Prophecy and Apocalyptic as Genres

"Then Jeremiah got another scroll ... They also added on this scroll several other messages of the same kind." — Jeremiah 36:32, NET¹

"The added section must be considered secondary because of its contents and, hence regarded as a post-exilic retrospective gloss." — Emanuel Tov, on Jeremiah 27:22, 1999²

Summary

Skeptics make three claims about the Bible's prophetic books.  They grew over centuries and were updated after the events they describe.  Some were written under false names.  And their symbols were messages to their first readers, not forecasts for later ones.  Each claim is partly true, and each matters for the claim that the Bible predicts the future.  But none of them reaches the texts that were copied before the events they are said to predict.  This article tests each claim against the books themselves, the manuscripts, and the best defenders of each side, and scores what follows by the eight tests of How to Test a Fulfilled Prophecy.

1. The prophetic books grew, and the Bible says so.  Jeremiah dictated a scroll, the king burned it, and Jeremiah dictated it again, and "several other messages of the same kind" were added.¹  The conservative Egyptologist Kenneth Kitchen agrees that "prophetical books were not first written down on one specific occasion."³  Growth by collection is common ground.  The question is what was added, by whom, and when.

2. Jeremiah survives in two editions, and the second adds at least one promise after it came true.  The Greek Jeremiah is about one-seventh shorter than the Hebrew and puts the oracles against the nations in a different place.  Two Hebrew fragments from Qumran, 4QJerᵇ and 4QJerᵈ, copied in the first half of the second century BC, side with the short form.²  In the longer Hebrew edition, a threat that the temple vessels would go to Babylon gains a promise that God would "bring them back and restore them to this place."¹  The short edition has only the threat.  Emanuel Tov, the leading textual critic of the Hebrew Bible, calls the addition "a post-exilic retrospective gloss."²  It is the clearest case in the manuscripts of a biblical prediction completed after the event.  Both editions are older than Christianity, so the finding touches no prophecy about Jesus.

3. Isaiah 40-55 probably comes from the end of the exile, on evidence that does not assume prophecy is impossible.  These chapters speak to exiles in person, treat Jerusalem as already in ruins and Cyrus as already on the march, and differ in style from chapters 1-39.  S. R. Driver, who set out the classic case in his Introduction, wrote that the late date does not reduce the chapters "to a vaticinium ex eventu," a prophecy after the event, because they were plainly written before the fall of Babylon they announce.⁴  The defenders of one Isaiah, from E. J. Young to John Oswalt, rest on the book's title, on readers who called it Isaiah's (Ben Sira about 180 BC, the Great Isaiah Scroll, the New Testament), and on the possibility of prediction.  Their testimony shows how the book was read by about 180 BC, not who wrote each part.  Oswalt's charge that critics beg the question is fair against some of them and does not touch the argument from the prophet's standpoint, though that argument is not proof.  Who Wrote Isaiah 40-66? weighs the question in full.

4. Other books show later additions too, and editing is not forgery.  Isaiah 2:2-4 and Micah 4:1-3 are the same oracle.  Zechariah 9-14 has its own headings and a different world.  The headings of the prophetic books speak of the prophet in the third person.  A later hand that collects, heads and extends a prophet's words is different from a writer who invents a prophet or rewrites a failed prediction.  Only the second would count against prophecy, and the manuscripts show it only in Jeremiah's second edition, most clearly at 27:22.

5. Apocalypses such as 1 Enoch, 4 Ezra and 2 Baruch were written under ancient names and review past history as prophecy.  John Collins, whose definition of the genre scholars still use, explains why the two go together:  the false name "makes possible the ex eventu prophecies of history between the time of the alleged seer and that of the actual author."⁵  Whether Daniel 7-12 belongs to this pattern depends on its date, which Dating the Book of Daniel argues.  The genre cannot settle the date on its own, because Collins himself grants that if Daniel's predictions are real, its likeness to the apocalypses "would be greatly diminished."⁵

6. Whether ancient readers saw through the false names is disputed, and the evidence leans against a harmless convention.  The evangelical Richard Bauckham argued in 1978 that a pseudonym was not fraud but a confession that "the age of prophecy has passed."⁶  Joyce Baldwin answered that the theory asks a pseudonym both to deceive no one and to add authority, and "cannot have it both ways."⁷  The sealed-book device, in which a seer is told to hide his book until the end (Daniel 12:4, 1 Enoch 1:2), only explains a late appearance if readers believed it.  Early Christians did condemn the forgeries they caught.

7. The Bible's own rule makes many predictions conditional, and that cuts both ways.  Jeremiah 18:7-10 says a threatened nation that repents may be spared, and Jonah's Nineveh was.  The Reformed scholar Richard Pratt concludes that "with rare exception, Old Testament prophets did not speak of what had to be, but of what might be."⁸  By the method article's symmetry rule (Test 6), a prediction that may lawfully fail cannot count much when it succeeds.  Apocalyptic is the opposite case:  it frames its timetable as fixed, "for what has been decreed must occur" (Daniel 11:36), so the conditional rescue is not available for its misses.

8. Apocalyptic symbols were decoded for their first readers.  Daniel's angel explains the beasts as kings and kingdoms, and names Media, Persia and Greece.  Revelation calls on "the one who has insight" to calculate the beast's number, and "Nero Caesar" spelled in Hebrew letters adds up to 666.  A second-century variant, 616, matches the Latin spelling of Nero without its final n.  But Irenaeus, the earliest writer to discuss the number, did not know the Nero reading.  The first-audience reading fits the books' own words ("the time is near," Revelation 1:3) better than readings that find modern nations in the symbols.

9. What follows for prediction.  These findings remove some items from the evidence:  Cyrus by name, if Isaiah 40-48 dates from the 540s BC;  Daniel 11, if Daniel dates from the 160s BC;  the return of the temple vessels in Jeremiah 27:22;  and every conditional oracle that is counted only when it succeeds.  They do not touch the strongest cases.  Every addition that can be dated happened before the oldest surviving copies, and those copies are older than Christianity.  Isaiah 53 and Daniel 9, measured against the life of Jesus, pass Test 1 whichever theory of their composition is true.

