Sandeep and I are writing a paper on torture. We are trying to understand the mechanics and effectiveness of torture viewed purely as a mechanism for extracting information from the unwilling. A major theme we are finding is that torture is complicated by numerous commitment problems. We have blogged about these before. Here is Sandeep’s first post on torture which got this whole project started.
A big problem is that torture takes time and when the victim has resisted repeated torture it becomes more and more likely that he actually has no information to give. At this point the torturer has a hard time credibly commiting to continue the torture because in all likelihood he is torturing an innocent victim. This feeds back into the early stages of the torture because it increases the temptation for the truly informed victim to resist torture and pretend to be uninformed.
In light of this it is possible to say something about the benefits of adopting more and more severe forms of torture, waterboarding say. A naive presumption is that a technology which delivers suffering at a faster pace would circumvent the problem because it makes it harder to resist temptation for long enough.
But this logic is backwards. Indeed, if it were true that more severe torture induced the informed to reveal their information early, then this would only hasten the time at which the torture ceases because the torturer becomes convinced that his heretofore silent victim is in fact innocent. So credible torture requires that those who resist the now more severe torture must find compensation in the form of less information revealed in the future. In the end the informed victim is no worse off and this means that the torturer is no better off.
Once you account for that what you are left with is that there is more suffering inflicted on the uninformed who has no alternative but to resist. And this only makes it more difficult to continue torturing once the victim has demonstrated he is innocent. That is, the original commitment problem is only made worse.
Cheap Talk 
25 comments
Comments feed for this article
January 20, 2010 at 9:26 am
Alicia
Lets imagine a situation Torturer T has Victom V.
V has the following potential things he could be knowledgeable k or ignorant i. He could in either state say something s or stay quiet q.
From V perspective if tortured
as torture increases in intensity/duration
if k s increase in likelyhood
if i s also increases in likelyhood.
the rates of increase however are an issue because for T ks is the only desirable outcome. kq,iq,is are all negative outcomes for T with is ( false info being worst).
It becomes a question if ks increases at a rate such that the negative effect of IS is mitigated.
January 20, 2010 at 1:13 pm
Marc
s could be true (t) or false (f)
kst is the only desirable outcome
Then there is
ksf
kq
iq
isf
ist (he says he knows nothing)
January 20, 2010 at 12:26 pm
Rajiv Sethi
Jeff, if you’re working on this seriously I think it’s important to consider certain psychological factors: for example, the greater the degree of torture that has been inflicted, the less willing the torturer will be to face the realization that the victim is innocent. The same thing applies in death penalty cases – it’s very hard to accept that a mistake has been made if it can’t be corrected; the psychologically easier path is to convince oneself that the victim was guilty after all.
January 20, 2010 at 12:51 pm
Ozornik
Substitute “torture” with “interrogation” – no, even better: substitute it with “questioning”, and see for yourself that the logical template doesn’t change much.
Ergo: Questioning of terrorists is ineffective; we should do away with it.
P.S. Next step: substitute “terrorist” with “crime suspect”.
January 20, 2010 at 2:58 pm
IndyReader
@ Ozornik
If you can’t see the problem with your analogy here, then you need to stop discussing this issue.
January 21, 2010 at 12:50 am
I agree
wow. IndyReader is correct. It is scary what people try to pass as logic.
January 20, 2010 at 12:59 pm
98D
If the problem were merely one of the infliction of physical duress, then it might be that simple. But when that duress, combined with skillful application of advanced interrogation technique, induces increasingly psychologically regressed states – then the capacity to rationally resist the same amount of duress in the future will diminish.
An interrogator wouldn’t merely measure “duration of exposure multiplied by severity of pain without revelation of additional intelligence” as a proxy for “likelihood the interviewee has no additional information”, instead he or she would continue to apply pressure and look for indicators of deeper and deeper states of regression. If they can avoid being fooled, then once regression is confirmed and the capability of recalcitrance is at a minimum, the possession of additional information can be determined in an expedient manner and the session ended.
A knowledgeable interviewee exposed to such a process would know, in advance, that the interrogator will not cease until satisfied with confirmation resulting from the successful achievement of a regressed state. They would also know that such satisfaction will arise when enough duress has been applied for a long enough period to weaken the rational capacity to resist.
