Torture in a Just World

If the world is just, only the guilty are tortured. So believers in a just world are more likely to think that the people who are tortured are guilty. Perhaps especially so if they experience the torture closely and so feel a greater need to overcome cognitive dissonance. On the other hand, those farther away from the experience of torture may feel less need to justify it and they may be more likely to identify the tortured as victims. The theory of moral typecasting suggests that victims are also more likely to be seen as innocents (a la Jesus).

The theory is tested in a lab setting by Gray and Wegner. Experimental subjects are told that “Carol”, really a confederate, may have lied about a dice roll and that stress often encourages people to admit guilt. Subjects then listen to a torture session as Carol’s hand is plunged into a bucket of ice water for 80s. Subjects are then asked how likely is it that the torture victim was lying (1 to 5 with 5 being extremely likely). There are two intervention variables: 1) some of the subjects meet the torture victim before she is tortured, this is the close condition and some do not (distance condition) and 2) in some torture sessions the victim evinces pain (pain) and in others not (no pain). The key figure is shown below:

torturegraphThe most striking result is that in the close condition, the evincing of pain was associated with an increased judgment of guilt, consistent with torture causing cognitive dissonance which is relieved by a judgment of guilt (restoring the just world). But in the distance condition, the evincing of pain was associated with a decreased judgement of guilt, consistent with pain increasing the identification of the tortured as a victim and therefore innocent (a la moral typecasting).

Closeness in the experiment was reasonably literal but may also be interpreted in terms of identification with the torturer. If the church is doing the torturing then the especially religious may be more likely to think the tortured are guilty. If the state is doing the torturing then the especially patriotic (close to their country) may be more likely to think that the tortured/killed/jailed/abused are guilty. That part is fairly obvious but note the second less obvious implication–the worse the victim is treated the more the religious/patriotic will believe the victim is guilty.

The theory has interesting lessons for entrepreneurs of social change. Suppose you want to change a policy such as prisoner abuse (e.g. Abu Ghraib) or no-knock police raids or the war on drugs or even tax policy. Convincing people that the abuse is grave may increase their belief that the victim is guilty. Instead, you want to do one of two things. Among the patriotic you may want to sell the problem as a minor problem that We Can Fix – making them feel good about both the we and the fixing. Or, you may want to create distance – The problem is bad and THEY are the cause. People in the North, for example, became more concerned about slavery once the US became us and them.

I think research in moral reasoning is important because understanding why good people do evil things is more important than understanding why evil people do evil things.


Interesting. If this theory were broadly applicable, I wonder what explains the great malaise we felt in response to Abu Ghraib, but the general acceptance (except among some strong liberals and libertarians) of Guantanamo Bay. Perhaps it's that the former was done out of spite, whereas the latter was believed to have some objective purpose in the "war on terror".

What really fascinated me is that I consider myself to be a fairly patriotic person, but it's particularly that which makes me more sensitive towards harsher America policies, i.e. because I identify so closely with my state (or rather, culture, if there is any one, such entity) I want it to do that which I perceive as good.

I would say the main thing difference is that in Abu Ghraib there was undeniable evidence of abuse. All of the claims of abuse at Guantanamo Bay have been disputed.

Also, one can oppose specific abuses while still favoring the wartime holding of enemy prisoners.

'All of the claims of abuse at Guantanamo Bay have been disputed.'

Well, not exactly all claims -

'A new lawsuit seeks to force the U.S. government to make public “extremely disturbing” videotapes of a Saudi national whose abuse at the Guantanamo Bay prison has been called “torture” by a former Bush administration official.


In their lawsuit filed Monday, Lawrence Lustberg and Sandra Babcock seek to shed light on the treatment of their client Mohammed al-Qahtani, who was captured in Afghanistan during the hunt for Osama bin Laden in 2001 and was whisked to Guantanamo Bay, where government investigators later identified him as a man who had planned to participate in the Sept. 11, 2001 terror attacks.

The case of Qahtani first came to light in 2005 when Time magazine published secret log files from Guantanamo that detailed harsh interrogation techniques on the Saudi suspect.

In February 2008, he was charged with war crimes and murder, but on May 11 of that same year those charges were dropped. The reasons at the time were not made public.

