Pisgah Ceramic Production in the Southern Appalachians

1

Pisgah Ceramic Production in the Southern Appalachians

Ashley Schubert (University of Michigan) 71st Annual Meeting of the Southeastern Archaeological Conference

Greenville, South Carolina, November 2014 Abstract: The Pisgah culture of Western North Carolina (AD 1000-1500) has historically been characterized as a Mississippian phenomenon in part due to the addition of new decorative rim treatments on the ceramics, along with coeval changes in the built environment and subsistence practices. While certain aspects of ceramic production reflect continuity with earlier Woodland groups, the shift in decorative treatments points to extra-local interaction and influence. By considering intersite differences between assemblages recovered from the Cane River site (31Yc91), the Garden Creek site (31Hw1), and the Warren Wilson site (31Bn39), regional variation in participation with Mississippian social political relationships is better defined. Introduction to Pisgah

The process of Mississippianization across prehistoric communities in the Midwest and

Southeast has a long history of being defined through trait-list approaches. Initially, the

Mississippian expansion was defined as a configuration of specific and co-occurring features

such as wall-trench houses, substructure platform mounds, and shell-tempered pottery (Griffin

1985; Blitz 2010; Cobb 2003; Smith 1990). Although this political-economic shift started with a

rapid series of changes in the American Bottom ca. A.D. 1050 (Kelly 2000; Pauketat 1994), the

overall tempo and extent of Mississippianization across the Southeast is now regarded as a

heterogeneous transformation in which varied Mississippian cultural practices were integrated

within local social structures (Smith 1990; Ward and Davis 1999; Woodall 1999; Meyers 2002;

Cobb; Griffin 1992; Griffin 1943; Beck and Moore 2002). Similarly, for the Mississippian phase

of the prehistoric Cherokee occupation in the Appalachian Summit of North Carolina, called the

Pisgah phase (AD 1000-1600), some new behavioral traits emerge and transform the landscape

from the previously smaller, more generalized site settlement pattern of the Woodland period.

How Mississippian practices were selected and incorporated into local traditions across

the Appalachian Summit is still largely unknown. In order to better understand the variation in

2

Mississippianization, my research considers the material correlates of household and community

practices across three major Pisgah sites during the later end of the Pisgah period (AD 1200-

1600) – Garden Creek, Warren Wilson, and Cane River, all located in central Appalachian of

North Carolina (Figure 1).

The Warren Wilson, Garden Creek & Cane River sites

Research in the 1960s, with the Cherokee Archaeological Project, initially defined the

Woodland and Mississippian periods in the Appalachian Summit based off two sites with long

occupational histories: the Garden Creek site and the Warren Wilson site (Ward and Davis 1999;

Dickens 1976; Keel 1976). The Pisgah period was categorized as South Appalachian

Mississippian based on the presence of the “diagnostic duo,” – platform mounds and rectilinear

complicated stamped pottery- traits that co-occur together for the first time during the Pisgah

phase (Keel 1975:14). (More specifically, at the Garden Creek site). Another diagnostic feature

of the Pisgah culture is the collared, or thickened, rims (and their associated decorations), which

have no antecedent in the Appalachian Summit, leading researchers to suggest it was introduced

“along a broad frontier of expanding Mississippian culture” (Dickens 1976:200).

Garden Creek. The large mound and village site of Garden Creek (31Hw1) is located on

the floodplain of the Pigeon River in Haywood County. There are three mounds and several

village middens spanning over a 12 acre area (Dickens 1976; Wright 2014). The Pisgah

component of the Garden Creek occupation covers about 5 acres and consists of at least one

mound with an associated village area and midden (Dickens 1976). The Pisgah mound (Hw1) is

the largest at Garden Creek and consists of at least two construction stages and floors (Dickens

1976; Cobb). The mound served as a ceremonial substructure and consisted of two paired earth

lodges covered by a rock mantle surface, and subsequently more floor surfaces that were the base

3

for later ceremonial buildings and a log palisade. This initial capping of important public

buildings with earthen mantles, which in turn became the foundations of successive episodes of

alternating layers of structures and soil, is a widespread Mississippian earthwork building

practice adopted across multiple South Appalachian Mississippian groups (Cobb). Beneath and

adjacent to the mound was a village surface containing a palisade with a bastion, a village

midden, and two probable house patterns (Dickens 1976). The Garden Creek Mound No. 1 and

its associated village component provide a comparative sample of ceremonial and domestic

activity areas with clear evidence of Mississippian cultural practices.

