Jach's personal blog

(Largely containing a mind-dump to myselves: past, present, and future)
Current favorite quote: "Supposedly smart people are weirdly ignorant of Bayes' Rule." William B Vogt, 2010

Arrogance Labels

I've tried writing about this before, not here, and unsuccessfully, but I think I have a better picture in mind now that I can share. This is about the label "Arrogance", and public perceptions.

On one extreme, you have Christians who believe arrogance is a sin, and they arrogantly confess their humility. They're used as evidence by rational folk for the dangers of religious thinking and hypocrisy.

On the other extreme, you have Objectivists who believe arrogance is a virtue, and they're used as evidence by rational folk for the dangers of arrogance and stupidity.

(Okay, I don't know if any other rational folk believe that, so s/rational folk/myself (who may or may not be rational (you decide!)).)

But what is arrogance, really? It's a label, and like other labels it requires someone to place it. It takes a particularly humble person to recognize arrogance within themselves, and correct for it, it seems that this label is mostly applied to other people, not oneself. It's also most commonly used as a signal for disagreement, and even sometimes used as a counter-argument. "Atheists are just as arrogant as fundamentalists!"

Is there something resembling a solid structure to the adjective "arrogant" that we can definitively suggest a person is arrogant or not? I tend to think of arrogance as a projected attitude of overconfidence, so here's my take: an arrogant person is one whose internal subjective probability for some X is probably skewed in favor of X more than it rationally should be by Bayes' Theorem. The phrase "probably skewed" emphasizes that this judgment is based on another observer with their own internal probabilities affected by the "arrogant person's" personality projection, and secondly in most human cases "by Bayes' Theorem" is substituted with "by my own internal probability." Thus a moderate religious person may think an atheist is arrogant in denouncing God because their internal probability of God existing is quite high, especially if the atheist projects an attitude of certainty about it, even if by Bayes' Theorem after taking into account positive and negative evidence along with Occam's Razor seems to indicate the atheist is correct and justifiably certain. Of course, most atheists don't hold sophisticated models of probability theory (and I'm still learning, myself), so they may be just as culpable as a religious person.

Here it is important to note that humility and arrogance are not opposites: humility is supposed to be the natural state. Thus a humble person, in the same spirit, is one whose internal probability for some X is probably about right to what it rationally should be. Insert rationalist desiderata about desiring a map that reflects the territory, and it's easy to see why a humble state is better than an arrogant state. Underconfidence is the opposite of arrogance, and that too is not desirable.

So, are the Christians right? Is arrogance a sin, or at the very least, should arrogance be avoided? I think it's clear that having irrational probability estimates should be avoided, not at all costs, but wherever feasible, since underconfidence can be just as deadly as overconfidence, and so it's clear that arrogance is not a virtue, nor is it ever "justified" since by its definition it doesn't accurately reflect the territory. But like all moral questions, it's not so clear cut, not so black and white, to say arrogance is a sin and should always be avoided. We have a large capacity for arrogance, after all, it survived evolution.

I've read of economics professors who state that a rational person is risk-adverse. While I disagree, because a rational person should win, and sometimes winning requires risks, this can help illustrate my point that arrogance is not always undesirable. If it's good to be risk-adverse, it's going to be hard to become rich. There's another rule of economics, which says "the greater the risk, the greater the potential returns." And so by being overconfident, by taking risks, you have a larger potential for greater returns. When many people are taking risks, at least one of them will probably reap the higher rewards. Even though it's incredibly stupid to enter a multi-million dollar lottery, someone still ends up winning. And society needs these people, because without them we can't advance.

Let me further illustrate with Einstein. Einstein's Special Relativity (not to mention General Relativity) was an arrogant theory, but it turned out to be correct, and we're better off because of it. Sure, we would have gotten there eventually without him, if we only followed the risk-adverse model and found overwhelming evidence that non-special-relativity models are broken, but how long would that have held us back? And what about Grand Unification Theory? While I don't have much taste for string theory, at least they're trying, instead of waiting who knows how many years for better tools to be constructed. General Relativity was born out of theory, verified by experiment. Ideas born out of experiment are necessarily correct from the outset, but typically take much longer to reach.

Inversely, underconfident people are useful as well. No, not skeptics, who aren't confident in others, but unconfident people, who aren't confident in themselves to the point of underconfidence. These are the programmers who like to have tests for everything, or (hating tests) like to code major recipes in a REPL before integrating into a larger project. Without these people, certain sanity checks would be overlooked even by humble people, and that can lead to disaster just as surely as them being overlooked by arrogant people. The problem with underconfidence is that it can allocate too little time to progress and can indeed poison itself by considering another model that is probably no better, possibly worse, than the current one.

What's my point in all this? That arrogance is sometimes okay, underconfidence is sometimes okay, and humility, while the best for an accurate model, is nevertheless no surefire solution toward desirable consequences. We only care if a bridge engineer was arrogant, humble, or underconfident when the bridge collapses and kills people. Were he arrogant, we attribute his overconfidence to the bridge collapsing, and go after him personally. The media destroys him. Were he humble, we may investigate to see if there was a case for negligence, he'll probably find it hard to get work anywhere again, we'll chastise him to be more careful and other engineers as well, to be extra careful. Were he already extra careful, were every test conceivable run, we call it a tragedy, and begin questioning the very models themselves rather than a person's misuse and abuse of the models.

My sub-point is a repetition from the beginning: arrogance labels are typically applied to other people, and not by a rational process. If someone calls you arrogant, it may be worth considering whether you have an accurate map, and it would be a humble thing to do so in most cases, but don't let it get to you. I believe that projected attitude is still the biggest factor in a human's assignment of "arrogant" or "not arrogant" to another, and if you come across as certain, even if it's rational to do so, then let them call you arrogant and let you not be bothered.

Posted on 2010-11-01 by Jach

Tags: psychology, religion, thought


Trackback URL:

Back to the top

Back to the first comment

Comment using the form below

(Only if you want to be notified of further responses, never displayed.)

Your Comment:

LaTeX allowed in comments, use $$\$\$...\$\$$$ to wrap inline and $$[math]...[/math]$$ to wrap blocks.