The claim
The book of Nahum announces the destruction of Nineveh, capital of the Neo-Assyrian empire. Nineveh fell in 612 BCE to a coalition of Babylonians and Medes. Both halves of that sentence can be examined, and they have to be examined separately, because they are not the same kind of statement: one is about an event in the ground and the archives, the other is about when a poem was written.
This is the strongest case in the archive, and it is deliberately the first dossier written. If the method cannot be shown working here, it cannot be shown working anywhere. It is also the case where the temptation to overstate is strongest, which is why the verdict below is not the top grade.
The dating problem
Everything turns on this section, so it comes before the evidence rather than after it.
The earliest limit is externally anchored
Nahum 3:8–10 addresses Nineveh by pointing at a city that has already fallen: No-Amon, that is, Thebes in Upper Egypt. The passage speaks of that fall in the past tense and uses it as the argument — if Thebes could fall, so can you. Thebes was sacked by Ashurbanipal in 663 BCE, an event recorded in the Assyrian royal inscriptions of that reign.
That gives a floor. The text, or at least this part of it, cannot have been composed before 663 BCE. This limit does not depend on any view about prophecy: it comes from a comparison between what the poem treats as already done and when that thing was done.
The latest limit is where the argument becomes circular, and most treatments hide it
The usual ceiling given for Nahum is 612 BCE, the fall of Nineveh. The reasoning is that the poem announces the fall as something still to come, so it must have been written before the fall happened.
That reasoning assumes exactly what is under examination. A text written after an event and cast in the voice of an announcement would read the same way. Using the future tense of the poem to date the poem before the event, and then citing the event as fulfilment of the poem, is a closed loop. It is the single most common error in popular treatments of this passage, and this site will not commit it.
The honest upper limit — the latest date that can be set without assuming the conclusion — is the earliest surviving manuscript witness. For Nahum that is Qumran: the pesher on Nahum (4Q169) and the Minor Prophets material, from the last two centuries BCE. That is roughly five hundred years after either candidate date, so the manuscripts cannot arbitrate between them. Proposals in the published literature run from shortly after 663 BCE down to the Maccabean period.
What the field actually holds
The majority position places the composition between 663 and 612 BCE — that is, after Thebes and before Nineveh. It is a real majority and it is held for reasons, chiefly that the poem shows no knowledge of how Nineveh actually fell and describes a siege in generic terms.
But the dissent is not fringe. Some scholars place the composition at or immediately around 612, as a text produced in the moment rather than in advance; others accept an early core with later editorial work. That dissent is a reason the top grade is withheld here, not a footnote to be skipped.
What the external record says
The Fall of Nineveh Chronicle — tablet BM 21901 in the British Museum, catalogued as ABC 3 — is a Babylonian administrative chronicle covering the tenth to the eighteenth regnal years of Nabopolassar, roughly 616 to 608 BCE. It records the campaign against Nineveh in Nabopolassar’s fourteenth year, 612 BCE: a siege running from the month of Simanu to the month of Abu, and the fall of the city.
What makes the chronicle usable is not that it agrees with anything. It is that it was produced by the Babylonian administration that took the city, in a genre concerned with dating events for record-keeping, by people with no interest whatever in a Judahite oracle. It is independent of the tradition that produced the text under examination, which is the only property that matters when the question is external attestation.
What the two records actually share, stated exactly
Comparison is only worth anything if the claims being compared are named. Here they are.
- That Nineveh was destroyed. Both. This is the substance of the match, and it is not a small thing: Nineveh was the capital of the dominant power of its century, and its destruction was not an obvious outcome to predict.
- That the destruction came from a coalition rather than a single power. The chronicle names the Babylonians and their allies. Nahum 2–3 describes attackers without naming them.
- That the city was taken by siege. Both, though the poem’s siege imagery is conventional and could describe almost any Near Eastern siege.
And here is what is not shared, which matters just as much:
- The chronicle dates the fall precisely, by regnal year and month. The poem gives no date at all.
- The chronicle names Nabopolassar and his allies. The poem names no attacker.
- The poem’s most quoted specific detail — the river gates opened and the palace dissolved (Nahum 2:6) — has no counterpart in the chronicle, which says nothing about a river.
So the match is real and it is general. A reader who wants the poem to have described the mechanism of the fall in advance will not find that here, and should be told so plainly rather than left to discover it.
Where this dossier could go wrong
- Letting the strength of the attestation carry the dating. The event is superbly attested. That is not evidence about when the text was written. The two questions are independent and this dossier keeps them apart on purpose.
- Treating the agreement as more specific than it is. The chronicle records a military outcome; the oracle is a poem of judgement. Both say the city fell. Beyond that, the overlap thins fast.
- Quoting the chronicle from a popular retelling. The reading has to come from the critical edition — Grayson’s Assyrian and Babylonian Chronicles and Glassner’s Chroniques mésopotamiennes — with line references, not from a website that has copied a website.
- Assuming Nahum is a unity. If the book has an early core and later editing, then “when was Nahum written” may not have one answer, and a verdict that assumes it does is answering a question the text may not pose.
The verdict
2 · Probable. The external attestation is about as solid as this field ever gets: an independent, hostile-neutral, closely dated administrative record of the exact event, held in a public collection under a permanent identifier that anyone can follow. The dating is broadly accepted but carries relevant dissent, and the upper limit of the range cannot be fixed without partly assuming the conclusion.
The top grade is withheld deliberately, and this is the case where the dating clause of grade 1 applies for exactly the reason it exists. Nahum announces the fall rather than reporting one it postdates, so the verdict is measured under the prophecy criterion, which asks that the text be datable before the event by something other than the announcement itself. Here the ceiling of the dating range rests on the reading of the poem. Awarding grade 1 would mean treating a broad consensus as if it dissolved a circularity that it does not dissolve. If the flagship case in the archive cannot get full marks, that is information about the scale, and the scale is worth more than the marks.
Related: the book overview of Nahum, the Neo-Babylonian timeline, and the method note on how a text is dated.