MANNAFEST

The Evidence Room · The Long Range

Did the prophets really predict the future?

Some prophecies are far weaker than apologetics websites admit — this page leads with the worst one and prints the verse that embarrasses it — and a small number survive every honest objection, which is why they need no inflated numbers.

This is the argument most likely to have been oversold to you. If you grew up in church you have probably heard that the Bible contains thousands of fulfilled prophecies and that the odds against it are astronomical. If you left church, that claim is probably one of the reasons.

So this page is built backwards from the usual one. It leads with the hardest case rather than the easiest, prints the verse that embarrasses the standard argument rather than omitting it, and retires by name two claims that have been repeated for a century and are not true.

MannaFest believes the prophets spoke for God and that some of what they said could not have been known by ordinary means. This page is written on the conviction that a case which needs padding is a case that has not been made — and that the padding is what people notice.

Three ways this argument routinely goes wrong come first, because you will need them to read everything after.

Walking it through

  1. Three ways this argument goes wrong

    The first is circularity. Never offer a fulfillment whose dating is the very thing under dispute. Daniel is the standing example: if you argue that Daniel predicted Antiochus IV in detail, and your reader's whole position is that Daniel was written after Antiochus IV, you have not given him a reason — you have restated your conclusion in a louder voice. A prophecy is only evidence if the text can be shown to predate the event on grounds the other person already accepts.

    The second is the tally. You will not find a number on this page — no two thousand prophecies, no three hundred messianic predictions. Those counts are produced by counting loosely, and a hundred weak items do not add up to one strong one. A small set treated deeply, with both columns printed, is worth more than a list nobody checks.

    The third is type confusion, and it is the one honest readers of the New Testament have to face. The New Testament's own citation practice includes at least three different things, and it uses one word for all of them. There is direct prediction. There is typology, where an earlier event is a pattern that a later one fills out. And there is application, where a text is brought to bear on a situation it was not about.

    Matthew's use of Hosea is the clearest case. Hosea 11:1 is not a prediction at all — read the whole verse and it is plainly about the exodus, God calling Israel out of Egypt, in the past tense. When Matthew says Jesus went to Egypt so that this might be fulfilled, he is not claiming Hosea foretold it. He is claiming Jesus is walking Israel's story again and doing it rightly. That is typology, it is a recognized and ancient way of reading, and saying so out loud costs nothing. What costs something is pretending Hosea was a prediction, because any reader who opens Hosea will catch it in ten seconds and will assume, reasonably, that he has been handled.

  2. The hardest case first: Tyre — DEBATED

    The Tyre oracle in Ezekiel 26 is the most-cited fulfilled prophecy in popular apologetics, and it is the one that has done the most damage to the credibility of the argument. It leads this page for that reason.

    Here is what the oracle says. Many nations will come up against Tyre, wave after wave. Nebuchadnezzar, king of Babylon, is named and brought against her. Her walls will be broken and her towers thrown down. Her stones, timber, and dust will be laid in the midst of the waters. She will be scraped bare — a bare rock, a place for the spreading of nets.

    Here is what happened. Nebuchadnezzar besieged Tyre for thirteen years, roughly 586 to 573 BC. Tyre at that time was two settlements: a mainland town and the real city on an island offshore. He did not take the island. The siege ended in some negotiated arrangement, and Tyre continued.

    Two and a half centuries later, in 332 BC, Alexander the Great took the island — and he did it by demolishing the mainland ruins and pushing the rubble into the sea to build a causeway out to the walls. That causeway silted up over the centuries into a permanent isthmus. Tyre is not an island anymore. You can walk out to it.

    And then Tyre was rebuilt. It was a working city in the New Testament: Jesus travels to its borders in Mark's Gospel, and Acts records a church there whose disciples walked Paul down to the beach to pray with him. Tyre is a Lebanese city today, with tens of thousands of residents and Roman ruins that tourists visit.

    So the maximal reading of the oracle — permanent desolation, never rebuilt, bare rock forever — is simply false, and every reader with a search engine can establish that in a minute.

  3. The Bible's own candor about Tyre

    Here is the part that almost never appears on Christian websites, and it is in the Bible.

    Sixteen years after the Tyre oracle, Ezekiel returns to the subject and says outright that Nebuchadnezzar got nothing out of it. His army worked until their heads were rubbed bald and their shoulders were worn raw — and had no wages. God then gives him Egypt instead, as payment for the service.

