Babylon

The city taken by the Persians had been extensively rebuilt during the two generations before Cyrus. Nebuchadnezzar II (605–562 BCE) used the resources of the Neo-Babylonian empire to create a capital covering about 850 hectares. A double fortification circuit and moat enclosed both sides of the Euphrates. A stone bridge connected the two halves, while the Processional Way linked the principal ceremonial buildings to the monumental Ishtar Gate.^[1] Glazed-brick bulls and mušḫuššu dragons marked this route; at the northern edge, a fortified palace projected from the wall.

At the centre stood Esagila, the temple of Marduk, and the adjacent ziggurat Etemenanki, later associated with the Tower of Babel. Esagila combined chapels, courtyards and administrative spaces within the institution that organised much of the city's cult and economic life. The ziggurat and temple were therefore both monuments and working parts of an urban religious system.^[1]

Excavation alone did not identify this plan. Babylonian topographical texts supply names for buildings and gates, while royal inscriptions and stamped bricks assign much of the monumental programme to Nebuchadnezzar's forty-three-year reign.^[1] The Achaemenids inherited a recently reconstructed capital whose palaces, temples, streets and official traditions were already integrated. Their reuse of that fabric explains why Persian rule could be institutionally consequential without producing an equally conspicuous new building programme. Kuhrt places Babylon within a wider programme that rebuilt old cities and repaired their shrines after Nebuchadnezzar's conquests. What distinguished the capital was the concentration of fortification, palace, temple and ceremonial works within one urban plan. Its scale drew on the resources of the Neo-Babylonian state, while its layout remained legible through indigenous topographical descriptions.^[1]

For the Greeks, Babylon was the greatest city in the world, and Herodotus's description of it, written while the city lay under Persian rule, fixed its image for two millennia:

Round it runs first a fosse deep and wide and full of water, and then a wall of fifty royal cubits' thickness and two hundred cubits' height.

Herodotus, Histories 1.178, trans. Godley^[2]

On the wall-top, he says, there was "space enough between for the driving of a four-horse chariot", and "There are an hundred gates in the circle of the wall, all of bronze, with posts and lintels of the same".^[2] Within, he describes the two great enclosures: the royal palace in one half of the city, and in the other, "still to this day the sacred enclosure of Zeus Belus": Bel-Marduk, whom he renders in the Greek manner as Zeus. "In the centre of this enclosure a solid tower has been built, of one furlong's length and breadth", with further towers rising "from this, and from it yet another, till at last there are eight": Etemenanki, seen through Greek eyes, its stages climbed by a spiral ramp, its summit shrine holding a couch and a golden table. "But no image has been set up in the shrine", he adds; the god's golden cult-image sat in a lower shrine, a detail that would later carry a great weight of argument (below).^[2]

The description is precious and cannot be taken at face value. The walls Herodotus gives would enclose an area many times the 850 hectares the excavators traced, and his dimensions have long been treated as exaggerated, whether by his informants or in the telling.^[3] He never names Nebuchadnezzar: the great works of the city he assigns to two queens, Semiramis and Nitocris, figures of legend rather than of the king-lists: "there were two that were women", he writes of Babylon's builders.^[2] Whether he saw the city himself remains debated. Asheri and his co-authors consider a visit probable and note that stretches of the walls remained visible.^[4]

The Persian takeover was prepared inside Babylon itself, in the estrangement between the last native king and the priesthood of Marduk (see Nabonidus). Nabonidus (556–539 BCE) promoted the moon-god Sin toward a precedence over Marduk and spent ten years of his reign away at the Arabian oasis of Teima. For Babylon the city, his absence had one consequence above all: the akītu, the New Year festival that was the supreme event of its civic and religious calendar, could not be celebrated, since it required the king. The Nabonidus Chronicle notes year after year that the king stayed in Teima, that Nabu did not come to Babylon, that Bel did not go out, and that the festival did not take place.^[5] The pro-Persian polemic known as the Verse Account put into Nabonidus's own mouth a vow subordinating the festival itself to his temple for Sin at Harran:

When I have completed what I am building I shall take his (sc. Sin's) hand and set him in his dwelling. Until I have achieved this and fulfilled my desire, I shall omit the festival, the New Year Festival I shall order to cease.

