I recently found an article from 2011 making a point I’ve long made myself, that the entire notion of a “presumption of naturalism” being axiomatic to history and the sciences is both an error made by some historians and scientists and an apologetic bluff by Christian apologists—and that, instead, naturalism is an evidence-based conclusion in the sciences reached by long experience, and thus is theoretically revisable; it is also based on evidence, and therefore cannot be “swapped out” by simply changing one’s faith commitment or “preferring” a different axiom. I recommend the whole thing: Gregory Dawes, “In Defense of Naturalism,” International Journal for Philosophy of Religion 70.1 (2011): 3-25.

This reminded me it will be worth making a more detailed summary for folks to cite of why this is a correct assessment, both to educate historians and scientists who need to stop feeding Christian apologists such a handy talking point, and to call out those apologists for getting the entire historico-scientific enterprise wrong, and thus completely failing to realize or concede why history and the sciences have de facto conceded ontological naturalism: supernaturalist belief systems simply aren’t tenable anymore.

The Dawes Argument

Dawes points out that the only actual axiom of the sciences that isn’t negotiable is what he calls “the procedural requirement of history and the sciences,” which is simply “the demand that any claims about human beings or the world they inhabit should be supported by reference to some publicly-accessible body of evidence.” This is equivalent to what in Proving History I formally describe as the “basic principle of rational-empirical history,” which indeed I assert as axiomatic to any serious historical method, that “all conclusions must logically follow from the evidence available to all observers” (Proving History, pp. 20-21).

And Dawes rightly points out that this is all that is required of history and the sciences, and yet this “does not, in principle, exclude reference to divine agency” or indeed anything supernatural or bizarre. Every explanation has a shot at becoming evidentially respectable and thus restored to the status of serious consideration. It simply has to pass this one procedural test. Supernaturalism has simply never passed that test. And that is why it is rejected as a non-starter in all professional fields of knowledge. This is not “bias against the supernatural.” It’s simply a restatement of an observed fact: supernaturalism has simply never worked before; so it’s unlikely to ever do. If supernaturalists want to change this conclusion, they have to do so with evidence. Not disingenuous complaints about persecution and bias.

Thus, Dawes concludes, “modern scientists and historians” are “justified in their commitment to natural rather than supernatural explanations” and that commitment “is neither a priori nor non-negotiable.” It can be revised. But only “if theologians were to produce adequate theistic explanations” of any phenomena “and show that these were preferable to any proposed natural explanations” would historians and scientists ever be warranted in changing their commitment. It is simply that this has never happened. And yet had things turned out differently, “the metaphysical naturalism of history and the sciences would then be overturned and God would become part of the working ontology with which scientists and historians operate.” Or likewise anything bizarre or supernatural.

Dawes then rightly argues that this means scientists and historians are really, for all their hemming and hawing, actually de facto metaphysical naturalists. They aren’t really just “methodological” naturalists; saying they are is more of a soft way to avoid having to anger their believing peers; or for believers working at secular universities to avoid getting fired for being a kook or denounced as atheists by their believing peers. In actual fact the reason they are “methodological” naturalists is that centuries of accumulated evidence supports no other set of explanations as plausible. Supernatural explanations, like all crank explanations (from aliens to psychics to wizards to Vast International Jewish Conspiracies), are just too unlikely to be true to ever warrant spending any time or resources testing or considering them. And that’s just that. When pressed, even most believers will admit this.

As Dawes eloquently puts it:

[T]he metaphysical naturalism of history and the sciences [is] a kind of working ontology, a set of assumptions about what kinds of entities are likely to exist. [And] while historians and scientists do operate with an ontology of this kind, it should be regarded as nothing more than a provisional commitment, justified by reference to the history of these disciplines. It is provisional in that it is defeasible: it could (in principle) be overturned. It would be overturned if the theologian were to present a series of successful theistic explanations of the kinds of facts in which scientists and historians are interested.

Such explanations would conform to the procedural requirements of history and the sciences, in that they would appeal to publicly accessible bodies of evidence. They would posit the existence and action of God as the most adequate explanation of the facts to which they appeal. But until religious believers do this, the metaphysical naturalism of modern historians and scientists requires no defense beyond the practice of their disciplines.

