In the ancient Near East, the right to rule is older than rule. From the first villages of the Fertile Crescent, around 9500 BCE, ritual specialists held a standing that reached across separate communities, an authority no chief had granted them, and when rulers appear in the record, six thousand years later at Uruk in the south of what is now Iraq, they build on it. Uruk made such a figure its en, its lord and the husband of its goddess. The war-leaders who followed, the lugal, took his office for themselves: Sargon of Akkad set his daughter in the moon god’s household at Ur, and his grandson Naram-Sin was made a god at his own city’s petition. Egypt, where the river made one territory of the whole country, fused priest and king into a single god-king at the start and kept him for three thousand years. Legitimacy came first, and every early form of rule was built by capturing it, localising it or imitating it.
The usual account runs the other way. Rulers come first. Grain that can be stored and taxed, wars fought over land that cannot be left, and a class that controls both produce the state, and religion follows as the instrument that makes the arrangement acceptable. Marx and Engels put it in one sentence: “The ideas of the ruling class are in every epoch the ruling ideas”. Robert Carneiro explained the state through war between peoples with nowhere else to go, with no role for belief at all (Science, 1970). Elman Service described the early chiefdom as “universally … theocratic” and still gave leadership “causal priority”, with “economic and sacred reinforcements developing in turn” (Origins of the State and Civilization, 1975). Timothy Earle argued that chiefs build monuments and ceremonies to give their ideology a body they can own (How Chiefs Come to Power, 1997). The numbers seem to agree. Seshat, a large database of past societies, finds that belief in gods who punish wrongdoing arrives after the largest rises in social complexity, and its best model for that rise is better farming combined with military technology (Whitehouse et al. 2023; Turchin et al. 2022). And Göbekli Tepe in south-eastern Turkey, for a quarter of a century the favourite example of a temple built before anyone settled, has turned out to be a village.
The other side has older allies than it is usually given. A. M. Hocart argued in 1936 that the machinery of government “was blocked out in society long before the appearance of government as we now understand it”, as an organisation for ritual (Kings and Councillors). Michael Mann, whose history of power gives ideas no special credit, found the Sumerian temple serving “first as an intervillage diplomatic service” and called it “the first state of history”. In the early civilisations, he concluded, “militarism clearly came later, building on top of existing forms of regional organization” (The Sources of Social Power, vol. 1, 1986). Marshall Sahlins went furthest: “Kings are human imitations of gods, rather than gods of kings” (“The original political society”, in Graeber and Sahlins, On Kings, 2017, p. 23). Each of them stops at the temple. None says where the temple’s authority came from before there was a city to house it, or why separate cities honoured the same gods. That step was a network of ritual specialists. Long before any of those communities became a city, it moved between communities with different leaders, kept the peace between them, carried technique and custom, and judged their disputes, and every later stage of rule was built on what it held.
David Graeber and David Wengrow reach part of the way from another direction. They read early political life as seasonal and reversible: on the Great Plains, the camp police, warriors appointed to keep order during the summer hunts and gatherings, had no power for the rest of the year, and “their sovereignty was no less real for its periodicity” (Wengrow and Graeber, “Farewell to the ‘Childhood of Man'”, Journal of the Royal Anthropological Institute 21(3), 2015: 597–619), an argument they carried into The Dawn of Everything (2021). Graeber reports the conclusion that Hocart “was right: kingship was originally a ritual institution”, and doubts it: “This seems highly unlikely to be true in every case” (On Kings, pp. 72–73). The network model agrees with them that ritual authority came first and that command was long kept seasonal or contained. It departs on duration. Their authority dissolves when the season ends; the model needs an office that lasted, held from year to year by specialists who moved between communities.
Oannes, the Seven Sages and the Offices Carried to Uruk
The Babylonians had their own account of where their institutions came from, and it began in the sea. Berossus, a Babylonian scholar writing in Greek around 281 BCE, described a creature called Oannes, with the body of a fish and a man’s head and feet, who came out of the Erythraean Sea, the Greek name for the waters south of Mesopotamia, in the first year of the first king. Before him, the people of Babylonia “lived without rule and order like the beasts of the field”. He “gave them an insight into letters and sciences, and every kind of art”, taught them to build houses, found temples and compile laws, and returned to the sea each night. The book survives only in later Greek quotations, but the cuneiform tradition behind it survives on its own. A tablet copied at Uruk in 165 BCE, the Uruk List of Kings and Sages, pairs each of seven kings who reigned before the Flood, the deluge that Mesopotamian tradition set at the start of its history, with a sage, an apkallu: “In the time of king Ayyalu: the sage was U-An”, which is Oannes. The sage of the king Enmerkar, the list says, “brought down Ishtar from the heavens into the Eanna”, the goddess’s temple at Uruk. Ishtar is the Akkadian name of the goddess the Sumerians called Inanna. Incantation texts describe the first sages as fish-like creatures of Ea, the Akkadian name of Enki, the god of the deep fresh water whose city was Eridu.
An older Sumerian myth, Inanna and Enki, tells a related story. Inanna sails from Uruk to Eridu, where Enki, drunk at their banquet, gives her the me, the divine offices and arts: en-ship, kingship, a priesthood and the crafts of the smith and the scribe. She loads them into the Boat of Heaven and carries them home to Uruk.
