fuzzy post-it

this is something i realized bothered me about gamergate, but in recent days i’ve discovered it applies far more broadly

usually, “begging the question” means making an argument that assumes the conclusion, then uses it to prove the conclusion. it’s a flavor of circular reasoning.

the particular subflavor i’ve commonly seen in gamergate looks like this:

leigh alexander said “hood men”, therefore she is a racist.

at first glance, this looks like X → Y. airtight logic.

but wait, it actually is not. it is possible for the left-hand side to be true, without the right-hand side being true. if, for example, “hood men” were not a slur, or if leigh alexander were black. you know. hypothetically.

so let’s rephrase that argument a little more accurately:

leigh alexander is a racist. leigh alexander said “hood men”, confirming she is a racist.

see, the argument only makes sense if you are already assuming the conclusion, and see the new piece of evidence as confirmation of it.

this doesn’t say that X implies Y. it says that X is compatible with Y. but that’s true for zillions of things. “the sky is blue, therefore the earth is surrounded by a thick shell of water.” that’s completely ridiculous, but if you already believe the conclusion, it sounds perfectly reasonable.

so it’s not really X → Y, or “X implies Y”. it looks more like: Y → (X → Y). “i’m assuming Y, so X confirms Y.”

which leads me to this fun fact:

Y → (X → Y)
¬Y ∨ (X → Y)
¬Y ∨ (¬X ∨ Y)
(¬Y v Y) ∨ ¬X
true ∨ ¬X
true

it turns out that this style of argument is a tautology — by the very strict and simplistic rules of symbolic logic, it can never be wrong. in fact, X can be anything at all! which means that if you already have a conclusion in mind, literally anything can be used to back it up. all you have to do is phrase it in a way that sounds good to other people who already agree with you and tada, “proof”.

in the light of dozens of “i knew it” tags, this seemed especially relevant.