David, Solomon, and the United Monarchy

Summary

In 1992 Philip Davies published a book arguing that the biblical David was a literary invention.  A year later a basalt fragment turned up at Tel Dan with the letters bytdwd on it.  The argument since then has not really been about whether David existed.  It has been about how much he ruled.

• The existence question is effectively closed, and it closed by inscription. The Tel Dan stele, an Aramaic victory text from the second half of the ninth century, has an unnamed king — almost certainly Hazael of Damascus — boasting of killing a king of Israel and a king of the "House of David."¹ That is roughly a century and a half after David's traditional dates, which is exactly the interval at which a real dynastic founder's name would still be the standard label for his kingdom. Assyrian scribes did the same thing to the northern kingdom, calling it Bit-Humri, the House of Omri, long after Omri was dead.
• The reading has a real objection and it is not fatal. There is no word divider between byt and dwd, and the stele uses word dividers elsewhere.² Dwd on its own can mean uncle, beloved, or kettle. Philip Davies proposed that bytdwd is a place name; George Athas suggested it means Jerusalem; Lehmann and Reichel suggested a divine epithet.³ The answer is that compound place-names and dynastic names routinely lose their dividers once they lexicalise — bytlhm, Bethlehem, is written the same way — and that the surrounding lines name a king of Israel and a second king killed alongside him, which is what the biblical narrative has at the death of Jehoram and Ahaziah. The majority position is that it means the House of David. It is a majority, not a unanimity.
• The Mesha Stele is often given as a second attestation. It should not be, yet. Line 31 of the ninth-century Moabite inscription is broken, and André Lemaire read bt[d]wd, House of David, in 1994. Finkelstein, Na'aman and Römer examined new high-resolution photographs of the stone and of the paper squeeze taken before the stele was smashed, and reported three things: "The taw that follows the beth in Lemaire's rendering … does not exist"; a vertical sentence-divider stroke stands before the waw, meaning the waw begins a new sentence; and the whole stretch between the beth and the waw was already eroded when the squeeze was made.⁴ They read a three-letter personal name and cautiously propose Balak. Lemaire and Delorme answered in 2022 with Reflectance Transformation Imaging and say the missing letter is a dalet after all, with independent support from Michael Langlois.⁵ This article has read the Balak paper in full and only press reporting of the reply, and takes no side. Cite Tel Dan; treat Mesha line 31 as an open question.
• Jerusalem in the tenth century was inhabited. That is now settled by carbon, not argument. A 2024 study in PNAS reports 103 radiocarbon dates from five excavation areas in Iron Age Jerusalem and states the result plainly: "We provide concrete evidence for a widespread human presence in the city during the 12th to 10th centuries BC, as short-lived material from all the areas studied was dated to these centuries."⁶ For twenty years the minimalist claim was that tenth-century Jerusalem was an empty or near-empty village. It was not.
• But the same study dates Jerusalem's monumental growth to the ninth century, not the tenth. Conclusion 2: "The city's westward spread out of its ancient core is assigned here to at least as early as the 9th century BC, as attested by the radiocarbon dating of the construction of monumental Building 100 … and the large-scale terracing activities of the bedrock done prior to its construction."⁶ Building 100's construction is constrained to 900–850 BC.⁶ That is after Solomon on any chronology. The honest reading of this paper is that it moves Jerusalem's urbanisation earlier than Finkelstein wanted and later than the biblical building programme requires.
• Khirbet Qeiyafa is the strongest single find for a tenth-century Judahite state, and it is contested on identity rather than on date. A 6-acre fortified city in the Elah valley, casemate wall, houses built against the wall, radiocarbon on burnt olive pits placing it about 1020–980 BC.⁷ The fauna is goats, cattle, sheep and fish, without the pigs and dogs that mark Philistine sites, and the domestic plan is not Philistine.⁸ Garfinkel reads it as Judahite. Na'aman and Koch read it as Canaanite; Finkelstein and Fantalkin argue it belongs to a north Israelite entity, writing that "There is no evidence for arguing that Jerusalem, Hebron and Khirbet Qeiyafa were the main centres of 10th century Judah."⁹ Everyone agrees somebody built a real fortified city there in the tenth century.
• The strongest recent development is a methodological one, and it comes from a mainstream archaeologist rather than an apologist. Erez Ben-Yosef's radiocarbon work moved the Aravah copper industry at Timna and Faynan from the Egyptian New Kingdom to the tenth century BC — livestock dung so well preserved his team first took it for modern, with imported feed carried more than a hundred miles.¹⁰¹⁰ The polity that ran that industry left almost no architecture. His conclusion in Vetus Testamentum: "the total reliance on stone-built archaeological features in the identification of social complexity … has resulted in skewed historical reconstructions," a flaw that "cuts across all schools of biblical archaeology," since "both 'minimalists' and 'maximalists' have been using stone-built architectural remains as the key."¹¹ So "no palace, therefore no kingdom" is not a valid inference. Note what this does and does not do: it dissolves an argument against a tenth-century state without supplying evidence for one, and it dissolves it for Finkelstein's north Israelite entity just as much as for Judah.
• The "Solomonic" monumental architecture is the weakest part of the traditional case, and Mazar's own chronology does not rescue it. The six-chambered gates at Megiddo, Hazor and Gezer were long credited to Solomon from 1 Kings 9:15. Finkelstein's Low Chronology moves them to the ninth century — "Accepting the Low Chronology means stripping the United Monarchy of monumental buildings."¹² Amihai Mazar's Modified Conventional Chronology is the standard conservative counter, but read in his own words it does something different from what it is used for: it merges the tenth and ninth centuries into a single Iron IIA, which "makes life more difficult for those who wish to utilize archaeology for secure historical interpretations" and means "any comprehensive archaeological synthesis of the Solomonic period is tentative and may be interpreted in more than one way."¹³ He rejects the Low Chronology as "strict and one-sided" without handing the buildings back to Solomon — at Megiddo one of Yadin's two palaces survives only to its foundations and the other's published pottery cannot be dated within Iron IIA.¹³ Yadin had already moved the Megiddo "stables" to Ahab.¹⁴
• Shishak's own inscription does not mention Jerusalem, and that cuts both ways. Shoshenq I's topographical list at Karnak names roughly 150 places from a campaign around 925 BC.¹⁵ Jerusalem is not among them, which is awkward for anyone treating the list as confirmation of 1 Kings 14:25. But the list is also concentrated north and south of Judah rather than in it, which is what one would expect if, as 1 Kings 14:26 says, Rehoboam bought the city off with the temple treasury rather than being sacked. Neither side gets a clean win.
• Finkelstein concedes more than he is credited with, and the disagreement is narrower than the labels suggest. In his own words: the Tel Dan inscription "leaves no doubt that David and Solomon were historical figures," and "many of the David stories in the books of Samuel—mainly the heroic tales and the description of his life as a bandit on the fringe of the Judean highlands—contain genuine, early historical memories." What he denies is the scale: "the great biblical story of the United Monarchy is left with no material evidence."¹⁶ Both he and this article accept David, accept Tel Dan, and reject the empire of 1 Kings 4:21. What is left in dispute is the size of the polity. He is also, on the editors' own summary of that volume, not a minimalist — he "disagrees with both the maximalists and the minimalists."¹⁷
• The empire of 1 Kings 4 has no external support — but Kitchen shows it is the right kind of empire for the period, which this article previously did not allow. Nothing in the archaeology or in the Egyptian, Assyrian and Aramaean records attests a Levantine empire of that extent. Kitchen's answer is that "the scale and nature of the wider realm of David and Solomon are not unique and belong to a specific period of history, namely, ca. 1200–900 — neither earlier nor later."¹⁸ Between the Late Bronze collapse and Neo-Assyria there were several mid-sized hegemonies — Tabal, Carchemish, Aram-Zobah — which "took over defeated opponents and their territories as they stood, without much territorial change," unlike the province-building Assyrian and Persian empires.¹⁸ On that reading 1 Kings 4:21 claims tribute and vassalage, not administration — a polity that reorganises nothing and leaves almost nothing to dig. It remains unattested; it is now harder to explain as a late invention, because the political form is period-specific. And the maximal English reading is wrong on everyone's account, Kitchen's included.
• Honest weaknesses on our own side. The Tel Dan reading rests on a fragment recovered in secondary use, and the Mesha reading on one damaged letter. Eilat Mazar's "palace of David" was dug by an archaeologist who chose the spot from 2 Samuel 5:17 and announced the identification in advance; four senior Israeli archaeologists published against her dating and called her approach biased.¹⁹ And this article's key radiocarbon source is a Tel Aviv–Weizmann collaboration whose own conclusions place the monumental city a century after Solomon.

