Models of Long-Distance Exchange in Hunter-Gatherer Societies

Exchange is defined as the practices, obligations, and social structures created and maintained by the acts of giving and receiving (Mauss Reference Mauss1954). Exchange is a catalyst for bonding among individuals, groups, and societies and is a critical means of building broader communal identities and alliances (Sahlins Reference Sahlins1972). Hunter-gatherer exchange systems have long elicited anthropological and archaeological study, in part because nonagrarian groups often operate in a socioeconomic paradigm where property rights, trade relations, and conceptions of ownership are different from those found in societies dependent on agricultural or pastoral production (Kelly Reference Kelly2013; Sahlins Reference Sahlins1972; Testart Reference Testart1982; Woodburn Reference Woodburn1982). Research shows hunter-gatherers typically engaged in multiple networks of exchange, each differentiated by scale, intensity, and periodicity (Fitzhugh et al. Reference Fitzhugh, Colby Phillips, Gjesfeld, Whallon, Lovis and Hitchcock2011; Zvelebil Reference Zvelebil2006). Exchange between individuals living in proximity include the movement of food, utilitarian tools, and basic raw materials given with little expectation of return beyond the norm of reciprocity (Testart Reference Testart1982; Woodburn Reference Woodburn1982). Hunter-gatherers also often develop relations with neighboring groups with whom they exchange information and objects as a means of insurance when local resources drop in productivity (Bahn Reference Bahn1982; Weissner Reference Weissner, Leacock and Lee1982; Whallon Reference Whallon2006; Whallon et al. Reference Whallon, Lovis and Hitchcock2011).

Whereas local and regional social networks are important, our focus is on exchange that spanned much greater distances. When studying hunter-gatherers, it is important to delineate long-distance exchange from long-distance movement of objects. The long-distance movement of objects often occurs through direct procurement in which highly mobile hunter-gatherer groups acquire raw materials at their source and travel vast distances before depositing them (e.g., Binford Reference Binford1979). In these scenarios, hundreds of kilometers are not onerous journeys but are instead parts of everyday life. However, when groups become less mobile and the landscape populates with communities claiming more territorial control, directly acquiring materials becomes more difficult as travel is curtailed (Bettinger Reference Bettinger, Ericson and Earle1982; Kelly Reference Kelly2013). To compensate for the increased difficulty in acquiring materials, many groups establish social relations with one another through which important raw materials or finished objects could flow (e.g., Sassaman et al. Reference Sassaman, Hanson and Charles1988). This is often the point at which communities begin establishing long-distance exchange networks.

Typically, this long-distance exchange is iterative; objects and materials travel through many hands as they move across the landscape. Described as “down-the-line” exchange (Renfrew Reference Renfrew, Sabloff and Lamberg-Karlovsky1975, Reference Renfrew, Earle and Ericson1977), this system is perhaps better understood as a series of smaller networks, each connected to one another through relatively small steps. Down-the-line exchange networks are relatively common in many smaller-scale societies as they move important materials across vast distances, yet they rarely result in close ties or alliances between far-flung communities because a web of intermediaries connects these groups.

In contrast to iterative or down-the-line exchange, archaeologists rarely consider the possibility that more direct networks linked hunter-gatherer groups over long distances (however, see Zvelebil Reference Zvelebil2006). Unlike down-the-line exchange, direct exchange involves far fewer intermediaries and creates a much more immediate relationship among groups (Renfrew Reference Renfrew, Sabloff and Lamberg-Karlovsky1975, Reference Renfrew, Earle and Ericson1977). Viewed as an aspect of complex societies, long-distance, direct exchange typically entails the presence of specialized traders, emissaries, or religious pilgrims who move objects, information, and even customs and belief systems across vast distances (Wynne-Jones Reference Wynne-Jones2010). In nonindustrial groups, direct exchange rarely moves bulk or common items but instead often involves the movement of highly valued “prestige items” between elites (Hayden Reference Hayden1998). Also described as “status goods,” or “elite goods,” prestige items are often necessary for marriage or mortuary events, are used as symbols of status and identity, and can be accumulated as amassed wealth (Friedman and Rowlands Reference Friedman, Rowlands, Friedman and Rowlands1977). The distribution and movement of prestige goods are often politically important as they materialize social and economic relations between people on local, regional, and interregional scales and can be critical in forming powerful alliances (Demarrais et al. Reference Demarrais, Castillo and Earle1996). While more common goods are consumed or widely distributed, prestige goods often stay in circulation for extended periods and, when eventually deposited or destroyed, are often used as burial items or offerings during ritual events (Hayden Reference Hayden1998). Although some items attain prestige because they entail intensive labor or require complex production technologies, many are considered special because they are made from exotic materials that originate from great distances (Hayden Reference Hayden1998; Peregrine Reference Peregrine, Chase-Dunn and Hall1991); however, these qualifers are not mutually exclusive.

