture are the same people who would use Sarin on civilians. There is simply no middle ground. The concept of innocence doesn't come into play anyway. Wars are mass murders stemming from greater social and political processes; a conscript soldier guarding a fuel depot is not any more "guilty" apriori than his civilian relative.

There is more or less intellectual consensus that wars cause suffering, and attempts are made to reduce the misery. Propagating torture is not serving a good cause here. You essentially argue that a "lesser" crime is acceptable in context of a "greater" crime going on anyway. If one's going to kill the victim might rape her anyway.

That is not what I'm saying at all. What I am saying is that compared to collateral damage the discussion about torture is completely out of proportion. Of course the concept of innocence comes into play. I'm not talking about war. I'm talking about drone strikes against terrorists which also kill their wives and children vs torturing a terrorist. The drone strikes are an almost unimaginably greater evil, and yet time and time again you have people taking a hard line stance against torture yet are willing to make a cost-benefit analysis in the case of drone strikes or special forces operations.

Note that I am not taking any pro or contra torture position. I'm merely pointing out the moral inconsistency.

Given that one of the most commonly discussed and unpleasant torture methods involves simulating the experience of drowning, I've often wondered why so few people consider inflicting that exact same experience on many enemy sailors by means of missile or torpedo to be even remotely bending the rules of engagement. Granted, the person that fires the fatal shot probably isn't an incorrigible sadist and may even be at imminent risk of a similar fate themselves, but they're still inflicting that suffering with a similar level of intentionality and certainty to a CIA operative with a bucket...

Hurting enemies while they're fighting, presumably to make them stop being able to hurt you or whomever you're protecting, has always been granted a special moral free pass.

But if there is no fight occurring, and the enemy is subdued and unable to cause harm, then the situation has changed a bit, into a greyer area.

I've also met a guy who has used water-boarding in the field, and he didn't seem like a sadist. He felt pretty bad about using it, though he thought it effective and necessary.

Torture is harming with intent. Collateral damage is hurting by neglect. People pay attention to intent when judging.

The intent of e.g. a drone strike is to kill the target. So even in terms of intent that is worse. The thing is that nobody even cares about that: arguments against drone strikes are virtually always about collateral damage of innocent people. Compare and contrast with the discussion about torture where people are not even willing to make a cost benefit analysis.

> The intent of e.g. a drone strike is to kill the target. So even in terms of intent that is worse.

For you and me yes. For people who believe targets are "bad people" worth killing it's noble intent.

Well, lets talk about people who are not like that, but people who think drone strikes are might be okay in certain circumstances. My point is that those are the same people who say that torture is not okay under any circumstance. That is the inconsistency I'm pointing out.

I think in case of drone strikes those people can claim that people (other than target that should in their opinion be killed) are killed by accident.

In case of torture you can't claim that individual you torture is somehow accidentally hurt. Probsbly those people would be fine with torture if they thought hurting the tortured individual is desired outcome.

It's not really an accident though if it's a likely outcome. The point is simply this: on the one hand people accept a substantial probability of killing innocent people and intentionally killing guilty people if the upsides are good enough, but these same people are unwilling to accept mere torture of guilty people no matter the upsides.

Maybe it's something about drones being remote. Torture was shown so many times on tv that it's closer to home. Not many pictures of people blown to bits by invisible thing in the sky.

The question why people are so bad at intuitively weighing harm is a different but interesting question. There have been some studies that suggest that there is an inverse relationship. When asked how much money they would give to save 2 lives people offer less than they do when asked to how much they would give to save 1 life. Not less per life, less total. As the number of lives goes up, what they offer goes down. This may have something to do with abstracting the harm. With torture it's easy to understand what's happening. With a single death it's harder to fathom how much is lost. The thousands that die in traffic each year hardly register.

The incompetence of humans to intuitively make moral judgements is precisely why we should think rationally about morality rather than emotionally.

There isn't much evidence for torture being used except as a form of punishment. Even the famous medieval torture was intended as a part of the punishment and to deter crime, rather than finding valuable information. Torture was only used on people convicted by other means, because it was technically illegal to punish anybody who didn't plea guilty.

