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

by Stephanie Dalley

  Following the sack of Babylon by Sennacherib, the New Year festival was not celebrated, as is stated laconically in the Babylonian Akītu-Chronicle:

  For [8] years during (the reign of) Se[nnacherib], for 12 years (during the reign of) Esar[haddon]—20 years altogether—Bel (Marduk) stayed in Baltil (Ashur city) and the akītu-festival did not take place.20

  Instead, the ritual was performed in Assyria. Quite overtly, an Assyrian version of part of the Epic of Creation used syncretism with Anshar to put an Assyrian deity, whether Ashur or Ishtar, as the hero-god in place of Babylon’s patron god Marduk, making appropriate changes in the early genealogy of the primeval gods. At the culmination of the epic in its Babylonian version Marduk supervised the building of Babylon by the gods; but with Babylon out of the question and an Assyrian version with Marduk replaced by an Assyrian deity, the city name too must have been altered in the revised text. Presumably Ashur city took that role when Ashur was the hero-god, and Nineveh when Ishtar of Nineveh took the leading part. Written in the reign of Sennacherib, the small part of the Assyrian version that is extant shows that he made deliberate modifications to the composition as part of his policy to elevate Assyria to supreme status, using techniques that were already in use. Although the wording changed in those particulars, the myth was essentially the same and had the authority of past scholarship.21

  The change was associated with the building in Nineveh of an akītu-house, a temple of the New Year festival, for Ishtar of Nineveh in her city.22 In this instance the manipulation of a traditional text shows clearly a central plank in Sennacherib’s policy of making Nineveh supersede Babylon. With his new version recited at the New Year festival in Nineveh, and with that adaptation reflected on the great bronze doors of his New Year festival temple at Ashur city, there was nothing surreptitious about his reform. It allowed Ashur to take the place of Marduk as Bēl ‘Lord’ at the city Ashur, and Ishtar of Nineveh to take the place of Marduk as Bēl ‘Lord’ at Nineveh. The change was visible to the public, not just embedded in text, and is no longer thought to indicate a short-lived or unsuccessful reform.23

  A different kind of explanation, not linked directly to the reign of Sennacherib but pointing to a tradition that he would have known, comes from evidence that various Babylonian cities in southern Mesopotamia could be called ‘Babylon’. It was an indigenous tradition dating from at least the 12th century BC, and perhaps even earlier. The first inkling came from several lexical texts—the dictionaries of ancient Babylonia—dated to that time or earlier. In two of them the city of Borsippa, which lies quite close to Babylon on the Euphrates, was called ‘Babylon the second’ or ‘another Babylon’. Also, a lexical text which listed gates, temples and quarters of Babylon city gave the information that the city of Eridu, located near the sea in the far south of the country, was likewise a ‘Babylon’.24

  This phenomenon can be explained through a historical development. More than a millennium before Sennacherib lived, king Hammurabi of Babylon conquered powerful rivals and became master of the lands from the Arabian Gulf to the borders of Assyria. He was sixth in a dynasty that had been until then quite parochial, but nevertheless had succeeded in establishing a reasonably stable line of city rulers, whose patron god Marduk was hardly important beyond his city. Hammurabi’s conquests, and his ambitious programme of law and education that accompanied them, enabled him to promote his city. Even before then he had associated his kingship with the oldest city, Eridu, by holding his coronation ceremony there. Either then or later in his reign he manipulated the theology of Babylon to assimilate Marduk with the god of Ku’ara, a city near Eridu, in the deep south. This god was Asalluhi, a god of magic and of the powers of healing closely connected to magic in the world of that time. In this way a powerful new dimension was added to the character of Babylon’s agricultural god. At that time or a little later, Babylon began to incorporate the gods of those cities into its own walls, by setting up branch temples. This move allowed the area of Babylon in which a branch temple stood to bear the name of the parent city, so that districts in Babylon were named ‘Eridu’, ‘Ku’ara’, and so on. In exchange the parent city might be called a ‘Babylon’.

  This strategy for promoting Babylon and securing the goodwill of older cities is particularly clear if we look at variation in the Sumerian King-list. A century or so before the lifetime of Hammurabi several rather different versions were composed. The aim was to show an unbroken line of kingship bestowed by the gods, securing legitimacy for the current ruler, so each version of the list was slanted towards a particular city. One version gives Eridu as the city which first received kingship from the gods, and another gives Ku’ara as the place to which kingship first descended from heaven. A third variant maintained that Babylon itself was the first city with a king.

