Subartu
Editorial Board
Sultan Muhesen, Joaquin Cordoba, Marc Lebeau, Lucio Milano, Walther Sallaberger, Philippe Talon
With the support of the following institutions: Universita Ca' Foscari di Venezia, Universitě Libre de Bruxelles, Universitad Autonoma de Madrid, Ludwig-Maximilians-Universität München
Subartu is edited by the European Centre for Upper Mesopotamian Studies
Manuscripts arc to be submitted to: Marc Lebeau, ECUMS-Brussels. 41 Boulevard A. Rcycrs, lite 6, B-1030 Brussels. Belgium
Order forms to be mailed to: Brcpols Publishers. Bcgiinhof 67. B-2300 Turnhout. Belgium
Pascal BUTTERLIN
(editor - éditeur)
A propos de Tepe Gawra, le monde proto-urbain de Mésopotamie
About Tepe Gawra: a proto-urban world in Mesopotamia
BHEPOLS
Subartii XXIII
Fig. 9: Tepe Gawra et son environnement. Mission archeologique franijaise de Mari.
Photo Andre Parrot 1953
Fig. 10: Tepe Gawra dans les champs du piemont. Mission archeologique franeaisc dc Mi
Photo Andre Parrot 1953
14
II. Religion, Function, and Social Networks: Tepe Gawra in the Late Fifth and Early Fourth Millennia BC
Mitchell S. Rothman, Widener University
Resume
Rothman dans cetarticle revient precisement sur le developpement de I'espace sacred Tepe Gawra. A I'aide d'analogies anthropologiques, il revient sur le developpement pragressif d'un centre religieux et presente line nouvelle analyse dit niveau XIII. II s'interroge sur les divers aspects de la religiositea Tepe Gawra, apartir de la grille d'analyse qu 'il aforgee etdeveloppe point par point ces aspects en insistant notamment sur I'iconogra-phie. II developpe une vision fondee sur I analyse de divers reseastx de relations tisses amour did site de Gawra et les sitae dans la discussion sur les contrastes de developpement entre le nord et le sud mesopotamien.
In the earlier history of archaeology, the site was considered a world unto itself. What happened at Uruk-Warka or Susa was emblematic of a larger, but unseen and little considered world. Such comparisons that were offered usually were related to diffusion or migration, the spread of ideas, artifacts, technology, and peoples. However, as culture history and anthropology played a larger role in our understanding of the modern and also ancient worlds, a realization emerged that, in fact, antiquity was not made up of isolated pockets of sites. It rather constituted inter-related networks of varying scale: local, regional, and even international.1 A network is defined as a system made up of many one-to-one interactions.2 Those sets of interactions that are the most frequent define the core network, but as interactions extend beyond the core, broader and larger networks are created. A polity is a network with frequent interactions at least on a political axis; a trading network usually extends well beyond the polity. Networks of information, military action, economic exchange, and religious or social rituals can be of very different sizes than the polity.
In this context, cities and towns - centers - were not simply larger agglomerations of people, but places within larger social, economic, and political networks where certain functions were found; for example, central marketing, specialized craft production, Great Tradition religion, and leadership groups.3 Many of the residents of large centers do work that people in towns or even villages do, but these additional functions were also found there. People were drawn to the center because of the special services it provided, and centers incorporated the countryside because of its needs, whether political, military, or economic. Kowalewski et alia assert that one of the prime functions of all centers is to mediate between their polity and other central places in larger networks, although how this role is manifested varies from place to place and time to time.4 At times centers permit subsidiary places to create their own networks outside the polity. At other times, leaders at centers try to close those contacts and make all outside connection flow through the center itself. One key, then, to understanding the evolution of ancient societies is determined by the nature of these networks and changes within them. It is a key that opens a door to incredible variability and analytical complexity.
In this article, I will raise questions about how we perceive the ancient world and analyze it. I will ask what roles religion, economic production and exchange, political leadership, and social structure played in the functioning of ancient societies, in the creation of networks of varying scale, and in the dynamics of change. I will offer one case, the piedmont and Jazira of northeastern Iraq, particularly Tepe Gawra of the late fifth and fourth millennia BC. variously called the Late Ubaid and Uruk, Gawran, or LC 1 -5 periods as an example.5
This time is such a critical period, because during it the first true cities,'' first political organizations recognizable as states,7 and the first more formal exchange networks—the Uruk Expansion—developed.8 At the same time, our emphasis on the rise of stales has perhaps distracted us from looking at the development
1 Nissen 2001.
2 Blanlon etui. 1997.
3 Blanton 1976.
* Kowalcvski elal. 19X3. 3 Rothman 2001a.
'' although Bultcrlin 2003 sees this period as proto-urbain. 7 Wright and Johnson 1975; Wright 1998. "Algazc 1993.2001.
15
Subartu XXIII
of other pre-state forms of social organization, which are equally important to understand, including that form social scientists call the chiefdom. In doing so, it is necessary to recognize that the term chiefdom encompasses great variety of societies in scale, focus, and degree of complexity.9
II. 1. Units of Analysis
Understanding the relations of city or town and countryside in its various network contexts first raises the significant question of what the units for our analysis will be. Algaze in accepting World Systems Theory assumes that the unit is the world-system itself, in this case, all the areas covered or even touched by the Uruk Expansion.10 As Hall and Chase-Dunn have written, "the fundamental unit of social change is the world-system and not the society".111 take a contrary view. As I have written elsewhere,12 the best analytical approach to understanding cultural change and the nature and role of any network must begin at the local level and work its way out to the largest networks in which people and institutions engage.
Regardless of how larger trends at a regional, international, or in these modern days, global scale affect people, they react to it in a local context. For example, the price of champagne may vary because of a global market, but the reaction to changing price must be dealt with in the vineyards and markets where the champagne is produced. The effect of price change will depend on many local factors: the value of the franc, how many people depend on the wine industry directly or indirectly, the size of its market, governmental subsidies of the industry, and the local culture.
In this same light, Nissen emphasizes that changes in local political and economic arrangements always occur in response to a particular set of needs determined by the local situation.13 He asserts, for example, that the appearance of Uruk cultural elements, especially administrative ones, in northern Mesopotamia represents the adoption by local populations of some innovation that is of use to them rather than an imposition by an outside force. So, writing is much delayed there as compared to Uruk-Warka. Frangipane writes that the best approach to understand north-south relations in Uruk period is to view the phenomena from the north.14
This approach emphasizes the characteristics and trajectories of development in the north, and only then reconstructs "the historical roots of their external relations".15 It is the nature of local development that determines the degree and kind of interaction outside of the local area, as I will discuss later on. Blanton el alia write, "The main shared assumption of economic analyses beyond the local system is that some features of any social formation may be the result of inter-actions across local-system boundaries. This is not to imply that extra-local interaction is necessarily always the major source of socio-cultural change"."' A second assumption they make is that the regional system is not just an expansion of a local system, even a dominant one. The various networks on a regional scale, like that represented by the Uruk Expansion, take on their own dynamics.
II.2. Religion and Complexity
In the early phases of the so-called New Archaeology, particularly in the United States, studying ancient religion was suspect. Too many early researchers had grasped at religious ritual or ceremonial acts to explain a world of artifacts and phenomena without first explaining how ritual and ceremony play a role in the larger evolutionary trajectory of societies. As Flannery writes, "archaeology has absolutely no coherent and consistent theoretical framework by means of which ritual or religious data can be analyzed and interpreted".17
At the same time, the importance of religious ideology as seen from modern cultures is obvious. All cultures subscribe to "ultimate sacred propositions", which through faith define the way their members see this and the sacred world.18 As Jung has written, these propositions arc metaphors or archetypes that define people's cognitive map of how the human world functions as much as it docs the way the divine world works.'9 Through the religious experience of ritual, people arc impelled into action. Through it their
9Earlc 1991. 1(1 Algaze 1993.
11 Hall-Chase-Dunn 1996, 11-12. l2Rothman 2001a, 2004.
13 Nissen 2001.
14 Frangipane 2001.
15 Frangipane 2001.
"' Blanton eral. 1997,5-6. 17 Flannery 1976, 330. '"Rappaport 1968,237-239. 19 Jung 1964.
