---

I. What the Doctrine Actually Claims

Fairness first. The Watchtower’s position, stated plainly: Daniel 4 records Nebuchadnezzar’s dream of a great tree cut down, with “seven times” allowed to pass before its stump could grow again. The Society reads this dream as a prophecy with a second, end-times fulfillment — not just about Nebuchadnezzar’s seven years of madness, but about the interruption of God’s rulership through the Davidic line in Jerusalem.

That interruption, they teach, began when Jerusalem fell to Babylon — a date they fix at 607 B.C. — and lasted “seven times,” calculated as 2,520 years (7 × 360 “prophetic” days, each day standing for a year, per Ezekiel 4:6 and the day-for-a-year framework borrowed from Revelation 12’s “1,260 days” equaling “3½ times”). Counting 2,520 years forward from October 607 B.C. — with no year zero — lands on October 1914. That, they say, is when Christ began ruling invisibly in heaven, with the outbreak of the First World War two months earlier serving as corroborating “sign.”

This calculation appears throughout Watchtower literature, including their own What Does the Bible Really Teach? explanation of the 606-year plus 1,914-year addition reaching 2,520.

The entire structure depends on one load-bearing fact: that Jerusalem fell in 607 B.C., not any other year. Move that date, and 1914 moves with it.

II. Where 607 B.C. Actually Comes From

This is worth tracing carefully, because it changes what kind of claim 607 actually is. It did not arrive as an independent finding, checked against primary sources and then found to fit. It arrived as an inheritance.

The “seven times as 2,520 years” framework did not originate with the Watchtower at all. It was popularized by nineteenth-century Second Adventist writers — John Aquila Brown as early as 1823, later E. B. Elliott, Robert Seeley, Joseph Seiss, and Nelson Barbour — decades before Charles Taze Russell adopted it. Elliott himself floated 1914 as one possible end date among others, including the French Revolution. The number 2,520 was already fixed in Adventist circles; what remained undetermined was the starting date the count would begin from. 607 B.C. is the date that was eventually settled on and carried forward.

Early Bible Students initially used 606 B.C., not 607. The date changed by a year at some point in its early history — a fact worth noting on its own terms, without asserting why the change was made, because it shows the figure has not been fixed by independent verification since it was first proposed.

A second date shows the same pattern: the return-from-exile date used in Watchtower chronology is 537 B.C., not 536 B.C. — even though historians had already established Babylon’s fall to Persia in 539 B.C., which is the fixed point the return date is calculated from. A one-year difference between 536 and 537 is small on its own, but it is the second instance of a secondary date landing exactly where the 607-to-1914 count requires, rather than being fixed independently and then found to agree with it. Two such instances is a documented pattern, stated here without claiming to know what any individual writer intended by it.

III. What the Primary Evidence Actually Shows

A. Babylonian court records

Neo-Babylonian chronology does not rest on guesswork. It rests on more than two thousand dated cuneiform business and administrative documents from the reigns of Nebuchadnezzar, Neriglissar, and their successors — a body of evidence so extensive that Yale Assyriologist Raymond Philip Dougherty called it the necessary “ultimate criterion” for settling Neo-Babylonian chronological questions, as early as 1929, decades before the current debate sharpened.

These records place the fall of Nebuchadnezzar’s regnal years, the reign of his successors, and the surrounding decades with precision — though it is worth being exact about what they do and do not directly attest. No Babylonian record of the destruction of Jerusalem itself has yet been found; the Babylonian Chronicle series that would cover those years is broken precisely at that point. What the tablets do fix with precision are the anchor dates around the event — most importantly the capture of Jehoiachin, dated in the Babylonian Chronicle to the second of Adar in Nebuchadnezzar’s seventh year, which converts to March 597 B.C. Once that anchor is fixed, the biblical text’s own internal count from Jehoiachin’s captivity to Jerusalem’s fall — eleven years, per 2 Kings 24:12 and 25:2 — places the destruction in 587/586 B.C., not 607. This is a derived date, not a directly attested one, and it is worth saying so plainly rather than overclaiming.

