Interpreting the South Flats Earthwork (20MU2): insights gained from geophysical surveys.
Michigan's prehistoric earthen enclosures are among the least understood archaeological sites in the state. This paper explores the function of the South Flats Earthwork (20MU2) using remote sensing strategies to reveal aspects of its internal structure not available through excavation. Ground penetrating radar and magnetometry were employed to explore this fragile archaeological site. Results, coupled with excavation data, suggest that South Flats was the work of a small-scale society and a locus of storage and food exchange through kin ties.
Presented here are the results of surface geophysical surveys at the South Flats enclosure (20MU2) in Muskegon County, Michigan. South Flats has been excavated twice, once by George Quimby in 1937 and later by Crand Valley State University (GVSU) in 2006 (Quimby 1965; Gaff and Brashler 2011). Even though previously studied and reported on, this earthwork was subjected to geophysical surveying in 2010 to collect more information related to site structure than was available from excavation alone. We intentionally chose this option after conducting limited excavations and restoration of the enclosure in 2006 because the site is small and fragile; we wanted to learn more without further impacting the site. The results of the geophysical surveys are presented here and we offer interpretations based on the results of the survey coupled with original work done regarding the site that was not included in the original report. We first briefly situate our work in landscape archaeology and mobility studies. Next, we review and summarize previous interpretations of Michigan's earthworks and place them in the context of other earthen enclosures in the eastern United States. We then discuss the geophysical surveys at South Flats and the results of that work. The geophysical data coupled with the excavation data collected at South Flats is an example of how archaeologists can gain site-wide information about the structure of the site not evident through excavation alone, which in turn can enhance interpretations about site placement and function. Finally, we discuss South Flats in the context of current discussions about landscape, mobility and social interaction in the Late Prehistoric of Michigan.
LANDSCAPE, MOBILITY AND SOCIAL INTERACTION
Understanding the landscape placement and internal arrangement of a structure such as South Flats, and its relationship to group dynamics and mobility is critical to understanding the site's function and meaning. Approaches to understanding human interactions with landscape began with work in the early twentieth century by geographer Carl Sauer (1925) among others and subsequently a variety of ethnographers such as Basso (1996) Thornton (2008), and numerous archaeologists writing mostly beginning in the late twentieth century and continuing into the present (Ashmore and Knapp 1999). Several summaries of the numerous approaches taken by landscape archaeology more thoroughly discuss its history and theory and are nor summarized here, except to acknowledge that landscape approaches in archaeology today are diverse (Knapp and Ashmore 1999; Ashmore and Blackmore 2008). How archaeologists employ landscape archaeology is influenced to some extent by the intellectual traditions which frame questions and interpretations based on theoretical positions of the investigator (Ashmore and Blackmore 2008). We acknowledge our predominantly processualist approach and see understanding sites such as South Flats through the lens of ecological and environmental archaeology and landscape, rather than by searching for meanings and phenomenology of landscape, though we see both approaches as worthy of discussion.
It is from this perspective that we engage our understanding of the modest monument that is South Flats. We seek to understand it in terms of the group of peoples who constructed it and gave its location meaning through its placement in the physical environment, the adaptive strategies evident at the site and its regional context, and how it might inform us with regard to social, political and economic dimensions of Late Prehistoric society in west central Lower Michigan. Other approaches with broader geographic scope and different theoretical underpinnings conceptualize and interpret the built environment of Late Prehistoric Michigan in somewhat different ways, which we only briefly touch on below. An example of these kinds of landscape approaches in Michigan archaeology is the pioneering work of Howey (2006), Howey and O'Shea (2006) and Howey (2012) with their interpretation of earthen enclosures, which we briefly describe in the section that follows.
In addition to the importance of landscape, we also seek to understand South Flats from an understanding of discussions of relative mobility among hunter-gatherer-fishing and horticultural societies. Early studies of mobility (Lee and Devore 1968) and others obscured the diversity of mobility strategies employed by hunter-gatherer groups and relied on a typological approach to understand the continuum from mobile to sedentary among groups (Kelly 1992). Binford (1980) developed the concept of mobility by differentiating between residential mobility, when an entire local group moves together from one location to another, and logistical mobility, where individuals from within a larger group travel to specific locations for specific resources or tasks/activities and then rejoin the group. From these two strategies, Binford sets out two ideal hunter-gatherer settlement strategies to be employed as conceptual tools, not ideal types, which include collectors who move as a residential group from one location to another, employing logistical forays to bring resources back to the residential camp, and foragers, who make frequent moves in the form of short logistical forays. Binford (1982, 1983) supplemented this model with the concept of what Kelly (1992, 45) suggests might he termed territorial or long term mobility where groups "move among a set of territories." This overall model has been developed significantly by others examining a range of hunter-gatherer societies (Kelly 1992; 1995). A critical element of mobility based models is the concept of risk and how mobile and not-so-mobile groups manage risk through their movements across the landscape in subsistence activities as well as socio-political arrangements and alliances that can buffer subsistence or other risk behaviors (Spielman 1986; O'Shea and Halstead 1989). These can include cooperative interactions regarding food resources which can be further cemented by kinship and other social and exchange agreements (Spielman 1986). Binford, Kelly, Spielman and others' discussions of mobility and risk buffering are the basis of significant work by several archaeologists working in Michigan. For the current study, the most important is the work of Holman and Lovis (2008).
Writing primarily about northern Michigan, Holman and Lovis note that "All of the ethnographically documented native people resident in Michigan during the early years of direct and consistent European contact, around A. D. 1650 were mobile to some degree, even after the late adoption of horticulture" (Holman and Lovis 2008, 280). They further propose that constraints on mobility varied from region to region with hunter - gatherer groups occupying territories that were "essentially filled to capacity," and some of these groups had contact with horticultural groups, primarily to the east from Ontario and to the south in Indiana and Ohio. They suggest that the groups living in Michigan during Late Woodland/Late Prehistoric period immediately preceding European contact also relied on seasonal mobility and used strategies employed by historically known groups (Holman and Lovis 2008). After examining archaeological, ethnographic and ethnohistoric data, they describe "Chippewa" and "Ottawa" patterns more fully developed than original constructions of the 1970s (Fitting and Cleland 1969) detailing in particular risk buffering and mobility strategies employed by these groups both in the early historic period and previous late prehistoric societies.
Holman and Lovis (2008) suggest the Chippewa hunter-gatherer-fisher pattern relies on residential mobility, moving entire communities seasonally up and down the coast and into the interior of the Canadian Biotic Province and the transition zone. They suggest this pattern is similar to that followed by Mackinac Phase peoples (ca. AD 1000-1200) in northern Michigan. They link the Ottawa Pattern to the Juntunen Phase (ca. 1200-1400) in the Upper Great Lakes during which maize becomes significant with a settlement system based "on growing crops and maintaining an active Great Lakes fishery" (Holman and Lovis 2008, 299). Groups with this pattern participated in a semi-sedentary way of life with mobility occurring at points in the cycle of the group when harvests and fish were in short supply. Groups then engaged in logistic mobility, through contacts with other groups for exchange and hunting in shared territories that were acknowledged as such through kin and other ties. In both the Ottawa and Chippewa patterns, mobility-either residential or logistical these options are critical to survival and are reinforced by both exchange and kin ties between groups which serve to make the movements between environments and territories fluid and sustainable (Holman and Lovis 2008; McClurken 1988).