Verse Account of Nabonidus, col. ii, trans. Kuhrt^[6]

The Verse Account is victors' literature, composed after the conquest to blacken the fallen king, and its portrait of Nabonidus as sacrilege incarnate served Cyrus directly. The lapse of the festival itself, though, is attested by the Chronicle, and it mattered later for a second reason: the akītu was already in abeyance long before any Persian could be blamed for its absence (below).^[5]

Cyrus's army took the city in October 539 BCE after the battle at Opis and the surrender of Sippar; his troops entered Babylon on 12 October, and the king made his own ceremonial arrival seventeen days later (see the fall of Babylon). What the event meant for the city is best read in the Chronicle's own claim about the weeks in between, while the soldiers ringed the temple precinct: "Interruption (of rites/cult) in Esangil or the temples there was none, and no date was missed."^[5] Continuity of cult across the change of masters is the Chronicle's message, as it is the message of the Cyrus Cylinder; both texts come from the Marduk priesthood that had made its peace with the conqueror, so neither is a neutral record. Private Egibi contracts continue across Cyrus's accession year. Tolini describes a brief interval of uncertainty in which the Egibi and Nur-Sin families withdrew to Borsippa; Itti-Marduk-balāṭu made his will there and later entered Cambyses's service.^[7]

The new king was folded into Babylonian forms from the first days. Kuhrt follows Kessler's argument that Cyrus, on his entry, underwent a ritual investing him with the Babylonian royal regalia at the temple of Nabu ša ḫarê, the shrine "where the royal insignia were bestowed on the Babylonian ruler".^[8] At the New Year of 538 his son Cambyses received the sceptre in the same temple, installed as sub-king of Babylonia in an arrangement that lasted barely a year. The episode illustrates the early Achaemenid adoption of Babylonian forms of royal office.

Cyrus retained the country's administrative machinery under a new governor. The identity of that governor is complicated by several similar names. The Nabonidus Chronicle calls the Gutian commander who entered Babylon Ugbaru or Gubaru and records his death in November 539. Administrative documents attest a different Gubaru, installed later as governor and still active in the 520s; Kuhrt treats the two men as distinct.^[8] Tolini dates the later governor's full title, governor of Babylonia and Transeuphratene, to Cyrus's fourth year (535).^[7] Xenophon's Cyropaedia adds a literary Gobryas who defects to Cyrus, while the Persian Gobryas who joined Darius in 522 belongs to another context (see the Seven). Mitchell concludes that the surviving accounts cannot be fully synthesised.^[9] The distinction between the two historical office-holders rests on chronology and document type rather than on spelling alone: a narrative chronicle records the commander's death, whereas legal and administrative tablets continue to attest the later governor. Greek forms of the name cannot by themselves collapse those dossiers.

The governor of Babylon and Across-the-River controlled a territory broadly corresponding to the former Neo-Babylonian empire.^[10] Local Babylonian officials continued below the highest level, and Kuhrt finds no evidence for unrest during the first Persian generation.^[1] The governor's authority also entered private contracts: penalty clauses invoke the ḫīṭu ša Gūbari, the "penalty of Gūbaru", in a form Sandowicz compares with the royal penalty formula ḫīṭu ša šarri.^[11] Under Darius the combined governorship passed to Uštanu, known in Greek as Hystanes.^[10]

Babylon rebelled twice during the crisis of 522–521. Nidintu-Bêl and Arakha each took the name Nebuchadnezzar and claimed descent from Nabonidus before Darius defeated them (see the accession of Darius and the Behistun inscription).^[12] Herodotus later said that Darius pulled down the walls, but the archaeological and literary commentary indicates dismantling rather than their complete removal.^[4]

Babylonia and Syria–Palestine initially formed one large province, Akkadian Bābili u Ebir-nāri, "Babylon and Across-the-River" (see the satrapy system). The arrangement broadly preserved the territory of the Neo-Babylonian state under a Persian satrap, while Babylon remained a royal residence and local Babylonian officials continued below the governor.^[12] The title placed the Euphrates boundary inside one administration rather than between two provinces. A later reference to a pīḫatu of Babylon and Transeuphratene shows that the combined territorial designation retained an administrative afterlife.^[12]

