« Tim Harford's Dear Economist | Main | The Conquest of Nature »

Torture and neuroeconomics

Studying some neuroeconomics has made me even more opposed to torture than I was in the first place.  Yes I will make an exception for the ticking nuclear time bomb.  But if we are torturing a very very bad person, I don't see the torture as satisfying justice.  The part of the brain which suffers is not the same as the part of the brain which planned the crime.  Yes neuroeconomic data are hard to interpret.  But under one view, there is a sheer production of pain which is severed, to some extent, from the individual personality of the criminal.  It is almost as if we are creating a new suffering entity which consists of little more than pure suffering.

I don't think this is the most important argument against torture, but it is one additional consideration.  Retributive justice does not weigh on the pro-torture side of the scale as much as one might think.

Matt Yglesias surveys what the recent "torture compromise" really means.

Posted by Tyler Cowen on September 22, 2006 at 05:30 PM in Law | Permalink

Comments

Is torture even looked upon as a form of punishment here in the "western world"? The only pro-torture argument I've heard in a good while is that of it being an effective interrogation method when all else fails.

Also, do you have any introductory books (or books at all, for that matter) on neuroeconomics, to recommend? The selection at Amazon appears somewhat scarce, but I'd like a good introduction before hitting the academic papers.

Posted by: Robert at Sep 22, 2006 5:25:41 PM

Honestly, I am surprised and, yes, shocked that you seem to take it for granted for torture to serve as a punishment. There was something like the 8. Amendment in your country's Constitution, correct? But this document seems to be largely forgotten nowadays anyway.

The only forgiveable exception from the absolute prohibition of torture might be to avert a clear & present danger. Even in those cases, the effectiveness of torture is arguable. At any rate, there cannot be a procedural way to allow this within a State of Law - it must be on Presidential prerogative, where the responsibility rests.

Posted by: Chris at Sep 22, 2006 5:43:51 PM

Talk about missing the point.

No one in the West talks about torture as retribution or punishment. It's job is to get the individual being tortured to spill the beans.

Duh.

Posted by: Honest Bob at Sep 22, 2006 5:47:03 PM

People don't talk about torture as punishment but they assuredly think about it.

Let's put it this way. The President likes to talk about "coercive interrogation tactics (waterboarding, obviously)" being used against Khalid Sheik Mohammed. First, I bet what is going through the minds of the President's supporters is, "the guy planned 9/11. Who cares what happens to him?" Second, if you think I am wrong then consider a scenario where KSM is held by the CIA and then his hypothetical 8-year-old daughter is brought into the interrogation room with him and is waterboarded while he watches. Are you as enthusiastic about torture now, even if seeing his own daughter tortured forces KSM to reveal a planned terror plot against the U.S.?

One of the things that makes the torture of terrorists easy to contemplate is the fact that they are loathsome individuals who (one might think) deserve no better. If the goal of torture is solely to extract information and if this is considered morally permissible, then it shouldn't matter if the person being tortured is completely innocent of anything. I can imagine torturing a loved one being even more effective for some people than torturing the person with direct knowledge.

I don't think anyone (besides a couple of yahoos) thinks of torture exclusively as punishment. But to suggest that isn't a consideration for most people is unrealistic.

Posted by: Mark at Sep 22, 2006 6:13:04 PM

I'm wondering why Tyler wants to turn subunits of the brain into persons with moral standing. Why, Tyler, why?

Posted by: Rhadamanthus at Sep 22, 2006 6:47:16 PM

Is this an argument against rewards, too? Does pleasure stimulate a different
area of the brain?

Posted by: Jthaddeus at Sep 22, 2006 6:47:35 PM

That "torture" is not "punishment" is a matter of definition, but their distinction has nothing to do with whether they are unpleasant to the recipient because punishment is supposed to be unpleasant just as torture is and therefore they share the dimension of unpleasantness. Torture is however by definition different from punishment because torture is to coerce an action in the future and punishment is to punish an action in the past.

An example: when the American boy was caned in Singapore, that was not torture even though it was extremely painful for the boy, because it was to punish an action in the past rather than to coerce an action in the future. In contrast several of the activities which Bush is being criticized for would be mild as punishments, and are criticized not because they are unconscionably painful but because their purpose is to coerce an action (spilling the beans).

Posted by: Constant at Sep 22, 2006 6:59:44 PM

Robert is correct.

You are comparing apples and oranges. No one is talking about torturing terrorists merely out of spite. Almost everyone rejects that. However, when you cannot get a terrorist detainee to talk, using the rough treatment often proves an effective way to get that part of the brain that wants to tell secrets to do so.

