Are obsidian subsources meaningful units of analysis?: temporal and spatial patterning of subsources in the Coso Volcanic Field, southeastern California

Abstract

Archaeologists frequently assign artifacts to chemically discrete subsignatures of major obsidian sources. While the technical ability to do so has been demonstrated, it remains to be shown that such information is behaviorally meaningful. Indeed, some analysts choose not to make such determinations under the presumption that the data are not anthropologically relevant. Using a case study from the Coso Volcanic Field, which has at least four distinct subsignatures, we examine this problem and conclude that subsource identification can be useful and quite interesting. This is particularly so when large datasets encompassing spatially expansive areas can be assembled and statistically analyzed.

Introduction

As chemical analytical techniques have improved in the natural sciences, the quality and quantity of archaeological data resulting from their application has increased exponentially. Obsidian sourcing, or fingerprinting, has greatly advanced our understanding and knowledge of obsidian use in western North America. For example, in eastern California this technique is regularly used to reconstruct prehistoric patterns in exchange, mobility, and land use, among other topics (e.g., [1], [3], [4], [10], [13]). Recent analytical advancements now allow archaeologists to even recognize related but discrete signatures, or subsources, of obsidian derived from distinct source zones (e.g., [7], [15], [16], [19], [21]). While possible analytically, the interpretive value of identifying such subsources has not been explored.

Since at least the 1930s [12], the Coso Volcanic Fields have been recognized as a regionally important source of obsidian by archaeologists. While the pioneers Jack and Carmichael [17], [18]recognized a single discrete chemical source, in more recent studies others have noticed the presence of distinct sub-signatures ([11], [14]; see [6]for reservations and potential problems). Although more chemically similar to one another than to other obsidian source zones, such as Casa Diablo or Fish Springs, these subsources can be differentiated using most modern analytical techniques, such as X-Ray Fluorescence (XRF) and Instrumental Neutron Activation Analysis (INAA). To date, four main Coso sources are consistently recognized by analysts, including Sugarloaf Mountain, West Sugarloaf Mountain, Joshua Ridge, and West Cactus Peak, though others may be present as well. These subsources represent different flows of obsidian (Fig. 1shows the location of the study region and the different Coso “source zones”).

It was originally thought that differentiating subsources might be useful in hydration studies since it is known that different sources hydrate at differentrates [11], [14]. Recently Gilreath and Hildebrandt [13]have suggested that the West Sugarloaf and Sugarloaf Mountain subsources hydrate at the same rate (see also [22], [23]). Analyses presented in Table 1, which shows hydration data for different temporally diagnostic projectile point types collected in the China Lake region (see [5]), support this result and suggest that all four subsources hydrate at a similar rate. There are no statistically significant differences in the average hydration readings for temporally diagnostic projectile points, though the sample sizes are small for Joshua Ridge and West Cactus Peak. Thus, in our estimation, differentiating subsources does not assist or refine our ability to tell time with Coso obsidian artifacts.

Thus, while the Coso subsources can be chemically differentiated, a more important question involves the significance of doing so. In other words, does it really matter whether a piece of obsidian comes from West Sugarloaf instead of Sugarloaf Mountain? Does the effort expended on systematically identifying subources tell us anything interesting about prehistoric behavior that we did not already know? As discussed below, we believe the answer to this question is yes.