Archaeological Confirmations of the Old Testament

Summary

This article is an inventory rather than an argument.  It lists what has actually been confirmed, states a published criterion for what counts as confirmation, and includes the claims that failed alongside the ones that held.

• Fifty-three people named in the Hebrew Bible have been identified in ancient inscriptions. Lawrence Mykytiuk of Purdue published the list in Biblical Archaeology Review in 2014 and extended it in 2017, with full documentation of the biblical references and the inscriptions for each.¹ They range from Egyptian and Mesopotamian kings down to junior officials of Judah.
• He uses a criterion, and the criterion is the important part. A firm identification requires three matching identifying marks — typically the person's name, his father's name, and his title or office. Two marks give "no more than a reasonable hypothesis," because Hebrew names repeat and many men named Hananiah had a father named Azariah.¹ This is a falsifiable standard, applied consistently, and it is why his list is usable where most "the Bible was right again" lists are not.
• What a confirmed name establishes is narrow, and Mykytiuk says so. Verifying that a person existed does not verify what the Bible says he did. What it does establish is that the person, his father and his office were as described — which shows he was in a position to do what he is said to have done. That is a real result and it is not the same as corroborating the narrative.
• The strongest single items are the ones where a hostile or indifferent foreign source names an Israelite. Shalmaneser III's Kurkh Monolith on Ahab at Qarqar in 853 BC. The Black Obelisk on Jehu's tribute in 841. Sennacherib's prism on Hezekiah shut up "like a bird in a cage" in 701. The Babylonian Chronicle on the capture of Jerusalem in March 597. Ration tablets on Jehoiachin in Babylon. These come from administrations with no interest in the biblical story.
• The confirmations cluster heavily in one period, and this shape is itself informative. Of the 53, almost all belong between the ninth and fifth centuries BC. Exactly one belongs earlier: David, from the Tel Dan stele, and that is a dynastic label rather than a contemporary notice. There is no confirmed person from the patriarchal narratives, the Exodus, the conquest, or the period of the judges. The Hebrew Bible's historicity, measured this way, gets steadily better as it gets later, which is what one would expect of any ancient tradition and is not what a triumphalist presentation implies.
• A parallel list of "almost real" people exists and it should be cited alongside. Mykytiuk keeps a second category — identifications that are reasonable but not firm because only two marks match. Shebna the royal steward, from the Silwan tomb inscription, is one: "There are only two marks (traits) of an individual, and these do not include his complete name, so this identification, though tempting, is not quite firm."¹ Balaam son of Beor appears in a plaster wall inscription at Deir ʿAllā, in the same Transjordanian region the Bible places him — but the inscription is from about 700 BC and the relationship is to a shared folk tradition, "which is not necessarily historical."¹
• Site-level confirmations are real and are a different kind of evidence. Hezekiah's tunnel and its inscription; the Siloam Pool; the Lachish siege ramp and Sennacherib's own relief of the siege; the Babylonian destruction layer in Jerusalem; the Lachish ostraca; the Ketef Hinnom amulets. These confirm settings and events rather than the theological claims attached to them.
• The forgery problem is the largest structural threat to this whole category, and the field's own record is bad. In 2003 the Israel Antiquities Authority declared the James Ossuary, the Jehoash Inscription, the Ivory Pomegranate and other high-profile inscriptions forgeries. After a seven-year trial, Oded Golan and Robert Deutsch were acquitted of all forgery charges in March 2012 — the court finding that the prosecution had not proved forgery beyond reasonable doubt, not that the objects were genuine.² Nothing was settled, and the items remain unusable for exactly that reason. Every one of them is unprovenanced.
• Honest weaknesses. A list of confirmed names is the easiest kind of apologetic to assemble and the easiest to over-read. It says nothing about the Pentateuch, nothing about miracles, and very little before 900 BC. Some entries depend on contested readings — Belshazzar's title, the Mesha stele's line 31, the Tel Dan word divider. And the same field that produced this list also produced the Ebla "Sodom" claim, the Tall el-Hammam airburst, and the pig-bone ethnic marker, all withdrawn.

Verdict: from the ninth century BC onward, the Hebrew Bible names real people who really held the offices it assigns them, and this is established by a published criterion rather than by enthusiasm.  Before the ninth century, almost nothing in this category exists.  A skeptic who says "no biblical figure has ever been confirmed" is wrong by fifty-three.  An apologist who presents the list as confirming the Bible's narrative or its theology is claiming more than the list carries.

