Summary

A ziggurat or ziggurat is a stepped Mesopotamian religious edifice, also present in Elam, consisting of a stack of several high terraces probably supporting a temple built on top. The term comes from the Akkadian ziqqurratu(m) (feminine, sometimes abbreviated to ziqratu, in Assyria siqurratu or sequrattu, in Sumerian ideograms U6.NIR), derived from the verb zaqāru, "to raise", "to build high". It can therefore be translated as "the very high". This is a type of monument characteristic of Mesopotamian civilization, whose memory survived long after its disappearance thanks to the biblical story of the Tower of Babel, inspired by the Babylonian ziggurat.

Since the discovery of the great Mesopotamian capitals, it has been possible to analyze many of these buildings, although not all remain intact. Many are very dilapidated and look like hills, while others have completely disappeared. The Mesopotamian civilization also left few descriptions of them, either in texts or images. Some ziggurats (above all those in Babylon) are mentioned by Greek authors (Herodotus and Ctesias). While their general appearance is now fairly well known, there are still some grey areas as to their meaning and function, in the absence of any explicit text on the subject.

From terraced temples to ziggurats: the question of origins

Historically and architecturally, ziggurats are commonly seen as heirs to the cult buildings built on terraces in Lower Mesopotamia. This lineage has long been criticized, but now seems to be accepted. However, the boundaries between buildings described as ziggurats and those that preceded them are not defined in the same way by specialists.

These buildings appeared in the course of the 5th millennium B.C., consisting of the raising and widening of high brick terraces, supporting monumental edifices identified as temples, with a tripartite plan characteristic of this period, which undoubtedly concentrated the main rituals of divine worship; they are attested throughout the Mesopotamian area, and also in Susiana. The earliest example of a building erected on a terrace that could be interpreted as a temple is attested at Eridu during the Obeid period, around 5000 A.D. It consists of four successive constructions (levels IX to VI) of increasing size over time and with a tripartite plan, common at the time, but situated on a platform more than one meter high. This type of low-terrace construction is common in protohistoric Mesopotamia (notably in the buildings of Uruk IV and III, second half of the 4th millennium, during the Recent Uruk period and the Djemdet-Nasr period), a context in which certain constructions are distinguished by the fact that they are erected on an increasingly high terrace (roughly more than two meters high). Such is the case of the Uruk edifice called "Anu's ziggurat" by the site's excavators, a high terrace supporting a remarkably well-preserved temple (the "White Temple"), already preceded by similar edifices dating from the Obeid period. The best-preserved high terrace temple was unearthed at Tell Uqair in Lower Mesopotamia. It dates from the end of the Uruk and Djemdet Nasr periods (late 4th millennium). It consists of two superimposed terraces, the first with a curvilinear façade while the second is rectangular, on which is built a building interpreted as being a temple, still partly preserved.

In the 3rd millennium (period of the Archaic dynasties), a temple on a terrace was apparently built in Uruk's second sacred district, the Eanna, but its ruins are covered by the later ziggurat and therefore poorly known. Another similar building from the same period is the "Oval Temple" at Khafadje, in the Diyala valley, of which all that remains is the rectangular terrace decorated with pilasters, measuring 25 × 30 metres, still 4 metres high, with a perpendicular staircase leading to the temple at its summit, which has now completely disappeared. This edifice takes its name from the two oval enclosures isolating it from the rest of the city. Another temple-supporting terrace, whose foundations have now disappeared, was uncovered in the south at Tell Obeid. Other terraced temples from the same period are attested in Upper Mesopotamia and Syria, notably at Tell Brak and Tell Mozan.

Some sites on the Iranian plateau in the 3rd millennium feature monumental constructions comprising several superimposed terraces: at Tureng Tepe, Tepe Sialk and Konar Sandal in Iran, and as far afield as Mundigak in Afghanistan and Altyn-depe in Turkmenistan. Although sometimes still referred to as "ziggurats" by their excavators, there is no evidence that these buildings are related to Mesopotamian terrace temples, from which they diverge in many respects. In any case, the links between the two types of construction remain little studied, not least because little is known about Iranian terraces.

When did the first buildings that could be described as ziggurats appear? In terms of terminology, the term ziqqurratu(m) only appears at the beginning of the 2nd millennium, after the construction of the first buildings of this type. The pre-Ur III texts of several kings of Lagash, notably Gudea, mention constructions designated by the Sumerian term GI.GÙ.NA (or GI.GUNU4), which should perhaps be identified as temples on terraces, and more precisely the temple built on the terrace, since this term was later used in Akkadianized form gigunû to designate the temple overlooking the ziggurat. The term may originally have referred to reed sanctuaries (meaning GI) erected on terraces. Constructions resembling ziggurats (i.e. multi-storey buildings) appear on cylinder seals as early as the end of the Uruk period and the time of the Archaic Dynasties, as well as in the Akkadian period, but there is no confirmation that these were indeed terraced temples, as they could have been degree altars or other cult constructions.

Following the typology inherited from Lenzen, it is customary to distinguish between buildings built on a single terrace and those erected on several storeys, which would be ziggurats in the strict sense of the term. A. Parrot also seems to focus on the number of terraces: as soon as there are three, it would be a ziggurat, which would make those of the Ur III period the oldest (assuming they did have three terraces, see below). However, according to the same author, the earliest ziggurats date from earlier periods, as he considers the stepped constructions depicted on archaic cylinder seals mentioned above to be ziggurats. However, the distinction between ziggurats and earlier monumental terraced temples may appear artificial, insofar as the relationship between them is difficult to contest and makes them similar in nature. In fact, some of these terraced temples from the final period of the Archaic Dynasties may have been described by their excavators as ziggurats, as in the case of Lenzen's "ziggurat of Anu" at Uruk (because it has two storeys), or those at Kish. In any case, it is more commonly accepted that the constructions built in the major religious centers of Sumer by Ur-Nammu of Ur and his successor Shulgi around 2100 are genuine ziggurats, even if they do not use this term (or any other precise term) to designate them, and that they must therefore be distinguished from the ancient temples on terraces, even if some maintain that there were ziggurats before.

