So R. Joseph Hoffmann has published three essays arguing against mythicism. I don’t have a bone to pick with mythicism, but I do have bone to pick with bad arguments. Especially self-contradictions.
Here is the bit of self-contradiction that demonstrates that they see Bayes’ Theorem as some manner of sorcery instead of modeling correct thinking when dealing with uncertainty:
But I think the basic factuality of Jesus is undeniable unless we (a) do not understand the complexity of the literature and its context, or impose false assumptions and poor methods on it; (b) are heavily influenced by conspiracy theories that–to use a Humean principle—are even more incredible than the story they are trying to debunk; or (c) are trying merely to be outrageous. To repeat Morton Smith’s verdict on Wells, the idea that Jesus never existed requires the concoction of a myth more incredible than anything to be found in the Bible.
The use of any single “theorem” to deal with the values discussed here beggars the credible.
(speaking of poor methods…)
Did anyone notice it? No? It was his reference to a Humean principle. The same Humean principle that is a Bayesian principle, which he then denigrates at the beginning of the very next paragraph. Ironically, if Bayes’ Theorem doesn’t apply, then neither does Hume’s argument that he appeals to since they are the same thing.
David Hume and Thomas Bayes were contemporaries. Hume used logic to arrive at his conclusion, while Bayes used math to arrive at his formula (which necessiates Hume’s conclusion.). If math doesn’t apply, then neither does logic; Bayes is no less applicable to historical questions than a logical syllogism.
If someone thinks that when doing historical analysis, extraordinary claims requires extraordinary evidence, they are a Bayesian. If someone thinks that falsifiable historical hypotheses are better than unfalsifiable historical hypotheses, then they are a Bayesian. Bayes theorem models all correct probabilistic thinking. If historians are dealing with uncertainty, and using probabilistic language, they should know the rules of probability.
Stephanie Fisher writes:
[Bayes’ Theorem] is completely inappropriate for, and unrelated to historical occurrence and therefore irrelevant for application to historical texts
Of course, she is wrong. Unless we have 100% confidence in every single argument and evidence in history, then probability theory will necessarily apply. It doesn’t matter if you are using percentages to the nth decimal place since it’s not about mathematical accuracy but about making sure your conclusions (which are necessarily probabilistic statements in history) follow from your premises (which are also probabilistic statements). Even if you use educated (or even uneducated) guesses, you still have to follow the rules of probability so that your conclusion follows from your premises.
I reiterate: If historians are using probabilistic statements and educated guesses, they still have to know the rules of probability. To say that probability theory doesn’t apply is to say things like Occam’s Razor and falsifiability don’t apply. And if falsifiability doesn’t apply, then that’s not even pseudoscience. That’s religion.
So the self-contradiction, the irony, is that Hoffmann (and Fisher) contradict themselves when they claim that Bayes theorem doesn’t apply. It does apply, you just don’t understand it. Sure, you can use Bayes’ theorem incorrectly just like you can use formal logic incorrectly, but the sweeping statement that it doesn’t apply is to shut yourself out of correct thinking. The contradiction I’ve hopefully pointed out is that they already use Bayesianism intuitively when they think correctly, even for mundane everyday things. They just need to use it more explicitly when doing scholarship, which is Carrier’s point.
Does Hoffmann really think that he needs mathematical precision to the nth decimal place to conclude that if a student of his misses a week of class that the student was probably goofing off instead of having been abducted by aliens? I would hope not: Welcome to Bayesianism.