The research themes described below offer guidance by which southeastern Paleoindian sites may be found, examined, and evaluated for NRHP and NHL status. They discuss the major research questions facing Paleoindian researchers in the region, and describe the kinds of information needed to answer these questions. What follows is intended to complement, and supplement, the discussion of research themes presented in the national context. Many of the questions and approaches raised in what follows, it should be noted, are derived from existing southeastern Paleoindian historic context studies. Some overlap between specific themes occurs, since many of the processes involved are interrelated.
The importance of the Folsom discovery for southeastern archaeology, accordingly, lay in the fact that the projectile points found with the extinct bison had a distinctive basal flaking pattern, or fluting, that made this kind of artifact an unambiguous marker of an late Pleistocene age site. Prior to this time, of course, fluted points had been found in various parts of the Southeast, but their great antiquity was not recognized. Fluted or more accurately, basally thinned lanceolate points were, in fact, advanced as a distinctive artifact type in Mississippi, where they were described as Coldwater points in Calvin Brown's (1926:132-134) volume The Archaeology of Mississippi. The Coldwater type is used to describe a distinctive, presumably Late Paleoindian lanceolate point found in northeast Mississippi and adjacent parts of Arkansas, it should not, however, be used to describe local fluted points (e.g., McGahey 1981, 1996:354; McGahey: personal communication 2000).
By the mid-1930s fluted projectile points were being recognized in a number of parts of the Southeast, and interpreted as artifacts left behind by early occupants. Most of these points were surface finds, and the papers describing them (e.g., Bushnell 1935:35, 1940; Wauchope 1939) were the first of what has proven to be an increasingly extensive and important descriptive literature on the regional Paleoindian projectile point record (as summarized in Anderson 1990a, 1991; Anderson and Faught 1998, 2000; Anderson and Sassaman, eds., 1996; Brennan 1982; Mason 1962; and Williams and Stoltman 1965). The vast majority of fluted points found to this day in the Southeast, in fact, are from surface context, and having good locational and descriptive information about them is essential to understanding variability within these early occupations.
For the first two to three decades after 1926, fluted points in the Southeast tended to be described as Folsomoid or Folsom-like by local archaeologists. By the 1950s, southeastern fluted points tended to be referred to as Clovis or Clovis-like, since most eastern variants lacked the full length fluting and fine retouch characteristic of true Plains Folsom points. It was about this time as well that a number of distinctive subregional variants began to be named, such as the Cumberland, Redstone, and Wheeler types (e.g., Cambron 1955; Cambron and Hulse 1964:30, 99; Kneberg 1956; Lewis 1954:7). In recent years, as data on thousands of points has accumulated from across the region, appreciation for the range of variation evident has increased, although efforts to untangle it are still in their infancy. Given the thousands of points from the region that have been measured, in fact, it is surprising how little analysis has actually been done with this primary attribute data from a regional scale (Meltzer 1984, and Morrow and Morrow 1999 are important exceptions), although appreciable excellent work has done as part of site, locality, or state level analyses (e.g., Anderson et al. 1990; Breitburg and Broster 1995; Daniel 1998; Daniel and Wisenbaker 1987; Goodyear 1974; Morse, ed., 1997; to cite a few examples).
Following the Folsom discovery, some Paleoindian materials were found comparatively early on in secure excavation context in the Southeast. An early stone tool industry characterized by heavily weathered chert was found during the New Deal era excavations at Macon Plateau. While numerous points, tools and debitage dating to the later Paleoindian and Early Archaic Dalton, Taylor/Bolen, and Palmer/Kirk complexes were found, only a single fluted point was recovered, in 1935 (Kelly 1938; Waring 1968:237). A more extensive Paleoindian assemblage, characterized by eight fluted points and some 280 unifacial tools, was found in surface and excavation context at the Parrish Village site (15HK45) in western Kentucky, excavated by WPA crews from 1938 to 1940 (Freeman et al. 1996:395-396; Rolingson and Schwartz 1966; Tankersley 1990a; Webb 1951). These excavations remain among the most extensive ever undertaken in the Southeast that had as a goal the recovery of Paleoindian materials.
