Late Chalcolithic Mesopotamia: towards a definition of sacred space and its evolution
It is usually admitted that the development of monumental architecture is one of the major aspects of the urbanisation process in Late Chalcolithic Greater Mesopotamia. From the beginning, this development has been linked to the institutionalisation of religious practice at the dawn of history. This theocratic view of Mesopotamian history is well known, its roots lie in Deimel’s temple state theory. Formulated during the 1920s to explain 3rd millennium city State development, this theory has been applied to 4th millennium architecture during the 1930s; at Uruk as in Gawra or tell Brak, the monumental tripartite buildings excavated were interpreted as ‘temples’, and all the iconographic data interpreted along the ‘King Priest theory’. Deimel’s ghost is still operating in the processual approach of those developments, because neo evolutionist theories have largely viewed the development of complex societies as the result of the development of a religious elite of priests, integrating in huge religious compounds and festivals the people both of the city and its countryside. The precise nature of this integration remains disputed, but it is commonly accepted that priests built up a redistributive system to control and manage the incipient urban economies. This largely theocratic view has been challenged during the 1980s, following the debates about the evolution of tripartite architecture, for the most during the Ubaid period. Curiously, Ubaid monumental architecture has been the main focus of attention in those discussions and the following ‘Uruk’ period has attracted less attention, on that specific topic. Discussion has focused largely on the Uruk expansion, the local developments prior to this phenomenon and its chronology. Even if the classical vision of the priest kings of Uruk times has been challenged by philologists (Steinkeller 2001), the debate has remained centred around the idea that Uruk was first of all a religious centre, theatre of the sacred marriage represented on the Uruk vase. My point here is not to challenge the idea that Uruk or Susa were religious centres or that such festivals occurred at Uruk or elsewhere during the 4th millennium, but to establish on what criteria we can discuss the notion of sacred, in societies which were the cradle of huge mutations, from LC 1 to LC 5 period.
First of all it is useful to present an overview of those discussions and their conceptions about what religion should mean during the proto-urban period. Our second step will be focused on the kind of scenography which appears through Late Chalcolithic monumental architecture and the last one will be to contextualise those observations and define a kind of mental map of Late Chalcolithic religiosities.
Theocracy and neo-evolutionism
Writing something about the sacred during the late 4th millennium is definitely an impossible challenge. Charvat has delineated carefully some orientations about the spiritual world developed during ‘Uruk’ times and its revolutionary nature (Charvat 2002, 150–158). And even if much has been written on the topic, we have to recognise that we are dealing with some recurrent assumptions about what this religiosity should be and what kind of agency we should expect to reconstruct. First of all, we are dealing with very different cultural traditions in Greater Mesopotamia, since the Late Neolithic period (Fig. 6.1). Even if those regions belong to a shared cultural community from the Ubaid period on, that does not mean that there existed some common ‘religion’, or a common set of beliefs, even during the so called ‘Uruk expansion’. Late Chalcolithic religious history has been largely dominated by some key issues defined for the most part during the Uruk excavations, and we are still largely influenced by that way of thinking.
The first impediment is linked to the king priest theories. As we have noted in our introduction, Deimel’s hand is everywhere in this way of dealing with Late Chalcolithic societies, especially at Uruk itself. Following the excavations at Uruk, in Eanna and Anu Ziggurat sectors, archaeologists recognised ‘temples’, with the clear assumption that tripartite architecture was in itself proof of the existence of religious monumental architecture either on high or low terraces. Heinrich (1982), in his monumental synthesis, introduced a hint in this way of seeing things, distinguishing temples and ‘Kulthäuser’, that is the buildings in Eanna which were thought to be festival houses, or reception houses in the case of the buildings and the real temples (the white temple for instance). The Eanna buildings were interpreted as the predecessors of the later Eanna precinct, crowned by the massive ziggurat and dedicated to the goddess Inanna, whose engraved symbols were recorded on various objects (Blocher 2013). The sector named ‘Anu Ziggurat’, only because it was situated near the later Bît Resh monument dedicated to the god Anu, was thought to be the high temple of another deity, the differences between the two precincts being assigned two religious differences (Nissen 1988, 100–101). It was conceived as the other centre of a twin city, Kullab, being under the protection of an unknown god and Eanna of Inanna. Those centres would have merged together in prehisoric times (Nissen 2001). Uruk was seen as the matrix of future City-States, the supposed order prevailing in later Early Dynastic times being exported towards the late 4th millennium, that is 800 years before.
Fig. 6.1: Proto-urban centres in Greater Mesopotmia (P. Butterlin).
The results of the excavations at Tepe Gawra were also interpreted along the same line (Tobler 1950; Rothman 2002; 2009). The tripartite buildings discovered there were interpreted as another set of archaic temples, as would be, some years later, the whole of the Eridu sequence (Safar and Lloyd 1981) or the Uqair Buildings. It appeared clearly that tripartite architecture had a very long story and that the Uruk buildings were only the latest stages of a long evolution, beginning in Ubaidian times. That meant that the socio-religious order prevailing during the late Uruk period was already at stake at Eridu or Susa earlier on. Therefore, the sealings figuring masked figures were interpreted as representations of ‘shamans’, early religious leaders of proto-urban communities like Susa A. Those shamans would have been later replaced by king priests figured as master of the animals, and performers of rituals. The idea that those centres were directed by king priests at the head of the society has been enduring, especially since iconography has been interpreted on the same line. The various representations of the ‘king priests’ and, first of all, the famous Uruk vase, were conceived as the ‘mise en scène’ of ritual activities and the celebration of a theocratic order (Butterlin 2003, 74–77; Winter 2007).
