on a few selected sites chosen for extensive excavation because they contained well-stratified deposits and were rich in diagnostic artifact types. Classic sites of this type in Polynesia, excavated during the 1950s and 1960s, include the Pu’u Ali’i and Waiahukini sites in Hawai’i, the Wairau Bar site in New Zealand, and the Ha’atuatua and Hane dune sites in the Marquesas.

A significant shift in archaeological research priorities, accompanied by a reorientation of field methods, began in the early 1960s. This was, in essence, a broadening of the Polynesian research agenda beyond narrow concerns with culture-historical sequences and the long-standing question of “Polynesian origins”; the agenda would now encompass questions of cultural change and evolution, of the nature of prehistoric societies and political systems, and of their ecological and economic contexts. This shift in research orientation naturally did not occur in a vacuum, and it was part of the broader reorientation in Anglo-American archaeology from a culture-historical to a “processual” approach (Trigger 1989). To a large degree in Polynesia, this involved not a complete rejection of the older culture-historical orientation but rather a broadening of the research agenda to incorporate extensive efforts at the reconstruction of prehistoric culture.

In Polynesia this shift can first be detected in Suggs’s Marquesan research, in which, despite a continued emphasis on key stratified sites, the research questions encompassed such issues as demographic, economic, and sociopolitical change in Marquesan society. More influential, however, was the introduction of the “settlement-pattern” approach to Polynesian archaeology by roger green, who had been trained in this approach by its main proponent in the United States, gordon willey of Harvard University. Green first applied a comprehensive, settlement-pattern survey methodology in his study of the ‘Opunohu Valley on Mo’orea Island (Society Islands), in which all sites in a valley landscape were recorded and treated as a record of nonportable artifact variability (Green et al. 1967). As Green summarized the perspective of settlement-pattern archaeology, “[With] increasing concern with delineating the social aspect of the data recovered from sites… the day has passed when such monuments or their structural features can afford to be treated only as contexts for portable artifacts and not as artifacts in their own right” (1967, 102).

A settlement-pattern orientation soon came to dominate archaeological research throughout Polynesia, particularly as investigations were expanded to such islands and archipelagoes as Samoa (Green and Davidson 1969, 1974), New Zealand (Groube 1965), the Marquesas (Bellwood 1972; Kellum-Ottino 1971), Easter Island (McCoy 1976), and Hawai’i (Kirch and Kelly 1975). Although initially designed to elucidate aspects of precontact Polynesian sociopolitical organization, such settlement-pattern studies soon came to include a strong research orientation toward economic and ecological questions. Thus in Hawai’i, for example, much work in the 1970s and early 1980s was focused on the field evidence for variability in prehistoric agricultural systems, both dryland field systems and irrigated pondfield terraces (Kirch and Kelly 1975; Yen et al. 1972). This research was by no means limited to surface survey and mapping, and it included new methods of excavation and analysis, such as the interpretation of agricultural soils (Kirch 1977). Similar concerns prompted major research projects elsewhere in Polynesia, such as the Palliser Bay research of B. Foss and Helen Leach (1979) and the study of the Polynesian outlier of Tikopia by P. V. Kirch and D. E. Yen (1982).

This extension of settlement-pattern archaeology to encompass economic and ecological aspects of precontact Polynesia societies also began to open up issues of the dynamic relationships between Polynesian populations and their island ecosystems. Under the older culture-historical paradigm, the island environment had been viewed largely as a static context for human settlement. The settlement-pattern approach, by contrast, put humans on the land as active agents of change. The accumulation of much zooarchaeological evidence for changes in island faunas, combined with interdisciplinary work between archaeologists, geomorphologists, palynologists, and other natural scientists, led to a considerable rethinking concerning the