John Oliver discusses torture

[marplots]

Ok, now how about if there is only a 50% chance of the bomb going off at all, so that if you did nothing, there's a 50% chance of no harm? Is it worth pressing the button, even though it's only reducing the chance of death to the people trapped in the room from 50% to 0%?

I'm starting to really dislike those people in the room.

 

[Belz...]

Ok, now how about if there is only a 50% chance of the bomb going off at all, so that if you did nothing, there's a 50% chance of no harm? Is it worth pressing the button, even though it's only reducing the chance of death to the people trapped in the room from 50% to 0%?

Coldly, I would say yes.

 

[Belz...]

This is why I'm picking on you in this thread; you claim to want to discuss facts, but all you've done is construct hypotheticals...

That is a lie. I've said that discussing facts is more interesting than discussing personal morals, and I've done more than construct hypotheticals. I didn't even construct one in the post you quoted. Perhaps you have me confused with someone else, which would be hilarious since you claim to focus on me.

 

[jimbob]

There is no interrogation technique which is perfectly reliable. And of course torture can (and often does) work. That's why we train people to try to resist it.



Indeed. Especially since the interviewers are often looking for information which they will be able to confirm (such as the location of a target). Tell them what they want, the torture stops. Tell them a lie, it only stops for as long as it takes to determine that you lied, then it starts again.

There are multiple reasons to avoid using torture. The "fact" that it doesn't work isn't one of them. It does work. The idea that it doesn't is a lie people tell themselves to try to avoid the difficult moral decisions by pretending that they're simple.

I'd flip the argument the other way up.

As you said, there are historical examples of torture working (the one of Guido Fawkes is an obvious one).

However there are plenty of examples where it didn't work or was counterproductive.

http://www.theguardian.com/commentisfree/2014/dec/12/us-torture-shaker-aamer-cia-guantanamo

Or this from the Communists in Vietnam:

http://www.newsweek.com/cia-torture-report-vietcong-vietnam-war-292041

The CIA is hardly the only spy service to grapple with blowback from making prisoners scream. Even leaders of Communist Vietnam’s wartime intelligence agency, notorious for torturing American POWs, privately knew that “enhanced interrogation techniques,” as the CIA calls them, could create more problems than solutions, according to internal Vietnamese documents reviewed by Newsweek.

In many cases, torturing people wrongly suspected of being enemy spies caused “extremely regrettable losses and damage,” says one of the documents, released to little notice in 1993 by Hanoi’s all-powerful Public Security Service (PSS).

In another part, the report explains how the agency eventually began to arrest its own people, and before long, false confessions began destroying the PSS itself. “A reconnaissance team member who one day went out to arrest someone would suddenly be arrested the very next day,” the documents said. “An investigating officer who one day was sitting in his chair conducting interrogations would the next day be forced to stand in that same office to be interrogated himself.”

As well as the several cases where the Gestapo failed to get information from people they captured.


In the 1980's I read a book about MKULTRA, "The Search for the Manchurian Candidate

It listed several attempts - mainly using non-physical torture to determine whether intelligence was good or deliberately misleading. They got the prisoners to admit to lying, but they also said whatever they thought the people wanted, which was useless

ETA: which means an argument for its use based on its efficacy is not that strong. There are plenty of other reasons why it should not be used even if it was effective - which I think is where you are coming from.

 

[Ziggurat]

ETA: which means an argument for its use based on its efficacy is not that strong.

That is true. The flip side of the fact that there are no good studies showing torture is less effective than other methods is that there are also no good studies showing it's more effective. The people who get their panties in a bunch when I say torture works haven't been able to think clearly enough to notice that, though (I'm not calling you one of these people, just to be clear).

There are plenty of other reasons why it should not be used even if it was effective - which I think is where you are coming from.

Also true.

 

[sunmaster14]

Coldly, I would say yes.

I agree. The next step of course is that those responsible for waterboarding KSM would argue that they were faced with that situation, although the probabilities and numbers of people involved were different. I wouldn't necessarily agree with them (I probably wouldn't actually given my view that, for obvious reasons, the intelligence community generally exaggerates threats), but it's difficult for me to put myself in their shoes.

 

[Beelzebuddy]

As you said, there are historical examples of torture working (the one of Guido Fawkes is an obvious one).

I don't know about that one. When Zig mentioned Guy Fawkes, it got me curious enough to read up on it. This is, they got the names of a number of "conspirators" from torturing him, who were given show trials and all executed at the same time. It's likely (from a spied-upon conversation) that at least one of the names was a real conspirator, but they weren't exactly trying to weed out false positives there or anything. We can see from Salem a few years later that actual crimes, knowledge, or plain facts aren't necessarily involved in such coerced implications.

