Jump to ContentJump to Main Navigation
Gathering at Silver GlenCommunity and History in Late Archaic Florida$

Zackary I. Gilmore

Print publication date: 2016

Print ISBN-13: 9780813062716

Published to Florida Scholarship Online: January 2017

DOI: 10.5744/florida/9780813062716.001.0001

Show Summary Details
Page of

PRINTED FROM FLORIDA SCHOLARSHIP ONLINE (www.florida.universitypressscholarship.com). (c) Copyright University Press of Florida, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in FLASO for personal use (for details see http://www.florida.universitypressscholarship.com/page/privacy-policy). Subscriber: null; date: 21 June 2018

Late Archaic Depositional Narratives at Silver Glen

Late Archaic Depositional Narratives at Silver Glen

Chapter:
(p.83) 3 Late Archaic Depositional Narratives at Silver Glen
Source:
Gathering at Silver Glen
Author(s):

Zackary I. Gilmore

Publisher:
University Press of Florida
DOI:10.5744/florida/9780813062716.003.0004

Abstract and Keywords

Chapter 3 provides a history of Archaic period landscape modification at the Silver Glen complex in northeast Florida. Primary focus is directed toward the various depositional traditions and strategies that were implicated in the construction and repeated reshaping of the complex by Archaic hunter-gatherers between ca. 8000 and 3000 B.P. The beginning of this interval marks the earliest evidence for large-scale digging and deposition at Silver Glen Springs, an event that prefaced more than three millennia of subsequent preceramic engagement with this location. These developments are discussed in terms of the historical precedents they set for the complex’s earliest pottery-making inhabitants after ca. 4500 B.P. It is argued that the complex’s two largest shell mounds were intentionally sited atop preexisting mortuaries in order to draw on the authority of the ancestral past, even while implementing profound future-oriented transformations of the material and social landscape. Also discussed is an assemblage of huge roasting pits that were likely used to provision mound-centered feasts. Throughout chapter 3, focus centers on identifying the various times, places, peoples, and materials that were bundled together in particular depositional episodes and linking these together to create a coherent account of the complex’s development through time.

Keywords:   deposition, landscape, shell mounds, Silver Glen complex, feasting, pits

As already noted, the Silver Glen complex is one of four known Orange period mound centers in Florida’s middle St. Johns River valley. Only sporadically visited by archaeologists in the twentieth century, Silver Glen was the focus of seven recent seasons of field research by the University of Florida’s St. Johns Archaeological Field School. These investigations uncovered an elaborate array of cultural deposits that provide an almost unbroken record of approximately seven thousand years of pre-Columbian indigenous history. During the Late Archaic period, Silver Glen witnessed the construction and use of an assortment of temporally overlapping contexts by communities of various size and composition. The sheer scale of some of these contexts points to the complex’s status as an important place of gathering, and their diversity provides a virtually unparalleled opportunity for investigating the variety of mound-centered interactions engaged in by Orange period hunter-gatherers. In this chapter, I attempt to take advantage of this potential by constructing a narrative history of the multiscalar depositional events that contributed to the creation of Silver Glen as a widely influential place of regional integration.

The Environmental Setting

The physical environment experienced by Florida’s Archaic inhabitants was heavily influenced by the state’s dynamic geological history. As summarized by Thomas Scott (1997, 2011), the Florida platform (roughly half of which forms the current terrestrial portion of the state) has a geologic basement composed of Mesozoic and pre-Mesozoic rocks. At the beginning of the Paleocene (approximately 65 million years ago), or perhaps earlier, carbonate sediments came to dominate deposition on the platform, eventually resulting in a thick mantle of limestone, dolostone, and evaporites (p.84) that covered virtually its entire expanse. This carbonate layer houses the massive Floridan aquifer system, the source of most of Florida’s countless artesian springs and much of its potable water (Miller 1997). As discussed above, springs were important cultural and ecological resources for pre-Columbian peoples throughout the state, and their initial flow has been posited as the primary impetus behind a mid-Holocene influx of migrants into the St. Johns valley (Miller 1992, 1998). Florida’s carbonate lithology is also responsible for its distinctive karstic terrain, which features abundant solution features such as sinkholes, caves, conduits, and fractures. Near the beginning of the Miocene (approximately 25 million years ago), a dramatic uplift of the Appalachians resulted in a marked increase in siliciclastic sedimentation across Florida that continued up to the Quaternary, albeit interspersed with additional periods of substantial carbonate deposition. Miocene to Holocene sediments now blanket the entire platform, ranging from less than 1 meter thick in some areas to more than 300 meters in others (Scott 1997, 2011). These heterogeneous near-surface deposits were the primary source of the various clays, quartz sands, and heavy minerals used by pre-Columbian potters.

Florida’s largest river, the St. Johns, cuts through these surficial deposits along a 500-kilometer course that roughly parallels Florida’s Atlantic coastline from southern Brevard County to Jacksonville. The St. Johns is unusual among rivers in the Northern Hemisphere in that it flows from south to north. It has an extremely low gradient (less than 0.02 meters per kilometer) (Miller 1998: 28), resulting in low flow velocities and a smaller than expected discharge rate given its substantial breadth. This low gradient also means that the river is highly responsive to sea level changes and is tidally influenced over most of its length. Its exceedingly slow-moving currents have resulted in an abundance of backwater environments and expansive lakes (Miller 1998: 67). These productive aquatic environs are especially characteristic of the middle portion of the St. Johns River, which stretches from Lake Monroe in the south to Lake George in the north. Sometimes referred to as the “St. Johns offset” (Pirkle 1971; White 1970: 104), this segment of the river juts to the west and cuts into the underlying limestone (Schmidt 1997). It is fed by a large number of sizeable springs and features a litany of old channels, lagoons, and lakes that speak to this area’s complicated hydrological history. Paleoclimatic data and archaeological site distributions suggest that the general parameters of the river’s present hydrological regime were likely in place by at least seven thousand years ago (Randall 2010; Randall et al. 2011: 15; Watts et al. 1996).

(p.85) The Silver Glen complex is distributed along a 1-kilometer-long spring run at the intersection of Lake, Marion, and Volusia Counties in the middle St. Johns valley. This spring run is fed by the first-magnitude (indicating a discharge rate of more than 100 cubic feet of water per second) Silver Glen Springs and empties into Lake George, Florida’s second-largest body of freshwater. Currently, land rises steeply from 1.0 to 4.0 meters in absolute elevation both north and south of the run before levelling off, although this topography has been greatly altered by anthropogenic activities, including ancient shell deposition and modern shell removal (Randall et al. 2011; Sassaman et al. 2011; see below). The soils surrounding the complex are diverse and range from excessively drained to poorly drained (USDA 1975, 1979). They support a diverse array of native vegetation, including various oaks, pines, cypress trees, gum trees, and saw palmetto, along with a host of other trees, shrubs, and grasses. Upland forests and wetlands bordering the spring run are home to many mammals, birds, reptiles, amphibians, and gastropods that were exploited by the complex’s indigenous inhabitants (Blessing 2011a). The unique hydrological conditions presented by the confluence of the spring run, lake, and tidally influenced river provide habitat for a wide range of aquatic fauna, including manatee, otter, alligator, turtle, and scores of species of both freshwater and marine fish (Bass and Cox 1985). Importantly, it also supports abundant shellfish populations, including Viviparus georgianus (banded mystery snail), Pomacea paludosa (Florida apple snail), and freshwater bivalve (Unionidae), which form the primary constituents of freshwater shell matrix sites along the St. Johns River.

The Silver Glen Complex: Archaeological Overview and Background

The Silver Glen complex, as originally configured (figure 3.1, top), comprised an extraordinary array of pre-Columbian cultural features that included some of the largest shell deposits in Florida. The remains of these features are arranged along the southern and northern banks of the spring run. Although the complex is broken into three separate sites—8LA1, 8LA4242, and 8MR123—its division is more a function of archaeological reporting procedures than any hard-and-fast boundaries in the distribution of cultural remains. Because of its massive size and archaeological complexity, the largest of these sites, 8LA1, has been divided into two areas, 8LA1-East (8LA1E) and 8LA1-West (8LA1W), to ease communication (following Sassaman et al. 2011). 8LA1W is further subdivided into four (p.86)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.1. Maps showing the reconstructed (top) and current (bottom) topographies of the Silver Glen complex and the locations of various sites and excavation loci.

Topographic reconstruction adapted from Randall 2014.

(p.87) distinct excavation loci (Locus A–Locus D). Site 8LA1 is confined to the area south of the run on the property of the Juniper Club, a private organization, while 8MR123 and 8LA4242 are located north of the run, within the boundaries of the Ocala National Forest.