Verdict: David was real, his dynasty was real and named by its enemies within about 150 years, and by 1000 BC somebody in the Judaean hills was building fortified cities.  Everything beyond that is contested and some of it is unsupported.  The minimalist position of the 1990s — no David, no dynasty, no tenth-century state, Jerusalem an empty village — has been falsified point by point, most recently by carbon.  The maximalist position — a Solomonic empire from the Euphrates to Egypt, with monumental building programmes at Megiddo, Hazor and Gezer — has not been vindicated by the same evidence.  The picture that fits the data is a real, modest, tenth-century highland kingdom centred on Jerusalem, whose founder's name stuck to the dynasty, and which grew into a substantial state during the ninth and eighth centuries.

Part 1 — What is actually claimed

The biblical material about the united monarchy is not one claim but several, and they need very different amounts of evidence.

That David existed.  A minimal claim.  One inscription settles it.

That he founded a dynasty ruling from Jerusalem.  Also minimal, and also inscriptional.  A neighbouring king calling Judah "the House of David" is exactly the evidence one would want.

That Jerusalem was his capital and was a real settlement in his time.  An archaeological claim about a specific, heavily built-over ridge.

That he and Solomon ruled a unified kingdom over both Israel and Judah.  Harder.  This requires a political fact that leaves no direct trace unless someone writes it down.

That Solomon built the temple, the palace, and fortifications at Hazor, Megiddo and Gezer.  An archaeological claim about datable buildings.

That the kingdom stretched from the Euphrates to Egypt.  A claim about an empire, which should be visible in the records of the states it supposedly dominated.

The mistake made by both sides is to treat these as a single package that stands or falls together.  They do not.  The first two are now well supported.  The last is not supported at all.

