Adze Studies in the Pacific

Stone tools represent one of the most commonly preserved stylized artifacts in Polynesian sites, and adzes in particular have been used since the turn of the century to assess relationships among island settlements. Initially typological in nature (Duff 1959; Figueroa and Sanchez 1965; Emory 1968), these studies have increasingly looked at the process of stone tool manufacture, especially for adzes (Leach and Leach 1980; Leach and Witter 1987, McCoy 1990), and at the petrographic and chemical characteristics of the stone itself as markers of resource use patterns (Cleghorn et al. 1985; Ayres and Beardsley 1988; Ayres et al. 1997; Weisler 1990, 1993, 1997; Beardsley et al. 1991, 1996; n.d.; Lass 1994). Finished tools as well as the various byproducts of their manufacture all represent important indicators of the procurement and distribution systems of interest to archaeologists.

The Rapanui (Easter Island) Stone Adz Project

Rapanui's significance in anthropological study of cultural evolution and adaptation is also closely linked to and archaeologically reflected in the unusual elaboration of stone working on the island. Its fame is largely tied to the production and transportation of stone statues; however, understanding the basic stone tools used in day to day life is necessary to resolve questions about many facets of prehistoric technology, production, and social and political organization, as well as the manufacture of the stone images.

Hard basaltic rock from quarries on Rapanui used to manufacture adzes and axes for shaping wood and stone in the traditional technology represents the focal point of this research. Tool-quality stone samples, debitage, chips from adz bits, and broken adzes, are available from archaeological sites around the island for comparison to the lithic source material in quarries. In addition to adzes, high quality stone also was used sometimes for producing stone fishooks and knives. We are developing data on a sample set of stone tool manufacturing materials from selected archaeological sites and stone quarries. This study includes a set selected for detailed geochemical analysis. We plan to analyze additional samples for refining the data base and for documenting artifact distributional patterns.

THE ISSUES

This paper explores issues of the social and economic organization of lithic procurement. Such organization affects how we interpret regional exchange networks, socially bounded territories, and elite redistributive efforts. High quality, flakable basalt used for carefully shaped adzes, chisels and other tools represents an important material resource in regional procurement and distribution. Knowledge of source locations and chemical composition data, though incomplete, are becoming more widely available for Pacific Island contexts.

Spatial and technofunctional analyses of lithic materials are in their infancy for Rapanui, but we expect they will be a major avenue to understanding past life on the island.

Neither obsidian or fine grained basalt is known ethnographically to be an integral resource in maintaining the status hierarchy; that is, their use by the elite in redistribution is not documented. It remains for archaeology to develop a coherent and comprehensive model for resource use that will allow us to establish if this was the case for prehistoric Rapanui society.

A Kin-Based Geographic Model shows a distribution that is primarily a function of acquisition and exchange networks based on family or simple reciprocal ties. In this model households acquired stone through exchange networks similar to those through which other nonlocal, patchy resources were acquired. On Rapanui, these diverse resources would include obsidian, white trachyte, red scoria, coral, as well as perhaps localized plant resources such as totora reed. This distributional pattern is not necessarily based on a formal redistributive process or ranked status hierarchy.

The Centralized Redistribution model is one where the movement of stone is facilitated and controlled by an elite or group. This is characteristic of chiefdoms and of the use of scarce resources.

THE PRIMARY SOURCE DATA

The quarry site complex, Rua Tokitoki Quarry, is located inland from La Perouse Bay on Rapanui's north coast and is in the eastern, Hotu Iti section of the island. The site or series of exposures covers an area of approximately 0.4 square kilometers and ranges in elevation from approximately 40 to 75 m. Outcrops and a number of pits, apparently for extraction of fine quality stone, and extensive debitage piles are visible.

The pits providing access to buried lava are as much as 8-10 m across, 2-3 m deep and are surrounded by the flakes and other refuse from flaking hard lava to produce tool blanks. Based on the number and size of quarry areas and the debitage from production, it is likely that thousands of stone adzes were prepared here.

Sociopolitical Context

The La Perouse quarry complex falls within the traditional Tupahotu "tribal" area within the larger entity called Hotu Iti (Routledge 1919) and also the "Matanui" group (Hotus et al. 1988). Our hypothesis is that the Tupahotu were directly involved in adz and other tool production and that they served as the source producers of these tools designed to be exchanged with other groups on the island. This hypothesis can be tested as more distributional data become available on the spatial patterns of the tool-quality rock found as adzes or as debitage.

ANALYTICAL RESULTS

Twelve groupings containing multiple samples were distinguished based on the macroscopic and microscopic examinations of 100 samples taken from 11 sites found in 9 survey quadrangles (Quad). These ranged from very coarse material (Group 4) with distinctive petrographic features to very fine materials with black/dark groundmass that differ only slightly from one group or subgroup to another (See Table 1). Of these, Group 2 includes the quarry materials sampled from Rua Tokitoki and identifies the important geographical spread of this distinctive and prime adz quarry material. Groups 6 and 9 have the widest geographical distribution. The small set of Quad 3 samples from the island's southwest corner shows the most distinctive set of materials--50% of the samples belong to GRP 9--and none are from Rua Tokitoki.

The basalt adz-quality materials on Rapanui have considerable potential for provenance determination at the quarry level or at the "Group" level. As with all fingerprinting efforts based on elemental or other data, sampling and subpopulation boundary problems are significant issues. This pilot study has not yet addressed intra-quarry variability at Rua Tokitoki--this will require substantial additional field observation and sampling--however, the Rua Tokitoki material is sufficiently distinctive that it can be differentiated petrographically from other materials. While not all group definitions are confirmed, in all, at least 15 distinct source materials have been distinguished. The study does demonstrate that further study could successfully differentiate several source areas for adz-quality stone and identify a small number of geographical locations where distinct stone was acquired.

We offer the hypothesis that the Rua Tokitoki quarry area was exploited primarily in late prehistoric-early historic times (ca. AD. 1300-1850's) and that--based on the scale of production and the provenance determinations made in this research--it provided stone tool blanks for use beyond the local area. Thus, inter-kin group (sociopolitical unit) exchange is represented. We find no evidence to support the idea that Rua Tokitoki, although it was a massive quarry undertaking, represents the main source of adz material as stated by Métraux. It is clear that no single quarry provided all or even most of the Rapanui adzes--in part because of functional differences among these tools--and available data suggest considerable local or regionally-available stone was used. This determination argues against any elite control of stone adz materials.

Tool stone distribution is significant for any discussion of lithic provenance issues related to use of stone for purposes other than flake stone tools, including building, statuary, and varied uses of non-basaltic rock such as volcanic glass, pumice, scoria, trachyte, and coral rock. Each of these uses represents a somewhat different set of issues for provenance determination and interpretation, but taken together we would expect to get overlapping patterns of resource extraction and use that would provide a stronger basis for hypotheses regarding social interaction and stone as commodity.

STONE ADZES

A set of stone adzes, including small versions, called chisels, that represent woodworking tools from Easter Island. These tools were hafted to an L-shaped handle to form the complete tool and reflect the long-standing practice of carving on the island. These are primarily for shaping wood, but some adzes and pick-like tools were used for carving and finishing stone used in building and sculpture. The two longest specimens show side views; the others illustrate the back or "base" view of the tool. Several methods of manufacture are represented, in order of application, percussion flaking, pecking, and grinding; fully finished tools (toki) are typically ground. Types from left to right: 4D, 2A, 4D, 3H? (upper), 2A (lower left), 2A/3D (lower right). Specimens originally examined in the Father Sebastian Englert Museum, Rapa Nui.