From ʿApiru to Hebrew

Some time around 1340 BCE the ruler of Jerusalem, a man named Abdi-Heba, wrote to Pharaoh in Akkadian, the diplomatic language of the age, and complained. His scribe was trained in cuneiform and glossed some words with their Canaanite equivalents. In letter after letter the complaint is the same. The land of the king is being lost, and it is being lost to the ʿApiru. “Why do you love the ʿApiru but hate the mayors?” he asks in one letter. In another: “The land of the king deserted to the Ḫapiru.” In a third he names the culprit closest to home, his rival in the hills to the north: “Are we to act like Labʾayu when he was giving the land of Šakmu [Shechem] to the Ḫapiru?”¹

The word Abdi-Heba used for his enemies had been in circulation for five centuries and would stay in Egyptian use for two more. It named a condition: people who had left or lost the town and the lord they were born under, and who now lived as hired soldiers, bonded workers, raiders or refugees. Three centuries after Abdi-Heba, the hill country he was losing held a people who called themselves Israel, one named group among several in the highlands, and who were called, in their own oldest laws and by their Philistine neighbours, ʿibrîm, Hebrews.²

The Hebrews emerged in part from the ʿApiru of the Late Bronze Age Levant, and they turned the ʿApiru’s status name, which meant a change of allegiance, into the name of a people Inference. Southern pastoralists joined them in the highlands Inference; whether they brought YHWH, and whether YHWH came from the south at all, is contested Contested. Their god belonged to a people and was worshipped through standing stones and a tent, without statues Inference. Their law came to be written in a script that had grown up outside the cuneiform schools Inference. After the Late Bronze collapse they settled the hill country beside Canaanite villagers Inference. Centuries later Israel’s scribes compressed that origin into the history of one family, using eponyms for groups and placing late law in ancient mouths. They shaped that history to make each settlement of the land legitimate, the last of them the return from Babylon Inference. Under Assyria and Babylon the same scribes drew on the language of the imperial loyalty oath for the covenant, made YHWH the only god, and wrote texts that refuse human kings; whether that refusal remembers the highland ʿApiru is untested Inference, untested.

George Mendenhall proposed in 1962 that Israel began as ʿApiru who rejected the Egyptian-backed rulers of the Canaanite cities, and Norman Gottwald built the idea into a sociology in 1979.³ The criticism disposed of their peasant revolt, their equation of the two words by sound law and their single founding moment. Amélie Kuhrt stated what remains: the hypothesis “does not depend crucially on the linguistic link.”⁴

What the scribes wrote about the ʿApiru

Every surviving word about the ʿApiru was written by someone else. They appear in the archives of palaces that hired them, the contracts of households that bound them, and the letters of kings who feared them. No text written by an ʿApiru describes being one.

The term appears in more than 250 texts, from the nineteenth century BCE to the twelfth, and from Anatolia and Iran to Egypt.⁵ It is written several ways: syllabic Akkadian ḫa-bi-ru or ḫa-pi-ru, the logogram SA.GAZ, alphabetic Ugaritic ʿprm, Egyptian ʿpr.w. The earliest appear in the Old Assyrian merchant texts of Kanesh in Anatolia, where they are prisoners and palace staff.⁶ At Mari on the Euphrates, in the eighteenth century, the term had become a status a man could enter. Terru, a king of Urkesh forced out of power, wrote that he had fled, using an idiom Daniel Fleming translates “I have left to live as an outsider” (ana ḫapirūtim attaṣi).⁷

At Nuzi in the fifteenth and fourteenth centuries, ʿApiru men and women bound themselves to wealthy citizens by contracts that Moshe Greenberg described as partaking “of the character of both slavery and adoption.”⁸ At Alalakh, census lists record units of ʿApiru soldiers attached to towns, and a clay prism from Tikunani on the upper Tigris counts 438 ʿApiru fighters by name in the service of its king.⁹ The ʿApiru were never one people. Greenberg’s survey found Akkadian and West Semitic names in an early Babylonian list, mostly non-Semitic names at Alalakh, a mixture at Nuzi, and an ease of absorption that “indicates that they were not distinguished by ethnic unity.”¹⁰ The statue of Idrimi from Alalakh tells, in a royal composition of disputed date, how an heir to a throne “resided among the displaced people for seven years” (the people written SA.GAZ) before taking back his kingdom.¹¹¹²

The Amarna letters of the mid-fourteenth century, sent by Canaanite and Syrian rulers to Pharaoh, are where the ʿApiru become a political subject. Fleming counts the term in more than fifty of the 350-plus letters.¹³ Rib-Hadda of Byblos writes of “the war of the ʿApiru,” and of towns and whole lands that “join” them, and he uses the word as an insult for Abdi-Ashirta and his son Aziru of Amurru, a highland polity with no city at its centre.¹⁴ In the south the most frequent target is Labʾayu of Shechem. In one of his own letters Labʾayu protests that he is no rebel, and in Fleming’s reading he writes that he “did not know that my son was going around with the ʿapiru.”¹⁵ Fleming argues that Labʾayu’s ʿApiru were a population with a standing relationship to his house, and that the scribes treated them, in three letters, almost as a polity, marking the word with the sign for a place.¹⁶

The same letters show that the word described employment as well as enmity. The ruler of Damascus, Biryawaza, a loyal vassal, assures Pharaoh that he is ready to serve “along with my troops and my chariots, along with my kin, along with my ʿapiru, and along with my sutû.”¹⁷ Amenhotep II’s Memphis stele lists 3,600 ʿApiru among the captives of a campaign, entered separately from 15,200 Shasu, Egypt’s word for the pastoralists of the steppe.¹⁸ A papyrus of Ramesses II’s reign, Leiden 348, orders: “Issue grain to the men of the army and (to) the ʿApiru who are drawing stone for the great pylon of the [...] of Ramesses.”¹⁹ As late as the reign of Ramesses IV, in the 1150s, a quarrying expedition to the Wadi Hammamat included 800 ʿApiru.²⁰ Fleming concludes that they “need not be detached from the social order.”²¹ The armed bands in the hills are one part of the record. Seti I’s smaller stele from Beth-Shean reports that the ʿApiru of Mount Yarmuta had risen against the people of a neighbouring town.²² Nadav Naʾaman gives the definition: ʿApiru “is a designation for people who were uprooted from their original political and social frameworks and forced to adapt to a new environment and way of life.”²³ Some were absorbed into palaces and households. Some, in the hills of Canaan, were never absorbed Inference.

