Daniel's Prophecies: Prediction or Postdiction?

The book of Daniel is the ace up the sleeve of every prophecy apologist. The statue in Nebuchadnezzar's dream that maps onto four successive empires. The seventy weeks that count down to the messiah. The detailed descriptions of wars between kings of the north and south that line up with Hellenistic history with uncanny precision. If Daniel really was a Jewish exile writing in Babylon in the 6th century BCE, the predictive accuracy is staggering — perhaps the single strongest evidence for supernatural revelation in the entire Bible.

But Daniel was not written in the 6th century BCE. The scholarly consensus — held by the overwhelming majority of critical scholars for well over a century — is that the book of Daniel reached its final form in the 160s BCE, during the persecution of the Jews by the Seleucid king Antiochus IV Epiphanes. The visions of chapters 7–12 were written then; older court tales about Daniel, already collected, became chapters 1–6. The "predictions" are not predictions at all. They are postdictions — history written after the fact and dressed up as prophecy. And the evidence for this is so strong, so multifaceted, and so convergent that rejecting it requires you to ignore linguistics, historiography and canon history at once.

Let me walk through the evidence, and the strongest replies to it.

The Tell: Where the Predictions Stop Being Accurate

This is the most important piece of evidence, and it's the one I always lead with because it's the most intuitive. If you wanted to determine whether someone was genuinely predicting the future or pretending to predict the future by writing about events that had already happened, what would you look for? You'd look for the moment when the accuracy breaks down. A genuine prophet should be equally accurate about all future events. A faker will be devastatingly precise about everything up to his own time and then suddenly get vague or wrong about what comes next.

Daniel 11 is the test case. It contains a detailed account of wars between the "king of the north" (the Seleucid dynasty) and the "king of the south" (the Ptolemaic dynasty). John J. Collins, in his Hermeneia commentary on Daniel — the standard critical commentary in English — walks through these verses and demonstrates that they correspond to known Hellenistic history with remarkable specificity. The marriages, the military campaigns, the betrayals, the shifting alliances — they track the actual historical record from the late 4th century through the early 2nd century BCE with the kind of detail that would make a historian jealous.

And then we hit Daniel 11:40–45. The text describes the final campaign of the contemptible king — identified by critical scholars as Antiochus IV Epiphanes — predicting that he will sweep through Egypt, conquer Libya and Ethiopia, and then pitch "the tents of his palace between the sea and the glorious holy mountain," where "he will come to his end" (Daniel 11:45, World English Bible). But that's not what happened. Antiochus IV died in 164 BCE in Persia, during a campaign in the eastern part of his empire, not between the Mediterranean and Jerusalem. He died of illness, not in a climactic battle.

Here's the thing. The predictions are stunningly accurate up to about 167 BCE — the desecration of the Jerusalem temple by Antiochus, the "abomination of desolation," the suspension of the daily sacrifice. These are confirmed historical events. But from that point forward, the text gets it wrong. Antiochus doesn't die where Daniel says he'll die. The end of the world doesn't follow immediately, as Daniel 12:1–3 predicts. The resurrection of the dead doesn't happen.

The traditional answer to this is older than modern criticism, and it deserves to be stated in its own terms. The third-century philosopher Porphyry made exactly the argument I have just made. Jerome, in the prologue to his Commentary on Daniel, reports him as holding that "whatever he spoke of up till the time of Antiochus contained authentic history, whereas anything he may have conjectured beyond that point was false, inasmuch as he would not have foreknown the future." Jerome, answering him, conceded the history — "Up to this point the historical order has been followed, and there has been no point of controversy between Porphyry and those of our side" — and disputed the reference. From 11:21 on, he wrote, "those of our persuasion believe all these things are spoken prophetically of the Antichrist who is to arise in the end time," with Antiochus "to be regarded as a type of the Antichrist." Modern conservative commentators usually move the hinge later, to 11:36. Thomas Constable, whose freely available Notes on Daniel are a widely used compendium of the conservative commentary tradition, puts it plainly: "There is a similar break between verses 35 and 36 of chapter 11. What was predicted in verses 2 through 35 has happened. What follows in verse 36 through 12:4 has not yet happened." He notes that older conservatives such as Keil and E. J. Young drew the line at the same verse. On this reading, the verses that fail for Antiochus were never about him. They await a future king.