To use people's PIN hypothetical from upthread, yes, it's quite likely that someone can be tortured into giving up their real pin number, after a few false ones, of course. But the thing is, if you happen to incorrectly input the PIN, and you go back and say it isn't working and torture them some more, they'll still give you more numbers to make it stop, and you'll have no way of knowing that the truth passed you by entirely.

That's the thing that the "torture doesn't work" people are getting at. What makes or breaks a tool of intelligence isn't whether or not it ever signals a true positive (if that's all you care about, flip a damned coin, it's right half the time) but that it provides a level of accuracy (minimizing false negatives AND false positives) which makes it worth bothering with. I don't think anyone would deny that torture is a ridiculously inaccurate method of gathering information, which is why all the hypotheticals we've been discussing in this thread are things we can test right away, like PIN numbers or bomb locations. The only times that torture accurately provides information is when you already know the right answer and can discard false positives, or you don't much care what the answer is and can accept false positives, or you're just doing it to be sadistic.

That is why it doesn't work: because it only "works" when you don't need it to work.

 

[Belz...]

I agree. The next step of course is that those responsible for waterboarding KSM would argue that they were faced with that situation, although the probabilities and numbers of people involved were different. I wouldn't necessarily agree with them (I probably wouldn't actually given my view that, for obvious reasons, the intelligence community generally exaggerates threats), but it's difficult for me to put myself in their shoes.

Ah, I see your point.

Gee, it's almost as if we're having a conversation, here.

 

[magellan]

... That we know of.
Sometimes withholding op details trumps crowing about a success.

Of course.
There are a lot more ticking bombs that don't go off than the few we are allowed to know of.

ticking bombs that went of:
Boston
Madrid
London
Oklahoma city bombing
Tokyo subway 199(7 i believe)
Antalya 1995 (PKK, minor, they blew up some trash bins. But it left an impression. I was on holiday there back then. Again: don't nail me on the year) [they are good guys fighting ISIS now]

seems to me the key to operational success is to keep it a secret.
Once you know that there is a ticking bomb it's easy to avoid.

 

[Ziggurat]

I don't think anyone would deny that torture is a ridiculously inaccurate method of gathering information

You have no evidence that it's more inaccurate than other methods. People can lie under non-torture interrogation, and you can't know when they're lying, just like with torture.

which is why all the hypotheticals we've been discussing in this thread are things we can test right away, like PIN numbers or bomb locations. The only times that torture accurately provides information is when you already know the right answer and can discard false positives

This problem is no different with torture than it is with any other interrogation technique. You never know if you got the right answer until you test that answer. And being able to test the answer is not equivalent to having the answer by independent means.

 

[sunmaster14]

I don't know about that one. When Zig mentioned Guy Fawkes, it got me curious enough to read up on it. This is, they got the names of a number of "conspirators" from torturing him, who were given show trials and all executed at the same time. It's likely (from a spied-upon conversation) that at least one of the names was a real conspirator, but they weren't exactly trying to weed out false positives there or anything. We can see from Salem a few years later that actual crimes, knowledge, or plain facts aren't necessarily involved in such coerced implications.

To use people's PIN hypothetical from upthread, yes, it's quite likely that someone can be tortured into giving up their real pin number, after a few false ones, of course. But the thing is, if you happen to incorrectly input the PIN, and you go back and say it isn't working and torture them some more, they'll still give you more numbers to make it stop, and you'll have no way of knowing that the truth passed you by entirely.

That's the thing that the "torture doesn't work" people are getting at. What makes or breaks a tool of intelligence isn't whether or not it ever signals a true positive (if that's all you care about, flip a damned coin, it's right half the time) but that it provides a level of accuracy (minimizing false negatives AND false positives) which makes it worth bothering with. I don't think anyone would deny that torture is a ridiculously inaccurate method of gathering information, which is why all the hypotheticals we've been discussing in this thread are things we can test right away, like PIN numbers or bomb locations. The only times that torture accurately provides information is when you already know the right answer and can discard false positives, or you don't much care what the answer is and can accept false positives, or you're just doing it to be sadistic.

That is why it doesn't work: because it only "works" when you don't need it to work.

No, it also accurately provides information when you can verify an answer trivially, but there are millions or even quadrillions of possible answers, only one of which is true, so that you can't come up with the answer yourself. Do you really think that is so implausible? That's how almost all computer password systems work actually. You used a password to login to ISF, but it's impossible for the admins at the ISF to determine what your password is (at least it should be).

 

[marplots]

That's the thing that the "torture doesn't work" people are getting at. What makes or breaks a tool of intelligence isn't whether or not it ever signals a true positive (if that's all you care about, flip a damned coin, it's right half the time) but that it provides a level of accuracy (minimizing false negatives AND false positives) which makes it worth bothering with. I don't think anyone would deny that torture is a ridiculously inaccurate method of gathering information, which is why all the hypotheticals we've been discussing in this thread are things we can test right away, like PIN numbers or bomb locations. The only times that torture accurately provides information is when you already know the right answer and can discard false positives, or you don't much care what the answer is and can accept false positives, or you're just doing it to be sadistic.