The date of division into Babylonia and Beyond-the-River has changed with the evidence. Stolper's 1985 study allowed a split as early as 503 and considered it certain by 480; his later examination showed the combined satrapy still intact in 486, placing the change under Xerxes.^[10][13] Boiy likewise dates separate provinces from Xerxes's reign.^[12]

Older scholarship included the division among punishments imposed after the revolts of 484. Kuhrt instead interprets it as an administrative subdivision of a territory too large for one provincial government.^[1][14] The change redefined a governor's jurisdiction, but neither its date nor its form independently demonstrates a punitive policy toward Babylon or its temples.

Babylonian legal documents identify rulers through date-formulae. Under Cyrus these normally include "King of Babylon" and "King of the Lands" (šar Bābili šar mātāte), and Cambyses used the Babylon title during his brief co-regency.^[10][12] Because the formulae occur in everyday contracts rather than occasional royal monuments, their sequence can show when scribal usage changed. It also shows variation: an older formula might persist beside a newer one rather than disappearing by decree.

Earlier scholarship interpreted the title's declining use under Xerxes as a deliberate demotion after the revolts. The documents indicate a gradual change. Formulae naming Xerxes king of Persia and Media, Babylon and the Lands continue from his fifth to sixteenth years, and older forms recur under Artaxerxes I.^[10] Kuhrt and Sherwin-White therefore reject an abrupt policy change tied to the revolts. Waerzeggers presents the fading title as part of longer imperial integration, in which separate territorial titles became less necessary.^[15][14]

The rebels could deploy the same language. A tablet in the Metropolitan Museum dates by Šamaš-erība as "King of Babylon and of the lands".^[14] His use of the formula shows its continued political force, but also cautions against treating any single title as a transparent statement of administrative policy.

Babylonia is documented by tens of thousands of legal and administrative tablets from temples and family firms. The major temple households—Esagila at Babylon, Ebabbar at Sippar and Ezida at Borsippa—controlled estates, herds, prebends and dependent labour. A prebend gave its holder a hereditary share in temple service and income; the resulting networks joined cult, property and civic office. Temple archives consequently record much more than ritual.

Jursa's analysis, summarised by Pirngruber, links market activity to taxation. The state demanded workers for armies and building projects, but temple households did not always have enough dependent labour to meet the drafts. They sold produce or wool, raised silver and hired substitutes.^[16] Monetisation was therefore a practical condition of the tax system, not simply evidence of private commerce. Demand for hired labour could support an urban labour market, while some of the silver collected by the crown returned through local expenditure.^[16]

Obligations could also require attendance at the imperial court. Records show Borsippa priests travelling to Susa for labour and tax duties and to represent their interests.^[17] Their journeys place local temple personnel inside imperial administration rather than at its margins. Dandamaev and Lukonin identify further imposts on harvests, livestock, market activity, crafts and persons.^[18] The tablets thus document both the scale of temple resources and the distinct mechanisms by which the crown drew on them.

Persian rule also established an agrarian system in Babylonia, best documented at Nippur, in which crown land was parcelled into holdings owing service and tax. Its elementary unit was the "bow land" (bīt qašti). Holders, often groups of related men, owed at least nominal military service and the ilku, paid mainly in silver; service could in practice be commuted into payment. Holdings were grouped into ḫaṭrus under a šaknu, who allocated land and collected the dues.^[10] The arrangement joined military obligation to agricultural production without requiring every holder to serve personally.

Large business houses operated between these holders, the temples and the crown. The Egibi family of Babylon had traded since the seventh century, and its archive extends from the Neo-Babylonian kings into Xerxes's reign.^[7] In the later fifth century the Murašû firm at Nippur leased or managed service land, supplied agricultural inputs and credit, converted produce into silver, and transferred rents and taxes to landholders and the crown.^[10] Its archive is evidence for a particular land-for-service environment, not a complete model of Babylonian commerce.

