Monday, September 10, 2007

Open letter to an Interrogater

I've been having an argument with an interrogator who turns out to be really lousy at making his case. In general, I agree with him about torture. Where he loses me is his insistence that torture never works.

I finally just wrote an open letter to him.


Hi there!

This is the Evil One, with an open letter for (or at least about) you.

Since you’ve declared me evil, and declared that I “support torture”, I really don’t expect this letter prompt you to attempt a civilized discussion, or to make any impact on you at all. I believe that you, with your declaration, have set about to define yourself as holding the moral high ground, and absolve yourself from mounting a reasoned defense of your case. I don’t blame you – thinking is hard work, and it takes practice.

This letter is really for the spectators. I intend to lay out my case for the dark side and let readers judge for themselves. On the off chance that you care to reply, I’m leaving comments open.

For my part, I don’t think you’re evil, merely foolish. I believe you when you say you hate torture, but I believe the arguments you have been making hurt your case and make it more likely that people will be tortured in the future. As good as your motives are, you lack wisdom.

First, let it be noted that for the most part, I agree with what you write. I merely believe that you carry your argument to extremes, and this, as I’ve mentioned, hurts the very cause you are fighting to promote.

I have two major objections to the arguments you raise against torture: silly definitions and your “utilitarian argument”.

1) Silly definitions

First, you define torture in a manner that is silly, useless, and which actually demeans real victims of torture. In your comment here, you define torture as “Any physical or mental coercion. Any.”

By my reading, this definition includes a huge number of acts which I don’t consider torture. A parent who puts his child on a “time-out” is coercing that child to observe the “time-out”, otherwise the child would not observe it. A police officer who pulls over a motorist for a traffic infraction is coercing that motorist – he doesn’t pull over because of his love of stopping and chatting with random police officers. An interrogator questioning a detainee is forcing that detainee to stay in the room with him, or at least at the detention facility. I doubt that, if given a free, un-coerced choice, that detainee would remain at the detention center.

Indeed, I can’t imagine any society that can function without some form of coercion to enforce whatever laws it deems necessary. Indeed, I don’t see how anyone can interrogate a person unless he can coerce that person to hang around long enough to be interrogated. So when you offer a definition of “torture” that forces me to choose between any sort of working civil order and a Hobbesian anarchy, I guess I have to stand on the side of “torture”.

You have introduced, at least by implication, another definition of “torture”. On several occasions, you have, in effect, challenged people with: “how would you like it?"

If you are willing to have it done to yourself, or your loved ones, if they are suspected/accused, then, just maybe, you have a morally defensible (though disgusting) position.

link

The implied definition is, “torture” is anything you would rather not experience.

Unfortunately, while everything we both agree is torture is covered by that definition, so are any number of things we disagree over. Furthermore, so are any number of things I’m sure neither one of us would call torture. For example, in the army, you were trained to kill people. In battle, trained soldiers go into the field with the full expectation they will kill other soldiers, and with the full awareness that they may themselves be killed.

In my opinion, your definition of “torture” clouds the issue. When the same word applies equally to the tearing off of fingernails, the gouging out of eyes, and placing a child on a “time-out”, the word loses most of its meaning and all of its moral and emotional impact. When “torture” means anything as innocuous and routine as being pulled over by a traffic officer, a reasonable person may conclude “torture” isn’t that bad.

Your definition of “torture” also insults everyone who has been subject to real torture. By lumping the treatment Senator McCain received at the Hanoi Hilton together with routine treatment of prisoners here in the States, you cheapen McCain’s experience.

Now, it’s entirely possible that I’ve completely misinterpreted your definition of “torture”, and you do not intend to include every form of coercion. However, when I’ve given my take on your definition in the past, you have declined to either clarify or withdraw it.

I assume you stand by it, and what I infer from it. If you don’t, comments are open.

I don’t have a cut-and-dried definition of “torture”.

I note the United States Code defines “torture” in the following way:

Section 2340. Definitions

As used in this chapter -

  1. "torture" means an act committed by a person acting under the color of law specifically intended to inflict severe physical or mental pain or suffering (other than pain or suffering incidental to lawful sanctions) upon another person within his custody or physical control;
  2. "severe mental pain or suffering" means the prolonged mental harm caused by or resulting from
    1. the intentional infliction or threatened infliction of severe physical pain or suffering;
    2. the administration or application, or threatened administration or application, of mind-altering substances or other procedures calculated to disrupt profoundly the senses or the personality;
    3. the threat of imminent death; or
    4. the threat that another person will imminently be subjected to death, severe physical pain or suffering, or the administration or application of mind-altering substances or other procedures calculated to disrupt profoundly the senses or personality;

Now this definition suffers from a problem you were trying to avoid with your definition: What do we mean by “severe”? While we’re at it, how long is “prolonged”, how soon is “imminent”, and how profoundly is “profoundly”? Any reasonable definition, used by reasonable people in the real world is ultimately going to be a process of line-drawing. You have chosen to draw your line such that “severe” means “any”, “prolonged” is “any”, “profoundly” is “any”, and “imminently” is “at any time in the foreseeable future”. I don’t think this is a useful definition.