Hence, “It follows that if” their arguments “were generally accepted as sound arguments, the existence of spiritual beings would become part of our science.” Ergo, the only reason such entities haven’t become part of our science is that there aren’t sound arguments for them. I would suggest this is a harsher thing to admit to, uncomfortable for any ardent believer, and awkward even for the secularist—for admitting it would offend too many friends and peers, and plunge them into endless arguments with what are really, honestly, cranks; an exercise most professionals would rightly prefer to avoid as a waste of their time. So to avoid this blowup, they invent nonsense about “Oh, don’t worry, I’m just a methodological naturalist.” Just to quiet the kooks down.

And yes, if you are uncomfortable with the conclusion that all accumulated evidence to date establishes supernatural explanations are implausible, you’re a bit of a kook. Or even a raging one, depending on your level of fanaticism. You can perhaps avoid the charge by retreating into a soft theism that accepts the conclusion, by building a barricade around your theology with such slogans as “God doesn’t intervene in nature” or “God doesn’t want us to believe because of evidence” or “God’s ways are untestable and unknowable” or “God can only be evident to the eye of faith” and such like. Dawes gives some examples of that kind of retreat. And points out their flaws. Though they are more flaws of avoiding the truth than denying it. (See also the example I examine of Brian Huffling.)

At this point I was particularly amused by Dawes having found a devastating admission in the writings of the fanatical Christian apologist Alvin Plantinga. Plantinga outright said: “the best arguments for the public rationality of Christian belief are not particularly successful—at any rate they don’t show that Christian belief is likely with respect to public evidence.” Ouch. But rather than join liberal theologians in retreating to a “safe place” of simply asserting God didn’t or wouldn’t leave us any historical or scientific evidence, Plantinga goes full kook by declaring we don’t even need evidence to believe in God. That’s how badly the evidence has gone for his worldview. Even he admits it.

Hence Dawes would rather we be more honest, eventually referring to attempts to define science axiomatically as “methodologically” naturalist in courts of law to get the Trojan horse of “Intelligent Design” out of state schools:

[A] successful legal strategy is not the same as sound philosophy. In any case, its success may be short-lived. In the long term, such arguments merely hand ammunition to one’s opponents, who can suggest (as we have seen) that this definition of science begs the question, is itself a confessional commitment, or is simply bad philosophy. …

It is, therefore, a mistake to say that “for science to be science, by definition it can pursue, identify, and entertain only natural causes” …. What we should say is that for science to be science, by definition it can “pursue, identify, and entertain” only those causes whose existence can be argued for on the basis of publicly-accessible evidence. This procedural requirement is the only non-negotiable commitment of history and the sciences.

Of course, admitting that would blow up the world. It would be declaring war on religion. And calling conservative Christians delusional. It’s a political conundrum. But intellectually, Dawes is right.

Supernatural Explanations Are Simply Bad Explanations

I recently discussed the same point already in respect to the history of science, and how history itself fails to find any reliable evidence of the supernatural. This is why we can’t claim things like “miracles are real because Jesus was raised from the dead.” He wasn’t. Miracle-free explanations of all the pertinent evidence are abundantly available that do a far better job of explaining that evidence. And since this same outcome has resulted for every investigable case yet investigated, no substantial prior probability can arise that would ever warrant wasting time testing supernatural explanations, any more than “aliens did it” explanations or “pychic powers did it” or “the Illuminati did it” or any crank notion whatever. What theists won’t admit is that their explanatory model has simply become crank. And this was solely a consequence of the way the evidence went. History has proved their beliefs inaccurate and obsolete.

The supernatural has simply never been a good explanation of anything (see, for example, my article on Bayesian Counter-Apologetics). Even conceptually it begins as an extraordinary claim rather than an ordinary one (see, for example, my article on The Argument from Specified Complexity against Supernaturalism). And yet no extraordinary evidence has ever been adduced for it, as logic requires (see, for example, my article on why Extraordinary Claims Require Extraordinary Evidence).

As I wrote long ago in Defining the Supernatural, this need not have been the case. If we lived in the world described by the Harry Potter novels, or Jonathan Strange & Mr. Norrell, or the film Constantine, we could have vast evidence confirming supernatural explanations are sometimes valid, and they would thus enter the repertoire of viable, testable scientific explanations of observations. Indeed whole sciences would arise to study them. What is remarkable is that we don’t find ourselves in such a world. Every attempt we’ve made to test if we are has failed to find any evidence of it. And this has been going for hundreds of years. The odds that we are deceived in this are next to nil by now. And that’s simply because that’s the way the evidence went. It’s not an axiom or a presumption or a faith commitment. It’s a result.