ʿApiru and ʿibrî

The Hebrew word ʿibrî looks like ʿapiru. The objection is phonological and it is strong. Ugaritic writes the word ʿprm in an alphabet that cannot confuse p and b, and Egyptian writes ʿpr.w. The cuneiform sign bi is therefore read pi here, and the root is ʿ-p-r; “cross over” would need ʿ-b-r. Rudolf Borger in 1958, and Anson Rainey after him, added a second block: the cuneiform spellings show that one vowel of ʿapiru was long, and no regular development turns that form into ʿibrî. Rainey’s verdict, in 2008, was categorical: “There is absolutely no relationship!”²⁴ Fleming, who is more sympathetic to an ʿApiru background for Israel, agrees on the word. “Hebrew,” he writes, “does not derive from the old West Semitic term ʿapiru, though it may come from the same root and be indirectly related.”²⁵ He has proposed instead a derivation from the Mari term ḫibrum, used for the mobile part of a tribe; he set the two words side by side in 1998, and the full study he announced in 2012 has not appeared. That etymology would give “Hebrew” a tribal origin and bypass the ʿApiru, and Ki-Eun Jang’s 2023 article puts the comparison in print without reference to them.²⁶ The derivation from ʿapiru therefore stays contested, with Naʾaman and Oswald Loretz arguing for a transfer of the term and Rainey, Borger and Fleming against a sound-law descent Contested.²⁷ The thesis rests on use and context, which is weaker evidence than a sound law.

What survives the phonology is the use. Naʾaman’s 1986 study traced how ʿibrî behaves in the Bible, and the behaviour is that of a status word.²⁸ The oldest law in the Torah opens: “When you acquire a male Hebrew slave, he shall serve six years; in the seventh year he shall go free, without payment” (Exod 21:2). A Hebrew here is a man in bonded service, the arrangement Nuzi’s ʿApiru made with their patrons Inference. Deuteronomy rewrites the law as “your brother the Hebrew, or the Hebrew woman” (Deut 15:12, lit.), turning the status into kinship. By Jeremiah’s time the word needed glossing: the prophet explains that no one should keep “a Judean, his brother” as a slave (Jer 34:9, lit.).²⁹

The narratives keep the old sense in the mouths of outsiders. Potiphar’s wife calls Joseph “a Hebrew” in contempt, and the Egyptians will not eat with “the Hebrews” (Gen 39:14; 43:32). In 1 Samuel the Philistines call Saul’s men Hebrews coming “out of the holes,” and they ask of David’s band, “Who are those Hebrews?” (1 Sam 14:11; 29:3). In 1 Sam 14:21 “the Hebrews who had previously sided with the Philistines” go over to Israel in the battle. The Hebrew of that verse is uncertain and its sense is disputed; Naʾaman reads the Hebrews here as Israelites, others as a separate group.³⁰ Read as a separate group, it describes a body of fighters serving one master and then joining Saul, which is what the Amarna letters describe ʿApiru doing Contested. About half of the Bible’s uses of the word are spoken by foreigners or to them.³¹ Jang singles out the Hebrews of 1 Samuel 13–14 as exceptions to that pattern and finds them defined by their “collective capability of choosing and transferring political allegiance.” She compares them with the Mari ḫibrum and with the Philistine commanders’ question about David’s men in 1 Sam 29:3.³² Fleming had read the Amarna ʿApiru in similar terms, as whole populations whose allegiance could pass to Abdi-Ashirta or to Labʾayu.³³

Fleming also notes a Mari verb, ḫabārum or ʿapārum, whose labial is ambiguous in cuneiform, which “always has to do with departure from one’s home to take up residence in a different place, especially a different political entity.”³⁴ If that verb stands behind ʿapiru, the category meant something like “those who changed allegiance” Inference. Jang’s reading of the Hebrews of 1 Samuel reaches the same sense from the biblical side, and the two together point to allegiance as the content of the name Inference. Israel heard the same name as a crossing. In 1 Sam 13:7 the narrator plays on the word: “Some Hebrews crossed (ʿibrîm ʿābrû) the Jordan.” In Josh 24:2 the ancestors lived bəʿēber hannāhār, literally “beyond the River,” and Genesis makes Eber, whose name means “the far side,” the ancestor of all Hebrews (Gen 10:21). When Jewish translators in Alexandria rendered “Abram the Hebrew” in Gen 14:13, they wrote ho peratēs, “the emigrant” in the standard English translation, literally one who crosses over.³⁵

Egyptian evidence fixes the other end of the change. On the Merneptah stele of about 1208 BCE, Canaanite cities carry the sign for a foreign land, and “Israel is laid waste, having no seed” carries the sign for a people.³⁶ Egypt also went on using “ʿApiru” for labour gangs half a century later, so the two labels coexisted and named different things.

