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

by Stephanie Dalley

  The passage leaves one in no doubt about the king’s personal enthusiasm and commitment to bronze-casting techniques.

  Later use of a water-raising screw in the ancient Near East (apart from Egypt) is known from the upper Euphrates region, at Ayni, between Carchemish and Samsat, an area which had long been a part of the Assyrian empire. There a Roman inscription showed that Vespasian and Titus had constructed a screw, opus cochliae, complemented by a rock sculpture showing the Euphrates personified; but the method of rotation there is not indicated (see Figure 24).38

  Fig. 24 Sketch of an eroded rock sculpture at Ayni on the upper Euphrates, set beside a Latin inscription recording the installation of a water-raising screw in the reign of Vespasian. The river-god reclines with a pot from which water flows.

  One possible reference to rotating a screw by foot in Egypt may be mentioned. From a Roman wall-painting we know that a screw set at quite a low angle could be rotated by a man standing on the cylinder fitted with foot-holds, and treading it to turn it.39 This, it has been suggested, is the mechanism behind the image given in the biblical text of Deuteronomy 11:10: ‘Egypt … where you sowed your seed and watered it by tread, like a vegetable garden.’40 There is a more specific and amusing account of the treading process in the writing of Philo Judaeus, a philosopher of Alexandria, who was contemporary with Josephus. He delighted in a reversal of the normal use of hands and feet in the operation of a screw:

  Compare the screw, the water-lifting device. There are some treads around the middle onto which the farmer steps whenever he wants to water his fields, but of course he would keep slipping off. To keep from continually falling, he grasps something sturdy close by with his hands, and clings to it, suspending his whole body from it. In this way he uses his hands as feet and his feet as hands, for he supports himself with his hands which are generally used for working, and he works with his feet which normally serve as supports.41

  But this method of rotation would work only for relatively small, light machines, and Sennacherib’s inscription certainly does not refer to treading. In the years since a water-raising screw cast in bronze was understood to make sense of the technical words in Sennacherib’s inscription, several suggestions have been made as to how the screw was rotated. Clearly the treading method is not described, for the text speaks of ropes, chains and cables or wires; but whether those words should be taken literally, or whether they are a rhetorical triplet for a single item rather than three separate ones, is impossible to decide. Sennacherib evidently did not try to describe a mechanism with a pulley—for which evidence is described below—nor did he attempt to describe a wheel-type of mechanism with cogs. The word for the wheel in Akkadian is well known to be mugerru; as for cogs, which are required to alter the direction of force from an awkward angle to one that maximizes strength, one would expect to recognize the word for teeth in conjunction with the word for wheel; but neither word is there. To the objection that cogs are known only from much later times—the chance discovery of the Antikythera mechanism, described below, pushed back the date for the invention of differential gearing by many centuries—one may point to the discovery by excavation in north-west Iran, in two tombs at Marlik, of what appears to be parts of a gear mechanism, described by the excavator as ‘Bronze Gears(?)’, datable to between 1250 and 883 BC.42

  A word of explanation is in order, to give some background from the astounding Antikythera mechanism. Although wooden gears in the form of ‘teeth’ cut from the radius of a disk were certainly known by the 3rd century BC,43 sophisticated, differential gearing was thought to come very much later, until a clump of bronze, badly encrusted and corroded, was found by sponge divers underwater in a shipwreck off the Aegean island of Antikythera. It was dated to the 1st century BC. Retrieved at the beginning of the 20th century, it could not be investigated until new techniques, especially high resolution X-ray tomography, were developed to reveal its 30+ gears, a crank, and inscriptions amounting to a manual with an astronomical, mechanical and geographical section. Machines of similar complexity are not known from elsewhere for the next 1,000 years.44 The level of complexity in the machinery took everyone by surprise, and has shown that scholars had underestimated the possibilities of complex gearing in antiquity. We simply do not know how far back in time it may go. Therefore we cannot say for certain that a simple cog was not known in Sennacherib’s time, but we can be sure that it is not mentioned in his inscription in connection with the rotation of the screw. It is unlikely that the ropes and chains of Sennacherib’s text were linked to a cog mechanism that was not explicitly mentioned in the prism passage.