Now, in the real world, there are a variety of techniques that have been approved for use by interrogators, including the notorious “water-boarding”. So far, red-hot pincers, bamboo slivers under nails, and bastinado are not on the “approved” list. They’re on the “yes, it’s torture” side of the line, while the “attention slap”, extreme temperatures, prolonged standing, and water-boarding are on the “no it’s not” side.

My take on the issue: I’m not sure where I draw the line between “torture” and “not torture”. Indeed, any line I draw would probably be a fuzzy one, with some things being “mostly torture” and others “only a little bit torture”. Indeed, since people differ in their sensitivity and reaction to various things, I’d probably wind up having to draw a different line for every person on the planet.

Ultimately, I believe reasonable people can disagree over where to draw the line. You apparently don’t.

2) Your “Utilitarian Argument”

You have repeatedly advanced a “utilitarian argument” against torture. [Add links] You repeatedly assert that “torture doesn’t work”. This argument fails on two counts. Firstly, utilitarian arguments are very dangerous. They’re only as good as your data.

The book “Freakonomics” made the case that liberalized abortion laws were responsible for the drop in crime rates in recent years. The children aborted are the ones that would have been more likely to grow up poor and in single-parent homes. These are the very groups that are more likely to commit crimes.

Now, John Lott writes in his book, “Freedomnomics”, argues that abortion has actually caused an increase in crime.

What is the utilitarian argument with respect to abortion? I don’t know. The final results aren’t in yet. They may never be.

The point remains, if you base your argument on a moral point on utilitarian grounds, then your morality is only as solid as your data, and you may find your utilitarian argument urges the adoption of a policy you detest, and having raised the utilitarian argument in the first place, you have no grounds to object.

Secondly, your utilitarian argument is wrong – torture does work, especially if we allow your silly definition to stand.

Interrogation can’t possibly work unless the person being interrogated is required (that is, “coerced”) to remain where the interrogator can question him. As you define “torture”, interrogation is impossible without torture.

You have taken me to task for “arguing in bad faith” because “it’s impossible to prove a negative”. Fine, I take you at your word. Your “utilitarian argument” is not provable.

Furthermore, it’s very easy to disprove. All I have to do is cite one case to show that “torture” does, at least sometimes work. And since you define “torture” in such a way that even the methods of interrogation you approve of are impossible without it, we now have absolute proof that “torture” does, in fact, work.

However, let’s now bring in reasonable people with real-world definitions of “torture”, where there is a line drawn between “torture” and “rough treatment”. (And possibly, a line between “rough treatment” and “not being nice”.) We still have cases where you concede that rough treatment, even that which crosses the line into torture, does work. It obtains information which is true.

Why? Because as I keep saying, using torture as a means of collecting information doesn't work. As a system, it fails. Someone might tell the truth, but the amount of non-truth which enters the system buries it.

There are those who pretend that's not the case. That somehow we can sift the truth from the lies; without having any troubles. That somehow the dedicated bad person, who is willing to plant bombs, bury people alive, whatever fantasy of justification the torture mongers want to trot out, will somehow break when his body is beaten, his flesh is torn, his mind is assaulted with terrors, the electrodes are supplied with current, the water rises past his nose and mouth, his bones broken, his sleep deprived, his environment changed, etc., etc., etc., ad naseam.

link

Sooner or later my honest report (assuming I break) will be lost in all the crap (all the more so if there is more than one person being tortured, the interrogators will start to manufacture corroboration; and when the story changes, so too will the false corroboration change to match it, because the answers are expected, and the source will be guided to them).

link

It obtains information which is true, “but”, you say, this information is contaminated. Under torture, you point out, a person will say anything he thinks his interrogators want to hear, in order to stop the torture. How, you ask, do we differentiate between accurate information and inaccurate?

In other words, we have three issues to consider: The effect of incentives on human behavior, the signal to noise ratio of any information obtained, and the use to which any information obtained is to be put.

First, the use. You are right in saying the use of torture would contaminate all information obtained from a subject, if the information were being used to build a legal case. The thing to remember is that the war on Jihadists is not a law enforcement action. It is a war. During war, soldiers are called on to deal with the enemy in ways police officers would never be allowed to deal with criminals. No soldier, for example, reads a captured enemy his or her Miranda rights, for example.