This is indeed one of the clearest examples of progress in philosophy. There simply is no honest rebuttal anymore to the Argument to Naturalism. Science has de facto embraced this conclusion of philosophy just as it embraced other conclusions of philosophy like empiricism. Science is in fact built on top of this philosophical architecture now. For a reason.

The Observed Base Rate of the Supernatural Is Zero

I similarly showed in Proving History that the presumption of naturalism in history today is not actually a presumption but a conclusion from past experience: it is an observation of the actual base rate of success for supernatural explanations (pp. 114-17). Which so far has been zero. That doesn’t entail the prior odds are zero (because all confirmed observations won’t include all possible things that have happened, only at best a more-or-less random sample of them); but it does entail the prior odds must be so extraordinarily low as to not warrant considering—until we have some truly exceptional evidence that changes that.

It didn’t have to be that way. If the supernatural actually existed, it would have had an observable base rate by now, and thus would no longer be an extraordinary claim, and thus would no longer require extraordinary evidence to prove. The need of extraordinary evidence is an empirical consequence of the thorough failure of supernaturalism to explain anything better than naturalism already does. It’s a consequence of history—the history of a world with no observable supernatural phenomena in it.

As I explained in my discussion of Crank Bayesians, when faced with some phenomenon we want to explain—like what causes lightning or genetic complexity—we look at what has actually turned out to be the explanation of everything else similar through all past time; and what has turned out to be false. We then can have a rough frequency estimate of how often it’s been “magic and ghosts” (of which gods are only a subset), and how often it’s something decidedly not like either. So far, after hundreds of years of millions of proper investigations, nothing has ever turned out to be explained by magic or ghosts. Everything has turned out to be explained by mindless physical phenomena; especially countless things we used to claim were caused by magic or ghosts. So based on our background knowledge, “magic and ghosts” are literally the least likely explanation of anything. That’s not an opinion. That’s not a mere belief. That’s a material fact of a vast array of frequency data, accumulated by thousands of experts over hundreds of years. In other words, it’s an established fact of reality.

If we count up all the things in history we at some point couldn’t explain, or thought was explained by magic or ghosts, and then securely found out what the actual cause was (so that it is now approximately a universally accepted fact of science or history), how many of those things turned out to be magic or ghosts? If the answer is zero (and it is), and the number of those things is in the millions (which have reached that degree of investigation, so that it is now a known fact of the world what causes them, not just a mere belief or speculation), then the prior probability the next thing you ask the cause of will have been caused by magic or ghosts is logically necessarily millions to one against. And if the number of such things is in the billions, it’s billions to one against; if in the billions of trillions, then billions of trillions to one against. There is no rational escape from this consequence. And that’s even just for the simplest supernatural hypotheses; most actual such hypotheses incorporate enormous arrays of unproven assumptions, rendering them even more improbable.

Which is why gods, as a sub-category of magical ghosts, must have even smaller priors than that, as a logically necessary fact. Since the set of all “magical and ghostly causes” includes gods, as sub-category of a very specific kind of magical ghost, it follows that the prior probability that some magical or ghostly cause is to credit is necessarily greater than the prior probability that a specific magical or ghostly cause is to credit. Claiming otherwise is well known as the conjunction fallacy. So, the claim that “either a mindless but magical Tao force caused the universe or a god caused the universe” must always be more probable than the claim that “a god caused the universe.” It is logically necessarily always the case that P(God) + P(Any Other Supernatural Explanations) > P(God).

As I wrote long ago:

Historians appeal to natural explanations because they know that such explanations have always worked before, in every case that could be thoroughly tested. I repeat: they have always worked in every other case that could be thoroughly tested. This is no small point, and it stands as very compelling evidence in favor of natural explanations. And the case is secured when there are other good reasons that support a particular natural explanation.

The healing miracles of the New Testament are a case in point. They all look like psychosomatic events, not magic. Jesus never does anything truly magical, like remove cholera from the land or cure an amputee (except when Luke adds to Mark’s story of a severed ear an account of Jesus curing it—a claim unknown to Mark and thus not even in Luke’s source, which makes it more likely fabricated than genuine). Rather, Jesus only ever cures a few blind, paralyzed, and “demon possessed,” all well-known psychosomatic conditions—possibly the conditions Christian missionaries themselves most commonly “cured,” thus the Gospels have Jesus establish the precedent. Naturalism thus leaves us with far more probable explanations of the data than any supernaturalism can muster.