16
A propos de Tepe Gawra
attitudes toward human goals, other people, their effect on the world, and their values are formed and reinforced along with other aspects of their lives.20 People strive to make their religious beliefs coherent with their perception of the secular world, at times creating new religious movements in response to economic and political conditions (for example, the origins of Islam).21
Religion, unfortunately for archaeologists, is much more difficult to understand than almost any other cultural system, because it functions differently in a variety of social groupings and for various purposes. Redfield points out that in most societies a kind of peasant religion often exists alongside the formal religious systems where political development of rank or stratification occurs (Small and Great Traditions).22 Often Small Tradition religion is full of magical ideas, healing potions, and shamanistic beliefs. Great Tradition religions are those of priests. The distinction here is very important for early Mesopotamian society, as we shall sec.
Shaman is a term that derives from the religious practitioners of East Asia. It applies, however, to a whole class of religious practitioners that existed in pre-Christian Europe, the Middle East, Africa, Native America (north and south), and parts of Asia. Shaman are part-time specialists, whose power derives from an ability to communicate directly with divine spirits and bring the gods' mystical powers to bear on human problems: disease, warfare, or social strife.23 They often use animal familiars, call on animal spirits, or even take animal form. As practitioners called on only in times of need, they do not work out of formal religious buildings.
Priests, on the other hand, are trained formally in schools to communicate with God or the gods. They are masters of liturgy and ritual, often full of secret knowledge, some of which may have shamanistic overtones. They tend to work out of formal institutions and are part of the established political structure, including leaders and factions. It is for their ritual that people build churches, mosques, synagogues, and temples.
By their nature, priestly religions are associated with complex society, because political and economic leaders often sanctify themselves in the minds of the people using religion. This means both that leaders benefit from religious ideology to re-affirm their political ends, and also that the religious archetypes of a given time confirm the current social order.24 In sanctification, by associating themselves with the gods, leaders shield themselves from attack and enhance their authority.25 The wall paintings from the Mari palace associate the throne room of the king with the life-giving properties of water and the gods all in one frame. Moorey has proposed that the death pits from among the Royal Tombs at Ur represent another part of this process.26 He argues that those who enacted the annual mating ritual of the god Dumuzi and goddess Inanna as guarantors of the fertility of the land were given god-like status in death. Records of the time indicate that the players in this ritual were often members of the king's immediate family. Jacobsen27 sees the advent of kingship as bringing a new relationship with the gods and with new characteristics of both the king and the gods, that of "majesty and energy". This does not mean that kings were not believers themselves. The Hammurabi stele, apparently placed in the squares of towns in his realm, is really messages to the gods. When one remembers that the normal residents, judges, and even kings could not read them, clearly they are not law codes. Rather, the preamble speaks of divine justification for Hammurabi being king because of his efforts to establish justice, protect the weak, build temples, and expand the gods' realms. They are letters to the gods.
The transformation in cultures from a shamanistic or folk religion into a Great Tradition religion also is reflected in the development of towns and cities. "The city may be imagined as that community in which orthogenetic and hetcrogenetic transformations of the folk society have most fully occurred. The former has brought about the Great Tradition and its special intellectual class, administrative officers and rules closely derived from the moral and religious life of the local culture, and advanced economic institutions, also obedient to these local cultural controls. Insofar as the city has an orthogenetic role, it is not to maintain culture as it was; the orthogenetic city is not static: it is the place where religious, philosophical and literary specialists reflect, synthesize and create out of the traditional material new arrangements and developments that are felt by the people to be outgrowths of the old. Insofar as the city has a heterogenetic role, it is a place of conflict of differing traditions, a center of heresy, heterodoxy, and dissent, of interruption and destruction of ancient tradition".28 A case in point for the distinction of Great and Small Traditions may be the temples of some southern Mesopotamian cities, where temples existed on the high and low mounds
2(1 Geertz 1973a, b.
21 Wolf 1951.
22 Redfield 1953.
23 Lessa-Vogt 1979, 301-2.
24 Winter 1987, 1992.
25 Drennan 1976.
26 Moorey 1977.
27 Jacobsen 1976.79-80.
28 Reel field and Singer 1954. 58-59.
17
Subartu XXIII
contemporaneously.29 Alternatively, the two temples may reflect social stratification and possibly a hetero-genetic aspect of early kingship.
Religion also codes a less often mentioned social element: its integrative function. The Kede of modern Nigeria provide an example.30 The Kede were organized into a segmentary state that exists within a multiethnic region. Although there is a Kede homeland, traders and later colonists often lived in areas where they were at least temporary minorities. The way they maintained their societal integration and also brought other ethnic groups into their sphere of control was through religious ritual. One such ritual has to do with the river, the source of transport and water for agriculture. This ritual is supposed to be for the whole river community, Kede and others. "The priest climbs a rock in midstream and throws a stone toward the bank: the spot where the stone falls is believed to mark the line to which the river will rise thai year".31 The high chief is not permitted to attend this ritual, although he has adopted it as his own. The chief often provides the goods needed for rituals even ones he cannot attend. Therefore, "a single belief and a common cult comprise both rulers and ruled, notwithstanding the religious barriers which otherwise separate the two, and add to the political dependence [...] another, spiritual, dependence"'.32 The Holy Roman Empire surely would not have been an empire without the Roman Catholic Church as a unifying force.
In sum, religious ritual and symbolism are indicators of systemic cultural perceptions and of culture change in more than just ideology. They serve as well as one dimension of measuring social, political, and even economic conditions and dynamics. Religious networks—networks through which religious ideologies are shared—may overlap, but still they define core areas.
II.3. Measuring Religious Ideology and Change in the Late Fifth and Early Fourth Millennia BC
As so often is the case, theorizing about religion is easier than measuring it so we can investigate the ancient past. Ancient societies leave fewer directly interpretable material remains for religious ideology than for economic or even political behavior. Analysis of ancient religion is fraught with the danger of imposing our own conceptions on a very different past. Because it is a matter of the native's perceptual framework and because we are limited to material remains, we need to find those material correlates of religious or more generally ideological perceptions. For Mesopotamia in the late fifth and fourth millennia this means temples, seal design, sculpture, perhaps painted pottery design.33 and funerary goods. In other words, it requires us to enter the world of art historians, but from an anthropological or archaeological viewpoint.
The most direct element in this search is certainly the temple: dedicated religious spaces or buildings. What, however, constitutes a temple? It must be more than an architectural form, because in saying that a building must house primarily religious ritual because of its shape alone, wc leave ourselves open to error. Public buildings need not be temples. From the perspective of Tcpc Gawra and more broadly northern Mesopotamia34 buildings with primarily religious functions (sec figs. 1 and 2). temples, must have:
• a tripartite plan with easy access from front or side of the building into an assembly area (often through double doors or an open courtyard),
• a sacrificial platform in the central room,
• niches, usually in the back wall of the main room, facing the front doors, or an altar.
• a large basin, often double rimmed, sunk in the floor of one of the adjoining rooms, usually one near the back wall with relatively open space, and
• carefully plastered and painted white or red walls.
I developed these criteria based on Gawra buildings, but Jawac!35 also cites the first three criteria as good indicators of religious temples from the Ubaid lo historic periods in Mesopotamia, as docs Kubba.3'' Implicitly, all three of us would add another criterion: few domestic or craft artifacts may be present in
2" Oatcs 1986. 382.
30 Nadcl 1940.
31 Nadel 1940,191.
32 Nadcl 1940.191.
" Sec forthcoming Charvat volume, The Iconography of Pristine Statehood.
34 Jawacl 1965.
35 Jawad 1965. 3f' Kubba 1987.
18
A propos de Tepe Gawra
the main sanctuary, although non-religious activities and buildings can exist in the temple compound. Using a partial set of criteria leads to erroneous assignments of function. That has clearly been the case at Gawra and at other sites. For example, Speiser37 asserts that there were four temples in level VII1C, a Southeastern (see fig. 1), Northeastern, Central, and Western temple. Analysis of the contents38 of these buildings showed that the Northeastern "Temple" was the home of an important official (chief?) with a shrine on the first floor, but only the Southeastern Temple was actually a special function religious building. The Western Temple was clearly a public building, but its remains show no signs of any religious function (fig. 1). To the contrary, it appears to be a place where grain rations were given out, flintknap-ping was done, and some goods were sealed. Contrary to other publications,39 there is no evidence of any temple from Gawra XII or XIA/B, or of any building with significant ritual purpose. By this same token, one wonders about the "temples" of Gawra XIII (fig. 2). The Northern Temple, the most complete of these buildings which were built in succession (that is, not occupied synchronously), certainly fails the criteria for a temple listed above. Its entry way is an indirect one on the long, northern side (the image in figure 2 is flipped vertically to match with other temples shown). This is hardly appropriate for easy public access. Its interior spaces lack the private areas, potentially for priestly worship or storage of supplies and sacred symbols. Contemporaneous temples from Eridu, which have similar form, have such spaces in profusion. What is specifically lacking is any offering table or altar, present in all the other clearly religious buildings in figure 2: the Ubaid Eridu VII and VIII temples, the LC 5 or Late Uruk White Temple from Uruk-Warka, and the LC 3/4 or Middle Uruk Eye Temple from Brak. A preliminary analysis of the functions of the Gawra XIII buildings also suggests that these buildings were not temples (see appendix).