The Pottawatomi pattern, originally conceived of as large permanent horticultural peoples who practiced some logistical mobility through hunting (Fitting and Cleland 1969) has been found difficult to support based on recent work at the Late Prehistoric Moccasin Bluff Site in southwest Michigan (O'Gorman 2003, 2007; O'Gorman and Lovis 2006). Maize does occur in southwestern Michigan, but appears to be found in the context of intrusive Mississippian communities such as Wymer West Knoll (O'Gorman 2007), which appear to be near the northern border of Mississippian expressions in Michigan. Evidence from excavated Late Woodland sites in west central Michigan, where South Flats is situated, provides abundant evidence of cache pits and storage in the Late Prehistoric, but to date, little evidence of horticulture north of the St. Joseph drainage in the far southwestern portion of the state (Brashler, et al. 2000). Despite systematic flotation at several sites in the Grand and Muskegon river drainages dating to the Late Woodland, no evidence of horticulture has been identified to date.
Holman and Lovis (2008, 300) conclude that there is considerable similarity between the contact period and late prehistory in the flexibility expressed by the residential and logistical mobility strategies they document. Their argument is grounded in extensive research related to site location, artifacts and subsistence remains for the northern Great Lakes area and we suggest that it has applicability to Late Prehistory in much of the western Lower Peninsula as well. We now briefly consider Late Prehistoric earthen enclosures in Michigan and their wider context.
MICHIGAN'S LATE PREHISTORIC EARTHWORKS
Michigan's Late Prehistoric earthworks have long been a source of discussion in the archaeological community. Estimates from early work in Michigan by Greenman (1927) and Hinsdale (1931) suggest that perhaps as many as 80 enclosures once were found in the lower peninsula of the state. Relatively few of these have been systematically excavated and estimates suggest fewer than ten remain relatively intact (Zurel 1999), many having been destroyed by agricultural activity and other twentieth century construction. The enclosures are highly variable in size and shape including circular and horseshoe shaped forms and ranging in diameter from about 12 m to over 100 m. Dates for Michigan's enclosures fall primarily between AD 1350 and AD 1650 in the period of the later Late Woodland/Late Prehistoric (Zurel 1999).
Michigan's Late Prehistoric earthen enclosures are distributed close to the northern and western limits of these and similar kinds of constructions, which occur in Indiana, Ohio, Pennsylvania, New York and southern Ontario. Many of these have been interpreted as fortified villages, mortuary locales and/or ceremonial centers (Mainfort and Sullivan 1998; Neusius, et al. 1998). In Indiana, for example, Oliver Phase Late Prehistoric enclosures such as Strawtown and Scranage are associated with horticultural villages defined by palisaded earthen enclosures (White et al. 2003). In Ohio enclosures such as the Ripley site are associated with village and mortuary activities (Neusius et al. 1998) and other sites associated with Fort Ancient villages. To the east in Pennsylvania and West Virginia are sites associated with Monongahela and Fort Ancient, in New York and southern Ontario are Proto-historic Iroquoisan sites. Further south in Illinois and Kentucky, stone walls were employed which were thought to be defensive structures with similar intentions as some kinds of earthen enclosures (Milner 2004). Few circular enclosures have been identified in Wisconsin, though Goldstein (1995) suggests that effigy mounds, dating somewhat before the Late Prehistoric enclosures in Michigan, may have functioned in similar ways (symbolic, prescribing sacred space) as some enclosures.
In Michigan, enclosures have been characterized as fortified villages, ceremonial or council circles, animal traps, and enclosed garden areas (Zurel 1999). Excavations at the Mikado (Carruthers 1969), Walters-Linseman (Cornelius and Moll 1961), Whorley (Speth 1966) earthworks indicated the diversity in shape, contents and perhaps functions of several enclosures. Milner and O'Shea (1998) suggest that northern Michigan enclosures were places for trade and interactions at ecological and social boundaries. Zurel (1999), in a summary of Michigan's earthworks and their interpretations, notes that they usually have minimal amounts of material culture, and that whatever artifacts and features exist are concentrated within the earthwork and not outside them. He further notes two groups of earthworks. In eastern Michigan, earthwork sites are at the boundary between the Canadian and Carolinian biotic provinces. In southwestern Michigan, earthworks are found in places where the Prairie Peninsula enters the Carolinian biotic province in that part of the state. Writing about west-central Michigan, Brashler et al. (2000) note that earthworks in this area are the most poorly understood part of the Late Woodland.
Most recently, based on previous research at the Missaukee Earthworks in the 1920s, 1960s and her own work, Howey (2012; 2006) emphasized the landscape, symbolic, and monumental qualities of earthen enclosures in Michigan and described a system of inter- and intra-tribal relationships revolving around the placement of enclosures. Howey concludes that Missaukee Earthworks was "a structured ritual precinct for periodic aggregation of coastal fisher-horticulturalists and inland foragers during Late Prehistory (ca. AD 1200-1420 ...) (2012,118). Howey further suggests both integrative and symbolic functions for Missaukee drawing on the Algonkian creation story of Bears Journey (Howey 2006, 2012; Howey and O'Shea 2006). In her model, Howey suggests that South Flats was an example of an enclosure constructed by or related to coastal tribal fisher-horticultural societies who were integrated with the inland tribal entities at Missaukee and nearby occupation sites.
Key in her argument is the understanding that coastal groups were the locus of societies that had become horticulturalists, or alternatively intensive fisher people and were providing maize and fish to groups occupying the interior. To more fully understand the internal structure of South Flats, we chose to amplify our test excavations with two non-invasive geophysics techniques, ground penetrating radar and magnetometry. These two techniques complement each other and have been used extensively in understanding the internal structure of Middle Woodland enclosures and communities in Ohio, but have not been employed in efforts to understand Michigan's earthen enclosures to our knowledge.
Interpretations of the internal qualities of earthworks are strongly limited by the fact that archaeological testing and excavation is based on a sampling strategy; that is, entire sites are almost never excavated. In addition to a reliance on sampling, oftentimes personnel and rime constraints limit the amount of archaeological work that can be done at a site. Since sites are rarely excavated in their entirety, maximizing the knowledge of a site's internal structure becomes paramount to arguments of complexity. We intentionally chose to follow up our initial investigations using the two noninvasive technologies to enhance our knowledge of the site's internal structure without further disturbing the site.
