ven if supernatural forces intervene in the world materially, it is impossible for humans, operating as we do on the physical plane, to prove that material effects, even very remarkable ones, stem from supernatural causes that exist outside the physical plane. Therefore, even if supernatural forces do indeed exist, asserting that I know they exist must necessarily be false.
To illustrate this, think of a miracle from the Bible – let’s say the parting of the waters by Moses. The story reads like a myth, not unlike other myths from various cultures across the globe throughout history. But let’s imagine that scientists found concrete physical evidence that the parting of the seas actually occurred. Wouldn’t such evidence prove the existence of the god of Moses?
The answer is no. It does not prove the existence of Yahweh any more than it proves the existence of Thor. Jews and Christians might ask, “Why would Thor intervene on behalf of Moses?” Believers in Thor can simply answer, “Thor works in mysterious ways.”
Or if not Thor, perhaps some other deity parted the seas, or a different supernatural force. Perhaps it was a magic UFO. Perhaps it was subconscious wish-energy generated by the angst of the harried Jews, exerting a telekinetic force on the waters.
If we had concrete physical evidence proving that the seas parted, but no concrete physical evidence suggesting a cause for that effect, then we’d have to admit that the cause was a mystery; it would not follow that Yahweh exists.
The intervention of a supernatural force within the physical plane is impervious to investigation by physical creatures such as humans, since our investigation cannot penetrate the curtain to the other side (the spiritual plane, Heaven, or what have you). Therefore, we cannot conclusively attribute such intervention to one supernatural force over another.
Nor can we even say with certainty that some supernatural force must be at play, since the physical evidence could instead be caused by something natural that we simply don’t understand yet. Even if the seas parted, we could not be sure that Yahweh, Thor, or telekinesis was behind it; instead we’d have to admit that the cause was unknown.
Furthermore, it is unscientific to speak of the statistical probability of a supernatural force existing. When we flip coins or roll dice, myriad material forces – complex, overlapping, and often undetectable to our senses – confound our ability to foresee what’s coming. Probabilities help us predict outcomes by finding patterns in like past events. We know that flipped coins have even chances of landing heads and trails, because we’ve flipped coins numerous times before. A supreme being’s existence is an event that either happened once or not at all; there are no like past events we can reference to ascertain a pattern. Moreover, since there is no way of tracing a material effect to a supernatural cause, there is no observation of nature that could possibly suggest a probability for the existence of the supernatural.
I’m not suggesting that supernatural forces absolutely must not exist. They might. But even if they do exist, it’s impossible to know that they do.
Now, let us carry this point to its logical conclusion. If you wish to base your world view on reason and evidence, then as it is impossible to prove a supernatural cause for a material effect, there are never good grounds for accepting any supernatural force of any kind.
And so the most rational stance as to the existence of a supernatural deity, regardless of the deity, is this: a willingness to admit that we cannot rule it out, but an unwillingness to personally adopt such beliefs, as they are unsupportable by reason and evidence. This stance might be called implicitly atheistic agnosticism, or more simply, agnostic atheism.
-  Here I take the position called frequentism, which defines probabilities in terms of frequencies derived from trials. Probabilities do not apply to unique events because, by definition, there are no identical trials. However, we often speak informally of probabilities as rough, subjective indications of our confidence or doubt in propositions. These are called subjective Bayesian probabilities, and they allow us to use other kinds of evidence than past trials to measure degrees of personal belief. I see many reasons to doubt God exists, and I can count up these reasons and conclude that the chances of God’s existence are, say, one in 10,000. It is tantamount to saying, “I don’t know for sure, but I strongly doubt that God exists.” ↩