What can we tell about the ritual of the LC 1-3 (Terminal Ubaid to Early Middle Uruk) from the temples we do have? If they are gathering places, "assembly halls" as Nissen40 suggests for the building in Uruk-Warka's Eanna district, they are for pretty small congregations. The middle halls with altars of the Gawra VIII, IX, and presumably X temples measured approximately 50 m2 (164 ft.2). For the Gawra XI/XA temple, the size drops to approximately 25 m2 (82ft.2). These are smaller than the size of the living room of a modest American house today. Parenthetically, the middle hall of the White Temple is approximately 150m2 (492 ft.2), of Eridu Temples VII and VIII approximately 62.5 m2 (205 ft.2). Packed tightly, Gawra VII, IX, and X temples could squeeze maybe 25 people standing, the XI/XA temple could accommodate fewer than that. In other words, these are not meeting places for congregations of people in significant numbers. As Margueron has said,41 most public ritual was probably conducted outside the temple building. The hollow bottoms of hut statues42 and the statues of gods may indicate that they were carried on sticks like statues of saints during Mexican religious fiestas in public squares.
If the temples were for smaller groupings, what was happening there? Was it just for the priests? The structure of these temples and the direct access of the middle rooms with their altars or offering tables from the large doors to the street would argue against that. If there were a kind of holy of holies, one could imagine that even in ancient times it would be shielded from the street. For Gawra what we know most about ritual comes from the V1IIA temple. The fire that destroyed V1IIA burned so hot it turned sun-dried mudbrick into baked brick where it fell. We therefore have extraordinary preservation (fig. 1). The centra] room in addition to an altar contained two censers, clay horns, and a grinding stone in the shape of a human fallus43 (a memo or pestle) with faint traces of a red powdery substance on the tip and a metate (mortar or rough grinding surface). The altar as everything else had signs of burning, but so did the altar of the XI/XA temple, which had not burned. In one side room of the VIIIA temple were piles of animal bone, in the opposite one, stored grain, a seal and sealings, an animal figurine, macehcad, and a metal piece. The large bowl sunk in the floor was actually in a special room outside of the temple entrance in VIIIA. It would not require much of a stretch to say that these bowls were likely for ablutions, that is, ritual purification. This all certainly adds up to a place for ritual sacrifice with the people doing the sacrifice coming from the outside.
37 Speiser 1935.
38 Rothman 2002a, 1301".
39 Margueron 1986 a.
TO Nissen 2001, 154-55.
41 Margueron, Joumcc consacrce a Tcpc Gawra symposium. Univcrsitc de Vcrsaillcs-Saint Queniin. Paris. 3 June 2003.
42 The ideological function of many hut statues (sec Rothman 2002a, 65-66) is somewhat supported by the images of double volutcd creatures on seals (figure 3).
43 Speiser 1935. Plate XLV1. b.
19
Subartu XXIII
There are many different meanings to rituals that fall under the rubric of sacrifice.44 Without a written record it is all but impossible to say which people cognized in these prehistoric times. One window we do have into prehistoric perceptions is the imagery they left, and in some cases, their mortuary practices. For northern Mesopotamia, the vast number of images comes from seals and sealings.45 There are also painted designs on pots. In this case, those pots may be of special note, because from Gawra XII to VIII, most pottery was unpainted, and painted pottery was most often recovered from areas with ritual or leadership functions.
The approach many46 suggest for interpreting these design elements is a semantic or linguistic one. Seals are like speech acts, meant to communicate certain messages in certain forms to particular audiences. Those multiple audiences, aside from various human social groupings, include the gods, much as the Hammurabi Stele is not really a law code directed at judges, but a plea of support from the god. Through it Hammurabi claims to be fulfilling his royal responsibility to dispense justice. As anthropological linguists have long claimed, semantics and speech acts are a very good entry way into the mental, perceptual aspects of culture.47 In this case the subjects of the designs are like words that carry meanings within a particular cultural system. Whereas art historians tend to focus on the grammar of art (form, rendering, and style), I am more interested in the connotation and denotation of words. As I already stated above, this analysis, though essential to try, is fraught with interpretive danger. DittmanV attempt to link seal design at Susa to administrative function is too literal, in my opinion, but it is an attempt of the kind I am making.
What is so distinctive in the subjects of the Late Ubaid into LCI/earliest LC2 (Gawra XII-XIA) Ubaid-Uruk transitional period is the presence of what are clearly shamanistic figures (fig. 3. a-e). Horned figures at Gawra and Susa also associated with snakes and eyes, the hollow-bodied, animal-faced figures at Gawra and Degirmentepe, and natural scenes all speak to a shamanistic tradition. Such a tradition is typical of less complex forms of social organization and not an institutionalized temple tradition. "The shaman enacts his [or her] roles in small-scale, multifunctional communities whose religious life incorporates beliefs in a multitude of deities, daemons, nature spirits or ancestral shades-societies that Dürkheim might have described as possessing mechanical solidarity f...] and segmental organization."49 By mechanical solidarity Dürkheim means a society with a lower degree of interdependence, unlike a state or complex chiefly society whose organic solidarity meant that perturbations in one societal segment would dramatically affect other economic and political segments. As 1 have argued elsewhere.5" trade and access to exotic resources in Gawra XII fit a model of mechanical organization with strong extended family groupings. These kinship groups, however, cooperate in central grain storage and trade. The glyptic of LCI across Mesopotamia speaks to this message, as may the sprig ware of this same period (fig. 3. n). In fact, the shamanistic and wild nature themes are shared in the terminal Ubaid and LCI across a mountain front from the Zaaros to Degirmentepe in the Taurus. The Susa examples both in glyptic and painted pottery51 arc from Acropole 27-25. They have clear ties to the intcrmontanc valleys of Fars and Kurdistan near the later Silk Road.52 Except for one example from Oueili53 these themes are missing from the Tigris Euphrates lowlands.
Because the situation in the south and northwest is so different than the northeast (see below) what we might be seeing here in the LCI and early LC2 are different networks of religious ideology. The claim of an Ubaid expansion has long been established in the literature. However, if there is such a distinct difference in religious ideology, we need to re-examine what that expansion means, if it is not more than just the diffusion of the slow wheel and related painted design on pottery. Even that is questionable, because the percentages of "Ubaid" pottery among the ceramic corpus at northern sites are actually low.54
The Early LC2 period that follows indicates a change in the glyptic and the perceptions of the Gawrans (fig. 3, f-j). "Most important, the repertoire of human figures expands from the single shamanic figure to extended groups of figures engaged in a variety of activities."55 These depictions often include buildings and also the hut figurines (fig. 3, i, m), the former of which was found in association with a building I interpreted as a shine
44 Bowen 1998, chapter 6, "Sacrifice Contested".
45 Rothman. 2002b.
4f'Rothman, 1994,2002b; Forest 1996; Charvat, forthcoming. 47 Hymcs 1962, 1964; Bolingcr 1975; Charon 1998. 4* Dittman 1986. 4" Turner 1997, 81.
50 Rothman 2002a. 2001b.
51 Voigt-Dyson 1992, 132. 32 Voigt - Dyson, ibid.
53 Pittman 2001,412.
54 Stein -Ozbal 2001.
55 Pittman 2001.417.
20
A propos de Tepe Gawra
room.56 It is toward the end of this period that residents built a temple facing out onto the plain below. This religious institution I have interpreted as part of an expansion of the Gawran polity, integrating nomadic pasto-ralists and farming populations. Parallel centers with temples appeared at Qalinj Agha, Grai Resh, and Nuzi, in a pattern reminiscent of the temples of chiefs of the Hawaiian Islands in pre-colonial times.57 This use of religion is meant to close the local network and funnel it through the center or town, which Tepe Gawra had become.
The Late LC 2 period saw a major increase in the role of the temple at Gawra. The glyptic continue in the pattern of early LC 2 (fig. 3, k-m). Hut statue spirits were depicted, as were worshipers. The dancing, presumably ritual men found on a seal (fig. 3,1) are paralleled by a bowl with a similar theme (fig. 3, p) associated with a shrine room over a deep tomb58. Animals, now frequently domesticated varieties, appeared in significant numbers (fig. 3, k, o). One repeated design from the trash of the Gawra IX temple shows caprids standing in front of the facade of a building, perhaps the temple itself.
