I've been reading Jaynes again this week, and he's just very, very convincing. And so I'm trying to read everything you wrote through these Bayesian glasses, bu

I've been reading Jaynes again this week, and he's just very, very convincing. And so I'm trying to read everything you wrote through these Bayesian glasses, but sadly, I'm not successful. Jaynes is rather critical of Fisher's hypothesis testing, on the ground that you can't accept or reject an hypothesis on its own; you need an alternative to compare it to, and that alternative needs to make definite predictions. I don't see what the alternative to your null hypothesis is (the negation of the null hypothesis does not make definite predictions)