Prior to the twentieth century, the most visually imposing parts of the complex would have been two gigantic, multilobed shell mounds—one at the mouth of the spring run and the other surrounding the spring’s primary vent. These mounds were noted repeatedly by nineteenth-century naturalists and antiquarians, who marveled at their impressive scale and apparent antiquity (Bartram 1942: 44; Le Baron 1884: 774; Wyman 1875: 38–39). Jeffries Wyman (1875: 39) describes the huge U-shaped mound at 8LA1E as forming three sides of a hollow square with a maximum height of 20 to 25 feet (6 to 7.6 meters) and exhibiting a deep, shell-free interior. His excavations there in the 1860s yielded “an abundance of pottery and other evidence of man’s agency” (Wyman 1875: 40). Of the mound at 8MR123, Wyman (1875: 39) states that it formed an “amphitheatre” of shell around the spring’s source. Two other prominent aboveground shell deposits apparently went unnoticed by nineteenth-century visitors to the complex. These deposits, Locus A at 8LA1W and 8LA4242, formed roughly linear ridges situated on either side of the run, approximately midway between the spring pool and Lake George. Locus B and Locus C at 8LA1W are two topographically subtle, yet elaborately stratified, areas of shell deposition located on ridge noses overlooking the run west of Locus A. Locus D is a subsurface shell-free midden still farther to the west.

Like most of the other large shellworks in the middle St. Johns valley, the mounds at Silver Glen were severely impacted by shell mining in the first half of the twentieth century. South of the spring run, mining began in 1923. At 8LA1E, shell was dug out down to the level of the water, and in some cases below, and sold for a sum of $17,000 (Sassaman et al. 2011: 1). At the same time, the shell ridge at 8LA1W’s Locus A was reduced to a series of discontinuous escarpments and isolated remnants (Sassaman and Randall 2011: 121). In the northern half of the complex, mining between the 1920s and 1940s removed virtually the entire aboveground portions of the 8MR123 mound and the ridge at 8LA4242 (Randall et al. 2011: 18). Fortunately, intact subsurface deposits remain at most of these locations. The only substantial shell deposits to escape mining activities were those at Locus B and Locus C, which are manifest only subtly on the surface.

Perhaps because of its denuded postmining condition, the Silver Glen complex received only scant archaeological attention during the twentieth (p.88)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.2. Adaptation of Alden Potter’s (1935) sketch map of Silver Glen Springs showing the extent of shell deposits and shell removal at that time.

(p.89) century. Some general observations were made of the mound at 8MR123 by visitors from the Civilian Conservation Corps as it was being mined, and a rough sketch map was produced (Potter 1935; figure 3.2). Alden Potter (1935) and Allen Taylor (1935) also noted a number of mostly Archaic period artifacts as well as the presence of human burials at the site. In 1990, two 1 × 1 meter test units were excavated into remaining basal strata at 8MR123 by Florida State University researchers led by Rochelle Marrinan, revealing approximately 3 meters of intact mounded deposits. This work yielded four radiocarbon assays that suggest that at least the lower component of the mound was erected during the preceramic Thornhill Lake phase (Marrinan 1990). In the 1950s, John Goggin and others collected pottery samples from the ground surface and water surrounding 8LA1E that are now housed at the Florida Museum of Natural History. Otherwise, the southern half of the complex had remained completely unstudied by archaeologists until recent investigations by the University of Florida’s Laboratory of Southeastern Archaeology (LSA).

Between 2007 and 2013, LSA researchers and participants in its St. Johns Archaeological Field School, directed by Kenneth Sassaman, conducted extensive surveys and detailed stratigraphic excavations across the entire complex (Randall et al. 2011; Sassaman et al. 2011). I was directly involved in these investigations for five of the seven field seasons, during which time I designed and oversaw the project at 8LA1W’s Locus B. The remainder of this chapter relies on the results of this field research in outlining a historical narrative of the depositional strategies and events that factored into the making and remaking of the complex during the Late Archaic.

Depositional Events and the Making of the Silver Glen Complex

As noted above, Silver Glen contains archaeological deposits that span most of the past seven thousand years; however, radiocarbon data (appendix A) indicate that most of the complex’s substantial shell deposits and its most prominent aboveground components were emplaced during the Archaic period. A summary of the Archaic depositional sequence in this location goes as follows: The first large-scale shell deposition began as early as 6500 cal BP and culminated in the Early Mount Taylor shell ridge at 8LA1W’s Locus A. During this same period, a morphologically similar and likely coeval ridge was erected directly across the spring run from Locus A at 8LA4242, although its precise chronological position is unknown. As occupation of these paired structures was winding down circa 5700 cal BP, (p.90) a Thornhill phase sand and shell mortuary was established at 8MR123 and a small residential settlement—one that would be serially reoccupied over the next thousand years—was set up at 8LA1W’s Locus B. An additional preceramic mortuary was founded at 8LA1E, but little is known regarding its configuration and age. During the Orange period, the mortuary deposits at 8LA1E and 8MR123 were built up and appended with huge quantities of shell, resulting in the enormous multilobed mounds described by Wyman and others. During the same period, hundreds of massive roasting pits were dug, filled in, and encased with shell at Locus B and Locus C, and numerous apparently ephemeral, circular encampments were occupied around the complex.

Here, I concentrate primarily on the complex’s development during the latter part of this sequence. In particular, I consider the different depositional strategies engaged in by the complex’s Orange inhabitants; the temporal, spatial, and social relationships that these strategies achieved; and their effects on the composition and structure of the Late Archaic communities tied to Silver Glen. To understand these processes, however, we must first grasp the state of the preexisting, late preceramic Silver Glen landscape as it would have been encountered at the outset of the Orange period.

Mount Taylor Precedents

By Orange times, the area surrounding Silver Glen Springs, as with much of the middle St. Johns valley, was far from pristine. It was, rather, teeming with the material remains of millennia of past domestic and ceremonial activities in the form of eye-catching mounds and ridges, buried features and deposits, and countless scattered artifacts. The remarkable concentration of these remains at Silver Glen and similar places was surely one of the most significant attractors for the vast quantities of Orange period people and materials that converged there.

Early Mount Taylor Shell Ridges

Two of the most obvious preceramic structures that would have confronted early Orange visitors of Silver Glen are the extensive shell ridges at 8LA1W’s Locus A and 8LA4242. Positioned directly across from one another, each consists of a roughly 75-meter-wide shell ridge stretching for approximately 200 meters along the edge of the spring run. Their original heights are unknown, although the Locus A ridge retains a total of 3 meters of intact vertical deposits (Randall 2013; Sassaman and Randall 2011: 121). (p.91) As discussed briefly in chapter 1, excavations at Locus A indicate that the first Mount Taylor activities there involved the digging of numerous large pits along the run (Randall and Sassaman 2012). The entire pitted area was then capped off with a mantle of tan/brown sand. Atop this sand, a series of individual house platforms were constructed and occupied. Over time, the sand and shell nodes resulting from these residential activities began to overlap, eventually resulting in a complex mounded palimpsest of intersecting surfaces and domestic deposits. Radiocarbon assays from upper and lower Locus A strata show that the bulk of the ridge’s aboveground depositional sequence played out relatively rapidly sometime between 6300 and 5940 cal BP (Randall 2013). After this interval, there are few signs of continued Archaic period activity at Locus A, suggesting more than a millennium of abandonment prior to the onset of the Orange period.

Thornhill Lake Phase Mortuaries

Immediately preceding the onset of Orange period activities at the complex were Thornhill Lake phase deposits at 8MR123. The sprawling mound that surrounded the spring pool prior to historic shell mining began as an earthen mortuary deposit. Multiple in situ burials were uncovered during mining operations in this area, and human remains have been routinely encountered since that time (Potter 1935; Randall et al. 2011: 35–38). The original dimensions and height of the mortuary mound are unclear; however, recent excavations reveal that it was composed primarily of brown and yellow sand that was then capped with a lens of dense freshwater shell. Charcoal samples yielded radiocarbon age estimates of 5850–5590 cal BP for the sand and 5590–5320 cal BP for the shell cap, situating these deposits relatively early in the Thornhill Lake phase (Randall et al. 2011; Randall 2013). During this same period, additional large quantities of shell were deposited around the spring pool and along the run (see detailed descriptions in Randall et al. 2011). In at least some spots, these shell deposits would grow to reach a height of more than 3 meters above the surrounding terrain (based on Marrinan 1990).

The discovery of a human interment in shell deposits at 8LA1E hints at the possible presence of an additional preceramic mortuary facility in that portion of the complex (Sassaman 2011b). The location of the burial, which was exposed by erosion and subsequently reburied, places it within what were, prior to shell mining, the basal deposits of the northern ridge of 8LA1E’s U-shaped mound. The burial is now situated within an expansive area of dense concreted shell that parallels the southern bank of the spring (p.92) run. This deposit’s great depth when the mound was intact, its highly concreted condition, and its resemblance in size and shape to intact Mount Taylor ridges suggest that the site’s eventual U-shaped monument was preceded by a roughly linear preceramic mortuary.