Recent advances in the deployment of surface geophysics technologies like Ground Penetrating Radar (GPR) and magnetometry now make it possible to get a glimpse of a whole site within the time, labor, and financial constraints of most typical archaeological projects. Kvamme (2003) argues that techniques like these are useful in generating data in some contexts for understanding intra and inter site structures within specific sites. He further argues that GPR is a technique that should not be limited to use in open landscapes (plowed fields and arid lands). These techniques are "ideally suited for detecting the very kinds of cultural features that traditional landscape archaeology approaches cannot: buried architecture, dwellings and other constructions that give structure and meaning to human occupations" (Kvamme 2003, 454). The geophysical data presented below, following a brief summary of the South Flats enclosure based on the earlier excavations, allows us better insight into the construction and use of this enclosure in a way not possible with excavation alone. Finally, the results of both the geophysical work and the excavation results, which we summarize below, allow us to make a modest contribution based on South Flats in context of other sites including enclosures and the landscape of Late Prehistoric Lower Michigan.
THE SOUTH FLATS ENCLOSURE
The site known as the South Flats enclosure is in the Muskegon State Game Area in Muskegon County, Michigan. Michigan Department of Natural Resources. South Flats is located within the Muskegon River drainage, an area that was logged intensively during the 1800s (Alexander 2006). The site does not appear to have been plowed, and it was probably protected, in part, when a private hunting club bought the property in 1929. The South Flats Gun Club held the land until the Michigan DNR took it over in the 1970s. Other than minimal construction of hunting blinds and stations, there was no development on the property. After taking over management, the Department of Natural Resources opened the area for recreational purposes, primarily hunting. Doing so has led to minor damage to the earthwork, mostly from off-road-vehicle traffic. The DNR has erected barriers to prevent further impact, a mostly successful measure.
The site is situated in the transition between the Carolinian and Canadian biotic provinces as shown in Figure 1. South Flats overlooks a broad floodplain from a high bluff that is approximately 25 m above the Muskegon River floodplain. It is situated on a peninsula that is not particularly distinguished by its location on the landscape given that there are other similar high bluffs to the east and west. The peninsula, like much of the upland area in west Michigan lacks relief and is defined by a small intermittent drainage immediately to the south and west of the site and a more significant drainage, Mosquito Creek approximately one km to the northeast. Several small streams and creeks run through the floodplain to merge with the Muskegon River below. The floodplain is frequently wet and swampy, flooding intermittently and drying out to some extent during late summer and fall. Vegetation today at South Flats is dominated by pine and oak forest. This ecological setting is consistent with descriptions of white pine-white oak vegetation drawn from early General Land Office (GLO) surveys carried out in the 1800s (Michigan's Natural Features Inventory 2013). This vegetation type dominates the lower Muskegon drainage along with the abundant cedar swamp and riparian zones associated with the river, Lake Muskegon and Lake Michigan. South Flats' location would have provided relatively easy access to a variety of riparian and upland habitats as well as easy transportation to Lake Michigan, which is roughly 20 km southwest.
The well-drained sandy soils at the site are part of the dry outwash plains in the area that are not conducive to agriculture, and fall within the Rubicon Soil Series (USDA SCS 1968, 29). At the surface, the soil tends to be 7-8 cm thick (10 YR 2/1) with another 5 cm eluviated zone (10 YR 5/1). Subsoils at 70 cm below surface range between 7.5 YR 4/4 and 10 YR 6/6 on the Munsell color scale. These soil descriptions are consistent with what was observed in the field during the 2006 archaeological work (Gaff and Brashler 2011). The minimal historic disturbance, and well-drained sandy soils make the South Flats locale an ideal location for geophysical surveying. The vegetation cover, pine and oak between 80 and 100 years in age, posed some challenges to the survey work and indeed, would have potentially discouraged some from attempting remote sensing at the site.
George Quimby was the first archaeologist to work in the area. In 1937, he was employed to carry out archaeological work at two sites, a village or campsite located nearby and at South Flats (Quimby 1993). During his work at South Flats, he excavated six units concentrated in the northern part of the site that cut across the earthwork or near it. Quimby (1965) reported that he found little in the way of artifacts and no evidence for prehistoric posts. Quimby concluded the earthwork was either a Middle Woodland ceremonial site or a Late Woodland settlement with a palisade despite not finding any posts (Quimby 1965).
Other than Quimby, and a brief survey by the Office of the State Archaeologist (Mead 1986) the only other archaeological work conducted at the South Flats enclosure was done by Grand Valley State University as part of an archaeological field school in 2006 (Gaff and Brashler 2011). That project had three elements. The first part of the project involved establishing a site grid and collecting data for generating topographic maps. Given the remote location, an arbitrary datum was established southeast of the earthwork and a site grid oriented on magnetic north. Almost 400 data points were collected using a Nikon Pulse Laser Station. The grid served as a reference for the second part of the project--shovel testing inside and outside the enclosure. Shovel tests were excavated outside the 30 m wide earthwork in 10 m intervals and inside it at 5 m intervals, with all shovel tests taking place at grid intersections. The earthwork itself was not subjected to testing. To increase data quality over standard shovel testing practices where small holes are shoveled out to a fixed depth, shovel tests at South Flats were excavated as 50 cm x 50 cm units with arbitrary levels of 10 cm when cultural or natural strata were not evident. The third part of Grand Valley's archaeological project was to re-excavate Quimby's units that had been left open. These excavations were irregular and were not oriented on the Grand Valley State University grid. The technique used was to remove levels from each of Quimby's units as well as to cut back the walls a few centimeters to expose fresh profiles and plan views of the earthwork. At the end of the project, all units were backfilled and the earthwork restored to conform to the existing undisturbed landscape.
Results of these excavations produced evidence from two features with radiocarbon dates of AD 1400-1640 and 1440-1650, both of which are two sigma calibrated dates reported in Gaff and Brashler (2011). In addition, 399 ceramic body sherds, five rim sherds and five Late Prehistoric triangular projectile points were recovered, two points from Feature 1 and three from Feature 2. Small quantities of floral and faunal remains (mostly mammal) were recovered, again, primarily from the two features, and a single piece of flat native copper was recovered from Feature 1. No cultigens of any kind were recovered. Further supporting the lack of evidence for maize at the site are the initial studies by Hurd (2013) and Teall (2013) on residue from one of the rims from Feature 1. No evidence of either maize phytoliths or maize starch was identified.
Two possible large posts were noted in 2006, but no wall trench or post pattern was evident in the construction of the earthwork, nor was there a pattern to the placement of the posts. If these two large and somewhat amorphous features do indicate posts, their location is some distance from each other on the edge of the inside of the enclosure, so these would have been spaced widely around the enclosure in stead of in lines of post molds typically found ar sites associated with prehistoric warfare (Milner 1999; Keeley 1996). Keeley (1996) and others suggest that in some cases embankments alone can be indicative of fortifications. That the embankment at South Flats is less than 0.5 m in height argues against the site being used for physical defense. Its position on a high bluff above the river also might be interpreted as defensive, but access to the site from the interior is a simple walk across level ground, and there are many similar bluff top settings along this stretch of the river. Finally, the possible posts were much larger (0.5-0.7 m) in diameter and somewhat amorphous compared to typical post sizes found in fortified embankments elsewhere in the eastern U. S. If they are intentionally placed posts or large trees left as markers, a more reasonable interpretation might be that they are associated with some ritual or calendrical function. Both of these functions have been attributed to isolated or infrequent post features occurring at Late Prehistoric sites in the Midwest. The two possible posts associated with the enclosure do not however fall on any solstice, equinox or cardinal direction access that can be detected.
