Theistic Satanism: Home > Philosophy > Existence of the supernatural

The existence of the supernatural

by Diane Vera

Copyright © 2003 by Diane Vera. All rights reserved.

  1. Definitions
  2. Naturalism as a pragmatic assumption in science
  3. Some apparent limitations of science
  4. The paranormal and skepticism

Below are my thoughts on why the existence of a supernatural realm should not be considered at all unlikely. This article does not attempt to justify a belief in a God or gods in particular. Regarding the latter issue, see Why believe in a God or gods? and Reasonable (though not completely rational) theism.

  1. Definitions

    By "supernatural," I mean "beyond the potential reach of science." I will refer to the set of entities within the potential reach of science as "the natural world."

    An "entity" is any thing which exists or is purported to exist.

    By "science," I mean the set of academic disciplines which practice what is now known as scientific methodology.

    By "reach of science," I mean the set of entities whose nature can be understood by science. Obviously it's not possible for science to investigate everything in the universe directly (let alone everything in the multiverse, if you accept the many-worlds hypothesis). The question is whether science can potentially understand the nature of everything that exists. By "nature," in this context, I mean the set of laws that something is subject to, plus the set of mechanisms by which it works.

    There is, of course, a probably-huge difference between the potential reach of science and the current reach of science. Were this not the case, the scientific enterprise would be on the verge of going out of business.

    Thus, for example, if the potential reach of science were limited to macroscopic objects -- which was the current reach of science back in Isaac Newton's day -- then the behavior of atoms and subatomic particles would be "supernatural" by my definition, because it's radically different from the behavior of macroscopic objects as predicted by Newtonian mechanics. Obviously, atoms are not supernatural. Their behavior has been accounted for, to a large extent, by quantum physics, a theory which is very weird by everyday standards, but nevertheless a product of scientific methodology.

    By a "paranormal" phenomenon I mean an event, observed by at least one person, which both (1) appears to be outside the reach of current science (not necessarily outside the potential reach of science) and also (2) is difficult if not impossible to verify scientifically, e.g. it cannot be repeated at will in the lab, nor does it leave behind any unmistakably weird evidence that would be convincing to people other than those who witnessed the event. (There have been many experiments, e.g. the Michelson-Morley experiment, which demonstrated things beyond the reach of the current science of their day, but the results of these experiments were not "paranormal" because they could easily be verified by anyone with the necessary equipment.)

    Metaphysical naturalism (also known as ontological naturalism) is the belief that everything that exists is part of the natural world, and hence that no supernatural realm exists.

    Methodological naturalism (also known as epistemological naturalism) is the belief that human knowledge can be maximized by assuming that the natural world is all that exists -- even if it's not, in fact, true that the natural world is all that exists.

    More generally, a pragmatic assumption is an assumption that is believed to be worth acting upon regardless of whether the assumption itself is true.

  2. Naturalism as a pragmatic assumption in science

    Nearly all scientists accept methodological naturalism, at least within science itself. A key part of scientific methodology is the idea that science should consider only naturalistic explanations and not consider supernatural explanations.

    Why? Because only natural explanations lend themselves to the methods of science. For example, "God did it" is not a testable, falsifiable hypothesis. And, because "God did it" -- even if true -- does not lend itself to the methods of science, it is also unlikely to lead to any further hypotheses which lend themselves to scientific methodology either. Thus, supernatural explanations -- even if true -- are a scientific dead-end. Only naturalistic explanations can be scientifically fruitful. Therefore, science must always look for possible naturalistic explanations.

    But does this, in itself, imply the actual nonexistence of any and all supernatural entities? On that question, scientists disagree, and so do philosophers.

    (For more information about methodological naturalism and why it is needed in science, even if metaphysical naturalism is not true, see these articles on my methods of science page.)

    Some would argue that the many successes of science, with its methodological naturalism, imply the truth of metaphysical naturalism -- or at least a strong likelihood that metaphysical naturalism is true. I disagree.