Whereas the long-distance exchange of prestige goods is more commonly associated with agrarian, state-level societies, its earliest emergence is occasionally found in hunter-gatherer communities, in which case it is usually seen as influencing a transition from egalitarian to trans-egalitarian societal structures (e.g., Hayden Reference Hayden1998). Typically, within trans-egalitarian groups, the long-distance acquisition of prestige goods is closely tied to ritual and ceremonial gatherings that include feasting (Hayden Reference Hayden, Dietler and Hayden2001). The connection between ritual aggregation and prestige goods has been well documented, and many argue that the political potency of prestige goods increases during large-scale gatherings because these events are venues where social roles, relations, and histories are made visible, explicit, and at times, alterable (e.g., Aldenderfer Reference Aldenderfer1993).

Archaeologists often view copper as a prestige item, especially in North America where communities began exchanging smaller items long distances during the Late Archaic. Although there are some areas, such as the mid-Atlantic, where copper exchange does not appear alongside increases in hierarchical relations, in other portions of the Archaic Eastern Woodlands the creation and movement of copper is seen as indicative of increasing power imbalances and a rising importance of interregional exchange (e.g., Pleger Reference Pleger2000).

Copper Exchange in the Archaic Woodlands

The Archaic period in the Eastern Woodlands of North America was a time of shifting ecological and environmental conditions. Beginning at the onset of the Holocene and ending with environmental conditions stabilizing near modern levels, the Archaic period began about 11,500 cal BP and continued until about 3,200 cal BP (Sassaman Reference Sassaman2010). Human population size in the Eastern Woodlands increased dramatically during the Archaic period as groups adopted increasingly specialized subsistence practices focused on locally available resources (Anderson Reference Anderson, Sassaman and Anderson1996). In tandem with rising population size, social groups increased in size and definition, and many adopted settlement strategies that required less mobility (McElrath et al. Reference McElrath, Fortier, Emerson, Emerson and McElrath2009; Sassaman Reference Sassaman2010). With decreased mobility, increased population size, and a larger “filling in” of the landscape, the acquisition of raw materials often required exchange through intermediaries (e.g., Sassaman et al. Reference Sassaman, Hanson and Charles1988). Exchange may have been further promoted as a means of finding mates and forming broader alliances that could be counted on during periods of subsistence stress and intercommunal violence (Bender Reference Bender1985). Research shows that ritual activities and large-scale ceremonial gatherings also increased in frequency and intensity through the Archaic (Claassen Reference Claassen2015), and objects and materials were likely exchanged during these events to create or strengthen alliances.

Documentation of Archaic copper exchange is facilitated by chemical sourcing that provides an accurate means of finding the origin point of objects and raw materials (Rapp et al. Reference Rapp, Allert, Vitali, Jing and Henrickon2000). Most research has focused on copper-working in the Great Lakes, a practice that began as early as about 7000 cal BP (Beukens et al. Reference Beukens, Pavlich, Hancock, Farquhar, Wilson, Julig and Ross1992). The people who worked it were known as the “Old Copper Culture,” and their earliest copper objects were heavy-socketed tools, including projectile points and axes, generally considered to be utilitarian (Binford Reference Binford1962). The Old Copper Culture spread throughout present-day Indiana, Illinois, Michigan, Wisconsin, and Ontario by the Late Archaic (after ca. 3600 cal BP), at which point archaeologists define it as the Red Ochre and Glacial Kame complexes (Sassaman Reference Sassaman2010:89–90). Red Ochre and Glacial Kame groups more commonly fashioned copper into beads, necklaces, and other objects of personal adornment, which they traded more extensively than their predecessors (Goad Reference Goad1978).