The strongest objection to an ʿApiru origin is that Merneptah’s Israel had another root. Egyptian texts from the fourteenth to the twelfth centuries place the Shasu in the southern and eastern steppe, living in tents, and Egypt counted them apart from the ʿApiru. Most recent work on Israel’s origin looks to them. Rainey argued that early Israel came from them, and read the Karnak relief of Merneptah’s campaign as showing Israel in Shasu dress.³⁷ Avraham Faust argues that Merneptah’s Israel was largely made up of Shasu pastoralists, joined by Canaanites and perhaps a small group from Egypt; Manfred Bietak and Gary Rendsburg trace Israel’s core to the Shasu of Edom and Seir Contested.³⁸ On their reading the ʿApiru are unnecessary. The answer is that the highland coalition may have held both: southern herders who gave it much of its pastoral custom, and uprooted people from the collapsing city-state order who gave it its social name Inference. Whether the god came with the herders is contested Contested. Bietak and Rendsburg, who reject an identification, leave the door open: “While some ultimate connection cannot be discarded, given the range of dates and places recorded for the ʿApiru/Ḫabiru, it is clear that not all these people can be Hebrews in any way, shape, or form.”³⁹

Excursus, a hypothesis about older roots: the Amorite tribes and the priests before them

The connecting claims below are INFERENCE, and nothing elsewhere in the argument depends on them.

One line of thought, argued at length in the essay Priests Before Kings, carries the ʿApiru back to the empire of Akkad and to an older class of ritual specialists. The ʿApiru are first attested in the nineteenth century BCE. Akkad, the first empire to claim the whole of Mesopotamia, fell around 2154 BCE. No text links the two, and the Sumerian word SA.GAZ, older than either, meant “robber” in the third millennium.⁴⁰

What the gap does hold is the Amorites. The kings of Ur built a wall against them, which Shu-Sin named “Muriq-Tidnim,” “Holding Tidnum at a Distance,” around 2034 BCE, and Sumerian city scribes mocked the Amorite who “lives in a tent, exposed to wind and rain.”⁴¹ By the eighteenth century Amorite dynasties ruled Babylon and Mari, and the Mari archive shows tribal confederations governing through elders and assemblies alongside their kings.⁴² One clan could be split into “the residents of Appan” and the ḫibrum ša nawêm, “those transhumant in the steppe.”⁴³ Treaties were made by “killing a donkey foal.”⁴⁴ The vocabulary of this world reaches the Bible: Mari nawûm and Hebrew nāweh, “pasture camp”; šāpiṭum and šōpēṭ, “judge”; naḫālum and naḥălâ, “inheritance.”⁴⁵

The hypothesis is that the Amorite tribal world, with its councils, covenants and pastoral law, is the matrix from which the Late Bronze highland ʿApiru drew their custom Inference. Within that world a class of specialists may have carried law and ritual between tribes and cities without belonging to either Inference. Priests Before Kings makes such a class older still: the Sumerian en, a priest-ruler whose authority rested on ritual knowledge, was captured by the warrior kings, the lugal, and Sargon of Akkad’s installation of his daughter Enheduanna as en-priestess at Ur shows the capture.⁴⁶ If an uncaptured remainder of that class survived among the tribes, the Levites, landless and dispersed among the tribes, would be its Israelite descendant Inference.⁴⁷ The test of the hypothesis is a positive marker across the gap: a Middle Bronze text naming ʿApiru or SA.GAZ in a ritual or legal role. None is known.

El, YHWH and the tent

The name Israel contains the god El. Mark Smith draws the inference: “The original god of Israel was El.”⁴⁸ The Torah preserves the memory. In Exodus 6:3 God tells Moses: “I appeared to Abraham, Isaac, and Jacob as El Shaddai, but I did not make Myself known to them by My name YHWH.” Genesis attaches El epithets to local sanctuaries: El Elyon at Salem, El Olam at Beersheba, and at Shechem the altar Jacob names El, God of Israel (Gen 14:18; 21:33; 33:20). These are gods of places.

YHWH came from somewhere else. The oldest poems in the Bible bring him from the south: from Seir and “the country of Edom” in the Song of Deborah, from Sinai, Seir and Mount Paran in Deuteronomy 33, from Teman in Habakkuk 3 (Judg 5:4–5; Deut 33:2; Hab 3:3). An inscription at Kuntillet ʿAjrud in Sinai, around 800 BCE, invokes “YHWH of Teman.”⁴⁹ Two Egyptian topographical lists, at Soleb under Amenhotep III and at Amara West under Ramesses II, name a “land of the Shasu of Yhw,” and these lists are often cited as the earliest trace of the god. Faried Adrom and Matthias Müller find no positive evidence that Yhw there is a divine name, and Racheli Shalomi-Hen, reviewing all the Egyptian material, concludes that Egyptian scribes did not know Yhw as a god.⁵⁰ Christian Frevel argues that the Omride kings of the northern kingdom introduced YHWH in the ninth century and that Judah received him from the north Contested.⁵¹ The southern origin of YHWH remains the majority view, and it now rests on the biblical poems and the “YHWH of Teman” of Kuntillet ʿAjrud.

The Bible’s own account names the people who brought him. Moses is tending the flock of Jethro, “priest of Midian,” when he meets God at Horeb, “the mountain of God” (Exod 3:1). Jethro later leads a sacrifice that Aaron and the elders of Israel join (Exod 18:12). The Kenites, descendants of Moses’ father-in-law, go up with Judah into the Negev (Judg 1:16), and the Rechabites, whom Chronicles counts among the Kenites, still refused in Jeremiah’s day to build houses, sow seed or plant vineyards, and lived in tents “upon the land where you sojourn” (Jer 35:6–7; 1 Chr 2:55). This cluster of memories is the basis of the Kenite or Midianite hypothesis, revived by Joseph Blenkinsopp in 2008.⁵² Fleming rejects the hypothesis. The biblical texts, he concludes, show “a persistent and perhaps surprising sense of kinship with such pastoralist neighbors,” and they “do not indicate that these were the first peoples to worship Yahweh”; on his account “Yahweh was not taken from outsiders.” He reads Yhw in the Egyptian lists as one of the herding peoples of the inland Levant, listed “without a particularly southern location,” and YHWH as the god of a people without kings Contested.⁵³ Juan Manuel Tebes rejects any model in which one migrating group carried the god north, and reads the southern evidence as a long process of cult practice among the peoples of the Negev, southern Transjordan and the northern Hejaz Contested.⁵⁴ Both accounts strike at the claim that southern herders brought the god to the highland coalition; that claim now rests on the biblical poems and Kuntillet ʿAjrud alone Contested.