Edward Shorter is a historian of medicine in the modern age, and in his book From Paralysis to Fatigue: A History of Psychosomatic Illness in the Modern Era (1992) he describes a problem faced by any historian of psychosomatic illness:

With the exception of those in the last chapter, the patients described in this book are all dead. Is it certain that their symptoms were not caused by an organic disease? Retrospectively, it is not. There is only the presumption of psychogenesis, based on (a) the history of the illness, such as paralysis after seeing a frog in the road, and (b) the response to what was essentially placebo therapy, such as hydrotherapy or administration of a laxative. These two circumstances give certain symptom patterns a flavor of psychogenesis. (p. 4)

In other words, we know what psychosomatic illnesses look like: the symptoms are caused by things we have since confirmed don’t have those effects (like, seeing a frog causing paralysis); and the symptoms are removed by things we have since confirmed have no actual medicinal effect (like a charismatic faith healing act). If Jesus or missionaries emulating his myth really could miraculously cure wounds and ailments, it is simply improbable that all they would ever be seen curing are established psychosomatic conditions using established psychosomatic methods. By contrast, if Jesus or Christian missionaries weren’t supernaturally empowered, then the data are exactly what we’d expect to see.

This is the way every claim goes. And worse, a great many times we even outright confirm evidence of fraud, fabricated or exaggerated accounts, or wholesale mythmaking (I provide an extensive bibliography in Sense and Goodness without God, §IV.1, pp. 212-13, but just for a short list, see: Polidoro, Secrets of the Psychics; Blackmore, In Search of the Light; Nickell, Entities and Looking for a Miracle and The Science of Miracles; Cornwell, The Hiding Places of God; Schumaker, Wings of Illusion; Hines, Pseudoscience and the Paranormal; and Randi, Flim Flam! and Encyclopedia of Claims, Frauds, and Hoaxes of the Occult and Supernatural). By contrast, never do we ever discover a genuinely supernatural event or effect. For half a century the million dollar Randi prize was never claimed.

That should not be what turned out. A supernatural world should have been well documented by now. We surely should have proved even just one instance of it by now. Not zero. Only if the world lacks the supernatural would we be so likely to have found none. And found none we have. The effect this has on vanishing the prior probability of supernatural explanations, to below even that of “aliens did it” or “psychics did it” or “the Illuminati did it,” I lay out in in The End of Christianity, regarding “evidence of supernatural design” in the universe (pp. 282-84). I make a similar point in my talk on Miracles & Historical Method.

Imagine you can take all claims of the supernatural, all miracle claims, all claims of gods or ghosts or wizards intervening in the world, and sort them on a table. We sort them into two piles: all the ones that we haven’t and can’t properly investigate so as to test what really caused them (owing to the evidence no longer being accessible, for example); and all the ones that we did thus properly investigate. The first pile we can’t use to argue anything, because any item in it is as likely to have any one cause as another—we can only tell how relatively likely different explanations for it are from examining the second pile, the one containing everything we did get to investigate. So we look at that second pile. How commonly did our investigations find a real supernatural cause? Never. Zero. What did we find? In every case, without exception: some natural cause—from fraud to error to mythmaking. So the odds anything in the first pile are going to turn out differently is dismally small, approaching zero.

Sorry. Our World Turned Out to Be Godless.

I had already made all these points fifteen years ago when I published my first book, Sense and Goodness without God: A Defense of Metaphysical Naturalism, a book that still holds up pretty well today (despite only a few needed corrections or updates). Part IV of the book is on the overall reason why we can and should reject as implausible everything supernatural (pp. 209-90). The reason is not faith-based or axiomatic or presumptive. It’s entirely empirical and evidence-based. Adapting here one of the lines of argument I wrote up there:

In the 1st century we had no idea what caused lightning, but we had a cogent supernatural explanation: divine or demonic agency, ranging from the anger of Zeus to the combats of evil spirits in the clouds. Yet at the same time scientists of the 1st century proposed that lightning was caused by friction between colliding clouds, by analogy with colliding flint stones. This explanation, though still speculative, was much closer to the truth. It only lacked the yet-to-be-discovered sciences of electricity and pressure.

In such a way, scientists throughout history have found that what we call “natural” explanations keep working, unlike supernatural explanations: the stars and planets actually followed predictable routines that had nothing to do with human events, tested drugs cured the sick more often than spells, agriculture flourished under scientific care but floundered under prayers and magic. Then scientists found that atomic and other “naturalistic” explanations for every phenomenon had a much wider explanatory power than supernatural theories, predicting more things, more successfully. Thus, they correctly guessed they were on to something, and stopped accepting “supernatural” explanations because they constantly failed and never had any evidence to sustain them.

