The Mystery of the Hanging Garden of Babylon: An Elusive World Wonder Traced

by Stephanie Dalley

  Fig. 64 Parthian lintel found in the South-West Palace at Nineveh, in the 19th century. Its design was thought to be Assyrian. Length 1.83 m, ht. 0.26 m.

  The Parthian ‘lintel’ is not long enough to fit over an original doorway in the South-West Palace; 4 m, 3 m and 2 m are the approximate widths of different original doorways there, so that even the narrowest would not allow for the overlap required to support the ends of the stone (see Figure 64). An original doorway was perhaps narrowed, or the carved stone was designed for a different use. In any case, the existence of such an object shows without a doubt that the site of the South-West Palace of Sennacherib was occupied, in part at least, by a social elite, for mere squatters do not install large, sculptured stones.

  Had the palace been utterly ruined for many centuries, the Assyrian panels of bas-relief sculpture would have been buried, hidden beneath metres of debris, and so not available for re-use by that hypothetical ‘well-to-do Sassanian or Arab’. A part of Sennacherib’s palace was excavated much more recently, in 1968, and in Hellenistic levels were found ‘floor and foundations made of limestone and marble’, as well as columns and re-used bas-relief panels ‘that once graced the Throne-Room’.30

  Almost a thousand years intervene between Assyrian decline in 612 BC and the Sassanian period, when Ammianus Marcellinus referred to events of AD 359 and when Philo the Paradoxographer was writing. In theory each stratum of occupation should be found in place in between previous and subsequent layers. In practice whole levels are often missing even though texts show clearly that the city flourished during those times.31 These difficulties are compounded with the much wider problem of dating pottery and other finds. The uncertainty applies when a late addition was made to a panel of sculptures in the South-West Palace, where the surface was only partly rubbed smooth for a new cavalryman on horseback.32 The figures bear no relationship to the surrounding scene; the horseman bears an intriguing resemblance to figures on the frieze of the Parthenon in Athens, as well as to earlier Assyrian ones. If some access to the palace buildings was possible after 612, a wider range of time opens up for dating the addition.33 If a pair of figures without an Assyrian context—such as a particular campaign—was added, the alteration suggests a lack of royal engagement or centralized control that might be characteristic of the post-empire period.

  Excavators were surprised to detect renovation carried out soon after 612 on the temple of Nabu, not far from Sennacherib’s palace. On a paved road alongside it, wall-panels of Sennacherib and Ashurbanipal had been re-used.34 Campbell-Thompson and Hutchinson, who excavated many decades after Layard, were at a loss to account for what they found just on top of a major destruction level with ash. ‘It is not easy to explain the presence of three repairs on the E. corner. All are subsequent to the destruction … Indeed, there are many repairs on the SE side.’ A similar situation was found for the so-called palace of Ashurnaṣirpal II on the south-east side of the Nabu temple, where the same excavators reluctantly admitted to some apparent restorations that followed the destruction.35 The administration responsible for repairing the temple of Nabu may not have cleared and repaired every point of damage, but one often visits an old city or an ancient castle to find that some of its rooms and gates have been cleaned and refurbished, whereas others are quite ruined and no longer functional. Four levels of occupation following the destruction, and probably earlier than the Seleucid period, were discovered by later excavators in two other areas.36

  More recently a fresh analysis of reports and objects found in the great temple of Ishtar of Nineveh suggested that repairs originally assigned to the Parthian period were actually made soon after 612, comparable with those made to the temple of Nabu. Conversely, other signs of repairs or modifications, originally dated to the Assyrian period because so much Assyrian material was re-used there, are now thought to be Parthian in date.37 Possibly, therefore, the cult of Ishtar of Nineveh was never discontinued but was maintained without a break to Parthian times. Several Roman helmets of Parthian or Sassanian date were found in the Ishtar temple, likely to be dedications made by soldiers, owing to her nature as a war-goddess (‘battle is a game for her’), as well as terracotta figurines with erotic themes,38 stored or dedicated there owing to her nature as goddess of love and fertility.

