PROJECT SUMMARY

This summary is based on a larger, more detailed article that has been submitted for publication to the Journal of the Polynesian Society.

Shortly after Christian missionaries arrived, the population of Samoa was overwhelmingly concentrated along the coast (Watters 1958). Synthesizing information from a variety of previously published sources, Watters estimated that 96% of the population of lived within a mile of the sea; the remaining 4% lived in isolated inland settlements, or in one of just a few known inland villages.  However, prior to 1840 there were numerous mountain ridgetop settlements, many thought from oral traditions to be refuge sites (Davidson 1969; Smith 1899).

These sites are often attributed by Samoans to a legendary period of warfare with the Tongans, said to have occurred about the 13th century A.D. (Golson 1969). Archaeologists have since investigated fortified inland and/or mountain ridgetop sites and have also suggested that they possibly date to this period in time (Best 1992, 1993; Buist 1967; Frost 1978), even while sometimes simultaneously challenging the use of oral tradition to explain them (see especially Best 1992). However, little independent data has come forth to test the basic hypothesis that mountain residential sites were occupied by Samoans due to internecine warfare, or especially, Tongan occupation of coastal sites. The first step in evaluating these claims would be to establish the authoritative chronology of the mountain settlements, thus confirming whether or not the sites were occupied in the correct time frame. Then, alternative explanations must be considered.

In 2004, archaeologists from Texas A&M University undertook a single-season exploration of the mountain residential sites in American Samoa in order to resolve the question of their chronology. Specifically, we wanted to determine the date of their initial construction. Because they are reported to have been occupied until the historic period, we were less concerned with establishing the radiocarbon chronology for their abandonment. It was also hoped that the excavations would reveal whether the sites were stratified, possibly indicating multiple occupations.

Three large previously known mountain residential sites were selected for inclusion in the study: Lefutu (AS-21-002), Old Vatia (AS-24-002), and Levaga Village (AS-25-027).

During excavation, special attention was paid to charcoal and other organic material suitable for radiocarbon dating. Frost’s (1978) descriptions of the sediments she encountered at the mountaintop sites in the 1970s indicate that the discovery of charcoal would not be a problem.

Samples were submitted to Beta Analytic, who performed the pretreatments, and radiometric or AMS analyses. Four of the dates ostensibly mark the period of initial construction of the foundations, and by extension the occupation of the mountain settlements (Beta-194326, 193194, 193195, and 193196). A chi-square of these dates reveals that they are statistically the same at the 95% confidence interval (df=3; T=2.1; c2(.05)=7.8). The pooled mean conventional age (calculated with OxCal 3.9) is 680±20 B.P. I interpret these results to mean that the three sites were first established at roughly the same time; certainly within 100 years of each other, and most probably within 50. If the pooled mean conventional age is calibrated, 56.7% of the variance at 2 sd is captured within the period A.D. 1270-1310. I suggest that this is the most like age during which the sites were initially built. This interpretation is further enhanced if the curve intercept data are consulted. Though now considered less reliable than the probability method, the intercept method in Calib 3.3 (Stuiver and Reimer 1993) corroborates these results: A.D. 1300 (Beta 194326), A.D. 1300 (Beta-193196), A.D. 1290 (Beta-193194), and A.D. 1280 (Beta-193195). Bear in mind that the calibration itself is an interpretation and is subject to change as calibration curves are refined.

The data presented here indicate that mountain residential sites began to appear on Tutuila, with some variance, around 1300 A.D. Large-scale changes were taking shape in Samoa by the beginning of 14th century. In
addition to the establishment of mountain villages we see the continuation of industrial quarrying activity and the final acts of exporting Tutuilan basalt around the Pacific (Best, et al. 1992; Clark 1993, 1996; Clark, et al. 1997; Leach and Witter 1987, 1990). If ethnographic accounts are accurate, this was a time of great social transformation as well, as the period of Tongan domination was either ending, or peaking, at this time. Subsequently, there was a profusion
of monumental architecture, including star mounds, fortifications, great walls, and the like, associated with increasing socio-political complexity, between A.D. 1300 and 1830 (Green 2002; Meleisea 1995). This is the Samoan culture that was in full bloom when European explorers were first encountered, and may rightfully be termed, “Classic Samoan.” Archaeological investigation of this period is vital to an improved understanding of Samoan prehistory and its place in Polynesian cultural development.