How to Talk About Antimemes

by lsusr1 min read22nd Dec 201912 comments


Social & Cultural DynamicsRationality
Personal Blog

Every society is wrong about lots of things. Ours is no different. It's difficult to notice when your society is wrong because we learn most things by copying the people around us. By definition, you can't identify the things the people around you are wrong by copying those same people. For this reason, the rules for discussing antimemes differ from the usual rules of rational philosophy.

Rational dialogue is based on mutually agreed-upon facts. The disputed territory is what we can conclude from these facts. But antimemes are definitionally those things that most people are wrong about. So two people are not likely to agree upon the facts. There are several ways around this challenge, all of them bad.

  1. You can state antimemes as you see them. This is bad because two random people will usually disagree about most of the facts. This provokes counterarguments instead of refutations. You end up debating about individual antimemes instead of discussing antimemetics.
  2. You can eschew facts altogether. This is bad because you can't climb the ladder of abstraction down to the bottom rung. A variant of this approach is generalizing from fictional evidence which is just a fancy way to cover up the eschewment of facts.
  3. You can use dead antimemes from other societies and time periods. This is bad because dead antimemes are no longer antimemetic and because it's hard to distinguish symbiotic wars from dead antimemes.

Convincing examples of antimemes are hard to come by even though antimemes are not themselves rare. It's simply more unlikely than not that two random people will recognize exactly the same antimemes. Paul Graham and I probably agree on the antimemetic properties of Lisp but the majority of programmers do not recognize this antimeme for what it is.

When you're talking to an individual person you can restrict the antimeme examples to the small intersection you both agree upon. This is impossible when writing publicly. There's no single antimeme every reader will agree with you about. If you identify several antimemes then it's near certain every reader of your article will disagree with at least one of your examples.

I talk a lot about antimemes so here's a new rule for comments on my articles.

If someone gives multiple examples to support an argument and you agree with any of the examples then just ignore the examples you don't like. If you disagree with what remains then refute that central thesis instead.