Credit made the system work but could redistribute land. Cardascia regarded a debtor's permanent loss of mortgaged property as uncommon, while Dandamaev considered it frequent; Stolper presents the disagreement without resolving it.^[10] Marduk-rēmanni of Sippar shows a different local combination in Darius's reign, joining crown leases, trade, lending, travel and probable tax farming.^[19] Together the archives show imperial taxation implemented through local firms and Babylonian legal forms, with commercial opportunity and coercive collection operating in the same institutional field.

Herodotus's tribute list assigns Babylonia 1,000 talents of silver, the largest assessment imposed on any one satrapy in his account, together with an annual levy of five hundred boys as eunuchs. He also says that Babylonia supported troops stationed in the country and provisioned the king and court for four months each year.^[18] These obligations combined cash, people, food and military supply; the silver figure alone does not describe the province's total burden. Assessment, provisioning and satrapal income record different relationships between Babylonia and the crown.

Herodotus attributes a separate income to Tritantaechmes, satrap of Babylonia: a daily artaba of silver, 16,800 horses and Indian hounds supported by four tax-exempt villages.^[20] Dandamaev and Lukonin regard the numerical scale as fantastically exaggerated.^[18] The figures therefore cannot be converted directly into provincial output or compared as audited accounts. Within Herodotus's representation, however, Babylonia is the empire's greatest fiscal contributor (see the peoples of the empire). Assessment is not the same measure as provincial wealth: it may reflect taxable capacity, royal demand and the shape of Herodotus's district list. The passage is therefore evidence both for perceived pre-eminence and for the Greek construction of imperial wealth, not a balance sheet from which Babylonian output can be calculated.

Jursa's quantified study of the "long sixth century" changed the economic picture of Babylonia. From the late seventh century the evidence indicates an integrated regional economy, with canal transport, regional specialisation, cash-crop agriculture and substantial market activity.^[16] Babylon's size and demand helped draw produce toward the metropolis, while low transport costs on the navigable canal network connected specialised districts. Pirngruber links this regional integration to cash-crop production and the monetisation required to convert rents and taxes into silver.^[16]

The fiscal system that encouraged these exchanges also imposed labour drafts and pressure to sell crops soon after harvest. Different archives emphasise different parts of the process: Stolper's study of the Murašû land-for-service sector foregrounds extraction, while Jursa's urban and temple evidence foregrounds growth.^[10][16] These are not necessarily incompatible conclusions, because the same demand for silver could expand markets and increase institutional dependence.

The chronology limits any long-term model. Most growth data belongs to the Neo-Babylonian period and early Darius. Documentation becomes much thinner after 484; Pirngruber finds only three fifth-century diary fragments that contain prices.^[16] The later course of the Babylonian economy cannot therefore be reconstructed with the same precision as its sixth-century expansion, and the archive break may exaggerate the appearance of economic change.

In the summer of Xerxes's second year, northern Babylonia rose. Waerzeggers reconstructs the sequence from the archival context of tablets dated by two rebel kings. The earliest known text recognising Šamaš-erība, a man of uncertain origin, is dated to day 4 of month V in 484 BCE. Ten days later Bēl-šimânni was recognised at Borsippa and Dilbat, while Šamaš-erība remained king at Sippar.^[14] The two men opposed Persian rule and may also have fought each other. This reconstruction comes from ordinary contracts whose date-formulae recognise a ruler, not from a narrative of the revolt. Each tablet establishes that a king was acknowledged at a particular place and date; their accumulation supplies the sequence and geography. Silence can mark the absence of recognition, the loss of an archive, or both.

Šamaš-erība eventually appears at Sippar, Babylon, Kiš and Borsippa, while Bēl-šimânni is attested only at Borsippa and Dilbat; Tolini considers a battle between them probable.^[7] The last known rebel-dated tablet is from day 11 of month VIII. After it, the cuneiform record from the affected northern cities falls sharply silent.^[14] Southern cities are not known to have joined. Waerzeggers's reconstruction places the restoration of Achaemenid rule by the end of the year.^[19]

The chronology was disputed for much of the twentieth century. San Nicolò and Böhl placed Bēl-šimânni in 484 and Šamaš-erība in 482; Boiy's 2004 synthesis still presents that two-year scheme.^[12] Kuhrt's 1995 survey, following Briant, placed a revolt in 481.^[1] Waerzeggers's single-year reconstruction, scrutinised by Oelsner and now generally accepted, places both rebels in 484.^[14]

The clearest evidence for the suppression's effects is the abrupt termination of surviving archives. Waerzeggers records thousands of texts from the years before 484 and only a small number after it; the closures extend from Sippar to Dilbat and include both temple administrators and private families.^[19] The pattern is concentrated in Xerxes's second year, in the rebel cities of the north, and among people tied to the major temples.

