Site Occupation
Occupation at 38HR496 apparently began during the Early Archaic subperiod and continued through the Mississippian period. Further use of the site area continued during the Historic period in the form of farming and logging and in modern times as a spoil dump. Based on the artifact assemblage, the Archaic occupation was intermittent. Evidence of the Archaic occupation includes Early Archaic scrapers, Middle Archaic Morrow Mountain projectile points, and Late Archaic Savannah River projectile points and Thoms Creek ceramics. For the most part, these diagnostic artifacts were recovered at the lowest excavation levels.
The Woodland occupation, however, appears to have been substantial over time. Several of the research questions dealt with the Woodland occupation at 38HR496. This occupation is represented by abundant diagnostic ceramics. However, the post features that were thought to date to the Woodland occupation following testing may actually be associated with the Mississippian occupation. In addition, few other post features were identified and no house patterns could be discerned. Consequently, we cannot advance any definitive data on either Woodland or Mississippian houses. The pit features yielded Woodland dates from charcoal recovered from flotation samples but contained Mississippian ceramics and many contained corn. Although Joseph et al. (1998) warn against downplaying the importance of maize agriculture during the Late Woodland, the corn cupules recovered from one of the pit features was dated to AD 1260 to 1400, firmly within the Mississippian Period.
Ceramic Vessels. The number of vessels identified in the 38HR496 ceramic assemblage increase from the Late Archaic through Middle Woodland period but drops dramatically during the Late Woodland, before peaking during the Mississippian Period. Pee Dee decorative styles account for 48 vessels. If we assume that the presence of more vessels equates to the presence of more people, these analytical results strongly suggest that 38HR496 was most intensively occupied during the Mississippian Period and least intensively occupied during the Late Woodland subperiod. However, this decrease in site occupation in the Late Woodland may not be accurate. The Hanover ceramic series has traditionally been assigned to the Middle Woodland, but the date range of its production has not been firmly established for the northern South Carolina coast. Joseph et al. (1998) conducted extensive research on ceramic wares recovered in the North Coastal Zone. They ultimately place Hanover ware production between AD 200 to 800. As they did not recognize a Mississippian Period for this zone, these dates place Hanover wares in the latter half of the Middle Woodland in their chronology. Since their work, we have confirmed the presence of a formal Mississippian presence in the zone and therefore view the termination of the Woodland Period at around AD 1100. This places the Hanover wares into the Late Woodland period. In North Carolina, Herbert (1999) speculates that Hanover wares were being produced up to AD 1200 in the southern North Carolina coastal area. The continuation of production of Hanover wares into the Mississippian Period cannot be rejected until firmer dates establishing the extensive production range are obtained from preserved contexts with Hanover ceramics.
Mills (1989) speculated on the relationship between vessel assemblages and site occupation periods. She presents several formulae based on vessel class. According to Mills (1989), site occupation of short duration should be reflected in a ceramics assemblage that has little variation in vessel forms. Conversely, occupation of a site of long duration would result in a wide variety of vessel forms and stable proportions of each class (Mills 1989). The 38HR496 vessel assemblage contains only 23 vessels for which form could be distinguished: 15 open mouthed bowls and eight jars. The 38HR496 vessel assemblage has so little variability that, if we used Mills’ hypothesis, we would have to assume that occupations at the site were of short duration. The presence of post and pit features may negate this assumption. Regardless, we can hypothesize that 38HR496 was likely occupied seasonally during the Archaic and Early/Middle Woodland periods and more permanently during the later periods.
Feature Interpretation. Features 606, 609, 617, 621, 628, 629, and 631 are all pits ranging from 75 to 95 cm in diameter and 35 to 53 cm in depth. Each of these features were noticeable in plan view due to soil discolorations but did not have sharp edges. They were all basin-shaped and five of them yielded small amount of corn although the radiocarbon dates obtained from charcoal recovered from feature fill dates these pits to the Middle to Late Woodland periods. A similar feature type was identified at Wachesaw Landing. Trinkley et al. (1983) describe their Feature 1 as measuring 121 cm in diameter and 48 cm in depth. The photographs of their feature look very similar to the five pit features at 38HR496. The Wachesaw Landing Feature 1 had a small corn cob pit along its southern edge and contained cobs in the feature fill as well as trade beads and small animal bones. Feature 10 identified at site 38HR309 was also very similar to the pits excavated at 38HR496. This feature measured approximately 70 cm in diameter and 40 cm in depth. Feature 10 fill contained corn and other botanical remains as well as Middle Woodland vessel fragments (Joseph et al. 1998).
Subsistence
Unfortunately, the degree of preservation of faunal material was poor, hindering our ability to fully examine subsistence practices. However, botanical material indicates exploitation of mast resources (i.e., hickory, acorn, and black walnut). The lack of shell middens suggests that either shellfish were not being exploited or that they were not being transported to the site. However, a wide range of other aquatic and terrestrial resources are abundant in the site vicinity and were likely exploited. The Waccamaw River and Socastee Swamp would have provided numerous fish and turtle species. The uplands would have supported deer, small mammals and edible plants. And we recovered corn, indicating that agriculture was being practiced. The archaeobotanical analyses also identified a number of weedy species that would have grown up in disturbed settings.
Lithic Resources and Tool Production
The lithic artifacts analysis confirmed that this site had retained a degree of stratigraphic integrity. Diagnostic projectile points were recovered at appropriate stratigraphic levels. The most interesting aspect of this analysis is its highlighting of the use of presumably locally available material. These raw materials were collected from riverine sources (i.e. cobble/gravel deposits). Geological and archaeological evidence indicate that these cobble deposits were widely available, although availability of material likely varied over time. Indeed, evidence suggests that different materials may have been available in certain areas over time due to exposure of underlying geologic strata due to the meandering and incising of waterways such as the Waccamaw River. Raw materials may also have washed down the rivers into the Coastal Plain from outcrops located in the Piedmont and along the Fall Line.
Joseph et al. (1998) have suggested knappers became less discriminating regarding the quality of their raw material. At 38HR496, this trend may be reflected by the dramatic decrease in the use of metavolcanic material over time. For flaked artifacts, metavolcanics dropped from a high of 799 in Level 3 of the excavation units to 269 in Level 1. It is most likely that increased sedentism over time necessitated the exploitation of readily available river gravel beds and that the reduction and tool production techniques were modified to accommodate this raw material. To exploit the local lithic resources more efficiently, the inhabitants at 38HR496 adapted their tool making strategy. Bipolar percussion was used, in conjunction with common flint knapping techniques, to create stone tools from the river cobbles. The bipolar technique is represented by the hammerstone/anvils, bipolar cores, and bipolar flakes recovered from the site. Also, tool production appears to have become more expedient over time. A number of researchers have suggested that the biface industry decreased over time (e.g., Jeske 1992) and this view is borne out in the increase in humpback bifaces and expedient flake tools in the 38HR496 lithic assemblage.
Conclusion
Data recovery at 38HR496 has provided a wealth of information about prehistoric life along the Waccamaw River, particularly on the Mississippian occupation. This site is one of the largest Mississippian sites identified to date on the northern South Carolina coast. Considering that a decade ago a Mississippian presence was not believed to have been present, a great deal of research is still needed to fully understand the extent of the Mississippian presence in this area. It is our hope that one day we will have enough data to be able to define the Waccamaw Mississippian chiefdom.