The Sumerian story of the Flood puts a priest-king at the break in history. Its hero, the man the gods warn and save, is “Zi-ud-sura the king, the gudu priest”, king and temple priest in one person. After the flood he sacrifices oxen and prostrates himself before the gods, and the text calls him king each time. The story survives on a tablet from Nippur written around 1600 BCE, so it is a late memory of a priest-ruler carrying the order of the world across the flood.
These texts were written down long after the events they describe, the sage tablets between two and a half and three thousand years after writing began at Uruk, so none of them can witness prehistory. They witness what Mesopotamians believed. In every version the arts of rule and craft reach the city from outside, through sages or from an older shrine. No sage ever rules. Each serves “in the time of king X”, and the king needs him. The tradition remembered an office beside the throne that knew what the throne did not, and a city that got its institutions by taking them from an older centre.
Göbekli Tepe, Karahan Tepe and Kfar HaHoresh
The archaeology of the first villages shows communities that shared far more ritual than they could have shared rulers. Göbekli Tepe, near the city of Şanlıurfa, was built and rebuilt between roughly 9500 and 8000 BCE: circles of T-shaped limestone pillars carved with foxes, boars, snakes and vultures. Its first excavator, Klaus Schmidt, called it “a regional centre where communities met to engage in complex rites” and held that nobody lived there (Documenta Praehistorica, 2010). His successors at the German Archaeological Institute have since found houses, rock-cut channels, a probable cistern, thousands of grinding stones used on cereals, and burials. They write that “a revision of the popular scientific view is now unavoidable” (Clare 2020).
The houses put the network in the first villages, among settled people, and the rest of the evidence puts it there too. Göbekli is one of more than a dozen sites in the hills around Şanlıurfa, now studied together as the Taş Tepeler, the “stone hills”, and the same ritual grammar recurs across them: T-shaped pillars, benches, life-size human heads and statues, men shown holding their phallus. At Karahan Tepe, about forty-six kilometres east of Göbekli, builders cut a chamber into the bedrock and left ten pillars standing in it, each shaped as a phallus (Karul 2021). At Sayburç, a village of about 8500 BCE, a bench in a communal building shows a man holding his phallus between two leopards (Özdoğan 2022). These were separate settlements with their own households, and they shared a vocabulary of images carved in stone at monumental scale.
The same pattern runs through the Levant, the lands along the eastern Mediterranean. At Kfar HaHoresh in Lower Galilee, a site of the same period where nobody lived, the excavator Nigel Goring-Morris found plastered skulls, arranged human bones and a pit holding at least eight aurochs, wild cattle, sealed under a human burial. On his reading it was a mortuary and cult centre serving the villages of the surrounding lowlands. Across the region, communities removed the skulls of some of their dead, kept them and remade them: grooved and drilled for hanging at Göbekli, modelled with plaster faces at Jericho in the Jordan valley and at Tell Aswad near Damascus, once at Çatalhöyük in central Anatolia. Ofer Bar-Yosef and Anna Belfer-Cohen named this whole area of shared building plans, tool technology, skull practice and exchange the “PPNB interaction sphere”, after the archaeological period, the Pre-Pottery Neolithic B (1989); Eleni Asouti has called the construct too abstract to explain much (2006). Its obsidian, the volcanic glass prized for blades, moved through what its modern analysts describe as clusters of local exchange joined by a few long links (Ibáñez et al. 2015).
Coordination across wide areas without any sign of command continued long after. Between about 6100 and 5200 BCE the Halaf culture spread a recognisable style of painted pottery, with regional variants, along with round houses and stamp seals, from the Mediterranean coast to the foothills of the Zagros mountains, and no capital has been found.
Three explanations fit a shared ritual grammar. Villages that intermarried would share a mythology; neighbours who admired each other’s monuments would copy them; or a class of specialists who carried the rites from community to community would reproduce them. The analysts of the obsidian network favour the first and tie long-distance exchange to marriage. The specialists explain more of the combination. Intermarriage spreads stories, but cutting a ten-pillar chamber into bedrock in a grammar shared with sites a day’s walk and more away takes people who know the grammar, and Kfar HaHoresh served several villages and belonged to none. The current excavators of Göbekli now write of “(archaeologically speaking) invisible decision-makers” among its people and list ritual experts as one possibility, with storytellers and hunters (Clare 2024). On this model the specialist was reciter, interpreter and judge at once. In a culture without writing, law is kept by people who remember it, explain it and apply it to a quarrel, and those three tasks fall to one person, a bard-priest whose standing came from what he knew.
Ritual authority is also the one form of standing that people without chiefs are known to tolerate. Christopher Boehm compared egalitarian hunter-gatherers and found them equal by effort: the group watches anyone who tries to dominate and pulls him down by criticism, ridicule, disobedience and, when those fail, what he calls extreme sanctions, an arrangement he named a reverse dominance hierarchy (Hierarchy in the Forest, 1999). A man who gives orders and backs them with force is what such a group exists to stop. A man whose standing rests on knowing the rites, the songs and the custom, and who commands no one, threatens none of them, and his authority can spread across many groups without any of them giving up its equality. The first authority above the single community would therefore be a ritual one. Foragers also fought. At Jebel Sahaba, a cemetery beside the Nile in northern Sudan, 41 of the 61 people buried carry healed or unhealed wounds, which the latest study reads as repeated raids and ambushes between groups between about 16,600 and 11,400 BCE (Crevecoeur et al. 2021). Raiding of that kind needs a leader for the length of a raid, and violence between groups did not by itself produce a permanent war office. It did create the demand the ritual specialist met. Groups that raid each other need someone both sides will hear when a killing has to be settled, and a person who commands neither side and carries the custom both share is that person.