  Eridu, as a name for a district of Babylon, was the most sacred area which encompassed the great temple and ziggurat of Marduk. We do not know if this happened in Hammurabi’s time, for our evidence comes from later in the second millennium. What is certain is that Babylon incorporated ‘Eridu’ and ‘Ku’ara’ into its innermost citadel area as districts bearing those city names. The great Epic of Creation refers to this tradition of incorporation in the portentous declaration of Marduk, whose words celebrate the building of Babylon as the first city to be built, following the defeat of chaos:

  The Lord invited the gods his fathers to attend a banquet,

  In the great sanctuary which he had created as his dwelling.

  ‘Indeed, Bab-ili (“gate of gods”) is your home too!

  Sing for joy there, dwell in happiness!’

  Eridu and Ku’ara are not the only two city-names that were given to districts in the heart of Babylon city. Another is Kullab, the name of a part of Uruk, ruled by the legendary hero-king Gilgamesh. In the case of Uruk, we know for certain that cultic rituals appropriate to that city were replicated in Babylon.

  From a lexical text and from variants of a king-list, as well as the names of areas within the city, it is certain that Babylon enfolded those other great cities: Borsippa, Eridu, Ku’ara, and Kullab-Uruk. The declaration spoken by Marduk in the Epic of Creation had a solid basis in the temples and rituals of Babylon, to which the names of its districts refer. This situation was a source of ambiguity which occasionally needed to be clarified by special emphasis in expressions of Babylonian kings: ‘The land of Babylon which is within Babylon’, and ‘the land of Kumar (Ku’ara) which is within Babylon’.

  An example of a name given to more than one city as a result of reforms comes from Egypt. Several centuries earlier than the time of Sennacherib, it dates from the New Kingdom, and is linked to the changes made in the time of Akhenaten when not only the city he built at Amarna, but also Thebes and probably Memphis too, were referred to in the same period as ‘Horizon of the Aten’.25 This was an epithet that did not replace the traditional name of each city. As a deliberate act, the naming would have given a sense of cohesion to an even more extenuated kingdom than that of Assyria and Babylonia. By analogy we may suggest that ‘gate of gods’ bab-ilī was an epithet for certain cities in Babylonia and Assyria, although there is no direct evidence for its use in any Assyrian inscriptions from the time of Sargon and Sennacherib onwards.

  Unexpectedly specific evidence that a ‘Babylon’ was located in Assyria came to light very much later, from medieval texts on astronomy. An astronomer named Azarqiel,26 working around AD 1070, incorporated much older information into the ‘Toledo tables’. His text referred explicitly to ancient observations taken from Mesopotamian astronomical data, supposedly still available in the Middle Ages, giving a standard figure for the longest day of the year, claiming it came from ‘Old Babylon’. Such a figure is specific to the latitude on which the observations were originally taken, and it is demonstrably not that of Babylon but somewhere in Assyria, in the vicinity of Nineveh. The records showed that three sets of observations attributed to Old Babylon, ‘Other/Second Babylon’, and ‘New
Babylon’ respectively, were made on different latitudes. 36.06 degrees for ‘Old Babylon’ suits Nimrud, 33.04, 32.32 and 32.04 for ‘Other/Second Babylon’ suit Sippar, Babylon and Borsippa respectively, 31.18 and 30.56 for ‘New Babylon’ suit Uruk and Ur respectively.27

  Cuneiform sources showed that Azarqiel’s information could be correct. However enticing this may be, it has its weaknesses. An exact match for Nineveh is not found there, neither for Nineveh nor for Babylon. Promising though these latitude-linked observations are, the gap in time between the earliest mention of the Hanging Garden in the 2nd century BC and the Toledo tables of the 11th century AD was a huge one, and one could argue that the observations were attributed spuriously to Babylon to give them the authority of great antiquity, and that the latitudes were either coincidental or faked.