    Read that again slowly. The prophet who delivered the Tyre oracle is on record saying the campaign did not pay out. Scripture reports the shortfall itself. It did not need a critic to discover it.

    That is a remarkable thing for a book to do, and it is the reason this exhibit is on the page rather than off it. A text engaged in propaganda does not file this verse. Whatever else Ezekiel is, he is not curating his own record.

  4. Tyre: both readings at full strength

    There are two serious ways to read this, and neither is stupid. Both are printed here without a verdict.

    THE TELESCOPING READING. The oracle itself announces many nations, plural, coming like successive waves — that is verse 3, before Nebuchadnezzar is ever mentioned. And there is a grammatical hinge that defenders point to and that is genuinely there in the Hebrew: in verses 7 through 11 the subject is he, singular, and it is Nebuchadnezzar; at verse 12 it shifts to they. On this reading Nebuchadnezzar is the first wave and the oracle was never confined to him. Alexander, casting the mainland's stones and timber and dust into the sea to reach the island, is a later wave doing precisely what verse 12 describes — and doing it in a way no one in Ezekiel's century could have designed as a forgery, because nobody builds a prophecy around a causeway. Add to this that ancient judgment oracles used sweeping, total language as a recognized convention. When an Assyrian king says he left no survivor, historians read the genre and do not accuse him of a false prediction; the same courtesy is owed to a Hebrew prophet.

    THE CRITICAL READING. Kris Udd, writing in Tyndale Bulletin 56.1 (2005) — an evangelical journal, open access, and the fairest single treatment of this text — argues that the oracle's terms as delivered were not fulfilled by Nebuchadnezzar, and that Ezekiel 29:17-20 is exactly what it looks like: an inspired adjustment. God says the campaign did not pay and assigns a different payment. On this reading the telescoping move reads the plural pronouns harder than the text will bear, and the appeal to hyperbole is doing convenient work — if total language can mean partial outcomes whenever the outcome was partial, the prophecy cannot fail, and a claim that cannot fail is not a claim.

    Both readings are held by people who believe Ezekiel spoke for God. This page does not adjudicate between them, and the exhibit says DEBATED on its face.

    Why does this lead the page? Because the airbrushed version — Tyre destroyed forever, never rebuilt, scraped bare exactly as predicted — has been the standard apologetic move for a century, and every informed reader already knows it is wrong. Publishing the candid version, including the verse that embarrasses the claim, is the single act that buys this entire wing its credibility. A reader who sees Ezekiel 29:18 printed here by us will believe us later, when we say something holds.

  5. Babylon, and two claims we are retiring by name

    Babylon is the second exhibit and it needs the same treatment.

    What actually happened: Cyrus took Babylon in 539 BC essentially intact and without a general battle — the Cyrus Cylinder and the Babylonian chronicles agree that the city came over, and Cyrus presented himself as a restorer of its temples rather than its destroyer. There was no fiery overthrow. Babylon then declined slowly over centuries, losing status under the Seleucids as Seleucia drew off its population, and was effectively abandoned by roughly the second century AD. And in the 1980s Saddam Hussein rebuilt portions of it, laying bricks stamped with his own name in imitation of Nebuchadnezzar.

    So the honest form of the Babylon argument is about a long decline that did in fact end in desolation — not about a sudden overthrow, and not about a site that was never touched again.

    TWO CLAIMS THIS SITE IS RETIRING, and we are naming them so nobody has to wonder whether we simply hadn't heard them.

    First: the story that Bedouin will not pitch a tent at Babylon out of superstitious dread, offered as fulfillment of Isaiah's line about the Arabian not pitching tent there. This is a nineteenth-century travel anecdote that entered the apologetic bloodstream through Halley's Bible Handbook and has been repeated ever since. It is a traveler's impression, not a datum, and it is not a fact about the present. We will not use it.

    Second: the claim that Babylon's stones were never reused. They demonstrably were, extensively, for centuries — that is a normal fate for a brick city near settlements that needed bricks. The claim is false and repeating it discredits everything printed beside it.

  6. Cyrus named — with the dating question stated plainly

    Isaiah names Cyrus, calls him God's shepherd and his anointed, and says he will order Jerusalem rebuilt and the temple's foundation laid. Cyrus of Persia did exactly that in 538 BC.

    If the passage was written in the eighth century BC as tradition holds, that is a named individual roughly a century and a half before his birth, and it is difficult to explain naturally. That is the strong form of the argument and it deserves to be stated.