By the late 1940s, excavations began to occur in the region's rockshelters and floodplains, with the goal of locating deeply stratified deposits, and hence the recovery of assemblages dating to specific and progressively more ancient periods. Important work was done at sites such as Hardaway in North Carolina (Coe 1964; Daniel 1998), Russell Cave in Alabama (Griffin 1974; Miller 1956); and the Stanfield-Worley Bluff shelter in Alabama (DeJarnette et al. 1962), and at similar sites in adjacent areas, such as at the Modoc Rock Shelter in Illinois (S. R. Ahler 1993; Fowler 1959). This work accelerated in the ensuing decades, particularly with the rise of CRM archaeology in the 1970s. Important research directed to or documenting Paleoindian components has occurred in floodplain settings in the Little Tennessee River of eastern Tennessee (Chapman 1985), the Haw River of North Carolina (Claggett and Cable 1982), and in rockshelter deposits at Dust Cave in Alabama (Driskell 1996). Once again, projects in nearby areas, such as at Koster in Illinois (Struever and Holton 1979), St. Albans in West Virginia (Broyles 1966, 1971), and Rodgers Shelter in Missouri (S. A. Ahler 1971; McMillan 1971; Wood and McMillan 1976) have contributed markedly to our understanding of late Pleistocene or immediate post-Pleistocene, Holocene occupations in the Southeast.
Also beginning in the late 1940s, with McCary's (1984, 1991; Hranicky 1989) pioneering work in Virginia, the systematic recording of all known fluted projectile points began to occur in various parts of the region. This work, typically conducted within individual states by avocational and professional archaeologists working in tandem, has accelerated markedly within the past two decades. Fluted point recording projects are currently ongoing in almost every state in the Southeast. At present, almost 6000 Paleoindian points have been documented in the region by these surveys (Table 4), and detailed measurement data has been recorded for over two thirds of this sample, and is typically readily available from the researchers generating it in each state.
Since the early 1970s, extensive survey and excavation work has been conducted across the Southeast under the mandate of federal environmental legislation. Millions of acres of land have been examined, and almost 200,000 archaeological sites have been recorded (Anderson and Horak, eds., 1995). A great many Paleoindian sites have been found and excavated and, perhaps most importantly, detailed technical reports on the work have been prepared. Crucial to effective cultural resource management archaeology is making determinations of site significance, that is, knowing how a site can provide information that can help us better understand the past. The NHL research themes discussed below offer a framework by which significance may be evaluated, by suggesting specific areas where research is needed.
Any artifact-rich site securely dated to more than ca. 11,500 rcbp/13,450 B.P., and ideally yielding unequivocal diagnostics, will be crucial to addressing this theme. As such, properties like Cactus Hill, Page-Ladson, Saltville, and Topper, if convincingly shown to be pre-Clovis in age and culture, will be of profound importance, and unquestionably eligible for NRHP and NHL status. Of critical importance to exploring the theme of Peopling Places, therefore, is determining "What constitutes early Paleoindian sites in the Southeast?" A related question would be "What constitutes later Paleoindian sites in the Southeast?"
To better understand initial occupations, we also need to know what specific landforms, sediment types, and microenvironmental settings were used by Paleoindian populations in the Southeast. Are such settings sufficiently distinct or unusual that they can be used to predict the probability of finding early materials? What field methods are appropriate for these settings, to maximize the possibility of discovering and evaluating early components? Geoarchaeological research in the Southeast in recent years has shown that Paleoindian sites may be found in specific floodplain terrace settings, around Carolina Bays, and in areas of extensive colluviation (e.g., Brooks and Sassaman 1990; Brooks et al. 1996; Goodyear 1999a). Site survey data has been used to resolve areas favored by early populations as well, such as in the vicinity of prominent shoals (O'Steen et al. 1986). Fall Line locations across the Southeast have long been known to possess major Paleoindian assemblages, as have areas around chert or other high quality stone outcrops. As Paleoindian components are identified in the Southeast, GIS technology has proven a highly effective means of evaluating and refining Paleoindian settlement models, as demonstrated in northeast Arkansas (Gillam 1996a, 1996b, 1999).
Tesar (1996:27-34) has provided an excellent overview of where sites throughout the Paleoindian era are likely to be located in the Florida area (see also Dunbar 1991). Of particular interest, Tesar (2000:28) also suggests that "in marine and estuarine settings shell offered an alternative cutting tool with which to fashion bone and wooden artifacts. The biodegrading of organic based artifacts has favored an interpretive bias focused on stone artifacts." Early coastal sites may thus be particularly difficult to discern, if lithics were not as commonly utilized.
