Exile, Babylon, and Persia

Summary

For most of the Old Testament, the argument is about whether the events happened at all.  From about 600 BC the argument changes character, because Babylonian and Persian administrations wrote things down on clay and clay survives.  This period is where the biblical narrative can actually be checked against contemporary foreign documents, and it is where it does best.

• A Babylonian chronicle dates the fall of Jerusalem to a specific day, and the biblical account matches. The chronicle of Nebuchadnezzar's early years records that in his seventh year he "besieged the city of Judah and on the second day of the month of Addaru he seized the city and captured the king. He appointed there a king of his own choice, received its heavy tribute and sent to Babylon."¹ That is 16 March 597 BC. 2 Kings 24 has Nebuchadnezzar take Jerusalem, deport Jehoiachin, and install Mattaniah as Zedekiah in his place. The chronicle does not name Jehoiachin or Zedekiah, but the sequence — siege, capture, king deposed, replacement installed, tribute taken — is the biblical sequence, from a Babylonian scribe with no interest in Judah's national story.
• The deported king turns up on ration lists in Babylon. Tablets excavated near the Ishtar Gate between 1899 and 1917 record oil rations issued from the royal stores. One reads "10 (sila) to Ia-ku-u-ki-nu, the son of the king of Ia-ku-du, 2½ sila for the 5 sons of the king of Ia-ku-du." Another names "Ya'u-kīnu, king of the land of Yahudu."² That is Jehoiachin, king of Judah, drawing a royal allowance in Babylon with five sons. 2 Kings 25:29–30 says he "ate daily in the king's presence for the rest of his life" and "was given daily provisions by the king."³ A ration list is precisely what that describes.
• A settlement called "the town of Judah" existed in Babylonia and its residents' contracts survive. About 200 cuneiform tablets, dated 572 to 477 BC, come from āl Yahudu, the Town of Judah, and neighbouring places, and they carry Judean personal names in Judean legal transactions.⁴ The exile was not a literary theme; it was a tax district. But the provenance of these tablets is bad and this must be said: they are unexcavated, entered the antiquities market by unknown means, possibly through looting, and sit largely in private collections.⁵ This is a real weakness and anyone citing them should say so.
• Two officials who threw Jeremiah into a cistern have left their seals. Jeremiah 38:1 names "Gedaliah son of Pashhur" and "Jehucal son of Shelemiah" among the officials who demanded the prophet's death.⁶ Eilat Mazar's City of David excavations recovered a bulla reading Yehuchal ben Shelemyahu in 2005 and a second reading Gedalyahu ben Pashhur in 2008, a few yards apart.⁷ Two minor characters, named with patronymics, both attested in clay from the right city and the right stratum. This is the kind of correspondence that is very hard to manufacture.
• A Judean officer's letter from the last months before the fall matches a verse in Jeremiah. Lachish Letter 4, an ostracon from the destruction level, ends: "may (my lord) be apprised that we are watching for the fire signals of Lachish according to all the signs which my lord has given, because we cannot see Azeqah."⁸ Jeremiah 34:7 says the Babylonian army was attacking "Jerusalem and the cities of Lachish and Azekah … because they were the only fortified cities of Judah that were still holding out."⁹ A field officer who has lost sight of Azekah is an officer watching the last outposts fall in the order Jeremiah describes. This is an undesigned coincidence of the strongest kind, because neither document is trying to corroborate the other.
• All three Babylonian officials named in Jeremiah 39:3 are attested in cuneiform, and one of them was identified in 2007 in a museum drawer. Michael Jursa found Nebo-Sarsekim on a Sippar receipt acquired by the British Museum in 1920 — a payment of 1.5 minas of gold to a temple, dated to Nebuchadnezzar's tenth year, 595 BC.¹⁰ The match is not just the name: Jeremiah's title rab-sārîs is literally "chief eunuch," which is the tablet's own title for him.¹⁰ Nergal-Sharezer in the same verse is the Neriglissar of classical sources, Nebuchadnezzar's son-in-law and later king; Nebuzaradan appears on a court-official prism from about 570 BC.¹⁰¹¹ Irving Finkel of the British Museum: "A throwaway detail in the Old Testament turns out to be accurate and true."¹²
• The Cyrus Cylinder does not mention the Jews, and apologetic use of it should stop overstating. The cylinder is real, it is contemporary, and it describes a policy: "I returned the images of the gods, who had resided there, to their places and I let them dwell in eternal abodes. I gathered all their inhabitants and returned to them their dwellings."¹³ The cities it names are Assyrian and Babylonian — Aššur, Susa, Agade, Ešnunna, Zamban, Me-Turnu, Der.¹⁴ Jerusalem, Judah, Israel and the Jews appear nowhere on it. What it supplies is the policy background that makes Ezra 1 plausible — a Persian king who restored deported populations and their cults as a matter of imperial practice. It does not confirm Ezra 1, and calling it "the Cyrus decree" is wrong.
• The exile numbers in Jeremiah are small, specific and awkward, which is a point in their favour. Jeremiah 52:28–30 gives an official tally: 3,023 in the seventh year, 832 from Jerusalem in the eighteenth, 745 in the twenty-third, "in all, 4,600 people."¹⁵ 2 Kings 24:14 says ten thousand. The two figures do not agree, and Jeremiah's look like an administrative extract — odd numbers, three separate deportations, a stated total. A national story being invented does not usually deflate its own catastrophe to 4,600 people.
• Belshazzar and Darius the Mede belong to a different article, and the honest report on them is split. Daniel 5 names a king of Babylon that no classical source knew and cuneiform recovered in the nineteenth century — a genuine and impressive hit. Daniel's Darius the Mede has no room in the Babylonian record between Belshazzar and Cyrus, and this is the hardest historical problem in the Old Testament. Both are worked through in dating Daniel.
• The seal corpus from Jerusalem's last administration is larger than two bullae, and it splits cleanly on provenance. Excavated: "Eliakim, assistant of Jehoiachin" on jar handles from royal estates, Ja'azaniah "royal official" from Mizpah, and a Gedaliah "over the house" seal impression from Lachish.¹¹ From the market: Baruch son of Neriah "the scribe" and "Jerahmeel the king's son" — the two most emotionally satisfying items in the corpus, since Baruch's carries name, patronymic and profession. This project does not use them. The provenance rule bites hardest exactly where one would most like it not to.
• Honest weaknesses. The year of Jerusalem's destruction is still disputed between 587 and 586 BC and the chronicle covering it is lost. The Al-Yahudu tablets are unprovenanced. The Persian-period archaeology of Yehud is thin, and the province was small and poor — which sits awkwardly with the scale of the restoration described in Ezra and Nehemiah. And the correspondences above are with the narrative books; they establish the setting and the personnel, not the theological claims.

