by Eckart Frahm
Emergent Landscapes of the Early Bronze Age
The most prominent feature of Near Eastern cultural landscapes is the mound (variously tell, tepe, or höyük). Mounds are the cumulative result of centuries or even millennia of sedentary inhabitation using predominantly mud brick architecture. In the Neolithic Period, settlements were small and transitory, with occupation generally lasting only a few generations before communities relocated. This pattern lasted until the Ubaid Period, when settlements became more permanent and the resulting settlement mounds began to reach considerable heights (Akkermans and Schwartz 2003: 159–60). Throughout this early phase, communities split before growing demographically large, and, consequently, sites were small. Most experiments in settlement agglomeration appear not to have been durable, such as the extensive settlement at Khirbat al‐Fakhar (Al‐Quntar et al. 2011).
In the Ubaid Period, communities developed a durable spatial mindset on the proper way to settle: in a nucleated form, over a long term, and preferably set atop a pre‐existing mound, whether continuously occupied or not. One factor in this shift must have been economic: the emergence of widely‐recognized rules for land tenure, whether at the household or the community level, to regulate how the settlement’s agricultural and pastoral resources were managed and transferred. Settlement stability was not, however, entirely economically motivated; it is likely that generalized and shared cultural attitudes about settlement had developed. Such attitudes were responsible for individuals and groups choosing to remain on tells, or selecting abandoned ones for the location of new settlements. This general mindset underlaid specific meanings and significance that were attributed to individual places, now lost to us in the absence of written records. The new spatial mentality appears to have been strongest in the late third millennium BCE, when almost all settlement occurred atop tells, and then to have broken down over the course of the second millennium BCE, finally replaced in the Iron Age with a radically different spatial logic (see below).
The pattern of tells changed radically in the middle of the third millennium BCE, when a series of large settlements formed across the northern arc of the Fertile Crescent. This settlement landscape included new forms of land use that left a remarkably deep imprint on the landscape. This process was not the region’s first steps toward urbanism; Tell Brak had already coalesced into a 130‐hectare city by the middle of the fourth millennium (Ur et al. 2007, 2011). But while Tell Brak, and Khirbat al‐Fakhar before it, were isolated phenomena, urbanism in the Early Bronze Age was widespread throughout northern Mesopotamia.
The most prominent elements of this demographic shift were a series of spatially extensive, densely occupied settlements that ranged up to 120 hectares in size. Most of these cities expanded from already ancient tells to include broad lower towns. For example, at Hamoukar, a 15‐hectare tell dated to the fourth millennium was resettled around 2600 BCE, and a 90‐hectare lower town to its south was settled; within this area, 98 hectares were occupied between ca. 2600–2000 BCE (Ur 2010b: 104–9). Similar growth patterns occurred at Tell Mozan, Tell Leilan, Tell al‐Hawa, Tell Taya, Tell Khoshi, and Tell Baqrta, all of which expanded in excess of 60 hectares. Population estimation is a particularly uncertain science (Postgate 1994), but the largest of these cities may have been home to 10,000 to 15,000 persons. Excavations at these sites revealed remarkable concentrations of political and economic power: monumental temple and palace institutions, writing and administrative technologies, craft specialization and mass production, and considerable disparities in status and wealth (recently reviewed in Stein 2004; Ur 2010a; Matney 2012).
In some regions, the urbanization process took place at the expense of settlements in the hinterland. In the Wadi al‐Murr, for example, the urbanization of Tell al‐Hawa could be explained entirely by the abandonment of villages in its hinterland; site numbers were reduced, but the total settled hectares remained roughly constant (Wilkinson and Tucker 1995: 50–3). Elsewhere, the appearance of towns and cities included growth in both site numbers and total settled hectares (for example, around Hamoukar and Tell Beydar; Ur 2010b: 104–9; Ur and Wilkinson 2008: 307–8). In these cases, Early Bronze Age cities probably benefited from immigration or nomadic sedentarization.