    First, an assumption can be scientifically fruitful yet still turn out to be false.

    For example, it was once thought that it would eventually be possible to make accurate long-range weather predictions. Thanks to the advent of chaos theory, it is now known that accurate long-range weather predictions are impossible. (For more about this, see What is Chaos? by Matthew A. Trump and Chaos Theory and Fractals by Jonathan Mendelson and Elana Blumenthal.) However, the earlier belief in the eventual possibility of accurate long-range weather predictions helped to motivate (and obtain research grants for) a lot of valuable meteorological research that might never have been done otherwise. Thus, belief in the eventual possibility of accurate long-range weather predictions (which I'll refer to henceforth as the EPALRWP) was scientifically fruitful in the sense of leading to greater scientific knowledge. Still, the EPALRWP turned out to be false.

    Besides being scientifically fruitful, naturalism has another key feature in common with belief in the EPALRWP. Both naturalism and the EPALRWP are optimistic estimates of the possible reach of human scientific knowledge. Naturalism assumes that everything that exists can, at least in principle, be explained in ways that lend themselves to scientific investigation by humans (or by human-created machines, if/when AI gets off the ground). Thus, naturalism is the most optimistic possible view of the potential longterm success of the scientific enterprise.

    Why are naturalism and the EPALRWP both scientifically fruitful? For essentially the same reason that optimism in general can be fruitful. If you believe you can do something, then you're far more likely to try it, and hence far more likely to succeed at it, than you would be if you believed you couldn't. In the case of naturalism, if you assume that everything has a good, elegant and testable scientific explanation, then you are more likely to look for -- and hence more likely to find -- good, elegant and testable scientific explanations.

    The scientific fruitfulness of methodological naturalism does not imply the truth of metaphysical naturalism, any more than the pragmatic fruitfulness of optimism in general implies that we live in a perfect pollyanna world. Most of the time, in most aspects of life, reality fails to live up to the most optimistic possible predictions.

    This observation doesn't, in itself, prove that metaphysical naturalism is false. The points to note here are that (1) naturalism is a necessary pragmatic assumption in science, yet (2) the pragmatic value of a given assumption, even its value in the search for truth, does not prove the assumption itself to be true.

    An even more dramatic example of the latter point can be found in the history of mathematics, another rigorous approach to seeking truth. It was once believed that all true mathematical statements could be proven from a handful of axioms. This idea was very fruitful as long as it lasted. However, in 1931 it was blown away by Gödel's Incompleteness Theorem, which proved the existence of some true-but-unprovable statements. Since then, it has been shown that there are many other unsolvable problems, including undecidable propositions and incalculable numbers. For an easy-to-read introduction to this topic, see The Omega Man by Marcus Chown in New Scientist magazine, 10 March 2001.

  3. Some apparent limitations of science

    Science has explained a lot. And there is good reason to hope that science will continue to explain a lot more. But will science ever be able to explain everything?

    One good reason to doubt this is that science, itself, has discovered some apparent limitations of science.

    One well-established example is the Heisenberg Uncertainty Principle. For an easy-to-read explanation, see the section titled The Heisenberg uncertainty principle in the Intro to Quantum Mechanics by Todd Stedl.

    More dramatic examples can be found in cosmology. For a good, clear explanation of current cosmology, see James Schombert's lectures for an astronomy course taught at the University of Oregon. A complete index of all the lectures can be found at about two thirds of the way down this page. The most interesting part, the discussion of the Big Bang, begins in lecture 15, but I recommend reading the entire series. The lectures contain quite a few statements like the following:

    Physics of the early Universe is at the boundary of astronomy and philosophy since we do not currently have a complete theory that unifies all the fundamental forces of Nature at the moment of Creation. In addition, there is no possibility of linking observation or experimentation of early Universe physics to our theories (i.e. its not possible to `build' another Universe). Our theories are rejected or accepted based on simplicity and aesthetic grounds, plus their power of prediction to later times, rather than an appeal to empirical results. This is a very different way of doing science from previous centuries of research.