To further understand the internal structure of the site, a crew from Grand Valley State University returned to South Flats in 2010 to conduct surface geophysical surveys. While it is customary to do archaeological testing after a geophysical survey in order to ground truth survey results, in this case, the geophysical survey was carried out after excavation to look for and better estimate the number of possible features at the site. If additional work should be planned for the site, the geophysical information will be extremely critical in guiding additional excavation of this fragile site.
Despite the site's seeming simplicity, the density of material within the enclosure and two features, suggested that some subsurface complexity was yet to be discovered. The geophysical surveys made it possible to determine the potential of other features being found within and close by the earthwork. The use of geophysics offered the opportunity to "see" the entirety of the site without having to damage the enclosure through more excavation. Geophysical data could also be used to better understand the role that cache pits and storage play at enclosure sites. Grand Valley State University's 2010 geophysics project had the potential to yield data contributing to the discussion about earthworks and Late Prehistory settlement in Michigan. To the best of the authors' knowledge, this represents the only use of ground penetrating radar and magnetometry at an earthwork site in Michigan.
Geophysical investigation of archaeological sites is common practice (Conyers and Goodman, 1997; Kvamme 2001; 2003) and these methods have been applied to earthwork structures in several field and laboratory studies (Dalan and Walters, 1996; Dalan and Bevan, 2002; Stone, 2003). The aim of the geophysical surveys at the South Flats Earthwork 20MU2 was to locate any previous excavations within the archaeological grid, to identify for future excavation potential pit features similar to those found in the 2006 excavations containing burnt material at the base, to determine possible entry ways through the apparently continuous earthwork structure, and to identify other anomalous areas for further investigation. To achieve these goals, two geophysical instruments (magnetometer and ground penetrating radar) were used to survey the site. Geophysical surveying was performed using a GSSI (Geophysical Survey Systems Inc.) SIR 3000 GPR system and a Geometrics G-858 Cesium Vapor Magnetometer in dual sensor mode.
Magnetic survey methods are ideal for locating burned areas, hearths, kilns, and fired bricks due to thermoremanent magnetism (Kvamme, 2001). Additionally, the natural magnetic susceptibility of the soil will be altered when trenches, ditches, or other voids are back-filled (Kvamme, 2001). This alteration will be recorded as a magnetic anomaly in magnetic field/gradient surveys. A Geometries Model G-858 Mag Mapper Cesium Vapor Magnetometer was used with dual sensors at 0.7m spacing to survey the shallow subsurface magnetic properties of the survey area (Figures 2 and 3). The G-858 Cesium Vapor Magnetometer is a highly sensitive magnetic surveying instrument. This device can he used to locate foundations, ferrous artifacts, and burn pits at archaeological sites. The maps generated for this investigation show local highs and lows which are representative of buried or surface ferrous objects that enhance the earth's natural magnetic field as well as alterations of the background magnetic field from disturbances of the soil profile. Some objects exhibit a positive (pink/orange) and negative (blue/purple) trend. For a ferrous object, this represents the dipolar nature of the material and indicates that the object is probably oriented sub-horizontally with respect to its magnetic dipole. If the anomaly shows one pole, the magnetic dipole of the object is probably oriented vertically. Burn pits have also been observed as monopolar anomalies in previous surveys (Sherrod, et al. 2003). Objects at both the surface and at depth can cause magnetic anomalies, but the magnitude of the anomaly due to a small source is impacted significantly by proximity to the sensors.
Ground penetrating radar (GPR) is another geophysical method that can be very useful in the field of archaeology. A GPR system emits an electromagnetic (EM) pulse into the subsurface and measures the time it takes for the signal to reflect back to the antenna as well as the amplitude of that signal when it reaches the antenna. Reflections of the EM pulse are caused by changes in the electromagnetic properties of the subsurface such as electrical conductivity, relative permittivity, and magnetic permeability. Large changes in these properties result in strong amplitude reflections. Strong reflections can be expected from the interface between soils and rocks and from stratigraphic boundaries of intense electrical property contrast such as sands (high resistivity) to clays (low resistivity). However, the amplitude of the GPR signal becomes greatly diminished when it passes through low resistivity units such as clays. Therefore, clay-rich zones will be apparent in the GPR profiles as sections of high signal attenuation. The sandy nature of the soil at this site was ideal for GPR penetration. In general, solid metallic objects in the subsurface may be identified as strong, ringing reflections. A discrete object (including tree roots) may produce hyperbolic reflections in the survey profile while the center will he represented as a vertical sequence of positive and negative reflected peaks.
Geophysical Data Collection at South Flats
A geophysical survey grid was established with reference to the reinforcing rods that were found onsite in 2006, originally put in place by Quimby. This geophysical grid is not oriented in the true cardinal directions. Instead, the "north" axis is approximately 59[degrees] east of magnetic north. Coordinates within the geophysical grid increase to the southeast and to the northeast. Exact coordinates for the reinforcing rod within the geophysical survey grid are (25SE, ONE) and (25SE, 42NE). The geophysical surveys were performed on this grid with reference to the west corner of the grid as 0 m southeast and 0 m northeast (OSE, ONE). The Geometrics G-858 Cesium Vapor Magnetometer was used to obtain magnetic field data. This instrument was used with two sensors (in vertical gradiometer mode). The area was surveyed along approximately southwest to northeast survey lines incrementing in a southeasterly direction every 1 m from OSE to 44SE. The data were collected with a 10-readings per second frequency in bi-directional survey mode. Fiducial marks were inserted every meter from ONE to 42NE.
The GSS1 SIR 3000 was used with a 400MHz antenna to collect the GPR data. The scan length was set at 65 nanoseconds and each profile was stored as an individual file. The operator pulled the antenna along the survey line in continuous collection mode. Lines were surveyed over the entire survey area in a similar grid pattern to that used in the magnetometer survey. Line spacing and fiducial marking for the GPR survey was identical to that used for the magnetometer survey.
Geophysical Survey Results
Magnetometer data were reconciled and plotted using Oasis Montaj as shown in Figures 2 and 3.
A slight horizontal magnetic gradient of approximately 12nT difference between the northwest and southeast side of the survey area was noted and filtered from the data. The vertical magnetic gradient was obtained by subtracting the magnetic field reading of the top sensor from the magnetic field reading of the bottom sensor and dividing by the sensor separation (0.7m). By rotating the map and applying a transparency to it, the magnetic data were overlain on the archaeological site maps. Figure 2A shows the outline of the archaeological site, the excavation pits created by Quimby in 1937 (A through F) and the location of the two features from the 2006 excavations (F1 and F2). The vertical magnetic gradient plot is overlain on this base map in Figure 2B. Figure 3A shows the location of positive and negative shovel test pits from the 2006 excavations, with the total magnetic field from the bottom sensor overlain on that base map in Figure 3B. A positive shovel test pit is one in which any kind of human signature (artifacts or soil stains) was discovered while a negative shovel test pit is one in which there is no evidence of human activity. These images allow for a rapid identification of anomalous responses, which were caused by known previous excavations and those that are unlikely to have been caused by such disturbances.
