Hypothetically, if there were a supernatural realm, would it be reasonable to expect that we would be able to study it using the scientific method? On the other hand, is it possible that the supernatural could be real but not accessible to scientific investigation? If so, what other non-scientific methods might allow us to have a justified true belief (i.e., 'know') in the existence of the supernatural?

Related: Is parapsychology a science?

  • 2
    For one we might actually be studying the supernatural every time we do Science, it just depends on your definition of the natural and supernatural really. Say people who meet ghosts of past ancestors might consider this natural and not meeting them or their disappearance as unnatural, just like how scientists consider working with probabilistic models of quantum mechanics i.e., phenomenon like entanglement, superposition and tunnelling might be considered actually supernatural by some who don't know these things about nature.
    – How why e
    Commented Jun 15 at 2:03
  • 5
    Radio waves are pretty supernatural. A handheld device lets you communicate with no exchange of anything material or that you can sense at all. Show me something more astounding than that? Gravity waves. Ok. You got me there... Neutrinos causing supernova explosions, which are how we end up with the iron that makes our blood work. Boy, so many super (beyond) natural things exist! Why chase ghosts when real accessible miracles are everywhere? Study them. Be curious.
    – Scott Rowe
    Commented Jun 15 at 3:59
  • 3
    If wishes were horses, beggars would ride. Scientific method is different for physics, biology, psychology, it is adapted to each subject to meet its needs. The only common core are vague platitudes: be thorough, test your claims, observe and do not make things up, reason soundly, check biases, etc. If we study something we would want to do those things, whether we call it "scientific" or not. But what substance is there to saying it about we know not what?
    – Conifold
    Commented Jun 15 at 7:15
  • 3
    It all boils down to what "supernatural" and "real" (and their intersection) mean.. As Conifold pointed out, "science" can be quite broad and adaptable to subject matter under study. In what sense, for example, something "real" and "supernatural" is not also broadly "natural"?
    – Nikos M.
    Commented Jun 15 at 13:24
  • 2
    Well, I haven't done any scientific research on it, but I have a belief, that I consider justified, that this is a question intended to push a personal agenda, and I'm also pretty sure that it didn't come to me from any supernatural realm.
    – JonathanZ
    Commented Jun 15 at 13:33

8 Answers 8


Well, as someone who rejects the supernatural, I'd put forth the argument:

Yes and no. No, because by definition, the sciences cannot study the supernatural because modern science rejects the existence of the supernatural on metaphysical grounds. By definition, it would not be possible. Contemporary science generally embraces methodological naturalism which is presumed to reject the supernatural categorically. From WP:

Methodological naturalism, the second sense of the term "naturalism", (see above) is "the adoption or assumption of philosophical naturalism … with or without fully accepting or believing it.”[25] Robert T. Pennock used the term to clarify that the scientific method confines itself to natural explanations without assuming the existence or non-existence of the supernatural.

Yes, if the supernatural is simply a phenomenon that needs to be reduced to the natural. Therefore, if the sciences were to come across an apparent miracle, it would be, by virtue of its own process, obligated to presume that there is a natural explanation. Let's say some being appears to be an angel lands on the lawn in DC. The sciences would begin to build a case. Does it emit light? Does it have a heat signature? Does it have mass? And if the case being made go to the point where the being began displaying miraculous qualities (could fly, can speak any language, can get objects to levitate), an effort would be made to provide scientific explanations (it's an advanced being with anti-gravity technology, it uses a quantum supercomputer, it comes from an advanced civilization).

This brings us to the linguistic nature of the question and the famous three laws of Arthur C. Clarke:

  1. When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.
  2. The only way of discovering the limits of the possible is to venture a little way past them into the impossible.
  3. Any sufficiently advanced technology is indistinguishable from magic.

It's the last one that applies here. Obviously, the implicit metaphysical premise is that the linguistic category 'supernatural' simply means 'natural, but beyond the current capacity of natural explanation'.

