On his Dangerous Idea blog Victor Reppert reprints an old exchange between us (we have been going at it for over 35 years now):
This is one that appears in my paper on Hume on Miracles, Frequencies and Prior Probabilities.
Science is unavoidably naturalistic, or atheistic if you prefer. Science operates in terms of scrutable, independently testable entities that operate in accordance with knowable regularities. Supernatural beings, on the other hand, are essentially mysterious; claims made on their behalf are not independently checkable, and there are no “laws of supernature” governing their behavior. Furthermore, “explanations” in terms of supernatural entities are inevitably post hoc and untestable. In other words, proponents of supernaturalistic theories can glibly account for things we already know, but become strangely silent when asked to predict something new, something that would allow their theory to be tested.
Even though the locus of discussions of miracles is historical rather than scientific, if it is the case that supernaturalist hypotheses are inevitably untestable, this would mean that supernaturalist claims cannot be genuinely supported by evidence. But some points can be made in response to this position. First of all, I see no in principle impossibility in “laws of supernature.” One cannot, of course, generate deterministic laws governing divine conduct, but one cannot generate such laws concerning the behavior of subatomic particles, either. One can, of course, form probabilistic expectations concerning the conduct of subatomic particles, but, as we have noted, one can generate probabilistic expectations concerning divine conduct as well. It would disconfirm belief in the Christian God if Jim Bakker were to die and rise again on the third day, ascending into heaven a few weeks later. The “laws” of supernature that Christians or other theists are inclined to postulate may not be as detailed as the laws scientists hope to discover in nature, but they leave theistic claims open to confirmation and disconfirmation.
 Keith Parsons, “Is there a Case for Christian Theism?” in J. P. Moreland and Kai Nielsen, Does God Exist: The Great Debate (Nashville: Thomas Nelson, 1990) p. 189.
I basically agree with what I wrote twenty one years ago, but I think I should clarify and expand some things. I think that hypotheses postulating powerful supernatural entities are in principle testable. A template for such a test is found in I Kings Chapter 18 where you have the famous experimentum crucis between Elijah and the priests of Baal. You recall: Elijah makes a sacrifice for Yahweh and the priests make one for Baal. The priests call all day for Baal to send fire and burn the sacrifice, to no avail. Then Elijah steps up and calls upon The Lord, and fire falls from heaven to consume the sacrifice. Elijah then leads the people in a massacre of the priests of Baal (hey, every good OT story has to have at least one massacre).
Now, of course there would be many intractable practical problems with setting up such a test. For instance, you would probably need to bring in James Randi to make sure that no trickery was involved. However, I see no reason why, in principle, such a test could not be done nor why a supernatural hypothesis could not be confirmed in this manner. The “inevitability” I mention in the above quote I would now take as referring not to the in-principle impossibility of such a test, but to the predictable refusal of religious apologists to submit their hypotheses to such testing. We could have a test like in I Kings now live on TV and broadcast worldwide. Fire falling from the sky would pretty much shut up pesky atheists like me (if, again, I could be sure that it was the real deal and not that somebody had dropped napalm). Why, then, instead of resorting to recondite philosophical arguments, do religious apologists not avail themselves of the obviously superior choice of performing a public experiment? Nullius in verba! Do the experiment! Give atheist smartasses their comeuppance! (No massacres, OK?)
Well, of course, no modern day apologist, however brash and confident his claims, will conduct such an experiment. He will probably decline mumbling something about how you should not put The Lord to the test. No such scruples bothered Elijah, and I hope that it is not too cynical to assert that were they confident of the sorts of results Elijah got, today’s theists could set aside their qualms also. Why don’t they have such confidence? This gets back to what I said about “laws of supernature.” The biggest practical problem facing a latter-day, would-be conductor of an Elijah-type demonstration is that there do not seem to be any detectable regularities governing miracle events that would provide a basis for predicting experimental results. God will send fire from heaven when and only when he damn well pleases, and when he will please seems completely unpredictable. No detectable laws appear to govern the performance of miracles. Technically, we know of no true nomological conditionals of the form “If a condition of type C were to obtain, then God would perform a token of miracle-type M.” Of course, God could tell us when to expect a miracle, but that does not seem to happen either.
Now Victor thinks that there might nevertheless be knowable (non-deterministic) laws of supernature that permit the confirmation or disconfirmation of theism. Unfortunately, the example he gives is completely unhelpful. I cannot at all see how Jim Bakker rising from the dead and ascending into heaven would disconfirm Christianity. Rather than quibble, though, I can just say “Ok, then, put up or shut up!” Less truculently: If there are laws of supernature that could support something like a scientific experiment, then name the laws and provide at least a sketch of the experimental design. Let’s do it! Enough of the endless logic chopping; let’s do the experiment.
Of course, as he indicates, Victor will reply that he did not mean that there were discoverable laws of supernature of sufficient detail and specificity to provide the basis for an actual experiment. We cannot say when and where God will perform his next miracle (e.g., the Houston Astros winning next year’s World Series). We can only make much vaguer predictions, maybe like this: Any world that God creates will have sentient creatures to be in communion with God. Obviously, there is no way to test such a “prediction” scientifically, but it could be (and is) used as the basis of a philosophical argument for theism. Philosophers can argue out whether the existence of sentient creatures in this world confirms the existence of God.
However, such a reply concedes my original point in the passage quoted, which concerned science, not philosophy. My argument was that trying to support theism with science is dicey, in part, because theistic hypotheses just do not seem amenable (in practice if not in principle) to confirmation or disconfirmation with the sorts of experiments and tests regularly employed in the natural sciences. As far as I can tell, Victor here has said nothing to challenge that claim.