TheJach.com

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

I'm slowly working through finishing The Waking Dream, which was assigned in a college class and I read maybe 70% of back then. The thing about it all that sticks with me is the idea of how differently our ancestors used to think about the world, and how remnants of that mode of thought in fact still are common today.

Rationality in the epistemic sense is about seeing the world as clear as you can, mathematically, given what you've seen of the world. This doesn't mean perfection. Rationality is about observing one white swan, then two white swans, then three, and therefore surmising that there could be more swans. And if there are more swans, you should expect the next one's color to be white, with odds of 80% (under Laplace's Rule, $\frac{s+1}{n+2}$) against not-white. A hundred white swans seen, you can expect the next to be white at 99%. This doesn't mean the next one won't be black!

Suppose you did make this into a bet, with two people and yourself as the intermediary. One of them only reasons rationally, the other one reasons in their own irrational way. Both have only seen 100 swans. The rationalist bets $99 that the next swan you bring out is white, the irrationalist bets$1 that it is black, not merely non-white. The rationalist should be pretty happy -- there's a 1% chance the next could be any other color besides white, thus there's even less than 1% chance that it's any specific color! Why is the irrationalist taking this poor bet? They claim because they saw a frog devour 10 flies that morning and that is a symbol that black is coming.

You reveal the next swan: it's black! The irrationalist has won.

If this sort of reasoning could be repeated, it would cleary be better on average to use the sort of reasoning the irrationalist does and read bug signs. It would be more instrumentally rational even.

Unfortunately it doesn't replicate well, but that doesn't mean it never "works", if "works" is really how we can describe it. A better description might just be "got lucky in this particular jump beyond what you can be rationally sure of." But that requires nothing more than imagination.

Suppose you come across a dead bird (or rather its remains) in the road. You can't inspect the bird, all you know is it's dead and it's in the road. What killed it? A rationalist would say it is probably from a car. But could be some predator animal, from the set of predators known to be around the area but most likely not a Wild Dog from Africa.

Turns out it's neither a car nor a predator. It wasn't a heart attack. It wasn't old age. It wasn't from eating a poison. It wasn't some kid throwing a rock at it. It wasn't some other bird colliding with it. It wasn't a lightning strike. It wasn't a particularly strong wind blowing debris into it. It was, in fact, a bullet, shot recreationally from a close by farm and unintentionally hitting it!

All of these were in the possibility space, but it took imagination to bring them out, and a rationalist would admit that each has some small probability of being the cause but wouldn't put much weight on any of them (and indeed wouldn't consider some of them or the vast amount of other possibilities I left out). Now suppose someone had guessed this correctly! We might call them amazing.

Imagination is a fun thing. It helps you to explore a possibility space without relying strictly on what's in front of you or your personal memory. But it also can be dangerous and let you imagine impossible things. Imagine a cube of light. It's not contained by anything, supported by anything, diminishing over time, just a cube of white light floating in the air. You can imagine it, but it's not possible.

Let me start getting to my point. It may help to read this before then. But basically, look at Einstein. Somehow, mysteriously, he got two big theories right, with their initial big experiments coming later, and their toughest big experiments coming much later, that happen to agree with them. How? I can only say it was through an irrational "hunch".

This is very much a Traditional Rationality point. His irrational hunches provided great value to everyone. It would be a mistake to demand everyone working on such problems to justify every little step they make, to justify why they're even pursuing one particular line of thought versus any of the others that are equally justifiable from the evidence everyone can see and understand.

So my point is just the title. Sometimes irrationality can be good. You shouldn't have to justify everything you do with pure rational Bayesian reasoning. Our brains aren't fully introspective enough for that.

Irrational modes of thought may be a great way to enjoy the world. There's no rationality to the idea of spinning a globe (on two axes so you don't have hemisphere bias) and stopping it with your finger and deciding to try and visit the first land mass your finger hits. This doesn't make it a bad idea. It doesn't make it a great idea either, but you can see that it's not clear that it will usually lead to either bad or good results.

Religion and faith are irrational. This doesn't mean religious people are bad, or that they're irrational in all aspects of their lives.

The thing it does mean is that religious advocates should stop trying to frame their faith in the language of rationality, at least without the context of prediction. There is no rational reason to believe in any of that nonsense. There are plenty of irrational reasons. But that alone doesn't tell you whether it's good or bad for you to Believe, or people in general. If you're already an atheist devoted to only keeping beliefs that pay rent, it's worth considering whether it makes sense to pay the sanity cost to reap the benefits of a deadbeat Belief that is shared by lots of people you like and who won't like you as much if you don't share it.