Faced with such an inevitability, it would make no difference whether one was knowledgeable or not, or spilled the beans earlier or later. If no credible trade of a cessation in pain exchanged for a limited amount of information could be arranged, then an interrogator would simply apply enhanced duress to everyone suspected of possession critical information until satisfied nothing more can be extracted.
January 21, 2010 at 10:43 pm
Padraig
But there is a fly in your ointment.
Forcing individuals to regress in order to weaken the will to resist is a little like prying a splinter out with dynamite – sure, you’re gonna get the splinter out, but you’re gonna do a lot more than just that. Similarly, inducing regression through repeated trauma has the effect of weakening the captive’s ability to recall, express, and even to keep the truth separate from fiction. In other words, you don’t just crack open a captive’s skull in order to take a look inside – you have to get them to tell you the truth, in a situation where you can make no credible promise not to torture them if they play ball. If torture’s gonna work, it has to work rationally: otherwise we’re just lost in the mystic quandary of how pain + person = truth.
January 20, 2010 at 1:27 pm
Marian
At this point the torturer has a hard time credibly commiting to continue the torture because in all likelihood he is torturing an innocent victim.
People who volunteer for performing torture on someone are probably sadists and will continue just for fun.
That is how it worked in every historical case I know of. Including my country, Czechoslovakia, in the Stalinist times.
January 20, 2010 at 1:35 pm
Andy
The only advantage possible gained by torture then is through information asymmetry. Your post assumes the tortured party knows in advance what will be done, and thus can wait out the torture and seem ignorant.
The most effective torture methods, it seems, would be a great and imposing threat with little need for follow through. The greater the threat, the more information the victim will divulge before discovering the improbability of the threat being carried out.
Any success waterboarding had before (the threat or feeling of drowning) is now gone (anyone who can read or listen knows it is a method of torture and not a method of execution)
January 20, 2010 at 2:06 pm
Greg
The underlying logic of your argument seems to be that the torturer wants to be fair, and thus will stop torturing a victim whom he believes to be innocent. But doesn’t he win the torture game if he commits to continue torturing until he gets verifiable, truthful information? If the victim has no information, too bad for him, but if he does have information, this commitment takes away his incentive to withhold it in hopes that the torture will stop before he reveals the information.
This discussion reminds me of something that Solzhenitsyn (I think) said in one of his books: that the Soviets under Stalin tortured until the victim confessed, but the Germans under Hitler tortured until the victim told the truth. The Germans, in any case, seem to have quite successful in extracting truthful information from their victims.
January 20, 2010 at 3:03 pm
IndyReader
And you would be able to confirm the difference (your Stalin/Hitler diff) how?
I love how you just justified the use of torture by now including the US in this amazing group of failed states, which we went to war against based on principle; ironic, no?.
January 20, 2010 at 2:12 pm
steve
The Germans got most of their info from snitches and, earlier in the war, from skilled interrogators who did not torture.
Steve
January 20, 2010 at 2:17 pm
Doc Merlin
In terms of actual usefulness, Greg is correct. Torture for confessions or political reasons are unlikely to yield fruitful information. Also, torture doesn’t start with torture.
From the literature, this seems to be the way it goes (this is very approximate):
it starts usually with mild questioning, and then they are prolonged over the course of multiple days.
The questioning turns into threats.
A sympathetic individual is introduced to the prisoner to gain their trust.
The questioning turns into mild physical abuse (sleep dep and slapping the torso).
Prisoner given a chance to “boast” about their exploits.
The physical abuse and sleep dep are extended and accompanied by loud music, uncomfortable positions, and cold, to raise stress levels.
Prisoner given another chance to “boast” about their exploits and implicate others.
Lastly, water-boarding is used.
Most break way before water boarding, and give verifiable information much before that point.
Ozornik also has a point; your analysis would also apply to questioning by police. Yet from the data we know that many people confess to police.
January 21, 2010 at 10:52 pm
Padraig
And we also know that people confess falsely to the police, even in the absence of torture. This was precisely the problem with judicial torture – an overemphasis on confessions instead of empirical evidence. It might be that true confessions are more likely than false confessions, but false confessions have real negative implications for informational torture – that is, it takes time and resources away from real leads, and can lead to a spiraling chain of denunciations and torture. Look at the Salem Witchtrials – the chain of accusations ended up involving the Royal Governor’s wife before it ended.
