Quote:

Originally Posted by youngalexander The usual garbled twaddle AA? You get more like your lowercase each post.

A lot of debates about the historical Jesus on the web contain anticipation of Richard Carrier's book, formerly promised to decide the matter of whether or not Jesus existed using Bayes' Theorem, but now promoted by Carrier in this video, apparently, just to strike down using Bayes' Theorem some bad arguments for the historical Jesus without promoting better arguments. This would probably be more realistic.

Proving an argument for or against the historical Jesus using Bayes' Theorem would not be easy. Bayes' Theorem is useful for repeated observations of very similar phenomena (related algorithms of hypothesis testing are used by data processing software routinely in my field), but almost anything in history happens only once, ancient historical evidence is almost entirely written language, and judgments of probability based on linguistic evidence come out of a set of mental computations that require either a human brain or a supercomputer installed with a futuristic AI program. Knowledge of the probable meaning of any phrase anywhere requires not just the meanings of the individual words in the phrase but knowledge of the grammar of the whole language, the literary context, and the social context. All of that data (a data set that happens only once) feeds and combines into a monstrous mental process that has not come close to being distilled into a known mathematical algorithm of any sort, let alone something as simple as Bayes' Theorem.

Richard Carrier is a smart guy, but he has no background in statistical science or mathematics, presumably since high school (there are other academics who have made the philosophical application of Bayes' Theorem the focus of their careers for decades). It is a problem not just for ancient history but in almost any relevant debate. Bayes' Theorem is not even appropriate for the Riddle of Epicurus, because Carrier's argument presumes that human beings are closely-analogous to the gods (with the same perspectives of good and evil).

And I don't accept Carrier's claim that all of our reasonable thinking is a form of the application of Bayes' Theorem. If our mental processes of reasoning can be statistically modeled, it would require a helluva lot more than Bayes' Theorem, no matter how many terms you include in the equation, because not all (nor even most) of our best judgments of probabilities are based on a populated array of similar data points. I suspect that the most useful thing Richard Carrier can do with such ideas is to start a personality cult. He is already well on his way, apparently.