Gavvra's continuous occupation from the Ubaid period ended after level VIII. Level VIII's glyptic represents a continuation of the images of the LC2, but elsewhere, particularly in the south, the trend toward images of centralized control and institutionalization of religion intensified (figure 3, q-s). This is the beginning of the so-called Uruk Expansion. "Now anonymous workers and domesticated or herded animals are the dominant subject matter. The horned shaman was transformed into a nude human hero who masters the forces of nature. Soon, this figure is replaced by the 'priest-king' of the Late Uruk phase."59 "Although the data are meager and poorly controlled, a fragmentary chain of evidence suggests that we should seek the origins of this change [cylinder seals and drilled design]—a change that undoubtedly reflected profound reorganization in the society—in southern Mesopotamia and Khuzistan."60 As evidence of the sources of sealing clay also suggests,61 the tightly controlled polity of Gawra XI/XA-IX has given way to a much more open system by LC3 Gawra VIII.
However, although priestly temples appeared in the Jazira and adjoining piedmont, the role of the temple north and south may have been very different. By the third millennium when the temple cult begins to emerge from the fog of prehistory, the gods of the south are clearly perceived in terms of the natural environment. "Thus in the south we find a group of city gods related to the marsh life and its primary economies, fishing and hunting."62 The economic independence of the temples as large land-owning enterprises also appears different than in the north, where in the Ebla texts "mentions of temples as institutions are notable in their absence."63 The southern towns of the Ubaid were consciously built around the temple. Large central temples dominated the settlement landscape. Temples in the north tended to be smaller and scattered throughout the major settlements.64 In addition, where we find large temples in the Euphrates area, they tend to be physically integrated into the palace complex, such as the ones at Arslantepe.65 Although I agree with Stein that temples were "not the initial nuclei around which the northern cities developed,"661 am not so sure that they were not "ideological attractors" in another way. In an area where de-centralized, presumably kin-based groupings played such a fundamental role, the integrative role of religion may have been quite essential. To assess that, we must look at the political structure and functions of northern polities. More study of ritual's role in the mobilization of labor for various activities is also essential.67
II.3.1. Religion, Economics, Politics, and Change
I take it as axiomatic that groups become promoted and hierarchies of leadership evolve not as inevitable results of events, but rather of conscious manipulation by actors in the drama. Changes give potential leaders opportunities to manipulate those within their grasp for labor, goods, loyalty, tribute, prestige, or ritual participation. They are coordinators, who later become policy-makers, regulators, and kings. So, the major distinction between kinship or community and administrative forms of social organization represented in the tribes, chiefdoms, and states model must be seen as representing modalities of political action. As Tony Wallace writes, "Kinship structures define boundaries of membership by relationships of marriage and parentage. Community structures define boundaries of membership primarily by common residence
56 Rothman 2002a, 104.
57 Earle 1991.78.
58 Rothman 2002a, 114.
59 Pittman 2001,427.
"> Pittman 2001,419-20.
61 Rothman-Blackmail 1990.
"2 Jacobscn 1976, 25.
63 Stein 2004, 75.
64 Stein ibid.
65 Frangipanc 1997. "f-Stein 2004, 76.
" Stein 1994, 4If.
21
Subartu XXIll
in a territory (or, metaphorically, by common adherence to a profession). Administrative structures define boundaries of membership largely by contract (as in bureaucracy) or by traditional claims to personal service and depend on a hierarchical branching structure of super- and subordination to channel communication downward of plans and upward of reports".68 Clearly. Wright and Johnson's definition of the state fits Wallace's definition of an administrative form of social organization: "a state will be defined as a society which is primarily regulated through a differentiated and internally specialized decision making organization which is structured in minimally three hierarchical levels, with institutionalized provision for the operation and maintenance of this organization and implementation of its decisions."69 As Service writes, "Chiefdoms are particularly distinguished from tribes by the presence of centres which coordinate economic, social, and religious activities. The great change at the chiefdom level is that specialization and redistribution are no longer merely adjunctive to a few particular endeavors, but continually characterize a larger part of the activity of the society."7" Wright71 takes Service's idea and expands on it. citing the great range of societies that fall within the chiefly type of society from simple to complex.
However, to say chiefdom or state, even if we could determine the structure of society, is not enough. We might accept a major role for southern city leaders in the Uruk Expansion and external exchange as keys to understanding larger trends in Greater Mesopotamia during the fourth millennium BC. However, we must not see these places as cookie cutter forms. Leaders have different goals and options in their relation to various segments of society. We again must see a possible multiplicity of leadership groups or elites. Even if strong leaders are at the top of the clear political and social hierarchies, custom and belief may restrain the power and role of leaders. That is to say, in what Blanton et alia call an "exclusionary power strategy",72 political actors seek to amass for themselves wealth, status, or political authority. An opposed but equally common "corporate strategy" shares authority, decision-making, and rewards across "groups and sectors of society in such a way as to inhibit exclusionary power." According to them the exclusionary strategy in economic terms can be seen as a "network strategy" in which those who wish to garner wealth, status, and political authority must establish ties outside of their local polity. These ties are represented in exchange relationships in a wider network outside their own local exchange sphere. This strategy is especially clear when leaders attempt to control the distribution of rare and highly valued goods or their manufacture for export. This type of strategy may well be a common phenomenon in many of the northern and eastern sites that exhibit the most emulation of Southern Uruk artifacts.
Arslantepe appears to have been a classic case of network strategy, as that determinedly local site clearly had based its amazingly centralized palace system on its role as a production center, especially for metals, and as a trading intermediary.7-1 It was not a case of imperialism or of colonialism. The site was never "taken over" by Southerners, but its leaders' attempts at a network strategy brought both Southern artifact styles and administrative technology—as well as Transcaucasian manufacturing technology and wares—into the site.
II.3.2. The Case of Tepe Gawra
So, we must start with local conditions for the data set I offer here. For Tepc Gawra and the piedmont and Jazira plains, the realities are formed by three environmental and topographic factors. This area of northern Mesopotamia was an agriculturally rich, rainfall zone. Gawra particularly sits by a natural spring at the intersection of plains and the first foothills of theZagros. Both wild animals and sheep and goal pastoralists—this area is attested in modern times as part of the lowland pasture for migrating pastoral nomads—provide its residents with rich animal resources: milk, meat, libers, and sinews. Gawra further is situated by one of few natural passes through the Jebel Maqlub onto the Iranian Plateau. Its early role in transporting lapis lazuli, ultimately from Badakshan in Afghanistan, is probably a result of its fortuitous location.7,1 Gawra is also not far from the Tigris along one of its navigable tributaries. To its north lies a treasure-trove of resources including obsidian, copper, gold, lead, semi-precious stone, soft carving stone, wood, and so forth.
Chronologically, the relevant levels to discuss here, XII-VI1I. began at about 4400 BC in the Terminal 'Ubaid or LCI Period.75 Levels X1A/B and XI/XA lit in the same time horizon as (he Early Uruk in the South, SAR early LC2 or Gut's Gawra A, 4100-3900 BC. Levels X and IX were occupied in the Early Uruk Period, late LC 2 or Gawra B, 3900-3750 B.C. The date of Gawra Vlll's three major re-buildings is early
68 Wallace 1971. 1.
69 Johnson 1973, 2. 7,1 Service 1962, 143.
71 Wright 1994.
72 Blanton etal. 1996. p. 2.
" Frangipanc - Palmicri 1983, Frangipanc 1993.
74 Hermann 1968.
75 Rothman 2001a: tabic 1.1, 2002a: chapter 3.
22
A propos de Tepe Gawra
Middle Uruk, LC3, or Uruk A at about 3750-3650 BC. In other words, it ended well before the construction of Habuba Kabira, the Uruk-influenced Ninevite IV, or Arslantepe VIA (so, maybe contemporaneous with early VII or VIII). Gawra VIII was nonetheless occupied at the beginning of the first Uruk Expansion somewhere between 3700 and 3600 B.C.
