Continued erosion of the Sigatoka Sand Dunes on the western coast of Viti Levu, Fiji has exposed discrete assemblages of ceramics associated with all phases of Fijian prehistory. Excavations here in 2000 investigated stratigraphically separated occupation floors associated with Fijian Plainware and Navatu phase components, respectively radiocarbon dated to between ca. 450-550 C.E. and 550-650 C.E. The excavations and analysis of recovered data allow for a clarification of previous misunderstandings of the mid-sequence occupation at the site as well as its associated uses and features. These data further bear upon the Plainware/Navatu phase transition for Fiji as a whole. In the Lau Islands of southeastern Fiji this transition is described as abrupt and attributable to influences or a population movement from Vanuatu. Mid-sequence ceramic and other data from Sigatoka illustrate a similar break that potentially represents different cultural traditions. Keywords: Fiji, Sigatoka, excavation, ceramics, migration.
Few other sites in the history of Oceanic archaeology have spawned as much research investment as the Sigatoka Sand Dunes on the Coral Coast of Viti Levu in Fiji (Figure 1). Continuous erosion on the face of this extraordinary geological feature over the past half-century has exposed a wide array of archaeological materials dating from the initial period of Fijian settlement up to and including the historic era. Numerous archaeologists have conducted surveys, surface collections, or excavations here, and each has used his or her results to create or refine our understanding of the site and Fijian prehistory. As was illustrated in pivotal excavations at the dunes in the mid 1960s by Lawrence Birks (1973), the ability to delineate cultural complexes at Sigatoka is greatly enhanced by large volumes of temporally diagnostic ceramics, including restorable vessels, by a rapid burial of archaeological remains and by consequential separation of cultural strata by sterile lenses of sand.
[FIGURE 1 OMITTED]
Yet, and in spite of the large number of projects, the Sigatoka Sand Dunes archaeological record remained enigmatic, if not problematic, at the turn to the twenty-first century (see Marshall et al. 2000). Notwithstanding the literally hundreds of thousands of ceramic sherds recovered from the site and the widespread occurrence of human burials (Best 1989; Visser 1994), only limited evidence for sustained occupation was present (Birks 1973; Burley 1997; Burley and Shortland 1999; Hudson 1994). Indeed, prior to 2000 architectural or habitation features were rarely encountered, faunal material of any substance was nonexistent, and the nonceramic artifact assemblage amounted to little more than the occasional adze or grinding stone (Birks 1973:47-50). And in a recent review of Sigatoka archaeology by Marshall et al. (2000:5-8), even the usefulness of the Sigatoka ceramic sequence for modeling Fijian culture history was challenged.
In this paper, I present initial results of an excavation project undertaken on the eastern end of the Sand Dunes in May and June of 2000. Recovered materials resolve many of the incongruities as given, and they provide a new and important data set by which to reconfigure an understanding of mid-sequence occupations at Sigatoka. The data come from stratigraphically separated habitation floors, each with architectural features, distinctive ceramic suites, nonceramic artifact assemblages, and faunal remains. Ceramics from the lower component correlate with what I will call the Fijian Plainware phase, with continuity in vessel form and technology extending back to the Lapita period in Fiji. The upper component relates to the Navatu phase, it having a different ceramic assemblage as defined by the presence of distinctive new vessel forms, a new range of decorative applications, and, possibly, a variant manufacturing technology. Plainware and Navatu occupation floors are very closely spaced in time at Sigatoka, and diagnostic ceramics for each are found elsewhere on the dune associated with what Birks (1973) defined as the "Level 2" paleosol. …