How should we define the supernatural?

I was initially inspired by @PdotdQ’s argument that methodological naturalism is tautological (see Side Comments on Christians in Science - #47 by PdotdQ) . Specifically, I was intrigued by this premise:

I disagree with this. I haven’t fully decided yet, but surprisingly, I think atheist blogger Richard Carrier has some good observations when he defines the supernatural (Richard Carrier Blogs: Defining the Supernatural):

In short, I argue “naturalism” means, in the simplest terms, that every mental thing is entirely caused by fundamentally nonmental things, and is entirely dependent on nonmental things for its existence. Therefore, “supernaturalism” means that at least some mental things cannot be reduced to nonmental things. As I summarized in the Carrier-Wanchick debate (and please pardon the dry, technical wording):

If [naturalism] is true, then all minds, and all the contents and powers and effects of minds, are entirely caused by natural [i.e. fundamentally nonmental] phenomena. But if naturalism is false, then some minds, or some of the contents or powers or effects of minds, are causally independent of nature. In other words, such things would then be partly or wholly caused by themselves, or exist or operate directly or fundamentally on their own.

Thus, Carrier’s defining characteristic of the supernatural is something which is has a fundamentally irreducible mental (or personal) aspect. Because personal agents defined in this way may act in unpredictable ways (even if they can be influenced by the environment and external factors), they cannot be analyzed by science, which requires regularities. Thus, we cannot scientifically analyze and predict the will of God.

Now, atheists might think that this definition of the supernatural renders it implausible, anti-science, outdated, superstitious, and so on. But the idea that there are irreducible aspects of reality is thriving in many circles today. There are many philosophers, even non-religious ones such as Thomas Nagel and David Chalmers, who regard consciousness as irreducible to material explanations. Chalmers, in particular, advocates taking conscious experience as fundamental (Facing Up to the Problem of Consciousness). I think it is not a huge step from there to agreeing that the notion of irreducible mental agents is not a ridiculous possibility. It doesn’t prove the existence of the supernatural, but it renders it plausible if we have other grounds to believe that it does exist (as Christians might affirm). Similarly, there are many people who believe that we have irreducible, libertarian free will.

What do you all think of this definition of the supernatural? Does it cohere with how Christians view the supernatural aspects of God, angels, demons, and human souls? What is missing?

I certainly don’t think Carrier captures all aspects of the supernatural in the Bible; for example, miracles are commonly regarded as supernatural phenomena as well, even though they have little to do with the existence of mental aspects. But I think miracles are a different category of supernatural compared to human souls and God - phenomenologically they are just interruptions of the regular natural order.

Side note: Carrier actually reminds me of Augustine, who in The City of God argues that sin is a fundamentally not the result of external factors, but a human’s irreducible will wanting to do evil:

“If the further question be asked, What was the efficient cause of their evil will? there is none. For what is it which makes the will bad, when it is the will itself which makes the action bad. And consequently the bad will is the cause of the bad action, but nothing is the efficient cause of the bad will.”
Augustine, The City of God, Book XII


For the record, methodological naturalism, on its own terms, is tautological. That is one clue is incorrectly named.

Just to be clear, I do not hold that methodological naturalism is tautological, but was summarizing a position that I encountered before.

In my view, its tautological-ness in the end depends on the various definitions of natural/supernatural/science etc and the particular form of MN that one is considering.

I am more sympathetic towards this view (paragraph copy pasted from Sean Carroll):

Science should be interested in determining the truth, whatever that truth may be – natural, supernatural, or otherwise. The stance known as methodological naturalism, while deployed with the best of intentions by supporters of science, amounts to assuming part of the answer ahead of time. If finding truth is our goal, that is just about the biggest mistake we can make.


Not if you define the supernatural as above: a personal entity that is fundamentally irreducible. Modern science always attempts to explain phenomena by reducing them to lower-level mechanisms. Thus, it cannot invoke irreducible personal agents, such as “God caused the ball to fall, not gravity.”

Given Carroll’s definition of science, I cannot imagine how one can scientifically prove that the Resurrection, for example, was supernatural. At most, one can say as a scientist that the Resurrection breaks regular observed patterns (that dead people stay dead). But breaking of a pattern doesn’t prove supernaturalism; we observe breaking of established patterns often in science and attempt to explain those at a deeper level.

In general, Carroll tends to argue that theism is false (or unnecessary) because God doesn’t add much as a scientific hypothesis. This forms the context of this statement. He basically commits the error of reducing God and supernatural entities to “God-of-the-scientific-gaps”. He also seems to espouse a mild scientism, by implicitly assuming that science is able to find out the truth at all times. Christians, in contrast, hold that science is not enough: revelation is also needed.