In 2009, a Bush administration official revealed the reason to Bob Woodward of the Washington Post:

"We tortured Qahtani," Susan J. Crawford said. "His treatment met the legal definition of torture. And that's why I did not refer the case" for prosecution.'

" the evincing of pain was associated with an increased judgment of guilt"

I'm reminded of Trial by ordeal. If Carol wasn't guilty, the cold water wouldn't hurt her.

Or, you may want to create distance – The problem is bad and THEY are the cause. People in the North, for example, became more concerned about slavery once the US became us and them.

But we already have that in the U.S. right? It's not North vs South but Democrats vs Republicans. When there's a Republican president, Democrats feel distant and object to torture. When a Democrat takes over then Democrats switch to justifying while Republicans start objecting (and libertarians, who are never in power and therefore never lack for the necessary distance, always object--but they're marginal and it doesn't much matter). Unfortunately, I don't see a lot of reason to hope that the whole country is going to start objecting to middle-of-the-night SWAT raids, rendition, drone strikes, etc -- at least not at the same time.

When a Democrat takes over then Democrats switch to justifying

The problem with this equivocation is that far from all Democrats and liberals justify the behavior of Democratic politicians. One cannot spend five minutes in the online progressivesphere without hearing an argument over drones.

I don't mean to bash Republicans with their relative fealty to conservative politicians who violate conservative ideals. But Haidt does this lazy equivocation in his book too, and it really detracts from the analysis.

Exceptions don't a political movement make. I'd suggest that the Republicans lost influence and voters over time due to the torture issue. Was it enough to swing power in 2006? Maybe a bit, but there were other issues as well. Similar with the Democrats.

If you are twisting a narrative to elicit support for a policy you just need enough support to get it done. There are always individuals who won't buy into the narrative.

Further, how many voted for Obama in 2008 on his promise to close Gitmo?

Corey: The Democratic establishment, including the great majority of elected leaders and the party as a whole, have backed Obama to the hilt on his war on terror excesses, right? Very much in the same way that the Republican establishment backed Bush, and presumably for the same reasons.

Now to be fair, while the Democratic establishment often criticized Bush for the horrible things he did in the war on terror, they never really did much to stop any of that stuff, and that's about the same thing they're doing with Obama. So mainly, the difference is whether they say nice or mean things about the stuff they're not about to actually get involved in stopping.

Why would it be just even if only the guilty are tortured? Hell of an assumption.

It might be that no just world is such that someone is tortured in it, but if there were a world that was just and in it someone was tortured, then that torture would have to be just. In fact, I am inclined to agree that no just world includes torture. Thus we are envisaging what is sometimes called a 'counter-possibility'. A counter-possibility is something that is not really possible, but presents the appearance of possibility (perhaps because it is allowed by salient principles -- even if ruled out by hidden principles). In this case, the appearance of possibility is supported, no doubt, by acceptance of the principle that it is not always unjust to inflict suffering. (Witness training, medical procedures, layoffs, assessment, or just punishment (if there is any)). We often have to make resort to counter-possibilities in making sense of others' beliefs. So we don't really disagree, except that I think that Tabarrok needn't really think that a just world might include torture, while you suggest he is assuming this.

So we don’t really disagree, except that I think that Tabarrok needn’t really think that a just world might include torture, while you suggest he is assuming this.

Well, he might not, but his first sentence is, "If the world is just, only the guilty are tortured." He could have written, "If the world is just, then no innocents are tortured."

I also wonder whether these results actually mean anything. Notice that even though the range of possible answers is 1 to 5 the graph only covers roughly the vertical distance from 1.75 to 3.75 and that the error bars are only +/- one s.e.

Redraw it covering the full range, and show the 95% confidence interval and the differences will appear much smaller, and the error bars will all overlap.

Less here than meets the eye, I'd say.

Since we aren't really at war, what the government is doing amounts to some far off tortures and murders, like some nut job across the country. It is high evil, but we are busy. If we were really at war, we'd be more interested and then wouldn't care because we'd be more patriotic.

