Sam Harris on profiling

For a recent blog post on profiling, CFI’s John Shook accused Sam Harris of being a “bigoted hate monger.” The reaction in the rest of the atheist internets has been almost all negative, though not quite as harsh. Let’s have a look at what Harris said:

Although I don’t think I look like a jihadi, or like a man pretending not to be one, I do not mean to suggest that a person like me should be exempt from scrutiny. But other travelers fit the profile far less than I do. One glance at these innocents reveals that they are no more likely to be terrorists than walruses in disguise. I make it a point to notice such people while queuing for security at the airport, just to see what sort of treatment they receive at the hands of the TSA.

While leaving JFK last week, I found myself standing in line behind an elderly couple who couldn’t have been less threatening had they been already dead and boarding in their coffins. I would have bet my life that they were not waging jihad. Both appeared to be in their mid-eighties and infirm. The woman rode in a wheelchair attended by an airport employee as her husband struggled to comply with TSA regulations—removing various items from their luggage, arranging them in separate bins, and loading the bins and bags onto the conveyor belt bound for x-ray.

After much preparation, the couple proceeded toward the body scanner, only to encounter resistance. It seems that they had neglected to take off their shoes. A pair of TSA screeners stepped forward to prevent this dangerous breach of security—removing what appeared to be orthopedic footwear from both the woman in the wheelchair and the man now staggering at her side. This imposed obvious stress on two harmless and bewildered people and caused considerable delay for everyone in my line. I turned to see if anyone else was amazed by such a perversion of vigilance. The man behind me, who could have played the villain in a Bollywood film, looked unconcerned.

[snip]

We should profile Muslims, or anyone who looks like he or she could conceivably be Muslim, and we should be honest about it. And, again, I wouldn’t put someone who looks like me entirely outside the bull’s-eye (after all, what would Adam Gadahn look like if he cleaned himself up?) But there are people who do not stand a chance of being jihadists, and TSA screeners can know this at a glance.

Needless to say, a devout Muslim should be free to show up at the airport dressed like Osama bin Laden, and his wives should be free to wear burqas. But if their goal is simply to travel safely and efficiently, wouldn’t they, too, want a system that notices people like themselves? At a minimum, wouldn’t they want a system that anti-profiles—applying the minimum of attention to people who obviously pose no threat?

As far as I can tell, Harris is right that would-be hijackers and bombers on American planes at this point in history are most likely to be Muslim. There are other ideologies that motivate terrorism, but that’s the main one we (in America, and many other countries) are facing right now.

The problem that I actually see with what Harris says is that “anyone who looks like he or she could conceivably be Muslim” isn’t a profile. Even the wheelchair-bound old woman in Harris’ example could conceivably be Muslim (perhaps of a liberal sect that doesn’t require women to wear headscarves.)

All the same, Harris’ point about anti-profiling seems right. Broad anti-profiles (like just “female”) would be a mistake because they’re too easy to exploit (terrorists respond by recruiting more women), but I doubt there’s much risk of terrorists specifically recruiting 80 year old wheelchair bound women to slip bombs past airport security.

  • http://cafeeine.wordpress.com Cafeeine

    but I doubt there’s much risk of terrorists specifically recruiting 80 year old wheelchair bound women to slip bombs past airport security

    That is a naive view, in my opinion. Sure, Al-Qaeda might not go for recruiting an 80-year old wheel-chaired retiree for a bombing now, but the moment this category openly becomes an anti-profile, they become a prime target for trojan-horsing. This should have been obvious to Harris.

    • jacobfromlost

      I think the reality is that it so very unlikely that Al Qaeda could recruit an 80 year old, wheel-chair bound woman that simply entertaining the idea is irrational. That’s Sam’s point.

      We seem to be taking story logic from “24″ and “Homeland” and trying to apply it to reality. It might make a great plot twist in a show, but imagine yourself in Al Qaeda’s present position, and tell me how you recruit an 80 year old woman in a wheel-chair…or why you would want to (in the context of reality and all the specifics of Al Qaeda’s ACTUAL circumstances)? If the counter argument is “because security wouldn’t be screening the 80 year old”, then you have to explain how, exactly, you imagine such a woman being recruited, despite not fitting the profile in other respects. If she DID fit the profile in other respects, she would be profiled. If she DIDN’T, I don’t see any reasonable method for recruitment. Al Qaeda doesn’t have the resources, means, or know-how to do something like that.