There's plenty of evidence, if you are not specifically determined to look the other way. Waterboarding was used not to deter crime or to punish.

But isn't inability to remember much because of pain/shock the point? That fragments of actual memories of places been and instructions given might be a little stronger than a torture subject's capacity to invent alternatives, or even recall why they were so keen not to give that information in the first place. Useless to establish guilt or complex plans, but quite possibly better than nothing for a few elusive details, assuming that "nothing" is what more acceptable interrogation techniques have yielded.

Of course, it's even easier for the subject to simply agree with the torturer if the torturer actually has a narrative they want agreed to, but in those case actually obtaining information was obviously never the reason for the torture in the first place...

This is not how torture works in real life. At least, not the one made by security forces.

Both of my parents were tortured, both faced it several times. My father was in prison twice and my mother was never in prison but took directly to military places. But not only they were tortured, several people from the circle of family friends were tortured too. One of them now works in a state program trying to help the people who were tortured to continue with their lifes, even if they were tortured in 1974-1980.

While the 1973 was made with support of US and brazilian governments, there was no support when the firsts arrests came. Most of people faced strong pain under different methods, including electricity, beating to almost dead, removing nails, etc... but it was only after the US experts arrived that torture became effective.

Torture is about removing will. A common practice after that was not only rape by military, but they put dogs to rape woman. They took their fingers off and put them in cells full of shit so they had to avoid the infections. Put them in really small spaces until they were lost.

This were process made by days, they didn't get into shock status, they were destroyed as human beings. When they go out, torturers didn't ask them: "hey, how do i find this person"... they just took a bunch of pictures and check the reactions. Or they put them in a car and go around the city and let them see somebody they knew.

By the way, when you reach this point you don't care about torture anymore like to try to avoiding it letting the torturer hear anything they want to hear. You are broke, you don't have will, the only thing you can wish is to be dead. You don't resist anymore.

Torture was so effective that political organizations who were fighting the dictatorship adopted the policy of when somebody is arrested, all their known people had to run and disappear in 24 hours.

Here's the reason that it doesn't work. Imagine that your village has a 1000 people and someone is torturing you to find out "who is your cell leader." At first you are resisting, and so you name random people in the village. After you've named 50 people or more, you break and name the real person. The torturers though don't know that this is the real person, so they'll keep torturing you, and you'll just go through every name in the village to make the torture stop. How can they separate the real person from the ones you lied about either out of resistance or to make the pain go away?

In some cases this could work. They want to find out where you hid the guns, and every time, they go and dig till you break and give them the right place, and then they do find the guns.

What you've said sounds like a reasonable argument why torture doesn't work in that case. My point wasn't that torture does work (I have no idea about that), just that the research outlined by O'Mara doesn't address situations like this (and so the research isn't enough to prove that "torture is ineffective" as is suggested).

Usually you ask for verifyable information. CIA officers do not need to be reminded what a lie is.

I think the comments of the FBI revealed wrt. the CIA torture programs shows that CIA officers a) don't know how to interrogate people, and b) do need to be reminded what a lie is.

I agree that if what you want is some concrete information that the subject is reluctant to give up ("Where have you buried the guns I need in order to kill the rest of your family?") -- you could probably get at that via torture. I don't know if that would typically be the fastest or most reliable way to get that information (Consider, eg: Staging an escape, allowing the subject, helped by an under-cover agent, point you to said guns).

I think people get hung up on the "bomb plot theory" too much. My impression is that while such simple cases make for great TV, that's not the kind of real-world questions intelligence organizations really are concerned with. So if we say: "ok, use torture to save the children!" -- if that was taken at face value -- it means: "Disband most (all?) large-scale operations that allow torture today, such as the CIA programs."

I also think there's a bit of a hang-up on "effectiveness". If you're fighting an illegal war, kidnapping and killing people that are clearly innocent -- it's hard to accept an argument from "effectiveness". At least if there's no indication that atrocities would stop if you "won".