Posted by: mike at Sep 22, 2006 7:08:56 PM

People do, in fact, talk about torture as a punishment. Very well regarded legal minds do. (He did later repudiate the idea, but not on moral grounds.)

If someone of Eugene's obvious brilliance and training thinks like this, that leads me to believe the same impulse that leads to prison rape jokes happen in the heads of many, many other people.

Posted by: fishbane at Sep 22, 2006 7:28:30 PM

People who talk about torture as a kind of punishment need to explain why all punishment is not torture. After all, all punishment is intended to be unpleasant to the person receiving it. If the difference between torture and punishment is not in their respective goals, then what is the difference?

Posted by: Constant at Sep 22, 2006 7:36:59 PM

There is a very important distinction between torture (which, most commenters identify as inflicting pain to extract information) and cruel and unusual punishment (which is primarily retribution). I have an article on the economics of the latter, which can be accessed here for those who are interested.

Posted by: EclectEcon at Sep 22, 2006 8:24:48 PM

Constant,

I would define torture as the deliberate infliction of extreme pain or suffering on another person. Torture certainly can be punishment: look at how England punished Guy Fawkes, for instance. All punishment is unpleasant but legitimate forms of punishment do not deliberately seek to induce extreme pain in another person.

Pain may be incidental to punishment (prison mattresses may cause extremely painful back aches in some people) but it may not be the end goal of punishment. Prison authorities cannot stick spikes in mattresses belonging to people who do not have back aches solely to ensure they cross a certain extreme pain threshold.

Posted by: Mark at Sep 22, 2006 8:40:43 PM

Whether or not people talk about this, IMHO, torture as punishment is directly relevant to what went on at Abu Graib, where torture degenerated from an interrogation tactic to punishment to pure sadism. I agree with Tyler that it creates an experience of pure suffering disassociated from the personality or actions of the victim, although my perspective is more Christian than neurological.

Posted by: DK at Sep 22, 2006 8:42:28 PM

The argument that torture is completely different from punishement because torture is a means to extract information fails because torture.is.not.a.useful.method.to.extract.information. Is not, has never been, and will never bee. It is a great method to get confessions, and to get people to say whatever they think you want to hear. And that renders it completely useless as a method to receive actually truthful information. Signal-to-noise ratio and all that jazz.

The Soviets had figured this out 75 years ago and used torture accordingly, but somehow the tough guys on the right side of our political spectrum seem unable to comprehend it. I have rarely been as frustrated with this country as these days.

Posted by: Commenterlein at Sep 22, 2006 9:11:45 PM

I would rather be whipped than executed. Apparently, then I would rather be tortured than punished. It seems to follow, then, that torture is not necessarily worse than punishment, at least not from the point of view of the person being subjected to it.

I would also rather be caned than be imprisoned for one year. Again, punishment appears worse than torture.

I am of course adopting a definition of torture as being the deliberate infliction of extreme pain. Caning and whipping are torture under that definition, regardless of their function.

Posted by: Constant at Sep 22, 2006 11:23:33 PM

"All punishment is unpleasant but legitimate forms of punishment do not deliberately seek to induce extreme pain in another person."

Actually, historically, that is not true. Look up "corporal punishment."

According to Wikipedia (usual caveats, but I think the burden of proof is on those who dispute the following):

"Corporal punishment is the deliberate infliction of pain intended to correct behavior or to punish. Historically speaking, most punishments, whether in judicial, domestic, or educational settings, were corporal in basis."

Posted by: Constant at Sep 22, 2006 11:29:52 PM

Whether or not people talk about this, IMHO, torture as punishment is directly relevant to what went on at Abu Graib, where torture degenerated from an interrogation tactic to punishment to pure sadism.

Torture, in that circumstance, was not inflicted by the MPs to personally gather information. Supposedly, the MIs wanted them "softened up" for conventional interrogations and were left to their own imaginations as to what sort of pain and degradation to inflict. There is a good argument for not allowing such reckless and unstructured forms of abuse to occur. After all, you are applying torture, in that instance, before you have even begun to find out what you might accomplish using conventional interrogation and before you've even evaluated what a subject is likely to know. Such torture is not going to be merely functional, but sadistic also.

However, I don't think that makes a case against using torture to facilitate interrogation under the more controlled circumstances we've seen the CIA use it to get information of Khalid Shaikh Mohammed and company.

Posted by: mike at Sep 23, 2006 12:07:56 AM

The question is, Tyler, whether there is a possibility to "inflict the punishment" on a specific part of the brain. I have my doubts, because the planning unit is most likely one without much attentionn to emotions (it's the rational plane of the killer).

