2 Kings is the stretch of the Bible where checking is most possible. From the ninth century onward, nearly every king it names can be looked for in somebody else’s records — and most of them are there.
What the book is
Twenty-five chapters, ending in ruin. Elijah is taken up and Elisha succeeds him; the two kingdoms alternate as before; Jehu exterminates the house of Ahab. Assyria arrives and does not leave. Samaria falls in chapter 17 and the narrator stops to explain at length why. Then Judah alone: Hezekiah besieged by Sennacherib, Manasseh, Josiah’s reform after a book is found in the temple, and finally Babylon — deportations, the walls broken, the temple burnt. It closes with a short paragraph about a captive king given a seat at another king’s table.
How it is put together
The machinery is that of 1 Kings: accession synchronisms, reign lengths, the fixed verdict about high places and the way of Jeroboam, and the citation of annals that have not survived. After chapter 17 there is only one sequence left to keep.
Two theological set-pieces stand out. Chapter 17 explains Samaria’s fall as the consequence of two centuries of apostasy. Later the loss of Judah is laid at Manasseh’s door in terms so absolute that Josiah’s reform cannot undo it, which sits awkwardly beside the enthusiasm of the Josiah chapters themselves — the main textual reason for the proposal of two editions.
Two blocks stand apart from the framework. The Elisha stories are prophetic wonder-tales in a register the annalistic notices never use, generally taken for an older collection incorporated whole. And the reform of chapters 22 and 23 makes demands — one sanctuary, the Passover kept centrally, the country shrines destroyed — that correspond closely to Deuteronomy, which is where most modern accounts of Deuteronomy’s origin begin.
The dating question
The last event recorded is the release of Jehoiachin, in the middle of the sixth century BCE, so the finished work can be no earlier, and nothing in it requires a later date. The live questions are internal: whether an earlier edition ended with Josiah, and how much older material the annalistic notices preserve. An old Jewish tradition names Jeremiah as the writer. See how we date a text.
The regnal numbers
The book dates each accession by the regnal year of the opposite kingdom’s king, and gives each reign a length. In principle that is a double system which should check itself. In practice it does not: add the Judahite reigns across the span from the division to the fall of Samaria, do the same for Israel, and the totals disagree — with each other, and at points with the synchronisms.
Four devices are standardly used to reconcile them, each defensible:
- Co-regencies. A son ruling alongside his father, the shared years counted twice. Attested in the region, and implied by the text in a place or two, but nowhere stated as a rule.
- Accession-year counting. Whether the part-year in which a king takes the throne counts as his first or as his predecessor’s last — and whether both kingdoms used one convention throughout.
- Different calendars. A year beginning in spring and a year beginning in autumn will shift a synchronism by one.
- Corrupt numerals. Numbers are the easiest thing to miscopy, and the Greek text carries different figures in places.
Complete chronologies have been built that reconcile every figure using these together, and they work — but they work because each device is applied where it is needed, on choices the text does not authorise, and different systems yield different absolute dates. Where an Assyrian or Babylonian record supplies a fixed point, the fixed point wins and the arithmetic is bent to meet it. It is a real difficulty, and the reason two reference works can date the same king differently.
Where the external record touches it
The Mesha stele — found east of the Dead Sea in the nineteenth century, later broken and reassembled with the help of a paper impression taken while it was whole — carries the words of Mesha king of Moab. He records that Omri king of Israel oppressed Moab, that Omri’s son continued it, and that Moab threw off the yoke and took its towns back. It establishes a king of Moab, a king of Israel called Omri, and Israelite control of Moabite territory and its loss. It does not confirm 2 Kings 3, which tells of the same revolt from the other side and ends inconclusively after a sacrifice on the city wall. Mesha writes that Israel perished forever, which it did not. Two victors’ accounts of one war is what we have.
The Black Obelisk, from the Assyrian city of Nimrud, is a four-sided stone carved in registers with captions. One register shows a man bowed to the ground before Shalmaneser III, and the caption names Jehu, calling him of the house of Omri. It establishes an Israelite king of that name paying tribute to Assyria in the ninth century; it does not confirm the account of his coup, and whether the prostrate figure is the king or an envoy is disputed. One detail is worth carrying away: Jehu had exterminated Omri’s line, and Assyrian scribes called him Omri’s son regardless. The label was shorthand for a kingdom, not a claim about parentage — the same caution to bring to the phrase “house of David”.
From there the coverage thickens. Assyrian inscriptions name Menahem among tribute-payers and record the installation of Hoshea in place of Pekah; the king the Bible calls Pul is the same man it elsewhere calls Tiglath-pileser. On the fall of Samaria the sources disagree: 2 Kings has Shalmaneser besieging the city three years, while Sargon II claims its capture in his own inscriptions. Whether Sargon finished a siege his predecessor began or annexed the credit afterwards is unresolved.
Sennacherib’s campaign against Hezekiah is the best-documented episode in the book. The Assyrian annals, on inscribed clay prisms, record the taking of many fortified towns of Judah, heavy tribute, and Hezekiah shut up in his city like a bird in a cage; the biblical account records the tribute too, and then an angel destroying the Assyrian camp. Both agree Jerusalem was not taken, and neither explains the withdrawal in terms the other would accept. Reliefs from Sennacherib’s palace at Nineveh show the siege of Lachish, and the ramp they depict is still on the ground there.
Hezekiah’s tunnel is often overstated, so state it precisely. The tunnel exists, runs from the spring outside the walls to a pool within, and can be walked today. An inscription cut into its wall describes the two teams of diggers hearing voices through the rock and breaking through to each other. The letter forms belong to the late eighth century. The inscription names no king. That the tunnel is Hezekiah’s rests on the date of the script together with 2 Kings 20:20 — a reasonable inference, and an inference, not the stone’s own testimony. See what external attestation means.
Babylonian chronicle tablets date Nebuchadnezzar’s capture of Jerusalem and the removal of its king to a particular month of a particular year, with a precision the biblical account does not attempt. The chronicle series then breaks, so the destruction of 586 is covered by no surviving tablet. From the Judaean side there are the letters written in ink on potsherds in the burnt ruins of Lachish, discussed under Jeremiah.
And then the last four verses. Jehoiachin is brought out of prison, given a change of clothes, a place at the Babylonian king’s table and a daily allowance for life. Administrative tablets excavated in Babylon list rations of oil and barley issued to the king of Judah and his sons. The book’s last quiet detail is the one an archive happens to corroborate.
Where to start
Read chapter 17, then 18 and 19, then 25: the fall of the north, the siege that failed, and the fall of the south. Then read 25:27–30 slowly. Anything we have had to revise is recorded at corrections.