I am near to completing my forthcoming rewrite of Sense and Goodness without God (which I will be retiring soon, so if you want to collect the original edition before it goes out of print, better get it now!). And one of the things I am doing is removing things that have gone out of fashion or matter less and replacing them with newer, more relevant material. Some of what I have cut isn’t useful anymore. But some could be, just not useful enough to clutter a book on positive philosophy for atheists. One of those things is my section on prophecy. After twenty years, my conclusion is, if you still think prophecy is a thing, you are so behind the curve that you need to absorb sensible science and philosophy first. And that’s what my new book provides. It doesn’t need to lower the bar to argue with the epistemic equivalent of illiterate witchdoctors. There are other books for that.

So I pulled that section out and am putting it here. As such it is a briefer and tighter version of my older articles on this. So this is really the only thing you need read to get the gist of why you should not still be falling for claims to miraculous prophecy in the 21st century. We throw women around the moon now. You can catch up.

There are more recent articles that cover other aspects of this that are still worth reading, like Behold Babylon USA! on the silliness of conspiracy-theory retcons of the Book of Revelation (where I explain what the author of Revelation was actually talking about and how his prophecy actually failed, spectacularly), which I also have done for the Book of Daniel (in How We Know Daniel Is a Forgery and Debating the Authenticity of Daniel). Likewise Everything You Need to Know about Coincidences covers the basic problem that the human brain is bad at understanding how common random coincidences are or how to determine the probability of one, which explains a lot of fallacious folly today about prophecy. I also addressed a non-existent prophecy in Is Christianity Exploding in Asia? A Critical Thinking Test, which is another way for prophecies to be false: to not even exist. And as I recommended last month, The Case Against Miracles covers a wide range of the folly of biblical prophecy-mongering pretty well.

But here’s something on the underlying methodology of evaluating prophecy as a kind of miracle. For theory of miracles generally, see articles in my supernatural category, but especially Naturalism Is Not an Axiom of the Sciences but a Conclusion of Them, with Daniel Bonevac’s Bayesian Argument for Miracles and Theism & Atheism: Miracles. And since almost all prophecy claims are actually conspiracy theories, you might want to read up on how Christianity Is a Conspiracy Theory. But now, to theory of prophecy.

Miraculous Prophecy Defined

In general use, a miracle is any event that more probably has a supernatural than a natural cause. A kind of miracle in history that is an animal all its own is prophecy. Unlike mere marvels, a prophecy is itself an ordinary statement, and refers to otherwise normal historical events. What makes the prophecy miraculous is its ability to predict those events. Thus, at first glance you don’t need to rely on anyone’s claims that a miraculous prophecy happened: you can check for yourself. If the prophetic statement truly predates the event foreseen, then you have all the evidence you need: proof of a miracle right before your eyes—that is, if such a thing could not happen by accident or through natural means, and only some superhuman intelligence could be at work. And that’s always the rub.

Of course, some “natural” psychic power might be to blame—or astrology, or casting bones or reading the inside of turtle shells or something. For the difference between the natural and the supernatural see Defining Naturalism: The Definitive Account. But “natural” miracles would still count as paranormal events. And there is no evidence of those, either. Psychics are always full of shit. And there is no physics warranting the belief that stars and planets dictate history through an as-yet-undiscovered “historion” particle, or that tachyons from the future inscribe secret codes onto reptile skeletons. And usually the claim is it’s god or magic or angels or something. So we’ll focus mainly on that.

But whether someone is saying a bodiless magical God or stellar historions permit preternatural predictions of the future, it’s an extraordinary claim either way (see William Lane Craig’s Duplicitous Denial That Extraordinary Claims Require Extraordinary Evidence). What would empirically verify it either way is if the evidence of what happened were such that it was more probable as the outcome of the preternatural thing than as just a normal, natural coincidence or scam. But that cannot derive from just comparing the probability of the outcome, because the preternatural things proposed are already extraordinarily less probable than the natural, normal things. So you have to overcome not only the probability of the evidence on naturalism, you also have to overcome the probability of naturalism. Because gods or historions are simply always the least likely explanation of anything.