      • Kevin

        To make it hit home harder, let’s say that terrorist recruiters target those with depression since it is easier to convince someone who is suicidal to commit suicide. To use their same logic, the recruiters will then just target people who are completely content with their lives and try to convince them to commit suicide. The logic is almost too comical.

        • jacobfromlost

          This reminds me of another example of how inept Al Qaeda is, and how we are our own worst enemy in this respect:

          After bin Laden was killed, there were immediately rumors and conspiracy theories rampant on the net claiming he wasn’t dead, that Obama staged it, etc. Standard stuff. Obama knew this could be a problem, considered releasing the pictures, decided against it as it would be counterproductive, and wouldn’t be “proof” to conspiracy theory nuts anyway. Obama also said bin Laden would never be seen again, never make another video, etc. That’s pretty much all Obama could do to prove bin Laden was dead (if people wouldn’t accept the facts of all of the American officials involved in the whole operation because they are part of the “conspiracy”.)

          But what did Al Qaeda and bin Laden’s family immediately do? They confirmed bin Laden was dead, again and again, sometimes overtly and sometimes by complaining about how it was done! Every time some conspiracy nut started claiming he was not dead, I’d simply say, “So why does his family say he is dead? Why does Al Qaeda say he is dead? What is the purpose of them claiming he is dead?” Within a couple of week’s, that conspiracy theory vanished, at least to a very, very small number–entirely, I think, because the bin Laden family and Al Qaeda almost immediately admitted it.

          All Al Qaeda had to do was collectively say, “That wasn’t bin Laden”, or even just blatently say, “None of the events the US is claiming happened.” That would have been enough to perpetuate the conspiracy theories and fear, no matter what US officials said, and caused Obama all kinds of problems. But they just are not organized enough or smart enough to do something like that, and that would have been the simplest form of terrorism, costing them nothing, and causing us all kinds of problems (not the least of which is making bin Laden into a kind of inverse-martyr; he might be dead, AND might be able to kill us all indefinitely into the future so we better spend billions securing ourselves against ghost attacks).

      • eric

        imagine yourself in Al Qaeda’s present position, and tell me how you recruit an 80 year old woman in a wheel-chair

        You recruit the mother or grandmother of one of your current mosque-goers. You recruit a doctor to get her a wheelchair (not hard – Al-Zawahiri himself was an M.D.). And, like I believe was done for all of the 9/11 terrorists, you get one of your Imams to tell her than its acceptable to Allah that she dress and act western for the mission.

        Or you just find someone who fits the excluded category and plant a bomb on her.

        Your challenge points out a common problem with making an excluded category; you assumed that the category of wheel-chair-bound, western-looking, elderly women with strong terrorist sympathies is too small to worry about. That’s probably actually a good assumption. But terrorists don’t have to recruit from that group to use the excluded category to get a bomb on the plane. They basically just need an older woman, and can fake the rest. IOW, you vastly overestimated the unlikelihood because your assumptions about what they had to do to access that category were wrong.

        The same problem could occur with any excluded category: you might vastly overestimate the unlikelihood because you make some flawed assumption about who terrorists must recruit to gain access to that group. Here’s another historical example: pregnant women. Surely they wouldn’t be bombers! Well, in fact, Al Qaeda already tried that. And they didn’t need to find, recruit, and brainwash some pregnant woman – they just had her husband plant a bomb on her.

        We prevent becoming a victim of our own wrong assumption by random screening. It protects us not just from threats we asses as highly likely, but the truely high likelhood threats that we may have assessed incorrectly.

        • jacobfromlost

          Eric: You recruit the mother or grandmother of one of your current mosque-goers.

          Me: How? How do you find Al Qaeda members in a mosque? It’s impossible for me to believe that tons of everyday Muslims in the US are actually Al Qaeda, and “asking around” is not effective as that is the fastest way to have the FBI knocking on your door, or tapping all of your communications.

          Eric: You recruit a doctor to get her a wheelchair (not hard – Al-Zawahiri himself was an M.D.). And, like I believe was done for all of the 9/11 terrorists, you get one of your Imams to tell her than its acceptable to Allah that she dress and act western for the mission.