Preceramic Locus B

The final area of preceramic deposition that is likely to have impacted subsequent Orange ideas and activities is located at 8LA1W’s Locus B. Between approximately 5800 and 4600 cal BP, Locus B was the site of repeated small-scale settlements and abandonments by Thornhill Lake phase hunter-gatherers. Stratigraphically, this occupational sequence is manifested as a series of crushed-shell living surfaces with intervening thin layers of dark, organic-rich sands that are largely shell- and artifact-free (figure 3.3A; see also Gilmore 2011). At least one of the surfaces appears to have been formalized through the laying down and burning of clay-rich sediment prior to the commencement of shell deposition. Numerous pit features of various sizes and shapes descend from the habitation surfaces. Both the pits and surface deposits contain abundant material residues of everyday living, including tools and ornaments made from stone, bone, and shell, along with charcoal, vertebrate faunal remains, and paleofeces (figure 3.3B,C; Gilmore 2011). Once again, the last of these habitation deposits was capped with a thick layer of mostly whole freshwater gastropod and bivalve shell. The culmination of these various depositional activities was a relatively discrete shell node, similar in many respects to those found at the Hontoon Dead Creek Village site (described in chapter 1). Reaching approximately 1 meter in height at its apex, this anthropogenic deposit is responsible for most of the modest topographic relief currently visible at Locus B.

In summary, by the beginning of the Orange period, the upland areas both north and south of the Silver Glen Springs run were covered with the remains of thousands of years of preceramic habitation and ceremony. Among the most conspicuous ruins were two expansive shell ridges, two large-scale mortuary facilities, and the accumulated residues of a repeatedly occupied settlement. Although it is possible (and perhaps likely) that none of these places were utilized without interruption into the Orange period, they all played active roles in the momentous events of that era. (p.93)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.3. Preceramic Thornhill Lake phase domestic deposits and artifacts from 8LA1W’s Locus B: (A) photograph and generalized stratigraphic drawing of the north profile of Test Unit 46; (B) select stone and bone tools and ornaments; and (C) a burned-out lightning whelk (Busycon contrarium) vessel.

Photographs by the author.

(p.94) Orange Interventions

The material vestiges of Silver Glen’s preceramic places factored significantly into the interactions and decision making of the complex’s Orange period inhabitants. In some locations, such as Locus A and 8LA4242, ancestral deposits were apparently actively avoided. Not to be confused with ignorance of or indifference toward the past, the dearth of Orange activity at such prominent points on the landscape instead likely reflects the lingering social power of such places and corresponding efforts to maintain an appropriate cultural boundary or buffer around them (sensu Munn 1996). Several other Mount Taylor shell ridges and mounds in the middle St. Johns region also remained seemingly untouched even as substantial Orange deposits accrued only a few tens of meters away (Randall 2007; Randall and Sassaman 2010).

This treatment of defunct historical places contrasts sharply with that bestowed on sites such as 8LA1E, 8MR123, and to a lesser extent Locus B, where the past was directly and persistently engaged by Orange peoples via repeated acts of excavation, circulation, and deposition. The perpetrators of these events drew on Silver Glen’s by-that-point abundant historical resources to help gather different people, places, and times together into an arena of common engagement. In doing so, they durably altered the complex’s material and social landscape and significantly impacted the historical trajectory of the broader region.

Monumental Transformations (8LA1E and 8MR123)

The most dramatic landscape transformations instigated by Orange groups at Silver Glen involved the construction of the massive shell mounds at 8LA1E and 8MR123. As already noted, both mounds were severely impacted by historic shell mining, which removed the vast majority of their aboveground deposits. Nevertheless, premining descriptions of these monuments, along with extensive testing of their intact remnants, have revealed a great deal regarding the mounds’ respective configurations, contents, and social histories.

Today, few obvious signs exist of the 8LA1E mound that once dominated the Silver Glen landscape. However, a comprehensive auger survey conducted in 2007 as part of the St. Johns Archaeological Field School largely confirmed Wyman’s (1875) characterization of the mound’s configuration and horizontal extent (figure 3.4; Sassaman 2011b: 37). Eighty-four auger tests were conducted at 20-meter intervals across all open portions of the (p.95)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.4. Map of 8LA1E showing the locations of auger tests and test units, along with the current distribution of shell in relation to the estimated boundaries of the mound prior to twentieth-century shell mining.

Figure adapted from Sassaman 2011b.

site, and observations were recorded on the presence/absence of shell, the depth and condition (crushed, whole, burned) of shell, and the presence/absence of nonshell midden (Sassaman 2011b: 38–39). These tests revealed the footprint of a sprawling U-shaped shell construction paralleling the spring run in its longest dimension and opening toward the west. Composed primarily of freshwater gastropod and bivalve shell, the mound is estimated to have stretched for 300 meters along the run and to have approached 200 meters in width. According to Wyman (1875: 39), it originally exceeded 8 meters in height at its apex in the northeastern corner of the site. In stark contrast to the mound itself, its interior, which Wyman (1875: 39) described as forming a “deep valley” between the two ridges, was found to be virtually shell-free.

Auger tests along the banks of the Silver Glen Springs run and Lake George, in the area originally encompassed by the mound’s north ridge, (p.96) revealed shell deposits as deep as 2.5 meters below the modern surface. The concreted state of these basal deposits and their position well below the current water table suggest that they were deposited during preceramic times, perhaps as part of a Mount Taylor (mortuary?) ridge that fronted the spring run (Sassaman 2011b). Excavation of four test units and numerous core samples in the mound’s northeast corner failed to locate any undisturbed deposits. They instead uncovered a complicated amalgam of shell, sand, and artifacts that were most likely displaced and redeposited during the course of shell mining (Sassaman 2011b). Excavated artifacts from this area consist mostly of ornately decorated Orange pottery sherds (discussed in detail in chapters 4 and 5), hundreds more of which have been collected from the surrounding ground surface, nearby tree throws, and the shallow waters of the adjacent spring run and lake. Seven radiocarbon assays obtained from soot adhering to the exterior surfaces of Orange sherds and from charred Spanish moss fibers preserved within their fabric indicate that Orange deposition at 8LA1E’s north ridge began as early as 4600 cal BP and persisted until at least 3800 cal BP (appendix A; Sassaman 2003b).

Since 2007, various archaeological techniques have been brought to bear on the study of 8LA1E’s south ridge, including augering/soil probing, ground-penetrating radar (GPR) survey, and test unit excavation (Sassaman 2011b). Initial auger tests demonstrated that unlike the north ridge, where Orange deposits were emplaced over a preexisting Mount Taylor ridge, the south ridge was constructed directly atop the Late Archaic ground surface (Sassaman 2011b: 40). This was corroborated through excavation of ten 2 × 2 meter excavation units along the south ridge that showed an expansive, yet discontinuous, layer of shell overlying shell-free sand. Excavations uncovered few artifacts in the shell deposits save for a modest amount of mostly plain Orange pottery. Underneath the shell, a maze of mineralized roots, including entire palm root balls, were encountered in the underlying sand (figure 3.5). These are thought to have resulted from water percolating through overlying shell deposits and bringing down calcium carbonate minerals that covered and ultimately preserved the structures of subterranean root systems. The abundance of both linear hardwood roots and basketball-sized root balls may be an indication that many trees were deliberately felled in this area in preparation for the emplacement of shell (Sassaman 2011b: 90). Charcoal from a basal pit feature uncovered near the center of the south ridge yielded an age estimate of 4060–3830 cal BP (Sassaman 2011b: 60), suggesting the possibility that the south ridge may have been a relatively late addition to the 8LA1E monument. Because the (p.97)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.5. Mineralized root masses beneath the shell deposits at 8LA1E’s south ridge: (A) in plan view and (B) in vertical cross section.

Photograph courtesy of the Laboratory of Southeastern Archaeology.

(p.98) ridge was truncated by shell mining, its maximum height is impossible to know; however, Wyman (1875) is clear that it did not reach the stature of its northern counterpart.

Even fewer details exist regarding Orange contributions to the 8MR123 mound north of the spring run. Orange activity at the mound is clearly indicated by the abundant fiber-tempered sherds that have been recovered from the surface around the spring pool, a majority of which, like those from 8LA1E’s north ridge, bear elaborate incised decorations (see chapters 4 and 5). However, while test unit excavations into surviving remnants of the mound (Marrinan 1990; Randall et al. 2011) have yielded extensive evidence for large-scale shell deposition during the preceramic Thornhill Lake phase, intact Orange deposits have been intersected in only one location, near the eastern margin of the spring pool. There, excavations revealed a highly concreted shell deposit that did not yield any Orange artifacts but did return a radiocarbon age estimate of 4410–4080 cal BP, well within the Orange period (Randall et al. 2011: 107). Ten meters away, a subaqueous radiocarbon sample obtained with a bucket auger at a depth of approximately 2.5 meters below the current surface returned a temporally overlapping age estimate of 4520–4240 cal BP (Randall et al. 2011: 107). These data imply that more than 2 meters of shell were deposited at 8MR123 during the Orange period, at approximately the same time that large-scale deposition was also taking place at 8LA1E. At 8MR123, Orange depositional efforts, when added to the preexisting sand mortuary and expansive shell deposits, culminated in the vast and visually imposing “amphitheater” of shell described by Wyman (1875: 39).

