Letting it go: when you shouldn’t respond to someone who is wrong

I’m requesting that people follow a simple guide when determining whether to respond to a post. This simple algorithm should raise the quality of discussion here.

  • If you care about the answer to a question, you will research it.

  • If you don’t care about the answer, don’t waste people’s time by arguing about it, even if someone’s post seems wrong.

  • If you don’t care and still want to argue, do the research.

Why should you follow these rules?

Fairness.

It takes very little effort to post a contradictory assertion. You just have to skim a post, find an assertion (preferably one that isn’t followed or preceded immediately by paragraphs of backing evidence, but that’s an optional filter), and craft a sentence indicating that that assertion is wrong or flawed. Humans can do this almost by instinct. It’s magical.

Refuting a contradiction takes effort. I typically spend at least five minutes of research and five minutes of writing to make a reply refuting a bare contradiction when I have already studied the issue thoroughly and know which sources I want to use. I go to this effort because I care about these statements I’ve made and because I care about what other people believe. I want to craft a reply that is sufficiently thorough to be convincing. And, I’ll admit, I want to crush my opponents with my impeccable data. I’m a bit petty sometimes.

If I haven’t researched the issue well—if my sources are second-hand, or if I’m using personal experience—I might spend two hours researching a simple topic and ten to fifteen minutes creating a response. This is a fair amount of time invested. I don’t mind doing it; it makes me learn more. It’s a time investment, though.

So, let’s compare. Half a second of thought and two minutes to craft a reply containing nothing but a contradiction, versus two hours of unpaid research. This is a huge imbalance. Let’s address this by trying to research people’s claims before posting a contradiction, shall we?

Trust.

You are convinced that someone’s argument is flawed. This means that they have not looked into the issue sufficiently, or their reasoning is wrong. As a result, you can’t trust their argument to be a good example of arguments for their position. You can look for flaws of reasoning, which is easy. You can look for cases where their data is misleading or wrong—but that requires actual effort. You have to either find a consensus in the relevant authorities that differs from what this other person is saying, or you have to look at their specific data in some detail. That means you have to do some research.

Community.

If you want people to stick around, and you’re brusquely denying their points until they do hours of work to prove them, they’re going to view lesswrong as a source of stress. This is not likely to encourage them to return. If you do the legwork yourself, you seem knowledgeable. If you’re careful with your phrasing, you can also seem helpful. (I expect that to be the tough part.) This reduces the impact of having someone contradict you.

Advancing the argument.

From what I’ve seen, the flow of argument goes something like: argument → contradiction of two or three claims → proof of said claims → criticism of proof → rebuttal → acceptance, analysis of argument. By doing some research on your own rather than immediately posting a contradiction, you are more quickly getting to the meat of the issue. You aren’t as likely to get sidetracked. You can say things like: “This premise seems a bit contentious, but it’s a widely supported minority opinion for good reasons. Let’s take it as read for now and see if your conclusions are supported, and we can come back to it if we need to.”

Bonus: “You’re contradicting yourself.”

Spoiler: they’re not contradicting themselves.

We read here a lot about how people’s brains fail them in myriad interesting ways. Compartmentalization is one of them. People’s beliefs can contradict each other. But people tend to compartmentalize between different contexts, not within the same context.

One post or article probably doesn’t involve someone using two different compartments. What looks like a contradiction is more likely a nuance that you don’t understand or didn’t bother to read, or a rhetorical device like hyperbole. (I’ve seen someone here say I’m contradicting myself when I said “This group doesn’t experience this as often, and when they do experience it, it’s different.” Apparently “not as often” is the same as “never”?) Read over the post again. Look for rhetorical devices. Look for something similar that would make sense. If you’re uncertain, try to express that similar argument to the other person and ask if that’s what they mean.

If you still haven’t found anything besides a bare contradiction, a flat assertion that they’re contradicting themselves is a bad way to proceed. If you’re wrong and they aren’t contradicting themselves, they will be annoyed at you. That’s bad enough. They will have to watch everything they say very carefully so that they do not use rhetorical devices or idioms or anything that you could possibly lawyer into a contradiction. This takes a lot more effort than simply writing an argument in common modes of speech, as everyone who’s worked on a journal article knows.

Arguing with you is not worth that amount of effort. Don’t make it harder than it needs to be.