          Me: You are acting as if there are Al Qaeda sleeper cells all over the US, and that any random Imam will tell any random old woman it is ok to blow herself up in plane. I think the lack of any attack over the last 10 years illustrates Al Qaeda is not everywhere.

          Eric: Or you just find someone who fits the excluded category and plant a bomb on her.

          Me: You’re skipping over the hard part. Finding such a person and planting a bomb on them IS VERY, VERY HARD TO DO. It would be much easier just to shoot a shoulder fired rocket at the plane from the edge of the runway. Why would you go through all the trouble of planting a bomb on an old woman in a wheelchair, when far simpler ways to blow up the plane are available.

          Eric: Your challenge points out a common problem with making an excluded category; you assumed that the category of wheel-chair-bound, western-looking, elderly women with strong terrorist sympathies is too small to worry about. That’s probably actually a good assumption.

          Me: That’s not the only assumption I made. The other was that it would be far to difficult GIVEN AL QAEDA’S ACTUAL SPECIFIC CIRCUMSTANCES NOW to do such a thing.

          Eric: But terrorists don’t have to recruit from that group to use the excluded category to get a bomb on the plane.

          Me: But I’m talking about actual Al Qaeda as it exists. They don’t have the resources, means, or know-how to do what you are saying. The most they could do is blow up a plane, and there would be much easier ways to do it from the ground.

          Eric: They basically just need an older woman, and can fake the rest. IOW, you vastly overestimated the unlikelihood because your assumptions about what they had to do to access that category were wrong.

          Me: I don’t think they are wrong, though. It would be far easier to shoot a plane down from the ground, yet they haven’t done that. In fact, they haven’t done ANYTHING, and there are quite a few things that would be much easier to do.

          Eric: The same problem could occur with any excluded category: you might vastly overestimate the unlikelihood because you make some flawed assumption about who terrorists must recruit to gain access to that group.

          Me: How do you know YOU aren’t making flawed assumptions about the resources and means that Al Qaeda has to carry out any attack? You seem to assume there are many members of Al Qaeda in America. If so, why have they not done anything in 11 years?

          Eric: Here’s another historical example: pregnant women. Surely they wouldn’t be bombers! Well, in fact, Al Qaeda already tried that. And they didn’t need to find, recruit, and brainwash some pregnant woman – they just had her husband plant a bomb on her.

          Me: In America? We’re talking about American airline security here.

          Eric: We prevent becoming a victim of our own wrong assumption by random screening. It protects us not just from threats we assess as highly likely, but the truely high likelhood threats that we may have assessed incorrectly.

          Me: I don’t agree with your assessment that my assessment is wrong. All the data available suggests otherwise.

          If we allow our imaginations to run THAT wild, why can’t we assume a dozen pilots will be turned to Islam, then to Al Qaeda, then assume they will crash their planes into landmarks, so close all the landmarks and force pilots to take lie detector tests every time they fly?

          I think bin Laden’s recently released letters support me, not you.

          • Yaron

            @jacobfromlost – Your response to Eric basically reads as a claim that it’s hard for a terrorist organization these days to find people willing to carry bombs into airplanes.

            Which is true, but of course hard is not impossible. And is beside the point for the purpose of the discussion. If the discussion was should any people be screened and checked in any way before boarding planes, then it would have been a valid point worth discussing.

            But the issue is not whether anyone in particular may be a terrorist carrying a bomb, but whether a specific sub-group of people (e.g. elderly women in wheelchairs) are less-likely enough to be so that they can be ignored while general purpose screening of people still goes on.

            And the overall difficulty of recruiting people in general… is not relevant.
            If you assume that a random young person may be, then that person has a good chance of having a grandmother, with a decent chance that this grandmother is alive. Ergo, if you accept that some people may be terrorists carrying bombs, then it follows that old people in wheelchairs may also be.
            And if the policy will be not to screen old people in wheelchairs, then for any plan to use a young person to carry a bomb, the grandparents of that person will be used instead, since the chances of success will be much higher.

            And it may even be easier to convince elderly people to become suicide bombers, since they have less of a life to expect to anyway.
            Remember that old people in general aren’t less religious than young people, or less politically driven than young people. So the same amount of willingness to die for the cause, combined with less personal cost when doing so, means that recruiting “someone”‘s grandma won’t be much harder (or may be easier) than recruiting “someone”.