If a light switch in your house suddenly, on its own, flips to “on,” always the least likely explanation is “God did it.” It’s always extraordinarily more likely it was physics (e.g. worn springs plus vibration or changes in temperature or moisture) or a prank (e.g. someone is goofing you). You therefore have to rule those out before you can rule God in, and that means you need evidence that makes those things less likely than God. Which is hard, because God is very unlikely—and even less likely to be pranking your light switch. So you need some pretty damned good evidence. This is actually a straightforward Cartesian Demon problem. That We Are Probably Not in a Simulation entails we are even less likely in thrall to a god. Because a sim is inherently less likely than a buddy pranking you, and gods are inherently less likely than a sim. And if it’s improbable we’re in a sim, it’s even more improbable gods are around fiddling with things instead. Unless you have damned good evidence. And no one ever does.

Think of what you would need to disprove your roommate was pranking you with the light switch, just to establish it was a worn-out switch spring—now realize what it would take to rule out that your well-in-evidence roommate pranked you but that a magical celestial entity pranked you instead. So it takes a lot of evidence to rule out obvious commonplace causes of remarkable things. And if you can’t rule them out, you can’t rule in even less common causes. This is as true when comparing worn switches with roommate pranks as when comparing roommate pranks with alien monsters. The reason Christianity Is a Conspiracy Theory is that there are so many way more probable but obviously false conspiracy theories that can explain all the same evidence, and you don’t have any evidence for them either.

Proving a Prophecy Miraculous

In the old Christian propaganda manual In Defense of Miracles, Robert Newman presented five criteria by which a prophecy could be proved to be a supernatural miracle, and his criteria are as applicable to psychics and astrologers and turtle-shell readers as to prophets. His criteria are quite good and I will repeat them here, quoting him verbatim:

1. It must actually be in evidence that “the [prophetic] text clearly envisions the sort of event alleged to be the fulfillment.” So, we have to rule out prophecies so ambiguous they can refer to anything and therefore will usually succeed even by accident.
2. It must actually be in evidence that “the prophecy was made well in advance of the event predicted.” So, we have to rule out making up a prophecy after the fact and then lying about when the prophecy was uttered.
3. It must actually be in evidence that “the event actually came true.” So, we have to rule out made-up fulfillments—which means not prove the fulfilling events didn’t happen, but fail to prove that they did (a crucial distinction).
4. It must actually be in evidence that “the event predicted could not have been staged by anyone but God.” So, we need evidence that is less probable on “someone around then faked it” than “an alien entity actually did it.”
5. It must actually be in evidence that “the event itself is so unusual that the apparent fulfillment cannot be plausibly explained as a good guess.” So, we have to show that “alien entities exist and do things like this” (or “astrology works, and in just this way” or whatever) is more probable than chance accident (cosmic monsters vs. worn switch springs: which happens more often?).

This is a structurally sound approach. Clarifications may be needed. For example, for Point (5), not only can’t the prophecy be an easy guess, it can’t be inevitable either. In other words, if a prophet was likely to say some particular thing for cultural or traditional reasons, then any “hits” (that is, outcomes that match what was said) cannot be distinguished from accidents. This is not a “good guess,” because this is not a guess per se, but an accidental correspondence. For example, if I was liable to say over and over again that a bird will land on my head, because it was a chant I inherited from my father, and then a bird landed on my head, it cannot be said that I predicted this. It cannot even be said that I guessed it. My reason for issuing this prediction had an entirely separate cause, and because it is always repeated, its success can have nothing to do with prophetic powers.

We see this phenomenon in Judaic prophetic tradition, when the doom of the Jews or their enemies—the “punishment” of the storm god Yawheh—is constantly proclaimed for cultural reasons, often as a threat aimed at encouraging piety and morality. When prophecies like these come true they cannot usually be counted as a prediction, because they would’ve been made anyway, and the predicted outcome is not all that improbable to begin with, given hundreds of years.