Neo-evolutionism has not fundamentally modified this situation. In fact, the theocratic model was embedded in Service’s model from the beginning. Tracking the great divide between chiefdom and states was merely tracking two different steps in a theocratic evolutionary line, from religious chiefs to king priests. This way of dealing with the archaeological record has been challenged, especially during the 1980s, around the now classical discussion about tripartite Ubaid architecture: on what precise criteria could we define a ‘temple’? Aurenche, Forest and Margueron represent different critical approaches of the theocratic way of interpreting the data (Aurenche 1981; Forest 1996; 1999; Margueron 1992; 2009).
A good example of those discussions is the debate about the Gawra sequence: thoroughly excavated during the 1930s by American excavators (Speiser 1935; Tobler 1950), it has long been considered as the best example of a rank society at the time of the chiefdoms and complex chiefdoms of the 4th millennium BC (Forest 1983; 1996; Rothman 2002; Frangipane 1996). With its 19 levels, it provides an extraordinary sequence, spanning from the 6th to the mid-4th millennium BC (Fig. 6.2). Its material and the global structures of its different levels came under close scrutiny during the 1980s; at the beginning of our century they induced a new set of studies gathered in a collective book published in 2009 (Butterlin 2009). Without any doubt, Tepe Gawra is the mirror of evolutionism and its failure as an euristic device, especially when we speak about ‘chiefdom’ and ‘complex chiefdom’ and the role of religion in those processes.
Tepe Gawra seems to be an ideal case: the tell was completely excavated until level IX, and half of the surface of the prehistoric village was unearthed. It produced an impressive and apparently continuous superimposition of villages, offering the opportunity to follow the evolution of a prehistoric community in Upper Mesopotamia, from the Halaf period to the middle of the 4th millennium. An impressive set of tripartite building has been excavated and they were considered as ‘temples’ (Tobler 1950).
From that point on, scholars first focused on the interpretation of the inner organisation of this community, identifiying clear breaks in the sequence: after level XVI, level XIII, and thereafter, level XII and the following levels (Butterlin 2005). The sequence provides a complete evolutionary set, from an egalitarian society to the simple and complex chiefdom, especially on the various phases of level VIII. But looking more precisely at these approaches, one may see easily that there are considerable discrepancies between those interpretations: where Frangipane identifies a simple chiefdom, Forest sees a ‘communauté domestique agricole’ based on nuclear families (Frangipane 1996; Forest 1996). Most importantly, the role played by ‘religion’ in this matter is disputed. The crucial point is to know here if and when religious architecture appears at Gawra. For example Rothman distinguished in each level (during LC period) a new step on the evolutionary ladder, from the simple chiefdom to the theocratic complex chiefdom. He proposes to interpret the differences observed between the tripartite buildings as a functional one: all the tripartite buildings are not temples but only the in antis ones should be considered that way, from level XI on. Before, there are no traces of religious architecture and so level XI should be considered in northern Mesopotamia during LC 2 period as the great differentiation between religious and secular architecture. Forest, on the other hand, as Margueron denies the idea that there existed even one temple in Gawra. All the monumental architecture is secular and its evolution is interpreted as an inner evolution of the chiefdom or as the result of exterior influence notably from the south. Those buildings were houses of chiefs.
Behind this discussion lies a stratigraphic issue which was underestimated. Since the 1980s, many debates have dealt with the standard stratigraphic division of the Gawra sequence. I have produced a new section of the site, recording the actual altitudes of the summit of the walls and their foundations. This kind of study allows us to identify the numerous stratigraphic discrepancies and to re-evaluate the data. It is now well known that the whole sequence is an artificial reconstruction; Forest, Rothman and I tried to solve the main problems, wherever it was possible. One might suggest that the existence of shrines or ‘temples’ could have been an excellent focus and justify the recurrent occupation of the site or some spots of the site. But this is not the case.
Fig. 6.2: Tripartite architecture of Upper Mesopotamia. Left: Arslantepe VII, Grai Resh, Brak and Hamoukar; right: Tepe Gawra VIII, A–C, composition at same scale (P. Butterlin).
Against Rothman’s assumption about religious architecture, Margueron adopts a phenomenological approach towards religious architecture based on a fundamental assumption (Margueron 2009). The sacred space is a stage, where the encounter between the deity and the faithful or the devotee occurs. It is materialised by the place of epiphany and the place of offering. It explains the stability of this kind of place. At Gawra, those ‘temples’ constituted a matter of discussion and Margueron successfully demonstrated that none of those temples was built upon a former one (Margueron 2009). If one puts on a plan all these temples, it is astonishing to see that they ‘moved’ around the whole settlement all along its history. It is quite the opposite at Eridu for instance, from Ubaid times on and in the later temples in the City States. This lack of continuity at Tepe Gawra, even in the monumental space, gives the impression of a constant redistribution of power in those ‘chiefdoms’, whether this power is religious or not in its roots.