January 21, 2010 at 4:35 am
Filip Spagnoli
If you’re interested in the time aspects of torture, there’s also the wild goose chase element.
January 21, 2010 at 9:39 am
Anderson
People who volunteer for performing torture on someone are probably sadists and will continue just for fun.
Well, there’s a self-selection effect; normal people become nauseated and opt out, so that in addition to the higher % of sadists who volunteer, they’re also the only ones left after a while.
Re: “Doc,” many people do confess to police, and many of those confessions are false. And I’m not sure what “literature” Doc is referring to for the efficacy of torture. 24?
January 21, 2010 at 12:44 pm
dave
This article presumes that the point of torture is to extract meaningful information from the subject. However, a study of a number of historic situations in which regimes routinely employed torture reveals, that in many cases, a ‘confession’ (the explicit goal) had little do with one’s actual guilt. The subject who resisted, and professed innocence or ignorance merely ensured a continuation of torture: the subject who confessed immediately was guilty, and therefore deserving of torture as punishment. Torture serves a purpose beyond the desire to interrogate the subject: it serves to terrorize a greater number by example, to generate a fear sufficient to extract information/confessions from those beyond the subject, based solely on the fear that it could happen to them. The return of a single broken, beaten, terrified man can no doubt suppress entire dissenting groups: this, i think, is the point of torture, this evident fear that extends beyond the subject, who is made a living example.
January 21, 2010 at 6:15 pm
Top Posts — WordPress.com
[…] The Limits of “Enhanced Interrogation Techniques” Sandeep and I are writing a paper on torture. We are trying to understand the mechanics and effectiveness of torture […] […]
January 22, 2010 at 9:45 am
Tracy W
At this point the torturer has a hard time credibly commiting to continue the torture because in all likelihood he is torturing an innocent victim.
Ummm, is this a big problem? How many cases are there in human history where a person has started torturing someone, but their sense of ethics prevented them from continuing at some point? I thought the evidence was fairly clear that the initial decision to torture breaks the moral boundary against hurting another person, and once that’s been broken, the torturer has little problem continuing, and the innocence or guilt of the victim is irrelevant.
Now okay, most cases of torture I have read have been accounts by the victims of torture, or people outraged on the victim’s behalf. But the Milgram experiment results don’t seem to support the idea that the torturer has a hard time credibly committing to continue the torture – and in those cases the torturer “knew” the victim was innocent from the start. (The “victim” was an actor, the set up was that the experimental subject thought they were there to study the effects of pain on learning, the real set up was to see how much pain the subject would inflict faced with an authority figure.)
Basically, this premise here is inplausible.
January 22, 2010 at 12:17 pm
Morgan Warstler
The entire calculus changes if you simply add Lie Detectors to the equation.
They generally work… and they are becoming much harder to beat. Include in micro-expressions as well as a double fail safe to remove any doubt.
Question the terrorist while hooked up to a lie detector and IF he fails AND IF he fails micro-expressions WATERBOARD him.
If his story changes under “torture,” sit him back down and test his new story.
—–
This pretty much solves psycho-game-theory problems.
If the interrogator sees the terrorist pass the lie detection and the micro expression, the case for waterboard is decreased.
—-
It it pretty obvious we are moving this way, soon we’ll sit the poor sap down put a brainwave scanner on him, hook him up to a lie detector, point the machine visions camera at him and ask him questions and when he lies he’ll be given some horrible electric shock, waterboarded, etc.
February 15, 2010 at 12:13 am
52 Taco Nights » Blog Archive » #08 – New Brighton, MN – The Burns’s’s’s
[…] one reason or another Shannon Burns has interrogated me for the past 5 years. I am not sure why she is collecting the information nor do I know what she does with the answers. […]
August 3, 2013 at 5:33 am
www.the-baker-chick.com
Hello to all, it’s truly a pleasant for me to go to see this web site, it consists of useful Information.
August 4, 2013 at 12:42 pm
http://strongbrands.wordpress.com/2013/04/01/branding-kevin-ware/
Does your website have a contact page? I’m having trouble locating it but, I’d like to send you an email.
I’ve got some suggestions for your blog you might be interested in hearing. Either way, great site and I look forward to seeing it develop over time.
September 4, 2014 at 9:25 am
flooring west linn
Hello my family member! I want to say that this post is amazing, nice written and come with almost all vital infos.
I would like to look more posts like this .