  • And thus we now have invisibility cloaks.
    – mudskipper
    Commented Jun 15 at 18:44
  • @JD If science presupposes philosophical naturalism, then wouldn't it be epistemologically self-inconsistent for a scientist to believe in the supernatural? See the two examples I reference in this question.
    – user66156
    Commented Jun 15 at 18:44
  • 1
    @Mark Well, I don't know if you've people, but the following facts apply. 1. Scientists operate with bounded rationality. 2. Most scientists have no substantial education in the philosophy of science. 3. There is no monolithic 'science' in practice. Scientists are free and almost always have a personal understanding of what science is to them...
    – J D
    Commented Jun 15 at 19:22
  • 1
    There are plenty of scientists who segregate naturalistic thinking from their supernaturally enriched religious faith. Gould is famous for defending the utility of that with his en.wikipedia.org/wiki/Non-overlapping_magisteria A person can be of two minds, in the extreme quite literally. Consider DID. The typical scientist doesn't have the metaphysical vocabulary to account for a sophisticated account of science and religion.
    – J D
    Commented Jun 15 at 19:22
  • @Mark Oh. You have a better version of the question. I'll riff on it in a bit. Maybe something in the text will appeal to you.
    – J D
    Commented Jun 15 at 19:24

The question, while interesting, is actually a clever linguistic trick.

What happens is natural. The only exception is impossible things. When they happen, the universe includes magic. If it was understandable by science, it wouldn't be magic.

If, on the other hand, Harry Potter magic (or the Christian God) can actually be understood consistently with science, then it is not magic, just "sufficiently advanced technology."

  • 2
    Pretty much everything we do now would look like magic 300 years ago. Well, back to communicating with people around the world by tapping on a little flat thing... I have important things to think about.
    – Scott Rowe
    Commented Jun 15 at 12:38
  • 2
    After seeing it twenty times I've stopped finding the trick "clever".
    – JonathanZ
    Commented Jun 15 at 15:13
  • 2
    @Miss_Understands - The engineers I knew in graduate school thirty years ago were already dreaming about how to create audio/video streaming services, high power semiconductors, knowledge systems or neural networks, other inventions, which have become integrated into technology today. The network implications (complex system models) were not understood by most of us back then. I did not anticipate the capabilities of neural network systems to develop so rapidly but my model for a neural network was the early perceptron. Some sources argue AI is overhyped. Neural systems are also mysterious! Commented Jun 15 at 16:09
  • 1
    @JonathanZ "It's really only an interesting psychological phenomenon."
    – Scott Rowe
    Commented Jun 15 at 16:51
  • 2
    @Miss_Understands Not really. When we saw communicators and tricorders in Star Trek, we recognized them as science fiction, not fantasy. They were pretty obvious extrapolations of the technology we already had. But if you showed them to someone in medieval times, they would have no reference point to extrapolate from. The cloest thing they could relate to was witchcraft.
    – Barmar
    Commented Jun 15 at 20:02

Science very specifically (by definition) includes experimentation. I'm quite pleased that Google (or Oxford Languages) even chose to underline "experimentation". I didn't do it... look:

Definition of science

So, the question: "If the supernatural were real, would we able to study it scientifically?" is asking...

"If the supernatural were real, would we be able to study it through observation, experimentation and the testing of theories against the evidence obtained?"

Now let's be sure we are correctly interpreting the word "supernatural".

Definition supernatural

Noting: "a manifestation or event attributed to some force BEYOND SCIENTIFIC UNDERSTANDING or the laws of nature".

Uh-oh... We have hit a snag.

By definition "supernatural" must be "beyond scientific understanding"... otherwise, it does not meet the criteria to be called "supernatural".

I think based on definition of the word "supernatural"... the complete, correct and conclusive answer to your question is "No".

"Supernatural" and "scientific" are by definition not compatible.

Unless of course we decide to flout convention and make-up some new definitions for science or supernatural??

  • 6
    Lots of things that are currently within the realm of scientific understanding used to be beyond it. Microbiology was beyond scientific understanding before the invention of the microscope, for example. Stipulate a similar instrument and what was previously beyond scientific understanding becomes a field of research. It would no longer be supernatural to researchers armed with a wooscope, but it would still be the same phenomena that we here and now call "supernatural."
    – Chris Cox
    Commented Jun 15 at 11:48
  • 4
    Today's "beyond scientific understanding" is tomorrow's "been there, done that". Most people now have the 2-way wrist TV of my childhood. I had no idea that would happen.
    – Scott Rowe
    Commented Jun 15 at 12:32
  • 2
    @Mark since the scientific method is rooted in observation, then if it's not accessible to science, that means it's not observable. If it's not observable, then how in the world would someone come to believe it in the first place? Through guessing?
    – TKoL
    Commented Jun 15 at 14:25
  • 2
    @Mark I did not not anticipate how silly this would get, my mistake. I don't see why demons would matter though.
    – TKoL
    Commented Jun 15 at 15:25
  • 1
    Then the question is really about demons, not levitation. Just because something is hard to arrange or inspect doesn't mean it can't be studied by science. We know the composition of stars in other galaxies. Just find a way to get any observations of demons. If you can't, then there is nothing to study. If it isn't consistent, then what are we talking about?
    – Scott Rowe
    Commented Jun 15 at 16:56

Other answers are saying supernatural phenomena are defined to be non-scientific. I want to add that, if something is natural and real, it doesn't imply it could be studied scientifically.