Those roots have particularly been discussed by Frangipane who has proposed to make the distinction between the sources of power in northern and southern Mesopotamia or Susiana (Frangipane 1996; 2007). The precocious development of high terraces and monumental architecture in Ubaidian Mesopotamia and Susiana has been linked to the way some households among the extended households of southern Mesopotamia managed to stand as a benchmark for the whole community (Pollock 1999). This process had already occurred during the Ubaid 0–2 phases and it would explain how those families concentrated through kinship the material resources in a staple finance system. This system would have been successful and highly dynamic in Southern Mesopotamia, whereas it resulted in a monumental failure in Susiana (Hole 2010). Ritualised exchange would have been the key for the development of high ranking families having a special kind of prestige, as proposed by Adams and later by Stein. This kind of organisation would have been fundamentally different from the horizontal egalitarian societies in the north, even after the Ubaid expansion. There, a different kind of elite, whose power was not founded on staple finance but wealth finance was developing.
This would surely explain the huge differences one observes between the different parts of Greater Mesopotamia later on during the Late Chalcolithic period. Recent researches have focused on defining different paths towards urbanisations from the Ubaid 3–4 on, to the end of the 4th millennium, following the definition of a new chronological framework, at Santa Fe in 1998 (Rothman 2002). Religion, and we mean here institutionalised forms of religiosity, seems to be deeply embedded in the development of Southern Mesopotamia and Susiana during the Ubaid times/Late Susiana. It is best materialised through the development of monumental high terraces. It is not the case in Northern Mesopotamia, where a different path towards urbanisation occurs, with possibly different religious traditions, as expressed for example by the widespread distribution of eye idols in this region during the LC 3 period (Stein 2012). It is only during the LC 3 or 4 period that some buildings could be interpreted as ‘temples’, at Arslantepe and Tell Brak (Eye temple). And those differences are particularly striking when we discuss those famous tripartite monumental buildings and their evolution.
Religious or political scenography in proto-urban societies?
Whatever the situation during the late Neolithic in Greater Mesopotamia, everybody agrees that from the late Ubaid times on, the development of tripartite monumental architecture is one of the major aspects of the urbanisation of this part of the world and a major step in religious history. It is agreed that it created a new relationship to the sacred, embedded in its institutionalisation through central agencies and their architectural expression. The main impediment is here quite simple: since a temple is fundamentally in Mesopotamian tradition the house of the god, and since tripartite architecture became during the Ubaid period the usual way of building houses, how is it possible to distinguish the house of the god from another one, especially the chief’s one? It is the classical dilemma and paramount in this process is the development of integrated monumental complexes, whether as complex compounds or high terraces, where tripartite units are only one part of the whole scenography, but still the object of the main focus. It is possible to make among all those buildings some differences, and I will argue here that those differences are some clues to understand the early differentiation between political and religious prestige buildings.
First of all, let us come back to tripartite buildings themselves and their evolution. Much attention has been devoted to Ubaid architecture but proto-urban tripartite architecture has not been studied so intensely. The publication of the whole data of Uruk (Eichmann 1989; 2007, and for an overview 2013), recent excavations in northern Syria, in Brak, Hamoukar or Tell Feres have produced new data and allowed us to study the whole documentation in a unified manner (Butterlin 2009; 2012; Stein 2012). First of all, tripartite architecture is basically the typical layout of a house, from Ubaidian times on. It is still the case during the LC 1 period, at Tepe Gawra or Degirmentepe for instance. The main difference is the disappearance at this stage of the cruciform layout of the central space which was typical of Hamrin and Gawra Ubaidian architecture. When one compares the surfaces, it appears clearly that the typical village house at that time has an area of about 90–100 m2, with an average area of 35 m2 for the central place. It is less for example than at the Tell Madhur house. The great houses at Degirmentepe or Gawra XII have an area of c. 175 m2 and a central place of 50 m2. Those houses are considered to be the houses of the chiefs of those northern mesopotamian villages. Two household levels are clearly present and there are no traces of any religious activity in those houses.
At that time, nothing compares to the scale achieved in Southern Mesopotamia where a typical monumental tripartite architecture appears during the Ubaid 3–4, at Eridu, Uruk and possibly Oueili. There, the tripartite houses have an average surface of 280 m2 and a central space of c. 95 m2. Those buildings situated on terraces belong to a completely different tradition. During the LC 2 and 3, at a time where no information is available in Southern Mesopotamia, we observe the following situation in the north (Fig. 6.2): first of all in the tigridian region, tripartite architecture ceases to be the ‘normal’ way of living and is now reserved. At Gawra, one finds only two or more monumental houses with a great stability of the surface of the central space (around 27 m2), and a progressive monumentalisation of the houses, around 190 m2, for the greatest ones. With the development of the typical in antis layout and a surface around 190–200 m2, we can define a monumental module in the Tigris region. Are those buildings temples? This is still debated but we are dealing with a typical scale, a northern monumental module. In the Khabur basin, tripartite domestic architecture is still widely used as in the Sinjar. Interestingly, the LC 3 houses are smaller than during the LC 1 phase, c. 70 m2 in surface with a central space of 20 m2. It is precisely at this time that a different type of tripartite architecture appears for the first time, at Arslan tepe (Frangipane 2004; 2007) and Tell Brak (Eye temple): those buildings are outliers in the whole sequence in the north with 400 m2 and 575 m2 of general surface and central spaces of 123 m2 and 108 m2. If we compare the Eye temple to what we know of tripartite architecture in the north, it appears that the distance between them is particularly marked. The existence of storage rooms in the northern half of this building is quite specific, and unusual.