  So little is known about the circumstances of Nineveh’s sack in 612 that we are not even sure who kept control of the heartland of Assyria afterwards. But the Babylonians retained some control, for a contemporary chronicle tells that the conquering king Nabopolassar honoured the city after its capture by receiving tribute there, presumably in a palace. He would not have done so if nothing but smoking ruins and dusty rubble had remained, with danger from collapsing roof-beams. At the very least, this detail shows that one part of one building was still in sufficiently good condition to be chosen for a royal ceremony.

  The 14th year (of Nabopolassar’s reign, i.e. 612 BC—after the siege of Nineveh had been successful for the Babylonians) … the king of Akkad [and his troops] moved as far as Naṣibina. Pillage and banishment … and the Rusapeans were brought to Nineveh before the king of Akkad.39

  Nabopolassar was distraught at the damage done to the wonderful city. Nabonidus later recorded his predecessor’s reaction to the sack:

  (As for) the king of Babylon—the work of Marduk, for whom pillage is an abomination—he (Nabopolassar) did not lay a finger on the cult of any of the gods, but went around with his hair unkempt (in mourning), and slept on the ground as his couch,40

  an admission which shows the intention had been to prevent devastation, and that some cults were still viable in the city. Thus archaeological evidence of repair to the temples may be linked to that confession in a Neo-Babylonian text.

  No clay records specifically dated later than 612 BC have been found at Nineveh. However, the Assyrian city of Dur-Katlimmu41 on the Habur river has recently yielded unquestionable evidence that it remained important, revealing late Assyrian monumental buildings where Assyrian-style administration persisted without a break for the next fifty years, into the reign of Nebuchadnezzar II.42 Duplicates of Sennacherib’s prism and cylinder inscriptions, describing how he created his palace garden, were placed in several different cities, raising their chances of long-term survival.

  Fig. 65 Impression from a cylinder seal with a Late Elamite inscription, probably late 7th–6th century BC, found above the temple of Ishtar at Nineveh. Chalcedony, 2.2 × 1.75 cm.

  A few undated clay tablets bearing correspondence in the Elamite language were found at Nineveh, and a study made of the seal designs impressed on them suggested that they quite likely dated after 612 (see Figure 65).43 This suggestion, though challenged,44 matches a new understanding that the Elamite capital at Susa had recovered by 625 from an earlier destruction, confirmed by new evidence from a rich tomb and the analysis of its contents.45 Cyrus the Great has proved to be an Elamite rather than a Persian, so his conquest of Babylon reflects the strength of that Elamite revival.46 On some of the sculpture panels in both the South-West Palace and the North Palace, selective damage done to disfigure pro-Assyrian Elamites and the images of those Assyrian kings who had fought the Elamites implies the rooms were still accessible after the sack, perhaps some time later.47 It now seems possible that Greek historians implicitly included Elamites when they referred to ‘Medes’, emphasizing Cyrus’ family connection with the Median royal line.48

  As at Nineveh, so at nearby Nimrud, archaeologists expected to find that the city had been abandoned in 612, for they dug with Xenophon’s Anabasis in hand.49 When one reads Mallowan’s quite detailed account, one finds a discrepancy and a tension between what was found and what was expected. He found that the ‘ruins’ of the temple of Nabu were reoccupied between around 500 and 300 BC,50 and a sequence of six ‘not unprosperous villages’ arose upon the citadel,51 around and on top of the temple of Nabu, continuing the ceramic and burial customs of Assyria. Its people built large kilns for the manufacture of glass; they impor
ted the best wine (for strength and bouquet) all the way from Thasos in the northern Aegean,52 and buried their dead with silver coins and with antiquities such as cylinder seals. From the subsequent period when Seleucid kings ruled, a remarkable hemispherical glass bowl of laced design, mainly clear but with yellow and brown elements, is dated from similar finds elsewhere, as a fashionable item probably imported from Alexandria in the late 3rd or 2nd century BC.53 Renewed excavations in the 1980s found no less than three levels of post-Assyrian occupation in a large building complex inside the wall of the lower city, and ‘all produced pottery identical with that of the latest Assyrian occupation from when they cannot have been far removed in time’.54 A reasonable level of reoccupation had soon followed the disaster of conquest; in particular, ‘an important post-Assyrian settlement is attested by a large house’ in which Achaemenid pottery was identified.55

