Motivating a Semantics of Logical Counterfactuals — LessWrong