Part 1: Who wrote Isaiah 40-66?

Chapters 40-66 of Isaiah speak to exiles in Babylon and name Cyrus, the Persian king who took Babylon in 539 BC, about 160 years after Isaiah's last dated oracle.  Since the late eighteenth century most scholars have assigned chapters 40-55 to a prophet of the exile and 56-66 to writers after the return.  Defenders of one author, from Oswald Allis to John Oswalt, answer that the late date rests on "the prior conviction that Isaiah of Jerusalem could not have known the future in any supernatural way."⁹

Who Wrote Isaiah 40-66? weighs both cases in their makers' words.  Its verdict is that chapters 40-55 most probably come from the 540s BC, and that the arguments which carry that date do not assume prophecy is impossible.  The prophet speaks to the exiles in person, treats Jerusalem's fall as past, and points to "former things" already fulfilled as proof of his new predictions.  Williamson notes that this argument "presupposes an acceptance of the validity of predictive prophecy."¹⁰  But the case is not proved.  It rests on a rule that prophets speak to their own generation, and Oswalt grants that the single-author view makes the book "unique" rather than impossible.⁹  The outside witnesses, from Ben Sira and the Great Isaiah Scroll to the New Testament, show how the book was read from about 180 BC on, not who wrote it, and the style statistics split.

For prediction the result is narrow.  On the late date, Cyrus by name drops out of the evidence.  Isaiah 53 does not:  it was copied before the birth of Jesus whatever its date, so its Test 1 against the Gospels is untouched.

Part 2: Jeremiah's two editions

What the manuscripts show

Jeremiah's book tells part of its own history.  In the fourth year of King Jehoiakim, about 605 BC, Jeremiah dictated his oracles to the scribe Baruch, who "wrote down in a scroll all the Lord’s words."¹  The king cut the scroll up as it was read and burned it.  Jeremiah dictated it again, and the chapter ends:  "They also added on this scroll several other messages of the same kind."¹  So the book's first edition grew in the prophet's own lifetime, and the book says so.

The manuscripts show that growth went on.  The Greek translation of Jeremiah, made in Egypt in the last centuries BC, is much shorter than the standard Hebrew text.  Emanuel Tov puts it at "one-seventh" shorter in one study and "one-sixth" in another.²¹¹  A word count in this site's Old Testament text article puts the Hebrew text behind the Greek at about 16.9% shorter than the Masoretic text, the lowest ratio of any book.  The Greek also puts the oracles against the nations in another place.  In the Hebrew they are chapters 46-51.  In the Greek they follow 25:13, which reads like a heading for them:  "everything that Jeremiah has prophesied against all the nations."²¹

For a long time scholars could argue that the Greek translator simply cut his text.  J. Gerald Janzen's study of every Greek omission, published in 1973, concluded that "the evidence does not support the commonly held theory that the translator abridged his Vorlage," meaning the Hebrew text he translated.¹²  The Dead Sea Scrolls then settled it.  Two Hebrew fragments from Cave 4, 4QJerᵇ and 4QJerᵈ, contain parts of chapters 9-10, 43 and 50.  Tov dates them to "the first half of the second century BCE" and finds that they "resemble the LXX of Jeremiah in the two major features in which the reconstructed Vorlage of that translation differs from MT, namely, the arrangement of the text and its shortness."²  Other Qumran copies follow the longer text.¹¹  So two Hebrew forms of Jeremiah were being copied in the same century, and the short one is not a translator's abridgement.

Which edition came first

Tov, Janzen and most textual critics hold that the short form is the earlier one.  Tov calls it "an earlier edition of that book (edition I; by the same token, MT is called edition II)."²  The longer edition adds words, phrases and whole sections, "the largest of which are 33:14-26 and 39:4-13."²  Janzen calls the Masoretic text "late" and "developed."¹²

The idea of two editions is older than the scrolls, and it began on the conservative side of the question.  Janzen traces the modern debate to Johann Gottfried Eichhorn in the 1780s, "who proposed that the two text traditions originated in successive editions of the book by Jeremiah himself."¹²  Kenneth Kitchen pictures the book growing from small scrolls of oracles kept by date into larger collections, and writes that "Jeremiah 36 shows us the beginning of the process."³  Gleason Archer's conservative survey makes the same point with the Greek in view.  The short edition behind it "was published in the prophet's own lifetime and first disseminated in Egypt," and later "Baruch made a more comprehensive collection of his master's sermons and rearranged the material in more logical order.  The MT doubtless preserves this posthumous edition of Baruch."¹³  On every one of these views, a prophet's book could exist in more than one authorised form.  The disagreement is about who made the second form and when.

One promise added after it came true

Most of the additions in the long edition repeat or expand what the short edition already says.  One of them changes a prediction.  In Jeremiah 27, the prophet tells the people not to believe prophets who say the temple vessels taken in 597 BC will soon come back.  He adds that the vessels still left in Jerusalem will go to Babylon too.  The Greek text, translating the short edition, ends there:  "they shall go into Babylon, says the Lord."¹⁴  The long Hebrew edition continues:  "They will remain there until it is time for me to show consideration for them again.  Then I will bring them back and restore them to this place."¹  Ezra 1:7 records that "King Cyrus brought out the vessels of the Lord’s temple which Nebuchadnezzar had brought from Jerusalem."¹

Tov judges the added promise to be "a post-exilic retrospective gloss," because it is "anticlimactic" in a speech about false prophets and its wording follows Ezra.²  His textbook calls it "a vaticinium ex eventu," a prophecy written after the event, and lists three more places where the long edition shows post-exilic touches:  25:14, 27:7 and 29:6.¹¹  He does not call the editor a forger.  In his view the editor "felt that he could not leave the words of Jeremiah without correction."²

This is the clearest place where the manuscripts let a reader watch a biblical prediction being completed after the fact.  It is a small case.  The return of the vessels is not among the prophecies apologists usually cite, and Jeremiah's seventy-year prophecy (25:11-12, 29:10) is in both editions.  But it shows that the mechanism skeptics propose, a later hand bringing a prophecy into line with what happened, did operate inside the transmission of at least one prophetic book.