That figure is an inference, and the field has been burned by inferences of this kind. In 1930 the biblical scholar Martin Noth built a sacred league of twelve Israelite tribes around a shared shrine from shared tradition and a Greek analogy, and the league dominated scholarship for decades before it was largely abandoned. A network inferred from shared carvings risks the same fate, and the evidence that would decide it can be collected. If ancient DNA from the people buried in the special buildings of separate sites shows unrelated individuals from outside the local lineages recurring across communities, and if objects in those buildings come from distant workshops, the network was there. If the dead in every special building are local kin, intermarriage and emulation explain the pattern and this stage of the argument fails.
Çatalhöyük is the case the model has to survive, because a town that large without priests looks like evidence that settlement never needed a ritual office. The best-studied town of the Neolithic, occupied for over a thousand years from before 7000 BCE, had no temple and no priests’ quarter. Every building its first excavator, James Mellaart, called a shrine has turned out to be a house, and ritual was kept in the houses of lineages, the “history houses” of his successor Ian Hodder. A settlement that grew without taking a network office into itself kept its religion at home, which is consistent with the model, and that is as much as Çatalhöyük can be made to say in the model’s favour.
Ubaid Temples from Eridu to Tepe Gawra
Between the Halaf and the first cities lies the Ubaid period, named after a small mound near Ur and dated from about 6500 to 3800 BCE, and in it ritual moves out of the house into a building of its own. Pottery and building styles called Ubaid spread from the southern plain up the rivers into northern Mesopotamia, Syria and south-eastern Anatolia, and down the Gulf coast, where the painted bowls turn up among herding communities (Carter and Philip, Beyond the Ubaid, 2010). No capital has been found for any of it.
The building that travels with the style has a tripartite plan: a long central hall flanked on both sides by rows of smaller rooms. At Eridu the shrines raised on the spot of Enki’s later temple reach this form in the last Ubaid building phases, VII and VI, in the late fifth millennium, with niched and buttressed outer walls; the sequence below them begins in the sixth millennium and has a gap in its middle levels. Far to the north, at Tepe Gawra near modern Mosul, the excavators found three temples of closely similar plan standing together in level XIII, of the same late Ubaid date (Gil Stein, in Beyond the Ubaid, p. 25).
At Tell Brak in north-eastern Syria the pattern reaches the threshold of the city. Max Mallowan excavated there in 1937 and 1938 a temple rebuilt several times between the first half of the fourth millennium and about 3000 BCE, and found in it thousands of small alabaster figurines carved as bodies topped by large eyes, offerings to a god whose name is lost. The Eye Temple stood in a settlement that grew into a city slightly earlier than Uruk in the south.
On the model, this is the archaeological form of capture by settlement: a travelling office tied down to one place. In the Halaf villages and at Çatalhöyük, ritual is kept in houses. In the Ubaid it gets a building of its own, built to a plan recognised across the region, standing in a particular village and rebuilt on the same spot for centuries. A specialist who had moved between communities now had a house in one of them, and the house belonged to the god. The thousands of eyes at Brak are offerings on a scale beyond one household, from people who came to that building because it was the one that counted.
The editors of the standard volume on the period read the same evidence more cautiously. Robert Carter and Graham Philip reject the idea of an “Ubaid culture” with definable boundaries and find a horizon with “no undisputed core” (pp. 4–5), a set of forms shared, in Stein’s words, “to varying degrees by a diverse set of regions”, spread by exchange and imitation as much as by any common institution. The tripartite plan was also used for ordinary houses, and two later analysts of Tepe Gawra have argued that its buildings called temples served other purposes. Stein reads the southern Ubaid as run by “corporate leadership strategies”, with community temples that may have organised the surplus of irrigation through ritual (p. 32), which makes the temple a building of the whole village before it is anyone’s seat. The model needs one further step from that reading: a building that belonged to everyone and was run by the people who knew its rites became, at Uruk, the household of a lord.
The En of Uruk
The cities of southern Mesopotamia, on the plain of the lower Tigris and Euphrates, made up the land of Sumer, and their language, Sumerian, is related to no other known language. To the north lay Akkad, whose people spoke Akkadian, a Semitic language related to Hebrew and Arabic. Writing begins in the south, at Uruk, around 3300 BCE, in the great household of Eanna, Inanna’s precinct. The archaic tablets from Eanna are overwhelmingly accounts, about eighty-five per cent: rations, fields, livestock, workers, with the rest made up of lists of signs and words used to train scribes (Englund 2004). Inanna’s reed-bundle emblem writes her name in the oldest of them. The highest office the accounts use is written with the sign EN. On a slightly later field tablet from Jemdet Nasr, north of Uruk, the EN receives two-thirds of five fields and five other officials share the rest. Nobody in the archaic accounts is called lugal in the later sense of king. The most copied of the archaic lists of professions opens with another title, NAMEŠDA, which vocabularies written much later translate as “king”, and what it meant in 3300 BCE is unknown.
