Mesopotamia · seventh to sixth century BC · Greek and Latin witnesses from Berossus (c. 290 BC) to the late-antique Philo
Was the hanging garden of the Greek wonder-lists a real garden at Babylon, the Assyrian king Sennacherib's garden at Nineveh misplaced by the Greeks, or no single structure at all?
What happened. The story, as Greek writers told it, is that a king of Babylon, the great city on the river Euphrates in what is now southern Iraq, built an artificial mountain beside his palace for a wife who came from the hills of Media, in north-west Iran, and pined for them: stone terraces on brick vaults, planted with full-grown trees, watered from the river by machinery hidden inside the structure. By the first century BC Greek writers counted it among the seven wonders of the world. Five ancient descriptions survive, none by an author who writes as if he had seen it: the earliest, by Berossus, a priest of Babylon writing in Greek about 290 BC, names Nebuchadnezzar II, king of Babylon from 605 to 562 BC, as the builder; two others, by Diodorus, a Greek historian of the first century BC, and Curtius, a Roman historian of the first century AD, say only "a Syrian king", meaning an Assyrian one, a king of the empire to the north; and two, by Strabo, a Greek geographer writing about AD 20, and a short treatise on the seven wonders that circulates under the name of Philo of Byzantium, a Greek engineer of the third century BC, though the treatise itself was written between the fourth and sixth centuries AD, name no builder. The garden has never been found: Robert Koldewey, the German archaeologist who excavated Babylon from 1899 to 1917, thought he had found it in a block of vaulted rooms that turned out to be a storage basement, and no later excavation has produced a garden.
Why it is a mystery. Nothing in the Babylonian records settles it either: no cuneiform text of Babylon's own kings (cuneiform is the wedge-shaped script of ancient Mesopotamia, the land between the rivers Tigris and Euphrates, pressed into clay) mentions a garden there, and Herodotus, the Greek historian who described the city about 430 BC, says nothing of it. Three answers have been given: the traditional one keeps the garden at Babylon and gives it to Nebuchadnezzar, whose own inscriptions describe a new palace, the North Palace, built on terraces "as high as a mountain" beside his father's, so that on this view a planted terrace of that palace was exaggerated by later writers. Since 1993 Stephanie Dalley, an Assyriologist at the University of Oxford (Assyriology is the study of the cuneiform texts of Mesopotamia), has argued that the Greeks had the wrong city: the garden was built by Sennacherib, king of Assyria from 705 to 681 BC, beside his palace at Nineveh, the Assyrian capital on the Tigris near the modern city of Mosul, three hundred miles north of Babylon and a century earlier; his inscriptions describe a palace garden and a bronze water-raising device that she reads as a water screw (the spiral pump that the Greeks credited to Archimedes of Syracuse four centuries later), his canals and aqueducts survive in part, two carved stone panels (reliefs) from Nineveh show a garden, and the Greeks, who did confuse Assyria with Babylonia, moved the garden south. Some specialists accept this (Renger 1998 and Veenhof 2001, as reported by Reinhold Bichler and Robert Rollinger, ancient historians at Innsbruck whose 2005 article examines Dalley's thesis at length; Meijer 2017, with hedging); many keep the garden at Babylon, among them Bichler and Rollinger, Timothy Potts, an archaeologist of the ancient Near East who reviewed Dalley's book in 2013, Ariel Bagg, the historian of Assyrian water engineering, in 2014, and Olof Pedersén of Uppsala University, whose 2021 book synthesises the archaeology of Babylon; and a third answer, held in some form by several historians of Babylon (Amélie Kuhrt, Paul-Alain Beaulieu, Bert van der Spek and Michael Seymour), is that no single garden lies behind the story, which combined a Mesopotamian royal park, a Babylonian palace terrace and Greek engineering fantasy.
How it was judged. The kings' own inscriptions (Sennacherib's prisms, six- or eight-sided blocks of clay inscribed on every face, and Nebuchadnezzar's palace texts) were read in the standard editions, RINAP 3/1 for Sennacherib and RINBE 1/1 of 2024 for Nebuchadnezzar, together with the five Greek and Latin descriptions in the original languages, the excavation record at both cities and the arguments on both sides as published (Dalley's 1994 article and 2013 book only in part). The evidence was sorted into seven clusters, groups of items that rise and fall together, and each cluster was given a probability under each answer; with flat priors, meaning the three answers started level, the arithmetic gives Babylon 57%, no single structure 38%, Nineveh 5%. Three findings carried the most weight: every ancient witness who names a city names Babylon; Berossus places the garden in Nebuchadnezzar's palace, in a paragraph that follows Nebuchadnezzar's own inscription item by item; and the one passage in Sennacherib's inscriptions that describes his garden, repeated with small variants from prism to prism, speaks of water drawn by bucket and says nothing of terraces. Sennacherib's garden at Nineveh is real and well documented, and it is the closest ancestor of what the Greeks describe, but the evidence does not support Dalley's claim that it is the garden they described. Because three clusters turn on readings that specialists dispute, the headline is Babylon 55%, no single structure 35%, Nineveh 10%; these are judgements expressed as numbers, conditional on the framing and the weights, and a quotation of Berossus by another ancient writer that lacked the garden sentence, or an Assyrian text or relief showing the garden on vaulted tiers, would move the result towards Nineveh.