A messianic oracle missing from the first edition

The largest addition is Jeremiah 33:14-26.  It promises "a righteous descendant of David" and says that "David will never lack a successor to occupy the throne over the nation of Israel."¹  Christian writers read the passage as messianic.  Many critics deny it to Jeremiah because the Greek lacks it.  Tov does not:  "there is no sound reason for this scepticism," because the passage is full of Jeremiah's own expressions, and "the burden of proof is on those who deny the section to the prophet in whose name it has been transmitted."²  He suggests that "authentic material probably continued to circulate among the prophet's followers even after ed. I was completed."²  William Holladay, in the Hermeneia commentary, disagrees.  He dates the expanded edition "toward the end of the fifth century B.C.E." and finds that 33:14-26, "with its twin concerns of a future for Levitical priests and Davidic kingship, appears to emerge out of the same milieu as that of Zechariah 10-14 and of the Chronicler."  He also finds passages from Nehemiah's time, about 445 BC, already in the short edition.¹⁵  So the passage may be Jeremiah's own words added late, or a later writer's.  Either way it entered the book before the Qumran copies were made.

What two editions show, and what they do not

The two editions show four things.  Prophetic books kept growing after the prophet's first collection.  The additions went in under the prophet's name, which Tov sets beside the ancient practice of writing in another's name:  "pseudepigraphal authorship and revision were common practice in antiquity."²  At least one addition updated a prediction after it came true.  And the editors did not purge the book's other difficulties.  Tov himself judges that editor II "did not distort significantly the message of the prophet as handed down to him."²

They do not show that Jeremiah's predictions in general were written after the events.  The short edition already contains the seventy years, the fall of Jerusalem, and the oracles against the nations.  Its Qumran copies were made centuries after the exile, too late to prove a date before it, so Jeremiah's own predictions rest on internal evidence, the subject of Prophecies About Nations and Cities.  And both editions are older than Christianity.  Nothing in the history of Jeremiah's text bears on any claim that the prophets foresaw Jesus.

Part 3: Later additions in other books

Headings written by someone else

Most prophetic books open with a heading about the prophet, in the third person:  "This is the Lord’s message that came to Hosea son of Beeri during the time of Uzziah, Jotham, Ahaz, and Hezekiah."¹  Amos's heading dates his oracles "two years before the earthquake," which only someone writing after the earthquake could say.¹  Hans Walter Wolff, in the Hermeneia commentaries, calls Hosea's heading "the work of a Judaic redactor" and sees in Amos's dating clause "the first traces of this early redaction."¹⁶¹⁷  Conservatives do not need to resist this.  Isaiah himself tells his followers to "tie up the scroll as legal evidence ... and give it to my followers," and Jeremiah's book was written down by Baruch.¹  A prophet's words were collected and headed by his circle.  The question is whether the circle added new oracles in his name.

One oracle in two books

Isaiah 2:2-4 and Micah 4:1-3 are the same oracle, almost word for word:  the nations stream to the temple mount, and "they will beat their swords into plowshares, and their spears into pruning hooks."¹¹  Someone borrowed.  John Oswalt lists the options (Micah borrowed from Isaiah, Isaiah from Micah, or both from a common source) and finds that the passage "fits more naturally" in Micah.  He adds that "the possibility of literary dependence need not cause concern among those who take a high view of inspiration since the Synoptic Gospels have evident dependencies."¹⁸  Wolff goes further.  He finds the vocabulary of the shared oracle late, "These word-statistics point clearly to the time of the Persian Empire," and sets its earliest date at "the completion of the postexilic temple in spring of 515 B.C.E."¹⁹  If he is right, the oracle is an addition to both books.  If Oswalt is right, it is an older saying placed in two collections.  Either way it is a promise about the far future with no fixed date, so it neither gains nor loses anything as a prediction.

Amos's happy ending

Amos is a book of judgment on the northern kingdom.  It ends with a promise:  "In that day I will rebuild the collapsing hut of David," and the exiles will come home and "never again be uprooted."¹  Wolff sees "the hand of a redactor not in evidence elsewhere in the book of Amos."¹⁷  Shalom Paul, the Jewish scholar who wrote the later Hermeneia volume on Amos, grants that "there is common agreement among most commentators that the last verses are from the hand of an exilic or postexilic theologian-redactor," but finds the arguments for lateness "all of which, however, are seriously open to question."²⁰  Francis Andersen and David Noel Freedman, in the Anchor Bible, name the step critics take from the date to the genre:  once the "fallen" booth is read as the fall of Jerusalem in 587 BC, "it is only a short step to labeling these utterances as simulated prophecies after those events."²¹  They decline to take it.

Zechariah 9-14

The last six chapters of Zechariah begin with a new kind of heading, "This is an oracle," which recurs at 12:1 and at the head of Malachi.¹  They mention no Zechariah, no date, and no temple building.  Matthew cites one of their verses, about thirty pieces of silver, as words of "Jeremiah the prophet."¹  In the seventeenth century the Cambridge scholar Joseph Mede took Matthew at his word.  Chapters 9-11, he wrote, "are indeed the Prophecies of Jeremy, and that the Jews had not rightly attributed them," and some of them "should in likelihood be of an elder date than the time of Zachary;  namely, before the captivity."²²

Since then the proposed dates have spread across six centuries.  Andrew Hill, surveying them in 1982, gives "a terminus a quo of the 7th or even 8th century b.c." and "a terminus ad quem of approximately 150 or 140 b.c."²³  Bernhard Stade dated the chapters to the Greek period, partly because Zechariah 9:13 names Greece.²⁴  That is the kind of argument Oswalt warns against, since it dates a text by what it seems to predict.  Hill's own method does not:  using nineteen grammatical features of late Hebrew, he found the chapters "linguistically similar to Haggai, Zechariah 1-8" and dated them "between ca. 515 and 450 b.c., a date of ca. 515-475 b.c. being preferred."²³  Carol and Eric Meyers, on historical grounds, put them in the first half of the fifth century and judge that assigning them "to the Hellenistic world is unwarranted."²⁵  The Persian-period date has the better support.  Whichever date is right, chapters 9-14 were written centuries before Jesus, so their place in the messianic texts, discussed in Messianic Prophecy:  An Overview, does not turn on it.