We also know from older surveys—no modern survey has been done in this area—that there were a series of small sites in Gawra's area. All of these sites overlapped some of the time of Gawra XII-VIII- for example Arpachiyah was occupied in LCI and into early LC2- and some outlived the fire that burned level VIII. For example, Khirbet Ba'wiza continued to be occupied from the LC3 until the Late Uruk, LC5 or Uruk B, and Khurkruk was occupied from the LC 2-4.76
We think of Nineveh as the great center of this area. However, there is a real possibility that Nineveh was not a major center, perhaps not even occupied in the critical Ubaid-Uruk transition period (LCI) and in the earliest Uruk, Gawra A, or early LC2 period. If, as I argue, Gawra served as a small center in these piedmont zones away from the Tigris, Qalinj Agha and Yorgan Tepe (Nuzi) also fulfilled the same central functions in the circumscribed piedmont areas south of the Greater and Lesser Zab rivers. None of these small centers ever equaled the larger settlements on the open plains near riverine or major overland routes like Nineveh, Hamoukar, Tell al-Hawa, or Brak. Travel barriers circumscribed them all in the winter.77
II.3.3. Changing Functions and Town Plans
So what role did Gawra play through this long span of time? What sorts of networks was it a part of? How do I know that Gawra was a center and what kind of center it was? How does the town plan reflect the importance of Gawra's various roles? To do this it was necessary to determine the various functions—sets of activities like craft production, housing, religion, or governance—of the site. To do that I had to reconstruct the town plans.78 The settlement structure and traffic patterns of the town plans may themselves be variables to measure function and change.
Evidence of the rale of Gawra as a center, again a site which has unique functions not repeated elsewhere in its locality and serving a multi-site polity, is already evident in Gawra XIII, a Late 'Ubaid settlement of approximately 4600 BC.79 (figs. 4 and 5) Three public buildings, each too large to serve just the residents of the Gawra mound, were built at the eastern edge of the mound. Excavators recovered a series of discarded clay locks and other artifacts in the well of one of those buildings (see appendix). These clay locks, sealed with incised stamps, were the basis of administrative control of goods, as presumably only those authorized to break the clay lock on doors, around the mouths of jars or sacks, and on baskets could access goods.80 Although some seals even this early were surely used by individuals, in such contexts they were surely institutional. As discussed above, these public buildings may be social centers rather than temples per se.
The next major level at Gawra, XII,81 is parallel to Susa A with its rank-marking vases and massif brick platform. At Gawra there are no such massive buildings. There are no temples in the half of the mound exposed (fig. 6). In fact, the remains of the town are typified by a series of parallel large, possibly two-story buildings with the same layout.82 Each, including the largest, the so-called White Room building, has a large central room or mittelsaal like the traditional Anatolian salon. They are all multi-function spaces with evidence of cooking, of craft, agricultural tools, and religious artifacts. Even the largest of these buildings, the one with the White Room, housed a similar set of functions. A large number of burials, especially young children and infants, were found under the floors of each. This I interpret as evidence of extended family dwellings also found at Late 'Ubaid Tell Madhur and Abada. An analysis of the distribution of imported material, including obsidian nodules, shows an even distribution among these houses.83 Two elements mark Gawra XII as more than a village. There are two large buildings, one which appears to have been a grain storage facility, with broken sealings and a seal, grinding stones, and piles of the precursor to the beveled rim bowl, the Wide Flower Pot. I call them ration bowls, but these ceramic forms are associated with a wide variety of purposes from bitumen mixing to baby burials to construction fill. The other is a series of rooms near the entryway with an area of bins behind them for sorting items, presumably the craft goods, including items of imported lapis, gold, and obsidian blades found in the small storerooms. The
76 See Rothman 2001b, figures 10.3-10.7.
77 Rothman 2002a, 11; Rothman 2001b.
78 To see how this was done sec Rothman 2002a, Chapter 4. Contrary lo Forest, this effort is not "desesperee". 79Tobler 1950.
80 Rothman 1994.
81 There is a level XIIA, but it provides little information. 83 See Rothman 2002a, figure 3.6.
83 Sec Rothman 2001b.
23
Subartu XXIII
quantity of stored goods, the central grain silo, and the number of manufacturing tools belie interpretation as a simple village. It is my belief that Gavvra already was taking on some of the functions of a small town center, one dominated by craft production and exchange, as I will discuss below.
Gawra XII was burned. In XII dead bodies were found in the street with stones in their backs. The large Round House building in XIA, which was a fortress with grain storerooms, was also burned. Clearly, the functions of Gawra in this terminal 'Ubaid (LC 1) and early LC2 periods were sufficiently specialized and noteworthy to create the competition that led to their demise. As Blanton etal write.84 societies in which a group emerges to administer such central functions are politically volatile and may be open to attack from inside or outside the polity. If my proposition that a large mobile population - pastoral nomads - was one key element of the Gawran polity, as it certainly was at Mari, a strategy to incorporate them and keep them from destroying the polity was necessary. I believe that trying to generate a strategy to accomplish this catalyzed the evolution of the Gawran site. It was not until Level XI/XA that the key was in place.
Gawra XIA/B was in many ways a continuation of level XII. now rebuilt after the destruction (fig. 6). Contrary to Tobler,85 no temple structures were excavated in this level. The building Tobler identifies as a temple may, however, have some specialized productive function, based on the number of spindle whorls, bone needles, and blades of a type used in weaving found there. That same building had a large pile of Wide Flower Pots by a large oven. One wonders whether it was not a place where large numbers of meals were cooked and given out for rations, but that is pure speculation. The same range of domiciles, craft making areas, and so forth that describe XII also typified level XIA/B. Unlike level XII. level XIA/B had a number of smaller domiciles in addition to large, multi-family houses. The clear status and role of the extended family may have been waning at this time. The most spectacular functional area of XIA/B was the Round House, the large thick-walled round structure in the center of the town. This building was constructed near the very end of the occupation of this level. Stratigraphic analysis, especially from the southeastern quadrant of the site, shows that before the Round House and the gate and tower structures in the northeast quadrant were built, an earlier stratum of small houses existed, which we have named XIB. The Round House served as a granary and fortress for the occupants of Gawra. In general, the construction of buildings indicates that in addition to the gatehouse, the walls of houses along the mound edges formed a kind of defensive wall. As I already said, it did not save the residents of XIA from the fire.
The next major step in the evolution of complexity is evidenced in phase XI of XI/XA (fig. 7). A large temple was built with its entrance facing out east into the countryside and hills, not into the interior of the town. The craft activities evidenced in XII and XIA/B, including thread spinning, cloth making, stone tool knapping, wood carving, and small clay object making, continued. In Gawra XI/XA these activities became specialized. That is, many more were produced than could possibly be consumed by the Gawrans themselves, and they were segregated into a special facility. A separate wood working shop, weaving shop, and area for firing small clay figures possibly for ritual were recovered from special-function buildings or areas. A public building whose purpose seems largely administrative was also built. Houses were now small, not the multi-room extended family houses of XII, whose form seems transformed into the architecture of public buildings. One of the significant changes in this period seems to be somewhat of a change in strategy. Seals and sealings did not appear in domiciles, only special function buildings. As a town center, the political actors at Gawra appear to be consciously drawing the wider population into the center. At the same time, an analysis of sealing clay shows that the leaders of the Gawran polity were primarily tightening their grip in their polity network.86 As already stated. I propose that this population includes pastoralists, as I saw in modern Dezful, Iran, in the 1970's. For them, Gawra would offer a market for wool, meat, cheese, and whatever products or raw materials they managed to pick up on their seasonal migrations. They would, in addition, be brought into the religious orbit of Gawra, whose temple ritual included animal sacrifice. That sacrifice from the evidence of the temple of VI1IA is primarily of sheep and goat. As Blanton ct alia propose, the use of knowledge and ritual, a powerful political tool, tends to be locally directed.87 All indications are that the Gawra of XI/XA is oriented that way.
Gawra X and IX, late LC2 or Gut's Gawra B, represent a new strategy for the administrators of Gawra (fig. 6). If XII (fig. 7) and XIA/B (fig. 8) concentrated on grain storage, and XI/XA on craft manufacturing and religion, X and IX were focused largely on religion. The same craft activities continue to be present on the Gawra mound, but they arc dispersed into private houses. The temple and a large administrative building dominate the site in both levels. Remains of administrative activity concentrate around the two main functions of the town. This is the time when the richly endowed tombs from Gawra emerged. More on that later.
84 Blanton etal. 1996.
85 Tobler 1950.
86 Rothman - Blackmail 1990.
87 Blanton etal. 1996.
24
A propos de Tepe Gawra
Gawra VIII was occupied in earliest LC3 or Gut's Uruk A after the first incursions of Southerners into colonies (fig. 8). For those who have not read the new SAR volume, our chronology and analysis implies that there were three stages in the supposed expansion:
• the first is the arrival of small trading posts, like those at Hacinebi;
• the second fully southern outposts like Sheikh Hassan, and
• the third a full migration of southerners, whether for trade or not, at sites like Habuba Kabira, Jebel Aruda, and perhaps Tell Brak.