This surface of 400–500 m2 is astonishing in the north. Those buildings were usually considered as temples, with a layout inspired from the south. Indeed, this surface is precisely what I proposed to name the Uruk standard monumental buildings (Butterlin 2012): at Uruk, both at Eanna and at the Anu ziggurat, this scale is the most attested one, with an average surface of 400 m2 for those buildings (Fig. 6.3). There are two other steps, what I call second rank and giant buildings. Nothing indicates in all those buildings that we are dealing with temples. At Uruk, it is usually accepted, as in Arslantepe or Brak, that those buildings were used ceremonially, for assemblies and ritual consumption. It is the same case at Gawra, but the general layout is not the same. This theme is central in the discussion around commensality and rituals of feastings. The main question is of course to know for whom those reception places were conceived, and what kind of general layout was present.
The main difference that can be observed first of all concerns the circulation system: at Gawra from level XI on, the in antis layout marks a complete shift from the previous traditions, where lateral access prevailed. This means that we are dealing with a very specific type of reception, with a direct an unique access to the central space. At Uruk, apart from the Anu ziggurat sequence, the prevailing system is the multiple access one, with a growing number of lateral doors, as the building becomes bigger. Usually, the central space presents two openings on each small side which are to be interpreted as windows, not doors.
A second point is the general layout of the central space. Since the late Neolithic, the main hall was the special focal point of those buildings, a space of reception and prestige, usually with a fireplace. This is still the case with various specific layouts. At Uruk, the presence of a second hall of reception, the so-called Kopfbau and a very distinctive supposed fireplace, usually compared to a keyhole, characterise the tripartite buildings. Those keyholes are also attested in the pillared ‘halls’, which seem to be simplified versions of tripartite buildings, reducing more and more the wings. At Uruk, those reception halls can be divided into three classes: a standard one, typically 5–6 m and a large one, 16–18 m long. These reception rooms are di-symetrically organised: usually with a pair of doors around the fireplaces and a double pair on the other side of the room. The organisation is completely different at the white temple, with a symmetric layout; we will come back later to that point.
A second rank is characterised by a much wider central space 8–11 m large, and with various lengths, up to 37.50 m or 62 m for the Kalksteingebäude. Here the same di-symmetrical organisation of space prevails as is the case for the temple C. Three fireplaces are present in the central place, two in the western half of the room and one rectangular in shape, in the eastern part. This central space is clearly divided in two equal parts, the eastern one with a huge fireplace and access to the staircases and a western one, with multiples doors and two fireplaces. This means a kind of specialisation and, in some way, a polarisation of space. I would argue that the eastern part was the place of honour, and the western one the place where the visitors were welcomed, and could get out easily. This is, finally, the main purpose of all those doors, to give instant and quick access to specific parts of the building. How many people gathered there remains difficult to assess, the important point is that the existence of many reception places means that different assemblies occurred at the same time in different spaces.
This last remark drives us to our second point. All this organisation of space is highly differentiated and organised along a central idea, a hierarchised and differentiated system of reception, which is the typical way of creating some kind of distance. This effect was achieved through different scenographies which have been identified long ago: buildings situated on top of high terraces belong to completely different kind of complexes as those integrated in complex combinations of structural units, as is the case for most of the tripartite buildings excavated at Eanna. Are those differences linked to different kind of rituals as proposes Nissen? It is possible but the conception of space is utterly different.
At Eanna, it is possible to identify in the later stages of the sequence, now GS 16, four different compounds. Three of them comprise one building of second rank, and one or two standard buildings, organised around courts and in one case associated with a pillared hall. The interesting point here is the variation around a common structural scheme: elementary units are combined in different ways. As far as we know, those compounds are enclosed and welcomed gatherings. We ignore what kind of assemblies occurred there but the very fact that those complexes coexisted means that more than one assembly occurred at the same time, and that those assemblies were divided along several compounds, but concentrated in one place. All this evocates some kind of confederal system associating different social units. It is most probable that huge ceremonies or feasts occured there, with redistribution operations. It is well known that the sealings and the few archaic tablets recovered in those complexes are linked to these operations. Those complexes can therefore be considered as a very specific step in the history of commensality (Pollock 2012): ultimate expression of the ‘tripartite tradition’ of receptions, they were the laboratory of a new society.
Fig. 6.3: Uruk monumental architecture, Uruk Eanna and Anu ziggurat (Courtesy R. Eichmann 2007) and colonial tripartite architecture, composition at same scale (P. Butterlin).
The next level in Eanna, level 15 is all the more interesting: the old complexes are levelled, apart from the complex of building C which is enlarged, to the west. A new court and three building are added to a system which appears as a reception complex. Two of those new buildings are halls: the Hallenbau and the Pfeilerhalle, the latter being strategically situated between the two courts. It appears as a monumental gateway decorated with stone cone panels. The adornment of the 12 polygonal pillars of this building has been interpreted as a zodiac, giving to this building an almost cosmological meaning. It could well be the case but a more fruitful approach is to understand the symbolic value of the colours used.