Given the extremely denuded postmining condition of the 8LA1E and 8MR123 mounds, we cannot fashion a traditional linear narrative of their construction and use. These mounds are close to what Geoff Bailey (2007: 203) refers to as “true palimpsests,” meaning that most of the evidence related to their histories of use has been removed by recent activity. Nevertheless, an enlightening, albeit partial, account can be pieced together of the various Orange period constituents that were gathered in these locations and the relationships that were achieved in the course of particular depositional events. The fact that portions of both mounds were grafted onto preexisting mortuary facilities suggests a deliberate attempt to connect with ancestors through renewal of a deep-seeded depositional tradition. Throughout the Mount Taylor period, mounds and ridges in the middle St. Johns valley were almost invariably superimposed over former places of everyday living, a clear sign of these sites’ enduring social importance (p.99) to descendant communities (Randall 2010; Sassaman and Randall 2012). During the Orange period, a substantial shift occurred, as most preceramic sites were apparently actively avoided even in cases when the landscape surrounding them was intensively occupied. The only exceptions were a few select places, such as Silver Glen, where temporal boundaries were collapsed and the past was once again engaged through strategic deposition at a historically significant location. The architects of the first Orange period additions to 8LA1E and 8MR123 were, in this way, strongly influenced by the material remains of what came before them. They also, however, set an effective precedent for future activities in these places, which, over the next five centuries, involved social gatherings during which copious amounts of shell and pottery were deposited onto the ever-growing monument and into the water alongside it.

The respective temporalities involved in the construction of the two mounds are unknown; however, the sheer scale of the structures speaks to the extraordinary nature of the social gatherings that they must have facilitated. This is especially evident when the mounds are considered in relation to contemporary places of everyday living. As discussed in chapter 1, most aspects of Orange settlements point to a dominant tradition of small coresidential groups and/or high levels of residential mobility. By comparison, the Orange period mounds at Silver Glen stand out as disproportionately massive and permanent features on the Late Archaic landscape. Their construction would have required the assembly of unusually large and diverse groups of people, repeated returns to the same place over a very long period of time, or, perhaps most likely, some combination of the two. The unique social conditions fostered by the mounds are evinced by high frequencies of elaborately and boldly decorated pottery when compared to other kinds of contemporary sites, a situation that has also been noted in reference to other Orange mounds in the region (e.g., Gilmore 2011; Sassaman 2004a; Saunders 2004a, 2004b). This disparity is discussed in detail in chapter 4; suffice it to say here that these sorts of context-restricted stylistic embellishments may be a sign of the relatively high degree of cultural diversity and ritual significance implicated in mound-centered interactions at Silver Glen (e.g., Mills 1999, 2007; Spielmann 2004).

A final point to consider regarding Silver Glen’s Orange period mounds is the “patently dualistic” (in the words of Sassaman and Randall [2012: 71]) configuration of 8LA1E. The organization of social space into a U is a common cross-cultural means of bringing together two discrete groups into a common sociality, with leadership positions typically situated at the (p.100) closed end of the U (Grøn 1991). 8LA1E’s two parallel ridges are asymmetrical and artifactually distinct, with the north ridge exhibiting a higher maximum height and width, a higher frequency of Orange pottery, and a larger proportion of decorated vessels than its southern counterpart. Pointing to these factors as well as several seemingly anomalous cultural transformations that accompanied the adoption of pottery in the middle St. Johns valley, Kenneth Sassaman and Asa Randall (2012) hypothesize that 8LA1E’s U shape may in fact reflect a dual social organization resulting from the coalescence of indigenous (Mount Taylor) people and extralocal migrants, likely originating from the Atlantic Coast. In this scenario, the mound’s asymmetries may relate to the differential abilities of established and newly arrived groups to amass the social and material resources involved in mound construction (following Russo 2004; Saunders 2004b). These possibilities are considered further in subsequent chapters.

In summary, despite the poor current condition of Silver Glen’s Orange period shell mounds, clearly they functioned as important nexus points for extended Late Archaic communities. They were gathering places where broad-scale relations were condensed and negotiated. The symbolic renewal of preceramic depositional traditions and the superimposition of Orange mounds over old mortuaries would have been effective strategies for connecting the present to the past and thereby helping legitimize the dramatic social and material transformations that were under way across the region at the time. The unprecedented scale of the mounding projects undertaken points to the contribution of innumerable individual depositional events, each of which drew on memories of the past while simultaneously helping establish a durable material framework that would persist far into the future.

Pit Events and Place-Making at Locus B

At the same time that vast quantities of shell were being heaped on the surface at 8LA1E and 8MR123, equally impressive amounts were interred in subterranean pits at 8LA1W’s Locus B (Gilmore 2011). Fortunately, the relative inconspicuousness of Locus B’s shell deposits meant that they were spared the destructive effects of shell mining suffered by other parts of the complex. As a result, Locus B provides the basis for a much more nuanced archaeological account than that possible for the just-discussed shell mounds.

Locus B (figure 3.6) occupies a relatively flat, well-drained ridge nose approximately half a kilometer southwest of 8LA1E and 80 meters south of the (p.101)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.6. Map of Locus B at 8LA1W showing the locations of excavated test units and surveyed GPR grids.

(p.102) spring run. As noted above, it consists of a small, roughly crescent-shaped shell node and the extensive subsurface cultural deposits that surround it. The shell node opens toward the spring run and rises only about a meter above the surrounding landscape at its highest point. A close-interval auger survey in 2008 showed that the aboveground shell at Locus B was deposited atop a relatively flat sand surface and that the site’s current topographical configuration is almost entirely the product of anthropogenic activities conducted during the Late Archaic period (Gilmore 2011: 174). Between 2007 and 2013, 81 square meters of test unit excavations were conducted at Locus B, revealing well-preserved and highly stratified deposits that register the site’s dynamic depositional history and shifting relationship to the broader Silver Glen landscape.

Three distinct depositional patterns were found to have existed at Locus B over the course of the Late Archaic period (figure 3.7; Gilmore 2011: 249–252, 2015a). The earliest substantial deposition in this area of the complex was centered in the location of the shell node. As already discussed, this portion of Locus B was utilized during the late preceramic Thornhill Lake phase as the site of a small-scale, intermittent settlement. The deposits making up this component include a series of stacked occupational surfaces lined with thin layers of freshwater bivalve shell mixed with occasional vertebrate fauna and a variety of tools and debitage made of stone, bone, and marine shell. Many small pits were dug down from these surfaces and infilled with similar materials. Just west of the shell node, dense deposits of vertebrate faunal and paleofeces are also thought to date to this same period. The broad variety and relatively high frequency of artifacts and features suggest a domestic, residential use of Locus B during this interval.

The second depositional pattern began coincident with the appearance of Orange pottery circa 4600 cal BP, transforming Locus B from a small settlement to a specialized shellfish-processing locale replete with scores of massive, overlapping pits (table 3.1). These pits are mostly distributed across a gently sloping area just off the western edge of Locus B’s shell node, although isolated examples have also been uncovered farther to the east and west. Similar Orange period pits also occur near the high point of 8LA1W’s Locus C and in the uplands northwest of the 8MR123 shell mound (Randall et al. 2011); however, the relatively limited excavations in these areas make direct comparisons to Locus B difficult. The pits at Locus B are, for the most part, densely packed, and they frequently intersect, many apparently having been dug one on top of another (figure 3.8). The scale of these features dwarfs virtually anything found either before or after in (p.103)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.7. Schematic showing the Late Archaic depositional sequence at 8LA1W’s Locus B.

Figure originally published in Gilmore 2015a; reproduced courtesy of the University of Alabama Press.

(p.104)

Table 3.1. Late Archaic pit features at 8LA1W’s Locus B

Feature

Shape

Length × width or diameter (cm)

Depth (cm)

Culturehistorical affiliation

Deposition pattern

14C age estimate (cal BP)a

1

Basin

25 (dia.)

14

Orange

2

4

Basin

45 × 27

10

Orange

2

14

Basin

17 × 12

8

Orange

2

15/65/95

Basin

223+ × 215+

85

Orange

2

4410–4100

16

Basin

100+ × 90+

80?

Orange

2

17

Amorphous

45 × 22

20

Thornhill?

1

25

Amorphous

38 × 30

28+

Thornhill?

1

26

Cylinder

103+ × 82+

65+

Orange

2

4520–4300

27

Cylinder

100+ (dia.)

50+

Orange

2

33

Cylinder

100 × 66

84

Orange

2

4230–3980

34

Basin

87 × 83

12

Orange

2

35

Basin

82 × 74

16

Orange

2

36

Basin

120+ × 80+

~90

Orange

2

3980–3830

37

Basin

55 × 34

19

Orange

2

4080–3850

38

Cone

120 × 72+

94

Orange

3?

