An obvious problem of today's global system is the lack of political control over the globalized economy. While identifying it as a problem, we certainly also realize that this lack of political control is exactly what has made the international economy so dynamic for thousands of years. The economy operates at the highest hierarchical level of the global system, knowing no boundaries, while political control is restricted to the lower levels. Attempts at establishing political order at the global level have not, so far, been very successful. The UN is composed of states with internal autonomy and has, therefore, very limited power for intervening.

It is relatively easy to distinguish the structure of a global system from above. It is composed of center-periphery relationships thereby embracing a number of societies, in different positions and of different types. A continuous evolution of the global system has taken place during the last 5,500 years. At the same time, decentralizations of industrial production and capital accumulation have produced recurrent regional shifts, accompanied by local collapses. As we usually study situations where the global system is already at work we are able to take certain processes and mechanisms as given. I shall in this chapter go back to the very early process of social evolution in southern Mesopotamia where we can follow the primary evolution of the state within the framework of an emerging global system. Here it is possible to study the transition from 'non-coercive' to 'coercive' power (cf. Clastres 1974), the establishment and transformation of hierarchical levels, the use and extension of rhizomes (structures that spread horizontally between political units), and perhaps the primary separation of economy and politics as well.

The systems theory model presented by Barel (1973) is a useful point of departure. A social system is conceived as a hierarchy of levels in which some form of exchange must take place between them in order for the hierarchy to be maintained. Levels must be open, but only partially so, in order to maintain its specificity.

il n'y a pas de hierarchie sans echanges entre les niveaux hierarchiques et ces echanges ne peuvent intervenir que si les niveaux sont partiellement

[153]

ouverts. Inversement, un niveau ou un systeme ne peut etre completement ouvert, sans perdre sa specificite, c'est-a-dire sa qualite de niveau ou de systeme. ('there is no hierarchy without'exchanges between the hierarchical levels, and these exchanges can only take place when the levels are partially open. Conversely, a level or a system cannot be entirely open without losing its specificity, that is, its quality as a level or a system.) (Barel 1973:168)

This also suggests that the point where output/input takes place can be controlled and monopolized. Hierarchization is a reversible process. An external/ horizontal relation can be transformed into an internal/vertical relation. There can also be dissolution and a return to the former pattern.

Before trying to apply this model to the process of social evolution in Mesopotamia during the period circa 4500-2000 BCE, I shall illustrate it with reference to Central African kingdoms of the Kongo type (Ekholm-Friedman 1972, 1977, 1985).

The Kongo system is a useful starting point as it represents the simplest form of a hierarchical, social system as represented by Barel. At the top of the pyramid there was a king with monopoly over external trade. Under him we find a political hierarchy of chiefs. The village chief gave tribute to the district chief, who in turn gave to the provincial governor, who gave to the king. In the opposite direction products were transferred from the top down, which makes tribute rather look like an exchange between territorial levels. Before the arrival of the Europeans at the end of the fifteenth century, there existed external exchange between social systems of this type, the nature of which we know almost nothing about since external trade after the contact was immediately redirected toward the Europeans. Central African kings realized the importance of their monopoly over external trade for their own superordinate position, and in the early literature we can see them worrying about the fact that the Europeans refused to respect it. When the Europeans bypassed the central power and began to trade directly with chiefs at lower territorial levels, the kingdoms accordingly disintegrated.

The case of West Central Africa illustrates that external/horizontal relations ('trade') are easily transformed into internal/vertical ('tribute'). Large kingdoms initially expanded when the inflow from the top increased due to the new trade with the Europeans. It was easy for a chief to accept the status as client. It was to his own advantage because of the transfer of resources to him from above. This strengthened his internal authority. Every local unit, at every territorial level, maintained its autonomy. The kingdoms themselves were composed of long chains of interrelated local groups, or units, with a minimum of hierarchical press, from higher to lower levels. In this type of social system the influence from the higher level is so weak it does not even look like a hierarchy or 'vertical structure.'

In a Kongo system the hierarchy is composed of kinship-organized basic units (in principle isomorphic) plus the network that binds them together. In this way the hierarchical levels are nothing but the chain of superordinate and subordinate basic units. Their structural interrelation is, in the folk model, represented as

[154]

segmentation. In the beginning, says one of the origin myths, a group of people settled in the capital. The rest of the country was empty. The original inhabitants multiplied, and as a consequence of local overpopulation there was an out-migration, or decentralization, of 'sons' to the provinces, and from there to the districts, and so on (Ekholm-Friedman 1972). It is unlikely that the kingdom of Kongo originated 'from above.' The king came last and not first. Local groups were already there, implying the emergence of a hierarchical network via matri-limality, prestige goods and external exchange.