    Now the part that has to travel with it. Since the late eighteenth century the majority critical position has held that Isaiah 40-66 comes from a different hand, writing during or near the exile — the so-called Deutero-Isaiah — which would place these chapters in Cyrus's own era or close to it. The arguments are internal: the audience of those chapters is addressed as already in exile, the vocabulary and imagery shift, and the theological emphases change. Conservative scholars answer that the unity of the book is attested in every ancient witness including the Great Isaiah Scroll from Qumran, which contains the whole book as one work with no seam, and that the critical case ultimately assumes what it sets out to prove — that predictive prophecy of this kind does not happen.

    This page does not settle that. What it will not do is build a probability claim on top of an unsettled dating. If you present the naming of Cyrus as astronomically improbable, every informed reader hears you assuming the dating that is precisely in dispute, and the whole page loses him. The correct form is: on the traditional dating, this is striking; on the critical dating, it is not a prediction; the dating is contested; here is where you can read both.

  7. Daniel: two columns, and no verdict from us

    Daniel is the great battlefield — four kingdoms in a statue and in beasts, and the seventy weeks — and it is the standing example of the circularity trap. So it is presented here as two columns, neither adjudicated.

    THE MACCABEAN CASE, stated at its strongest. Daniel 11 tracks the wars of the Ptolemies and Seleucids with a precision that is unlike anything else in prophetic literature — marriages, betrayals, campaigns, a specific desecration of the temple — and it tracks them accurately right through 167 to 164 BC. Then, at 11:40-45, it describes the king's end in terms that do not match how Antiochus IV actually died. That combination is the argument: detailed accuracy up to a point and divergence immediately after is the classic signature of a text written at that point, looking backward with confidence and forward with guesswork. Add the Greek loanwords among the musical instruments in chapter 3, and the placement of Daniel among the Writings in the Hebrew canon rather than among the Prophets, which may suggest it arrived after the prophetic collection had closed.

    THE EARLY-DATE CASE, stated at its strongest. Kenneth Kitchen's study of Daniel's Aramaic concluded it belongs to the Imperial Aramaic of the Persian administrative period rather than to the later Western Aramaic one would expect from a second-century Palestinian author. Eight copies of Daniel were found at Qumran, the earliest dated on paleographic grounds to around 125 BC, which sits awkwardly close to a supposed composition around 164 — a book written within living memory would have had little time to be copied, distributed, and received as Scripture by a sectarian community. And there is Belshazzar. Daniel names him as the ruler in Babylon at its fall; Greek historians did not know he existed, and for centuries this was cited as proof that Daniel was ignorant fiction. Cuneiform records recovered in the nineteenth century established that Belshazzar was Nabonidus's son and functioning co-regent — which is exactly why Daniel 5 has the king offering Daniel the third place in the kingdom rather than the second. That detail is not the kind of thing a second-century author could have reconstructed; the memory it preserves had been lost by the time the Greek historians wrote.

    AND A REAL DIFFICULTY FOR THE EARLY DATE, listed because leaving it out would be exactly the dishonesty this page is against: Darius the Mede. Daniel describes him receiving the kingdom at Babylon's fall, and there is no external identification of him at all. No inscription, no chronicle, no king list. Proposals exist — that he is a throne name for Cyrus, or that he is Gubaru the governor — and none of them has documentary support. It is an unresolved problem, and it sits on the same page as Belshazzar, which is a resolved one.

    This site holds a traditional early date for Daniel. It states that as its position, not as a demonstrated result, and it prints the case against.

  8. The probability question, handled once and then dropped

    Sooner or later somebody quotes you a number: the odds of one man fulfilling eight messianic prophecies are one in ten to the seventeenth power. It comes from Peter Stoner, a mathematics professor, in Science Speaks (Moody Press, 1958; revised 1963). Here is what the calculation actually is, stated the way Stoner himself stated it — which is more honest than the way it is usually passed along.

    Stoner assigned each prophecy an estimated probability, multiplied them together, and then divided by the number of people who have lived, arriving at one in ten to the seventeenth. The multiplication is arithmetic and nobody disputes it. The question is where the individual probabilities came from, and Stoner tells you: they were, in his words, “a combination of the estimates given by this class... and estimates given me later by some twelve different classes of college students, representing more than 600 students,” which he then weighed and deliberately made more conservative than his students had.