            All your comments on the difficulty of reaching the original “someone” are, again, not relevant to the discussion. They’ll only be relevant if the issue was “do we need to screen anyone at all?”. But, again, once it’s accepted that some people should be screened, it doesn’t make sense to exclude old people, since the difficulty of finding and recruiting these old people is not more difficult than recruiting the young people who are not considered for exclusion.

  • RW Ahrens

    My biggest criticism of anyone who touches this subject this way is that they have failed to ask the Israelis how they do it.

    For over thirty years or more, they’ve been dealing with the issue of their enemies trying to hijack aircraft. They’ve gotten pretty darn good at it, because they don’t profile looks, they profile behavior, which is apparently a better indicator of what somebody has in mind.

    The problem with that is the education that is required to make your screening personnel understand how it works without their personal biases about looks getting in the way.

  • Kevin

    The I think the crucial factor that makes Harris’ point far from being easily dismissed is the current state of affairs. The main argument against profiling is that it is incredibly not effective, it produces way too many false positives and tramples too many civil rights. However, the current process produces way too many false positives too. If profiling is marginally effective (which is debatable), even in the slightest, it will make the current process better. Also, its not like you can argue on 4th amendment grounds because the administration has declared that it is reasonable to search anyone who wishes to board a plane. It seems that to stay consistent we would have to either start profiling or to require more cause to search someone at an airport. Personally, I prefer the latter.

  • http://www.nd.edu/~dhicks1 Dan Hicks

    As far as I can tell, Harris is right that would-be hijackers and bombers on American planes at this point in history are most likely to be Muslim. There are other ideologies that motivate terrorism, but that’s the main one we (in America, and many other countries) are facing right now.

    1) The direct inference from the premise that many terrorists are Muslim to the conclusion that many Muslims are terrorists is an instance of the base rate fallacy. About 0.8% (8 in 1,000) of the US population is Muslim; let’s suppose there are about 3,000 terrorists (1 in 100,000) and 90% of them are Muslim. Then

    P( terrorist | Muslim ) = P(T|M) = P(M|T) * P(T) / P(M) approx.= (9 E -1) * (1 E -6) / (8 E -4) = 0.001125, or about 0.11%.

    In other words, someone who’s Muslim in the US is almost certainly not a terrorist.

    2) The premise itself is arguably false, depending on how strong `most likely’ is. Prior to Sept. 11, the deadliest terror incident in US history was the Oklahoma City Bombing, committed by a white sympathizer with the militia movement. Last year, the FBI arrested four men associated with the militia movement for plotting to buy explosives and manufacture ricin. Five of the FBI’s current Ten Most Wanted are white non-Hispanic; the other four are white Hispanic.

    • swb

      Dan,
      To your first point — Harris/Hallq, in my reading of things at least, have only asserted (in your words) “that many terrorists are Muslim” but I haven’t seen where anyone is concluding that “many Muslims are terrorists” anywhere, as you assert — so I find your base rate fallacy claim to be a bit erroneous. Besides, I don’t think anyone needs a probabilistic calculation to prove to themselves that the vast majority of Muslims aren’t terrorists — that’s simply common sense to any reasonable person.

      To your second point — you may be correct. Overall, non-Muslim terrorists might very well outnumber — and thus be a much larger potential threat in the US — than Muslim terrorists, but that is a different discussion. Harris/Hallq are simply addressing the TSA screening procedure which attempts to protects us from a singular, very specific form of terrorism (hijacking), not terrorism as a whole. I think the relevant data to examine here would be Muslim hijackers vs non-Muslim Hijackers, not to simply examine all domestic terrorists.

  • http://peicurmudgeon.wordpress.com/ peicurmudgeon

    I think the current stats on racial profiling in NYPD’s stop and frisk policy as described by Ed Brayton:
    http://freethoughtblogs.com/dispatches/2012/05/03/nypds-racist-stop-and-frisk-policy/

    are as good a proof as any that profiling does not work.

    • http://www.facebook.com/chris.hallquist Chris Hallquist

      I agree the stop and frisk policy is awful. That doesn’t show profiling in general is wrong.

  • mnb0

    Come visit my hometown, Moengo Suriname, and tell me who is a muslim and who isn’t. My “girl”friend (well, she’s 55) only wears a headscarf when going to her mosque. And she has been member of the board for a couple of years …..
    Still she is just Sunnite and not member of some sect.