There are multiple origin points for Archaic period copper as groups in the northern Appalachians and eastern Canada are now known to have been more involved in copper production than previously assumed (Figure 1; Lattanzi Reference Lattanzi2008; Levine Reference Levine, Levine, Sassaman and Nassaney1999). Although not as well understood as their counterparts in the Great Lakes, metallurgists in these other regions also engaged in exchange networks, including groups living in the Canadian Maritimes who produced objects that eventually moved as far south as Poverty Point, a massive earthwork site in northern Louisiana (Hill et al. Reference Hill, Greenlee and Neff2016).

Figure 1. Location of selected copper production centers, Late Archaic shell rings, and other selected archaeological sites.

Archaeologists have long viewed Poverty Point as an aberration, in part because of its scale. Dating to the Late Archaic, Poverty Point consists of six concentric earthen ridges, at least five mounds, and a series of “woodhenges” measuring between 20 and 200 m across (Gibson Reference Gibson2007; Spivey et al. Reference Spivey, Kidder, Ortmann, Arco, Gilmore and O'Donoughue2015). The material culture recovered from Poverty Point also demonstrates its uniqueness. Excavations and pedestrian surveys have recovered dozens of copper objects, including awls, thin sheets, beads, pendants, nuggets, and at least one plummet (Bell Reference Bell1956; Webb Reference Webb1982). Copper is only one example of nonlocal materials found at Poverty Point: nonlocal stone used to make bifaces and other tools are measured in metric tons (Gibson Reference Gibson2007), and the frequency of nonlocal stone vessels at Poverty Point “eclipses all other assemblages known from the Eastern Woodlands” (Sassaman Reference Sassaman2010:62).

Material culture arrived at Poverty Point through a series of associated locales, likely including two shell rings located near the mouth of the Mississippi River (Figure 1). Cedarland and Claiborne are the westernmost manifestation of shell ring construction, a phenomenon spanning the Gulf and Atlantic coasts from Mississippi through South Carolina (Russo Reference Russo, Roksandic, de Souza, Eggers, Burchell and Klokler2014). Shell rings are defined as large (50–200 m across) circular or arcing midden deposits consisting of shellfish, vertebrate fauna, botanical remains, and material culture that encircle broad, shell-free plazas (Russo Reference Russo, Roksandic, de Souza, Eggers, Burchell and Klokler2014). Because of their scale, formality of deposition, and widespread construction, archaeologists have debated the function of shell rings, but most agree that they were used, at least occasionally, as a place for large-scale regional gatherings (Russo Reference Russo, Roksandic, de Souza, Eggers, Burchell and Klokler2014; Sanger Reference Sanger2017; Sanger and Ogden Reference Sanger and Ogden2018; Sassaman Reference Sassaman2010; Saunders Reference Saunders, Roksandic, de Souza, Eggers, Burchell and Klokler2014; Thompson and Andrus Reference Thompson and Fred T Andrus2011). At Claiborne, excavations recovered two copper bracelets and a pendant, a cache of soapstone vessels, and a piece of galena, all placed in a pit at the apex of the crescent-shaped ring (Bruseth Reference Bruseth and Byrd1991:18; Sassaman Reference Sassaman2010:62–63). The soapstone vessels date to about 3200 cal BP, roughly coincident with the abandonment of Poverty Point (Sassaman and Brookes Reference Sassaman and Brookes2017).

Poverty Point, Cedarland, and Claiborne mark the southernmost extent of Archaic copper exchange in the Eastern Woodlands and are unlike most sites in the Lower Southeast (defined as Florida, Georgia, South Carolina, Alabama, Louisiana, and Mississippi) where Archaic period copper is virtually absent (Figure 1; Goldstein and Meyers Reference Goldstein, Meyers, Kuijt, Quinn and Cooney2014; Sassaman Reference Sassaman2010). The southernmost extent of notable quantities of Archaic period copper is the lower Ohio River Valley where communities labeled as the Shell Mound Archaic may have acted as intermediaries between northern metal-producing communities and groups further to the south as they amassed copper from the Great Lakes and marine shellfish, likely from the lower southeast Atlantic coastline (Goad Reference Goad1978; Sassaman Reference Sassaman2010:46).