Part 2 — The inscriptions

The Tel Dan stele

Gila Cook of Avraham Biran's team found Fragment A in July 1993;  Fragments B1 and B2 followed in June 1994.  Biran and Joseph Naveh published them in the Israel Exploration Journal in 1993 and 1995.  The stone was not in its original setting — it had been reused in later construction — which matters for dating the stele but not for its authenticity, since it came out of a controlled excavation.²⁰

The text is Old Aramaic.  An unnamed king, generally identified as Hazael of Damascus, describes killing two kings.  The reconstructed lines read:

… king of Israel, and [I] killed [Ahaz]iahu son of [Jehoram kin-]g of the House of David, and I set [their towns into ruins and turned] their land into [desolation] …²¹

The two names are partly restored, and scholars are divided on the restorations.²²  What is not restored is bytdwd.  Those six letters are on the stone.

Why the objection is worth taking seriously.  The inscription uses word dividers.  If byt dwd meant "House of David," the objection runs, one would expect a divider.  And dwd has other meanings.²

Why it does not carry.  Three reasons.  First, compound toponyms and dynastic names commonly lose the divider once they function as a single name;  Bethlehem is written bytlhm without one.  Second, the alternatives have to explain a text in which the king of Israel and the king of bytdwd appear as a pair, which is what the biblical account of Jehoram and Ahaziah has, and which is not what a place-name or a divine epithet naturally produces.  Third, the "House of X" dynastic formula is standard in the region: Assyrian records call the northern kingdom Bit-Humri, the House of Omri.

Niels Peter Lemche and Frederick Cryer went further and argued from chisel marks that the inscription was a modern forgery.  This has been generally rejected because the fragments came from a controlled dig, and Lemche himself conceded that the forgery arguments "need to be much more forceful if they are to prove beyond doubt that the inscription is the work of a forger."²³

The Mesha Stele

The Moabite Stone, from about 840 BC, is the other candidate.  It certainly names Israel and Omri.  Line 31 is broken, and the argument is over what stood in the break.

The physical situation matters and is rarely explained.  The bottom of the stele was shattered in 1869, but a paper squeeze had been taken beforehand.  The stone was then reassembled with plaster filling the gaps.  So there are three things to look at: the surviving original stone, the plaster restoration, and the squeeze.

André Lemaire, examining the squeeze in 1994, proposed restoring bt[d]wd — House of David — and read it as evidence that David's dynasty controlled southern Moab.  Kitchen, Rainey, Aḥituv and Weippert adopted the reading.⁴

Finkelstein, Na'aman and Römer published against it in 2019, and their objections are not only epigraphic.  Working from a new Louvre high-resolution photograph of the squeeze taken for a 2018 exhibition, they report:

1. The taw that follows the beth in Lemaire's rendering of bt[d]wd does not exist.
2. More importantly, before the waw of b[--]wd a vertical stroke appears, that—like many similar strokes in the stele—marks a transition between two sentences …
3. The letter after the waw is indeed a dalet, the left side of which is slightly damaged.⁴

If the vertical stroke is real, the waw starts a new sentence and cannot be part of a name beginning with beth, which dissolves the reading entirely.  They also note that "the two letters after the beth were already eroded when the squeeze was produced," so the squeeze cannot supply them.⁴

Their three historical objections are the stronger half of the argument.  First, the Tel Dan bytdwd names a kingdom, whereas Lemaire's Mesha reading has to name a dynasty — a different use.  Second, why would Mesha refer to an enemy by a collective term when he names Omri by name in lines 4–5?  Third, and most substantively: "not only is there no evidence for the assumed possession of southern Moab in the late 9th century BCE by the weak Jerusalem dynasty, the assumed Judahite conquest of the area south of the Arnon River contradicts the narrative in 2 Kings 3, which recounts a failed campaign of the kings of Israel and Judah to this territory."⁴

That last point deserves attention from anyone using the Mesha reading apologetically.  It says the "House of David" restoration puts Judah in control of a region the Bible itself says Judah failed to take.

Lemaire and Jean-Philippe Delorme answered in 2022 in Biblical Archaeology Review, reporting Reflectance Transformation Imaging of both the stone and the squeeze and concluding that the damaged letter is a dalet and the reading stands, with independent support from Michael Langlois.⁵

Where this leaves the reader, and this article.  RTI is a genuine methodological advance over photographs, and two epigraphers working independently is meaningful.  But this article has read the 2019 paper in full and has only press reporting of the 2022 reply, and the 2022 reply as reported addresses the letter and not the three historical objections.  The Mesha stele should not be cited as a second attestation of the House of David until that exchange is read in full.  The Tel Dan stele does not need it.

The Karnak list

Shoshenq I — biblical Shishak — campaigned in the Levant around 925 BC, and his topographical list at the Bubastite Portal at Karnak names roughly 150 places.¹⁵

Jerusalem is not on the list.  This is a genuine problem for the naive apologetic use of the relief as "confirmation" of 1 Kings 14:25.  The list also concentrates on places north and south of Judah rather than in the Judaean highlands.

But the biblical account does not describe a sack of Jerusalem.  1 Kings 14:26 has Rehoboam handing over the treasuries of the temple and the palace.  A city that pays is a city that is not stormed, and pharaohs listed conquests, not payments.  The strongest thing that can be said is that an Egyptian campaign in the southern Levant at approximately the right date is independently attested, and that the biblical and Egyptian records neither confirm nor contradict each other in detail.