We also have to rule out selection bias. How do we know that the compilers of the Bible, for example, didn’t just pick the books or passages that got lucky and throw away the ones that didn’t? This would artificially bloat the apparent success rate of the Bible, by disguising the more numerous failures, hence making mere guesses seem miraculous. This relates to Point (2): to meet that criterion, we also have to be sure no one tampered with the documents later to make them successful after the fact. As for the Bible, how can we really be certain no one did that? Cold readers (fake psychics) rely on making a lot of predictions and you forgetting or ignoring the failed ones and then, in result, being impressed by the successful ones, when really, they just said a bunch of plausible things until you reacted positively. They didn’t actually read your mind. Likewise, prophets doing that aren’t really predicting the future. They’re just stumbling into success by throwing so much the spaghetti at the wall some has to stick by chance anyway.

Which is often best accomplished by being ambiguous. Hence Point (1): we must be especially wary of retrofitting. In other words, if a prediction is suitably vague, then any future event that can be made to fit the prediction can be claimed a success. This is another common trick well known to those who investigate psychics. In effect, if any event that will fulfill the prediction is likely to happen anyway, then it cannot be called a miraculous prediction—even if it was neither staged nor a guess.

For example, if I said that Zimbabwe would suffer under an evil ruler who would start a war with a great nation, that would never become a miraculous prediction, since that might happen anyway. Since I have not specified a time, even if Zimbabwe’s history meets with such an event in the next two or three thousand years you can claim I predicted it—but given thousands of years, who would be surprised? And even if the event were to happen tomorrow my prediction would not be miraculous, since I allowed that it might happen any time, meaning that my prediction was deliberately loose enough that I could claim any such event at any time a success; and there are many ways just such a thing could naturally happen in Zimbabwe at almost any time—because not just the time but the details are ambiguous. What counts as a ‘great nation’, for example? Or a ‘war’? Or even ‘start’? Or ‘evil’? Or ‘ruler’? Or, hell, even ‘Zimabwe’? A common trick is to stretch the sense of such words until almost literally anything would count, which essentially ensures success, because the probability of accidental matches phenomenally increase when we increase the number of circumstances that would match. So you can’t be flippant about Point (1).

Nor can you be flippant about Point (4). What if that prophecy actually caused a Zimbabwean autocrat to start the predicted war? Instead of predicting the future, the prophecy caused it. That’s not miraculous. It’s natural causation. In the end, you can’t state these five principles, and then weasel out of them by some device and claim to have proved a miracle. You actually just advertised that you can’t.

Ezekiel on Tyre

One of the most popular cases, frequently cited as one of the best examples of a miraculous prophecy in the Old Testament, is a pair of predictions made by the Prophet Ezekiel about the fate of the two principal cities of Phoenicia—Tyre and Sidon—who were among the most loathed neighbors of Israel.

On the one hand, in Ezekiel 28:22–23 the prophet issues the most vague of predictions about Sidon that could never be counted as miraculous. That a city should be sacked and its people slaughtered in antiquity is already a highly likely event, and that a prophet should declare such a doom upon an enemy city is likewise commonplace, so that the one is certain to follow the other by chance alone. Indeed, there is an already expected correlation between the notoriety of a city, such that it would even be thought of for such cursing, and the likelihood of it attracting enough attention that it would be a likely target for sacking. This is an example of an inevitable, and perfectly natural, correspondence between a prophecy and an event. It thus fails on Point (5), as well as Point (1), for its temporal and material lack of enough specificity to ever be miraculous.

On the other hand, Ezekiel 26:3–14 predicts that the city of Tyre will be attacked by many nations, its walls torn down and its rubble cleared away, and it will be a bare rock. Then “out in the sea she will become a place to spread fishnets” and will never be rebuilt. The passage specifically predicts that King Nebuchadnezzar of Babylon will do this, and his army will throw the stones, timber and rubble into the sea. Is this a miraculous prediction?