Verdict: from 597 BC onward the biblical narrative is anchored in contemporary foreign documents at a level of detail that no other part of the Old Testament reaches.  A dated Babylonian chronicle, ration tablets naming the deported king and his sons, seals of two named officials, a soldier's letter matching a prophet's list of surviving cities, and a settlement of Judeans doing business in Babylonia.  This does not make the whole Old Testament historical by extension — the argument does not transfer backwards, and anyone using it that way is overreaching.  What it does establish is that the writers of Kings, Jeremiah and Ezekiel were describing a world they knew, with real people in it.

Part 1 — The fall of Jerusalem

597 BC, dated to the day

The Babylonian Chronicle series is a set of terse year-by-year records kept by Babylonian scribes.  The chronicle covering Nebuchadnezzar's early years — ABC 5, sometimes called the Jerusalem Chronicle — is a contemporary administrative document with no theological interest whatsoever.  Its entry for the seventh year:

In the seventh year, the month of Kislîmu, the king of Akkad mustered his troops, marched to the Hatti-land, and besieged the city of Judah and on the second day of the month of Addaru he seized the city and captured the king.  He appointed there a king of his own choice, received its heavy tribute and sent to Babylon.¹

Addaru 2 of Nebuchadnezzar's seventh year is 16 March 597 BC.  This is one of the few events in the Old Testament that can be dated to a single day.