White, red and black were the main colours used in cone mosaics. Charvat has linked those colours to both symbolic and social components of the Uruk society (Charvat 2002, 145). He has argued that white was the colour of the gods, mixing purity, fertility and divinity. The white coating of the walls and the floors of the white temple are in this case of special significance. Red and black stand for two of the estates he identifies in Uruk society: red for the power, LUGAL, KINGAL, and black for the commoners (LU, GURUSH). Let us add in the case of the Pfeilerhalle, that the western façade of the building, oriented towards the external part of the compound was adorned in black and white as the eastern part, looking towards the inner part, with temple C, was adorned in red and white mosaics. If red was indeed the symbol of archaic power it could make some sense, in a complex which appears as a protopalace. The great court situated to the south belongs also to this new complex and we could argue that the last stage of the Eanna complex saw the development of a huge protopalatial ceremonial complex, built upon the ruins of a different system.
Could we be speaking of some kind of temple/palace complex, where redistributive and ritualised operations occur? It is the idea put forward by Frangipane about the Arslantepe VI A complex and it points out how difficult it is to understand those complexes. At Arslantepe, tripartite architecture occurs at level VII, with the already mentioned temple C, centrally situated upon a low terrace on the hill. The discovery of a lot of mass produced bowls on the soil of this building leaves no doubt about its redistributive functions, along with seals. Its religious function remains unclear and we have already noticed that its layout and its scale fit more with the southern mesopotamian tradition than a northern one. It could therefore have been a ‘temple’. The important point is the abandonment of the tripartite layout of monumental buildings at the end of the 4th millennium and the progressive development of an integrated complex, layer VI A, where bipartite monumental buildings become the main architectural structural units. This extraordinary complex articulated around a street has provided a fabulous insight on the complex redistributive system at work at Arslantepe, with minor influence coming from Southern Mesopotamia. The ritualised way of the redistribution processes occurring at Arslantepe must have been embedded within a set of religious beliefs and practices which belong to a different tradition from the one in the south. The paintings discovered on the walls of the main corridor are clear testimony of a completely different symbolic universe. Temple C of the preceding period was replaced by a bipartite building and it is difficult to consider that every bipartite building was conceived as the house of a god. Those buildings situated on the southern slopes of the hill near a gate were clearly some interface where differentiated receptions/distributions occurred, as appears in the building B and nearby storehouse.
To sum up, this study of tripartite and monumental architecture seems to confirm the idea expressed by Steinkeller that there existed a very early division between religious and secular powers in those states (Steinkeller 1999). This difference appears especially through different kinds of scenographic layouts and integration devices of elementary units which all derive from the tripartite archetype, adapted to specific cultural and political situations.
Time and space: a new religious mind?
It is a common idea to recognise a profound ideological mutation accompanying the urban revolution. The presentation of the religious dimensions of those mutations is usually centred around the idea that the religious roots of the traditional agrarian city-State of Mesopotamia were founded at Uruk, through the institutionalisation of the relationship between the king priest and the godess. This revolution is best expressed through the famous Uruk vase, materialisation of a new vision of the world and the social order. The monumental centre of Uruk has been interpreted along this theocratic view: the Eanna monuments were seen as the stage of the sacred marriage festivals and the ‘Anu complex’ could have been linked also to those rituals. It remains difficult to link the various names given to temples in the archaic texts to actual buildings, all the more because those texts have not been found in situ, and are attributed to the later phases of the Uruk period, that is Uruk IV a or III, that is LC 5 and later (Szarzynska 1992). She records 25 names of temples in the archaic texts of Uruk, the two main terms designating them being èß and é. The various reorganisation of the Eanna precint and the Anu complex are attributed to changes in religious conceptions combined to political changes. Those changes do not appear in the textual data, either because they occurred before the development of the earliest scripture (especially at the Anu ziggurat) or because the actual data is not clear enough to understand those changes.
Once we abandon the idea that tripartite monumental architecture is necessarily devoted to religious activities, how is it possible to establish a distinction between religious or secular buildings? Is this distinction useful at all in this context? Forest argued that the whole series was secular and that the only real temples were at Uruk the Riemchengebäude and the Steingebäude (Forest 1999). The tripartite buildings were merely communal buildings. This view is certainly excessive but it points out the weakness of the theocratic views. My proposition here is that it is possible to make a distinction between different types of monumental settings, those devoted more specifically to political purposes being the integrated complexes articulating various tripartite buildings, halls and courts. On the contrary, isolated tripartite buildings on top of terraces or situated in specific compounds constitute a different set of complexes. This is specifically the case for the Anu complex, the Uqair temple and possibly the Jebel Aruda ‘temples’. At this point, we can propose two lines of arguments. The first is the particular organisation of this kind of tripartite building, the second is the evolution of those buildings or the complexes in which they were integrated. Here the question will be centred on the very specific space and time in which those complexes were integrated.
As we have seen, most of the tripartite monumental buildings excavated were organised around fireplaces both present in monumental and domestic architecture, for instance in the Euphrates colonies houses. Some of those buildings present more complex installations, and it is of the utmost importance to understand if those installations are relied to cultic or ritual activities. Two buildings both isolated either on top of a terrace or in a precinct are specific: the white temple and its predecessors, and the Steinstiftgebäude, the stone cone building. As we have seen, the tripartite monuments built upon the Anu ziggurat belong clearly in scale to the monumental standards of the Uruk period. What distinguished them is their specific layout and some peculiar installations. There was no kopfbau at the white temple, two doors only gave access to the building, one from the north and the other by the south. The whole plan of the building as building D and E before presents some singularities: the walls of the eastern half of the building are wider than those of the western part and this enlargement seems to have no structural meaning. The central space is not rectangular but slightly trapezoidal in shape and wider in the east than in the west. Two features are present in this central space: one stepped platform in the centre of the space and a podium in the northwestern part of the central place, accessed by a small staircase. One additional feature was a complex systems of gutters whose function remains uncertain. They were interpreted as ritual by Heinrich but it is still difficult to understand what was their purpose. The association of a stepped platform and a platform standing against the wall is the main argument to identify the building as a temple but the specific rituals performed there remain unclear.
