The coral-built towns of the eastern African coast figure prominently in discussions of African forms of precolonial urbanism (LaViolette and Fleisher 2005; McIntosh 1997). Earlier questions about the origins of urbanism - often invoking external inspiration - have been supplanted by recent research investigating urban function (McIntosh and McIntosh 1993). The crucial insight regarding what functions cities carry out for associated populations has helped to reinvigorate the study of African urbanism, and reveal the distinctive qualities of precolonial urban formations across the continent (Fleisher 2010; LaViolette and Fleisher 2009; McIntosh and McIntosh 1993; Pikirayi 2001; Wynne-Jones 2007c). SwaHili towns have been explored with reference to their wider hinterland, seen as market centers and ritual foci, the apex of a settlement hierarchy, and home to an emergent mercantile elite (Wright 1993:670).
On the Swahili coast, this conceptual turn is based on a series of regional surveys that set the coral-built towns within their larger settlement context. In all regions subjected to such analysis, the stonetowns have emerged as just one component of a much larger landscape of occupation. Foundational research of the 1980s and 1990s, which sougHt the earliest earth-and-thatch iterations of the stonetowns (Abungu 1989; Horton 1996; Sinclair 1987) proceeded concurrently with a series of surveys that mapped the distribution of settlement in the coastal regions (Chami 1994; Fawcett and LaViolette 1990; LaViolette et al. 1989; Schmidt et al. 1992; Wilson 1982); the overall picture that emerged was of coral-built towns developing out of settlements of impermanent architecture, which were themselves part of a network of similar settlements across the region. The African roots of the towns were thus firmly established. More
Recently, regional studies have explored areas around specific towns, with the objective of examining town-country interactions during the period of urbanization (Fleisher 2003, 2010; Helm 2000; Wynne-Jones 2007c). Within the context of the broader regions, it had been assumed that the settlement pattern reflected a simple relationship of increasing hierarchy, with the developing stonetowns emerging as political and economic foci (Wright 1993:665). These studies have begun to explore the types of relationship that might be manifest between developing centers and their regions, exploring centripetal processes that go beyond the economic. Fleisher (2003, 2010), for example, has suggested that the stonetown of Chwaka functioned as a ritual center and brought the community together under the banner of Islam, wIth the construction of the elaborate mosque manifesting the central position of that religion for a wider community. In Chwaka's surrounding region of northern Pemba, urban development was accompanied by a dramatic decline in countryside settlement, while in southern Kenya (Helm 2000) and southern Tanzania (Wynne-Jones 2007c), urban formation had less of an effect on the surrounding populations. In the latter examples, different kinds of economic, and possibly ritual, communities have been postulated that did not lead to a focus of regional population within the emerging coral-built centers.
SWAHILI URBAN SPACES Of THE EASTERN AFRICAN COAST
These approaches to urban function explore particular types of practice - whether political, economic, or religious - and gauge those forms of production in relation to surrounding populations. The focus of attention moves away from a search for definition to one of process. Despite recognizing the complexity and diversity of Swahili society, and the towns' relationships to broader populations, this turn toward urban function has somewhat neglected the study of urban planning wIthin the stonetown settlements. Yet, the seeds of an understanding are present, through an emphasis on interaction and activity. Therefore, we see the movement toward a more functional interpretation oF African urbanism as an important step toward thinking through the way human practice came to define (anD be defined by) urban spaces.
In this chapter, we begin by discussing what SmIth (2007), following Rapoport (1988), calls "high-level meaning," explored through ethnographic studies of the Swahili (see Fisher and Creekmore, Chapter 1 In this volume). High-level meaning relates to "cosmologies, world views and the domain of sacred" (Smith 2007:30),
Whereas middle-level meaning reflects the more worldly concerns of communities and elites in the manipulation of space and the "transmission of messages about identity, status and power" (Smith 2007:30). It is at this latter level that archaeologists have often explored the manipulation and use of space as a crucial resource through which power and authority were legitimized, constituting the "political landscape" of the ancient world (A. T. Smith 2003). Low-level meaning "concerns the recursive relationship between architecture and behavior" (Smith 2007:30) and is often sought in archaeologies of daily practice. Following the consideration of ethnographic data, we discuss archaeological attempts to decipher the meaning of town plans, arguing that these most commonly fall under the rubric of middle-level meaning, thus favoring the agency of elite members of society. Finally, we explore the way that previous researchers on the Swahili have emphasized the role of practice in the development of key institutions in Swahili towns, most importantly the stonehouse tradition, and argue that this practice-based understanding, which seeks low-level meaning in urban plans, has the potential to transform our understanding of Swahili urban spaces. As such, the levels of meaning give structure to our discussions. Yet, as Rapoport himself states (1988:325), levels of meaning should not be thought of as mutually exclusive categories, but as heuristic tools for subdividing a continuum of practical action. Here, we attempt to show that it is equally unhelpful to separate out different features of Swahili town planning, but that an approach that incorporates both ritual and worldly concerns, as well as the ways that people lived in and used spaces, can unite the disparate understandings already achieveD through ethnography and archaeology.
World History