That’s all sufficiently specific, and perhaps predated the event. That’s harder to prove than most pretend. There has been no reliable chain of custody, nor any verification of when the text came to exist—such as would be needed to authenticate it in a court of law, for example. And tampering and falsifying are far more common hidden activities than psionics. But even if we had the required legal authentication, what it predicted is not all that unusual, nor very unforeseeable, nor can we really rule out Nebuchadnezzar staging it to fulfill that very prophecy. So, really, we can drop mic and move on. No miracle can be confirmed here. But the situation here is worse in precisely the most significant way: the prophecy as stated did not come true. Which proves Ezekiel did not have any divine voice telling him the future. He was making shit up and just getting lucky on occasion. Which makes this worse than merely not being able to prove Ezekiel a miracle worker: it proves he was not. And that’s all it takes to prove the Bible isn’t divinely inspired. Just one blatant failure. Because a real God wouldn’t allow that into his book.

In any such case, once we have established to a reasonable extent that a document is authentic (which we can barely do here) and properly understood (which we can reasonably do here), the historian’s next task is to examine it as a source: who was Ezekiel and where and when did he record this prophecy? Ezekiel was a Jewish guru held captive by Nebuchadnezzar since the sack of Jerusalem in 597 BC. Already we see some problems. It is all too likely that Ezekiel is issuing propaganda flattering his captor to get on his good side, while wishing ill on an old enemy of the Jews. Moreover, Ezekiel could easily have intel about the king’s plans. At the very least he would get the scuttlebutt and see the preparations. His prophecy about Tyre was issued in 586 BC. Nebuchadnezzar began the siege of Tyre only a year later. So, really, this fails Point (5). This is obviously credible as just a good guess. But as I said, it’s worse than that, because it didn’t happen.

Tyre resisted Nebuchadnezzar’s every attack for over a decade, and finally came to terms with him in 573. He did not sack the city after all—forcing Ezekiel to retract his prediction in verse 29:18–19, and instead predict a victory against Egypt, after Nebuchadnezzar had already turned against that country. So what do we have here? We have a man who sees the world’s most powerful army preparing to besiege a city (or, based on other chronologies, even already besieging it) and then predicts it will be taken and destroyed—hardly something he could not guess would happen. Yet even his guess failed, and so did the prediction! A failed prediction can hardly be a miracle. So this even fails Point (3). Worse, it disproves Ezekiel is a prophet. He is a false prophet. And therefore by God’s law should have been executed for fraud (Deuteronomy 18:20-22). So Ezekiel wasn’t working miracles. He was a felony conman. Ooops.

Apologetics Is a Grift

In practice, apologetics means “defending a claim against all the evidence refuting it.” Which more often than not requires lying; making apologetics, really, an art of lying. Lying by omission is the most common method; but most apologists inevitably resort to outright lies, deliberate distortions, or creating a false flight. Even honest apologists can fall into this trap eventually (see Comparing Apologetical Methods: Jonathan Sheffield vs. Michael Licona), because the function of apologetics is not to determine the truth, but to shield a belief from the truth. And there really aren’t any effective ways to do that honestly. Apologetics is therefore inherently corrupting of one’s integrity and intellect. Few who devote themselves to it survive its poison.

Hence, as we could have predicted, to try and save this failure, apologists like Newman will spread false claims about the history of this event that make the prophecy seem miraculous. But that continues to violate Point (3). For instance, Newman claims that Nebuchadnezzar “took” the city in 573 BC. But we have no evidence of that. The city submitted to Babylonian rule without being sacked, as Ezekiel’s retraction of the prophecy confirms. Indeed, since it was a trade powerhouse with two outstanding all-weather naval ports, a conqueror would have been a fool to destroy it. More likely Ezekiel just gullibly amplified Nebuchadnezzar’s propaganda, rather than Nebuchadnezzar’s actual plans for the city. Moreover, Tyre has a long history of working out just such favorable agreements to get besiegers off its back. This was the key to its survival. So even a half-competent insurance speculator would have outperformed Ezekiel as a “prophet” here.