Editing or forgery?

The principle at work in all these cases is the one critics call vaticinium ex eventu, prophecy written after the event.  It is sound when a text is accurate in detail up to a point and then goes wrong or vague, because that turning point marks the writer's own time.  Daniel 11, the Animal Apocalypse and Jeremiah 27:22 are cases where it applies.  It is unsound when it is used to date a text by its accuracy alone, which is Oswalt's complaint and the method article's Test 1.

Three questions separate an edited prophetic book from a forged one.

1. Does the addition claim a false author? Most of the additions above are anonymous. Isaiah 40-66 and Zechariah 9-14 never name their writer. Driver noted that Isaiah 40-66 "nowhere claims either to be written by Isaiah, or to have originated in his age."⁴ The false attribution, if there is one, lies in the combined title, not in the text. Daniel 7-12 is different: it speaks in Daniel's own voice, so if it is late it is pseudonymous in the full sense, as Part 4 discusses.
2. Does the addition change a prediction after the event? Jeremiah 27:22 does, and Tov finds smaller touches of the same kind at 25:14, 27:7 and 29:6. The other additions in this Part add promises without dates, or headings, and could not be checked against events when they were written.
3. Do the misses survive? If editors had brought prophecies into line with history as a rule, the books would contain few failed oracles. They contain many. Ezekiel kept his own admission that Nebuchadnezzar's army "received no wages from Tyre for the work he carried out against it,"¹ and Prophecies About Nations and Cities finds the failed oracles against Tyre, Egypt and Babylon's violent sack all left in place. That is strong evidence that the editors of the prophetic books did not systematically rewrite them, and it is evidence a skeptic can check.

Part 4: Pseudonymity and the apocalypses

What an apocalypse is

In the last two centuries BC and the first two AD, Jewish and then Christian writers produced a kind of book that modern scholars call an apocalypse, from the Greek word for "revelation" that opens the book of Revelation.  In 1979 a working group of the Society of Biblical Literature, led by John Collins, surveyed the candidate texts written between about 250 BC and AD 250 and proposed a definition that is still the standard one:  an apocalypse is "a genre of revelatory literature with a narrative framework, in which a revelation is mediated by an otherworldly being to a human recipient, disclosing a transcendent reality which is both temporal, insofar as it envisages eschatological salvation, and spatial, insofar as it involves another, supernatural world."²⁶

The Jewish examples share more than that definition.  An angel explains symbolic visions of beasts, horns and numbers.  The seer is a famous figure from long ago:  Enoch, seventh in line from Adam;  Ezra or Baruch at the fall of Jerusalem;  the sons of Jacob in the Testaments.  And the seer often reviews history from his own day to the writer's, as if it were still future.  Collins explains why the last two features travel together.  "All the Jewish apocalypses are pseudonymous," he writes, and in the historical apocalypses the false name "has an essential role, because it makes possible the ex eventu prophecies of history between the time of the alleged seer and that of the actual author."⁵  An Enoch who lived before the Flood can "foresee" the Maccabees.  A writer under his own name in 165 BC could only report them.

History written as prophecy: the Animal Apocalypse

The clearest example is the Animal Apocalypse in 1 Enoch 85-90.  It retells biblical history as a dream about animals:  white bulls for Adam and the patriarchs, sheep for Israel, and beasts and birds of prey, among them ravens, for the nations that rule them.  It runs accurately through the exile and the Greek period to a moment when lambs grow horns and "there sprouted a great horn of one of those sheep."²⁷  R. H. Charles, the Oxford editor whose translation is still standard, wrote that "in the ‘great horn’ it is impossible to find any other than Judas Maccabeus," who led the Jewish revolt against Antiochus IV from 166 BC.²⁷  Then the vision turns.  The great horn is still fighting when God himself comes down, judges the shepherds and the blinded sheep, and builds a new house.  Charles drew the obvious conclusion:  "this Section must have been written before the death of Judas, 160 B.c.  The author, therefore, must have expected the Messianic kingdom to appear within twenty years or more."²⁷

Scholars of every school, conservative and critical, date it to the second century BC, not to Enoch.  It has every feature of a text written after the events it "predicts":  detailed accuracy up to a point in the writer's own time, then a turn to the end of the world that did not happen.  That is the turning-point test of the method article's Test 1, and the Animal Apocalypse is the control that Test 8 requires.  Any rule that would credit Daniel 11 with foreknowledge must also be tried on 1 Enoch 85-90, and the rule has to say why it gives a different answer.

The same books explain why an ancient seer's book was only now being read.  Enoch says his vision was "not for this generation, but for a remote one which is for to come."²⁷  Daniel is told, "close up these words and seal the book until the time of the end."¹  In 4 Ezra, written after the Romans destroyed the temple in AD 70, Ezra dictates ninety-four books and is told to publish twenty-four, the Hebrew Bible, and to keep "the seventy last" for "the wise among thy people."²⁸  Charles took the seventy as "the apocalypses which were secret books."²⁸  Read as history, the sealing story explains how an old book could surface late.  It only does that work if readers believed it.

Is Daniel an apocalypse?