Writing also changed what the old office could be. In a culture without writing, custom exists only in the memory of the people who recite it, so the person who recites the law, the person who explains it and the person who adjusts it to a new case are one person, the bard-priest of the first villages, and an adjustment leaves no trace: the next recitation is the law. Jack Goody and Ian Watt documented how such a tradition fits itself to the present (“The Consequences of Literacy”, 1963). Among the Tiv of Nigeria, the genealogies used orally to settle court disputes had diverged considerably from the ones British officials had recorded in writing forty years earlier. In Gonja, in Ghana, the records made at the turn of the twentieth century show the founder of the state, Ndewura Jakpa, with seven sons, each the ruler of one division; when the story was written down a second time, after two divisions had disappeared, he had five. Oral societies, as Walter Ong summed up their findings, “live very much in a present which keeps itself in equilibrium or homeostasis by sloughing off memories which no longer have present relevance” (Orality and Literacy, 1982). A written text fixes one version, and Uruk shows the effect at once: scribes copied the archaic list of professions almost unchanged for more than a thousand years, into the schools of the Old Babylonian period, after 2000 BCE. Once a rule is written down, any later change shows against the original, and a change that shows needs an authority behind it. Amendment becomes an act that someone must be entitled to perform, and the office that performs it, the legislator, separates from the office that interprets the text and from the scribe who keeps it. The archaic tablets of Uruk are accounts, with no laws among them, so none of this is visible in 3300 BCE. It unfolds later, when Urukagina, a ruler of Lagash around 2350 BCE, lists in writing the practices of his predecessors that he has abolished and the debtors he has freed, and when the court of Ur-Nammu, king of Ur around 2100 BCE, issues the first surviving collection of laws in a king’s name.
The same household produced the images of a recurring figure, bearded, wearing a headband and a skirt of netting, whom archaeologists have called the “priest-king”. He appears on about thirty objects of the period, among them the Warka Vase, a tall alabaster vessel carved with a procession of offerings to the goddess (Warka is Uruk’s modern name), and the Lion Hunt Stele. On impressions in clay from the carved stone cylinders used as seals, found in Eanna and in no other precinct of the city, he stands over naked prisoners with their arms bound while men beat them.
Piotr Steinkeller has argued from the later texts that this office, en-ship, “apparently was the original form of Sumerian kingship” (1999). The word named a lordship that gods and the first rulers held in common. The great gods carry it in their names: Enlil, lord of the air and chief god of the land, and Enki, lord of the earth and of the deep water. The ruler of Uruk carried it as the goddess’s husband. He spent at least one ritual night with Inanna in her house, the gipar, and in later centuries a king of Sumer could take part in that rite only if he held the en-ship of Uruk and bore the title “spouse of Inanna”. That is why the one word attaches to an office, to the person holding it and to a god. It did not mark divinity, since the sign written before a god’s name is a different one, DINGIR. Epic poems remember the rulers of the age as en: in Enmerkar and the Lord of Aratta, both Enmerkar of Uruk and his rival in the eastern mountains are called en. Each city had its god and the god’s household: Inanna at Uruk, Nanna the moon god at Ur, Enki at Eridu, Enlil at Nippur.
On the model, this is where the settled network became a lordship. The Ubaid temples give the settling an archaeological form; how their keepers became lords is still the least documented step. A permanent city needed a legitimacy that would outlast a season, and the obvious source was the people who already carried the rites and the custom between communities. The city took one of them in and made him its lord, a lord who shared his title with the gods. A priest who belonged to Uruk spoke for Uruk, kept his contact with the network as one city’s representative, and ruled at home with an authority no one around him could check, because he knew what they did not. Three things in the record point outward from the single city. Separate cities honoured one family of gods with no ruler over them all, and seals of the period list groups of southern cities together, read by some as a shared cult system and by others as a league (Steinkeller 2002). The myth of the me has Uruk take its offices from an older centre at Eridu. And when the first rulers’ own inscriptions appear, around 2550 BCE, a king of Kish named Mesalim settles the border between the rival cities of Lagash and Umma “at the command of Ištaran”, a god: the inter-city judgment the network once gave, now given by a king in a god’s name. Archaic accounts from several cities recording the same offices in one system would confirm the step.
The later tradition remembered the holders of that lordship as gods. The Sumerian King List, a list of cities and reigns compiled by scribes and copied in its fullest form under the kings of Isin, a southern city, around 1800 BCE, opens: “After the kingship descended from heaven, the kingship was in Eridug. In Eridug, Alulim became king; he ruled for 28800 years.” Eight kings in five cities rule for 241,200 years before the Flood. After the Flood, in the first dynasty of Uruk, the list writes three names with the sign for a god: Lugalbanda, who reigns 1,200 years, Dumuzid the fisherman, and Gilgamesh, who reigns 126. The marking is uneven, since the two rulers before Lugalbanda, Meskiaggasher, who “became en and lugal”, and Enmerkar, get no divine sign. The lords of Uruk who shared the goddess’s bed are remembered as gods themselves, with reigns no man could live. The section before the Flood is a late addition, missing from the oldest copies, and it records what scholars of the early second millennium believed: that kingship began at Eridu, Enki’s city, and that the first rulers lived as long as gods. Henri Frankfort concluded in 1948 that in Mesopotamia “the office, and not the office-holder, was of superhuman origin” (Kingship and the Gods). His formula describes the state the list’s compilers lived in, after the lugals had separated the office from the person who held it. By then kingship was a gift that “descended from heaven” and passed from city to city, and the king who received it was a man. The divine signs, the god-length reigns and the sacred marriage preserve the older arrangement, in which office, holder and god shared one lordship.
