How to perform science without using "methodological naturalism."

Some people here seem to believe that science operates under an impediment by restricting itself to what is often called “methodological naturalism”, as a result of which it is unable to address or answer the full scale of questions that might arise regarding the universe.

Why has no one then developed an improved scientific method that dispenses with MN? Can anyone give an example of how this would be employed, and what we could learn from it?


All papers must contain a disclaimer saying “Or God did it, we’re not sure”.

Next we’ll hear from the “Technologically and spiritually advanced aliens are telekinetically making it happen with their giant psychokinetic brains”-crowd, that they want those disclaimers in there too.


People have. It’s called ID and scientific creationism.

Personally I think MN works just fine in science.


Actually I think we have an example: The Genealogical Adam hypothesis.

That means the “improved MN” would basically consist of positing empirically untestable claims for entirely nonscientific reasons, then believing them anyway despite the impossibility of scientifically verifying them.

I’m reminded of this cartoon

The Genalogical Adam hypothesis is pretty much this in essence.
“There was an Adam supernaturally created!”
“Oh yeah? Prove it!”

Suddenly I recall a parable about an invisible dragon in somone’s garage.


What discoveries or knowledge have those revealed that could not have been learned from regular science?

1 Like

If I understand correctly, Geneological Adam is not intended as a means of obtaining knowledge, but as a ploy to make it easier for people to remain Christians if they accept evolution.

@swamidass ?

Methodological naturalism is unnecessary. All that science requires is that theories be testable. “Maybe an unknown supernatural power tampered with our experiment in a way that we can’t verify” is no better or worse than “Maybe an unknown natural phenomena tampered with our experiment in a way we can’t verify”.

Entertaining and totally absurd.

The GAE follows MN closely.

I’m pretty sure we established a while ago you are mainly poking fun when the GAE comes up.

That is not the purpose, but it is a result. The purpose is to honest about what the evidence does and does not say.

Here is how I understand methodological naturalism:

Suppose, as many believe, a person parted the Red Sea just by waving a staff at it.

Most of those who believe this happened say this is “supernatural” or a “miracle”, and therefore is an exceptional event that does not affect how we understand the “natural” world to operate.

According to MN, however, we dispense with the idea of the “supernatural” and, if there was good evidence that this occurred, we would either come up with an explanation for it consistent with how we understand the “natural” world operates, or change our understanding of the “natural world” to accommodate this phenomenon.

It is a curious phenomenon that we have never encountered a clearly demonstrable occurrence of this sort that could not be accounted for thru one of those two responses.

What you are describing is not MN. That is not how science works. Sorry.

1 Like

What am I getting wrong? If we witnessed a person part the Red Sea, and there was no doubt this had happened, would the only correct response from scientists be: “Miracle. Nothing for us to do here.”?

Nothing is testable if your tests are constantly being rigged by gremlins.


I think there are two separate aspects of methodological naturalism at play here, and not everyone is keeping them distinct:

  1. You are testing a theory which include the possibility of supernatural action.
  2. You are testing a purely naturalistic theory, and supernatural entities are interfering.

It’s the difference between testing for the existence of leprechauns vs allowing for the existence of leprechauns.


One of the critical features of scientific testing is repeatability. MN assumes the laws of nature are constant so tests may be independently repeated and verified. How are you going to have confidence in the test of, say, a new vaccine if you have to allow for a supernatural Loki God making it effective one trial and a deadly poison the next?

1 Like

Suppose a pair of undergraduates told their instructor that their lab experiment didn’t produce the data they expected because undetectable gremlins were rigging the experiment. Would their instructor need to ask them whether those gremlins were natural or supernatural before dismissing their claim?

I don’t think that is quite accurate.

Certain things are demonstrated to be repeatable and verifiable, and so are determined to be laws. There is no assumption involved. If they are violated, they are no longer considered to be laws. We don’t just assume that they are still laws and consider the violation to be some “supernatural” event.

Not even remotely. The whole point seems to be to make a claim that is consistent with scientific observations, despite having zero empirical evidence that justifies making the hypothesis in the first place. You have theological reasons for positing it, and then you try to construct the hypothesis such that it is consistent with the evidence. But the hypothesis is not motivated by evidence, and seems to yield no specific testable predictions.


I agree. Science proceeds from the observable to determine what the laws are. Nothing supernatural has ever been observed by scientists investigating nature, so there is simply no reason to account for it. MN exists because we live in a world where many people believe in the supernatural. If that weren’t so, MN wouldn’t be needed.

No, but that’s not particularly relevant because it’s about student honesty, not about methodological naturalism.

Suppose a scientist ran an experiment and got some results. Are those results valid if the possibility that undetectable gremlins might have been rigging the experiment is taken into account?