While Shell Mound Archaic communities likely played an important role within the broader exchange networks spanning the Eastern Woodlands, we know little about how Archaic period copper objects moved to even more distant nodes in the lower Southeast, including Poverty Point and its associated shell rings. Considering that they are long lasting and likely highly valued, an implicit assumption is that copper moved through many intermediaries, perhaps over many generations, eventually ending up in important locales such as Poverty Point, at which point they were purposefully deposited (Milner Reference Milner2004). Others argue that the relational networks spanning the Archaic, of which copper is just one, were based on much more immediate relations between far-flung communities (Sassaman Reference Sassaman2010). At Poverty Point, some researchers suggest people gathered during massive “trade fairs” where objects were exchanged (Jackson Reference Jackson and Gregg1991), whereas others argue the site was a religious aggregation point that drew people from across the Eastern Woodlands, perhaps through pilgrimages (Spivey et al. Reference Spivey, Kidder, Ortmann, Arco, Gilmore and O'Donoughue2015).

As already noted, copper was one of many materials exchanged across the Archaic Eastern Woodlands as other objects, including oversized blades, zoomorphic beads, and carved bone pins, traveled along unique yet intertwined social networks (see review in Sassaman Reference Sassaman2010:138–142). Communities deployed many of these exchanged goods during ritual gatherings and often interred them with their dead, evidence of the increasing visibility, and perhaps importance, of mortuary programs during the Archaic.

Archaic Mortuary Program

Mortuary programs became increasingly localized and circumscribed during the Archaic as communities established regional traditions across the Eastern Woodlands (Buikstra and Charles Reference Buikstra, Charles, Ashmore and Knapp1999; Charles and Buikstra Reference Charles, Buikstra, Phillips and Brown1983; Milner et al. Reference Milner, Buikstra, Wiant, Emerson and McElrath2009). Cremation was a widespread and relatively common practice in the northern portions of the Eastern Woodlands during the Archaic period, including Ontario and the upper Midwest (Converse Reference Converse1980; Donaldson and Wortner Reference Donaldson and Wortner1995), Newfoundland and Labrador (Fitzhugh Reference Fitzhugh1978; McGhee and Tuck Reference McGhee and Tuck1975; Robinson Reference Robinson1996), and in the northeastern United States (Regensburg Reference Regensburg and Stinton1983; Stanzeski Reference Stanzeski1996) and southern New England (Bolian Reference Bolian, Starbuck and Bolian1980; Dincauze Reference Dincauze1968, Reference Dincauze1975; Robinson Reference Robinson1996). Often carried out alongside other mortuary practices, including flexed and extended inhumations (Pleger Reference Pleger2000), cremation has a particularly deep history in the Great Lakes, where communities have incinerated their dead since the Paleoindian period (Mason and Irwin Reference Mason and Irwin1960).

The distribution of copper use and cremation overlaps to a large extent, as Archaic period cremations are found in low numbers as far south as the southern Ohio Valley (Webb and DeJarnette Reference Webb and DeJarnette1942) and the Tennessee River Valley (Chapman and Myster Reference Chapman, Myster and McNutt1991) with virtually none recorded in the lower Southeast (Sassaman Reference Sassaman2010:66–77). Instead, there we find a rich assortment of localized practices including pond burials, flexed interments in shell mounds, extended inhumations in sand mounds, bundle-burials, and crouched-position interments within pits (see review in Sassaman Reference Sassaman2010:66–77). The one region in the lower Southeast where mortuary data is notably absent is from the Georgia Bight, a portion of the Atlantic coast between Cape Fear, North Carolina, and Cape Canaveral, Florida, where only a few possible Archaic-age bodies have been recovered (Figure 1; Michie Reference Michie2000).

