Why Did Sennacherib Fail to Take Jerusalem?

"As for him (Hezekiah), I confined him inside the city Jerusalem, his royal city, like a bird in a cage." — Sennacherib, Rassam Cylinder, 700 BC¹

Summary

In 701 BC Sennacherib of Assyria invaded Judah, took its fortified towns, and shut King Hezekiah up in Jerusalem.  He did not take the city, and Hezekiah stayed on his throne.  The Bible says the angel of the Lord killed 185,000 Assyrians in one night and Sennacherib went home, where his sons later killed him.  Much of this can be tested, and it tests well.  The cause of the withdrawal cannot be tested at all.

1. The enemy's own records confirm the frame of the story (Part 1 and Part 2). Sennacherib's Rassam Cylinder, dated spring 700 BC, a year after the campaign, names "Hezekiah of the land Judah," claims 46 of his walled cities and 200,150 people, and says Hezekiah was confined in Jerusalem "like a bird in a cage."¹ It never claims the city fell. The tribute matches 2 Kings 18:14 on the gold, 30 talents, and differs on the silver. The siege of Lachish, where 2 Kings puts Sennacherib's headquarters, is carved on his palace walls and survives in the ground as a siege ramp with a Judean counter-ramp behind the wall.
2. The biblical account has three layers, and the first is the one the annals match (Part 3). A short record of surrender and tribute (2 Kings 18:13-16), the speeches of the Assyrian envoy, and the prophetic story of the night of the angel. The mention of "Tirhakah king of Cush" in 19:9 is not an error that forces a second campaign: Taharqa was a prince of about 20 in 701 on Kenneth Kitchen's chronology, and the title is the one he held when Kings was written. The case for a second campaign around 688 rests on an older chronology and on an Egyptian stela that names no enemy.
3. Hezekiah's survival was unusual, and this can be counted (Part 4). Of ten resisting rulers or ruling groups whose fates Sennacherib's own annals record, Hezekiah is the only one who kept his throne in his own city. The others fled, were deported, were killed or, in one case, were flayed.
4. No explanation of the withdrawal is attested (Part 5). Plague, the Egyptian army, a negotiated surrender, and trouble in Babylonia have each been proposed by serious scholars. Each fits some evidence. None is recorded by any source. A plague does not kill 185,000 men in one night: the fastest killers, pneumonic plague and cholera, take at least half a day to show symptoms. 185,000 is half again the largest Assyrian army on record.
5. The murder is confirmed, twenty years on (Part 6). The Babylonian Chronicle records that Sennacherib "was killed by his son" in 681 BC.² Esarhaddon's own inscription tells of his brothers fighting for the throne in Nineveh and fleeing abroad. The name of the elder son, Urdu-Mullissu, is very likely the biblical Adrammelech; who actually struck the blow is still argued.

Verdict.  Everything ordinary in 2 Kings 18-19 that can be checked is confirmed:  the invasion, the fall of Lachish, the blockade, the tribute, Hezekiah's survival, and the murder by a son.  The outcome was unusual by Assyria's own standards, and both sides' accounts agree that Jerusalem was not taken.  The cause of Sennacherib's retreat is beyond the evidence, and the number 185,000 is unsupported and very large.  A conservative reader keeps a real deliverance in a real war;  what has to be given up is any claim that history has confirmed the angel or the number.

Part 1: The sources and how close each is to the event

The Assyrian account was written within a year of the campaign.  The biblical account in its final form is about 140 years later but claims older sources.  Everything else is later still and depends on one of those two.

The campaign of 701 BC is Sennacherib's "third campaign" in his own numbering.  His scribes rewrote his annals each time a new campaign was added, so the same story survives in several editions.  The earliest is the Rassam Cylinder.  Its colophon dates it to "Ayyāru (II), eponymy of Mitūnu, governor of the city Isāna (700)," that is, spring 700 BC.¹  Mordechai Cogan of the Hebrew University notes that this date, compared with earlier prisms that report only two campaigns, is what fixes the Judah campaign to 701.³  The best-known edition, the Chicago Prism, carries copies dated 691 and 689 BC.⁴

The dates in the table come from the editions cited below.  The Bull inscription date is Cogan's.³  The date of Berossus's Babyloniaca is the one Russell Gmirkin uses.⁵  The Chronicle copy is dated "in the twenty-second year of Darius," which is 500 or 499 BC.²  The date of Kings comes from its last event, the release of Jehoiachin "in the first year" of Evil-Merodach, 562/561 BC, and Kings cites "the scroll called the Annals of the Kings of Judah" for Hezekiah's reign.⁶⁶  An earlier edition of Kings under Josiah, or of the Isaiah narratives under Hezekiah's successors, is widely proposed and would shorten the gap, but it is an inference from the text, not a dated manuscript.

Two features of the sources matter for everything that follows.

1. The Assyrian texts are royal propaganda. They report victories and do not report setbacks. So what they leave out tells little, and what they admit against their own interest tells a great deal.
2. Every later source except the Babylonian Chronicle depends on one of the two early ones or on a separate oral tradition. Josephus retells Kings, adds Herodotus, and changes the place of the siege to Pelusium "from the text of Herodotus," as his Loeb editor Ralph Marcus notes.⁷ Herodotus is an Egyptian temple story. Berossus survives only in excerpts of excerpts. So the case is really Assyria against Judah, with Herodotus and the Chronicle as outside witnesses.

