The obsidian artefacts came from two sites, FRL and FAO, located in West New Britain Province, Papua New Guinea. These sites are notable for having well-defined stratigraphy and relatively long time depth. The Bitokara Mission Site (FRL) was excavated by Specht (1988). The FAO excavations were part of the larger Garua Island Project (Torrence et al, in press). Both sites have a sequence of volcanic tephras with a well-developed soil in the upper horizon of each. The volcanic events have been dated by radiocarbon. The tephras have been sourced and cross correlated by macroscopic and geochemical techniques.
Stratigraphic units have been divided into four chronological periods:
- Older than W-K1 tephra (10,000 – 6000 BP)
- Between W-K1 and W-K2 tephra (5900 – 3600 BP)
- Between W-K2 and Dk tephras (3600 – 1100 BP)
- Post-Dk tephra (1100 BP – present)
Organic preservation of macroscopic plant material is poor due to the high acidity of soils and high rainfall, but organic residue on stone artefacts was not unexpected because phytoliths are extremely resistant to decay.
- 10 artefacts taken haphazardly from each arbitrary unit of 1m x 1m grid during excavation, using a metal trowel so that they were not handled in any way. With adhering soil, artefacts placed into plastic bags, air-dried, sealed and stored.
- Analysis began with microscopic examination. Use-wear and residues assessed concurrently using high and low magnification.
- To distinguish from post-depositional contamination, integrated analysis of residues with four main kinds of use wear is required. These are:
- Surface preservation of the obsidian tools and residues present was recorded.
- Phytoliths identified by comparison with reference collection.
- Integrity of residue on artefact tested by comparing phytolith assemblage on surface with that of surrounding soil. If residue due to usage then there should be differences.
- Phytolith assemblages included wide variety of grasses falling mainly within Bambusoideae and Panicoideae sub-families, herbaceous types (Compositae, Marantaceae, Cyperaceae), diverse arboreal types (Palm, Moraceae, Musaceae, Burseraceae, Annonaceae, etc) and unknown Dicotyledon types.
- The analysis of starch grains, often from plants that do not make phytoliths (e.g. Root crops such as taro and yams), expand our knowledge of plant use. The authors were surprised to note the survival of these given the aggressive techniques of phytolith extraction. The only current explanation is that the grains were silicified in some way.
- In a number of examples the phytolith data provided additional information on the types of plants which produced the use-wear.
- The earliest Periods 1 and 2 appear to have been characterised by both multi-functional and single purpose tools primarily for processing starch. Hafting was common, but unretouched flakes were also used on soft, starchy plants.
- Period 3 is characterised by change to more expedient tool use.
- By period 4 this trend has progressed further. In the majority of cases the very sharp, unretouched obsidian flakes were grasped with bamboo leaves. There is no evidence for starch processing in this period. This is important as ethnographic accounts only refer to uses such as shaving. Shell not stone tools used in the recent past for processing starchy tubers.
Kealhofer, L, Torrence, R and Fullagar, R. 1999. Integrating Phytoliths within Use-Wear/Residue Studies of Stone Tools. Journal of Archaeological Science 26: 527-546.