Warren Wilson. The Warren Wilson site (31Bn29), in Buncombe County, North

Carolina, was defined by Dickens (1976:18) as “characteristic of a medium-sized Pisgah village”

that contains solely domestic aspects of the Pisgah culture. Located on an alluvial terrace of the

upper Swannanoa River, cultural debris covers approximately 3 acres; the Pisgah component

consists of multiple palisade lines surrounding a circular village with a central plaza (Dickens

1976; Moore 2002). In total, at least seventeen domestic structures and seven palisade lines have

been identified, along with associated burials and features, in a relatively stable village layout

similar to that of other Mississippian hamlets and farmsteads (Moore 2002). Warren Wilson

represents one of the best understood (and published on) Pisgah villages in terms of its site

structure and assemblages (Moore 1981; Moore 2002; Dickens 1976; Ferguson et al. 1986;

Yarnell 1976; Simpkins 1984; Wing 1976). Unlike Garden Creek, Warren Wilson has no

discernible public or ceremonial structures and represents a smaller Pisgah site, thus providing a

richer data set for understanding the variability of Pisgah communities.

Cane River. The Cane River site (31Yc91), located on a rise above a tributary of the

Nolichucky River, is about 4-5 acres in extent based on palisade orientation. Limited salvage

4

excavations in 1989 and 1990 revealed one domestic structure, portions of several palisade lines,

and a wide array of features (Figure 2). Recent excavations in 2013-2014, guided by magnetic

gradiometer and ground-penetrating radar survey, has identified other portions of the palisade

lines, and uncovered additional feature and architectural assemblages across the site.

Additionally, two recent radiocarbon dates reveal that the site was occupied between ca. AD

1290 and 1610, making it contemporaneous with mound construction at Garden Creek (ca. AD

1423) and likely contemporaneous with the Warren Wilson village site (roughly dated to

between the twelfth and fifteenth centuries AD based on its ceramic assemblage). Cane River’s

intermediate village size and unusual upland location outside of a larger river valley provide a

unique perspective on Pisgah lifeways during this later sub-phase.

Utilizing the ceramic collections from Garden Creek and Cane River that I have recently

analyzed, compared to the published analyses of Warren Wilson, I hope to evaluate the nature

and extent of structural persistence or transformation associated with Mississippianization. This

paper will go over some of these preliminary findings with a sample of the data collected so far.

Previous Pisgah Ceramic Research

Former studies of Pisgah ceramics have discussed the diachronic shifts in ceramic

attributes within the Pisgah phase, subdivided into Early from ca. A.D. 1000 – 1250 and Late

from ca. A.D. 1250-1450 (Dickens 1976). Dickens was the first to note “ceramic changes from

Early to Late Pisgah are subtle. Rectilinear motifs are bolder, some curvilinear stamping is now

present, and check stamping (usually bolder) continues as a strong minority finish” (Dickens

1979:16). Later analysis by Dave Moore (1981) compared an earlier upland single-house

occupation (the Brunk site) to the later village occupation at Warren Wilson. This work

highlighted additional possible trends in temporal stylistic shifts between Early and Late Pisgah,

5

such as changes in collar shape (become larger), surface decoration (more narrow forms of

Rectilinear Design A in earlier, more curvilinear surface finishes in the later), and rim orientation

(more inslanted rims in the later phase). These trends possibly indicate temporal shifts in passive

ceramic construction or an intentional change indicating broader social mechanisms. It is also

possible these changes may be attributed to site or feature context, and the exact practices those

ceramics were associated with at the household or village level. That is why these broader

comparisons across multiple contemporaneous sites not participating in the same “local” Pisgah

system should be utilized and can further our understanding of the Pisgah phase social dynamics

within a broader settlement system.