Set beside 2 Kings 24:10–17: Nebuchadnezzar besieges Jerusalem, Jehoiachin surrenders, he is deported with his household and the elite of the city, tribute is taken from the temple and palace, and Nebuchadnezzar installs Jehoiachin's uncle Mattaniah as king under the name Zedekiah.

Four independent elements match: the siege, the capture, the removal of the king, and the installation of a Babylonian appointee, with tribute.  The chronicle names no Judean.  The Bible names them all.  Neither text depends on the other.

587 or 586 for the destruction?

The chronicle that would have covered Jerusalem's second and final fall is lost.  The date is reconstructed from biblical regnal figures cross-referenced with Babylonian chronology, and scholarship has been split between 587 and 586 BC for a century.  This is a one-year disagreement about a well-attested event, not a doubt about whether it happened.  It is worth stating because popular writing on both sides tends to present one figure as settled.

2 Kings 25:8–10 dates the burning to the seventh day of the fifth month of Nebuchadnezzar's nineteenth year, when Nebuzaradan "burned down the Lord's temple, the royal palace, and all the houses in Jerusalem" and the army "tore down the walls."¹⁶  The destruction layer in the City of David is real and is one of the fixed points used in the 2024 radiocarbon study of Iron Age Jerusalem discussed in David, Solomon, and the United Monarchy — that study treats 586 BC as "the historically secure stratigraphic horizon" and used it as a calibration anchor.

The Lachish ostraca

In 1935 and 1938, excavators at Tell ed-Duweir (Lachish) found ostraca — letters written in ink on potsherds — in the destruction debris of the gate.  They are Hebrew military correspondence from the last months of the kingdom.

Letter 4, from an officer named Hoshaiah to his superior Ya'ush, closes:

And may (my lord) be apprised that we are watching for the fire signals of Lachish according to all the signs which my lord has given, because we cannot see Azeqah.⁸

Jeremiah 34:7:

He did this while the army of the king of Babylon was attacking Jerusalem and the cities of Lachish and Azekah.  He was attacking these cities because they were the only fortified cities of Judah that were still holding out.⁹

An outpost officer reporting that Azekah's signal is no longer visible, in a period when Jeremiah names Lachish and Azekah as the last two fortified cities besides Jerusalem, is a correspondence nobody designed.  It is worth noticing what makes it strong: neither text is making a claim the other could corroborate.  Hoshaiah is reporting a signalling problem.  Jeremiah is listing the cities still holding out.  They intersect by accident.

One caution.  The standard reading — that Azekah has just fallen — is an inference.  The letter says only that Azekah's signal cannot be seen, which could have other causes.  The correspondence with Jeremiah remains striking either way, and it is stronger for being stated carefully.

Letter 3 in the same group mentions "the commander of the army Coniah son of Elnatan, has gone down to go to Egypt," and a letter "from the prophet" warning "Be on guard!"¹⁷  Jeremiah 26:20–23 describes a prophet, Uriah son of Shemaiah, fleeing to Egypt and being extradited.  The connection is suggestive rather than demonstrated;  the ostracon does not name its prophet.