What can I do about all the idiotic things coming out of DC? Vote? Funny. Get arrested? Funnier. I had a friend spend a couple years in Federal Prison on a drug war protest against an illegitimate law. You know what he got when he got out? Older. And if we were really at war I would be able to do even less.

What can I do about all the idiotic things coming out of DC?

"Irish democracy"

More regimes have been brought, piecemeal, to their knees by what was once called “Irish Democracy,” the silent, dogged resistance, withdrawal, and truculence of millions of ordinary people, than by revolutionary vanguards or rioting mobs.

Particularly in this case. Is torture worse than drone strikes? Arguably the only thing we got out of torture was targets, such as OBL's compound in Pakistan. So, by not torturing we probably just drone strike more targets with less intelligence and more collateral murders. What allows this race to the bottom is probably the seduction of one party vis-a-vis the other.

"I think research in moral reasoning is important because understanding why good people do evil things is more important than understanding why evil people do evil things"

I find this last paragraph a ridiculous end to an interesting post.
There isn't a cohort of human beings who self - identify as "evil people" who spend their lives doing "evil things." Whether you're a conservative Republican, an Islamist Jihadist, a bleeding heart liberal or mafia boss the one thing you think is that you are doing the best you can given your own (maybe only implicitly specified) way of looking at the world. Even when people boast about being bad it's essentially a cheap-talk status play to deter rivals, or - at worst - part of a slightly over-zealous rebellious growing up phase of trying on different identities.

What he's likely referring to is the bent towards abnormal psychology akin to the medical bent towards disease.

The sentence and word 'evil' have been used a bit too much to be just 'the mentally ill with violent outcomes' even if nowadays, we tend to redefine evil as an illness.

The truth is indeed that almost no one, not even Charles Manson, think of himself as evil...

Yeah, but they are evil. And we lock them up and throw away the key. But they are usually only able to kill a few to a few thousand people. They usually aren't able to cause the deaths of millions.

There are however a lot of human beings who
1) Self-identify as good people
2. Identify others as "evil people" because they do evil things.

If you think that only evil people do evil things, and you're self-identified yourself as a good person, then you may well not be as on guard against your own evil impulses as someone who self-identifies as good and thinks that good people can do evil things. The danger is not people who boast about being bad, it's people who boast (even just internally, to themselves alone) about being good.

Good point Tracy W. It has long been observed that the worst atrocities have been committed not by bad people trying to make others bad, but by good people trying to make other people good (in their eyes, natch).

Yes, the 'good' people are the ones to watch out for. They'll support anything, given a bit of rationalisation...

IMO, the best protection against that is to teach that 'us versus them' reasoning is just a bit of genetic programing misfiring in the context of the 21st century...

The torturer is always guilty.

"research in moral reasoning is important because understanding why good people do evil things is more important than understanding why evil people do evil things." absolutely

"If the world is just, only the guilty are tortured." - I assume you are just postulating this for the purposes of this argument and not offering it as a moral or ethical value statement.

"If the world is just, only the guilty are tortured": by golly, that's fair old bollocks from which to start an argument. If you are an American soldier, you torture people for information (unless you're indulging yourself for fun). If you were a Roman Catholic, you tortured heretics so that they recanted and thereby saved their souls.

He's talking about people and their cognitive dissonance.

It is called the 'just-world fallacy,' and it runs very deeply through a certain broad current of American culture/religion.

I get the same impression every time a hippie starts talking about 'Karma'.
Or a Chomskyite starts talking about 'blowback'.

In some people's universe, when something bads happens to you, it's probably because you deserve it.

I think the point of Alex's post is that, while the starting point is indeed bollocks, it's bollocks that people have a terrible tendency to believe.
As the examples you give may well support.

thanks prior_approval

It looks like a case of trying to find a subset of the data for which there is an 'effect' and build a theory around it. Note how (judging from the figure) neither the close/distant nor the pain/no pain conditions would be 'significant' on their own. And note that the confidence bars are based on 1 standard error instead of the more 'standard' 2 which would probably overlap (implying no significant difference within both the close and distant groups). And most importantly of all, what kind of 'effects' are we talking about: all estimates are between 2 and 3 (on the 5-point scale). Even if the interaction is significant, the scale of the effects is quite removed from the scale of the theory which they are supposed to support.
P.S. terrible figure: does not include the full range of the response (y-axis); the bars are misleading (they could as well be filled top-down rather than bottom-up) and displaying the four bars one after the other does not help the crucial comparison (the interaction, which would have been clearer if the close and distant group estimates were plotted in the same space).