My difficulty with that answer is that the text gives no signal of a change of subject. Daniel 11 marks every new ruler with a new introduction ("In his place a contemptible person will stand up," 11:21), and the king who "will do according to his will" in 11:36 gets none. He is still the "he" of 11:21–35, and 12:1 dates the deliverance and the resurrection to "that time" when he falls. Jerome states the objection himself: the critics ask, he writes, "why the prophetic discourse should abruptly cease mention of these great kings and shift from Seleucus to the end of the world." His answer was that the chapter skips elsewhere: only four Persian kings follow Cyrus in 11:2, and "many who came in between were simply skipped over." But 11:2 compresses two centuries into a verse, while 11:21-35 is the most detailed stretch of the chapter, detail Jerome himself grants is "appropriate to the person of Antiochus." Following one king that closely and then leaving him, without a word, for a king centuries away is not skipping in the way 11:2 skips. A change of referent placed exactly where the history stops matching is not evidence against the postdiction hypothesis. It is what the postdiction hypothesis predicts.

The whole pattern is exactly — exactly — what you would expect if the author was writing during the Maccabean crisis of 167–164 BCE. He knew the history up to his own time because he had lived through it. He then projected forward, predicting Antiochus's imminent defeat and the eschaton, and got it wrong because he was guessing about the future, not reporting divine revelation. That is the standard critical argument, and Collins's commentary works it through verse by verse, reading the whole of 11:21–45 as Antiochus, down to the forecast of his death that did not come true.

Carol Newsom, in her Old Testament Library commentary on Daniel (written with Brennan Breed), makes the same point. The detailed correspondence between Daniel 11 and Hellenistic history is not a sign of supernatural foresight. It is a sign of contemporary knowledge. The author knew these events because they had already occurred.

Daniel Is Not in the Prophets

Here's a fact that surprises most Christians: in the Jewish canon, Daniel is not classified as a prophetic book. The Hebrew Bible is divided into three sections — Torah (Law), Nevi'im (Prophets), and Ketuvim (Writings). Daniel is placed in the Ketuvim, alongside Psalms, Proverbs, and Esther. Not alongside Isaiah, Jeremiah, and Ezekiel.

Why does this matter? Because, on the usual critical reconstruction, the collection of the Nevi'im was closed before Daniel was written. If Daniel had existed in the 6th century BCE — contemporaneous with Ezekiel and the later portions of Jeremiah — there would be no obvious reason to exclude it from the prophetic collection. But if Daniel was composed in the 160s BCE, after the prophetic canon was already fixed, then its placement in the Ketuvim makes perfect sense. It was a late arrival. That is not to say that Jews of the Second Temple period refused to read Daniel as a prophet. Plainly they did. The Qumran Florilegium (4Q174), a pre-Christian text, introduces a line of Daniel 12 as "written in the book of Daniel the prophet," and Josephus, writing late in the first century CE, says Daniel received revelations "as to one of the greatest of the prophets." What the Ketuvim placement records is not how readers regarded Daniel, but where the collection finally shelved him.

Christian Bibles rearranged the order of the Old Testament books, placing Daniel among the prophets, between Ezekiel and Hosea. That ordering is Christian, but reading Daniel as a prophet is not a Christian invention. Matthew's Jesus speaks of the abomination "spoken of through Daniel the prophet" (Matthew 24:15, World English Bible), and he inherited the habit from the Jewish readers just quoted.

Defenders of the traditional date read the Ketuvim placement differently, and their explanation deserves a hearing. Constable gives the standard one: "The Jews placed this book in the Writings section because Daniel was not a prophet in the sense in which the other Hebrew prophets were. He functioned as a prophet, and he wrote inspired Scripture, but he was a government official, an administrator in a Gentile land, rather than a preaching prophet." I should also concede what the other side is entitled to here: the canon argument is the weakest strand of the critical case. The placement is first attested in a rabbinic list, the baraita in the Babylonian Talmud (Bava Batra 14b) that puts "Daniel and the Scroll of Esther" among the Writings, centuries after the book was written. And John Barton has argued that in the Second Temple period there was no closed collection of "the Prophets" at all: everything outside the Torah could count as prophecy, and Daniel, like David, was read as a prophet. What survives is a pattern rather than a proof. A book that was read as prophecy within a generation, yet ended up outside the prophetic collection when the rabbis fixed it, is consistent with a late arrival. It does not establish one on its own.