A resourceful group took wives, by bridewealth in the form of prestige goods, from client groups which thereby were transformed into 'son-groups' through avuncolocal residence. At a certain age, sons were transferred to their mother's brothers in the wife-giving groups (Ekholm-Friedman 1972, 1977). This created homogeneity with respect to language and culture. However, no aristocracy appeared. There were tendencies in this direction, as sixteenth and seventeenth century documents note that the king sent his own men to the provinces as governors. This may be interpreted as an embryo of class distinction. The process was, however, interrupted when the area was increasingly involved in slave trade and the larger political units disintegrated.

The pyramid is, in fact, nothing but a geographical surface, and its base is an open field for expansion from above. It would be more correctly represented if it was collapsed so the top is placed at the same level as the base. In this way one sees that 'hierarchical' relations rather imply horizontal alliance. The relationship between super- and subordinate group is strikingly egalitarian. The lower levels maintain much of their internal autonomy. As noted, the relationship implies exchange and resembles trade much more than tribute. The provincial governors even received more prestige goods at the king's court than he brought with him when he handed over the so-called tribute. It is in the very logic of the system that all chiefs are interested in finding their own patron, e.g. attaching diemselves to someone with even greater resources. In doing so their own power is not undermined, but strengthened both internally and relative to their subordinates in the hierarchy.

From outside it might be reasonable to question the pyramidal form, but it is unambiguously supported in the cosmology. According to precolonial religion, life-force must come to the individual from above, e.g. outside. At the highest level and furthest back in time, at maximum distance from man, there is God - as Ancestor or as a kind of Big Bang. At lower levels, closer to man, there are ancestors of various dignities, and at the point where the divine world enters the earthly world, the king is situated, God's representative on earth. Between the king and Ego there is finally a hierarchy of political/territorial chiefs, and at the bottom, Father just above Ego (Ekholm-Friedman 1991:145-57).

The word 'Mesopotamia' means 'between the rivers' and refers to the land between the rivers Euphrates and Tigris, the main part of it is in today's Iraq.

[155]

The alluvial plain of southern Mesopotamia is divided into a southern-most part, known as Sumer in the third millennium, and Akkad in its northern part. Northern Mesopotamia consists of the lands east and west of the Tigris further to the north. The time interval covere of here is from about 5000 to 2200 BCE, when a major crisis occurred in the area. This period is usually divided as follows:

Ubaid (4500-3500 BCE)

Uruk (3500-3000 BCE)

Early Dynastic (3000-2350 BCE)

The Akkadian empire (2350-2200 BCE)

The first two periods have traditionally been denned according to their pottery, making the temporal distinction irrelevant historically. There is a clear cultural continuity cross-cutting the periods, as well as structural breaks that do not coincide with this traditional division.

The process of social evolution in the area was by no means linear. There were disturbances at the end of Uruk, a major crisis occurred about 2200 BCE that affected a large region, including the eastern Mediterranean, Egypt and the Indus valley. In Mesopotamia the latter crisis resulted in an interruption of civilized life, where hardly any form of civil disorder was absent (Bell 1971:7). After a short period of 'anarchy and confusion' the area was invaded by a 'barbarian tribe from the Eastern mountains' (Westenholz 1979:113). A reorganization then took place about 2050 BCE with the establishment of The Third Dynasty of Ur.

Since writing did not appear until the Late Uruk period, our information about the Ubaid and most of the "UAlk periods is limited to archaeological material. Archaeologists emphasize that the conditions for prehistoric investigations in the soudi are unfavorable due to heavy silting and a shifting pattern of lagoons and river courses (see Oates and Oates 1976:121). The first script was pictographic, on clay tablets, concerning mostly economic matters. By the very end of the fourth millennium pictograms also took on the sound value of words, an invention that made possible the identification of the language as Sumerian. From this point in time the area is called Sumer and its inhabitants Sumerians. The earliest texts come from the city of Uruk about 3100 BCE (Nissen 1986). The clear cultural continuity between Ubaid and Uruk (see Adams 1981:59) has been taken as an indication of the population being Sumerians even before Uruk. A frequendy used example of this continuity is the clear resemblance between the temples at Eridu during the Ubaid and the Uruk periods. The Uruk temple is larger and more elaborate, but its general design is similar to the earlier one. The gradually higher temple-base in the later period was a result of building the new temple on the top of the old one.

Around 2350 BCE, Sumer was conquered by the Semitic-speaking norm, under Sargon of Akkad, and was included in a short-lived empire. During the Akkadian period, royal officials used their own language to the exclusion of

[156]

Sumerian (Larsen 1979:78). This state of affairs was to some extent reversed in later Ur, when Sumerian was again the official language (Oates 1979:43). Around 1800 BCE Sumerian ceased to be spoken at all (Cooper 1973).