Part 2 — The exiles

Jehoiachin's rations

Robert Koldewey's German excavation of Babylon between 1899 and 1917 recovered administrative tablets from vaulted storerooms near the Ishtar Gate.  Among them are oil-ration lists.  Three of them carry the same name in different spellings:

Babylon 28122: "…t[o] Ia-'-u-kin, king…" Babylon 28178: "10 (sila of oil) to …Ia-'-kin, king of Ia[…] 2½ sila to [… so]ns of the king of Ia-a-hu-du" Babylon 28186: "10 (sila) to Ia-ku-u-ki-nu, the son of the king of Ia-ku-du, 2½ sila for the 5 sons of the king of Ia-ku-du"²

Another lists him among foreign craftsmen — carpenters from Arvad, woodworkers from Byblos, Greek craftsmen — with "10 sila to Ia-ku-u-ki-nu, the king of Judah's son."¹⁸

What this establishes.  A man called Jehoiachin, identified as king of Judah, was in Babylon drawing rations from the royal stores, with five sons.  The tablets date to roughly 595–570 BC.¹⁹

What the Bible says.  2 Kings 25:27–30 has Evil-Merodach release Jehoiachin from prison in the thirty-seventh year of his exile, seat him above the other captive kings, and provide "daily provisions by the king for the rest of his life."³

The ration tablets are earlier than the release, so they do not confirm that specific episode.  What they confirm is the underlying situation: a deported Judean king maintained at Babylonian royal expense, with his household, as a matter of administrative routine.  Which is the situation both 2 Kings and Jeremiah describe.

Āl-Yahudu, the Town of Judah

About 200 cuneiform tablets record legal and commercial transactions — leases, loans, sales, tax receipts — from a place called āl Yahudu, "the town of Judah," and from neighbouring settlements, some with names like Tel-Aviv and Tel-Melah that also appear in Ezekiel.⁴  The earliest is dated 572 BC, about fifteen years after the destruction of the temple;  the latest is 477 BC under Xerxes.²⁰

The picture they give is of a Judean community that was resettled as a group, kept its own naming traditions across generations, and functioned as an administrative unit — a šušānu Yahūdāya, a unit of Judean state dependents, used for taxation and conscription.²¹  Judean names with the theophoric element -yahu appear throughout.

The problem, stated plainly.  These tablets were not excavated.  They surfaced on the antiquities market by unknown means, possibly through looting, and they are held mostly in private collections.⁵  That means their archaeological context is gone and their authenticity rests on internal criteria — the plausibility of the Akkadian, the prosopography, the coherence of the archive.  Specialists who have published them regard them as genuine, and the internal case is strong.  But an unprovenanced archive is a weaker witness than an excavated one, and a reader should know that.

A second caution about distribution.  Of the roughly 200 texts, only fourteen come from the Neo-Babylonian period and only three from Nebuchadnezzar's reign.²²  The bulk are Persian-period, mostly under Darius I.  So the archive documents the Judean community in Babylonia far better than it documents the deportation itself.

Part 3 — The bullae

A bulla is a lump of clay pressed with a seal to close a document.  The document rots;  the clay survives, especially if the building burns.  Jerusalem's destruction layer has produced a large number.

Two are directly relevant.

Jeremiah 38:1 names four officials who heard Jeremiah's message and demanded his death: "Shephatiah son of Mattan, Gedaliah son of Pashhur, Jehucal son of Shelemiah, and Pashhur son of Malkijah."⁶  Jeremiah 37:3 has Zedekiah send "Jehucal son of Shelemiah" to the prophet.

Eilat Mazar's City of David excavation recovered:

• a bulla reading Yehuchal ben Shelemyahu, in 2005;
• a bulla reading Gedalyahu ben Pashhur, in 2008, a few yards away and in the same First Temple stratum.⁷

Why this is strong.  Both are minor characters.  Neither is a king.  Both are given with patronymics, which is what makes the identification checkable — the name plus father's name is a much rarer combination than a name alone.  And they were found in the right city, in the right period's stratum, within a few yards of each other, which is what one would expect of two officials serving the same administration.

Why it is not decisive.  Hebrew names repeat, and neither bulla carries a title.  A different Gedaliah son of Pashhur is possible.  The strength of the case comes from the combination of two of the four names in Jeremiah 38:1 turning up together, not from either one alone.