  After the television documentary on the Hanging Garden was shown,45 various people wrote in with suggestions for the method of rotation, but all of them involved cogs or pumps. However ingenious their proposals, they cannot be adopted because there is no support for them, neither in the inscription of Sennacherib nor in the Greek sources, nor in the technologies known to have existed at that time. In the case of the water wheel,46 not only is the lack of the word for a wheel significant, but even more telling is the objection that such machines are eminently visible in the landscape; they would surely be shown on Assyrian sculptures just as shadufs are, and would have been mentioned by the Classical authors; their high visibility is impossible to reconcile with the words of Strabo: ‘raising the water … although no-one outside could see it being done’. My own early suggestion of a crank handle, shown in the drawing I published in 1994, incited derision and is likely to be wrong: there is no evidence for that mechanism at such an early date, and no support for it in Sennacherib’s inscription. Despite several ingenious speculations the mechanism for rotating the water-raising screw remains unknown.

  Some scholars have suggested other ways of understanding the technical passage in the prism inscription, since from their point of view there is no question of using Classical sources, which they considered to be legendary and fictional, to understand the mechanism described in the cuneiform text. They turned back to earlier attempts to extract a meaning solely from the words used in Sennacherib’s prism inscription, and envisage, rather non-specifically, a framework of beams and pillars with ropes and chains, perhaps resembling the cerd. The two flaws in such an argument are as follows. As already mentioned, the shaduf, like the screw, is used to raise surface water from lower to higher ground, not to raise water to the surface vertically from a deep well. By contrast, the cerd is used to raise water from a well, and requires buckets.47 The phrase ‘instead of a shaduf’ implies that the new mechanism does the same job as a shaduf. Besides, to bring water up from a deep well was a familiar task on the citadels of Mesopotamian cities, and not a matter for description in a royal text. As physical evidence for drawing well water, a well-head with a pulley and rope is shown on the Assyrian panel of stone bas-relief found at Nimrud;48 and pulleys were retrieved from a deep Assyrian well, also at Nimrud.49 Thus there was no reason to invent anything for drawing well water.

  A potential difficulty in the prism inscription arises from the disjointed nature of the text where mention of the castings and the garden occurs. This is the incoherent sequence given in the relevant part of the text: quarrying of stone and making colossal sculptures; casting of bronze by the new method to make ‘cylinder and alamittu-palm’ and pillars and colossal bronze sculptures; details of metal sheathing on cedar pillars, inlay on wooden pillars, bas-reliefs around walls and calling them a wonder; provision of water from cables, ropes and chains and setting up ‘cylinder and alamittu-palm’ over cisterns instead of shadufs; making ‘those palaces’ beautifully; raising the surrounds of the palace, calling it a wonder for all people; creating a garden to be like the Amanus mountains, planting it with exotic plants. Thus the mention of castings and machines for raising water is not adjacent to the mention of the garden. The disjunctions are due to the cut-and-paste method used for compiling long inscriptions of that kind. Similar disjunctions have been recognized in other inscriptions, and are disc
ussed in detail in Chapter 7.

  One might expect that the technical terms used by Sennacherib in connection with the water-raising device would be found in one of the great lexical lists of cuneiform tradition. Those lists are forerunners of dictionaries, in which words are grouped in several different ways. We have sections dealing with ropes and chains; with the parts of a shaduf;50 and alamittu is found in several sections on date palms.51 Since many such lists were not updated in the first millennium, we would not expect them to list machinery that was invented after the time they were compiled.52

  Luckenbill’s translation ‘park’ for kirimāhu led to the suggestion that the garden was in fact a hunting park, which would have required too much space to be located on the citadel, and as we have seen, the famous garden was on the citadel. The usual word for a hunting park is ambassu, which Sennacherib did not use. He selected kirimāhu, a rare literary word derived from Sumerian, which can be understood to mean ‘high garden’, according to its setting high up on the citadel, or metaphorically ‘great garden’ in the sense of important rather than large. Sargon was the first to coin the word for his palace garden, whereas hunting parks are known much earlier from the inscriptions of previous Assyrian kings. Therefore the word ‘park’ has been replaced by ‘high garden’ in this book.