Stepped building is not a new architecture in the ancient Near East, as it has been implemented and repeated by the ancient Egyptians since the era of the third dynasty (c. 2700-2600 BC). According to O. Kaelin, there were Egyptian influences on Mesopotamia, with King Ur-Namma adopting the idea of tiered construction, thus breaking with the traditions of temple architecture, and introducing tiered construction to Mesopotamia with the first building, after several generations and centuries of establishing interest in the Egyptian model.

Terrace temples from the Third Dynasty of Ur

Whether or not the kings of the Third Dynasty of Ur (21st century) were the "inventors" of ziggurats, this period was in any case decisive for the future success of the terraced temples that came to be known as ziggurats. From the beginning of the Third Dynasty in Ur, the major cult centers of Lower Mesopotamia were gradually equipped with these buildings, which were designed according to the same model, although they were not strictly identical: these ziggurats had a rectangular base, with a large central perpendicular staircase establishing a symmetrical principle in their appearance, and were built within an enclosure. From an architectural point of view, therefore, they take up the heritage of earlier terrace temple traditions and reformulate it into a well-thought-out type of building.

These constructions were apparently initiated by the dynasty's founder, Ur-Namma (2112-2094), and continued by his son and successor Shulgi (2094-2047). At least four ziggurats were built in the main religious centers of Sumer, where the dynasty originated: Ur and Nippur. It would appear that there was also one at another important site of the period, Tell Drehem (Puzrish-Dagan), where a characteristic mound of ziggurat ruins has been identified (but not excavated), and perhaps at another major place of worship, Larsa.

These buildings are built on the same principle: terraces, three according to the most common reconstruction based on Leonard Woolley's work at Ur, or two according to the alternative proposed by Schmid, stacked on top of each other and probably supporting a temple, reached by two lateral staircases parallel to the base and a large central perpendicular staircase, but with a different orientation. At least two of them (those at Eridu and Uruk, and the other two with traces of earlier constructions) follow in the footsteps of earlier terraced temples, which would seem to confirm the relationship between the two types of building. The Akkadian term ziggurat does not appear at this time, while the Sumerian term appears only as the ceremonial name of the Eridu terrace temple, É-U6.NIR (and it is perhaps from this that this name designated all buildings of the same type). So, even for this period, it's probably anachronistic to speak of "ziggurats".

These buildings required the development of new construction techniques and the mobilization of many workers. If we look at the context in which these buildings were built, we can see that they were part of a policy of major works implemented by the rulers of this veritable empire, which dominated the whole of Mesopotamia at the time, and was served by a bureaucratic apparatus and a host of dependents whose numbers had never been seen before. This explains why these four ziggurats were built according to an almost standardized model, as it were "in series". More broadly speaking, the kings of Ur III placed particular emphasis on the religious aspect of their role, highlighted in several royal hymns and by their "divinization", and the building of the ziggurats under this dynasty is undoubtedly to be placed in this ideological context.

Ziggurats in the 2nd and 1st millennia

After the collapse of the Third Dynasty of Ur around 2004, the construction of terraced temples continued under the impetus of the kings of Amorite origin of the Lower Mesopotamian states of the early 2nd millennium, who were accustomed to adopting the traditions inherited from their prestigious Sumerian predecessors. It is often difficult to know to whom to attribute the construction or reconstruction of ziggurats excavated on sites, which archaeology can generally only date to a general period. Foundation inscriptions and other royal texts commemorating the construction or restoration (the difference is sometimes difficult to grasp from the texts) of a ziggurat can help to pinpoint the date of work observed during excavations, but these are only available for a minority of cases. In any case, it was around this time that the term ziggurat first appeared.

The Amorite kings who were most active in ziggurat construction were those of the first Babylonian dynasty. The most famous of these, Hammurabi (1792-1750), rebuilt the Ebabbar ziggurat at Larsa (dedicated to the sun god Shamash), as reported in a later inscription by Nabonidus, who in turn restored this edifice. But it was his son Samsu-iluna (1749-1712) who left texts on the construction of ziggurats: a foundation inscription commemorating the construction of that of the Ebabbar of Sippar (another great sanctuary of the sun-god), also celebrated in the name of his eighteenth year of reign; and another reporting work on the ziggurat of Kish dedicated to Zababa and Ishtar. The first state of the Babylon ziggurat is also attributable to one of these rulers, as are probably those of the nearby cities of Borsippa and Akkad. Further north, ziggurats may have been built in Central Mesopotamia in the kingdom of Eshnunna, and certainly in Upper Mesopotamia around the beginning of the 18th century at Tell Rimah (probably the ancient Qattara) and at Assur for the city's tutelary god, Assur (often confused with the great god Enlil). The best candidate for the construction of these two buildings was King Shamshi-Adad I.

