The Cultural History of Bannerstones in the Savannah River Valley

Abstract

In the early twentieth century, the late William H. Claflin Jr. of Belmont, Massachusetts, made regular forays to the middle Savannah River Valley of Georgia and South Carolina to collect artifacts from the area and occasionally dig at sites such as Stallings Island. Most of his artifact collection is now curated at the Peabody Museum of Archaeology and Ethnology, Harvard University. Among the relatively well provenienced items are 140 bannerstones from several locations in the middle Savannah. Informed by current archaeological knowledge of the region, the Claflin bannerstones register the history of group formation, interaction, and alliance over a span of ca. 1,300 years (5500-4200 cal. B.P.). During this time, three successive phases are marked by changes in morphology, production, and distribution. Transitions between phases coincide with major changes in the distribution and alignment of regional populations. Summaries of the chief features of each phase are followed by observations on the historical circumstances accounting for such change. Bannerstones are stone artifacts that were pecked and polished into various shapes and drilled lengthwise or widthwise with a hole roughly one centimeter in diameter. The use of such items has long been a topic of debate in eastern North American archaeology (Baer 1921; Knoblock 1939; Kwas 1982; Moorehead 1917; Webb 1957). The term bannerstone itself evokes a sense of ceremonial importance. Indeed, bannerstones occur in burials and caches of the Archaic period (e.g., Chaflin 1931; Charles and Buikstra 1983; Doucette 2003; Douglass 1882; Fowler 1957; Hassan and Farnsworth 1987; Moore 1882; Webb 1946; Webb and DeJarnette 1942; Webb and Haag 1939) and they sometimes bear remarkable design qualities and exaggerated or hypertrophic forms (Sassaman 1998). But they also occur in simple, rough forms and are found broken in middens with the usual domestic refuse. Thus bannerstones appear to have had a dual existence in both and ritual realms of Archaic life (Kwas 1982). The mundane identity of bannerstones is the spear-thrower or atlatl weight, following from the functional inferences made by William S. Webb (1957). Having observed drilled prismoidal stones with other spear-thrower parts in Shell Mound Archaic graves in the mid-South (Webb 1946:221; Webb and DeJarnette 1942:270-286; Webb and Haag 1939:51-58), Webb not only inferred the function of these stones but also developed a model to explain the evolution of elaborate forms as a response to demands for increased spearthrowing efficiency (Webb 1957). Modern understanding of the mechanics of spear throwing undermine Webb's model (Cotterell and Kamminga 1990:168-169). Spear throwers without attached weights were used by hunters worldwide and were most likely used throughout the Holocene in the New World before (and perhaps even after) bow and arrow technology appeared. Thus, although drilled stones and other media were no doubt, attached to spear throwers by some Archaic hunters in the East, and were probably used to launch spears, the extraneousness of this attachment, combined with the otherwise elaborate, hypertrophic forms, warrants use of the term bannerstone, which is what we favor. Bannerstones are a uniquely Eastern Archaic trait with tremendous spatial, temporal, and formal variation. Documenting the variation of bannerstones has been an ongoing project since the days of Webb. Knoblock (1939) and Lutz (2000) assembled the largest compendia on bannerstones, and Kwas (1981) provided one of the first analytical approaches to address temporal variations. Several of us have been working on the social implications of bannerstones in the south Atlantic slope (Blessing 2006; Sassaman 1996, 1998; Sassaman and Randall 2004), Bowen (1994) continues to amass a database for Ohio and vicinity, and Burdin (2004) recently compiled the data for Kentucky. Bannerstones are affiliated most directly with populations of the Middle and Late Archaic periods and probably do not predate 8500 cal. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 10 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.