The nature of initial human settlement in the Southeast is largely an enigma at present. Assemblages that appear to predate 13,450 B.P. have been found at a number of locations, such as Cactus Hill (44SX202), Little Salt Spring (8So18), Saltville (44SM37), Page-Ladson (8Je591), and Topper (38AL23). Unfortunately, the dating of some of these sites, and in some cases the artifacts themselves, is somewhat equivocal at present (although, as we shall see below, recent evidence from three of them, Cactus Hill, Topper, and Saltville, increasingly supports an early dating). While there are thus tantalizing hints of early human occupation in the region, well dated sites with extensive artifact assemblages have yet to be found [with the possible exception of Cactus Hill, where work is ongoing, and where increasing evidence suggests a very early occupation (McAvoy et al. 2000)]. No diagnostic artifacts are currently known that unambiguously identify pre-11,500 rcbp (13,450 B.P.) assemblages in the Southeast and, indeed, few sites have been excavated that are widely accepted as dating to this period. Our sample of sites of this time period is thus in no way representative or complete, even to major type.
That human colonization and settlement of the Southeast occurred prior to 11,500 rcbp, however, must be considered possible, given the general acceptance of the dating of the Monte Verde site in Chile at ca. 12,000 to 12,500 rcbp (ca. 14,000 to 14,750 B.P.) (Bonnichsen and Turnmire 1999; Dillehay 1989, 1997; Meltzer et al. 1997). If people could have reached the southern cone of South America by some time prior to 12,000 rcbp, they probably could just as easily have reached the southeastern United States by this time or soon thereafter. Whether they were successful and survived, or died out (i.e., representing "failed migrations"), is currently unknown. The spotty nature of the archaeological record from this era, over both time and space, suggests the latter, or else extremely small populations, possibly occupying portions of the region now largely inaccessible, such as on the continental shelf.
When the first people arrived in the Southeast remains unknown, although a number of sites have yielded evidence suggesting initial occupation might have begun up to several thousand years prior to 11,500 rcbp (13,450 B.P.). About this time or shortly thereafter, however, Clovis assemblages occur widely across the region. Whether fluted point assemblages were present prior to this is currently unknown, although some data from the Southeast hints at such a possibility. Traditionally, sites occurring prior to the widespread appearance of fluted points are called "pre-Clovis," a term that can continue to be used quite effectively to describe possible pre-11,500 rcbp occupations in the Southeast, at least until specific assemblages or artifact categories can be recognized and named. While diagnostics remain elusive, there are indications at sites like Cactus Hill in Virginia and Meadowcroft in Pennsylvania that large and small blades, and possibly triangular and lanceolate point forms, may come to be recognized as a diagnostic indicators of extremely early, pre-Clovis occupations (Adovasio et al. 1999:427-428; McAvoy and McAvoy 1997).
In light of the paradigm shift represented by the widespread acceptance of the antiquity of the Monte Verde site, a reevaluation of early assemblages should be made in the Southeast. Thus, while controversy has surrounded the dating of the Meadowcroft rock shelter deposits from southwestern Pennsylvania (e.g., Haynes 1992:367, but see Goldberg and Arpin 1999:340 and Adovasio et al. 1999, who appear to have effectively refuted arguments against the dating), more attention should be paid to the stratigraphic relationships evidenced at the site, and the viability of the Miller Lanceolate as a possible pre-Clovis point type (Adovasio et al. 1978, 1990, 1999; this is particularly important in light of the recent discovery of similarly unfluted "Early Triangular" points and a blade industry in apparent pre-Clovis context at Cactus Hill in Virginia). Equally important, increased effort should be made to look for early assemblages in stratigraphic contexts that are traditionally ignored in the Southeast, such as below late Pleistocene/early Holocene alluvium or colluvium (Goodyear 1999a, n.d.). Direct physical examination of individual specimens should also help resolve questions of their antiquity. Thus, the Natchez pelvis, which was found with Late Pleistocene megafaunal remains and initially thought on the basis of fluorine testing to have great antiquity, was recently AMS dated to 5580±80 rcbp (AA-4051) (Cotter 1991; M. Smith 1993:63). A goal of all such work should be the development of criteria for the relatively easy recognition of early, pre-Clovis age Paleoindian assemblages in the Southeast.