There are certain distinctive features, mainly due to ecological conditions, that remained more or less the same diroughout. The southern part of the alluvium is a lowland with a nearly total lack of sufficient rainfall to ensure agricultural production. It was consistendy dependent on irrigation for agriculture, a technique that was certainly introduced from outside. Irrigation farming existed in central Mesopotamia as early as 5500 BCE. The first farming settlements in the south appeared about 5000 BCE. With irrigation the river valleys provided exceptionally fertile agricultural land, enabling permanent settlement and a relatively high population density. Yields were high in relation to both land and labor inputs (Adams 1981:243). But irrigation was a tricky business. The rivers sometimes caused flooding, irrigation works could easily be damaged, they needed constant maintenance (Oates and Oates 1976:125), and the technique itself caused problems of salinization. Barley (Hordeum vulgare) and einkorn wheat (Triticum monococcum) were cultivated, and the later preference for barley can be explained by its greater resistance to salt (Algaze 1993:106; Huot 1989:26).

The presence of lagoons and marshes provided a rich and varied environment with an abundance of fish and birds. Marsh reeds were utilized as building material for huts and granaries, and the river clay for sickles, pots, plates, and jars. There is early evidence of date palm, 'without doubt one of the most useful plants known to man' (Oates and Oates 1976:121), flax, tamarisk and poplar (Hout 1989:26). Uruk texts testify to the importance of both sheep and goats, while catde and pigs seem to have been of more importance during the Ubaid period (Hout 1989:27).

A frequendy cited feature of the area is its lack of crucial raw materials. It had no metals, timber or stone. It was completely dependent on imports for many of its 'civilized' products. Its bronze metallurgy needed copper and tin and maybe even wood for fuel. The temples were mainly built of mud bricks, i.e. of local material, but to some extent of imported stone. Stone was also needed for a number of implements and for specific items, such as sculptures and cylinder seals (Larsen 1979:76).

In Sumer the monumental buildings were temples and associated administrative structures, not royal graves as in Egypt. The land in its entirety was conceived as belonging to a specific god, and the temple was his house. As the temples were usually built of mud-brick and not of more time-resistant stone, the remains bear, unfortunately, no resemblance to what was there in the past.

The location of the area within the larger region was certainly of great importance for its development. The two rivers, of which the Euphrates was the more navigable, were decisive communication links. Sailboats appeared at a very early time. Knapp (1988:45) notes that 'Rivers . . . and their tributaries served as trade and transport routes ... Besides carrying imported commodities, the rivers provided for the movement of people, redistribution of goods and supplies, and

[157]

transport of military contingents. 'During early Ubaid, trade was mainly carried on with the north, and the area was thereby in the position as a somewhat disadvantaged hinterland. However, when trade contacts developed with central Iran and via the Persian Gulf (cf. Wright 1969:103), the area found itself in a-central position within a wider regional exchange network.

When trying to understand the process of change from Ubaid until the crisis of 2200, it is essential to take both vertical and horizontal structures into account. A global system (Algaze 1993) emerged by the later half of the Uruk period, in which southern Mesopotamia occupied a position as core area. The Uruk system expanded geographically by colonization and by the establishment of control over important trade routes. Over time this global system underwent fundamental changes, as all global systems do, in both its horizontal structures and its core area. It is essential, even if it is not at all easy, to study how local/vertical processes are related to horizontal processes within the global system. The main difficulty lies in applying a holistic view, treating what happens as a unified process, in spite of the cognitive necessity of breaking it up into two different perspectives, which to some extent must be studied separately.

I shall in this section try to demonstrate that a structural change of crucial importance, the transition from a one-level system to a two-level system, occurred during Ubaid. This change provided the very basis for the cultural development that followed during Uruk.

The Ubaid period is, in both northern and southern Mesopotamia, characterized by a homogeneous settlement pattern with small villages or hamlets scattered throughout the country. The sedentary communities were 'widely and fairly evenly dispersed sites' (Adams 1981:58; for the same pattern in the north, see Akkermans 1989:341, 349). The various Ubaid settlements also exhibit a 'remarkably homogeneous material culture' (ibid.). The resemblance, including ceramics, burial practices and certain architectural features, suggest that close contacts existed among them. The small and dispersed sites found in northern and southern Mesopotamia were, in other words, linked to one another by networks of exchange. Among the widespread exchange items of the earliest period are obsidian, flint and bitumen. The south was the more marginalized hinterland within the regional exchange network.

The change came, however, in the south. Knapp (1988:42) notes that 'During the Ubaid period, modest farming villages began to grow into large population centers, and the temple or temple complex originated.' Some sites grew considerably larger. Even if most of them were still small as before, some of them were more than ten ha (Adams and Nissen 1972). The other significant feature mentioned in the quotation above is the temple.