The title had equivalents across the region. Akkadian bēlu and West Semitic baʿal, cognate words for “lord” and “owner”, did the work Sumerian en did, and scribes writing Akkadian used the sign EN to write bēlu. A further link between en and the Semitic god El has been suggested and has never been tested; the model does not need it, since it rests only on the highest title for a lord meaning the same thing across the region.
Lugal, the Assembly and Nippur
Robert Carneiro’s theory, the strongest form of the kings-first account, explains the rise of the war king. States form, he argued, where farmland is hemmed in, by land that cannot be farmed or by neighbours who hold the rest, conditions he called environmental and social circumscription (Science, 1970). People who cannot move away from a stronger neighbour must fight him or submit, and war under those conditions produces conquest, subject villages and a ruler who commands the army. Southern Mesopotamia was such a place, irrigated land with desert beyond it, and by the third millennium its cities fought over fields and canals; the border between Lagash and Umma that Mesalim settled was one such quarrel. Once fighting is cheaper than moving, a war leader becomes necessary, and the lugal is that leader. Carneiro’s theory accounts for him and says nothing about the en, who was already there and whose standing was the lugal’s first problem.
The war role left the en and became an office of its own, and the land records below suggest it came from outside the temple. Lugal is written with two signs meaning “man” and “big”. The first rulers who call themselves lugal in their own inscriptions, from around 2600 BCE, are kings of Kish, in the north of the plain near later Babylon, among them Enmebaragesi and Mesalim, and kings of Ur. The King List itself remembers the two titles as separate things joined: the first ruler of Uruk after the Flood, Meskiaggasher, “became en and lugal”.
Thorkild Jacobsen gave the lugal an origin in 1943. The earliest Sumerian cities, he argued, were governed by an assembly of free men presided over by a council of elders, which could grant kingship to one member for a limited time in an emergency (“Primitive Democracy in Ancient Mesopotamia”, Journal of Near Eastern Studies). One of his texts was the poem Gilgamesh and Aga, in which Gilgamesh puts a demand from Kish first to “the elders of his city”, who advise surrender, and then to the able-bodied men, who refuse. Critics have pointed out that the poem is known from copies made centuries later, that Gilgamesh is already “the lord of Kulaba”, a district of Uruk, before either body speaks, and that the pairing of elders and young men may be a literary device (Dina Katz). Where assemblies can be documented, they are real. Daniel Fleming found elders, assemblies and “the town” governing beside the kings in the eighteenth-century BCE palace archive of Mari, and Andrea Seri found the same bodies settling disputes and disposing of property under Hammurabi.
The assemblies have a place in the model, beside the en. Settlement produced two kinds of standing. The settled specialist held the god’s household, its rites and its fields. The kin groups held their own land and their own men, and the elders spoke for them. In the north, land belonged to families; the large temple estates were a southern institution, strongest at Lagash and Umma (Gelb, Steinkeller and Whiting, Earliest Land Tenure Systems in the Near East, 1991). Even an emperor of Akkad had to buy the land of the northern families. The black stone obelisk of Manishtushu, Sargon’s son, records the king buying about 3,430 hectares in four districts around Kish, each parcel “from several sellers of the same kinship grouping”, with forty-nine citizens of Akkad standing as his witnesses. A war leader chosen by an assembly of such men came from the lay side of the city. He could command the men, and he had no standing with the gods. He had to acquire it.
Piotr Michalowski has objected that en, lugal and ensi, the title used at Lagash, were local words for the same thing, the sovereign. The objection is fair, and the localities still tell. The cities with the largest temple households, Uruk and Lagash, named their ruler for the god’s household or as the god’s steward, and Kish, where families held the land, named him the big man. Ur, in the south, also used lugal, which makes the line a tendency with exceptions.
At Lagash the tension can be watched. Around 2350 BCE a ruler named Urukagina, or Irikagina, set up inscriptions listing the abuses of his predecessors: “the oxen of the gods plowed the ruler’s onion plot”, and the god’s best fields had become the ruler’s onion and cucumber beds. His remedy was to make the city god Ningirsu the owner of the ruler’s household and fields, and the goddess Bau the owner of his wife’s estate. He declared freedom, ama-gi, for debtors, and swore he “would not surrender the orphan or the widow to the powerful”. Ningirsu, he said, had taken him “by the hand from among the multitude”. He called himself ensi in his first year and lugal from the next, so the ruler who handed his household to the god took the war leader’s title in the same years. Historians have read the reforms as propaganda; at least one measure has now been shown to have been carried out (Schrakamp and Zólyomi 2025).
Nippur shows what was left of the network’s older function, a city that conferred rule and never held it. It was the city of Enlil, the chief god of the land, and of his temple, the Ekur. No dynasty of Nippur appears in the King List, and Nippur never ruled anyone. Its possession made a king the king of the land, and every power of the plain fought for it: Kish, Uruk, Akkad, Ur, Isin, Larsa. The earliest explicit statement that Enlil grants kingship comes from Lugalzagesi of Uruk, around 2350 BCE: “When Enlil, king of all lands, gave Lugalzagesi the kingship of the Land”. Xianhua Wang has argued that Lugalzagesi may have invented that role for Enlil (2011); if so, a king wanted that theology and built it. A later chronicle, the History of the Tummal, lists king after king, from Enmebaragesi of Kish to the first king of Isin, who rebuilt Enlil’s temple and his wife’s shrine at Nippur. The kings of Ur were crowned at Nippur.