It is the same case when dealing with the Steinstiftgebäude. Without analysing this most particular building, now fully published (Eichmann 2007, 364–378), two features are of particular interest: the so-called basin (Room 4) and the complex installations present in the central space (Room 9), a fireplace, again a gutter and post-holes, seven around the fireplace and ten in the southwestern half of the central space of the building. Those latter are organised in three rows and their function remains unclear. The basin, defined by the coated bitumen soil, occupied the whole space 4: it was at least 1 m deep and its purpose remains unclear. The specific function of this building is not as clear as for the white temple, and it remains singular in the whole sequence. Its specific way of building, interpreted as a kind of purification ritual, the construction after the destruction of the building of the Riemchengebäude, whose rich content was interpreted as the product of a ritual deposit, the objects stemming from the stone cone building, have contributed to the idea that it was a very specific ritual building linked with water. This particularity lead Lenzen to the idea that it was devoted to the god Enki but this still remains a guess (Eichmann 2007, 378).
It has long been recognised that there existed a completely different conception of architectural space between the large integrated complexes excavated at Eanna and the Anu ziggurat for instance. The difference lies in the way the buildings are combined and set up, it lies also in a different conception of the regeneration of those complexes. It is obvious at Eanna for instance that the whole sector was regularly an object of huge readjustements, the former buildings usually being levelled to the soil. It is impossible to know how long some of those buildings lived but it is important to notice that apart for Building C, the other tripartite buildings were short lived, whole complexes being levelled and replaced by different ones. This instability is one good criteria to consider those buildings and complexes as secular and it stands in opposition to the dynamics operating in the high terraces complexes. The distinction high low-terrace is not so easy to establish from the beginning (Lenzen 1941). The distinction is obvious at Uruk, or Uqair; it is not so clear during the Ubaid period. The future high terraces at Uruk or Eridu began as low terraces.
When we compare the proto-urban development of those high terraces some interesting further points appear. For instance, at Uruk and Susa, the high terraces were not the first terraces or massifs built on the site. At Susa, the first monumental building is the funerary massive and at Uruk Anu, an Ubaidian terrace upon which stood tripartite buildings was erecetd before the construction of the ‘Anu ziggurat’ (Fig. 6.4). The link between those Ubaidian terraces, which were progressively enlarged and the Anu ziggurat is not clear. Eichmann has suggested that their later levels existed at the same time as level Z 20 and 19 of the ziggurat. The main point is that this ziggurat was built later and that the two terracces coexisted for a while during the Ubaid 4 period. In fact, the same process occurs at Susa: foundation of a first monument, coexistence of both monuments and, later on, abandonment of the first monument and development during the Uruk period of the high terrace. It is not possible here to argue further on that point, let us just say that the comparison could go further since the two sequence seem to be contemporary, from late Ubaid to Uruk times. Between those stages, signs of abandonment and destruction are obvious both in Uruk and Susa and those phases seem to indicate a discontinuous history on critical spots of those proto-urban centres (Fig. 6.5).
At Uruk, the older terrace was covered by a new monument the so-called Steingebäude whose history and function are still debated (Eichmann 2007). Considered by its excavators as a funerary monument used during the sacred marriage as the repository for the dead king, this monumental subterranean structure has been considered by Forest with the Riemchengebäude as the only real temple in Uruk. This is certainly misleading since there are no traces of ritual installations in either building, but it is clear that the Steingebäude was linked in some way to the high terrace. This building was not short lived as previously thought by the excavators, and even if it is difficult to know how long it lived, it seems that while the terrace was slowly growing, the stone structure remained untouched, until its massive and complex filling, and sealing by the later ziggurats. The building is centrally built upon: a central space is bordered by two concentric corridors, delimitated by massive walls made of limestone and moulded concrete blocks. It seems difficult to admit that they were only supporting a light structure as usually proposed. If the central space of this monument was a place of display, as suggested by the central rectangular platform set in it, it could have been a place of exposition, with a subterranean level and a first level, made of bricks, where people could look from the first level onto the subterranean place of exposition. It is of course tempting to see there two steps of the sacred marriage, but that remains only a guess. It has been interpreted as a ‘gueule d’enfer’ by Szarzynska (1981) and Charvat has proposed to see there a primitive ‘Giparu’, the reed mat set on the central platform being the nuptial bed of the pontifical couple EN and NIN (Charvat 2002, 101). The five post-holes set in the platform which represent, to his mind, the centre of the world and its four cardinal points, would give to this space an ‘archetypal’ function. It would have a mythical place linked to the beginnings of the world and a critical spot in Uruk’s religious topography. The relationship between this huge monument and the tripartite buildings which were regularly rebuilt upon the high terraces is one of the keys to understanding the rituals performed there. But, interestingly, the Steingebäude is not the oldest building on this spot. The developement of high terraces was therefore not the first step of the proto-urban development but seems to be merely the sign of a promotion, political and religious.
