Part 2 of my series on the new Macmillan reference Theism & Atheism: Opposing Arguments in Philosophy: my discussion of the Argument from Miracles, which turns that argument on its head. Far from being evidence for theism, the collective evidence regarding miracle claims throughout history is actually evidence for atheism. (For a description of the entire volume and my role in it see Part 1.)

Miracles

Half this chapter was written by Arif Ahmed of the University of Cambridge; the other half by me. Ahmed tackles the questions, “What are miracles? Is it reasonable to believe that such events either could or did occur? And is belief in miracles really a matter of belief in this or that event, or is it a way of interpreting the events?” Including a defense and update of Hume. After his excellent survey of those questions, I then show that, in fact, seen in historical context, “The miracle stories found in the texts produced by hundreds of religions in recorded history do not increase the probability of theism. In fact, examined in the aggregate, they decrease the probability of theism,” in my section titled “At Present, Miracle Claims Undermine Theism” (pp. 219-26).

I have made this point before. Here I produce a more detailed case for the conclusion, including sections on “the verdict of rational historical empiricism” (laying out the reasons why mainstream historians now reject miracle claims and are correct to do so), “the real origin of miracle claims” (laying out the only causes of miracle claims historians have verified actually occur) and “why apologetics is ineffective” (laying out why attempts to bypass all of this are uniformly irrational). I also refute at the methodological level Craig Keener’s new attempt to defend miraculism, and explain how our position in no way derives from an arbitrary “bias against the supernatural,” but from a rational, statistically valid response to the actual evidence of science and history to date.

My overall conclusion:

It is not only that we lack the evidence we should expect to have found by now for any of the kinds of miracles commonly claimed (whereas we have all the evidence we need that the other causes of miracle claims not only exist but are frequent). It is also that we lack the evidence we should expect to have found by now for miracles that are actually impressive, as in indicative of an overwhelmingly superhuman power, not merely a superhuman power that is conveniently trivial enough to go unnoticed by all but a few.

These two observations are very improbable on any theistic explanation of miracle claims, but are exactly what we expect on any sound atheistic explanation of miracle claims. Therefore the actual evidence surrounding miracle claims is evidence for atheism. Not the other way around. Theists can only get the opposite result by leaving most of the pertinent evidence out, or inventing improbable ad hoc excuses, which is an irrational method. Hence theism is irrational.

Likewise:

Living gods don’t need ancient poorly attested miracles as evidence of their creeds. Living gods can work living miracles. The reliance, therefore, on long dead tales to support the existence of living gods, is a fallacy of the first order. It would only be necessary in a world without gods. Which is why we can know such is the world we live in.

Likewise, anything claimed a miracle today either dissolves on inspection or test, or evades all inspection or test. A fact only false claims would require, and that no morally responsible God would allow. Again, that’s evidence for atheism, not theism. (See Miracles & Historical Method and Resurrection: Faith or Fact as examples.)

Schnall on Miracles

Ira Schnall of Bar-Ilan University in Israel attempts to make the contrary case but in so doing exemplifies every point we made in our earlier section about how theism can only be defended irrationally. Schnall’s chapter is described as:

In this chapter, we first examine what a miracle is supposed to be from a theistic point of view. Then we consider whether, or to what extent, reports of miracles are to be believed. Finally, we deal with the role of miracles in theistic religions, and in particular, whether miracles can establish the truth of theism.

Straightforward goals. And this is all worth reading for anyone who wants to correctly understand a theist’s position on this—and many atheists don’t. For example, too many atheists get completely wrong what theists define miracles as and actually believe them to be.

Schnall coherently defines a miracle as “an event that involves temporary suspension of the natural order by a supernatural agent.” And he elaborates on several different theological ways of understanding that statement. He allows that natural but providentially arranged coincidences can also be called “miracles” in a figurative sense, but he proceeds to exclude those kinds of miracle from the remainder of his discussion. (But if you want to explore my take on those, see Everything You Need to Know about Coincidences.)

Schnall then spends a lot of time trying to rebut Hume, of course. But he never addresses any modern revisions of Hume. His chapter was thus already obsolete before he even submitted it to Macmillan’s editors. His argument against Hume amounts to saying “that the biblical accounts of God’s revealing to us that He will perform miracles, even if we are suspending judgment as to their veracity, serve to lend some credence to subsequent miracle reports” (that’s a direct quote). This is irrationally fallacious from start to finish:

• Revelations, Schnall subsequently admits, are also miracles we’d have to suspend judgment on (else we argue in a circle). And it is not logically possible for a premise on which we have suspended judgment (which entails assigning its truth an epistemic probability of 0.50 or 50%) to increase the probability of a conclusion by any amount at all. It therefore lends no credence to it, not “some.”
• A generic statement that “God will perform miracles” has no logical connection to any specific miracle claim; it therefore tells us exactly nothing about which miracle claims are true, much less that “most miracle claims will be true,” even though only the latter can get from “revelation” to a conclusion that any given miracle claim is more likely than not true. Hence such a statement even if believed still lends no credence to any particular miracle claim’s truth, not “some.”
• There is abundant evidence that claims attributed to God are fake, being invented or merely imagined by human beings, and frequently false; therefore it is irrational to think such claimed revelations “lend some credence” to anything, much less claims that are already extremely improbable.