For the purposes of comparing my analysis of the ceramics from Cane River and Garden

Creek, feature assemblages from both Dr. Dickens and Dr. Moore’s analysis of the Warren

Wilson ceramics were utilized. For the Garden Creek and Cane River assemblages, all sherds

were coded according to vessel portion, thickness, and diameter; technological attributes related

to production such as temper material, size, sorting, and presence/absence of mica; as well as

stylistic attributes such as exterior and interior surface treatments, presence/absence of usewear

and color. Additionally, all rims were coded for profile, rim form, lip form, decoration,

thickness, and vessel diameter. In this paper I will discuss the technological attributes and

stylistic attributes but will leave attribute analysis of the rims for another time. There are several

issues worth mentioning regarding differences in sampling and analytical methods. Dickens’

reported data (1970) included every recovered sherd from excavations, while Moore’s reported

data (1981) only includes those sherds larger than 1” for analysis. This can provide quite a

difference in identification of rectilinear complicated stamped varieties (Moore 1981:49),

potentially inflating the abundance of Rectilinear Design A compared to Rectilinear Design B.

6

For my sampling strategy at Garden Creek and Cane River, only ceramics larger than 25 mm

were analyzed, thus aligning my methods with Dr. Moore’s procedure. Another issue in data

collection was the variety of identified exterior surface treatment types. Dickens only identified

8 categories of exterior surface treatment, in addition to “unidentifiable”, while Moore separated

out his decorative types into 19 categories. For my own personal analysis I had 20 categories.

Therefore many of the broader comparisons between site assemblages will involve collapsing the

stylistic categories from Moore’s work and my own to facilitate comparisons with the larger

dataset collected by Dickens.

The Ceramic Patterns at the Warren Wilson Site

Roy Dickens was the first to provide detailed and comprehensive descriptions of the

Warren Wilson site Pisgah ceramics, sampling 30,144 sherds from excavations across plowzone,

feature and burial contexts (1976:172). For comparisons to new data collected from other Pisgah

sites, I utilized Dickens’ ceramic analysis of 2,204 sherds from 24 features at Warren Wilson.

Later analysis by David Moore (1981) sampled 2,850 sherds from two features at Warren

Wilson. After combining the results from these two samples from Warren Wilson (n=8,034),

from across 26 different feature contexts, the body surface treatments show 55.9% are rectilinear

complicated stamped, 0.9% are curvilinear comp stamped, 6.9% are check stamped, 1.7% are

plain, 1.2% other and 33.4% unidentifiable (Figure 2). For the complicated stamped that could

be assigned a specific variety in the two features analyzed by Dr. Moore (1981:46), 21.4% were

the broad form of Rectilinear Complicated Design A, 1.2% were the narrow form of Rectilinear

Complicated Design A, with 29.8% of the 1st variety of Rectilinear Complicate Design B making

up the largest proportion, 3.2% the 3rd variety, and 4.2% were Rectilinear Complicated Design C

(Figure 3).

7

The Ceramic Patterns at the Garden Creek site

Mound 1 at Garden Creek served as a ceremonial substructure and consisted of two

paired earth lodges and subsequently two more floor surfaces that were the base for later

ceremonial buildings and a log palisade. Rim sherds were analyzed from both Earthlodge 1 & 2

(n=64) and from Floor 1 (n=199) and Floor 2 (n=17) of the mound. One previous radiocarbon

date from an adjacent feature, assumed to be a borrow pit for clay fill utilized in building either

Floor 1 or Floor 2, yielded a date of A.D. 1435 +/_ 70 years (Dickens 1970:78). However that

remains the only radiocarbon date from the Pisgah occupation at Garden Creek, so the duration

of mound construction or chronology of other adjacent off-mound features is unclear.

In total, when comparing body surface treatment across all recently analyzed Garden

Creek rim sherds from mound contexts (n=280), 16.7% are rectilinear complicated stamped,

0.7% are curvilinear complicated stamped, 1% are check stamped, 37.4% are plain, 37.7% are

other and 6.4% are unidentifiable (Figure 4, Figure 5). Condensing the Floor ceramic

assemblages together and Earthlodge sherds together demonstrates that the earlier levels of the

mound (i.e. the Earthlodges) have different ratios of Rectilinear Complicated stamped designs

compared to the later Floor levels, with Rec. Com. Design A being more prevalent. The Floor

assemblages had a few Curvilinear Complicated Designs and a higher ratio of Rectilinear Design

B to Design A (Figure 6).