  As for the confusion of rivers: in Babylonia it was relatively easy to confuse the Tigris and Euphrates because they had many branches, and were linked by an extensive network of canals bringing slow-moving water across flat and featureless land. To an Ionian, a Greek or a Hebrew for whom swift, separate rivers between hills or mountain ranges were normal, the terrain would have been quite alien. But in the north of the country, where Nineveh lay on the Tigris, irrigation canals are not part of the scenery, and the Tigris is easy to distinguish. Since the two regions are so different, it is almost certain that Herodotus made reference to the channelling that leads from Bavian through Jerwan to Nineveh when he described the changes that ‘Nitocris’ made to turn a straight watercourse into a winding one which passed an Assyrian village three times. He misleadingly attributed it to ‘Babylon’,28 as described above, just as Diodorus Siculus wrote that Ninus founded on the Euphrates (not the Tigris) the city to which he gave his own name, Nineveh.29 It was the confusion of naming Nineveh as a Babylon that caused the confusion between the two rivers.

  Why should more than one queen be called a Semiramis, in addition to her own name? The answer to the question lies, as with confusion in naming kings, in the concept of the archetype, represented by that Sumerian word me. Queenship, like kingship, was a concept represented by insignia, an ordinance sent to mankind from heaven. The name of a very famous example of a historical queen came to be identified with the archetype, and could be used for later famous queens, as we have seen for Nebuchadnezzar. An accretion of legends is attached to the name ‘Semiramis’ in Greek texts, and the use of the name for more than one woman can be explained through that concept. She was variously credited with leading campaigns with her husband ‘Ninus’, and with building works in Babylon, among them the famous Hanging Garden: Diodorus Siculus wrote that she founded a large city in Babylonia on the Euphrates including the temple of the Babylonian Zeus and the Hanging Garden (he does not actually name the city), and Quintus Curtius Rufus wrote that Semiramis, not Bel, founded Babylon.

  The original ‘Semiramis’ was a historical queen at a time when Nimrud, not Nineveh, was the main royal residence. If you were an Assyrian early in the 8th century BC you would have known about Sammu-ramat, daughter-in-law of Shalmaneser III, wife of Shamshi-Adad V, and mother of Adad-nirari III, because she was the most powerful woman in the world of that time. You would know that she in person, contrary to the custom of queens at that time, joined her son in a campaign to Arpad in the vicinity of modern Aleppo with the result that her own name was inscribed on a royal stela, as partner in heroism with her son the king. That stela was set up on the border of Assyrian territory on the upper Euphrates, and was discovered in recent times.

  Boundary stone of Adad-nirari, king of Assyria, son of Shamshi-Adad king of Assyria (and of) Sammu-ramat, the palace woman of Shamshi-Adad king of Assyria, mother of Adad-nirari, strong king, king of Assyria, daughter-in-law of Shalmaneser, king of the four quarters.

  When Ushpilulume king of Kummuh (Commagene) incited Adad-nirari king of Assyria, and Sammu-ramat the palace woman, to cross the Euphrates, I fought a pitched battle with them: Atar-shumki son of Adramu from Arpad, together with 8 kings who were with him, at Paqarahubunu. I robbed them of their camp. To save their lives they dispersed.

  In that year this boundary stone was erected between Ushpilulume king of Kummuh and Qalparunda son of Palalam king of Gurgum.30

  The term ‘palace woman’ meant ‘queen, official consort’ at that period because the word normally translated as ‘queen’ was reserved for goddesses. The inscription shows without a doubt that Sammu-ramat campaigned with her son, which suggests that the campaigns later ascribed to Semiramis by Ctesias and others may have had some link, however tenuous or garbled, with a genuine event.

  The extent of her fame during her lifetime is confirmed by the existence of another stela, inscribed only with her name and titles, found during excavations in the city of Ashur on the Tigris, and first published in 1913:

  Statue31 of Sammu-ramat the palace woman of Shamshi-Adad king of the universe, king of Assyria, mother of Adad-nirari, king of the universe, king of Assyria, daughter-in-law of Shalmaneser, king of the four quarters.32

  The stela lacks any kind of human image. It stood in a forest of similar stelae, in a space between two city walls, perhaps as a result of clearing out a temple in much later times; the other stelae bore the names of kings and high male officials.