The terminations were not perfectly simultaneous. Tolini's table places the end of the Egibi archive of Babylon in month V of Xerxes's first year and the Ebabbar temple archive of Sippar in month III of his second.^[7] The disappearing families tend to hold prebends or temple offices and bear established city names. Other men continued whose livelihoods depended more directly on Persian power; Waerzeggers interprets the distribution as evidence for opposing political alignments in 484.^[19] Southern archives largely continued, and the later Murašû firm at Nippur belonged to a different, Persian-facing notability.^[7] At Sippar, Jursa interprets the surviving pattern as an administrative cleanup undertaken by officials who intended the institution to continue, not as the physical burning of an archive.^[21]

The institutional consequences remain disputed. Kuhrt interprets the evidence as the removal of concentrated political power from a traditional temple elite and its transfer to people dependent on Persian nobles, while maintaining the continuity of cult.^[22] Jursa argues that replacement within hereditary priesthoods would necessarily have disrupted cultic and social practice.^[17] Both positions connect the archive break to reprisals against the northern temple elite, but they differ over the extent of disruption. The sparse post-484 record does not resolve that difference.

Later classical accounts associated Xerxes with the ruin of Esagila and Etemenanki and with the removal of Marduk's golden image. Herodotus 1.183 became one foundation of this reconstruction, but his passage distinguishes the god's seated image in the lower shrine from a separate solid-gold statue of a man in the outer enclosure:

and in the days of Cyrus there was still in this sacred demesne a statue of solid gold twelve cubits high. I myself have not seen it, but I tell what is told by the Chaldeans. Darius son of Hystaspes purposed to take this statue but dared not; Xerxes his son took it, and slew the priest who warned him not to move the statue.

Herodotus, Histories 1.183, trans. Godley^[2]

Herodotus calls the removed object an andrias, a human figure, rather than the god's cult-image, an agalma. Kuhrt and Sherwin-White therefore conclude that the notice "does not refer to the statue of Marduk"; they also show that gradual titulary change and the uncertain history of the akītu do not document an abrupt punitive abolition of Babylonian kingship.^[15] Later Greek accounts placed Xerxes's supposed destruction within a legitimating contrast between Xerxes the sacrilegious destroyer and Alexander the restoring conqueror. Kuhrt and Sherwin-White treat Arrian's rebuilding notice as part of that Greek representation rather than independent evidence for destruction.^[15] Boiy likewise calls the alleged vindictive reaction "rather Greek fiction than historical reality" and finds no cuneiform indication of a changed policy toward the city's population or cults; the temple's deterioration cannot be securely attributed to deliberate destruction by a Persian king.^[12]

The akītu was celebrated in Nisan and linked the city, Marduk and the king. Kuhrt and Sherwin-White reconstruct a sequence in which the king confessed the proper discharge of his duties, received a ritual blow from the chief priest, and took Marduk's image from Esagila to the shrine of destiny, where the gods assembled. He then led the procession outside the city to the akītu-house. Tribute and captives could be displayed along the route, connecting the civic calendar, the cult of Marduk and the material reach of empire.^[15]

An older interpretation made participation constitutive of Babylonian kingship: a ruler became legitimate by "taking the hand of Marduk", so the festival's supposed cessation under Xerxes was taken to mean that he had abolished the title and cult. Kuhrt and Sherwin-White reject that inference. They read "taking the hand" as a summary reference to royal participation and find no securely attested Persian-period performance with the king present.^[15] The apparent celebration in 538, when Cambyses entered the temple of Nabu, may instead record his installation as sub-king: the recorded day was one on which the festival sequence would normally have taken the king away from Babylon to fetch Nabu's image.^[15] The Chronicle already records repeated non-performance during Nabonidus's absence. Early Achaemenid participation in Babylonian royal forms is well attested, but the festival's later non-attestation does not by itself demonstrate the abolition of kingship or destruction of the temple.