Babylon, more likely than not, at about 55%. No single structure behind the story, about 35%; Nineveh, about 10%.
The case for Babylon rests on the unanimity of the witnesses that name a city, on Berossus, and on what Sennacherib's own text does and does not say, and each of those is a reading that a specialist could contest. It is a reading of the sources by non-specialists.
The wonder was more likely than not a real raised garden at Babylon, on the terraced North Palace that Nebuchadnezzar II says he built "as high as a mountain". Four of the five ancient descriptions put the garden at Babylon; the fifth, Philo, names no city, king or river, so every witness that names a city names Babylon. Josephus, the Jewish historian who wrote in Greek at Rome about AD 95, quotes Berossus twice, in two different works, and the garden sentence is in both quotations and in the later one by George Syncellus, a Byzantine chronicler of about AD 800; so the sentence was already in the text of Berossus that Josephus had, which was the summary made by Alexander Polyhistor, a Greek compiler of the first century BC. The surviving copies therefore rule out an addition made after Josephus but cannot show whether Polyhistor added it, and that Abydenus, a Greek writer on Assyrian history of about the second century AD who did not draw on Josephus, is an independent second witness is likely rather than established. A Babylonian tablet of the Persian period (539 to 331 BC, when Persian kings ruled Babylon) assigns gardeners and water-drawers to the "new palace" at Babylon, so whatever the wonder was, that palace had a watered garden.
All three figures are conditional on the three hypotheses as framed and on the weights given to seven clusters of evidence; they are judgements about evidence rather than measurements, and Nineveh is given 10% rather than the 5% the weighing returns because three of the seven clusters turn on readings that specialists dispute. Granting Dalley all three at once, that Berossus' garden sentence is an interpolation (a sentence inserted by a later hand), that Sennacherib's device was a screw and that Greek confusion could move a garden from the Tigris to the Euphrates, puts Nineveh ahead, 55 against 27. To make Nineveh the clear favourite would take, in addition, an Assyrian text or image showing the garden on vaulted tiers, of which none is known.
Sennacherib's garden is documented in his own prisms, with canals, the stone aqueduct at Jerwan, in the hills north-east of Nineveh, that carried one of them, and two palace reliefs, and the Greek tradition did confuse Assyria with Babylonia. But his prism says the water was drawn "by bucket" over wells, with ropes, wire chains and cables; the device itself, which the inscription lists among the palace's cast-metal sculptures as "tree trunks and date palm(s) of copper", is read by Dalley as an Archimedean screw, a conjecture that neither the lexica (the dictionaries of the language) nor the editors accept; "a wonder for all peoples" is said of the palace rather than of the garden, on the Chicago Assyrian Dictionary's reading of the phrase siḫirti ekalli as "the entire palace" (CAD S 236; Dalley reads "the surroundings of the palace"); "a replica of Mount Amanus" (a mountain range in what is now southern Turkey) is a formula three Assyrian kings used; and no Neo-Assyrian text (one from the Assyrian empire of the ninth to seventh centuries BC) calls Nineveh "Babylon". The blank at Babylon is real, though: no inscription of Nebuchadnezzar's contains the word for garden and nothing identifiable has been excavated, which is why "no single structure" keeps about a third.
Seven clusters of evidence, each given a probability of being observed under each of the three hypotheses, and flat priors: that weighing gives Babylon 57, no single structure 38, Nineveh 5, and the shares above lift Nineveh to 10% because the inputs are judgements. A starting weight tilted in Nineveh's favour still ends at Babylon 56, no single structure 38, Nineveh 7. Dropping any single cluster leaves Babylon ahead of Nineveh, though without Berossus the literary-construct reading leads at 53%, and dropping the location cluster (the witnesses' unanimity about the city) altogether, which means granting Dalley the whole confusion argument, lifts Nineveh to 27% (a different scenario from the one in which Babylon falls to 27 when Dalley is granted her three contested readings at once).
The evidence is of three kinds, all in print. First, what the kings themselves wrote: the clay prisms on which Sennacherib's scribes recorded his building at Nineveh, read in RINAP 3/1, the standard edition of his inscriptions, and the inscriptions in which Nebuchadnezzar II, king of Babylon from 605 to 562 BC, describes his palace at Babylon, read in RINBE 1/1 of 2024, both with the Chicago Assyrian Dictionary, the standard dictionary of Akkadian. Second, the five Greek and Latin descriptions, read in the original languages with published translations. Third, what excavation has found at both cities, with the two Nineveh reliefs and the surveys of Sennacherib's canals.
