A TYPOLOGY OF FLUTED POINTS FROM FLORIDA DAVID K. THULMAN Department of Anthropology, George Washington University, Washington, D.C. 20052 Email: dthulman@gmail.com VOL. 60 (4) THE FLORIDA ANTHROPOLOGIST DECEMBER 2007 Florida’s Paleoindian Period is receiving increasing scrutiny and interest in recent years (Hemmings 1999; Thulman 2006b; Dunbar 2007; Dunbar and Vojnovski 2007; Faught 2004), but the effort is hampered by the lack of intact sites of undisputed Paleoindian age and an inventory of Early, Middle, and Late Paleoindian toolkits from Florida. Most of what we think we know has been inferred from out of context artifacts or by analogy to other areas of North America (Thulman 2006a). One area that has undergone recent reevaluation is the typology of Paleoindian points 1 in Florida (Dunbar and Hemmings 2004; Farr 2006; Schroeder 2002), which builds on the earlier comprehensive typology of Florida points by Ripley Bullen (1975). Efforts here focus primarily on fluted, lanceolate points with ground basal edges, which were presumably all made during the Late Pleistocene period. The purpose of this article is to create some hypothetical groups of fluted point forms from which we can model cultural chronology and continuity during this period in Florida. Bullen (1975) identified two fluted forms: the Clovis, which is always fluted, and the Simpson, which is occasionally fluted. In their recent publication, Dunbar and Hemmings (2004), looking more closely at the variation in Florida lanceolates, added a Waisted Clovis form that is always fluted and included the Suwannee, which is occasionally fluted. Farr (2006) and others recognized that fluted forms from outside Florida are also occasionally found here, such as the Redstone, Quads, and Dalton points. Regardless, all these researchers would probably agree that Clovis comes in two varieties – waisted and straight-sided (or slightly excurvate) – and Suwannee and Simpson points are occasionally fluted. Two major groupings of points were identified in this analysis from a dataset of 72 fluted, basally-ground lanceolate points from Florida: Straight-sided and Waisted. The straight-sided group includes narrow, medium, and deep- based subgroups, and the waisted group includes narrow and spatulate subgroups. The details of these categories are described below. These groups were discerned using both qualitative and quantitative data on the size and ratios of attributes of the points, such as minimum width or the ratio of minimum width to maximum width, hafting technique, and resharpening trajectory. Like Dunbar and Hemmings (2004), this article avoids the word “type,” which implies a degree of finality, and uses “group,” “subgroup,” and “form” to emphasize the hypothetical nature of this typology. It also avoids using formalized and loaded terms such as “Clovis” in these descriptions. I do not assert that these subgoups necessarily represent the types that would have been recognized by Florida’s Paleoindians, only that these are reasonable groupings of these data. My approach to typology assumes that the Paleoindians made these points based on a particular culturally-appropriate mental model or template (i.e., the artifact design) and a degree of acceptable variation in mind, which would tend to produce highly patterned distributions of points that vary around a central tendency. These mental models would have included all aspects of the manufacture, use, resharpening, and ultimate discard of the point, including how the point was shaped, flaked, fluted, and hafted. Only points that retained intact bases and flutes were used in this analysis, because these would retain the original mental model. It is assumed that over time, point forms (and any other artifact form) change as people make changes to the mental model. Typically, changes are made when the maker perceives that a change to the design is needed, probably a technical challenge that needs to be resolved, such as the haft does not fit well on the shaft. 2 These changes are usually made to relatively small aspects of the artifact design, such as the size the ears on the base or the ratio of the width of the base to the width of the waist. Over time these design changes are compounded, and distinct artifact forms are produced by different groups of people that archaeologists perceive as regional traditions (Thulman 2006b). With enough examples in proper chronological context, the evolution of a form could be reconstructed. Without these examples, we can nonetheless posit continuity between “types” by speculating how the form changed. The Data The data used in this typology were gathered as part of a larger effort to collect and preserve information about the distribution of Paleoindian points in Florida, which are defined as lanceolate-shaped, bifacial, chipped stone artifacts with ground bases. Virtually all of the data were gathered from private collectors, who generously made their collections available to me. In total, nearly 1000 scanned images of Early, Middle, and Late Paleoindian points from north-central Florida were collected. Several attributes were measured with the digital measuring tool in Photoshop (Figure 1), and the data were recorded for statistical analysis. In addition to the scans, data on thickness, grinding, and location were also collected. Seventy-two of the fluted points were used to develop this typology. All of the fluted points were collected as individual specimens, which means they did not come from an undisturbed stratigraphic context. The data only includes finished artifacts, which do not provide the kind of