+1 As tested on normal subjects, this is . Given how close the bars are, are there characteristics of jailers you would want to recruit that would separate the bars in actual performance. Or, to put it another way, are there characteristics of jailers that make it more likely they would inflict pain and follow orders without the need to feel that the person is guilty. Or, are there authoritarian types that do not question authority and accept an authorities statement that the person is guilty. Interesting followup.

Bingo. There is no way the authors hypothesized only an interaction.

People get defensive when they feel that they are being personally accused of involvement or complicity in some wrongdoing. That includes when some in-group a person identifies with is defined as being responsible (i.e. Republicans, Americans, Democrats, liberals, conservatives).
So it makes total sense that if you want to keep people from getting defensive, you either minimize the severity of the wrongdoing (it's a small problem easily fixed), or identify it as the responsibility of some OTHER group - preferably an outgroup.

I think the uglier side of this is that there's a whole "shaming" dynamic, where instead of actually trying to change the other side's mind, which would entail one of the above, people purposely try to maximize the severity of the wrongdoing AND assign it to the opposing group. I.e. REPUBLICANS are responsible for TORTURE! You're all EVIL GUILTY PEOPLE! Which had the net effect of getting Republicans to defend torture. Useful if your goal is to score political points. Not useful if your goal is to end torture.

Torturers always defend torture.

People in power always defend what they are doing.

Well, until roughly a dozen years ago, nobody in power in America ever defended torture as policy. Times change, apparently.

So what your doing is calling Republicans names.
Because calling people names is such an effective way of getting them to change their policies.

Are you more interested in ending torture, or calling people torturers?

I find disturbing the comments that imply a liberal vs conservative difference on these issues. I am not aware of any significant changes that took place when O replaced B. This business of trying to put a political spin on non political issues is a symptom of a deeper illness in our society. Issues of this import should not be decided by which side cheers loudest as in a sports contest. There are vastly more important things to be decided than can be reasonably addressed by taking a side based on political (or other) leaning.

Oh well, I'll wander off to Kunstler's blog, the title of which says it all.

The only comments are that there is no difference.

The obvious conclusion is that we should be trying to decrease religiosity and patriotism as obstacles to moral progress.

Or increase them - both Jesus Christ and the American Constitution oppose torturing other humans as wrong, one morally, the other legally.

This would be the same Jesus who preached that sinners would be tormented for all eternity, I take it. force wheresoever necessary.

Well, Jesus wasn't about force, and as for the words in the Constitution against torture - well, they seem to have lost their force against a couple of centuries.

How about "If the world is just, nobody is tortured."

Yes, but the world isn't just. Torturing is. And some people need to believe the world is just and thus have to shoehorn the torturing into their just world view. This isn't really about torture per se.

Lovely work Alex. Thank you.

Ticking time bomb justification for torture of course presumes guilt, but let's say guilt is certain, and the issue is saving lives.

ZD30 starts off with torture as the focus, so the litmus test on torture is does ZD30 make the case for or against torture.

The guilt of those tortured is never questioned.

And two hours after the start of the torture, the ticking time bomb threat of bin Laden has been dealt with.

Therefore, torture works.

Except, the real world time line is an eight year span of "time bombs going off" and killing people.

Did torture really save a single life?

If the torture is ok if the victim is guilty, then torture of the guilty is always justified because they are guilty. That was a fundamental debate circa the migration out of Europe to all places, and it colored the conflict between "Americans" and British authorities and their agents, and also colored the conflicts between religious communities and within them, eg the "Puritian" enforcement of moral purity.

One of the remaining debates two centuries after the explicit prohibition on torture - no cruel and unusual punishment - is the death penalty, solitary confinement, and life sentence of 60-70 years (15 year old kid who dies at 75 after 60 years in prison).

The debates all center on "does it work"? And it goes a bit further, does it have economic utility.