GPR data were processed in Reflexw to generate three-dimensional images for analysis. The earthwork structure itself may be clearly observed in the two shallowest timeslices produced from these data, as shown in Figures 4A and 4B. The deeper time slices show high amplitude responses at depths of 0.4m (Figure 4C) and 0.7m (Figure 4D). Although the subsurface was ideal for GPR in its high sand content, most images contain a significant amount of noise at shallow depths from tree roots and previous excavations. The timeslices of Figure 4 show the location of trees within the survey area and Figure 5 shows the location of trees adjacent to the survey lines presented as profiles. Anomalous responses relating to the earthwork structure appear to be confined to the upper 1.0m of the subsurface (assuming a subsurface velocity for the GPR signal of 0.1 m/ns as determined from diffraction hyperbolae in the profiles).
Interpretation of Geophysical Data
The northern edge of the site contains significant magnetic noise from the 1937 excavations of Quimby and the reconstruction of the earthwork that occurred during the 2006 field season (to replace the material excavated by Quimby). Additionally, the rebar that was put in place by Quimby produces a magnetic response, which not only produces noise at the northeast and southwest edges of the plot, but also allows the geophysical survey grid to be aligned with certainty on the 2006 archaeological grid. The large, positive monopolar anomaly centered at 10SE, 23NE in the geophysical survey grid is a very significant disturbance of the magnetic field within this area. The 2006 shovel test pit records indicate that there had been a recent lire at the surface. Thus, the anomalous response is caused by modern sources and may mask more subtle anomalies, relating the prehistoric earthwork time frame, at depth.
Despite these disruptions to the background magnetic field, there are several small anomalies that appear to have no correlation to the old excavations, reconstructions, or background fields. The southern tip of the survey has four distinct magnetic high responses, which may be related to the earthwork structure as they are also apparent as strong GPR reflections within the upper 1m of the subsurface (located at (28SE, 4NE), (33SE, 12.5NE), (39SE, 19.5NE) and (42SE, 7NE) in the geophysical survey grid system). These occur both on the structure and outside the structure, and may be of interest for future excavation efforts as fewer artifacts were recovered outside the structure than inside it. The western-most of these magnetic anomalies is somewhat distorted by the magnetic anomaly generated by the rebar on the southwestern edge of the site. These anomalous responses are highlighted with a dashed gray line in Figures 2 and 3 and are roughly east-west in orientation. Additionally, the series of bipolar anomalies found in an arced line within the structure from (8SE, 13NE) to (25SE, 29NE) in the geophysical survey grid system may be of interest. This series of features is marked in the magnetic maps with a dashed gray line. The eastern-most anomaly in this series matches the location of the second pit feature that was discovered in 2006 and used to obtain a radiocarbon date of 370 [+ or -] 50 BP (calibrated AD 1440-1650) (Gaff and Brashler 2011). With the exception of Arthurburg Hill, an Early Woodland enclosure (Garland and Beld 1999), the South Flats dates comfortably fit with dates for Late Prehistoric enclosures (Howey 2012).
Several of the anomalous magnetometer responses that correlate to positive shovel test pits are at locations where charcoal fragments were found within the test pits. In the process of excavation, much of the charcoal was removed from the subsurface. Therefore, all such units have a depleted quantity of charcoal compared to their original amount. Feature 1, discovered during the 2006 excavations and located at (17SE, 17NE) within the geophysical survey grid has a very modest magnetic response. However, a significant quantity of charcoal and burnt logs was recovered from this excavation unit. The excavation and charred material at these locations may have altered the magnetic signature of the burnt area. Feature 2, located at (26SE, 30NE) within the geophysical grid, contained much smaller quantities of charcoal compared to F1 according to the 2006 excavation records. The geophysical response shown near this location is very large, but notably similar to the magnetic response from the previous excavations performed along the northern boundary of the earthwork. Further, this feature was not completely excavated, or, the bipolar magnetic anomaly at this location may be caused by the excavation, and not the effects of the feature discovered at this location.
The outline of the earthwork structure shown in the timeslice image of Figures 4A and 4B illustrates the applicability of GPR in this environment. Additionally, the subsurface at the location of pit Feature 2, as described from the 2006 archaeology field season, shows a very distinct hyperbolic GPR response at a depth just under 1.0m, as shown in Figures 4C, 4D, and 5. The proximity of this reflector to the excavation unit indicates that the source of the anomalous response is related to either something directly adjacent to that excavation unit or to the walls of the excavation unit itself. However, the strong magnetic anomaly at this location gives rise to a need for further investigation. There are similar GPR features beneath other magnetic anomalies. Figure 5 shows two such hyperbolic responses at (7SE, 12NE) and (7SE, 22NE). The first of these corresponds to the western edge of the series of bipolar magnetic anomalies described above in an arced line across the survey area. This reflector is located near, but not directly adjacent to, a negative shovel test pit. The second corresponds to the edge of one of the 2006 excavation units, which resulted in a positive shovel test pit. Several of the prominent features in both the magnetic and GPR results may be caused by the remnants of previous excavations performed at this site. Likewise, the large trees on site contributed to a noisy response in several of the GPR profiles. This can be seen in the timeslices presented in Figure 4 as several of the high amplitude responses are centered around the red dots representing trees in this image.
Data from the magnetometer and GPR suggest possible subsurface features of archaeological interest at several locations across the site, both inside and outside the earthwork structure. These results suggest that there are additional archaeological features within the earthwork that were not excavated in 2006. However, similar to the excavation results, the density of features is light. Sampling with shovel testing inside the earthwork revealed two features (F1 and F2). The geophysical surveys produced six possible features, two within the enclosure, one on the edge of the enclosure and three immediately outside, producing a more thorough coverage of the earthwork's interior and so is in line with the results from shovel testing. While these new anomalies have not been ground truthed, they present signatures similar to the features previously identified, and their location was supported by two independent geophysical techniques. Thus, we are reasonably confident that we have located most if not all of the features within and near the earthwork, though we do not know the exact nature of the unexcavated ones. The important fact is that there appear to he few features or possible features present within and adjacent to the enclosure.
One of the inferences to draw from these data is this low density of features at the site. There is no evidence for overlapping features or dense clusters of features indicative of either long term or repeated use--a conclusion supported by excavation data (Gaff and Brashler 2011). Also noteworthy about these results is that the number of possible features combined with the two verified from excavation equals a number of features easily created by a limited number of people in a relatively short period of time or very sporadically perhaps over a one to a few generations, allowing for the standard deviations in the overlapping radiocarbon dates. The historic record indicates that it was well within the means of a small family group to create several features in a short amount of time (e.g., Wilson 1987).