Instead:

• The only way to warrant believing that any particular miracle claim is more probably true than false (and thus the only way to add “some” credence to any miracle claim) is if we have a large enough pool of miracle claims that have already been reliably tested and confirmed. There is no way to bootstrap your way from having only untested claims, to any base rate of such claims being true.
• Even if we had any tested-and-verified miracle claims (and in actual fact we have exactly none), their base rate might still be small. So it could still be that most miracle claims are nevertheless false. Which would mean that of any untested miracle claim, we should then assume it’s probably false. The exact opposite of “lending it some credence.” And that’s even in a world where we have confirmed a lot of miracle claims are true. And we live in a world where none have.
• Given that hundreds of miracle claims have been investigated reliably and none were confirmed to be true, the actual base rate of true miracle claims on testing alone is quite small, and possibly zero. When we add in all the other things that have to be true for which there is no evidence (e.g. the existence of even the supernatural at all, much less highly bizarre and specific versions of it, and yet stranger and more specific versions that would even bother performing the specific miracle claimed and do so only rarely), epistemically, the highest possible base rate of true miracle claims becomes so small as to render them all inherently absurd.

As Ahmed explains in his own chapter, presciently refuting Schnall, Hume was ultimately right: only testimony whose falsity is more improbable than the claim testified to can justify believing that claim—to any degree at all. There is no logically valid way around this (see my demonstrations in Proving History). By not addressing any of the modern revisions of Hume, Schnall completely gets wrong the actual Neohumean argument against miracles. The entire issue is one of relative frequencies: What happens more often? That when people make wildly extraordinary claims they are lying or mistaken? Or that when people make wildly extraordinary claims they are reporting the truth? The latter simply isn’t what usually happens. Human testimony therefore is not likely to ever satisfy the requirement for rationally warranting belief in the miraculously extraordinary.

There are exceptions to this rule; which Schnall acknowledges. Such as a large number of demonstrably reliable witnesses, whose collective testimony would indeed be less probable if their testimony was false, than if the extraordinary event testified to happened. But for this to be logically valid, we have to have this testimony from them (not merely second or third hand) and we must have actual evidence of their reliability (which means not merely their honesty, but also their competence to not be fooled or mistaken). Those conditions have never been met for any miracle claim ever in the entire history of the human race. And that is why no miracle claim today is credible. Circumstances could change that conclusion. But they haven’t yet. And that’s that.

Confusing Events with Their Causes

In addition to having no rational defense of belief in any miracle, because of ignoring all advances in the relevant philosophy since the 18th century—a classic case of bad philosophy, leading to irrational armchair conclusions, which typify even peer reviewed defenses of theism, thus demonstrating theism is irrational—throughout his chapter (but for one subsection that doesn’t really illuminate the matter) Schnall regularly confuses “events” with “causes.” He rarely makes the crucial distinction between whether an observation is true (“The Bible says the Apostles saw Jesus fly into space” or “I saw Aleister Crowley cure a paralytic with a crystal wand”) and whether that observation’s explanation—as in, what caused it—is even supernatural, much less specifically “God.”

One has to first establish an observation is even true before one can then evaluate what its most likely cause is.

For example: That I saw Aleister Crowley cure a paralytic with a crystal wand could be entirely true and entirely credible—but it does not follow that he actually did what I saw. We can thus prove I really saw that. But we have a lot more work to do before we can believe that what I saw is what really happened.

We already know from countless investigations that those kinds of events turn out to be cons or delusions; and we’ve never once confirmed otherwise. Therefore we should simply assume that’s what they all are, until we get a well-investigated example that’s thus proved not to be. Schnall completely ignores the epistemic requirement of ruling out cons and delusions, and never acknowledges the material fact that the base rate of them is so high as to require our ruling them out in any particular case before ever believing any miraculous event has even occurred, much less what caused it, which is an entirely different question.