The enormous pressures that this urban settlement system placed on its landscape resulted in dramatic transformations visible even today in the archaeological landscape. The necessity of feeding large urban populations placed strains on the traditional dry‐farming based agro‐pastoral economy, with two main effects. Settlements chose to extensify cultivation by bringing more land under the plow. This process can be documented in the landscape via shallow linear features that represent the remains of ancient trackways (Wilkinson 1993; Ur 2003). These tracks are mostly invisible on the ground, but can be mapped using aerial and satellite photographs. They are overwhelmingly associated with sites of the Early Bronze Age urban phase across northern Mesopotamia (Ur and Wilkinson 2008: 310–11). They also occur in northern Iraq, where dating them is complicated by a lack of archaeological surveys (Altaweel 2008: 65–9; Ur et al. 2013). These tracks became depressed as farmers, shepherds, and their animals traveled through cultivated land, where their movements were constrained by fields on either side. Where land was uncultivated, movement was unrestricted, and depressed tracks did not form; hence, the presence of tracks is a proxy indicator for the presence of fields (Wilkinson 1994).
Another response of farmers was to intensify, by introducing nutrients into the soil via manuring. Organic refuse was collected along with other domestic debris and composted in settlements to be spread out upon the fields later. What remains of this practice in the landscape are the incidental bits of inorganic debris, which have been kept in the topsoil by millennia of succeeding agriculture (Wilkinson 2003: 117–18). The landscapes surrounding the cities of Hamoukar (98 ha), Brak/Nagar (70 ha), and Tell al‐Hawa (66 ha) have dense scatters of potsherds, which are the surviving evidence of intensive agriculture in their immediate hinterlands (Wilkinson and Tucker 1995: 19–23, Ur 2010b: 65–76, Ur et al. 2011, Ur 2015).
Together, radial trackways and manure zones describe inner intensive and outer extensive zones of cultivation that reach their greatest extent around the large cities. They are not, however, limited to large cities; smaller radial systems and manure zones are found around towns and even small villages of only a few hectares. In the Beydar region, small villages may have been cultivating at a rate in excess of the needs of their estimated population, and possibly even to an extent greater than the villagers could have undertaken themselves, which raises interesting questions of labor mobility (Ur and Wilkinson 2008: 313–15).
Pastoralism was also important, although it is more difficult to quantify spatially. The increased cultivation of barley around these cities and towns may have been for animal consumption as fodder (Charles and Bogaard 2001: 319). Fodder production would also explain why many settlements appear to have been cultivating far more land than their estimated human populations would have required. The conversion of former pasture areas between settlements into cultivated land therefore may have been offset by an increased emphasis on settlement‐based flocks.
The Early Bronze Age cultural landscape was thus a very full one, modified extensively by human communities. Sedentism and settlement nucleation reached unprecedented levels not to be seen again until the Neo‐Assyrian Period (and, even then, only in a few political capitals; see below). The agro‐pastoral economies of these urban settlement systems operated at high intensity, as farmers brought outlying territory under the plow and attempted to enhance the yields of already‐cultivated lands closer to their settlements.
Despite the intensive and potentially overextended agricultural economy, and the monumentality of settlements and landscapes, we should not assume that the hand of a centralized administration lay behind these developments. Monumental palace and temple complexes did exist, but there is no evidence that they managed or inspired this expansion of
agricultural production, or that they coerced people to nucleate at urban sites. There is no unequivocal evidence, for example, for centralized storage of cereals or animals (Ur and Colantoni 2010). The trackways, over 6000 kilometers of which have been recorded in northeastern Syria alone, are not part of planned communication routes, but rather emerged through the uncoordinated but purposeful actions of farmers, shepherds, and their animals. The motivation for agricultural intensification must be sought at the household level, possibly as new commensal strategies assumed central importance for creating and maintaining social relationships (Ur 2009). The Early Bronze Age urban landscape appears to have been the unplanned result of widespread rules and attitudes about land tenure, household based surplus production, and the social roles of communal meals.
Further evidence comes from the patterns of movement revealed by the preserved trackways. Most simply radiate outward from settlements and fade out beyond the fields, but some connect with trackways radiating from nearby settlements to create networks (Figure 1.2). In no cases were there direct tracks between cities, or between capitals and subsidiary towns (e.g., between Brak and Beydar). Movement through the landscape, even that of political elites, respected local systems of agriculture and land tenure (Sallaberger and Ur 2004).
Figure 1.2 EBA sites and linear trackways in the region of Tell Brak and Tell Beydar (based on Ur 2010b Map 3).
