Material Analysis

The First Season of Excavations at A’asu, Tutuila, American Samoa

Adzes | Composite Edge Scrapers | Other Stone Tools

Adzes are generally thought of as the quintessential Polynesian stone tool. In Samoa most sites with lithic assemblages are characterized by the presence of adzes, while other flaked tool types are less frequent. At A’asu, however, adzes were the least numerous tool type encountered. Adzes are defined as cutting implements with the cutting edge running transversely to the long axis of the haft (Buck 1930). This simple definition seems to incorporate a variety of functional types, including what might be called adzes, chisels, and even axes and end scrapers in other typological systems. In Samoa and most of the rest of Polynesia the term adze incorporates all such tools on formal grounds, but does not imply function. A variety of functions are usually imagined or inferred from ethnographic accounts, including tree felling, canoe-shaping, scraping, and many other uses.

All stages of adze production, except quarrying, are present at A’asu, as the collection has specimens ranging from blanks to finished products. Bedrock grinding facets (foaga) have not yet been discovered at A’asu, but their existence is not ruled out. A rare hammerstone (0226) also suggests that stone tool manufacture is taking place on site. Furthermore, debitage is the most numerous lithic type present in all levels.

Two weak correlations were also noted: 1) The older buried adzes were composed of an olivine basalt (Basalt II), while surface artifacts were composed of darker basalt (Basalt I); and 2) the degree of finish in terms of grinding, is greater on the surface finds than the buried ones. Grinding may be described as either a functional or stylistic attribute of finished adzes. There clearly is a functional component to a ground finish in that the blade is essentially sharpened, balanced, and performance improved by the process. However, it is not clear that the degree of grinding present is necessary to achieve the level of performance desired by the manufacturer. Excess grinding may have been stylistic attribute associated with social status, artifact chronology, material type, intended use, geography, or some combination.

Adzes 0279 and 0255 both have been intentionally thinned at the butt, potentially for hafting. This feature, sometimes referred to as a tang, occurs on 50% of the completed adzes from our collection with a preserved poll, confirming Green’s (after Palmer 1969; Green 1974) suspicion that this is a formal and well-established practice in the post ceramic period in Samoa.

Composite Edge Scrapers top of page

Until recently, flake tools were not well -described in the archaeological literature for Samoa, making it difficult to come to any conclusions as to whether or not formal flake tool catagories were appropriate. The recent publication of a proposed lithic typology of Samoan flake tools was sorely needed, and should be elaborated as needed (Clark et al. 1997).

AASU-242, formed on a large, triangular flake, provides an example of a tool that shows attributes of both Clark’s class 7b (burin) and 2c (rounded scraper). Usewear and remodification occur unifacially around the entire edge of the flake, except for the platform. The distal edge is retouched to form a steep sided, rounded edge, making an arc of approximately 59°. The rounded edge is 4.4 mm thick, and the edge angle (steepness) is 66.5°. The retouch and unifacial edge damage occurs only in the direction of the dorsal surface, indicating that it was scraping along a surface with its ventral face forward. In this specimen, the proximal edge of the tool is pointed, exhibiting minor edge damage, indicating that it may have been used as an engraver.

Other Stone Tools top of page

There are four additional specimens with intentionally rounded scraper edges (AASU-231, -266, -273, and -300). Of the remaining two, both are weakly rounded, but the evidence is less clear (AASU-243 and -297). Edge angle (steepness) may turn out to be a key attribute in developing a typology of scrapers. What constitutes the difference from a steep-sided vs. normal edge angle? Further research on larger collections may be able to quantify the answer, but of the seven scrapers in this small assemblage, four exhibit an edge angle greater than or equal to 58º (AASU-231, -242, -266, and -300), while the other three have edge angles less than or equal to 30º (AASU -243, -273, and -297). This substantial bi-modal distribution warrants further investigation. While edge angle is, in part, a function of a thicker flake, the hypothesis that it is an intentional product cannot be ruled out. Of all scrapers only one exhibited grinding (AASU-297).