How the settlement of the continent may have proceeded, given the changes in vegetation that were occurring at the regional scale, has also been the subject of some fairly extensive modeling in recent years (Steele et al. 1998). In brief, the early emergence of hardwood forests across much of the Southeast would have made the area attractive to early populations, something clearly indicated by the fluted point distributions. Such analyses can also be conducted at the regional and locality scale, as Gillam's (1996a, 1996b) research in northeast Arkansas has shown, where population distributions over much of the Paleoindian period was shown to be, at least in part, constrained by lithic raw material locations.
Creation of Social Institutions
Where distinctive Paleoindian artifact categories exist, such as the Cumberland type, typological classification and distributional analyses are relatively easy to accomplish. A bewildering array of Paleoindian projectile point types and variants, not all mutually exclusive, have been defined in the Southeast. Almost all of these, furthermore, are intuitively rather than quantitatively based, and extremely difficult for differing researchers to consistently sort (Morse 1997:134). Collections analysis should help standardize our systematics, and help resolve identifiable tool forms, whose age can then be determined through excavation.
While we have a good handle on the geographic distribution of Paleoindian fluted point types, since most state surveys record these artifact categories, our information on the occurrence of later Paleoindian point forms, particularly Daltons and side notched forms, is much more spotty. Locational and measurement data are not systematically recorded for these point types in many southeastern states. This is because, quite simply, so many Dalton and side notched points are known to exist in some areas that the task of recording them appears quite overwhelming, with the result that effort is typically directed to earlier or less common artifact types, like fluted points, or fluted and unfluted lanceolates. This selective approach to data recording must change if we are to understand what the stylistic variation observed within Dalton or the early side notched horizon actually means. If we include some of the San Patrice varieties, there are currently over a dozen named Dalton types and variants (Ensor 1987; Goodyear 1974, 1982; Justice 1987; Morse 1997). Likewise, while Dalton points are described as common in many areas, at least compared with presumably earlier Paleoindian forms, because primary data has not been systematically collected, exactly how common they are, or where they occur, is not well known. Other than in northeast Arkansas, where the distribution of Dalton assemblages have been explored for decades (e.g., Gillam 1996a, 1996b, 1999; Morse 1971b, 1973, 1997a; Redfield 1971), we currently know less about the geographic occurrence of Dalton points in the Southeast than we do about fluted points. The same is also true about our knowledge of the occurrence of side notched points.
Expressing Cultural Values
Likewise, what role did specific artifact categories play in defining cultural identity? There is no question that Paleoindian populations in the Southeast placed great value on their tools of stone, bone, and ivory. The workmanship on many specimens is superb, reflecting a level of expertise rarely achieved by the flintknappers of subsequent periods. The aesthetic appeal of these artifacts to modern archaeologists and collectors alike no doubt helps to explain the widespread interest in these early peoples. Concern for exemplary craftsmanship was a major and widely shared Paleoindian cultural value, even in the production of everyday stone tools. Was the Paleoindian fascination with high quality lithic raw materials, accordingly, solely due to the needs of a highly curated toolkit, as Goodyear (1979, 1989) has argued? Or was it also shaped by the ceremonial potential of artifacts made from these materials, as exemplified by the presence of elaborate specimens in caches and burials, or their role in facilitating ceremony and interaction, as Walthall and Koldehoff (1998) have suggested for Sloan points in the central Mississippi Valley, markers of what they call a possible "Cult of the Long Blade?" Visiting quarry areas, accordingly, may have been as much about promoting interaction as about procuring high quality stone, if groups knew they could find others at these locations at certain times of the year (Daniel 1998:194-195; 2001). Likewise, by procuring high quality stone, and using it to make elegant tools, these peoples were reinforcing a basic aspect of their culture.
How Paleoindian use of the landscape shaped their culture is a research topic to explore under the theme "Expressing Cultural Values." The distribution of Paleoindian sites and assemblages can be examined from such a landscape perspective, much in the way predictive models of site location are developed. Gillam (1996a, 1996b, 1999), for example, has shown that settlement changed in northeast Arkansas over the course of the Paleoindian era, with later sites found in a much wider array of settings. Some aspects of land use continued more or less unchanged, however, since both fluted point and Dalton assemblages indicate extensive use of the same lithic raw material sources.
