Upper Mesopotamian cIties were highly nucleated, built-up settlements with wall-to-wall architecture and little open space. Extramural areas contained ields, pasture, burial grounds, and small villages. Given that the growth of these cities was rarely restricted on all sides by natural features, how do we explain this nucleation? The universal presence of city walls, sometimes matched with a glacis and ditch or moat (as aT Titris, Mozan, Al-Rawda, Beydar, and Chuera) suggests that protection from theft or conflict was one motivation for nucleation. Yet, even if violence were the exception rather than the rule, nucleation makes infrastructure less expensive to build and maintain by reducing the length of city walls, streets, and drainage systems. Another impetus for nucleation is land tenure, which may have limited sprawling settlement on communal or crown land. Finally, city walls have important symbolic value, were often constructed in the earliest period of the city, and may have fostered
Nucleation by encouraging compaction within the walls. In the following discussion, I explore these explanations for nucleation.
Defense
The existence of substantial defensive structures at large cities, small towns, and even small villages (e. g., Cooper 2006b:70) indicates strongly that violence was a real threat. In addition, city walls are symbols of power and they project power by making it possible for a city to send its army or militia out to confront or threaten others, leaving smaller forces behind to defend the homeland (Pauketat 2007:122, 131). Violence was clearly a part of life in these societies, especially in the later third millennium (Sallaberger 2007:422-423). The Ebla texts include accounts of conflict between Mari and Ebla, a forced treaty between Ebla and Abarsal, tribute gifts sent to Ebla from threatened cities, and weapons exchanged with allies (Archi and Biga 2003; Merola 2008; Sollberger 1980). Archaeological evidence for violence at these cities is not as plentiful as lists of tribute from subject polities, but excavations reveal destruction layers at several sites, weapons in burials, and victory iconography from Mari and Ebla (Akkermans and Schwartz 2003:269). Other examples of violence or suggestive of violence include a mass burial, changes in regional settlement patterns, and local urban structure (abandoned villages and suburbs), a shift from extramural to intramural burials, and construction of new defenses at Titri§ (Algaze et al. 2001:68-70), and possibly the burning of the sacred structures of Steinbauten I-II at Chuera (Klein and Orthmann 1995:75; Orthmann 1995a, plans 6, 13, 1995b:32; Moortgat 1962:35; Pruss 1998).
Symbolism
Aside from their role in defense, city walls also have important symbolic value that is recorded in texts, images, and their physical manifestation (Ristvet 2007:184). Upper Mesopotamian city walls deined the city, marking a clear perimeter with gates mediating entry and exit. A barrel cylinder from Mashkan-shapir, a second-millennium city in Lower Mesopotamia, celebrates the building of the city wall as a deining act in establishing the city (Steinkeller 2004:135-146). Third-millennium texts from Beydar list
State workers - in this case, shepherds - under gate names, indicating that in at least some cases, gate names represented quarters or neighborhoods (Sallaberger 2004:18-19). In many examples, city walls were buIlt early in the life of the city, as at Kazane, Mozan, Chuera, Beydar, Leilan, Al-Rawda, and others. These may be instances of conspicuous consumption and expressions of power, as described by Trigger (1990:127), although the labor and resources required to build a city wall that benefits all residents may involve more negotiation among corporate groups than the centralization of power described by Trigger (see for example the range of professions involved in building a city wall, as described in Ristvet 2007). Thus, the city wall may have first marked officially the establishment of a city, and later constricted growth and intensified nucle-ation as residents packed into the area behind the walls in order to be part of the city. In general, extramural settlement is rare, but it is documented at Al-Rawda anD Titri§.
PRODUCTION OF SPACE IN MESOPOTAMIAN CITIES
Although defense, symbolism, or economizing construction can be considered sensible reasons to nucleate, another possible factor is land tenure. Tony Wilkinson combines archaeological survey, analysis of period textual sources, and analogy with recent historical practice in the region to argue that Bronze Age villages probably had a communal land-tenure system. In this system, use rights shifted regularly and land did not belong to a single family (Wilkinson 2010:59). This type of land tenure explains in part the ever-increasing height of tells, which were repeatedly occupied because building outside the village on communal land was discouraged. If this system was maintained during the urbanization process, then tightly packing into a nucleated city would limit the amount of former field or pasture land that would have to be taken over by the city during the development of a lower town around the tell. Although powerful city rulers may have claimed ownership over all the land in a given region (Wilkinson 2010:57), at least initially one would expect former land tenure practices to present a roadblock to the development of low-density settlement with copious open space between residential areas. A complicating factor affecting the use of land around cities is squatters, or those constructing housing, gardens, and other features in unused open space without necessarily possessing use or tenure
Rights. These activities may be temporary, but if tolerated, their product may become a permanent part of the city (see Neuwirth 2005 For modern examples).
Wilkinson notes that texts record villages, their land, and their inhabitants changing hands as a single entity owned by the state or private families (Wilkinson 2010:58; see also Steinkeller 1993:125126). This practice could leave local land-use or land-tenure patterns intact, despite the knowledge that an overlord was the official owner. The case of city land ownership is more complicated. Although private home ownership within cities is attested by texts from the second millennium (Stone 1987; Van De Mieroop 1999), the division of extramural land is less clear. Highly generalized summaries of textual evidence for land exchanges across the third to second millennia throughout Mesopotamia suggest that traditional systems of communal ownership were eroded as a few landowners amassed vast holdings while the size of the average family's plot shrank (Zaccagnini 1999:339). In other cases, ruling institutions laid claim to all land and redistributed it to citizens according to their social rank (Dahl 2002). Inherited property may be absent from most texts, which chronicle sales, leaving open the possibility of larger, privately owned plots not listed in texts (Zaccagnini 1999:339). Texts from Beydar suggest that a handful of state officials controlled most of the land and male labor in its region (Sallaberger and Ur 2004:57-58; Widell 2003:723). However, this evidence is selective, and Jason Ur argues that the center controlled only part of village land (Sallaberger and Ur 2004:57, footnote 13). In any case, long-lived pathways between fields endure in some areas as "hollow ways" on the landscape, indicating continuity in the structure of fields adjacent to cities (Casana 2013; Ur 2003).
In sum, nucleation in Upper Mesopotamian cities may derive from a combined need for security and efficient infrastructure, which generated constrictive, highly symbolic walls. Nucleation is also compatible with traditional notions of land tenure that discourage dispersed housing, and building structures side-by-side along narrow streets is consistent with principles of defensible space (see defensible space section later in this chapter). It is difficult to determine with certainty which of these factors was primary. I favor an interpretation based on a pre-urban preference for close-knit living in clustered neighborhoods, combined with shared land tenure in which the defensive value, if not the symbolic value, of circuit walls varied over time wIth the degree of conflict.
World History