If we accept, albeit without ground the features in the interest of preservation, the two anomalies within the enclosure, with similar signatures as the two excavated features represent two similar features we then have a total of four features whose function is critical to understanding the activities occurring within the enclosure. Based on the two excavated features, several possible interpretations can be made. First, they are morphologically similar to pits labeled as cache pits that occur with some frequency throughout the western and northern lower Peninsula which have been described by several authors (Dustin 1966; Hambacher et al. 1995; Dunham 2000; Howey and Parker 2008). Dunham (2000,225) characterizes these enigmatic features as the most abundant archaeological feature in Michigan. Recently investigations conducted under the auspices of the Federal Highway Administration and the Michigan Department of Transportation at 200T283 by Commonwealth Cultural Resources Group (CCRG) resulted in excavation of a large number of these pits. Many of these features are similar to those from South Flats in terms of their structure and evidence of burning (Hambacher 2012; Hambacher personal communication). At 200T283 and elsewhere, cache pit features are manifest at the surface as circular depressions and are almost always empty. They occur in clusters near occupation sites as well as near enclosure sites (Howey 2006; 2012; Gaff and Brashler 2011).
There is considerable diversity in the internal structure of archaeologically and ethnohistorically described cache pits (Densmore 1929; Dunham 2000; Howey and Parker 2008), including pits that were used once and others that were possibly used twice. In an oft-quoted passage in attempts to understand caching behavior in the Great Lakes region, Densmore describes the storage of food among the historic Chippewa in the following way:
... it was the custom to store food obtained during the summer in caches or pits dug near the village. The food kept perfectly, the pits were never disturbed, and this method of storage was safe and practical. The women of two or three families usually combined the work of storing food, and often put rice, sugar and vegetables in separate pits. Seed potatoes and seed corn were stored in a similar manner. A food cache was usually about six feet deep and was lined with birch bark. The rice and sugar were in makuks and after they were in place the spaces between them were filled with hay. When the pit was nearly filled a covering of birch bark or hay was added. Beams of wood were laid across and the whole was covered with a mound Beams of wood were laid across and the whole was covered with a mound
Howey and Parker (2008) note both similarities and differences between this description and excavated archaeological cache pits. For our purposes, we argue that the two features excavated at South Flats are morphologically similar to both her description (though not in terms of contents) and some of the cache pits excavated at 200T283 by CCRG. Here pits appear to have been lined with wood, which was burned either before storage took place, or possibly after stored food was removed, so that the pit could be used again. Similar features were identified at another locale on the Grand River excavated by GVSU during the summer of 2012. What is striking about the South Flats features, however, is that while they are morphologically like cache pits elsewhere, and contain evidence of burning seen at other cache pit sites in Michigan, the two South Flats pits have refuse debris including animal remains, triangular projectile points, diagnostic rim sherds, and in the case of Feature 1, a piece of copper. These materials (copper and triangular points) do not occur elsewhere within the enclosure, or in its near proximity with the exception of a single triangular point from within the earthwork recovered in a shovel test. The two points in Feature 1 were made from glacial till chert. Two points from Feature 2 were made from Bayport chert (from the Saginaw Bay area) and a glacial till chert. Other raw materials present on the site include Deer Lick Creek chert, Lambrix and Norwood, chert common to the west side of Michigan throughout the later Late Woodland (Luedtke 1976).
The placement of points and copper artifacts, combined with the pit morphology and distribution of other materials inside and near the enclosure suggests the possibility that the pits were excavated as storage features initially, burned and then used to incorporate refuse from a meal or meals, with symbolically charged objects including copper and diagnostic lithics. There are, of course other possible explanations, (roasting pits filled in with refuse and symbolic objects) but the possibility that the pits served both storage and subsequently refuse/ceremonial deposit functions seems the most parsimonious given the internal structure and distributions identified through excavation enhanced by the possible presence of two additional features identified through the geophysical survey.
If we accept thar there are four such pits within the enclosure together with what is known from the two excavated pits and the results of the geophysical surveys, these pits could have been created by a limited number of people in a relatively short time, or perhaps over several short term uses by groups separated by any where from a season, to a decade or more. Again we chose not to ground truth the other two features to consciously preserve the limited fragile remains of the enclosure so our interpretations are somewhat limited by our 50 percent sample.
Another significant contribution of the geophysical survey, revealed in the GPR timeslice (Figure 4), is confirmation of the location and the size of the enclosure. Due 10 vegetation, recent erosion (ORV traffic) and the earlier excavation, the exact boundaries of the enclosure were not verifiable with the limited test excavations conducted in 2006. We now see the enclosure external limits between 35 and 40 m in diameter. This remains one of the smallest documented enclosures in Michigan, both in terms of its small enclosed space as well as slight relief and estimated volume of earth moved to construct it (ca. 300 cubic meters). Furthermore, we do not see conclusive evidence in the geophysics data for an entryway, which we speculated may have existed based on what appeared to be a slight gap in the above ground earthwork (Gaff and Brashler 2011). Thar gap does not appear in the geophysics data. The anomalies within the enclosure and the two excavated features do not appear patterned in any way though they form an arc through roughly the center of the enclosure running roughly from east to west (Figure 3.) Finally, one of the other anomalies identified in the survey appears to occur within the earthwork itself, while the other three occur near the exterior of the enclosure. The anomaly in the enclosure is difficult to explain, though perhaps it represents a third deep post-like feature, though this remains a speculation at best.
Summary and Conclusions
In 2011, Gaff and Brashler proposed that earthworks emerge as a reflection of patrilineal bands augmenting social networks in a period of subsistence uncertainty as group territories filled and indigenous inhabitants experienced pressure from populations expanding from the east and south. The scale of storage or for that matter of any activity at South Flats represented by a handful of pits at most, is consistent with mobility patterns seen by small groups of hunter-gatherers characterized by the Chippewa pattern as described by Holman and Lovis (2008). These groups practiced residential mobility with logistical forays to secure resources, meet with other groups, perhaps feast and exchange kin. The numbers of people meeting at South Flats must have been few based on the small number of features and light amount of debris. The presence of diagnostic lithics and lithic raw materials predominantly from the west Michigan coastal area is in line with the Chippewa model of groups moving up and down the coastline seasonally perhaps encountering other groups in shared territories at places like South Flats. Though small in number, the meeting of south with north is also supported by the presence of a single Moccasin Bluff sherd and a Juntunen rim in Feature 1. The absence of maize and fish, and the presence of mammalian fauna at the site is in keeping with hunter gatherer groups meeting, exchanging and perhaps feasting at a place marked by a modest monument.
The location of South Flats, an almost inconspicuous monument on one of many conspicuous high bluffs overlooking the Muskegon is significant, but in what way we are hard pressed to decipher. The South Flats enclosure is an unusual site placed in a location 20 km from the lakeshore, with meaning that does not seem inherent in its surroundings. In this regard, it is like some other enclosures located in surprisingly usual settings, and unlike others such as Missaukee situated on a geological feature of some prominence. It is beyond the scope of this paper, however, to speculate in regards to the meaning of the site location other than to suggest that it does not appear to be the work of horticultural or fishing peoples gathering to exchange those foodstuffs, but the work of small groups of hunter-gatherers.