A perhaps cynical concept is that of scapegoat. This is the sacrificial lamb. It grows out of paying property for a transgression to provide some justice.

Torture, execution, life in solitary, to a degree only needs a victim who can be despised to satisfy a part of society who want some to pay, guilty or not. When people convicted decades ago are exonerated, with the presumption of guilt driving the blind pursuit of conviction and execution without regard to the evidence being absolutely clear in the record, those involved in the prosecution too often insist the person was and still is guilty. When the DNA evidence connects the crime to a person already serving life who confesses and who the evidence supports, the prosecutors refuse to admit they were wrong, which means the real criminal can never be prosecuted because the defense points to the prosecutor's own claim that the right person has already been convicted. Victims want the innocent person executed because that is who they have blamed for their suffering for two decades.

Hey, how long have economists concluded the drug laws don't work - they fail the utility test. Yet how many economists actively engage in the political-economy of decriminalizing drugs and their trade, developing something like the alcohol and tobacco policies for them? I bet most economists conclude that those who use drugs are guilty of being bad people in some fashion, and thus deserving of "torture". But does 90% of taxpayers need to be "tortured" paying for the "torture" of the kid of someone we don't know. If our own kid gets swept up in a drug crime and subjected to the "torture" of jail, court, prison, criminal record, and a life sentence of denied job opportunity....

This topic simply comes down to why people fail to make rational "market theory" "utilitarian" decisions on so many policies that society implements. Why do people reject the clear evidence that human activity can cause deadly environmental harm? California suffered severe environmental harm and ended up passing lots of laws that Texans claim are unneeded. But how many Texans would chose to live in parts of China where the decision makers totally agree with the Texans and you need very expensive breathing equipment to avoid harm just from walking down the street.

People are not naturally rational actors.

Every comment for this thread:

"The problem is that stupid people with misplaced moral certainty inflict evil in the name of justice," I say with misplaced certainty about the universal applicability of my own moral values and a full willingness to "reeducate" those who disagree with me and to constrain those who don't see the light.

Torture is about torture.

There are those opposed to torture, and those who defend it. And there are those who practice it.

America used to represent a society that had always rejected torture, proudly. Now, like most other societies throughout human history, it has used torture as an instrument.

Call me nostalgic for the past. The one where no one felt any need to restrain torture as a policy, since it was rejected out of hand as being thoroughly and completely un-American.

What? Ever heard of the Wickersham Commission?

The main takeaway of Wickersham report re:torture is that shockingly this bad thing (torture) is widespread. It's completely in accordance with what prior_approval is claiming (general societal rejection of torture).

The Indian Wars, then? Slavery - which included a serious element of torture to break non-compliant slaves?

I just don't get the white-wash of US history going on here.

Wait, *why* would it be that in a "perfectly just" world, only guilty people are tortured? Since when does guilt of anything provide moral warrant for torture?

Shouldn't the line be a bit more like, "for those who believe torture is justifiable, it is (usually) justifiable only against people who are guilty. (Though this is somewhat tautological *if* we restrict torture to its instrumental, and therefore any victim of torture is by definition guilty of concealing something they have no right to conceal) Therefore, believers in a just world who also believe torture is justifiable, will be more likely to think that those tortured are guilty." Or, to the empirical point, "in general the more guilty someone is (seen to be) the less concerned (some/most) people are about their moral status."

The empirical point about what drives people's reactions to torture doesn't need to say anything about when torture is or isn't justified. I'm not sure what the point is of starting things off with a rather risible, but materially irrelevant, moral axiom.

Well, America as a society which has lost its instinctive rejection of torture could be one explanation.

Several decades ago, this discussion, as you have noticed, would have been rejected, since torture itself was rejected as having any utility, a belief originating from the very founding of the United States.

And I think that a lot of the reason for that is that because certian people were more interested in expressing their hatred of Republicans and/or Americans more generally, they pushed a lot of people into a position of defending torture. Particularly in the 2003-2004 timeframe, the level of anti-American virtriol being expressed by the anti-war movement pushed a lot of people into defending the actions of the US, and at first that was mostly to say that these things that were being described as torture really weren't torture. (i.e. waterboarding isn't really torture, etc._ By taking the other tack - that this was a problem and it could be fixed, that would have been more effective. But instead it had to be this OMG! AMERICANS ARE EVIL! SEEE WE TOLD YOU SO! LOOK EVIL AMERICANS! message which antagonized Americans and made them pay LESS attention to it. Which was ultimately more about people on the left "proving" how evil America was for ultimately reasons that are totally unrelated to torture. They weren't trying to fight torture, they were trying to get people to hate America.