Skeletal remains recovered at Daw's Island, South Carolina, are the best understood and most firmly dated Archaic period human remains from the Georgia Bight (Figure 1). Excavations recovered three nearly complete burials while surface collections resulted in the recovery of at least five additional individuals (Michie Reference Michie2000; Rathbun et al. Reference Rathbun, Sexton and Michie1980). The intact remains were found eroding from a midden, characterized as a thin lens of dark soil, oyster shell, mammal bones, fiber-tempered pottery, antler and stone projectile points, engraved bone pins, and steatite objects. The midden dates to about 3900–3700 cal BP (Michie Reference Michie2000:43), and the remains were all in flexed positions and found either on their side or back. Other than those from Daw's Island, there are very few reported Archaic period remains from the Georgia Bight (Russo Reference Russo2006:46–47) and only a few from other shell rings.

Because of the size and layout of the shell rings, early archaeologists and antiquarians assumed they were burial sites, but excavations failed to find more than a few scattered pieces of human bones (Moore Reference Moore1897). More recent research has likewise found only a few human remains (Marrinan Reference Marrinan1975; Trinkley Reference Trinkley1980), often intermixed with the shell and vertebrate fauna that make up the arcing middens. The low occurrence of Archaic period human remains at shell rings and other coastal sites has led many archaeologists to assume coastal residents of the Georgia Bight handled their dead in a manner that makes them difficult, if not impossible, to find archaeologically, including burial at sea, open-air burials in nearby forests and marshes, or cremation (Elliott and Sassaman Reference Elliott and Sassaman1995; Marrinan Reference Marrinan1975; Trinkley Reference Trinkley1980). One of the few human cremations reported from the lower Southeast that may date to the Late Archaic was found at White's Mound in Richmond County, Georgia (Figure 1), where Phelps and Burgess (Reference Phelps and Burgess1964:200) uncovered a large pit containing fragmented and burned human bones. Although the bones were exposed to heat, there was no sign of burning within the pit (Phelps and Burgess Reference Phelps and Burgess1964:200). Based on the high levels of heat and fragmentation, Phelps and Burgess (Reference Phelps and Burgess1964:200–202) suggest the remains are evidence for either cannibalism or crematory practices (also see Trinkley Reference Trinkley1980:46). Unfortunately, the remains and their associated materials have never been directly dated. This is problematic, as there are Woodland period ceramics intermixed with the skeletal remains, which may suggest the pit and burned human remains are not Late Archaic (Phelps and Burgess Reference Phelps and Burgess1964:200).

Although it is possible that coastal peoples living in the Georgia Bight disposed of their dead in ways that make remains difficult to detect, it is also likely that archaeologists have not been looking in the right locales. Russo (Reference Russo2006:50) highlights that to this point, the vast majority of research conducted on coastal Late Archaic sites, particularly shell rings, has focused on shell middens and that despite their clear importance, the interior portions of shell rings, particularly the direct centers, have only occasionally been tested (Sanger and Ogden Reference Sanger and Ogden2018). This is surprising, given that later Woodland period circular shell middens in the lower Southeast occasionally contain burials in their centers (Bense Reference Bense1994). Although few ring centers have been investigated, when they are tested, they often contain features that archaeologists find difficult to explain. For example, Calmes (Reference Calmes1967:9–11) excavated the centers of both Skull Creek shell rings and found large circular pits, including one in the larger ring that contained a “concretion-like” material. Similar large pits have been found in the centers of Stratton Place (Trinkley Reference Trinkley1980:256), A. Brush Krick (DePratter Reference DePratter1979), and perhaps Lighthouse Point (Trinkley Reference Trinkley1980). Invariably, archaeologists describe these features as simple processing or roasting pits, paying little attention to the objects found within them, which occasionally include calcined bone (Calmes Reference Calmes1967; Trinkley Reference Trinkley1980). It is possible that some of these previously excavated pits may include remains that, as our findings document below, are difficult to recognize as human because they have been calcined and perhaps pulverized into small fragments.

Study Area

Excavations in the center of the McQueen shell ring uncovered a burial pit filled with more than 80,000 calcined bone and teeth fragments, a piece of worked copper, and other associated material culture (Figure 2; Sanger and Ogden Reference Sanger and Ogden2018; Sanger et al. Reference Sanger, Hill, Lattanzi, Padgett, Larsen, Culleton, Kennett, Dussubieux, Napolitano, Lacombe and Thomas2018). In our prior publications (Hill et al. Reference Hill, Lattanzi, Sanger and Dussubieux2019; Sanger et al. Reference Sanger, Hill, Lattanzi, Padgett, Larsen, Culleton, Kennett, Dussubieux, Napolitano, Lacombe and Thomas2018), we detailed analyses demonstrating that the copper band is made of materials that originated in the Great Lakes region. As there is no evidence of local copper working in the lower Southeast during the Archaic, we assumed the copper band was also manufactured in the Great Lakes region.