4140–3830

41

Basin

Indet.

Indet.

Orange

2

42

Cylinder

100+ (dia.)

Indet.

Orange

2

45

Cylinder

71 × 58

70

Orange

2/3?

48

Basin

230 × 135+

42

Thornhill

1

4860–4650

49

Basin

54 × 23

20

Thornhill

1

50

Cylinder

45 × 43

31

Thornhill

1

4840–4580

51

Cylinder

140+ × 100+

102

Orange

2

52

Basin

67 × 27

65

Orange/St. Johns?

2/3?

54/55

Basin

227+ × 134+

50+

Orange

2

4140–3900

64

Basin

50 × 43

16

Orange

2

66

Cylinder

117 × 79

66

Orange

2

90

Cylinder

56 × 47+

75

Orange

4150–3980

91

Basin

74 × 64

29

Orange

94

Basin

72 × 58

Indet.

Orange

2

104

Cylinder

151+ × 70+

106

Orange

2

4080–3880

121

Basin

70 × 60

13

Orange

187

Cylinder

128 × 43+

85

Orange

2

196

Basin

104+ × 100+

33

Orange

2

197

Basin

114+ × 49+

27

Orange

2

198

Basin

77 × 59+

33

Orange

2

199

Cylinder

106 × 90+

46

Orange

2

4810–4450

(a) Based on the total two-sigma range of AMS radiocarbon assays (see appendix A).

(p.105)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.8. Horizontal distribution of Late Archaic pit features in the block excavations at 8LA1W’s Locus B. All elevations listed in meters below datum.

(p.106) the complex’s eight-thousand-plus years of prehistoric occupation. Most examples measure between 70 and 120 centimeters in diameter, and many exceed 1 meter in depth. The largest is well over 2 meters wide and more than 1 meter deep, exhibiting an estimated volume of over 2.5 cubic meters. Pit shape varies considerably, ranging from broad, deep basins to narrow, straight-sided shafts (Gilmore 2011, 2015a).

The fill in the Locus B pits also varies widely. Several of the pits have lenses of bright orange oxidized sand, charcoal, and concreted whole mussel shell lining their bottoms, suggesting that at least some of them were used for roasting shellfish. None contains a substantial quantity of either vertebrate fauna or artifacts, save for modest amounts of fragmented and undecorated Orange fiber-tempered pottery. While all the pits contain some quantity of shell, the frequency, composition, condition (that is, the degree of crushing, burning, and weathering), and structure of shell deposits are quite heterogeneous. Some are filled primarily with sand and contain only a trace of shell, while others appear to have been infilled in one massive depositional episode. The most striking features, however, are those containing layer after layer of shell of different types and conditions, a situation indicative of multiple discrete depositional acts. Deposition into one large pit, Feature 38 (figure 3.9), for example, began with a thick layer of dense whole, banded mystery snail (Viviparus georgianus) that over time became concreted. A 20-centimeter-thick layer of crushed and burned mussel shell was then deposited before the pit was finally topped off with sand and another layer of whole and crushed mystery snail. In another example, Feature 104 (figure 3.10), infilling began with the deposition of a 20-centimeter-deep layer of mixed shell (including apple snail [Pomacea], mystery snail, and bivalve) followed by a layer of unusually large whole apple snails. A thin lens of mostly shell-free sand was then either emplaced or simply allowed to accumulate in the still-open feature. Subsequently, another layer of whole apple snail was laid down, followed directly by a layer of whole paired and unopened mussel shells. On top of the mussels was a thin stratum of very dark, almost black, organically enriched sand and, finally, a layer of lighter brown sand. While virtually every pit contains a unique fill sequence, the constituents of individual layers are replicated across pits, suggesting that they may have been combined according to particular “recipes” or “grammars” in different locations; however, no consistent pattern has yet been discerned.

Shortly after the cessation of large-scale pit digging (circa 3900 cal BP), a large quantity of whole Viviparus shell was deposited across the surface (p.107)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.9. Drawing and photograph showing the excavated cross section of Feature 38 from 8LA1W’s Locus B.

Photograph courtesy of the Laboratory of Southeastern Archaeology.

Late Archaic Depositional Narratives at Silver Glen

Figure 3.10. Drawing and photograph showing the excavated cross section of Feature 104 from 8LA1W’s Locus B.

Photograph courtesy of the Laboratory of Southeastern Archaeology.

(p.108) of Locus B, an event that marked another major transition in the site’s history (figure 3.11). This “shell cap” (the third depositional pattern at Locus B) forms a 30–50-centimeter-thick, mostly homogeneous stratum of un-consolidated shell that in many places contains little or no soil matrix. Like the pits below it, this stratum contains only sparse vertebrate fauna and artifacts aside from a small amount of fiber-tempered pottery. In contrast to the undecorated sherds from the pits, however, many of those recovered from this overlying deposit exhibit the curvilinear incisions and punctations typical of a relatively rare variety of Orange pottery called Tick Island Incised (Griffin 1945). The overall homogeneity of the deposit, the lack of evidence for trampling, and the paucity of vertebrate fauna all indicate that this layer of shell was emplaced relatively rapidly, probably in the course of one or a few large-scale depositional acts. Importantly, the shell cap is virtually coextensive with the pits underlying it and in some places appears to have infilled open pits, in effect turning what must have been a pocked, uneven surface into a relatively flat and smooth one. This mantle of shell is not unlike the ones that have been found covering discontinued Mount Taylor habitation sites (Randall 2010; Sassaman 2010; Sassaman and Randall 2012) and perhaps constitutes yet another renewal of a long-lived tradition of ritually marking a transition in the use of a place by capping it with clean, whole shell.

Unfortunately, because they are ubiquitous in many regions and may appear largely interchangeable upon cursory examination, pit features are, more often than not, lumped together and given little weight in archaeo-logical interpretations. Pit fill, in particular, which is generally assumed to be unrelated to a feature’s primary function, is prone to being dismissed as mere secondary refuse, a result of casual discard into a convenient receptacle. As John Chapman (2000b: 61) notes, the “humble pit” represents a class of feature that has been “much maligned, ignored, or otherwise maltreated” in many archaeological narratives.

With regard to Locus B, the tendency to undervalue pit deposits is exacerbated by a general reluctance on the part of many regional archaeologists to attribute a cultural significance to shellfish beyond their status as an abundant subsistence resource (e.g., Crothers and Bernbeck 2004; Marquardt 2010a, 2010b; Trinkley 1985). Because of this firmly entrenched perspective, pit deposits such as those uncovered at Locus B are unlikely to be investigated for any purpose beyond the reconstruction of Late Archaic dietary habits (cf. Blessing 2015; Gilmore 2015a). In contrast, I would argue that acts of pit digging and deposition represented more than the mindless (p.109)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.11. Excavated vertical profile of the (2-meter-wide) east wall of TU58/59 at 8LA1W’s Locus B showing a stratified Orange period pit capped with dense banded mystery snail (Viviparus georgianus) shell.

Photograph courtesy of the Laboratory of Southeastern Archaeology.

repetition of subsistence-related behaviors, conducted for the same reasons and in the same manner independent of context. Instead, like all social projects, they reflect the historically conditioned decisions of knowledgeable and intentional actors occupying particular material and social settings. As such, it should come as little surprise that, at least occasionally, pit-related practices constituted important historical events with substantial roles in processes of cultural categorization and meaning production.

(p.110) Multiple attributes of Locus B pits suggest that they held significance beyond their practical utility as, initially, shell-roasting facilities and, subsequently, refuse containers. Most obvious is their size. While several contemporary shell-matrix sites in the broader region include shell-filled pits (e.g., Blessing 2015; Janus Research 1995; Saunders 2004a; Trinkley 1985), the extraordinary size and frequency of those at Locus B set them apart from other documented feature assemblages. The sheer scale of digging and shellfish processing suggested by the pits seems out of proportion with the everyday subsistence requirements of small kin-based hunter-gatherer groups occupying a diverse and productive environment. With no evidence suggesting, and presumably little need for, long-term storage at the site, an alternative possibility is that the pits were geared toward the rapid production of great amounts of food, perhaps for consumption at the periodic feasting events hypothesized to have taken place at the nearby shell mounds (i.e., 8LA1E and 8MR123).

The content of the pits also suggests a meaning beyond the purely mundane. Unlike some earlier instances of ritualized deposition in the same region (e.g., Endonino 2008; Wheeler et al. 2000), the Locus B pits are not marked by an abundance of unusual or exotic objects, save for one modified deer mandible that was likely part of a mask and a few marine-shell disk beads found scattered across multiple features. They are instead distinguished more by a paucity of many materials frequently found within general midden deposits throughout the region (e.g., Russo et al. 1992; Sassaman 2003a; Sassaman and Randall 2011), including vertebrate fauna, lithic/marine-shell tools and debitage, and paleofeces. As shown in figure 3.12, the frequency of bone recovered from 260 liters of analyzed fill from the Locus B pits is only a tiny fraction of that observed in contemporaneous domestic midden deposits from the same region. This virtual absence of vertebrate fauna supports the notion that the contents of the Locus B pits do not represent a random sample of general midden materials but instead constitute an intentionally selected subsample, one largely restricted to shellfish remains. If the pits had instead been infilled through casual acts of refuse disposal, one would expect them to contain a more complete cross section of the diverse materials employed in everyday domestic tasks.