The physical evidence from Qumran is the traditional side's best argument from the manuscripts. Eight fragmentary copies of Daniel turned up in the caves, and Frank Moore Cross dated the oldest, 4QDanc, to the "late second century BC." Gerhard Hasel, defending the traditional date, pressed the point: "Inasmuch as Daniel was already canonical at Qumran at about 100 BC, how could it have become so quickly canonical if it had just been produced a mere half century before?" It is a fair question, but it has an answer. A book written to be read in a crisis, by a community that expected the end within a few years, is exactly the kind of book that gets copied and cited fast, and 4QDanc, which preserves chapters 10-11, is a copy of exactly such a crisis text. The critical editor of the Qumran Daniel scrolls, Eugene Ulrich, drew no conclusion against the Maccabean date from the fragment. As Hasel himself reports, Ulrich described 4QDanc as "only a half century later than the composition of the book of Daniel."

The Language Problem

The book of Daniel is written in two languages. It begins in Hebrew (1:1–2:4a), switches to Aramaic (2:4b–7:28), and then returns to Hebrew (8:1–12:13). This bilingual structure is itself unusual and has generated extensive scholarly discussion. But for dating purposes, the relevant question is: what kind of Aramaic does Daniel use?

The Aramaic of Daniel is Imperial Aramaic, the standard literary Aramaic of the Persian empire, and it stayed stable enough over several centuries that the language alone cannot pin the book down. Conservative scholars press this hard, and on this point they are largely right. K. A. Kitchen, in the 1965 Tyndale volume that remains the classic statement of the traditional case, called Imperial Aramaic "practically undatable with any conviction within c. 600 to 330 BC." He also noted that Daniel's word order differs from that of two Aramaic texts found at Qumran, the Genesis Apocryphon and the Job Targum, "both of them embarrassingly close in time to a supposedly second-century Daniel." But both Qumran texts are local Palestinian Aramaic, and Kitchen's own explanation of the difference is that Daniel's Aramaic "belongs to the early tradition of Imperial Aramaic," the formal literary standard, rather than to those "later and local, Palestinian derivatives." That is a difference of register and tradition, not of date, and Kitchen conceded that Daniel's early-looking features "could be argued to be survivals till the second century BC." The honest summary is that the Aramaic is compatible with the Hellenistic date the other evidence points to, and does not independently establish it. The weight falls on the vocabulary.

More damning are the Greek loanwords. In Daniel 3:5, among the list of musical instruments that signal the moment to bow before Nebuchadnezzar's golden statue, we find qarna, mashroqita, qatros, sabbeka, pesanterin, sumponeyah. Three of these are Greek loanwords: qatros (lyre), pesanterin (psaltery), and sumponeyah (symphony or bagpipe). Qatros derives from the Greek kitharis, pesanterin from psalterion, and sumponeyah from symphonia.

Greek loanwords in a text supposedly written in 6th-century Babylon present an obvious problem. Greeks were present in the Near East long before Alexander's conquests in the 330s BCE, as traders, mercenaries and craftsmen, but Greek words in the Aramaic of that period are another matter. Kitchen, the ablest advocate of the early date on this point, could cite from the pre-Alexandrian Aramaic papyri one certain Greek term, stater, a coin name "clearly labelled 'Greek money'," and one probable second. Daniel 3 has three names of musical instruments in a single verse. The presence of Greek musical terminology in a supposedly pre-Hellenistic text points toward composition during the Hellenistic period, when Greek language and culture had thoroughly penetrated the eastern Mediterranean and Near East.

Kitchen's reply is that the argument proves too much. "The idea that Greek words and influence could not affect the Near East or appear in Aramaic before Alexander the Great must be given up," he wrote, and he set out the evidence for Greek traders on the Syrian coast from the eighth century BCE and for Greek mercenaries in the Babylonian forces around 605–585 BCE. That is true, and it disposes of the crude version of the argument. But the question is not when Greek had these words (the kithara is as old as Homer, as Kitchen notes). It is when Aramaic borrowed them, and for the period before Alexander the evidence is a single coin name. Kitchen is right that the words are few, three against some twenty Persian ones, and that instruments tend to travel under their own names. That is the profile you would expect from a Hellenistic-era writer using a conservative literary Aramaic whose older loanword layer is Persian. Three words cannot settle the date alone. They point the same way as the history.

