The claim
In 701 BCE Sennacherib, king of Assyria, campaigned in Judah, took its fortified cities, and besieged Hezekiah in Jerusalem. The biblical account, in 2 Kings 18–19 and again in Isaiah 36–37, says the siege was lifted without the city being taken and the Assyrian army withdrew after a catastrophe in its camp. Sennacherib’s own inscriptions describe the same campaign, name Hezekiah, describe confining him in Jerusalem, list the tribute extracted — and do not record taking the city.
This is the rarest situation in the field: two substantial accounts of a single campaign, written by the two sides, both surviving, disagreeing about the outcome, and overlapping enough that the disagreement can be read rather than guessed at.
It is not, however, a case about prophecy, and the dossier does not treat it as one. The text under examination was written after the event it describes. What is being tested here is attestation: does the external record support the biblical account of what happened, and where does it diverge?
The dating problem
The Assyrian text
The prism inscriptions were produced as foundation records within a decade or so of the campaign, in Sennacherib’s own reign. They are close to the event in time, and they are not neutral: royal inscriptions are a genre with known conventions, composed to be buried in a wall and to present the reign in a particular light.
The biblical text
2 Kings belongs to the Deuteronomistic History, and its final form has a firm floor that comes from the book itself: 2 Kings ends with the release of Jehoiachin from prison in Babylon, an event of about 561 BCE. The book in the form we have it is therefore no earlier than the mid-sixth century — some 140 years after the campaign.
That does not make the account late in every part. There is a long-standing argument that 2 Kings 18:13–16, which records submission and tribute in flat annalistic prose with no theological framing, derives from an earlier Judahite source much closer to the event. If that is right, the biblical record preserves two accounts of the same campaign side by side, and one of them reads very much like Sennacherib’s.
Because the text under examination is later than the event, nothing here depends on a claim of prediction, and the circularity that dominates the dating of prophetic books does not arise.
What the external record says
Sennacherib’s account of the third campaign survives on several clay prisms. The best known is the Taylor Prism, BM 91032 in the British Museum, one exemplar of the inscription edited as no. 22 in the Royal Inscriptions of the Neo-Assyrian Period series. The Judaean passage states that Sennacherib besieged and captured forty-six of Hezekiah’s fortified cities, took a large number of people and livestock as spoil, and shut Hezekiah up in Jerusalem, his royal city, like a bird in a cage. It then lists the tribute sent after him to Nineveh, including thirty talents of gold and eight hundred talents of silver.
The second source is the sculptured relief sequence from Sennacherib’s palace at Nineveh showing the siege and capture of Lachish. The panel carrying the royal epigraph that names Sennacherib and Lachish is BM 124911. Lachish itself has been excavated, and the Assyrian siege ramp and the destruction layer are in the ground — which makes this one of the very few episodes where a royal inscription, a monumental relief and an excavated site can be laid beside a biblical narrative.
Where the accounts agree
- That the campaign happened, in the reign of Hezekiah, and that Sennacherib conducted it in person.
- That fortified cities of Judah were taken and their populations deported.
- That Lachish was central to the campaign — the reliefs make it the set piece, and 2 Kings 18:14 has Hezekiah send his submission to the king of Assyria at Lachish.
- That Jerusalem was invested but is not claimed as captured by either side.
- That Hezekiah paid a substantial tribute in gold and silver.
Where they diverge, and what the divergence is worth
The gold matches; the silver does not. 2 Kings 18:14 gives three hundred talents of silver and thirty talents of gold. The prism gives eight hundred talents of silver and thirty talents of gold. The gold figure is identical in two independent records, written in different languages by the two sides of the same war. The silver figure differs by a factor of nearly three.
Both halves of that are informative, and the second half is usually left out. The agreement on gold is hard to explain except by both accounts drawing on a real transaction. The disagreement on silver is a genuine discrepancy, and the explanations offered for it — different talent standards, tribute paid in instalments, Assyrian inflation of the figure, later biblical adjustment — are all plausible and none is established. This dossier does not pick one.
The outcome is framed incompatibly. The prism ends with tribute and a subdued vassal. 2 Kings 19:35–36 ends with the Assyrian army struck down and Sennacherib withdrawing to Nineveh. These are not the same story, and no reading makes them one.
Where this dossier could go wrong
The first temptation is to treat the Assyrian silence about capturing Jerusalem as a confession. It is not. Royal inscriptions do not narrate failures, and a campaign that ends in tribute rather than conquest is reported as tribute, which is a success in that genre. Sennacherib’s text concedes nothing by stopping where it stops; it is doing what such texts always do. Reading it as an admission imports a modern expectation of full disclosure into a document designed to be buried in a wall.
The second temptation runs the other way: to treat the biblical account of the army’s destruction as a pious explanation invented to cover a defeat. That is possible, and serious historians hold it. But it is not established by the prism either, and the classical tradition preserves a separate story of an Assyrian withdrawal caused by a disaster in the camp, which means the biblical version is not the only ancient account of an abnormal ending.
Third, the chronology of Hezekiah’s reign is disputed, and a minority of scholars have argued for two Assyrian campaigns rather than one. A dossier that does not mention that is presenting a settled picture where the field has an unsettled one.
Fourth, and most important for how this page should be read: the prism is one exemplar of an inscription that survives in several copies, and the copies differ. The source record attached here is pinned to a specific object by catalogue number precisely so that a reader can check which exemplar is being quoted rather than taking “the Sennacherib prism” on trust.
The verdict
5 · Contradicted or partial. This grade is not a judgement that the biblical account is unreliable. On the contrary — measured as history, 2 Kings 18–19 comes out of this comparison unusually well: the king, the campaign, the fall of the fortified cities, the centrality of Lachish, the investment of Jerusalem, the tribute and the gold figure all stand up against a hostile independent record.
The grade attaches to the traditional reading, which is the thing this archive examines. That reading takes 701 BCE as a story of deliverance in which Jerusalem is spared and Assyria is broken. The external record shows a campaign that devastated Judah, stripped it of its fortified cities, deported its population and ended with its king paying tribute — and the biblical text itself records that submission, in 2 Kings 18:14–16, in its own words. The divergence is not only between the Bible and the archive. It is inside the Bible, between what the text says and what the text is usually remembered as saying.
Grade 5 on this scale covers exactly that case: the external record diverges from the traditional reading, including where the biblical text itself records the divergence. It is the right grade here, and it is worth noticing that it lands on the case where the evidence is strongest. A scale that only awards high grades when the evidence is good and low grades when it is thin is not measuring anything.
Related: the book overview of Isaiah, whose chapters 36–37 carry the parallel account, and the method note on what counts as attestation.