As noted above, the primary constituent of the pit deposits is shell that varies considerably in terms of both size and condition and was combined in a unique manner in every excavated pit, often resulting in elaborate stratified fill sequences. At a roughly coeval site in South Carolina, Michael Trinkley (1985) interpreted similarly layered (although substantially (p.111)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.12. Graph showing the relative abundance of vertebrate faunal remains from the pits at 8LA1W’s Locus B and contemporary domestic middens from the middle St. Johns basin.

Blue Spring data were obtained from Sassaman 2003a, and Groves Orange Midden data were obtained from Russo et al. 1992.

smaller) pits as containing the remains of successive meal dumps. In this scenario, the pits were used repeatedly for roasting shellfish that were removed, consumed, and then disposed of back into the pits before the next batch was processed. Presumably, this sequence was repeated until a pit was topped off and another had to be dug. At least three factors render this interpretation inadequate for explaining the layered pit fills at Locus B. First, evidence for roasting (heat-oxidized sand, burned shell, large charcoal lumps), where it exists, occurs in only a single layer lining the bottom of a pit. Given Trinkley’s hypothesis, one would expect a layer of thermal alteration between every individual shell stratum. Second, three of the excavated pits at Locus B contained strata composed of whole, unopened, and unburned bivalve shells, indicating that they were neither cooked nor consumed. And finally, pit deposition at Locus B appears to have taken place rapidly, with very little time elapsing between the deposition of the first layer and the last. Almost all the shell layers within the pits sit directly on top of one another with no intervening sediment accumulation or soil formation. Moreover, none of the more than two dozen massive pit features either excavated or encountered in profiles at Locus B show any evidence of having collapsed in on themselves. My firsthand experience excavating test units into the site’s soft unconsolidated sand indicates that, if left open, pits would have been subject to structural failure during the first substantial rainfall. The fact that not a single excavated example did fail suggests (p.112) that they were infilled almost immediately, not over a protracted period as implied in the “meal-dump” scenario.

An interpretation more consistent with the archaeological evidence is that the pits were infilled soon after they were dug, perhaps as part of a single continuous depositional process involving a variety of materials selected from several different sources. The short period of time indicated between pit digging and pit filling at first seems at odds with the highly weathered condition of the shell composing some pit deposits. In addition, the diverse combinations of shellfish species, along with frequent disparities in the extent of weathering, burning, and crushing between layers in the same pit, make it unlikely that all the shell in the pits underwent the same process of collection, roasting, and consumption prior to immediate deposition. These characteristics instead point to a diverse array of prede-positional taphonomic histories in which some materials were deposited soon after harvesting, some after they were processed and consumed, and still others only after they had been left out in the elements for some length of time. The fact that materials with such diverse histories ended up in the same features negates the likelihood of casual discard and instead suggests that they were intentionally selected for a particular purpose and combined in meaningful ways. Individual shells composing the pit-deposit layers could possibly have derived from specific important events such as feasts or other communal ceremonies and thus may have had to be dealt with in a particular manner (see Walker 1995). These residues may have been stockpiled for some period of time before their inclusion in a pit. If so, then the layered pits may have served as “bundled” histories (as discussed in chapter 2), used for linking particular events together and ordering them to form complex historical narratives. As with all bundles, it was not so much the inherent value of the substances themselves (in this case, individual shell deposits) but rather the symbolic transposal achieved via their combination that made them meaningful.

In effect, these deposits would have constituted inverted, subterranean shell mounds, homologous to the countless aboveground monuments that marked and structured the Late Archaic landscape. However, unlike aboveground mounds, which rely on being seen and interacted with for much of their effect, the underground “mounds” at Locus B were completely obscured even as they were constructed. One might question what the point was of building a subterranean monument that no one would ever see. Part of the answer to this question may be provided by Suzanne Küchler (1999) and others (e.g., Gillespie 2008; Hendon 2010: 113; Mills 2008), (p.113) who argue somewhat paradoxically that the memory of an event can be heightened or reinforced through symbolic acts of forgetting, which include the destruction or concealment of associated objects. Based on this idea, the deposition and burial of residues from important occurrences such as feasts may have functioned to memorialize these events, as well as the places where they transpired. This is the basic idea used by Julian Thomas (1999b: 72, 2000: 80) to explain the common Neolithic practice of siting monuments atop assemblages of old infilled pits, some of which had been dug generations earlier. However, in this case, while the burial of shell deposits may well have enhanced particular memories, the distributional data show that many of the Locus B deposits were in fact reexposed and viewed as new pits were dug that intercut old ones. These would have been encountered by Late Archaic people in much the same way that they are by modern archaeologists—in profile, with the entire sequence of deposits made visible and begging for interpretation. Far from ancillary, this aspect of the Locus B pits holds the key to understanding their eventfulness.

As already discussed, the features in question were sited in a location that had been used previously by preceramic people as a small-scale settlement. The first Orange period pits excavated at Locus B would have intersected deposits from this earlier occupation (figure 3.13), granting their diggers access to a relatively distant past and perhaps adding another layer of meaning to the massive-scale roasting activities taking place there. As this practice was repeated through time, Orange pits also began to intersect each other, exposing material reminders of more recent people and occurrences (figure 3.14). Eventually, as these features covered the site, the encountering of old infilled pits must have become the expected outcome of, and probably even added motivation for, continued digging. The cumulative effects of these repeated material engagements can be seen by examining diachronic changes in pit deposition practices. When dated pits are placed in chronological order, a pattern emerges that shows increasing depositional complexity through time (figure 3.15). While the earliest pits appear to have been filled in one or a few distinct episodes, the later ones (e.g., Features 38 and 104) exhibit more elaborate sequences of shell and earth. In addition, most of the pits with the most complex layered fills exhibit no basal oxidation or any other evidence that they were ever used for roasting.

As soon as the first pit was excavated and infilled at Locus B, it would have exerted a structuring influence on all subsequent digging in that location because of its enduring material presence and the memories it (p.114)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.13. Excavated vertical profile of the (2-meter-wide) south wall of TU63 at 8LA1W’s Locus B showing an Orange period pit feature intersecting preexisting Mount Taylor cultural deposits. Photograph courtesy of the Laboratory of Southeastern Archaeology.

Figure originally published in Gilmore 2015a; reproduced courtesy of the University of Alabama Press.

facilitated. As time elapsed and old pits were uncovered with growing frequency, pit deposition may have become an increasingly deliberate effort to write a particular history into the Locus B landscape, with the knowledge that it would eventually be uncovered by subsequent digging. By the final stages of large-scale pit digging at the site, many pits appear to have been dug for the explicit purpose of receiving shell deposits. At this point, pits were no longer just a means of memorializing other events such as mounding and feasting ceremonies. They instead became important events in their own right, gaining influence by citing (sensu Jones 2005) already well-established shell-mound traditions but altering them in important and strategic ways. Unlike their aboveground counterparts, which were (p.115)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.14. Excavated vertical profile of the (4-meter-wide) north wall of TU38/39 at 8LA1W’s Locus B showing a series of massive, overlapping Orange period pits.

Photograph courtesy of the Laboratory of Southeastern Archaeology.

(p.116)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.15. Chronologically ordered (from top left to bottom right) vertical profiles of Orange period pits from Silver Glen Run’s Locus B.

(p.117) susceptible to being observed and experienced by anyone within a certain distance of them, the buried mounds at Locus B would have allowed the site’s Orange period inhabitants to regulate the timing and circumstances of their opening (much like a traditional bundle or modern time capsule), possibly in ways that heightened their impact. Like more traditional monuments, while memorializing and relying on the authority of the past, the Locus B pit deposits were oriented primarily toward achieving a particular future by preconfiguring a lasting point of reference into the landscape. Depositional practices, in this context, were part of a deliberate strategy, or “intervention” (Sassaman 2012), geared toward the production of future memories (cf. Eves 1996).

Circular Living: Silver Glen’s Orange Habitation Spaces

The immense shellworks at 8LA1E and 8MR123, along with the sprawling assemblage of hypertrophic pits at Locus B, provide tantalizing clues as to the nature of Silver Glen’s public gatherings and projects, events that drew large quantities of people and materials into the complex. These locations, however, have yielded very few signs of the kinds of routine activities associated with day-to-day habitation. Even if Orange period Silver Glen was a designated ceremonial center, sparsely populated outside of special occasions, people would still have needed to sustain themselves for the duration of their visits. Consequently, even in the absence of long-term occupation, one would still expect to find substantial traces of domestic activity such as food remains, hunting implements, cutting and processing tools, and the like. So far, the best evidence for such activities has come in the form of circular arrangements of materials and features found scattered across less conspicuous areas of the complex.