During the second half of the 2nd millennium, new ziggurats were built, while previous ones continued to be maintained. In Babylonia, one of the two Kassite kings named Kurigalzu (probably the first, in the early 14th century) erected one in his new namesake capital, Dur-Kurigalzu (Aqar Quf). Other Kassite kings restored ziggurats, such as one of the two Kadashman-Enlil at Nippur and Marduk-apla-iddina (1171-1159) at Borsippa, and that of Babylon was enlarged, perhaps under the reign of Nebuchadnezzar I (1126-1105). The Assyrian kings of the same period built several ziggurats at the same time as they restored the existing ones at Ashur and Nineveh. Two ziggurats known from excavations were built in the double temple of Anu and Adad at Ashur (making a total of three ziggurats identified by archaeology in this city), and another at Kar-Tukulti-Ninurta, a new city founded by Tukulti-Ninurta I (1245-1208). An inscription by the Assyrian king Salmanazar I (1275-1245) reports the restoration of several temples, including ziggurats, such as those dedicated to Ishtar at Arbèles (Erbil) and Talmussu (exact location unknown), for which no further evidence is available. At the same time, several temples were built in the Elamite kingdom (in south-western Iran today), starting with the new city founded by King Untash-Napirisha (1345-1305), Dur-Untash (Chogha Zanbil). Inscriptions from the same kingdom mention the existence of two other ziggurats during this period, in the great city of Susa and perhaps also on the site of Chogha Pahn.

The Assyrian kings of the first half of the 1st millennium repeatedly restored the ancient ziggurats, and some of them built new ones, notably in the capitals they built for their kingdoms, this type of edifice being indispensable for a large Mesopotamian city: Assurnasirpal II (883-859) had one built at Kalkhu (Nimrud) around 870, Salmanazar III (858-824) may have initiated another at Tell el-Hawa, and Sargon II (722-705) had one erected at Dur-Sharrukin (Khorsabad) at the end of the 8th century. A text by the latter king refers to his restoration of a ziggurat dedicated to Adad in Nineveh, the second to be attested on this site. Around the same time, Kudurru, governor of Nippur, took part in the upkeep of the ziggurat in the city of Der. King Assarhaddon (680-669) had the second ziggurat built or restored at Uruk, in the sanctuary of Anu. Several large ziggurats in Lower Mesopotamia were restored or enlarged by the Assyrian and Babylonian rulers of the first half of the 1st millennium. The ziggurat at Babylon, Etemenanki, is the best-known of these, and was remodeled between the 7th and early 6th centuries by the Assyrian kings Assarhaddon and Assurbanipal (669-627), then the Babylonian kings Nabopolassar (626-605) and Nebuchadnezzar II (605-562), marking the culmination of this type of construction. The ziggurat at Borsippa is the most studied by archaeologists. The inscriptions of Nabonidus (556-539) indicate that he restored those of Ur, Larsa and also the two in the sanctuary of Ishtar in Akkad, the location of which has not been determined.

So Mesopotamian texts show that there were ziggurats, traces of which have not been found even on intensively excavated sites such as Nineveh and Susa, where they were probably levelled during antiquity (or during excavations in the latter case). Lists of ziggurats classified by city have been found on two tablets whose copies date from the Neo-Assyrian and Neo-Babylonian periods, but which are probably copies of earlier texts. They list 22 and 23 ziggurats respectively for sites in Lower Mesopotamia alone, including some cities with important sanctuaries that have not been extensively excavated (Kutha, Dilbat, Marad). However, they sometimes contradict archaeological discoveries, notably giving several ziggurats for sites where only one has been identified, as at Nippur, and their interpretation is therefore difficult. In all, there may have been some thirty ziggurats in Mesopotamia (twenty with certainty) and three in Elam, built between the end of the 21st century and the 8th century, bearing in mind that others may have existed in unexcavated sites and not attested by texts. In any case, the observation of certain sites during surveys has sometimes revealed the presence of mounds resembling the ruins of ziggurats (Drehem, for example). Finally, archaeologists sometimes disagree as to whether a building is a ziggurat or a temple on a terrace, due to the problems of definition mentioned above, particularly in the case of buildings excavated long ago and dating from the Archaic period (3rd millennium): this is the case for the first "ziggurat of Anu" at Uruk, which corresponds more to a temple on a terrace, or for the two Archaic-period "ziggurats" at Tell Inghara in Kish, whose ruins are not sufficient to characterize them as such (even though this site probably had a ziggurat).

The end of ziggurats

The ziggurats of Babylonia continued to be maintained at least until the fall of the Babylonian kingdom in 539. These buildings followed the fate of the Mesopotamian religious tradition, which slowly withered away in the second half of the 1st millennium. The last major work on ziggurats was carried out in Uruk during the Seleucid period, in the middle of the 3rd century, when the ziggurat of Eanna was redeveloped and that of the new cult complex of the god Anu was built from the ruins of an older one. At the same time, it seems that the Babylonian ziggurat and the main ziggurat at Ashur continued to play a cult role. They ceased to function in the Parthian period, around 100 A.D. It is clear that most ziggurats gradually fell into ruin after the fall of the Mesopotamian empires, no longer being maintained. Some were converted for a time into fortresses during the Parthian period, at Nippur, Borsippa and perhaps Assur. All ziggurats were eventually abandoned, as were most of the large ancient Mesopotamian cities in which they were located, and their bricks were often used as building materials by the populations living near them. This has not prevented some ziggurats from remaining impressive despite the test of time, and still arousing the imagination of travellers (at Borsippa, Dur-Kurigalzu, Chogha Zanbil), while others have completely disappeared, probably as a result of urban redevelopment (Nineveh, Susa).

Shapes and dimensions

A ziggurat is a massive building built on a vast terrace (kiggallu) serving as a foundation, and composed of several (two or three to seven) solid terraces of square or rectangular plan stacked and set back from one another, forming levels (rikbu), the top floor supposedly supporting a temple. Metrological texts on the dimensions of ziggurats seem to indicate that they were based on symbolic numbers, which is at least evident in the case of the number of storeys. In practice, the shapes of these buildings varied, making them a relatively heterogeneous whole despite their similar morphology.

