Sources - Papers
This article was printed in New Perspectives on Cahokia Views from the Periphery edited by James B. Stoltman published in 1991 by Prehistory Press.
The Diamond Bluff Site Complex and Cahokia Influence in the Red Wing Locality
Roland L. Rodell
Cahokia influence in the northern reaches of the upper Mississippi Valley is recognized among a handful of sites in the Red Wing Locality that have Middle Mississippian-inspired traits. Within this group of sites is the Diamond Bluff mound and village site complex. Although known to local antiquarians during the 19th century (cf. Svec 1985), Diamond Bluff was not identified with Mississippian culture until the late 1940s, when Middle Mississippian related artifacts were recovered from the site area (Lawshe 1947; Maxwell 1950a, 1950b). Since then Diamond Bluff has periodically been referenced in writings pertaining to Middle Mississippian in the upper Mississippi Valley (Gibbon 1974, 1991; Griffin 1960; Hall 1962; Kelly 1991; Stoltman 1986; Williams and Goggin 1956).
The Diamond Bluff site complex is comprised of at least two large village areas and a large mound group that includes a variety of mound types characteristic of Late Woodland mound building traditions in the Upper Midwest. What has made Diamond Bluff interesting (as well as confusing) is that not only have Late Woodland and Oneota artifacts been recovered from both mound and village contexts, but mixed with them are Middle Mississippian-inspired artifacts. Most notable are the Ramey Incised motif on locally made ceramic jars and artifacts identified with the Southeastern Ceremonial Complex. These Middle Mississippian-inspired artifacts are the signature traits of the Silvernale Phase and are found at a number of sites in the Red Wing Locality, most notably the Bryan, Silvernale, and Energy Park sites. Since these sites are discussed elsewhere in this volume, this presentation will focus primarily on the archaeology of the Diamond Bluff site complex.
The objective of this paper is to address two areas that are specific to understanding the archaeology of the Diamond Bluff site complex and the theme of this volume. First, a review of archaeological investigations of Diamond Bluff is presented. The site complex has been of interest to antiquarians and archaeologists for over a century, from the early mound surveys to more recent excavations and surface collection investigations. A major portion of this segment of the paper will focus on the 1948 Wisconsin Archaeological Survey (WAS) excavations (Maxwell 1950a, 1950b). Recently, through the generosity of Dr. Moreau Maxwell of Michigan State University and students who participated in the WAS project, the University of Wisconsin-Milwaukee Archaeology Laboratory obtained field notes, maps, profiles, and photographs of the WAS excavations. These materials have provided insightful information about the mound excavations and testing of the village area. Regrettably, the artifacts from the WAS excavation are not available for study. The illustrations of pottery motifs that are presented below are taken from pencil and paper rubbings that accompany the WAS notes, while rim profiles are derived from Hall (1962).
The second part of the paper will address the more elusive problem of the Middle Mississippian presence at Diamond Bluff and in general the Red Wing Locality. The thesis favored here is that the Middle Mississippian presence in the Red Wing Locality was limited to contact between Cahokian traders and the indigenous population sometime in the early to mid 11th century A.D. The impetus that led to this contact was the expansion of the Cahokian "symbiotic-extractive exchange system" (Gibbon 1974:133) into an area that was an established locality in a regional Woodland trade network. For the indigenous Red Wing population, Cahokian contact initiated the Silvernale Phase that is recognized archaeologically by a limited variety of Middle Mississippian-inspired traits found among Late Woodland and Oneota remains. What has remained an unresolved issued is the cultural identity of the Red Wing population at the time of Cahokian contact (e.g., Gibbon 1991; Stoltman 1986). While evidence from Diamond Bluff, as well as the general Red Wing Locality, is thin, it would appear that Late Woodland peoples were utilizing the area at the time of initial Cahokian contact.
Before proceeding, some comment regarding the different site names for Diamond Bluff is in order. Depending on the source, the reader may find one or more names used to refer to the site setting. They include the "Diamond Bluff mound group," (cf. Svec 1985, 1987), the "Mero group" (Brower 1903) and "Mero site" (Lawshe 1947; Wendt and Dobbs 1989), and the "Trimbelle River Group" (Squier 1914) and "Trimbelle site" (Wends and Dobbs 1989). The "Trimbelle site" is identified in the State Historical Society of Wisconsin site code files as 47-Pi-93, while the other names are associated primarily with site area 47-Pi-2. The name Mero refers to one of three families that owned land on the Diamond Bluff terrace near the end of the 19th century. Trimbelle comes from the name of the stream that enters the Mississippi Valley at the terrace. Without exception these names identify either specific site phenomena, such as mounds, or only a portion of the overall site setting. The name Diamond Bluff is derived from the small community about a mile north of the site area. It is also the name of the terrace on which the site complex is located (Martin 1916:151), and it is the name assigned by one of the first investigators, T. H. Lewis, to the large mound group that dominated the terrace (cf. Svec 1985). Due to the intricate makeup of the site area, the name preferred by this writer and used here is the Diamond Bluff site complex.
The Diamond Bluff site complex (47-Pi-2, 93) is located in the upper Mississippi Valley at 44° 38' north latitude and 92° 36' west longitude. The Diamond Bluff terrace is on the Wisconsin side of the valley and is approximately 5 miles directly north of Red Wing, Minnesota. The site complex is found over much of the 220 acre (89 hectares) terrace, which is 1.3 miles (2.1 km) long, up to 0.4 miles (0.6 km) wide, and elevated 60 feet (18.3 meters) above the surrounding floodplain. The gently undulating terrain on the ter- race is comprised of twelve soil types with textures ranging from silt loams to sand (Haszel 1968).
In the Red Wing Locality the Mississippi Valley is about 3 miles wide. The valley is defined by steep bluffs-made of Ordovician sandstone and dolomite (Paull and Paull 1977)-that rise over 350 feet above the floodplain. Reconstructions of early 19th century vegetation patterns reveal the close proximity of a variety of plant communities Finley 1976, Marchner 1974). In the Mississippi floodplain hardwood forests of soft maple, ash, basswood, elm, willow, and cotton wood were prevalent. Many of the elevated river ter races, like the Diamond Bluff terrace, were covered by oak and prairie communities, as were extensive areas of uplands on the Minnesota side of the valley (Marchner 1974). On the Wisconsin side, the uplands had sugar maple-basswood-oak forests and tracts of oak openings (Finley 1976).
Several streams converge in the Red Wing Locality, including the Trimbelle River that enters the Mississippi Valley at the Diamond Bluff terrace, and the Cannon River on the Minnesota side of the valley opposite the Diamond Bluff terrace. Also from Minnesota, the Vermillion River enters the Mississippi River at Red Wing, and 15 miles upstream is the con-fluence of the Mississippi and St. Croix Rivers. Wisconsin's Chippewa River enters the Mississippi 25 miles downstream from Red Wing.
In an 1869 presentation on prehistoric earthworks in the Red Wing Locality, Dr. W. M. Sweney mentions a large mound group on the Wisconsin side of the Mississippi Valley: "...there are vast numbers of mounds dotted over the sandy plain between the river and bluffs. Some of these deviate from the regular cir-cular form, being composed of a main body oblong shape..., resembling the prostrate position of a bird with wings outstretched" (from Brower 1903:63). Given the more detailed survey data of subsequent investiga-tors, Diamond Bluff is the one mound group on the Wisconsin side of the valley that would have fit this description. Although Sweney does not indicate how many mounds were on the Diamond Bluff terrace, it is estimated that there must have been around 500.
