I always thought the talking snakes argument was very weak, but being confronted by a very weird argument from a young-earth creationist provided a great example for it:
If you believe in evolution, why don’t you grow wings and fly away?
The point here is not about the appeal to ridicule (although it contains a hefty dose of that too). It’s about a gross misrepresentation of a viewpoint. Compare the following flows of reasoning:
Christianity means that snakes can talk.
We can experimentally verify that snakes cannot talk.
Therefore, Christianity is false.
and
Evolution means people can spontaneously grow wings.
We can experimentally verify that people cannot spontaneously grow wings.
Therefore, evolution is false.
The big danger in this reasoning is that one can convince oneself of having used the experimental method, or of having been a rationalist. Because hey, we can scientifically verify the claim! - Without realizing that the verified claim is very different from the claims the discussed viewpoint actually holds.
I’ve even seen many self-proclaimed “rationalists” fall into this trap. Just as many religious people are reinforced by a “pat on the back” from their peers if they say something which is liked by the community they are in, so can people feel motivated to claim they are rationalists if that causes a pat on the back from people they interact with the most.
I know someone who told me that she hoped President Trump wouldn’t successfully legalize rape.
I opined that, perhaps, that might not be on his itinerary.
She responded that, of course it was, and proved it thus:
She is against it.
She is against him.
Therefore he is for it.
I bring this up to sort of angle at ‘broadening’ the talking snakes point. The way people do arguments is kind of how they do boggle. Find one thing that is absolutely true, and imply everything in its neighborhood. One foot of truth gives you one mile of argument.
As soon as you have one true thing, then you can retreat to that if anyone questions any part of the argument. Snakes can’t talk, after all.
She responded that, of course it was, and proved it thus: She is against it. She is against him. Therefore he is for it.
That sounds so extremely stupid that I have to ask: Is that literally, in so many words, what she said, or is it possible that she said something (still presumably stupid but) a bit less stupid, and there’s a slight element of caricature in your presentation of what she said?
I’m distilling, sure. The actual text was something like:
Jane: I’m just so concerned and terrified about that awful man being elected. I shudder to think what he’ll do.
Walter: (distracted) nods
Jane: I doubt he’ll manage to legalize rape, but...
Walter: (tuning into conversation fully) wat?
Dan: There’s a whole government, they’ll stop him.
Walter: I don’t think Donald Trump is trying to legalize rape.
Jane: You think the best of people, but you have to open your eyes. He is in bed with all kinds of people. He mocked a retarded person on national tv. He is taking money from the russians.
Walter: Even if that’s true, I still don’t think he’s for legalizing rape.
Dan: Listen, if you were right, then he never would have called all Mexicans rapists, or proposed that we wall them out!
Walter: nods, tuning back out
Conversation continues in a precious bodily fluids direction.
It stuck in my mind as a perfect example of the whole “arguments are soldiers” thing. Lesswrong is always more correct than I imagine.
It helps that you shared the dialogue. I predict that Jane doesn’t System-2-believe that Trump is trying to legalize rape; she’s just offering the other conversation participants a chance to connect over how much they don’t like Trump. This may sound dishonest to rationalists, but normal people don’t frown upon this behavior as often, so I can’t tell if it would be epistemically rational of Jane to expect to be rebuffed in the social environment you were in. Still, making claims like this about Trump may be an instrumentally rational thing for Jane to do in this situation, if she’s looking to strengthen bonds with others.
Jane’s System 1 is a good bayesian, and knows that Trump supporters are more likely to rebuff her, and that Trump supporters aren’t social allies. She’s testing the waters, albeit clumsily, to see who her social allies are.
Jane could have put more effort into her thoughts, and chosen a factually correct insult to throw at Trump. You could have said that even if he doesn’t try to legalize rape, then he’ll do some other specific thing that you don’t approve of (and you’d have gotten bonus points for proactively thinking of a bad thing to say about him). The implementation of either of these changes would have had a roughly similar effect on the levels of nonviolence and agreeability of the conversation.
This generalizes to most conversations about social support. When looking for support, many people switch effortlessly between making low effort claims they don’t believe, and making claims that they System-2-endorse. Agreeing with their sensible claims, and offering supportive alternative claims to their preposterous claims, can mark you as a social ally while letting you gently, nonviolently nudge them away from making preposterous claims.
Generally speaking, when a person says “X”, they rarely mean X. They usually mean one of the things they associate with X, usually because it is what their social group associates with X.
For example, when someone says “make trains run on time”, they mean Hitler. (Even if the quote actually comes from Mussolini. No one cares about Mussolini, but everyone knows Hitler, so for all practical purposes, Mussolini is Hitler, and fascists are nazis. If you don’t trust me, just ask a random person whether Hitler’s followers were fascists.) Why don’t they simply use the word “Hitler” when they want to talk about Hitler? Because “Hitler” actually means: a bad person. So when you want to talk about Hitler specifically, as opposed to talking about a generic bad person, you must say “make trains run on time”.
This is how normies talk all the time. If you take how normies talk, and add a lot of penises, you get Freudian psychoanalysis. If instead of penises you use holy texts, you get kabbalah. This explains why both of them are so popular among normies who want to know the deep truths about the universe.
This comment was very insightful, and made me think that the young-earth creationist I talked about had a similar motivation. Despite this outrageous argument, she is a (relatively speaking) smart and educated person. Not academic-level, but neither grown up on the streets level.
