We regularly see the unsophisticated claim that, since some claimed event is improbable based on known natural causes, it probably has a supernatural cause.
Probability ( of a SuperNatural Cause ( for CLaimed ( EVent ))) = 1 – Probability ( of a Known Natural Cause ( for CLaimed ( EVent )))
… or P(SNC(CL(EV))) = 1 – P(KNC(CL(EV)))
The inadequacy of this formulation of supernatural causation is immediately clear. We have ignored the probability of the claim being true or false. Eye-witness claims are subject to an examination of the honesty and cognitive reliability of the witness. Only the most dishonest of supernaturalists attempt to omit this variable. So let’s revise our formulation.
Probability ( of a SuperNatural Cause ( for CLaimed ( EVent ))) = 1 – Probability ( of a Known Natural Cause ( for (True ( CLaimed ( EVent )))) – Probability ( of the Falsity of ( CLaimed ( EVent ))
…or P(SNC(CL(EV))) = 1 – P(KNC(CL(EV))) – P(F(CL(EV)))
Is this an adequate formulation? Not yet. We have not yet addressed the probability that the cause of the event may be an unknown natural cause. In centuries past, epileptic fits had no known natural cause, leading supernaturalists to claim the cause was therefore something supernatural such as demon-possession. So we need to remedy this omission.
Probability ( of a SuperNatural Cause ( for CLaimed ( EVent ))) = 1 – Probability ( of a Known Natural Cause ( for ( True ( CLaimed ( EVent )))) – Probability ( of the Falsity of ( CLaimed ( EVent )) – Probability of ( UnKnown Natural Cause for ( True ( Claimed ( EVent )))
…or P(SNC(CL(EV))) = 1 – P(KNC(CL(EV))) – P(F(CL(EV))) – P(UNNC(T(CL(EV))))
This last essential variable in our formulation of the probability of a genuine supernatural event is often missing in discussions of the probability of supernatural events. How is P(UNNC(T(CL(EV))) assessed? It is assessed by examining the track record of supernatural claims in the past for which there was no natural cause available as an explanation at the time. Epilepsy has already been introduced. We now know the material mechanism behind the once materially inexplicable phenomenon of seizures. We have acquired a degree of epistemic humility. The fact that there is no current material explanation is no excuse to conclude the cause is supernatural.
So what has been the track record of the claims of supernaturalists? How many times has a conclusion that an claimed event is supernatural been later shown to either 1) have a natural cause or 2) to have been fabricated?
The balance of the ultimate wins/losses of supernatural claims over history will give us degree of confidence we can honestly place in the probability of future claims of supernaturalists.
The following is a short list of some of the events supernaturalists once claimed were clearly supernatural.
– Statistically more rain in a particular region due to the petitioning of a god by those living in that region.
– Seizures that were posited as evidence of demonic possession, leading to the torture and abuse of the epileptics.
– Diseases, lightening strikes and accidents posited as the punishment of some deity.
– The transfer of traits from parent to child prior to the discovery of DNA.
– The existence of thought before the discovery of neurons.
– The movement of celestial bodies prior to the discovery of natural laws of motion.
– Greater wisdom, altruism and self-control given to believers of a particular god to degrees not naturally possible prior to the ability to statistically assess whether such claims were true.
A proper formulation of the probability equation for supernatural events, plus an assessment of the track record of supernaturalists’ claims demonstrates just how justified we are in dismissing future claims of supernatural causation as highly improbable. This is basis of the methodological naturalism that undergirds the accelerating success of science today.