The potential of increased trade for Gawrans would certainly affect the functions and governance of the site and its polity. The sealing clay analysis mentioned earlier shows that the network had opened to trade with rulers being more interested in trading through regional networks than controlling their local one.88 The fewest number of private domiciles from the time of XII onward are found on the VIII mound. Only potentially one or two houses may exist within the confines of administrative buildings in VIIIC and B, and those that are in VIIIA are of very large size with large enclosed courtyards and specialized craft areas. In these craft areas seals and beads, as well as obsidian artifacts were being produced in large numbers. A significant number of obsidian nodules and obsidian cores were recovered from level VIII, and most of the blades had no signs of use wear. In addition, residents built a central storehouse, very reminiscent of Shaikh Hassan 7.89
Clearly, Gawra at level VIII was primarily a specialized center with little evidence of a significant population.
As already stated, Gawra VIII ended in a cataclysmic fire. A number of other key sites appear to have been abandoned very close to the same time: Grai Resh to the West near Sinjar, Qalinj Agha to the south, Musharifa and Rifan north on the Tigris, and nearby Shaikh Yosef.90 Nineveh appears to have begun a major growth spurt. Tell al-Hawa appears to have shrunk some in size, although its role as the central place on the North Jazira continued.91 The time of the major migrant cities of Habuba Kabira and Jebel Aruda, LC 5, saw major growth at Nineveh. The presence of clay sickles in level 35 could indicate a southern presence, although not necessarily the southern take-over Algaze proposes. This is also the period of the hollow ways around Tell al-Hawa, visible ruts that may be well-traveled roadways. Wilkinson and Tucker write, "Late Uruk sites were apparently significantly related to routeways. The correspondence between the development of urban centres and routes suggests that communications were stronger and settlement growth may have been stimulated by these links and perhaps trade along them."92
II.3.4. Trade, Ritual, Leadership, and Social Structure
All of the foregoing raises questions about the relationship of trade, religious ritual, leadership, and social structure within various networks. Both from a data and interpretive point of view, these are difficult questions to answer, and sometimes even to frame.
Gawrans certainly saw major changes between the founding of level XII and the end of VIII. As I have written elsewhere, mapping of seals and sealings, representing in this case the action of administrators, on the functions described above shows an increase in administrative complexity.9' Concomitant with that we see ever increasing social differentiation in graves. Both the types of graves changed, and the marking of status through grave goods, many from imported materials, increased.94 How then do we understand why these changes happened?
Traditionally, to answer this, we look first at political and economic organization and settlement pattern. Because we see in the origin of cities and social organization of the kind we term states precursors of our own time, we tend to emphasize them. Service's step evolutionary model of bands, tribes, chiefdoms, and states lies behind this procedure. However, its critics rightly say that sometimes we are too intent on naming social organizational types and too little interested in explaining why they evolved, how they worked, and why they changed. We concentrate too much on leaders and their elites rather than seeing multiple groups of influence and action, what scholars call heterarchy.9?
BS Rothman - Blackmail 1990. 89 Boese 1995/6. ""Rothman 2001,381-383.
91 Lupton 1996.
92 Wilkinson - Tucker 1995. 86. " Rothman 1994.
94 Pcasnall 2002.
95 Stein 1998.
25
Subartu XXIII
The distinction of state, chiefdom, and tribe is nonetheless an important one. because it gives us a way to compare societies. By comparing similar phenomena, we begin to tease out what is distinct about how these societies were organized, functioned, and changed. Service's types are heuristic tools, not a reality any contemporary would understand. If we could talk to the ancients, none would tell us. "1 live in the Gawra complex chiefdom", nor "I am a citizen of the Uruk city-state".
Still, what effect does trade have on forming and empowering leadership groups? In other words, what role do coordinators play in this activity and how does it provide new opportunities for leaders' self-promotion? If Algaze is right, it was at least for southern Mesopotamia the building block of economic prosperity and elaboration of royal control, but did they "profit" because they controlled it or profited from it? Alternatively, the issue may not be control at all. These goods and the leaders' role in obtaining them may also be important as a kind of propaganda. Renfrew writes "we cannot discuss commodities or the development of economy without considering such embedded social concepts as value and demand".96 As Pollock writes,97 goods or raw materials become imbued with value, because they carry social messages about the status of the individuals who own and display them, whether in life, or in death. The same can be said for the earliest use of copper. Renfrew writes,98 "In most cases early metallurgy appears to have been practiced primarily because the products had novel properties that made them attractive to use as symbols and as personal adornments and ornaments, in a manner that, by focusing attention, could attract or enhance prestige." Part of the ideology of early kingship in Mesopotamia was the idea that like the gods, rulers were providers of what is valuable, often rare and from a great distance away.99
Does trade require formal leadership? For goods in down-the-line trade10" very little organization is required, and very little productive activity is needed to meet the demands of trading partners. This is exchange in which goods move from the source area to nearby places. The recipients keep a significant percentage of the materials or goods and pass some on down the line. By the time these materials or goods reach the end of the line there are very few.
Once a society moves beyond down-thc-line exchange an ideological shift may be required in order to create a society where production is for more than subsistence or use as gifts for tribal members.101 The role of leaders may be to mobilize the labor in its local networks. Certainly, seal images and other evidence suggest that the rulers of Uruk-Warka ran a major textile weaving enterprise under their direct control, most likely with corvee labor.
At the same time, in Gawra XII when the distribution of seals and scalings and of functions give little indication of formal administration, Gawrans were involved in considerable exchange. The gold, lapis, and obsidian could have come through down-the-line exchange, but a much larger exchange network of pottery has been documented. For example, sprig ware manufactured at Shelgiyya up the Tigris was coming into Gawra, and impressed wares from Gawra XII to XI/XA were part of a separate exchange network in which Tell Brak and Gawra were involved.102
In other words, social stratification need not be catalyzed by trade per sc. but follows or is concomitant with it. My own prejudice is, as Renfrew writes.1"3 [regarding] "the relationship among three important variables: a developing system of production and exchange: the circulation of goods of prime value (especially in the early stages); and the emergence of prominent social ranking \...] the obvious inference is that the three can most readily develop together through a kind of multiplier effect, where each mutually enhances the others."
The development of regulating, hierarchical administrations is not a necessary result of production and exchange systems, although nascent leadership groups can use trade symbolically and economically to enhance their power and prestige.
In one sense, what was most important to the ancient Gawrans is clear in their town plans, particularly the center of the mound. In Gawra XII it was a common grain storage and buildings for storing trade goods in XIAB the Round House, in XI/XA craft production, in X and IX religion, and in VIII storage of craft and religious artifacts.
96 Renfrew 1986, 142^13.
97 Pollock 1983.
98 Renfrew 1986, 146-147. '•" Jacobscn 1976.
"'" Ren few elal. 1969.
11,1 Appaduri 1986.
102 Rothman - Blackmail 2003.
>m Renfrew 1986, 163.
26
A propos de Tepe Gawra
II.4. Changing Networks and Evolution
The foregoing presents the outlines of a picture of the ancient Mesopotamian past. I believe that we must see the ancient reality as complex, confusing, and messy as it was. There was no unilineal evolutionary path. There are no cookie cutter models that will fit each circumstance, although there are repeated patterns that may help us explain some developments.
Network analysis can help us understand the functioning of a system especially as it approaches and then becomes a state. According to the tenets of network analysis, an administrative form of social organization (as opposed to a kinship-based system) is a network in which leaders establish certain kinds of interactions with the individuals and groups who are their subjects. These interactions are not open-ended, because leaders have the authority to control them, depending on the basis of their rule. Those bases could include storing food for later redistribution, guaranteeing trade, controlling the production of goods, defending their territory, positioning themselves as the surrogate for the divine, or other ideological posturing. Based on the needs and potentials of the system, and leaders interference in its functioning, people's behaviors are to some degree limited and directed, institutions arise as the concrete entities through which activities are carried out, and so forth.
At the same time, we must not confuse our heuristic models with the reality the ancients experienced. As Drennan writes,104 "It seems highly unlikely that the natives of Mesoamerica [or Mesopotamia] accepted the social conventions granting higher status to a restricted set of individuals because they realized that is was necessary in order to maintain a system of economic symbiosis, which in turn enabled them to enjoy various economic advantages of specialization".
