Is certainty dangerous?

When I re-read Sam Harris’ The End of Faith several months ago, I noticed a passage that at the time I thought Harris could not possibly have meant, and (I thought) was probably just carelessly repeating from the more ecumenical religious writers. Here’s the passage:

Once a person believes–really believes–that certain ideas can lead to eternal happiness, or to its antithesis, he cannot tolerate the possibility that the people he loves my be led astray by the blandishments of unbelievers. Certainty about the next life is incompatible with tolerance in this one.

On this, I wrote:

Harris gets something importantly right here, that the belief that infidels are damned is very dangerous. But the way he says it is just begging to have believers complaining about atheists being so certain and intolerant.

The right response to this sort of criticism is that it’s the content of the beliefs that matters, not the level of certainty with which they’re held. Better to be certain the Earth revolves around the Sun, than think that God probably wants you to kill infidels. And “tolerance,” sadly, has become a meaningless feel-good word that Harris would’ve been wise to avoid.

To elaborate: if someone only thinks there’s a 50% chance that Christians go to Heaven and non-Christians go to Hell, that’s enough to motivate some pretty extreme measures to get people to convert to Christianity and stop people from apostatizing. On the other hand, nothing of that sort would follow from being certain that beliefs won’t have any effect on the afterlife, even if there are some small bad effects from that kind of certainty.

I initially assumed that Harris would agree if this were pointed out to him. But now one of Natalie’s posts has made me realize that some people really do, as their considered opinion, think certainty is dangerous. I’m very puzzled by this. I mean puzzled–I’m not using that as a euphemism for anything dismissive.

I can give one reason I have for being skeptical of the alleged dangers of certainty. It’s that it seems to me that it’s totally normal to be certain about some things. We don’t normally think there’s anything insidious about answering “yes” to, “are you certain?” I touched on this in the Not so Different posts:

It’s better to respond to believers who complain about atheists being so certain by asking them, okay, how certain are you (if you’re not a Mormon) that Joseph Smith didn’t really dig up a set of golden plates, containing the writings of a series of ancient Native American prophets, who were genuinely guided by God, and then translate them accurately into English with supernatural help? And if you’re going to claim no one can be 100% certain about anything, what percentage chance do you assign to the possibility that Joseph Smith really did all that stuff?

Now there actually are some interesting philosophical arguments as to why no one can be 100% certain about anything, but I get the feeling that most people who say that aren’t motivated by a deep understanding of the arguments. They’re just repeating something they vaguely remember from the one philosophy course I took in college. And even if you thought seriously about the issue and think those arguments are right, the most you can say is that while nothing is 100% certain, some things are just so astronomically improbable that our brains can’t even keep track of the tiny probabilities, so their chance of being true may as well be 0%. That’s my response to people who complain about atheists being so certain.

Now maybe in these cases we really should say we’re not 100% certain, we should say we’re ninety-nine-point-however-many-nines-you-feel-like percent certain. I’m not denying that. But I have a hard time believing anything of great importance rides on that distinction. It’s an arcane philosophical dispute, and I can’t believe coming down on the wrong side of such a dispute would be dangerous.

But the thing I’m really hoping to hear from people on in the comments–is this a question people find interesting? Is it worth beating my head against a bit more? Do many people share Natalie’s point of view? Or would I be better off just shrugging at this one?

How selfish are voters?
Catholics: why aren’t you Protestant?
Avoiding divorce doesn’t make you a traditionalist
Slavery abolition and animal rights: the biggest problem
  • Ace of Sevens

    I just wanted to be the first to ask if you’re sure that certainty is OK?

  • Ace of Sevens

    Now that that’s out of the way, the danger of certainty isn’t so much direct harm it causes as the fact it’s irrational and unconnected to how likely a belief is to be true. The psychological mechanisms that allow one to be certain tend to encourage extreme beliefs in general, which is bad.

    • ash

      Depends on “who” is being certain. If someone like Dawkins says he is absolutely certain about something, we all assume the caveat…

  • Brad

    I don’t think that you’ve interpreted Natalie’s article correctly. She’s knocking down the straw man of “Post-modernism means all knowledge is equally uncertain”, not supporting it:

    While we had people like the existentialists examining that wasteland of lost values, and contemplating the total absence of certainty and certain meaning, and how to cope, we also had people like Foucault trying to find new ways of understanding and reasoning, of dealing with acceptance of the fact that all positions are inherently subject positions, and conditioned by a context. To survey the plurality of positions and understand The Truth not as a singular goal but as a process and dialogue, a discourse.

    This is not the same thing as absolute relativism. Nor is it incompatible with skepticism. They are, in my mind at least, easily harmonized, and deeply so. Although certain individual post-modern thinkers did indeed assume a position of absolute relativism, or the idea that there is no truth at all, that’s not the position of post-modern thought in a general sense. Post-modern thought simply teaches you to be careful to acknowledge the biases and assumptions you’re bringing to the situation, how the context will affect your interpretation, awareness that there may be unconsidered variables at work, that other perspectives are worth considering, and to recognize the risk and danger in certainty, or losing sight of the possibility you’ve got things wrong.