From archaeological excavations (which generally only uncovered the base of the building), it appears that ziggurats were built on a square or rectangular base: in the south, the first ziggurats of the Ur III period were rectangular, but over time the square shape seems to have triumphed, while in the north they were systematically square, as at Chogha Zanbil in Elam. On the other hand, there is no uniform orientation. These bases vary in size. The smallest have sides of around 30 m: 31.50 × 19 m at Tell Rimah, 36.60 × 35 m for the twin ziggurats of Anu and Adad at Ashur in the Medio-Assyrian period (reduced to around 24 × 21.30 m in the Neo-Assyrian period), 31 × 31 m at Kar-Tukulti-Ninurta, 37 × 30 m at Sippar, etc. The largest on the ground are those at Chogha Zanbil in Elam. The largest on the ground are those at Chogha Zanbil, with 105.20 m sides, and the Anu complex at Uruk in the Seleucid period, which would have had a base of 110 m sides. Babylon in its final state has a square base of around 91 m on each side. Between these extremes, we find ziggurats with a base whose sides generally vary between 40 and 60 metres: 43.10 × 43.10 m at Khorsabad, 51 × 51 m at Kalkhu, 60 × 60 m for the ziggurat of Ashur in the city of the same name, 43.50 × 40.30 m at Larsa, 56 × 52 m at the Eanna of Uruk, 57 × 39.40 m at Nippur, 62.50 × 43 m at Ur, and up to 67.60 × 69 m at Dur-Kurigalzu.

Access to the upper levels of the ziggurats was via staircases. In southern Mesopotamia, the bases of these staircases appear on their foundations: a main staircase runs perpendicular to the building, and is flanked by two other staircases set against the monument and parallel to the wall. At Ur, where the first level of the building was excavated, they were joined by a square doorway (Woolley's "gate-tower"), which must have been the starting point of the access route to the top of the building. As the upper floors have not been preserved, it is impossible to know how the upper temple was accessed, since it seems that the main staircase did not reach the top of the building. Assyrian ziggurats, on the other hand, have left no traces of such staircases. This is due to the peculiarity of their architectural environment: they were generally attached to a temple integrated into the same complex and located on a common platform: access was probably via a staircase whose base was located inside the temple or on its roof, and which disappeared with the deterioration of the buildings. This is what analyses of excavations at Tell Rimah, Assur (although less obvious for the main ziggurat), Kar-Tukulti-Ninurta and Kalkhu suggest. The case of the Dur-Sharrukin ziggurat, for which Victor Place described an ascent along a spiral ramp, is problematic since it is thought to be an isolated case, and the reliability of this excavator's records is debatable. On the best-preserved ziggurat at Chogha Zanbil, ascent to the second floor was originally via four internal vaulted staircases running perpendicular to the building, located in the middle of each side of the second floor. External or internal staircases would then give access to the other floors.

The number of storeys in the ziggurats is uncertain in most cases, as the erosion of the ziggurat tops and the lack of representations for earlier periods prevent confirmation in most cases. If we follow Woolley's proposals, there would have been three of the first ones built at the time of Ur-Nammu, but this has been disputed by Schmid, who proposes simply two storeys. Chogha Zanbil, the best preserved, would have had four or five, not counting the high temple. It is clear from texts and images that the Babylonian ziggurat had seven, which seems to be the maximum and also a number with strong symbolic value, perhaps seen as the ideal number of floors for a ziggurat in recent times. A late school tablet showing an elevation plan of a seven-storey ziggurat with their dimensions is undoubtedly an abstract exercise representing an idealized ziggurat, and another tablet from Nippur does the same. In the north, Medio-Assyrian and Neo-Assyrian seals tend to depict four-storey ziggurats, but this type of simplified representation is difficult to exploit for architectural reconstruction. A Neo-Assyrian bas-relief found in Nineveh, preserved only by a copy, depicted a four-storey ziggurat, probably in an Elamite city, topped by a tall temple with horns.

It is therefore impossible to know how tall these buildings were, and only an estimate is possible. The ruins of the one at Dur-Kurigalzu still stand 57 metres high, and it could originally have risen to almost 70 metres, this edifice being one of the largest ziggurats. At Borsippa, the ruins were 47 metres high at the time of excavation, at Kalkhu they may have been around 25 metres above the plain and may have been around 40 metres high in ancient times, and at Chogha Zanbil they are still 25 metres high and may have originally reached over 50 metres. The one in Babylon is said to have been 90 metres high, according to the Esagil Tablet, a document describing the dimensions of the edifice. But the reliability of this text is called into question, given that it seems to give symbolic figures, as its aim is to explain the cosmological function of the edifice and not necessarily to describe it as it really is. All this explains why the height of the Babylonian ziggurat is disputed despite the available sources, and the latest estimates are lower (around 60 metres).

The high temple

It is generally believed that the top floor of ziggurats included a temple, most often called a gigunû. In the case of the Babylonian ziggurat, we find šahūru (a term that may in fact refer only to its terraced roof), or rarely bīt ziqrat ("temple (lit. house) of the ziggurat"). That of Anu's ziggurat in Uruk during the Seleucid era bears a temple name, É.ŠÁR.RA ("House of the Whole").