The Bible’s god also travels. When David proposes a temple, the oracle refuses in the god’s voice: “From the day that I brought the people of Israel out of Egypt to this day I have not dwelt in a house, but have moved about in Tent and Tabernacle” (2 Sam 7:6). The tent shrine has Bronze Age parallels. Fleming published a Mari text describing a large public tent built on wooden frames, qersū, cognate with the qərāšîm, the boards of the tabernacle.⁵⁵ At Timna in the southern Arabah, after the Egyptians abandoned their Hathor shrine in the twelfth century, later users probably roofed the court with cloth and defaced the Hathor images, and they reused Egyptian altars in a row of standing stones.⁵⁶⁵⁷ And the tent was used by empires as well: the closest formal parallel to the tabernacle’s plan may be Ramesses II’s war camp at Qadesh.⁵⁸ The tabernacle tradition shares with the Mari and Qadesh tents a portable tent on a frame and with Timna a shrine of standing stones probably roofed with cloth, and the evidence does not say which of them it learned from.

The god of a people still had a place among other gods. The Septuagint and a Qumran manuscript of Deuteronomy preserve the older reading of Deuteronomy 32:8–9: when the Most High divided the nations, he fixed their borders “according to the number of the sons of God” (4QDeut^j; LXX), and YHWH’s portion was Jacob. Emanuel Tov judges that the Masoretic “sons of Israel” “reflects a conscious intervention in the text by a scribe uncomfortable with the polytheistic picture.”⁵⁹

Images were in use at Israelite shrines. At Kuntillet ʿAjrud the blessing is “by YHWH of Samaria and by his asherah.” A tomb at Khirbet el-Qom near Hebron blesses a man named Uriyahu by YHWH and “his asherah.” An Iron IIA temple at Tel Moẓa, a few kilometres west of Jerusalem, used human and animal figurines in its cult, and Hezekiah broke up a bronze serpent “that Moses had made,” to which the Israelites had been making offerings (2 Kgs 18:4).⁶⁰ Seung Il Kang, following Christoph Uehlinger, holds that “the biblical prohibition on images is largely a creation of biblical authors after the fall of Jerusalem” Contested.⁶¹ Tryggve Mettinger distinguishes de facto aniconism, the mere absence of images, from programmatic aniconism, their conscious repudiation.⁶² The first is old in the Levant and visible in the desert’s standing stones; Uzi Avner counts more than 450 prehistoric shrines with unshaped stones in the Negev and eastern Sinai.⁶³ The second is late. The thesis claims only the first for early Israel: worship without statues, through stones and an empty tent, carried by a people whose god moved with them Inference. Deuteronomy gives the ban a desert reason after the fact: “you saw no shape when YHWH spoke to you at Horeb out of the fire” (Deut 4:15).

Writing outside the cuneiform schools

The oldest alphabetic inscriptions were cut by Semitic speakers working for Egypt, at Wadi el-Ḥôl in the Western Desert and at the turquoise mines of Serabit el-Khadim in Sinai, where the best-read words are lbʿlt, “for the Lady,” and mʾhbʿlt, “beloved of the Lady.”⁶⁴ These early alphabetic writers honoured an Egyptian goddess under a Canaanite title, at a state mine.⁶⁵

In the Late Bronze Age the official writing of Canaan was cuneiform, learned through years of copying lexical lists in scribal schools; the Amarna letters themselves, including Abdi-Heba’s complaints about the ʿApiru, are in a Canaanite-inflected Akkadian that reflects that training.⁶⁶ The linear alphabet lived alongside it in small things.⁶⁷ Alice Mandell’s study of the Amarna scribes finds that the alphabet “was on the margins of institutional life in Canaan in the second millennium B.C.E.” When Egyptian administration withdrew and the palaces collapsed, it “filled this scribal vacuum.”⁶⁸ Naʾaman reads the same evidence differently and traces the alphabet’s spread to Egyptian centres and royal courts; on his reading too, what it stood outside was the cuneiform school Contested.⁶⁹

At ʿIzbet Ṣarṭah, a small settlement on the western edge of the hill country, a potsherd of the twelfth or eleventh century carries four lines of clumsy practice and then the alphabet in order: someone was learning to write in a village of the type in which Israel emerged.⁷⁰ Seth Sanders argues that what came out of that learning was new in kind. The early alphabet in the southern Levant served for dedications and ownership marks; in the ninth and eighth centuries Hebrew became a written vernacular, and the Bible became, in his phrase, “the first text to address people as a public.”⁷¹

The limit is literacy. Jack Goody and Ian Watt, whose 1963 essay founded the argument that the alphabet democratises, excluded Palestine from it: “Anything like popular literacy ... is not found in the earliest societies which used the Semitic writing system.”⁷² Handwriting analysis of the Arad ostraca, letters to a Judahite fort around 600 BCE, found a high probability of at least twelve writers in a garrison of perhaps twenty to thirty, which the authors take as “widespread literacy in the late 7th century BCE Judahite military and administration apparatuses.”⁷³ That is the literacy of officials. Christopher Rollston’s summary is blunt: “we have no evidence for the common folk writing and reading.”⁷⁴ Deuteronomy itself commands that the words be written “on the doorposts of your house and on your gates” (Deut 6:9), and it also has priests read the scroll aloud every seventh year “in the presence of all Israel,” men, women, children and resident foreigners, “that they may hear and so learn” (Deut 31:11–12). The doorposts brought the text into the house; the reading of it stayed a public act at the place God would choose, and the evidence suggests that most households could not read it.