When considering the mound contexts again as a whole, within the rectilinear

complicated stamped that could be identified by variety, the broad form of Rec. Comp. Design A

was the most common (at ~18%) with the 1st variety of Rec. Comp B (~16%) and narrow form

of Rec. Comp. Design A (~14%) also being very prevalent (Figure 7). Given that the ratios of

Design B to Design A switch between earlier and later mound contexts, with the narrow and

8

broad forms of Design A being approximately equal in representation – this three-way split in

exterior surface decoration across all mound levels is to be expected.

In total, comparing body surface treatment across Garden Creek rim sherds from 91

feature contexts (n=778), 16.3% are rectilinear complicated stamped, 2.6% are curvilinear

complicated stamped, 0.8% are check stamped, 31.6% are plain, with 15.2% as “other” and

33.5% as unidentifiable (Figure 8, Figure 9). Within both feature and mound contexts, “other”

surface decorations included incised surface motifs such as those on cazuela bowls, cord-marked

surfaces (which made up 20% of the surface treatments sampled from the mound contexts), net-

impressed, burnished and indeterminate stamped exteriors. This accounts for some of the

observed patterns in exterior surface treatment and the large proportion of “other” types at

Garden Creek, which tends to stand out when we compare these three sites side-by-side. Within

the complicated stamped that could be identified by variety within the feature assemblages, the

broad form of Rec. Comp. Design A was by far the most prevalent type (~29%). The narrow

form of Rec. Comp. Design A was the second most common (~13%) followed by the Rec.

Comp. Design B and C varieties (Figure 10).

The Ceramic Patterns at the Cane River site

In total, when comparing body surface treatment across recently analyzed Cane River

sherds from 35 feature contexts (n=4,683), 43.7% were rectilinear complicated stamped, 0.1%

were curvilinear complicate stamped, 0.6% were check stamped, 20.9% were plain, 7%

unidentified, and 27.6% “other” (Figure 11, Figure 12). Of the rectilinear complicated stamped

that could be identified by variety, the broad form of Rec. Comp. Design A was the most

common, with the narrow form being the second most common. Rectilinear complicated

9

stamped C was rare, while the B varieties were equally represented (1st, 2nd and 3rd varieties)

around 1% each (Figure 13).

Inter-site Comparisons/ Results

When considering all three assemblages together, aspects of technological production do

not show many differences. Although, when noting characteristics of the temper, I also noted

whether mica was present in the paste for each sherd at Garden Creek and Cane River.

Unfortunately this data wasn’t collected for the Warren Wilson ceramic assemblage. There is a

slightly higher amount of mica temper present in Garden Creek’s feature contexts (48.7%) and

Cane River’s feature contexts (30.9%) compared to the mound sample at Garden Creek, which

has mica recorded in the temper in only 24% of the assemblage (Figure 14). It is interesting to

note that the Garden Creek and Cane River feature contexts also tend to have thicker vessel walls

on average (by 1mm, comparing averages to mound contexts). Therefore the difference in

proportion of mica temper utilized might be a functional aspect of production for larger, thicker

vessels deposited more often in off-mound areas, as opposed to differences in clay sourcing.

However this correlation is weak given the sample size and lack of data for Warren Wilson.

Variation in internal surface treatment is noticed within the Cane River feature

assemblage, which has the majority of vessel interiors burnished (78%), although this was also

the common practice at Warren Wilson. Moore noted that in his sample from features at Warren

Wilson, 45% had smoothed interiors and 55% were burnished (1981:52). Cane River also had

37% of the interiors smoothed and 15% that were rough (Figure 15). How this may correlate to

exterior surface treatment has yet to be considered, but it is worth noting that burnishing at

Warren Wilson was noted to be especially common on shallow bowls (Moore 1981:52). At

Garden Creek the mound contexts had more smoothed interior surfaces (38.9%) and rough

10

interior surfaces (42.5%) than burnished, while the feature assemblage was pre-dominantly

smoothed (44.7%) and burnished (36.4%) interiors.

Overall differences between exterior surface treatment in these assemblages, broken

down into the mound and off-mound assemblages at Garden Creek, the feature contexts at

Warren Wilson, and the feature contexts at Cane River, show the dominant form of body surface

decoration at Warren Wilson and Cane River is rectilinear complicated stamping. While this is

also present at Garden Creek, the plain and “other” designs make up larger percentages than the

complicated stamped designs. This may be due to the fact that no body sherds were included in

the sample from Garden Creek-- I limited my analysis to rim sherds, which allowed me to get a

large data set on rim attributes as well as exterior surface treatment. However often the sherd

would be broken off below the rim, at the shoulder, making “indeterminate stamping” one of the

most commonly assigned decorative treatments. I collapsed/condensed “indeterminate stamping”

into the “other” category for this inter-community comparison, and it is highly probable that

“indeterminate stamping” encompasses both complicated stamped varieties, as well as simple

stamped surface treatment.