So, I think to split up the brain and act as if a part of the brain is person all by himself, doesn't get us very far (not even in the jurisdictional sense). Either we assume that those are characteristica of one person, which means that they are intra-dependent and thus if one fails, all get punishment, because they appear to the outside world as ONE.

This isn't very libertarian, but as long as we can still concern ourselves with the "Ego" of a person, as one and not as a fragmented hive in the brain, it is still coherent.

Posted by: Max at Sep 23, 2006 6:11:07 AM

All punishment is deliberate infliction of suffering. Folks who want to distinguish between punishment and torture but who do not want to talk about their distinct goals, are reduced to making claims about what is or is not overly severe suffering. Well, in answer to that, who's to say what is or is not overly severe? If one person says X is not too severe but Y is too severe, and someone else says Y is not too severe, by what principle can we decide who is right?

(Remember that all punishment is deliberate infliction of suffering - that's what makes it a deterrent; obviously if it did not make the subject suffer it would not be feared and it would not be a deterrent)

Here's a way to distinguish between moral and immoral infliction of suffering. Moral infliction of suffering does not violate the rights of the subject; immoral infliction of suffering does. What are those rights? Well, it's easier to describe the limits of those rights. But let's use examples: a person has no right to kill another person. Therefore the intended victim has the right to deliberately inflict suffering on the would-be murderer in order to stop the murderer. An example of this would be kicking him in the balls.

Do people have a right to kick the balls of would-be murderers who are trying to kill them? It is deliberate infliction of extreme suffering, so the folks who have been arguing that torture is wrong and who have been defining it as "the deliberate infliction of extreme pain or suffering on another person" must logically conclude that, no, the targets of would-be murderers do not have the right to kick their attackers in the balls. I, in contrast, would say they do.

Now here's the opposite example: a person has the right to remain silent. I'm sure people are familiar with the right to remain silent. Because of this right, it is wrong to coerce a person into speaking. Torturing someone to make him talk is an example of such coercion.

Posted by: Constant at Sep 23, 2006 6:14:33 AM

As soon as you start trying to seperate "the you that is punished/tortured" and "the you that commited a crime", you are headed off to nowhere. Am I not the same person I was yesterday because of all the cells that have died off and been replaced, or the different contents in my brain? Whether or not I am really the "same person" I feel that I am. I accept my actions in the past as being committed by me and I want to avoid being punished in the future. That makes inflicting pain on me following criminal acts an effective way of preventing me from performing such acts.

People often say "torture doesn't work", but I haven't seen any evidence for that. Apparently it worked for the French in Algeria (militarily, if not politically).

Posted by: TGGP at Sep 23, 2006 11:34:19 AM

I did not use the word "suffering" precisely because it is a very broad term and is susceptible to semantic games. Extreme physical pain, on the other hand, has a very uncontroversial definition.

Constant is either ignoring or misinterpreting the word "deliberate" which is an extremely important part of the definition of torture. The example of prisoners getting backaches illustrates this: their pain is not deliberate but rather incidental. As for a murderer threatening to kill you, kicking him in the balls is a legitimate strategy because the aim of that action is not the infliction of extreme pain, but rather the incapacitation of the murderer. If you could hit him in the side of the head and knock him unconscious without any pain, that might be a preferable strategy that achieves the same ultimate objective.

And of course I am quite aware of the history of corporal punishment as well as the history of its abolition in the Western world (and most of the non-Western world) and the reasons for that abolition. When the government becomes directly involved in the business of deliberate human cruelty, government tends to become dominated by uncontrollable sadists and sociopaths. The distinction between inflicting pain for reasons of supposed national security and inflicting pain for fun is extremely fragile.

Posted by: Mark at Sep 23, 2006 8:26:19 PM

Mark,

If you did not use the word "suffering" precisely because it is a very broad term and is susceptible to semantic games - if that is not the reason you used the word "suffering" - then why did you use the word "suffering"? Now, it may be that you claim you did not use the word "suffering", and you are explaining why you did not use the word "suffering". However, in fact, you used the word "suffering".

I do wish you would write less ambiguously. That way I would not have to waste time trying to figure out statements which are either incomprehensible or demonstrably false. It's a pain in the neck to have to spend extra time on things like this, especially when - as may be the case here - the other person won't even spend the time to look at his own previous statements.

You, or someone else named "Mark" with a similar email address, wrote:

"I would define torture as the deliberate infliction of extreme pain or suffering on another person."

Setting that aside...

"As for a murderer threatening to kill you, kicking him in the balls is a legitimate strategy because the aim of that action is not the infliction of extreme pain, but rather the incapacitation of the murderer. If you could hit him in the side of the head and knock him unconscious without any pain, that might be a preferable strategy that achieves the same ultimate objective."

