The early Woodland period in North Carolina prehistory has been robed in mystery and misunderstanding throughout its study. Plagued by early concepts of wholesale cultural replacement (Coe 1964; Caldwell 1958), the study of the origin of Woodland traditions in the southeast has focused on outside influences from the north and west. Moreover, concepts such as "discontinuity" and "cultural hiatus" established by Coe’s work in the North Carolina piedmont have tended to bias our interpretations of the Archaic-Woodland transition. Others (Gardner 1983) have maintained that Coe’s early work at sites such as Hardaway and Doerschuck failed to recover any true early Woodland remains. What has been evasive is the information that bridges the Archaic to the Woodland, and more specifically, the site that embodies the modes of cultural change between the two. The E. Davis Site (31Fy549) may be the snapshot of such a transition.

Located in Winston-Salem, NC, the E. Davis Site was discovered in the spring of 1986 during a compliance project for the NC Department of Transportation. This site was identified with a single 50cm by 50cm testpit which came down upon a concentration of cultural material lying just 4cm below surface. Contained within this feature were five fabric-impressed pottery sherds, a single small-stemmed projectile point, fire-cracked rock, and 10.5g of charcoal. Subsequent radiocarbon analysis of the charcoal sample yielded a date of 2170 +/- 80 B.P. (Beta 17859), corrected to a range of 410 B.C. to A.D. 10 (Klein et al 1982). At present, these are the earliest dates recorded for an early Woodland occupation within the north-central North Carolina piedmont (Abbott et al 1986; Davis 1987).

Follow-up testing of 31Fy549 performed by NC Department of Transportation archeologists concluded that while the site may be of significant value, its location beyond the limits of the proposed corridor made it unlikely that it would be disturbed during construction. Therefore, the recommendation was for avoidance over excavation (Lautzenheiser 1986). Permission was sought from the landowner by Wake Forest University Archeology Laboratories to conduct more extensive investigation at the site. Work at 31Fy549 began in the spring of 1987 and continued intermittently through the spring of 1989. During that period, a total of 27 square meters were excavated uncovering numerous subsurface features, lithic and ceramic artifacts, and varied ethnobotanical remains. In the spring of 1997, Wake Forest University returned to the E. Davis Site and uncovered an additional 10 square meters, delimiting site boundaries and filling in areas of sparse excavation.

The E. Davis Site is located in the interriverine uplands of the northwest North Carolina piedmont at an elevation of 275m above mean sea level (AMSL). Lying within a climax oak-hickory forest, the site is situated on a long and narrow terrace measuring approximately 50m north to south and 20m east to west, bounded to the west by a precipitous drop and to the east by an ascending slope to the neighboring ridge. At the base of the slope to the west is an unnamed tributary of Minorcas Creek fed by a springhead located only 70m south of the site. The soils at the site belong to the Pacolet Series (specifically Pacolet fine sandy loam) which are formed from weathered granite, mica, gneiss, and schist resulting in moderately to highly acidic conditions. General stratigraphy for these soils consists of 10-18cm of yellowish-brown to greyish-brown fine sandy loam underlain by red to yellowish-red clay to sandy clay (Zimmerman 1976). Explicitly bounded on all sides by natural formations, the site is relatively small (15m by 10m) and surprisingly pristine. Unlike most upland sites, E Davis has not been subjected to plowing or tremendous erosion. In fact, colluvium from the ridge to the east has protected the subsurface integrity of the site.

Basic stratigraphy at the site consists of a thin root mat approximately 3cm thick, over a sandy clay soil (apparently colluvium) which extends to a depth of approximately 30cm. Compact red clay underlies this and extends to an unknown depth. Artifacts and features were encountered immediately beneath the humus layer at approximately 5cm BS.

The wealth of information provided on the early Woodland by the E. Davis site can be approached in three classes: subsurface features, lithics, and ceramics.

The seven features discovered during excavation can be divided into two classifications. The first group comprised of four features (3-3, 6-3, 9-7, 11-4) is described as small rock-filled pits. Generally encountered at 10cm BS, all of these features were initially represented as loose concentrations of stones, roughly 40-60cm in diameter, with variable amounts of charcoal present. At 20cm BS, these features assumed a regular, circular pattern averaging 25 cm in diameter extending an additional 15cm into sterile subsoil. In one case (11-4) the bottom was tapered to a point, while in the remaining four the bottoms were flat. The fill of the pits contained concentrations of heat-altered quartz mixed with charcoal, although the walls of the features exhibited no signs of burning. Lithic and ceramic material was present in small quantities. Similar "pits with rocks" have been recorded from early Woodland sites in eastern Tennessee (Lafferty 1981; Chapman 1975; Lewis and Kneberg 1957) and northwest Georgia (Crook 1984), but their use has been a matter of some confusion. Lafferty (1981) suggests that these features represent drying pits where hot stones are scooped from a fire (along with incidental charcoal) and placed in shallow depressions. Lewis and Kneberg (1957) suggest these features served as supports for cooking vessels where ceramic pots were placed upright in the pits and surrounded by hot stones. Alternatively, Crook (1984) develops a scenario where hot stones are placed in the shallow pits and are thermally shocked with water to cleave cobbles and assist with the lithic reduction process. While exact determination of function of these features at the E. Davis site is difficult, it is important to note that the size of these basins argues against their use for storage (see Caldwell 1958) and favors their use as part of some sort of indirect cooking process.