The Historical Errors

One caution first. Collins regards the court tales of chapters 1–6 as older than the visions: traditions about Daniel that were collected before the Maccabean crisis, not composed during it. They are still not eyewitness records. They are legends about a figure set in a court their tellers knew only at a distance, and that is all the argument below needs.

If Daniel were a genuine 6th-century document written by someone with firsthand knowledge of the Babylonian and early Persian courts, we would expect it to get the basic historical facts right. It doesn't. And the errors are not minor.

Start with the biggest problem: Darius the Mede. Daniel 5:31 records that after Belshazzar's feast and the fall of Babylon, "Darius the Mede received the kingdom, being about sixty-two years old." Daniel 6 then describes Darius the Mede's administration, his decree, and Daniel's encounter with the lions' den. The problem is that Darius the Mede does not exist in any historical record outside the book of Daniel. Babylon fell to Cyrus the Great of Persia in 539 BCE. There was no Median interregnum between the Neo-Babylonian and Persian empires. The Median Empire had already been conquered by Cyrus before he took Babylon. No ancient source — not the Babylonian chronicles, not the Cyrus Cylinder, not Herodotus — mentions a Darius the Mede ruling Babylon. Xenophon's Cyropaedia does give Cyrus a Median king above him, but it calls that king Cyaxares, and the Babylonian records know nothing of him either.

H. H. Rowley's Darius the Mede and the Four World Empires in the Book of Daniel (1935) is still the classic critical study. Rowley examined the identifications proposed up to his day, and even D. J. Wiseman, defending the traditional date, grants that Rowley "very ably discussed many of these theories and shown that Darius the Mede could not be identified with Cambyses, Gobryas, Astyages or Cyaxares." Defenders of the traditional date have since offered new candidates. John C. Whitcomb (1959) revived the governor Gubaru. Wiseman himself proposed in 1957, and defended in the 1965 Tyndale volume, that "Darius the Mede" is Cyrus under another name, reading Daniel 6:28 as "the reign of Darius, even (namely, or i.e.) the reign of Cyrus the Persian." And Steven Anderson (2014) and Rodger Young (2021) have revived Xenophon's Cyaxares. None has won general acceptance. The Gubaru and Cyaxares proposals need a Median king whom the Babylonian chronicles leave no room for. Wiseman's avoids that problem, but only by making "Darius the son of Ahasuerus, of the offspring of the Medes" (9:1) the same man as Cyrus the Persian, son of Cambyses. That means treating "Ahasuerus" as a royal title and "the Mede" as a claim through Cyrus's mother, and it means the book gave one king two names without ever saying so. Wiseman himself called it a view that "must remain only a theory and be further tested." The most parsimonious explanation is that the storyteller, writing centuries after the fall of Babylon, confused the historical sequence and invented a Median ruler. A genuine 6th-century courtier would not have made this error.

This is where Darius matters most. The critical reading counts the four kingdoms of chapters 2 and 7 as Babylon, Media, Persia and Greece, which requires a Median kingdom between Babylon and Persia, and that is exactly the kingdom the book's Darius supplies. The traditional reading, already present in 4 Ezra and classically stated by Jerome, combines Media and Persia and makes the fourth kingdom Rome: "the fourth empire," Jerome wrote on 2:40, "which clearly refers to the Romans." Its defenders point to 8:20, where a single two-horned ram stands for "the kings of Media and Persia." But the narrative of chapters 5–6 treats the Median kingdom as a distinct reign, with "Darius the Mede" (5:31) followed, on the plain reading of 6:28 rather than Wiseman's, by "Cyrus the Persian." And the persecutor of chapter 7, the "little one" among the horns of the fourth beast (7:8), most naturally matches the "little horn" of chapter 8 (8:9), which the book itself places among the successors of "the king of Greece" (8:21–23).

And then there's Belshazzar. Daniel 5 calls Nebuchadnezzar Belshazzar's "father" again and again (5:2, 11, 13, 18) and calls Belshazzar "his son" (5:22). He wasn't. Belshazzar was the son of Nabonidus, the last king of Babylon, who took the throne from outside Nebuchadnezzar's line. Belshazzar did serve as co-regent while his father was away at the oasis of Tayma, so the author got that much right — Belshazzar was a real person who exercised royal authority. But calling him Nebuchadnezzar's son is a historical error that suggests distance from the actual events. Constable gives the traditional reply: "Nebuchadnezzar was Belshazzar's grandfather rather than his 'father,' but the original language commonly used 'father' in the sense of ancestor." The usage is real. But the premise needs a marriage between Nabonidus and a daughter of Nebuchadnezzar that no Babylonian source records, and Daniel 5 presents the relation as plain paternity, repeated for emphasis. The simplest explanation is a storyteller who knew Nebuchadnezzar as the great king of Babylon and Belshazzar as its last ruler, and joined them.