One conservative claim about the list should be rated low, and its history is instructive.  Champollion read one name-ring — number 29 — as "Judah the Kingdom," and took it as proof that Shoshenq had taken Jerusalem.  In 1888 W. Max Müller showed it should be read yad-ha-melek, "Hand of the King," and that became the standard reading.²⁴  The claim is nonetheless being revived: Gavin Cox, in the Journal of Creation in 2022, proposes a new hieroglyphic translation of ring 29 (transliterated y-w-d-h-m-r(w)-k) indicating tribute from Judah.²⁵  A reading overturned in 1888 and re-proposed in a creationist journal in 2022 without acceptance in Egyptology is not evidence.  It is recorded here because the claim circulates and a reader should know what happened to it.

One conservative point about the list is fair, though.  Cox observes that Judean towns paying tribute through Jerusalem would not appear individually in a list of conquests, since the tribute was collected centrally.²⁵  That does not turn the list into corroboration.  It does mean the absence of Jerusalem cannot be pressed as a contradiction of 1 Kings 14:26, which describes payment rather than capture.

Part 3 — Jerusalem in the ground

This is where the argument has been sharpest, because Jerusalem is the hardest site in the region to excavate.  It has been continuously occupied for three thousand years, it is built on bedrock that Iron Age foundations reach down to, and large parts of it are under houses, roads and holy places.

The old argument

Finkelstein's position for two decades was that tenth-century Jerusalem was a small village, that the "united monarchy" was a memory of a later northern kingdom projected backwards, and that Judah did not become a state until the eighth century.  The evidence was negative: no tenth-century monumental architecture had been found.

Eilat Mazar and the Large Stone Structure

In 1997 Eilat Mazar reasoned from 2 Samuel 5:17, which has David going down to the stronghold, that the palace should lie north of the oldest settlement core, and in 2005 she began digging there.  Under Byzantine and Second Temple layers she found large foundations, later shown to include walls up to 7 metres thick, and she identified them as David's palace.²⁶

Her dating argument is stratigraphic: Iron Age I pottery below the foundations, Iron Age II pottery above, therefore construction between the two.²⁷

The criticism is serious.  Israel Finkelstein, Ze'ev Herzog, David Ussishkin and Lily Singer-Avitz published jointly in 2007 arguing that the walls do not belong to one building at all — that the regular western walls line up with a Hasmonean-period structure including a ritual bath, while the eastern remains are a separate and flimsier entity — and that Mazar's approach was biased.¹⁹  Finkelstein's verdict on the dating was that "all one can safely say is that its various elements post-date the late Iron I/early Iron IIA and predate the Roman period."²⁸

This project's assessment.  Mazar's find is real and important;  her identification of it as David's palace is not established, and the manner of the announcement — a site chosen from a verse, an identification declared at the outset, funding from an ideologically committed foundation — is the kind of thing that ought to lower a reader's confidence regardless of which side they are on.  The structure should be cited as a substantial early Iron Age building in the City of David whose extent, unity and precise date are disputed.

What the carbon says

The strongest evidence now available is the 2024 PNAS study by Regev and colleagues — a Weizmann, Tel Aviv, IAA and Bar-Ilan collaboration — presenting 103 radiocarbon dates from five excavation areas.⁶

Its method deserves a note, because it solves a problem that has bedevilled this whole field.  The radiocarbon calibration curve has a flat stretch, the Hallstatt Plateau, across part of the Iron Age;  a single date in that stretch calibrates to a range 300 years wide, which is useless.  The team got around it by dense sampling of stratified sequences, combined with microarchaeology to verify that each sample really belonged to its context, and by anchoring against two historically fixed events: the mid-eighth-century earthquake and the Babylonian destruction of 586 BC.  The 586 anchor was checked against 100 single-ring measurements from calendar-dated tree rings between 624 and 572 BC.⁶  The result, in their words, is that "instead of calibrated ranges spanning 300 y, we succeeded in reaching decadal precision."⁶

Their four numbered conclusions are worth reading as a whole, because each side quotes one of them.

1. We provide concrete evidence for a widespread human presence in the city during the 12th to 10th centuries BC, as short-lived material from all the areas studied was dated to these centuries.  However, these contexts were often lacking diagnostic ceramics.

2. The city's westward spread out of its ancient core is assigned here to at least as early as the 9th century BC, as attested by the radiocarbon dating of the construction of monumental Building 100 in excavation Area 10 (GPL) and the large-scale terracing activities of the bedrock done prior to its construction.

3. Major changes in the city's urban plan were shown to follow the devastating earthquake of c. 760 BC …

4. Ongoing development of the city is documented between the 8th and the 6th centuries BC, until the destruction in 586 BC.⁶

Point 1 is the one conservatives should note: tenth-century Jerusalem was inhabited across the whole excavated area, and the reason this was missed is stated in the paper — those contexts often lacked diagnostic pottery, which is what earlier surveys relied on.  Point 2 is the one skeptics should note: the monumental phase begins in the ninth century, with Building 100's construction constrained to 900–850 BC.⁶

Neither point is a victory.  Together they describe a real tenth-century town that became a monumental capital about a century later.