The first step in the process of change was the appearance of a new form of centralization. Instead of each local kin group managing its own businesses, a number of different kin groups, probably fishermen as well as cultivators, established a

[158]

higher unity, the function of which was cooperation and coordination to the benefit of all. This structural novelty, which emerged during the later part of Ubaid and was manifested in the appearance of the temple, showed a great potential for social evolution.

The early Ubaid social system consists of two structural elements; small local units plus an egalitarian exchange network linking them to one another. I shall call this structural form a one-level system in order to distinguish it from the two-level system that appeared later.

The earliest Ubaid houses, mostly found in the north, are large-scale buildings where no distinction was made between domestic and sacred. At Tepe Gawra the earliest (tripartite) buildings have ovens or bins and were, as noted by Akkermans, 'intended in the first instance for living' (1989:343). There were other types of buildings as well, seemingly dedicated to serve special functions, as storage or stable. The same pattern has been reported from other excavated sites in central Iraq, e.g. Tell Abada (Jasim 1989) and Tell es-Sawwan (Margueron 1989).

The reason we know less about houses in the south is that domestic houses were constructed of perishable reed (Hole 1989:167). Ecological changes in this area have rendered prehistoric investigations exceptionally complicated. At Oueili, near Larsa, there is, however, evidence from early Ubaid of the same kind of large-scale construction found in the north. The houses exceed 200 square meters and have a 'complex multicellular plan' with evidence of communal functions (Huot 1989:32).

Some of the buildings from this period are rectangular with buttresses while others are of tripartite or bipartite design. The character of the houses reflects an extended household economy, centralized and coordinated but still within the borders of the local kin group, in no need of the kind of management and control that came later. The vast granaries (eighty square meters) found at Oueili 'suggests perhaps communal storage,' Hout says, and adds 'But this storage seems to function without any particular means of management or control' (1989:39). This is centralization, in the form of 'communal storage,' but without the type of administration we associate with the later temple complex.

According to Akkermans (1989:349) all the settlements in northern Mesopotamia were rather small, most of them with only one or two houses. The largest, such as Tepe Gawra and Tell Adaba were not more than one ha, which would include fifty to one hundred inhabitants. Thus, in the earlier period each local unit was economically more or less autonomous, there was very little specialization, it had the form of an extended household, and the buildings had communal functions of various kinds besides being places to live.

It is worth noting that centralization existed before the temple complex. Mesopotamian agriculture, be it irrigated or not, necessitated storage and thereby a more extensive cooperation than what is found, for example, in modern West

[159]

Central African cultivation where each producer prefers to have his/her own fields controlling his/her own product. The 'communal storage' during early Ubaid may be seen as an embryo to the later higher unity.

True temples did not appear until the later part of Ubaid, earlier in the south than in the north. In both areas they were built according to the tripartite plan that evidently originated in the north where it was reserved exclusively for temples. In the south the earliest temple is the one at Eridu, which replaced an earlier structure consisting of a one-roomed rectangular building with interior buttresses. Opinions diner on whether the earlier structure was a true temple, or whether there was a break in the building tradition (Akkermans 1989:344f). It could as well has been a house of the type found at Oueili.

The new social system that emerged was composed of two hierarchical levels, a lower level of local kin groups, and a higher level represented by the temple. The difference between the two systems may not seem very significant. Earlier, centralization and coordination among nuclear families belonging to one single local kin group, and now, centralization and coordination among several kin groups. But the change is crucial. As long as the point of centralization coincides with the power structure of the social group, it just underlines or preserves this structure, curbing the forces of change. What eventually alters the actual power structure is when a point of centralization is established 'outside' of the existing social groups.

The identified sites grew larger, and it is important to note that they only contained the sedentary part of the population. We should, with Adams, visualize around each -center 'smaller, less sedentary groups who depended primarily on their herds or on fishing"('1981:59). There is no real evidence of these groups in the archaeological material. But we should also keep in mind that cattle and pigs predominated in Ubaid, not sheep and goats that are usually associated with herding. Herding became an important branch of the Sumerian economy only later. There are, in other words, reasons to question the role played by herdsmen in late Ubaid society. We know from the archaeological material in the north that cattle and pigs were kept in special buildings. In one such building at Abada in early Ubaid, a bitumen-lined basin has been found which was probably used to water domestic animals (Jasim 1989:83).

The new form of centralization was expressed in religious terms. As Gauchet (1977:33) demonstrated with material from contemporary primitive societies, humans tend to represent higher organizational units in religious terms, as an external point of reference. In Ubaid the political initiative must have come from below, from the power-holders at the local level, since the higher level originally was 'empty.' Their fusion in the form of a common religious center implies an alliance, aiming primarily at the maintenance of their own power and dominance. The relationship among the various subunits was egalitarian, which can be deduced from the lack of archaeological evidence for social stratification. David and Joan Oates relate the increasing settlement size of the period with 'an increas-

[160]

ing need for some form of centralized control' and then remark that there is, surprisingly enough, yet no evidence for social stratification (1976:124f).