Daniel 7-12 has the features of the genre:  an angel interprets symbolic beasts and horns (7:16, 8:16), the visions review the kingdoms from Babylon to Antiochus IV, the seer writes in the first person from the sixth century BC, and the book is sealed until the end (12:4).  Most critical scholars therefore class Daniel with 1 Enoch as an apocalypse and draw the conclusion Collins draws:  "the pseudonymity of Daniel has long been established beyond reasonable doubt."⁵

Conservative scholars reply that the classification is doing work it cannot do.  Joyce Baldwin, in the Tyndale commentary, granted that Daniel is "the example par excellence of apocalyptic in the Old Testament, and yet, on most definitions of the term, it proves to be an exception," since chapters 1-6 are court stories about real kings with nothing symbolic about them.²⁹  Her objection was to the reasoning:  "does it necessarily follow that the book of Daniel must be pseudonymous merely on the ground that it is an apocalyptic kind of book?  It is the methodology which is questionable."²⁹  In a companion article she noted that the critical certainty rests "in the last analysis" on "the content of those chapters which seem to foretell future stages of world history, and in particular chapter 11."⁷

Collins concedes the logical point.  Conservatives "insist that it is not pseudonymous and does not contain ex eventu prophecy.  If this were so, the affinities with the apocalypses would be greatly diminished."⁵  So the genre label follows from the date, and the date has to be argued on its own evidence.  Collins's own case is not that prediction is impossible.  "For the critical scholar," he writes, "the issue is one of probability," and he gives his reasons:  "There is no apparent reason, however, why a prophet of the sixth century should focus minute attention on the events of the second century," and the book is accurate about Hellenistic history but "notoriously confused" about the Babylonian and Persian periods it claims as its own.⁵

That evidence belongs to Dating the Book of Daniel, which concludes that the question is undecided and that the book splits:  chapters 1-6 look older than the Greek period, while chapters 8-12 are contested, with the literary evidence favouring about 165 BC and a 2025 radiocarbon date on the Qumran copy 4Q114 favouring earlier.  This article adds one point to that case.  The genre argument is not independent evidence, as Collins grants, but the Animal Apocalypse is.  It was written in the 160s BC, about the same events, in the same symbols and the same shape, and its accurate history stops at the same moment as Daniel 11's.  Daniel's early date needs a reason why two books so alike came about in such different ways.  A defender can answer that the Animal Apocalypse imitated an older, genuine Daniel, as later apocalypses imitated earlier ones.  That reply explains the shared form.  It does not explain the shared stopping point, which is the turning-point argument that Dating the Book of Daniel weighs.

Were the first readers deceived?

Scholars give three answers.

The first is that a false name was a convention every reader understood.  Richard Bauckham, writing as an evangelical in 1978, argued that pseudonymity was "a device expressing the apocalyptist’s consciousness that the age of prophecy has passed:  not in the sense that he fraudulently wishes to pass off his work as belonging to the age of prophecy, but in the sense that he thereby acknowledges his work to be mere interpretation of the revelation given in the prophetic age."⁶  Emanuel Tov makes a similar remark about Jeremiah's editor:  "pseudepigraphal authorship and revision were common practice in antiquity."²

The second is Joyce Baldwin's.  She pointed out that the convention theory asks the false name to do two incompatible jobs:  "On the one hand we are asked to believe that this was an accepted literary convention which deceived no-one, and on the other that the adoption of a pseudonym, which presumably went undetected, increased the acceptability and authority of a work.  Those who contend that Daniel was written under a pseudonym cannot have it both ways."⁷

The third comes from the opposite side.  Bart Ehrman, a New Testament scholar and agnostic, surveyed what Greek, Roman and Christian writers actually said about false authorship.  "The first thing to note is that in virtually every instance in which an ancient author mentions forgery, he condemns it."³⁰  In his technical study he defines a forgery as "a writing that makes a false authorial claim, with the apparent intention of deceiving its readers," and argues that "this purposeful intention to deceive the reader is what made a falsely named writing a forgery."³¹  Ehrman's evidence comes mostly from Greek and Christian writers, not from the Jewish circles that wrote 1 Enoch and Daniel, and the Jewish evidence is thinner.

What the ancient readers of these books said fits Baldwin and Ehrman better than the convention theory.  The letter of Jude quotes 1 Enoch 1:9 as a prophecy:  "Enoch, the seventh in descent beginning with Adam, even prophesied of them."¹  Around AD 200 Tertullian defended the book as Enoch's own.  He knew it was "not received by some," but their doubt was whether a book "published before the deluge" could have "safely survived that world-wide calamity," and he answered that Noah was Enoch's great-grandson and that "Enoch possesses a testimony in the Apostle Jude."³²  The doubters doubted the book's survival, not its author's name.  In the ancient discussions found for this article, the question is always whether such a book is genuine, never whether its name was a device everyone saw through.

When Christian leaders did catch a book written in an apostle's name, they rejected it for that reason.  About AD 200 Serapion, bishop of Antioch, wrote of the Gospel of Peter:  "we receive both Peter and the other apostles as Christ;  but we reject intelligently the writings falsely ascribed to them, knowing that such were not handed down to us."  He had first allowed it to be read and changed his mind only when he heard that its readers were "involved in some heresy," so his rejection was not on authorship alone.³³

One favourite example should be dropped.  Tertullian also reports that a presbyter in Asia was removed from office for writing the Acts of Paul, and this is often cited as an ancient punishment for forgery in an apostle's name.  Ehrman himself points out that the presbyter wrote stories about Paul, not a work claiming to be by him, so the case shows that invented history was condemned, not that false authorship was.³¹

What the genre means for prediction

If a book is pseudonymous and reviews history as prophecy, everything it "predicts" up to the writer's own time is history, and only what lies beyond that time can be tested as prediction.  In the Animal Apocalypse that part failed.  If Daniel 7-12 is such a book, the same holds for Daniel 11:40-45, as Daniel 2, 7, and 11 shows.  If Daniel is a sixth-century book, its likeness to the apocalypses proves nothing against it.  Either way, the genre does not decide the question.  Test 1 does.

