Theism and evidence

No atheist should ever let any theist start talking about evidence. To talk about the evidence for God's existence is poison for the atheist, because the existence of God is, for the theist, a metaphysical proposition; the statement, "God exists," is true "come what may." Calling the existence of God metaphysical is simply descriptive, not pejorative; the naturalist's statement, "A statement about the world is true if and only if (and only to the extent that) it is supported by perceptual evidence," is equally metaphysical.

Kazim discusses this issue regarding the Craig-Ehrman debate about "evidence" for the resurrection, correctly noting that the debate is not about the quality of the evidence but the prior plausibility of the resurrection. Craig correctly notes that if you find the resurrection initially plausible, you will require little evidence of low quality to believe it true. But the believer does not find the resurrection merely "plausible", he believes it to be 100% true. Bayes theorem clearly shows that if you believe some proposition with certainty, no amount of evidence will persuade you to the contrary.

Craig notes:

In order to show that that hypothesis [the resurrection] is improbable, you’d have to show that God’s existence is improbable. But Dr. Ehrman says that the historian cannot say anything about God. Therefore, he cannot say that God’s existence is improbable. But if he can’t say that, neither can he say that the resurrection of Jesus is improbable. So Dr. Ehrman’s position is literally self-refuting.

Craig is correct here: the supernatural resurrection of Jesus (he makes the distinction clear in the next sentence) is neither probable nor improbable, in just the same sense that "all gnorts are kerfibble" is neither probable nor improbable. The resurrection of Jesus is, for the believer, a metaphysical statement, not susceptible to evidentiary proof or disproof.

Any theist who says that he believes on the basis of skeptical evidence is mistaken or lying. They are usually mistaken: theists typically do not understand how skeptics employ evidence. More importantly, theists do not understand that they employ evidence differently than skeptics: they are constantly astonished when a skeptic persists in her disbelief after the theist has presented his "evidence".

That Craig is using evidence differently than the skeptic is made clear by his argument that the truth of the resurrection of Jesus is "evidentially" supported only if the prior probability is high (and it's prior with certainty). For the skeptic, the prior probability is irrelevant. First, we can turn any Bayesian evidentiary argument around and calculate the prior probability the evidence is sufficient to overcome. The lower the prior probability, the stronger the evidence. Second, Bayesian evidentiary arguments increase the posterior probability exponentially as the amount of evidence increases linearly: if I'm ten times more skeptical than you, it takes only one more item of evidence to convince me.

More importantly, the skeptic does not look for the presence of confirmatory evidence: she looks for the absence of disconfirmatory evidence. The laws of physics are interesting, for example, because they specify an needle of confirmatory evidence in a haystack disconfirmatory evidence. If I drop a rock, and it does pretty much anything other than accelerate directly towards the center of the Earth at 10m/s², then the theory of gravity is falsified. In order to understand the "supernatural" hypothesis of Jesus' resurrection, then, we must know what disconfirmatory evidence it specifies. If it specifies a needle of disconfirmatory evidence in a haystack of confirmatory evidence, it's not very interesting. And, of course, to the extent that the believer is certain of the resurrection "come what may", the "hypothesis" specifies no disconfirmatory evidence whatsoever and is not — in the skeptic's sense — even a hypothesis.

It's no fault at all to use "evidence" in a different way than the skeptic... so long as one is honest and clear about distinguishing the different uses. If the theist says that he uses evidence in thus-and-such a way, which is different from how skeptics use evidence in this-and-that way, then there's no problem.

If this distinction were make clear and explicit, however, the entire field of evidentiary apologetics would simply collapse. The ordinary theist can be easily forgiven for misunderstanding skeptical evidentiary arguments; we can — with more difficulty — attribute an individual professional's misunderstanding to incompetence. But when an entire field rests on obfuscating an equivocation between two distinct and incompatible senses of a key term we can conclude only a pattern of intentional mendacity.

Furthermore, we must ask: why have theists redefined a key skeptical term in the first place? Many theologians declare outright that skeptical evidence — and the inexorable conclusion that no gods exist — is simply irrelevant to theism. Why even bother to engage in "evidentiary" apologetics when you have to completely redefine "evidence" to get the whole enterprise off the ground? Again, we can conclude only a pattern of intentional mendacity. Evidentiary apologetics exists to dishonestly undercut skeptical evidentiary arguments against the existence of gods by redefining "evidence" under the feet of the skeptics. And this dishonest endeavor is required because the metaphysical and fideistic arguments are spectacularly unpersuasive.

For these reasons I believe it is just as fruitless to debate evidentiary apologists as it is to debate creationists on the quality of the evidence. The only fruitful topic of debate concerning evidentiary apologetics is: "What precisely do you mean by 'evidence'?" And the only strategy for the skeptic is to show that any proposition — however outrageous — can be supported according to the theists' definition of "evidence". Every debate with an evidentiary apologist should start with the question: Why should we believe the parrot is dead?