When considering the identified varieties of rectilinear complicated stamped and

curvilinear complicated stamped designs, the Garden Creek Mound contexts had by far more

diverse representation compared to the Garden Creek and Cane River Feature contexts that had

Rectilinear Complicated Design A varieties being the clearly preferred surface treatment. Earlier

assemblages in the mound (those associated with Earthlodge 1 and 2) had Rectilinear Design A

as the predominant variety, while later contexts (Floors 1 and 2) preferred Rectilinear Design B.

While we have a feature dated to AD 1435 that may have been used to construct Floor 1 or 2, the

length of time for mound construction and how it relates to these off-mound occupation areas

11

isn’t really clear. However it is quite possible that the within-mound shifts of rectilinear

complicated stamped varieties indicate temporal trends in ceramic production over time within

this later Pisgah sub-phase.

Discussion and Future plans/questions: Mississippian socio-political relationships

Future directions for analysis of these assemblages will include rim attribute analysis, as

well as considerations of feature context and composition at each site. This would allow for

greater interpretations of intra-site variability and how that variation may affect the inter-site

comparisons discussed here. Additionally, I have plans to radiocarbon date more features at

Cane River - as well as samples from Warren Wilson and different mound levels and off-mound

features at Garden Creek. This will really help to corroborate these possible inter-phase

temporal trends. Also, by dating some of these features and fine tuning a ceramic seriation, it

will be easier to determine what aspects of community variation is due to contemporaneous

social practices or diachronic change. Lastly, the assemblages excavated from Cane River in

2013 and 2014 were not included in this analysis as data collection is still ongoing.

Acknowledgements: Funding was provided by the University of Michigan’s Museum of Anthropological Archaeology Radiocarbon Years Before Present Fund, the Griffin Scholarship, and the Griffin Endowment. Many thanks go to Roy Dickens and Dave Moore for their work on the Pisgah culture. Also a big thank you to Brett Riggs, Stephen Davis, Vin Steponaitis and the Research Laboratories of Archaeology at UNC Chapel Hill for giving me lab space and access to the Garden Creek collections this past summer. It was a joy to be a part of your community, albeit for a brief time! I would also like to thank Rob Beck, Scott Ashcraft, Elaine Dellinger and the Yancey History Association, and the Michigan friends and colleagues for all their help and encouragement.

12

Figure 1. Pisgah communities in the Appalachian Summit of North Carolina that are discussed within this paper.

13

Figure 2. Exterior surface treatment percentages across all Warren Wilson feature contexts (Dickens 1976, Moore 1981)

Figure 3. Complicated stamped type (rectilinear or curvilinear) and variety across two features at Warren Wilson (Moore 1981)

55.9%

0.9% 6.9%

1.7%

33.4%

0.1%

Warren Wilson Feature Contexts (n=8034)

Rec Comp

Curv Comp

Check

Plain

Unident

Other

1.2%

21.4%

29.8%

0.7%

3.2% 4.2% 0.25% 0.7% 0.5%

Warren Wilson Complicated Stamped (n=812)

Rec Comp A Narrow

Rec Comp A Broad

Rec Comp B 1

Rec Comp B 2

Rec Comp B 3

Rec Comp C 1

Rec Comp C 2

Curv Comp A

Curv Comp B

14

Figure 4. Exterior surface treatment percentages across all Garden Creek mound contexts [note the rectilinear complicated stamped varieties are separated out into Design A (RCA), Design B

(RCB) and Design C (RCC)]

Figure 5. Exterior surface treatment percentages across all Garden Creek mound contexts [the rectilinear complicated stamped varieties are not delineated here]

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

Garden Creek Mound Contexts (n=280)

Exterior Surface %

16.7% 0.7%

1%

37.4%

6.4%

37.7%


Rec Comp

Curv Comp

Check

Plain

Unident

Other

15

Figure 6. Exterior surface treatment percentages of the Garden Creek mound contexts broken down by earlier (Earthlodge 1 & 2) and later (Floors 1 and 2) structural fill zones