That is why it doesn't work: because it only "works" when you don't need it to work.

You are demanding too much of the method. First of all, as Zuggurat pointed out, the same criticism applies to any interrogation method used. It even applies when someone is entirely willing to tell you the truth - they can still be mistaken or ill-informed or biased.

And that's why the second part - consilience - rules the day. We do it in areas completely divorced from torture - we look for converging lines of evidence to give us confidence that can't be obtained with one line of evidence alone.

I reject the picture where "tell me anything and the torture stops" is the rule. You'd instead use torture as a punishment or a threat for lying, not in a "make it stop now" way. I'd wager that the most effective torture is the torture that never actually happens at all.

 

[marplots]

I'm not well informed on the use of torture in the Vietnam war to extract information from US and allied troops. Was it done? Did it work? I imagine our returned POWs would know whether or not they "gave up the info" when tortured.

This, of course, would be torture with a specific purpose divorced from extracting a false confession or for sadistic reasons.

I did find this account:
http://www.defense.gov/news/newsarticle.aspx?id=25312

And while the victim brags about creative lying, there's a subtext saying he did "give it up."

Swindle said the experience with the ropes from his second day in captivity was repeated four or five times before he was moved to the infamous Hanoi Hilton prison camp in late December of that year. The lies were all about coping. Prisoners learned many different ways to cope in their tortured captivity. "You give in," he said. "So don't think you don't give in. But it's how you do it and what you give them.

"You're trying to minimize what they gain from you, because we have a very strong sense of how they use propaganda," Swindle explained.

Does that count as working? Depends, as usual on what you expect torture to accomplish.

 

[jaydeehess]

Yes, like saying that if you go to the grocery store tomorrow and take a carton of milk to the cashier, you'll be asked for some money. It's a hypothetical but it's so likely to be true that I would hesitate to call it that. I don't know about you, but I would give my PIN number under my scenario.

It all rounds in one's expectation of the result of acquiescence to the interrogator's demands. In your scenario I assume the expectation is that once they get what they want the captive will eventually be released. A terrorist captive understands he's not going to be released simply because he cooperates with his torturers. In fact the act of torture more likely instills the thought that the captors will continue until he is dead, that the end result of his capture will be death sooner or later.

It also depends on the harm done by devulging the demanded information. In your scenario the loss connected to giving out your PIN is a loss of funds. Your hypithetical captive is not being asked to condemn his compatriots to death or capture. Bit of a difference there. Your hypothetical captive is not being asked to ignore his entire raison d'etre, and world view.

 

[jaydeehess]

I agree that it would almost certainly do more harm than good if used as a general screening device. In the case of a tough nut like KSM, who was thought to possess extremely useful information and was uncooperative, it was arguably justified.



Well, still, there's a world of difference between a young, uneducated soldier and a terrorist mastermind like KSM. Not even counting the difference in moral culpability, which is relevant only so far as a moral analysis is concerned.

I see now that I am a page or two behind the present conversation but I have limited time today.
Not sure if anyone pointed out that in KSM's case, it didn't work. He lied, but Cheney counts that as a win for some odd reason. Yes he gave information that was acted upon but a wild goose chase is hardly in the 'it worked' category even if it did result in a work project for the FBI.

 

[ehcks]

It's simple really.
If you wanted to keep a name secret but torture were applied, you'd say the name. Lots of other stuff as well, but you'd talk.
The notion that torture never works is ridiculous. An interrogator has to sift truth from BS anyways regardless of the methods used.

This argument continues to read like "Because a broken clock is right twice a day, we should keep using it."

 

[Upchurch]

Of course.
There are a lot more ticking bombs that don't go off than the few we are allowed to know of.

{snip}

Just to clarify, the time bomb is only part of the "ticking time bomb scenario". The two other critical elements is having a suspect who has the critical information necessary to either find or defuse the bomb and enough time to act on the information once you have it. This is, of course, assuming that you have the correct person and can get the accurate information and act upon it within the allotted time.

There is a lot of unrelated and unlikely factors that have to work together just right to obtain a true "ticking time bomb scenario" in real life. That's why I call it unrealistic. TVs and movies are the only place, under the eye of a script writer and director, where all these elements can come together correctly.

 

[theprestige]

Ah, I see your point.

You had my curiosity.

Gee, it's almost as if we're having a conversation, here.

But now you have my attention.

 

[Ziggurat]

This argument continues to read like "Because a broken clock is right twice a day, we should keep using it."

Who here has argued that we should use torture?

 

[Upchurch]

Who here has argued that we should use torture?

Monketi Ghost and sunmaster14 (there are many other versions of his pro-torture argument in other threads)