Sargon’s Daughter and Naram-Sin’s Horned Helmet
Sargon of Akkad, who reigned from about 2334 to 2279 BCE, made the first empire in the region out of the cities of Sumer and the north, and he moved on the priesthood early. His Akkadian name, Šarru-kīn, is read “the king is established” or “the king is legitimate”, which is what a ruler without a lineage would need to assert. A text from the Assyrian empire of the seventh century BCE, written some 1,500 years after him, gives him a birth to match. His mother is an ēnetu, the Akkadian title of a high priestess of the en kind, and his father is unknown: “My mother was a high priestess, my father I knew not”; she set him “in a basket of rushes, with bitumen she sealed my lid”, the river carried him to “Akki, the drawer of water”, and Akki raised him as his son. The legend shows only what later tradition believed: that the first great lugal had a mother from the en office and a father nobody could name. A calcite disk from the gipar at Ur, the residence of the moon god’s priestess, names her as Enheduanna, zirru-priestess, wife of Nanna and daughter of Sargon, who also styled himself king of Kish; its relief shows her attending a libation. On the disk her title is zirru, a title of the en, and the art historian Irene Winter judged that the office existed before her (2010). Hymns preserved under her name may have been attributed to her later. A Genealogy of Identities, an earlier essay on this site, follows what Akkad’s capture of the en office meant for the cities it overwrote.
The appointment shows the step the lugals took. A lugal held force and had no sacral standing of his own, and he could not leave lords who shared the gods’ title above him. Sargon moved the en-ship into his own family and changed what it was. At Uruk the en had been the ruler and the goddess’s husband; at Ur the en was the king’s daughter, wife and servant of the moon god, holding a priestly office that served the god and answered to the king. Naram-Sin put his own daughter, Enmenanna, into it, and for nearly six hundred years kings kept it in their families, normally for the reigning king’s daughter, chosen by divination. The last known holder, Enanedu, was the sister of Rim-Sin of Larsa, and the office seems to have lapsed when Babylon crushed the revolt of the southern cities around 1739 BCE.
Naram-Sin, Sargon’s grandson, who reigned from about 2255 to 2218 BCE, put down a revolt of the cities of Sumer and the north. An inscription on a statue base found at Bassetki in northern Iraq records what followed: because he had “secured the foundations of his city”, its citizens asked the great gods of eight cities, from Ishtar of Uruk to Enlil of Nippur, that he be made the god of Agade, the capital that gave Akkad its name, and they built him a temple there. On his victory stele he climbs a mountain wearing a horned helmet, the headgear of gods, and in the second half of his reign his name is written with the divine sign. In the south, the Akkadian kings built their crown land largely by confiscating temple estates. A fragmentary victory stele of a Sargonic king, probably Naram-Sin, records about 1,570 square kilometres of land in the province of Lagash, confiscated and handed out to his dependants (Gelb, Steinkeller and Whiting 1991).
Deification did not hold. Later Sumerian poets wrote The Cursing of Agade, in which Naram-Sin attacks Enlil’s Ekur with ladders and axes and Enlil sends “the Gutians, an unbridled people, with human intelligence but canine instincts … and monkeys’ features” to destroy his city. In fact the Akkadian kings rebuilt the Ekur, and bricks of Shar-kali-sharri, Naram-Sin’s successor, say so; the poem is Nippur’s verdict on the dynasty that had made itself divine. Kings became gods in emergencies, by petition, for a century or so at a time: Naram-Sin after the revolt, then Shulgi of Ur and his successors after a crisis of their own. Ur-Nammu, founder of that Third Dynasty of Ur, was not deified in his lifetime, and Hammurabi of Babylon never wrote his name with the divine sign.
Under the Third Dynasty of Ur, which historians call Ur III and date from about 2112 to 2004 BCE, the capture was completed in administration. Its kings became the owners in practice of the southern temple estates; the temples remained the domains of their gods in theory and were run by the state through its provincial governors (Gelb, Steinkeller and Whiting 1991). The first law collection to survive, the Laws of Ur-Nammu, comes from this court around 2100 BCE, possibly from his son Shulgi, and its prologue repeats Urukagina’s promise: “the orphan was not delivered up to the rich man; the widow … to the mighty man”. The reform text of Lagash and the first surviving code both come from rulers of Sumerian cities, where the temple households had been strongest, and both claim the god’s justice for the ruler.
Egypt’s God-King as the Control Case
If the model is right, a country in which the network and the territory were one would never have needed to capture anything. The Nile valley is a single narrow corridor of farmland from Aswan to the Delta, and the communities along it shared one river, one flood and one calendar. Egyptians called the southern valley Upper Egypt, because it lies upstream, and the Delta Lower Egypt. A ritual office that served the whole valley could rule it, and in Egypt priest and ruler appear together from the first. The country was united around 3100 BCE; its history then runs through the Old Kingdom, the age of the pyramids (about 2686 to 2160 BCE), the Middle Kingdom (about 2055 to 1650) and the New Kingdom (about 1550 to 1069), with periods of division between them.
At Hierakonpolis in Upper Egypt, excavators have uncovered a ceremonial complex built around a plastered oval court, in use from about 3500 BCE into the First Dynasty. Its rubbish pits held the bones of crocodile, hippopotamus and turtle alongside cattle and fish, and the excavators read them as dangerous animals captured and killed in a display of order imposed on chaos, the central theme of later kingship (Friedman). From the beginning of the dynasties, the king’s oldest name identifies him with the god Horus. Frankfort’s judgment, softened since, still frames the case: “Pharaoh was of divine essence, a god incarnate; and this view can be traced back as far as texts and symbols take us. … His coronation was not an apotheosis but an epiphany.”
