Posted: Jan 25, 2012 2:07 pm
by logical bob
Mus Ponticus wrote:Carrier doesn't say that all valid arguments use Bayes' theorem, he's saying that you can use Bayes' theorem to describe any valid arguments.

I can see that's what he says, but I'm trying to get past his clumsy terminology to see what he means. If he thinks "any argument that violates Bayes’ Theorem is invalid. All valid historical arguments are described by Bayes’ Theorem" makes sense then when he says "are described by" he must mean "do not violate". Although he's trying to give the appearance of a tight logical argument in this paragraph he isn't clear on what he means by violating a theorem or describing an argument. I was paraphrasing him in the hope of making his point at least coherent.

I'd be willing to accept that an argument that used probabilities in a way contradicted by a correct application of Bayes' Theorem would be invalid. It simply does not follow from this Bayes' Theorem has any bearing on all valid arguments, for the reason I gave.

The bigger point is that history doesn't handle probabilities in a mathematical way and has never claimed to do so. A mathematical treatment of probability requires that the probabilities be expressed numerically. Who's attempting to put numbers of the chances of Brother of the Lord being a biological relationship rather than a title? Nobody. So what exactly is the role of a process that requires numbers as input?