Finkelstein's reply

Israel Finkelstein, whose position the study's first conclusion undercuts, posted observations on it in May 2024.  His opening is not what a reader might expect:

This article is important, as it reports the first massive (103 samples), systematic project of radiocarbon dating Iron Age remains in Jerusalem.  I focus here solely on the archaeological/historical issues.  Let me start by a comment on contexts: The reader of the article must acknowledge differences in context-reliability of samples sent to radiocarbon dating.  Short live samples from a sequence of superimposed floor accumulations, connected to architectural remains are the best candidates for radiocarbon dating.²⁹

His objection, in other words, is about context reliability rather than about the measurements: 103 samples do not all come from contexts of equal quality, and the best kind — short-lived material from stratified floor sequences tied to standing architecture — is the scarcest.  That is a fair methodological point and it applies to the study's first conclusion, which rests partly on scattered twelfth-to-tenth-century dates from contexts that the paper itself says "were often lacking diagnostic ceramics."⁶

This project has read only the first paragraph of Finkelstein's three-page response, which is what academia.edu serves without a login, and does not know how far he presses the objection.  What can be said is that the leading critic of tenth-century Jerusalem read this paper and called it important.

Part 4 — Khirbet Qeiyafa

Discovered in 2007 and excavated by Yosef Garfinkel and Saar Ganor, Khirbet Qeiyafa sits on a ridge above the Elah valley, on the border between Philistia and the Judaean hills, about 30 km from Jerusalem.  It covers about 2.3 hectares, six acres, inside a 700-metre casemate wall built of field stones, some of them weighing eight tons.⁷

Its importance is chronological.  Four burnt olive pits dated at Oxford, plus pottery, place the site at roughly 1020–980 BC on Garfinkel's refined chronology.³⁰  It is fortified, it was built and abandoned quickly, and it predates the fortifications at Lachish, Beersheba, Arad and Timnah.³¹  If it is Judahite, then a highland polity capable of building a planned fortified border city existed at exactly the time the biblical narrative places David.

The Israel Antiquities Authority put the implication bluntly: "The excavations at Khirbat Qeiyafa reveal an urban society that existed in Judah already in the late eleventh century BCE.  It can no longer be argued that the Kingdom of Judah developed only in the late eighth century BCE or at some other later date."³²

The dispute is over whose city it was.

Not Philistine, on two grounds that are actual evidence rather than inference.  The private houses abut the city wall, an arrangement not used in Philistine cities.³³  And the faunal remains are goats, cattle, sheep and fish;  Philistine sites are marked by pig and dog consumption.⁸

Canaanite, argued by Nadav Na'aman and Ido Koch from similarities with nearby Beth Shemesh.

North Israelite, argued by Finkelstein and Fantalkin, who write that "There is no evidence for arguing that Jerusalem, Hebron and Khirbet Qeiyafa were the main centres of 10th century Judah … Between the two possibilities for the territorial affiliation of Khirbet Qeiyafa with a highlands polity—Judah or an early north Israelite entity—the latter seems to us the more attractive one."⁹

Note what that concedes.  Finkelstein and Fantalkin are not disputing that Qeiyafa is a tenth-century fortified city built by a highland polity.  They are disputing which highland polity.  That is a much smaller disagreement than the one they were having in 2001, and the movement has been in one direction.

Garfinkel's own claim is much stronger than this article endorses, and the reply to it is the interesting part.  Writing in Biblical Archaeology Review in 2011 under the title "The Birth and Death of Biblical Minimalism":  "The argument that Judah was an agrarian society until the end of tenth century B.C.E. and that David and Solomon could not have ruled over a centralized, institutionalized kingdom before then has now been blown to smithereens by our excavations at Khirbet Qeiyafa."³⁴  His reasoning is that the fortifications required organised labour beyond a chiefdom, and that a trained scribe posted to a remote border town implies a state that could train and station scribes.  Finkelstein and Silberman had themselves written that "the power of the chief was limited … The economic and military capacity of a chiefdom was severely limited."³⁵

Notice what Finkelstein and Fantalkin did not say in reply.  They did not argue that Qeiyafa was a chiefdom's work.  They argued it was a north Israelite state's work.  The dispute has moved from "chiefdom or state" to "which state," and that shift is the honest measure of what has happened to the minimalist position since 2001.

Aren Maeir, excavating Philistine Gath nearby, gives the properly cautious version: "The finds have not yet established who the residents were."³⁶

Two further claims should be held loosely.  Garfinkel identifies the site with biblical Sha'arayim, "two gates," on the strength of a second gate.  Yehuda Dagan disputes the identification, and Na'aman doubts that Sha'arayim means "two gates" at all, noting that the -ayim ending in place names is generally not the dual.³⁷  And the Qeiyafa ostracon, an inked five-line inscription, has produced several mutually incompatible readings and should not be used to carry an argument.

Part 4b — Timna, and the argument that changes the terms

The most interesting recent development in this debate does not come from Jerusalem, from an inscription, or from an apologist.  It comes from a copper-mining camp in the Aravah desert, and it is an argument about method.

What was found

Erez Ben-Yosef of Tel Aviv University has excavated the Iron Age copper-production sites of the Aravah — Timna on the Israeli side, with Thomas Levy's Khirbet en-Nahas and the Faynan district on the Jordanian side.  The sites were long assigned to the Egyptian New Kingdom, thirteenth and twelfth centuries BC.  High-precision radiocarbon on organic material, including remarkably preserved livestock dung, moved the industry's floruit to the tenth century BC.

Ben-Yosef's own description of the moment: "We thought maybe some nomads had camped there with their goats a few decades ago … But the [radiocarbon] dates came back from the lab, and they confirmed we were talking about donkeys and other livestock from the 10th century B.C.  It was hard to believe."¹⁰

The dung also carried seeds and pollen well enough preserved to reconstruct the animals' diet, and the feed had been brought from more than a hundred miles north, near the Mediterranean coast.¹⁰  Industrial-scale smelting in a desert, supplied over that distance, implies an organisation capable of running it.

