Considering the Source: The Importance of Raw Material Characterization and Provenance in Obsidian Use-Wear Studies

The reason for the choice of a specific type of obsidian for tool manufacturing is generally assumed to be related to the location of the source, the appearance of the material, and/or the sociopolitical relationships that may restrict its procurement, rather than the inherent qualities of the type of obsidian. This research examines the use of Monte Arci obsidian (types SA and SC) using macroscopic and low power techniques. From these two types of obsidian, an experimental set of tools was made and used on materials available in Sardinia during prehistory. The results of this comparative experiment show different types of obsidian have dissimilar wear patterns and vary in effectiveness when used on the same material. This information was applied to the analysis of expediently Considering the Source: The Importance of Raw Material Characterization and Provenance in Obsidian Use-Wear Studies 1 Department of Anthropology, University of South Florida, Tampa, FL 33620, USA. [email protected] L’ossidiana del Monte Arci nel Mediterraneo. Nuovi apporti sulla diffusione, sui sistemi di produzione e sulla loro cronologia, Atti del 5° Convegno internazionale (Pau, Italia, 27-29 Giugno 2008), a cura di C. Lugliè. NUR, Ales, 2010 T. J. Setzer and R. H. Tykot 72 made, informal, small, flaked obsidian tools excavated from the Neolithic site of Contraguda, and from Duos Nuraghes. Interpretations are made about site function and the role of obsidian in different time periods. 1. The choice of a lithic material Archaeologists have researched the function of lithic artifacts for over a century. In order to identify the purpose of these artifacts, conclusions have been drawn based upon the form of the tool, ethnographic studies, residue analysis, and use-wear research. Functional studies are frequently used to obtain clues about the activities that occurred at a site and explore prehistoric lifeways. The results from functional studies not only contribute to the understanding of the choices prehistoric people were making regarding the materials used for lithic tools, but they also give ideas to others conducing use-wear research. While the use of a tool has long been accepted as one link in the chaîne opératoire model, which is used to understand human decisions, this research demonstrates the importance of considering the other links when forming experimental designs and answering research questions. Without the correct identification of the source of the lithic material (at least in the case of obsidian) used by prehistoric people, and the inclusion of this factor in the experimental design, one is in danger of forming incorrect interpretations of the use of the tool, function of a site, and general human behavior. Although the properties of the individual raw materials must have led to the deliberate selection of these resources, little research has been done on the relationship between use and other attributes, such as flaking properties and raw material variability. For example, Greiser and Sheets (1979) compared the wear patterns of different lithic materials, such as variations between flint and obsidian; however, they did not research the differences between physicochemically varying obsidian. Others, such as Kamminga (1978, 1982), have found considerable mechanical variation within some types of rock, such as quartzite, yet little variability in other lithic materials, including obsidian, which has a limited range of usefulness due to its fine texture and brittle nature (Hayden 1979). Schiffer (1979) noted that materials from a single source vary physicochemically, and experimental studies should produce results that are applicable to all lithic materials. However, Spear (1980) concluded that in general the wear on obsidian was quite similar to that of chert and he found that it is quite possible to determine the direction of use, as well as if the tool was used on hard or soft contact materials. This research (Setzer 2004; Setzer and Tykot 2005) examines the use Considering the Source 73 of obsidian tools from the sites of Contraguda and Duos Nuraghes. In particular, this research addresses the question of variability in Sardinian A (SA) and Sardinian C (SC) obsidian from Monte Arci and the usefulness of the two types when performing certain tasks. Observations were also made regarding the wear patterns occurring on these obsidians. 2. Contraguda The open air site of Contraguda sits on a hill in the Coghinas Valley of Sardinia, about 20 km from the north coast of Sardinia and 3 km north of the town of Perfugas in Sassari. Used during the Late Neolithic, or the Ozieri period, Contraguda extends over several hectares and is the largest Ozieri settlement known on Sardinia. Not only is the site one of the largest open-air sites from this time, but it is also one of the only openair sites with obsidian artifacts. While most of the contemporaneous sites that have produced obsidian artifacts on Sardinia have been almost exclusively rock shelters and caves, Contraguda provides archaeologists with a different perspective on the lifeways during the Late Neolithic on Sardinia. This site was first identified in 1980 during an archaeological survey that was conducted to identify and catalog archaeological features at the site. In 1992, Boschian et al. (2000-2001) conducted a systematic investigation of Contraguda. Five radiocarbon dates obtained from this site place it in the middle of the fourth millennium BC, with the calibrated dates ranging between 4050 and 3770 BC. However, the samples that provided these dates were not from the same context as the obsidian tools, which appear to be from later in the Ozieri period. The tools examined in this research were found at this site outside of a feature, called Structure A-B. This structure is composed of a series of small, interrelated walls, which form rooms. Structure A-B is of unknown purpose, and the quality of the construction varies. One hundred ninetytwo obsidian tools were found in strata beneath the plow zone associated with this structure. 