The remaining three features (3-4/5, 9-3, 9-4) are difficult to describe as anything other than extremely shallow depressions with high concentrations of fire-cracked rocks and other cultural materials. Generally oval in plan view, these features averaged 90cm by 110cm with only slight discoloration of the soil. The edges were indistinct and typically defined by the limits of the artifact concentration. None of these features extended beyond 25cm BS. Again, similar artifact lenses have been described in work from eastern Tennessee (Lafferty 1981; Chapman 1975, 1977), but function has been indeterminate. In all cases at the E. Davis site, these shallow features were found in proximity to the small basins described earlier. If this association is more than coincidental, one could develop a scenario where these artifact concentrations reflect some sort of maintenance activity related to the function of the neighboring basins.

The lithic assemblage from E. Davis reflects the specialized use of the site. Of the over 1300 pieces of debitage recovered in situ, 63% are of felsite while the remaining 37% is comprised of mainly native quartz. Cores are relatively rare (n=8) while tertiary retouch flakes account for over 90% of all lithic material representing late-stage tool production and maintenance. Formal tools consist of one large heavily battered biface, five modified flake bifaces (possibly preforms), and 15 projectile points or fragments. Of the 15 projectile points, there are 2 Guilford points (mid to late Archaic), one large stemmed point (possibly Savannah River or Kirk Stemmed), one ovate biface (rudimentary stemmed), 2 small straight stemmed points, 5 small contracting stemmed points, and 2 Yadkin triangular points (early Woodland). Of particular interest among the projectile points are the small straight and contracting stemmed points. While classified by some as Gypsy points (Davis 1987; Rogers 1989; see Oliver 1981, 1983, 1985), this typology is loosely applied to a wide range of early Woodland point forms. Similar points have been recovered at the Haw River Sites (31Ch8) (Claggett and Cable 1982) and the Wakefield Sites (Gunn, personal communication) in the eastern piedmont of North Carolina.

In regards to the contracting stemmed points, their similarity in form to the mid-Archaic Morrow Mountain II points described by Coe (1964) has been noted elsewhere (Claggett and Cable 1982), but even Coe admits confusion over the stratigraphic distribution of these points asserting they may represent a transitional Archaic-Woodland form that may have survived to a later date (Coe 1964:43). It is interesting to note that using Thomas’s (1978) four-variable discriminant analysis and Shott’s (1997) later two-variable analysis of metric attributes, the seven small stemmed points fall within the range of arrow points and not dart points. Although Thomas’s function fails to correctly identify 14 out of 100 points in his original study and Shott’s only slightly better, the breakage pattern represented on most examples does not contradict this conclusion. Placing the introduction of the bow and arrow at such an early date is not without precedence (Odell 1988; Lewis and Kneberg 1957) and if it is in fact plausible, the implications for subsistence patterning and subsequent social shifts are significant to southeastern prehistory.

The ceramic assemblage from E. Davis falls into the Yadkin series described by Coe (1964). Of the over 1000 sherds recovered, 37% are fabric impressed, 14% are cordmarked, 1% are plain, and the remaining 48% are unidentifiable. All sherds are tempered with coarse grit with some crushed quartz present, ranging in size from 2 to 6mm. Paste in all cases is compact and well-mixed. Wall thickness ranges from 6-8mm. Based on exterior surface treatment, color, temper, and paste characteristics, the minimum number of vessels represented is eight (Rogers 1989). Of these, six have been partially reconstructed allowing for some determination of size and shape. Orifice diameters are tightly clustered around 18cm with all forms seemingly representing small straight-sided bowls. The ceramics from E. Davis are typical of the early Woodland in North Carolina. Comparisons to the assemblage recovered at 31Ch8 shows relative homogeneity of surface treatment and vessel form during this period (Rogers 1989). Comparisons with roughly contemporaneous pottery types in surrounding areas, such as Swannanoa ceramics in the Appalachian Summit region (Keel 1975), the Watts Bar ceramics of eastern Tennessee, and the Kellogg Phase ceramics of northern Georgia, reinforce Caldwell’s (1958) assertion of the regional distribution of cord and fabric marked ceramics during the early Woodland.