Why it matters more than the usual apologetic use of it

The obvious apologetic move is to connect Timna to 2 Samuel 8:13–14 and 1 Kings 9:26–28 and call it Solomon's mines.  That move is available and it is not what makes this important, because nothing found at Timna names an Israelite.

What makes it important is that the polity running these mines left almost no architecture.  Ben-Yosef drew the methodological conclusion in Vetus Testamentum in 2019, and it is aimed at everybody:

the total reliance on stone-built archaeological features in the identification of social complexity in the vast majority of recent studies has resulted in skewed historical reconstructions.  Recognizing this "architectural bias" and understanding its sources have important implications on core issues in biblical archaeology today, as both "minimalists" and "maximalists" have been using stone-built architectural remains as the key to solving debated issues related to the geneses of Ancient Israel and neighboring polities (e.g., "high" vs. "low" Iron Age chronologies), in which—according to both biblical accounts and external sources—nomadic elements played a major role.¹¹

He is explicit that the flaw "cuts across all schools of biblical archaeology."¹¹

What this does and does not do for the united monarchy

It does undercut a specific inference.  The Low Chronology's force comes largely from moving monumental buildings out of the tenth century, on the reasoning that a state without monumental architecture is not a state.  Ben-Yosef's Aravah polity ran an international copper industry, controlled territory, and organised long-distance supply, while leaving the kind of remains that a survey would read as pastoral camps.  So "no palace, therefore no kingdom" is not a valid step, and that is a mainstream finding rather than an apologetic one.

It does not supply a Davidic kingdom.  An argument that a state could be archaeologically invisible is not evidence that this one existed.  It converts a negative argument into a non-argument;  it does not create a positive one.  Ben-Yosef's own extension of the idea — that David's monarchy may have been substantially tent-dwelling and equally complex — is a hypothesis he offers, not a result he demonstrates.

And it should be applied evenhandedly.  If absence of architecture cannot refute a tenth-century Judahite state, it also cannot refute a tenth-century north Israelite one, which is what Finkelstein and Fantalkin propose for Qeiyafa.  The argument dissolves a class of inference for everybody.

The Ophel pithos inscription

One further item belongs here and should be reported with its limits.  In 2012 Eilat Mazar found seven broken pithoi in a bedrock fill beneath a monumental building in the Ophel, one carrying a fragmentary eight-letter inscription just below the rim.  Douglas Petrovich published an analysis in Palestine Exploration Quarterly, reading it as a wine docket — "[In the firs]t [(regnal) year]: pseudo-[win]e from [the garden of ??]" — and dating it on script and vessel type to the tenth century.³⁸

Rate this low.  Eight letters, fragmentary, with a proposed reading that supplies most of the words in brackets.  The inscription's language has been read as Hebrew, as Canaanite, and as neither;  Petrovich's is one reading among several.  It is worth knowing that the oldest inscription from Jerusalem is probably tenth-century.  It is not worth building on.

Part 5 — The chronology fight

Everything above sits on top of a technical dispute about when Iron Age IIA begins.

Finkelstein's Low Chronology, published in 1995 and 1996, moves stratigraphic dates down by up to a century.  Its effect on this question is direct: the monumental buildings traditionally credited to Solomon become ninth-century, the work of the Omride dynasty.  Finkelstein states the consequence himself: "Accepting the Low Chronology means stripping the United Monarchy of monumental buildings, including ashlar masonry and proto-Ionic capitals."¹²

Amihai Mazar's Modified Conventional Chronology places Iron IIA from the early tenth to the mid-ninth century, which keeps the relevant material in the tenth without the problems of the older high chronology.³⁹  Finkelstein's low chronology has drawn sustained criticism from Amnon Ben-Tor, William Dever, Kenneth Kitchen, Doron Ben-Ami, Raz Kletter and Lawrence Stager.⁴⁰

Mazar's own account of what his chronology does is more candid than the use conservatives make of it, and it is the most useful thing in the whole dispute.  In the volume where he and Finkelstein debate each other directly, he writes:

The outcome of the Modified Conventional Chronology is that both the United Monarchy and the Omride dynasty are included in a single, more lengthy archaeological period, denoted Iron IIA.  This revised periodization in turn creates a greater flexibility in the interpretation of the archaeological data, and makes life more difficult for those who wish to utilize archaeology for secure historical interpretations in the southern Levant during the tenth and ninth centuries B.C.E. … The modified chronology would mean that any comprehensive archaeological synthesis of the Solomonic period is tentative and may be interpreted in more than one way.  And finally, it rejects the strict and one-sided Low Chronology as suggested by Finkelstein.¹³

Read that twice.  The MCC does not restore the Solomonic attributions.  It merges the tenth and ninth centuries into one archaeological period, which means that for most buildings one cannot decide between Solomon and Ahab.  Mazar says so explicitly, and gives the reason: at Megiddo the southern of Yadin's two "Solomonic" palaces survives only to its foundation courses, and the published pottery from the northern one "does not allow such a subtle dating inside the boundaries of the Iron IIA period."¹³

So the conservative use of Mazar against Finkelstein is half right.  He does reject the Low Chronology as "strict and one-sided."  He does not thereby hand back the Solomonic buildings;  he makes their attribution undecidable on present evidence.