The investigation conducted by T. H. Lewis provides us with the most detailed record of the Diamond Bluff mound group. Lewis surveyed the Diamond Bluff terrace in the spring of 1887 and recorded 396 mounds (Figure 12.1). At the time of the survey, Lewis estimated that nearly 150 mounds had already been destroyed by cultivation (Svec 1985). Therefore, his total may represent only 75% of the mounds observed by Sweney. Among the extant mounds were 308 conicals that ranged in height from 1 to 9 feet and diameters up to 100 feet. There were 77 linear mounds extending to lengths of 184 feet and heights of 7.5 feet. The 8 oval mounds attained heights of 7 feet and lengths of up to 80 feet. Finally, the effigy mounds included a bird effigy at the southern end of the terrace that had a wing span of 92 feet, and "wolf" and "panther" forms at the north end of the terrace. The effigies averaged 141 feet in length and 2.5 feet in height (Svec 1985).
The value of Lewis's investigation can not be overestimated. Unfortunately, however, his notes on Diamond Bluff provide no information on artifacts or village sites (Svec 1985). There can be little doubt that Lewis observed artifact debris while mapping the mounds, especially since much of the site area was cultivated. Apparently Lewis believed in a moundbuilder race unrelated to Native American Indians and, therefore, reasoned that the Indian artifacts found scattered among the mounds did not belong to the mound builders and should not be considered part of the survey data. According to Charles Keyes, "...there is nothing...to show that Lewis had any interest in fragmentary relics of any kind, ...his interest even in perfect relics...was rather moderate" (1928:103).
In 1903 J. V. Brower published a report of mound explorations in the upper Mississippi Valley that focused on the Red Wing Locality where, "at Diamond Bluff, an extraordinary mound region was encountered" (1903:47). Brower estimated that approximately 300 mounds-including the "panther" and "wolf" forms-were still visible on the terrace. Although an accompanying sketch map of the Red Wing Locality provides little detail, it does show two village sites associated with the Diamond Bluff terrace (Brower 1903). One is near the north end of the terrace, while the other is located at the south end off the edge of the terrace (Figure 12.2). Unfortunately, there is very little discussion in the text about the Diamond Bluff mounds, and there is no mention of the habitation sites.
Within a decade of Brower's survey, George Squier of Trempealeau, Wisconsin, visited the Diamond Bluff terrace (1914). Squier describes the Diamond Bluff mound group not only as having some of the largest mounds he had encountered, but the largest mound group recorded during his investigations (1914:138139). Squier's sketch map of the Diamond Bluff group depicts only 100 mounds on the north half of the terrace, of which approximately 60 were under cultivation (Figure 12.3). It is of specific interest that the map does not clearly show the "panther" effigy on the north end of the terrace. This is curious since the mound is in an area that was most likely in pasture. Squier did note that one of the mounds suggested an effigy, but its form "was doubtful for the outlines are by no means distinct" (1914:139). The fact that later investigators readily identified the "panther" effigy, suggests that Squier's visit to Diamond Bluff was brief and his map hastily prepared (Svec 1985).
Sometime around 1940, portions of the Diamond Bluff terrace were surface collected by members of the Minnesota Archaeological Society. An account of the survey notes the presence of a fairly large mound group at the north end of the terrace that includes an effigy (Lawshe 1947). Presumably, the effigy is the "panther" mound mapped by Lewis. Furthermore, Lawshe refers to two village sites on the north half of the terrace where over 2000 artifacts were collected. "Two spots seem to have been occupied on the site (terrace?), one at the northern end and the other near the mounds, the mound site being larger and apparently longer occupied" (Lawshe 1947: 77; parentheses and emphasis added). These artifact scatters may or may not represent two distinct occupation areas. The area identified as near the mounds would be the north village site depicted on Brower's map (Figure 12.2), and subsequently the focus of excavations by the Wisconsin Archaeological Survey and the University of Wisconsin-Milwaukee. It is possible that this village extends into the area identified as the northern end, although many of the artifacts could have just as easily come from mounds destroyed by cultivation.
The artifacts recovered by the collectors included simple notched and unnotched triangular points, scrapers, a variety of ground stone items, modified mussel shells, a copper ornament shaped like a mace, and a small engraved "short-nosed" god mask made of marine shell. Most of the pottery is tempered with crushed shell and displays a variety of curvilinear and rectilinear motifs (Lawshe 1947). A brief description identifies many of the pottery types as belonging to the "Oneota Aspect," although one of the illustrations is of a rolled rim that is labeled "resembles Cambria" (Lawshe 1947:74). Although no specific reference is provided, most likely Lawshe was comparing the sherd to "Cambria pottery type C" from the Cambria site in the lower Minnesota River valley, which Wilford suggested was Aztalan-Middle Mississippi influenced (1945:39).
By 1948, only 47 mounds remained intact on the north end of the Diamond Bluff terrace in an area used for pasture. Maxwell notes that remnants of mounds were still visible in the adjacent cultivated field (1950a). Among the extant mounds were 30 conicals, 16 ovals, and the "panther" effigy, with only four of the mounds having evidence of major disturbance (Maxwell 1950a, 1950b). Figure 12.4 illustrates 43 of the mounds and is derived from a base map of the WAS project area (see Maxwell 1950b:429). Six mounds were selected for excavation, and test pits were opened in the village area southeast of the existing mounds.
The WAS excavation took place during the summer of 1948 and included a field school under the supervision of Dr. Moreau Maxwell of Beloit College and Mr. Chandler Rowe from Lawrence College. Before discussing the excavations, some comment is necessary regarding the available information. The quality of the field notes varies considerably, not only collectively but also in regard to specific excavations. For example, the notes for Mound 4, and to a lesser extent Mound 26 (the "panther" effigy), are more detailed and informative than the notes for the other mound excavations and village area. Although some individuals recorded the presence of debitage, faunal remains, etc., their remarks are inconsistent, often brief, and lack detail. During the project, excavation fill was not screened nor were matrix samples collected. Therefore, information regarding the composition of mound fill, density of artifacts, kinds of artifacts, and faunal or floral remains is incomplete. By noting these shortcomings it is not my intent to be critical of the WAS project, its participants, or field procedures, but simply to indi-cate the limitations and problems with the data that are available. Overall, the field notes and photographs have been a valuable source of information that have provided insights on the WAS project.
With a base diameter of 50 feet and height of 7 feet, Mound 4 is the largest conical among the extant mounds (Maxwell 1950b:438). When Lewis measured the mound in 1887, the diameter was 44 feet and the height was 8.5 feet. Its size was second only to one other conical that had a wider diameter (Svec 1985). The discrepancy between Lewis' s measurements and those recorded by the WAS may be due to the methods of measurement. Over the span of 61 years, however, soil erosion and slumping- accelerated by grazing animals would tend to decrease mound height while increasing its width at the base.