During the entire course of social evolution there are hierarchical levels in the form of a higher unity, devoid of any real political power, constituting more of an alliance between political power-holders at a lower level. In our contemporary world, the UN is a clear example of this type of higher unity. In order for a hierarchical level to take on political power, it must usually either control external exchange or be opposed to an enemy of the same dignity, or both. The world of today cannot be united politically since none of these mechanisms prevail. The only condition that could alter the situation is an encounter with aliens from outer space, cooperating with the global elites, or attacking the earth.

Social evolution often seems to take place in two steps with respect to hierarchical levels. First a higher unity in the form of a symbolic space is created, and then this 'empty' space can be filled with economic and political content. The process of change in southern Mesopotamia encompasses an early transformation of this type. From being nothing but a higher unity in the Ubaid period, this level becomes a real power center in the Uruk period. Interestingly enough, the higher unity does not thereby disappear. It reappears during both Uruk (Diakonoff 1973:186) and Early Dynastic, now among political units of a different kind, still as an elite strategy for maintaining power.

The temple originally represented cooperation and coordination among the various subunits in projects oriented both internally and externally. It played a decisive role in both external and internal exchange. In fact, these phenomena are aspects of one and the same system and must consequently be understood in the light of one another. We know that a point of centralization is easily transformed into something more substantial. When resources are channeled via a higher hierarchical level, it often leads to an elaboration within this special space. A rather peculiar example of this mechanism is found in modern African states where the inflow of resources from outside via the point of centralization has led to the emergence of super-wealthy state classes disconnected from the rest of the population. In ancient times the evolutionary potential of this structure was considerable. When a point of centralization was established 'above' existing social groups, it was beyond their reach and could gain enough autonomy to develop on its own. It then transformed the former power structure, giving rise to a new type of society. The possibilities lie in its capacity to use external flows for internal development.

It is usually argued that the temple originally had only ceremonial functions. We know, however, that the temple was also a center for the collection and distribution of agricultural products, and that it was directly involved in food production. Waines (1987) points to the symbolic importance of bread as the distinctive feature of culture and civilization in medieval Iraq. The temples produced food 'for the gods whose care and feeding was a matter of daily concern and among the

[161]

items deemed 'fit for the gods' was bread, baked in the temple ovens as special prayers were uttered at each stage of its preparation' (ibid.:259; ref. to Oppenheim 1954:191). The large quantities of fishbone found in the early temple at Eridu are interpreted as offerings (Oates 1979:124), but also as exchange. They point, according to Adams (1966:50) to 'a very early beginning of ritualized patterns of either offerings or exchange (emphasis added) in which at least the products of the specialized group of fishermen were made available to a considerably wider segment of the population.' There is no reason why we should expect a clear demarcation line between offerings to the gods and exchange among those involved. Cereal and fish constitute two of the most important export articles in later periods.

The temple was the god's home. It is easy to interpret the fact that it did not appear until later Ubaid as an indication of a true religious innovation, founded in a structural innovation. The city 'belonged' to the god in the sense that 'the focus of loyalty was the city,' says Joan Oates, then quoting Noah Kramer: 'People were identified as citizens of this or that city, and not with a clan or some other kin-related group' (1977:474). A god who owns the city does not belong to a kinship-organized society, as the one that preceded late Ubaid. In these types of societies the gods usually appear as ancestors, related to the living by blood and descent, in the same manner as the living are related among themselves. Here it is place and not descent that unites them. The god who owns the city belongs to a society where a higher unity is established among different, non-related groups of people. It is worth noting that no collections of fishbones (or anything similar) have been reported from the earlier Ubaid houses. Other categories of people were simply not involved.

When it comes to the political organization of late Ubaid, the 'primitive democracy' model suggested by Jacobsen (1957) is quite convincing, even though it is based on mythical material and can only be 'observed' with any certainty in later periods. According to this model there was a general assembly (unken, or 'circle of the people'), which in this context probably refers to all the free, adult men. This word occurs in the earliest texts (Oates and Oates 1976:135). There was also a more restricted group of elders, a 'council of elders.' The assembly elected among themselves an en ('a lord'), as a kind of chairman with administrative functions. It probably also elected a war-leader in periods of external conflict (Jacobsen 1957:103). Such a 'chairman of the assembly' is mentioned in the Uruk tablets circa 3100 BCE (Nissen 1986:328).