The link between this history and the ʿApiru is an inference with no direct witness Inference, untested. The ʿApiru appear only in cuneiform and Egyptian texts written by their employers and enemies, and no alphabetic inscription names them. The test would be an alphabetic text from a context independently identified as ʿApiru. The script that served the post-collapse highland societies was the one that had lived outside the cuneiform schools, and Israel wrote its law in it Inference.⁷⁵

Egypt remembered as one family’s bondage

The Torah’s story of Israel in Egypt compresses movements that ran across most of the second millennium. A painting in the tomb of Khnumhotep II at Beni Hasan, around 1890 BCE, shows a group of ʿAamu, Asiatics, arriving with their animals and weapons under a leader titled “ruler of foreign lands.”⁷⁶ A papyrus now in Brooklyn lists the servants of a Theban household around 1750 BCE: about 45 of 95 have Asiatic names, and one of them is a woman whose name corresponds to Shiphrah, the name of a Hebrew midwife in Exodus 1:15.⁷⁷ At Avaris in the eastern Delta a Levantine population grew through the late Middle Kingdom and produced the Hyksos kings, whom Ahmose expelled around 1550 BCE; twelve centuries later the Egyptian priest Manetho, as quoted by Josephus, said the expelled shepherds went “through the wilderness, for Syria,” and built Jerusalem.⁷⁸ A frontier official of the late thirteenth century reports admitting “the Shasu tribes of Edom” past the fortress of Merneptah in Tjeku “to the lakes of Per-Atum,” Pithom, “to keep them alive and to keep their flocks alive.”⁷⁹ Ramesses III boasts that he destroyed the people of Seir “among the Shasu tribes,” razed their tents, and gave them as slaves to the temples.⁸⁰ Bondage in Egypt was a common fate of Levantine herders and labourers for half a millennium.

No evidence shows a single large exodus. Aren Maeir asks where the traces of a large group crossing the desert are, given that archaeologists “can identify the camps of small groups of ephemeral prehistoric hunter-gatherers”; Kadesh Barnea has no substantial occupation before the end of the Late Bronze Age.⁸¹ Ronald Hendel argues that the oppression remembered is the Egyptian empire itself, which ruled Canaan for roughly four hundred years and regarded all of it as Pharaoh’s property; former slaves from Egypt “joined the former ‘slaves’ who had never been in Egypt,” and together they became Israel.⁸² Bietak and Rendsburg write that “numerous eisodi and exodi occurred,” and that the tradition “is stylized as the descent to Egypt by a single extended family.”⁸³

One group may have carried the Egyptian strand. Moses, Phinehas, Hophni, Merari and Pashhur have Egyptian names, and they are all Levites; Richard Elliott Friedman argues that the Levites were the group that left Egypt and brought the exodus story and the god YHWH into Israel. Benjamin Sommer, reviewing the book, notes that Friedman’s source analysis is idiosyncratic and that some scholars find Egyptian names outside the tribe of Levi, while agreeing that they cluster there Contested.⁸⁴

The festival that commemorates the Exodus may be older than the story Contested. Julius Wellhausen argued in 1878 that Passover began as the shepherds’ offering of the firstborn of the flock, and that “the exodus is not the occasion of the festival, but the festival the occasion, if only a pretended one, of the exodus.”⁸⁵ Genesis and Exodus both frame Israel’s worship as a herder’s rite offensive to the empire: “all shepherds are abhorrent to Egyptians,” and “what we sacrifice to YHWH our God is untouchable to the Egyptians” (Gen 46:34; Exod 8:22). These are late texts, and they record Israel’s self-image as a pastoral people at odds with an imperial cult Inference.

Shechem and the highland villages after 1200

Egypt’s grip on Canaan loosened through the twelfth century. Ramesses VI is the last pharaoh whose name is attested there, on a statue base from Megiddo; excavations at Jaffa found the Egyptian fortress destroyed twice in the second half of the twelfth century, the last time around 1125 BCE, and the excavators read the destructions as local resistance to Egyptian rule.⁸⁶ As the garrisons left, settlers took over the hill country. Surveys from the Jezreel to the Beersheba basin found about 30 settlements of the Late Bronze Age and about 250 of Iron I.⁸⁷ The new villages were small, unwalled, without palaces or temples. Their houses followed the plan later called the four-room house, and their bone deposits have almost no pig.⁸⁸ None of these markers proves ethnicity. Israel Finkelstein, who once saw a sharp contrast with Canaanite pottery, later concluded that the pottery and the building plans of the Iron I sites “tell us nothing about ethnicity,” and William Dever writes that “early Israelites look ceramically just like Canaanites.”⁸⁹ Pig avoidance fits pastoralists and dry hill country as well as it fits a taboo, and it was common outside Philistia in the lowlands as well as the hills.⁹⁰

Finkelstein’s model is the main rival to the thesis, and at full strength it needs no ʿApiru. The settlers were local pastoralists who had lived in and around the hills through the Late Bronze Age, part of a long cycle in which herders settled when conditions allowed and dispersed when they did not. The village plans and the missing pigs are what settling herders leave. The ʿApiru of the texts were armed bands and displaced people around the cities and in their pay, and nothing ties them to the village builders.⁹¹ Brendon Benz offers a second rival: on his reconstruction the Late Bronze polities of the highlands, Labʾayu’s Shechem among them, were already run by many centres of power, and Israel continued them.⁹² The answer is that both models describe who built the villages and are silent on what they were called. Fugitives from collapsing towns who gathered in the hills were the kind of people an Amarna-period scribe had called ʿApiru Inference, and the clearest text placing ʿApiru in the hills places them at Shechem: Abdi-Heba accused Labʾayu of “giving the land of Shechem to the Ḫapiru.”⁹³ On this reading the ʿApiru were one strand of a mostly local population, which is Ann Killebrew’s “mixed multitude” Contested.⁹⁴ The answer has a weak point. Abdi-Heba used the word against rivals he disliked, and in one letter he complains he is himself treated like an ʿApiru.⁹⁵ The word was a weapon in his letters, and some of its force in the Shechem accusation may come from that use.