    So the inputs were averaged classroom intuitions, refined toward caution by the author. Note that Stoner disclosed this in the book. It is not a scandal, and he is not a fraud — he labeled his own method plainly. It becomes a scandal only when a website prints 10^17 as though a mathematician had measured something, with the sourcing stripped off.

    The same goes for the endorsement that usually travels with the number. The American Scientific Affiliation reviewed the book and its statement covers the probability material as being, in general, dependable and accurate. The statement also says explicitly that it does not extend to “issues of Biblical exegesis or historical accuracy.” Apologetics sites quote the first half and drop the second, which converts a narrow endorsement of the arithmetic into a scientific seal of approval on the interpretation of the prophecies — the very thing it was written to exclude.

    That is the entire treatment of probability on this page. No number is used in the argument that follows, because the argument does not need one.

The objections, at full strength

Each objection below is stated the way the people who hold it state it. If you hold one of these and think it is put unfairly here, that is a failure on our part — tell us.

It was written after the fact

The objection

Vaticinium ex eventu — prophecy after the event — is not an ad hoc excuse invented for the Bible. It is a well-documented literary practice across the ancient Near East and the Greco-Roman world, found in Akkadian prophecy texts, in Egyptian material, and throughout Jewish and Christian apocalyptic. Ancient authors wrote in the name of a revered figure from the past and had him foretell events the author had already lived through. It is the standard scholarly explanation for detailed 'prediction' precisely because it explains the data so cleanly, including the signature pattern: sharp accuracy up to the author's own moment, then vagueness or error immediately after. Daniel 11 displays that pattern exactly. And the explanation requires nothing miraculous and nothing dishonest by ancient standards — pseudonymous composition was a recognized genre, not a forgery.

The response

This is the strongest objection to the whole category and it is why this page refuses to argue in a circle. Three honest responses, none of them a knockout. First, the explanation is powerful precisely because it is unfalsifiable in the general case: any prophecy whatever can be redated after its fulfillment, so the framework can never lose. That is a reason to demand independent dating evidence rather than to dismiss the objection. Second, independent dating evidence is exactly what the messianic material has and what Daniel is fighting about. The Great Isaiah Scroll and the pre-Christian Greek translation of the Hebrew Scriptures put the text of the servant passages, the birthplace oracle, and the seventy weeks in circulation demonstrably before the events they are applied to — that is not a Christian claim about a Christian document, it is a manuscript fact about Jewish documents preserved by non-Christian hands. Third, where the dating genuinely is contested, as with Daniel and Deutero-Isaiah, this page does not use the passage as evidence. That is a real cost and it is paid above.

A man who knows the prophecies can fulfill them on purpose

The objection

This objection needs no hostility toward Jesus and no conspiracy. A first-century Jew who believed himself to be the Messiah would have known the messianic texts as well as anyone alive, and some of them are enactable at will. Ride into Jerusalem on a donkey. Cleanse the temple. Speak in parables. Go to Jerusalem at Passover. Say the words of Psalm 22 from the cross. The Gospels do not merely permit this reading — they assert it. They say Jesus sent two disciples ahead specifically to fetch the colt, and Matthew and John attach the Zechariah oracle to the act explicitly. The text itself describes a deliberate enactment. Everything downstream of that is a man doing what the script called for.

The response

The concession here should be total on the enactable items, and the Gospels make it for us. Zechariah 9:9 was deliberately enacted; the text says so and there is nothing to defend. A prophecy a claimant can arrange proves nothing about foreknowledge — at most it shows intention, and intention is what the Gospels are reporting. So this page does not count enactable prophecies as evidence at all. What it counts is the residue: the things outside a claimant's control. He did not choose his birthplace or the census that put his family there. He did not choose the manner of his execution — crucifixion was a Roman decision made by other men, and the Psalm 22 description of pierced hands and feet and divided garments predates Roman crucifixion by centuries. He did not choose to be born within a window that put his death before the destruction of the Second Temple, which the seventy weeks and Malachi's expectation of the Lord coming to his temple both require. Deliberate enactment is a real filter, and this page runs the list through it and keeps what passes.