This is why scientists even in antiquity pursued only “natural” theories (as I document now in The Scientist in the Early Roman Empire). And in doing so they got very close to the truth, articulating explanations for sound, light, evolution, weather, poison, disease, and getting far closer than any theologian ever came to what actually turned out to be true. Unfortunately, this brilliant discovery was thwarted by the rise of Christianity, which all but put science on hold for a thousand years (and Yes, the Dark Ages Really Were a Thing), relying instead on “supernatural explanations” and the adage that it was vain to study nature when we ought to be saving our souls instead. It was not until the Renaissance, when pagan science was rediscovered, that the bias in favor of what we call the “natural” was taken up again. Then lo and behold, every century since has seen unprecedented progress in our knowledge and mastery of the world (that it was indeed pagan values the Christians had to recover to reboot science I demonstrate in my chapter on the point in The Christian Delusion).

So there is some merit to the presumption of naturalism in scientific practice, what scientists will now call “methodological naturalism.” It has proven valid as a rule of thumb, one that is more likely to produce success, saving us from wasting time and resources on blind alleys. Still, exactly as Dawes would later argue, it is not an absolute law. There may yet be evidence of the supernatural just around the corner. But the evidence against this possibility now is so vast, and the evidence for it so feeble, you may as well count on catching Big Foot or finding a Leprechaun’s gold.

This is merely a global application of a principle essential to any successful quest for the truth: the burden of proof for any claim is on anyone who makes a claim contrary to established facts, or grounded in fewer established facts than any competing hypothesis. For when facts are established, with widespread and multi-faceted corroboration, the odds are clearly on the side of their being correct when pitted against anything contrary. This is why the burden of proof weighs heavily on those who advocate belief in anything supernatural—a burden they have so far never even come close to meeting. To the contrary, whenever closely scrutinized, they come up with nothing.

Instead, all the evidence we have so far, about us and the world, is best explained by metaphysical naturalism (or in more recent parlance, “ontological naturalism”). Naturalism is therefore probably true, and that means there probably is no god, no angels or demons, no ghosts, spirits or souls, no psychic or magical powers, no miracles or infallible scriptures, but instead only a blind, mechanical universe. Science and history proceed as if this is the case because all the evidence science and history have actually accumulated over the centuries exhibits nothing else being the case. We thus reject theistic and supernatural explanations for the same reason we reject alien interventions, magical explanations, or vast megaconspiracies: not because of any “bias against the supernatural” (or “against aliens” or “against megaconspiracies”) but because of the complete failure of such paradigms to ever produce an effective or demonstrable explanation of any phenomenon in the world, by any rational method.

The same conclusion can be reached by several other independent lines of observation, which further corroborates its reliability. For instance, there are no minds without complex physical brains, therefore it’s not likely there can be a divine mind, since there is clearly no gigantic brain for it. Whereas if God can have a mind without a physical brain, it’s inexplicable why we need them. It’s far more probable such a god would create beings with minds like His, minds that could not be damaged or destroyed, rather than minds needlessly dependent on something so fragile as a brain. By the same reasoning, there is no plausible reason why an Almighty would need billions of years and trillions of galaxies to accomplish his ends through long, deterministic causal processes. But that is exactly what we should expect if there is no god, but only nature. There is a reason only naturalistic cosmological theories pass scientific review. Theists simply don’t have any cogent explanations of the actual facts, beyond mere assertions and speculations. And assertions and speculations are not science.

Conclusion

Gregory Dawes makes several points with which I agree: (1) methodological naturalists are really de facto ontological naturalists (and if theists, they have to compartmentalize so as not to think about the actual reason this methodology is required); (2) history and the sciences are ontologically naturalist because the historical accumulation of evidence in every field has evinced no other way we can honestly expect the world to be (supernatural explanations have simply turned out to be a really bad bet: never in evidence, always outdone by better explanations, and really bad at actually accounting for all the data); and (3) the only actual axiomatic presumption in all our sciences is that of only allowing rational inferences from publicly available evidence, a requirement supernatural explanations could easily meet—if anything supernatural actually existed. It’s only because they don’t pass this test that supernatural explanations are now deemed unscientific and ahistorical. And for the larger consequences of this observation, see Why A Neo-Aristotelian Naturalism Is Probably True.