Shechem is where the Bible puts the covenant that made Israel. Judges names the god of its temple Baal-berith and El-berith, “lord” or “god of the covenant,” and remembers the city’s old ruling house as “the entourage of Hamor, the father of Shechem,” a name meaning “donkey” (Judg 9:4, 28, 46); Genesis and Joshua call them the kin and the children of Hamor (Gen 33:19; Josh 24:32). At Mari a treaty was “killing a donkey.” The chain from the Mari rite to Shechem’s donkey clan and its covenant god is attractive and unproven: each link stands alone, and none is confirmed by an inscription Inference.⁹⁶ Whether the city’s Middle Bronze fortress temple stood through the Late Bronze into Iron I, and when it fell, the excavation reports do not settle Contested.⁹⁷

Joshua 24 sets the covenant at Shechem and describes Israel as made by choice. Joshua tells the assembled people that their ancestors “lived beyond the Euphrates and worshiped other gods,” and orders them to “choose this day” whom to serve: the gods their ancestors served “beyond the Euphrates,” or “those of the Amorites in whose land you are settled” (Josh 24:2, 15). Noth, Mendenhall and Gottwald read the chapter as the memory of a covenant that incorporated new groups. Thomas Römer reads it as a Persian-period composition Contested.⁹⁸ If it is early, Israel remembered being formed by joining. If it is late, it shows how Persian-period authors chose to describe their ancestors: as people who had served other gods and crossed over to this one, which is evidence for the self-image and none for the Iron Age Inference. Shechem also shows how a family story absorbed group histories. Zev Farber traces three biblical versions of its taking: by Gideon’s son Abimelech, by Jacob “with my sword and bow” (Gen 48:22), and by Simeon and Levi in the story of Dinah (Gen 34).⁹⁹ Genesis tells the third version at length, and it makes the conquerors two of Jacob’s sons.

Genesis as a land title

Genealogies in oral societies record present relationships more than past events. Robert Wilson’s comparative study found that they expand, contract and reorder as alliances change, and that their functions are domestic, political and religious.¹⁰⁰ Wilson argued that the form of a genealogy follows present use; he did not argue that genealogies are invented. Read that way, the patriarchal genealogy is a political map. Finkelstein and Thomas Römer argue that the early Abraham traditions belonged to “the eponymous hero of the population of the southern highlands” in the later Iron Age, and that Abraham was joined to the northern Jacob as grandfather only later.¹⁰¹ Fleming argues that much of the Bible’s narrative is the inheritance of the northern kingdom of Israel, reinscribed by Judah after Israel’s fall.¹⁰²

The stories that fix the land claims are legal in form. In Genesis 15 God makes a grant of land “from the river of Egypt to the great river, the river Euphrates” (WEB) and lists the peoples who hold it. It also explains the delay: Abraham’s descendants will be strangers in a foreign land four hundred years, and they will return in the fourth generation, “for the iniquity of the Amorite is not yet full” (Gen 15:13, 16, 18–21, WEB). Moshe Weinfeld showed that the Abrahamic covenant follows the form of the ancient Near Eastern royal land grant.¹⁰³ Genesis 23 is a deed of sale: Abraham weighs out four hundred shekels to Ephron the Hittite for the cave of Machpelah at Hebron, in the presence of the Hittites, “all who entered the gate of his town” (Gen 23:10, 18). Genesis 33:19 is a second deed, for the plot at Shechem bought from the sons of Hamor for a hundred qəśîṭâ, the plot where Joseph’s bones were later buried (Josh 24:32). The two purchases sit in the two heartlands, Hebron for Judah and Shechem for the Joseph tribes. The curse that makes Canaan “a servant of servants to his brothers” (WEB) falls on Canaan, the son who did nothing; it was his father Ham who saw Noah’s nakedness (Gen 9:25).

The texts carry the marks of their real dates. The Philistines, who arrived in Canaan in the twelfth century, are already in Gerar in Isaac’s time (Gen 26:1). Abraham sets out from “Ur of the Chaldeans,” a people first attested in Assyrian records of the ninth century, whose dynasty ruled Babylonia from 626 BCE (Gen 11:31). The list of Edom’s kings is introduced as the kings who reigned “before any king reigned over the Israelites,” a phrase written after Israel had kings (Gen 36:31).¹⁰⁴ Editors also protected lineages. In Judges 18:30 the priest of the idol at Dan is Jonathan, grandson of Moses; the Masoretic text inserts a raised letter nun that turns Moses (Mšh) into Manasseh (Mnšh).¹⁰⁵

Most critical scholars place the Torah’s final shaping in the Persian period, and the strongest legitimation devices fit that period: the borders of Genesis 15 resemble the Persian province “Beyond the River,” and Abraham’s origin in Mesopotamia mirrors the returning exiles.¹⁰⁶ The prophet Ezekiel, in the 580s, quotes the people left in Judah after the Babylonian conquest: “Abraham was one, and he inherited the land; but we are many. The land is given us for inheritance” (Ezek 33:24, WEB). He rejects their claim. The Abraham story was already working as a land title, and two communities were already fighting over it. The Torah legitimated each settlement in turn: the highland villages, the Judah of Josiah, and the return from Babylon Inference. The text that survives carries mainly the last of these. Yonatan Adler’s survey of the archaeological evidence for Torah observance finds it widespread only after the Hasmonean revolt of the second century BCE.¹⁰⁷

Esarhaddon’s oath and the exile

In 672 BCE the Assyrian king Esarhaddon made his subjects swear an oath of loyalty to his chosen heir, Assurbanipal. In 2009 excavators at Tell Tayinat in southern Turkey found a copy sworn by the governor of an Assyrian province, in the inner sanctum of a temple, and Jacob Lauinger, who published it, argues that it was displayed there.¹⁰⁸ Deuteronomy reads in places like a loyalty oath to YHWH on the Assyrian model. Its curse list makes the sky over the disobedient bronze and the earth iron (Deut 28:23), the same assignment found in the treaty’s curses, and Deuteronomy 13 orders that a brother, son, daughter, wife or friend who entices one to serve other gods be denounced and killed, “Your hand shall be first on him” (WEB). Hans Ulrich Steymans argues that Deuteronomy 28:20–44 follows the order of the treaty’s long curse section, and Bernard Levinson and Jeffrey Stackert argue that Deuteronomy 13 revises the treaty clause requiring subjects to report any plotter, family included.¹⁰⁹¹¹⁰ On this reading Judah’s scribes transferred the oath from the Assyrian king to their own god, sometime between 672 and the “scroll of the Teaching” that the high priest Hilkiah found in the temple in Josiah’s eighteenth year, 622 BCE (2 Kgs 22:8). The book presents its laws, written centuries after the time it describes, as Moses’ speech in the land of Moab before the crossing of the Jordan (Deut 1:1–5).

