Originally posted by Apologia4JC18 in the thread What is the difference between "atheism" and "Fundy Atheism"? on December 21st 2005:

Quote Originally posted by John Powell
POWELL:
If you came up with that on your own, Apologia4JC18, then I compliment you for your insightful thinking. Michael Horner used something like your approach in his debate with Farrell Till. I criticized Farrell for not adequately replying to Horner's criticisms. My method is an effort to avoid the criticisms of Horner and others who oppose the ECREE principle while preserving the proper ways that supporters have used it.

I posted the dictionary definitions of "extraordinary" and "ordinary" in hopes that you would see that the terms, as I use them, relate reasonably well to the standard definitions. Your response seems to be based merely on those dictionary definitions. You are attacking a straw man version of my method.

In the rest of my post I showed what is more precisely meant by the terms (in my method). The extraordinary tied to the claims is "beyond what is ordinary" in a different sense than the extraordinary tied to the evidence. That significant difference is how I hope to avoid the infinite regress problem you point out.

An extraordinary claim is a claim that is "beyond what is ordinary" with respect to the frequency of it having been correct in the past. In particular, an extraordinary claim is one for which there are few, if any, accepted similar events that have occurred. The "Jesus resurrected" claim is an extraordinary claim because there aren't any similar events to it that you and I agree have occurred.

Extraordinary evidence, on the other hand, is evidence that is "beyond what is ordinary" with respect to its reliability. Extraordinary evidence is evidence that is markedly superior to the ordinary say-so evidence of a typical person. Generally speaking, seeing it for yourself or reading it in a science journal counts as extraordinary evidence.

A reasonable restatement of the principle "Extraordinary claims require extraordinary evidence" might be "Seldom correct claims require extra reliable evidence."

If your extraordinary claim were "My dog can converse like a human" then me visiting your home and having a normal human conversation with your dog would constitute extraordinary evidence. No infinite regress. Note, however, that my claim to having had a normal conversion with your dog would not, in general, constitute extraordinary evidence to other skeptics but would merely constitute ordinary say-so evidence to them. They might have to see for themselves the talking dog.

So, please read and respond to the rest of my previous post to you and to what I posted here. If there is anything in this material that you dispute, question, or don't understand then point it out. If, after doing so, you think I should reply in greater detail to the scenario you posted above then I will.

John Powell
Thank you for the clarification. I actually agree with this formulation, but I have two questions.

First, with respect to the existence of God: it has always seemed to me that the denial of God could be made only on the supposition that certain contingent beings have not had causes (if all contingent beings have had causes, then it seems to follow necessarily that an eternal, uncaused cause of all contingent beings must exist).

But it also seems like the claim that atheists must make amounts to an extraordinary claim, according to your definition. Certainly, no one has verified the existence of an uncaused contingent being. It also seems to be true that the claim that any particular contingent being has arisen without a cause is "seldom correct", to say the least.

Wouldn't the denial of the causal premise (and therefore, the denial of God) then count as an extraordinary claim requiring extraordinary evidence?

Secondly, the principle seems to turn on what one takes to be reliable. If one claims A as reliable evidence for B, someone could ask "How do you know A is reliable?" If one claims C as evidence for the reliability of A, one could ask "How do you know C is reliable?" and on down the line, ad infinitum. Presumably, there must be a stopping point, or there could never have been an end point (as there must be).

Thus the question seems to be, "How can our choice of starting points be justified when it is presumably the case that any number of such points could yield a coherent set of propositions?"
Great point