    Our physics can explain most of the evolution of the Universe after the Planck time (approximately 10-43 seconds after the Big Bang).

    However, events before this time are undefined in our current science and, in particular, we have no solid understanding of the origin of the Universe (i.e. what started or `caused' the Big Bang). At best, we can describe our efforts to date as probing around the `edges' of our understanding in order to define what we don't understand, much like a blind person would explore the edge of a deep hole, learning its diameter without knowing its depth.

    In other words, the actual cause of the Big Bang is unknown, and probably unknowable. Yet the article also says:

    One thing is clear in our framing of questions such as `How did the Universe get started?' is that the Universe was self-creating. This is not a statement on a `cause' behind the origin of the Universe, nor is it a statement on a lack of purpose or destiny. It is simply a statement that the Universe was emergent, that the actual of the Universe probably derived from a indeterminate sea of potentiality that we call the quantum vacuum, whose properties may always remain beyond our current understanding.

    If the actual cause is unknown, then on what grounds can it be said that the universe was "self-creating"?

    As far as I can tell, the only grounds are that to say that the universe was not self-creating would be contrary to methodological naturalism. That the universe was "self-creating" is an assumption astronomers must make in order to have any hope of arriving at a scientific theory of the origin of the universe.

    But methodological naturalism is merely a rule of the scientific game. It does not imply that it's true that the universe was "self-creating." Nor, even if it is true, does it imply that astronomers necessarily ever will arrive at a complete scientific understanding of the origin of the universe. However, in order for astronomers to give it their best shot, they must proceed on the most-likely-to-be-fruitful assumptions, rather than settle for an obviously dead-end hypothesis like "God did it."

    Unfortunately, as the above-quoted lectures effectively admit, even the naturalistic theories of science may be leading us to some scientific dead-ends too. Instead of being created by a God, the universe is said to have been "probably derived from a indeterminate sea of potentiality that we call the quantum vacuum, whose properties may always remain beyond our current understanding." In other words, something which may turn out to be almost as unknowable as "God."

    Also interesting is Lecture 21, on the "anthropic principle." In particular, note the following statement: "Since the many-worlds hypothesis lacks the ability to test the existence of these other universes, it is not falsifiable and, therefore, borders on pseudo-science" -- another example of how even methodological naturalism does not always lead to testable hypotheses.

    Another reason to doubt that science will ever be able to explain everything is that much of science is undergirded by mathematics, in which it has been proven that there are many problems that are in principle unsolvable. This means, at the very least, that science will never have a mathematically complete theory of everything.

    Admittedly, the existence of many undecidable propositions in math does not, in and of itself, imply that there exists anything utterly beyond the potential reach of science. Nor do any of the other above-mentioned limitations of scientific knowledge constitute fundamental limitations of scientific methodology itself. For example, although precise measurement of the motion of tiny particles is impossible, a lot can still be inferred about those particles based on sufficiently large amounts of evidence.

    Thus, as far as I am aware, scientific methodology has not yet encountered its own analog of Gödel's Incompleteness Theorem or the arithmetical hierarchy of undecidable propositions. (If science were to encounter such a thing, it would prove the existence of a supernatural realm, in my sense of the word "supernatural.")

    Still, is there any good reason to believe that everything that exists is within the reach of science? I think not. Too much of what we have learned during the past century points to the idea that human knowledge is extremely limited, on all fronts. Why should the overall scope of science itself be any different? Perhaps it is, but there are no good a priori reasons to assume so, other than as a useful pragmatic assumption within the scientific enterprise itself.

  4. The paranormal and skepticism

    So far, I've shown only a likelihood that a supernatural realm exists. I have not established the truth of any specific supernatural claims. Indeed, it is probably not possible to establish specific supernatural claims, at least not with anything approaching certainty, because the supernatural is, by definition, beyond the reach of science, and thus probably beyond the reach of any reliable approach to knowledge.