  • Tim Groc

    It is a difficult one.

    I do wonder how many of those people on FTB who have criticised Harris in the last couple of days, would act any differently if they were the ones who had to “profile”. Deep down, they would make decisions based on appearance, behaviour, clothing, gender, etc. even though they might openly deny it.

    It seems to me you either profile or you don’t. If we don’t, then don’t complain about long waits in the airport lounge.

    I don’t expect to see any of FTB’s more famous jet-setters complaining about waiting in airports. If I do, I’m going to shout “hypocrite”.

  • piero

    Drug-traffickers have no qualms abour stuffing a newborn with drugs. An elderly couple could be blackmailed. But these are very unlikely scenarios. Let’s say that the chances that an elderly couple having being blackmaioed to carry a bomb are one in a million. That’s comparable to the chances that the plane will crash due to mechanical failure or human error.

    In short: anyone could carry a bomb and destroy a plane. But the relevant question is: how likely is it? What percentage? We all know that some people will die from an adverse reaction to a vaccine, and that percentage (based on current evidence) is far higher than the probabiity that an elderly couple from Vermont have either become radical Muslims or have been blackmailed to do as told.

    An empire that needs to keep its people in a permanent state of paranoia is in its death throes. My only hope is that the new empire proves to be less blood-thirsty and less petrol-thirsty.

  • jg29a

    In the replies, a lot of people are coming up with the very first step of a game theoretical analysis (terrorist groups will see that certain types of people are screened less intensely, and therefore make efforts to recruit from those types), but what’s clearly required is an iterative model. Anti-terrorist agencies have intel on recruitment, training, and long-term strategy of terrorist groups. They adjust to terrorist groups’ adjustments, and then the terrorists adjust again, and so on. It’s very complicated, of course, but many decades of excellent research have gone into modeling this sort of phenomenon.

    The main problem in the responses to Harris is the leap from the obvious point that terrorist groups will adjust to the circumstances of profiling, to the frankly ludicrous idea that the iterative adjustments might lead to an equilibrium wherein every identifiable demographic presents a nearly equal risk.

    • Kevin

      This reminds me of a natural game of rock-paper-scissors. I don’t remember the exact species but the joist is that species “scissors” beats species “paper” which beats species “rock.” When there are many rocks, scissors are selected against and papers is selected for. When there are many papers, rocks are selected against and scissors are selected for. This cycles, almost by generation. Unlike the game rock-paper-scissors, the equilibrium is different because it takes a long time to adapt to the new situation because to take advantage of the new environment it requires a new generation to be produced.

      Similarly, to take advantage of the new profiling requirements, terrorists require recruiting and training a new generation of terrorists. The question becomes how much intelligence do we have of their recruiting process. If it is complete, it would be like playing rock-paper-scissors while being able to see what the other person is going to play. Sure, they may switch strategies, but we would know the appropriate profile to switch to. If not, then the proper equilibrium would probably look more like the species than the game.

  • MatthewL

    Profiling is a problematic practice for all sorts of reasons both pro and con. It seems that the only way it can really work effectively is with highly trained and carefully selected agents.

    While this makes sense in Israel where there is high risk in a small theater it doesn’t seem realistic in the much larger and lower risk world of air travel in the US. Even if we had the budget and inclination for such an approach we would be unlikely to find and train enough suitable candidates. I suspect this is part of the reason we have chosen to rely more heavily on technological methods (along with lobbying on the part of the suppliers of course).

    I also think it’s bad policy to give a free pass to low risk passengers. This allows the privileged class (such as successful, middle aged white males like me) to disregard the difficulties others have to bear.

    The best solution for cases involving the elderly, disabled, young children and so on would be to establish procedures to effectively screen them while accommodating their special needs. Perhaps this is something that should be addressed under the Ameicans with Disabilities Act.

  • Pingback: Bruce Schneier: get rid of secondary screenings at the airport | The Uncredible Hallq

  • Pingback: instant performer

  • Pingback: Is Ovarian Cyst Miracle Legit

  • Pingback: picture fram

  • Pingback: water ionizer

  • Pingback: minecraftlikegames.com

  • Pingback: water ionizer

  • Pingback: Darian Braun Lawsuit

  • Pingback: yellow october


CLOSE | X

HIDE | X