Egypt had no priesthood outside the state for a king to capture. In the Old Kingdom most high officials held both administrative and priestly titles, and no clear line divided the two (Bárta 2017). Temple staff served in five rotating groups, the phyles, a month or so at a time, as duty rosters from the pyramid temples at Abusir record. Ann Macy Roth has argued that the phyles began as kin groups of Predynastic Upper Egypt, possibly entered by initiation, which the first kings turned into rotating royal service (1991); if she is right, an older organisation was absorbed whole at unification. In doctrine the king was the only officiant. The temple walls show him, and nobody else, offering to the gods, and every priest acted in his name.
Egypt did take something from the network to its east, and what it took was imagery. Uruk’s merchants had planted settlements as far as Syria; none is known in Egypt. Yet at the moment kingship formed, the Egyptian elite acquired cylinder seals with no Egyptian ancestry, the niched brick façade that also frames the king’s name in his oldest title, and the scene of a hero holding two lions apart, carved on the ivory handle of the Gebel el-Arak knife. The long-necked felines on the ceremonial slate palette of Narmer, the king who appears at unification, come from the same repertoire. Specialists see these as prestige goods and motifs reaching Upper Egypt through the Delta and the Levant, with no sign of instructors from Uruk (Campagno 2013). Egypt imitated the look of the Uruk ruler and built its own office.
The Narmer Palette shows war at unification: the king smiting a kneeling enemy, rows of decapitated bodies. Egyptologists now read it as an ideological statement of the king imposing order rather than a record of a single conquest, and its ideology shows the god-king fighting the war himself. Egypt produced no separate office of war leader to fight it for him. Its Old Kingdom armies were levies raised for a campaign and led by the king’s officials, as the Sixth-Dynasty official Weni describes. The Middle Kingdom kept permanent garrisons in its chain of fortresses in Nubia. A standing army with chariots came only after Egypt faced a rival kingship on its own soil. From about 1650 BCE a dynasty of Levantine descent, the Hyksos, ruled the Delta from Avaris. Isotope studies of the burials there show immigrants settled for generations before their rise, a takeover from within (Stantis et al. 2020). The kings of Thebes, the great city of Upper Egypt, fought them out by about 1550 BCE, and the New Kingdom that followed was a military state, with frontier bases, chariotry and campaigns as far as Megiddo in Canaan.
Its kings were war leaders, and one of them moved on the sacral order. Akhenaten, who reigned from about 1353 to 1336 BCE, struck at Amun, the god of Thebes and of the dynasty that had driven out the Hyksos. From about his fifth year he had Amun’s name and images hacked out, taxed other gods’ temples to fund his own, the sun disc Aten, and moved the court to Amarna. The attempt did not survive him. The Restoration Stela of Tutankhamun, who came to the throne a few years after him, describes the temples “from Elephantine to the lagoons of the Delta” fallen into ruin, and records that the king “initiated priests and prophets, children of officials from their towns”. No document shows the priests of Amun resisting Akhenaten, and the story of their plotting goes back to the anticlerical assumptions of early Egyptology.
The sacral office then outlasted the throne. At the end of the Twentieth Dynasty, Herihor, a general who was also vizier and High Priest of Amun, took the full royal titles at Karnak in Thebes. Through the Twenty-first Dynasty, from about 1069 to 945 BCE, Upper Egypt was governed from Thebes by High Priests of Amun who were also army commanders, and its important decisions were put to the oracle of Amun, presented as the true king, while kings at Tanis in the Delta held the north. The pharaoh himself stayed a god for as long as Egypt had temples. Greek kings of the Ptolemaic dynasty and the emperors of Rome were carved on temple walls in the pharaoh’s crown, offering to the gods; the last emperor shown as pharaoh in a temple relief is Decius, at Esna, around 250 CE.
Seshat’s Punishing Gods and the Evidence That Would Refute the Model
The Seshat findings are the strongest objection, and they measure a different thing. Seshat codes whether a society believed in high gods who punish wrongdoing, and in its analysis those gods arrive after the largest rises in complexity (Whitehouse et al., Religion, Brain & Behavior, 2023), though a rival analysis of the same data reverses the order (Beheim et al. 2021). The thesis concerns an office that carried rites and settled disputes, and on that Seshat’s own results are consistent with it. In Turchin’s analysis of the database, the number of levels in a society’s religious hierarchy varies independently of its military, administrative and settlement hierarchies, which move together (Turchin et al. 2022), as they should if priesthood and command come from two separate sources. In the Near East the record of shared ritual begins thousands of years before the record of the war king, and the war king, when he comes, spends his first generations acquiring what the sacral office held.
The model makes predictions beyond the region, and anyone can test them on other early states. Where settlement localises a ritual network, a centre should survive that confers rule and never rules. Every unifier should move on the sacral office early, by putting his kin in it, rebuilding its shrines or taking its estates. Royal deification should follow a crisis of legitimacy and fail to last. A polity whose territory and network coincide should fuse priest and king from the start and raise no separate war office until it meets a rival kingship. A warrior king’s attack on a sacral institution older than his dynasty should be reversed within a generation or two, and after a fused god-king falls, the next holders of force should go to the sacral office for legitimacy. In the records of any region, religious and military hierarchy should vary as separate lines.