This emergent landscape was potentially unsustainable, however; the combination of high population density, urban nucleation, intensive agriculture, and variable climate placed these settlement systems at high risk of collapse. Agent‐based computer modeling suggests that villages and towns could survive most droughts (Wilkinson et al. 2007: 65–6), but large population centers were especially vulnerable. Large cities could be sustained under normal conditions of climatic variation if their neighboring towns and villages could be convinced or coerced to contribute agricultural surplus, but, in the face of multi‐year droughts, this overextended system was liable to collapse (Wilkinson 1994). Initially, it was proposed that urbanism and political entities had collapsed on account of an abrupt aridification event, variously attributed to volcanoes, meteorites, or global changes in atmospheric circulation (Staubwasser and Weiss 2006; Weiss et al. 1993). This model has been critiqued in recent years in favor of new models that recognize variation in local settlement trajectories (Kuzucuoğlu and Marro 2007; Wossink 2009; Danti 2010; Ur 2015).
Imperial Landscapes of the Neo‐Assyrian Period
By the start of the seventh century BCE, the landscape of northern Mesopotamia had been transformed in ways that would have rendered it unrecognizable to an Early Bronze Age urban dweller. At the most basic level, the settlement landscape of cities, towns, and villages with broad catchments of intensively cultivated fields between them had disappeared. In its place was a nearly even distribution of small villages or hamlets. On the other hand, a handful of cities had grown to tremendous sizes. The walls of Nineveh, for example, could contain seven of the largest Early Bronze Age cities. With the shift towards larger cities, the Assyrian cultural landscape set a pattern that would become typical for the great empires that succeeded it (Wilkinson and Rayne 2010; Adams 2005).
This transition unfolded in the second millennium, after all of the great Early Bronze Age cities were either abandoned (e.g., Leilan and Hamoukar) or substantially transformed (e.g., Tell Brak). The descendants of the former urbanites now migrated with their animals as part of a pastoral lifestyle that is well documented in the Mari tablets (Fleming 2004; Durand 2004) but exceedingly difficult to discern in the archaeological record (Lyonnet 1996). Cities of the Middle Bronze Age in northern Mesopotamia were fewer and uniformly smaller than their Early Bronze Age predecessors. The memory of the earlier cities remained, and many were deliberately resettled and even refortified, but urban populations never regained their former density. At Tell Leilan, for instance, the lower town was largely “hollow” and therefore presented a blank slate upon which royal palaces and other large institutions could be inscribed (Ristvet 2008 fig. 3). Late Bronze Age (Mitanni and Middle Assyrian) cities were also small and infrequent, with the notable exception of Kar‐Tukulti‐Ninurta, a 500 ha planned city that hinted at the future direction of urban settlement (Dittman 1990). A variable pattern of ruralization describes most of the western part of the region, with some isolated Late Bronze Age towns and cities (Ristvet 2008; Wilkinson 2002; Szuchman 2009; Ur 2010b: 157–160); no comparable systematic data exists for the Assyrian core along the Tigris River, however.
The settlement landscape of the early first millennium BCE was dominated by the great capital cities. The original political capital and enduring religious center was the old city of Ashur, which, at 70 ha, was within the range of Early Bronze Age urban sites. The political center moved to a series of increasingly large new foundations: Aššurnaṣirpal II founded the city of Kalḫu (Nimrud; 360 ha); Sargon II founded Dur‐Šarrukin (305 ha); Sennacherib expanded Nineveh to 750 ha (Stronach 1994; Oates and Oates 2001). These planned imperial capitals were three to seven times larger than even the most populous cities of the Early Bronze Age, and would have required a much larger agricultural catchment to sustain them (for Nineveh, see Wilkinson 2003: 128–30). The provincial capitals were also large, but well within the range of Early Bronze Age urban sites (e.g., Dur‐Katlimmu at 110 ha; Tušḫan at 35 ha; Kühne 2011; Matney et al. 2011).