"because certian people were more interested in expressing their hatred of Republicans and/or Americans more generally, they pushed a lot of people into a position of defending torture. "

This has nothing to do with anti-Americanism or politics. Americans who accepted torture did so because they were assured that it only happened to terrorists. America's increased acceptance of torture is more or less in accordance with Alex's linked study -- slap an imaginary Al Qaeda face on the victim and a lot more people were ok with it. The second you ask these people whether the police should be able to waterboard their child if he is a suspect in a crime they will scream bloody murder.


You know how a lot of news reports about the Tea Party, especially partisan ones from the left, imply that the Tea Party is mostly made up of stupid racists? And how this is basically used to pre-emptively declare that whatever complaints they voice are invalid and shouldn't be heard? It looks to me like that smear is identical to what was done with the antiwar movement and the anti-torture movement. A whole lot of us thought that the Iraq war was a really bad idea, and we were routinely called traitors or blame America first liberals or accused of being unpatriotic, and that was used as a reason not to listen, to give almost no coverage to antiwar rallies, etc.

Now, I'm sure there were people opposing both the Iraq war and our formal policy of committing war crimes because they hated America. Just like I'm sure there were people opposing Obama and going to Tea Party rallies because they hate blacks. But mainly, this kind of smear is about shutting people up, or deciding up front that you don't have to listen to what they say.

> why good people do evil things is more important than understanding why evil people do evil things

IOW, why 90% of evildoers are more important to study than 10% of evildoers

My question is why are psychologists OK with using deception in their experiments? Because the research ends justify the means?

Because Human Research Review Boards say it's ok?

Because most of their subjects are psychology students.

It's a great opportunity for moral preening, but I think that none of the commenters so far has looked at the experiment.

It depended upon tricking 80 Harvard students into thinking they were participating in a different kind of study. They washed out 6 participants for 'suspicion' but do any of you believe that in 2013, with every intro psych class on earth discussing the Milgram experiments, and using a group of the most cynical people on earth (kids who make it into Harvard), that a single one of the participants was really fooled? Maybe these were even undergrad psych students (though the average age was 24). I can't tell whether the participants were paid, or received class credit, but I expect that the only people who washed out were the people who either didn't want to do it or were too dumb to pretend they weren't 'suspicious' of the real nature of the experiment.

From the study:

Eighty-eight participants (54 females, 32 males, 2 unspecified,
Mage = 24) were recruited from on-campus sources. Six participants
were excluded for suspicion, leaving a total of 82 participants.

Participants in both conditions were extensively debriefed for
suspicion using a ‘‘funnel-debrief,” as suggested by Bargh and
Chartrand (2000). As mentioned earlier, six participants were excluded,
but the vast majority believed the experimental set-up.

Excellent points.

I am surprised that Alex relied on an experiment like this for anything. Can't every study describe in some detail the demographics of the people involved and how they were solicited? Something like "80% of the participants were students at the University; 20% were offbeat locals who emailed after seeing our flyer at an off-campus coffee shop. All were given [$50/ice cream/class credit] for their participation [regardless of whether][unless] they were excluded for 'suspicion'."

And commenters seem to be making a wild leap from the study's subject of guilt to the very different subject of 'War on Terror' torture. Almost every complaint about torture by the US and its allies presumes that the victims are themselves guilty of crimes but that despite this guilt it is a terrible thing to torture them for information about their confederates and plans.

That is very different from the more common problem of many countries' normal criminal laws, where police have long tortured/stressed suspects for the purpose of getting information to show that they are guilty of a completed crime.

I am now going to exclude myself for suspicion.

It looks like the close group also shows a larger variance. Is there any explanation of that?

Very interesting. Made even more so given the media attention to the pope.

Has anybody tried using electronic cigarettes?

Comments for this post are closed