A general caution about this class of evidence.  Bullae are small, portable, valuable, and heavily forged.  These two came from a controlled excavation, which is what makes them usable.  Unprovenanced bullae — and the market is full of them, including some famous ones — should not be cited.

Part 3b — Jeremiah 39:3, and three officials in cuneiform

One verse deserves separate treatment, because it is the densest concentration of checkable detail in the exilic narrative.

Jeremiah 39:3 describes the moment Jerusalem's wall was breached: "all the officials of the king of Babylon came and took seats in the Middle Gate: Nergal-Sharezer of Samgar, Nebo-Sarsekim a chief officer, Nergal-Sharezer a high official and all the other officials of the king of Babylon."

Nebo-Sarsekim.  On 5 July 2007, Michael Jursa of the University of Vienna, working through the British Museum's roughly 130,000 cuneiform tablets on Babylonian officials, identified this name on a small receipt from Sippar acquired by the museum in 1920.  The text reads in full:

[Regarding] 1.5 minas [0.75 kg] of gold, the property of Nabu-sharrussu-ukin, the chief eunuch, which he sent via Arad-Banitu the eunuch to [the temple] Esangila: Arad-Banitu has delivered [it] to Esangila.  In the presence of Bel-usat, son of Alpaya, the royal bodyguard, [and of] Nadin, son of Marduk-zer-ibni, Month XI, day 18, year 10 [of] Nebuchadnezzar, king of Babylon.¹⁰

Year 10 of Nebuchadnezzar is 595 BC, eight years before the fall of Jerusalem.

Why this is a strong identification.  It is not just a name.  The title matches exactly: Hebrew rab-sārîs, which the NIV renders "chief officer," is literally "chief eunuch," which is the tablet's own title for Nabu-sharrussu-ukin.¹⁰  Name plus office is two of Mykytiuk's three identifying marks, and the period agreement supplies the third;  he includes Nebo-Sarsekim in his firm list.  See Archaeological Confirmations of the Old Testament.

And the provenance is clean, which matters after Part 5's forgery discussion in the sibling article: the tablet came out of a nineteenth-century Sippar excavation into the British Museum in 1920 and sat unread in a drawer for eighty-seven years.  Nobody made it to order.

Irving Finkel of the British Museum, who has no stake in the Bible's reliability, put the implication this way: "If Nebo-Sarsekim existed, which other lesser figures in the Old Testament existed?  A throwaway detail in the Old Testament turns out to be accurate and true."¹²

The other two officials in the same verse are also attested.  Nergal-Sharezer is known from cuneiform as the son-in-law of Nebuchadnezzar, married to his daughter Kashshaia, who later became king of Babylon himself — the Neriglissar of the classical sources.¹⁰  And Nebuzaradan, the captain of the guard who burns the city in 2 Kings 25:8–9, has been identified on a prism listing officials stationed around the king.  Donald Wiseman's account, read directly, is more careful than the apologetic summaries of it:

First a group of at least nine high-officials (masennu), of these Nabû-zêr-iddinam the 'Chancellor' or 'chief baker' (rab nuhatimmu) is to be identified with Nabuzaradan (rab tabbahim) present at the final siege of Jerusalem (Jer. 39:9–10).¹¹

The philology is the good part: Hebrew rab tabbahim, conventionally "captain of the guard," is literally chief of the slaughterers or cooks, and Akkadian rab nuhatimmu is chief baker.  The two titles line up as well as the names do.

But Wiseman states an objection in the same paragraph that apologetic uses drop.  Whether the text is a court list at all is disputed: "Against this is the absence of offices and names known from other texts and the inclusion of officials from districts far from the capital."¹¹  And his own conclusion, several pages later, is that it is not a court list: after the officials the text names kings of Tyre, Gaza, Sidon, Arvad and Ashdod, and "The whole would therefore seem to be a document drawn up to record a procession to commemorate a special occasion."¹¹  He adds that Jehoiachin's absence from the surviving portion "may suggest that their names have been lost in the break" — an honest way of saying the document does not name him.

