Thoms Creek Plain

exterior viewinterior viewview of paste

from David Anderson's Thoms Creek Type Description


The type Thom's Creek Plain was first formally described by Phelps (1968:21), based on a sample of 176 sherds from White's Mount (9RI4) and the Boy Scout site (9BK6) along the Savannah River drainage in Georgia. A second formal type description, based on a sample of 4369 sherds from 13 sites along the South Carolina coast, was prepared by Trinkley (1976a; 1980b:17). The two descriptions serve to document variability in the ware along the central Savannah River and in the South Carolina Sea Island area, respectively. A thorough review of taxonomic and classificatory efforts associated with Thom's Creek ceramics has been provided by Trinkley (1980b). The Thom's Creek series, similar in several respects to the Stallings series, originated with Griffins's (1945) description of a non-fiber tempered punctated ware at the Thom's Creek site on the upper Congaree River near Columbia, South Carolina. Thom's Creek Punctate was the only ware recognized in the original type description, although Griffin (1945:470) noted that there were three plain sherds "probably belonging to the type called "Thom's Creek Punctate;" this marked the first (informed) recognition of what would later be called Thom's Creek Plain.

Although apparently postdating Stallings wares along the Savannah (e.g., Stoltman 1974:91; Phelps 1968; Trinkley 1980a: 45-48), stratigraphic and radiocarbon data from the South Carolina area indicates a long period of overlap or coassociation for the two series (Trinkley 1976a, 1980a, 1980b:19; Sassaman 1993). Changes in decorative treatment over these wares may follow similar trajectories. Evidence for an early appearance of Thom's Creek Plain, predating the decorated Thom's Creek types, is currently equivocal, but is suggested at a few sites (Trinkley 1980a:63, 287). An increase in plain finish with increasing excavation depth occurs at Fig Island (Trinkley 1980a:63), and this pattern was also noted at Mattassee Lake (Anderson et al. 1982).

Using data from several coastal South Carolina Sites, Trinkley (1980a) has suggested that decorative variability between Thom's Creek phase sites may reflect a temporal dimension:

It is possible to suggest that the pottery be seriated such that Thom's Creek Plain is the oldest pottery (acknowledging that it will be found at all sites representing undecorated portions of decorated vessels) followed by Thom's Creek Reed Punctate. At the time reed punctating was losing popularity, Thom's Creek Shell Punctate was gaining popularity. Awendaw Finger Pinched appears to follow the shell punctate style. Based on the more prominent occurrence of the minority ware Awendaw Finger Impressed pottery from Lighthouse Point and Stratton Place, I am tempted to suggest that this pottery represents the last expression of the Thom's Creek potters (Trinkley 1980a:287).

The primary geographic distributions of these decorative motifs—Thom's Creek Reed Punctate in the interior and along the southwestern coast of South Carolina, Thom's Creek Shell Punctate in the central coastal area (to about Charleston Harbor), and Awendaw Finger Pinched in the northern coastal area (from Charleston Harbor to the Santee River)—additionally suggested to Trinkley (1980a:291, 314-315) a gradual population movement northward along the coast. Many of the Thom's Creek sites in the interior of the coastal plain may, in this view, date to an early part of the Thom's Creek phase, and may "represent part of an early seasonal cycle between the coast and the interior which is not found as frequently later" (Trinkley 1980 a: 291).

An alternative explanation for the observed distributions, based on inferences about Late Archaic sociopolitical organization, has been proposed by several investigators, notably Widmer (1976a:43), Michie (1979:49), and Anderson et al. (1979:94-95). In this view, the differential distributions of the Thom's Creek wares may correspond to the territories of discrete sociopolitical entities:

It is suggested that late Archaic artifact distributions delimit the boundaries of relatively endogamous, probably tribal level social groups. At least two, and possibly three, such groups are hypothesized to exist in the Sea Island area of South Carolina, characterized by Stallings wares in the southwest and Awendaw ware in the northeast, with a possible third group between them. . . . A separate group may have occupied much of the interior of the Coastal Plain, characterized by Thom's Creek ceramics. . . . group endogamy is inferred from the relatively discrete ceramic distributions. If exogamous spouse procurement and exchange occurred, greater intergradation and stylistic overlap might be expected (Anderson et al. 1979:94-95).

Strictly speaking, Thom's Creek Plain can only be identified at the assemblage level, and not sherd by sherd. The ware tends to intergrade with later plain wares. An appreciable proportion of many assemblages appear to be almost temperless, with inclusions larger than l mm uncommon. This aspect of the paste is a diagnostic attribute of the ware; conscious selection or milling of clay sources may have been practiced.

The exterior surfaces of the Thom's Creek Plain vessels are almost uniformly undifferentiated, although parallel scraping marks are common on specimens from the Sea Island area, from smoothing with a marine shell. The vast majority of the Thom's Creek Plain rims are incurvate or straight, and only rarely excurvate in form. Lips are typically rounded or flattened, and sometimes exhibit decorative treatment, usually narrow incising or cord marked or simple stamping. Some manufacturing differences are evident over its range; material from along the Savannah tends to be thicker with a coarser (sandier, less compact) paste than is noted on materials from further to the northeast, along the Edisto, Santee, and Pee Dee Rivers.

Sorting Criteria

Plain surface finish. Paste, general surface finish and rim and lip forms similar or identical to that noted for Thom's Creek Reed Separate Punctate. May be confused with later Woodland Plain with which it tends to intergrade (see Background discussion for Woodland Plain). Sherds may be from undecorated portions of decorated vessels.


Found throughout the Coastal Plain and Fall Line areas of extreme eastern Georgia, South Carolina, and southwestern North Carolina. Infrequently noted above the Fall Line, typically only along major drainages.

Chronological Position

Late Archaic period, Thom's Creek Phase (2000BC- 1000BC).

Primary references

Phelps (1968:21); Trinkley (1976a; 1980b:17).