Two intersecting 5-foot-wide trenches were cut through the center of Mound 4. Profiles of the mound show basket loads and soil boundaries described as seepage lines, and at the base an area of midden bordered by a mottled loam. Some of the loads contained concentrations of organic remains and artifacts described as village refuse. Field notes indicate that basket loading was more evident than what is depicted on the composite profiles of the east and west walls (Figure 12.5 ). There were also numerous seepage lines that are most likely the boundaries of basket loads modified by soil processes. There is no evidence of multiple building episodes. Instead, it appears that the mound was constructed during a single event, or over a short period of time.
Below the humus layer that covers the mound, most of the mound fill texture is described as sandy loam. The textural variation within the sandy loam class, along with the introduction of midden during construction, would create significant textural differences within the matrix of the mound. The development of seepage lines is probably related to these textural differences and the movement of water through the mound. The differences in soil texture will either facilitate or impede the movement of water. As water moves downward, contact with a finer textured soil would tend to redirect its movement laterally. Depending on soil textural differences, the effects of these processes through time may enhance or breakdown the boundaries between basket loads.
At the base of the mound are two areas described as mottled loam and midden that appear to be a remnant A soil horizon bordering a concentration of midden. A plan view of the Mound 4 excavation reveals that nearly 150 square feet of midden were exposed at the base of the mound (Figure 12.6). When the borders of the midden are connected, the estimated area increases to roughly 225 square feet. The nature of the midden is not clear; however, with a village site adjacent to the mounds and basket loads containing village refuse, it is not unreasonable to assume that some of the mounds were constructed over a preexisting habitation area. By extrapolating and connecting the limits of the midden, the shape and orientation of the projected area corresponds closely with the 200 square foot house feature excavated by UWM in the nearby habitation area (Alex 1974; Van Dyke 1983; and below). This, however, remains speculative since there is no evidence of submound features suggesting storage pits, hearths, or postholes, nor are there any specific details regarding the recovery of artifacts from either the midden or the mottled loam.
The human remains uncovered in Mound 4 (Figure 12.7) are believed to represent five adults and one juvenile (Maxwell 1950a, 1950b). Poor preservation made it difficult or impossible to determine sex or disposition of the burials. The only identifiable elements were fragments of larger bones, portions of the skull, and teeth. Since there are no indications in the profiles of an intrusive pit, it can only be concluded that the interments took place during construction of the mound. What appeared to be one adult in an extended position was encountered in units 89 and 90 L3 approximately 3 feet below the surface. The remains consisted of two femurs, a portion of the cranium and molars. Three interments representing two adults and one juvenile were in the north trench. In unit 93 L3, right and left humeri, a clavicle fragment, along with maxilla and mandible fragments were uncovered approximately 4.5 feet below the mound surface. In close proximity were a humerus and teeth identified as belonging to a juvenile. In unit 92 L3, a femur was uncovered in the fill close to the mound surface. The human remains found in the east trench were interpreted as two individuals. Approximately 2 feet below the mound surface in unit 91 L1, a celt, burnt bone, and teeth were uncovered, while in unit 91 CA two unidentifiable bone fragments believed to be human were found in the fill approximately one foot below the mound surface.
Artifacts were found throughout the mound fill, although little information is available on proveniences. With the exception of the celt noted above, all the other artifacts were simply part of the general mound fill. In addition to the celt, the other lithics include a stemmed projectile point and a small triangular point. Among the 245 fragments of pottery were 188 grit-tempered and 57 shell-tempered sherds. This included 16 rim sherds representing four vessels. Two of the shell-tempered vessels and one grit-tempered vessel appear to be local variants of Ramey Incised. Both the shell-tempered jars have bolstered lips, relatively straight necks, and angular shoulders (Figure 12.8: E). The grit-tempered jar has a rolled or bolstered lip, a straight neck, and angular shoulder with a scrolled motif (Figure 12.8: F). The fourth vessel is an aberrant grit tempered jar that has vertical tool incisions on the interior of the rim and a combination of punctates, diagonal and horizontal incised trails on the rim exterior (Maxwell 1950b; also see Hall 1962: plate 41, R & S).
When Lewis mapped this mound in 1887, it was 60 feet long, 38 feet wide, and had a height of 5.5 feet (Svec 1985). By 1948 over half of the mound had been destroyed and the remaining portion was pitted. The remnant measured 28 feet in length, 33 feet in width, and did not exceed 2.5 feet in height. The texture of the mound fill ranged from sand to sandy loam. There was also evidence of charcoal in some of the basket loads used to construct the mound. Among the artifacts recovered in the fill were two small, shell-tempered sherds with smoothed surfaces, two small, cordmarked, grit-tempered sherds, and a small cube of galena (Maxwell 1950b:3). No human remains were found in the mound nor was there evidence of prehistoric disturbance after construction of the mound.
Mound 8 is described as a low conical with a diameter of 30 feet and height of 2 feet. These dimensions correspond closely with the 28 foot diameter and 2 foot height recorded by Lewis (Svec 1985). Although no evidence of human remains or diagnostic artifacts were found in the mound (Maxwell 1950a:440, 1950b:3), field notes indicate some of the basket loads contained concentrations of organic material. In particular one reference notes an abundance of charred material that appeared to be corn stalks and wood. There is also a reference to lithic debris in the mound fill.
In 1887, Mound 15 had a length of 62 feet, a width of 25 feet, and height of 4.5 feet (Svec 1985). The WAS recorded a length of 66 feet, a maximum width of 40 feet, and a height was slightly over 4 feet. On top of the mound and near the center was a pit roughly 8 feet in diameter and approximately 2 feet deep. Two small, grit-tempered sherds and bone fragments believed to be human were found in the mound fill. Information about the mound fill is limited to brief notations about pockets of clay that were periodically encountered above a relatively uniform clay layer. The latter bordered a small primary mound that was encountered at approximately 2.5 feet below the surface. The primary mound was only partially exposed and excavated during which a human premolar was found. The dimensions of the primary mound were estimated to be approximately 4 feet in diameter and 2 feet in height. Beneath the primary mound was a rectangular-shaped feature measuring 2 x 3 feet with a depth of 0.75 feet. No cultural remains were recovered from this feature.
When Lewis mapped the "panther" effigy in 1887, it had a length of 180 feet and body width of 29 feet from the hump on the back to the underside (Svec 1985:75, table 2). These dimensions correspond closely with a length of 175 feet and body width of 34 feet recorded by the WAS. The mound was excavated by two intersecting trenches cut through the upper body and head of the effigy (Figure 12.9). Notes provide few details on the mound fill. It is assumed that the general composition of the fill was similar to that found in the other excavated mounds. In one of the profiles a dark loam is identified just above the subsoil horizon, suggesting the presence of a buried A soil horizon similar to that in Mound 4. Also, midden is noted as being present in the mound fill. Among the artifacts reported mixed in the fill were a triangular projectile point, some grit-tempered sherds, and two sherds with angular shoulders from which the temper (shell?) had been leached out. Near the base of the mound at approximately the original ground surface, shell-tempered pottery sherds from an angular shouldered jar were found (Maxwell 1950a 441).
Approximately 2 feet below the mound surface, unidentified burnt material, a shell tempered jar, and human remains were encountered (Figure 12.9 ). The fragmentary cranium, maxilla, and mandible were identified as belonging to a juvenile. It appears that the burial and jar were interred during mound construction. The excavators found no evidence of disturbance or an intrusive pit that could account for the burial after the mound was completed. The shell-tempered vessel is described as having a scalloped lip and a slightly everted rim. On the angular shoulder are a series of lobes, and above each lobe are six incised concentric loops. Below the shoulder the cordmarked surface is smoothed over (Maxwell 1950b).