Part 1 — The criterion

Most lists of "biblical figures confirmed by archaeology" are unusable because they have no threshold.  A name on a seal that matches a name in the Bible is not an identification;  it is a coincidence waiting to be assessed.

Mykytiuk's rule is that a firm identification needs three identifying marks in common between the inscription and the biblical text.  Typically these are:

1. The person's name.
2. His father's name (patronymic).
3. His title, office, or another distinguishing detail — "king of Judah," "the scribe," a town, a role.

Two marks are not enough.  His worked example is Hananiah son of Azariah, from a blue stone seal published in 1883:

This seal reveals only two marks (traits) of an individual, the names of father and son, therefore the identification it provides can be no more than a reasonable hypothesis … One must keep in mind that there were probably many people in Judah during that time named Hananiah/Hananyahu, and quite a few of them could have had a father named ʿAzariah/ʿAzaryahu … Therefore, it would take a third identifying mark of an individual to establish a strong, virtually certain identification.¹

Why this is the right kind of rule.  It states in advance what would count, it can be applied by a critic, and it produces a shorter list than the author would like.  Mykytiuk's excluded categories are as informative as his included ones: he leaves out "Shalman" in Hosea 10:14 because too little is known of the biblical figure to attach him to anyone, and "Osnappar" in Ezra 4:10 for the same reason.¹  He also excludes identifications that rest on copied manuscripts rather than excavated inscriptions — including Josephus on Ethbaal, Jezebel's father, "because Josephus' writings do not come to us from archaeology."¹

That last exclusion is stricter than most apologists would apply and it is correct.

Part 2 — The list

The 53, grouped as Mykytiuk groups them, with reign or floruit dates as he gives them.¹

Egypt (5).  Shishak / Sheshonq I (945–924);  So / Osorkon IV (730–715);  Tirhakah / Taharqa (690–664);  Necho II (610–595);  Hophra / Apries (589–570).

Moab (1).  Mesha (early to mid-ninth century).

Aram-Damascus (5).  Hadadezer;  Ben-hadad son of Hadadezer;  Hazael (844/842–c. 800);  Ben-hadad son of Hazael;  Rezin (to 732).

The northern kingdom of Israel (9).  Omri (884–873);  Ahab (873–852);  Jehu (842/841–815/814);  Joash (805–790);  Jeroboam II (790–750/749);  Menahem (749–738);  Pekah (750?–732/731);  Hoshea (732/731–722);  and Sanballat "I," governor of Samaria under Persian rule (mid-fifth century).

The southern kingdom of Judah (14).  David (c. 1010–970);  Uzziah / Azariah (788/787–736/735);  Ahaz (742/741–726);  Hezekiah (726–697/696);  Manasseh (697/696–642/641);  Hilkiah, high priest under Josiah;  Shaphan, scribe under Josiah;  Azariah, high priest under Josiah;  Gemariah, official under Jehoiakim;  Jehoiachin (598–597);  Shelemiah, father of Jehucal;  Jehucal, official under Zedekiah;  Pashhur, father of Gedaliah;  Gedaliah, official under Zedekiah.

Assyria (6).  Tiglath-pileser III / Pul (744–727);  Shalmaneser V (726–722);  Sargon II (721–705);  Sennacherib (704–681);  Adrammelech / Arad-mullissu, Sennacherib's son and assassin;  Esarhaddon (680–669).

Babylonia (7).  Merodach-baladan II;  Nebuchadnezzar II (604–562);  Nebo-sarsekim, his official;  Nergal-sharezer, his officer;  Nebuzaradan, a chief officer;  Evil-merodach / Amel-Marduk (561–560);  Belshazzar, son and co-regent of Nabonidus (c. 543?–540).

Persia (6).  Cyrus II (559–530);  Darius I (520–486);  Tattenai, provincial governor of Trans-Euphrates;  Xerxes I / Ahasuerus (486–465);  Artaxerxes I (465–425/424);  Darius II (425/424–405/404).

The argument the list supports, stated by its best advocate

A list of confirmed names invites the question "so what?"  Kitchen gives the sharpest version of the answer, and it is about the regnal chronology of the divided monarchy rather than about any single name.