That Deuteronomy draws on the language of Assyrian loyalty oaths is now widely accepted; Juha Pakkala opens his study by calling the influence of political and loyalty treaties on the book “uncontested.”¹¹¹ How direct the borrowing was, when it happened and what it was for remain contested Contested. Laura Quick traces the curses of Deuteronomy 28 to a West Semitic curse tradition carried in Aramaic and finds no direct literary link to Esarhaddon’s text, and no Judahite copy of the treaty has been found.¹¹²¹¹³ William Morrow accepts that a loyalty oath alluding to the treaty lies behind Deuteronomy 13 and 28, first sworn to a king of Judah and later taken into the book. He concludes that the two chapters can no longer carry the reading of Deuteronomy as a covert answer to Assyria.¹¹⁴ Pakkala dates the theology built on the treaties after the fall of Jerusalem, as “revolutionary theological conceptions that were possible only after the older order and its main institutions had collapsed in 587 B.C.E.”¹¹⁵ If Pakkala is right, Assyria supplied the form and Babylon’s conquest the occasion.

Esarhaddon’s treaty ordered each sworn subject: “you shall teach it to your sons,” and “In the future and forever Aššur will be your god, and Assurbanipal, the great crown prince designate, will be your lord.”¹¹⁶ Deuteronomy’s creed names one god and commands, “Impress them upon your children” (Deut 6:4, 7). On the transfer reading the echo is deliberate; Quick and Morrow give reasons to doubt it Contested. Deuteronomy 6 orders the words written on the doorposts, məzûzōt, the word for the doorposts that carried the Passover blood: “put it on the two doorposts and the lintel of the houses,” and “the blood on the houses where you are staying shall be a sign for you” (Exod 12:7, 13; Deut 6:8–9). Deuteronomy puts the written covenant where the blood had marked the houses in Egypt Inference.

The first texts to deny that other gods exist date to the exile or shortly before it. Second Isaiah, writing as Cyrus of Persia approached Babylon, has YHWH say: “I am YHWH and there is none else; beside Me, there is no god,” and “I form light and create darkness, I make weal and create woe” (Isa 45:5, 7). Smith dates the shift: “Texts dating to the Exile or shortly beforehand are the first to attest to unambiguous expressions of Israelite monotheism.”¹¹⁷ Practice lagged behind the texts. In the fifth century a Judean garrison at Elephantine in Egypt kept a temple of YHW and swore by other gods beside him.¹¹⁸

Kings refused: Gideon, Samuel and the law of the king

After Gideon defeated Midian, the men of Israel offered him a throne for himself, his son and his grandson. He refused: “I will not rule over you myself, nor shall my son rule over you; YHWH alone shall rule over you” (Judg 8:23). The verb is māšal, “rule,” and the word “alone” is the translators’ addition.¹¹⁹ His son Abimelech hired a band of men with seventy shekels from the temple of Baal-berith, killed his brothers and was made king at Shechem. Jotham, the one brother who escaped, answered from Mount Gerizim with a fable. The olive tree, the fig tree and the vine each refuse to “wave above the trees,” because each has fruit to give. Only the thornbush accepts, and it offers the trees its shade: “come and take shelter in my shade; but if not, may fire issue from the thornbush and consume the cedars of Lebanon!” (Judg 9:8–15). Martin Buber called the fable “the strongest anti-monarchical poem of world literature.”¹²⁰ Shade was the language of real kings. Hammurabi closes his laws with “My benevolent shade is spread over my city,” and Lamentations remembers the captured king of Judah as the one “in whose shade we had thought to live among the nations” (Lam 4:20).¹²¹ A thornbush offering shade parodies that royal claim Inference.

In 1 Samuel the elders ask Samuel to “appoint a king for us, to govern us like all other nations,” and YHWH tells him, “it is Me they have rejected to rule over them” (1 Sam 8:5, 7). Samuel then sets out “the practice of the king who will rule over you.” The king will take sons for his chariots and his fields and daughters for his kitchens, give the best fields and vineyards to his courtiers, and take a tenth of the grain and the flocks, “and you shall become his slaves” (1 Sam 8:11–17). Deuteronomy’s law of the king puts a written text above the king. It forbids him many horses, many wives and much silver and gold, forbids him to send the people back to Egypt, and requires that a copy of the Teaching be written for him by the Levitical priests, to be read all his life, so that he will not “act haughtily toward his fellows” (Deut 17:14–20). It gives him limits and a reading duty and assigns him no task in war, justice or worship.¹²²

Genesis tells the story of serfdom from its other side. In a famine Joseph sells Pharaoh’s grain to the Egyptians for their money, then for their livestock, then for their land and their persons: “Take us and our land in exchange for bread, and we with our land will be serfs to Pharaoh.” The land “passed over to Pharaoh,” everyone except the priests owed him a fifth of every harvest, and the people thanked Joseph for it: “You have saved our lives! We are grateful to my lord, and we shall be serfs to Pharaoh” (Gen 47:19–25). A few chapters later a new king sets taskmasters over the Israelites “to oppress them with forced labor,” and they build “garrison cities for Pharaoh” (Exod 1:11). Kings uses the same two terms for Solomon. He “imposed forced labor on all Israel” and built “garrison towns,” and when the northern tribes broke from his son they stoned Adoram, “who was in charge of the forced labor” (1 Kgs 5:27; 9:19; 12:18). The word the Egyptians use for themselves, ʿăbādîm, is the word Samuel uses for the king’s subjects (1 Sam 8:17), and Leviticus reserves it for God: “they are My servants, whom I freed from the land of Egypt,” and “the land is Mine; you are but strangers resident with Me” (Lev 25:23, 55). Leon Kass reads Joseph’s policy as the template a later Pharaoh turns on Israel: “In Israel, things will be different—indeed, reversed.”¹²³