Part 2: What the Assyrian records and the ground confirm

The invasion, the battle with Egypt, the fall of Lachish, the blockade of Jerusalem and Hezekiah's payment are confirmed by the side that had every reason to tell it differently.

The campaign in Sennacherib's words.  The Rassam Cylinder runs through the campaign in order.  Luli of Sidon fled overseas.  Sidqa of Ashkelon was deported.  The people of Ekron had put their pro-Assyrian king Padi in chains "and handed him over to Hezekiah of the land Judah in a hostile manner."  They called on "the kings of Egypt (and) the archers, chariots, (and) horses of the king of the land Meluḫḫa," that is, Kush.¹  The armies met near Eltekeh:  "With the support of (the god) Aššur, my lord, I fought with them and defeated them.  In the thick of battle, I captured alive the Egyptian charioteers (and) princes."¹  Cogan places the battle at Tell esh-Shallaf, "about 3 km west of the city of Rehoboth."³  Sennacherib then took Ekron, "killed the governors (and) nobles who had committed crime(s) and hung their corpses on towers around the city," and "brought out Padî, their king, from the city Jerusalem."¹

Then the passage on Judah:

(As for) Hezekiah of the land Judah, I surrounded (and) conquered forty-six of his fortified walled cities and small(er) settlements in their environs, which were without number ... I brought out of them 200,150 people, young (and) old, male and female ... As for him (Hezekiah), I confined him inside the city Jerusalem, his royal city, like a bird in a cage.  I set up blockades against him and made him dread exiting his city gate.  I detached from his land the cities of his that I had plundered and I gave (them) to Mitinti, the king of the city Ashdod, and Padî, the king of the city Ekron, (and) Ṣilli-Bēl, the king of the land Gaza, (and thereby) made his land smaller.¹

Three details in this account are worth noticing.

1. The numbers are scribal, not fixed. The editors record that most copies of the Rassam Cylinder read 46 cities, but "ex. 1 has ... 'thirty-six of his cities' and ex. 2 has ... 'fifty-six of his cities.'"¹ The 200,150 people is a booty count in a genre that likes large numbers: the same annals claim 208,000 people carried off on the first campaign.⁴ Taken as people counted rather than people deported, it is not impossible for a kingdom of Judah's size, but nothing checks it.
2. "Like a bird in a cage" is a stock phrase. The RINAP editors note that the simile "is used for the first time in an Assyrian context by Tiglath-pileser III," and that it marks a king "confined to his city" and also "the cowardly nature of that ruler."¹ It describes a blockade, not a capture. No edition of the annals, from 700 BC on, claims Jerusalem fell.
3. The tribute came later and far away. "After my (departure)," Hezekiah sent to Nineveh "30 talents of gold, 800 talents of silver," luxury goods, "chariots, shields, lances, armor ... bows and uṣṣu-arrows," his daughters, his palace women, and the "elite troops whom he had brought inside to strengthen the city Jerusalem."¹ 2 Kings 18:14 has "300 talents of silver and 30 talents of gold," and puts the demand while Sennacherib "was at Lachish."⁸ The gold agrees exactly. Cogan converts the Assyrian figures with the light talent of 30 kg: "Hezekiah's payment was 900 kg of gold and 24,000 kg of silver."³ The silver disagrees by a factor of 2.7. Neither text explains why; different talents, or different parts of one payment, are both possible and neither is proven.

Lachish on the palace walls.  Sennacherib gave the siege of Lachish a whole room of his South-West Palace at Nineveh.  The panels are now in the British Museum.  Over the scene of the king on his throne, the caption reads:  "Sennacherib, king of the world, king of Assyria, sat in (his) armchair and the booty of the city Lachish passed before him."⁹  The RINAP editors place it on BM 124911, Room XXXVI, slabs 12-13.  Austen Henry Layard, who excavated the room, described "as many as ten banks or mounts" thrown up against the walls and battering rams rolled up them.¹⁰  No relief shows Jerusalem taken.  Christoph Uehlinger has suggested that one throne-room slab may show Jerusalem with Hezekiah on its walls, which would fit a city blockaded but not stormed.¹  Lachish, not Jerusalem, was the victory Sennacherib put on display.

Lachish in the ground.  The city of Level III at Tell ed-Duweir was burned.  Its date was once argued.  James Starkey, who led the British dig of the 1930s, "tentatively assigned" the burning to Nebuchadnezzar's campaign of 597 BC.¹¹  Olga Tufnell, who published the Iron Age results in 1953, moved it to Sennacherib:  the pottery and the "royal" jar stamps "weight the evidence in favour of Sennacherib's campaign rather than that of Nebuchadrezzar more than a century later."¹¹  David Ussishkin's Tel Aviv University excavations of 1973-1994 settled it.  Oded Lipschits, who disputes Ussishkin on other points, writes that the excavations "put an end to this debate.  Ussishkin established that Lachish Level III was indeed devastated by Sennacherib in 701 B.C.E."¹²

Ussishkin's team excavated the Assyrian siege ramp and found something the reliefs do not show:  "opposite the Assyrian siege ramp, an unsuspected Judean counter-ramp intended to bolster the city's defenses against the Assyrian onslaught."¹³  On a 2002 site visit, Ussishkin's figures as reported by Biblical Archaeology Review were these:  the ramp's remnants "rise 40 feet," the arrowheads found in that corner of the mound number "more than 800," and the site yielded "430 l'melekh handles, more than any other site in ancient Israel."¹⁴  These are the jars many scholars connect with Hezekiah's preparations for the siege.

