«Pottery’s Place in the Gathering Histories of Florida’s Late Archaic Monuments Zackary I. Gilmore University of Florida Student Paper Competition ...»
Pottery’s Place in the Gathering Histories of Florida’s Late Archaic Monuments
Zackary I. Gilmore
University of Florida
Student Paper Competition Entry
Southeastern Archaeological Conference, Tampa Florida
Abstract: Recent theorizing in materiality emphasizes the ability of things and places to gather
various (human and nonhuman) actors and position them in particular social arrangements
(bundles, entanglements, intersections, etc.). The histories of Florida’s Late Archaic mounds are composed of such gathering events. Organizational and scalar transformations in moundcentered gatherings corresponded with the appearance of the region’s earliest pottery technology.
Using stylistic and sourcing data (petrographic and NAA), I argue that novel material and social conditions associated with pottery disrupted existing exchange relations and reordered mounding traditions, culminating in the establishment of large-scale festival centers along the St. Johns River.
Gilmore – SEAC 2013 Introduction: Gathering and Bundling Things-in-Motion A recurring theme in recent materiality discussions is the idea that everything is in constant motion (e.g., Hicks 2010; Hodder 2012; Ingold 2011a; Pauketat 2013a). In this view, people, materials, and other entities all exhibit distinct pathways of movement. Through time, individual pathways crosscut others forming complex, ever-changing webs of association, what Ingold (2011a, b) terms “meshworks.” The different categories of material culture (i.e., artifacts, features, and places) studied by archaeologists constitute nodes of intersection within these webs.
They are material gatherings of multiple moving parts, variously bringing together people, materials, techniques, traditions, memories, and beliefs (Heidegger 1971; Ingold 2010). The upshot of this line of thinking for archaeologists is that it is no longer tenable to view material culture items as singular, stable objects with fixed associations and representational meanings.
An effective alternative is to view archaeological materials as relational “bundles.” Traditional ceremonial bundles, made by animistic societies worldwide (Zedeño 2008), are composed of multiple items held together by a wrapping. According to Zedeño (2008:364), while each bundled object “has its own properties and realms of interaction, when two or more objects are combined, their interactive capabilities integrate to become a new object—the bundle—that is more than the sum of its parts.” In other words, a bundle’s affective power or agency is derived not from the essential qualities of any individual component but from the relationships established by combining and positioning different materials within a particular context (Pauketat 2013a, b). Recognizing the bundle-like qualities of all people, things, and places, Pauketat (2013a, b) and others (Hodder 2012; Keane 2005) have usefully extended the concept to include virtually all manner of material phenomena.
As convergences or gatherings of distinct things in motion, bundles constitute important
inflection points within larger historical webs of relationships (Pauketat 2013a:27). It is important to keep in mind, however, that bundles are never static. Bundles of all kinds change hands and shift between contexts, along the way acquiring associations with new people and events. They are subject to processes of accumulation, dissolution, recombination, and decay.
Consequently, rather than enduring as stable objects, bundles might better be envisioned as unfolding through time as series of successive gathering events (sensu Casey 1996; Hicks 2010;
Strathern 1990, 1992).
Although innovative and potentially transformative, the bundle concept is only informative insofar as it is grounded in real empirical data regarding the movement and intersection of various materials in the past. One of the most effective ways to acquire such data is by employing geological and chemical sourcing techniques. Using these techniques to study distinct “moments” of bundling and linking them through time, archaeologists can construct illuminating biographies of materials and practices and track their shifting social entanglements.
Gathering Places in Late Archaic Florida While all things and places may be considered bundles or gatherings to the extent that they bring together disparate elements, archaeologists have identified some places with a particular affinity for gathering and arranging constituents in large quantities and from across vast spatial scales (Adler and Wilshusen 1990). In fact, recently collected data show that many of North America’s largest and most iconic sites, including those at Chaco (Cameron and Duff 2008; Lekson and Cameron 1995), Cahokia (Alt 2006; Pauketat 2004; Pauketat and Alt 2004), and Poverty Point (Kidder 2010, 2011; Sassaman 2005; Spivey et al. n.d.), all served as centers of regional integration where disparate groups and materials were drawn together for purposes of exchange, ritual, and monument construction. The convergences and interactions fostered by
these places produced new types of social formations and affected the historical trajectories of entire regions.
Late Archaic mounds in northeast Florida (Figure 1) were also important gathering places, where wide-ranging people and things-in-motion converged and were bundled together.
As with all bundles, though, the material and social relations that they fostered were tenuous and subject to historical transformations and disjunctures. One major disjuncture occurred coincident with the appearance of the region’s earliest pottery and involved major alterations in the composition, scale, and organization of mound-centered gatherings.
The Late Archaic Period in northeast Florida consists of the late pre-pottery Thornhill phase and subsequent Orange period. The Thornhill phase (ca. 5600-4600 rcy B.P.) is characterized by a general continuation of earlier Mount Taylor period domestic practices but with a novel mortuary tradition centered on the construction of sand burial mounds and increased participation in extraregional exchange networks (Beasley 2008; Endonino 2008). Thornhill sand mounds were constructed throughout the middle St. Johns River valley and along the adjacent Atlantic coastline. They frequently served as mortuary monuments with excavated examples producing varying numbers of interments (Randall and Tucker 2012; Sears 1960). Sand mound deposits frequently contained exotic objects and materials, often in association with particular burials. These included chipped stone from north and west-central Florida (Endonino 2007, 2010), Strombus gigas celts with origins in south Florida (Wheeler and McGee 1994; Wheeler et al. 2000), bannerstones from the piedmont of Georgia (Randall 2010; Sassaman and Randall 2007), and groundstone beads from the Midsouth (Endonino 2010; Randall 2010).