    None of that teaches us to abandon critical thought, or stop thinking there’s any point in weighing one argument above another. It’s only the shallowest, most unsophisticated readings that would interpret the overall thrust of existential and post-modern thought to be “well, everyone’s, like, you know, entitled to their, like, opinion and stuff, dude”. Instead, the questions post-modernism encourages us to ask are of deep importance to being able to develop as truly critical thinkers. It teaches us to consider factors most people miss, and to be able to apply skepticism and critical inquiry inwardly, as well as to the overall context, instead of limiting it to the analysis of a particular presented idea. Consider the presentation of the idea as well, and who is doing the presenting, and your own interpretation. It creates a much stronger and healthier skepticism, not a “but on the other hand, but on the other hand, but on the other hand” paralyzed octopus. At no point does post-modernism, as a movement, definitively assert that one hand can’t be more right than the others.

    (Italics are Natalie’s, Bold is mine)

    Stepping back to address your question directly, as someone relatively new to skeptical thought, I definitely think its important to maintain a proper skeptical approach, even to my new-found views. This doesn’t mean I don’t think they are more right than my prior beliefs.

    Regarding Sam Harris’ quote, I think I lean toward your interpretation. I just don’t think that he’s really saying what he appears to be saying.

  • iknklast

    Personally, I suspect uncertainty about a deeply felt belief might be more dangerous, in some ways. If you’re uncertain, but you really want to believe, and feel it’s important to believe, you will want at all costs to avoid someone questioning your belief, because it will add to the uncertainty, and hence the angst you’re feeling. So, you will have a huge investment in silencing critics.

    When someone perceived their opponent (in this case, atheists) as dangerous to something truly important (belief), it seems to me they are much more likely to become dangerous themselves in protecting that belief.

  • Matthew

    I think we are dealing with two different kinds of certainty here.

    The kind Chris is talking about is rational certainty arrived at by consideration of evidence and open to adjustment if contrary evidence presents itself. The other is certainty based on faith without regard for evidence or rationality which (as recent research shows) hardens in the face of contradiction.

    The former is useful and even necessary to operate in day to day life and I think is pretty much uncontroversial, but I am inclined to agree with Sam Harris that the other is inherently dangerous regardless of its social or political efficacy. As soon as a particular set of beliefs are placed beyond reproach it becomes inevitable that people will go to extremes to protect them.

  • Peter Hurford

    As a matter of probability theory, you can’t be 100% certain about something and ever change your mind about it again. There are some things I’m very sure of (heliocentricism), but I don’t want to say absolutely sure, because the protect of never changing my mind ever again no matter what makes me uncomfortable. Thus 99.999999999% is great, and actually significantly different from 100%.

    • Kevin

      I think this is the crucial difference. Someone who is not certain can be ‘rehabilitated’ using evidence and reason, but someone who is absolutely certain cannot be. So, while in the short-term, the person thinking there is a 50-50 chance of God sending non-believers to hell may cause more harm, they can ultimately be reasoned with (assuming they are not absolutely certain that it is 50-50). On the other hand, someone who is absolutely certain cannot be. If you are considering the long-term, then it would make sense that absolute certainty is much more dangerous than the content of beliefs.

  • scenario

    There’s a difference between certainty and absolute certainty. Certainty says that you are sure beyond any reasonable doubt but are still willing to listen to new evidence. Absolute certainty says you are so certain, why bother to listen to any new evidence?

    • Matthew

      Yes. And these are not different as a matter of degree (0.000001% or whatever) but as a matter of kind. Certainty based on faith is fundamentally different from certainty based on evidence. The latter is amenable to adjustment whereas to the former it is anathema.

      Perhaps the proper mathematical perspective would be to describe the faith based form as infinitely certain.

  • Matthew Morse

    There’s a relationship between the question asked here and Libby Anne’s recent post on faith. “Certainty”, as used by Natalie, and “faith”, as used by Daniel, have specific and narrow meanings. In both cases, the meaning is belief without evidentiary support for, or even in spite of contradictory evidence against, the underlying claim.

    Having a high degree of confidence in your beliefs isn’t the problem. The problem is when your certainty or faith prevents you from considering the possibility that you are wrong, and causes you to ignore evidence that challenges your beliefs.

    The certainty of your belief that the Earth goes around the sun is supported by the large quantity of evidence for this fact. You are justified in your certainty, and if someone proposes an alternative model, you are correct to ask that the model account for the evidence you have before you take it seriously. If someone provides some evidence that appears to contradict the motion of the Earth, you are correct to view that evidence skeptically.

    But if you reject that model or that evidence out of hand and refuse to even consider it, at some point you may be more committed to your belief than to the truth, and that is when certainty becomes dangerous.