The problems keep compounding. Daniel 4 describes Nebuchadnezzar being driven from human society to live like an animal for seven "times" (Aramaic iddanin, traditionally read as seven years). No Babylonian or Greek source records this event. The Prayer of Nabonidus found at Qumran (4Q242) describes Nabonidus — not Nebuchadnezzar — being afflicted with illness and dwelling in Tayma. Many scholars believe the Daniel story is a garbled version of traditions about Nabonidus that were mistakenly transferred to the more famous Nebuchadnezzar. This kind of error is consistent with an author working from distant, confused oral traditions — not from personal experience.

Daniel 9:25–26 — The Seventy Weeks

Now let me apply all of this to the passage that generates the most apologetic excitement: the prophecy of the seventy weeks.

Daniel 9:25–26 reads, in the Jewish Publication Society's 1917 translation (JPS 1917): "Know therefore and discern, that from the going forth of the word to restore and to build Jerusalem unto one anointed, a prince, shall be seven weeks; and for threescore and two weeks, it shall be built again, with broad place and moat, but in troublous times. And after the threescore and two weeks shall an anointed one be cut off, and be no more…" (A word on translation: I quote JPS 1917 here, not the World English Bible this site uses by default, because the JPS follows the Masoretic accents and keeps the "seven weeks" apart from the "threescore and two weeks"; the World English Bible runs them together and prints "the Anointed One" as a title.)

Christian interpreters have traditionally calculated the seventy weeks (understood as "weeks of years" — 490 years) as counting from a decree to rebuild Jerusalem to the crucifixion of Jesus. The math has been done in various ways, with different starting points (Cyrus's decree, Artaxerxes' decree to Ezra, Artaxerxes' decree to Nehemiah) and different endpoints, and the calculations never quite work out without some creative adjustments to the chronology.

But the parsimonious reading — the one that fits the chapter's own context — is that the anointed one "cut off" in verse 26 is Onias III, the rightful high priest in Jerusalem, pushed out of office in 175 BCE and murdered in 171 BCE. Onias was the last legitimate Zadokite high priest. His removal and murder at the hands of pro-Hellenistic factions was a traumatic event for traditional Jews, and it came only a few years before the full-blown persecution under Antiochus IV Epiphanes. Collins, in his Hermeneia commentary, identifies the anointed one "cut off" with Onias III and notes that this identification fits the chronological and historical framework of the passage far better than a reference to Jesus.

The "going forth of the word to restore and to build Jerusalem" (JPS 1917) most likely refers to the oracle in Jeremiah 25:11–12 and 29:10, which predicted seventy years of Babylonian domination. Daniel 9 opens with Daniel reading Jeremiah's prophecy and praying for its fulfillment. The seventy weeks are a reinterpretation of Jeremiah's seventy years — multiplied by seven to yield 490 years — as a way of explaining why the restoration promised by Jeremiah had not yet fully materialized. The endpoint is not the distant future. It is the author's own time: the Maccabean crisis.

This reading makes sense of the entire chapter. It makes sense of Daniel's prayer of confession on behalf of Israel. It makes sense of the angel Gabriel's response explaining the timeline. And it makes sense of the fact that the cutting off of the anointed one is followed, within a few years, by the desecration of the temple — the abomination of desolation — which is the defining event of the Maccabean period.

The Maccabean reading has an arithmetic problem too, and it should be said plainly. James Montgomery, whose 1927 International Critical Commentary defends it, conceded that on this reading "the 62 Weeks would then take us down some 65 years too far," and that "we can meet this objection only by surmising a chronological miscalculation on part of the writer," an author who "was not embarrassed, in the absence of a known chronology," in squeezing 434 years between the return from exile and Antiochus. Montgomery, following Schürer, showed that Jewish Hellenistic chronographers made errors of exactly that size: the historian Demetrius, writing before 200 BCE, put seventy years too many between the return and the accession of Ptolemy IV. So the difference between the two readings is not that one adjusts and the other does not. It is what each has to adjust. The Maccabean reading needs an ancient writer to have been wrong about Persian-period chronology, as his contemporaries demonstrably were. The Christian reading has to choose among several decrees, sometimes redefine the length of the year, and then give the seventieth week of 9:27, with its halted sacrifice and its "one who makes desolate," some referent other than the one event in the period that matches it exactly: the suspension of the temple sacrifice in 167 BCE. One reading fits the passage's own context and its endpoint. The other does not.