By gathering geographically and temporally diverse suites of materials into an arena of common engagement, Thornhill mounds bundled broad stretches of space and time. In doing so,
they positioned these items in a way that singled out particular individuals. This is particularly clear at Bluffton where an entire mound was constructed around a single central burial (Sears 1960). In other places, certain burials included elaborate grave goods, including literal bundles of foreign and ritually significant materials (Moore 1894; Randall 2010:187). One person from Thornhill Lake, for example, was interred with a bannerstone on either wrist and a stone pendant near the neck and was surrounded by “great quantities” of shell and stone beads (Moore 1894:168). Disparities among burial treatments indicate that not everyone had equal access to these exotic materials. So, while mound-centered gatherings during the Thornhill phase undoubtedly involved large numbers of people, bundles were configured so as to highlight the status of only a few.
At ca. 4600 rcy B.P., the Thornhill sand mound tradition was abruptly halted. At this point (marking the start of the Orange Period [4600-3600 rcy B.P.]), most of the scores of extant sand and shell mounds along the middle St. Johns were abandoned and seemingly intentionally avoided (Randall 2010; Sassaman and Randall 2012). Active mounding, which had been virtually ubiquitous throughout the river valley, was consolidated and restricted to only four known locations. In all four, preexisting mortuary mounds were built up and appended with massive quantities of freshwater shell. What resulted was a series of huge, multilobed mound centers, reminiscent in some ways of contemporaneous coastal shell rings but at an even larger scale. In contrast to their Thornhill phase counterparts, Orange mounds did not receive any burials and saw a marked reduction in the incorporation of exotic materials. Consequently, while the status of Thornhill sand mounds as widely influential gathering places is difficult to question, the social scale of participation in Orange shell mound interactions has remained less clear.
What Orange mounds do contain are copious amounts of techno-stylistically diverse
fiber-tempered pottery. Orange pottery deposition in general appears to have been at least broadly structured, with decorated sherds showing up disproportionally at mounds compared to contemporary non-mounded sites in the same regions (Gilmore 2011; Sassaman 2003, 2004;
Saunders 2004a, b). Large quantities of elaborately incised vessels were deposited in mounds themselves and possibly into the water alongside them. This resulted in new kinds of mound bundles that contrasted significantly with those that preceded them. Chemical and mineralogical sourcing of Orange pottery provides an effective means of determining the social scale of these bundled associations and revealing the historical details of the transformations they facilitated.
Silver Glen Run (8LA1/8MR123) The Silver Glen Run (SGR) complex (Figure 2) is the northernmost of four known Orange mound centers in the middle St. Johns Valley. Located along a spring run that drains into Lake George, the complex consists of an elaborate array of shell-free and shell-bearing components including mounds and other deposits that span the Middle and Late Archaic, as well as later periods (Randall et al. 2011; Sassaman et al. 2011). Its Orange component includes two massive shell mounds (8LA1E and 8MR123) as well as a specialized shellfish processing locale replete with hundreds of large roasting pits (here labeled Locus B) and a possible habitation area or “village.” More than a dozen radiocarbon assays indicate that use of all of these different contexts overlapped temporally during the Orange Period.
Given their coevalness and close proximity in space, the stylistic and technological differences exhibited among their respective Orange pottery assemblages are striking (Figure 3).
Whereas more than 75 percent of vessels from 8LA1E’s north ridge and the 8MR123 mound exhibit rectilinear incisions, fewer than half of those from the south ridge, Locus B, and the 8MR123 village are decorated. Moreover, among the Locus B vessels that are decorated, there is
an unusually high incidence of curvilinear incisions and/or punctuations characteristic of the Tick Island Incised style of Orange pottery, a rare type at both mounds. In addition to surface treatment, mound vessels also exhibit larger average rim diameters and wall thicknesses and higher frequencies of exterior sooting and mend-holes than those from non-mound areas (Gilmore 2011).
Pottery Sourcing - Neutron Activation Analysis (NAA) and Petrography Recently, a combination of sourcing techniques, including NAA and petrography, were applied to Orange pottery from five distinct SGR contexts with the goal of identifying the social scale of the complex’s Late Archaic social gatherings. A total of 288 pottery samples were submitted for NAA. This analysis yielded chemical composition data related to 33 elements, three of which were ultimately excluded to avoid artificially biasing the results. Principle components analysis and simple bivariate plots were used to divide the samples into chemical composition groups, the validity of which were evaluated using mahalanobis distance. These composition groups, presumed to reflect distinct raw material compositions, were then compared to raw clay data in hopes of linking SGR pottery to particular source locations.
SLIDE13 Although preliminary, three distinct chemical composition groups (CG1-CG3) have been distinguished within SGR’s Orange assemblage. The first two groups have generally similar chemical profiles with CG1 (n=33) exhibiting slightly higher levels of most elements when compared to CG2 (n=46). The only exceptions are manganese, arsenic, zirconium, and hafnium, all of which show higher concentrations in CG2 than CG1. An additional 15 samples exist that cannot be assigned to either CG1 or CG2 but possess a chemical signature that falls somewhere in between the two groups. Comparatively, CG3 (n=73) is characterized by substantially lower concentrations of most elements, and is especially deficient in iron. This
group, however, exhibits the highest levels of sodium, antimony, and uranium of any of the composition groups. Of the 126 unassigned samples, 77 are likely members of one of the three groups but cannot be statistically included, while another 49 are outliers with a 0.1 percent chance of belonging to any of the defined groups. These composition groups are shown here in a bivariate plot of principle components 1 and 2. Similar relationships between the groups are apparent in biplots of certain individual elements. While there is substantial overlap among groups in these diagrams, their distinctiveness is confirmed by mahalanobis distance.