And Mazar's own picture of the kingdom is a moderate one.  In the editors' summary of his position: the United Monarchy was "a state in the early stage of development far from the imperial state portrayed in the Bible.  Its capital was more like a medieval European burg surrounded by a medium-sized town."  He also holds that the Tel Dan inscription "clearly eliminates … the notion that David and Solomon were the mere invention of later biblical authors."⁴¹  That is close to this article's own verdict, and it is worth noting that the best-regarded middle position in the field arrives there independently.

Finkelstein's reply is an accusation of selection, and it should be recorded: he argues that "a more thorough reading of the Tel Rehov radiocarbon data (that is, taking into consideration all readings rather than a selection of results) can be interpreted as supporting the down-dating of the Iron Age strata."¹⁶  This project cannot adjudicate that without the underlying dataset, and says so.

What a non-specialist should take from this.  The chronology dispute is genuine, technical, and unresolved, and it is decided by radiocarbon at sites like Tel Rehov rather than by anybody's view of the Bible.  The one thing that should not be done is to treat either chronology as established and then present its consequences as archaeological findings.  Much popular writing on both sides does exactly that.

It is also fair to record that the movement over thirty years has been toward Mazar and away from the strongest form of the low chronology, and that Finkelstein's own later reconstruction concedes a great deal.  In The Forgotten Kingdom (2013) he proposes that Saul built a polity reaching from Jezreel to Hebron, centred on Gibeon, which Shoshenq destroyed.  Dever's response is that a polity of that extent "is a united monarchy in its own right," which if right means the disagreement is now about the capital and the king's name rather than about whether a tenth-century state existed.⁴²  Dever's language about Finkelstein is intemperate and this article does not endorse the tone, but the structural point stands.

Part 5b — What Finkelstein actually concedes

This article has been treating Finkelstein largely through his critics.  Read directly in the volume where he and Mazar argue side by side, his position is narrower than its reputation and concedes more than conservatives usually allow.

The reference to the Davidic dynasty … in the Tel Dan inscription of the ninth century B.C.E. leaves no doubt that David and Solomon were historical figures.  And there is good reason to accept that many of the David stories in the books of Samuel—mainly the heroic tales and the description of his life as a bandit on the fringe of the Judean highlands—contain genuine, early historical memories.  These in turn were transmitted orally and put in writing not before the late-eighth century B.C.E.  But the great biblical story of the United Monarchy is left with no material evidence.  In the tenth century B.C.E., places such as Megiddo in the north still featured Canaanite material culture.  The kingdom of David and Solomon was no more than a poor, demographically depleted chiefdom centered in Jerusalem, a humble village.¹⁶

Three things in that paragraph are worth separating.

He grants the existence question outright — "leaves no doubt" — and grants it on the same inscription this article treats as decisive.

He grants that the Samuel material contains "genuine, early historical memories," and specifies which: the bandit narratives, David on the fringe of the highlands.  That is a substantial concession about the tradition's antiquity, and it is one the popular presentation of Finkelstein never mentions.

He denies the scale, and only the scale.  "The great biblical story of the United Monarchy is left with no material evidence."

So the disagreement between Finkelstein and this article is narrower than the labels suggest.  Both accept David and Solomon as historical.  Both accept the Tel Dan inscription.  Both deny the empire of 1 Kings 4:21.  What remains is the size of the polity — chiefdom or modest state — and where its capital sat on that scale.

And on that, one of his supporting claims has since been tested.  He wrote in 2007 that Jerusalem was "a humble village."  The 2024 radiocarbon study reports "concrete evidence for a widespread human presence in the city during the 12th to 10th centuries BC, as short-lived material from all the areas studied was dated to these centuries."⁶  Not a capital of an empire — but not a village either, and the finding post-dates his statement.

One more thing the editors of that volume record, which bears on the method article: Finkelstein "disagrees with both the maximalists and the minimalists as to the facts of history."¹⁷  He is routinely described as a minimalist and he is not one in the Copenhagen sense.  The label obscures more than it conveys.

Part 6 — How big was the kingdom?

Here the article has to say something a conservative reader may not like.

1 Kings 4:21 describes an empire.  "Solomon ruled all the kingdoms from the Euphrates River to the land of the Philistines and to the border of Egypt."  1 Kings 4 and 10 go on to describe an administrative apparatus, a standing chariot force, and international trade on a scale that would make tenth-century Israel a major regional power.

Nothing supports this.  Not the archaeology of the Levant, not Egyptian records, not the Assyrian records that begin to describe Levantine states in the ninth century, not the Aramaean inscriptions.  The first extra-biblical notice of an Israelite king in a major imperial record is Ahab at Qarqar in 853 BC, in Shalmaneser III's Kurkh Monolith, and Ahab is a ninth-century northern king, not a tenth-century united one.

The available answers.  One is that the language is ordinary ancient royal hyperbole, the same genre in which Egyptian pharaohs claim to have subdued all foreign lands and Assyrian kings claim to rule the four quarters.  This is a legitimate reading and it is not special pleading — the same discount is applied to Egyptian and Assyrian texts as a matter of course, and refusing to apply it to Hebrew texts alone would be inconsistent.  The other is that these chapters are later idealisation.  A third, weaker option is that a brief hegemony left no trace, which is possible but unfalsifiable and should not be leaned on.

