@Meaningness @cwage Okay, why do we need quantification?
-
-
Replying to @GrumplessGrinch
@GrumplessGrinch@cwage If you want to say “any time you see a snark, it might be a boojum” you need quantification.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@GrumplessGrinch@cwage Because this is a universally quantified (“for all”) statement; it’s not just about a particular snark.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@GrumplessGrinch@cwage PT can only say P(boojum(s0)|snark(s0))>0, for some particular snark s0.1 reply 0 retweets 1 like -
Replying to @Meaningness
@GrumplessGrinch@cwage Statistical inference (which is *not* the same as PT) gives you limited quantification, which is what you want.2 replies 0 retweets 0 likes -
Replying to @Meaningness
@GrumplessGrinch@cwage Statistical inference starts with a model (which falls out of the sky—this is why PT + logic < epistemology).1 reply 0 retweets 1 like -
Replying to @Meaningness
@GrumplessGrinch@cwage The model is a parameterized, universally quantified statement, e.g. forall x P(boojum(x)|snark(x)) = c.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@GrumplessGrinch@cwage Statistical inference then lets you estimate the parameter c given evidence.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@GrumplessGrinch@cwage The model itself is a logical statement, not a probabilistic one; it is true or false.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@GrumplessGrinch@cwage You could try to assign a degree of belief to the model, but that wades straight into the A_p tarpit I started to >1 reply 0 retweets 0 likes
@GrumplessGrinch @cwage > deconstruct on #LessWrong before I got interrupted and backburnered that project.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.