It sometimes happens with fanatics that people are certain despite being wrong. But that is no basis to generalise about a link between certainty and error. Generally they are opposite.

I agree with scotchbooks here Robert. When you are certain of something, you are less likely to change your mind when you come across disconfirming data, all other things being equal.

If you have a formula by which certain beliefs are measures critically, you are still better off with a touch of uncertainty, to 'keep the door open'. The door doesn't need to lead to false beliefs. Instead, it can be a necker-cube style perception shift that wouldn't have happened otherwise. If you are certain of the truth of something, you can still benefit by gaining a better conceptual understanding, which would be limited if you are locked in a cell of certainty. Sometimes parts of a belief seem to go retrograde during a perspective shift, but the concept as a whole benefits.

Actually Robert, there seems to be considerable evidence that feelings of certainty "stem from primitive areas of the brain and are independent of active, conscious reflection and reasoning." For the evidence and arguments in favor of this I refer you to the work of Robert A. Burton, a UCSF neuroscientist, as detailed in his book "On Being Certain: Believing You Are Right Even When You're Not" from which the above quote was selected.

Once we become certain of the truth of a political or religious belief it becomes almost invulnerable to rational or logical refutation. This will become Haidt's position as well when he discusses what can be done to mediate the bitter disputes splitting our people and political parties today.

Incidentally, even scientific theories such as those of Darwin are never finished products or facts to which certainty can be attached. They are, at best, highly probable explanations with at least a wink of doubt. Consider the 5 sigma standard for discovery of the Higg's boson. Still some doubt.

Robert's certainty about certainty makes it harder for me to dismiss his certainty! Maybe that in itself points to an original evolutionary advantage to our feeling of certainty (and thanks, by the way, to scotchbooks for bringing up Robert Burton's book, which we read here several years ago). It's easy to imagine that getting that feeling of absolute conviction would be a spur to decisive action that could aid in survival. In other spheres of life, such as social leadership, there could be an advantage to anyone who who displayed such a quality, because in times of peril or need we look to people who tell us they are absolutely sure of what to do. We may feel our own evidence is too limited for us to know ourselves, but we know something should be done, and the guy who shows this certainty, well he must have it for a good reason, or so we may believe. Perhaps this cocksure leader screws up, perhaps he doesn't, but we seem not to be able to help give credence to his feeling of certainty, unless we have strong evidence of our own, or an opposing leader who is certain in his own right.

I too can believe that certainty is not primarily a forebrain, rational product. I think when we talk of wanting to see a "core of conviction" in a leader, we're talking about the elephant. We want to see that this person has that solid, instinctual quality of always knowing who he is and what is the right thing to do. We want that to remain even in the face of rational arguments that might sway a less anchored personality. Or at least I would want to argue that we are conflicted about certainty, valuing it in people yet also valuing the ability to change course when disconfirming evidence presents itself.

scotchbooks wrote:Actually Robert, there seems to be considerable evidence that feelings of certainty "stem from primitive areas of the brain and are independent of active, conscious reflection and reasoning." For the evidence and arguments in favor of this I refer you to the work of Robert A. Burton, a UCSF neuroscientist, as detailed in his book "On Being Certain: Believing You Are Right Even When You're Not" from which the above quote was selected.

As with Haidt, I followed the discussion on Burton here with interest, although sadly not finding time to read the book, but chipping in on the summary comments from my perspective.

It is worth going back to the recent discussion here about Plato to explain my view on certainty. Plato held that we can only be certain of knowledge that arises from intelligence, and that beliefs which arise from appearances are never certain. This means that political and religious convictions which lack an intelligible rational frame should not be classed as certainty, despite the fervour with which they are held.

Intelligence tells us that only core scientific facts can be indubitably certain. Astronomy is the paradigm, with our knowledge of facts about the universe sitting together so coherently that doubting them arises only from stupidity or ignorance. But at the frontier, science acknowledges its uncertainty: the fact that we cannot explain dark matter does not mean we should entertain doubt about whether the inverse square law of gravity predicts planetary positions.

For any other claims, they need to match the gold standard of astronomical knowledge to be classed as certain. Religion and politics just don't cut it, due to the obvious existence of large scale delusion and error in these fields.

Once we become certain of the truth of a political or religious belief it becomes almost invulnerable to rational or logical refutation. This will become Haidt's position as well when he discusses what can be done to mediate the bitter disputes splitting our people and political parties today.

What I don't like here is the apparent elision from the observation that social certainty is illusory to the claim that scientific certainty is similarly unreliable. This line of thinking leads to a mad sort of solipsism where we are not even sure the universe exists.

Incidentally, even scientific theories such as those of Darwin are never finished products or facts to which certainty can be attached. They are, at best, highly probable explanations with at least a wink of doubt. Consider the 5 sigma standard for discovery of the Higg's boson. Still some doubt.

The fact that science is incomplete does not imply that it contains no certain knowledge. Uncertainty about the Higgs boson should not imply uncertainty about the existence of atoms. Atoms absolutely exist as a necessary condition of experience. It is a fallacy to say that not knowing everything means we know nothing.

A short but excellent article written by Haidt that lays out clearly and succinctly his position can be found here: http://scholar.google.com/scholar_url?h ... i=scholarr

Incidentally, I must object to the claim that Hume had no sense of moral duty. Hume introduced the notion of utility into moral theory; we approve of moral acts in part because they have utility – they are useful. But, in a section titled Why Utility Pleases (Section V of Hume, 1751), he argues that we approve of such useful actions because of our ability to sympathize with the recipients of those actions. He also believed that the highest merit a human can achieve was benevolence.