All these buildings have disappeared as a result of erosion of the buildings supporting them, making it impossible to be certain of the existence of such an edifice at the summit of each ziggurat, even if this solution is generally retained, especially since these buildings are derived from the ancient terraced temples, which are temples in their own right, with several rooms (including a cella), measuring 22.30 × 17.50 m for the "White Temple" of the Anu ziggurat at Uruk, and around 18 × 22.50 m for the Tell Uqair temple. In the case of the ziggurats, the question is also whether this was a temple in the strict sense of the term, or rather a smaller chapel. Interpretation often depends on the number of floors reconstructed: for Ur, L. Woolley's reconstruction places this construction on a supposed third floor, which would only leave room for a smaller building, while for H. Schmid (after E. Heinrich), the high temple would be built on the second floor (which measures around 36 meters long and 26 meters wide), leaving space for a multi-room temple.

The best way to find out about these buildings is to refer to the few texts that mention the one at Chogha Zanbil, and above all to the documentation relating to the one in Babylon in the 1st millennium, namely a recently rediscovered stele on which the building is depicted with its high temple, accompanied by a plan of it and above all texts, such as Herodotus' description, or the Esagil Tablet. According to the latter, the temple measured 25 × 24 meters, and would have reached a height of 15 meters. Access was via doors on each side, leading to six cellae (papāhu) arranged around a central covered courtyard. Built with cedar beams, its outer walls were covered with blue-glazed bricks. Herodotus says that there were no statues, only "a large, well-furnished bed, and near this bed a golden table". So, at least the rich furnishings of the gods were there, as in normal cellae. An image of the ziggurat, probably Elamite, carved on a bas-relief from Nineveh, shows that the upper temple was decorated with two horned animal heads on at least one of its sides, and the Babylonian Creation Epic perhaps indicates that the ziggurat in that city also featured horns.

Building materials and techniques

The ziggurats were built using the most popular building material of Mesopotamian civilization: clay bricks. Stone is only used where it is available, in Assyria, to build the bases and facings of these edifices, as was observed for the base level of the Kalkhu ziggurat. Clay bricks can be rectangular or square, laid on edge or flat, and in different types of bonding (masonry or panneresse). The central core of ziggurats was made of mud bricks, the vast majority of bricks used in their construction. It was generally framed by a casing of fired bricks, which were much stronger and less permeable to water. This formwork is around 1.50 metres wide in the ziggurats of the Ur III period, but reaches 15 metres in the ziggurat of Babylon in its final state. The last ziggurat to be built, that of Anu in Uruk in the Seleucid period, has a base built around a core dated to its Neo-Assyrian state, but surrounded by a brickwork separated by notable thicknesses of mortar, constituting a more crude construction than traditional ziggurats. The Chogha Zanbil ziggurat is unique in that it was not built according to a horizontal principle of stacked terraces, but according to a vertical principle, starting with a block forming the upper floor around which the lower levels are built, all levels starting from the ground.

The walls were generally decorated on the outside with pilasters and redans, and were slightly curved to compensate for the effects of perspective (entasis). In addition to this decoration, the walls of the Uruk ziggurat from the Ur III period appear to have been covered with a light-coloured plaster that was intended to give the building a lustrous appearance. At Tell Rimah, the west façade of the ziggurat was decorated with twisted half-columns, following a pattern also found in the temple.

The staircases and floors of the upper floors are also generally made of fired bricks. Glazed bricks may have been used from the 1st millennium B.C. onwards for some of the upper temples, as mentioned above, and possibly for the upper storeys. In any case, fragments of such bricks have been found on the ruins of some ziggurats, such as the one at Nippur. At Dur-Sharrukin, if we follow Victor Place's description of the ziggurat, the walls of the building's floors would each have a specific color, but this testimony has been disputed. However, other testimonies from ancient excavations at Borsippa and Ur also point in the direction of specific colors assigned to certain floors, which does not appear for other sites.

The mass constituted by the millions of bricks agglomerated in a ziggurat posed a number of physical problems: weights, thrusts, settling, lateral slippage, as well as problems of infiltration and water run-off. Mesopotamian builders therefore implemented various processes to ensure the durability of these edifices. Bitumen was used to waterproof the base of ziggurats. Rainwater running off the upper floors was evacuated by "gutter-drains" made of unbaked brick. Layers of reeds, spaced at regular intervals between the bricks, formed a chain that prevented the bricks from slipping. Some ziggurats (Uruk, Borsippa, Dur-Kurigalzu) also included an anchorage of braided reed ropes running the length of the ziggurat. Tree trunks could also be placed in the brick mass (as at Chogha Zanbil and Borsippa), serving as chain links or reinforcement, connecting the mud-brick core to the fired bricks of the outer layer. Small tunnels were also left in the ziggurat, no doubt to allow the brick mass to dry out, or to compensate for variations in brick size due to heat or humidity. The builders of the Tell Rimah ziggurat employed different techniques, no doubt for the same purpose: a 90 cm gap separates the heart of the building from the outer walls, and a central vaulted chamber, empty and inaccessible, was left inside the ziggurat. This same procedure is used in other Assyrian ziggurats, such as Kalkhu.

The ziggurat in urban space and landscape

Like the main monuments built by the ancient Mesopotamians, the ziggurat is located in a city. It is generally part of the city's central district, where its main political and religious buildings are located. More precisely, it is often located in a true "sacred quarter", which forms a whole, with worship areas, stores, kitchens, workshops, residences and administrative services. Ziggurats are located close to the main low temple, often in a large walled courtyard. In some cases, such as Ur, there is no independent low temple with which the ziggurat is associated; the main cult rooms (offering preparation areas and divine apartments) are probably located in the enclosure bordering the tower, or in the ziggurat temple (see below).