The books of Kings give reign lengths and cross-references between the two kingdoms — Menahem accedes in Uzziah's thirty-ninth year, Pekahiah in his fiftieth, Hoshea in Jotham's twentieth — dozens of interlocking figures.  Assyrian records give independent dates for several of the same kings.  Kitchen's point is what the alternative requires:

a later writer's imagination that just happens (miraculously!) to coincide almost throughout with the data then preserved only in documents buried inaccessibly in the ruin mounds of Assyrian cities long since abandoned and largely lost to view.³

How much weight this carries.  Real weight, and less than the exclamation mark implies.  The interlocking regnal data are genuinely hard to invent, and they were checkable only after Assyriology recovered the eponym canon in the nineteenth century.  But three qualifications belong with it.  The reconciliation is not automatic — it requires coregencies and dual dating, which is what Thiele's system is for, and critics have called that machinery ad hoc.  The synchronisms are for the divided monarchy, the period where this section's list is dense anyway, and they say nothing about earlier material.  And a Judean scribe working in the eighth or seventh century would have had access to court annals, so "later writer" need not mean "writer with no sources."

The project's chronology article works through the seventeen cuneiform synchronisms and Thiele's system in detail, including where the numbers still fail to reconcile.

What the shape of the list shows

Foreign kings dominate.  Twenty-nine of the fifty-three are non-Israelite rulers.  That is not a weakness — the Bible naming the right pharaohs, Assyrian kings and Persian emperors in the right order is a genuine result — but it should be counted for what it is.  Egypt, Assyria, Babylon and Persia left far more records than Judah did.

The Judahite officials are the most interesting entries.  Hilkiah, Shaphan, Gemariah, Jehucal, Gedaliah and their fathers are minor characters, several of them named only once or twice.  Finding them is much harder to explain as coincidence than finding Nebuchadnezzar.  Two of them — Jehucal son of Shelemiah and Gedaliah son of Pashhur — appear in a single verse, Jeremiah 38:1, and were found as bullae a few yards apart in the same stratum.  See Exile, Babylon, and Persia.

And the period distribution is the fact nobody advertises.  Fifty-two of the fifty-three fall between roughly 950 and 400 BC.  The single earlier entry is David, and that comes from the Tel Dan stele's dynastic label bytdwd, written about a century and a half after his death — see David, Solomon, and the United Monarchy.

There is no confirmed individual from the patriarchal narratives, the Exodus, the conquest, or the judges.  Not one.  This should be stated plainly by anyone using the list, and it has an ordinary explanation — the second millennium in Canaan is far more poorly documented than the first, and none of these figures was a king with an archive.  But an ordinary explanation is not the same as a confirmation, and the shape of the evidence is what it is.

Part 3 — The "almost real" list

Mykytiuk keeps a second, weaker category, and citing the first without the second is the commonest misuse of his work.

Shebna the royal steward.  A rock-cut tomb inscription at Silwan near Jerusalem plausibly belongs to the official of 2 Kings 18:18 and Isaiah 22:15–16.  Mykytiuk's assessment: "There are only two marks (traits) of an individual, and these do not include his complete name, so this identification, though tempting, is not quite firm."¹

Balaam son of Beor.  A plaster wall inscription from Tell Deir ʿAllā, dated to about 700 BC, names Balaam son of Beor — in the same Transjordanian region where Numbers 22 places him.  This is genuinely striking.  But Mykytiuk's summary of the scholarship is careful: many scholars conclude that the inscription's Balaam and the Bible's belong to "the same folk tradition, which is not necessarily historical."¹  A shared tradition at 700 BC is not an attestation of a thirteenth-century person.

Hananiah son of Azzur.  Two marks, no third.  A reasonable hypothesis and nothing more.¹

The existence of this second list is what makes the first one credible.  A scholar who publishes his near-misses alongside his hits is applying a standard;  one who publishes only hits is compiling a brief.

Part 4 — Sites, inscriptions and events

Name identifications are one category.  Confirmations of settings and events are another, and they are treated in detail in the sibling articles.  A short inventory, with the honest rating attached.

Firm, from excavated contexts.