3. Duos Nuraghes Duos Nuraghes is part of the Borore Group of sites, which is an elliptically shaped cluster of sites occupying about 40 sq km of Borore and the adjacent Birori communes. Long noted for its fertile volcanic soils and numerous springs, the region has a long history of mixed farming with a reliance on sheep. Research at Duos Nuraghes (Webster 2001) has uncovered ceramic jars, vessels, metal tools, and approximately 300 pieces of obsidian. Obsidian hydration dates (Stevenson and Ellis 1998) T. J. Setzer and R. H. Tykot 74 suggest that obsidian was re-used throughout the occupation of the site and that some contextual disruption of the site has occurred. Obsidian hydration dates from this site range from the late 35th century BC to the early second century AD. Due to the destructive nature of obsidian hydration dating, the tools examined in this study were not dated, and are addressed by their deposition into various strata. 4. The experiment The goal of the experimental portion of this research was to produce a set of tools comparable to the artifacts found at Contraguda and Duos Nuraghes, use them to process various materials that were likely used in prehistoric times at the sites, examine the wear patterns on the experimental set, and compare the use wear on the experimental tools to those found on the assemblages from the sites, thus allowing the function of the prehistoric tools to be interpreted. In addition, the experiment addressed the possibility of the choice of a type of obsidian based upon variation in function. The tools used in this experiment, as well as those analyzed from the site of Contraguda and Duos Nuraghes, are expediently made, flaked tools without retouch. First, both in situ and secondarily deposited samples of obsidian were obtained from the Monte Arci region for this research. Type SA obsidian was collected from the Conca Cannas region of the western side of Monte Arci, and SC obsidian was collected on the eastern side of Monte Arci. One large nodule of SC obsidian and two smaller pieces of SA obsidian were selected for the production of experimental tools. One hundred fifty experimental tools were produced using direct hard hammer percussion methods. From this set, 80 were selected based on attributes such as size, sharpness of edges, and morphology. In other words, the pieces that resembled tools from the artifact assemblages were selected. This sample set was divided into SA and SC obsidian tools. These experimental tools were numbered and classified by type, based on definitions taken from Andrefsky (1998). The types used in this study were flake, flake shatter, non-flake debitage, and blades. These tools were numbered and the type of obsidian was noted, along with topographic features, edge morphology, and both macroand micro-edge wear. The topography included the general nature of the edge (e.g., flat, undulating, or ridged) as well as other attributes that were present on the edge, such as percussion ripples and feathering. Morphological features of the used edge were also recorded. These features included the angle, length, thickness, profile, and shape of the edge. The angle measurement of the used edge was taken at the midpoint of the used edge 1 mm back from Considering the Source 75 the edge of the tool using a goniometer. The length of the used edge was measured using a pliable piece of wire as a guide. The measurement of the thickness of the tool was also taken from the midpoint of the used edge. The profile is a measurement of the plan of the use edge, which could be convex, straight or concave. This measurement is a ratio of the perpendicular distance of the working edge and its chord, or linear distance between the extremities of the working edge. Sixty of these tools (30 made from the SA obsidian and 30 made from the SC obsidian) were used in a controlled experiment, while volunteers in a blind experiment used the other 20 tools (ten made from SA and ten made from SC obsidian) in order to assess how accurately the use-wear patterns were interpreted. The use-wear materials were chosen based on the categories outlined by Shea and Klenck (1993), who categorized them in terms of yielding and resistance (Table 1). The yielding classes are soft, medium (semi-rigid), and hard (rigid), while the resistance categories were animal (non-siliceous), vegetal (moderately siliceous), and inorganic (highly siliceous). The aim of this portion of the experiment was to produce wear patterns that could be compared to those on the artifacts allowing for the function of the prehistoric tool to be attributed to one of these categories, rather than a specific plant, animal or other material. The materials processed by the obsidian tools were chosen based on their availability during the Neolithic in the Contraguda region. First, the control set was used and analyzed. Macroscopic and microscopic use wear was recorded on these tools. For the purposes of this analysis, macroscopic wear is that which is seen without any magnification, and microscopic wear is that which is seen at 50x magnification. The identification of the most likely function of the tool was based on the type of use wear, its frequency, size, and distribution. In addition, processes such as edge rounding were also considered to aid in diagnosing tool function. Hardness of materials worked Soft Medium Hard