The Coats-Hines site (40WM31) is a potential pre-Clovis site located in Franklin, Tennessee. The site rests, geographically, at the convergence of the Central Basin and Western Highland Rim. The site was discovered during the construction of a nearby golf course when a salvage team uncovered a mature female mastodon.. The site was later excavated in 1994-1995, during which time two additional mastodons were uncovered, in direct association with lithic artifacts. Preliminary radiocarbon dates reveal the site was deposited during the late Pleistocene epoch at roughly 12,000^(14)C yr BP. During the summer of 2012, the site was excavated with the goal of determining the depositional setting of the site and geographic region, as well as establishing the antiquity of the archaeological remains. The site geology was determined through field interpretation and texturing, micromorphological analysis, laboratory particle size analysis, and radiocarbon dating. Sedimentation at the site is a combination of cherty colluvium from upslope as well as alluvium. Four chronostratigraphic sequences of sedimentation were determined to have occurred during the last glacial, the Pleistocene-Holocene transition, the Holocene, and modern time periods. The volume, distribution, and composition of the nine defined stratigraphic units are dependent on the fluctuations occurring in the climate during these time periods. The climate changes and rates of deposition occurring at Coats-Hines were correlated to similar sites in the region. The Coats-Hines site was surveyed along the wet-weather drainage that bounds the site during in the spring of 2013. A channel unconformity was discovered, likely dating to the Pleistocene-Holocene transition and providing context to the 1994/1995 excavation.

This dissertation presents new data on early Paleoindian stone technologies in the Southern Plains periphery. Analyses of lithic artifact assemblages show that significant technological changes occurred between the transitions from pre-Clovis to Clovis and from Clovis to Folsom/Midland. After an initial introduction to the problems in chapter one, a detailed technological description of the pre-Clovis assemblage from the Debra L. Friedkin site, Texas is presented. Site-scale and general technological comparisons to Clovis reveal similarities and differences. I conclude that the pre-Clovis assemblage at Friedkin cannot be considered Clovis, but could represent an ancestral technological assemblage. The third chapter presents the analysis of Clovis bifaces from the Hogeye site, Texas. I identify patterns in the biface reduction process and suggest that these patterns could be use to distinguish between regional Clovis cultural signatures and the idiosyncrasies of individual Clovis flintknappers. The fourth chapter compares Clovis and Folsom/Midland technologies and site-use at a single site, the Debra L. Friedkin site, Texas. I show that while late-stage biface reduction and point production were the focus of both occupations, Folsom/Midland groups also reduced some early- or middle- stage bifacial cores. More broadly, the Friedkin site shows that Clovis and Folsom/Midland settlement along Buttermilk Creek varied. Ultimately, this dissertation provides new evidence of possible Clovis origins, documents Clovis biface reduction signatures, and identifies site-use and technological similarities and differences between Clovis and Folsom/Midland. Defining and comparing early Paleoindian adaptations and technologies is key to understanding how humans dispersed into North America and how they adapted to new and changing environments during the last Ice Age.

The archaeological literature warns against trusting the context of artifacts found within a vertisol due to the constant mixing of sediments caused by the shrink/swell properties of clays. These churning processes were thought to be the defining characteristic of vertisols until only the past few decades. It is now apparent that vertisols vary drastically based on a wide spectrum of variables and are fully capable of forming without churning processes. The Buttermilk Creek Site, Block A represents a prime example of a minimally developed vertisol. In addition, the site itself is a heavily occupied lithic quarry that has been almost continuously inhabited since Clovis and possibly Pre-Clovis times. This thesis takes a detailed look at the sediments and distribution of lithic artifacts from Block A of the Buttermilk Creek site to address the two following research objectives: 1) to determine if the archaeological context within the floodplain sediments at Block A has been disturbed by post-depositional processes, and 2) to identify discrete occupation surfaces within the vertic floodplain sediments at the site. These objectives are addressed using a variety of methods, including: 1) plotting the stratigraphic position of diagnostic artifacts, 2) determining the size distribution of debitage and artifact quantities throughout the floodplain deposits, 3) examining the distribution of cultural versus non-cultural lithic material, 4) recording the presence or absence of heat alteration in the deposits, 5) creating maps showing the degree of fissuring across the site, 6) analyzing differences in patination on artifacts, and 7) analyzing the presence of calcium carbonate on artifacts from all levels. Results from these analyses show that, despite the classification of sediments at Block A as a vertisol, vertical displacement of artifacts is largely absent. Chronologically ordered diagnostic points, consistently size sorted artifacts, and a lack of constant mixing of calcium carbonate throughout the profile suggest that artifacts found as deep as 20 cm below the Clovis-aged horizon represent intact cultural horizons. These oldest components found in Block A may represent some of the earliest known evidence of people in the New World.

This thesis covers more than thirty years of the author's research into the Paleoamerican period of the Middle Atlantic Region of North America, including the last 19+ years of focused work on the Cactus Hill site (44SX202) and replication of the Paleoamerican occupation discovered there. Using a landform and geology based predictive model derived from the Paleoamerican occupation at Cactus Hill, the author directed preliminary archaeological testing in three other areas of the same Nottoway River Valley, where Cactus Hill is located. These areas were the Barr site, located 11 miles (18 km.) downriver from Cactus Hill; the Chub Sandhill Natural Resource Conservation Area, located 19 miles (30 km.) downriver from Cactus Hill; and the Blueberry Hill site (44SX327), located approximately 1,000 feet (300 meters) east of Cactus Hill. The latter two produced OSL dated, pre-Younger-Dryas landforms, as predicted. The Rubis-Pearsall site (44SX360), located in the Chub Sandhill preserve also produced a buried Paleoamerican, Clovis age cultural level confirming the model. In addition to the OSL dates, Blueberry Hill also produced a distinct and apparently discrete activity surface with a possible pre-Clovis age Cactus Hill point at the same depth as the Paleoamerican levels at Cactus Hill and Rubis-Pearsall.