Figure 2. (A) Location of shell rings on St. Catherines Island, (B) Location of burial pit within the McQueen shell ring (shell arc outline based on thickness of shell deposits), (C) Plan view of burial pit marked by dark lines signifying the upper and lower extent and dotted lines showing the assumed western extent of a pit that has not been mapped, human and nonhuman bones are marked by circles, and topographic lines indicate small sand hill. Elevation of the pit extent and hill are based on arbitrary datum elevation at 5 m.

The McQueen shell ring is one of two contemporaneous shell rings on St. Catherines Island, located near the center of the Georgia Bight (Sanger and Thomas Reference Sanger, Thomas, Thomas and Sanger2010; Figure 2). Dating to about 4300–3800 cal BP (Supplemental Table 1), the McQueen and St. Catherines shell rings are among the oldest known sites on the island and were deposited shortly after sea levels stabilized near modern levels (Bishop et al. Reference Bishop, Meyer, Vance, Rich, Bishop, Rollins and Thomas2011; Sanger and Thomas Reference Sanger, Thomas, Thomas and Sanger2010). Both rings are similar in size, with middens measuring roughly 70 m across and encircling shell-free plazas spanning 30–40 m (Figure 2). Measuring only 0.2–1 m in height, the circular midden at McQueen is not as tall as the midden at the St. Catherines shell ring that measures 1–1.5 m. Both ring sites appear to have been occupied year-round, although population levels may have fluctuated throughout the year with numbers increasing during seasonal gatherings (Colaninno Reference Colaninno2012; Quitmyer et al. Reference Quitmyer, Jones, Reitz, Quitmyer and Thomas2012; Sanger et al. Reference Sanger, Quitmyer, Colaninno, Cannarozzi and Ruhl2019). Evidence suggests that shell-ring site residents engaged in relations with their inland neighbors, likely facilitated by periodic gatherings, during which materials, including stone for knapping, were exchanged (Sanger and Ogden Reference Sanger and Ogden2018; also see Gilmore et al. Reference Gilmore, Sassaman and Glascock2018).

Geophysical surveys provided the impetus for excavating in the McQueen shell ring plaza, as they revealed a unique magnetic signature in its center (Mahar Reference Mahar, Thompson and Thomas2013). Located roughly 30 cm below the surface, only the very upper edge of the burial pit was impacted by hand plowing during the eighteenth and nineteenth centuries. The pit measured almost 1.5 m across and 1 m deep and was filled with a dark sandy loam, distinguishing it from the surrounding yellow-brown sand (Figure 2). Conical in shape, the pit and surrounding soils showed no signs of thermal alteration, which typically manifests as reddish, highly consolidated sands. Excavations also revealed the presence of a buried sand hill, measuring 30 cm tall, next to the pit (Figure 2).

Pit contents included highly fragmented bones that were typically calcined and very small. We were unaware that some of these remains might be human until the assemblage was hand sorted in the laboratory and their identification was confirmed by project bioarchaeologists. Acting in accordance with requirements for all Native American remains from St. Catherines Island, David Hurst Thomas conveyed these findings to Georgia's Council on American Indian Concerns. The resulting agreement facilitated continued excavations and the analyses presented here without sharing any photographs of the human remains.

Excavators encountered several very small (1–5 cm) fragments of copper about halfway down the pit alongside a larger, nearly intact copper band. Although we do not share any photos of human remains, we include an image of the copper object recovered alongside the cremains (Figure 3). The copper band measures between 38 mm and 19 mm wide, 160 mm long, and 1 mm thick and has no engravings, impressions, incisions, or other embellishments on either face. It has not been exposed to heat and therefore was not placed in the fires alongside the cremated bones.

Figure 3. Photo of copper band from the McQueen shell ring. Photo by Matthew Sanger.

