A puzzle with no confirmed site
No remains have been confirmed as the Hanging Gardens, and whether they stood in Babylon, Nineveh or only in later storytelling is unresolved. This article therefore treats the gardens as a thought experiment: if a planted, terraced structure like the one in the Greek and Roman descriptions had been built in Mesopotamia, what could ancient builders have used? Everything below is about what techniques were available, not proof that they were used.
The challenge is substantial. A garden raised above ground needs a structure strong enough to carry soil and trees, a way to keep water from soaking and weakening the structure, and a way to lift water from a river or canal to the top. Each of these problems had known solutions in Mesopotamia.
Building the structure
Mesopotamia lacked stone and good timber, so a raised garden would almost certainly have been made of brick. Ancient writers who describe the gardens mention terraces on vaulted supports. Mud brick is cheap and good in compression, but it weakens when wet. Baked brick is stronger and more water resistant but costs more, because fuel was scarce.
Mesopotamian architects already built very large masses of brick in ziggurats, city walls and platforms, so a terraced structure of brick was not beyond their skill. Ziggurats are in some sense terraced mountains, and comparable terraced building appears elsewhere, for example in the Machu Picchu terraces of the Andes, though built by different means and for different purposes. Brick vaults and arches were also known in Mesopotamia by the second millennium BCE, though they were used mostly for drains and tombs and not for wide halls. A planted terrace on vaults would have been an ambitious use of the technique.
Waterproofing
Keeping the soil’s moisture away from the brick was the central problem. Mesopotamian builders knew a good material: bitumen, a natural tar that seeps from the ground in parts of Iraq. It was used to waterproof bathrooms, drains, boats, and the joints in baked brick. Some descriptions of the gardens mention asphalt, reed matting and lead sheets as layers beneath the soil. Reed matting is a plausible ingredient, since reeds were abundant and were used with bitumen in other waterproofing. Whether these classical details are reliable is unknown, but the combination matches known practice.
A well-designed garden would also need drainage, so that excess water did not build up and weigh down the structure. Planted roofs and terraces in the modern world also use layered waterproofing and drainage, so the problem itself is a real one.
Soil and plants
The depth of soil matters. Large trees need more depth than shrubs or vines, and the weight of wet soil is considerable. A cautious builder might have limited deep soil to areas over thick walls or piers and used lighter plants elsewhere. Classical writers speak of trees growing in the gardens, which would have been the hardest part to support. We do not know how true these claims are, and some scholars suspect the descriptions are exaggerated.
Lifting water
The most debated element is irrigation. Mesopotamian farmers used canals and ditches to bring river water to fields, a system with deep roots in Sumerian irrigation farming, but a raised garden needed water lifted vertically. Several devices were available at different times, and the dating of each is important:
- The shaduf, a counterweighted pole with a bucket, is attested in Mesopotamia by the late third millennium BCE in seal images. It lifts water by a small height with human power, so it could be used in a series of steps.
- Rope-and-bucket systems with animal power were also possible, though specific evidence for Babylon is thin.
- The water wheel and the chain of pots, in which pots on a rope or chain are turned over a wheel, are usually dated later, to the Hellenistic and Roman periods, and so would fit a late tradition better than an early garden.
- Water-raising screws: the Greek tradition connects the screw with Archimedes in the third century BCE. Stephanie Dalley has argued that Assyrian bronze-cast devices of the seventh century BCE may have been screws, which would push the date earlier. This reading is debated.
The idea of a Euphrates garden depended on a river that rises and falls with seasons, with highest water in spring and early summer after the snowmelt in the north. A reliable system would need to adapt to those changes, perhaps by drawing from a canal or a cistern filled when water was high.
Comparison with Assyrian water engineering
Assyrian royal inscriptions record large water projects. Sennacherib’s canal and aqueduct systems at Nineveh, including a long stone aqueduct at Jerwan, show that Mesopotamian engineers could move water across land by gravity. A garden watered from a stream higher up, as shown in Assyrian palace reliefs, avoids the need for lifting devices. This is part of the reason some scholars find the Nineveh theory attractive. The Roman aqueducts that came later used similar principles on a larger scale, but there is no direct link.
Weighing the hypotheses
Taken together, a brick-and-bitumen terraced garden watered by manual lifting devices or by gravity-fed canals appears technically possible for a Mesopotamian builder. That is not evidence that it happened. The practical difficulties, including the weight of soil, constant maintenance and the effect of seasonal floods, suggest that any garden was likely more modest than the later descriptions. The question should remain open until excavation or newly read texts show where, and whether, such a structure stood.