It is certainly enticing to ponder the possibilities of the early introduction of the bow and arrow in the southeast, but I think a more attainable goal of this paper is the discussion of the early dates for the ceramics at E. Davis and their relative late occurrence when compared to surrounding traditions. When considering the transition from the Archaic to the Woodland, prime importance is placed on the adoption of ceramic production. Often described as the late Archaic with pottery (Dragoo 1976), the early Woodland marks an arbitrary heuristic break between earlier hunting and gathering subsistence patterns and impending horticultural settlement. As such, attention is placed more on dissimilarities than similarities between the two. Yet studies shows that early ceramics probably contributed little to the overall economic productivity of those groups who chose to produce it (Sassaman 1993; Rice 1987; Claggett and Cable 1982). It is likely that other containers, which may or may not appear in the archeological record, could adequately serve the same functions as ceramics.

Although the earliest North American ceramics come from coastal South Carolina and Georgia (with the earliest dates ranging from 4465 +/- 95 to 3110 +/- 110 B.P.), these late Archaic wares, such a Stallings and St. Simons, are fiber-tempered and closely resemble earlier forms found in steatite bowls. Slightly later, Thom’s Creek sand-tempered ceramics appear along the South Carolina coast and northward into southeastern North Carolina. It is not until 1500 years later that Caldwell’s (1958) Northern Tradition pottery, such as Vinette I from New York, makes its appearance in the record. From there, he proposes a southward expansion of these cordmarked ceramics reaching the southeast by 800 B.C. This northern tradition finds its expression in Virginia as Seldon Island Cordmarked dated to 750 B.C. (Gardner 1986). By that time however, Eastern Tradition fabric impressed ceramics had already found a foothold elsewhere in the southeast. Watts Bar ceramics from eastern Tennessee emerge during the first millennium B.C. and Kellogg Phase Dunlap ceramics from northwestern Georgia have been dated to at least 636 B.C. (Baker 1970) and possibly as early as 1000 B.C. (Garrow 1975).

It is evident from the above synopsis that ceramic production was well entrenched in the southeast considerably earlier than its acceptance in the central piedmont of North Carolina. It is certainly naive to assume native groups in this area had no knowledge of ceramic production while those around them were actively producing it. Rather one must assume that these groups did not possess a perceived need for such technology or could not conveniently incorporate its production into their subsistence and settlement scheduling. I assert here that their refined seasonal scheduling would have allowed them the time and materials for ceramic production much earlier than is witnessed in the archeological record. Fall camps located in the clay-rich uplands, laden with mast crops and exploitable fauna would provide the time and resources for groups within what Caldwell (1958) terms an "acorn economy" to produce pottery. With that in mind, one must consider elements of perceived need as catalyst for production.

The size and shape of the vessels from E. Davis argues against utilitarian benefits which generally center around their use as either storage containers or cooking vessels. The economic needs of indigenous early Woodland groups in the central Piedmont were aptly satisfied through a highly regulated seasonal-based schedule embedded within a "forest efficiency" model (Caldwell 1958). This sort of settlement/subsistence patterning results in an internally stable economic and social environment. Moreover, the success and efficiency of the forest adaptation leads to a conservatism which precludes the acceptance of other seemingly more advanced social and economic systems and their associated technology (Rice, et al 1972). However, what we may be witnessing with the initial adoption of ceramic use is an assimilation of discrete Archaic social groups into a larger social system bound by the use of pottery as a marker of social identity. Yet, the conservative nature of these groups prevents ready acceptance of other trappings of Woodland culture, mainly agriculture which does not appear in the region until about A.D. 700. The reasons as to why mobile, likely kin-based bands, would care to pursue incorporation into a larger and foreign social system are difficult to ascertain. It may be entirely possible that these threshold Archaic-Woodland groups with established territoriality maintained some control over the upland environment. Outside groups may then secure access to this territory through the exchange of ceramics or the specific technology for its manufacture. If this is the case, the use of ceramics need not force any change in economic strategy, but rather formalizes alliance between neighboring groups. Archeologically speaking, we should expect to find a hybridization of Archaic and Woodland assemblages at these points of contact. The E. Davis Site with is simultaneous use of stemmed and triangular points and the presence of early ceramics may represent such a synthesis

.

Abbott, Lawrence E., Jr., R. Jackson Marshall, and Elizabeth H. Hull

1986 A Cultural Resources Reconnaissance of the Silas Creek Parkway Extension Project Area, Winston-Salem, N.C., M.S. on file. Archeology Laboratories, Wake Forest University.

Baker, Steven

1970 Observations from the Lum Moss site: an interpretive field report. Southeastern Archaeological Conference, Bulletin 13:115-118.

Caldwell, Joseph R.