We ran direct accelerator mass spectrometry (AMS) radiocarbon dates on a human cranial vault fragment found 2 cm above the copper band and a femur fragment immediately adjacent to it. Both dates are statistically identical and show the cremated individuals died about 4100–3980 cal BP, contemporaneous with the formation of the shell arc (Supplemental Table 1, lab numbers OxA-32446 and UCIAMS-130901; Sanger et al. Reference Sanger, Hill, Lattanzi, Padgett, Larsen, Culleton, Kennett, Dussubieux, Napolitano, Lacombe and Thomas2018:E7674). The material culture recovered from the pit supports the radiocarbon findings, as they include fiber-tempered pottery and stemmed projectile points, which are diagnostic of the Late Archaic (e.g., Sassaman Reference Sassaman1993). At least one projectile point was fractured by high levels of heat, presumably as it was exposed to the same flames as the surrounding bones (Sanger and Ogden Reference Sanger and Ogden2018).

Discussion: Movements of Ideas and Objects

The use of cremation at McQueen is unlike what is typically found in the Late Archaic lower Southeast and may even be a unique discovery, as no similar finds have been published. However, an absence of data does not prove that other cremations are not present at other Late Archaic sites in the lower Southeast. There are several ways to interpret these findings: (1) cremation is more common in the lower Southeast than previously assumed, and the discovery of cremains at McQueen is part of this broader yet underappreciated pattern; (2) residents of McQueen independently invented crematory practices; or (3) the practice of cremation originated elsewhere and was brought to McQueen. We find little evidence for the first two hypotheses. Cremation was a relatively foreign practice in the lower Southeast, and we propose that it came to McQueen alongside the copper band through direct trade lines between the Georgia Bight and further to the north, perhaps extending as far as the Great Lakes, and likely was facilitated by communities living in the Ohio and Tennessee River Valleys.

We find our first hypothesis (that cremation is a more widespread tradition) unconvincing, as there are very few instances of cremation in the region, and the few that have been documented, such as White's Mound, are problematic. Instead, the few Late Archaic period burials that have been recovered, such as those found at Daw's Island, suggest that inhumations are a means of handling the dead in the coastal lower Southeast. As already discussed, the paucity of Late Archaic period human remains from the coastal lower Southeast is notable and could be evidence of crematory practices that resulted in highly fragmented cremains that either are difficult to recover or are misidentified when recovered. In this regard, it is possible that additional cremations were performed at other shell rings but have been ignored or inaccurately identified as food remains located in roasting pits (e.g., Trinkley Reference Trinkley1980). It is difficult to determine whether cremation was a more widely held practice when this hypothesis is based on negative evidence (i.e., the lack of human remains). Further excavations and reassessment of prior finds, especially from other shell rings, are needed to determine whether additional Late Archaic period cremations are located in the lower Southeast.

Without this additional evidence, the burial pit at McQueen appears unique, perhaps reflecting an independent invention by the residents of the Georgia Bight. Containing at least seven individuals, the McQueen cremation does not hold all the individuals that would have perished over the hundred or more years in which the ring was in use. Our geophysical surveys and excavations did not encounter any similar burial pits, and we are relatively confident that the one in the center of the ring is the only one on site. As such, the dead interred in the center of McQueen may have been handled differently than other deceased individuals, perhaps because of who they were or how they died. As they were burned in the flesh and the burial pit does not appear to be excavated multiple times, the individuals buried at McQueen almost certainly represent a group of people who died at, or approximately at, the same time. The deaths of at least seven people in such a short period could reflect the presence of a relatively large population who came together to bury their recently departed. Alternatively, the co-occurrence of these deaths could reflect an anomalous event in which a group of people perished, perhaps through disease, violence, or accident. Perhaps the inauspiciousness of their deaths required handling these bodies differently, creating an impetus to invent a new tradition with few, if any, parallels in the region.