Perhaps the most important contribution to be drawn from the geophysical work at South Flats is that it amplifies what is known of the interior structure by suggesting that there were probably no more than four features most likely used for both storage and refuse disposal for events that most likely took place within or very near the enclosure. The use of magnetometry and GPR for this project demonstrates that these techniques can be used to better understand the internal content and structure of earthen enclosures in forested environments, something not heretofore accomplished in Michigan archaeology.
Many questions remain unanswered about Michigan's earthworks and only more research into them will yield answers. More work is needed to understand the caching behavior of Michigan's prehistoric inhabitants, including an exhaustive review of the ethnographic literature related to caching and storage. A systematic comparison of a large number of pits from multiple sites is critical, as is a more careful analysis of their placement on the landscape. The employment of geophysical strategies such as magnetometry and ground penetrating radar in this research is one example of how problems associated with these interesting sites can be approached. The research described here has shown that archaeologists and geophysicists working in an interdisciplinary fashion can generate new kinds of data to answer on going questions about the Late Prehistoric period in Michigan.
We gratefully acknowledge the comments provided by several anonymous reviewers. In addition, we appreciate the conversations with several colleagues, especially Robert Mainfort and Michael Hambacher regarding earthworks and the interesting ongoing discussions concerning Michigan's late prehistory.
ALEXANDER, J. 2006. The Muskegon: The Majesty and Tragedy of Michigan's Rarest River. East Lansing, MI: Michigan State University Press.
ASHMORE, W. A. AND C. BLACKMORE. 2008. Landscape Archaeology. In Encyclopedia of Archaeology, ed. Deborah Pearsall, 1569-1578. Philadelphia, PA. Elsevier Science.
ASHMORE W. AND A. B. KNAPP. 1999. Archaeologies of Landscape: Contemporary Perspectives. Maiden, MA: Blackwell Publishers.
BASSO, K. H. 1996. Wisdom Sits in Places: Landscape and Language Among the Western Apache. Albuquerque, NM: University of New Mexico Press.
BINFORD, L. R. 1980. Willow Smoke and Dogs' Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation. American Antiquity 45(1): 1-17.
--. 1982. The Archaeology of Place. Journal of Anthropological Archaeology. 1 (1): 5-31.
--. 1983. Long-term Land Use Patterns: Some Implications for Archaeology. In Working at Archaeology, ed. L. R. Binford, 379-386. New York: Academic Press.
BRASHLER, J. G., E. B. GARLAND, M. B. HOLMAN, W. A. LOVIS, AND S. R. MARTIN. 2000 . Adaptive Strategies and Socioeconomic Systems in Northern Great Lakes Riverine Environments: The Late Woodland of Michigan. In Late Woodland Societies, ed T. E. Emerson, D. L. McElrath, and A. C. Fortier, 543-579. Lincoln, NE: University of Nebraska Press.
CONYERS, L.B., AND D. GOODMAN. 1997. Ground-Penetrating Radar. Lanham, MD: Rowman & Littlefield.
CORNELIUS, E. S. AND H. W. MOLL. 1961. The Walters-Linseman Earthwork Site. The Totem Pole 44(9).
DALAN, R.A., AND B.W. BEVAN. 2002. Geophysical Indicators of Culturally Emplaced Soils and Sediments. Geoarchaeology 17 (8): 779-810.
DALAN, R.A., AND H.W. WALTERS, JR. 1996. The Application of Soil Magnetic Techniques for Determining the Original Form of Prehistoric Earthworks. SEG Annual Meeting Expanded Technical Program Abstracts with Biographies. 66: 790-793.
DENSMORE, F. 1929. Chippewa Customs. Smithsonian Institution Bureau of American Ethnology Bulletin 86, Washington, D.C: Smithsonian Institution Press.
DUNHAM, S.B. 2000. Cache Pits: Ethnohistory, Archaeology, and the Continuity of Tradition. In Interpretations of Native North American Life: Material Contributions to Ethnohistory, ed. M.S. Nassaney and E. S. Johnson, 223-260. Gainesville FL: University Press of Florida.
FITTING, J. E. AND C. E. CLELAND. 1969. Late Prehistoric Settlement Patterns in the Upper Great Lakes. Ethnohistory 16(4): 289-302.
GAFF, D. H. AND J. G. BRASHLER. 2011. The South Flats Earthwork, 20MU2. Midcontinental Journal of Archaeology 36(2): 195-226.
GARLAND, E. B. AND S. G. BELD. 1999. The Early Woodland: Ceramics, Domesticated Plants, and Burial Mounds Foretell the Shape of the Future. In Retrieving Michigan's Buried Past, ed. by J. R. Halsey, 125-146. Bloomfield Hills, MI: Cranbrook Institute of Science.
GOLDSTEIN, L. A. 1995. Landscape and Mortuary Practices: a Case for Regional Perspectives. In Regional Approaches to Mortuary Analysis ed. L. A Beck 101-121. New York: Plenum Press.
GREENMAN, E. E. 1927. The Earthwork Inclosures of Michigan. Ph.D Dissertation Univerisity of Michigan, Ann Arbor, MI.
HAMBACHER, M. J. 2012. Staging the Landscape: Storage and Mobility during the Late Prehistoric Period in the Lower Grand River Valley of Western Michigan as Seen from 200T283. Paper presented at the 2012 Annual Meeting of the Eastern States Archaeological Federation, Perrysburg, OH.
HAMBACHER, M. J., M. B. HOLMAN, K. ECAN, AND B. A. SMITH. 1995. Camp Cache and Carry: The Porter Creek South Site (20MN100) and Cache Pits at 20MN31 in the Manistee National Forest. The Michigan Archaeologist 41 (2-3):47-118.
HINSDALE, W.B. 1931. Archaeological Atlas of Michigan. Michigan Handbook Series 4. Ann Arhor, MI: University of Michigan Press.
HOLMAN, M. B. AND W. A. LOVIS. 2008. The Social and Environmental Constraints on Mobility in the Late Prehistoric Upper Great Lakes Region. In The Archaeology of Mobility: Old World and New World Nomadism, Costen Advanced Seminar 4, ed. Hans Barnard and Willeke Wedrich, 280-306. Los Angeles, CA: Costen Institute of Archaeology, University of California-Los Angeles.
HOWEY, M. C. L. 2006. Ritual, Resources, and Regional Organization in the Upper Great Lakes, A.D. 1200-1600. Ph.D. dissertation, Department of Anthropology, The University of Michigan, Ann Arbor, MI.
--. 2012. Mound Builders and Monument Makers of the Northern Great Lakes, 1200-1600. Norman, OK: University of Oklahoma Press.
HOWEY, M. C. L. AND J. M. O'SHEA. 2006. Bear's Journey and the Study of Ritual in Archaeology. American Antiquity 71(2): 261-282.