Most scientific methods depend on the thing to be studied being repeatable. Some counterexamples:

  • How much water did I (this StackExchange user) drink last weak? After I forget this information there isn't likely any clue to recover it.
  • What's the historical maximum total weight of the species of bacteria being the last to extinct before year 1000? There are hardly any record of it.
  • How many people are there in average on the moon between the first human to reach there and the last human to either leave or die? We don't have a better way to guess other than just wait until humans get extinct or the moon get destroyed.

But they are all real and have precise correct answers.

So, even without the supernatural part, there isn't enough condition in the question to make something possible to study scientifically.


If the supernatural were real, would we be able to study it scientifically?

Yes. Take astrology as our example. If astrology were a real science, this is what would happen:

Scientists would notice that astrology was overpredicting a given personalty trait, by comparing their theoretical predictions against the real incidence of that trait in the human population. Further calculations would show that the incidence of that trait in the population could be explained by the presence of a planet at X and Y coordinates in the sky at the time of birth.

Scientists would then turn their gaze to that point in the sky, and there the planet would be. The discovery and presence of the new planet would reconcile theory and observation.

Similar analyses could be done for areas of study that depend upon otherworldly assumptions.


A broad definition equates the supernatural with the domain of miracles. The supernatural domain deals with miraculous beings and miraculous events.

  1. Typical examples for supernatural phenomena are religions, see religion-science from the Stanford Encyclopedia:

    “One way to distinguish between science and religion is the claim that science concerns the natural world, whereas religion concerns the supernatural world and its relationship to the natural.”

  2. Phenomena, which are claimed to exist in a supernatural domain, cannot be studied scientifically. Because the method of science needs reliable observations of the phenomena from its domain of investigation.

    I do not know any method that “might allow us to have a justified true belief (i.e., 'know') in the existence of the supernatural”. History confirms this negative report.

  • So, beliefs of supernatural things would be 'unjustified' by definition. If we could justify them they would be natural. Why do people want to study or assert things that are not natural? That's the big question on everyone's mind...
    – Scott Rowe
    Commented Jun 15 at 12:35
  • 2
    @ScottRowe I think your question has already found different answers, e.g., the supernatural serves as a parallel world for wishful thinking or for legitimizing secular institutions.
    – Jo Wehler
    Commented Jun 15 at 13:01
  • Yes, well, science seems to assert that there are numerous parallel worlds, so worlds with and worlds without supernatural elements suggest themselves. Now, which one are we in? :-)
    – Scott Rowe
    Commented Jun 15 at 13:58
  • 1
    @ScottRowe The answer may depend on the person: Some people live within the natural world, others live in a supernatural world. While some schizophrenics live in the natural world on weekday, but on Sunday morning within a supernatural world :-)
    – Jo Wehler
    Commented Jun 15 at 15:17
  • Gosh, I recently encountered someone here whose profile said basically that.
    – Scott Rowe
    Commented Jun 15 at 16:48

If there was any sure way if getting to see a real miracle, it would be advertised as such by reputable secular sources, and made a business. Schools would teach spells like in Harry Potter, supermarkets would sell magic ingredients, governments would produce laws on which magic is allowed and tax it's application.

The Catholic Church as an example has existed for almost 2000 years, teaching superstitions and investing any and all reports of such. They could monetize of selling tickets to do/see miracles (and use the profits to do good of course, or spend most of it the way they usually do on "administration"). Since they don't, you can be sure that there is not enough to be investigated. If prayers worked for healing, there would be hospitals without doctors, where prayers were applied to injuries. If speaking to the dead or resurrection worked, police would use it to ask victims of murder about their murderer. If levitation worked, they'd teach in in highschool classes, along with broom riding.

The closest you can get to believing in supernatural is homeopathy magic as opposed to the placebo effect. This superstition somehow seems to be difficult to get rid off, as a society, despite education.

If cities built on clouds existed, they could be scientifically researched. If monsters under the bed existed, they'd be made into a zoo. If the Easter bunny existed, it would be caught and bred. Of course if any of those fairy tales and superstitions were real, they could be researched.