In the major cities of Lower Mesopotamia, ziggurats were located in a vast complex generally isolated from the rest of the urban space by an enclosure delimiting a sacred perimeter, to which only cult personnel usually had access. This seems to have been the case systematically from the 1st millennium onwards. For example, an entire architectural complex covering 350 × 300 meters in Uruk is organized around the ziggurat, with sanctuaries at its foot, several courtyards surrounded by wide walls in which rooms used as outbuildings for cult activities (kitchens, warehouses, workshops, chapels, etc.). In Babylon during the same period, the enclosure surrounding the ziggurat and its complex measured around 400 meters on each side, and was separated from the associated low temple, the Esagil, located to the south. Elamite ziggurats are also located in a sacred quarter, which would be called (in Elamite) kizzum at Susa and siyan-kuk at Chogha Zanbil, surrounded by an enclosure, the former possibly containing a sacred grove (husa). In Upper Mesopotamia, ziggurats are often attached directly to the low temple with which they are associated, and in all likelihood access to their upper storeys was via a staircase from the temple, which has not been found due to the deterioration of these buildings. Two Assyrian ziggurats, however, have been isolated from the other buildings: the Dur-Sharrukin ziggurat, for which the restoration of access to the upper storey poses a problem, as we have already seen, and the ziggurat of the god Ashur in the city of the same name.

It seems that several cities had more than one ziggurat: this is certainly the case at Ashur, where traces of three such buildings have been found (including a group of twin ziggurats attested only at this site), and at Nineveh, where two ziggurats are mentioned in the foundation inscriptions, in Uruk, with those of the Ishtar and Anu complexes, at least in the Neo-Assyrian and Seleucid periods, and perhaps in other Lower Mesopotamian cities, if the ziggurat lists are to be believed.

By their mass and elevation, and despite their isolation within enclosures, ziggurats were meant to dominate the city in which they were built. In Lower Mesopotamia, the flat terrain made them visible from miles away. In Upper Mesopotamia, where the terrain was more irregular, they were built on the kind of acropolis that formed the main quarter of large cities, combining palaces and temples. They therefore overlooked the rest of the buildings, all the more so when they were located near the edge of the hill, as at Kalkhu and Assur (for the main ziggurat). The ziggurats were therefore prominent features of the urban landscape of Mesopotamia's great capitals and sacred cities. Even today, the relatively well-preserved ruins of ziggurats dominate the sites where they are found.

A prestigious monumental construction

Because of their mass and spectacular appearance, and the means employed to build and preserve them, ziggurats are among the most important monuments built by the ancient Mesopotamians. Their construction was a task taken on by the rulers, who put their administration and workforce into action. As with palaces, great temples and city walls, ziggurat constructions are described in building inscriptions, which emphasize their monumental aspect and the symbolic importance that their construction held for kings and their personal prestige. The Sumerian term for these buildings, É.U6.NIR, can be translated as "House of Admiration". The names given to certain ziggurats emphasize the respect these buildings inspired or their spectacular appearance: the "House-foundation adorned with terror" (É.TEMEN.NÍ.GÙR) at Ur or the "House-mountain exalted" (É.KUR.MAH) at Kish. The mountain metaphor is, however, a common one for Mesopotamian temples in general. Certain ziggurats or parts of ziggurats were sometimes worshipped: The ziggurat temple at Nippur, for example, was one of the recipients of offerings in the time of Ur III, alongside its tutelary god and his throne, and certain Neo-Assyrian and Neo-Babylonian texts indicate that ziggurats, in addition to receiving offerings in the same way as other objects of worship such as the weapons of the gods, were deified (their name was then preceded by the deity's determinative).

An attempt was made to estimate the material and human resources required to build the ziggurats. Mr. Sauvage estimated the quantity of bricks needed to build the second floor of the Ur ziggurat at over 7 million (both unbaked and baked). According to him, the construction of this floor would have required some 95,000 working days for bricklaying, and 50,000 working days for other tasks, i.e. 95 and 50 days respectively if 1,000 workers were employed, a number attested in the case of temple construction in the same period. A text from the Neo-Babylonian period tells us that over 8,500 people were employed in the construction of the Sippar ziggurat, which is a considerable number. For the same period, J. Vicari estimates that the Babylonian ziggurat comprised 36 million bricks (but this depends on the size attributed to it), which he says could have been built by 1,200 men in 1,250 days, a theoretical calculation insofar as this edifice is in fact an extension of an earlier, smaller ziggurat and therefore did not require such labor; while M. Sauvage estimated that it would have taken 1,500 workers (including a good 1,000 masons) around 330 days to build it, without taking into account other materials (wood, reeds) and stewardship.

It's likely that the administrators in charge of these sites adjusted the number of personnel mobilized according to the expected construction time and their means. They didn't need specialized personnel to prepare the bricks. Workers may not have been available all year round, due to agricultural obligations, maintenance of other public buildings such as canals, etc., making it difficult to estimate the time needed to build or restore a ziggurat, not to mention any unforeseen circumstances. Another problem was finding specialized personnel, the master masons, who could have a wide range of skills and were therefore indispensable to the site. However, we know nothing about the architects who designed and supervised the construction of these buildings, as the sovereign systematically presented himself as their designer.

In the end, building a ziggurat doesn't represent a huge amount of work, and not necessarily much more than another monument, given that a large temple required around 20 million bricks (not counting its outbuildings). A royal palace or city wall, on the other hand, required far more resources.