Likewise: That the Bible says the Apostles saw Jesus fly into outer space may itself be entirely true—after all, the Bible does say that. But it does not follow that what the Bible claims is itself true. We have no statement from any Apostle that they saw that; the author of that statement (“Luke”) never even says he met any Apostle, much less heard them say that; no prior source says that, not even Luke’s known sources (“Mark”; “Matthew”; or even, so far as we know, “Q”), not even the only source we have who actually says he met any Apostles (Paul); and tales of that kind are so routinely invented for heroes and demigods that we have no reason to believe this is any different. So the claim itself is not even believable. And that’s the fate of most miracle claims; the rest go the way of Aleister Crowley’s magic wand.

It’s all the worse that even if somehow we could prove someone did claim to have seen something like a man flying spontaneously into outer space (and we never have), the frequency of lies, delusions, and hallucinations is known to be so high, particularly among the devoutly religious, that we have no reason to believe they are even telling the truth or that what they saw actually happened. And even if we could get to that stage (and we’ve never gotten anywhere near that stage for any miracle of such kind), we still have to ask what caused it. After all, the Acts of Peter claims Simon could fly into space using sorcery. How do we know Jesus didn’t do so as well? And that’s already an absurd explanation. The intervention of extraterrestrial technologies is already more likely than that, and already itself absurd—whereas “hallucination” and “lying” are not absurd.

Thus no matter how improbable those explanations are, they remain vastly more probable—in fact by far the most probable explanations known of any actual such observation. And most claims of such an observation can’t even be established as actual! The evidence required to verify such a fact is thus far greater than Schnall recognizes. So he never confronts the central reason that miracles are not believable: the fact that they have never been confirmed to occur. The observations alone, even just the claim of one, already have far more probable causes, causes which are not “God performed a miracle.” And that’s why we never get to any warranted claim that God performed a miracle. (And yes, we could have sufficient evidence, even scientifically, to believe that He did: see my discussion and examples in Defining the Supernatural; and my new example below.)

By ignoring all of this and never addressing any of it, Schnall demonstrates how irrational his commitment to miracles being real actually is, and thus how irrational theism is. He at least avoided the most typical method used by theists to justify believing in miracles: lying about what their sources even actually say (see, for example, Did the Apostles Die for a Lie? and Then He Appeared to Over Five Hundred Brethren at Once!; or to compare with Schnall see my old critique of In Defense of Miracles). But even when Schnall struggles to stick to the truth, he abandons everything humanity has learned from science as to the required methods of verification and test, and gets to no rational basis for his conclusion. Yet ends up believing his conclusion! And that’s how we know theism is irrational.

How You Could Do It

Take theism out of it so you can avoid any religious bias that might cloud your reason. Let’s say instead that someone claims to have discovered sorcery actually works. And we gather countless reports of effective spells, including ancient accounts, because now we are posed the task of determining which of those accounts are of real magic, and which mere fables or fictions. We can use as a model the novel Jonathan Strange & Mr. Norrell by Susanna Clarke, which depicts exactly that scenario.

What we would need here, is exactly identical to what theists need for miracles. But they never do it, they never get it, and thus never have any rational basis for believing what they do. Which establishes, once again, that theism is irrational. Only theisms that ditch the claim of miracles can survive this criticism; but no theisms of that kind are even internally coherent. Why believe in a God who never does anything whatever? Isn’t that by definition not a God? “Well, I don’t believe in any miracles, I just believe God magically created the universe.” Which is a miracle. So you can’t actually ever “ditch” the claim that miracles happen and still justify believing in any real God; so you are simply always stuck having to meet the same rational epistemic requirements for believing in God as would justify believing in sorcery.

In Clarke’s novel, Mr. Norrell is faced with the need of proving to important persons in British government that he is now a sorcerer who can cast spells, that “magic has returned to England.” Jonathan Strange, an intellectual among those persuaded by Norrell’s demonstrations, then starts practicing this magic himself, and soon encounters the question of which ancient tales of magic are real or false—in order to ascertain what spells he might be able to discover or recreate. This research leads him eventually to discover an enormous hidden network of paths through and to every mirror on earth, by way of another world, a parallel dimension, with anciently eroding “roads” from place to place, called “The King’s Roads,” after the legendary Raven King of yore who supposedly created that network by some since-lost fantastic magic. This all transpires in the early 19th century, overlapping the Napoleonic Wars, in which this newly discovered magic, now under royal patronage, is soon deployed as a weapon.

Let’s suppose this actually happened. What would it take for us to believe it? We can imagine two different scenarios, both alluded to as possible in the novel: one, by which this magic would wane again and thus no longer work; the other, by which this magic, once unlocked, remains extant.