The immediate purpose of kicking someone in the balls is indeed to inflict extreme pain. The pain is in turn a means to incapacitate him. The same is true of pepper spray and mace. One deliberately inflicts pain - as a means to a further end. In the case of torture, the pain is as a means to the further end of getting the subject to spill the beans. You want to excuse defensive kicking in the balls by pointing out that the ultimate goal is defense. Fine, but then the same argument can be used to excuse torture to extract information.

As for corporal punishment, that is another topic (though perhaps in your mind it is not). However, I'll mention that your criticism of corporal punishment can be extended to all punishment: the distinction between imprisoning someone for long stretches of time to deter activity and imprisoning someone for long stretches of time to achieve malicious satisfaction is a fragile one.

Finally, much of the controversial torture is intended to produce, not extreme pain, but shock, fear of death, and the like. The immediate goal of water-boarding is not to cause pain but to make the subject feel as though he is about to die. The immediate goal of the bizarre sexual humiliations in Abu Ghraib is not to produce pain but to produce profound humiliation. And so forth.

So in that dimension as well, your definition of torture is quite selective. If you are trying to defend the policies of this administration, your quite restrictive definition is the definition you might employ.


Posted by: Constant at Sep 23, 2006 8:51:55 PM

I didn't remember using the word suffering. My bad: consider it erased from my previous post. I don't consider my definition exhaustive; it is just that I am pressed for time and want to give a clear, albeit incomplete, definition that cannot be subject to endless semantic arguments. Moral philosophers get paid to come up with such definitions; I don't.

Blog comments are not the ideal medium for discussing such subtle issues. I would recommend anyone interested in the torture debate (which really should be anyone with a pulse) to read Thomas Nagle's essay "War and Massacre." It is by far the most cogent modern piece of writing on morality during wartime I have read.

I agree that corporal punishment is a different issue but you were the first to bring it into the discussion. The morality of various forms of criminal punishment is a difficult area but requires much more than a couple of paragraphs to do it justice.

Finally, I will note that what I mean by "deliberate" is that your actions and intentions track the level of pain in the victim. If you are in a situation where you have to defend yourself and you have the choice of incapacitating him using an extremely painful method versus incapacitating him using a painless method, the choice of the painful method is morally indefensible. If in practice there is no painless method available, then painful self-defense is morally justified because the defender's intentions do not track the pain of the victim. It is intentions that matter -- not goals -- in defining torture. Again, you should read Nagle on this point as I cannot completely flesh out the distinction in a couple of sentences.

Posted by: Mark at Sep 23, 2006 10:43:41 PM

One final thought I would like to add about the murderer example is that inducing extreme pain in order to stop the attack is fundamentally different from torture. The distinction is that in the case of self-defense, I am taking advantage of a simple biological fact: the human body does not behave the same way when it is in pain. I use pain as a means of incapacitating the murderer; if there were other means available, though, I would be obligated to use them first.

In the case of torturing someone to extract information, something far more sinister is going on. I am instead using pain to lower another person's inhibitions and, in fact, strip away his ability to be an independent moral agent. I am forcing the person to make choices he might not make under other circumstances, not merely inducing a purely physical reaction. I think this is another reason why we should consider the use of extreme pain for the purposes of coercion to be wrong. I realize this may be too Kantian for some tastes; I just throw it out as an observation.

Posted by: Mark at Sep 23, 2006 11:50:36 PM

To me, the solution seems relatively straightforward. We don't want to give the US government authority to torture anybody it chooses (or to rendition anybody it chooses to torturers) because that can go wrong pretty quickly. Basically, you can get innocents telling torturers what the torturer wants to hear, and that's detrimental to everything we're trying to accomplish. In addition, it's not necessarily all that effective with your super-ignorant tribal low-level guys, who probably have no idea how swell we are. Pleasantly surprising the rank-and-file might be a more effective interrogation method.

But if you have a guy like KSM, who clearly does know relevant stuff, then torturing him seems okay by me. The fact that he's really bad and helped kill 3,000 people probably contributes to my sanguine attitude on the matter.

So why don't we have a Congressionally authorized waterboarding list? The President submits names, and Congress votes on authorizing waterboarding for each name. That way, we keep the waterboarding to the hard-core upper-level guys who fit the following criteria:

a) we already know that they know stuff, so we're not risking having some poor innocent goat farmer making a bunch of false confession

b) they will not respond to nice methods.

Who's with me?

Posted by: Keith at Sep 24, 2006 4:24:18 PM

Post a comment