B. VAT 4956 — the astronomical anchor

This is the single most important primary text in the whole debate, and it is worth walking through carefully, because the Watchtower has published a direct, technical defense of 607 B.C. against it — making this the one place where a careful reader needs the details, not just the conclusion.

VAT 4956 is a Babylonian astronomical diary explicitly dated to “Year 37 of Nebuchadnezzar.” It records the moon’s and five visible planets’ positions against fixed stars throughout that year, plus one lunar eclipse — the kind of data that can be checked against modern astronomical computation with very high precision, since planetary and lunar positions are not a matter of interpretation once the observation is fixed.

Assyriologists Abraham Sachs and Hermann Hunger, working independently of any Watchtower framework, computed that all thirteen sets of lunar positions and the recorded eclipse match the year 568/567 B.C. as Nebuchadnezzar’s 37th year — which places his 18th year, the year Jerusalem fell, at 587 B.C.

The Watchtower’s 2011 two-part Watchtower article response argued instead for 588 B.C. as the 37th year (yielding 607 for the 18th), on the grounds that the fifteen sets of planetary observations on the tablet are ambiguous enough to set aside, and that the lunar data alone fits 588 better once an intercalary month is factored in.

The strongest version of this argument was made not by the Watchtower’s anonymous editorial staff but by Rolf Furuli, a Norwegian scholar of Semitic languages, and it deserves to be stated fairly before it is answered. Furuli’s case is that astronomical diaries were a copied genre, sometimes transmitted long after the observations they record, so a diary’s stated regnal year cannot simply be trusted without independently checking the raw celestial data — and that when this is done for VAT 4956, the lunar positions, which he considers less prone to scribal corruption than the planetary ones, fit 588/587 B.C. as well as 568/567, once a corrected reading of the “9th of Nisan” line and the placement of an intercalary month are applied. This is a genuine philological and astronomical argument, not a fringe claim, and it is worth taking seriously as exactly that.

The rebuttal to that argument, laid out in detail by chronologist Carl Olof Jonsson in his review of the pro-607 case, is that the 588 B.C. reading only works by discounting the planetary data as unreliable and by relying on a disputed reading of a single cuneiform sign (a “9” that earlier researchers read as “8”) to force one lunar position into alignment — while the 568/567 B.C. reading requires no such adjustments and is corroborated by at least nine other independent astronomical tablets from the same period, all converging on the same chronology. A single tablet reading massaged to fit a predetermined date is a very different kind of evidence than nine tablets independently agreeing.

C. Convergence across sources

The 587/586 B.C. date is not an isolated Assyriological claim. It aligns with the Nabonidus Chronicle, Ptolemy’s Canon of Kings, and Egyptian synchronisms for the same period — an intersecting web of independent record-keeping traditions (Babylonian court scribes, Hellenistic astronomers working from older king lists, Egyptian regnal records) that had no reason to coordinate with one another and every reason to record their own kings’ and neighbors’ reigns accurately for their own administrative purposes.

IV. Scripture Itself Points to 587, Not 607

This is the section that matters most, because it removes the false choice the Watchtower presents: accept secular chronology, or accept the Bible. That is not the actual choice on the table.

Rodger C. Young’s peer-reviewed study “When Did Jerusalem Fall?” (Journal of the Evangelical Theological Society, 2004) makes the case from the biblical text’s own internal chronology, cross-checked against the fixed Babylonian anchor date for Jehoiachin’s captivity (597 B.C., established independently of any biblical claim) rather than requiring a fresh astronomical argument of its own. This is not an appeal to Scripture “alone” — it is an appeal to Scripture read against the one Babylonian date that is not in dispute. Young shows that Ezekiel’s own dating — his vision dated to “the twenty-fifth year of our exile” and explicitly cross-referenced as fourteen years after the city fell (Ezekiel 40:1) — cannot be reconciled with a 586 date, only with 587, once the reckoning conventions Judah’s court recorders actually used (non-accession-year counting, Tishri-based regnal years) are applied consistently across 2 Kings, Jeremiah, and 2 Chronicles. Young himself acknowledges that scholarship remains genuinely split between 586 and 587; this piece follows his case for 587, but the more important point for a Witness reader is that both figures sit a full two decades from 607, and neither requires the tablets to be wrong.