A high temple

Ever since the first explorations and excavations of ziggurats in Mesopotamia, speculations have been made as to their function. The first analyses made by explorers and archaeological site excavators (Niebhur, de Sarzec) were utilitarian: they were elevated buildings where people could shelter from the heat and mosquitoes that abound in the wetlands of Lower Mesopotamia. For Victor Place, the Khorsabad ziggurat is an astronomers' observatory. This was not their main function, but it's still possible that they were used to observe the heavens, especially as Mesopotamian "astronomers" were priests. Subsequent interpretations focus on the religious sphere: it has been proposed that the ziggurat is a funerary construction (Hommel), or a symbol of the cosmos or the earth in miniature (Rawlinson, Jensen, Lagrange), or a divine throne (Lethaby, Dombart).

Indeed, there's no doubt that ziggurats are buildings with a religious function: they are located in a sacred space, are dedicated to a deity, and bear a ceremonial name in Sumerian like other Mesopotamian temples, beginning with the term É, meaning "house", as a temple is considered to be the residence of a deity.

If the interpretation of the ziggurat as a temple is accepted, it remains to be seen what the true meaning of an elevated temple is. It is generally considered that ziggurats, or rather the temples above them, are not "ordinary" temples, which are those on the ground. It soon became clear that, while architecturally these buildings were indeed continuations of archaic terraced temples, functionally they were not. German architect and archaeologist Walter Andrae was the first to put forward a theory on the subject. He sees the ziggurat as a building designed to support an elevated sanctuary (Hochtempel) linked to a nearby sanctuary at ground level (Tieftempel), since ziggurats are generally adjacent to a traditional temple. According to him, the "high temple" would be the divinity's habitual earthly residence, and he might occasionally descend to his low temple. But textual documentation does not indicate that ziggurat temples had such an important ritual role (see below).

In addition, a recent proposal, based in particular on excavations at the Borsippa ziggurat, is that the ziggurat should be interpreted as a temple in its entirety, and not just a support for the temple high atop it.

A cosmological interpretation: a link between Heaven and Earth

A. Parrot has taken up and extended Andrae's reflection on the symbolism of ziggurats as a link between the divine and human worlds. He sees these edifices as a manifestation of man's desire to rise (the etymology of the word used to designate them), no doubt to be closer to the divine world whose dwelling place is Heaven, as illustrated by the addition of more and more terraces to ziggurats over the centuries. Subsequent interpretations have focused on the cosmological aspect of the ziggurat as a link between Heaven and Earth. This is reflected in the names of some ziggurats: Larsa, "House of Heaven and Earth" (É.DUR.AN.KI), Babylon, "House of Heaven and Earth" (É.TEMEN.AN.KI), and Sippar, "House of Heaven" (É.KUN4.AN.KÙ.GA). An inscription by King Nabopolassar relating to the construction of the former proclaims: "Marduk, my lord, commanded me concerning Etemenanki, to ensure its foundation in the bosom of the Lower World and to make its summit rival the Heavens". The image of the sky to be reached can also be found in the inscriptions of its distant predecessor Samsu-iluna, as well as in Genesis, which relates that the builders of the Tower of Babel wanted its "top to touch the sky", and for it to be a "gateway to the heavens".

This interpretation requires an explanation of the Mesopotamian conception of the world. According to various texts, notably mythological ones, the ancient Mesopotamians believed that the World was made up of Heaven (AN), and the whole of Earth and the Lower World (KI), which had been separated at the beginning of Time by a creative deity or group of deities, whose identity varied according to local traditions. Heaven was home to the main deities of the Mesopotamian pantheon, the Anunnaki, while the Lower World is the equivalent of the Underworld. Between the two lies the earth's surface, where humans live. The ziggurat could therefore symbolize a kind of link between the two great parts making up the World, or even a passage from one to the other, as the name of the Sippar ziggurat indicates. In any case, the Babylonian Creation Epic (Enuma Eliš) makes the Babylonian ziggurat the center of the World, on the spot where the god Marduk created Heaven and Earth after defeating the primordial deity Tiamat. Principles derived from this cosmology may have governed the construction techniques used to erect the ziggurats, notably the layout of the foundations and staircases, but this point remains to be clarified in the light of less explicit textual and archaeological sources. This does not preclude combining this interpretation with older symbolic analyses, such as that which sees the ziggurat as a reproduction of a sacred mountain, an important symbol in Mesopotamian religion as a source of life and, above all, contact with the divine world.

The ziggurat's significance as a point of contact with Heaven seems particularly important. The name of Borsippa's ziggurat, "House of the Seven Sages of Heaven" (É.UR.IMIN.AN.KI.A), refers to its seven floors, which may themselves refer to the seven "wandering" astral bodies known at the time (Sun, Moon, Mercury, Venus, Mars, Saturn and Jupiter). One cosmological interpretation, moreover, seeks to relate the ziggurat's stages to these stars: the most complete form of the ideal ziggurat (at least in recent times) would be to have seven stages, like the seven astral bodies known to the Mesopotamians. Still on the subject of observing the sky, the Uruk ritual of the Seleucid period, which takes place at the top of the city's ziggurat, is linked to the appearance of several stars that can be seen there at night. The name of the high temple of this ziggurat, É.ŠÁR.RA ("House of Totality"), may be linked to the fact that the entire sky could be observed and approached from here. It also refers to the old interpretation of these buildings as observatories for the sky, which cannot be completely abandoned.

Ritual function

While the ziggurat has a religious function, and it's generally accepted that it symbolizes a kind of link between the human and divine worlds - although this is never explicitly stated in ancient texts - its ritual function is poorly documented and little studied, and the few hypotheses on the subject are highly conjectural.