Also in the north trench- the exact proveniences are not provided- two more discoveries were made. A second vessel was recovered in darkly stained soil at a depth of approximately 1.5 feet below the mound surface. It is described as a small, globular-shaped, shell-tempered jar with a plain surface and loop handles (Maxwell 1950a:441, 1950b). Toward the center of the trench and at the pre-mound ground surface, a submoundpit with an oval shaped orifice was uncovered. Near the top of the feature was a shell-tempered sherd with a bolstered or rolled-like lip, a straight neck, and a Ramey-like motif on the shoulder (Figure 12.8: A). Within the pit were the skeletal remains of an adult female and a child (Maxwell 1950a:442, 1950b). In the mound fill above the feature there was no evidence of an intrusive pit through the mound leading to the feature. Therefore, prehistoric excavation of the pit and introduction of the shell-tempered sherd apparently took place sometime prior to construction of the mound. Although submound burial pits have been found beneath effigy mounds (e.g., McKern 1928), it is also possible that the pit was a pre-mound feature associated with the village area.
Number 38 is described as a large oval mound measuring 75 feet long, 50 feet wide, and 7.5 feet high. These dimensions compare favorably with the measurements of 80 x 48 x 7 feet, recorded by Lewis in 1887 Svec 1985). Excavation uncovered human skeletal remains in the upper levels of the mound in three locations. Bone preservation was poor, and it could not be determine if three or four individuals had been found. The fragmentary remains were identified as adults and consisted of some teeth, along with mandible, skull, femur, tibia, and scapula fragments. One of the interments appeared to have been in a semiflexed position (Maxwell 1950a:3). There were no pottery vessels or tools associated with the human remains, although field notes indicate that a small piece of mica was found with one of the burials. The notes also indicate evidence of basket loading the presence of deer bone and charcoal flecks in the fill, and that the sandy loam mound fill was darker than that found in the other excavated mounds. Artifacts recovered in the mound fill included two grit-tempered and two shell-tempered sherds and a small triangular point.
In addition to the mound excavations, the WAS conducted some excavation in an adjacent village area. This is the same village site noted on Brower's 1903 map (Figure 12.2) and encountered by Lawshe (1947). Maxwell reports that two Oneota village sites were located near the extant mounds (1950a:428, 430); however, the notation of two sites served more to distinguish surface finds in the plowed area from the artifacts recovered in the test pit excavations near the edge of the terrace. The field notes have no information on the surface collections in the plowed field other than a listing of some of the artifacts. The walkovers in the cultivated area were conducted when rain prevented work on the excavations (Maxwell, personal communication). Among the artifacts listed are a number of triangular-shaped projectile points and shell tempered pottery sherds.
Excavation in the village area was initiated when midden material was discovered eroding from a ravine at the edge of the terrace approximately 640 feet southeast of Mound 6 (Figure 12.4). Six 5 x 5 foot test units were opened near the ravine in an area that had not been disturbed by plowing. Initially, the units were excavated in arbitrary 6-inch levels. This was soon changed to 3-inch intervals when it became apparent that strata could not be clearly defined below the humus layer. Three features were recorded (Figure 12.10). All are described as having straight sides and rounded bottoms and containing village refuse.
The diameter of feature 1 was approximately 3 feet, with a depth of 2 feet. The excavation recovered several shell-tempered sherds and a triangular projectile point. Feature 2, along the edge of the bank, was represented by two overlapping pits that measured 7.5 feet in length with a maximum width of 5 feet No information is available on depth. The feature(s) contained a large milling stone, shell tempered pottery, charcoal fragments, along with mollusca, fish, and faunal remains. Three large bones are not described in detail, but notes indicate that they appeared to be bison remains. Feature 3 was approximately 3 feet in diameter with a depth of over 1.5 feet. It contained faunal remains and shell-tempered and grit-tempered pottery.
Other artifacts recovered in the village area included "thumbnail" scrapers, abrading stones, and shell-tempered pottery (Figure 12.8: B. C, D). Maxwell reports that about one year after the WAS project a gravel removal operation near the village excavation area exposed more features. Mixed with Oneota pottery were, "...shell- tempered, angular shouldered rim sherds with horizontal rims, rolled lips, and scrolls and curvilinear elements..." indicating the village also had a Middle Mississippian-related component (Maxwell 1950b:4).
In summarizing the WAS excavations, Maxwell (1950a, 1950b) recognizes three components at Diamond Bluff: Late Woodland Effigy Mound, Middle Mississippian, and Oneota. Regarding the specific nature of these relationships Maxwell concludes, "...that all of the mounds sampled were constructed after the arrival of elements of Mississippian culture" (1950b:4). Specifically, the effigy mound was built by Oneota or peoples in direct contact with Middle Mississippian after the introduction of Middle Mississippian traits (1950b). Although both Oneota and Middle Mississippian traits were found in the village area, it was unclear if the two were contemporary or if Oneota developed out of Middle Mississippian (Maxwell 1950a: 443).
In 1974 the University of Wisconsin at Milwaukee conducted an excavation on the Diamond Bluff terrace under the direction of the late Dr. Robert Alex (Alex 1974; Van Dyke 1983). The UWM excavation was located near the edge of the terrace approximately 130 feet west of the WAS village excavations. Attempts to relocate the WAS test pits were unsuccessful. It is possible that the WAS excavation area no longer exists because the quarrying operation noted above may have destroyed that portion of the site.
The UWM crew opened 138.75 square meters (1497.4 sq. ft.) of the site area by excavating n 10 centimeter levels (Figure 12.11). Over 40 features were uncovered some that extended to depths of over 1.5 meters along with a semi-subterranean house and a problematic house (Alex 1974). The floor of the semisubterranean house was encountered approximately 50 centimeters (20 inches) below the surface. Its semi-rectangular shape measured approximately 5 meters by 3.75 meters, encompassing an area of approximately 18.75 square meters (202 sq. ft.). Although bordered by several post holes, there was no evidence of wall-trench construction. The entrance was not excavated, but is believed to have been at the north end. Associated with the structure were three pit features and a hearth. The problematic structure was in an area identified by numerous post holes and midden. The exposed midden covered approximately 4 square meters (43 sq. ft.) and was bordered on the west side by post holes that extend a distance of nearly 6 meters (19.7 ft.). The time constraints of a field school prevented completion of the excavation in this area (Alex 1974).
The lithic artifacts include a variety of cherts, silicified sandstones, and some Knife River Flint. The most common tools types are unnotched and side-notched arrowheads, "thumbnail" or humpback scrapers, sandstone abraders, hammerstone cobbles, and a variety of bone tools. The pottery assemblage is dominated by shell tempered wares that include Oneota types and Middle Mississippian-inspired Silvernale wares (Figure 12.12). Grit-tempered pottery comprises only a small fraction of the assemblage.
