Graduation Year


Document Type




Degree Granting Department


Major Professor

Brent R. Weisman

Committee Member

Nancy Marie White, Ph.D.

Committee Member

Thomas J. Pluckhahn, Ph.D.

Committee Member

Beverly G. Ward, Ph.D.

Committee Member

Steven Reader, Ph.D.


chert provenience, cost paths, geographic information systems, Middle Woodland, use-wear, weights of evidence


The research undertaken in this dissertation was designed to explore how the institutionalized social inequalities in prehistoric Woodland society are reflected in the differences in the procurement, in the life history, and the final discard locations of the flaked chert stone tools from the Crystal River site (8CI1). The Woodland period (1000 BC to AD 1000) was a time of both stability and change in Native American society. Many of the core institutions such as subsistence, hunting and ceramic technology, and residence remained relatively constant while religious and political institutions underwent dramatic changes. This study focuses on how these social inequalities were manifested in the chipped stone tool assemblage from this site.

The Crystal River site is an Early to Middle Woodland-period mound complex located in coastal Citrus County, Florida. Dedicated as a National Historic Landmark site in 1991, the Crystal River site is internationally known and respected. Despite extensive work at the site conducted by Bullen and others during the 1940-60s, little was actually published about the material remains excavated from the site. Work resumed on the site in the 1980s and has continued as required by park maintenance and repair issues. Since 2007, remote sensing and other non-invasive technologies have been employed to advance research further at the site. This research returned to the flaked stone materials recovered during the periods 1903-1964 and 1984-2001 to illuminate site activities better without additional ground-disturbing activities.

Multiple techniques were employed to develop the data sets that were used to investigate the research questions addressed in this study. The GIS-based weights-of-evidence procedure was used to predict the locations of chert outcrops within a 50 km study area. This model validated the existing quarry cluster method of determining the provenience of Florida cherts. A cost-path analysis was used to identify those chert sources that would have been most accessible to the site's inhabitants. These techniques defined a series of coastal chert outcrops that form the newly-proposed New Coastal quarry cluster.

A chaîne opératoire or operational sequence approach was adopted for the analysis of the chipped stone assemblage. A waste flake analysis, a hafted biface classification, and a raw material provenience classification were conducted for all flaked-stone materials. Use-wear determinations were made using both low-power (10-70x) and high-power (50-400x) magnification analysis techniques. A life history approach was taken to the hafted biface assemblage and hafted biface retouch index (HRI) values were determined for all hafted bfaces and biface fragments.

The provenience analysis demonstrated that the majority of the chert used by the inhabitants of Crystal River came from outcrops and quarries south of the site along the coastal marshes and the western margins of the Brooksville Ridge. These resources are all within a short canoe trip from the site. Two life history trajectories are suggested for the chipped stone tools from Crystal River. The majority of the chert was obtained from local sources. The second life history was defined for a small subset of the hafted bifaces that were transported from quarries located outside the core subsistence catchment of Crystal River site.

Four research hypotheses were developed to test propositions related to the ways in which institutionalized social inequalities are reflected in the patterning of the chipped stone artifact assemblage at the Crystal River site. Although only some of these hypotheses were supported, the results of this investigation do support much of the research that has previously been conducted with the lithic assemblages from Woodland mound complexes in Florida. Chert acquisition is heavily reliant on local lithic sources. Chert procurement appears to be embedded in the collection of other resources. Stone tool use at the site follows the typical expedient flake tool/local raw material pattern that has been documented for other Middle Woodland sites in the region. There was no evidence to suggest that thermal alteration was used to enhance the quality of either the local cherts or those brought to the site from more distant sources. The analysis identified two distinct life histories for at least part of the stone tool assemblage. Many of the hafted bifaces, formed tools and flake tools recovered from the site were made from local cherts. These tool where likely made, used, and discarded at Crystal River. Some of the hafted bifaces and flake cores were made from cherts found on the outer edges of the 50 km study area defined for this investigation. These items were brought to the Crystal River site, used, resharpened, and broken in transit, and finally replaced by new tools at the site. The broken fragments of these tools were discarded in the midden debris to eventually become part of the archaeological record from this now-famous site.