In other words: the Bible does not need 607 B.C. to be true. It needs careful attention to how ancient Judean court scribes counted regnal years — and when that attention is paid, Scripture and the Babylonian tablets agree with each other, on 587.

This is worth stating directly: rejecting 607 B.C. is not rejecting biblical chronology, and it is not rejecting Jeremiah. It is rejecting one specific calculation whose documented history is inheritance from earlier calculations and internal adjustment, not independent verification against primary chronological evidence.

V. Why This Matters Theologically

If 607 B.C. is wrong, 1914 has no chronological foundation. And 1914 is not a peripheral date for Jehovah’s Witnesses — it is the year they teach Christ began ruling invisibly, and it is the necessary first link in a chain the Watchtower has built explicitly on top of it: 1914 establishes the invisible presence; the “generation” doctrine (itself repeatedly reinterpreted) is anchored to that presence; and the appointment of the “faithful and discreet slave” in 1919, which the Governing Body identifies with itself, is dated and justified specifically in relation to the 1914 timeline — the Watchtower’s own account states that Christ, together with the Father, inspected the condition of God’s people from 1914 to early 1919 before the appointment was made, tying the appointment’s date directly to the earlier one. Each link depends on the one before it. This does not mean every aspect of the Governing Body’s authority collapses if 607 falls — the organization could, in principle, retain other grounds for its role. But the specific claim that 1914 marked a verifiable, divinely revealed turning point loses its only stated chronological basis, and that basis is what this article has examined.

Take away 607, then, and what remains is not a minor date correction. It removes the specific chronological claim that the Governing Body’s authority has been built on top of — the claim that they, uniquely, were positioned to recognize when the “last days” began. Someone who works through VAT 4956 and the biblical chronology carefully is not thereby doubting Christ’s kingship, and reaching a different conclusion than the Watchtower does not make that person dishonest. It means asking a narrower, answerable question: whether one specific nineteenth-century calculation — inherited from earlier Adventist sources, then adjusted internally more than once as the surrounding math required — actually tells him what the organization says it tells him.

VI. What It Was Like to Wait

As 1914 drew near, the confidence in print began to waver. Early Watch Tower writings that had once called the date settled and certain grew noticeably more tentative in the final months, hedged with language acknowledging the writers were not infallible and that things might not fully develop as expected. When the year passed without the visible transformation so many had anticipated, the entire prophetic framework needed rewriting — a process that took decades to settle into anything like its current form.

This is worth sitting with for a moment, not to mock people who were doing their honest best with the light they had, but as a documented historical fact worth setting alongside today’s confidence: the chronology that felt settled to Bible Students in 1913 required rewriting within two years, regardless of how certain it had sounded in print beforehand. That does not by itself prove today’s date is wrong. It is a reason, though, to weigh the astronomical and documentary evidence above on its own terms, rather than on how settled any single presentation of the date sounds — since this history shows those two things can come apart.

VII. The Door That Stays Open

None of this is offered as a trap, and it is not a triumph. Questioning a specific calculation is not the same as questioning God, and it is not a betrayal of the sincere desire that first drew someone to study the Bible seriously — a desire the Watchtower itself, whatever else is true, actually cultivated in its people. If VAT 4956 and the Bible’s own internal chronology both point away from 607 B.C., that is not bad news about God. It may be an invitation to look further back than 1879 — to a faith older than any single organization’s calculation, examined and argued over and still standing, held by men and women who died defending Christ centuries before Charles Taze Russell wrote a word. That door costs nothing to walk through, and no one on the other side of it is waiting to shame anyone for having asked a hard question first.