Although an extensive analysis of faunal remains has yet to be conducted, elements from a variety of terrestrial mammals have been identified. They include bison, beaver, elk, squirrel, muskrat, white-tail deer, raccoon, rabbit, black bear, gray and red fox, badger, skunk, wolf, and dog (Van Dyke 1983). Examination of a small sample of fish remains revealed several species known to inhabit both the main channel and backwaters of the Mississippi flowage (McInnis 1977). An analysis of maize from three excavation units identified North American Popcorn, Midwest 12-row, and Eastern 8-row. The samples also contained wild plum seeds (Cutler and Blake 1976).
Human remains represented by several cranium fragments, complete and fragmented mandibles, and loose teeth were recovered from midden filled pit features. There are three mandibles with cut marks on the posterior aspect of the ramus and two mandibles that display cut marks on the anterior of the horizontal ramus. A number of parallel cut marks are also visible on cranium fragments of occipital and parietal bones. Among the several individual teeth are three canines, one incisor, and one premolar that are notched near the tip of the root suggesting they were worn as pendants on a necklace (Rodell n.d.).
Five radiocarbon dates are available from the UWM excavation area (Table 12.1). Three of the radiocarbon dates were obtained from wood charcoal recovered in feature 42, a pit in the floor of the semi-subterranian house. The two radiocarbon dates from feature 1 were also derived from wood charcoal (Figure 12.11). Both features were filled with midden debris. The uncorrected dates (Bender, et al. 1978) have been recalibrated by the method proposed by Stuiver (1982; also Stuiver and Reimer 1986). Recalibration of conventional dates adjusts for past fluctuations in atmospheric C14 and is based on dendrochronological determinations.
The dendro-age dates indicate that in the millennium before A.D. 1250 conventional radiocarbon ages err toward being too old (Stuiver 1982:5). The conversion of an uncorrected date may intercept the dendro-age calibration curve more than once resulting in more than one age per corrected date as is evident with three of the dates in Table 12.1. In these cases all of the corrected ages must be considered (Stuiver and Reimer 1986).
Within a decade of the UWM excavation, the Diamond Bluff terrace was subject to more archaeological investigation. In 1980 and 1981 the State Historical Society of Wisconsin (SHSW) surveyed the north half of the terrace in the site area listed as 47-Pi2 (Penman 1981:12-13, 1984:35-36). The SHSW collections include lithic debris and tools, along with shell-tempered and grit-tempered pottery. Shell-tempered pottery was recovered in the general area where mounds had existed (Penman 1984:36). This suggests that the village excavated by the WAS and UWM extended several hundred feet eastward from its confirmed location along the west side of the terrace edge, or that the mounds had contained shell-tempered pottery either in the general fill, as intentionally buried vessels, or both.
Following the SHSW investigations, the Institute for Minnesota Archaeology (IMA) conducted a series of surface collections over the entire terrace (Wends and Dobbs 1989). The IMA survey relocated 192 of the 396 mounds mapped by Lewis and identified four concentrations of artifact scatters that are viewed as being habitation settings. Two of the sites are listed as Woodland occupations. They are site 47-Pi-93, located at the southern tip of the terrace within an area that had a dense concentration of mounds, and Mero No. 3 situated in the center of the terrace between two clusters of mounds. The sites designated Mero No. 1 and Mero No. 2 are bordered by mounds and are identified as Silvernale villages.
The IMA surveys (Wends and Dobbs 1989) have provided some insights on the spatial relationships of habitation areas and mounds on the Diamond Bluff terrace (Figure 12.13). The two Woodland site areas 47-Pi-93 and Mero No. 3 are represented by thin scatters of lithic debris and cord marked, grit-tempered pottery. Mero No. 3 covers roughly 1.7 acres (0.7 hectares) and is in the approximate center of the terrace in an area where apparently no mounds had been built. This site may represent a small habitation site or an activity area associated with the other site areas. Site 47-Pi-93 is at the south end of the terrace, covers roughly 3.5 acres (1.4 hectares), and is where several mounds existed. While 47-Pi-93 may represent an activity area or habitation setting, it is equally likely that this artifact scatter may represent the remains of mound fill and burial offerings strewn about by plowing.
The artifact scatters identified as site areas Mero No. 1 (47-Pi-2) and Mero No. 2 are located on the west side of the terrace. In comparison to the Woodland sites, Mero No. 1 and No. 2 have greater densities of artifacts over larger areas, and the assemblages are represented by mixtures of Woodland, Silvernale, and Oneota artifact types. Both sites are estimated to cover about 17 acres (6.9 hectares). Wendt and Dobbs (1989) note the limits of Mero No. 1 and No. 2 are marked by distinct declines in artifact densities, and that these boundaries appear to correspond with the mounds that once bordered the respective site areas on three sides. There is, however, good reason to suspect that some mounds were constructed over a previously existing Oneota habitation area, given the findings of the WAS mound excavations near the site area 47-Pi-2 (Mero No. 1). As presented above, at the base of Mound 4 (Figure 12.6), a concentration of debris described as village midden was uncovered, and at the base of Mound 26 (effigy) the submound pit containing the human remains can arguably be interpreted as a village pit burial. Pit burials in habitation settings are reported at the Bryan site (Gibbon 1979: 9-10) and are known at other Oneota sites (Hall 1962; Kreisa 1987; Mc Kern 1945). Whether or not the pattern of the Mero No. 1 and Mero No. 2 habitation areas bordered by mounds began during the Woodland period, or is indicative of the Silvernale period occupation, or is the result of later Oneota settlement will only be resolved by continued investigation.
2Stuiver and Reimer 1986.
With the kinds of data presently available from the Diamond Bluff site complex, and other sites in the Red Wing Locality (Dobbs 1984; Gibbon 1979), what can be said regarding both the nature of Middle Mississippian contact and the identity of the indigenous Red Wing population at the time of contact? In Oneota studies there has been a long standing debate revolving around the origins of Oneota culture and its relationship to Middle Mississippian culture (e.g., Griffin 1960; Gibbon 1972, 1974; Stoltman 1983, 1986).
One position favors the emergence of the Oneota culture in situ and prior to expansion of Cahokian influence into the northern Mississippi Drainage around circa A.D. 1050 (Gibbon 1982). Supporters of this position point out that, ". . . radiocarbon dating revealed that Oneota arose as a coherent cultural complex too early to be a 'simple' offshoot of more complex, Mississippian type cultures to the south, mainly . . Cahokia" (Brown 1982:107). While there is agreement here that Oneota culture cannot be viewed simply as a by-product of Middle Mississippian culture, the statement that "Oneota arose as a coherent cultural complex" prior to the spread of Middle Mississippian influence in the 11th century is misleading. In the upper Mississippi Valley, many of the traits that would become the hallmarks of the Oneota cultural complex large village sites, shell-tempered globular jars with chevron motifs, triangular points, humpback scrapers, catlinite pipes and tablets, sandstone abraders, a riverine adaptation, bison hunting, maize horticulture, etc. do not appear collectively as cultural units before the time of the Cahokian trade network. Furthermore, it has been argued that Oneota symbolism, displayed on pottery, in copper ornamentation, etc., was greatly influenced by Middle Mississippian (Been 1989; Hall 1991).