Jacobsen (1957:104) emphasizes the provisional nature of the assembly:

Viewed as a whole the most characteristic element of the Primitive Democracy pattern is probably its provisional and ad-hoc character. It is called upon to function in emergencies only ... The assembly deals only with specific crisis for which it was called . . . When that emergency had passed we must assume that the larger unity temporarily imposed on the community vanished with it and left the ordering of society to the numerous minor ... power structures.

[162]

The higher unity existed so far only as a form of cooperation and coordination among egalitarian kin groups for their common benefit. Their cooperation made them stronger than they would have been individually, externally as well as internally. The existence of a general assembly may be taken as an indication of rather egalitarian relationships within the kin groups and among free, adult men, but to the exclusion of women and unfree individuals. Initially the initiatives were entirely concentrated at the lower level and the higher unity seemed unequivocally to serve their common interests. However, in creating this form of centralization they let loose forces of change that soon were out of their control.

During Uruk an amplification of the higher level took place, transforming it from an 'empty' higher unity into a power center, with a dominant class and its workers, and a surrounding community sector. Among the first expressions of this process are concentrations of high status goods within the temples (Knapp 1988:43). The temple grew increasingly larger and it took on both economic and political functions (Adams 1966:125). It became an encompassing organization where no distinction was made between the sacred and the secular or between political and economic activities (cf. Knapp 1988:69). It was a temple/state controlled economy, similar to modern socialism even if not so dominating. The state thus developed within the religious domain. The head of the temple had both religious and political authority. The separation between temple and palace, between 'church' and 'state,' came later, as did the separation between economy and polity.

The temple was now active and creative in a number of fields. It introduced and developed various crafts, organized agricultural production, specialized production in workshops, was in charge of storage and redistribution, and charted long-distance trading expeditions. It also managed irrigation and construction works. All the spectacular innovations of this era took place at the temple, including writing, keeping of accounts, plough, wheel, wheel-made pottery, cylinder seals, metallurgy and stone sculptures. The temple/state elite grew in size and power, usurping the functions of the lower level.

A new type of political unit emerged. A dramatic population increase occurred in both Sumer and the Susiana plain (Khuzistan) to the east, in today's Iran, which at the time was closely related to Sumer. This increase was mostly a result of the inflow of people from surrounding areas. Characteristic of the Uruk period was the clustered settlement pattern that diverged from the evenly dispersed sites in Ubaid. There is a gradual transformation from one to the other (Adams and Nissen 1972:11). Algaze describes Uruk as 'a small number of centralized cores in fierce competition,' even suggesting that the high level of conflict compelled them to form these enclaves (1993:115). The decentralized political organization of the area allowed the major sites to expand economically, giving rise to increasing competition, and likely confrontations, which eventually encouraged larger agglomerations. This picture belongs, however, to late Uruk when the general

[163]

development in the area had put the various centers in 'fierce competition' with one another.

The political unit was now not only much larger but also of a very different form than the former two-level system with its 'empty' higher unity. The Uruk social system, found in both Sumer and the Susiana plain (seeJohnson 1975), is described as a three- or four-tiered hierarchy: cities, smaller towns, villages and hamlets.

The internal structure of the Uruk system has been efficiently analyzed by John-son and Wright in a series of works (Johnson 1975; Wright and Johnson 1975:279). Their term 'local exchange' is very useful as it points to the intimate relationship between internal development and external trade. Johnson (1975:285) says explicitly that local exchange and trade are 'complementary processes' as 'long-range trade provides economic links between more or less independent settlement systems, local exchange provides similar linkage within individual systems.' This view of the combination of the two types of exchange reveals the rhizomic nature of what I have called 'vertical structure,' and offers a theoretical point of departure for the understanding of the relationship between a country's domestic economy and its export production. If the two structures are seen as complementary we may not fall into the trap of conceiving export production of minor importance just because it represents a small part of the economy as a whole. This is valid for societies in general, not just this particular area in the fourth millennium.

Johnson and Wright discuss this in terms of control hierarchy, or a hierarchy of information processing, arid- here they are close to the systems theory model suggested by Barel discussed earlier. The development at the temple level led to a pronounced differentiation between higher and lower levels, referring not only to the relationship between the city and the rural sector, but also to the relationship between the temple and an outer community sector within cities themselves. We must keep in mind that most of the townspeople were peasants. At the end of Uruk there was a rapid development of various types of craft production in temple workshops. In the previous period craft production was a dispersed activity. Johnson's consideration of the Susiana plain shows that craft production was now concentrated in workshops in the major setdements. We see a shift in ceramic production from 'small scattered shops to larger centralized shops' (Wright and Johnson 1975:279). Uruk ceramics are wheel-made, a technique that was invented in this area and in this period. Wheel-made ceramics were mass-produced and of superior artistic quality. Ceramic kilns have been found in the major settlements in Susiana (Middle Uruk), while 'diere is no clear evidence of Uruk ceramic production at any omer site in the area' Johnson 1975:92). Something was evidently produced centrally and then distributed in mass-produced ceramic containers. Uruk pottery has been found over a very large area, an indication of Sumer's mass-production of export goods in dlis period. There was also a flow of other

[164]

products (e.g. lithic) from large central workshops in Susiana to rural settlements (Johnson 1975:109,112).