  Archaeological evidence for the Achaemenid period presents notorious difficulties which extend far beyond Nineveh. Apart from the two ceremonial centres at Persepolis and Pasargadae, in which the Persian kings imported foreign styles along with foreign craftsmen, the new rulers had no tradition of civic architecture to impose upon conquered cities, unlike the Greeks whose theatres, stoas and gymnasia herald Greek presence throughout the ancient world. In Assyria palaces had been built so monumental and durable that it was easy to make them available for a new administration, sweeping aside debris and using local workmen to repair roofs and doorways in the traditional way. Even in Babylon and Uruk, where a high standard of cultural and civic life was maintained, one is hard-pressed to identify a specifically Achaemenid building. As Willem Vogelsang has remarked, the Persian empire ‘does not seem to have had much effect upon the material culture of the East’, and the old cities, with their grand buildings still more or less preserved, ‘often provided the administrative infrastructure for the newly emerging empire’.56

  When Cyrus II built his ceremonial capital at Pasargadae around 545 BC, it was the palace architecture of Nineveh that he imitated, not that of Babylon or Susa. He could send craftsmen ‘to the partly ruined palaces of Sargon and Sennacherib in order to make close copies of the original models’.57 Likewise when Darius built Persepolis around 515 BC, nearly a century after the fall of Nineveh, ‘most of the motifs on the reliefs on façades, staircases and doorways can be paralleled in Assyria … the overwhelming impression of influence [is] from Assyria’, specifically from the South-West Palace of Sennacherib at Nineveh.58 Had the palaces at Nineveh been totally ruined in 612, the brick structures would have been quite eroded by the time those Achaemenid palaces were planned nearly a century later, and so would not have inspired imitation by successful and ambitious emperors. For royal-sponsored architects and designers to have viewed the palaces and gateways at Nineveh before embarking upon the design of Pasargadae, with its composite creatures dominating the main gateways, a certain degree of local security must have been relied upon. The designers and artists would have visited Sennacherib’s palaces at Nineveh in order to copy and adapt marvels from the most famous palaces in the world.

  As a major city of the Achaemenid period, Nineveh’s size and importance is vehemently proclaimed by the biblical prophet Jonah, whose Hebrew book was probably composed at that time. But its testimony seems to be cancelled out by the Anabasis of Xenophon. He described how the battle of Cunaxa in Babylonia in 401 BC led to the death of Cyrus the Pretender and the consequent retreat northwards of his Greek soldiers. Those remnants of ‘10,000 men’ marched through hostile territory that had once been the heartland of Assyria, passing ruined and abandoned cities with non-Assyrian names. He wrote the account more than thirty years later.

  It is generally admitted that there are unsatisfactory aspects of the story as a description of true events.59 Reacting to an earlier account now lost in which his involvement was described less creditably, Xenophon is thought to have remodelled events to exaggerate his own heroic role and to refute charges of bribery, corruption, of serving purely for personal enrichment.60 Three months are not accounted for in his journal-like account, a ‘Great Lacuna’.61 Almost none of the places he named in Mesopotamia can be identified from contemporary texts, and some known cities that he might have mentioned are absent from his account. He did not even name the place where the vital battle took place—only elsewhere is it called Cunaxa. The Median Wall that he described cannot be located in the appropriate area.62 Snow fell and disappeared with unreal rapidity in the highlands of Anatolia. He omitted to mention that his troops, working their way up the Tigris valley on its eastern side, had to cross the Lower Zab, a river of significant size.

  On reaching the region where the three great royal cities of Assyria—Nineveh, Nimrud and Khorsabad—were located, he spoke of deserted cities named Larissa and Mespila. Larissa was a common Greek word for any citadel but Mespila is found neither in Akkadian nor in Greek.63 Most scholars assume that those are alternative names for Nimrud and Nineveh, but they are not so named in any other source. There is no reason why Xenophon should not have known, at the very least, the name of Nineveh, since it was used in a wide variety of literature both during his lifetime and later: as Ninus or Nineve to Ctesias, later to Strabo,64 Diodorus Siculus and Tacitus,65 and much later to Ammianus Marcellinus.66 As Ninue it was known in the Aramaic stories of Ahiqar and Ashurbanipal and Shamash-shum-ukin which were popular in Ptolemaic Egypt; as Nineveh to the authors of the Book of Jonah; and as ‘Ninus, the time-honoured capital of Assyria’ to the biblical Greek story of Judith. The Assyrian name of Nimrud as Calah is also preserved in Strabo’s mention of Calachene.67 Those two city names are not the only ones to have survived through hard times and changes up to the present day; many others bear witness to continuous residence.68 Elsewhere loss of a city’s civic life leads to renaming when it is revived at a much later date.