Information obtained through sufficiently rough treatment should, and probably would, be excluded from any criminal proceeding. However, if, somehow, rough treatment caused someone to divulge the location of a “ticking time-bomb”, this information would be used to find and disarm the bomb, or at least evacuate the area. No one would use it to arrest the bomb. (More to the point, this information would probably not be allowed in court as evidence the subject of the interrogation was involved in a conspiracy to plant and set off that bomb.)

Second: Torture is designed to create an incentive. In order to prevent or avoid torture, the subject will, in theory, change his behavior. If the change in behavior means giving up useful information, then that’s what the interrogators want. However, a problem arises when the subject doesn’t have useful information, or has a strong desire to withhold it and/or give up false information. In the face of a strong counter-incentive, we have to realize that the information we get may be false. Indeed, I recently posted a blurb about false information that had been obtained, not from torture, but in order to obtain a reward. From this, we may conclude that non-torture doesn’t work.

A point you continually gloss over is that every subject of any method of interrogation must be presumed to have a strong incentive to withhold information and to mislead interrogators. When I say you gloss over it, you argue as if information obtained from what you call torture is never checked for accuracy or consistency with information obtained from other sources. In so doing, you imply that you have access to techniques which yield zero false information. In other words, you imply your preferred techniques yield 100% signal and 0% noise.

If this is, in fact, your claim, then I don’t believe you. In fact, I have such strong doubts that any technique exists which will yield 100% signal and 0% noise that I believe anyone claiming one does is either incredibly stupid or lying.

I will give you the benefit of the doubt and assume you admit to non-zero false information rates in all known techniques of interrogation.

Therefore, using the same line of argument you use to show “torture doesn’t work”, I could “prove” that your favorite techniques don’t work. All I would have to do is highlight every piece of false information you uncover, ignore the true information, and disregard any steps you take to differentiate between the two.

Presto! Interrogation doesn’t work, and we can reassign all interrogators to the motor pool.

Now, if you were willing to discuss this topic in a reasoned fashion, we could discuss such topics as which methods of interrogation are better. Reasonable people could try to uncover which methods have what signal to noise ratios. Adults in a reasoned discussion could address questions like the following:

  • Do your preferred methods have a materially higher ratio than the methods you despise?
  • Do your preferred methods yield information as quickly as the methods you despise? (Sometimes information may be needed in a hurry – e.g., the “ticking time-bomb” scenario.)
  • Do different methods of interrogation yield different sets of accurate information? (People willing to divulge certain secrets under one technique might, in theory, only divulge others under a different method.)
  • What moral price are we willing to pay for any given type of information?

And that last one is the question that you really should be asking. Your silly definition of “torture”, and you “utilitarian argument” are nothing more than ways to try to avoid dealing with the fact that sometimes we do have to choose between two bad scenarios. There isn’t always a perfectly good solution, and adults have to come to grips with that fact.

1 comment:

Squiptryx said...

I'm a friend of Karl's, with generally opposing political views, and I teach logic, so I thought I'd comment on the logic of his post. (Oh, and I'm also Evil, but in a good way.)

1. Point One: Coercion

Here I agree with Karl. It makes no sense to define torture as any form of coercion. First, as Karl says, there are undoubtable instances of coercion that cannot reasonably be called torture. Duh. Furthermore, there are instances of torture where ith only coercion involved is keeping the victim in place to be tortured. There have been plenty of instances of torture for sick personal reasons of the torturer.

2. How would you like it? I wouldn't like it, but I would agree to a certain set of social rules that would allow me to be tortured under certain circumstances. There have been instances of parents who became mentally deranged and became dangerous to their children. Suppose I became deranged and locked up my own daughter with no food and water, and limited air, in a location only I knew. My present, sane, self ernestly desires that such a future, deranged, self be tortured if that this the only way to guarantee that my daughter be saved. And if my daughter died because this future deragned self wasn't tortured, I doubt I would be able to live with myself.

3. I agree with Karl that the "official" definition he quoted has a line-drawing problem, but I think it's a very reasonable starting point. I might draw the line differently however. To me, anything with the potential to produce significal permanant mental or physical impairment is definitely torture, while momentary, non-damaging pain isn't. (I've been married twice. It's amazing how much pain you can recover from. Eventually.) This definition might, for instance, imply that waterboarding isn't torture.) However, we might want to include waterboarding as torture for purposes of discussion.

4. Utilitarianism clearly justifes torture uder some definitions. If we can reasonably imagine cirumstances where the (rare, and strictly limited) inflicting of suffering by security services prevents more suffering than it causes, then that infliction is justified, whether we call it torture or not.

4. You can prove a negative. It's just silly to say you can't. It's easy to prove there's no square circles, and only a little harder to prove there's no griffins or hippogriffs (darn!). Or do you think that there's a reasonable possibility that unicorns and pookahs exist?

Gotta go shopping - Ciao Bella!