In this process, a growing differentiation took place between the temple as such and the rest of me society, the community sector. The latter became increasingly dependent on me temple's goods and services. In exchange its members gave of their produce and labor. This new type of relationship is described as redistribution, exchange or symbiosis.

The Archaic texts of Uruk (circa 3100), the first of a series of texts, derive probably from a single large economic unit within the precinct of Eanna dedicated to the temple of the goddess Inanna (Nissen 1986:324). These texts constitute our main written source of knowledge about late Uruk society and economy. The somewhat later texts from Ur belong to the very end of Uruk or the beginning of the Early Dynastic period and can be considered as well. The texts reveal a strikingly hierarchical structure of various activities, which certainly contradicts the idea that late Uruk society would have maintained its earlier, more egalitarian spirit. In his analysis of the Uruk texts, Nissen (1986:329-30) claims that the sign NAAI stands for 'the leader of the unit of a given activity. If interpreted this way, there were leaders of 'the city,' 'law,' 'troops,' 'plow,' etc. The texts ofUr contain a picture of the military organization; 'a list of soldiers under sergeants (ugula) formed into a company (un-sir-ra) under colonels (nu-banda) one of which seems to be in supreme command' (Jacobsen 1957:107ff). There is also information on high-ranking officials, such as 'the chairman of the assembly' and different grades of priests.

Agriculture as well as fishing played a crucial role in the Uruk economy, in the form of export production to be sure, but also for internal consumption. Tyumenev (1973a:72) points to the fact that the earliest signs indicate bread-baking and brewing, which seems quite reasonable given me temple as the house of a god mat needed constant feeding. The early texts also show herdsmen, hunting scenes, wild birds, and an 'abundance of pictographs of various fish testifies to the considerable role of fishing in the economy of the temple.'

As the temple complex grew larger, an increasing number of workers were employed cultivating land, in irrigation and construction, and in workshop activities. All these workers had to be fed. In one of the Uruk texts daily rations are listed of bread and beer for about fifty individuals. Another part mentions barley and fish. There is also archaeological evidence of these rations in the form of a widely dispersed, mass-produced container called 'beveled-rim bowl' that is supposed to have been a ration measure (Gates and Oates 1976:1291). Sumer could never have been a major exporter of food without advanced technology and craft production, in other words without its position as center in a global system. It has always needed the most advanced technology. In ancient times it also needed 'armies of laborers.'

Every peasant certainly owned his own tool, such as a spade, hoe and flint sickle (Crawford 1991/1997:45). But the more complicated means of production, such as plows, seed funnel and draught animals (oxen) were provided by the temple (Crawford 1991/1997:44). In the texts of Ur a 'House of the Ploughs' is

[165]

mentioned (Tyumenev 1973b:74). A number of specialists are listed, such as carpenters, gardeners, cooks, bakers, coppersmiths, jewelers and potters (Nissen 1986:329; Tyumenev 1973b:74). Some of the crafts were organized in groups under the direction of a master craftsman. Copper production was carried out in Uruk, and there was a rapid increase in the amount of metal used at the end of the fourth millennium (Crawford 1991/1997:131). Harder alloys did not appear until the Early Dynastic period.

A large part of the Uruk texts deals with textiles (Nissen 1986:330). We know that early and throughout its history, Sumer was a great exporter of textiles. We have archaeological evidence for its trade with surrounding areas. There is, however, very little information in the Uruk texts referring to these foreign areas. When talking about Sumer's exports it must be kept in mind that there is no clear archaeological evidence and this lack of evidence is commonly interpreted as an indication of perishable export products as well as of perishable processing equipment (Crawford 1973:232). A way of trying to circumvent the problem is to look at later periods, assuming that Sumer exported about the same articles earlier. In this manner Algaze (1993:4) identifies the main export articles as food, such as cereal, dried and salted fish, oil, dates, and industrial products, such as textiles, leather work and in later times items of metal. Leemans (1960:114) emphasizes Mesopotamia's constant role as a producer of agricultural products, besides the production of 'industrial articles of relatively high value such as garments or articles of fine craftmanship' (Leemans 1960:115). This would be valid from Uruk until Ur III:

Southern Mesopotamia itself was an agricultural and cattle-raising land which produced almost nothing else ... As materials of exchange ... the soil of southern Mesopotamia could only offer its agricultural produce such as barley, dates and sesame, or the products of cattle-raising such as butter, cheese and leather. (Leemans 1960:115)

These articles were exported in large quantities by boat north along the Euphrates and south via the Persian Gulf.