These great urban centers dominate archaeological discussion because of their excavation histories and their artifacts, which presently fill the world’s great national and imperial museums. These cities were, however, few and far between. The Assyrian countryside was remarkably rural, especially when compared to the Early Bronze Age urban phase. Early reconnaissances failed to notice this dispersal because they focused on high mounds, the quintessential and most easily recognizable site form. Recent full‐coverage systematic survey across Iraq, southeastern Turkey, and northeastern Syria has revealed a fully settled Neo‐Assyrian landscape of small towns, villages, and farmsteads. Most of these settlements were small (two hectares or less) and are now low mounds, on the order of one to two meters high. When earlier sites were resettled, it tended to be on a reduced scale; for example, a three‐hectare village appeared on the northeastern corner of Hamoukar’s massive Early Bronze Age lower town, and a one‐hectare farmstead sat on the northern fringe of Brak’s 130 ha fourth millennium city (Ur 2010b: 112–14; Ur et al. 2011). In some cases, isolated temple structures appear to have been erected by the state within the ruins of former cities, for example at Tell al‐Rimah and Tell al‐Hawa (Wilkinson and Tucker 1995: 61). Assyrian towns emerged as extensive low mounds in the shadows of older mounds, and, consequently, have been largely overlooked by archaeologists, who have favored excavations at tell summits (Wilkinson et al. 2004).
The rural pattern of Assyrian settlement was remarkably widespread. On a regional scale, the filling of the landscape has been demonstrated almost everywhere that systematic observations have been made (reviewed in Morandi Bonacossi 2000, Wilkinson and Barbanes 2000, Wilkinson et al. 2005). The Neo‐Assyrian period saw the greatest expansion, in terms of the number of sites, as demonstrated by surveys around Tell al‐Hawa, Hamoukar, Tell Brak, Tell Beydar, and Erbil (Figure 1.3; Wilkinson and Barbanes 2000; Wright et al. 2006–07: 13; Ur 2010b; Ur and Osborne 2016). The Cizre plain, the last major alluvial plain upstream from the Assyrian capitals, experienced an identical settlement expansion (Parker 2001), as did the lower Khabur valley (Morandi Bonacossi 1996; Kühne 2010).
Figure 1.3 The evolution from nucleated to dispersed settlement in the Hamoukar and North Jazira Project areas (based on data from Wilkinson and Tucker 1995, Ur 2010b). A. Urban settlement and trackways in the later EBA, ca. 2600–2000 BCE; B. Rural settlement in the Iron Age (early 1st millennium BCE).
The evolution of the Assyrian settlement landscape occurred in three primary ways. As described above, major sites of the Bronze Age were resettled, alth
ough almost always at a much more modest scale. Furthermore, the “vacant” spaces in between these earlier settlements were filled in, in a manner that suggests a conscious attention to the interfaces of the former settlements’ catchments. Such infilling is well demonstrated for the Wadi al‐Murr (Wilkinson 1995: 145–7) and the region of Tell Beydar (Wilkinson et al. 2005 fig. 12). Finally, lands previously considered too marginal for agriculture were now heavily settled. In the Early Bronze Age, most such lands would have been considered too dry for cultivation and better used as a pastoral resource, whereas, under Assyrian control, they were filled with nearly the same continuous scatter of small settlements as the wetter plains. The lower stretches of the Khabur River, near its junction with the Euphrates River, had been sparsely settled below Dur‐Katlimmu throughout the second millennium, but experienced a remarkable expansion in occupation in the eighth century (Kühne 1995; Morandi Bonacossi 1996). Simultaneously, the steppe around the Wadi Ajij, a region that presently has less than 200 mm of rainfall annually, underwent an explosion of village settlement (Bernbeck 1993; Kühne 2010), as did the steppe around the Jebel Abd al‐Aziz (Hole and Kouchoukos, in press). The steppe around Hatra likely experienced a similar expansion (Ibrahim 1986).
The demographic transition was accompanied by, and probably closely related to, a dramatic interference in the natural hydrology. Neo‐Assyrian kings paired the construction of new capitals with the excavation of massive irrigation systems that would bring water to their hinterlands (Figure 1.4), and would irrigate a broad expanse of Assyria that had previously been subjected to the vagaries of rainfall (extensively reviewed in Bagg 2000). The Assyrian kings boasted about their irrigation projects prominently in royal inscriptions, some of which were inscribed on or adjacent to the irrigation systems themselves, allowing for an approximation of the chronology.