It is difficult to reject the possibility of independent invention, but there are notable parallels in how people at McQueen and the Great Lakes handled their dead that suggests a shared tradition. Making the link between the Great Lakes and the coastal lower Southeast is not altogether surprising, considering that Late Archaic Great Lakes burials occasionally contain shell artifacts thought to originate from the coast (Abel et al. Reference Abel, Stothers, Koralewski, Prufer, Pedde and Meindl2001; Del Castillo Reference Del Castillo1986; Donaldson and Wortner Reference Donaldson and Wortner1995; Stothers and Abel Reference Stothers and Abel1993). There is also some evidence to suggest that the mortuary practices of groups in the Great Lakes and Upper Midwest extended further to the south, including among Shell Mound Archaic communities in the Ohio River Valley and perhaps as far south as the Tennessee River Valley (Chapman Reference Chapman1977; Chapman and Myster Reference Chapman, Myster and McNutt1991; Goldstein and Myers Reference Goldstein, Meyers, Kuijt, Quinn and Cooney2014; Trevelyan Reference Trevelyan2004).

Within all these regions (the Great Lakes, Ohio River Valley, and Little Tennessee River Valley), we find mortuary practices that are also found at McQueen, including the cremation of the dead in places other than where they were buried and the interment of copper alongside the dead (Abel et al. Reference Abel, Stothers, Koralewski, Prufer, Pedde and Meindl2001; Donaldson and Wortner Reference Donaldson and Wortner1995; Goldstein and Myers Reference Goldstein, Meyers, Kuijt, Quinn and Cooney2014; Pfeiffer Reference Pfeiffer1977; Stothers and Abel Reference Stothers and Abel1993). In the Ohio and Tennessee River Valleys, the co-occurrence of copper and cremation is relatively uncommon during the Late Archaic, but when they do occur, interments are often placed on a nearby hilltop, a practice also found in the Great Lakes (Donaldson and Wortner Reference Donaldson and Wortner1995; Goldstein and Meyers Reference Goldstein, Meyers, Kuijt, Quinn and Cooney2014). St. Catherines Island is a relatively flat landform, yet residents of McQueen placed their burial pit next to a very small hill located in the center of the ring (Figure 2). Such hills are rare, and our excavations did not encounter any similar features anywhere else at the site. As such, this small topographic similarity could be evidence of another connection between McQueen and communities closer to the Great Lakes. However, we understand that a 30 cm tall pile of sand is of a much smaller scale than the burial contexts often used in the Great Lakes. The point of emphasis here is not necessarily any topographical similarity, but rather the ritual significance of a visually distinct feature of the land where the cremated remains were interred.

We find additional similarities in how communities in the Great Lakes and McQueen handled the dead that are not as well represented or well reported in the regions found in between, including how both groups cremated bodies while still in the flesh (Donaldson and Wortner Reference Donaldson and Wortner1995; Pfeiffer Reference Pfeiffer1977). Great Lakes groups also cremated both men and women as well as individuals of all age groups (Converse Reference Converse1980; Donaldson and Wortner Reference Donaldson and Wortner1995; Ritchie Reference Ritchie1949; Ritzenthaler and Quimby Reference Ritzenthaler and Quimby1962), similar to the demographic profile found at McQueen. The stratigraphy of the McQueen burial is also similar to those at the Great Lakes where Late Archaic burials are occasionally layered (Donaldson and Wortner Reference Donaldson and Wortner1995). Cremations conducted by Late Archaic Great Lakes communities occasionally include faunal remains (Abel et al. Reference Abel, Stothers, Koralewski, Prufer, Pedde and Meindl2001; Donaldson and Wortner Reference Donaldson and Wortner1995). While we cannot definitively state that the red stains found on the bones from McQueen were the result of people placing ocher on the cremains, this seems a plausible conclusion. The use of red ocher is a hallmark of burials in the Great Lakes (Abel et al. Reference Abel, Stothers, Koralewski, Prufer, Pedde and Meindl2001; Donaldson and Wortner Reference Donaldson and Wortner1995; Ritzenthaler and Quimby Reference Ritzenthaler and Quimby1962).

Finally, copper is typically associated with deep-water creatures and spirits known as manitous in the Great Lakes, where copper is often described as originating when a warrior dove to great depths to remove a piece of an underwater beast (Trevelyan Reference Trevelyan2004). The pygmy sperm whale vertebra found in the McQueen cremation is a unique archaeological find not replicated elsewhere on the island and reflects a purposeful deposit of a deep-water creature alongside the copper band and cremains. Whether this vertebra is meant to echo stories of the manitous is impossible to determine, but we have few other theories about why this bone was placed in the burial pit.