The published arguments on both sides were read, though the 1994 article and 2013 book of Stephanie Dalley, the Assyriologist at the University of Oxford who since 1993 has argued for Nineveh, could be read only in part; her statements of 1993, 2003, 2008 and 2017 and two reviews of the book were read in full. Five earlier studies are known through other studies and the notes of the Sennacherib edition; Sennacherib's canal inscriptions at Bavian, the rock inscription at the head of his canal system, and at Jerwan were read in older translations; and the Philo text, a short treatise on the seven wonders written between the fourth and sixth centuries AD, in a poor scan.
The method had two steps. Each passage was read in its edition and compared with what Dalley and her critics say it means. Then the evidence was sorted into seven clusters, groups of items that rise and fall together, and each cluster was given a probability under each of the three answers, so that the weighing is written down and anyone can change it. Starting the three answers level (flat priors), the arithmetic gives Babylon 57 per cent, no single structure 38 and Nineveh 5. The weighing was then re-run with other starting weights, with each cluster left out in turn, and in two scenarios that favour Dalley.
Figure 1 shows how the probability divides between the three answers: Babylon 55 per cent, no single structure 35, Nineveh 10. Figure 2 shows how the descriptions of the garden reached us. Two lines are independent of each other: the Babylonian priestly tradition (Berossus, a priest of Babylon writing in Greek about 290 BC, and Abydenus, a Greek writer on Assyrian history of about the second century AD) and the Alexander-historian tradition (Clitarchus, who wrote a popular history of Alexander about 300 BC, and the unnamed source of Strabo, a Greek geographer writing about AD 20). Both lines put the garden in Babylon.
Three findings weighed most: every ancient witness who names a city names Babylon; Berossus places the garden in Nebuchadnezzar's palace, in a paragraph that follows Nebuchadnezzar's own inscription item by item; and the one passage in Sennacherib's inscriptions that describes his garden speaks of water drawn by bucket and says nothing of terraces. Because three of the seven clusters turn on readings that specialists dispute, the headline lifts Nineveh to 10 per cent. The figures are conditional on the framing and the weights: a starting weight in Nineveh's favour ends at Babylon 56, and granting Dalley her three contested readings at once puts Nineveh ahead, 55 against 27.
Already established by others. The verdict agrees with Bichler and Rollinger (2005), Bagg (2014, who rejects the Nineveh thesis and treats the classical location at Babylon as reliable), Potts (2013) and Pedersén (2021), who leave the garden at Babylon. Pedersén writes that "most scholars have accepted the identity of Nebuchadnezzar's North Palace and the palace described by Berossos" (2021, 125); that is his assessment, and no survey of specialists' opinions was made for the report. Nineveh has had published support besides Dalley (Renger 1998 and Veenhof 2001, as reported by Bichler and Rollinger; Meijer 2017, hedged; Foster 2004, a study on the Nineveh side known by its title and opening only), and the editors of Sennacherib's inscriptions take no position, filing the identification as an "opinion" with "critical reviews" (RINAP 3/1, n. 44). Rollinger (2013) had already argued that the Greek "hanging" reflects Nebuchadnezzar's own "high as a mountain"; the North Palace identification is his, the Leipzig Assyriologist Michael Streck's (2018) and Pedersén's; and the bucket argument against the screw is Bagg's.
What historians have said. The tradition since antiquity puts the garden at Babylon, where a century of excavation has found nothing; the block of vaulted cells in Babylon's Southern Palace that Koldewey identified as the garden between 1899 and 1917 turned out to be a storeroom, and nobody now defends it. Since 1993 Stephanie Dalley has argued for Sennacherib's garden at Nineveh, in articles of 1993 and 1994, a 2003 article with John Peter Oleson, a historian of ancient technology, on the water screw, a 2008 chapter on "Babylon" as a name for other cities, her 2013 book and a 2017 chapter. Some specialists accept the identification (Renger 1998 and Veenhof 2001, as reported by Bichler and Rollinger; Meijer 2017, hedged), many keep the garden at Babylon (Bichler and Rollinger 2005; Potts 2013; Bagg 2014; Pedersén 2021), and several Babylon specialists (Kuhrt, Beaulieu, van der Spek, Seymour) hold something like the view that no single structure lies behind the story. The editors of Sennacherib's inscriptions take no position.
The decisive passages, read in the standard editions, show the Nineveh case resting on three contested readings at once: that Berossus' garden sentence is an interpolation, that Sennacherib's copper tree trunks were a screw, and that Greek confusion could carry a garden from the Tigris to the Euphrates. Each looks improbable on the evidence, and Berossus, the one witness who had the Babylonian records in front of him, puts the garden in Nebuchadnezzar's palace. That is where Bichler and Rollinger, Potts, Bagg and Pedersén already stood. The weighing gives that position explicit shares, Babylon 55%, no single structure 35%, Nineveh 10%, with the figures the re-runs give when the weights change (Babylon 56% from a prior in Nineveh's favour, 27% when Dalley's three contested readings are all granted), and adds the observation that Sennacherib's "wonder" is said of the palace rather than of the garden; it does not show that the position prevails in the field.