Astronomy is the paradigm, with our knowledge of facts about the universe sitting together so coherently that doubting them arises only from stupidity or ignorance.

I must be stupid or ignorant. Or perhaps I have a different conceptual definition of what 'certainty' is, or what 'doubt' is.

Uncertainty about the Higgs boson should not imply uncertainty about the existence of atoms.

The casual simplicity of words makes some things unquestionable. But what are you referring to when you say "atom"? Are you certain you fully understand what an atom is? Isn't the word merely a placeholder term for the collective system of even smaller units - quarks? Can you not entertain the possibility that atoms are a higher-level manifestation of a currently misunderstood phenomenon at a scale so small we can't detect it? If the way you define the word 'atom' is different from the truth, you can still be certain that "something of the type" exists. But then, of what use is such a claim to science?

Certainty is a silly position. What harm is there to doubt the existence of atoms, if only to see what sort of creative hypotheses pop into your head? Is this a path that's prohibited to students of physics? My problem with certainty isn't so much it's status with regards to knowledge. The problem is the walls it erects in your head.

In the end, we could still be playing out a script in the videogame of an extremely advanced civilization. Do you believe the chance is infinitesimal, or that it's certaintly impossible? There's an infinite difference between the two.

That is a great example, interbane.

In common use, we are safe to say that atoms "certainly" exist. But that's true... what exactly do you mean by atom? We are roughly sure what atoms are, and we might point to a diagram with a knot of blue and red balls surrounded by what look like tiny planets zooming around in orbit. But we know that electrons, the orbiting bits, don't actually orbit like that. There definitely is uncertainty about what exactly atoms are. Protons are composed up quarks and so far we think that's the end of it, but does that mean we can dismiss data that tells us there might be something even more basic than the quark?

This is always why i like to frame my knowledge in terms of confidence. I am very confident that our general picture of the atom is accurate. I am more confident in that than i am that my car will be in the driveway in the morning. But that does admit the possibility that i'm wrong. That there are details i don't know.

Saying that you are extremely confident, rather than certain, does not mean you have to treat the atom model as being equally likely as the legitimacy of fairy dust. It just admits the addition of new details so far unknown, or corrections to the theory which had so far gone overlooked.

Think of Einstein's improvement on Newtonian gravity. These are essentially the same thing for everyday occurances. But when dealing with big space, big time, and high energies, it makes a big difference which formulation of the equation you use.

This is really a matter of being didactic it think. I still have no problem telling people that they are definitely wrong when they are definitely wrong. Admitting to the fallibility of human thought does not cede the high ground to those who claim baseless certainty.

There are some things that don't work well with people's brains.

One good example is the different between a face and a name. We can remember a face for years after a single brief exchange, but somehow manage to forget what the person's name is within that very same exchange. The name is an abstraction.

In other words, we're good with experiential pattern recognition, but bad with abstraction. The difference is incredible. We can remember an hour's worth of gossip, but not remember the simple 7 digit phone number used to make the call. The number is an abstraction.


Another area where it seems people are wired a specific way is to think dichotomously. Yes/no, true/false.

Add a dash of complexity to your thought style, and you have an couple of extra points on the gradient. Perhaps "I don't know" is situated halfway between certainly true and certainly false.

But how many people do you know have even more points on the spectrum. Different levels of "higher confidence"? How many people weigh the effect of more than two causal inputs influence on any given phenomenon?

I don't think we're optimized to think in terms of confidence levels or gradients. Yes/no, true/false style thinking is dominant. It's all too human to polarize and jump on bandwagons.

Interbane, relevant to your comment: There is an area of the brain called the fusiform face area that preferentially processes faces. Recognizing individuals is crucial to survival in most species; you need to remember who is friend and who is foe, who is kin and who is not. So not surprisingly, there is special circuitry to handle that. Damage to the FFA yields prosopagnosia--inability to recognize faces, even one's own face in the mirror.

Anthropologist Robin Dunbar has argued that, in humans, gossip replaced mutual grooming for building social bonds. Primates and other social species spend a very large proportion of their time grooming each other. We don't do that, but (as Dunbar has argued) we spend an equivalent amount of time gossiping, and for the same reason.

Regarding yes/no dichotomy: It turns out that if you do studies where people are required to make yes/no decisions, they find it more difficult than giving a graded response. So, for example, pollsters usually ask you if you to rate on a scale of 1 to 10 how strongly you agree with a political statement or policy. If you ask people to make yes/no decisions, they will try to hedge, and you'll get something like "Well, I think so, yes, if the circumstances are right."

We do tend to polarize, though, when it comes to things that identify which group we belong to. For example, in some studies, the experimenters had people choose whether they preferred Kandinsky or Klee paintings. Then they had them engage in various tasks, like dividing money among participants, or scoring essays, and so on. Even when group membership was identified on the basis of something as ridiculously simple and made-up-on-the-fly as artist preference, people show extreme in-group bias, favoring others that are members of their own group by giving them more money or better scores.

denisecummins wrote:A short but excellent article written by Haidt that lays out clearly and succinctly his position can be found here: http://scholar.google.com/scholar_url?h ... i=scholarr

Incidentally, I must object to the claim that Hume had no sense of moral duty. Hume introduced the notion of utility into moral theory; we approve of moral acts in part because they have utility – they are useful. But, in a section titled Why Utility Pleases (Section V of Hume, 1751), he argues that we approve of such useful actions because of our ability to sympathize with the recipients of those actions. He also believed that the highest merit a human can achieve was benevolence.

Thanks for posting the article. For those without the time or inclination to read The Righteous Mind, it's a good precis of the book. Haidt added a sixth moral foundation in the book: the Liberty/oppression foundation.