Kitchen's reply, which changes what the verse is claiming

The strongest answer to all of this is Kitchen's, and this article owes it a hearing because it does not defend the reading being attacked.

the scale and nature of the wider realm of David and Solomon are not unique and belong to a specific period of history, namely, ca. 1200–900 — neither earlier nor later.¹⁸

His argument is that between the collapse of the Late Bronze great powers and the rise of Neo-Assyria there was a window in which several mid-sized regional hegemonies existed — he works through Tabal/Tarhuntassa, Carchemish, and the Aramean centres at Zobah and Damascus — and that the biblical description of David's realm belongs to that class rather than to the class of Assyrian or Persian empire.¹⁸

The structural point is the good one.  These mini-empires, he argues, "took over defeated opponents and their territories as they stood, without much territorial change;  thus such units could readily break away as natural political entities when the central power weakened."  He contrasts this with "the vast Neo-Assyrian, Neo-Babylonian, and Persian empires, where (increasingly) new large-scale provinces became the norm."¹⁸

Why this matters for 1 Kings 4:21.  On Kitchen's reading the verse is not claiming provincial administration from the Euphrates to Egypt.  It is claiming hegemony over vassals and tributary allies who kept their own territories and rulers — which is what the same chapter describes when it has kings "bringing tribute" rather than governors reporting.  A hegemony of that kind reorganises nothing, builds nothing outside the core, and would therefore leave very little for an excavator.  That connects directly to Ben-Yosef's architectural-bias point:  the trace one is looking for may not be the trace such a polity leaves.

Three things it does not do.

It does not supply evidence.  Tabal, Carchemish and Aram-Damascus are attested in external sources.  David's hegemony is not.  Kitchen has shown that the type existed in the window, which removes the objection that the biblical picture is anachronistic — and an argument that a thing is possible is not an argument that it happened.

It concedes the deflation.  A mini-empire of vassals and tributaries is a considerably smaller claim than the one 1 Kings 4 and 10 read like in English, and Kitchen is agreeing that the maximal reading is wrong.

And the boundary phrase still overreaches on any reading.  "From the Euphrates to the border of Egypt" as a tributary sphere is still a very large claim for a polity whose capital, on the best current radiocarbon, was not monumental until the ninth century.

This article's revised position.  1 Kings 4:21 remains unsupported by external evidence.  But Kitchen has shown that the political form it describes is specific to 1200–900 and not to any later period, which makes the passage harder to explain as a late writer's invention than this article previously allowed.  The honest statement is that the verse describes a recognisable and correctly-dated kind of realm, that no evidence attests this particular one, and that the maximal English reading is wrong on everyone's account including Kitchen's.

Common objections

"David is a literary invention, like King Arthur."  The comparison is a commonplace of the minimalist literature.  It fails on one point: no Arthurian equivalent of the Tel Dan stele exists.  There is no ninth-century enemy inscription referring to "the House of Arthur," and there is one referring to the House of David.

"The Tel Dan stele might be a forgery."  Argued by Cryer and Lemche from chisel marks;  rejected because the fragments came from a controlled excavation, and softened by Lemche himself.²³

"Bytdwd doesn't have to mean House of David."  True, and the lack of a word divider is a real observation.  But the alternatives have to explain the parallel with the king of Israel in the same sentence, and the "House of X" dynastic formula is standard in the region.  Most epigraphers accept the reading.

"Tenth-century Jerusalem was an empty village."  Falsified in 2024 by 103 radiocarbon dates showing occupation across all five excavated areas in the twelfth to tenth centuries.⁶

"So the Bible's account of Solomon's Jerusalem is confirmed."  It is not.  The same study puts the city's monumental expansion in the ninth century and constrains Building 100 to 900–850 BC.⁶

"Khirbet Qeiyafa proves the Davidic kingdom."  It proves a tenth-century fortified city built by a highland polity with non-Philistine diet and architecture.  Whether that polity was Judah, an early north Israelite entity, or a Canaanite one remains disputed by serious people.

"The Solomonic gates at Megiddo, Hazor and Gezer settle it."  Their date is exactly what the low-chronology dispute is about.  Yadin himself moved the Megiddo "stables" from Solomon to Ahab.¹⁴  Citing them as settled evidence assumes the answer.

"Minimalists apply a standard they would use nowhere else."  There is something to this — few historians would deny Alexander's existence on comparable evidence — but it is a weaker argument than it looks, because the disputed claims here are not about existence but about the scale of a state, and scale is exactly the kind of claim archaeology is good at testing.  On the existence question the minimalists lost, and on the empire question they have not been answered.

What would settle it

• A tenth-century monumental inscription from Jerusalem. Anything royal and datable would do more than all the stratigraphic argument combined. Judah's securely dated royal inscriptions currently begin later.
• Resolution of the Iron IIA chronology by radiocarbon. The Regev method — dense stratified sampling, microarchaeological context verification, historical anchoring — is the template. Applying it at Megiddo, Hazor and Gezer would decide the "Solomonic" gates.
• Excavation under the modern city. The single largest constraint on this question is that most of Iron Age Jerusalem cannot be dug. This is unlikely to change.
• A published faunal and epigraphic profile from Qeiyafa's contested contexts, adjudicating Judahite against north Israelite.
• Any tenth-century Levantine text mentioning a king in Jerusalem. Egyptian, Aramaean or Phoenician. None is known.

Sources and further reading

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