Why Do We Believe

post by Screwtape · 2023-02-17T20:58:17.387Z · LW · GW · 3 comments

Summary: The fundamental question of rationality is, why do you believe what you believe? In this exercise, we ask that question repeatedly in an attempt to figure out our foundations.

Tags: Repeatable, small, experimental

Purpose: The purpose is twofold. Firstly, we want to grow comfortable and familiar with figuring out where our beliefs come from. Second, we want to do some sleuthing around beliefs we maybe aren’t sure about.

Materials: Paper and pencils are useful. A device capable of accessing wikipedia is invaluable.

Announcement Text: “I ask you the fundamental question of rationality: Why do you believe what you believe?”

There are lots of things we think are true. Many of them are actually true! In this meetup, we’re going to state some beliefs about the world and then step through why we hold that this is true – or if we disagree, why we hold that this is false! We’ll list our reasons, and then consider whether those reasons are credible. The intention is to practice on easy targets, but that doesn’t mean that the easy answers of “well everyone knows that” or “its obvious” are going to pass!

Description: 

1. As the organizer, state a fact about the world. Examples include “The moon is made of blue cheese.” “The earth is basically a sphere.” “Cars run on faerie power.” “Things fall downwards because of gravity, which accelerates objects at about 9.81 m/s^2.” “I am Abraham Lincoln.” “Humans take oxygen from every inhalation and exhale carbon dioxide.” Mathematical nerd-sniping can be good too if you have a group that will enjoy it.

2. Next, each person writes down why they believe the fact to be true or false. This may be “because I did this experiment myself” or “because these people told me” or something else. You can write down more than one reason! Once you’re done, find a partner. 

3. With your partner, take turns trying to poke holes in each other’s reasons. (“Maybe the experimental evidence provides an alternate explanation, like X.” “Maybe the person who told you that isn’t really that trustworthy.”) Write down the possible holes. For each reason with holes, repeat step two, writing down why this hole doesn’t convince you or why it does convince you this isn’t a good reason.

4. When you and your partner are satisfied, go to step 1 with a different fact.

There isn’t a way to lose Why Do We Believe. If you ever realize that you are wrong, you instantly change sides to the ‘winning’ team with the truth. However, you can still strive to play well! Come up with the strongest possible reasons, and the strongest possible holes in your evidence.

If you want to win bonus points however, go out and perform an experiment. It doesn’t need to be a big official experiment to count!

Variations: 

Setting yourself up somewhere with more space to move about and more materials designed for replicating physics experiments sounds like fun. Someday I want to run a "So you think the world is flat? Let's prove it!" meetup because playing with lasers and sundials sounds like a good summer evening. The glaring flaw here is that seeing the physics experiment with their own eyes might not be the key reason someone thinks this might be wrong. 

If your group has a common interest other than rationality, that interest is a good source of facts. Investigate how best to draw perspective correctly, how best to make a magic deck curve out in monoblack, or what the effective distance of a hockey slapshot actually is.

Hardmode: make the fact a controversial one, where there is genuine uncertainty in your own heart and people are likely to have assumptions about the answer. I recommend against doing this as your first few examples, possibly not on the first time you run this. In a martial arts class you do not start to learn a new technique by sparring with it.

Notes: Try not to drive the reasons people come up with. You might expect certain answers, and it’s cool to anticipate the test someone will want to run so you can have supplies ready, but there’s no point testing reasons that nobody objects to. 

Timing can vary wildly. If it is much too short (say, less than two minutes) that’s a sign people might want to try harder with poking holes in reasons. Still, “I believe it because everyone around me says its true” can be an honest answer and that doesn’t take long to resolve. You care more about getting the answers people actually believe than you care if those are good reasons or not. If people are excited about a chain of evidence or have dived into devising an experiment, that’s great! Don’t feel like the whole group has to stay on the same questions at the same time.

Credits: Recommended reading is Scientific, Legal, and Rational evidence [LW · GW] and Fake Explanations [? · GW] but honestly this is an outgrowth of “what do you think you know and how do you think you know it?” all of which I got from Eliezer Yudkowsky.

3 comments

Comments sorted by top scores.