Changing Perspectives on the Archaeology of the Central Mississippi Valley

2. An Examination of Mississippian-Period Phases in Southeastern Missouri

2 An Examination of Mississippian-Period Phases in Southeastern Missouri Gregory L. Fox MODERN (post-1940) archaeological research in southeastern Missouri (Figure 2-1), and indeed in the central Mississippi River valley, has focused primarily on classificatory-historical problems (e.g., Lafferty and Price 1996; Phillips et al. 1951; Williams 1954), though more-recent studies by Dunnell (1982, 1988), Dunnell and Feathers (1991), Feathers (1990b), Holland (1991), O'Brien (1994a), Teltser (1988, 1993), and others have attempted to go beyond those classificationoriented studies and better document the tremendous variation that exists in the region's archaeological record. Still, the majority of archaeologists working in the area are culture historians, and they continue to base their conclusions on intuitive and untested interpretations of archaeological phenomena, many of which were formulated three or more decades ago. Two classificatory units, phases and pottery types, are integral to the interpretive framework that has grown up around the archaeology of the Mississippian period in southeastern Missouri. Understanding not only the history behind these units but also their limitations and how they have been applied is integral to assessing the accuracy of various cultural-historical interpretations. This chapter is a critical examination of phase constructions that have been proposed for southeastern Missouri. Although I focus specifically on that area, I suspect the findings have broad applicability to the central Mississippi Valley. My examination of Mississippian-period phases in southeastern Missouri was stimulated by two realities. First, despite forty years of archaeological research in the region, the initial classifications of cultural phenomena made by Stephen Williams in the early 1950S (Williams 1954) have remained essentially unchanged. Second, researchers working in the area (e.g., Klippel 1969; Marshall 1965; J. R. Williams 1967, 1968, 1972) have difficulty assigning many archaeological components to the existing phases. Stephen Williams recently asserted that his original definitions of the four phases in southeastern Missouri-Cairo Lowland, Nodena, Malden Plain, and Pemiscot Bayou-were based on more data than were generated from his excavations at Crosno (23Mh), located in Mississippi County, and from his surface collections of other sites in southeastern Missouri.1 A review of the literature, examination of artifacts in museums and private collections, and discussions with other archae- 32 Fox Kilometers ! ! 0 30 N t Figure 2-1. Map of southeastern Missouri showing physiographic features and locations of archaeological sites mentioned in the text. Mississippian Phases in Missouri I 33 ologists were also involved. Regardless of the impact of these activities, archaeological interpretation in the region is still based on the phases Williams (1954) presented in his dissertation. Over the succeeding decades, archaeologists (e.g., Klippel 1969; Marshall 1965; Smith 1990; J. R. Williams 1967, 1968, 1972) embraced the four-phase system and added a few pieces here and there, but the system has not been evaluated through a detailed comparative analysis, a consistent application of chronometric controls, or the use of statistical methods. Here I use statistical procedures rather than intuitive groupings to reexamine the type-frequency information , derived from surface collections and excavations, which Williams (1954) and subsequent researchers (e.g., Chapman and Anderson 1955; Klinger 1977b; Lewis 1982; Marshall 1965; Smith 1990; J. R. Williams 1968) used to define Mississippian-period phases and to assign assemblages to those phases. But what is a phase? Is it a group of similar assemblages-"similar" being based on some qualitative feel that assemblages are somehow similar-or is it a class of assemblages that share a set of necessary and sufficient traits that serve to distinguish any particular phase from all other phases (e.g., Dunnell 1971; Rouse 1964)? Williams's formulations for southeastern Missouri might be interpreted either way. Since he never explained this element of his conceptual scheme, it is necessary to evaluate his phases as both groups and classes. Williams's Phases: Groups or Classes? In his dissertation, Williams (1954) provided the first modern categorization of Mississippian-period phases in southeastern Missouri. For all intents and purposes, subsequent phase designations that have been proposed for the area (e.g., Marshall 1965) are simply variations on that original system (Table 2-1). The Cairo Lowland phase, originally based on Williams's (1954) excavations at Crosno and on his review of museum collections, can be considered the archetypical phase in the region, since all other phases are defined primarily on the absence of Cairo Lowland ceramic markers and secondarily on the presence of other pottery types in...