  Also at odds with the picture given by Xenophon is the evidence from a letter of Arshama, royally born satrap of Egypt in the late 5th century BC, referring to his emissary travelling through Assyria with written permission to claim rations from Arshama’s estates. The regions are named, and they include several parts of the Assyrian heartland, one of them a district near Nineveh (see Figure 66).69 Arshama’s administrators would have repaired and maintained the sturdy infrastructure that delivered water to his farms and orchards, given the incentive of royal patronage. The Assyrian water channels would have been too valuable to neglect.

  It would have added to the verisimilitude of his account, had Xenophon mentioned the Assyrian cities by their known names. Both Mespila and Larissa are said to have belonged to the Medes before they were conquered in circumstances which Xenophon attributes in legendary or novelistic fashion to a persistent cloud and to a thunderstorm respectively. The extant fragment B of the novel Ninos and Semiramis, preserved on a papyrus dating to the early 1st century AD,70 echoes some of the motifs in the Anabasis, in particular about Armenians, crossing rivers, ice and snow in mountain passes, a similarity encouraging the view that the Anabasis shares with Xenophon’s Cyropaedia some of the character of a novel.71 At that time the novel was gaining popularity—a new form of prose literature—weakening the link between fiction and factual narrative. A growing fashion for composing prose fiction, coloured with touches of genuine historical background, may explain why Ctesias’ Persica, preserved only in fragments, seems to mix historical fact with legend.

  Fig. 66 Sketch map showing locations in which Arshama, satrap of Egypt under Darius II, owned estates, c.425 BC.

  Xenophon understood that part of Assyria belonged to Media, for which no other evidence can be found. His view belongs to a scheme for world history in terms of the rise and fall of great empires, which was popular at that time. Since the supposed Median empire had been the most recent to fall, there is a suspicion that a schematic literary-historical pattern has influenced his composition, presenting a picture of an almost empty land following the dramatic end of Assyrian imperial power. In presenting Assyria as a la
nd lacking visible monuments, Xenophon implied that it was outside known history, in accordance with the general idea that public monuments serve as markers of historical time; their destruction abolishes it.72 In fact the Assyrian heartland was certainly well within the Achaemenid Persian empire at the time, and had previous belonged to the Neo-Babylonian empire. The letter written by Arshama the satrap shows that literacy and careful accounting were still in force there; and those literate qualities belong to city life and centralized education.73

  When we turn to the Seleucid and early Parthian periods, Greek inscriptions, Hellenistic sculpture, and coins found in and around the ancient Assyrian buildings at Nineveh give datable evidence, but there is no trace of Greek-style civic architecture such as stoa, theatre and hippodrome. One may compare a similar lack in Babylonian cities under Seleucid rule, apart from Babylon itself.74

  Fortunately Layard found ‘the remains of several dishes and vases in serpentine and marble’, one fragment having an inscription of a priest of Amun from Ptolemaic Egypt,75 contemporary with Seleucid rulers, which testifies to Nineveh’s international ties. This evidence is augmented by a remarkable inscription, carved in well-formed Greek letters on a stone column, which was found in the temple of Nabu recording a dedication by Apollophanes son of Asklepiades on behalf of Apollonius, who is called ‘stratēgos and epistatēs of the city (polis)’. Its date is probably around 32 BC, so an earlier inscription on the stone column, which had been erased, presumably dates to the Seleucid period.76 Two of the names contain the name Apollo, who was regarded as the Greek equivalent of the Mesopotamian god Nabu.77 The rather vaguely recorded evidence for repairs to the temple of Nabu scarcely helps to reconstruct the use of the building after 612, but the date recorded on the column inscription, probably 32/31 BC, combined with the find-spot, allows the possibility that the temple of Nabu was still in use.