• The Babylonian capture of Jerusalem, 16 March 597 BC, dated in a Babylonian chronicle and matching 2 Kings 24 on four independent elements. Exile, Babylon, and Persia.
• Jehoiachin in Babylon, on oil-ration tablets from near the Ishtar Gate, with his five sons.
• Sennacherib's siege of Lachish, 701 BC — the siege ramp is in the ground and Sennacherib's own palace relief depicts the assault. Almost uniquely, an event is attested by the attacker's picture of it, the defender's ruin, and both sides' texts. Why Did Sennacherib Fail to Take Jerusalem?
• The Lachish ostraca, Hebrew military letters from the destruction debris, one of which intersects Jeremiah 34:7 by accident.
• Hezekiah's tunnel and the Siloam inscription, describing the two tunnelling teams meeting.
• The Ketef Hinnom silver amulets, carrying Numbers 6:24–26 — though the palaeographic date is disputed by six hundred years. See Moses and the Composition of the Pentateuch.
• Merneptah's Israel, about 1208 BC. See The Emergence of Early Israel.

The Assyrian side, with the texts.  Nine kings of Israel and Judah are named in Assyrian inscriptions;  five of the references — Menahem, Pekah, Hoshea, Ahaz and Hezekiah — have direct biblical parallels, and four concern events the Bible does not record.⁴  The two most-cited are worth quoting rather than summarising.

Shalmaneser III's Kurkh Monolith, on the battle of Qarqar in 853 BC, itemises the coalition against him:

1,200 chariots, 1,200 cavalry, and 20,000 soldiers belonging to Hadad-ezer of Damascus;  700 chariots, 700 cavalry, and 10,000 soldiers belonging to Irhuleni of Hama;  2,000 chariots, and 10,000 soldiers belonging to Ahab, the Israelite … 1,000 camel-(rider)s of Gindibu from Arabia … (all together) these were 12 kings.⁵

Two observations that cut different ways.  Ahab fields more chariots than anyone else in the list, which is a larger force than most reconstructions of ninth-century Israel would predict, and a number of scholars suspect the figure is inflated or corrupt — Assyrian royal inscriptions routinely exaggerate.  And the same passage supplies, incidentally, the thousand Arabian camels that bear on the patriarchal camel question.

Real but contested in what they mean.

• Khirbet Qeiyafa, a tenth-century fortified city whose ethnic affiliation is disputed.
• The Large Stone Structure in the City of David, whose extent, unity and date are disputed.
• The Late Bronze I wall at Jericho reported in the 2024 interim excavation report, for which no destruction date is given. See Jericho, Ai, and the Conquest Narratives.

Not usable.

• Anything unprovenanced. See Part 5.

Part 5 — The forgery problem

This is the largest structural threat to the whole category, and the field's record is poor enough that it should be stated first rather than buried.

What happened.  In 2003 the Israel Antiquities Authority convened expert panels which concluded that several of the most celebrated recent finds were forgeries: the James Ossuary ("James, son of Joseph, brother of Jesus"), the Jehoash Inscription (a royal text describing temple repairs), the Ivory Pomegranate (read as a temple object), the Three Shekels ostracon and the Widow's Plea ostracon.  Oded Golan and Robert Deutsch were charged.

What the court decided.  After a trial running from 2005 to 2012, Judge Aharon Farkash acquitted both men of all forgery charges in March 2012, finding that the prosecution had not proved forgery beyond reasonable doubt.  Golan was convicted of trading in antiquities without a permit.²

What that means, exactly.  An acquittal in a criminal forgery trial is a finding that forgery was not proved.  It is not a finding that the objects are authentic.  The scholarly dispute over each item continues, and the items remain unusable in argument — not because a court said they were fake, but because their provenance is unknown and their authenticity is contested.

The general rule that follows.  Every one of the disputed items came from the antiquities market.  Every item in Mykytiuk's firm list that could be checked came from an excavation or from a documented early acquisition.  The two bullae of Jeremiah's officials are usable because Eilat Mazar dug them out of a stratum;  a bulla bought in Jerusalem in 1990 with the same text would be worth nothing.

And the proportion should be kept straight.  Grabbe's survey makes the point that the forgery controversy distorts non-specialists' picture of the field: "headlines such as 'Forgery Hysteria Grips Israel: Is Everything a Fake?' can leave non-specialists with a rather distorted picture of what is consensus and what is genuine controversy.  Museums around the world are filled with artifacts relating to ancient Israel and Judah that no one questions.  It is the unprovenanced objects sold on the antiquities market that cause the problems."⁶

That is the correct summary in both directions.  The excavated corpus is not in doubt.  The market corpus is where every dispute in this section's Part 5 lives.

