Reflections on Paddle Stamped Pottery: Symmetry Analysis of Swift Creek Paddle Designs
Pluckhahn, Thomas J., Southeastern Archaeology
Swift Creek Complicated Stamped pottery, dating to the Middle and Late Woodland periods, is one of the most distinctive ceramic types in eastern North America. Stylistic analyses of Swift Creek pottery have focused mainly on the spatial distributions of individual paddle designs reconstructed from stamped impressions on pottery. These studies have proven useful for tracking instances of social interaction but offer little help in the identification of assemblage-level variation and broader social processes. Symmetry analysis offers one possible avenue for the comparison of Swift Creek assemblages. In this pilot study, I consider the symmetry of more than 200 reconstructed Swift Creek paddle designs. The results indicate a concern with mirror symmetry, rotational symmetry, or a combination of both. I next contrast the symmetry of sub-assemblages from the site of Kolomoki, a Woodland ceremonial center in southwestern Georgia. The analysis reveals consistency in the symmetry of assemblages from village deposits, while mound contexts are in many ways unique. I relate these trends to broader social trends, particularly the development of an active strategy of social incorporation.
Swift Creek Complicated Stamped pottery, dating to the Middle and Late Woodland periods, is one of the most distinctive ceramic types in eastern North America. Named for the type site near Macón, Georgia, that was excavated in the 1930s (Kelly 1938), Swift Creek pottery is found throughout Georgia and portions of adjacent states (Williams and Elliott 1998). The type is distinguished by complex curvilinear stamped designs representing both abstract geometric patterns and abstracted forms of natural phenomena (Broyles ca. 1960s, 1968; Snow 1975, 1998, 2003).
The elaborate designs on Swift Creek pottery long have been recognized as fertile ground for stylistic studies, from the pioneering work of Bettye Broyles (1968) to the more recent and ongoing efforts of Frankie Snow (1975, 1998, 2003; Snow and Stephenson 1998). Stylistic analyses of Swift Creek pottery have focused mainly on the spatial distributions of individual paddle designs. These studies have proven amazingly useful for tracking instances of social interaction between communities and, in some cases, the actual movements of people. However, they offer little help in the identification of assemblage-level variation and broader patterns of production, exchange, ethnicity, or identity.
Symmetry analysis offers one possible avenue for comparison of Swift Creek assemblages. Briefly, symmetry analysis consists of a technique for classifying the "underlying structure of decorated forms; that is, the way the parts (elements, motifs, design units) are arranged in the whole design by the geometrical symmetries which repeat them" (Washburn and Crow 1988:ix; Washburn and Crow 2004). Emphasis is on the way design elements are repeated, not the elements themselves. Symmetry analysis has been employed widely in the southwestern United States to compare assemblage-level variation in painted pottery (e.g., Hardin and Mills 2000; Hegmon 1995; Washburn 1999; Zaslow 1980). However, the technique has been used only sparingly in the eastern United States (for exceptions, see Denny 2001; Knepper, Bowen, and Petraglia 2004).
The potential advantages of symmetry analysis are minimally two fold. First, although there is some discrepancy in terminology among researchers (Rice 1987:263), symmetry analysis offers an established, replicable series of procedures that can be applied not only to pottery but also to other aspects of material culture. This allows for consistency of classification across classes of material remains and among different investigators (Washburn and Crowe 1988:32-33).
Perhaps more important, several studies suggest that symmetry and similar structural principles of design configuration and layout may be more sensitive to social factors and more effective media for symbolic communication than other attributes of design (Washburn and Crowe 1988:32, 40). …