A figure to use with care.  Tufnell reported pits holding "about 1,500 skulls," and in one chapter connected them with "those who did not long survive the siege."¹¹  Her own chapter on the ossuary deposit concludes that it "was made during the seventh century," after 701.¹¹  The 1,500 should not be quoted as the Assyrian death toll at Lachish until that conflict is settled from the excavation records.

What this confirms.  2 Kings 18:13 says Sennacherib "marched up against all the fortified cities of Judah and captured them."⁸  18:14 puts him at Lachish.  18:17 has him send officers from Lachish to Jerusalem.  All three are matched by Assyrian text or by the ground.  The siege of Jerusalem is also matched in a negative way:  the Assyrians say they blockaded it, the Bible says the envoys stood at "the conduit of the upper pool,"⁸ and neither says the city was stormed.

Part 3: The biblical account, Tirhakah, and the "two campaigns"

The biblical account has a short record of surrender and a longer prophetic story.  The record agrees with the annals.  The story names Tirhakah, which has led some historians to split the account into two campaigns;  the better evidence keeps it as one.

Three layers.  Since Bernhard Stade in 1886, critics have divided the narrative.  Mordechai Cogan and Hayim Tadmor, in the Anchor Bible commentary on 2 Kings, put it this way:  "The extensive account of Sennacherib's campaign to Judah divides into two discrete sources — A: 18:13b–16, a chronistic record;  and B: 18:17-19:37, prophetic narratives," a division "confirmed by most critics ever since."¹⁵  The prophetic part is usually split again into two parallel accounts.  For a reader the three layers are easy to see.

1. The tribute (2 Kings 18:13-16). Hezekiah sends to Sennacherib at Lachish: "I have violated our treaty. If you leave, I will do whatever you demand." He pays 300 talents of silver and 30 of gold, stripping "the metal overlays from the doors of the Lord's temple."⁸ This is the layer the annals match. It is also the layer least flattering to Hezekiah, which is a point in favour of its having come from royal records.
2. The envoy's speeches (18:17-19:9a). Three Assyrian officers come from Lachish "along with a large army."⁸ The chief adviser, the Rabshakeh, mocks reliance on Egypt, "that splintered reed staff," and offers terms that are exactly Assyrian policy: surrender, then "until I come and take you to a land just like your own."⁸ The annals' treatment of Ashkelon shows the threat of deportation was real. Kitchen reads the "splintered reed" as a sneer after Egypt's defeat at Eltekeh.¹⁶ Isaiah answers with a prophecy: "he will receive a report and return to his own land. I will cut him down with a sword in his own land."⁸ Then Sennacherib "heard that King Tirhakah of Ethiopia was marching out to fight him."⁸
3. The letter, the prayer and the angel (19:9b-37). Sennacherib writes to Hezekiah; Hezekiah spreads the letter before the Lord; Isaiah prophesies that the king "will not enter this city"; and "That very night the angel of the Lord went out and killed 185,000 in the Assyrian camp. When they got up early the next morning, there were all the corpses."⁸ Sennacherib goes home to Nineveh, and the chapter ends with his murder.

Isaiah 36-37 repeats layers two and three nearly word for word and omits the tribute.  2 Chronicles 32 abridges the story and says the angel "wiped out all the soldiers, princes, and officers," so that "Sennacherib returned home humiliated."⁸  So Kings is the base text, and the tribute record is the only part the Assyrians could confirm or deny.  They confirm it.

The Tirhakah problem.  Taharqa, the Kushite pharaoh whom the Bible calls Tirhakah, reigned from 690 to 664 BC.  In 701 he was not king.  So 2 Kings 19:9 calls him by a title he did not yet hold.  There are three answers, and each should be stated in its defenders' words.

1. Kitchen: a prince of about twenty, with a later writer's title. Kenneth Kitchen of Liverpool, author of the standard history of Egypt's Third Intermediate Period, reads the Kawa stelae, Taharqa's own inscriptions from Nubia: "Taharqa was a youth in Nubia when, aged 20, he was summoned north to Thebes with other royal brothers by the king, Shebitku – and they came north with an army."¹⁶ On Kitchen's chronology that summons fell about 702 BC. So: "Aged 20 or 21, Taharqa could well have been titular head of the expedition, with generals to advise him. One may note that the Assyrian accounts never state that the pharaoh came in person."¹⁶ The title is the narrator's, not the event's: "Taharqa was not king in 701 B.C., but he certainly was during 690-664 B.C., for a quarter of a century."¹⁶ His analogy: "Just as today one may say 'Queen Elisabeth II was born in 1926'— yet, in that year, she was not Queen." Taharqa does the same on his own Kawa stela IV, calling himself "His Majesty" for a time when he was only a prince.¹⁶ Kitchen adds that an inscription of Sargon II at Tang-i Var "shows clearly now that Shebitku was a power in Egypt from at least 706," so Shebitku could have sent Taharqa in 701.¹⁷
2. Macadam and Bright: a child of nine, so the verse belongs to a later campaign. M. F. L. Macadam's 1949 edition of the Kawa stelae, as Cogan and Tadmor summarize it, "suggested that Tirhakah was born in 710 and that at age twenty he was summoned to his brother Shebitku's side to serve as coregent for six years."¹⁵ John Bright drew the conclusion in his History of Israel: the stelae "tell us that Tirhakah was but twenty years old when he first came from Nubia to Lower Egypt to be associated with his brother. This means, on the dates followed here, that he was only nine in 701 (on no conceivable dating could he have been an adult at the time)."¹⁸ Bright therefore placed 2 Kings 18:17-19:37 in a second campaign about 688 BC. William Shea argued the same in the Journal of Biblical Literature in 1985 and, in Andrews University Seminary Studies in 1997, tied it to a new Taharqa stela from Karnak: "This cannot be the campaign of 701 because Tirhakah was not on the throne at that time and could not have been in Lower Egypt that early, according to his own stelae at Kawa."¹⁹ The window, he wrote, would be "the brief period of 688 to 686."¹⁹
3. An anachronism or a guess. Some critics take the name as the only pharaoh the writer knew. Henry Aubin reports Donald Redford's view: "Quite simply, Taharqa's was the only Pharaonic name known to the author of 2 Kings 19."²⁰