In 2010, a GPR survey was conducted across areas of the south ridge at 8LA1E, the technical parameters of which are detailed by Sassaman (2011b). In short, GPR was deployed in two ways and with two distinct goals in mind. First, a series of 30-meter north-south transects, spaced 10 meters apart, was surveyed in hopes of delineating the northern edge of the southern ridge of the 8LA1E mound. The results of these efforts were ambiguous, as the edge of the shell deposits was found to be nonlinear and discontinuous, likely because of historic mining and other recent landscape modifications (Sassaman 2011b: 69). Second, and most pertinent to the present discussion, GPR was also deployed across five contiguous grids (ranging from 20 × 30 meters to 7 × 10 meters in area) near the center of the mound’s south ridge. In this case, for each grid, the GPR unit was pulled (p.118)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.16. GPR results from 8LA1E’s south ridge showing “time slices” for five contiguous grids. The gray line indicates a projected arc of dense shell anomalies.

Figure adapted from Sassaman 2011b: 720.

(p.119) along north-south transects spaced 50 centimeters apart. The orientation of the transects was then turned 90 degrees, and perpendicular east-west data were collected in the same manner. The goal was to obtain a series of “Z slices” or “time slices” at different depths that would help shed light on the horizontal and vertical arrangement of subsurface shell deposits along the south ridge (Sassaman 2011b: 68–69).

Figure 3.16 shows the time slice at approximately 50 cmbs for the five contiguous GPR grids. The diagram’s various colors represent different degrees of reflectivity, with dark colors indicating low reflectivity and light colors high reflectivity. As shown in the figure, four clusters of highly reflective anomalies were found that each measure approximately 5–10 meters in diameter. Interestingly, the distribution of these anomalies is clearly patterned and appears to form an arc that, if projected out to form a complete circle, would exhibit a diameter of 80 meters and contain a total of sixteen to seventeen clusters, spaced some 15 meters apart (Sassaman 2011b: 73).

Ground truthing of these results was attempted through a combination of coring and controlled excavations. These efforts revealed that the GPR anomalies coincide with more-or-less discrete areas of dense shell that was deposited both atop the sandy substrate and into pit features (Sassaman 2011b). A likely—albeit admittedly unconfirmed at this point—explanation for the observed patterning is that these deposits resulted from one or more short-term encampments that were occupied near the onset of shell mounding in this part of 8LA1E. A single radiocarbon assay from one of the shell-filled pits suggests that this encampment would have been active circa 4060–3830 cal BP (Sassaman 2011b: 60). Excavated shell deposits contained modest amounts of vertebrate fauna and mostly plain Orange pottery sherds, but no clear domestic features (such as hearths, post molds, and house floors) were discovered. This could indicate a nondomestic origin for these deposits or, alternatively, may simply reflect the ephemeral nature of the occupation in question. In either case, a roughly circular organization of social space would be consistent with what we know about Orange cultural traditions from other locations, both within and outside the middle St. Johns region (e.g., Russo 2004; Sassaman 2003a; Saunders 2004a).

A similarly arranged residential space has been inferred north of the shell mound at 8MR123 (figure 3.17). Using data from shovel tests and stratigraphic excavations, Randall and colleagues (2011) documented an arcuate pattern of large Orange pottery sherds in this location with an interior diameter of approximately 70 meters. Inside this arc, only crumb (p.120)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.17. Map of site 8MR123 showing the location and configuration of Asa Randall’s (2011) suspected Orange settlement.

Map adapted from Randall et al. 2011: 206.

sherds were recovered, suggesting the possibility of an interior plaza area that was intentionally kept clean. The arc itself provides strong evidence for the presence of Orange architecture. Subsurface testing of the arc revealed a series of regularly spaced voids measuring approximately 25 to 30 meters. In a test unit sited just outside of one of these voids, numerous large Orange pottery sherds were found lying flat at the same elevation, along with a small shell-filled feature believed to represent an infilled posthole. Randall and colleagues (2011: 207) interpret these data as evidence that the test unit intersected the remains of an Orange period domestic structure of unknown dimensions. They hypothesize that this was but one of a number (p.121) of such structures that were arranged around small plazas evinced by the aforementioned voids. They further suggest that these individual house clusters were, in turn, organized around the much larger plaza mentioned above.

These data from 8MR123 provide some of the strongest evidence yet for sustained, village-like habitation of Silver Glen during the Orange period. In terms of size and circular layout, the inferred 8MR123 village closely resembles the just-discussed deposits at 8LA1E’s south ridge. Where the two clearly diverge, however, is in the presence of shellfish remains. Whereas the possible encampment at 8LA1E was defined largely according to the distribution of dense shell, the 8MR123 village is characterized by an almost complete absence of shellfish, a fact highlighting the need for greater attention to shell-free components at Orange period sites (Randall et al. 2011: 207). Unfortunately, the precise chronological relationship between these two contexts is unknown; however, a radiocarbon assay from a pit feature that was located near the 8MR123 village and contained incised fiber-tempered pottery returned an age estimate of 3560–3360 cal BP, placing it somewhat later in the Orange period (Randall et al. 2011: 128).

Yet another potential Orange residential area at Silver Glen is located in what is now an open, plowed “bait field” just northwest of the shell node at 8LA1W’s Locus B. In 2011, a GPR survey was conducted across two contiguous 30 × 30 meter grids (shown in figure 3.6) that together span most of the cleared portion of the field. As at 8LA1E’s south ridge, the GPR unit was deployed along grid transects in two perpendicular directions (north-south and east-west), and time slices were obtained for various depths below the Locus B surface. Two of these time slices (42–51 cmbs and 114/115–124/125 cmbs) are shown in figure 3.18. They reveal numerous highly reflective subsurface anomalies, many of which appear to be arranged into circles with diameters ranging from 12.5 to 17.75 meters. The best defined of these circular patterns is located in the southern half of Grid 1 and is clearly visible from approximately 30 to 75 cmbs. At least three additional circular arrangements are apparent across the two grids in time slices exceeding 100 cmbs.

Between 2010 and 2013, six 2 × 2 meter test units were excavated in the area of the shallowest circular pattern in Grid 1. Two of these units were positioned so as to intersect GPR anomalies at the northern and southern margins of the circle, while the others were arranged along perpendicular radii within the circle’s interior. Not surprisingly, the anomalies were found to consist of densely packed shell-rich pit features that were dug into a fine (p.122)

Late Archaic Depositional Narratives at Silver Glen

Figure 3.18. GPR results from the “bait field” at 8LA1W’s Locus B showing “time slices.” White lines indicate the outlines and diameters of circular arrangements of dense shell anomalies.

(p.123) sandy substrate. The exposed pits showed a range of size and morphology. While most were relatively small and consistent with typical domestic storage and/or food preparation, at least one (Feature 54/55) along the southern margin resembled the massive basin-shaped roasting pits found farther south at Locus B (see Gilmore 2011: 269). The pits also varied significantly in the amounts of shell and other materials that they contained. While virtually all pits held some quantity of shell, bone, and fiber-tempered pottery, none would be described as “artifact-dense.” A wispy shell layer covering the pits suggests that these features, like others at the site, may have received a shell cap at the end of their use-life. If this was indeed the case, however, much of the overlying shell must have been displaced by decades of recent plowing. Excavations in the interior of the circle revealed it to be virtually shell-free and devoid of any signs of Late Archaic occupation, although numerous features were uncovered that belong to later St. Johns period groups. Charcoal from two of the pit features (Features 54 and 90) yielded almost identical radiocarbon assays (4140–3900 cal BP and 4150–3980 cal BP, respectively) that overlap significantly with other areas of Locus B as well as the proposed encampment at 8LA1E’s south ridge.

While comparable in overall layout, the circular patterns of features at Locus B are only a fraction of the size of those projected at 8LA1E’s south ridge and the 8MR123 village. Even when considered in relation to the individual household clusters at 8MR123, the Locus B examples do not measure up, although the scalar contrasts in this case are much less dramatic. If the circles at Locus B are, in fact, indicative of Orange period residential activities, then the settlements that they represent must have been exceedingly small and temporary. Excavations revealed no obvious signs of architecture, and the restricted circumferences of the feature patterns would have allowed for only a few domestic structures, even if they were small and closely spaced. Moreover, the artifact densities within and around the pit features are relatively low when compared to the rich middens found at other Orange domestic sites in northeastern Florida (e.g., Janus Research 1995; Russo et al. 1992; Russo et al. 1993: 55–56; Sassaman 2003a). While the precise duration and periodicity of the activities responsible for the observed circular patterns at Locus B are unknown, the sporadic and overlapping distribution of the circles suggests repeated occupations of the site over the course of the Orange period. The limited size of the feature clusters, along with the low density of cultural materials contained within them, suggest that these occupations were relatively short-lived encampments rather than long-lived villages.

