Of Caves and Shell Mounds

12. Research Problems with Shells from Green River Shell Matrix Sites

12 Research Problems with Shells from Green River Shell Matrix Sites Cheryl Claassen SHELL DEBRIS FOUND in archaeological sites is typically treated as food debris attracting the attention of faunal analysts eager to engage in dietary and environmental reconstruction. In some cases shell is also present in archaeological context as ornaments or tools. Ornaments, ecofacts, and shell objects of unknown role are present in large numbers in the Archaic shell mounds of Kentucky 's middle Green River area. Instead of serving only to provide answers for standard archaeological questions, however, the nature of the shell debris of these sites actually stimulates many research questions not typically raised for shell-bearing sites. Problematizing the presence of both the freshwater and marine shell contained in these sites erodes the very foundation of their interpretation as either village sites or processing stations and of the shell as simply food debris and costume jewelry. Research Problems with Shell Ecofacts Many of the typical problems archaeologists address with freshwater shells in prehistoric sites have been performed with small shell samples from the DeWeese and Carlston Annis mounds in the Big Bend of Green River (Claassen, in prep.; Patch 1976, respectively). Bivalve shell density varies from top to bottom in these middens as does gastropod density. At DeWeese aquatic gastropods grossly outnumber land gastropods in the lower 40 cm of the midden, but the opposite situation prevails in the upper 2 m ofshell midden. A seasonality study (Claassen 1986) indicates that shells examined from one column in each site died in summer and fall. The suggestion that the shells were harvested seasonally , rather than year round, is one of many pieces of evidence now being cited to support an interpretation of these sites as seasonal aggregation points, rather than year-round settlements (Hofman 1985). Habitat Reconstruction One curious problem that was raised by the Shell Mound Archeological Project (SMAP) concerns the habitat from which the 60 species of Carlston Annis shells were harvested. Using modern habitat records, Patch (1976) concluded 1J2 Research Problems with Green River Shell Matrix Sites I :133 that these shells came from a quite shallow riffle-run habitat. Julie Stein (1982), however, concluded that Green River had been deep in the vicinity of the mound, with a silt-mud bottom. Suspecting that a broader sweep of modern habitat studies for freshwater bivalves would resolve the conflict, I enlarged on the number of sources used by Patch when I was analyzing the DeWeese bivalves (Claassen, in prep.). Fiftysix percent of the 32 species contained in a single DeWeese column were present in Green River in 1965 (Stansbery 1965), 53% of which could be found on mud or silt bottoms, and 88% could be found in deep water. Although these figures would seem to mean that far more of the Carlston Annis shells could have lived in a deep and muddy Green River than Patch recognized, the resolution of the problem is not so easy. Two of the four species that have not been recorded in deep water represent 47% and 48% of the valves found in the columns at Carlston Annis and DeWeese. Four possibilities are suggested: (1) that these two species do live in deep water in spite ofour inability to document such a habitat, (2) that the river reconstruction is incorrect, (3) that these animals have developed a shallow-water habitat preference in historic times, or (4) that both reconstructions are correct, meaning that the shells were not harvested from Green River. That people would transport shells over any significant distance may seem the least likely of the four explanations, but just such an interpretation has been offered in several other archaeological situations. Pacific coastal shells in large numbers have been recovered at mountain sites in California, over 10 miles from the sea, and in Peru, some 200 miles inland. More than 10,000 valves of the Gulf Coast bivalve Rangia cuneata have been recovered from the Arrowhead Farm site in Kentucky. Revised sea level charts still leave Archaic peoples of the Hudson River valley to carry tons of oyster shells 200 m or more then up bluff faces or by longer gradual routes to bluff tops. Likewise, the inhabitants of Globe Hill in West Virginia and of the Carlston Annis and DeWeese mounds in Kentucky had to haul shells up an incline. If one is willing to discard shells on top of an incline rather than at the creek, one might also be willing to walk over generally flat land...