  The term KIRI.MAH had great prestige in Sumerian, which is doubtless one good reason why Sargon and Sennacherib chose to use it in its Akkadian rendering. Way back in the third millennium BC it had been used to described a garden attached to a temple of the sky-god of Uruk, built by the Ur-Nammu who founded a great dynasty at the southern city Ur.53 Even earlier it was used by king Nanne, father of Mes-kiag-nunna, legendary king of the earliest dynasty at Ur, to denote a temple garden associated with the great goddess Ninlil at her cult centre Tummal, close to Nippur, a city regarded as the point where the umbilical cord that attached earth to heaven was located.54 By choosing to use the word, the two Assyrian kings were associating themselves with three of the greatest Sumerian cities of very ancient times. Alas, we know nothing about those two gardens, but we may suggest that they were set high up on citadel mounds, since that is where great temples were located at that time.

  The design and planning of Sennacherib’s South-West Palace and the garden are likely to have taken place simultaneously so that the required levels of terrain could be laid out to take into account the inflow of water from the aqueduct, its supply to the palace (with appropriate pipes) and its drainage through the lake at the bottom of the garden, and into the river Khosr at the foot of the citadel wall. By the late 8th century BC Assyrian palaces were provided with a comprehensive system of pipes and drains laid beneath floors and walls according to a design that would have been installed while the foundations were being laid, before floors and walls were put in place. Sennacherib would have inherited from his father the expert surveyors, engineers and builders whose experiences gained at Khorsabad were invaluable to him (see Figure 25). In all three of the main requirements—creating an artificial hill, bringing water from a far distance by diverting and linking several mountain streams, and installing terracotta pipes and baked brick drains—Khorsabad had been a brilliant testing ground for a work of architectural genius.55

  Sennacherib the king alone takes the credit for his palace, his garden and his water supply. A key ingredient for the inspiration and success of the project was royal patronage with the king in person taking a close interest, and probably playing a part in both general and detailed discussions. But the identity of the architect is unknown. In modern times the architect is separate from the quantity surveyor, the engineer and the clerk of works, owing to increased specialization which particularly divides theory from practice. Even among architects nowadays one speaks of ‘concept architects’ as different from ‘delivery architects’. Assyrian texts, whether public royal inscriptions or administrative records with restricted circulation, hardly ever mention those professions, let alone the individuals who practised them. Exceptional, therefore, is the mention of some of those professions in the royal inscriptions of Sargon and Sennacherib. In the context of his palace and garden at Nineveh, built ‘through the craft of clever master builders’,56 Sennacherib mentions the ‘master builder’ (šitimgallu). A long tradition of water engineers in Mesopotamia is represented by the profession sēkiru, who had his own seal, an indication of centralized authority, and was also known as ‘builder of watercourses’ ŠITIM ID.DA. The ‘man in charge of pipes/qanats’ (ša eli qanâte)57 was different in being a new profession introduced in Sennacherib’s time. There would have been a close relationship between architects, builders and water engineers, matching the need to plan the integral water supply and drainage systems for the public buildings on citadels.

  Fig. 25 Brick-built drains in Sargon’s palace at Khorsabad.

  Ground plans for major buildings and canals could be drawn out on plaster or on stone; the famous example is part of a statue of Gudea, ruler of Lagash in the late third millennium BC, who held on his lap the plan of the temple described in the accompanying inscription. Occasionally one finds a sketch of fields and canals on a clay tablet.