Newman’s answer is to offer an absurd ad hoc thesis: he makes up the story, told in no ancient source, that the inhabitants of Tyre resettled inland during the siege, forcing Nebuchadnezzar to “settle for very little plunder.” Not only is this false—for there had been a mainland suburb since before 800 BC, and it was not called Tyre but Ushu—it is extremely implausible: people are supposed to defeat Nebuchadnezzar by leaving a nearly-invincible island city, nestled behind a monstrous wall surrounded by the literal ocean as a mote, to resettle, with no fortifications, on the mainland, in open ground and with no port, while the Babylonian army apparently sits by and twiddles their thumbs?

The absurdity of this is palpable, never mind that it has no basis in any evidence. Perhaps sensing that such a lie might not work on his less gullible readers, Newman tried to rescue the claim in another ad hoc way as well: that Alexander the Great, centuries later, fulfills the prophecy. Because he famously used the rubble of the mainland city of Ushu to connect the island of Tyre to the mainland. But Ezekiel leaves no doubt that he means Nebuchadnezzar’s men will do this, not some other guy centuries later. The prophecy even incorporates charioteers in the victorious forces, yet chariots were no longer used in warfare by the time of Alexander.

So this is a classic case of retrofitting: changing what a prophecy said in order to make some other thing fulfill it. But then with centuries of history to play with, almost anything can be found to fit it, violating Point (1) again. It is worse than that, of course, since the prophecy as stated actually forbids attributing this event to anyone else but Nebuchadnezzar. We might as well call the Israeli rocket attack on the ruins of Tyre in the Arab-Israeli war of 1981 a “fulfillment.” And even worse, the eventual success of Alexander still did not fulfill any other details of Ezekiel’s prophecy—yet if we allow any detail to replace all the others, then the probability that someone, someday, would “succeed” in fulfilling the prophecy is, again, no longer miraculous.

Points (1) and (5) require not playing these kinds of games. So the fact that Newton had to play them is an admission that Ezekiel’s prophecy failed, and was therefore not only not miraculous, but worse, a high crime, for which Ezekiel should have strangled to death. And the Bible can’t be holy if it was written by criminals. Newton is effectively a sleazy defense lawyer trying to make all this go away with shifty rhetoric and browbeating howlers. Which ironically is how we know Christianity is false.

The failure of Alexander to actually fulfill the prophecy—not only temporally and aptronymically, but materially—drives Newman to tell or repeat more lies, insisting the mainland site was “scraped clean” by Alexander and has “never been restored” while “parts of the former island are used even today for spreading fishnets.” But that isn’t true. First, Ushu is also still inhabited and was never a bare rock (see the following satellite image: thanks to Alexander, Tyre is now a peninsula; the rest is Ushu).

Second, Ushu is not Tyre. Alexander’s successors built up Tyre as the powerful naval and merchant port it had always been, and it remained an influential city for over a thousand years. It has never been a bare rock. It still stands even to this day. The modern city of Tyre sits beside and atop the ancient ruins (many of which still stand), and has a population of over 70,000, twice what it was in Alexander’s day. It is now a major Lebanese financial center. No bare rock. It was not “forever uninhabited,” much less uninhabited for even a single day. It was perpetually rebuilt. And parts of the original city still stand.

So what about the fishnets? Authors like Newman are fond of citing a 19th (!) century tourist who saw fishnets stretched over the rocks of Tyre as proof of the fulfilled prophecy, ignoring the fact that fishnets have always been stretched over rocks in every city with a fishing industry since the invention of the net, and they were no doubt stretched across the rocks of Tyre long before Ezekiel was even born. Ezekiel can’t have miraculously predicted that. That would fail Points (1) and (5). But that is not what he claimed to have predicted anyway. Ezekiel said God said (and thus was lying): “I will scrape away her rubble and make her a bare rock [and then] out in the sea she will become a place to spread fishnets,” meaning, the entire island will only be a bare rock across which only fishnets will spread. Which didn’t happen. Failing Point (3).