Fig. 6.4: Uruk, Anu Ziggurat, proto urban levels (courtesy R. EIchmann 2007), composition at same scale (P. Butterlin).
It has long been said that both Uruk and Susa, two of the first great metropolis of the ancient Near East, were the result of a kind of synoikismos. Susa would have been a kind of confederal sanctuary (Hole 2007), developing at the expense of Chogha Mish which was the previous centre in eastern Susiana. Uruk would have been the result of the merging of Eanna and Kullab. This last point is not so clear, because if we follow our line of argument, we have to conclude that the Inanna temple, the Eanna, was actually situated from the beginning on the so-called ‘Anu ziggurat’. It remains difficult to know what kind of buildings were present at Eanna in the earlier stage of the Uruk period and the synchronisation of the Anu ziggurat and Eanna sequences remains a matter of debate. This enduring debate is centred upon the interpretation of the so-called Datierungschnitt in Uruk and its relationship with the eastern corner of Eanna. But there are good reasons to believe that the white temple is older then the Eanna precinct buildings (level IV) and could be contemporary with the older buildings, especially the Steinstiftgebäude. It seems that during the late Uruk period, the whole of the monumental centre at Uruk was integrated and connected through a complex system of terraces; a huge terrace called Alte Terrasse linking the Eanna with the later levels of Anu ziggurat (A 3–1). At this time, it seems that the protopalatial complex situated at Eanna was connected to the high terrace, through the old terrace. After the Uruk collapse, and the destruction of the building C complex, a new high terrace was built upon the old protopalatial complex, and it could be said that Eanna became Eanna at last. This huge shift after more than a millennium of careful reconstructions of the high terrace is, in itself, a complete change and must have been a revolution, both political and religious. The protopalatial complex has been levelled and it disappears with the king priest images. It is difficult to link this set of data to the textual data as mentioned above. But it is intresting to notice that there seems to be a shift in the archaic texts from Uruk IV to III: during the later stage of Uruk IV the main term applied to temples is simply èß, the term é being seldom used. During the Uruk III phase, that is after levelling of the level IV buildings and construction of the Eanna high terrace, the term é became more frequent along the use of more precise designations associating èß with deities notably Inanna (Szarzynska 1992, table 2, 287) and this shift is perhaps linked to those reorganisations.
Fig. 6.5: Uruk and Susa: chronocultural table (P. Butterlin).
Those briefs remarks give us a good insight on the highly dynamic societies which developed in Mesopotamia and Susiane from the 5th millennium to the 4th millennium. Instead of the regular and very continuous development usually described, we identify huge crisis, especially materialised upon the high terraces which were very sensitive points of representation, by levels of destructions, and huge shifts. Those shifts are testimony of a rhythm of evolution and cycles of regeneration. Besides the usual imagery of abundance, and the ideology of the cyclical regeneration of nature by the sacred marriages, lies another picture, punctated by major crises and reconstructions, whose religious aspects remain to be understood without any prejudice.
Bibliography
Adams, R. Mac. (1981) Heartland of Cities: Surveys of Ancient Settlements and Land Use on the Central Floodplain of the Euphrates. Chicago, Chicago University Press.
Blocher, F. (2013) Die Iconographie der Inanna/Ischtar im alten Orient. In N. Crüsemann, M. van Ess, M. Hilgert and B. Salje (eds) Uruk-5000 Jahre Megacity, 83–91. Begleitband zur Ausstellung Uruk 5000 Jahre Megacity Pergamon Puseum-Staatliche Museen zu Berlin, Publikation des Reiss-Engelhorn-Museen, Band 58. Petersberg, Imhof Verlag.
Butterlin, P. (2003) Les temps proto-urbains de Mésopotamie, contact et acculturation à l’époque dite d’Uruk en Mésopotamie, Paris, CNRS éditions.
Butterlin, P. (2009) Tepe Gawra et le monde proto-urbain de Mésopotamie. In P. Butterlin (ed.) A propos de Tepe Gawra, le monde proto-urbain de Mésopotamie, about Tepe Gawra, a proto-urban World in Mesopotamia, 1–15, Subartu XXIII.
Butterlin, P. (2012) Les caractéristiques de l’espace monumental dans le monde urukéen, de la métropole aux colonies. Origini XXIV, 171–191.
Butterlin, P. (2013) Les terrasses monumentales proto-urbaines et les centres proto-urbains de Suse à Uruk, études proto-urbaines 1. In J. L. Montero-Fenollos (ed.), Du village néolithique à la ville mésopotamienne, Bibliotheca Euphratica vol. I, 117–132.
Charvat, P. (2002) Mesopotamia before History, Routledge, London and New York.
Eichmann, R. (1989) Uruk, die Stratigraphie, Grabungen 1912–1977 in den Bereichen Eanna und Anu-Ziggurat. Augrabungen in Uruk-Warka Endberichte (AUWE 3), 2 vol. Mainz am Rhein, von Zabern.
Eichmann, R. (2007) Uruk Architektur I, Grabungen 1912–1977 in den Bereichen Eanna und Anu-Ziggurat. Augrabungen in Uruk-Warka Endberichte(AUWE 14), 2 vol. Rahden, von Zabern.