  To judge from the often rather irregular lines and angles of walls uncovered by excavation, much large-scale building was done by eye from a schematic understanding that was structured loosely around a cosmological symbolism. The symbolism is known from the vocabulary used in the inscriptions of the royal builder.58 Often modular units were repeated, with standard types that seem to imply the use of a pattern book;59 and the discovery in different cities of several palace courtyards paved with alternating black and white squares of river pebbles,60 precursors of mosaics, seems also to imply a centralized approach to design (see Plate 8). A metrological tablet of late Assyrian date gives the measurements of two cross-sections of a temple from gate to gate; other, later tablets show that the arrangement of a complex pilaster was described in terms of the number of standard bricks needed for shallow and deep pilasters, and for the stepped framework of a gateway (sometimes called rabbeted jambs), if the technical vocabulary of the cuneiform text has been understood correctly.61 The dominant requirement was to order enough bricks of appropriate sizes and shapes for an operation, which implies an arithmetical rather than a geometrical approach. We possess no manuals, no templates, no drawings of moulding profiles. It is likely that some of the rather rough mathematical tablets with multiplication tables found on clay tablets were intended for scaling up from a small sketch, but many other uses are possible. Much of the design would have been worked out on site for individual buildings, but far more planning would have been needed for the whole complex of Sennacherib’s palace, garden and aqueduct.

  For those critics who have suggested that the water-raising screw to which Strabo and Philo refer was a Hellenistic or Roman improvement, a late installation introduced into a revived version of the famous garden, it must be emphasized that the interpretation of Sennacherib’s inscription shows that the screw was part of the original design, and part of the reason why the garden was a World Wonder both in Sennacherib’s day and centuries later when Hellenistic authors wrote about it. The water-raising screw is miraculous in appearing to reverse gravity, to overcome the natural imperative of water to flow downhill.

  5

  Engineering for Water Management

  Consider Assyria, a cedar of Lebanon,

  Of beautiful branches, a shading wood, and of lofty stature;

  Whose crown was among the clouds.

  Water made it grow, the Deep made it tall;

  Her rivers she made flow around her planting,

  But her canals she sent forth to all the (other) trees of the field.

  Ezekiel 31: 3–41

  More than 2,000 years of accumulated expertise lay behind the engineering required for the supply of water to Nineveh, its orchards and its citadel with the Hanging Garden. An effective control of water was always crucial to the growth and continuity of cities and agriculture in Mesopotamia
. Although much of the know-how must have been local, improvements over the course of many centuries were sometimes adopted from foreign practices and expertise; the Assyrians were eclectic and appreciative of foreigners’ skills which they encountered in their far-reaching expeditions for trade and conquest, and had no inhibitions about importing experts from other countries or adapting anything foreign that was better than their own.

  In the case of Nineveh, the city had long played host to kings. There Tiglath-pileser I (1114–1076 BC) completed a palace which his father had begun to build, probably on the site where Sennacherib later built his South-West Palace, and made a garden on the citadel:

  beside that terrace I planted a garden for my royal leisure. I dug a canal from the river Khosr and [directed its water] into that garden. I brought the rest of that water into the outskirts of the city for irrigation. Within that garden I built a(nother?) palace…’2

  As conqueror of Babylonia, Tiglath-pileser I had a huge labour force at his disposal, ‘prisoners without number’, so much of the landscaping of the citadel was already in place when Sennacherib undertook to replace the old palace, garden and canal on a more ambitious scale.

  When the canals, weir, tunnel, dam and bridge-style aqueduct with corbelled stone arches, all designed and constructed by Sennacherib’s engineers, were traced and published in 1935, scholars in the field of Assyriology were amazed that such an extensive and well-planned scheme existed in the early 7th century BC (see Plates 11–13). It was not then fashionable to be concerned with the economics of ancient Mesopotamian cities; art and architecture, alongside texts, were the main focus of study. Since then new emphases on town-planning and supply have raised awareness of water engineering in early times, so that Sennacherib’s achievement can be set against a wider background of knowledge. A few installations dated earlier than his time are briefly described here, in order to show that certain advances can no longer be credited to Achaemenid Persians, or to Romans, as they have been in the past.