Evidence for Paleoindian mortuary behavior, another aspect of how people express their cultural values, is rare in the Southeast. The Sloan site is a spectacular exception, albeit one with minimal skeletal preservation. The earliest well preserved human remains from the region are from Warm Mineral Springs site in Florida, dated to between ca. 9000 and 10,000 rcbp (Cockrell and Murphy 1978), although whether these are intentional burials is unknown. Excavations at a number of wet sites in Florida, such as Windover (Doran and Dickel 1988; Doran et al. 1986), have found numerous well-preserved human remains, indicating that cemetery behavior dates at least as far back as the end of the Early Archaic period. The Sloan data indicates collective burial practices extend even further back in time, albeit in this case in a dry rather than wet or submerged settings. What the use of cemeteries implies in terms of group residential permanence in an area, territoriality, kin groupings, and so on are all subjects that must be considered. In a related manner, are there Clovis age ceremonial or burial artifact caches like Richey-Roberts/East Wenatchee (Gramly 1993) or Anzik (Lahren and Bonnichsen 1974) present in the Southeast? The answer is almost certainly yes, given their presence in later Dalton times at Sloan and other locations in the central Mississippi Valley (Walthall and Koldehoff 1998). Do additional Dalton cemeteries like Sloan exist awaiting discovery? Again almost certainly. We need to recognize that if such sites are found, they must be preserved and protected from looters or, if protection is not an option, quickly and carefully excavated.
Shaping the Political Landscape
Another topic that can be explored under the theme "Shaping the Political Landscape" is whether and why some portions of the Southeast were more heavily occupied during some Paleoindian periods than during others. There is no question that fluted point concentrations are almost invariably located along major transportation arteries, notably along or near major river channels, and in areas rich in floral, faunal, and lithic resources (Figure 3). Why were these areas more favored than others during this time? Likewise, why were some areas, such as the Appalachian highlands and portions of the Gulf Coastal Plain, minimally visited by peoples using fluted point technology? Preferences for certain landform types are also indicated during later Paleoindian times as well, at least within specific intensively examined localities (e.g., Gillam 1996a, 1996b; O'Steen et al. 1986). Unfortunately, the distribution of most Paleoindian point types at a larger, regional scale is largely unknown at present. Our understanding of where Dalton occupations or sites with side notched points occur across the region, for example, remains intuitively based (e.g., Justice 1987; Morse 1997), even though these are the most common Paleoindian point types known. When these artifact types can be mapped the way we can now plot fluted points, no doubt concentrations and voids will be found within the regional landscape that will profoundly influence our understanding of these occupations.
It is also important to ask how regional physiography, specifically the orientation of river drainages and the location of mountain ranges and shorelines, may have shaped group movement, interaction, and the rise of subregional cultural traditions during the Paleoindian (Anderson and Gillam 2000, 2001). Greater movement and interaction would have clearly been more likely in some directions than in others, and perhaps at some times rather than others (Anderson 1995a; Meltzer n.d.). Little interaction or movement, for example, might be expected across the Appalachian mountains, or between groups occupying the Atlantic and the Gulf Coasts, save in intermediate areas. The Feronia locality in south Georgia, at the interface between the Atlantic and Gulf coastal watersheds, is in an area ideally suited for interaction between Paleoindian groups occupying these two major regions. The presence of extensive Paleoindian remains in this area, some made on raw materials that come from appreciable distances, suggests such interaction actually occurred (Blanton and Snow 1986). Likewise, the restricted distributions of Cumberland points primarily to within the Tennessee and Cumberland river drainages, and Suwannee/Simpson points primarily to within the karst rich areas of the Florida peninsula (with an extension up the Atlantic Coastal Plain as far as southwestern South Carolina), suggests the peoples making these artifacts ranged within these areas and no farther. Within these habitual use areas, where do sites with these diagnostics themselves occur, and how does this compare to the occurrence of sites in other areas? Are distinctive subregional adaptations forming or present? Why do some point types occur primarily within areas no more than a few hundred kilometers in maximum extent, while other point types appear to occur much more widely across the region?
Finally, diachronic analyses will prove an increasingly important means of resolving how and why the social landscape evolved during the Paleoindian period. As noted previously, changes over time in artifact distributions and raw material fall off curves have been used to explore changes in group range and settlement organization over the course of the Archaic period (e.g., Sassaman et al. 1988). Similar approaches can be attempted during the Paleoindian era, once our temporal resolution improves. Related to this, we need to ask why fluted point assemblages continue fairly late in the Northeast, to perhaps as late as ca. 10,200 rcbp/11,850 B.P. (e.g., Bonnichsen and Will 1999; Lepper and Meltzer 1991; Levine 1990; Spiess et al.1998), when fluted points are clearly gone from across the Southeast by this time?