(p.124) If accepted as places of Orange period domestic habitation, the circular-patterned deposits at 8LA1E’s south ridge, the 8MR123 village, and the Locus B bait field offer important clues regarding the structure and historical significance of Orange period social interactions at Silver Glen. All three sites contain artifacts and features associated with routine domestic tasks but lack the dense midden deposits that result from long-term residence in the same location. Thus, rather than indicating sustained occupation, the data are more consistent with a series of scattered, ephemeral encampments such as one would expect to arise in the context of periodic aggregation events. At Locus B, these encampments apparently consisted of spatially distinct clusters of features surrounding small open plazas, while in the other two locations individual clusters may have been integrated into larger, more inclusive formations. As already noted, the consistently circular arrangement and clean interiors in all three cases contrast markedly with roughly linear Mount Taylor settlements such as that at Silver Glen’s Locus A. However, as Sassaman (2011b: 73) points out, they are strongly reminiscent of contemporary shell rings found along the Atlantic and Gulf coasts, perhaps signaling the importance of nonlocal influences at the complex (more on this in chapter 5). Regardless of their genealogical origins, though, what is important here is that these circular encampments, and the larger events to which they contributed, constituted important transformative developments in Silver Glen’s Late Archaic history.

Discussion: History in Place

The Late Archaic landscape at Silver Glen is not easy to reconcile with the slow-moving, gradualist narratives typically put forward for Archaic hunter-gatherers. Even a cursory examination of the complex in its current condition should quickly dispel any notion that its immense cultural deposits could have resulted from random or unreflexive discard behaviors or that they were dictated solely by ecological imperatives. To the contrary, when considered from a historical, human-scaled perspective, the picture that emerges at the complex is one filled with significant, historically contingent moments of transformation. These moments are inscribed in the landscape in the form of countless deposits of shell and other materials that Late Archaic peoples used to erect monuments, create spatial and temporal boundaries, and establish novel relationships and community structures. We can also see in various depositional events some of the strategies used by Orange period groups to construct their own historical narratives by (p.125) co-opting preceramic places and using them as effective resources for instituting new material and social realities. Thus, as the physical landscape at Silver Glen was modified, so too were the identity of the place and the relations among those gathered there.

From a distanced perspective, the productive natural environment presented by the Silver Glen Springs watershed was probably a necessary precondition for many of the Late Archaic cultural developments that transpired there. The spring run and lake almost certainly furnished many of the raw materials used in the construction of the complex’s monuments and other features, and the large aggregations argued to have taken place there are unlikely to have been possible in the absence of abundant local subsistence resources. However, at the scale of actual human decision making, ecological conditions alone are woefully inadequate for explaining the material evidence at hand. While the area surrounding Silver Glen is particularly resource rich, it is but one of many large springs feeding into the lake and nearby river channel, any one of which would have provided Late Archaic groups with similar ecological advantages. Moreover, Orange period hunter-gatherers engaged with the Silver Glen landscape in a manner that was highly systematic yet decidedly nonoptimal if guided primarily by subsistence pursuits. The elevated ridges at Locus A and 8LA4242, which served as very successful and long-term settlements for preceramic inhabitants of the site, were left virtually untouched. Meanwhile, massive amounts of shell were piled around the vent of the spring, an area that, based on current spring ecology data, would not have presented high-quality shellfish habitat (O’Donoughue 2011). What this suggests is that something other than food and efficiency must have been driving depositional decisions at the complex.

I would argue that this alternative motivating agent was history. Orange period people chose this particular location as a center of social gathering not because of the adaptive advantages it offered but because of the abundant and conspicuous historical resources available there. Drawing on more than three thousand years of shell mounding tradition, these groups deliberately superimposed their own mounds atop existing mortuaries, building them up and eventually creating some of the largest pre-Columbian monuments anywhere in Florida. The technological and compositional diversity present within Orange pottery assemblages from these monuments (detailed in chapters 4 and 5) indicate that mounding events attracted participants representing multiple cultural groups from across a broad geographic expanse. This is consistent with other evidence, including (p.126) the ephemeral nature of the encampments found strewn about the complex and the overall dearth of Orange period domestic debris, in suggesting that Silver Glen’s mounds were constructed in the context of punctuated moments of intensive occupation rather than through the protracted efforts of a residential population. By gathering copious amounts of shell, pottery, and other materials and depositing them in meaningful ways, mounding participants helped achieve durable connections between the past and present and among the various contemporary constituencies involved.

The relations established through deposition at Silver Glen’s mounds were reiterated in pit-centered practices at Locus B. By gathering diverse materials and holding them together in particular configurations, pits, in many respects, constitute depositional bundles par excellence. As Thomas (2012: 5) notes, “The filled pit is a stable context within which a series of biographies terminate and are ‘bundled together.’” At Locus B, hundreds of oversized pits bundle together layers of fill that acted as indexes of mound-centered consumption events. However, unlike traditional portable bundles that accrue power by changing hands and moving between places, pits are fixed in space but circulate through time as they are excavated through layer after layer of past deposits. Cross-culturally, digging into the earth is frequently thought to have involved the transgression of an important boundary between the past and the present or between this world and another (e.g., Chapman 2000b; Darvill 2012; Davies and Robb 2004; Knight 1989; Kunen et al. 2002; Pauketat 2008; Thomas 1999a). Rather than a neutral form of refuse disposal, then, pit deposition at Locus B may have constituted a quite literal “exchange with the ancestors” as old materials were removed and new ones put in their place (Chapman 2000b: 64). These transactions, clearly visible as intersecting strata in Locus B’s excavation profiles, would have permitted those gathered at Silver Glen to reexperience eventful sequences of the past while also providing a durable medium (much like aboveground shell mounds) for relating their own historical narratives to future diggers at the site. In this way, it is quite possible that the depositional sequences within Locus B pits reproduced, in microcosm, many of the historical and social relations achieved within the nearby mounds.

Importantly, the Orange period deposits at Silver Glen (both above-and belowground) had real historical consequences at multiple scales. As already touched on with regard to Locus B, each depositional act set a precedent for all that followed in the same location. Just as Silver Glen’s preceramic inhabitants left behind enduring material traces that affected (p.127) the decisions of the groups that followed them, so too did initial Orange deposits influence how subsequent activities and interactions at the complex would be executed and organized. This helps to explain how various cultural boundaries (such as the division between 8LA1E’s north ridge and south ridge) and discrete activity areas (such as the space encompassed by Locus B pits) at the complex could be implemented and maintained over a period of centuries without the inherent stability of either a permanent resident population or an institutionalized political structure. Over time, adherence to established depositional traditions resulted in a highly structured Late Archaic landscape at Silver Glen, one featuring many materially and functionally distinct contexts. In this way, the deposits themselves helped define parameters of interaction for those gathered at the complex.

At a larger scale, the massive mounding projects at 8LA1E and 8MR123—along with the large-scale pit digging, shellfish processing, and structured deposition conducted at Locus B—helped to redefine the entire complex as a locus of ritual and remembrance for Orange people, a place where the past could be accessed and employed as a social resource in the present. Writing in reference to Mississippian burial rituals, Timothy Pauketat (2010: 14) urges archaeologists to more seriously consider the lasting effects of such “emplaced remembrances” on the various people who attended them, those he refers to as the “missing persons” in Mississippian mortuaries. In the case at hand, the missing persons of interest include members of any number of far-flung communities spread across peninsular Florida. This means that at the same time these events were durably inscribed into the Silver Glen landscape, they were also incorporated into the bodily memories of geographically and socially diverse participants (sensu Connerton 1989). These memories would have provided the basis for macroscale communal relationships that would extend well beyond the complex itself and far outlast any individual gathering event. In this way, Orange period deposition not only affected material conditions in this particular location but also transcended the local by interjecting Silver Glen, as a specific kind of place, into the social memories of peoples subsequently dispersed throughout the broader region.

In short, rather than serving as a mere environmental proxy or a redundant record of broad-scale behavioral trends, Silver Glen’s cultural deposits provide an eventful, context-dependent history of Late Archaic social gatherings and interactions. The material consequences of these gatherings permanently shifted the frames of reference according to which the complex was experienced and remembered at both local and regional scales. (p.128) While some of the basic outlines of these events can be gleaned from the depositional strategies involved, many of their details are obscured by the complex’s currently denuded condition. Fortunately, additional material evidence exists that can help fill in many of the gaps. In the following two chapters, I turn to the most common class of (nonshell) material culture found within the complex’s Late Archaic deposits: fiber-tempered Orange pottery. Because of pottery’s virtually infinite malleability and penchant for preserving evidence for the various activities implicated in its manufacture and use, these vessels provide a valuable window into the actual practices conducted within Silver Glen’s various Orange contexts as well as the identities and relationships of the people who gathered within them.