Defining the Supernatural: A Recap

Continuing the discussion from How should we define the supernatural?.
The original thread has grown quite long and sometimes off-topic, so I decided to make a summary of some of the important points answering the original topic. Thank you for everyone who shared their thoughts. The original question was, how should we define the supernatural and natural? An auxiliary question: is the supernatural analyzable by science? I am going to classify the responses to this question several categories.

(Note that if I listed a name, it doesn’t necessarily mean they endorse that view fully, only that they brought it up in the thread.)

Regularism: the natural is anything that is regular and predictable, and the supernatural is anything which is not, although people have differing views on whether “supernatural” is a useful term.
Analyzable by science?: No.
Brought up by: @gbrooks9, @T_aquaticus, @jongarvey

Interactionism: anything that interacts with our senses and possibly tools that extend our senses is natural.
Brought up by: @PdotdQ, @cdods.
Analyzable by science?: Possibly. Following Sean Carroll, we shouldn’t limit the purview of science beforehand.

Scientific-historical: the supernatural was anything that we did not understand. What we have learned from science since we call “natural”.
Brought up by: @BruceS, @nwrickert
Analyzable by science?: Yes (assuming science will continue to progress).

Personalism/Mentalism: the supernatural is something which is fundamentally, irreducibly personal and/or mental.
Brought up by: @dga471, Richard Carrier
Analyzable by science?: No.

Theistic (Creator-created distinction): the supernatural is the uncreated acting upon the created. The natural is the created acting upon the created.
Brought up by: @T.j_Runyon, @swamidass
Analyzable by science?: ???

Sorry if I forgot to include your view or misunderstood it!


Comparing the different options

Regularism seems to be a popular view. However, as @T_aquaticus, @gbrooks9 and others have articulated, there are phenomena such as weather patterns, quantum mechanics (including radioactive decay) which are only predictable in aggregate (if at all), yet we don’t consider them to be supernatural. There are also a host of other examples brought up by @PdotdQ and discussed in Predictability Problems in Physics.

I would also add philosopher of science Nancy Cartwright’s work here, where she points out that even in physics, scientific laws are only completely valid in separate regimes - we only have a patchwork of laws connecting different domains, not a complete description of all reality at all levels. We might have to fall back upon the fuzzier notion of “predictable in principle”.

Finally, there also seems to be the problem that any spiritual truths we know entail predictability to some extent. For example, Christians believe that God is consistently faithful, loving, and providential. Thus, to some extent, God’s qualities are predictable and regular. But we don’t regard God as a natural being. I think there should be a way of rigorously distinguishing between God’s predictability and the predictability of nature, but more work needs to be done.

Interactionism is similar to regularism, with an emphasis on the senses. One problem (which @BruceS and @cdods picked up during the discussion) is that in science we rely on instruments which we do not interact directly with. Nobody has ever seen a Higgs Boson with their senses - only aftereffects of its existence. In fact, most modern physics experiments rely on a long chain of logical inferences from the phenomena, through the instrument(s), to the scientist herself.

Finally, there is the problem of miracles (such as the turning of water into wine, or the parting of the Red Sea), which people often regard as supernatural, yet in many cases they are observable by the senses. Are we content to bite the bullet and call these natural (or hypernatural, as @DaleCutler says)?

The scientific-historical definition seems to be popular among agnostics and atheists. If one starts with materialistic assumptions, then this view seems to be fully internally consistent. One critique is that this definition assumes that science will continue to be successful in the future to describe all aspects of reality. But how do we know this? Even in theoretical particle physics, where some are fearing that we are nearing a “particle desert” which will not be surmountable unless we build accelerators which are many orders of magnitude larger than what we have now, if at all. The future success of science seems to depend on an underlying starting assumption that everything in nature is regular and predictable in principle. This is not a philosophical conclusion, but an optimistic assumption.

Secondly, the scientific-historical definition is at odds with how theologians and religious scientists have historically understood the supernatural. Even in the Bible, we see affirmations of regularity and order in the cosmos. These are viewed as attributes of God’s action, not evidence that he doesn’t exist. Thus, it seems that this definition is improperly changing the understanding what the term “supernatural” is meant to refer to. It is a definition that is unrecognizable to anyone who does not already believe in the non-existence in the supernatural.

My own view of personalism hasn’t been discussed much, other than @swamidass pointing out that it could imply that humans are also supernatural. To that I agree, in the sense that our souls are probably supernatural and not fully analyzable by science. I also pointed out that one doesn’t need to hold humans to be fundamentally personal, they could be completely physical (although this might be in conflict with traditional theology).

Finally, @T.j_Runyon’s Creator/Created definition seems to restrict the supernatural to only direct actions of God, the only uncreated being. Everything else is natural. As some others have pointed out, are angels natural? How about the disciples of Jesus performing miracles? One way to wriggle out of this is to argue that in these instances, it is really God, and not angels or humans, who is acting. But presumably many people believe that God can also act through natural means (“Providence”). How do we distinguish natural providence from supernatural acts of God? Or is providence itself supernatural?

1 Like

What about my view?

Creation vs. Creator.

Supernatural is what the Creator directly does. Natural is what creation does.

1 Like

That is similar to @T.j_Runyon’s view. I’ll add it. I have also critiqued it.

I’m not even sure I have a view yet. I have ideas and I’m working through. Different elements from different views. Trying to bring them together in a coherent way. Regularism fits my thinking as well

1 Like

I certainly don’t believe that science describe all aspects of human reality.