  • J. Goard

    I think we have to maintain a distinction between the psychological concept of certainty and the mathematical one. We have extremely good evidence that the brain cannot represent even relatively enormous differences in extreme probability values, as demonstrated by people who capriciously throw out a number of nines in their “probability” of a godless world — 99.999999%, 99.9999999999% aren’t ten thousand times different psychologically, but rather represent a tiny psychological difference that either caused or didn’t cause me to punch the key a few more times. I’m with Chris: as long as we’re speaking about human thought, long before we get to such figures we might as well say “certain”.

    The important question is why an abstract epistemological question rarely seems to be trotted out when it comes to knowing the effects of putting out an electrical fire with water, or knowing the most direct route from L.A. to Memphis, or knowing the capital of Peru, but it’s a staple as soon as we’re discussing sky necromancer Yahweh impregnating a virgin with a son who was also himself. Once you concede the relevance of epistemological concerns within such an artfully delineated domain, you’re giving up the ship.

    • Richard Wein

      “I think we have to maintain a distinction between the psychological concept of certainty and the mathematical one.”

      Good point. Bayes theorem is a useful inferential tool, but it’s not a good model of how the brain works.

  • left0ver1under

    Certainty is the cause of all atrocities, whether motivated by religion, politics or ethnicity.

    Every group that has engaged in mass murder, war, slavery, genocide and other crimes “knew” they were right and could act with impunity. That was true of witch hunts, the crusades, the Nazis, the Soviets, the KKK, anti-abortionists, teabaggers, and many others. Bertrand Russell once noted that the reason communism and christianity oppose each other so strongly because they are so similar.

    The most open and tolerant societies were those who accepted and embraced doubt and differences of opinion. When you’re not 100% certain of your own views, you’re more likely to listen to others and allow them to disagree.

  • tonylloyd

    “But the thing I’m really hoping to hear from people on in the comments–is this a question people find interesting?”

    Yes, it’s very interesting. I am on the certainty is dangerous front, although this is a philosophical position:

    - We don’t need any reason to believe anything. We are born, grow up and acquire many of the beliefs of the culture, time and science that we are brought up in.
    - We don’t any reason to retain beliefs; we will continue believing them until we have reason to reassess them.

    So we adopt a proposition, p and keep holding it until one of two things happens:

    1. We find enough wrong with it to reject it outright and adopt ¬p, or
    2. We find a better proposition, p’

    In the case of 2 we keep holding p’ until we find a problem with it, and on.

    What effect does certainty have? The only effect it can have is to stop you rejecting or revising beliefs. Now a true belief will give no reason to be reassessed (a true theory will have no counter-evidence, etc. If the Bible were literally true then we would never experience any conflict between what it describes and the world)

    So we will not reject or revise true beliefs.

    So the only function of certainty is to stop us rejecting or revising false beliefs.

    I think you are correct about the content of beliefs. This is important, but it is not important in contrast to certainty; rather, to be truly dangerous, an idea needs to have bad content and certainty. A belief in Leprechauns can be held with as much certainty as you like, it won’t cause harm. If someone, however, believes that killing all atheists is a moral necessity then it’s very, very, important to change their mind. This you can’t do if they are certain about it.

  • Chris Hallquist

    Okay, these responses are helpful. I think what I’ve learned here is that people attach a lot of different meanings to the word “certainty,” and it’s worth being sensitive to that. In fact, I think I’ll take this as a less about being wary of loaded words in general.

    • Matthew

      Perhaps I’m bit obtuse about this but I still see only two meanings (with numerous applications). One is what I referred to as faith based, “absolute certainty” per scenario and others, and the other evidence based, which we might call “conditional certainty” to contrast with the absolute form.

      Are there others that I’m missing? Forgive me for dwelling on this but I really don’t see any other forms and I do find this taxonomy of certainty quite an interesting topic.

  • Chris Hallquist

    Well there are degree-of-confidence distinctions (99.9999% vs. 100%), distinctions based on the source of the belief (evidence vs. faith), as well as the issue of how ready you are to change your beliefs in response to new evidence. There may be some correlation between the different kinds here, but technically you could get 8 sub-types from that alone.

  • Scott

    You are making the usual assumption that Atheism is the same type of belief system as Theism or any belief structure.
    Lack of belief can not be put in the same category as affirming a faith position, especially a faith system that breeds blind certainty. Their are any billion number of beliefs I do not agree with. I don’t belief in the flying spagetti monster, or an invisible unicorn that follows me everywhere. Or Zues, Krishna, this follows as much for any belief. Atheism is a lack of belief not a belief. Therefore certainty is a mute point unlike say Believing that infidels must be cleanes by faith or fire!.

  • Calvin Wolfe

    “Moral certainty is dangerous” DAH!!!! Can you think of anything in this life that has no danger attached to it…….