Babylon was one of the places where the Persian court resided. Greek authors describe a seasonal circuit among Babylon, Susa and Ecbatana; Xenophon's version gives Babylon seven winter months:

Cyrus himself made his home in the centre of his domain, and in the winter season he spent seven months in Babylon, for there the climate is warm; in the spring he spent three months in Susa, and in the height of summer two months in Ecbatana.

Xenophon, Cyropaedia 8.6.22^[23]

The sources disagree on the seasons and durations, and the circuit is partly a Greek literary scheme. Briant nevertheless regards the institution of an itinerant court as secure. He classifies Babylon, Ecbatana and probably Susa as full royal residences with archives, treasuries and palaces, while they also served as provincial capitals.^[24]

Babylon differed from the newer palace centres in its urban fabric. A recent survey calls it the only principal residence that can be termed fully urban, with fortifications, administrative buildings, numerous sanctuaries, residential quarters and gardens; Pasargadae, Susa and Persepolis were more narrowly palatial or ceremonial complexes.^[25] Achaemenid kings used Nebuchadnezzar's palaces and left relatively little construction in their own names. The clearest Persian addition is the small "Perserbau" west of the Südburg, attributed from trilingual column-base fragments to Artaxerxes II; other proposed Achaemenid alterations remain disputed.^[12][25]

Persian- and Hellenistic-period tablets trace continuities in Babylon's urban geography and learned institutions. Documents retain the same names for city quarters across dynastic changes, allowing Boiy to compare the later city with its Neo-Babylonian predecessor.^[12] This does not prove that every neighbourhood retained the same population or function, but it does show that the urban frame remained intelligible to Babylonian scribes.

Esagila's scholars continued the astronomical diaries. Alongside observations of the sky they recorded weather, commodity prices, the Euphrates level and occasional political events, making the series valuable for both chronology and economic history.^[12] The latest cuneiform texts from Babylon are almanacs dated to 74/75 CE. In the early third century BCE the priest Berossus wrote his Greek-language Babyloniaca for the Seleucid court, another product of the temple's learned milieu.^[12] These records establish the persistence of specialist scribal practice, but their institutional origin also limits them: the survival of astronomical work cannot stand as a complete proxy for population, trade or the condition of every quarter.

Continuity did not mean stasis. Etemenanki was in disrepair by the end of the Achaemenid period, Esagila required work when Alexander arrived, and Seleucia-on-the-Tigris later drew political and economic activity eastward. The cuneiform record nevertheless documents settlement and scholarship after both Alexander's conquest and Seleucia's foundation. Boiy therefore reconstructs gradual contraction rather than immediate abandonment.^[12]

Babylon surrendered to Alexander after Gaugamela in October 331 BCE. Arrian reports that he ordered the restoration of temples said to have been destroyed by Xerxes and then sacrificed to Bel under Chaldaean direction.^[26] Babylon became a financial centre of Alexander's empire, with Harpalus supervising its treasury and revenues during the Indian campaign.^[12]

Alexander also ordered work on Etemenanki. Strabo, calling the ziggurat the tomb of Belus, records the scale of the unfinished project:

Here too is the tomb of Belus, now in ruins, having been demolished by Xerxes, as it is said. It was a quadrangular pyramid of baked brick … Alexander intended to repair this pyramid; but it would have been a large task and would have required a long time (for merely the clearing away of the mound was a task for ten thousand men for two months), so that he could not finish what he had attempted; for immediately the king was overtaken by disease and death.

Strabo, Geography 16.1.5, trans. Jones^[27]

Alexander died in Nebuchadnezzar's palace in June 323 BCE. The astronomical diary records the event in five words, "29: the king died": 29 Ayaru, 10/11 June 323.^[12] Babylon's economic importance diminished later as Seleucia grew and the Tigris displaced the Euphrates as the region's principal artery, but the change was gradual. Cuneiform tablets show that neither Alexander's conquest nor Seleucia's foundation produced an immediate end to the city, and Esagila's scribes continued to use the Babylonian calendar into the first century CE.^[12]