Eichmann, R. (2013) Frühe GroBarchitektur in der Stadt Uruk. In N. Crüsemann, M. van Ess, M. Hilgert and B. Salje (eds) Uruk-5000 Jahre Megacity, 117–127. Begleitband zur Ausstellung Uruk 5000 Jahre Megacity Pergamon Puseum-Staatliche Museen zu Berlin, Publikation des Reiss-Engelhorn-Museen, Band 58. Petersberg, Imhof Verlag.
Emberling, G. (2002) Political Control in an Early State: The Eye Temple and the Uruk Expansion in Northern Mesopotamia. In L. Al-Gailani Werr, J. Curtis, H. Martin, A. McMahon, J. Oates and J. Reade (eds) Of Pots and Plans: Papers on the Archaeology and History of Mesopotamia and Syria Presented to David Oates in Honor of His 75th Birthday, 82–90. London, Nabu Publications.
Forest, J. D. (1996) Mésopotamie, Naissance de l’Etat. Paris, Editions Paris Méditerranée.
Forest, J. D. (1999) Les premiers temples de Mésopotamie (4ème et 3ème millénaires). Oxford, British Archaeological Report S765.
Frangipane, M. (1996) La nascita dello Stato nel Vicino Oriente. Rome et Bari, Laterza & Figli.
Frangipane, M. (1997) A 4th Millennium Temple/Palace Complex at Arslan Tepe-Malatya. North South Relations and the Formation of Early State Societies in the Northern Regions of Greater Mesopotamia. Paléorient 23/1, 45–73.
Frangipane, M. (2004) Alle Origini del potere. Arslan Tepe, la collina dei Leoni. Rome, Electa.
Frangipane, M. (2012) Fourth Millennium Arslantepe: the Development of a Centralised Society Without Urbanisation. Origini XXIV, 19–40.
Heinrich, E. (1982) Die Tempel und Heiligtümer im alten Mesopotamien. Berlin, de Gruyter.
Hole, F. (2010) A Monumental Failure: the Collapse of Susa. In R. A.Carter and G. Philip (eds) Beyond the Ubaid, Transformation and Integration in the Late Prehistoric Societies of the Middle East. Studies in Ancient Oriental Civilization 63, 227–245.
Jacobsen, T. (1976) The Treasures of Darkness. New Haven, Yale University Press.
Lenzen, H. (1941) Die Entwicklung der Zikkurat von ihren Anfängen bis zur Zeit der III dynastie von Ur, AUWE 4, Leipzig.
Margueron, J.-C. (1992) Sanctuaires sémitiques. Supplément au dictionnaire de la Bible, Tome 11, 1104–1258. Paris.
Margueron, J.-C. (2009) Le temple et le sacré à la fin du cinquième et au début du quatrième millénaire à Tepe Gawra. In P. Butterlin (ed.) A propos de Tepe Gawra, le monde proto-urbain de Mésopotamie, about Tepe Gawra, a proto-urban World in Mesopotamia, Subartu XXIII, 39–77.
Nissen, H. J. (2003) Uruk: Key Site of the Period and Key Site of the Problem. In J. N. Postgate (ed.) Artefacts of Complexity, Tracking the Uruk in the Near East, 1–16. British School of Archaeology in Iraq.
Pollock, S. (1999) Ancient Mesopotamia. Cambridge, Cambridge University Press.
Pollock, S. (2012) Politics of Food in Early Mesopotamian Centralized Societies. Origini XXIV, 153–168.
Rothman, M. S. (2002) Tepe Gawra, the Evolution of a Small Prehistoric Center in Northern Iraq. Philadelphia, University of Pennsylvania.
Rothman, M. S. (2009) Religion, Function, and Social Networks: Tepe Gawra in the fifth and early Fourth Millennia B.C. In P. Butterlin (ed.) Tepe Gawra, Centre proto-urbain de Haute Mésopotamie et son monde. Subartu XXIII, 15–38.
Safar, F., Mustafa, A. and Lloyd, S. (1981) Eridu. Baghdad, Republic of Iraq, Ministry of Culture and Information. State Organization of Antiquities and Héritage.
Stein, G. (1994) Economy, Ritual, and Power in ‘Ubaid Mesopotamia’. In G. Stein and M. Rothman (eds), Chiefdoms and Early States in the Near East, 35–46. Madison, WI, Prehistory Press.
Steinkeller, P. (1999) On Rulers, Priests, and sacred Marriage: tracing the Evolution of Early Sumerian Kingship. In K. Watanabe (ed.) Priests and Officials in the Ancient Near East, 103–138. Heidelberg, Universitätsverlag C.Winter.
Szarzinska, K. (1981) Some remarks on the so called Steingebäude in archaic Uruk Warka. Akkadica 23, 45–49.
Szarzinska, K. (1992) Names of Temples in the Archaic Texts from Uruk. Acta Sumeroligica 14, 269–287.
Winter, I. (2007) Representing Abundance: the Visual Dimension of the Agrarian State. In E. Stone (ed.) Settlement and Society, Essays Dedicated to Robert McCormick Adams, 117–139. Los Angeles, University of California, Cotsen Institute of Archaeology.
Yoffee, N. (2007) Myths of the Archaic State, Evolution of the Earliest Cities, States and Civilizations. Cambridge, Cambridge University Press.