HOWEY, M. C. L. AND K. E. PARKER. 2008. Camp, Cache, Stay Awhile: Preliminary Considerations of the Social and Economic Processes of Cache Pits along Douglas Lake, Michigan. Michigan Archaeologist 54 (1-4): 19-44.
HURD, K. 2013. Extraction and Analysis of Starch Residues Utilizing a Heavy Liquid Technique from Native American Woodland Ceramics. Paper presented at the 2013 Student Scholarship Day, Grand Valley State University, Allendale, MI.
KEELEY, L. H. 1996. War Before Civilization: The Myth of the Peaceful Savage. New York: Oxford University Press.
KELLY, R. L. 1992. Mobility/Sedentism: Concepts, Archaeological Measures, and Effects in Annual Review of Anthropology 21:43-66.
--. 1995 The Lifeways of Hunter - Gatherers: The Foraging Spectrum. New York: Cambridge Universiry Press.
KNAPP, A. B. AND W. ASHMORE. 1999. Archaeological Landscapes: Constructed, Conceptualized, Ideational in Archaeobgies of Landscape: Contemporary Perspectives, ed. W. Ashmore and A. B. Knapp. Maiden, MA: Blackwell Publishers.
KOHL, J. G. 1985. Kitchi-Gami: Life among the Lake Superior Ojihway. St. Paul, MN: Minnesota Historical Society Press.
KVAMME, K.L. 2001. Current Practices in Archaeogeophysics: Magnetics, Resistivity, Conductivity, and Ground-Penetrating Radar. In Earth Sciences and Archaeology ed. P. Goldberg, V.T. Holliday, C.R. Ferring. 353-384. New York: Springer.
--. 2003. Geophysical Surveys as Landscape Archaeology. American Antiquity 68:435-457
LEE, R. B. AND I. DEVORE eds. 1968. Man the Hunter, Chicago, IL: Aldine.
LEUDTKE, B. E. 1976. Lithic Material Distributions and Interaction Patterns during the Late Woodland Period in Michigan. Ph.D. dissertation, University of Michigan, Ann Arbor.
MAINFORT, R. C, JR. AND L. P. SULLIVAN. 1998. Explaining Earthen Enclosures. In Ancient Earthen Enclosure of the Eastern Woodlands, ed. R. C. Mainfort and L. P. Sullivan, 1-16. Gainesville, FL. University Press of Florida.
MCCLURKEN, J. A. 1988. We Wish to Be Civilized: Ottawa-American Political Contests on the Michigan Frontier. PhD Dissertation, Department of Anthropology Michigan State University, East Lansing, MI.
MEAD, B. 1986. Original unpublished notes on field trip to Muskegon Game Area and Mounds. Norton Ms. on file, Bureau of History, Michigan Department of State, Lansing, MI.
MICHIGAN'S NATURAL FEATURES INVENTORY. 2013. Online resource Lansing, MI: Michigan State University Extension. Retrieved from http://mnfi.anr.msu.edu/data/veg1800.cfm.
MILNER, C. M. AND J. M. O'SHEA. 1998. The Socioeconomic Role of Late Woodland Enclosures in Northern Lower Michigan. In Ancient Earthen Enclosure of the Eastern Woodlands, ed. R. C. Mainfort and L. P. Sullivan, 181-201. Gainesville, FL. University Press of Florida.
MILNER, G. R. 1999. Warfare in Prehistoric and Early Historic Eastern North America. Journal of Archaeological Research 7(2): 105-151.
--. 2004. The Mound Builders: Ancient Peoples of Eastern North America. Thames and Hudson, London.
O'GORMAN, J. A. 2003. The 'Agricultural Village:' Myth and Recent Research at Moccasin Bluff. Michigan Archaeologist 49 (1-2): 91-100.
--. 2007. The Myth of Moccasin Bluff-Reth inking the Pottawatomi Pattern. Ethnohistory 54 (3): 373-406.
Neusius, S. W., L. P. SULLIVAN, P. D. NEUSIUS, AND C. M. MILNER. 1998. Fortified Village or Mortuary Site? Exploring the Use of the Ripley Site in Ancient Earthen Enclosures of the Eastern Woodlands, ed. R. C. Mainfort and L. Sullivan. 202-230, Gainesville, FL University of Florida Press.
O'GORMAN J. A. AND W. A. LOVIS. 2006. Before Removal: An Archaeological Perspective on the Southern Lake Michigan Basin. Midcontinental Journal of Archaeology 31 (1): 21-56.
O'SHEA J. AND P. HALSTED. 1989. Conclusion: Bad Year Economics in Bad Year Economics: Cultural Responses to Risk, ed. P. Halstead and J. O'Shea, 123-126. Cambridge University Press, Cambridge.
QUIMBY, G. I. 1965. An Indian Earthwork in Muskegon County, Michigan. Michigan Archaeologist 11 (3-4): 165-169.
--. 1993. A Thing of Sherds and Patches. American Antiquity 58(1): 7-21
SAUER, C. O. 1925. The Morphology of Landscapes. University of California Publications in Geography 2: 19-54.
SHERROD, L., D. LYNCH AND W. A. SAUCK. 2003. Geophysical investigations at the archaeological site of Fort St. Joseph, Niles, MI. Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmental Problems (SAGEEP '03). San Antonio, TX, April 6-10.
SPIELMAN, K. 1986. Interdependence Among Egalitarian Societies. Journal of Anthropological Archaeology 5:279-312.
STONE, J. 2003. Short Report: Cawrhorn Camps, North Yorkshire-a Photogrammetric Approach. Archaeological Prospection. 10: 153-157.
TEALL, E. 2013 Phytolith Recovery from Native American Woodland Ceramics Utilizing a Heavy Liquid Technique, Paper presented at the 2013 Student Scholarship Day, Grand Valley State University, Allendale, MI.
THORNTON, T.F. 2008. Being and Place Among the Tlingit. Seattle WA: University of Washington Press.
UNITED STATE DEPARTMENT OF AGRICULTURE, SOIL CONSERVATION SERVICE. 1968. Soil Survey: Muskegon County Michigan. Washington, D.C: U. S. Government Printing Office.
WHITE, A. A. R. G. AND D. MCCULLOGUH, MCCULLOUGH. 2003. Archaeological Investigations at Two Late Prehistoric Earthen Enclosures in Indiana. Reports of Investigations 301. Fort Wayne, IN: Indiana University-Purdue University at Fort Wayne.
WILSON, G. 1987. Buffalo Bird Woman's Garden. Minneapolis, MN: Minnesota Historical Society Press.
ZUREL, R. L. 1999. Earthwork Enclosure Sites in Michigan. In Retrieving Michigan's Buried Past, ed. J. R. Halsey, 244-248. Bloomfield Hills, MI: Cranhrook Institute of Science.
University of Northern Iowa
Kutztown University of Pennsylvania
JANET G. BRASHLER
Grand Valley State University
|Printer friendly Cite/link Email Feedback|
|Author:||Gaff, Donald; Sherrod, Laura; Brashler, Janet G.|
|Date:||Sep 22, 2013|
|Next Article:||Trends in michigan supreme court elections 2000-2012.|