Figure 7. Complicated stamped type (rectilinear or curvilinear) and variety across all Garden Creek Mound 1 contexts [note: no curvilinear complicated stamped varieties could be specified]

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Floors 1 & 2 (n=116)

Earthlodges (n=64)

14.2%

18.4%

16.3%

8.2%

6.1%

6.1%

Garden Creek Mound Complicated Stamp (n=49)

Rec Comp A Narrow

Rec Comp A Broad

Rec Comp B 1

Rec Comp B 2

Rec Comp B 3

Rec Comp C 1

16

Figure 8. Exterior surface treatment percentages across all Garden Creek feature contexts [note the rectilinear complicated stamped varieties are separated out into Design A (RCA), Design B


Figure 9. Exterior surface treatment percentages across all Garden Creek feature contexts [the rectilinear complicated stamped varieties are not delineated here]

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Garden Creek Feature Contexts (n=778)

Exterior Surface %

16.3% 2.6%

0.8%

31.6% 33.5%

15.2%


Rec Comp

Curv Comp

Check

Plain

Unident

Other

17

Figure 10. Complicated stamped type (rectilinear or curvilinear) and variety across all Garden Creek feature contexts

Figure 11. Exterior surface treatment percentages across Cane River feature contexts [note the rectilinear complicated stamped varieties are separated out into Design A (RCA), Design B


12.9%

29.5%

4%

1.4% 4%

5.4%

2.4% 2.7%

Garden Creek Feature Complicated Stamp (n=147)

Rec Comp A Narrow

Rec Comp A Broad

Rec Comp B 1

Rec Comp B 2

Rec Comp B 3

Rec Comp C 1

Curv Comp A

Curv Comp B

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Cane River Feature Contexts (n=4683)

Exterior Surface %

18

Figure 12. Exterior surface treatment percentages across Cane River feature contexts [the rectilinear complicated stamped varieties are not delineated here]

Figure 13. Complicated stamped type (rectilinear or curvilinear) and variety across all Cane River feature contexts

43.7%

0.1% 0.6%

20.9%

7%

27.6%


Rec Comp

Curv Comp

Check

Plain

Unident

Other

10.3%

25.3%

0.9% 1.5%

1.1% 0.8%

Cane River Complicated Stamped (n=2041)

Rec Comp A Narrow

Rec Comp A Broad

Rec Comp B 1

Rec Comp B 2

Rec Comp B 3

Rec Comp C 1

Rec Comp C 2

Curv Comp A

19

Figure 14. The percentage of the assemblages at Cane River and Garden Creek (mound and off-mound) that had mica within the temper (% present and % absent)

Figure 15. Interior surface treatment percentages across Cane River, Warren Wilson and Garden Creek (mound and off-mound) contexts

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

% Mica Present % Mica Absent

GC Feature Contexts

GC Mound Contexts

Cane River Feature Contexts

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Rough Plain Smooth Plain Burnished Eroded

CR Fea Interior Surface Treatment

GC Fea Interior Surface Treatment

GC Mound Interior Surface Treatment

WW Fea Interior Surface Treatment

20

Figure 16. Exterior surface treatment percentages across Cane River, Warren Wilson, and Garden Creek (mound and off-mound) contexts

0

0.1

0.2

0.3

0.4

0.5

0.6



Warren Wilson Feature Contexts (n=8034)


21

Bibliography

Beck, Robin A Jr., and David G. Moore 2002 The Burke Phase: A Mississippian Frontier in North Carolina Foothills. Southeastern Archaeology 21(2): 162–169.

Blitz, John H.

2010 New Perspectives in Mississippian Archaeology. Journal of Archaeological Research 18(1): 1–39.

Cobb, Charles R.

Mississippian microhistories and sub-mound moments. In The enigma of the event: moments of consequence in the ancient Southeast, edited by Zackary I. Gilmore and Jason M. O’Donoughue, pp. 1–42. University Press of Florida, Gainesville.

2003 Mississippian chiefdoms: how complex? Annual Review of Anthropology 32(1): 63–84.

Dickens, Roy S. Jr.

1970 The Pisgah culture and its place in the prehistory of the Southern Appalachians. University of North Carolina, Chapel Hill.