Part 5: Conditional prediction

The Bible's own rule

The prophets did not present every announcement as a fixed forecast, and one passage in Jeremiah states the rule outright:  "There are times, Jeremiah, when I threaten to uproot, tear down, and destroy a nation or kingdom.  But if that nation I threatened stops doing wrong, I will cancel the destruction I intended to do to it."¹  The same holds for promises:  a nation that turns to evil loses "the good I promised to do to it."

The book of Jonah shows the rule at work, and it shows that the first hearers understood it.  Jonah announced, "At the end of 40 days, Nineveh will be overthrown!"  The announcement had no condition.  The king of Nineveh still reasoned, "Who knows?  Perhaps God might be willing to change his mind and relent."  God did relent, and Jonah complained that he had known all along that God is "one who relents concerning threatened judgment."¹  A century after Micah foretold that "Zion will become a plowed field," the elders of Judah remembered that the city had been spared because Hezekiah repented:  "Did not the Lord forgo destroying them as he threatened he would?"¹  So conditional prophecy is not an apologist's invention.  It is part of how the genre worked, and the Bible applies it to its own oracles.

How far the rule reaches

Conservative scholars disagree about how many oracles carry an unstated condition.  Richard Pratt of Reformed Theological Seminary, in a 1993 lecture, sorted the Old Testament's predictions into three groups.  Some state their condition.  Some are "qualified by assurances," such as a divine oath, a ban on intercession, or the formula "I will not turn back."  The rest state no condition, and of these Pratt concludes that "conditions did not have to be stated explicitly to be operative."⁸  His verdict on the whole corpus is sweeping:  "with rare exception, Old Testament prophets did not speak of what had to be, but of what might be."⁸  Even the assured predictions, he adds, "are few in number and usually not very specific in their descriptions of the future."⁸

Apologists most often reach for the rule when an oracle failed.  Ezekiel foretold that Nebuchadnezzar would destroy Tyre and lay Egypt waste for forty years, and neither happened as worded, as Prophecies About Nations and Cities shows.  Robert Chisholm of Dallas Theological Seminary does not deny the misses.  His answer is that "Ezekiel’s prophecies were implicitly contingent from the beginning and that circumstances developed in such a way that God decided to alter his stated intention regarding both nations."³⁴

What the rule costs

The rule rescues the misses only by lowering the value of the hits.  The method article works this out as Test 6, the symmetry rule:  a condition rescues a miss only when the text states it or the record shows it was met, and when a condition is invoked to excuse a miss, the matching hits must be discounted by the same amount.³⁵  Pratt's own classes supply a rule that can be fixed before scoring.  Only his assured predictions are commitments that could not lawfully fail.  By his own account they are few and vague.  Every other oracle could come true or not without discrediting the prophet, so it cannot add much to a case for foreknowledge when it comes true.  Pratt grants that "most may be completely reversed."⁸  Then no hearer could know in advance which oracles would stand, and a hit tells little.  On the method article's arithmetic, if the chance of reversal were one-half and a shrewd observer could have guessed the outcome half the time, a fulfilled oracle would be worth exactly nothing as evidence.  Those two numbers are an illustration, not anyone's estimate.

Nineveh is the clearest example of the cost.  It is the Bible's own case of a prophet's announcement that did not happen, and the Bible counts Jonah a true prophet.  The same book cannot then be scored as a predictive hit for Jonah.

Apocalyptic plays by the opposite rule

The apocalyptic visions frame their timetables as fixed.  Daniel 11's king "will succeed until the time of wrath is completed, for what has been decreed must occur."¹  John Collins observes that the review of history as prophecy carries "overtones of determinism, the belief that the course of history has been set in advance," and that the oldest examples of it come from Babylon, "a culture well known for its deterministic religion."⁵  Richard Bauckham, writing as an evangelical, noted the tension inside Daniel itself:  beside "a passage like Daniel 11, with its deterministic emphasis" stands the prayer of Daniel 9, which assumes that God "responds in mercy to their repentance."⁶

This matters for scoring.  The conditional rescue fits an oracle addressed to a people who could repent.  It does not fit a decreed sequence of kings with a fixed number of days.  When Daniel 11:40-45 describes a last campaign and death of the king of the north that do not match what Antiochus IV did, the book's own frame leaves no room to say the decree was withdrawn.  The rescues on offer for those verses are a different event in the future or a non-literal reading, which Daniel 2, 7, and 11 weighs.

Part 6: Symbols and their first readers

Daniel decodes its own beasts

Apocalyptic symbols were not written to be mysterious forever.  Daniel's visions come with an interpreter, and the interpreter explains them.  "These large beasts, which are four in number, represent four kings who will arise from the earth."  "The fourth beast means that there will be a fourth kingdom on earth."  "The 10 horns mean that 10 kings will arise from that kingdom."¹  In chapter 8 the angel names names:  "The ram that you saw with the two horns stands for the kings of Media and Persia.  The male goat is the king of Greece, and the large horn between its eyes is the first king."¹  Which kingdoms the four beasts of chapters 2 and 7 are, and where the visions stop matching events, is the subject of Daniel 2, 7, and 11.  The point here is that the book tells its readers how to decode it, and its own decoding points to Persia and Greece.

Later readers decoded the same symbols again when the first decoding no longer fitted their world.  The author of 4 Ezra, writing after the Romans destroyed the temple, has an angel say that the eagle of his own vision "is the fourth kingdom which appeared in vision to thy brother Daniel;  [but it was not interpreted unto him as I now interpret it unto thee or have interpreted it]."²⁸  Charles called this "a particularly clear and interesting case of the way in which apocalyptic prediction was reinterpreted and reapplied."²⁸  A symbol that each generation can re-aim at its own enemy does not, by itself, predict anything.