Apologists still try to argue, of course, that Ezekiel “really” meant that Ushu would be leveled, not Tyre. But Ezekiel specifically says the nets and scraping will happen not on the mainland but “in the midst of the sea,” so he clearly did not mean Ushu, but Tyre. In fact, in the very next verse Ezekiel says “and her mainland” will have its population slaughtered. He thus clearly cannot have meant the previous verse to refer to that, but specifically only the island. Ezekiel always referred to the mainland site as among the “daughters” of Tyre anyway, never as Tyre herself. And that is to be expected. Ushu was neither rich nor powerful, since it had no ports—unlike Tyre, which had two ports situated to allow year-round sailing, making it one of the most powerful military and trade cities in the world. It would be silly to make elaborate claims about the fall of “mighty” Ushu. So resorting to this tactic would make the prophecy fail on Point (1).

I offer this as a paradigmatic case. When we examine other claims of miraculous prophecy in the Old Testament in the light of history, we quickly find that they don’t hold up. In fact, Newman never tells his readers that Ezekiel actually went on in verse 26:19 to predict that Tyre would be covered by the sea, and in 26:21 he says it would never be found again, two clearly failed predictions. There is definitely no miracle here. And yet this is supposed to be the best case in the Bible. I can say it is certainly a typical case. This is therefore just one more reason why you should give up believing in miracles.

Conclusion

There are many more examples like this in Tim Callahan’s Bible Prophecy, the Skeptic’s Annotated Bible, and the Secular Web’s Library on Prophecy. To test the miraculousness of any prophecy you must first establish, “Did they even say that?” Because the most common tactic of dishonest or delusional apologists is to change the plain meaning of the text into something else, failing on Point (1); and next to rely on a forgery that was written after the fact (like Daniel 9 falsely claiming to have “predicted” the death of Onias III centuries before, when it was actually written afterward), failing on Point (2); or asserting a fulfillment that insufficient evidence establishes as having happened, or that sufficient evidence even proves didn’t happen (as we just saw with Ezekiel), failing on Point (3); or trying to pass off an intentionally rigged event as a miraculous fulfillment (like Jesus getting himself killed specifically to fulfill Daniel 9), failing on Point (4); or claiming easily guessed or commonly declared outcomes are “miraculous,” failing on Point (5).

Did they even say that? Did they even mean that? Would they have said it anyway? Would it get naturally fulfilled anyway? Did it even happen? Or did it happen before they said it? Is it more likely the kind of outcome people make up? Is that outcome more common by chance than are alien monstrosities? And are we ignoring all the misses to only count the accidental hits? The fact that every example ever proffered fails at one or more of these checks is itself proof that Naturalism Is Not an Axiom of the Sciences but a Conclusion of Them. Had the world been different—had there actually been an oracular God—we should by now have many proven examples, not just a bunch of lies and fakes.

As I mentioned before with Islamic prophecy, the Bible could have done miraculous things, like accurately predict what we would find to be the speed of light, or the actual age of the Earth, or meticulously detailed the rise of Hitler, pegging it to era, technology, and concentration camps, or predicted the American Revolution, right down to an “unlikely combination of light winds and a dense fog that” would help “save the American army.” It could have accurately (not ambiguously) predicted the victory of Julius Caesar in Italy, establishing the Roman Empire, centuries in advance. It could have predicted microbes, radios, electronic computers, the printing press, or future supernovas—in fact, it could have predicted every visible supernova each fifty years for thousands of years, continually proving its miraculous accuracy, which could have no other explanation than something like God being involved. But of course bibles, as soon as made, could have been miraculously indestructible and unalterable, thereby proving God’s endorsement of their contents. And the Bible could have denounced slavery and other villainies, and commanded equal rights and democracy.

There are tons of ways the miraculous truth of the Bible could have been confirmed in evidence. It just wasn’t. And the problem is there are too many people who won’t admit that. And they tend to be ruining the world.