1958 Trend and Tradition in the prehistory of the eastern United States. American Anthropological Association, Memoir 88.

Chapman, Jefferson

1975 The Rose Island Site and the Bifurcate Point Tradition. Department of Anthropology, University of Tennessee, Reports of Investigations 14.

1977 Archaic Period Research in the Lower Little Tennessee River Valley. Department of Anthropology, University of Tennessee, Reports of Investigations 18. Coe, Joffre 1964 Formative Cultures of the Carolina Piedmont. Transactions of the American Philosophical Society, 54(5). Philadelphia.

Claggett, Stephen R. and John S. Cable, assemblers

1982 The Haw River Sites: Archeological Investigations at Two Stratified Sites in the North Carolina Piedmont. Report by Commonwealth Associates to U.S. Army Corps of Engineers, Wilmington District.

Crook, Morgan R.

1984 Cagle Site Report: Archaic and Early Woodland Period Manifestations in the North Georgia Piedmont. Georgia Department of Transportation, Occasional Papers in Cultural Resource Management 2, Atlanta, GA.

Davis, John D.

1987 Early Woodland of the North Carolina Piedmont: New Information from the E. Davis Site. Research report presented at the 44th Southeastern Archaeological Conference, Charleston, S.C.

Dragoo, Don W.

1976 Some Aspects of Eastern North American Prehistory: Review 1975. American Antiquity 41(1):3-27.

Gardner, William M.

1983 Comments on a Conference on Piedmont Archaeology. In Piedmont Archaeology, J. Mark Wittofski and Lyle E. Browning eds. Archaeological Society of Virginia, Special Publication 10.

1986 Lost Arrowheads and Broken Pottery: Traces of Indians in the Shenandoah Valley. Thunderbird Publications. Manassas, Virginia.

Garrow, Patrick

1975 The Woodland Period North of the Fall Line. Early Georgia 3(1):17-26.

Gunn, Joel

1997 personal communication.

Keel, Bennie C.

1976 Cherokee Archaeology: A Study of the Appalachian Summit. The University of Tennessee Press, Knoxville.

Klein, Jeffery, J.C. Lerman, P.E. Damon, and E.K. Ralph

1982 Calibration of Radiocarbon Dates: Tables Based on the Workshop on Calibrating the Radiocarbon Time Scale. Radiocarbon 24:103-150.

Lafferty, Robert H.

1981 The Phipps Bend Archaeological Project. TVA Publications in Anthropology 26. University of Alabama.

Lautzenheiser, Loretta

1986 Follow-up Archaeological Survey and Testing of the Silas Creek Parkway Extension Project Area, Winston-Salem, N.C. Forsyth County, R-2011. Report submitted to the North Carolina Department of Transportation, Planning and Research Branch.

Lewis, Thomas M. N. and Madeline Kneberg

1957 The Camp Creek Site. Tennessee Archaeologist 13(1):1-48.

Odell, George H.

1988 Addressing Prehistoric Hunting Practices Through Stone Tool Analysis. American Anthropologist 90:335-356.

Oliver, Billy L.

1981 The Piedmont Tradition: Refinement of the Savannah River Stemmed Point Type. M.A. thesis, Department of Anthropology, University of North Carolina, Chapel Hill.

1983 Formative Cultures – 20 Years Later: Refinement of the North Carolina Chronological Sequence. In Piedmont Archaeology, J. Mark Wittofski and Lyle E. Browning eds. Archaeological Society of Virginia, Special Publication 10.

1985 Tradition and Typology: Basic Elements of the Carolina Projectile Point Sequence. In Structure and Process in Southeastern Archaeology, Roy S. Dickens, Jr. And H. Trawick Ward eds. University of Alabama Press.

Rice, Prudence M.

1987 Pottery Analysis. The University of Chicago Press.

Rice, Prudence M., E. Pendleton Banks, and Robert E. Pace

1972 Contact Zones and Eastern United States Prehistory: Evidence from a Piedmont Rock Shelter. Southern Indian Studies 24:42-53.

Rogers, Rhea J.

1989 The E. Davis Site and the Origins of the Woodland in the Northwestern North Carolina. Paper presented at the 46th Annual Southeastern Archaeological Conference, Tampa, FL.

Sassaman, Kenneth E.

1993 Early Pottery in the Southeast. The University of Alabama Press.

Shott, Michael J.

1997 Stones and Shafts Redux: The Metric Discrimination of Chipped-Stone Dart and Arrow Points. American Antiquity 62(1)86-101.

Thomas, David H.

1978 Arrowheads and Atlatl Darts: How the Stones Got the Shaft. American Antiquity 43(3):461-472.

Zimmerman, James L.

1976 Soil Survey of Forsyth County, North Carolina. United States Department of Agriculture, Soil Conservation Service.

Library Index