Revelation's number

Revelation works the same way, and it says so.  It opens as a revelation of "what must happen very soon" and blesses those who hear it "because the time is near."  Where Daniel is told to seal his book, John is told the opposite:  "Do not seal up the words of the prophecy contained in this book, because the time is near."¹  Its most famous symbol is a puzzle set for its first readers:  "This calls for wisdom:  Let the one who has insight calculate the beast’s number, for it is man’s number, and his number is 666."¹

The puzzle uses a habit the first readers knew.  Greek and Hebrew wrote numbers with letters, so every name had a number.  A wall at Pompeii, buried in AD 79, carries the line "I love her whose number is 545," and Adolf Deissmann, who collected such texts, explained that "the name was concealed by resolving it into a number."³⁶  The number 666 is also special in itself:  it is the sum of the numbers 1 to 36, and 36 is the sum of 1 to 8.³⁷

The best-supported decoding is Nero.  Written in Hebrew letters as nrwn qsr, "Nero Caesar" adds up to exactly 666.  Some early copies of Revelation read 616 instead, and the Latin form of the name, "Nero" without the final n, drops exactly 50.  The textual scholar P. J. Williams sums up the case:  "the reading 616 can be explained as a variant of Nero Caesar, spelling the name Nero (נרו) as in Latin rather than Neron (נרון = Νέρων) as in Greek.  The omission of nun brings the total down by exactly 50 from 666 to 616."³⁸  The 616 reading is old.  Irenaeus knew it about AD 180, and the oldest surviving manuscript with the reading is a papyrus from Oxyrhynchus known as P115, which its editor dates to "the late third or early fourth century."³⁸  Two readings, each matching one spelling of one name, are good evidence that early copyists counted Nero.

But the case has limits, and they should be stated.  Williams points out that the Greek word for "beast," thērion, written in Hebrew letters, also gives 666, and without its final n also gives 616, so "the reading of P115 should not therefore be taken as giving exclusive support to the interpretation Nero Caesar."³⁸  The Nero sum depends on choices:  Hebrew letters for a book written in Greek, and the short spelling qsr for "Caesar."  The fuller spelling gives 676, and the Greek form gives 1,337.³⁷  And the earliest writer to discuss the number did not think of Nero at all.  Irenaeus tested the names Evanthas, Lateinos and Teitan, each worth 666 in Greek, and concluded that it was "more certain, and less hazardous, to await the fulfilment of the prophecy, than to be making surmises, and casting about for any names that may present themselves, inasmuch as many names can be found possessing the number mentioned."³⁹  He believed the vision had been seen "towards the end of Domitian's reign," a generation after Nero's death.³⁹

What first-audience readings mean for prediction

The symbols of Daniel and Revelation were addressed to their first readers, who were told how to decode them or invited to.  That fits the books' own words better than readings that find modern nations, rulers or technologies in the beasts and numbers.  Those modern readings fail Test 3, because no one read the texts that way before the events they now point to, and usually Test 2, because a symbol that has been fitted to Rome, the papacy and many later rulers in turn commits to none of them.  Eschatology and Israel counts fifteen dated end-time forecasts drawn from Daniel and Revelation over eighteen centuries, and none came true.

A first-audience reading does not settle whether the books predicted anything.  Revelation also promises the return of Christ and the end of the present order, which no reading places in the first century without changing what "return" means.  Those claims cannot yet be scored, for or against.  What the first-audience reading removes is the claim that Revelation's beast, or Daniel's horns, describe events of the last two centuries.

Part 7: What follows for prediction

Each finding above narrows the evidence for prediction in a different way.  None of them reaches the cases that pass Test 1 by manuscript, because every addition and every pseudonymous book that can be dated is older than the oldest copies, and those copies are older than Christianity.  The table sets out what each finding removes and what it leaves, by the method article's numbered tests.

Three cautions apply to the table.  First, "removes from the evidence" does not mean "proves false."  A late Isaiah 40-55 is still a book of forecasts, and some of them came true.  Second, the table runs in one direction only for a reason:  these findings concern how texts were composed and read, and composition cannot add foreknowledge to a text, only take away lead time.  Third, the strongest cases for prediction were never in danger from these findings.  Isaiah 53 was copied, and the seventy weeks of Daniel 9 were written even on the latest date, before the events Christians say they foretell, and their success or failure turns on the other tests, which their own articles weigh.

Objections and responses

1. "Critics date Isaiah 40-66 late only because they rule out prediction."  Some did.  John Oswalt shows that Bernhard Duhm took the title of Isaiah 14 as "proof of a postexilic date" once skeptics doubted prediction.¹⁸  But the main case does not rest on that premise.  Driver argued from the prophet's standpoint and said in so many words that the late date does not make the chapters a prophecy after the event.⁴  Radday began as a defender of one Isaiah and was moved by word counts.⁴⁰  What the standpoint argument does assume is a rule about how prophets speak, and a believer may reject that rule.  That is a real weakness, and it is the right place for the debate.

2. "Jesus and the apostles attribute Isaiah 40-66 to Isaiah, so the question is closed."  John 12:38-41 quotes Isaiah 53:1 and Isaiah 6:10 in consecutive verses, each as words of "Isaiah."¹  Oswald Allis counted about twenty quotations of Isaiah by name in the New Testament, "from both parts of the book."⁴¹  As a theological argument that is for readers who already accept the New Testament's authority, and whether naming a scroll by its title commits the speaker to a view of its authorship is itself disputed among believers.  As historical evidence it is testimony from the first century AD, later than Ben Sira and the Great Isaiah Scroll, and so it adds little to what they show.

3. "The Great Isaiah Scroll runs straight from chapter 39 into chapter 40, so the book was always one."  It shows that by the second century BC the book was copied and read as one book, which no one disputes.  Its only large break comes after chapter 33, and Kitchen draws from that break a two-part design of the whole book, not a date.³  A scroll copied four centuries after the exile cannot show who wrote what before it.