That means humans are supernatural. Is that really what you want to say?

Yes, in the sense that our wills are not fundamentally reducible to a deterministic, mechanistic set of atoms. This is also why when we are conducting a scientific experiment, we tend to want to isolate the system from human influence as much as possible.

Of course, there is a difference between humans as supernatural entities and God as supernatural entities. I haven’t fully thought out what the difference is, but one sure element is that God has powers to influence the natural, physical world as much as he likes, so we’re completely unable to put a bound on how much he can interfere with scientific experiments.

But this view of the supernatural doesn’t necessarily entail viewing humans as supernatural. You could still say that humans are fundamentally mechanistic and fully material (such that their actions and decisions are fully reducible to a set of deterministic rules); but God is not.

1 Like

I think I read Carroll’s paragraph differently. I read it more as an invitation: for every phenomena that is within the purview of science, science should be used to determine truth. I did not read it as claiming that everything can be studied with science. If supernatural causes can be studied by science, then science should be used to study it.

Perhaps the sentiment is better said as: start with a neutral view on the causes of things, do science, and if the cause turns out to be natural, so be it. If the cause turns to be supernatural, then so be it also.

1 Like

But how would you judge that the cause of something is supernatural while doing science? I am concerned that this has a danger of falling into ID-like arguments, i.e. if natural mechanisms seem unsatisfactory to explain something, invoke the supernatural.

1 Like

I believe here is where you need a solid definition on what is natural/supernatural. The point is that one lets science do its thing without cuffing it a-priori by saying “this is supernatural, and science shouldn’t touch it.”

1 Like

I’m certian this is an instinct to be resisted. Science is not the “generic” pursuit of truth.

Yes, which is why methodological naturalism, defined with secular language does not work. Historically, it was understood by another name, as a particular understanding of the Creator-Creation distinction. Science is meant to study creation, but not the Creator.


Isn’t this also sort of tautological? The whole debate is trying to determine what exactly is within the purview of science.

I agree, which is why I started this thread. I’m trying to see if we can create a definition of the supernatural which is not simply another God-of-the-gaps.

I certainly agree with you that this should be the attitude when you are acting as a scientist. But as a Christian, there seems to be many areas where scientific investigation would be out of place. As you’re a Catholic, I would think a good example would be the Eucharist. Surely it would be improper to scientifically investigate whether transubstantiation occurs, for example - first, it is claimed to be a metaphysical, not physical phenomena, and secondly, it seems to miss the point.

How do angels and demons fit into this scheme? They are also part of creation.

1 Like

If we found a way to study angels and demons, they would become part of the “natural” world, even if they were unseen. While we cannot determine their existence scientifically, they are consider superstitious or supernatural. Science studies creation, but this does not mean all of creation is within our view.


Is it? Perhaps it is, or maybe I am just not presenting it correctly. Maybe a better wording is that one should let science loose on everything. On some things, it will fail to generate any knowledge. On other things, it will generate some knowledge. This second subset of things are the ones I call the purview of science.

Within this second subset of things, parts of it is labelled natural and another part is supernatural (the division depends on your definition of natural and supernatural).

My understanding of Carroll’s paragraph is that since science can generate knowledge in this sub-subset that is labelled supernatural, it should be allowed to do so.

Perhaps my problem is that I don’t fully understand what is meant by “science” and where exactly the line with pseudoscience is drawn. This is why I think a lot of this depends on what one explicitly mean by science/natural/supernatural/etc. I am not sure, for example whether astronomy, social sciences, or mathematics is science or not.


So, can you elaborate more on your preferred definition of the supernatural vs. natural? You said that something is natural if it interacts physically with the world. But how does that work? How does it not also apply to God, who according to Christian belief has intervened several times to cause physical effects in the world?


Again, to clarify I am just summarizing a point of view that I encountered before, I do not hold it. I used to, but after some thinking I realized that I can come up with other definitions of supernatural vs. natural that sounds just as reasonable to me. This is why I used to think that MN is tautological, but now I mellowed to saying that its tautologicalness depends on one’s definition of supernatural/natural/science/etc.

Within my worldview, God would affect the physical world in a purely natural way. So indeed, the phenomena of say, the transfiguration on the mountain is a natural phenomenon with this particular definition of naturalness. God does not change the laws of physics to bend photons to make Jesus look bright, but rather those photons follow a purely natural evolution to arrive there. In a purely deterministic universe, this can be achieved e.g. through a particular choice of values on a Cauchy surface.

This does not mean that science can study the phenomena of the transfiguration. There are many natural phenomena that are non-repeatable and cannot be studied by science.