Testing them.  The second answer depends entirely on Macadam's chronology, and that chronology has not held.  Kitchen calls it "ingenious but erroneous," a scheme that made "Taharqa 9 years old in 701 B.C."¹⁶  Cogan and Tadmor, who are not defending the Bible's account of the angel, reached the same result:  after "a thorough reinvestigation by K. A. Kitchen of the chronology of the Nubian dynasty," Tirhakah "had already been of age in 701;  thus he could have been at the head of the Egyptian forces which engaged Sennacherib at Eltekeh."  They concluded:  "No disagreement exists, therefore, between 2 Kgs 19, which mentions Tirhakah, and the Rassam cylinder."¹⁵  They still call the title "admittedly anachronistic,"¹⁵ which is Kitchen's point put the other way:  the title belongs to the writer's day.

The second campaign has no witness.  Kitchen:  "no respectable evidence whatever has yet been produced from either Assyria or Egypt in favour of a second Palestinian campaign by Sennacherib."¹⁶  Cogan and Tadmor give the historical reason:  "the Assyrian monarch was not obliged to undertake a second campaign to the West after the impressive victory of 701 ... It is hardly any wonder that Sennacherib did not have to campaign for close to a decade until his death.  The world lay at Assyria's feet."¹⁵  Shea's Karnak stela does not fill the gap.  He himself reports that "there is no enemy or target city mentioned by name in the surviving portions of the text."¹⁹  A campaign is not placed in Judah by an Egyptian text that names no enemy.

The cost to each side.  The one-campaign reading costs the conservative reader one thing:  "king" in 19:9 is the narrator's title, not Taharqa's rank in 701.  That is no more an error than calling a prince by the title he later held.  The two-campaign reading costs the critic more:  it posits a war that neither Assyria nor Egypt records, to rescue a chronology that Kitchen, and Cogan and Tadmor after him, reject.

Part 4: Why did Hezekiah keep his throne?

Hezekiah led the revolt, held a pro-Assyrian king prisoner, and held out until the Assyrian army left.  Every other ruler in Sennacherib's annals who did that lost his throne, his freedom or his life.  Hezekiah lost territory and treasure and kept his crown.

The annals let this be counted.  The table lists every ruler whose resistance and fate the Chicago Prism, the Rassam Cylinder or the Nineveh prism of 694 BC record, from the first campaign to the sixth.

Sources for the table:  the Chicago Prism,⁴ the Rassam Cylinder,¹ and the Nineveh prism of 694 BC.²¹

The count.  Of ten rulers or ruling groups who resisted, nine fled, were deported, were killed or were flayed.  One kept his throne in his own city:  Hezekiah.  Rulers who submitted before a fight did keep their thrones.  The annals list eight "kings of the land Amurru," from Sidon to Edom, who "brought extensive gifts, four times (the normal amount), as their substantial audience gift before me and kissed my feet."¹  Hezekiah did not do that.  He paid "after my (departure)," and "he sent a mounted messenger of his to me to deliver (this) payment and to do obeisance."¹  In the annals' own terms, he submitted by messenger, after the king had gone.

What the scholars who edited these texts concluded.  Daniel Luckenbill, who published the Chicago Prism in 1924, read the length of the Jerusalem passage as "a sure sign that the victory claimed was not at all decisive."²²  He added:  "That his campaign was not a brilliant success we may legitimately infer from Sennacherib's own account of it in the annals."²²  Cogan, commenting on the later Bull inscription's language about Hezekiah, writes that it hints "at the outcome of the campaign in 701 BCE that was not in conformity with Assyrian military or ideological standards.  Rather than dethroning the king and dismantling his kingdom, Hezekiah was forced into submission and his kingdom destroyed."³

The other side of the ledger.  Cogan and Tadmor also call 701 "the impressive victory," and that is fair.¹⁵  Judah lost dozens of towns to three Philistine kings, paid 30 talents of gold, and sent its elite troops and its weapons to Nineveh.  Assyria had reasons to stop short of a siege of Jerusalem:  a siege was slow and costly, and a reduced, disarmed Judah paying tribute was a buffer against Egypt.  On this reading Sennacherib got what he came for and chose not to finish the job.  It explains why he could leave.  It does not explain why Hezekiah alone of the resisting rulers was spared the fate of Sidqa, or why the king who had held an Assyrian vassal in chains was allowed to submit by messenger after the army had gone.

So the unusual fact is this, and it is the strongest independent support for the Bible's picture.  Assyria's own records show that Sennacherib left Judah with Jerusalem untaken and its rebel king enthroned, against the practice those same records describe.  They do not say why.