The Genre Question

One more thing needs to be said, because it affects how we evaluate Daniel's "prophecies" even apart from the dating question. Daniel is not prophecy in the classical sense. It is apocalyptic literature — a genre that emerged in the Hellenistic and Roman periods, characterized by visions, angelic interpreters, symbolic imagery, pseudonymous authorship, and a deterministic view of history divided into periods or ages.

Pseudonymous authorship is the key feature here. Apocalyptic texts are regularly attributed to ancient heroes — Enoch, Moses, Ezra, Baruch — who supposedly received divine visions centuries before the real author's time. The fiction of antiquity is a genre convention. Whether its first readers saw through it is still debated. Bruce Metzger's classic survey of ancient forgery and pseudepigraphy is the usual place to start, and my own talk of "faking" above should be read with that debate in mind. But the convention itself is not in doubt. When the author of Daniel placed his protagonist in the court of Nebuchadnezzar and had him "predict" events that had already occurred, he was doing what the earliest Enoch books had already done in the third century BCE, and what 4 Ezra, 2 Baruch and the Apocalypse of Abraham would do after him, in the generations after Rome destroyed the temple in 70 CE (Collins, The Apocalyptic Imagination).

Newsom discusses this at length in her commentary: the pseudonymous framework is not incidental to the book's purpose. It serves a theological function. By presenting contemporary events as foreknown and predetermined by God, the author reassures his audience — Jews suffering under Antiochus's persecution — that their suffering is part of a divine plan and will soon end. The "predictions" are not meant to prove that God can see the future. They are meant to comfort a persecuted community by embedding their crisis within a narrative of divine sovereignty.

This is a fundamentally different purpose than what Christian apologists ascribe to the book. Apologists treat Daniel as evidence — proof that God exists because only God could predict the future with such accuracy. But the author wasn't trying to prove God's existence. He was trying to sustain faith in God's justice during a period when that justice was very hard to see.

The Verdict

Let me sum this up as directly as I can.

The book of Daniel was completed in the 160s BCE, and its visions were written then. The evidence converges from every direction: the accuracy of the "predictions" breaks down precisely at the author's own time, the book ended up outside the prophetic collection, the Aramaic and Greek loanwords are consistent with the Hellenistic period, the historical errors about the 6th century reveal an author writing at a great distance from the events he describes, and the genre conventions of apocalyptic literature explain exactly why the book takes the form it does.

The seventy-weeks prophecy does not predict Jesus. It describes the murder of Onias III and the desecration of the temple under Antiochus IV Epiphanes — events that the author had either witnessed or heard about firsthand.

The four-kingdom scheme does not predict the rise and fall of empires centuries in advance. It describes empires that had already risen and fallen by the 2nd century BCE, culminating in the Hellenistic kingdoms that were the author's contemporary reality.

The Maccabean date of the visions is not controversial in the academic study of the Hebrew Bible, even if the individual strands of the case — the Aramaic, the loanwords, the canon — carry different weight with different scholars, and I have tried to mark which are strong and which are weak. Collins, Newsom, Rowley, Goldingay, Portier-Young — the scholars who have spent their careers on this material — treat the Maccabean dating of the visions as established. Goldingay, an evangelical, shows that it is possible to hold the late date and the faith together. The sixth-century date survives mainly among conservative scholars such as Kitchen, Wiseman and Hasel, and in commentaries like Constable's, and its best arguments are the ones answered above.

That is my conclusion; you may reach your own. But I'd challenge anyone to engage with the evidence — really engage with it, not just wave it away with "God can do anything" — and come to a different conclusion. The predictions work because they aren't predictions. They're history. And the moment they try to be predictions — the moment they project forward from the author's own time — they fail. That pattern tells you everything you need to know.

And there it is. Prediction or postdiction? The evidence forces the choice, and the choice is clear.