Nor do I believe that the supernatural is what we do not understand. My view of naturalism is not what science knows, but rather what it accepts as a candidate for a scientific explanation.

Certainly there can be non-scientific, supernatural explanations. My approach to understanding the supernatural takes no position on their truth.

I do think that the boundary of the supernatural as implied by scientific practice can change, for example Einstein considered non-locality to be “spooky”. I’d put multiverse explanations as those that are currently controversial in that way, although I recognize many would be unhappy with calling them supernatural.

My view is better described as akin to the Pragmatic MN expressed by Boudry et al in “How Not to Attack Intelligent Design Creationism:Philosophical Misconceptions About Methodological Naturalism”. with the proviso that I take the success of science to be evidence to consider science to be effective at bounding acceptable scientific explanation.

It is important to distinguish wrong scientific explanations from explanations that are not even acceptably scientific. And I do think my approach may have some problems with that issue. But I have not given it enough thought to post about.


Thanks for sharing this paper. It would be a great one for us to discuss; I’d love to hear your thoughts on it. I read all of it and I think it really summarizes the common arguments for “intrinsic methodological naturalism” (or IMN, as they describe it) well. It clarifies some issues. But I think it doesn’t really undermine IMN completely; it only undermines IMN with regards to some low-level manifestations of supernatural events. I think everyone agrees that science has successfully ruled out some naive, simplistic ways of thinking about how the supernatural affects the natural; for example we can’t say that just praying for someone to be healed significantly increases the probability that they will be healed. Many theologians are also puzzled at how God sometimes seems to deliberately hide his presence.

But as Pennock and Scott and others (quoted in the paper) point out that an omnipotent God would have the capability to do many things in the natural world without us necessarily detecting it to be so. One could label this “pseudoscience” (as Boudry et al. do) but I don’t think theologians ever claimed this to be science at all.

There already is a thread on this paper but I’m happy to re-open if you want to discuss further.

On prayer: As I take you to you point out in your last paragraph, science can rule out scientific explanations on the efficacy of prayer or lack thereof. It cannot rule out a supernatural intelligent agent who chooses which prayers to grant and decides to thwart such experiments. Of course, one could argue theologically about whether that is in the nature of the God one believes in.

I agree that the label “pseudoscience” is incorrect for such supernatural explanations, with the proviso that supernatural explanations characterized as being science do qualify for this label.

I think there is a commonality with the genealogical approach to Adam: that approach is an explanation of how a belief could be true, and be compatible with science. But it is not a scientific explanation.

That’s probably enough for this thread.

1 Like


This is a very helpful distinction…

But it may need a footnote about metaphysically “created” things… like angels!

It seems to me that God’s providence, or simply ‘Providence’, is neither of those, or at least it is indistinguishable from the latter for the nonbeliever (hence the utility of the term ‘hypernatural’ for believers – at least a couple of us, anyway :slightly_smiling_face:).

Some occurrences, frequency of occurrences or several occurrences in sequence can be remarkable and meaningful enough to defy ‘normality’. (I did not say, ‘remarkable enough to defy nature’. :slightly_smiling_face:) Rich Stearns relates just such a sequence of multiple unaccountable ‘coincidental’ events in his best-selling book, The Hole in Our Gospel, that influenced him to resign from being the CEO of Lenox, Inc., manufacturer of high-end dinnerware, to become the president of World Vision, a large Christian NGO dedicated to caring for children and improving their lives. Improving their lives with basics of course, but also longer term, like giving sewing machines and training to mothers to become providers and entrepreneurs, or wells and long distance supply piping to isolated communities, as well as follow-through, equipment maintenance and training. (As a septuagenarian, I have multiple enough instances and several sequences of my own – or rather, that were given to me – that I took to documenting them in a ‘Co-instants Log’ about three and a half decades ago, with retrospective entries for two decades before that.)

The problem I see is that this could be used as an explanation for anything. Under those same rules, science could not rule out Leprechauns making mushrooms appear on my lawn when no one is looking. We would end up with an epistemology where anything is true. The only limit on what is true is our imagination.

This is also the general problem I see with defining the supernatural. The definitions seem to focus on finding excuses why we can’t detect the supernatural in an independent and objective manner. However, these definitions require us to adopt an epistemology that makes everything meaningless. To use an analogy, it is like changing the rules of a sport so that everyone wins, and in doing so it makes the very purpose of sport meaningless.

I agree and that is why I would not call theological explanations scientific.

I think that for someone who accepts theological explanations would start with the theology and constrain theological explanations to agree with that theology. So I imagine a Christian would rule out a capricious God to explain prayer experiments failures whereas an ancient Greek would not rule them out (nor would someone who believes in Leprechauns, I suppose)…

I also agree that one should not define if and only if conditions for supernatural or MN, but rather one should look to what is included and excluded from successful scientific practice to understand those concepts.

1 Like

I suspect that the disconnect may be due to people starting at different points. I start with premises and then work towards a conclusion. It would appear that others start at the conclusion and try to work out what premises they need to reach that conclusion.

1 Like

If you listen to podcasts, you may enjoy this one since it focuses on that issue for the epistemologies of science and religion.

I have not read the book which the author is being interviewed about, but based on the interview it discusses that issue starting with Galileo but not focusing on that episode only (despite the title).
The interview has a wide scope, but not much depth.


This is a crucial distinction for a site like PeacefulScience.Org.

Early on, Atheist/Agnostic participants were keen to present P.S. as a science discussion with religious implications.

But it is the opposite: it is a theological site with scientific implications.

1 Like