Discovering Architectural Patterning at a Complex Site

Abstract

To enable systematic interpretive studies of community organization and population, two techniques are advocated: (1) a probability sampling technique for the selection of at least a portion of a site's excavation units and (2) the systematic recording of architectural space. A substantive example is drawn from the investigations of a historic period town in northern Morocco, Qsar es-Seghir. Analyses of architectural remains produced information on the overall nature of land use, general layout of the community, and two alternative methods of estimating population.

Research Design and Sampling

“…Sometimes we can even expect new information to arise during the course of research that makes it desirable to revise our research design. Many researchers use "multi-stage" designs that use a preliminary or pilot stage to refine the research questions or methods and expectations about the data (e.g., Redman, 1973;Redman and Anzalone, 1980). Furthermore, research designs do not prevent us from recording anything we happen to notice during the course of making other observations.…”

“…In order to see and recognize any patterning in the spatial arrangement of structures and other features or facilities on the site, we need to have broad exposures, not a random sample of 1m x 1m squares or the like. In a very large settlement, we may still need to sample because of the extreme cost that would be involved in exposing the whole site, but then we might opt for a cluster sample, exposing large expanses in only one or two areas of the site that we have reason to believe might be representative of the whole or where we have reason to believe there is structurally interesting evidence (c.f., Redman and Anzalone, 1980). Sometimes a small pilot sample could be used to guide the planning of the larger cluster sample or a carefully planned purposive sample.…”

Recent Developments in Old World Archaeology

“…Bartel, working with a joint Yugoslav-American field team at Kraku'lu Yordan, a fourth century Roman site in northeastern Serbia, instituted a system of random sampling to guide the excavation program (Bartel et al 1979). Redman, at a medieval port fortress, Qsar es-Seghir, in Morocco, used a stratified, unaligned, systematic probability sampling technique to ensure even excavation coverage over the entire site and to facilitate expansion of investigations beyond the original sample (Redman et al 1979;Redman and Anzalone 1980). As a component of the Urban Survey Project of the Archaeological Exploration of Sardis, Sterud and Freedman employed a variety of random sampling techniques to investigate the surface areas of several localities within the city prior to excavation by the Sardis Expedition (Greenewalt 1979:1-4).…”