Mesopotamian kings called themselves shepherds. Hammurabi introduces himself as “the shepherd, selected by the god Enlil.”¹²⁴ Ezekiel takes the title from the kings of Israel and gives it to God: “Ah, you shepherds of Israel, who have been tending yourselves!” and “I Myself will graze My flock” (Ezek 34:2, 15). Ezekiel still promises a human ruler, “My servant David,” as the one shepherd (Ezek 34:23–24; 37:24), and Psalm 23 opens “YHWH is my shepherd; I lack nothing.” The shepherd title was still a royal claim in Babylon when Ezekiel wrote, and Jens Varhaug finds it absent from the self-description of the Persian kings.¹²⁵ That places the prophet’s transfer of the title in the exile Inference.

All of these texts reached their present form in Judah, most of them in the Deuteronomistic history of the seventh and sixth centuries or later, and the strongest reading of them is late. Timo Veijola assigned Gideon’s refusal, Jotham’s fable in its frame and Samuel’s warning to a late editor who drew, in Gerald Gerbrandt’s translation, “the theocratic consequence that absolutely no human kingship was possible.”¹²⁶ Ernest Nicholson reads Deuteronomy as the work of exiles in Babylonia, and its ban on a foreign king as an answer to the Babylonian great king.¹²⁷ On this reading the refusal of kings is the verdict of scribes who had watched the kings of Israel and Judah fall, and it preserves no memory of a people before the state Contested. The early side is weaker than its reputation. Frank Crüsemann, who did most to date the anti-royal texts early, placed them in the resistance of free landowning farmers to the new state, between Absalom’s revolt and the division of the kingdom, so his resistance happened inside the monarchy.¹²⁸ Veijola himself grants that the list in 1 Sam 8:11–17 and the fable reuse older northern material.

The evidence outside the Bible is thin. The Iron I highland villages have no palace and no temple, which fits a society without kings. Avraham Faust reads their uniform houses and plain pottery as the marks of an egalitarian ethos, and Lester Grabbe answers that the reading assumes the ethos it sets out to show.¹²⁹ By the ninth century the Mesha stele names Omri as king of Israel and the Tel Dan inscription names the house of David.¹³⁰ No Iron Age text from outside the Bible voices hostility to kings. The question would be decided by such a text, by a linguistic dating of Judg 9:8–15 to before the monarchy, or by a demonstration that 1 Sam 8 does not depend on Deut 17, and none of these exists. The chain of words from Pharaoh’s serfs to God’s servants is in the text. The claim that it carries a memory of the highland ʿApiru has no witness Inference, untested.

Editing and the problem of selection

One objection reaches every section. Every layer of the Torah was edited for its editors’ present. A reader looking for a people of outsiders, or for land claims, will find them, and late composition alone predicts the same pattern. The early stage of the thesis therefore rests on evidence the biblical editors did not control: the Amarna letters, Merneptah’s stele, the settlement surveys, and the Egyptian labour texts. Its biblical evidence is drawn from what those editors would have had reason to remove: a “Hebrew” slave law that treats Hebrews as a class, Hebrews who changed sides in a battle, ancestors who served other gods, a Midianite priest at Israel’s sacrifice, and a grandson of Moses serving an idol at Dan. Survivals that embarrass the editors are good evidence of an older layer. The criterion has failed before, and it does not prove the thesis.

What can be claimed, and how firmly

Attested: the ʿApiru were a status category of uprooted people across the Near East from the nineteenth to the twelfth century BCE, and the Amarna letters place them in the Canaanite highlands, at Shechem among other places. A people called Israel existed in Canaan by about 1208 BCE. The Bible uses “Hebrew” as a status word in its slave law and in the mouths of foreigners, and Israel read its own name as “those who crossed over.” The hill country filled with small unwalled villages after the Egyptian withdrawal. The received text ties Pharaoh’s serfs, Solomon’s forced labour, the king’s subjects and God’s servants together by shared words.

Strong: the first god of Israel was El. The Torah compresses centuries of movement between Egypt, Canaan and Mesopotamia into the history of one family; its genealogies work as charters and its patriarchs as eponyms, and its land stories take the form of grants and deeds. Deuteronomy presents seventh-century or later law as the speech of Moses, and much of the Torah’s present shape belongs to the Persian period. Deuteronomy draws on the language of Assyrian loyalty oaths, and programmatic monotheism appears with the Babylonian exile.

Inference: Israel was formed in part from the highland ʿApiru and turned their status name, a name for changing allegiance, into an ethnic name. Southern pastoralists joined them in the highlands. Early Israel worshipped a god of a people through stones and a tent, without statues. Its law was written in a script that grew up outside the cuneiform schools. Each settlement of the land, from the Iron Age to the return, produced its own layer of legitimation.

Contested: the derivation of ʿibrî from ʿapiru; the Shasu as Israel’s source; the dating of Deuteronomy’s treaty borrowing, and how directly it borrowed from Esarhaddon’s treaty; YHWH’s southern origin, and whether southern herders brought him; the date of the programmatic ban on images; whether the alphabet spread from the margins of the Late Bronze order or from its courts; whether the anti-royal texts of Judges and Samuel keep material older than the monarchy.

Untested: the link between the alphabet and the ʿApiru; the claim that Israel’s refusal of kings remembers the highland ʿApiru; and the hypothesis of an older priestly class, carried from the Amorite tribes into Israel’s Levites. Each predicts a text that has not been found.