So the identification of Nebuzaradan is good and the document is a procession list rather than a court register.  Both should be said.

What to make of a verse in which all three named foreign officials turn up in Babylonian records.  Less than triumphalism suggests and more than nothing.  Babylonian court officials are exactly the class of person Babylonian archives record, so finding them is not miraculous.  What it establishes is that whoever wrote Jeremiah 39 knew the names and the correct titles of specific mid-level Babylonian functionaries — the sort of detail that decays fastest in a tradition and that a later writer has no way to reconstruct.

Part 3c — The seals of Jerusalem's last administration

Wiseman assembles the epigraphic material from the years around the fall, and it is a larger group than the two bullae above.  It is also a group where the provenance test does real work, so it is set out with the provenance attached to each item.

From controlled excavation.

• "Eliakim, assistant of Jehoiachin" (l'lyqm n'r ywkn) — stamped jar handles from Tell Beit Mirsim, Beth-Shemesh and Ramat Rahel, from royal estates that Wiseman says "were left to supply wine to Babylon as they had previously to the royal household in Jerusalem."¹¹
• Ja'azaniah, royal official (y'znyh 'bd hmlk) — a seal from Tell en-Nasbeh, generally identified as Mizpah, which is where Gedaliah made his headquarters and where Jeremiah was brought (Jeremiah 40:6).¹¹
• Gedaliah — a seal impression reading lgdlyh 'sr 'l hbyt, "belonging to Gedaliah who is over the house," found at Lachish; Wiseman notes he "is generally identified with the prime-minister."¹¹ This is a different Gedaliah from the official of Jeremiah 38:1 whose bulla Eilat Mazar found, and the two should not be conflated.

Not from controlled excavation, and this matters.

• Baruch son of Neriah the scribe (lbrkyhw bn nryhw hspr) and "Jerahmeel the king's son" (lyrhm'l bn hmlk), the latter plausibly the man sent to arrest Jeremiah in Jeremiah 36:26.¹¹

The Baruch and Jerahmeel bullae are the most emotionally satisfying items in the whole corpus — Jeremiah's own secretary, named with his father and his profession, which is three identifying marks — and they come from the antiquities market, not from a dig.  Wiseman was writing in 1985, before the forgery controversies of the 2000s;  the Baruch bulla in particular has since been widely doubted.  This project does not use them, and lists them here only to make the point that the provenance rule bites hardest exactly where one would most like it not to.  See Archaeological Confirmations of the Old Testament on why.

A note on the deportation figure.  Wiseman records that Babylonian reprisals came in 582/1 BC when Nebuzaradan "took a further 745 Judeans into exile,"²³ which is Jeremiah 52:30's number for the twenty-third year.  A specialist in Neo-Babylonian history treating that odd figure as usable is worth noting, given how often small precise biblical numbers are dismissed.

Part 4 — Cyrus and the return

What the Cylinder says

The Cyrus Cylinder is a Babylonian foundation deposit written after Cyrus took Babylon in 539 BC.  It is a legitimation document: Marduk was angry with Nabonidus, Marduk chose Cyrus, Cyrus restored what Nabonidus had disturbed.  The operative passage:

From [Babylon] to Aššur and (from) Susa, Agade, Ešnunna, Zamban, Me-Turnu, Der, as far as the region of Gutium, the sacred centers on the other side of the Tigris, whose sanctuaries had been abandoned for a long time, I returned the images of the gods, who had resided there, to their places and I let them dwell in eternal abodes.  I gathered all their inhabitants and returned to them their dwellings.¹³

What it does not say

It does not mention Jerusalem, Judah, Israel or the Jews.  Not once.  The cities named are Mesopotamian.  Nor could it: the Judeans had no cult statue for Cyrus to return, which is the specific act the cylinder describes.