Who were the workers? There were a growing number of slaves in the economically expansive centers in the south. The population increase in the south was initially accompanied by depopulation and the abandonment of sites in surrounding areas, both in the north (Akkermans 1989:347) and to the east, (Lamborg-Karlovsky and Beale 1986:267). The temples seem to have held large numbers of slave-women, gin, as workers. In one of the Uruk tablets, 211 female slaves are recorded, denoted by an ideograph that is usually interpreted as 'woman from foreign mountainous country.' They were used in textile production and household activities. Female slaves are found in industrial textile production all over the Near East and eastern Mediterranean in ancient times. According to Tyumenev (1973b:73) there is no equivalent ideograph in the earliest tablets for male slaves, which, of course, does not necessarily mean that they did not exist.

[166]

Male slaves appear alongside female slaves on stone inscriptions from the end of Uruk. Chattel slaves in the form of war captives became increasingly important in the political economy of Mesopotamia in later periods (Gelb 1973), especially in large-scale irrigation and construction work. The temples could not have been built without 'armies of laborers.' It is estimated that the labor force required for the sub-structure of the Anu Ziggurat at Uruk was 7,500 man-years (Adams 1966:126). Among the workers were certainly also free and highly competent specialists, some of them from surrounding areas (cf. Kramer 1963:101). It seems probable that some of the foreign workers had to be bought or captured by force, while others may have come of their own accord.

'(I)t is towards the end of the "Uruk period" that the momentous change we have described takes place, 'says Frankfort (1971:73), when discussing why it is not always useful to divide the process into Ubaid, Uruk and Early Dynastic. The most spectacular development thus seems to have taken place in late Uruk. But late Uruk embraced both an early phase of rapid development and a later phase characterized by contraction and crisis. The serious problems toward the end of Uruk appeared when Sumer's life-supporting trade network was shattered. The very end of Uruk is characterized by internal warfare and by the destruction and abandonment of sites.

War evidently played a significant role during Uruk. There are cylinder seals depicting battle scenes and fettered war captives (see Postgatc 1994:241, 25; Adams 1981:63). But it is worth noting that no striking development took place in arms technology during Uruk. This came later, in the Early Dynastic period. The Uruk system was, as mentioned above, characterized by the autonomy of each major center. Even if trade and other forms of communication to some extent were effectuated between neighboring centers, they were essentially independent of one another. This means that each of them had the opportunity to encounter the surrounding world on its own, a situation that encouraged both external trade and warfare. There was a high degree of internal warfare in Susiana in late Uruk while the wars of the earlier period seem to have been primarily directed toward foreign areas Johnson 1975:157). This pattern could be explained with the contraction of trade in late Uruk, a condition that must have led to aggravated conflicts between the various polities. In southern Mesopotamia, city walls were erected already at the end of Uruk, a clear expression of the fact that 'an epoch of wars had begun' (Diakonoff 1973:186).

External trade and warfare are decisive evolutionary dynamics. The combination of these structures gave rise to a new political strategy in southern Mesopotamia. It emerged in late Ubaid and Uruk, and was then fully developed during the Early Dynastic period. Elites belonging to a number of centers turned outwards, both toward external markets and neighboring competitors. In this process they involved themselves in a new kind of political game that was solely their own. They became the dominant actors while the rest of the population was gradually reduced to mere instruments. The state with its class structure was born. Expansion fueled by the energy generated by interstate conflicts is a mighty force in social evolution, and when conflicts of this type disappear, by the victory of one

[167]

over another, or by the completion of the expansionist project, momentum is usually lost and vast empires may quickly disintegrate. Our experience as humans is somewhat odd on this point. When a satisfactory victory or level of freedom is finally reached, chaos is just around the corner.

This new political strategy made the production of weapons top priority, and from then on we have been caught in a constant and obsessive search for raw materials in order to feed the military complex. With the invention of bronze there was no return.

This chapter is the first half of a longer piece, to appear later. The second part deals with the evolution within the already formed city-state system itself whereas the first part deals with the emergence of this system from an earlier regional dynamic. The notion that the city-states did not emerge in a vacuum is a crucial aspect of the argument. The powerful states of Mesopotamia emerged within a larger regional organization which was the basis of city-state formation. The second part also deals extensively with the history of power relations, the differentiation of class relations, the separation of political and economic relations, the massive transformation of the labor force, the changes in representations of kinship and the emergence of imperial organization within the larger system of interregional exchange. A major aspect of this discussion concerns the spread of accumulative dynamics and productive structures from center to periphery over time, leading ultimately to shifting centers and pulsating empire formation.

[168]