What we do see is the formation of a number of linked networks emanating from the local polity, encompassing a wider geographical range and involving various cultural subsystems. The LCI or Late Ubaid/Ubaid-Uruk Transitional period saw a series of networks open to the world outside their narrow geographical range. Cultural exchanges involving religious ideals, artifact style, craft technology, seal shape and design, and economic exchange in exotic goods created commonalities across a wide area, mostly in the piedmont and intermontane areas of the Zagros and southern Taurus Mountains. At the heart of the northern and intermontane politics in this period were de-centralized systems with a very strong element of extended kinship. Some kind of coordinating organization was probably needed for centralized grain storage and craft manufacture. In southwestern Iran at Susa, the contemporaneous Susa A /Hflj-sj/indicates more complex chiefly social organization. Evidence of significant military activity is limited, although Gawra XII was burned, probably in an attack. Of the periods discussed here, LCI and LC3 are the ones when Gawrans were most receptive105 to new ideas.
Widespread networks of exchange continued into the LC2. However, if Tepe Gawra levels XIAB-IX are typical, networks funneled through centers were found throughout Mesopotamia. Religious ideology appears to have been an important element of these networks, used to bring together disparate populations. Extensive trade in such luxury items such as impressed pottery extended throughout the Jazira and into the Taurus, The attempted control by the centers made this a time when Gawrans were less able to acquire and utilize "foreign" ideas.
LC3 marked the beginning of the Uruk Expansion. A more open, regional focus is evident at Gawra VIII. Seal design and technological change in their manufacture appear to mark a shift in the exchange patterns and influence of southern Mesopotamia. The manufacture and no doubt exchange of obsidian blades also indicates the establishment of a new exchange network. It gave the evolving leadership of the Gawra and other sites a chance to establish a "network" strategy, as described by Blanton et alia.m Otherwise, Gawra was little touched by this first international trading system.
Throughout these shifting expanding and contracting networks, we see a complex picture of the Greater Mesopotamian region evolving through time.
Appendix, Tepe Gawra XIII
Level XIII was mostly excavated in the sixth season, again under the direction of E.A. Speiser. Unfortunately, Speiser did not continue recording the extraordinary stratigraphic detail that Bache had developed in seasons three to five. In addition, the architect's drawings (simplified in Fig. 4A and reproduced in fig. 4B) did not provide the detail that the earlier ones had for constructing stratigraphic schematics
104 Drennan 1976,346.
105 Winter 1977.
""' Blanton ctal. 1997; Rotluvmn - Pcasnall.
27
Subartu XXIII
(see Rothman 2002a). The original architect's drawings do have the architect's numbered points, but the surveyor's book that remains in the Archives of the University of Pennsylvania Museum does not cover this level. The drawings in Figure 4 do delineate the general stratigraphic position of XIII below level XIIA and above XIV. In the case of XIV, builders apparently scraped the mudbrick off the walls and built immediately on top of the stone foundations.
The drawings also suggest a slightly different reconstruction of the so-called Central and Northern Temples (fig. 4 and 5 here). The many doors of the Central Temple from Tobler's published drawing (1950, fig. XI) have never made much sense to me. The original drawings indicate a large room in the northern part of the East Temple and a complex doorway system into a large room that eroded off the mound. A second doorway into the Northern Temple from the north is also indicated.
All the excavators agree that the three buildings were not standing simultaneously. Each of the project participants have a different theory. Speiser believed that the Eastern Temple was built first, then the Central, then the Northern. The architect, E.B. Mueller suggested that the Northern and Central were built first, then the Eastern, and Tobler thinks that the Eastern was built first, then the Northern, and the Central last. As already stated we lack the points that would tell us the bottom elevations of the walls, the most critical. Two factors suggest to me that Speiser was probably correct. Although floors can be re-laid and re-laid, raising their elevation. However, given that the Eastern building was constructed directly on the stone foundations of level XIV, and its floors are among the lowest, that should make it the earliest. Second, the well that extends into level XIV and down is under the lowest floor of the Northern building, which has the highest floors and is overall the best preserved. The floors of the Central building are closer in elevation to the Eastern than the Northern. All are built not to overlap the others. By the same logic, pieces of a northern building in squares 5 and 6 S and Q probably belongs to XIII and not XIIA.
Most important, however, are the functions of the buildings. As already stated above, the hallmarks of later temples are missing from these three structures. In terms of artifacts, below is a table of all the registered artifacts (from registry books and chits). Lacking good provenience (elevation and context) information, they are grouped in squares associated with buildings (the well, most likely was associated with the Central building). Seals and sealings are prominent, however, most seals are found outside of any of the buildings among the smaller, thin walled structures in the east of the plaza, and northwest of the northern temple. Sealings came mostly from the well associated with the Central building, and a set from the Northern building. The remaining artifacts represent ordinary domestic activity: cooking pots, serving cups and serving bowls, awls, whetstones, and celts, not in a particular concentration. The only types of artifacts that do not fit a domestic or entertaining function are maceheads, figurines, pendants and amethyst balls. None have overtly religious functions.
So, what are these buildings? At this point I can only speculate, but given the development at Gawra in particular and in northern Mesopotamia in general of public, social buildings, these building probably filled those functions. From tholoi in the Halaf period to the Round House of LC 2 Gawra XIA, Fortress of XI and what I interpreted as a chiefly building in LC 3 Level VIII. these sorts of special function buildings are common. They were not formal temples.
Table I: Registered Artifacts from Gawra XIII
						Central	Temple		
6-0442				3J			painted cup		
6-none		4251	XIII	3J			bricks w/red paint	b2	
6-0552		3127	XIII	3K			painted foot	J7	under floor central temple
7-0074		1822	well	3M			sealing		
7-0079	38-16-65	1823	well	3M			jar		light green
7-0080		1822	well	3M			celt		
7-0098	Baghdad		well	3M			sealing		animals
7-0099	38-13-83		well	3M			sealing		shield
7-0)00	Baghdad		well	3M			sealing		row of men
7-0102	38-13-68		well	3M			weight		
7-0103	Baghdad		well	3M			bowl		painted sectors
7-0105	Baghdad		well	3M			jar		dark paint crude
7-0106	38-16-64		well	3M			jar		grecnware blotchy painl
7-0107	Baghdad		well	3M			knife		obsidian
28
A propos de Tepe Gawra
7-0108	discarded		well	3M			high sided bowl		gray brown paint bottom
7-0109			well	3M			bone awl		
7-0110			well	3M			bone awl		
7-0111	discarded		well	3M			large open bowl		red course
7-0112	38-13-59		well	3M			beaker		paint design? paint base
7-0113	38-13-61		well	3M			vase		fine 86mm tall paint design
7-0120	Baghdad		well	3M			sealing		4 shields
7-0121	Baghdad		well	3M			sealing		same as 7-80
7-0122	38-13-81a-c		well	3M			sealing		square shield
7-0123	38-13-81a-c		well	3M			sealing		round shield
7-0124	Baghdad		well	3M			sealing		teardrop shield
7-0125	Baghdad		well	3M			sealing		square shield
7-0126	38-13-73a-f		well	3M			sealings		shaman with wild animals
7-0127	38-13-85		well	3M			sealing		circle shield
7-0128	38-13-79		well	3M			sealing		circle animals
7-0146	38-13-74		well	3M			sealing		shaman & animals
7-0147	38-13-75		well	3M			sealing		birds gazelle fish etc.
