Peter Soderqvist

Ordinary happenings' credibility require ordinary evidence, for instance, fossilization is the ordinary evidence for ancient life forms, but that Christ are risen are an extraordinary claim, since it is a supernatural phenomenon, therefore it need something which is on a similar level with it, namely, supernatural evidence! If there is no supernatural evidence, Christ is not proven risen either; same it is with the rest of apparitions! Supernatural evidence in a world governed by natural laws is a contradiction in terms, and for the same reason, the believers have a hard time to prove their case!

bd-from-kg

RED DAVE:

As you remind us, Leonardos says:

But the skeptic should have no problem with this statement; in fact, I concur completely. What Christians don’t seem to be aware of is that the methodology of historians incorporates the principle of “extraordinary claims require extraordinary evidence”. So much so that an integral part of the standard methodology is to reject truly extraordinary claims out of hand. For example, reports that eclipses or earthquakes occurred in close proximity to significant events are routinely dismissed without further examination. This is even more true, of course, of reports of truly miraculous events. Such things tend merely to discredit the authors. But not as much as one might expect, because such reports are so very common in the ancient literature, even in works by relatively reliable authors. See Richard Carrier’s excellent essay, Kooks and Quacks of the Roman Empire.

bd-from-kg

Peter Soderqvist:

There is no logical reason why miraculous claims should require miraculous evidence. All that’s needed is evidence such that it’s more unlikely that the evidence would exist if the event had not occurred than it is that the event occurred. (See Bayes’ Theorem below.) Since we don’t know that miracles cannot occur, this is not (in principle) an “unmeetable” burden of proof. This subject is discussed admirably in Richard Packham’s The Man with No Heart: Miracles and Evidence

bd-from-kg

“Extraordinary Claims require Extraordinary Evidence” and Bayes’ Theorem

The rigorous statement of the “extraordinary claims/extraordinary evidence principle is Bayes’ Theorem. It states that:

P(H|E,B) = P(H|B) · P(E|H,B) / P(E|B)

Here

P(x) = probability of x
P(x|y) = probability of x given y
P(x|y,z) probability of x given y and z

H = the hypothesis (or here, the claim)
E = evidence for the claim
B = background (or prior) information

Notice that all of the probabilities here are “given B”, which is natural since we always have the “background” or “prior” information. So we can just drop the “B” on the understanding that all of our probabilities are relative to this background information. with this understanding the formula can be written:

P(H|E) = P(H) · P(E|H) / P(E) .

Finally, from the definitions and elementary probability theory we have:

P(E) = P(H) · P(E|H) + P(~H) · P(E|~H) .

So the formula becomes:

P(H|E) = P(H) · P(E|H) / [P(H) · P(E|H) + P(~H) · P(E|~H)]

Now instead of trying to figure out what all this means in the general case, let’s see what it tells us when the hypothesis H is highly improbable given only the background information (which we can generally think of as “what we know about how the world works”), and before we’ve seen the specific evidence for it. To simplify things, let’s also assume that the hypothesis is of such a nature that if it’s true, the evidence E is practically certain to exist.

Since we’re assuming that H is highly improbable (i.e., P(H) is small), P(~H) is practically 1 (since the two must sum to 1). And by our second assumption P(E|H) is also essentially 1. So the Bayes formula becomes (to a good approximation):

P(H|E) = P(H) / [P(H) + P(E|~H)] .

Thus in judging the credibility of a hypothesis or claim based on some specified evidence, we need only ask which is greater: the likelihood that the claim is true ignoring the evidence in question [i.e., P(H)] or the likelihood that the evidence would exist even if the claim were false (i.e., P(E|~H)]. If the latter is greater, the claim must be assigned a probability of less than 0.5 – that is, it’s more likely false than true.

For example, let’s see how to apply this formula to estimate the likelihood that Jesus rose from the dead.

[Note: Some people object to using the word “probability” in this kind of context since practically all of the “probabilities” involved are subjective estimates. They prefer to use the term “credibility” and use C(x) instead of P(x), etc. This doesn’t really change anything since the “credibility calculus” works exactly the same as the “probability calculus”.]

Remember that we need to have P(E|H) nearly equal to 1 to apply the simplified formula. Thus we can’t define E to be the exact, specific evidence available, but something like “evidence of such-and-such quality or better that Jesus rose from the dead”. It seems plausible to say that it is nearly certain that, if Jesus did rise from the dead, the evidence that he had done so would be at least as good as what we have, so we’re OK here. And of course H – the claim that Jesus rose from the dead – is highly improbable a priori – we know that it is very, very unusual (to put it mildly) for a man to rise from the dead.

Thus Bayes’ Theorem tells us that to estimate the likelihood that Jesus rose from the dead, we can simply compare the likelihood that a particular man (about whom we have no specific knowledge) should have risen from the dead, to the likelihood that he would have been reported to have risen from the dead in some anonymous writings produced decades after the fact (or that evidence of roughly this kind of quality should exist for the event).

I suppose that it may be possible (though I have no idea how) to arrive at the judgment that the former event is more likely than the latter. But the point is that the skeptic is not being unreasonable in demanding extraordinarily good evidence given that the odds against a particular man rising from the dead are at least hundreds of billions to one against, by the most conservative estimate. (Actually I’d put this probability somewhere between 1 in 100 trillion and 1 in a quadrillion. And that’s still probably conservative.)

At any rate, this should suffice to make it clear that “extraordinary claims require extraordinary evidence” just a simple, “layman’s” way of saying something that can be stated quite precisely and demonstrated rigorously in mathematical terms.

Primal

Well Plump DJ extraordinary claims require extraordinary evidence because they are more parsimonious and less conflicting with our background knowledge.


Example: Claim: Jesus rose from the dead.

In the case of the above claim I can assume A) It's true.

or B) It's false.

Now according to my experience no man rises from the dead, nothing I see die comes back. According to all I know of history, biology and medicine this is true as well.

So I can think one of two things more probable given such background knowledge 1) The person is lying/or wrong in his/her claim. 2) The even happened and all my previous knowledge is wrong.


Now I've seen many, many people lie or become confused. But I've yet to see one person rise from the dead.


So I go along with choice 1.

Basically that was chosen because there was a simpler/more parsimonious way to judge the claim. As is the case with all extraordinary claims.

Thomas Ash

Darn, that's exactly what I was going to say . Thief!

Seriously, good post, and welcome to IIDB - keep up the posting at that level (and keep on saying things I agree with !) and you'll be a great addition to this community.

Thomas Ash

One point I thought I'd make is that the claim religious people tend to deny is that claims require evidence, let alone that extraordinary claims require extraordinary evidence. What this means is simply that when a claim like the resurrection is made, what's lacking is any evidence to outweigh the prior scepticism we should give to this claim knowing what we do about the reluctance of people to rise from the dead (BUffy excepted .) So maybe we should fight that battle first, before going on to the next claim.