Developing the American Economy
Raw material source analyses comprise another important area for research. Where did the materials found on Paleoindian sites originate? Appreciable research has been directed to delimiting lithic raw material sources in the Southeast using trace element, petrographic, or fossil microfauna data (e.g., Anderson et al. 1982; Banks 1990; Daniel and Butler 1991; Goodyear and Charles 1984; Upchurch 1984). Such analyses are critical, since many materials are difficult or impossible to distinguish macroscopically. In Georgia, for example, there are cherts in the Piedmont that are identical in appearance to cherts from the Coastal Plain over 100 km away (Ledbetter et al. 1981). The Piedmont cherts lack the microfossil inclusions ubiquitous in cherts from the Coastal Plain, however, and a microscope is needed to differentiate the materials.
Expanding Science and Technology
Over the course of the later Paleoindian, point manufacture in the Southeast appears to have proceeded from predominantly fluted to basally thinned to unfluted types, and from straight sided lanceolates to broad, recurvate forms, and then to more straight sided or triangular beveled and increasingly more pronounced notched forms (Gardner 1974:18, 1989; Gardner and Verrey 1979; Goodyear et al. 1979:90-96; Morse and Morse 1983:60-65). Such patterning, if it can be refined through stratigraphic and absolute dating, can be used to infer the relative ages of specific points and assemblages. Throughout the Southeast it appears that a waisted lanceolate tradition-characterized by Cumberland, Quad, Beaver Lake, and Suwannee/Simpson types-was replaced by the lanceolate, lightly shouldered and notched Dalton tradition sometime around or after ca. 12,500 B.P./10,5000 rcbp. This trend is indicated everywhere save in Florida, where waisted Suwannee/Simpson forms are thought to have continued in use until they were replaced by side notched types. Throughout the region, these side-notched forms replaced the (for the most part unnotched) lanceolates prevalent for the preceding two millennia.
The dramatic increase in sites and assemblages over the course of the Paleoindian and into the Early Archaic periods in the region has been used to infer rapid population growth and landscape filling (Anderson 1990a, 1996; McAvoy 1992:157-163). This would have lead to increasing pressure on resources, and the need for new methods by which groups could exploit them. Later Paleoindian projectile points retain many characteristics of earlier assemblages, but they also evince evidence for extensive resharpening, suggesting a major difference in the use of these bifaces when compared with earlier points. This has been attributed to a need to kill and process large numbers of comparatively smaller animals (at least when compared with megafauna), such as white tailed deer (Goodyear 1974:14, 103, 1982; Michie 1973; Morse 1971b, 1973, 1975b, 1997b). This assumption, of course, leaves unanswered how earlier populations performed similar cutting and butchering tasks, whether megafauna were indeed their prey of choice, and whether the extinction of megafauna actually contributed to changes in these peoples toolkits. The gradual abandonment of the highly curated Paleoindian toolkit is believed to be directly related to emergence and increasing importance of foraging, generalist strategies over the region during the later Paleoindian and Early Archaic periods (Cable 1982a, 1996; Meltzer 1984, 1988; Meltzer and Smith 1986; Morse 1973, 1975b, 1997b). Is this view actually correct and, if so, how did the process proceed?
Transforming the Environment
How did changes in shoreline, particularly the fluctuations that occurred between ca 13,000 and 10,000 rcbp/15,630 to 10,450 B.P., and particularly changes that may have occurred during the Younger Dryas between 10,800 to 10,100 rcbp/12,900 to 11,650 B.P., affect Paleoindian adaptations? What effect did changes in vegetational communities have on human and animal populations, such as the emergence of a hardwood canopy over much of the region after ca. 11,000 rcbp/13,000 B.P.? Was there a change in technological organization, from logistically organized collectors to residentially mobile foragers, as has been suggested by some investigators (Anderson et al. 1996:6-7; Cable 1982a, 1996; see also Binford 1980)? Were residentially mobile foraging populations present at an earlier period in the lower Southeast, south of latitude 33° N, which appears to have been covered by a fairly homogeneous hardwood forest over much of the late Pleistocene (excluding Florida, which appears to have had fairly xeric plant communities away from streams) (Delcourt and Delcourt 1981, 1987; Jacobson et al. 1987; Watts et al. 1996; Webb et al. 1993)? Is the scarcity of Clovis sites in the Gulf Coastal Plain due to the fact that the hardwood canopy assumed to have been present in this area in the Late Pleistocene (Webb et al. 1993:448-450) was not particularly attractive to these peoples? Or are sites present, but just masked in some fashion? Were Paleoindian peoples exploiting the Coastal Plain living primarily out on the continental shelf, closer to the coastline (Faught 1996; Faught et al. 1992; Tesar 1996:27-34)? Is it realistic to expect that Paleoindian adaptations were the same in the dissicated karst terrain of Florida, on the Atlantic Coastal Plain, or along the major drainages of the Mid-South?