But that is not a matter for philosophy, nor science, nor skepticism. It's for mysticism, which is just a euphemism for superstition. Just as there are young earth creationists digging up the grand canyon to find human footprints crossing dinosaur footprints, and there are expeditions to turkey to find remains of Noah's ark, all reports of cultist experiences could be researched. It's just turned out to be a waste of time, too many times. Way too many times.

While at the same time, all of the arts, the natural sciences, the humanities, still offer vast areas of the unknown to be explored without anything supernatural. Nobel prizes waiting to be granted. Our knowledge of nature (or some gods great design if you prefer) waiting to be expanded.

So why keep asking to see a real magician pulling a rabbit out of a really empty hat, just because some.guy on YouTube claims they have experienced this?

This kind of question seems to link the meaning of life to the existence of the supernatural, like "My life only has meaning if there really are magicians pulling rabbits out of hats. If not, my life has been a delusion." This is not a philosophical question, but something for ex-cultist trying to reconcile with mainstream ask.

The important question to ask would be, could 100s, 1000s, millions of people be deluded an believing in a false supernatural theory? And the answer must obviously be YES, since not all of Judaism, Christianity, Islam, Buddhism, Shintoism, Taoism... Can be right, it proves large masses of people must be able to follow some false prophets. Regardless of whether any of them is right, the other ones are wrong, showing how easily millions of people can be misled.

Another more relevant question to ask would be: "If there had been zero supernatural events on earth in the last 10000 years, would there still be religions, cults, superstitions, and guys telling on YouTube that they have seen real ghosts?" And the answer is: yes, you betcha. Just like the previous question, the presence of so many different large religions believing in different superstitions and making up reports of their particular brand of miracles shows that humans will produce unreliable reports for various reasons.

  • Religion would certainly have a better case if there was just one, and everyone who asserted it agreed, across time.
    – Scott Rowe
    Commented Jun 15 at 17:01

Yes, but the scientific method is the wrong methodology. To scientifically study the supernatural, you want something similar to the historical method.

For this answer I have assumed the following definition:

That entity is supernatural which cannot be described in terms of the evolution of natural processes, yet which can at times affect such processes. Natural processes are those processes whose states are predicated on the past state of any or all other natural processes.

We study something very much like the supernatural when we study history or archeology; or study astronomy, which is just history through a telescope; or do investigations of recent events, which is just recent history.

In fact with a slight shift of definition one might accidentally or on purpose define the past to be supernatural: it supervenes upon the world we observe, but cannot be predictably influenced by reconfiguring the state of the present world.

Consider the constraints on studying history:

  • We cannot configure initial conditions and run the experiment to see what happens.
  • While the system may or may not happen to be relatively isolated, nothing we can do can make it more isolated from its surroundings.
  • Most information about the system (the past) is permanently missing.
  • The information which remains is in the form of relics - physical objects which are not part of the system, but which we can infer to have been affected by the system. Ruins and fossils for historians; or for astronomers the ancient light, perhaps distorted by gravity, which a telescope picks up from a distant galaxy.
  • Where records exist, they are often the most useful relics, so much so that we divide the study of our own past between the period in which records are available (history) and the period in which they are not (prehistoric archeology). Records give us access to information about the system which was still present when the record was made, but which is now long gone. On the other hand, records are only as reliable as the person who recorded them.

These constraints, perhaps with slight modifications, apply to the supernatural as well. Put another way: if we take something typically classed as supernatural, and we find that studying it doesn't have these constraints, we'll also find that it is a natural process as defined above.

To study with the constraints of historians, one should probably study with the methods of historians.

I grant that there are many other definitions of the natural and the supernatural. I just happen to think that they're all pretty much useless for questions such as this one. They group by where you'd shelf a novel featuring such an entity in the book store, or they assume a particular religious framework, or they attach to the current state of scientific knowledge, or they define either the supernatural or the natural out of existence. Useless, useless, useless.

  • This is a great explanation! It is true that we only study things that have already happened. Given that, we could include these historical methods in with the methods of science generally and just call history 'science'. It would have 2 kinds: stuff we can run again and stuff we can't.
    – Scott Rowe
    Commented Jun 15 at 12:26
  • 1
    @ScottRowe Works for me. If you can convince everybody to start calling the scientific method 'the experimentalist's method' instead, I'll vote for you for definition dictator.
    – g s
    Commented Jun 15 at 14:03

You must log in to answer this question.