Epistemic rationality having a value judgment is very context-dependent and individually subjective. This is the core of how it can sometimes be instrumentally rational to behave irrationally. The key area where irrational processes probably don't make much sense most of the time is, maybe ironic given my example of Einstein, Science. Science advances in a rather mechanical fashion when looking at the big picture, and it does that because on aggregate scientists believe that their scientific beliefs must be grounded in rationality so as to best describe the world with what we can observe. Experiments are ultimately the arbiter of truth, it doesn't matter how beautiful a theory is if there are no predictions it makes to experiment against, if it doesn't pay rent. (This is in contrast to theories making some predictions you can test, and some you can't, and some you can't even in principle. For example a theory of the universe says there is enough Space that one could travel beyond Earth's light cone and at that point one could never directly interact with Earth / vice versa again, but we don't suppose one stops existing or doesn't interact with things outside the light cone in the same sorts of ways we interact with things inside.)

Because science is so much about finding true things about the world (as opposed to true things about logical systems, i.e. theorems), about predicting future events and then watching them happen precisely as predicted, and has had so much success, with truth being thought of as such a Great Good Thing, faith advocates often say ridiculous things like "I know my faith is true". This means nothing more than "I know my faith is good". If you don't interpret it like that, just how exactly are you parsing the grammatical absurdity? This is like saying "I know my shirt is true" -- what exactly does it mean? Truth and falsehood are results of tests, there needs to be a claim to test before one can assert or verify its truth value. If "my shirt" in "my shirt is true" is a claim, what claim is it? Or is it a set of claims? The same question applies to "faith". Does my shirt exist? For certain definitions of exist, sure. Does my shirt feel nice? Sure. Does my shirt look red? Yeah. Is my shirt heavy? No. Am I really talking about the combination of the shirt fabric with the shirt seams, and so the claims I'm making are that the fabric is joined with seams? Or individually that the seams are of this particular pattern?

I like to think I'm fairly hardcore on trying to keep beliefs that I think I have reason to think are true, or at least that I can say have an arbitrarily high probability of being true, but I don't think everyone should be this way, and I'm not perfectly this way either. I have no explanation why I love the color green so much, or why I used to love the color red more when I was a kid. I don't need to try to justify it. Similarly people who are happy with their faith shouldn't waste time trying to justify it. Just accept it for what it is, a belief and feeling you can only keep (epistemically) irrationally but may provide nice instrumentally rational benefits.

Similarly to atheists, don't waste time mocking them for it. Of course it's irrational. Of course it's false. That doesn't matter. There may still be value, and if you really want to free someone from religion, you need to offer them either a substitute or a higher value. Just look at how Missionaries work, they don't convince people in the truth or rightness of their faith over others, they give a value sell on community or warm thoughts or orderly living or fighting the good fight against Great Satan or whatever.

Of course if predictions are made, and fail, feel free to call them out. Though if they succeed, be prepared to change your mind. If the world stopped for a day in space like was told to have happened in the Bible, while I might still be skeptical of the virgin birth and resurrection and so on, you'd have hooked me pretty hard that I'd start to believe something Godly actually exists.

I'm also not opposed to atheist outreach / combativeness on the evils of religion, because while religion can sometimes be a force of good it can just as easily (and perhaps more often) be a force of evil. The questions of morality are independent of a god actually existing, of the major supernatural claims of a faith actually being true. The only thing I'm opposed to is both sides engaging each other on the rather pointless issue of trying to justify a fundamentally irrational set of beliefs through a rational framework. The best argument religion has come up with is Aquinas' demiurge equivalent, and that was 800 years ago. It's impressive in the age of philosophy, but we are in the age of science, and we need more than just nice-sounding philosophy but evidence, and that means predictions. That religion has not made any successful predictions contingent on the demiurge being real in the last 800 years is evidence that it's a low probability theory. Don't bother engaging with it if you're an atheist, don't bother taking time out of your prayer to dish it out if you're a Believer.

Posted on 2017-04-19 by Jach

Tags: rationality, religion

LaTeX allowed in comments, use $\\...\\$\$ to wrap inline and $$...$$ to wrap blocks.