Joshua is the book where the biblical account and the excavated record diverge most visibly, and where the divergence has been argued about longest. It is also the book whose own sequel disagrees with it.
What the book is
Twenty-four chapters describing Israel’s entry into Canaan under Joshua son of Nun, the defeat of its cities, the division of the land among the tribes, and a covenant ceremony at Shechem. It opens the sequence — Joshua, Judges, Samuel, Kings — that runs from arrival in the land to exile.
The narrative is confident and rapid. The Jordan is crossed on dry ground, Jericho falls at a shout, a southern coalition is broken in one campaign and a northern one in another, and chapter 11 concludes that Joshua took the whole land and that the land had rest from war. Chapter 12 lists the defeated kings, in the manner of a royal victory inscription.
How it is put together
The book has two unlike halves joined at chapter 13. Chapters 1 to 12 are campaign narrative. Chapters 13 to 21 are administrative: tribal boundaries traced along ridgelines and watercourses, town lists, cities of refuge, settlements allocated to the Levites. Chapters 22 to 24 close with a dispute over an altar east of the Jordan and Joshua’s farewell. The lists behave differently from the narrative around them — they are what an administration produces, and the mainstream critical position is that they preserve real documents, with genuine disagreement about their period.
One feature matters for everything below: the narrative’s language of total destruction is formulaic, the same phrases recurring city by city. That is how ancient conquest accounts generally speak, including Egyptian and Assyrian ones about campaigns whose outcomes were more modest.
Where the external record touches it
No external document mentions Joshua or the campaign. The book names places rather than kings of empires, so the record it can be checked against is archaeological rather than textual — and it does not line up.
Jericho is the central case. The site has been dug repeatedly since the early twentieth century, and successive excavators reached opposed conclusions. An early campaign identified a collapsed defensive wall and connected it to chapter 6. A later and stricter excavation, working from pottery sequences, concluded that those fortifications belonged to a much earlier age, and that in the Late Bronze Age the settlement was small, weakly defended or unoccupied. That is now the majority reading, though still contested, partly because erosion has stripped much from the mound.
Ai is more awkward still. The site conventionally identified with it was a substantial Early Bronze Age city that then lay in ruins for centuries: on that identification there was no city to capture. The Hebrew name means something like “the ruin”, which some read as a sign that the story explains a conspicuous heap of stones rather than remembering a battle. Others argue the identification is wrong and propose sites nearby; that has not carried the field.
Not everything points the same way. Hazor, which chapter 11 singles out as head of the northern kingdoms and says was burnt, has a Late Bronze Age destruction layer with heavy burning and deliberate damage to statuary. That a great Canaanite city was violently destroyed in roughly the right period is not in dispute. Who destroyed it is: Egyptians, rival cities, incoming groups and internal upheaval have all been proposed, and a burnt layer carries no signature. Elsewhere the picture is mixed — some named cities show destruction, some none, several no occupation at the relevant time.
The other large fact from the ground is what appears rather than what is destroyed. In the early Iron Age many small unfortified villages spread across the central hill country, in areas previously thinly settled, and their material culture is broadly continuous with the Canaanite lowlands rather than foreign to it. This wave is generally accepted as the emergence of the population that becomes Israel. The disagreement is over where those villagers came from.
The competing models
Three classic reconstructions have dominated the discussion, and each deserves its strongest form.
- Military conquest. An outside group entered Canaan and took it by force, broadly as the book describes. Its strength is that it takes the text seriously and that Late Bronze Age destructions do exist. Its difficulty is the sites that are not there when needed.
- Peaceful infiltration. Pastoral groups moved gradually into the sparsely held highlands and settled, conflict coming later and locally. Its strength is the settlement pattern and the absence of a single horizon of destruction. Its difficulty is that it makes the conquest narrative almost wholly a later construction.
- Internal revolt. Israel emerged from within Canaanite society, as lowland populations withdrew from or rose against the city-state system and formed a new society in the hills. Its strength is that it predicts the cultural continuity rather than explaining it away. Its difficulty is that the revolt is inferred from theory, not found.
Most current work takes a fourth option overlapping the second and third: a gradual, largely indigenous emergence, drawing in displaced lowlanders and pastoralists and perhaps a smaller group carrying an exodus tradition from Egypt, whose story became everyone’s origin account. That is the mainstream position, and it too is a position — an inference from settlement patterns, not something a document states.
The tension with Judges 1
The most useful check on Joshua is not archaeological at all. It is the first chapter of the next book. Judges 1 works through the tribes and reports, with a repeated formula, that they did not drive out the inhabitants: not at Gezer, not at Beth-shean, Taanach, Megiddo or Dor. Judges 1 is not even consistent with itself on Jerusalem — verse 8 has Judah taking the city and burning it, verse 21 has the Jebusites still living there — and that internal seam is part of the evidence, not an embarrassment to be tidied away. The Canaanites remained, and in places were put to forced labour rather than destroyed.
That is hard to square with the summaries in Joshua 11 and 21, which state that the whole land was taken and not one promise had failed. It is also hard to square with Joshua itself: chapter 13 opens by observing that much land remains to be possessed, and chapter 15 concedes that Judah could not dislodge the Jebusites from Jerusalem. The tension runs inside the book as much as between the books.
Readings differ. The traditional harmonisation distinguishes a decisive initial campaign from a long and incomplete occupation, which the text can support. The critical reading treats the sweeping summaries as the later editorial frame and the piecemeal reports as older material it did not absorb. Both are positions. What is not in doubt is that the Bible preserves two accounts of the settlement that sit awkwardly together, and preserves them side by side.
Where to start
Read Joshua 11:16-23, then Joshua 13:1-7, then Judges 1 entire. Twenty minutes, and it puts the central problem in the text’s own words, before any spade is involved.
For what the silences above are and are not worth, see what external attestation means. Anything we have had to revise is recorded at corrections.