  Sammu-ramat is also named alongside her son in a dedicatory inscription for a new temple to the god Nabu, composed by the governor of Nimrud (ancient Calah).33 The text was written on a pair of life-size statues of divine attendants which were found in their original positions in the temple of Nabu:

  To the god Nabu, heroic, exalted son of Esagil, wise, splendid, mighty prince, heir of Nudimmud, whose command is supreme, skilled in the arts, trustee of all heaven and earth, expert in everything, wise, holder of the tablet stylus, learned in scribal art, merciful, judicious, who has the power to depopulate and resettle, beloved of Enlil lord of lords, whose might has no rival, without whom there can be no order in heaven, the merciful, compassionate, whose benevolence is good, dweller in Ezida, which is within Calah, great lord, his lord: For the life of Adad-nirari king of Assyria his lord, and the life of Sammu-ramat the palace woman his lady, Bel-tarṣi-iluma the governor of Calah … had (this statue) made and dedicated for his lengthy life, that his days might be long, his years many …

  This inscription and the two statues mark the promotion of a major Babylonian god in the heart of Assyria. They were still standing there when the temple was used during the Hellenistic period.34

  The epithet ‘son of Esagil’ links Nabu directly to Marduk whose temple in Babylon was called Esagil. At that time the kings of Assyria were on good terms with the kings of Babylon, a cordial relationship epitomized by a scene sculptured on a throne-base,35 showing Shalmaneser III clasping hands with the contemporary king of Babylon (see Plate 14). Therefore the historical Sammu-ramat would not have been involved in building work in Babylon.

  Those monuments recording the exceptional fame of Sammu-ramat are the ones that we know of; there may have been others. Long after her death, traces of her presence would still have been visible, even if the literacy required to read the cuneiform inscriptions was no longer available. Nimrud is the site where the memory of the great queen would have taken root, for Nineveh had not yet become the capital city; and the temple of Nabu at Nimrud was the main focus for memory, not least because of the statues that remained there until modern times. The ‘roughly carved limestone altar’ erected in one of its courts; ‘a considerable succession of potsherds’ found in another court, ‘the latest of which are Hellenistic’;36 and fine pottery beakers imitating the late Assyrian palace ware in the same building, show that the building had not been abandoned; its residents would have seen the statues every day. In this way the archaeological evidence supports the inference that legends based on the real Semiramis survived into Seleucid times.37

  In at least one of several ‘universal’ histories of Hellenistic times, written in Greek, the history of empires began with Ninus and Semiramis, and it
was Semiramis who conquered the first world empire alongside her consort, the eponymous founder of Nineveh.38 Hers was an empire that was to endure, according to the understanding of those historians, for 1,300 years. For that reason Alexander the Great was inspired ‘with a wish to rival Cyrus and Semiramis’,39 the first great empire-builders,40 for her fame outlived the downfall of Assyria and the following centuries of neo-Babylonian and Achaemenid rule.

  The name Sammu-ramat in the form Semiramis—by which the Greeks knew her—was used also for later historical Assyrian queens of great repute, causing much confusion among Greek historians who tried to trace the history of Assyria at a time when stories had already merged.41 The most significant of the great queens lived about a century later—Naqia the wife of Sennacherib, mother of Esarhaddon, and grandmother of Ashurbanipal and Shamash-shum-ukin who became kings of Assyria and Babylonia respectively. Naqia recorded her own building of a new palace at Nineveh for her son;42 and at Nimrud she made sacrifices in the temple of Nabu founded by her predecessor Sammu-ramat, where she associated herself with the historical queen in dedicating a statue of herself, decorated with gold. She placed it in close proximity to the inscription of Sammu-ramat, which would have been a century old at that time.43 She contributed publicly to the restoration of the great temple to Nabu in Borsippa near Babylon, supplying gold for making a crown for the god’s statue.44 Thus she was associated with acts normally acknowledged as deeds of the king alone. She owned estates east of the Tigris, on the border with Babylonia, in her own name.45 She made public donations for work on temples in Harran, and is known from a bronze sculpture in low relief to have shown herself as a public figure with her royal son (or grandson)—perhaps the ‘bronze likeness of Ninus and Semiramis and their officers’ which Diodorus says graced the palace walls (see Plate 15).46 The panel has an inscription engraved on it identifying her, and may well be the very work of art that lies behind his account.47 Such panels may be the kind of relic of past glory that evokes new stories, reinterpreting antiquities in the visible landscape to create new versions for local tradition.