Taking it from a more political point of view rather than neurological, Shakespeare did a political play on this subject which was summarily all about blow-back or stone cold revenge — "Titus Andronicus." (A movie by Julie Taymor of the play "Titus" is quite powerful)

In politics and war The dirtiest fighter makes the rules and now that the United States tortures people, I suspect we will be "one-upped" in the next war.

Indeed, once you resort to torture or other "questionable" means, you can't accuse or deny enemies from using them too; this sets new standards. The argument is generally "but they already do it", and in my opinion it fails to recognize 1) the large scale validation for every belligerent, 2) the race to "no-rule" warfare, as limits are one-upped once (de facto) accepted. So it's not because _some_ belligerents are willing to use banned means that everybody should start using them too.

Yes, many of my worries about what my government get up to are based on the fact that they lose all moral standing when others do the same, see Russia bombing Syria for a recent example.

> so unless serious brain damage occurred, the subject wasn't going to forget after torture.

Really? You have data on this? I'd love to see it.

However, there are multiple problems with your scenario.

The biggest being: what happens if the individual DOESN'T KNOW? He's going to tell you anything he can to get you to stop, and, if you could verify the crucial piece of information, you wouldn't be torturing him in the first place.

So, even if I concede that torture works when you know you have the right person (and I do not), the problem is that you can't distinguish whether you have the right person and whether he actually knows the information you want.

>>if you could verify the crucial piece of information, you wouldn't be torturing him in the first place.

This is not true. There are many things easy to verify but difficult to discover. For example a place the money is kept in a big house.

Yeah, I always hoped HN was the last place where you actually had to convince people that NP-complete problems exist...

Real world problems are often NP-hard, not NP-complete.

Well yes. But I think the parent was framing it with reference to the intelligence world conflicts. At least that's how I Read it.

Imagine you torture a guy to get the name of it's cell leader. You can't verify this easily. You have to capture that other guy, torture him and repeat the process. While in any of those steps you have the chance of getting false information without knowing it.

I would think most intelligence from a war zone would be of this type, i.e. hard to verify, and I think the parent was thinking along those lines as well.

Most of the people the US was holding at gitmo and other black sites around the world were simply people who were at the wrong place at the wrong time. But you never know that for sure. It's unconscionable that random innocents be tortured for they may yield some information. But allowing torture by using extreme cases like your's, in reality it is almost never that cut and dry.

I don't at the moment have time to read the article, but I will note that you do NOT need to have brain damage from the torture in order for it to interfere with recall.

Memory is state dependent. You recall things best when in a similar state to when you usually access that information. Placing someone in an extreme altered state can interfere with memory access of "normal" memories, without brain damage needing to occur.

The degree to which memory is state dependent was really brought home to me when I had a serious medical crisis some years back. The birth of my first child was very difficult and I usually have trouble remembering the details. During my medical crisis, I was in constant extreme pain for 3.5 years. During that time, I was able to remember the birth of that child with tremendous clarity after years of remembering it foggily.

You do not need to cause brain damage to interfere with mental processes like recall.

You greatly underestimate how much the brain shuts down in stress situations. People may not hear being talked to, recognize familiar people or notice severe wounds. It's the reason behind such rules such as no cleanup during training - it lead to people picking up spent ammunition in real gunfights.

During torture, the stress levels must be high enough so that the subject forgets that they don't want to tell you anything important, and at that point they may not be able to recall their own name. And of course people tell intentionally inaccurate, false or otherwise misleading information to avoid further torture.

So torture may work for those who somehow ended up on the enemy side and don't actually care that much one way or another, but not people who are likely to have valuable information. Even the Nazis knew that and instead tried to convict the interrogated they are civilized, well meaning people.

It's not clear that torture makes it especially likely to prevent lying about information, though.

It'd be nice if he studied that instead of making us assume it, though :-P

All information is hard to remember when you are in serios distress. Even your name after few nights of sleep deprivation doesn't sound obvious at all.

You will trash subject ability to be sure about anything well before you make him cooperaive.

Reply on news.ycombinator.com