Part 5: The explanations, tested

Five explanations have been offered for Sennacherib's withdrawal.  Four are natural and one is the Bible's.  Each is stated here by its defenders and then priced.  No source records any of them as the cause.

Plague in the camp

The claim.  Josephus was the first to turn the angel into disease, writing that "God had visited a pestilential sickness upon his army."⁷  John Bright judged that "the crippling of the Assyrian army by an epidemic is, in itself, in no wise improbable," and suggested "Perhaps the bubonic plague is in question."¹⁸  The historian of disease William McNeill counted it among the ancient epidemics of the Near East:  "the fatal visitation that 'slew in the camp of the Assyrians one hundred and eighty-five thousand' overnight, and caused the Assyrian king, Sennacherib, to withdraw from Judah without capturing Jerusalem."²³  Kitchen thinks "an epidemic or similar misfortune probably dictated an early return to Assyria and to Nineveh."¹⁶  Wolfram von Soden, as Cogan and Tadmor report him, noted "that plagues are reported in Assyrian sources as sometimes upsetting military campaigns," but "would not accept so large a number."¹⁵

The support.  Herodotus's mice.  In the Egyptian priests' story, field mice "ate up their quivers and their bows," the Assyrians fled, and "great numbers fell."²⁴  Mice can carry plague, and readers since Josephus have joined the two stories.  Cogan and Tadmor answer that the mice belong to Greek symbolism, not to the camp:  "mice were a symbol of the plague in Greek folklore and the god Apollo Smintheus (lit. 'mouse') was depicted with a mouse in his hand," and joining the stories ignores "the distance which separates Jerusalem and Pelusium."¹⁵

The price.  The text says "That very night," and "When they got up early the next morning, there were all the corpses."⁸  Disease does not work on that schedule.

1. Plague. The World Health Organization gives "an incubation period of one to seven days," and "Incubation can be as short as 24 hours."²⁵ The fastest form, pneumonic plague, "can be fatal within 18 to 24 hours of disease onset if left untreated."²⁵ So the shortest path from infection to death is about 42 hours, and most cases take longer.
2. Cholera, the fastest killer of a camp. WHO: "Symptoms appear 12 hours to 5 days after infection," and the diarrhoea "can be fatal within hours if untreated." But "Most people with cholera have no or mild symptoms."²⁶ The shortest path is about half a day, and deaths from one contaminated water supply would be spread over the four and a half days between the earliest and latest onsets.
3. The arithmetic. An epidemic can kill an army. It cannot kill one in a single night. For deaths to come all together, the onsets would have to fall within a few hours of each other, when the WHO ranges spread them over days. And because most cholera cases are mild and most bubonic plague cases survive, 185,000 dead would require several times that number infected.²⁷

So the plague explanation keeps the event and gives up the text's timing.  It is a natural account of an army struck down over days, which the narrator then told as one night.  That may be what happened.  But it is a guess about a cause no source names, and it does not remove the difficulty of the number.

The Egyptian army

The claim.  Henry Aubin argued in The Rescue of Jerusalem (2002) that the Kushite army saved the city:  "I believe my thesis—that Kushite Egypt turned back Sennacherib—is unshakable."²⁰  He drew large conclusions from it:  "The Kushites' intervention enabled the fragile, war-torn Hebrew kingdom to survive."²⁰  Kitchen gives a more modest military version.  After Lachish, Sennacherib's forces were split between Libnah and Jerusalem, and "rumour had reached him that Taharqa and the Egyptian army were aiming to attack him."¹⁶  Both rest on 2 Kings 19:7 and 19:9:  "he will receive a report and return to his own land," and "The king heard that King Tirhakah of Ethiopia was marching out to fight him."⁸⁸

The price.  Sennacherib says he beat the Egyptians and Kushites at Eltekeh and took their princes alive.¹  Royal annals do not admit defeats, but they rarely invent a named battle with named captives either.  Bright, who held a different reconstruction, pointed out that the battle at Eltekeh came early, "before Ekron, Eltekeh, and Timnah were taken."¹⁸  So the Egyptian explanation needs a second Egyptian army, or a regrouped first one, after Lachish.  Kitchen supposes exactly that:  the Egyptians "had possibly regrouped their own forces at a safe distance and now saw their chance to strike back at Sennacherib while his army was thus dispersed."¹⁶  That fits the Bible's "report" and the Assyrian silence.  No Egyptian or Assyrian text records it.

A negotiated surrender

The claim.  On this reading there was no crisis.  Hezekiah submitted, paid, and Sennacherib went home with his aims met.  Cogan and Tadmor speak of "the impressive victory of 701."¹⁵  Josephus told a version of this:  Sennacherib "on receiving three hundred talents of silver and thirty of gold, agreed to withdraw amicably," then broke his word, a sworn pledge that Ralph Marcus calls "a detail added by Josephus."⁷  The most detailed modern statement, Walter Mayer's "Sennacherib's Campaign of 701 BCE:  The Assyrian View" in Lester Grabbe's volume Like a Bird in a Cage (2003), could not be read for this article, so this position is stated here from the scholars who share it rather than from Mayer.

The price.  It fits the tribute and the annals perfectly.  It fits the Bible's first layer, 2 Kings 18:13-16, which has Hezekiah surrender and pay at Lachish.  What it does not fit is Part 4:  Hezekiah submitted by messenger after the Assyrian army had left, which is not how the annals describe a surrender before the king, and he kept a throne no other resisting ruler kept.

