the similarities and differences that exist among various societies.

It is the same story on the coast where the indigenous archaeologists (Abungu 1989; Mutoro 1987) have tried to show that coastal settlements were not foreign but local in origin. Felix Chami (1992) has shown that on the southern Tanzanian coast there was continuous occupation of sites starting from the first millennium a.d. to around the twelfth to fifteenth centuries. Furthermore, he shows the presence of grinders used for bead manufacture as well as marine shell and glass in sites that are about 55 kilometers away from the sea. In addition, there are similarities between littoral and hinterland ceramics, which Chami argues is indicative of both cultural and economic affiliation between the littoral and its immediate hinterland. This evidence contradicts the hitherto held opinion that the coast was culturally different from its hinterland (Horton 1984).

What is remarkable about the coastal studies by indigenous archaeologists is their attempt, not to rely on the standing structures hitherto used by most foreign archaeologists, but to use the nonvisible settlements, those thought to have been occupied by local inhabitants, including migrants and traders, as well. Notable among indigenous archaeologists is Paul Msemwa (1988, 1990) who has looked at refuse disposal within coastal settlements and used the results as a base in explaining the archaeological record of the area. It is also noteworthy that local pottery, which had previously been ignored when constructing coastal chronology—the earlier chronology was based on imported Arabic/Persian (Islamic) and Chinese pottery—is now being used as an important variable in chronological construction. The danger, however, of using pottery typology to date sites is not taking into account the problem of human scavenging activities. Old potsherds have frequently been found incorporated into later sites through human activities of recycling and reuse. Likewise, later human activities in older settlements occasionally mix up the archaeological record. For instance, at Kilwa, in southeastern Tanzania, carbon–14 dating contradicted the principle of superposition, with charcoal samples taken from lower levels being much younger than samples taken from the upper levels (Chittick 1966a, 1966b, 1967). It is noteworthy that Msemwa questions this disturbed nature revealed by carbon–14 dates and is looking at site formation processes as a way of developing an alternative chronology of coastal sites.

Almost all the studies that have been done in East Africa are bereft of theoretical discussion. The exceptions are Msemwa (1989), and Kiriama (1993). Msemwa used the general systems theory in modeling the use of the Rufiji River in Tanzania by the Warufiji people. In his study, Msemwa divided the river into ecosystems with each ecosystem being defined by a configuration of landscape and soil types, subsistence activities, distance from the sea, and means of communication. Using this model, Msemwa found that despite the fact that Mafia Island had all the advantages of the other coastal settlements, Mafia did not grow into an important center. Msemwa argues that this was a result of a lack of adequate resources, such as freshwater, and of the presence of malarial mosquitoes. He theorizes that emphasis should be placed on the relationship between resource availability and the patterning of settlement locations, particularly the nature of those resources.

Kiriama (1992) went further and used the food systems research paradigm to study not only coastal settlements but also their relationship with the hinterland. Since food systems research is concerned with changes in food systems over the long term, its application assists in understanding how settlement sizes were affected by the intensification or abatement of food resources or how food production in a particular region contributed to the migration and subsequent settlement of the area. Kiriama argues that the availability or lack of resources in the hinterland affected the setting up, sustenance, and eventual decline of some of the settlements on the coast. In other words, instead of looking at extraneous factors for the origins and subsequent decline of East African coastal settlements, researchers should look at developments in the hinterland.

Other foreign researchers have used the new archaeology paradigm in the region, not to try