This matters because the cylinder is routinely presented in popular apologetics as "the Cyrus decree," as though it were the text of Ezra 1 in cuneiform.  It is not.  Ezra 1:2–4 has Cyrus issue a proclamation naming the LORD God of heaven, Jerusalem, and the rebuilding of the temple.²⁴  No such document survives.

What it does establish

A policy.  Cyrus, in his own propaganda, presents himself as the king who reverses deportations and restores local cults.  That is exactly the imperial practice under which a return of Judean exiles and a rebuilt temple in Jerusalem would be unremarkable.

The correct claim is therefore modest and still worth making.  Ezra 1 describes an action that fits the documented policy of the ruler it attributes it to, in the reign it attributes it to.  Before the cylinder was read, Ezra's Cyrus looked like a Jewish fantasy of a benevolent pagan king.  He now looks like Cyrus.  That is real corroboration of a kind, and it is not the same thing as finding the decree.

The Persian period afterwards

The archaeology of Persian-period Yehud is thin.  The province was small, poor, and lightly populated, and Jerusalem in the fifth century was a modest town.  This sits awkwardly against the scale suggested by parts of Ezra and Nehemiah, and it is one of the honest difficulties of the period.

Independent Jewish documentation does exist from outside Judah.  The Elephantine papyri, from a Judean military colony in Egypt, include late-fifth-century letters that name Johanan the high priest in Jerusalem and the sons of Sanballat the governor of Samaria — both of whom appear in Nehemiah.  This article has not worked those documents up directly and states them only as a pointer.

How long the Persian period lasted is argued in Traditional Jewish Chronology.  The rabbinic count of Seder Olam Rabbah gives the Persian empire 52 years and dates the fall of Jerusalem near 423 BC.  Dated tablets, royal inscriptions, the Elephantine papyri and 38 observed eclipses support about 207 years and the 587 or 586 BC used in this article.

Common objections

"The Bible's exile narrative is a post-exilic construction."  Then it is a construction that got the day of Jerusalem's capture right, named a deported king who turns up on Babylonian ration lists with five sons, named two junior officials whose seals are in the ground, and listed the last two fortified cities in the order a Judean officer at Lachish was experiencing them.

"The Cyrus Cylinder proves Ezra 1."  It does not mention the Jews.  See above.

"The Al-Yahudu tablets prove the exile."  They are unprovenanced.  They corroborate strongly and they are not clean evidence, and both halves of that sentence should be said.

"The bullae could be different men with the same names."  Possible for either one alone.  The case rests on two of the four officials named in a single verse turning up in the same stratum a few yards apart.

"Jeremiah's 4,600 exiles contradicts 2 Kings' 10,000."  It does.  The most economical reading is that Jeremiah 52:28–30 preserves an administrative count of adult males or heads of household from three deportations, and 2 Kings 24:14 gives a round figure for the whole 597 group including dependents.  That is a harmonisation and it is not certain.  What should be resisted is the pretence that there is no tension.

"None of this confirms the religious claims."  Correct, and it should be conceded without hesitation.  Establishing that Jehoiachin was a real king drawing rations in Babylon says nothing about whether the exile was divine judgement.  What this material establishes is that the narrative books are reporting a real world with real people in it, which is a precondition for the theological argument rather than a substitute for it.

What would settle it

• The chronicle tablet covering 587/586. A continuation of ABC 5 would date the destruction of Jerusalem the way ABC 5 dates the capture, and end the 587-versus-586 argument overnight. Babylonian chronicle tablets are still being identified in museum collections.
• An excavated Al-Yahudu tablet. A single tablet from that archive recovered in a controlled dig would convert the whole corpus from strong-but-tainted to clean.
• A Persian administrative document mentioning Yehud's restoration. The Persepolis and Babylonian archives are enormous and incompletely published.
• Publication of more Elephantine material bearing on Jerusalem's fifth-century administration.

Sources and further reading

The literature reviewed for this article, including sources considered and not used, is in literature-review.md.