It would fail on clear evidence. A kingship that demonstrably arose first and later created a ritual network across peoples would refute it. So would a ritual network that first appears under an existing political unifier, with no earlier trace of shared ritual across his territory, or archaic accounts from Uruk naming a war leader who held more than the EN. The Neolithic test is the DNA one: if everyone buried in the special buildings of separate sites belonged to local lineages, and nothing in those buildings came from far off, the network goes, and the sequence loses its beginning.
Drought, the Many Kings and the Habiru
A break around 2200 BCE shows which order each region had built. Dust in a sea-floor core from the Gulf of Oman and in a stalagmite from a cave in northern Iran records a dry interval of two to three centuries beginning around 2250 BCE (Cullen et al. 2000; Carolin et al. 2019). Harvey Weiss argued in 1993 that the drought emptied the plains of northern Mesopotamia and brought down Akkad; others have shown continuity at some northern sites and see many causes, and how much weight the drought carries is still disputed. The break-up is undisputed. After Shar-kali-sharri, the King List asks, “Then who was king? Who was the king?”, and soon after hands kingship to “the army of Gutium”, the Gutians, whom later tradition remembered as raiders and as Enlil’s punishment. In Egypt the Old Kingdom ended in the same window, in the mid-twenty-second century BCE. A provincial governor, Ankhtifi, boasted in his tomb that “the whole of Upper Egypt died of hunger” while nobody in his province starved, a boast of a kind governors of the age repeated. Egypt’s crisis broke the central administration and left the office intact. In the words of the UCLA Encyclopedia of Egyptology, “the position of the king, however, was not put into question”, and the Middle Kingdom restored the god-king.
Mesopotamia restored nothing of the kind. After the brief empire of Ur III, the land broke into kingdoms, Isin, Larsa and Babylon among them, each with its lugal and its city gods, and the kings of the south claimed the favour of Enlil at Nippur. Genesis tells of a city and a tower built “to make a name for ourselves”, and its Table of Nations names “Babylon, Erech, Accad, and Calneh” as the kingdom of Nimrod, Erech being Uruk; both were written down in the first millennium and are Israel’s later summary of that world of kings, each its own small Babel.
Some people crossed out of it. From the nineteenth century BCE to the twelfth, texts from Anatolia to Egypt name a class of people called Habiru, or ʿApiru: people cut loose from their communities, living as outlaws, labourers and above all soldiers. Idrimi, son of a deposed king of Aleppo, says on his statue that he lived among them for seven years in Canaan, taking omens, before he won back a throne with their help. In the Amarna letters, written by the rulers of Canaan to Pharaoh in the fourteenth century BCE, Abdi-Heba of Jerusalem complains that “the land of the king deserted to the Hapiru”. Other letters show the same people serving the kings they threatened: Biryawaza of Damascus offers Pharaoh “my troops and chariots, together with my brothers, my ʿApiru and my Suteans”, the Suteans being another people of the steppe.
Israel came out of the same world, and its scripture remembers a problem like the one the Habiru posed: how to settle without becoming one more kingdom with a captured god. Whether the word Hebrew derives from ʿApiru is disputed, and no document connects the Habiru of the second millennium to the Israel first named on a stele of Merneptah around 1208 BCE; the connection is untested, and the biblical stories stand beside the Habiru as a parallel. The biblical memory gives the instruction in judging to an outsider priest. Jethro, Moses’ father-in-law, called the priest of Midian in Exodus and a Kenite in Judges, watches Moses judge the people alone from morning to evening and tells him to appoint capable men over thousands, hundreds, fifties and tens and “let them judge the people at all times” (Exodus 18:21–22). It also keeps, as late as the reign of Jehoiakim around 600 BCE, a family that refused to settle. The Rechabites told Jeremiah that their ancestor Jonadab had commanded them: “You shall never drink wine … Nor shall you build houses or sow fields or plant vineyards … but you shall live in tents all your days …” (Jeremiah 35:6–7). A late genealogy appears to count them among the Kenites (1 Chronicles 2:55). Frank Frick read them as a hereditary guild of metalworkers who kept their craft by staying on the move (1971); others see keepers of a nomadic ideal or an ascetic order. Either way, a lineage that carried its rule with it and owned no land was still honoured by a prophet of the kingdom of Judah around 600 BCE, and God’s promise to them in Jeremiah is that someone of their line will always be “standing before Me”.
Legitimacy is a belief held by the governed: the conviction that this person has the right to command and that obeying him is proper. A belief that deep is installed slowly, through repeated ritual and the word of people already trusted, and the network had installed this one across communities over thousands of years before any warrior ruled. Force compels obedience quickly and cannot produce that belief at the same speed. Each new holder of force therefore borrowed legitimacy from the office that already held it, by becoming the goddess’s husband, by appointing his daughter as the god’s wife, by rebuilding the god’s temple or by petitioning to be counted among the gods. History as Installation, an earlier essay on this site, follows the same process at the scale of whole civilisations.
Later rulers found other sources of legitimacy and never stopped drawing on the oldest one. Kings wrote law down and set it above themselves; officials came to claim that they protected the people from the king; the people became what the law protected; and at last states learned to run without kings at all. Since 1956 the national motto of the United States, printed on its banknotes, has been “In God We Trust”.