1976 Cherokee Prehistory: the Pisgah Phase in the Appalachian Summit Region. The University of Tennessee Press, Knoxville.

1979 The Origins and Development of Cherokee Culture. In The Cherokee Indian Nation: A Troubled History, edited by Duane H. King, pp. 3–32. The University of Tennessee Press, Knoxville.

Ferguson, Leland G., H. Trawick Ward, Daniel L. Sinpkins, Homes H. Wilson, David G. Moore, Roy S. Jr. Dickens, David J. Hally, Gerald F. Schroedl, Charles Hudson, and Theda Perdue

1986 The Conference on Cherokee Prehistory. In The Archaeology of an Appalachian Culture, edited by David G. Moore. Warren Wilson College, Swannanoa, NC.

Griffin, James B.

1943 The Fort Ancient aspect, its cultural and chronological position in Mississippi Valley archaeology. University of Michigan Press, Ann Arbor.

1985 Changing concepts of the prehistoric Mississippian cultures of the Eastern United States. In Alabama and the borderlands: from prehistory to statehood, edited by R. Reid Badger and L. A. Clayton, pp. 40–63. The University of Alabama Press, Tuscaloosa.

1992 Fort Ancient has no class: the absence of an elite group in Mississippian societies in the central Ohio Valley. In Lords of the Southeast: social inequality and the native elites of southeastern North America, edited by Alex W. Barker and Timothy R.

22

Pauketat, pp. 53–60. Archaeological Papers of the American Anthropological Association Number 3.

Keel, Bennie C.

1975 On South Appalachian Mississippian in the Appalachian Summit Area. Southeastern Archaeological Conference Bulletin 18: 12–18.

1976 Cherokee archaeology: a study of the Appalachian Summit. The University of Tennessee Press, Knoxville.

Kelly, John E.

2000 The Nature and Context of Emergent Cultural Dynamics in the Greater American Bottom. In Late Woodland Societies: Tradition and Transformation across the Midcontinent, pp. 163–175. University of Nebraska Press, Lincoln.

Meyers, Maureen S.

2002 The Mississippian frontier in southwestern Virginia. Southeastern Archaeology 21(2): 178–191.

Moore, David G.

1981 A comparison of two Pisgah ceramic assemblages. the University of North Carolina at Chapel Hill.

2002 Pisgah phase village evolution at the Warren Wilson site. In The Archaeology of Native North Carolina, papers in honor of H. Trawick Ward, edited by Jane M. Eastman, Christopher B. Rodning, and EA Boudreaux, Southeaste:pp. 76–83. Southeastern Archaeological Conference Special Publication 7.

Pauketat, Timothy R.

1994 The ascent of chiefs, Cahokia and Mississippian politics in native North America. The University of Alabama Press, Tuscaloosa.

Simpkins, Daniel L.

1984 An ethnobotanical study of plant food remains from the Warren Wilson site (31Bn29), North Carolina: a biocultural approach. University of North Carolina, Chapel Hill.

Smith, Bruce D.

1990 The Mississippian Emergence. Smithsonian Institute Press, Washington, D.C. Ward, H. Trawick, and Stephen R. P. Davis

1999 Time before history: the archaeology of North Carolina. The University of North Carolina Press, Chapel Hill.

Wing, Elizabeth S.

23

1976 Appendix B: Faunal Remains from the Warren Wilson Site. In Cherokee Prehistory: The Pisgah Phase in the Appalachian Summit Region, pp. 224–229. The University of Tennessee Press, Knoxville.

Woodall, J. Ned

1999 Mississippian Expansion on the Eastern Frontier: One Strategy in the North Carolina Piedmont. Archaeology of Eastern North America 27: 55–70.

Wright, Alice P.

2014 Inscribing Interaction: Middle Woodland Monumentality in the Appalachian Summit, 100 B.C. - A.D. 400. University of Michigan.

Yarnell, Richard A.

1976 Appendix A: Plant Remains from the Warren Wilson Site. In Cherokee Prehistory: The Pisgah Phase in the Appalachian Summit Region, pp. 217–229. The University of Tennessee Press, Knoxville.

Pisgah Ceramic Production in the Southern Appalachians

Documents

Transcript of Pisgah Ceramic Production in the Southern Appalachians