Trouble in Babylonia

The claim.  Bright put it beside plague:  "Two explanations are suggested:  that Sennacherib's army was crippled by an epidemic (II Kings 19:35), and that news came that his presence was required at home (v. 7).  The two are not mutually exclusive and both are plausible."¹⁸

The support.  The next year Sennacherib did go to Babylonia.  His fourth campaign defeated Mushezib-Marduk at Bittutu and installed his own son on the Babylonian throne.⁴  Grayson's summary of the Babylonian Chronicle says "Bel-ibni and his officers, apparently guilty of treason, were led away to Assyria."²  The Rassam Cylinder, dated spring 700 BC, does not yet mention that campaign, so the Babylonian trouble was building as Sennacherib returned from the west.

The price.  It costs nothing in the evidence, and it explains why a king might leave in a hurry.  It does not explain why he left Hezekiah enthroned rather than leaving a governor or a garrison, and no source connects the two.

The angel of the Lord

The claim.  2 Kings 19:35, Isaiah 37:36 and 2 Chronicles 32:21 say the Lord's messenger struck the camp.  The Bible gives no physical cause.

What can be tested.  The miracle itself cannot be tested by any historical method.  Its two checkable parts are the outcome and the number.  The outcome, Jerusalem untaken and Hezekiah enthroned, is confirmed by the Assyrians.  The number is not, and it is very large.

The 185,000, priced.  The largest army an Assyrian king claims to have mustered is Shalmaneser III's 120,000, in his fourteenth year:  "I mustered (the troops of my) extensive land in countless numbers (and) crossed the Euphrates in flood with 120,000 troops."²⁸  185,000 dead in one camp is 1.54 times that muster.  Sennacherib's only troop figure in his account of 701 is the 20,000 men he took from the booty:  "I conscripted 10,000 archers (and) 10,000 shield bearers and added (them) to my royal contingent."¹  And 2 Kings 18:17 sends to Jerusalem only the three officers "along with a large army," while the king himself stays at Lachish and then Libnah.⁸  If 185,000 died before Jerusalem, the detachment there was larger than any Assyrian army on record, and the main army was somewhere else.

Three proposals reduce the number, and none is established.

1. A rhetorical number. Cogan and Tadmor: "A similar use of exaggerated numbers in divine punishment can be found in 2 Sam 24:15, where 70,000 die in a three-day plague."¹⁵
2. A misread word. S. Feigen "thought that the number slain was only 185 men," taking the word for "thousand" as a misread abbreviation.¹⁵
3. "Thousand" as a unit. The Hebrew 'elep can mean a military unit as well as a thousand, a reading some scholars apply to the census numbers in Numbers. Applied here it would make the dead a number of units, not 185,000 men. It has not been worked out for this verse.

The conservative reader who takes the number literally has to accept that it is larger than any army Assyria ever claimed to field, and that it has no outside support.  The reader who takes it as rhetoric or a unit count keeps the event and gives up the figure.

Part 6: The murder of Sennacherib

2 Kings 19:37 says two of Sennacherib's sons killed him in a temple, fled to Ararat, and Esarhaddon became king.  Babylonian, Assyrian and Greek sources confirm a son as the killer, a fight among the brothers, flight abroad and Esarhaddon's succession.  The name of the elder son very likely matches the biblical Adrammelech.

The Bible's version.  "One day, as he was worshiping in the temple of his god Nisroch, his sons Adrammelech and Sharezer struck him down with the sword.  They escaped to the land of Ararat;  his son Esarhaddon replaced him as king."⁸  The text runs straight on from the retreat of 701 to this "one day," which was twenty years later.  Isaiah's oracle earlier in the story had said the king would be cut down "with a sword in his own land."⁸

The Babylonian Chronicle.  "On the twentieth day of the month Tebet Sennacherib, king of Assyria, ... was killed by his son (lit. his son killed him) in a rebellion."²  The Chronicle adds that "The rebellion continued in Assyria" until the second day of Adar, about six weeks, and that "On the twenty-eighth/eighteenth day of the month Adar Esarhaddon, his son, ascended the throne in Assyria."²  The year is 681 BC.  Eckart Frahm's summary, as Andrew Knapp quotes it:  "On 20th Ṭebetu (x.) 681 Sennacherib, as a consequence of his stubbornness, was killed by Urdu-Mullissu and his accomplices."²⁹

Esarhaddon's own account.  The new king never says outright that his father was murdered.  He blames his brothers:  "my brothers went out of their minds ... and they plotted evil, girt (their) weapons, and in Nineveh, without the gods, they butted each other like kids for (the right to) exercise kingship."  When he advanced, the rebels "deserted the army they relied on and fled to an unknown land."³⁰  The Bible's "land of Ararat" is Urartu, Assyria's northern rival and an obvious refuge.  Esarhaddon's "unknown land" does not name it, and does not contradict it.