The alternative position recognizes the emergence of Oneota culture during the 11th century A.D. as a result of Middle Mississippian interaction with Late Woodland peoples (Stoltman 1986). This would appear to be a more tenable position, especially in light of the empirical data and speculations on Oneota symbolism noted above. Gibbon (1972, 1982), however, has stressed the point that the emergence of Oneota lifeways involved a variety of processes that took place over a period of time. These processes culminated with Middle Mississippian contact during the eleventh century. There is, unfortunately, a poor understanding of these processes. And while there can be little doubt that the roots of the Oneota cultural tradition extend back to Late Woodland, the kinds of data relevant to this problem have remained elusive, resulting in interpretations based primarily on inferential evidence and speculation (Gibbon 1982:86). The Red Wing Locality is no exception to this problem. With the data presently available from Diamond Bluff and related sites in the area, the evidence for a pre-Silvernale Oneota occupation is based primarily on conjecture.
The argument for a pre-Silvernale Phase Oneota population in the Red Wing Locality relies upon radiocarbon dates that predate A.D. 1050 (Gibbon 1974:130, 1991). What has not been made explicit however, is if the early dates came from contexts clearly associated with early Oneota/pre-Silvernale remains, or if the dates were from mixed contexts containing Silvernale/Oneota remains. This is clearly a problem with the radiocarbon dates from Diamond Bluff (Table 12.1). The Diamond Bluff radiocarbon dates were derived from samples of wood charcoal that had been recovered from arbitrary 10 centimeter levels. Although it is not my intention to be critical of the UWM field methods, the excavation of features by arbitrary levels was not adequate enough to document the complex strata of the midden deposits, which presumedly were the result of both cultural and natural episodes of filling and mixing. The three dates from Feature 42 came from a 40 cm vertical segment. The two extreme dates A.D. 1001 (1031,1144,1446) 1157 and A.D. 1222 (1263,1273, 1275) 1281 are so far apart that their one sigma ranges do not overlap (at the two sigma range there is a 48 year overlap). Furthermore, their location in the stratigraphic sequence is in reverse order to their chronometric relationship. The same situation also exists for the two dates recovered from Feature 1, however, the one sigma ranges are in close harmony. In addition, a nearly complete Silvernale vessel came from the level with the date of A.D. 1033 (1158) 1215. Although the five radiocarbon dates from the Diamond Bluff village area (47-Pi-2) correspond to the time of Middle Mississippian influence and Oneota presence in the Red Wing Locality, it remains equivocal if the radiocarbon ages can clearly be assigned to specific components.
At Diamond Bluff there is evidence that Woodland groups periodically utilized the terrace from late Middle Woodland times through the Late Woodland period. Woodland culture at Diamond Bluff is most evident by the conical, linear, and effigy mounds that are characteristic of Late Woodland mound types in the upper Mississippi Valley. A Woodland presence is also recognized by other artifacts, most notably pottery. Although investigations indicate that the amount of Woodland pottery is small in proportion to the number of shell temper wares, cordmarked, grit-tempered pottery has been recovered in both mound and village contexts. According to Lawshe only a few of the pottery sherds collected by his party were tempered with sand or crushed stone, and had designs made by, "grass, beads and possibly ears of corn... (and) cord-wrapped paddle" (1947:83). Most of the pottery recovered by the WAS is described as being grit- tempered with nearly equal numbers of sherds having cordmarked or smoothed surfaces (Maxwell 1950b). From a total of 107 rims, 77 (72%) are grit-tempered and 30 (28%) are shell-tempered. None of the grit-tempered, cordmarked rims are attributed to specific types. The IMA survey identified concentrations of Woodland pottery on the Diamond Bluff terrace, although the number of sherds is small in comparison to the amount of shell-tempered pottery. Some of the grit-tempered pottery is identified as Late Woodland Madison Cord Impressed (Wends and Dobbs 1989). The Woodland pottery recovered by the UWM excavation represents only a small fraction of the ceramic assemblage. Although many of the sherds are simply grit tempered and cordmarked fragments, there are a few identifiable types. They include Angelo Punctate, a late Middle Woodland/ early Late Woodland type (Hurley 1974a) and two Late Woodland wares: Madison Cord Impressed, and Clam River ware (McKern 1963).
Madison Cord Impressed is a relatively common pottery type associated with Late Woodland mound sites (e.g., Hurley 1974b). The presence of Clam River ware, which has a more restricted distribution, is neither spatially nor temporally out of place at Diamond Bluff. Although this pottery type is best known among Late Woodland sites in the region bordering the St. Croix River (Caine 1974; Gibbon and Caine 1980:63; McKern 1963; Van Dyke and Oerichbauer 1988), approximately 100 miles north of the Red Wing Locality, Clam River pottery has been reported at the Plum Creek site (47-Pe-38) 25 miles east of Diamond Bluff (Ford 1982). Radiocarbon dates place the Clam River tradition during the Late Woodland period, ca A.D. 400 to 1100 (Gibbon and Caine 1980; Van Dyke and Oerichbauer 1988).
Among the lithic artifacts recovered from the Diamond Bluff terrace (Lawshe 1947; Van Dyke 1983; Wendt and Dobbs 1989), there are only a few tools that have been specifically identified as Woodland. Lawshe illustrates a variety of chipped stone and ground stone tools that may be Woodland artifacts (1947:76, 80, 82), and Maxwell reports that a projectile point recovered from Mound 26 was Woodland (1950a:442). In the UWM lithic assemblage there are some point types, as well as a significant amount of chert debitage, that could be of Late Woodland origin. The findings of the IMA survey indicate that Woodland groups may have relied more heavily on cherts than the later Oneota component (Wendt and Dobbs 1989).
In the Red Wing Locality investigators have reported that there were from 2000 to as many as 4000 mounds (Brower 1903; Schmidt 1941). Like Diamond Bluff, neighboring mound groups contained a variety of mound forms (cf. Winchell 1911). The Red Wing Locality was on the northwestern periphery of the effigy mound region, while to the north there was the Clam River mound tradition, and to the west were other contemporary mound building groups (Anfinson 1984; Gibbon and Caine 1980; Mallam 1976; McKern 1963). Given the generally accepted view that Woodland mound sites were the focus of annual gatherings where people conducted ceremonial, social and economic activities (Green 1986; Mallam 1976), it is suggested that the Red Wing Locality served this function for a variety of regional cultural traditions. The reason for the focus on Red Wing is not entirely clear. An important consideration, however, is its geographic location in relation to a number of major rivers and streams that converge in the area. In addition to the Mississippi River, the Cannon River provides a route to the Minnesota River Valley and access to west-central Minnesota (Wilford 1955:139). There is easy passage to Lake Superior, via the St. Croix and Brule Rivers, and there are major streams leading into the interiors of northern Minnesota and northern Wisconsin. It would appear then that the Red Wing Locality was a natural hub through which regional trade and information was channeled (Rodell 1989).
Middle Mississippian interest in the upper Mississippian Valley may have been initiated during the period A. D. 950-1050 (Kelly 1991). While no concrete proof for this exists in the Red Wing Locality, 70 miles downstream at Trempealeau, Wisconsin, there is some evidence, albeit tenuous at this time, of early Middle Mississippian influence. At Trempealeau there is a terraced, pyramidal, flat-topped mound and, from the unpublished notes of George Squier (Stevenson, et al. 1983), descriptions of red slipped pottery that are akin to pottery types associated with late Emergent Mississippian and early Mississippian phases at Cahokia. The finds at the Stull site (47-Tr-159), and the recent test excavations at the Squier Garden site (47-Tr-56) conducted by the Mississippi Valley Archaeology Center, have recovered small samples of red slipped rim sherds. Notwithstanding the speculative nature of the available data, it appears that the Trempealeau locality represents a Middle Mississippian presence in the upper Mississippian Valley that predates the Silvernale Phase in Red Wing (Green 1988; Squier 1905; Stevenson, et al. 1983). If, as suggested above, Red Wing was already the focus of regional trade and social interaction, the pull northward to the Red Wing Locality would have required little incentive on the part of the Cahokian trade network.
