The map being distinct from the territory, you must go outside your map to discount your probability calculations made in the map. But how to do this? You must resort to a stronger map. But then the calculations there are subject to the errors in designing that map.
You can run this logic down to the deepest level. How does a rational person adopt a Bayesian methodology? Is there not some probability that the choice of methodology is wrong? But how do you conceive of that probability, when Bayesian considerations are the only ones available to evaluate truth from given evidence?
Why don’t these considerations prove that Bayesian epistemology isn’t the true account of knowledge?
Why don’t these considerations prove that Bayesian epistemology isn’t the true account of knowledge?
Looks to me like you’ve proved that no one can ever change their beliefs or methodology, so not only have you disproven Bayesian epistemology, you’ve managed to disprove everything else too!
You are unwinding past the brain that does the unwinding.
A rational agent goes “golly, I seem to implement Occam’s Razor, and looking at that principle with my current implementation of Occam’s Razor, it seems like it is a simple hypothesis describing that hypotheses should be simple because the universe is simple.”
That is literally all you can do. If you implement anti-occamian priors the above goes something like: “It seems like a stochastic hypothesis describing that hypotheses should all differ and be complicated because the universe is complicated and stochastic.”
So, you cannot ‘run this logic down to the deepest level’ because at the deepest level there is nothing to argue with.
The map being distinct from the territory, you must go outside your map to discount your probability calculations made in the map. But how to do this? You must resort to a stronger map. But then the calculations there are subject to the errors in designing that map.
You can run this logic down to the deepest level. How does a rational person adopt a Bayesian methodology? Is there not some probability that the choice of methodology is wrong? But how do you conceive of that probability, when Bayesian considerations are the only ones available to evaluate truth from given evidence?
Why don’t these considerations prove that Bayesian epistemology isn’t the true account of knowledge?
Looks to me like you’ve proved that no one can ever change their beliefs or methodology, so not only have you disproven Bayesian epistemology, you’ve managed to disprove everything else too!
Counter example: I changed my epistemology from Aristotelian to Aristotle + Bayes + frequentism.
You are unwinding past the brain that does the unwinding.
A rational agent goes “golly, I seem to implement Occam’s Razor, and looking at that principle with my current implementation of Occam’s Razor, it seems like it is a simple hypothesis describing that hypotheses should be simple because the universe is simple.”
That is literally all you can do. If you implement anti-occamian priors the above goes something like: “It seems like a stochastic hypothesis describing that hypotheses should all differ and be complicated because the universe is complicated and stochastic.”
So, you cannot ‘run this logic down to the deepest level’ because at the deepest level there is nothing to argue with.