The Greek tradition.  Berossus survives here only through Alexander Polyhistor, as Eusebius's Chronicle preserves him in Armenian.  Polyhistor's Sennacherib "went on to rule for 19 years, until he died as a result of a plot hatched by his own son, Ardamusanus."³¹  A second Greek writer in the same book, Abydenus, has a garbled king list in which a later king "was slain by his son Adramelus."³¹  Neither mentions Judah, Jerusalem or any disaster to the army.  Josephus has Sennacherib "treacherously attacked by his elder sons Andromachos and Seleukaros."⁷

The name.  In 1980 Simo Parpola argued from a broken letter, ABL 1091 (now SAA 18 100), that the plotter was Esarhaddon's elder brother Urdu-Mullissu, and that this name lies behind both "Adrammelech" and Berossus's "Ardumuzan."  Knapp, who argues that Esarhaddon himself may have been behind the killing, still grants the name:  "The name that Parpola ended up with can be linked without much difficulty both to the Adrammelech mentioned in 2 Kings 19:37 // Isaiah 37:38 as one of the murderers of Sennacherib, and to the Adramelos/Ardumuzan named as the murderer by Berossos.  Parpola's discussion of the name here is almost certainly correct."²⁹  In 2021 Stephanie Dalley and Luis Siddall published a newly joined fragment of the same letter and re-read it:  the letter now seems to report "a supposed plot looking to frame the king's son, Urdu-Mullissu, in a conspiracy," which "might be a product of the pro-Esarhaddon machinations of the royal court."³²  So who struck the blow is open again.  That Sennacherib's son Urdu-Mullissu existed, was accused, and carries the name the Bible gives the murderer is not.

Where the Bible's notice came from.  Cogan and Tadmor think the verse's "factual notation on Sennacherib's assassination and the identity of his murderers and their place of refuge ... could only have been derived from a Babylonian chronicle."¹⁵  They also note that the Babylonian king Nabonidus later explained the murder the same way the Bible does, as divine punishment:  Marduk "made his own son murder the king of Assyria," for Sennacherib's sack of Babylon in 689.¹⁵

The score.

For a skeptic the murder does not count as a fulfilled prophecy, because Kings reached its final form long after 681 and Cogan and Tadmor think its notice comes from a chronicle.  It counts as evidence of accurate sources.  A writer who had the killer's name, the flight abroad and the successor right was working from good information, and the same writer is the one who reports the angel.

Objections and responses

"Sennacherib's annals say he won.  The Bible invented a deliverance to cover a humiliating surrender."  The annals and the Bible's first layer agree that Hezekiah surrendered and paid.  The Bible does not hide the surrender;  it records Hezekiah stripping the temple doors to pay.⁸  What the annals cannot hide is the outcome:  no capture, no deportation, and a rebel king left on his throne, which their own pages show was not the usual end.  Luckenbill, who edited them, read them as describing a campaign that "was not a brilliant success."²²  A deliverance story invented to cover a total defeat would not need to admit the tribute.

"Assyrian annals never record defeats, so the silence about the 185,000 is exactly what the Bible's account predicts."  That is correct as far as it goes.  But silence predicted is still silence.  It removes an objection;  it does not add evidence.  The positive evidence for something unusual in 701 is the Assyrians' own failure to depose Hezekiah.  The 185,000 rests on the Bible alone.

"The Babylonian Chronicle would have recorded an Assyrian disaster, and it does not."  Luckenbill argued this:  "That Sennacherib had not met with outright defeat is evidenced, I believe, by the silence of the Babylonian Chronicle, which was not slow to record Assyrian reverses."²²  The Chronicle's entries for these years are about Babylonia.  A disaster to an Assyrian detachment in Judah would not be a Babylonian event, and the Chronicle does not record the Judean campaign at all, not even the victory at Lachish.  It passes from "The first year of Bel-ibni," when Sennacherib "ravaged Hirimma and Hararatum," to "The third year of Bel-ibni," when he "went down to Akkad," with no entry for the second year, 701.²  Its silence is evidence against a battle lost in the field, which nobody claims;  it is not evidence about the camp.

"Josephus quotes Berossus confirming the plague on Sennacherib's army."  He does not.  Josephus says Berossus mentions Sennacherib, "he writes as follows," and the next section describes the plague.  The Loeb editor Ralph Marcus judges that "It is more likely, however, that a quotation from Berosus originally followed and has been lost from Josephus's text."⁷  The plague passage is Josephus's own retelling of 2 Kings.  The Berossus that does survive, through Polyhistor, says nothing about Judah.³¹  This claim should be dropped from apologetic use.

"Herodotus confirms the miracle."  Herodotus confirms that Egyptian priests, about 270 years later, told of a night disaster to Sennacherib's army followed by a retreat.  That is a real, independent memory that the campaign ended badly for Assyria.  But it is set at Pelusium in Egypt, credits an Egyptian god and an Egyptian priest-king, "the priest of Hephaistos, whose name was Sethos,"²⁴ and uses mice, which Cogan and Tadmor tie to Greek plague symbolism.¹⁵  It supports "something went wrong," not "185,000 died at Jerusalem."

"Tirhakah was not king in 701, so 2 Kings 19 is anachronistic and unreliable."  The title is anachronistic in the narrow sense that the narrator uses Taharqa's later rank.  Cogan and Tadmor, no defenders of the angel, accept that he "could have been at the head of the Egyptian forces which engaged Sennacherib at Eltekeh."¹⁵  The error that would matter, a pharaoh who could not have been there, is answered.  See Part 3.

"The two-campaign theory solves everything:  the Bible's victory belongs to a later war."  It solves the Tirhakah title and creates a war no source records.  It rests on Macadam's chronology, which Kitchen and Cogan and Tadmor reject, and Shea's Karnak stela names no enemy.  Bright himself saw a cost in his own reconstruction:  on the single-campaign reading Hezekiah, beaten, was given "much milder terms! This seems strange."¹⁸  That strangeness is real, and it is the same anomaly Part 4 counts.  A second campaign is one way to remove it;  an unrecorded event in 701 is another.