Mound building on the Diamond Bluff terrace was not limited to pure Woodland groups but continued after the introduction of Mississippian-inspired traits. As revealed by the WAS excavations some of the mounds were constructed after the introduction of Middle Mississippian symbolism, most notably the Ramey motif displayed on sherds recovered in mound fill and on a buried jar. In the American Bottom, the Ramey Incised motif is most closely identified with the Stirling Phase, circa A. D. 1050-1150 (Bareis and Porter 1984; Fowler and Hall 1974; Holley 1988). Emerson has presented a persuasive argument regarding the function of the Ramey Incised vessel as "utilitarian symbolic ware" (1989:67). Ramey Incised vessels were used in a variety of social functions, including rituals to "... reaffirm intragroup relations... (and)... integrate outsiders, traders, allies, and such into the social system" (Emerson 1989:67). If indeed Ramey Incised vessels were used in such observances, not only within the American Bottom but at other Midwest localities interacting with Cahokians, Emerson's thesis has implications for interpreting Mississippian in the Red Wing Locality.
In the Red Wing Locality a significant proportion of ceramic assemblages are dominated by Silvernale wares. In following the logic of Emerson's thesis (1989), Ramey Incised vessels were functionally part of everyday use-not just among a select group people but throughout the greater Red Wing community while symbolically the Ramey motif linked the Red Wing communities to the Cahokian "symbiotic-extractive exchange system" (Gibbon 1974:133; 1982:89). The Ramey Incised motif on locally manufactured ceramic vessels is by far the most common of the Middle Mississippian-inspired traits in the Red Wing Locality and is a hallmark of Silvernale ware (Anfinson 1979:183- 190; Wilford 1955). Among the Silvernale wares, the Ramey motif occurs on the classic Ramey-Powell jar form as well as on the basic globular Oneota vessel form. Of particular interest are the 77 grit-tempered rim sherds recovered by the WAS, 40 of which are described as having "Old Village" designs. Although this sample is small, it raises the possibility that the Ramey motif was adopted before other Mississippian ceramic traits such as shell tempering.
The context in which Ramey Incised sherds have been found in the Red Wing Locality is almost exclusively midden debris. A notable exception is the effigy mound (No. 26) at Diamond Bluff. As described above, two relatively complete shell-tempered vessels were recovered from this mound. The first vessel was a small undecorated jar with a smoothed surface and loop handles. Although plain wares are rare at Diamond Bluff, as well as in the general Red Wing Locality (cf. Gibbon 1979), it is not unusual to find them in early Oneota ceramic assemblages in other regions (cf. Hall 1962). The second jar had a scalloped lip, and incised concentric loops above a series of lobes on an angular shoulder. The motif of incised concentric loops is in the style of Ramey Incised, and is found on other vessels from Diamond Bluff and nearby sites (cf. Gibbon 1979). The lobed body effect is also found on a minority of vessels from the nearby Bryan site (Stortroen 1957: 41-42, Plate II; Gibbon 1979: Plate S18), as well as at Cahokia (Holley 1988:270, Figure 48-D). Although few in number, the lobed bodied vessels fit nicely within the local pottery sequence of Silvernale wares.
Also, from the Red Wing Locality there are a small number of artifacts relating to the Southeastern Ceremonial Complex. The Diamond Bluff site complex has revealed a limited variety of Southeastern Ceremonial Complex traits including a marine shell "god mask," the baton or mace-shaped motif depicted in copper ornamentation and incised on pottery (Figure 12.14), and chunky stones. A shell-tempered sherd with the mace motif was excavated by UWM from the floor of the semi-subterranean house. The shell mask, chunky stones, and copper artifact were recovered from the surface (Lawshe 1947), so it is not known if these items were from plowed down mounds, a village area, or some other context. At the Bryan site, small copper mace pendants (Gibbon 1979) and falcon symbolism, depicted by a bird motif or both a ceramic jar (Link 1982:34) and a bone pendant (Stortroen 1957: plate VII, figure B-4), have been found. Stylized falcons can also be derived from the feathered-Ramey scroll motifs on Silvernale wares (cf. Benn 1989; Hall 1991). Other artifacts found in the Red Wing Locality that are possibly related to the Southeastern Ceremonial Complex include small stones with engraved faces and a small pottery head and an ear spool (Gibbon 1979, 1991).
Like the Ramey motif, the artifacts identified with Southeastern Ceremonial Complex have ties to the theme of interregional relationships, prestige, and sanctioned authority (Brown 1976; Emerson 1989). A number of writers have interpreted the "god masks" as being symbolic of traders or trade alliances (e.g Gibbon 1991; Griffin 1967; Hall 1990; Kelly 1991). Hall has even suggested that the "god masks" signified ". . . a ceremonial relationship between the participants through a fiction of kinship, and that the ceremony was specifically used to establish friendly relations between otherwise unrelated groups" (1991:88).
In addition to the Ramey motif and the ceremonial artifacts, there are other artifact types in the Red Wing Locality that may be Middle Mississippian-inspired. Among them are a pyramidal flat-topped mound, notched triangular arrowheads, and semi-subterranean house forms (Gibbon 1991). A pyramidal flattopped mound between the Bryan and Silvernale sites was mapped by T. H. Lewis in the late 1800s (cf. Schmidt 1941; Winchell 1911:153). For whatever reason(s), Wilford did not include this mound type in his definition of the Silvernale Focus (1955:139-140). Other researchers, however, believe the association to be more than a coincidence (Griffin 1960; Hall 1962; Gibbon 1974, 1979). In fact, recent excavations conducted by the IMA at the Energy Park site have identified a Silvernale Phase component at the locality of this mound (Dobbs and Breakey 1987).
It is not uncommon to find small, side-notched, triangular arrowheads among the Silvernale Phase sites. At Diamond Bluff these points are made of cherts, silicified sandstone, and Knife River Flint (Lawshe 1947; Van Dyke 1983). Whether or not this point type is derived from Middle Mississippian influence or is the result of Plains influence is not clear since it was common in both spheres of influence (Kelly 1991; Kehoe 1966). From the Silvernale and Bartron sites there are isolated examples of tri-notched points that are more closely aligned to some Cahokian point styles (Gibbon 1991). In addition to the semi-subterranean house floor uncovered at Diamond Bluff (Alex 1974; Van Dyke 1983), the remains of similar structures were excavated at the Bryan site (Gibbon 1979). Like the notched points, it is also not clear if semi-subterranean house forms in the Red Wing Locality are of Middle Mississippian or Plains influence.
