A new member of the Unbelievable? Facebook group recently posted a lengthy demand for atheists to provide a summary of the applicable standards of evidence and criteria used for applying methodological naturalism to the question of God's existence. The subsequent thread included several explanations that it wasn't the job of atheists to state which of many gods they didn't believe existed, but the poster continued with what appeared a to be a veiled attempt to shift the burden of proof. Observing the protracted circularity I weighed in with the following:
It seems to me that the OP is a ploy to get an atheist to define God, to state what evidence would be apparent if such a god did exist, and then for the OPer to follow up with a statement that the particular god so defined is not the god believed in.
If I were to bite, this is how I would respond: we have two hypotheses — “a god exists” and “no gods exist”. We also have empirical evidence — we observe the world. We can compare the likelihood of the evidence we see given the two hypotheses. Given “a god exists”, how likely is the evidence we see? And given “no gods exist”, how likely is the evidence we see? Is the likelihood of the evidence we see more likely or less likely if “a god exists” rather than “no gods exist”?
Note that I’m not coming down on either side of this question, since that wasn’t what was asked. If we are to assess the evidence we observe using methodological naturalism we are not going to find anything supernatural, merely “unexplained”. If gods are supernatural, science cannot explain them, because methodological naturalism is restricted by definition to natural mechanisms.
Methodological naturalism can be used, however, to assess claimed supernatural effects, but will only be able to provide natural explanations (if any) of those effects. What it can’t do is to say “no natural explanations are currently available, therefore the explanation is supernatural.”
Some might conclude from the above that gods and other supernatural causes are entirely beyond explanation, and may even go further and say that since it’s impossible for science (using methodological naturalism) to explain anything supernatural, we therefore can have no knowledge of the supernatural.
So as far as knowledge of gods is concerned, we are left with “revelation” and nothing else.
This was largely ignored, but I reproduce it here because it summarises my position, and also because it is now unavailable on Facebook. After some further exchanges the thread appeared to be homing in on something, perhaps to a point where a degree of agreement might have been achieved. But eight hours later the entire thread had disappeared, so I posted a query:
What happened to Daniel Ray's thread querying the standards and criteria applicable to methodological naturalism? It was here last night, but this morning it's gone.
Even the thread above makes no sense, because after deleting the thread in question (and apologising for doing so) it appears that Daniel Ray's comments above have subsequently also been deleted. He may have deleted his account.
Either stand by your written word, or apologise for it. But don't delete it without warning, especially if doing so will delete everyone else's contribution — or as shown above, eviscerate a thread into nonsense.