The ziggurat has a ceremonial name of its own, and is distinguished from the lower temple where most of the divine worship takes place. There is very little evidence to suggest that the temples located at the top of ziggurats had a cultic role. Nevertheless, architecturally, there is a difference between ziggurats erected after the Ur III period, which function well around the low temple couple

More precise information comes primarily from the Marduk sanctuary in Babylon. We know from the Esagil Tablet that the high temple comprised six cellae, each housing statues of several deities: Marduk, Nabû and his consort Tashmetu, Ea, Nusku, Anu and Enlil. Opposite Marduk's cella was a chamber containing his bed and throne. Again, there is no mention of daily or routine rituals taking place in the ziggurat temple, although such a thing may have been assumed. Herodotus says that the divine chamber hosted a ritual union between the god and a local woman. Other Babylonian texts, too fragmentary to be fully understood, seem to refer to other rituals taking place in the upper temple: the lighting of a brazier during one of the rituals marking the kislimu festival; another rite apparently involving divine images, also taking place in several Babylonian temples; a cult calendar mentions a day dedicated to "Marduk and Zarpanitu of the Etemenanki", i.e. probably the manifestations of the divine couple in the temple atop the ziggurat; the mythological and ritual commentary nicknamed Marduk's Ordalia also seems to evoke rites taking place on the Babylonian ziggurat. It is clear from these meagre testimonies that the ziggurat temple in Babylon probably had only a minor ritual function.

Information on rituals that may have taken place in other ziggurats confirms this impression that these buildings played a secondary role in worship. Two Uruk texts from the Seleucid period describe two similar rituals taking place on the roof of the temple atop the ziggurat of the god Anu. One of these takes place at night and is apparently intended to ensure the continuity of light in a sacred fire, in connection with divine stars. During what seems to be the high point of the ritual, when the stars of the god Anu and his consort Antu appear, songs are sung, followed by sacrifices punctuated by the appearance of other stars, and the lighting of a torch carrying a sacred fire, which is then carried to other parts of the sanctuary. The ritual continues in the temple and the rest of the city until dawn. In these cases, it would be the height of the ziggurat that makes it the ideal place for part of this ritual linked to divinized stars, from which one can approach celestial plenitude, and light a torch that perhaps serves to capture the light of the night stars, and would symbolize the regeneration of the night, and

Elsewhere in Mesopotamia, texts from the 1st millennium BC (from Borsippa, Sippar and Assur) mention rituals and cult personnel in connection with the ziggurat or its temple, or the deified ziggurat. At the very least, this meagre documentation suggests that ziggurat temples had a ritual function in these periods.

Elamite documentation on ziggurats also provides clues, at least as to the existence of a ritual taking place in the sacred space surrounding that of Chogha Zanbil, at the foot of which a cult space has been uncovered, including offering tables and an ablutions basin. It can be linked to an Elamite work found in Susa and dated to the following century, a miniature representation of a "sunrise" ritual (sit šamši). Two priests perform a ritual between two buildings, which may well be a stepped altar and a ziggurat (or a second stepped altar), and near an ablutions basin. According to its name, this ritual would take place at dawn, whereas the Chogha Zanbil worship space is located on the side of the rising sun.

The architectural legacy of ziggurats after their demise appears to be nil. It is sometimes suggested that an Iraqi minaret such as that at Samarra takes its helical form from the ziggurat at Khorsabad, but it is more likely that it was the model of this minaret that influenced Victor Place's proposed reconstruction of the ziggurat, rather than the ziggurat itself inspiring medieval architects. Moreover, it's hard to imagine that the ancient edifice was sufficiently well preserved to serve as a model for the 9th century AD. The meagre architectural legacy of ancient ziggurats since their rediscovery by archaeologists can be seen in a number of modern, step-by-step buildings that were more or less inspired by them, notably in Germany and Austria.

The ziggurats' most notable posterity concerns just one of them, the Babylonian ziggurat, which in part inspired the authors of Genesis to write the myth of the Tower of Babel. This story explains how the inhabitants of Babel (a city inspired by Babylon) tried to reach Heaven, but were prevented from doing so by God, who confused them by multiplying their languages. The building also appears in ancient Greek literature (Herodotus), and has inspired numerous artists, primarily in Europe, right up to the present day.

The word "ziggurats" left its mark on a generation of players of the Warcraft 3 video game. Indeed, the phrase "We need more ziggurats!" is an expression often heard by players during the Undead campaign.

Sources

1. Ziggurat
2. Ziggurat
3. (en) A. L. Oppenheim et al., The Assyrian Dictionary Volume 21, Z, Chicago, 1961, p. 55-56.
4. a b c d et e Guichard 1998, p. 1393.
5. a et b Wicke 2017.
6. ^ Crüsemann, Nicola; Ess, Margarete van; Hilgert, Markus; Salje, Beate; Potts, Timothy (2019). Uruk: First City of the Ancient World. Getty Publications. p. 325. ISBN 978-1-60606-444-3.
7. Mierzejewski, A. (1983). Sztuka Starożytnego Wschodu I. Warszawa: Wydawnictwo Artystyczne i Filmowe. s. 30.
8. Mierzejewski, A. (1983). Sztuka Starożytnego Wschodu I. Warszawa: Wydawnictwo Artystyczne i Filmowe. s. 56.
9. Лурье И., Ляпунова К., Матье М., Пиотровский Б., Флиттнер Н. Очерки по истории техники Древнего Востока / под. ред. акад. Струве В.В. — М.—Л.: Издательство АН СССР, 1940. — С. 34. — 352 с. — (Научно-популярная серия под. общ. ред. акад. Вавилова С.И.).
10. Ру Ж. Великие цивилизации Междуречья. — С. 167.