While point types and house forms are suggestive of Plains influence, other artifact types leave little doubt that the Red Wing communities were in contact with Plains groups. In the Diamond Bluff lithic assemblage there is a small percentage of Knife River Flint, and pottery sherds identified as Mitchell Modified Lip (Alex 1974; Van Dyke 1983), a pottery type of the Middle Missouri Tradition. The pottery was recovered from the semi-subterranean house feature and appears to represent one vessel, although not enough sherds exist for a reconstruction. The sherds are tempered with a fine grit, have smooth surfaces, and thin trailed line decoration. The lone rim sherd is relatively short and everted outward and has a singular zigzag trailed line on the lip. Similar Plains and related Cambria pottery types have also come from the Bryan and Silvernale sites (Gibbon 1979,1991).
Found among the Red Wing sites are a variety of conventional Oneota artifacts (Gibbon 1979; Lawshe 1947; Van Dyke 1983). At Diamond Bluff, as well as other Red Wing sites, the style of the Oneota pottery is very similar to Blue Earth wares. The shell-tempered jars have medium to high everted rims, rounded shoulders, and chevron motifs (cf. Anfinson 1979:39-44; Gibbon 1979). In the UWM assemblage handles are rare although both the loop and strap types are present. Motifs illustrated by Lawshe (1947: 74, 87) include the target design and chevrons bordered by punctates. Similar motifs are also present in the UWM assemblage, although they are not as common as the Ramey motif that precedes adoption of the chevron (Hall 1962). Lip-rim decoration is rare. However, when present, it is most often found on the interior surface of the rim. The lithic assemblages include sandstone abraders along with numerous humpback scrapers and small unnotched triangular arrowheads made of cherts, silicified sandstone, and Knife River Flint. The lithic artifacts are not only characteristic of conventional Oneota assemblages, but more specifically they are indicative of the Silvernale Phase in the Red Wing Locality.
During the latter half of the 19th century the pioneering investigations on the Diamond Bluff terrace recorded a large and complex mound group comprised of a variety of mound types that are characteristic of Late Woodland culture (cf. Svec 1985, 1987). In the 1940s, two independent investigations found evidence of Oneota and Middle Mississippian-inspired artifacts on the terrace (Lawshe 1947; Maxwell 1950a, 1950b). In particular, the WAS project investigated both mound and village areas. Most revealing were the excavations in six mounds that clearly indicated some of the mounds were constructed after the indigenous population adopted the Ramey motif and shell tempering of ceramic jars. Subsequent studies include excavations in one of the village settings by UWM in 1974 (Alex 1974; Van Dyke 1983), Svec's reconstruction of the Diamond Bluff mound group (1985, 1987), and the surface collections by the SHSW (Penman 1981, 1984) and the IMA (Wendt and Dobbs 1989). The UWM excavation uncovered several midden-filled features and a semi-subterranean house floor. The five radiocarbon dates from these excavations fall within the period of the early 11th century to late 13th century A.D. The other projects have provided important information on the size of the mound group(s), artifact densities, and spatial relations of mounds and habitation areas on the terrace.
What remains poorly understood is the archaeological identity of the indigenous Red Wing population at the onset of the 11th century A.D. Late Woodland on the Diamond Bluff terrace is identified with the Effigy Mound and Clam River traditions. In addition to the mound types, Late Woodland is represented by a widely scattered but low density of artifacts on the terrace. Late Woodland utilization of the terrace was not continuous, but recurring over several centuries by relatively small groups that occupied the site on a seasonal basis for both social and economic reasons. In contrast, the major occupation of the Diamond Bluff site complex is represented by the Oneota cultural tradition.
While it is not known what attracted Middle Mississippian interests to the Red Wing Locality, it is suggested here that this area had been the focus of trade and interaction among regional Late Woodland groups. Initial attempts by Cahokia to establish a foothold in the upper Mississippi Valley may have taken place as early as A.D. 950-1000 at Trempealeau, Wisconsin. This, in turn, was followed by the establishment of trade relations with the Red Wing Locality in the 11th century A. D. (cf. Kelly 1991).
The evidence for a Middle Mississippian site-unit intrusion in the Red Wing Locality is, at best, equivocal. Archaeological investigations at Diamond Bluff and related sites have not revealed Middle Mississippian artifact assemblages comparable to Mississippian site assemblages to the south. Instead a Middle Mississippian "presence" is represented by a small fraction of artifact types, such as the marine shell "god mask," and Ramey symbolism. Presumedly, these were introduced to the local population by Middle Mississippian traders. A weak link in drawing any conclusions about Middle Mississippian in the Red Wing Locality is the absence of human skeletal data and burial assemblages that could be used for comparative study with other Mississippian, Late Woodland or Oneota populations. Although the WAS documented mound burial at Diamond Bluff, the kinds of information required for comparative study does not exist. Human skeletal remains recovered from the village area by UWM are limited to cranial fragments, mandibles, and teeth. The context of these bones and the presence of cutmarks on several of the fragments suggests the remains of trophies and not a resident burial population (Rodell n. d.). The study by Glenn (1974) of Late Prehistoric crania from the Red Wing Locality was limited to three individuals from the Bryan site. Although the sample was too small to have any statistical significance, Glenn suggested that the physical attributes of the crania resemble the Middle Mississippian type (1974:138).
The Silvernale Phase was a local response to the Cahokian trade network with the indigenous population emulating certain aspects of Middle Mississippian culture. Whether this population was Oneota (Gibbon 1974, 1979,1991) or Late Woodland (Stoltman 1986) is not entirely clear. The evidence from Diamond Bluff favors Late Woodland. For the duration of the Silvernale Phase, circa A.D. 1025 to 1200, the indigenous Red Wing population preserved much of its local autonomy. Initially only those traits of Mississippian culture needed to articulate with the Cahokian trade network were adopted. These artifacts were limited to items of the Southeastern Ceremonial Complex and Ramey symbolism, and they functioned to maintain the "magico-religious sanctions and guarantees" required to insure success of the Cahokian trade system (Gibbon 1974:136). As involvement with the Cahokian trade system declined or became modified, the Mississippian symbolism was modified to emphasize the regional character of cultural groups. Most notable being the falcon symbolism that Benn (1989) has argued became the unifying social identity of the Oneota cultural tradition (also see Hall 1991).
I would like to thank Dr. James Stoltman for inviting me to contribute to this volume. This chapter is a revised version of a paper presented in the symposium, "Pattern, Phase and Focus: a Symposium Honoring Will C. McKern," at the annual Midwest Archaeological Conference, October 16-18, 1987, Milwaukee, Wisconsin. I am grateful to Dr. Moreau Maxwell, and Elmer Denlinger, John Forde, Thomas Kehoe, Charles Jacobs, and Owen Miles. Through discussion and correspondence they provided information on the 1948 WAS excavation and graciously contributed field notes, maps, and photographs. Dr. Robert Salzer of Beloit College also provided information on the 1948 excavation. The University of Wisconsin-Milwaukee has provided access to artifacts, notes, and materials from the 1974 excavation.
Robert Boszhardt, James Stoltman and Allen Van Dyke provided comments on earlier drafts of this paper. Michael Kolb of the UWM Soils Laboratory patiently explained soil relationships and formation processes; and Ronda Ulrath drew the illustrations for Figures 12 and 14. I appreciate their efforts and contributions. Any errors or misinterpretations reflect my own shortcomings.
© 1999 The Institute for Minnesota Archaeology
Email us: firstname.lastname@example.org
Updated 29 Jun 1999