Changing Role of the United States in the World Community
Another important question that can be considered under this theme is the effectiveness of our existing chronologies and culture sequences for the Paleoindian period. The relative temporal placement of Paleoindian and Early Archaic projectile points and associated toolkits has been determined through excavations at stratified alluvial and rockshelter sites. Radiocarbon dates have sometimes provided absolute chronological controls to these sequences, but many of the dates that have been obtained are considered unacceptably early or late, blurring our resolution. Thus, while there is agreement that parallel sided Clovis lanceolates precede the more waisted or eared fluted or nonfluted lanceolates, which were in turn replaced by notched points, even approximate temporal ranges remain to be determined for many of these forms (e.g., see commentary on this problem by Griffin 1977:5; Meltzer 1988:15). While this relative sequence has been documented in stratigraphic columns from a number of locations in and near the Southeast, the only point forms that may be considered fairly well dated in the region are the terminal Paleoindian early side notched types. Even Dalton points, whose temporal range was once thought to be fairly well known, to between ca. 10,500 and 9900 rcbp/12,500 to 11,250 B.P. (Goodyear 1982), appear to date earlier and possibly later than this in parts of the region.
Models of the kind of fieldwork and data needed to answer these kinds of questions exist in the record of past research. Major excavations in the Southeast that have provided stratigraphic columns spanning appreciable portions of the Paleoindian and ensuing Early Archaic periods, and that have had a major role in shaping the culture sequence, include Hester in Mississippi (Brookes 1979); the Haw River and Hardaway sites in North Carolina (Claggett and Cable 1982; Coe 1964; Daniel 1998); Page-Ladson and Silver Springs in Florida (Dunbar et al. 1988, Neill 1958); Stanfield-Worley Bluff Shelter and Dust Cave in Alabama (Driskell 1994; Goldman-Finn and Driskell 1994), to name a few of the more prominent examples. Research at deeply stratified sites in areas just beyond the Southeast, at sites like Rodgers Shelter in Missouri (Wood and McMillan 1976) and St. Albans in West Virginia (Broyles 1966, 1971) have also provided important cultural historical information. When new examples of deeply stratified Paleoindian sites are found in and near the Southeast, like the recently discovered Big Eddy site in southwest Missouri (Lopinot et al., eds., 1998, 2000; Ray et al. 1998), or Wakulla Springs Lodge in Florida (Jones and Tesar 2000), they should receive extensive examination.
At the present, the identification of Paleoindian components across the Southeast suffers from considerable ambiguity, particularly in cases where supposedly well described and dated diagnostics like Clovis or Dalton points are lacking. As we have seen, however, appreciable morphological variation is subsumed under these types, much of which is poorly documented or understood at present. How much of this variation reflects temporal or cultural phenomena, and how much of it is due to constraints imposed by raw material, or the position of the artifact in a use-life cycle? Similar typological ambiguity pervades the use of many other Paleoindian forms, such as the Clovis Variant, Suwannee, Simpson, Quad, Beaver Lake, and Cumberland types. In an outstanding example of the type of research essential to untangling this variability, Breitburg and Broster (1995) examined a sample of 234 Cumberland and 654 Clovis points from Tennessee, explicitly documenting how these types differed from one another, as well as how point size within each category varied in differing parts of the state, something attributed to the distance they were from raw material source areas. Classificatory systems for Paleoindian artifacts in the Southeast should be based on quantitative analyses directed toward resolving and understanding the reasons behind observed morphological variation. While type and variety systems will continue to remain useful tools, such classification must be perceived as only a first, and to some extent limiting (since they constrain the study of variation) step in our analyses of Paleoindian assemblages.