The prophecies are vague enough to fit anything

The objection

This is how horoscopes work and everyone knows it. Write loosely enough — a ruler will fall, a city will suffer, a deliverer will come — and history will oblige you within a century. Then the fulfillment is chosen after the fact by people motivated to find one, from a menu of candidate events large enough to guarantee a match. Worse, retrofitting is invisible from the inside: nobody experiences themselves as stretching a text. Add the fact that later writers had the prophecies in hand while composing narratives about the events, and you have a mechanism for producing correspondence without any foreknowledge at all — the story shaped to fit the text, not the text foreseeing the story.

The response

The objection lands, and this page has already conceded most of what it demands: it has retired the Bedouin anecdote, refused the tally, printed the Tyre problem, and declined to build on contested datings. Vagueness is a real defect in a real portion of the material. But the objection is a filter, not a solvent — it removes the loose items and leaves the specific ones, and specificity is checkable rather than a matter of taste. Bethlehem is not 'a city.' Being sold for a named sum by a friend, and the money going to a named use, is not 'betrayal happens.' A date-window tied to a decree about rebuilding Jerusalem is not 'a deliverer will come.' The right response to this objection is not to argue with it; it is to hand the reader the specific ones and let him apply the objection himself, having first been shown the ones that fail it.

Nobody counts the misses

The objection

This is the selection problem, and it is fatal to any argument built on a tally. Israel and Judah produced an enormous volume of oracular material over centuries, of which we possess a curated fraction — curated, note, by people who already believed. If a thousand oracles were spoken and thirty came true, thirty hits is what chance predicts, and the misses were not preserved because nobody preserves a prophet who was wrong. The Bible itself concedes the phenomenon: Deuteronomy gives a test for false prophets, which presupposes there were plenty; Jeremiah spends whole chapters fighting prophets who spoke in the Lord's name and were not sent. Jonah announced Nineveh's overthrow in forty days and it did not happen. Any honest evaluation has to reckon with a denominator that no longer exists.

The response

The selection effect is real and it is the reason this page carries no total. You cannot compute a hit rate against a lost denominator, and any site that quotes one is quoting a number it cannot possibly have. Two things survive. First, the objection is aimed at the tally argument, and the tally argument is not being made here — a single unambiguous instance of knowledge that could not have been had is not weakened by there having been many prophets, in the way a batting average would be. Second, the biblical writers preserved the failures rather than deleting them. Ezekiel's own admission about Nebuchadnezzar's wages is printed above; Jonah's unfulfilled announcement is in the book that bears his name and is the whole point of its ending; Deuteronomy's test is there for anyone to apply to the collection it introduces. That is not a body of literature quietly grading itself. It is one that left the failing papers in the file — which is the only reason a page like this one can be written at all.

Go further

Every prophecy, with its status→

The prophetic timeline→

The seventy weeks→

The statue and the stone→

Four schools of Revelation→

Has the Bible been changed?→

Does life look designed?→

The apologetics room→

Every source we cite→

Sources: Kris Udd, “Prediction and Foreknowledge in Ezekiel's Prophecy Against Tyre,” Tyndale Bulletin 56.1 (2005) — open access; the critical reading of the Tyre oracle, printed here at full strength. · Peter W. Stoner, Science Speaks (Moody Press, 1958; rev. 1963) — cite-only; the 10^17 calculation and Stoner's own account of where his input probabilities came from, quoted from his text. · The American Scientific Affiliation's statement on Science Speaks — quoted at its actual scope, including the exclusion of “issues of Biblical exegesis or historical accuracy.” · K.A. Kitchen, “The Aramaic of Daniel,” in Notes on Some Problems in the Book of Daniel (Tyndale Press, 1965) — cite-only; the Imperial Aramaic argument for an early date. · The Cyrus Cylinder (British Museum, ANE 90920) — primary artifact; Cyrus's own account of taking Babylon peacefully and restoring its temples. · The Nabonidus Chronicle and the Verse Account of Nabonidus — cuneiform primary sources; the co-regency of Belshazzar. · The Great Isaiah Scroll (1QIsaᵃ, Qumran Cave 1) — primary manuscript; Isaiah as a single undivided book, pre-Christian. · The Septuagint, the pre-Christian Greek translation of the Hebrew Scriptures — primary; the messianic texts in circulation, in a second language, before the events. · Halley's Bible Handbook — named here only as the popular source of the Bedouin-tent anecdote this site is retiring. · Chuck Missler, Learn the Bible in 24 Hours (Thomas Nelson, 2002), and Daniel's 70 Weeks (Koinonia House) — cite-only; the seventy-weeks reckoning in the tradition this site stands in. Every cited work — the sources registry →