7-0148	38-13-76		well	3M			sealing		animals
7-0149	38-13-77		well	3M			sealing		horned animals
7-0155	38-13-65		well	3M			cooking pot		gray ware
7-0156	38-13-67		well	3M			groove steatite slab		
7-0157	38-13-62		well	3M			cannon spout bowl		burnished red
7-0158	Baghdad		well	3M			deep bowl		high sided dark paint
7-0160	Baghdad		well	3M			spouted jar		2 spouts mottled cooking
7-0162	Baghdad		well	3M			beaker		brown paint design
7-0165	38-13-82		well	3M			sealing		cross shields
7-0178	38-13-70		well	3M			bone awl		
7-0187	38-13-58		well	3M			beaker		fine paint triangle squares
7-0188	38-13-78		well	3M			sealing		square animals
7-0237	Baghdad		well	3M			beaker		vertical lines triangles red buff
						North	Temple		
5-1720	36-6-348	1463	XIII	4M	10P	7.04	sealings 3		deer in bush
5-1746	Baghdad	1548	XIII	4M	10P	6.57	bowl		
6-0397				3M			painted pot		"oven room c"
6-0401				4M			drain pipe		near N temple
6-0426	37-16-329	4520	XIII	3M			stone pot	e6	under floor N temple by cella
6-none		4502	XIII	4M			unpainted bowl	il	outside north temple
						East	Shrine		
6-0360	37-16-243	4252	XIII	4J			carved pot base	g8	triangles rectangles incense
6-0443			XIII	4J			obsidian pendant		
6-0445	37-16-225		XIII	4J			small painted pot		under e temple
6-0446			XIII	4J			painted vase		under e shrine
6-0447			XIII	4J			bowl		
6-none		4253	XIII	4J			clay peg	i7	
6-0308	37-16-332	1943	XIII	4K			whetstone		near SE wall temple on floor
29
Subartu XXIII
6-none		1947	XIII	4K			clay		w/impression of cloth
6-0353		1807	XIII	51			spouted cooking pot		se wall central room
6-0356	37-16-224	4272	XIII	5J			painted bowl	c8	
6-0367		4279	XIII	5J			sharpening stone	b9	gray limestone floor temple
6-0369			XIII	5J			painted cup		
6-0373	37-16-331	1806	XIII	5J			pol fragment	J5	floor room d central temple 3
6-0374	37-16-331		XIII	5J			seal crisscross		
6-0375			XIII	5J			small green ware cup	h6	long room from floor #3 temple
6-0381	37-16-325	1948	XIII	5J			votive bricks		long room burned on floor
6-0384			XIII	6G			rope grooved jar		
6-0403				6G			amethyst balls		
7-0116	38-13-57	1968	XIII	6G			small cylinder jar		cloth impressed exterior
			North	of	North	Temple			
6-0412	37-16-229		4510	40			painted jar		carinated jar
6-0449	37-16-453			40			seal		
6-0451				40			stamp seal		
6-0551		3135	XIII	40			red painted pot	a5	
6-none		4298	XIII	40			crystal pendant	al	
6-none		4505	XIII	40			bone awl	e5	
5-1714		1531	XIII	4Q	8P	6.29	ugly dish	e/f 8	
5-1721	36-6-339	1346	XIII	4Q	8P	6.45	human figurine		
5-1726	36-6-318	1492	XIII	4Q	8P	6.43	handled jar		painted
5-1734	discarded	1490	XIII	4Q	8P		red dish	c/d 7/8	on floor
5-1735	36-6-317	1539	XIII	4Q	8P	6.17	small jar		
5-1748		1550	XIII	4Q	8P	5.91	bone	b5/6	
5-none		1538	XIII	4Q	8P	6.40	pottery	d7	shaped like toadstool
6-0361	37-16-236		XIII	4Q			bottle		
6-0371		4271	XIII	4Q			pendant	a9	white stone
6-0389	37-16-354	4504	XIII	4Q			stamp seal	jio	from 50?
6-0434	37-16-230		4526	4Q			painted cup	e2	
6-0462				4Q			bow]		
6-0435				4Q			stamp seal		
6-0387			XIII	4Q			bead		
					Plaza	area			
6-0122		1754	XIII	5K			shallow bowl	b8	
6-0123		1753	XIII	5K			amulet		
6-0124			XIII	5K			painted spout bowl	well	
6-0132			XIII	5K			spouted pot	well	
6-0319			XIII	5K			goblet		
6-none		4141	XIII	5K			diorite cell		
7-0393		2150	XIII	6K	11M	6.30	macehead	g/h4	unfinished
	Building				East	of	North Temple		
6-0240	37-16-358	4134	XIII	5M			seal with shnman	c4	
6-0429		4522	XIII	5M			macehead	J6	
7-0320	38-13-56	2082	XIII	5M		7.02	celt		reddish stone
30
A propos de Tepe Gawra
7-0322	38-13-53	2082	XIII	5M		7.02	ring base bowl		paint squares design fine ware
7-0332		2082	XIII	5M		7.00	4 crude dishes		
4-0606		14		50	9N		steatite vial		
4-0607		19		50	9N		mortar		
4-0608	35-10-282	15		50	9N		steatite weight		
5-1653	Baghdad	1502	XIII	SO	9N	7.02	pendant		
5-1656	Baghdad	1443	XIII	50	9N	7.05	stamp seal		sprigs
5-1661	36-6-347	1507	XIII	50	9N	6.60	celt		
5-1697	disc	1528	XIII	50	9N	7.05	pot		
5-1711		1477	Xlll	50	9N	7.10	amulel	e/f 6/7	stone
5-none		1535	XIII	50	9N	6.85	celt	b6/7	ash
6-0391		4506	XIII	50			engraved bead	b5	
6-0399				50			small jat		
6-0422	37-16-330			50			palette		marble
6-0437	37-16-335			50			bored celt	c8	
6-0436				50			pottery nail		
6-0223	37-16-319		XIII	5Q			peg		
6-0396		4509	XIII	5Q			copper awl	d8	
6-0398	37-16-235	4292	XIII	5Q			pottery basket	h3	
6-0400	37-16-333	4293	XIII	5Q			basalt metate	e2	
6-0417	37-16-353			5Q			stamp seal		
6-0425	37-16-242			5Q			applique pot		
6-0450				5Q			game piece		
6-0455				5Q			bowl		
6-none		4300	XIII	5Q			awl	go	between walls
6-none		4300	XIII	5Q			bead	g6	between walls
6-none		4501	XIII	5Q			bird figurine	dl	
6-0390	37-16-382	4508	XIII	60			bead		diamond shaped
6-0413				60			plain bowl		
7-0291	Baghdad	2166	XIII	60		6.40	hollow foot	c9	stone
7-0292	Baghdad	2)66	XIII	60		6.40	leg of figurine	c9	clay
7-0295	Baghdad	2166	XIII	60		6.40	small bowl painted	c9	criss-cross and band design
7-0399	Baghdad		XIII	60		6.25	seal		herringbone
6-0386	37-16-336		XIII	60b			celt		
6-0372		4284	XIII	6Q			ram figurine	i2	
6-0393	37-16-237	4297	XIII	6Q			hoof pot	i7	
6-0452				6Q			pendant		
6-none		4536	XIII	6Q			pendant of gazelle	jl	
6-0491	37-16-233			6S			painted pot		
			no				association		
7-0293	38-13-55	2167	XIII	70		6.45	stone gadget	e/flO	small round base
7-0076	38-13-52	1821	XIIA	6E			jar		cord impressed
6-0466	37-16-266			7			painted bowl		
6-0471	37-16-228			?			painted cup		
31
Subartu XXIII
Trash
Level VIIIA
O
^    Level XI/XA
* set in bitumen floor
nictate and phallus shaped grinding stone
5 in
seal sealing
censer, jar stand, brazier
t0>S
jewel ry, ^ adornment
Q horns
^    hut stalues
grinders, mtillers
stored grain or vegetable
{3    small jars «32» ballista
(^™*^ serving wares celt, adze
._ cooking pot, |
(jjj> sieves, strainers     N     heavy knives
Wide Flower
spheres, weights metal
inaceheads I
J^mm^   animal bom
Fig. I: Tcpe Gawra Temples
Level XI/XA: Rothman 2002a Level X: Rothman 2002a Level IX: Rolhman 2002a Level VIIIA Rothman 2002a. 5.M
32
A propos cle Tepe Gawra
White Temple Unik Warka LC5 Late Uruk period
* scale unclear
Fig. 2: Ubaid, Southern, and Northern Temples and non-Temples
Gawra XIII, Toblcr 1950. XII
White Temple, Uruk-Warka: after Frankfort 1970. 4
Eridu Temple VII: Margucron 1986. j and k
Tell Brak Eye Temple: after Mallowan 1965. 34
"West Temple." Tcpc Gawra VI11C: Rothman 2002a, 5.73
33
Subariu XXIII
Fig. 3: Seals and Sealings
a. Pittman2001, 11.3, b
b. Pittman2001, 11.2, d
c. Rothman 2002a, plate 28, 477
d. Pittman2001, 11.4, a
e. Pittman 2001, 11.4, h
f. Rothman 2002a, plate 40, 1005
g. Rothman 2002a, plate 36, 1061
h. Pittman 2001, 11.6, f
i. Rothman 2002a, plate 43, 1578 j. Pittman 2001, 1.5, g
k. Rothman 2002a, plate 51, 2033 1. Rothman 2002a, plate 54, 2001 m. Rothman 2002a, plate 52. 2023 n. Rothman 2002a, plate 7, 322 o. Rothman 2002a, plate 20, 2229 p. Rothman 2002a, plate 18, 1949 q. Pittman 2001, 11.7, b r. Pittman 2001, 11.15, a s. Pittman 2001. 11.23. c
34
A propos de Tepe Gawra
Subartu XXIII
37
Subartu XXIII
38
A propos de Tepe Gawra
39