The worked example is the Baruch bulla, and it is the painful one.  A seal impression reading lbrkyhw bn nryhw hspr — "belonging to Berechiah, son of Neriah, the scribe" — is Jeremiah's own secretary with name, patronymic and profession.  That is three identifying marks;  on Mykytiuk's own criterion it would be a firm identification.  It came from the antiquities market, in the group of bullae published by Avigad, and its authenticity has been widely doubted since.  It is not on Mykytiuk's list, and it should not be on anyone's.  The same applies to the "Jerahmeel the king's son" bulla from the same group, plausibly the man sent to arrest Jeremiah in Jeremiah 36:26.

Compare what survives the test.  From controlled excavation in the same period: "Eliakim, assistant of Jehoiachin" on stamped jar handles from Tell Beit Mirsim, Beth-Shemesh and Ramat Rahel;  Ja'azaniah "royal official" from Tell en-Nasbeh;  and a Gedaliah "over the house" impression from Lachish.⁷  Less famous, less quotable, usable.

Why the market is dangerous in this specific field.  Forgers make what buyers want.  An object matching a biblical verse is worth vastly more than one that does not, so the market generates exactly the finds that would be most persuasive if genuine.  The base rate of forgery among "spectacular" unprovenanced biblical inscriptions is high enough that a spectacular match is evidence of forgery before it is evidence of anything else.

Part 6 — The claims that failed

An inventory that lists only successes is a brief, not an inventory.  Four flagship claims of biblical archaeology have been withdrawn since 1976.

• The Ebla tablets name Sodom and Gomorrah (announced 1976, withdrawn). "The present consensus is that Ebla has no bearing on … Sodom and Gomorrah."⁸
• The Nuzi tablets prove a second-millennium Abraham (withdrawn by conservatives in 1974).⁹
• A Tunguska-scale airburst destroyed Sodom (published 2021, retracted by Scientific Reports 2025).¹⁰
• Pig bones mark Israelite ethnicity (retracted 2013 by the team that made it famous).¹¹

Each is treated in the relevant sibling article.  The reason they belong here is that all four were, in their day, presented as archaeological confirmations of the Bible, and are still presented that way in popular writing.  A reader assessing the list in Part 2 is entitled to know the failure rate of the genre it belongs to.

The right response is not embarrassment.  It is to notice what distinguishes Mykytiuk's list from those four: a published criterion, a stated threshold, a parallel list of near-misses, and full documentation of the evidence for every entry.  That is what makes a list checkable.

Common objections

"Archaeology has confirmed the Bible."  It has confirmed that fifty-three named people existed and held the offices ascribed to them, plus a body of settings and events from the ninth century onward.  It has not confirmed the Pentateuch, the conquest, or anything theological, and it has almost nothing to say before about 950 BC.

"Archaeology has disproved the Bible."  Also false.  The specific things that have been disproved are particular claims made by apologists, not the biblical text.

"Confirming a king's name proves nothing;  of course the Bible knew who the kings were."  For Nebuchadnezzar, largely fair.  For Gedaliah son of Pashhur, an official named once in Jeremiah 38:1, not fair at all.  The list's strength is at the bottom, not the top.

"These lists are just apologetics."  Some are.  Mykytiuk's is not, because it states a threshold in advance, applies it against itself, and publishes the entries that fail.  Judge lists by whether they have a criterion.

"What about the James Ossuary?"  Unprovenanced, contested, and the 2012 acquittal did not establish authenticity.  It should not be used.

"Why is there nothing before David?"  Because the second millennium in Canaan produced almost no written records naming private individuals, and because Judah's own writing culture is attested from the tenth century onward.  This is a real explanation and it is also a real limit, and both halves should be said.

What would strengthen this inventory

• An excavated inscription naming any figure from the Pentateuch or Judges. The single largest gap. None exists.
• A tenth-century royal inscription from Jerusalem. Would move David and Solomon from one indirect attestation to direct ones.
• A published, applied provenance standard for the field, so that unprovenanced items are cited with a fixed marker rather than by each author's discretion.
• Independent re-examination of the disputed market items with methods both sides accept in advance. The 2003–2012 episode consumed a decade and settled nothing because no such agreement existed.

Sources and further reading

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