4. "If later editors added to Jeremiah, the prophetic books are forgeries."  Not by any useful definition.  A forgery makes a false claim of authorship in order to deceive.³¹  Jeremiah's editor added to a book that already carried the prophet's name, mostly by repeating and expanding what was there, and in Tov's judgment he "did not distort significantly the message of the prophet."²  The one addition that changes a prediction, Jeremiah 27:22, is a real case of updating after the event and should be scored as one.  It does not make the book's other oracles late.

5. "Daniel cannot be pseudonymous, because Jesus spoke of 'Daniel the prophet.'"  This is objection 2 again, and the same answer applies.  For a skeptic the question is the evidence for the book's date, which Dating the Book of Daniel sets out.  Baldwin's argument, that a false name either deceived its readers or was useless, is the stronger conservative point, and the ancient readers quoted in Part 4 support it.

6. "Conditional prophecy makes the Bible unfalsifiable."  Only if the condition is used in one direction.  The method article's symmetry rule (Test 6) charges every rescued miss against the hits.  Pratt's own classes show where the Bible is falsifiable:  the predictions it marks as sworn or irreversible.  A skeptic who wants a fair test should score those, and a defender should accept the result.

7. "666 is Nero, so Revelation was written under Nero and is only about the first century."  The Nero reading is the best one, and the 616 variant is strong evidence that early copyists read the number that way.  But it depends on choices:  the Hebrew alphabet rather than the Greek one in which Revelation is written, and the short spelling of "Caesar."  The fuller Hebrew spelling gives 676, and the Greek form of "Nero Caesar" gives 1,337.³⁷  Irenaeus, writing about AD 180, did not know the Nero reading at all.  The number says nothing about when Revelation was written, which is argued on other grounds.

8. "Futurist readings are true, and the first-audience reading denies prophecy."  The first-audience reading follows the books' own instructions:  Daniel's angel names the kingdoms, and Revelation says "the time is near" and "do not seal up the words."¹  A reading that finds a modern nation or technology in the symbols fails Test 3, since no one read the text that way before the event, and usually Test 2, since the symbol fits many candidates.  A futurist reading can still be held as theology.  It cannot yet be scored as a fulfilled prediction.

9. "The Animal Apocalypse shows that Jews wrote prophecies after the event, so all biblical prophecy is like that."  That overreaches as far as the opposite claim does.  Test 8 runs both ways:  the method that rejects 1 Enoch 85-90 must be passable by a genuine prediction.  Isaiah 53 and Daniel 9, measured against the life of Jesus, were copied before the events they are claimed to predict, whatever their date of composition.  Whether they pass the other tests is argued in their own articles.

10. "A ruler called Kurash ruled in Iran around 646 BC, so naming a Cyrus was not remarkable."  This is Kenneth Kitchen's argument for an eighth-century Isaiah:  "a still earlier “Cyrus” (Kurash) ruled in Parsua in 646."³  It removes an objection to the early date, but at a price.  If Cyrus was a name already borne by Iranian rulers, then naming one is less improbable, and the naming is weaker evidence of foreknowledge even if Isaiah wrote it.  The argument helps Test 1 and costs Test 7.

Issues and unanswered questions

1. The Great Isaiah Scroll at chapter 40 is described here from E. J. Young's 1958 account and Kenneth Kitchen's 2003 summary, not from the official edition (Ulrich and Flint, Discoveries in the Judaean Desert 32, 2010) or the photographs. The two accounts agree, but the column layout should be checked against the images.
2. The statistical studies of Isaiah have not been tested against each other. Radday and Adams and Rencher used different features, controls and questions. A modern stylometric study that separates "one author" from "a school steeped in one author" would decide more than either. None was found.
3. The conservative accounts of Jeremiah's two editions are thin here. Eichhorn's view that Jeremiah made both editions is reported through Janzen, and Kitchen's model of collected scrolls is general. A conservative commentary that answers Tov on Jeremiah 27:22 directly was not read.
4. The Jewish evidence on pseudonymity is thinner than the Christian. Ehrman's case rests mostly on Greek and Christian writers. David Meade's Pseudonymity and Canon (1986), the main argument that Jewish attribution claimed a tradition rather than an author, could not be obtained, and is known here only through others.
5. The Aramaic spelling of Nero's name. Delbert Hillers (1963) reported "Nero Caesar" spelled nrwn qsr in a document from Wadi Murabba'at. Neither his article nor the document's edition was read, so the article does not rely on it.
6. P115's date is given from Williams's report of the editor's judgement; the edition itself was not read.
7. Daniel's genre and date stand or fall together, as Collins concedes. If the radiocarbon date of 4Q114 discussed in Dating the Book of Daniel is confirmed by more samples, the comparison with the Animal Apocalypse in Part 4 loses much of its force.
8. Pratt's three classes of prediction have not been counted. A list of every Old Testament prediction "qualified by assurances," scored against events, would be the fairest test of prophetic foreknowledge the Bible's own rules allow. No one appears to have made it.

Verdict

The prophetic books grew.  Jeremiah says so, Kitchen agrees, and the two editions of Jeremiah prove it, including one promise added after it came true.  Isaiah 40-55 most probably comes from a prophet of the exile, on evidence that does not assume prediction is impossible, though the single-author view remains coherent and its defenders state its cost honestly.  The Jewish apocalypses were written under ancient names and review past history as prophecy;  whether Daniel is one of them depends on its date, which the genre alone cannot fix.  The Bible's own rule makes most oracles conditional, which weakens their hits as much as it excuses their misses, and apocalyptic timetables are framed as decreed, which excuses nothing.  Apocalyptic symbols were decoded for their first readers, and modern identifications of them fail the tests.

For the claim that the Bible predicts the future, these findings cut away the weakest evidence:  Cyrus by name on the late date, Daniel 11 on the late date, Jeremiah 27:22, conditional hits, and modern decodings of symbols.  They leave untouched the cases that were copied or written before the events they are said to foretell, which is where the argument for prediction has to be made.

Sources