"A plague explains everything naturally."  It explains an army losing men and going home.  It does not explain "that very night," since the fastest killers take half a day to two days to kill after infection and spread their deaths over days.  It does not explain the number, and no source names a disease.  A plague is a reasonable guess about a cause the sources do not record.

"185,000 is impossible, so the whole story is legend."  The number is a problem, priced in Part 5.  But the frame of the story is the best attested in the Old Testament:  the invasion, Lachish, the blockade, the tribute, the survival of the king, and the murder by a son are all confirmed.  An inflated number in an ancient text, if that is what it is, does not show that the event behind it did not happen.

"The prophecy that Sennacherib would die by the sword in his own land proves Isaiah was inspired."  Not to a skeptic.  Kings and Isaiah 36-37 reached their present form after 681, and Cogan and Tadmor think the murder notice was taken from a Babylonian chronicle.¹⁵  The murder confirms that the writer had good sources.  It does not, by itself, show the oracle was spoken before the event.

Issues and unanswered questions

1. Walter Mayer's chapter and Lester Grabbe's volume Like a Bird in a Cage (2003) were not read. They are the fullest statement of the "aims achieved" reading and of the critical consensus. No copy was found on archive.org or the open web. The negotiated-surrender section states that view from Cogan and Tadmor and from Josephus instead. The article should be checked against Mayer before publication.
2. Ussishkin's own excavation reports were not read. The Conquest of Lachish by Sennacherib (1982), his Tel Aviv articles of 1977 and 1990, and the five-volume final report (2004) carry the siege ramp's dimensions, the arrowhead and sling-stone counts and the burial evidence. The figures here come from his 1984 Biblical Archaeology Review note and a 2002 site visit reported by the same magazine.
3. Aidan Dodson's "The Rescue of Jerusalem: A View from the Nile Valley" (2020) is an Egyptologist's assessment of Aubin's thesis. Its PDF is behind a site challenge that could not be passed.
4. Grant Frame's 1999 edition of the Tang-i Var inscription is cited here through Kitchen. It is the evidence that Shebitku ruled by 706, and so that Taharqa could have come north before 701. It was not read.
5. Simo Parpola's 1980 paper and the full text of Dalley and Siddall (2021) were not read. Parpola is known through Knapp, and Dalley and Siddall through their abstract.
6. The Lachish ossuary. Tufnell's 1,500 skulls are tied to the siege in one chapter and dated to the seventh century in another. Until the excavation records settle it, the figure should not be used as a death toll.
7. The silver. The Bible's 300 talents and the annals' 800 differ, and no explanation has been tested here.
8. The 'elep reading of 185,000 has not been worked out, and nobody has shown what unit size it would imply or whether the reading fits this verse.
9. The date of Kings. An earlier edition under Josiah, which many scholars propose, would bring the Judean account within a century of 701. The case for it is argued in the Deuteronomy article and not repeated here.
10. The "fourteenth year." 2 Kings 18:13 dates the invasion to Hezekiah's fourteenth year, but the book's own synchronisms in 18:1 and 18:9-10 put his fourteenth year about 714/713, not 701. Proposed answers include a co-regency and a late editor's figure. The dispute does not change the Assyrian year of the campaign; it is set out in Mesopotamian Absolute Chronology.
11. No archaeological trace of the 185,000 is known. No mass grave of the period near Jerusalem has been identified in the sources read, and one would be hard to tie to 701 BC if found.

Verdict

History confirms nearly everything in 2 Kings 18-19 that it can reach, shows that the outcome was unusual by Assyria's own standards, and cannot reach the cause.

1. Confirmed. Sennacherib invaded Judah in 701 and took its walled towns. Lachish fell after a siege fought over a ramp and a counter-ramp. The Egyptians and Kushites came out and were beaten at Eltekeh. Jerusalem was blockaded and not taken. Hezekiah paid heavy tribute, with the gold the same in both accounts. Twenty years later Sennacherib was killed by a son, the rebels fled abroad, and Esarhaddon succeeded. The son accused is very likely the Bible's Adrammelech.
2. Unusual, and counted. Of ten resisting rulers or ruling groups whose fates Sennacherib's annals record, Hezekiah alone kept his throne in his own city, and he submitted by messenger after the army had gone. Luckenbill, Cogan and Bright all say the outcome was not normal. Herodotus independently remembers the campaign ending in a night disaster and flight.
3. Not attested. The cause of the withdrawal. Plague, a second Egyptian advance, a negotiated settlement and news from Babylonia each fit some of the evidence, and none is recorded. A plague does not kill in one night.
4. A problem to admit. 185,000 dead is half again the largest Assyrian army on record and has no outside support. It may be rhetoric, a unit count, or a literal claim that history cannot test.
5. What a conservative reader keeps and gives up. Keep: a real deliverance of Jerusalem in a real war, told by a writer with accurate sources, and an outcome the enemy's records confirm and cannot explain. Give up: any claim that Assyrian records, Berossus or Herodotus confirm the angel or the number, the "Berossus confirms the plague" argument, and the use of the Lachish 1,500 as a siege toll. The two-campaign theory need not be kept, because the Tirhakah problem has a better answer.

Sources