I’ve approached this question from a slightly different perspective.

The working definition of the “natural” for me has been “That which we can perceive with our physical senses.” More specifically, science would then be anything where we can evaluate hypothesis using our physical senses.

This will obviously leave some things, for example consciousness, where it may not be clear today how much falls into the realm of science (the natural) vs the supernatural. I would expect science to investigate these areas, but also also won’t be surprise if they find that somethings remain beyond the grasp of science, as they exist only in the supernatural.

@PdotdQ and @dga471

Your discussions so far don’t seem to touch the “natural” vs. “super-natural” dichotomy at all.

When Christian Alchemists were at the peak of their credibility, they believed they could accomplish important feats of transformation by their own mortal intelligence and insights.

One of the earliest big successes was the creation of Phosphorous, from urine. It even glowed!

One of the more lasting failures was their attempt to invoke angels and God to assist in manipulating nature. Some alchemists even set up a prayer tabernacle in their laboratory to help in what they thought was an obvious part of the “Natural Philosophy”.

The creation of phosphorous became incredibly reliable and predictable. Virtually any careful technician could do it.

But anything that involved invoking God, or even God’s tolerance, became notoriously spotty. Someone would claim success. Others would attempt to reproduce it; and fail. That would be a problem if the exercise was converting lead to gold, and you were trying to reproduce the process in front of the King and his royal guard!

“Super-natural” does mean, and should mean, events that are not reproducible or even predictable by humans.

“Natural” does mean, and should mean, events and processes that are either reproducible by humans or predictable by humans.

If it is regular enough to be predictable, who can doubt that it is part of the realm of natural science?

The chief gray area in the separation of Natural from Super-natural is in the unpredictable nature of Quantum-level events. Some things seem beyond prediction - - and yet there is no real suspicion that God would take special interest in the movement of a (say) a muon … compared to more significant things like the moral choice of a human. So while many sub-atomic events seem unpredictable, lots of scientists are willing to categorize such activity as inevitably “predictable” (at least in theory).

I think that we go wrong as soon as we ask for a definition. in the sense a set of necessary/sufficient conditions that further is presupposed by science.

Better I think is to say that MN is what emerges from the practices of successful science. This is what I take Sean C to be driving at. The supernatural is just what is currently excluded from successful science.

Mental phenomena and personal agency are addressed by science, although not physics. Philosophy too is needed. One place for philosophy is when norms enter, as in understanding what is a mental representation or what is meaning. Philosophy also enters when we try to relate vital human concepts like free will and moral responsibility to what science can and cannot tell us about their natures.

The problem I see in this is that we depend on science itself to build the instruments and provide the statistical models by which we claim to observe things. The Higgs boson would be one example.

I also suspect you used “physical” for a reason and that you have some candidates in mind for non-physical senses. .Is that fair?

There’s a lot to unpack, so I will try to break it up into a few thoughts.

  1. At least in my view, reductionism is separate from methodological naturalism (MN). If reductionism fails to produce an answer in some instances it does not invalidate the application of MN. On top of that, our inability to model very complex systems is a matter of ignorance. We can’t accurately model 3 or 4 body orbital systems, but I don’t think anyone considers them to be supernatural. We also can’t predict which nuclei of a specific isotope will decay next, but no one thinks that is supernatural. We can’t make long term predictions about weather patterns, but once again, no one thinks weather is supernatural. Does this also apply to the very complex human brain? If we had greater knowledge could we predict how humans think and react? Why do drugs and brain injury change behavior? I don’t know if we can answer all of these questions.

  2. Humans obey natural laws. Traditionally, the supernatural is defined as a violation of natural laws. When we find natural explanations for claimed supernatural miracles most people don’t consider it a product of the supernatural. When a magician makes someone move through the air using wires, do we consider that supernatural even though it was the product of a human mind?

  3. Miracles are notoriously hard to evidence. In the Bible we read of miracles that would be extremely easy to verify today. For example, Moses and his followers were guided through the desert by a pillar of smoke or fire, and food fell from the heavens. Jesus walked on water. Moses made a stick turn into a snake. Jesus turned water into wine. This seems to be quite different from the definition of supernatural being proposed, and these types of miracles don’t seem to happen in a way that makes them amenable to observation.

1 Like

I’ll say it.

The natural/supernatural dichotomy does not make any sense.

Long ago, people used “supernatural” to describe what was unknown. But we have learned a lot since then. And we have used “natural” to describe the world as we have learned to understand it.

It is as if the world started out coated with supernatural paint. As we learn more, we repaint it as natural. And the supernaturalists are finding themselves being painted into a corner. That corner gets smaller and smaller.

1 Like