I always thought the talking snakes argument was very weak, but being confronted by a very weird argument from a young-earth creationist provided a great example for it:
The point here is not about the appeal to ridicule (although it contains a hefty dose of that too). It’s about a gross misrepresentation of a viewpoint. Compare the following flows of reasoning:
Christianity means that snakes can talk.
We can experimentally verify that snakes cannot talk.
Therefore, Christianity is false.
and
Evolution means people can spontaneously grow wings.
We can experimentally verify that people cannot spontaneously grow wings.
Therefore, evolution is false.
The big danger in this reasoning is that one can convince oneself of having used the experimental method, or of having been a rationalist. Because hey, we can scientifically verify the claim! - Without realizing that the verified claim is very different from the claims the discussed viewpoint actually holds.
I’ve even seen many self-proclaimed “rationalists” fall into this trap. Just as many religious people are reinforced by a “pat on the back” from their peers if they say something which is liked by the community they are in, so can people feel motivated to claim they are rationalists if that causes a pat on the back from people they interact with the most.
I know someone who told me that she hoped President Trump wouldn’t successfully legalize rape.
I opined that, perhaps, that might not be on his itinerary.
She responded that, of course it was, and proved it thus: She is against it. She is against him. Therefore he is for it.
I bring this up to sort of angle at ‘broadening’ the talking snakes point. The way people do arguments is kind of how they do boggle. Find one thing that is absolutely true, and imply everything in its neighborhood. One foot of truth gives you one mile of argument.
As soon as you have one true thing, then you can retreat to that if anyone questions any part of the argument. Snakes can’t talk, after all.
That sounds so extremely stupid that I have to ask: Is that literally, in so many words, what she said, or is it possible that she said something (still presumably stupid but) a bit less stupid, and there’s a slight element of caricature in your presentation of what she said?
I’m distilling, sure. The actual text was something like:
Jane: I’m just so concerned and terrified about that awful man being elected. I shudder to think what he’ll do.
Walter: (distracted) nods
Jane: I doubt he’ll manage to legalize rape, but...
Walter: (tuning into conversation fully) wat?
Dan: There’s a whole government, they’ll stop him.
Walter: I don’t think Donald Trump is trying to legalize rape.
Jane: You think the best of people, but you have to open your eyes. He is in bed with all kinds of people. He mocked a retarded person on national tv. He is taking money from the russians.
Walter: Even if that’s true, I still don’t think he’s for legalizing rape.
Dan: Listen, if you were right, then he never would have called all Mexicans rapists, or proposed that we wall them out!
Walter: nods, tuning back out
Conversation continues in a precious bodily fluids direction.
It stuck in my mind as a perfect example of the whole “arguments are soldiers” thing. Lesswrong is always more correct than I imagine.
It helps that you shared the dialogue. I predict that Jane doesn’t System-2-believe that Trump is trying to legalize rape; she’s just offering the other conversation participants a chance to connect over how much they don’t like Trump. This may sound dishonest to rationalists, but normal people don’t frown upon this behavior as often, so I can’t tell if it would be epistemically rational of Jane to expect to be rebuffed in the social environment you were in. Still, making claims like this about Trump may be an instrumentally rational thing for Jane to do in this situation, if she’s looking to strengthen bonds with others.
Jane’s System 1 is a good bayesian, and knows that Trump supporters are more likely to rebuff her, and that Trump supporters aren’t social allies. She’s testing the waters, albeit clumsily, to see who her social allies are.
Jane could have put more effort into her thoughts, and chosen a factually correct insult to throw at Trump. You could have said that even if he doesn’t try to legalize rape, then he’ll do some other specific thing that you don’t approve of (and you’d have gotten bonus points for proactively thinking of a bad thing to say about him). The implementation of either of these changes would have had a roughly similar effect on the levels of nonviolence and agreeability of the conversation.
This generalizes to most conversations about social support. When looking for support, many people switch effortlessly between making low effort claims they don’t believe, and making claims that they System-2-endorse. Agreeing with their sensible claims, and offering supportive alternative claims to their preposterous claims, can mark you as a social ally while letting you gently, nonviolently nudge them away from making preposterous claims.
Generally speaking, when a person says “X”, they rarely mean X. They usually mean one of the things they associate with X, usually because it is what their social group associates with X.
For example, when someone says “make trains run on time”, they mean Hitler. (Even if the quote actually comes from Mussolini. No one cares about Mussolini, but everyone knows Hitler, so for all practical purposes, Mussolini is Hitler, and fascists are nazis. If you don’t trust me, just ask a random person whether Hitler’s followers were fascists.) Why don’t they simply use the word “Hitler” when they want to talk about Hitler? Because “Hitler” actually means: a bad person. So when you want to talk about Hitler specifically, as opposed to talking about a generic bad person, you must say “make trains run on time”.
This is how normies talk all the time. If you take how normies talk, and add a lot of penises, you get Freudian psychoanalysis. If instead of penises you use holy texts, you get kabbalah. This explains why both of them are so popular among normies who want to know the deep truths about the universe.
This comment was very insightful, and made me think that the young-earth creationist I talked about had a similar motivation. Despite this outrageous argument, she is a (relatively speaking) smart and educated person. Not academic-level, but neither grown up on the streets level.