Of those two scenarios, if we’re being honest, only the latter is analogous to God. Because God doesn’t “vanish” or “fall asleep.” If he performed miracles anciently, he should be doing so presently, indeed all the more, as the population in need of them is now a thousand times in size—so miracles should be thousands of times more frequent. You can explain your way out of that with a bunch of made-up “assumptions” about how God would behave differently than any other person in the same circumstances; but such “gerrymandering” your theory would only reduce the probability of that God existing, not rescue it from disproof as you might irrationally have thought.

But let’s set this aside and grant your irrational assumptions, no matter what they are. I’m confident you can make up all sorts of bogus reasons for God to act exactly as weirdly as you need him to in order to avoid ever being disproved—that’s how all irrational beliefs are always defended, so your doing that only further proves how irrational theism is. But since you are committed to being irrational, you won’t recognize this—because, by definition, only a rational person will—so I’ll coddle your irrationality this much for now.

What remains is scenario one: God performed tons of miracles in antiquity—parted seas, rained fire from heaven, turned people into salt, transformed sticks into snakes, raised the dead, turned water into wine, became incarnate, flew into space, mystically murdered thousands of pigs, erased the sun. On and on. But now he doesn’t. Because reasons. What would it take for us to believe God really did those things? The same exact things it would take for us to believe some wizards briefly revived magic in England during the Napoleonic Wars.

The records we’d have if that actually happened, especially as told in the novel, would be extensive. They would range from government documents and journalistic inquiries and published scientific papers recording carefully controlled experiments and observations, all the way to (just as for the holocaust) thousands of eyewitness testimonies. Not some claim that there “were” thousands of witnesses. Actually thousands of accounts actually written by those witnesses. These would include witnesses consulted independently and thus who couldn’t have coordinated their testimony, witnesses interrogated skillfully and thus whose accounts were vetted for their precision and credibility, and so on.

There would be soldiers from both sides of the war attesting to the magic used, and the outcome of battles would have been decided by it. There would be statesmen and spouses and people on the street who saw this stuff and wrote it down or had it written for them by journalists and historians and investigators who can verify actually having interviewed them and that what they recorded the witness approved as accurate. There would be skeptics attempting to disprove the magic and recording their failure. There would be vast numbers of hospital records of magical healings.

The probability of all those records being lies or hallucinations or mistakes would be extraordinarily small. So small it would be undeniable that some wizards really did turn the tide of history with magic, that England really did experience a few decades of genuine marvels effected by well-recorded magical procedures. Because the probability of some vast conspiracy to doctor those records, or any other incredible explanation of them, would be far less than the probability of the miracles thus attested to. Therefore the authenticity of those miracles would be as certain as nearly any other fact of history.

One might still have the question of what caused that magic to work. But if we maintain the thought experiment as defined, that what the novel depicts really did happen, then Strange and Norrell and others will have left extensive records of their researches into how and why magic works. The question might not have been answered to anyone’s satisfaction, but that would not matter: it would be clear the world we live in is imbued with the supernatural, even if it lies dormant from time to time. In such a world, claims of miracles would not longer be unbelievable.

One could then have had enough information surviving from antiquity to believe Jesus effected his miracles by known magics as explored and replicated by the likes of Strange and Norrell. Just as Strange verified the ancient tales of the Raven King’s sorcery by actually confirming its surviving effects. He found the King’s Roads. He found fragments of the King’s spellbooks, and verified their recipes worked and matched ancient tales of their effects. One could do the same for any God’s miraculism, through prayer or following around holy men or having teams everywhere ready to investigate any new marvel. If only that miraculism were real. For if it were, it could always in some way be replicated, tested, verified. Unless God deliberately chose to prevent you verifying it. Which would mean God wishes you not to know it’s real. And who are you to defy God’s will?

Conclusion

If we lived in Jonathan Strange’s world, that’s the kind of evidence we’d have. Likewise if we lived in the world imagined throughout the Old Testament, we’d have the analogous evidence by now. But that never happens, and has never happened. For any supernatural claim—whether of ghosts, magic, gods, demons, angels, psychic or mystical powers, faerie lands, heavens, hells, even just disembodied souls (for an elaboration of this point, see Part IV of my book Sense and Goodness without God; and for the epistemic logic of it, Proving History, index, “miracles”).

And that’s why miracles are never believable. If the world were the sort of place miracles really occurred, we’d have tons of solid evidence of that fact by now. Yet we have accumulated no solid evidence of it. None. To the contrary, all the sound evidence we’ve acquired so far has only found miracle claims to be lies, cons, errors, or delusions—or conveniently evading any reliable inquiry or test. A world with miracles in it would in all probability look differently than that. Which is what proves we don’t live in that world. Anyone who concludes otherwise is simply being irrational.

Continue on to Part 3