    Therefore, skepticism toward any and all specific supernatural claims is still justified. The mere likelihood that a supernatural realm exists does not, in itself, imply that we are justified in believing any or all supernatural claims willy-nilly.

    Likewise for paranormal claims. While at least some skepticism toward any and all specific paranormal claims is reasonable, is it reasonable to believe that no paranormal events ever occur?

    Is it reasonable to believe, for example, that if a given phenomenon has genuinely been experienced by at least one human, then it should always be possible, at the very least, to provide scientifically verifiable evidence of the phenomenon's existence?

    Clearly, the answer to that last question is no.  Even on an ordinary mundane level, there are plenty of things that happen to us every day, which, if challenged, we would be unable to prove.

    The idea that "if you can't provide good evidence that it happened, then it didn't happen" is a good methodological assumption for historians and other scholars, who otherwise would end up believing all manner of nonsense. It's also a good methodological assumption in a court of law. But this assumption, itself, is clearly not true -- yet another example of how an assumption can be scientifically (or otherwise) fruitful though untrue.

    But what if a given phenomenon is alleged to happen to a lot of people? Shouldn't it be possible to study that phenomenon scientifically, or at least to prove that it has happened? Not necessarily.

    Even on a mundane level, there are plenty of phenomena which happen every day, yet which, while not beyond even the current reach of science, are, at least, very difficult to study scientifically. Examples include creativity and humor. There are many aspects of human psychology that are extremely difficult to study scientifically because, although they clearly do exist, they don't lend themselves easily to experimental methodology, because of the subjectivity and unpredictability involved.

    As I said, these things are not beyond the reach of science, just difficult to study scientifically. But there is no reason why there couldn't also exist some aspects of human experience that are real yet even more unpredictable and/or veiled by even more subjectivity than creativity and humor are, so much so as to be utterly impossible to study via scientific methodology. Indeed, if a supernatural realm not only exists but also impinges on at least some people's lives in any way, then we would expect it to do so in precisely these sorts of ways. Otherwise, its manifestations would lend themselves to scientific study, and thus its nature probably would not be beyond the potential reach of science.

    Of course, we should still be careful about jumping to the conclusion that a given phenomenon is even paranormal, let alone supernatural in origin.

    For example, many "spiritual" experiences apparently have a natural explanation. See Spiritual experiences and the brain.

    Also, mere coincidences do not count as paranormal phenomena. Coincidences are more common than most people think. See, for example, the following articles about the Birthday Paradox:

    In general, I recommend a study of the skeptical literature, to become more aware of possible natural explanations for various odd phenomena:

    Alas, many of the "skeptical" writers on the above sites seem closed-minded. Many seem dogmatically committed to metaphysical naturalism. Nevertheless, their explanations of various phenomena are worth reading.

    Anyhow, some truly weird things have happened, both to me and to other people I know. Therefore, it seems likely to me that some truly weird things have happened to a lot of other people too.

    It cannot be known whether these events are supernatural in origin. Probably, at least some of them, and perhaps even most of them, do have a natural explanation. However, until a natural explanation for a given event X is actually found, "X is natural" ia an undecidable proposition.

    In the meantime, people are entitled to make their own guesses as to whether X is probably natural or not. For scientific methodological purposes, one should always assume that a given event is natural. However, as I will explain in my article Reasonable (though not completely rational) theism, different pragmatic assumptions may be useful for different purposes.

    If your aim is not to advance science but just to assess a seemingly paranormal situation in your own personal life, then you should still look for possible natural explanations first. But there is no reason for you to avoid contemplating the event's possible significance in terms of your own personal spiritual beliefs. Your spiritual beliefs, whatever they may be, aren't on anything remotely approaching as solid a footing as the findings of science. But, if your aim is merely to make your own best guess about your own personal situation, then your spiritual beliefs may well be better than nothing.

Back to: