1 LANDSCAPE PATCHES, MACROREGIONAL EXCHANGES AND PRE-COLUMBIAN POLITICAL ECONOMY IN SOUTHWESTERN GEORGIA.

1 LANDSCAPE PATCHES, MACROREGIONAL EXCHANGES AND PRE-COLUMBIAN POLITICAL ECONOMY IN SOUTHWESTERN GEORGIA.
1
LANDSCAPE PATCHES, MACROREGIONAL EXCHANGES AND
PRE-COLUMBIAN POLITICAL ECONOMY IN SOUTHWESTERN GEORGIA.
by
John Francis Chamblee
Copyright © John Francis Chamblee 2006
A Dissertation submitted to the Faculty of the
DEPARTMENT OF ANTHROPOLOGY
In Partial Fulfillment of the Requirements
For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2006
2
As members of the Dissertation Committee, we certify that we have read the dissertation
prepared by John Francis Chamblee
entitled Landscape Patches, Macroregional Exchanges, and Pre-Columbian Political
Economy in southwestern Georgia.
and recommend that it be accepted as fulfilling the dissertation requirement for the
Degree of Doctor of Philosophy
_______________________________________________________________________
Date: May 19, 2006
Paul R. Fish
_______________________________________________________________________
Date: May 19, 2006
Barbara J. Mills
_______________________________________________________________________
Date: May 19, 2006
Suzanne K. Fish
_______________________________________________________________________
Date: May 19, 2006
Gary L. Christopherson
_______________________________________________________________________
Date: May 19, 2006
Stephen A. Kowalewski
Final approval and acceptance of this dissertation is contingent upon the candidate’s
submission of the final copies of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and
recommend that it be accepted as fulfilling the dissertation requirement.
________________________________________________ Date: May 19, 2006
Dissertation Director: Paul R. Fish
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an
advanced degree at the University of Arizona and is deposited in the University Library
to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission,
provided that accurate acknowledgment of source is made. Requests for permission for
extended quotation from or reproduction of this manuscript in whole or in part may be
granted by the copyright holder.
SIGNED: John F. Chamblee
4
ACKNOWLEDGEMENTS
A National Science Foundation Dissertation Improvement Grant (# 0340675)
funded the bulk of this project. Other agencies also provided generous funding. I am
grateful to the Graduate College of the University of Arizona, the Historic Preservation
Division of the Georgia Department of Natural Resources, The William Shirley Fulton
Fund of the University of Arizona Anthropology Department and the Arizona State
Museum’s Raymond H. Thompson Endowment. Thanks to Dave Crass, State
Archaeologist of Georgia, for permission to survey the Chickasawhatchee WMA.
I thank my committee for their intellectual support and guidance. My major
Professor, Paul Fish, graciously gave me just enough rope with which to hang myself and
yet was always available whenever I needed help. I am similarly grateful to my long-time
mentor, Steve Kowalewski. Barbara Mills was consistently patient with my ugly sherds
of small quantity. Gary Christopherson actually came to the swamp to dig and map.
Suzanne K. Fish got me thinking about ecosystems again. Many colleagues also provided
great ideas – especially Jim Vint, Sammy Smith and Barnet Pavao-Zuckerman.
I appreciate support from the University of Georgia – especially from Mark
Williams and Dave Hally. Thanks to the Antonio Waring Lab of West Georgia and the
Archaeology Lab of South Georgia College for collections access. Frankie Snow, Keith
Stephenson and Dan Elliott provided key advice regarding fieldwork. In southwestern
Georgia, I received help from an entire community, including Steve Ruckel and Derek
Fussel from the Georgia DNR, as well as Lindsay Boring, Kay Kirkman, Woody Hicks
and Jean Brock from the Jones Center. I thank the many collectors and landowners – too
numerous to mention on one page – who helped me look for sites, allowed me to work on
their land, showed me their collections, or just provided their wisdom. Marvin Singletary
and Eugene Black provided great introductions to Southwest Georgia – past and present.
Foremost among the great collaborators I worked with in the field is Jamie
Waggoner, who spent nearly five months with me. Morgan Ritchie, Maren Hopkins, Jim
Chamblee, Eric Marks, Elaine Juzwick, Gary and Ruth Christopherson and Tisha Entz all
did a lot of digging and walking – each for least a month. Other folks who helped out
include Sammy Smith, Matt Hewett, Hope Chamblee, Kristian Klaene, Mary Ann
Miranda, Christina Hall and Stacey Thompson, My wife, Ruby Basham, kept us all fed
during the summer of 2004. Special thanks to Cherie Freeman for all the help in the lab.
Thanks to Jack Perryman, Doug Logan, Albert and Judy Newberry and everyone
at Edison First Baptist for their friendship. Finally, even with all this generous funding,
support, and advice, my research would not have been possible without David J.
Middleton. Dave allowed me to stay on his property for fourteen months. He also
introduced me to everyone he knew in southwestern Georgia and northern Florida who
might be able to help me – most of whom did. I cannot express my gratitude to Dave and
his family and will always treasure the time I spent at De Soto Springs Plantation.
5
DEDICATION
This dissertation is dedicated to my family. My wife, Ruby, has been a patient and loving
companion on a trip that was longer than we anticipated. My son, Jack, continues to be a
source of joy, pride, and respite. My parents, Hope and Jim, have been constant pillars of
support. Thank you.
6
TABLE OF CONTENTS
LIST OF TABLES .....................................................................................................................9
LIST OF FIGURES ...................................................................................................................10
ABSTRACT ..................................................................................................................12
CHAPTER 1: INTRODUCTION.................................................................................13
Theoretical Research Context...........................................................................16
Study Area and Research Design .....................................................................21
Chapter Summaries ...........................................................................................26
CHAPTER 2: RESEARCH CONTEXT......................................................................30
Agency and Structure in the Prehistoric Southeast: The Legacy of Chiefdoms
..............................................................................................................................31
Global and Local: Landscape Structure and Macroregional Interaction....40
Macroregional Interaction: The Thread of Continuity...................................41
Landscape Structure: The Interior Coastal Plain and the
Chickasawhatchee Swamp.............................................................................51
Summary: The Chickasawhatchee Swamp as a Test Case ............................59
CHAPTER 3: METHODS AND MIDDLE RANGE THEORY ...............................63
Archival Research, Existing Collections, and GIS Development ..................64
Existing Settlement Pattern Data and Previous Investigations ......................64
Classification of Existing Ceramic Collections .............................................67
GIS Data Assembly........................................................................................68
Field Methods .....................................................................................................73
Regional Survey.............................................................................................76
Intensive Mapping, Testing, and Excavations ...............................................83
A Note on Regional Formation Processes ...............................................84
Site Mapping ............................................................................................84
Shovel Testing ..........................................................................................86
Excavation................................................................................................88
Intrasite Processes: Artifact Deposition and Soil Analysis............................90
Ceramic Analysis ...............................................................................................92
Analytical Problems and Middle Range Theory............................................93
Chronology ..............................................................................................93
Formation Processes ...............................................................................98
Artifact Accumulation and Occupational Intensity..................................100
Ceramic Production and Exchange .........................................................102
Ceramic Classification and Categorization....................................................104
Relational Databases and Ceramic Classification ........................................105
Ceramic Analysis Summary ..........................................................................114
Chapter 3 Summary ..........................................................................................114
CHAPTER 4: UNDERSTANDING REGIONAL AND INTRASITE SITE
FORMATION PROCESSES........................................................................................116
Site Formation Processes and Survey Coverage Bias.....................................117
Survey Coverage............................................................................................118
7
TABLE OF CONTENTS-Continued
High Intensity Survey...............................................................................118
Low Intensity Survey ................................................................................121
Mechanized Farming and Artifact Variation .................................................126
Site Formation Processes at the Intrasite Scale ..............................................131
Hay Fever Farm .............................................................................................132
Windmill Plantation .......................................................................................142
Additional Excavated Sites ............................................................................147
Intrasite Pattern Summary..............................................................................148
Comparing Surface and Subsurface Assemblages .........................................148
Is Time on Our Side? The Effect of Repeated Disturbances on Artifact
Variation .............................................................................................................151
Regional and Intrasite Site Formation Processes: Summary and Conclusions
..............................................................................................................................155
CHAPTER 5: THE REGIONAL POLITICAL ECONOMY OF THE
CHICKASAWHATCHEE SWAMP............................................................................156
Regional Settlement History in the Chickasawhatchee Swamp ....................158
General Native American Ceramic Components...........................................158
The Woodland Period ....................................................................................161
The Middle Woodland Period..................................................................163
The Late Woodland Period ......................................................................168
The Mississippian Period...............................................................................172
The Middle Mississippian Period ............................................................174
The Late Mississippian Period.................................................................179
Summary of Regional Settlement Patterns ....................................................183
Intrasite Settlement Patterns: Four Examples................................................185
Windmill Plantation .......................................................................................185
Intrasite Patterns .....................................................................................186
Windmill Plantation in Regional Context ................................................190
Magnolia Plantation .......................................................................................191
Mapping ...................................................................................................192
Shovel Testing ..........................................................................................195
Excavations and Mound Cores ................................................................196
Ceramic Data...........................................................................................204
Magnolia Plantation in Regional Context ...............................................207
Red Bluff Earthlodge .....................................................................................208
Mapping ...................................................................................................209
Shovel Testing ..........................................................................................209
Excavations ..............................................................................................211
Ceramic Data...........................................................................................215
Red Bluff Mississippian Architecture: House or Earthlodge ..................216
The Red Bluff Earthlodge Site in Regional Context.................................217
Chickasawhatchee Knoll................................................................................218
8
TABLE OF CONTENTS-Continued
Hay Fever Farm .............................................................................................222
Settlement Zones and Cycles of Growth and Contraction.............................223
Landscape Variation and Regional Settlement...............................................226
Project-Wide Ceramic Variation and Density ................................................233
Summary and Conclusions................................................................................246
CHAPTER 6: THE CHICKASAWHATCHEE SWAMP IN MACROREGIONAL
CONTEXT......................................................................................................................247
Methods...............................................................................................................248
Macroregional Settlement Pattern Change .....................................................250
A Macroregional View of Intrasite Variation at Mound Centers .................270
Summary.............................................................................................................275
CHAPTER 7: SUMMARY AND CONCLUSIONS ...................................................277
Results .................................................................................................................278
Discussion............................................................................................................281
Future Research Directions ..............................................................................283
APPENDIX A: TABLE LISTING SITES, SITE AREAS, AND COMPONENTS OF
ALL CERAMIC SITES ................................................................................................286
APPENDIX B: COMPLETE PROVENIENCE LIST FOR PROJECT SURVEY
AND TESTING ..............................................................................................................290
APPENDIX C: CERAMIC CLASSIFICATION DATA, ACCORDING TO THE
LOT NUMBERS IN APPENDIX B .............................................................................305
REFERENCES CITED .................................................................................................334
9
LIST OF TABLES
TABLE 3.1,
TABLE 3.2,
TABLE 3.3,
TABLE 4.1,
TABLE 4.2,
TABLE 4.3,
TABLE 4.4,
TABLE 4.5,
TABLE 4.6,
TABLE 4.7a,
TABLE 4.7b,
TABLE 4.8a,
TABLE 4.8b,
TABLE 5.1,
TABLE 5.2,
TABLE 5.3,
TABLE 5.4,
TABLE 5.5,
TABLE 5.6,
TABLE 5.7,
TABLE 5.8,
TABLE 5.9,
TABLE 5.10,
TABLE 5.11,
TABLE 5.12,
TABLE 5.13,
TABLE 5.14,
TABLE 6.1,
TABLE 6.2,
Ceramic decorative types and temporal associations.........................................108
Vessel form classes............................................................................................110
Rim modifications and temporal associations ...................................................111
Observed and expected distributions of soil classes and vegetation categories in
low intensity survey transects, as well as chi square test results .......................123
Sherd size distribution data distributed according to field preparation techniques
...........................................................................................................................129
Soil Particle size and available phosphate data, according to stratigraphic level.
CAS 245, XU 1..................................................................................................136
Sherd size data, by stratigraphic level. CAS 245, XU 1 ....................................138
Available phosphate data, according to stratigraphic level. 9DU6, XU 2 .........144
Sherd size, by stratigraphic level. 9DU6, XU 2.................................................145
Sherd size data from surface collections with more than 15 sherds ..................150
Sherd size data from shovel tests at four test sites.............................................150
Sherd counts, site areas, and sherds/ha estimates from recently disturbed sites
...........................................................................................................................153
Sherd counts, site areas, and sherds/ha estimates from repeatedly disturbed sites
...........................................................................................................................153
Site area and component data, Unknown Native American Ceramic period
components ........................................................................................................159
Site area and component data, Woodland period components ..........................161
Site area and component data, Mississippian period components .....................173
Ceramic type distribution from excavation units and shovel tests, 9DU6.........188
Ceramic type distribution from excavation units and shovel tests, 9DU1.........205
Ceramic type distribution from excavation units and shovel tests, 9BX4 .........216
Ceramic type distribution from surface collections, CAS 89 ............................221
Ceramic type distribution from excavation units and shovel tests, CAS 245....222
Observed and expected distributions of soil classes in low intensity survey
transects, as well as chi square test results.........................................................230
Observed and expected distributions of reconstructed vegetation classes in low
intensity survey transects and among sites, as well as chi square test results....232
Sherd counts by decorative type, temporal designation, and vessel form .........236
Vessel thickness measures according to A) decorative type and temporal
designation and B) vessel form..........................................................................238
Rim modification counts....................................................................................239
Decorated ceramic relative frequencies .............................................................244
Mound and non-mound site frequencies in Georgia by time period..................251
Site area, shovel test frequency, and ceramic density data for mound sites in the
Fall Line Hills and Coastal Plain physiographic provinces ...............................271
10
LIST OF FIGURES
FIGURE 1.1,
FIGURE 1.2,
FIGURE 1.3,
FIGURE 1.4,
FIGURE 2.1.
FIGURE 3.1,
FIGURE 3.2,
FIGURE 3.3,
FIGURE 3.4,
FIGURE 3.5,
FIGURE 3.6,
FIGURE 4.1,
FIGURE 4.2,
FIGURE 4.3a,
FIGURE 4.3b,
FIGURE 4.4,
FIGURE 4.5,
FIGURE 4.6,
FIGURE 4.7,
FIGURE 4.8,
FIGURE 4.9,
FIGURE 4.10,
FIGURE 4.11,
FIGURE 4.12,
FIGURE 4.13,
FIGURE 4.14,
FIGURE 4.15,
FIGURE 4.16,
FIGURE 4.17,
FIGURE 4.18,
FIGURE 5.1,
FIGURE 5.2,
FIGURE 5.3,
FIGURE 5.4,
FIGURE 5.5,
FIGURE 5.6,
FIGURE 5.7,
FIGURE 5.8,
FIGURE 5.9,
FIGURE 5.10,
Project area location...........................................................................................15
Dougherty Plain physiography ..........................................................................22
Mound centers and study areas in the interior coastal plain ..............................23
Interior coastal plain regional ceramic sequences..............................................25
Comparative topographic profiles of piedmont and coastal plain landscapes ...61
Grouped soil classes on the Chickasawhatchee Wildlife Management Area ....71
Grouped vegetation classes on the Chickasawhatchee Wildlife Management Area
...........................................................................................................................73
Entity Relationship schematic for the project database .....................................75
High intensity (red) and low intensity (black) surface collection survey transects
...........................................................................................................................81
Sites located via high intensity (red) and low intensity (black) survey .............82
Collecting 5 cm3 soil columns from an excavation unit profile.........................89
High intensity survey areas and random test point distributions .......................120
Cumulative frequency distributions for elevation in high intensity areas..........121
Cumulative frequency distributions for distance to roads in high intensity areas
...........................................................................................................................125
Cumulative frequency distributions for distance to roads across the project area
...........................................................................................................................125
Row plowed field planted in corn......................................................................127
Site distributions according to plowing methods...............................................130
Sites used for sedimentological analysis............................................................133
Topographic context of Hay Fever Farm (CAS 245) ........................................134
Hay Fever Farm, XU 1, Final excavation level .................................................135
Stratigraphic distribution of available phosphate, CAS 245..............................137
Stratigraphic distribution of soil particle size distributions, CAS 245 ..............137
Stratigraphic distribution of ceramic frequencies, CAS 245 .............................138
Stratigraphic distribution of ceramic frequencies by sherd size, CAS 245........139
Topography and Inverse Distance Weighting surface of ceramic density, CAS
245 .....................................................................................................................141
One meter resolution aerial photograph of Windmill Plantation (9DU6)..........143
9DU6, XU 2, Final excavation level..................................................................143
Stratigraphic distribution of available phosphate, 9DU6...................................145
Stratigraphic distribution of ceramic frequencies by sherd size, 9DU6.............146
Stratigraphic distribution of ceramic frequencies, 9DU6 ..................................146
General Woodland period ceramic diagnostic distributions ..............................163
Middle Woodland period site distributions........................................................165
Middle Woodland period ceramic diagnostic distributions ...............................167
Late Woodland period site distributions ............................................................169
Late Woodland period ceramic diagnostic distributions....................................170
Middle Mississippian period site distributions ..................................................176
Topographic context of Magnolia Plantation (9DU1) .......................................177
Middle Mississippian period ceramic diagnostic distributions..........................178
Late Mississippian period site distributions.......................................................180
Late Mississippian period ceramic diagnostic distributions ..............................181
11
LIST OF FIGURES-Continued
FIGURE 5.11, Topography and Inverse Distance Weighting surface of ceramic density, 9DU6
...........................................................................................................................186
FIGURE 5.12, The Windmill Plantation mound........................................................................187
FIGURE 5.13, Mound A at Magnolia Plantation, facing SSE...................................................192
FIGURE 5.14, Topographic map and shovel test distributions of Magnolia Plantation (9DU1)
...........................................................................................................................193
FIGURE 5.15, Mound B at Magnolia Plantation, Facing NE....................................................194
FIGURE 5.16, Topographic map of Mound B...........................................................................196
FIGURE 5.17, East Profile Drawing, Mound B, XU 1, 9DU1 ..................................................198
FIGURE 5.18, East Profile Photo, Mound B, XU 1, 9DU1.......................................................199
FIGURE 5.19, South Profile Photo, Mound B, XU 1, 9DU1 ....................................................200
FIGURE 5.20, South Profile Drawing, Mound B, XU 1, 9DU1................................................201
FIGURE 5.21, Soil profile schematics from posthole tests of Mound B, 9DU1 .......................203
FIGURE 5.22, Topographic and site distribution context, Red Bluff Earthlodge (9BX4) and
Chickasawhatchee Knoll (CAS 89) ...................................................................210
FIGURE 5.23, Topography and Inverse Distance Weighting surface of ceramic density, 9BX4
...........................................................................................................................212
FIGURE 5.24, Profile Drawing, Level 4, XU 3, Structure 3, 9BX4..........................................213
FIGURE 5.25, Plan View Drawing, Level 4, XU 3, Structure 3, 9BX4....................................214
FIGURE 5.26, Fort Walton Incised (Lot 34) and Lamar Comp. Stamped (Lot 43) pottery, XU 3,
Str. 3, 9BX4 .......................................................................................................215
FIGURE 5.27, Site map of Chickasawhatchee Knoll, showing dog-leash collection unit locations
...........................................................................................................................219
FIGURE 5.28, Settlement zones during A) Woodland and B) Mississippian periods.......224-225
FIGURE 5.29, Site distributions and topographic context on the Chickasawhatchee WMA ....228
FIGURE 5.30, Site and soil distributions on the Chickasawhatchee WMA ..............................229
FIGURE 5.31, Sites and vegetation zone reconstruction on the Chickasawhatchee WMA ......231
FIGURE 5.32, Indicators of ceramic production irregularities..........................................240-243
FIGURE 6.1, Middle Woodland period settlement patterns ....................................................253
FIGURE 6.2, Late Woodland period settlement patterns ........................................................254
FIGURE 6.3, Early Mississippian period settlement patterns..................................................255
FIGURE 6.4, Middle Mississippian period settlement patterns...............................................258
FIGURE 6.5, Middle Mississippian period mound and soil distributions ...............................260
FIGURE 6.6, Late Mississippian period mound centers..........................................................263
FIGURE 6.7 Straight-line distances between Gulf and Atlantic Coast drainages and interior
settlement zones.................................................................................................265
FIGURE 6.8, Physiographic provinces and long-term mound distribution patterns ...............268
FIGURE 6.9, Physiographic provinces, Mississippian period mound distributions, and Late
Mississippian period non-mound site distributions ...........................................269
FIGURE 6.10, Ceramic richness and diversity at sites in the Fall Line Hills and Coastal Plain
physiographic provinces ....................................................................................273
FIGURE 6.11, Ceramic density data according to sherds/positive shovel test at ten study sites
within the Fall Line Hills and Coastal Plain physiographic provinces ..............274
12
ABSTRACT
Results from archaeological survey provide new insights into the origins of
variation among the prehistoric Native American societies that occupied the
Chickasawhatchee Swamp of southwestern Georgia. Through macroregional comparison,
these insights are broadly applicable to the Eastern Woodlands societies that existed
across the southeastern U.S. between A.D. 150 and 1600. Theoretical frameworks
concerning landscape ecology, inter-regional exchange, and agency and structure provide
the organizing structure for a multi-scalar view of change that contradicts earlier models.
Within the Chickasawhatchee Swamp, survey, mapping, and excavation data
present a complex regional settlement system. Within the swamp, a few large settlements
were occupied for the long-term, in spite of the absence of monumental architecture.
Smaller surrounding sites were periodically abandoned. At the swamp’s edge, several
subregions were organized around civic-ceremonial mound sites. At these edges, mound
sites and surrounding subregions were abandoned simultaneously. Instead of being driven
by changes in political complexity, residential mobility cycles were consistent through
time and related to the region’s heterogeneous landscape.
Macroregional spatial data comparing mound locations through time support data
from the Chickasawhatchee Swamp and confirm hypotheses relating mound construction
and transitional landscapes. New data emphasize continuity in inter-regional exchange
networks and contradict earlier views in which the emergence of hierarchical political
structures were a transformational process that fundamentally altered Eastern Woodlands
political economies. Temporal continuity and spatial variation are instead most evident.
13
CHAPTER 1: INTRODUCTION
Anthropologists pursue the origins of spatial and temporal variation among
societies. Many theoretical frameworks have been employed to further our understanding
of variation, but too often, the theoretical tail ends up wagging the empirical dog. In
archaeology, recent theories look carefully at variation without hampering archaeological
data with a burden of assumptive baggage. More recent studies shifted the emphasis of
social variation studies from unilineal models of social evolution (cf. Sahlins 1958 and
Service 1962) by emphasizing agent-driven leadership strategies (Blanton et al. 1996),
power (Earle 1997; Pauketat 1994), and the tension between the interests of social agents
and the unintended consequences of their behavior (Giddens 1979:44; Wiessner 2002).
This study takes up the debate concerning variation among the middle range
Eastern Woodlands societies that occupied modern day southwestern Georgia and nearby
areas during the Woodland and Mississippian periods (from approximately A.D. 150
through A.D. 1600). While acknowledging there are many factors influencing social
variation, the model of social change used here relates local ecological variation and
macroregional interaction to shifts in settlement patterns and ceramic production regimes.
The theoretical framework for the model places agent driven social change and
historically constituted institutional constraint in balance – without giving undue
emphasis to either phenomenon (Giddens 1979:95).
This is a multi-scalar analysis. Regional data from the Chickasawhatchee Swamp
(Figure 1.1) are contextualized by comparative and macroregional data from sites in
14
neighboring regions throughout the interior coastal plain and nearby physiographic
provinces. Intrasite data from the Chickasawhatchee Swamp and neighboring areas
supplement regional and macroregional analyses. Archaeological interpretations are
further contextualized with ecological data.
Overall, this study recognizes that, regardless of whether one considers prehistoric
social interaction or archaeological analysis, patterns that appear similar at one scale may
be different at another – and that the finer resolutions may not show the differences.
Together, multi-scalar analysis and a balanced view of agency and structure provide new
insights into the origins of variation among Eastern Woodlands societies. Specifically,
data from this study demonstrate the following:
1) Within the Chickasawhatchee Swamp, a heterogeneous and easily disturbed
ecological landscape structure reduced incentives for settlement aggregation and
instead encouraged a long-term settlement strategy driven by local settlement
shifts or sub-regional abandonment.
2) As a whole, the landscape structure of the interior coastal plain offers similar
disincentives for settlement aggregation as those seen in the Chickasawhatchee
Swamp.
3) Macroregional trends in the location and frequency of mound centers reflect
considerable variation according to broad-scale physiographic settings while
simultaneously exhibiting remarkable local continuity across the Woodland and
Mississippian periods
15
4) Prior research has de-emphasized or striking continuities between the
Woodland and Mississippian periods in the areas of settlement location,
settlement structure, and cycles of settlement pattern change.
5) Macroregional settlement pattern continuities suggest the long term importance
of macroregional exchange networks in constraining settlement pattern change.
In this chapter I present the key theoretical issues surrounding the research and also
introduce the study area and project scope. Following these discussions, chapter
summaries describe specific supporting arguments for the dissertation’s overall
conclusions.
FIGURE 1.1,
Project area location
16
Theoretical Research Context
Several domains of theoretical inquiry relate ecological variation and
macroregional interaction to the political economic shifts reflected by changes in
settlement patterns and ceramic distributions. Most important among these are the
relevance and importance of neo-evolutionary and agent driven models of social change,
the temporal and spatial scales at which different kinds of change are best understood,
and the relationships between social theory and theories and models developed in related
fields – such as ecology.
This discussion addresses these questions through a brief historical sketch of the
relationships between social theory, analytical scale, and human ecology in the study of
Eastern Woodlands societies. Problems with current theoretical models are juxtaposed
with the potential solutions pursued in this dissertation. Needless to say, these theoretical
domains require the support of additional theories and bridging arguments, but these
details are discussed in later chapters.
Regional studies of Mississippian period chiefdoms have yielded insights
regarding economic organization (Welch 1981), settlement structure and spatial
organization (Hally 1996; Steponaitis 1978), the cycling of chiefly leadership offices and
the abandonment of civic-ceremonial centers (Anderson 1994; Blitz and Lorenz 2006;
Williams 1994; Williams and Shapiro 1990), long distance exchange (King 2003), the
role of ideology and power (Pauketat 1994; Emerson 1997) and historic period shifts in
political economy (Worth 1988). To a lesser extent, studies of Woodland period political
economies also benefited from the regional perspective (Anderson 1998; Snow and
17
Stephenson 1998; Stephenson 1990; Stephenson et al. 2002). However, as noted by both
Anderson and Mainfort (2002) and Pluckhahn (2002), we know much less about these socalled “precursors” of Mississippian chiefdoms. For the Woodland period, this gap in
understanding is a result of both under-theorization (Cobb and Garrow 1996) and a lack
of data (Pluckhahn 2003).
Some of these difficulties stem from the often unacknowledged neo-evolutionary
implications of the chiefdom model itself. This particularly true of studies that apply
theories of human agency to Mississippian societies. While these theories undoubtedly
enrich archaeological interpretations, a focus on social inequality inevitably leads to the
investigation of short term processes such as chiefly succession, ideology, or leadership
(Pauketat 1994; Emerson 1997; Beck 2003). As a result, a balanced view of agency is
rarely achieved and the strategies of chiefly leadership are not fully contextualized by
structural constraints such as previously existing institutions – among the most important
of which are settlement, subsistence, and exchange systems.
Difficulties interpreting Woodland period societies are compounded a poor
understanding of inter-regional variation stemming from a persistent focus on the
regional and sub-regional scale. Regional studies recognize some variation from an
idealized chiefdom model. However, because regions are not addressed comparatively,
researchers overlook variation in the scale and distribution of settlements (but see
Anderson 1999; Knight 1997; King and Meyers 2002).
Regional studies often fail to adequately document the extent of interactions
between neighboring regions. Apart from the studies documenting the circulation of
18
prestige goods over long distances (for example, Muller 1997; Peregrine 1992), accounts
of inter-regional interaction between neighbors are rare. A compelling exception is
Anderson’s (1999) synopsis of Middle Mississippian collapse of Savannah River
chiefdoms. By adopting an inter-regional perspective, he presents correlations between
the abandonment of the Savannah Valley and changes among Late Mississippian
chiefdoms in the Oconee and Wateree Valleys.
Data comparability problems are one reason why other scholars may not have
followed Anderson’s example and conducted additional studies (Williams 1994 and
2000). Given the fact that Anderson and many others frequently focus on processes
spanning fewer than 75 years, comparability problems are significant. However, when
investigating long-term trends, it is often necessary to aggregate data, smoothing over
some short-term variation in service of understanding the long view.
A final limitation of short-term regional studies is an under-appreciation of the
role that local ecological diversity plays in shaping settlement. Human ecology models
are often applied across the Eastern Woodlands, blurring distinctions between diverse
ecological systems. And while there have been several recent calls to recognize the
importance of local diversity (Smith 2003; Wagner 2003), few studies have met this
perceived need through comparative studies relating political economy and ecological
diversity. One possible barrier thus far may be the fact that many ecological studies in
archaeology are species driven – related to specific processes that occur mostly at
localized scales (such as Scarry 1983).
19
This dissertation avoids some of the deficiencies often present in shorter-term,
regional studies that are driven totally, or in part, by neo-evolutionary theory. Broader
spatial and temporal scales, a focus on political and economic diversity, and the
comparison of landscape and archaeological data together more successfully describe
social variation and explain its origins. Using regional and macroregional data, a longterm, agent driven model of emerging complexity tacks between the “strategic conduct”
behind landscape use and utilitarian ceramic production and the “structuration” that local
anthropogenic landscapes and macroregional exchange networks impose (Giddens 1979:
54-59, 62-63, 66).
As noted above, this project addresses most or all of the Eastern Woodlands
occupation sequence, rather than focusing on a single period. This long-term view is
conducive to recognizing the constraining influence of long-term institutions. Theoretical
arguments for a longer view are supported by the contributions to method and theory
made by settlement pattern studies in the American Southwest (cf. Dean 1990; Fish et al.
1992) and Mesoamerica (cf. Kowalewski et al. 1989; Kowalewski et al. n.d.; Sanders et
al .1979), as well as existing calls for a southeastern renaissance of the long-term focus
(Elliott 2004; Kowalewski 1995 and 1996).
A multi-scalar spatial approach involving entire macroregions augments the longterm temporal view. Multi-scalar analysis is the process of integrating data from
investigations from sites, regions, and macroregions into a complete picture. Multi-scalar
analysis is commonplace among large scale surveys in highland Mesoamerica (cf.
Chamblee 2000; Kowalewski et al. 1989; Kowalewski et al. n.d.) Among the best recent
20
examples from North America are the aforementioned study of the Savannah River
Valley (Anderson 1999) and a parallel study on the Colorado Plateau (Wilcox 1999).
These studies stand out because 1) inter-regional interaction is a given 2) the authors
repeatedly change analytical scales to clarify patterns in the data, and 3) macroregional
and local patterns are related to the same sets of overall processes.
Macroregional analysis involves looking for patterns of change across multiple
regions by balancing the competing analytical goals of “retain[ing] variation as long as
possible” while simplifying and reducing variation to a “concept” (Kowalewski 2004),
such as market exchange, elite interaction, or warfare. Macroregional analysis requires
new approaches to method and theory (Feinman 1999:59). The few macroregional studies
that do exist (Hill et al. 2004; Smith 2002) rely on full-coverage, regional survey data – a
method that, for the southeastern United States, is often problematic over large areas, due
to the lack of surface visibility (Schiffer 1996).
While such data are still unavailable in the southeastern U.S., landscape ecologists
recently developed models that allow macroregional comparisons to include settlement
patterns, ceramic distribution shifts, and the landscapes in which they exist. These models
contradict earlier “steady state” models in which a given ecosystem had an “equilibrium”
or “climax” state that was altered by (usually human) disturbance (Gunderson and
Holling 2002; Holling et al. 2002). These new “alternative stable state” models recognize
that disturbance is part of an ecosystem and instead characterize landscapes according to
how they respond to structural changes (Beisner et al. 2006).
21
Together, multi-scalar, macroregional analysis and a long-term view make it easy
to emphasize variation, overcome neo-evolutionary assumptions, and employ a balanced
approach to strategic action and long-term structural constraint. To be deemed effective,
such a model must be tested in an appropriate setting and using an appropriate research
design. Said items are described in the next section.
Study Area and Research Design
The Chickasawhatchee Swamp’s ecological and prehistoric geopolitical setting
makes it an ideal place to investigate variation. Compared to areas used to develop earlier
models of Eastern Woodlands human ecology (cf. Smith 1978, 1989 and 1995), the
region is characterized by extreme heterogeneity. It is a mosaic of cypress bottoms,
upland long-leaf pine stands, and mixed pine and hardwood uplands in the center of the
Dougherty Plain –an area of karst topography surrounded by higher ground (Figure 1.2).
While the swamp is part of the interior coastal plain, it differs from the larger
physiographic zone in which it is embedded because of its lower elevation, greater
dependence on groundwater and more dispersed resource distribution structure (compare
Cammak et al. 2002 to Kirkman et al. 2000; see also Wagner 2003). Such ecological
diversity provides the means to ask whether heterogeneous landscape structures can lead
to differences in regional settlement patterns and political economy.
Although there was little systematic research in the Chickasawhatchee Swamp
prior to this project, several short studies suggest a settlement pattern history generally
comparable to other interior coastal plain regions. As seen elsewhere in the interior
22
coastal plain of Georgia (Fish 1976; Fish and Mitchell 1976; Fish and Fish 1977), an
abundant record of Archaic period land use is followed Woodland period settlement
consolidation of settlement and, in some regions or sub-regions, abandonment. These
preliminary data also suggest occupations from the early Middle Woodland through Late
Mississippian periods (see Figure 1.3). Several factors also set the Chickasawhatchee
Swamp apart and recommended it as a locus for intensively studying sources of variation.
FIGURE 1.2,
Dougherty Plain physiography
Brief or unpublished accounts by several archaeologists suggested more
Woodland and Mississippian period mound sites in the swamp than is typical elsewhere
in the interior coastal plain (Black 1977; Wauchope 1939; Worth 1989 – see Chamblee
and Snow 2002 for a summary). In addition, Hudson and his colleagues (Hudson 1997;
Hudson et al. 1984) and John Swanton (1985[1939]) both agree that Hernando de Soto
23
passed through the Chickasawhatchee Swamp in March of 1540. In spite of conflicting
interpretations regarding the actual polity through which the entrada passed (Capachequi
or Toa), both scholarly camps agree that the area was home to a large and visible
population. At the macroregional scale, the Chickasawhatchee Swamp’s central location
in the interior coastal plain also sets it apart. Located between several other well-studied
regions that vary greatly in terms of scale, complexity, and developmental trajectories
(Figure 1.4), the swamp is well suited for studies of inter-regional interaction (cf.
Schortman and Urban 1992).
FIGURE 1.3,
Mound centers and study areas in the interior coastal plain
Maximizing Chickasawhatchee Swamp’s promise as a study area required a
flexible research design that would accommodate both a long-term temporal perspective
and multi-scalar spatial analysis. Given the well-deserved reputation of regional surveys
for providing excellent long term views of change (Fish and Kowalewski 1990; Fish et al.
24
1992), pedestrian survey was selected as the centerpiece for project fieldwork. Intensive
site-scale mapping and testing programs supplemented the survey, established a
preliminary understanding of intra-site patterns at major sites, and provided additional
materials to build the regional ceramic chronology. Finally, a GIS built using data from
published reports and state site file offices provided a means for establishing a
comparative understanding of macroregional settlement pattern change.
The pedestrian survey covered 8.5 km2 of open fields and cleared areas, as well as
270 linear km of roads. A few weeks of reconnaissance survey with local landowners and
collectors supplemented the systematic surveys and established locations for some of the
largest and most significant sites. Crews located 259 archaeological sites, 238 of which
were previously unrecorded. One hundred and eight sites had Woodland or Mississippian
period components.
At the four sites the survey work revealed to be most significant and/or wellpreserved, crews executed shovel testing and mapping programs to get a preliminary idea
of site layout, site size, and intra-site artifact distributions. In areas with high artifact
concentrations or intact stratigraphy, crews excavated test units that increased our artifact
samples, provided additional information about formation processes, and lent key insights
regarding the construction sequences of mounds and structures.
Throughout the fieldwork, crews noted land use practices and their potential
impacts on site formation processes. Data from each scale of investigation were
integrated into a single relational database management system and GIS, thereby
maximizing analytical flexibility. These data were complemented by GIS data from the
25
Georgia Archaeological Site File and ecological and landscape data from a variety of
public sources. Database and GIS technology were integrated into both the fieldwork and
the later analysis, providing both maximum analytical flexibility and a firm empirical
grounding for later interpretive work.
FIGURE 1.4, Interior coastal plain regional ceramic sequences. 1) Worth 1998 2) Snow 1990
3) Worth 1988 4) Schnell and Wright 1993/ Blitz and Lorenz 2006 5) White 1981 6) Scarry 1985
7) Milanich 1994.
26
As with the fieldwork, data analysis and interpretation is multiscalar. Site scale
data focus on variables related to intensity of occupation and a limited analysis of
ceramic production. Regional data track changes in component size and site structure,
patterns of occupation continuity and abandonment, and the relationships between
settlement patterns and the possible long-term, anthropogenic changes in the landscape.
Macroregional comparisons emphasize the variables of site size and distribution and
ceramic frequency and diversity.
Chapter Summaries
The above summaries of research context, setting, and methods are, a good sketch
of the dissertations overall approach and implications. Unsurprisingly, many supporting
details remain. The chapter summaries below provide a more complete map of the overall
dissertation.
Chapter 2 presents the research context for the project, including theoretical
perspectives. Many concerns raised in this chapter are addressed in detail. A historical
history of theoretical developments in the southeastern United States, again notes the
impact that the chiefdom concept has had on efforts to integrate new theoretical
perspectives into research and the effects on Woodland period archaeology. An
alternative approach balances perspectives on agency and structure through multi-scalar
analysis. A narrative of the archaeology of the interior coastal plain and relevant nearby
regions highlights shifts in macroregional interaction. Finally, a discussion of new
theories from landscape ecology explains how such models can further the goal of
understanding variation in middle range societies through comparative research. A
27
description of the Chickasawhatchee Swamp’s ecology resource availability, emphasize
the potential of these new approaches.
Chapter 3 recounts the methods and middle range theories used in this study.
Definitions of archival research methods include approaches to GIS data collection, and
methods for documenting previous investigations and archaeological collections. Field
and laboratory method descriptions focus on regional survey, mapping, shovel testing,
and excavation, and soil sampling. A section on ceramic classification and categorization
describes methods and methods and outlines middle-range theoretic assumptions. An
overarching theme of Chapter 3 is the need to approach field investigations with the
understanding that data from multiple spatial scales must be integrated in single, flexible
analytical framework. This dissertation relies on standard techniques for statistical and
GIS analysis techniques, so discussions of quantitative method are abbreviated.
Chapter 4 presents site formation processes data from multiple scales. At the
regional scale, the chapter addresses bias introduced by differential survey coverage and
the destruction of sites by agriculture and silviculture. Regional data also show
differential visibility in full-coverage settings resulting from different field preparation
techniques and demonstrate similarities in artifact frequencies among recently disturbed
sites and those subject to repeated disturbance. At the intra-site scale, a comparison of
soil geochemical profiles, sedimentology, and the stratigraphic superposition of artifacts
provides insights into the nature of subsurface disturbance and provides new techniques
to detect such disturbance prior to beginning intensive excavations. Comparisons of
regional and intra-site data provide appropriate limits on the use of data.
28
In Chapter 5, data from regional survey, mapping, and testing in the
Chickasawhatchee Creek watershed provide a bottom up view of settlement variation.
Regional data describe settlement pattern change, continuity of occupation, and the
distribution of ceramic diversity. Intra-site data provide five examples of intra-site
variation, placed in context by the regional survey. Finally, region-wide analyses of the
relationships between settlement and ecology unite these data to create the outlines of an
economy characterized by long-term continuity in settlement patterns and ceramic
production. These data suggest a distributed economy that was not particularly
hierarchical, being instead dependent on the wild resources available in the mosaic of
small landscape patches across the project area.
Chapter 6 explores the political economy of the Chickasawhatchee Swamp
through macroregional analysis. Data from the Georgia Archaeological Site File show
long-term settlement pattern change according to physiographic province. Patterns of
concordant change strongly support the argument that the inter-regional interaction was
involved in the structuration of local political and economic shifts – especially during the
Mississippian period. These comparisons suggest long-term stability in settlement
strategies within a given area, while showing variation across physiographic provinces.
At the intra-site scale, shovel test data from sites in the Chickasawhatchee Creek and sites
in the fall-line hills show significant ceramic variation, suggesting differential
participation in the macroregional political economies and supporting the idea of a more
distributed economy in the interior coastal plain.
29
Chapter 7 is a summary and conclusion. This long-term view of variation in the
political economies of both the Woodland and Mississippian periods contributes to
broader arguments concerning variation among middle-range societies. Data suggest that
the Chickasawhatchee Swamp’s political economy was shaped by its landscape
heterogeneity and distributed resource base. Macroregional comparisons suggest that
such landscape patch dynamics are broadly influential – more so than previously
recognized. With regard to macroregional interaction, results strongly support Blanton et
al. (1997:vi) in their assertion that “global processes are, simultaneously and necessarily,
local ones.” By acknowledging that people interact at a scale beyond the region, it is
easier to see change as the outcome of global and local processes guided by strategic
action, and the historical constraints of unintended consequences.
30
CHAPTER 2: RESEARCH CONTEXT
This chapter presents the dissertation’s theoretical outlook and research goals. A
historical discussion of previous research contributions and the political economic and
environmental context highlight the need for theoretical contributions that recognize key
tensions between agency and structure. This chapter proposes a model that more
accurately describes political and economic variation through “bracketing,” or
independent views of strategic action and institutional change (Giddens 1979:95).
Methodological bracketing is achieved through a multiscalar research design that tacks
back and forth between broad and narrow geographical and temporal units (Kowalewski
1996:32).
Drawing on the work of Wiessner (2002), Kowalewski (1995, 1996 and 2004),
Muller (1997), and others (Anderson and Mainfort 2002; Nassaney and Cobb 1991 and
1996; Yoffee 1993), I argue that the prehistory of the eastern Woodlands of North
America is best understood through long-term, macroregional studies (Kowalewski
2004). Causes of regional shifts in political economy are varied (Earle 1997), but scope
of this study is limited to two variables – macroregional interaction and local ecological
variation. Due to its environmental and geo-political setting, the interior coastal plain of
southwestern Georgia is ideal for this analysis.
The first section of this chapter addresses the importance of long term,
macroregional approaches by demonstrating how short term, regionally focused, and
agent-driven views have limited intellectual progress in the investigation of eastern
Woodlands political economies. The following section considers the roles that
31
macroregional interaction and ecological variation can play in an overall approach to
balancing agency and structure. Throughout the chapter specific problems related to the
interior coastal plain and the Chickasawhatchee Swamp are woven into the narrative. A
summary of the issues presented here is addressed towards understanding how the broad
trends outlined in this chapter establish expectations for data from the Chickasawhatchee
Swamp.
Agency and Structure in the Prehistoric Southeast: The Legacy of Chiefdoms
The time periods of interest for this study span the Middle Woodland through
Late Mississippian periods (150 B.C. – 1600 A.D.). Archaeological evidence across this
broad sweep of time suggests dramatic shifts in social organization, including the
development of and collapse of the “Hopewellian Interaction Sphere” between A.D. 250
and 750 (Caldwell 1964), the expansion of Late Woodland populations between A.D. 700
and 900 (Cobb and Nassaney 1995; White 1981), the near simultaneous Mississippian
Emergence (Milner 1998; Pauketat 1994; Smith 1990), the reorganization of
Mississippian societies following the collapse of the largest centers by A.D. 1250 (Holley
1999), and European contact in 1540 (Hudson 1997).
Kowalewski (1996) is one of the most vocal advocates of the underlying
interconnectedness that runs through Eastern Woodlands societies across these
transitions. Through time, Kowalewski (2000:181) recognizes cyclic alternations between
“corporate” and “network” approaches to power in the political and economic strategies
of regional polities (cf. Blanton et al. 1996). His work, Pluckhahn’s (2002) recent multiscalar analysis of Kolomoki, make parallel research into subsistence patterns in the
32
eastern Woodlands (Scarry 1993; Smith 1995; Smith and Cowan 2003; Wagner 2003),
and Nassaney and Sassaman’s (1995) compilation of studies on macroregional interaction
stand out in their advocacy for or application of long-term, macroregional approaches.
Otherwise, research tends toward local and regional studies focused primarily on
the Mississippian period (e.g. Anderson 1994 and 1999; Emerson 1997; King 2003;
Moore 2002; Muller 1997; Pauketat 1994 Scarry 1996; Williams and Shapiro 1990), or,
less frequently, the Woodland period (e.g. Anderson and Mainfort 2002; Emerson et al.
2002; Nassaney and Cobb 1991; Williams and Elliott 1998). These studies have
contributed much to our understanding of Eastern Woodlands societies, but the lack of a
long term view has had negative effects. These effects have been compounded by the
neo-evolutionary intellectual legacy that is intertwined with the chiefdom concept
(Yoffee 1993; Yoffee et al. 1999).
The overwhelming focus on the Mississippian period has resulted in the need for
“far more problem-oriented and theoretically well-informed primary research… on the
Woodland period” Anderson and Mainfort (2002:540). This shortage of Woodland period
research is partially the result of the neo-evolutionary baggage of the chiefdom concept,
as scholars such as Peregrine exclude long term studies on the grounds that
“Mississippian societies represent the pinnacle of cultural evolution in eastern North
America” (Peregrine 1996:39 [my italics]).
When scholars overlook, or worse yet, endorse such linear thinking, the cycle is
perpetuated. In particular, Late Woodland period cultures are forced into “uniformity” on
the basis of an evolutionary stage model (e.g. Milanich et al. 1997:199). A similar
33
problem exists as scholars seek to trace out the spatial variation in complexity at the
“periphery” of the Mississippian world (cf. King and Meyers 2002). Milanich’s
(1996:255-256) typological treatment of the polities of Uzachile is a case in point.
Using the ethnohistoric documents from the de Soto expedition, Milanich
describes elite interactions between chiefs, or caciques, that closely mirror those of other
well-recognized chiefdoms – especially that of Ocute, in the present day Oconee River
Valley (compare Rangel 1993[1851]:264-266 with Rangel 1993[1851]: 272-273).
However, because the material culture of the area does not match the “Mississippian”
model, and perhaps also because of linguistic differences, Milanich (1996:256) argues
that Timucuan societies merely “acted” complex, “borrowing” appropriate chiefly titles
from adjacent Muskogean language families.
The borrowing of linguistic terms is a complex process, especially when the terms
in question reflect power relationships or ritual significance. The relationships between
Piman, Yuman, and Zuni languages during the Hohokam Classic period in Arizona are a
case in point (Shaul and Hill 1998). Zuni loan words are argued to be associated with
twelfth and thirteenth century increases in local shifts in Hohokam political hierarchies
and with increases in macroregional interaction (Shaul and Hill 1998:389).
However, these shifts did not simply reflect that some groups were merely
becoming more hierarchicial. Widespread religious practices, such as the kachina cult,
may have also helped to spread loan words without necessarily altering political
hierarchies (Shaul and Hill 1998:389). Returning to the southeastern case, it would be a
mistake to simply assert that “less complex” cultures borrowed from more complex ones,
34
unless one adopts, a prior, a model that demands the exclusion of cultures insufficiently
“complex” to merit the term “Mississippian.”
In recent publications, problems with the chiefdom concept become more
complicated. The most significant difficulty revolves around how to draw the line
between chiefdoms and everything else when one’s project area or period of study rests at
the edge of the so-called “Mississippian World.” Recent work by King and Meyers
(2002:114) illustrates this difficulty. While they attempt a theoretically neutral approach,
avoiding “terminology explicitly linked to any single explanatory framework,” their use
of the chiefdom model still leads to difficulties (King and Meyers 2002:113).
In particular, Meyers (2002) uses settlement patterns and subsistence in her
attempts to separate polities in that were located in present-day southwestern Virginia
into “chiefdom” and “non-chiefdom” classes. Although the data she presents provide a
compelling case for complexity outside the normally defined boundaries of
“Mississippian world,” she ultimately returns to Milanich’s (1996) aforementioned
model, in which groups in some river valleys are argued to be merely imitating the
behaviors of their more complex neighbors (Meyers 2002:188).
Apart from this conclusion, Meyers (2002:185-186) goes to great lengths to
demonstrate the importance of variation in middle range societies. She marshals
ethnohistoric and archaeological data to demonstrate the importance of salt and gorget
exchange, providing excellent support for the roles that global and local variables play in
structuring political economic changes (Meyers 2002). In the end, it is the chiefdom that
35
fails to account for variation in southwestern Virginia – not the data themselves, or the
broader arguments for which they are used.
In addition to problems accounting for variation, a shortage of Woodland period
research and the chiefdom concept itself lead to difficulties regarding the Mississippian
emergence and change through time in Mississippian period societies. Since its inception,
definitions of “chiefdom” presented by Sahlins (1958) and Service (1962) have
undergone almost constant revision. Modifications to the chiefdom model attempt to
improve its ability to explain not only the emergence of hereditary social inequalities, but
also variation in complexity and scale. Even as broader trends in anthropological theory
came and went, these two goals remained – trapped by the chiefdom’s place in an
evolutionary stage model.
Researchers initially explained the emergence of chiefdoms in “voluntarist,
adaptationist” terms (Earle 1997:68), arguing that genealogically-based leadership offices
grew from a need for increasing productive and redistributive efficiency as a response to
population growth (Service 1962:143-144). Over time the list of external stressors grew
to include warfare and competition (Carniero 1981; Fried 1967; Renfrew and Cherry
1986), increasing loads on information exchange networks (Earle 1987; Peebles and Kus
1977) and changes in subsistence (Kelly 1990; Welch 1990).
Approaches to variation founded on systems theory (e.g Carniero 1981;
Steponaitis 1978) proved successful in driving settlement pattern research (for example,
Anderson 1994; Hally 1994; Smith and Kowalewski 1980; Williams 1994a). But, as
these early models were criticized for being overly typological (Anderson 1994; Yoffee
36
1993; Feinman and Neitzel 1984; Yoffee et al. 1999), models grounded in multiple
theories of causation, elite interaction, and agency theory emerged (Anderson 1994;
Emerson 1997; Pauketat 1994; Peregrine 1992). Factors previously viewed as “external
stressors” are now seen in agent-based models as structural constraints on elite behavior.
Earle (1997:203-211) synthesizes these points of view, arguing that elites first
accumulate power in one of three realms of social interaction: physical force, ideology, or
political economy. Over time, elites manipulate their initial power base to consolidate
power in the other two realms. Earle’s model also emphasizes natural and historical
circumstances as constraining factors – a perspective consistent with other theories of
power in its emphasis on the dialectical relationships between agency and structure
(Giddens 1979; Blanton et al. 1996; Kowalewski 2000). Many southeastern scholars
accept Earle’s model, at least implicitly (e.g. Pauketat 1994; Schroeder 1997; Milner
1998; Scarry 1999; King and Meyers 2002; King 2003).
Debates focusing on the relative importance of structure, force, ideology, and
political economy in the emergence of chiefdoms overshadow questions of variation.
Specifically, investigations into leadership strategies (Pauketat 1994 and 2002; Emerson
1997) chiefly succession (Knight 1990; Anderson 1994), prestige goods exchange
(Peregrine 1992), and craft production (Welch 1981; Pauketat 1997; Anderson 1999), all
reflect an agent driven focus. This focus necessarily involves what Wiessner (2002:234)
refers to as an “omission of structure” that simultaneously imposes a “slate of simplicity”
on egalitarian social formations while obscuring “booms and crashes in the
archaeological record.”
37
As analysis of variation again becomes a major research focus (e.g. King and
Meyers 2002), and archaeologists push the geographic and temporal limits of knowledge,
the neo-evolutionary baggage of the chiefdom will need to be addressed by a more
balanced approach towards agency and structure. Processes that result in variation occur
at different spatial and temporal scales, so the most complete explanations will come
from studies that recognize the tension between agency and structure (Kowalewski 1995
and 1996; Wiessner 2002:234). Multi-scalar analysis provides the methodology to
separate strategic behavior of agents and structural constraints imposed through the
reproduction of institutions and then re-unite them in a complete analysis (cf. Giddens
1979:95).
Investigations into agent-driven behavior require data from areas with refined
chronologies, preferably at the level of local communities or sites (Pauketat 1998; Mills
2000). These data must come from intensive and focused excavation of specific sites.
Even when such data exist, the conclusions drawn from them are often equifinal. A recent
example is the string of recent publications addressing Mississippian leadership
strategies.
Pauketat (1994:184-185; 1997:11) and Emerson (1997:265-267) concentrate
heavily on the role of a universalizing ideology that supported a strongly centralized,
even authoritarian, political economy structure that had a profound, if short-lived, effect
on the entire population of Cahokia’s hinterlands. In contrast, Muller’s (1997:27)
historical materialist approach argues against such a structure, asserting that differences
between elites and non-elites are differences of degree, not of kind (Muller 1997:386 and
38
397). In the end, the scalar differences between Pauketat’s site distribution data and
Muller’s macroregional comparisons between distant sites cannot be reconciled to
achieve a resolution (compare Pauketat 1994:108-112 to Muller 1997:Figures 8.2, 8.7,
and 8.8). In addition, arguments like these are further confused when unrelated, but
similarly untestable theories are built around existing discourses concerning “variation
among chiefdoms” (e.g. Beck 2003).
In contrast to agency-driven investigations, archaeological research into structure
can be enhanced by the inherent long-term character of archaeological data. Long-term
structural constraints guide the strategic actions of agents and can even limit the scope of
their goals (cf. Bourdieu 1990:64-65; Giddens 1979:66-68; Kowalewski 2000:179).
Structure can be successfully investigated using data from limited test excavations,
regional survey, and state site file data (Anderson 1999; Kowalewski 1996; Williams
2006). However, archaeological investigations into structure in the southeastern United
States have been complicated somewhat by conflations of temporal, political, and
economic classifications.
Figure 1.4 presents traditional ceramic phase sequences for the regions around the
Chickasawhatchee Swamp. Except for the overlap between the Averett and
Standley/Rood phase sequences (Schnell and Wright 1993), these sequences do not
typically account for the ceramic complexity of the coastal plain. In their overview of the
interior coastal nine zones in the Georgia Coastal plain, Schnell and Wright (1993:46)
make an argument for a situation in which the ceramics suggest interaction between
“contemporary Mississippian and Woodland cultures.” While the wording of this
39
observation reveals a broadly culture-historical approach (see Kowalewski 1996:29 for
associated problems), the observation is a valid recognition of a long-standing problem in
archaeology.
Our understanding of temporal and spatial ceramic variation is always
complicated by the fact that we simultaneously use ceramics to order space and time and
understand behavior (Blinman 2000). The reverse is, of course, also true. Below, I
mention several other cases in which seemingly different “cultures” seem to co-exist at
nearby settlements (e.g. Elliott and Wynn 1991; Gougeon et al. 1996; Pluckhahn 1997).
Using ceramics to tease out these relationships would require ceramic analytical methods
based on some conception of technological style (e.g. Dobres and Hoffman 1994;
Hegmon 1998). Typically, such methods require ceramic samples from independently
dated contexts (Gosselain 1998; Mills 2002; Van Keuren 2001).
Rather than add to the confusion by creating another list of proposed phase
sequences and cultural affiliations, I will generally refer to differing blocks of time using
the period designations on the far left hand side of Figure 1.4. However, I also recognize
that, from A.D. 900 to 1200 (and, in some areas, beyond) there are multiple cases in
which so-called “Late Woodland” settlement clusters are contemporary with those from
the “Early Mississippian” period. Since a suitable resolution of this issue is beyond the
scope of this research, I am forced to compromise by occasionally using period
designations as shorthand for political economic strategies.
My hope is that, by raising these issues explicitly, and airing my misgivings, I can
contribute to a more long-term view of middle range variation in spite of my inability to
40
deal with this broader issue. As I hope to show below, there is, at the very least, a strong
argument for the hypothesis that, irrespective of production context, some ceramic
decorative techniques are associated with specific social venues for inter-regional
economic and political interaction. These associations form the basis for my reluctant use
of the terms “Mississippian” and “Woodland” political economy.
In spite of the above limitations, I demonstrate that, by focusing on structural
constraints in a given region and connecting them to concordant changes at the
macroregional scale (Kowalewski 1996:159), we can gain insights into trends guiding
variation. In the remainder of this section, I briefly present a subset of the neglected
questions of political economy that can be fruitfully investigated through a structuredriven approach. I also suggest ways in which these trends can be understood by focusing
on the crucial variables of interregional interaction and landscape variation.
Global and Local: Landscape Structure and Macroregional Interaction
Kowalewski (1995, 1996:32, 2000, 2004) has repeatedly argued that southeastern
United States has functioned as a single interconnected society, or “world system,” since
the Paleoindian period. This assertion is supported by data from Anderson (1995) and
arguments by Anderson and Sassaman (1996). Instead of attempting to demonstrate
Kowalewski’s argument, I take it as a starting point and show how broader contexts
provide avenues for investigating issues related to major shifts in regional political
economies.
I first consider the question of macroregional interaction across the Woodland
through Mississippian periods. Published studies from each of these periods show
41
consistent evidence for macroregional interaction – though the context of this interaction
may change. In considering landscape variation, I note the strong evidence for regional
variation in overall settlement models, particularly in the coastal plain (e.g. Pluckhahn
and McKivergan 2002; Wagner 2003; White 1981).
Macroregional Interaction: The Thread of Continuity
Evidence for macroregional interaction and the synchrony between broad scale
and local shifts in political economy comes primarily through the demonstration of
“concordant change” (cf. Kowalewski 1995; Smith 2002). Concordant change is defined
as patterns in archaeological data demonstrating that “during a relatively brief time
horizon social groups over a wide area alter their form and activities as part of a single
process” (Kowalewski 1995:159). Interaction and interdependence are inferred from
concordances in material culture shifts, rather than observed directly (Kowalewski
1995:159). Concordant changes occur at a variety of spatial and temporal and spatial
scales. By comparing long-term and short-term trends, it becomes possible to see the
tensions between agency and structure. In Georgia, structure takes the form of continuity
in macroregional interaction and exchange across changing contexts of political
economy.
The Middle Woodland period emergence and dissolution of the Hopewellian and
Swift Creek interaction spheres and the Late Woodland period expansion and
reorganization of settlement are major trends involving macroregional interaction.
Caldwell (1964:138) was among the first to note remarkable similarities among funeral
practices and burial items from the mid-continent to Florida. These similarities in funeral
42
practices coincided with “striking regional differences in the secular, domestic, and nonmortuary aspects” of material culture (Caldwell 1964:138). Jeffries (1976 and 1979)
record a similarly limited phenomenon at Tunacunnhee – as does Smith (1979) at
Mandeville. However, Anderson (1985) notes that evidence for participation in Hopewell
interaction along the Atlantic slope is limited to these areas.
In central Georgia and Florida, evidence suggests that limited participation in
Hopewell-related exchange came to an end by A.D. 350 (Pluckhahn 2003:182). In its
place, there apparently arose a different exchange system involving the exchange of both
eccentric and utilitarian vessels. Some items, such as the effigy vessels associated with
Kolomoki and McKeithen (Milanich et. al 1997:172-183; Sears 1956:58-65) have clear
funerary associations and can fairly be argued to be part of a ritual exchange network
(Pluckhahn 2003).
However, the widespread use and exchange of Swift Creek Complicated Stamped
utilitarian wares suggest additional interactions (Snow 1975; Snow and Stephenson
1998). Populations aggregated at mound centers such as Swift Creek and Hartford, as
well as several sites in the Lake Seminole all suggest that an exchange network was
established by around A.D. 250 or 359 (Jeffries 1994; Milanich et al. 1997; Snow and
Stephenson 1998; White 1981). These sites often have large U-shaped shell middens,
suggesting dependence on aquatic resources (Stephenson et al. 2002).
Stoltman and Snow (1998) combined petrography with Snow’s initial study of
imperfections within the paddles used to decorate Swift Creek pottery, demonstrating that
both pots and paddles were moving across the landscape. This being so, it is likely that
43
there were broader exchanges of material wealth and perhaps spouses involved in
contemporaneous exchange revolving around mortuary activities (Snow and Stephenson
1998:109-111; Stoltman and Snow 1998:152-153). This overall pattern is similar to the
finance-driven exchanges that Wiessner (2002) documents as associated with the ritual
exchanges involved in the Tee Cycle and Great War rituals.
The origins of these exchange systems and their accompanying spheres of
interaction are poorly understood, as are the reasons for their decline and the collapse of
this system between A.D. 600 and 750. Most explanations (e.g. Jeffries 1994; Peregrine
1992) treat the Middle Woodland period as a “precursor” to Mississippian polities and so
do not problematize the collapse. However, Pluckhahn’s explanation of the collapse of
Kolomoki is a significant exception.
In short, Pluckhahn (2002:218) argues that the ritually based politically economy
at Kolomoki and the mediating roles it may have played “collapsed under the weight of
its own success.” Though Pluckhahn does not mention the parallel, this explanation is
broadly similar to that provided by Wiessner (2002:246) for the collapse of the Great War
rituals among the Enga. Pluckhahn’s approach is also significant because it posits a
pattern of concordant change – associating the decline in importance of certain ritual
groups with documented Late Woodland period shifts in settlement structure.
Along with the decline in mound ceremonialism and the extent of a ritually-based
exchange network, Late Woodland trends suggesting discontinuity are population growth
(Kirkland 1994:72; Milanich et al. 1997; Pluckhahn 1994 and 2002:304; White
1981:650-654), and regional diversification (McElrath et al. 2000; Williams 2006). In the
44
interior coastal plain, Weeden Island wares and other decorated types fell out of
production, and plain, cord marked, simple stamped, and check stamped pottery rose in
frequency (Kelly et al. 1962; Milanich 1994:350; Stephenson 1990; White 1981:652; but
see Williams 1994b:135-136).
In spite of the regional diversity evident in Late Woodland ceramic traditions
(Williams 2006:185), evidence for spatial and temporal continuity exists. In particular,
Cobb and Nassaney (1995:211) note evidence for the exchange of shell ornaments and
drinking vessels, as well as utilitarian items, such as ceramics and lithics (see also Cobb
and Garrow 1996; Rudolph 1991). This kind of utilitarian exchange occurred during the
Middle Woodland period as well. These exchanges call into question arguments by
authors such as Peregrine (1992:40), who states the Late Woodland societies were
affected by a Middle Woodland period “dissolution of the highest level of the political
hierarchy.” Instead, we might wonder whether or not changes in political structure, or
leadership strategies (Blanton et al. 1996; Kowalewski 2000), involve changes in
archaeological visibility (cf. Nassaney and Cobb 1991:5) that – to those who wish to find
it – look like broader changes in settlement hierarchy.
In many areas, so-called “Late Woodland” societies are contemporaneous with
even the latest developments in the Mississippian political economy and European
contact (Elliott and Wynn 1991; Emerson et al. 2000; Milanich 1994:350-352; Pluckhahn
1996 and 1997; Worth 1996; Worth and Duke 1991). In particular, Pluckhahn notes
(1996) that evidence for the “Late Woodland” occupation at the Tarver Site is
contemporary with developments at Macon Plateau, only a few kilometers away. The
45
differences between the ceramic traditions at these neighboring sites are significant and
indicate that stylistic divergences are likely a matter of choice (Dobres and Hoffman
1994).
Along with population growth and settlement expansion, evidence for more
widespread warfare is an additional change that occurred in the context of continued
interaction – a factor that Kowalewski (1996:30) argues to be pivotal in the emergence of
Mississippian polities. The location of many Late Woodland/Early Mississippian
societies in a hilltop setting and the high frequencies of small triangular arrow points
support this point of view (Elliott and Wynn 1991; Worth 1996; Worth and Duke 1991).
Additionally, the co-occurrence in contemporary contexts of decorative types
alternatively characterized as Woodland or Mississippian supports the idea of intense
interaction between groups (Pluckhahn 1997:87; Stephenson et al. 1990:57; Worth
1996:55-57).
Evidence for warfare also is found at sites around the fall line near Macon and in
the Chattahoochee River. By A.D. 900 the large multi-mound center of Macon Plateau
had been established in central Georgia. Williams (1993:72) speculates that Brown’s
Mount, located on a high promontory, may have been a short-term defensive site
contemporary with Macon Plateau. Around A.D. 1100 a palisaded village, Cool Branch,
was established about the same time as Singer Moye (Blitz and Lorenz 2006). Similar
sites exist throughout northwestern Georgia (Cobb and Garrow 1996).
Overall, this evidence supports arguments by Cobb and Garrow (1996) that, for
much of the early portions of the Mississippian period, groups participated in the
46
macroregional Mississippian political economy to a varying degree – some adopting the
material trappings Mississippian practices wholesale, while others retained relative
independence through maintenance of regional styles in their utilitarian ceramics – not to
mention warfare. It is against this complex backdrop of population growth, settlement
expansion, and interregional interaction (both amicable and otherwise) that the
“Mississippian Emergence” took place.
Large polities, such as Cahokia, Moundville, and Macon Plateau existed primarily
during the Early Mississippian period and most had collapsed by A.D. 1250 (Knight
1997; Knight and Steponaitis 1998; Pauketat 1994; Williams 1994c). Some research
suggests that new exchange networks likely played a major role in the initial growth and
contraction of Mississippian polities (e.g. Brown et al. 1990; King 2003; Peregrine 1992;
cf. Wiessner 2002). However, given the evidence for warfare during this period, it is
certain that this played a role as well.
In the Middle Mississippian period, more local groups adopted material goods
reflective of participation in the larger political economy and practices such as mound
building and settlement consolidation expanded in a context of continued population
growth. However, the expansion of the Mississippian political economy marked the
beginning of cyclic abandonment at local centers across present-day Georgia. As noted in
the introduction, one consequence of this expansion was circumscription of the polities in
the Savannah River, the eventual abandonment of the area as a result of intensifying
competition and warfare, and a dramatic expansion in the neighboring Oconee Valley and
47
beyond (Anderson 1994:326-328; Kowalewski and Hatch 1991; DePratter 1994;
Williams 1994a).
In the Chattahoochee Valley, the overall Mississippian occupation sequence
consists of the cyclic abandonment and re-occupation of multiple centers through time
(Blitz and Lorenz 2006). Many centers are located a short distance from one another,
exhibiting the paired towns model first recognized by Williams and Shapiro (1990). Like
Williams and Shapiro (1990:164), Blitz and Lorenz (2006:140) argue that these paired
towns were sequentially occupied. But rather than citing environmental variables, the
latter authors argue for intra-polity competition, which they term the “fission/fusion
process” as an explanation (Blitz and Lorenz 2006:140). In contrast to some piedmont
areas, the Chattahoochee Valley was occupied throughout the Woodland and
Mississippian periods.
In the Apalachicola River Valley, platform mounds appear to have directly
followed the Late Weeden Island population expansion around A.D. 1000 (Brose and
Percy 1978:97; Scarry 1990:234-235). The persistence of Late Weeden Island ceramic
types, such as Wakulla Check Stamped, provides evidence for the continuity in the
Apalachicola Valley as well (Scarry 1990:236-237). Given the ongoing difficulties with
chronology in this area (Blitz and Lorenz 2006:104), I cannot speculate on the nature of
these continuities, but Milanich (1994:360) suggests that the expansion of Apalachicola
populations into the Tallahassee Red Hills eventually led to the development, by A.D.
1100, of Apalachee – one of the largest chiefdoms of the late prehistoric period (Scarry
1990:243).
48
On the northern edge of the interior coastal plain, along the middle Flint River
valley we see evidence of a valley-wide abandonment similar to those of the Georgia
piedmont (e.g. Hally et al. 1990). Two mound sites were established in the Middle Flint
River Valley around A.D. 1150. The area was abandoned by A.D. 1200, re-occupied a
century later, and abandoned again after the de Soto entrada in A.D. 1540 (Worth
1988:162-172). Similarly, Mississippian occupations of the Ocmulgee Big Bend Region
are suggestive of an expanding and contracting frontier. Late Woodland ceramics and
settlement patterns persisted until A.D. 1200. Mississippian polities that Stephenson and
et al. (1996:30-31) characterize as “marginal” then appear. By 1400, the only mound
center among these polities, Sandy Hammock, was abandoned (Stephenson et al.
1996:30). After a population downturn, Late Mississippian settlement shifted southwest
and concentrated along trails and rivers (Snow 1990:84-85, 90-91).
Taken as whole, the varying event sequences in Georgia regions are consistent
with Knight’s (1997) view of macroregional change among Mississippian chiefdoms.
Knight (1997:244) argues rapid growth, greater investment civic-ceremonial
construction, and more centralized control of iconography all in earlier polities. He
(Knight 1997:245) also argues that later polities appear to be shorter-lived. This view is
also consistent with Kowalewski’s (2000) arguments for shifts between corporate and
network strategies among earlier and later Mississippian polities. However, when
comparing the piedmont and the interior coastal plain, there are differences, as adoption
of the Mississippian political economy seems to have been later in coming and, generally
speaking, seems to have resulted in the abandonment of fewer entire valleys.
49
Across the Woodland and Mississippian period, the causes behind shifts in
settlement patterns and local political economies are multivariate – including the rise and
decline of certain ideologies, population growth, shifting subsistence patterns (discussed
below), and warfare. These causes work together to create dizzying variation in local
event sequences and political economies. However, running through every region in
every time are the threads of interregional connection. Irrespective of temporal and
spatial context, the evidence for connections between groups is strong.
Although preciosities are often the most frequently invoked evidence for interregional contact and were addressed here, the preceding discussions also suggest strong
associations between mound building and the practices of inter-regional contact. The
social contexts of mound building vary greatly through time (Hally 1996; King 2003;
Milanich et al. 1997; Pluckhahn 2003), but empirical data broadly support linkages
between mound-building, mound-ceremonialism and inter-regional contact (Jeffries 1976
and 1979; Milanich et al. 1997; Muller 1997; Pauketat 1994; Peregrine 1992; Pluckhahn
2003; Smith 1979). In a sense, mounds are the evidence of inter-regional contact, hiding
in plain sight.
In Elson’s (1998:22, 27, 56, and 67) comparative review of mound
ceremonialism, he briefly mentions historical and archaeological examples of
associations between the members of societies involved in trade and those using and
building mounds. However, he says little about the topic and even overlooks associations
between mound use and trade among Woodland and Mississippian period societies
(compare Elson 1998:34-36 with Hudson 1976:66 and 89). Broadly speaking,
50
associations between inter-regional exchange and mounds in the Mississippian period
were built around the office of the chief (King 2003:121-123). However, observations
that most items associated with the Hopewell interaction spheres were found in mounds
suggest a more long-term, structural link between the two (Caldwell 1964; Hudson
1976:66; Jeffries 1976 and 1979; Smith 1979).
The absence of mounds does not preclude inter-regional contact. In some areas,
specifically in the interior coastal plain, sites located near major historic trails suggest
some groups were involved in such exchanges without having built mounds (Kirkland
1994; Snow 1990). Nevertheless, mounds provide evidence for inter-regional contact is
specifically tied the shared beliefs and ritual practices associated with Hopewellian,
Weeden Island, and Mississippian material culture traditions.
In this light we see spheres of interaction, not merely as another trait to be added
to the list of qualifications necessary to be designated “complex” or “Mississippian,” but
as a structuring principle of eastern Woodlands macroregional interaction. That these
connections are evident from the Middle Woodland through the Late Mississippian
periods reminds us against framing the emergence of Mississippian chiefdoms against a
Woodland period “slate of simplicity” in which all inter-regional interactions did not
occur and local societies were egalitarian (Wiessner 2002:234). Throughout time, groups
in different regions participated in macroregional exchange spheres selectively and
strategically.
In part, their strategies may certainly have been defined by the affordances
provided by local systems. The differences between piedmont and interior coastal plain
51
settlement sequences support this hypothesis. In an effort to understand this possible
source of variation further, the subject of landscape variation is taken up in the next
section.
Landscape Structure: The Interior Coastal Plain and the Chickasawhatchee Swamp
It is a truism that landscape structure impacts settlement patterns (Fish and
Gresham 1990; Fish et al .1992; Kowalewski et al. 1989; Kvamme 1992; Warren 1990).
The reverse of this truism also holds (as demonstrated by Fish et al. 1990; Flannery 1969;
Smith 1978 and 1994; Spores 1969), but until recently, investigations of the recursive
impacts of anthropogenic changes in landscape structure have been limited by theoretical
and methodological constraints. Theoretically, archaeologists have been limited by the
unwillingness of colleagues in the ecological fields to accept the ubiquity of prehistoric
anthropogenic disturbance. Methodologically, we have been limited by our ability to
process large, detailed data sets concerning landscape structure.
Early, “multi-disciplinary” archaeological projects involved collaborators from
many fields – each investigating their own topics. Conclusions from geologists,
ecologists, and other natural scientists were used in formulating anthropological theory,
but the published evidence suggests a one-way interaction in which archaeological data
do not have much impact on conceptions of the landscape among the “hard sciences”
(Trigger 1989:280-282).
In North America, many ecological researchers, in particular, have been immune
to the idea of prehistoric impacts on the landscape. It is possible or even likely that their
blindness comes from the strength of the myth that much of the New World was an
52
unspoiled wilderness prior to 1492 (Denevan 1992). In ecology, the impact of
philosophers like Roderick Nash (1967) carved a central place for the very Euro-centric,
American, and middle-class concept of wilderness (Gomez-Pompa and Kaus 1992). The
wilderness myth successfully kept ecologists from thinking about Native American
anthropogenic disturbance, as they consistently cited an article built around the
unsupportable argument that “there is no ethnographical evidence that the Eastern
Woodlands Indians were sufficiently organized to carry out systematic burning of a large
area” (Russell 1983:85).
Recently, however this trend has begun to change. Biogeographers have
successfully used paleo-ecological reconstructions of the Georgia piedmont to
demonstrate that disturbance regimes changed from prehistoric fire-driven disturbance to
a post-Contact disturbance regime of human clearing in the context of overall fire
suppression (Cowell 1998). In these same studies, ethnohistoric and archaeological
evidence is used to support of the idea of anthropogenic fire (Cowell 1995 and 1998;
Denevan 1992; Goebel et al .2001; Kirkman et al. 2000). Cowell, (1998:86) in particular,
cites the likelihood of anthropogenic fire regimes to argue against using modern mature
ecosystems as models of prehistoric ecosystems – an argument supported by studies of
historic Creek land use (Ethridge 1996; Foster 2001; Foster et al. 2004).
At the same time, archaeologists have begun recognizing the importance of
anthropogenic landscapes and are finding archaeological evidence to support the role of
fire in creating large-scale disturbance (Heckenberger 1998; Heckenberger et al. 1999;
Heckenberger et al. 2003; Wagner 2003). These related developments have been driven
53
by the emergence of landscape ecology (Wagner 2003:127), the development of GIS
technology (Kramer et al. 2003; Noss 1987), and the growing recognition of disturbance
regimes and “alternative stable states” in ecological systems (Beisner et al. 2006;
Gunderson 2000; Holling and Gunderson 2002).
In her survey of these changes in anthropological theory Wagner (2003:133-135
and 154) notes the importance of disturbance regimes in explaining macroregional
ecological variation and of fire as an agent for anthropogenic disturbance. Moreover, her
overview of variation in disturbance regimes across the southeastern United States raises
important considerations for models of human/environment interactions in the interior
coastal plain.
Among the long-leaf pine stands of the Florida sand hills (an environment similar
to southwestern Georgia), Wagner (2003:140) notes strong associations between
permanent water sources, archaeological sites, and long-leaf pine stands that typically
thrive in a disturbance regime characterized by regular, low-intensity fires. This model
varies strongly from the floodplain weed theory summarized by Smith (1995) to explain
plant domestication and accepted by others as a standard model for human/environment
interactions. I do not dispute the argument that the floodplain weed theory is applicable to
explanations of domestication. But, in light of the variation in landscape structures and
disturbance regimes discussed by Wagner, general models for human/environment
interactions should be replaced with regional models of anthropogenic landscapes that
account for local variation in disturbance regimes and landscape structure (e.g.
Heckenberger 1998).
54
In this study, anthropogenic landscapes are considered in relationship to two land
use strategies – a general model for landscape clearing (e.g. Smith 1995 or Ethridge 1996
and Foster 2001) and the exploitation of key resources to maximize opportunities for
limiting risk and building or consolidating power (e.g. Earle 1997). These strategies are
generally complementary and create an anthropogenic landscape that serves to structure
shifts in political economy. Recursive relationships between landscape structure and land
use strategies can be defined through comparisons highlighting the capacity of different
ecosystems to resist and/or “bounce back” from disturbance.
The ecosystem properties of resilience and resistance are defined in terms of
processes and energy needed to shift between or maintain alternative stable states
(Chapin et al. 2002; Gunderson 2000; Holling and Gunderson 2002). Resistance, also
called engineering resilience, is the ability of an ecosystem to maintain similar overall
structure in the face of a disturbance (Chapin et al. 2002: 282; Holling and Gunderson
2002:28). By contrast, resilience, or ecosystem resilience, measures the magnitude of
disturbance at which ecosystems can still maintain their overall function, irrespective of
changes in structure. A resistant ecosystem will maintain overall structure, even if
functional outputs, such as biomass, net primary production, or nutrient outputs change.
By contrast, a resilient ecosystem will maintain constant levels of productivity, even in
the presence of significant changes in species composition and overall structure.
Many ecosystems in the Georgia coastal zone and interior coastal plain exhibit the
characteristics of resilient ecosystems. Landscape characterization studies of mixed longleaf pine and depressional wetland habitats suggest a landscape matrix characterized by
55
small, heterogeneous landscape patches (Kirkman et al. 1998). In these systems,
disturbance regimes suggest a low-intensity, ongoing interplay between regular fires and
flooding. Fire frequency and available water plays important roles in determining
ecosystem structure (Kirkman et al. 2000). These landscapes are characterized by high
species richness and biodiversity (Kirkman et al. 1998:7-8), traits of resilient landscapes
(Gunderson 2000:431).
Although detailed landscape characterization studies of the Georgia piedmont are
difficult to come by, two lines of evidence support arguments for a contrast between the
landscapes of the piedmont and those of the coastal plain. Kramer et al.’s (2003) recently
completed GAP analysis of Georgia includes measures of species richness across the
state. While the Fall Line, interior coastal plain, coastal zone are among the richest areas,
the piedmont – and especially the Savannah and western Oconee River Valleys are
among the least species rich areas. While modern ecosystem and disturbance regimes are
different from those of the prehistoric period (Cowell 1998), analyses of land-lot surveys
suggest a pre-settlement piedmont landscape that is more homogeneous than that of many
parts of the interior coastal plain (compare Cowell 1995:Figure 5 to Cammack et al.
2002).
In a resistant landscape, such as that of the piedmont, the establishment of cleared
land would have required greater initial investment, but probably would have yielded
more permanent results. Smith’s (1995) general model for floodplain agriculture fits well
with this resistant landscapes model and the overall disturbance regimes in the Georgia
piedmont, but less so with the interior coastal plain. A comparative view of landscape and
56
prehistoric settlement regimes highlights probable differences in land use strategies and
establishes expectations for settlement variation.
In most well studied prehistoric southeastern woodland regions, large patches of
floodplain were exploited for horticulture and, later, intensive agriculture. Wet-adapted,
open-air invasive species were developed into seed-based cultigens and then later
supplanted by corn (Smith 1995). In these relatively homogeneous landscapes,
anthropogenic fires could create permanent patches of disturbance, effectively converting
subsets of the eastern Woodlands to agro-ecosystem landscape patches. In this context,
isolated wetlands were exploited to produce additional resources that may have been used
to augment social status and attract followers (Pauketat 1994:61-62; Schroeder
1997:226). This overall production model is applicable to many regional cases in the
eastern woodlands (Smith 1978:484-486).
Mound centers emerge in areas of increased resource diversity (Hally 1994;
Junker 1999; Schroeder 1997). Increased diversity generally occurs along transition
zones between ecologically distinct landscape patches (Forman and Godron 1986:104105). In the Mississippian macroregion, mound centers located in such transition zones
are often preceded by large Late Woodland period (A.D. 750 - 1000) sites that appear
during episodes of population growth (Anderson 1994; Emerson 1997; Hally 1994; Kelly
1990; Milner 1998; Pauketat 1994; Scarry 1983).
According to Smith (1978:484-486), the use of riverine floodplains and adjacent
wetlands is a defining characteristic of the Mississippian mode of production. These
transition zones provide key opportunities for emerging leaders. In the lower southeastern
57
United States, wetlands are favorable locations for mound centers (King 1996; Smith
1994; Williams 1999; Williams and Shapiro 1990;). Although Pauketat and Schroeder
disagree over broader interpretations concerning sociopolitical development, they concur
that concentrated wetland resource zones were critical to the growth of chiefdoms in the
American Bottom. They argue that such zones provided resources that emerging elites
could have exploited to attract followers (Pauketat 1994:61-62; Schroeder 1997:226).
Schroeder (1997:210) provides specific support for this assertion, demonstrating
that nucleated mound centers grow where patches of deep wetlands, shallow wetlands,
and dry land exist in equal proportions within two kilometers of a center (Schroeder
1997:162-165). Beyond the two-kilometer catchments, these wetlands are spatially
constricted resources (Milner 1998: Figures 2.2-2.13; Schroeder 1997:35). There are only
a few transition zones between wetlands and dry ground because the wetlands are
embedded within a dry land system that forms the matrix, or dominant ecosystem type,
for the region (Forman and Godron 1986:159).
However, comparisons between chiefdoms in these typical riverine settings
described above and those in which wetlands are not spatially restricted provide insights
into how wetlands contribute to settlement pattern variation. Among Mississippian
chiefdoms in the Georgia coastal zone, Pluckhahn and McKivergan (2002:157) note
settlement dispersal. They hypothesize that this pattern resulted from the interruption of
continuous uplands amenable to settlement by expanses of swamp. Hypothesized
consequences of dispersal are greater household autonomy and less elaborate social
control mechanisms (Pluckhahn and McKivergan 2002:156-157).
58
I agree with Pluckhahn and McKivergan’s (2002:156-157) overall argument that
landscape structure shapes settlement patterns, but I argue that it is both the ubiquity of
wetlands and the lack of dry land that encourages settlement dispersal. Also accepted are
Pauketat (1994:61-62) and Schroeder’s (1997:226) arguments that, where wetlands are
spatially restricted, they can act as concentrations of diverse aquatic resources that elites
will seek to control for staple or wealth finance (e.g. Earle 1997:70-75), or ideological
power (e.g. Kikuchi 1976). In contrast, landscapes with dispersed wetlands make this
diverse resource base available to everyone and discourage their use as a power base. In
her survey of the present-day Lake Seminole area, White (1981:659) notes the heavy
exploitation of wetland resources at dispersed Late Woodland period sites.
In addition to the wide availability of wetland resources, the overall disturbance
regimes of the coastal plain would discourage the overall land use models similar to those
of the piedmont. Small landscape patches are more easily disturbed, but, in spite of
structural changes, maintain consistent ecosystem function in terms of production and
output. These contrasting responses to disturbance would be responsive to different kinds
of resource exploitation strategies.
Establishing permanent resource exploitation zones in the resistant landscapes of
the piedmont would require sustained efforts by large numbers of people. By contrast,
dispersed populations working small landscape patches would be most successful in
exploiting the patchy landscape of the interior coastal plain. In contexts where
populations did aggregate, sustained disturbances by large aggregate groups would result
in more rapid changes in ecosystem structure. While these changes might not impact the
59
objective measures of ecosystem productivity, the more rapid onset of changes would not
be sustainable in terms of human land use because of the rapid onsets of an alternative
stable state unsuitable for human exploitation (Holling and Gunderson 2002:44-45).
Together, the wide availability of wetland resources and the distributed nature of
upland landscape patches suitable land-based resource exploitation strategies encourage
settlement dispersal. Dispersed wetlands allow for risk-minimization behaviors, but
preclude the control of aquatic resources. At the same time, the ease in creating upland
disturbances discourages settlement aggregation. In this context, we would expect local
differences in the political economy similar to those observed by Pluckhahn and
McKivergan (2002). When such trends are investigated in the context of macroregional
interaction, the interplay between global and local processes becomes evident.
Summary: The Chickasawhatchee Swamp as a Test Case
This chapter has considered approaches for understanding variation in middle
range societies. I have argued that a long-term macroregional approach is necessary to
untangle some of the Gordian problems associated with the development of inequality
among Woodland and Mississippian period societies and to restore a more balanced view
of the tensions between agency and structure in driving social change.
In addition to demonstrating the importance of further divesting the chiefdom
concept from its neo-evolutionary heritage, I have also shown that inter-regional
exchange is one of the most significant constants among Eastern Woodlands societies. By
contrast, I have demonstrated the ways in which local variation in landscape structure can
provide the impetus for differential settlement and land use strategies.
60
The Chickasawhatchee Swamp is the ideal setting for investigating the interplay
between these complementary variables. The region is centrally located among nine other
well-studied regions of late prehistoric activity (Figure 1.3). Located in the center of the
karst geologic landscape of the Dougherty Plain (Figure 1.2), the swamp is the epitome of
a species-rich, patchy, heterogeneous, and resilient landscape – thus providing a contrast
to most well studied eastern Woodlands areas. Sample elevation profiles from the Oconee
River and Chickasawhatchee Creek drainages illustrate just how substantial these
differences are.
Figure 2.1 presents a comparative sample of elevation profiles from 20 km easttransects in both drainages. In addition to obvious differences in absolute elevation,
uplands, slopes, and drainage bottoms are much more pronounced in the Oconee Valley
and occur in a somewhat uniform, repeating pattern. Sharp elevation shifts strongly
impact landscape patch dynamics. In transects perpendicular to the watercourses,
contrasts exist between upland, slope, and bottomland ecosystems. However, along
parallel lines to the narrow waterways, ribbon-like landscape patches are uniform.
61
Piedmont and Coast Plain Topographic Profiles
200
180
Elevation (masl)
160
140
120
Oconee
100
Chickasawhatchee
80
60
40
20
0
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
Distance (km, east to west from the line's origin)
FIGURE 2.1.
Comparative topographic profiles of piedmont and coastal plain landscapes
In the Chickasawhatchee, the differences are more subtle. The extreme flatness of
the Dougherty Plain results in wide, slow, stream channels. The underlying karstic
geology also creates many sinkholes, which are later filled in and inundated to form a
patchwork of small, circular wetlands. This patchwork creates a very heterogeneous
landscape. Ecosystem types include regularly inundated bottomlands dominated by water
tupelo and cypress; occasionally inundated hardwood bottoms dominated by various oak
species; open-water, herbaceous ponds; large patches of longleaf pine; smaller patches
pine/hardwood mixed areas; and a few patches of fire intolerant upland hardwood stands
dominated by oak, sweet gum, and hickory (Cammack et al. 2002).
My own field observations suggest that the flooded areas are characterized by
moist, mucky soils and are completely unsuitable for human habitation. In the uplands,
62
each of the remaining patches is characterized by small size, dispersed nature, and thus,
their susceptibility to disturbance. The wetlands would have provided abundant aquatic
resources, but in the context of the larger coastal plain, intense burning in the uplands
would likely have encouraged the spread of long-leaf pine habitat (Wagner 2003), rather
than created an agro-ecosystem landscape (as in Smith 1995).
Given the study area’s geo-political and ecological setting, several expectations
emerge. The long-term narrative suggests that not all of the regions around the
Chickasawhatchee Swamp participated in macroregional interaction spheres. Thus, we
should expect that, during the Woodland or Mississippian periods, there was some
selective participation in these interactions in the Chickasawhatchee Swamp as well. The
landscape in the swamp suggests further variation.
Evidence for a different overall settlement strategy includes a dispersed settlement
pattern. Such evidence is necessarily multi-scalar – involving data from sites within the
Swamp, from the region as a whole, and from comparisons with neighboring areas.
Regional settlement differences alone are not sufficient. At the site scale, we should
expect less intensely occupied or smaller settlements. The exact means for testing these
hypotheses and establishing linkages with archaeological data are addressed in the next
chapter.
63
CHAPTER 3: METHODS AND MIDDLE RANGE THEORY
In this chapter, I present my approaches to archival and museum research, field
methods, and artifact classification, categorization, and analysis. Throughout the project,
I was mindful of the fact that data from multiple spatial and temporal scales would be
necessary to accomplish my goals of understanding regional and intra-site formation
processes, reconstructing regional settlement history and political economy, and placing
the Chickasawhatchee Swamp in macroregional context. In other words, my research
design was built around the premise that one needs a multi-scalar system in order to study
multi-scalar systems. The overall project design, to the extent possible, interweaves data
from multiple scales to form an integrated, multi-scalar whole. In many cases, the
answers to research questions appropriate to one scale are at least partially dependent on
the outcomes of research at greater and lesser scales.
The chapter opens with a description of archival research, geographic information
system (GIS) assembly, and the analysis of existing collections. Survey methods and
sampling techniques for intensive intra-site mapping and testing are then described. I then
discuss ceramic analysis and middle range theory, addressing issues of classification and
chronology, site formation processes, and the regional ceramic economy. Relational
database design formalized key assumptions underlying research and the issue is
addressed as appropriate. Methods used for macroregional comparisons are addressed at
the beginning of Chapter 6.
64
Archival Research, Existing Collections, and GIS Development
The initial phase of research for this project consisted of archival research,
ceramic classification of existing collections, and assembly of a GIS. Very little is known
about the project area, but several scholars recognized its potential importance. Both
Hudson et al. (1984) and Swanton (1985[1939]) hypothesized that Hernando De Soto
crossed the Chickasawhatchee Swamp in March of 1541, though these sets of scholars
disagreed as to whether this crossing involved contact with inhabitants of the polity of
Capachequi (Hudson et al. 1984) or Toa (Swanton 1985[1939]:173-174). Only 31
Woodland and Mississippian period sites were systematically recorded and only one
civic-ceremonial center (9DU6) had been mapped and tested (Chamblee and Snow 2002).
This section addresses each of the steps involved in the archival research stages of
the project. Archival research provided crucial preliminary understandings of regional
settlement patterns and ceramic and lithic variation. Analysis of existing collections and
field notes provided a preliminary view of intra-site structure and highlighted some of the
potentially more important sites in the project area. Assembly of GIS data helped raise
important questions regarding landscape variation.
Existing Settlement Pattern Data and Previous Investigations
Regional settlement pattern data were collected from the Georgia Archaeological
Site File and the Archaeology Laboratory of South Georgia College. In both cases, the
focus was on obtaining location and size data for as many sites as possible, while at the
same time identifying key civic-ceremonial centers as candidates for intensive mapping
and testing. Only 15 Woodland period and Mississippian period sites were recorded at the
65
Georgia Archaeological Site File and only 17 at South Georgia College. Three of the four
sites ultimately subjected to intensive testing were identified through archival research
(9BX4, 9DU1, and 9DU6).
In the 1930s, Robert Wauchope recorded few sites in the Chickasawhatchee
Creek drainage, including Magnolia Plantation (9DU1). Wauchope’s (1939) minimal
notes are summarized in the in a progress report for the WPA. Don Smith also recorded a
number of sites during the early 1960s. Tallassee Plantation (9DU22) is a large mound
center that has since been revisited by researchers focused on the Hernando De Soto
expedition (Hudson et al. 1984), but is now closed to researchers. Smith also first
recorded Windmill Plantation (9DU6). In December of 1962, Smith conducted 10 days of
excavations here, leaving collections unanalyzed at the University of Georgia
(Laboratory of Archaeology n.d.).
The only field notes from Smith’s excavations are those present on the field
collection bags. These notes are confusing, but suggest that Smith and others excavated
between five and ten test units of unspecified horizontal extent. The units were excavated
in six inch levels, usually to a depth of 18 in. Only one unit, EXU 10, was excavated
somewhat deeply, to a depth of 40 inches. Judging from the location of an unfilled trench
and associated backfill piles, most units were excavated into the mound’s eastern flank.
Assuming Smith used a method favored by his supervisor, Dr. A.R. Kelly, these units
were excavated in shallow, stair-step fashion down the slope of the mound (cf. Chamblee
et al. 1998:Figure 6).
66
In 1976, Frankie Snow and Chris Trowell, both of South Georgia College, visited
the Chickasawhatchee Knoll site in response to information that the site had been logged
and was being intensively collected by local hunters. Snow and Trowell made a general
surface collection of the entire site (Snow n.d.). A few days later, Snow and Trowell were
joined by Dr. Craig Sheldon and Ray Crook of the State Archaeologist’s office. Sheldon
et al. made a field map using a compass, tape, and level (Sheldon, personal
communication 2002) (shown in Chapter 5 as Figure 5.27). In addition to mapping the
site, this team conducted 17 randomly distributed collection circles, recorded on the bags
as “units,” across the site (West Georgia College n.d.). Surveyors exercised a 100%
pickup strategy within each five meter in diameter circle (Sheldon, personal
communication 2002). In addition to these collections units, Sheldon’s team created a
general surface collection comprised of many diagnostic sherds and a reconnaissance
collection of the steep, eastern slope of the site (West Georgia College n.d.).
In 1978, Snow and Trowell returned to the area and surveyed another largely area
of recently cleared timber (Snow n.d.). This was a 100% pickup, full-coverage transect
survey, recording 13 sites with ceramic components (Snow, n.d.). Between 1978 and
1994, Snow made several additional visits to the area and recorded other sites through
discussions with local collectors and informants. Included among these sites are the Long
Walk Site (CAS 162), Pot Handle (CAS 156), and a non-collecting visit to Magnolia
Plantation (9DU1) (Snow n.d.).
67
Classification of Existing Ceramic Collections
Ceramic collections from sites in Chickasawhatchee Swamp were curated at
South Georgia College, the State University of West Georgia, and the University of
Georgia. The 14 ceramic period sites housed at South Georgia College provided a sample
of about 1500 sherds. The 13 sites recorded in 1978 exhibited 544 sherds while 998
sherds were collected from Chickasawhatchee Knoll (CAS 89). All project area materials
from the State University of West Georgia are also from Chickasawhatchee Knoll,
including 1450 sherds from 17 of the 18 aforementioned collection units. The eighteenth
unit was comprised of lithics only. University of Georgia collections from Windmill
Plantation (9DU6) include 274 sherds, most of which are presumably from the mound’s
eastern flank.
I classified and categorized each of these collections before going into the field.
My goals were to sample the project area’s ceramic variation and build a classification
model for future analysis. This early attempt at ceramic classification and the final
analysis conducted on my field collections are comparable. However, as these collections
are reported with the overall body of dissertation data, a few differences are worth noting.
I classified existing collections analysis using a few decorative type categories
that, upon reflection, should either be discarded entirely, or reflect mis-classification. In
the former category are Lake Jackson plain and Lake Jackson Incised, both of which can
be subsumed into the category Lake Jackson Decorated. In the latter category are
Alachua Cob Marked and Leon-Jefferson Check Stamped. The former is likely Dunlap
Fabric Marked sherds and the latter should simply be lumped in with Wakulla Check
68
Stamped. I include these data as is in this dissertation, as I do not wish to discard my
original classifications without re-examining the collections. Data on vessel and sherd
size were not collected during the classification of existing collections.
GIS Data Assembly
Archival research on settlement patterns also included obtaining GIS data for
archaeological site locations. Other GIS data sets obtained from the state of Georgia and
the U.S. Geological Survey included 30 m resolution raster data on modern land cover,
30 m digital elevation models (DEMs) for the entire project area, three meter resolution
digital ortho-quarter quads around known mound sites, digital raster graphics of all
1:24,000 scale USGS topographic maps, and 1:24,000 scale vector data describing
modern roads, political boundaries, streams, and topography. All GIS data for the project
area are maintained in Transverse Mercator projection, UTM Zone 16, using the 1983
North American Datum (NAD_83).
The collection of GIS data was an ongoing process. Midway through the 2004
field season, ortho-rectified historical aerial photography became available and is used to
verify the existence or date of destruction for mound centers that could not be
independently investigated in the field (see Chapter 5). In addition, I collected digital
copies of several maps from Georgia Archives These maps show historic trails that likely
coincide with paths used prior to European contact. Late maps also show sources of
historic disturbance, such as the approximate date of a 19th century railroad construction
project that partially destroyed one of the mounds at 9DU1.
69
Local researchers from the Albany Office of the Georgia Department of Natural
Resources (DNR) and the Joseph Jones Ecological Research Center (henceforth, the
Jones Center) ultimately provided the most in terms of useful GIS data. The DNR
provided one meter resolution false color satellite imagery and up-to-date vector data
describing the roads on the Chickasawhatchee Wildlife Management Area (henceforth,
the WMA). In addition, Jones Center scientists had already undertaken a number of
projects on the WMA and provided two important data layers – a detailed, field-tested
soil map and a proposed ecological restoration model based on historic vegetation data
and soil chemistry. These layers in particular proved crucial in understanding the
relationships between local settlement and ecology and are described here in greater
detail.
Figures 3.1 and 3.2 provide detailed views of the soil data and the ecological
restoration model provided by the Jones Center. Both models reflect the intense
complexity of the Chickasawhatchee Swamp landscape and divide soil and vegetation
into a wide variety of categories. For the purposes of this project, many of these
categories are lumped into larger classes. The rationale for this simplification follows the
settlement pattern models of Johnson et al. (1988), in which broad classes of soil,
vegetation and slope were tested against site location data to develop predictive models to
guide future surveys. Johnson used three broad landscape categories. However,
Gremillion (2003) notes that, in regions with low relief, minor shifts in elevation can
result in dramatic changes in vegetation. Since the Chickasawhatchee Swamp is a low
70
relief area with dramatic ecological boundaries, I use only two physiographic zones,
uplands and lowlands, to stratify data for chi square testing.
The soil map is the result of direct field investigations. While Natural Resource
Conservation Service data did serve as a starting point, field checking revealed that, when
county-by-county soil maps were digitized and edge-matched, there were errors. In the
end, Jones Center researchers reclassified soils using field samples that focused on the
color and texture of the Bt horizon, as well as the thickness of A and E horizons and the
depth to soil saturation (Cammack et al. 2002). There are 11 soil classes, but, for this
study, soils are divided into upland and lowland categories and the upland into split into
three subgroups, based on the primary soil component (Figure 3.1). Upland soils consists
of clays (shown in dark orange), sands (in beige), and sandy clay loams (in light orange).
Bottomland soils consist of seasonally saturated organic bottomlands (in green), poorly
drained bottomland transitions (in grey), and fully submerged soils (in blue).
71
FIGURE 3.1,
Grouped soil classes on the Chickasawhatchee Wildlife Management Area
The Jones Center ecological restoration model is designed to approximate
conditions scientists believe may have existed in the swamp prior to American
colonization in the 1820s. This model was created by comparing general classes of forest
cover observed on the 1819 land-lot surveys to known vegetation regimes correlated with
the soil types described above (Cammack et al. 2002). Nineteenth century land survey
72
plots are suitable for several kinds of analysis related to historical ecology. The Georgia
land lottery system was established in a regularized grid. As these grids were established
in the field, surveyors marked lot edges in the field by blazing marks on “witness trees.”
Typically, there are12 such trees per lot. The general tree species (pine, oak, walnut, etc.)
each witness tree was recorded on plat maps for the surveyor maps and in surveyors’
notebooks. The locations of rivers, swamps, trails, contemporary Native American
settlements, and existing European settlements were recorded as well (Cowell 1995; Goff
1975).
In spite of its sophistication, application of the Jones Center restoration model
requires some caution. The model incorporates assumptions concerning ecological
succession patterns and soil/vegetation correlations that could be disrupted by human
activity (Katherine L. Kirkman, personal communication, 2004). So, some correlations
between aboriginal settlement and vegetation distributions in the model may reflect
model assumptions, rather than historical patterns. Nevertheless, as a means to provide
hypotheses from which to generate additional tests, the model is crucial.
The final model (Figure 3.2) shows landscape patches as divisible into six classes
that are lumped into upland and bottomland categories. Upland categories included
longleaf pine forest, mixed fire tolerant pine and hardwoods, and fire intolerant upland
hardwoods. The last category is particularly interesting because these areas occur as
islands surrounded by much moister terrain and vegetation. Bottomland categories
include oak dominated hardwood bottoms, cypress and tupelo dominated open water
swamps, and open water areas dominated by grasses.
73
FIGURE 3.2,
Grouped vegetation classes on the Chickasawhatchee Wildlife Management Area
Field Methods
Field investigations were conducted to obtain information on regional settlement,
intrasite variation, and, in a few places, site stratigraphy and deposition processes. Data
concerning ceramic distributions at various scales were collected using three techniques:
surface collections areas, shovel tests, and stratigraphic excavation unit levels. Each of
74
these units was employed with the awareness that data from each scale would have to be
quantitatively integrated.
Data integration of this kind is not uncommon. Archaeological research in
Georgia provides a dramatic example. Before Lake Oconee was built, a full coverage
survey of the flood pool was integrated with ecologically stratified survey transects and
numerous intensive intrasite excavation projects (Depratter 1983; Fish and Hally 1983;
Fish and Gresham 1990; Fish and Jeffries 1983; Smith 1995; Williams 1983). In later
years, these projects have been complemented by numerous upland surveys (Chamblee
1997; Elliott 1981; Williams 1993 and 2000), as well as reports concerning excavations
at every major mound center in the valley (Smith 1995; Williams 1990a, 1990b, 2003a,
2004; Williams and Shapiro 1990). Excavation projects were also undertaken within
prehistoric villages, special use sites, and farmsteads (Hatch 1995; Stephenson et al.
1993; Williams 1990c, 1992a, 1992b, 2003b, and 2005; Williams and Kowalewski 1989;
Williams and Shapiro 1990b). Over time, the long-term research design in the Oconee
Valley came to be driven by the understanding that the Valley represented a politically
independent, if demographically integrated part of the overall settlement system of the
late prehistoric Southeast (cf. Anderson 1994; Kowalewski and Hatch 1991; Smith and
Kowalewski 1980; Williams 1994a). Published research from the Oconee Valley brings
home the point that data from all parts of the settlement were necessary to understand the
complex changes in the region.
Although a small project like this one cannot reproduce or even approximate the
richness in data and interpretation provided by decades of research, the project does
75
follow the example of Oconee Valley researchers by seeking to investigate the
Chickasawhatchee Creek watershed as an integrated whole linked by processes at
multiple scales (cf. Kowalewski et al. 1989). While the work of integrating data from
various scales is often accomplished through painstaking reclassification or re-entry of
data, the relational database design for this project allowed for data integration on the fly
and made the process of ensuring data comparability part of the project’s normal work
flow (Figure 3.3). In the project database, data structures directly reflect field methods at
various spatial scales. These rules also reflect the understanding that archaeological sites
are altered many times after they are abandoned by their original occupants (cf. Schiffer
1996).
FIGURE 3.3,
Entity Relationship schematic for the project database
The specific steps taken to implement these guidelines are two-fold. Each method
for collecting data (surface survey, testing, and excavation) was given its own set of
tables and fields. These tables were tied back to a master table of archaeological sites.
However, regardless of the actual field methods employed, artifact collection units were
all treated the same in the sense that all ceramic classification records were stored in a
76
single table, regardless of how they were collected. Ceramic classification records, or
critical analytic units (CAUs) as Murphy (2002) calls them, correlate with the piles of
sherds we make in the process of classifying pottery from a single collection unit or
provenience. Each pile, or CAU, is tied back to its provenience, or collection unit through
a single lot number stored in the CAU table. Regardless of scale, every separate
collection unit is assigned a lot number. The storage of the lot number in the ceramic
records table and the tables linking lots to collection units ensure a permanent link and
referential integrity (cf. Arizona State Museum 2001).
In addition, while various collection units relate back to sites, these units do not
relate directly back to temporal components. Since ceramic components use artifact
analysis to model the estimated spatial extent of past occupations, these models are stored
in a separate table. Because all collection units and components are tied back to sites,
data from a variety of scales can be integrated or disaggregated to accommodate a variety
of analytical goals. Not only does this overall approach streamline analysis, it also
encourages researchers to think about their final data analysis while still in the field.
Regional Survey
Kowalewski (1990:79) argues that only full-coverage surveys can address the
broad range of theoretical problems relevant to studies of regional settlement patterns,
noting that “[s]pecification of spatial relationships, network characteristics, objective
complexity, and unanticipated variability is not feasible with current sample survey
methods.” Support for this assertion comes from the American Bottom where Schroeder
(1997: 253) notes that “[p]erceived clusters and gaps in site distributions are related in
77
part to differential survey coverage.” While these arguments are valid, they now apply
mostly to questions of settlement arrangement – that is the relationship of settlements to
each other. For a variety of reasons, questions of settlement distribution, or the
relationship between settlements and external variables, such as those related to the
landscape, are more manageable.
Advances in GIS technology, the recording of survey coverage limits with GPS
devices, and the availability of remotely sensed environmental data now provide fine
grained views of the relationships between survey coverage and the environment,
allowing for a detailed understanding of potential bias. With these data in hand,
distributional questions are accessible. Beyond these advances, sample data still provide
at least preliminary measures of settlement pattern data that are not dependent upon
relationships between settlements. Foremost among these measures are occupation size,
duration, and intensity. Although the data are in some ways in complete data, insights
into settlement pattern change is still possible when we acknowledge the potential for
undocumented complexity and variation.
Survey teams recorded 259 archaeological sites, 104 of which have ceramic
components relevant to this dissertation. Figures 3.4 and 3.5 show the location of all
survey transects and recorded sites in the project area. Regional survey strategies in the
Chickasawhatchee watershed represent a compromise between the sometimes competing
goals of full-coverage and reconnaissance survey. Good regional settlement pattern
studies need accurate characterizations of spatial and temporal site distributions sites and
of the relationships between site locations and landscape variation. By contrast,
78
macroregional comparisons often place increased emphasis on the upper levels of
settlement hierarchies, so it is critical that larger sites be located. These competing goals
were accomplished through a program of high intensity survey, low intensity survey and
site visits led by local informants.
Following the terminology of Wandsnider and Camilli (1992) this project was a
“site survey,” meaning that survey efforts were directed toward the discovery of large,
obtrusive clusters of artifacts, or sites. A site is defined as a scatter of at least five
artifacts surrounded by a 100 m buffer zone without artifacts. Exceptions to this rule are
sites with between one and four sherds and temporally diagnostic projectile points found
in isolation (of which there are only a few).
In all survey contexts, every team had a least one outdoor sports-grade, Garmin
GPS unit in constant operation. Linear track logs were used to record transects, and
points were used to mark component boundaries, overall site boundaries, and piece plot
(within a 10 – 30 ms) the location of temporally diagnostic projectile points. Surveyors
carried photocopies of 1:24,000 quad maps, as well aerial photographs, and plotted site
information directly on these copies.
At the end of each day, GPS data were downloaded and plotted into the project
GIS, printed, and annotated for future analysis. Annotations included summaries of field
notes and hand-drawn site boundaries. Collection strategies varied by artifact type. We
collected 100% of all ceramics and temporally diagnostic stone tools. However,
collections of expedient stone tools and production waste were limited to specimens
79
judged in the field to be representative in terms of size, source material, and treatment
(heated or not).
Red lines on Figure 3.4 and red polygons on Figure 3.5 indicate transects and sites
from high intensity surveys. Crews surveyed 8.5 km2 using full-coverage methods
common throughout Georgia (cf. Elliott 1981; Chamblee 1997). Recognizing the high
probability and of encountering very small lithic and ceramic scatters, as well as the need
to maximize diagnostic artifacts recovery, we employed sweeping 15 m transects, instead
of the more common 30 m separation (Schiffer et al. 1978:6). We marked site edges and
diagnostics with orange pin flags, laying down additional boundary nodes with each
transect. Because of the large size and complexity of many sites, we did not plot
boundaries with the GPS units or collect artifacts until all site and collection area
boundaries were fully established with pin flag markers.
Given available time and resources, it was not practical to follow Wandsnider and
Camilli’s (1992:184) recommendation that transects during site discovery be reduced to
five meters. However, once site boundaries were delimited, areas within the sites were
carefully swept at one to three meter intervals to ensure that smaller and less obtrusive
artifacts were collected – thus guarding against within-site bias against unobtrusive
artifacts (Wandsnider and Camilli 1992:180). Finally, crews made careful notes
concerning surface collecting conditions so that we could test the effects of the
“background noise” (cf. Wandsnider and Camilli 1992:182) on artifact frequency
distributions.
80
Black lines on Figure 3.4 and black polygons on Figure 3.5 indicate low intensity
survey transects. Only a portion of the Chickasawhatchee Creek drainage is under
cultivation. To compensate for the poor visibility in the region, we surveyed
approximately 270 linear km of roads and fire breaks. Visibility varied from excellent to
non-existent. Low intensity survey was carried out only by crew members with two years
or more of survey experience. Although we tried to handle site boundaries and artifact
collection areas as we did in the high-intensity survey, we had to estimate the site
boundaries extending beyond the transects. Estimations are conservative. Surveyors
based site boundary estimates on the shape of the landforms shown on the quad maps and
orthophotos. Generally, we did not extend site boundaries more than 30 m beyond the
road boundaries unless we encountered additional areas of disturbance or a continuous
artifact distribution that continued around the bend of a road.
Beyond the systematic surveys, I spent a total of three weeks touring the project
area with local collectors and landowners. These tours provided the means to establish
ties with local landowners, gain hands on experience concerning landscape variation, and
look specifically for large sites with monumental or intact above-ground architecture. As
noted by Schiffer et al. (1978:16), locals often know a good deal about the archaeology in
their region. In southwestern Georgia, where private property dominates and local
landowners and agricultural technicians are avid artifact collectors, this is doubly true.
This situation, when coupled with the pervasive damage to the archaeological record,
means that local information is often the best means of locating significant and intact
sites.
81
FIGURE 3.4,
High intensity (red) and low intensity (black) surface collection survey transects
82
FIGURE 3.5,
Sites located via high intensity (red) and low intensity (black) survey
83
While three of the four sites selected for excavation and testing were recorded in
the Georgia Archaeological Site File, these sites were re-located only with the help of
local assistance. A matter of greater importance was the process of obtaining permission
to work (Schiffer et al. 1978:9). In southwestern Georgia, where communities are
somewhat distrustful of outsiders, the process of obtaining permission to work was much
more intensive than expected and would have been nearly impossible without efforts to
connect with local social networks (see Elliott 2004).
Windmill Plantation is an example of the importance of relationships with local
landowners. The site was originally recorded without the benefit of USGS topographic
maps to pinpoint location. As a result, it was mis-plotted by more than a mile and was
located only after I and the landowner spent half a day traversing his property in search of
what he viewed as “likely spots” for a mound. In the end, we discovered the mound in
area that he thought to be a “long-shot,” but worth trying. Had it not been for the fact that
I was introduced to the owner of Windmill Plantation by a land manager with whom I
was already working, it is likely I would not have found 9DU6 on Windmill Plantation.
Together, the high-intensity and low-intensity approaches provided a good sample of
geographic and formal diversity at a low cost. However, the benefits of strong ties with
the community cannot be understated in terms of either archaeological data recovery, or
personal job satisfaction.
Intensive Mapping, Testing, and Excavations
Site-scale investigations produced intra-site data from a few key sites to better
understand the regional settlement system. At each site there were five goals: 1) define
84
the site boundaries; 2) create detailed maps of above ground architecture, if any; 3) build
a map of intra-site ceramic distributions; 4) use selective test excavations to understand
local and regional stratigraphic trends; and 5) obtain large representative ceramic samples
to assist in building a regional chronology. At each of the four sites where we conducted
systematic intensive fieldwork, methods were a compromise of these goals and available
labor. Broadly, the methods employed here follow those developed by Mark Williams
(1999) for rapidly assessing the boundaries, intrasite patterns, and artifact distributions of
sites likely to have significant deposits, such as mound centers.
A Note on Regional Formation Processes
Schiffer (Schiffer et al. 1978; Schiffer 1996:259) notes three variables that affect
the degree to which survey data are representative of regional archaeological settlement
patterns. As the preceding discussions suggest, the obtrusiveness of archaeological
materials, their visibility on the ground surface, and their accessibility to archaeologists
were all at play in shaping the sample provided by the final project data set. I addressed
the issues of obtrusiveness and accessibility through narrow survey transects and
intensive efforts to build relationships with local landowners and informants. The issue of
visibility is more complex and thus addressed in Chapter 4.
Site Mapping
Site maps were created with GPS receivers and survey instruments. Initial site
reconnaissance and confirmation of site boundaries were generally recorded with GPS
receivers. At sites investigated during 2003, I used an outdoor-sports-grade Garmin GPS
III to produce preliminary boundary maps that. In the summer of 2004, I used a
85
differentially corrected Trimble GeoXT, which provides precision within about a meter.
Topographic maps and detailed maps of above ground architectural features were
produced with a Sokkia Set 6F Total Station.
Total Station maps were developed using one of two strategies. At sites with large
mounds or structures, extra care was taken to map architecture. Readings were taken
approximately every meter across the entire mound and any disturbance or special
variation was recorded with additional, closely-spaced points. In these cases, topographic
features away from the mounds were mapped using 16 radial lines extending away from
the mounds. Along these lines, readings were taken every five meters. This technique was
applied at Windmill Plantation (9DU1) and the Red Bluff Earthlodge (9BX4), where two
sequences of radial lines were employed – one from Structure 2 and one from Structure 3.
At Magnolia Plantation, radial mapping away from the mounds was curtailed by
by the onset of deer hunting season, which brought to an end our permission to work.
Time and personnel constraints prevented the establishment of a single datum tying the
whole site together. Instead radial lines from each mound tie back to their individual
datums on the summits of Mounds A and B. Locations for these datums were later
recorded using the Trimble GeoXT. The final map of Magnolia Plantation (presented in
Chapter 5) is a composite of two total stations maps. The maps were pulled into a
composite using the GPS datum readings, then aligned correctly using angle
measurements between the mounds and the railroad via Total Station and using the
Trimble GPS. While this map should be refined in further investigations, it provides an
86
adequate view of intrasite patterns at Magnolia Plantation. At Hay Fever Farm (CAS245),
elevation readings along the radial lines were spaced one meter apart.
Shovel Testing
Recent objections to shovel testing notwithstanding (Kvamme 2003), the
excavation of systematic transects of small (between .3 and .5 m) test units across sites or
landscapes remains a trusted and trustworthy approach to documenting archaeological
site locations in the southeastern United States. A comprehensive shovel testing program
at Fort Gordon, Georgia, located 1,336 temporal components across just over 11,000
hectares, or at a rate of 8 temporal components per hectare (Benson 2003). Over the
numerous projects Mark Williams has undertaken to document mound sites in the
Oconee Valley and elsewhere (see the many references in the introduction to the field
methods section), a systematic grid of shovel tests to determine artifact frequency
distributions is always a first step in research. When data from these simple probes are
subjected to spatial analysis, patterns relevant to site structure often emerge (cf.
Pluckhahn 2003). This project was too small to undertake a region-wide program of
shovel tests, but a future program of regional shovel testing in areas unsuitable for
surface survey could contribute to a more unbiased view of settlement variation.
During this project, shovel testing at intensively investigated sites in primarily
directed towards defining site boundaries, detecting broad intra-site patterns and defining
areas of high artifact frequency – where test excavations would yield large ceramic
samples. Magnolia Plantation was the site where we excavated the first shovel tests. At
this site, we gained an idea of subsurface artifact distributions by excavating tests to
87
depths of one meter or more. After it became clear here, and at other sites tested later,
that artifact bearing layers did not extend into clay subsoils, shovel tests were excavated
to a minimum depth of 50 cm. In the rare instances that deposits extend further below the
surface, we excavated our tests to a depth roughly 20 cm below the occurrence of the last
artifact.
At each site, the first round of shovel tests was spaced 30 m apart. Transects
extended in radial lines in the cardinal directions. When these tests yielded high artifact
density zones along one or more of the initial lines, an additional suite of tests excavated
using an even sided grid arrangement with tests spaced 10 m apart. Once this grid
established general zones of high artifact density, two or three tests were sunk along the
midpoints between the tests in the 10 m grid. These last tests provided the data necessary
to select a two-by-two meter area for test excavations. At sites where we had permission
to work near the mounds, additional tests were excavated in five meter intervals around
the base of the mound to look for the mound-slope trash dumps commonly reported at
sites in the Georgia Piedmont (Williams and Smith 1984).
At Hay Fever Farm, shovel tests were excavated in a grid arrangement and tests
were spaced 30 m apart. At Red Bluff Earthlodge, tests were also excavated in a grid,
using 10 m spacing. At these sites, we substituted the grid system the for radial line
approach because it became clear during the course of mapping that radial lines transects
were unlikely to reveal areas of sufficient artifact density to justify test excavations.
Shovel test locations were recorded using a total station and the elevation readings
from the test locations were included in the topographic maps. An exception is Red Bluff
88
Earthlodge, where most shovel tests locations were recorded using compass and tape. For
each test, excavators recorded test depth, depth of soil changes, and depths of artifacts.
In radial-line based shovel test programs, site boundaries were determined on the
basis of negative shovel tests away from the site center. Criteria for a site boundary are
two-fold. First, two shovel tests in a row must be negative. Two additional tests were then
excavated on a perpendicular line to the last negative test. If all these proved negative, the
boundary was accepted. Otherwise, testing continued and the process was repeated. Grid
based shovel testing programs used multiple parallel negative tests to determine site
boundaries.
Excavation
Regional settlement pattern data provide hard evidence concerning settlement
patterns while also providing insights into possible trends in settlement hierarchy and
occupational intensity. However, any site hierarchy requires some measure of internal
site structure. Reference to subsurface investigations at sample sites provides clues about
settlement intensity and deposition processes. Excavations also provided additional data
concerning ceramic variation.
As noted in the preceding section, excavations were carried out in areas shown by
the shovel test transects to have the greatest artifact density. These excavations were
generally one-by-one or two-by-two meters in extent. We excavated each down to sterile
soil. At Mound B of Magnolia Plantation and Structures 2 and 3 at the Red Bluff
Earthlodge site, one-by-two meter excavations were conducted to document the
stratigraphy of intact surface architecture. Except when architectural units required trowel
89
excavation of features and piece plotting of artifacts, units were excavated with shovels
using 10 cm levels. Dirt was screened through ¼ in. hardware cloth. To compensate for
the lack of visible stratigraphy in non-architectural excavation units, we collected 5 cm3
soil sample columns from the sidewalls of excavation units at each site.
These samples were taken as one-meter-long columns from cleaned excavation
unit profiles. Samples were collected from the bottom up (cf. Vogel 2002) using a
custom-made steel soil cutter of appropriate volume (Figure 3.6). Each 5 cm3 sample was
bagged and labeled separately in brown paper bags. In sealed stratigraphic deposits
(mounds and structure floors) soil samples were taken for flotation at a later date.
FIGURE 3.6,
Collecting 5 cm3 soil columns from an excavation unit profile
90
Intrasite Processes: Artifact Deposition and Soil Analysis
The Dougherty Plain differs from other parts of the southeastern United States in
that the predominantly flat terrain, shallow groundwater, and sandy, sedimentary soils
create low energy hydrologic environments characterized by slow meandering streams
and large volumes of seasonally standing water (Golladay and Battle 2001; Hicks 1987).
Depositional processes in such settings create sedimentological profiles with stratigraphy
that seems undifferentiated when subjected only to visual inspection (Brooks and
Sassaman 1990). Researchers working in the coastal plain have developed field and
laboratory methods for differentiating stratigraphic deposits in such settings (Brooks
1992; Brooks and Sassaman 1990; Sassaman 1992).
These techniques generally entail excavations to create broad horizontal
exposures, as well as individual piece-plotting of most artifacts (Brooks and Sassaman
1990: 187). While archaeologists caution that we should not summarily “dismiss such
sites as mixed or disturbed,” (Brooks and Sassaman 1990:193), some aspects of their
techniques are not suitable for sites that may actually have been subjected to severe postdepositional disturbance. Since block excavations were not part of this study, a modified
subset of Brooks and Sassaman’s techniques were employed in a novel way as a means
of testing for subsurface disturbance.
While Brooks and Sassaman (1990) conducted their research with the goal of
understanding overall site formation processes and the sedimentological and human use
history of a given site, I focused on a narrow set of questions. To evaluate the usefulness
of excavated samples for ceramic seriation and to determine whether or not future
91
research would be merited at a given site, I was interested in whether or not the site
presented signatures of any kind that might indicate post-depositional disturbance. To
that end, A.J. Vonarx and I (Vonarx and Chamblee 2005) employ site particle size
analysis and available phosphate analysis on the 5 cm3 soil columns taken from the walls
of test excavation units.
As noted above, I collected appropriate soil samples from excavation unit profiles
at each site. Vonarx undertook all laboratory procedures using the Soil Analysis
Laboratory at Statistical Research Inc. (Vonarx and Chamblee 2005). Particle size
analysis was conducted on materials separated using the wet-sedimentation method
(Janitzky 1986), a more precise technique than those employed by Brooks and Sassaman
(1990). Since particle size data are better interpreted with controls from nearby non-site
contexts (Brooks and Sassaman 1990), and such contexts were not sampled in the field,
available phosphate analysis was added as an additional control.
Available phosphate analysis not only contrasts with other phosphate extraction
techniques in that it focuses only on the phosphate available for plant uptake, but also
because the technique extracts phosphate from soil components that, in the absence of
mechanical disturbance, are less subject to vertical migration through the soil (Mehlic
1984). Such a methodological difference is crucial in porous coastal plain soils. Vonarx
extracted and measured available phosphate using the Mehlic 3 extraction method
(Mehlic 1984).
Supplemental measures of soil disturbance include textural descriptions of soil
columns from county soil surveys and measures of ceramic relative frequencies and
92
ceramic frequencies according to a four part sherd size classification system below. The
full suite of techniques was applied to two-by-two meter excavation units at two sites in
the project area. Based on land use history as related by local land owners and managers,
one unit was excavated in an area judged to have possibly been disturbed, while the other
was excavated in a more stable setting. Results are reported in Chapter 4 and suggest that
the method is appropriate for separating disturbed and undisturbed contexts.
Ceramic Analysis
This project presents data collected from 6,556 sherds. Approximately equal
sample sizes came from my analysis of existing collections and that related to recent
fieldwork (3,306 and 3,292 sherds, respectively). Ceramic data are crucial in formulating
interpretations regarding site formation processes, regional and intra-site chronology,
intensity of occupation, and the social and economic contexts of ceramic production and
exchange all depend on these data. However, regional scale projects often depend on
ceramic samples that are not large enough or sufficiently fine-grained to provide insights
into micro-scale processes such as household ceramic production, learning frameworks,
vessel assemblage variation, and ceramic discard and accumulation rates. This project is
no exception, but small sample sizes and regional data still offer many research options.
This section summarizes analytical problems and middle range theoretical
approaches relevant to the domains of chronology, formation processes, artifact
accumulations and occupational intensity, and ceramic production and exchange. After
presenting these problems and options for addressing them, I briefly outline the basic
procedures and variables considered in ceramic classification and categorization. In the
93
course of describing my methods, I return to the theoretical problems I present,
highlighting the ways in which specific methods address specific problems. Ceramic
analysis results must be presented at multiple spatial scales and are addressed in the next
chapter.
Analytical Problems and Middle Range Theory
Chronology
In Chapter 2, I presented chronological outlines of the development changes and
published periodization for the regions surrounding the project area. The sherd-based
ceramic chronologies on which these narratives are based are seriations. While some are
surprisingly sensitive (cf. Willey 1998[1949]; Williams and Shapiro 1990), radiocarbon
dates are often the best independent chronometric measures. Thus the exact placement of
dates is not always secure, despite the fact that some regional chronologies consist of
equal-length phase designations in which is phases lasts no more than 75 or 100 years (cf.
Pluckhahn 2003; Smith 1995; Williams 1990c). However, through a variety of computer
simulations, Hatch et al. (1982) confirm what many know intuitively – that small sample
sizes set poor conditions for seriation. As only 5% of sherds from this project are
temporally sensitive, seriation is not an option.
Sample size issues are compounded by the fact that the project area does not fit
neatly into any geographically defined ceramic production tradition – such as the
Woodland period Swift Creek (cf. Williams and Elliott 1998) or Weeden Island
(Milanich et al. 1997; Willey 1998[1949]) traditions, the Mississippian period Fort
Walton (Scarry 1985; White 1985; Willey 1998[1949]), Roods/Bull Creek (cf. Blitz and
94
Lorenz 2006; Schnell et al. 1981), and Savannah and Lamar (Hally 1994; Wauchope
1966; Williams 2006) traditions. Instead, it seems to straddle a crossroads between all of
these previously defined ceramic analytical constructs.
To meet these challenges, I use an approach that leads to broader, but still
analytically secure and useful chronological designations. For each period, I assign
temporal associations using the presence or absence of diagnostic types that are restricted
to individual time periods. Specific designations are discussed below, with other
categorization variables. I restrict classification to the most specific temporal categories
(shown in Table 3.1) to assemblages with at least two diagnostics. If a specific time
period cannot be assigned, I use a more general category. Although resulting temporal
divisions are long and many sites cannot be assigned to any period, the combined effects
of low sample sizes and variation among chronologies in surrounding regions prevent
attempts to be more temporally specific. I discuss several of these interpretive difficulties,
before discussing site formation processes and ceramic production.
In the Middle Woodland Period, Pluckhahn (2002) is able to develop four periods,
whereas I only recognize two. Pluckhahn (2002:Table 2.2) notes high frequencies of
Swift Creek Complicated Stamped at Kolomoki during its most intense occupation. In
contrast, frequencies of Swift Creek sherds in the Chickasawhatchee drainage are quite
low. This discrepancy may have one of two causes. Middle Woodland period occupations
in the Chickasawhatchee Swamp may simply predate the occupation of Kolomoki. If this
is so, these occupations would have been followed by an abandonment that coincided
with Kolomoki’s ascendancy. However, it is also possible that the discrepancy results
95
from differential participation in the Swift Creek interaction sphere. In this scenario, a
shortage of Swift Creek sherds would indicate partial exclusion or reduced participation
in the broader spheres of exchange that involved Swift Creek pottery (Snow and
Stephenson 1998; Stoltman and Snow 1998). Without further research on this issue,
further divisions of the Middle Woodland period are uncertain.
I assign Late Woodland designations based on the presence of Weeden Island
types that do not appear at Kolomoki until close to its abandonment. Although check
stamped pottery is almost ubiquitous enough to be plain ware (White 1985:167), I also
use this ware as a Late Woodland diagnostic when I observe that one or more check
stamped sherds exhibit a small excised line just below the rim (cf. White 1981:652). Late
Woodland occupations present a chronological problem, as several sources suggest that
Late Woodland occupations in the interior coast plain were often contemporary with
nearby Mississippian occupations (Blitz and Lorenz 2002 and 2006; Brose and Percy
1978; Milanich 1994; Schnell and Wright 1993; Stephenson 1990; Snow and Stephenson
1998; White 1981).
In this project area, I observed very little evidence for the use of decorative
techniques that researchers in other regions identify with the Early Mississippian period
(see King 1996:37; Worth 1988:124). Apart from a single Etowah Complicated Stamped
sherd in the landowner’s collection at CAS 245, I saw no other complicated stamped
motifs or decorative techniques that could be attributed to the Early Mississippian period.
Lest support for the lack of an Early Mississippian period political and economic
transformation rest on the absence of a single type, it is also worth nothing that there is no
96
shell tempered pottery in the project area. Considering that Blitz and Lorenz (2002 and
2006) repeatedly emphasize the importance of shell-tempered pottery as an Early
Mississippian period marker, it is fair to say that the two lines of evidence concerning
pottery design support one another.
However, this does not mean that the project area was abandoned between A.D.
1000 and A.D. 1200. At present, I cannot determine with certainty whether any Late
Woodland period occupations might be contemporary with Early Mississippian
settlements in nearby regions. However, given the evidence from elsewhere in the interior
coastal plain presented in Chapter 2 (e.g., Stephenson 1990; Schnell and Wright 1993:3637), I assume that at least some of the Late Woodland period occupations in the project
area existed during what is normally called the Early Mississippian period.
Middle Mississippian diagnostics used in this dissertation are a subset of the types
used to define the Rood Phase by Schnell et al. (1981), Schnell and Wright (1993), and
Blitz and Lorenz (2006). In the following section, I discuss the importance of classifying
pottery using a system that separates attributes such as decorative design, rim
modifications, and vessel form. A prelude to this discussion revolves around problems
with the classification of Middle Mississippian period plain wares along the Atlantic
slope in general and in the Chattahoochee Valley, in particular.
Blitz and Lorenz (2006:235-236) classify many of their plain wares, including
some from Middle Mississippian excavation contexts, as “Lamar Plain.” Conventionally,
any “Lamar” designation connotes the Late Mississippian period (see Kowalewski and
Williams 1989; Smith 1995; Williams and Shapiro 1990; Williams and Thompson 1998;
97
Worth 1998). In contrast, Schnell et al. (1981:185-188) define a middle Mississippian
period type, Ingram Plain, which they argue to be the stylistic antecedent for Lamar Plain
– indicating that there may be difficulty distinguishing between the two. At
Cemochechobee, Ingram Plain sherds were recorded throughout the occupation sequence,
which Schnell et al. (1981:251) place between 900 and 1350 A.D. – in other words,
before the Late Mississippian period.
Although Blitz and Lorenz (2006: Appendix C) and Schnell et al. (1981:161)
make analytical separations among decorative types, vessel forms, and rim modifications,
there is some reason to suspect that the confusion surrounding the temporal separation of
plain wares stems from an incomplete separation of these attributes. Blitz and Lorenz
(2006) use Lamar Plain for a variety of vessel forms, while Schnell et al. (1981:188) only
used the Ingram Plain designation for open bowl vessel forms. However, open bowls are
used across most time periods (compare Wauchope 1966:Figures 27-36). In the end two
attributes seem to matter most: rim modification and vessel thickness.
Wauchope (1966:79, 80 and 85) describes a general trend of vessels becoming
thinner from the Middle Mississippian period to the Late Mississippian period, as well as
a reduced range of variation in the thickness. In addition, although Kowalewski and
Williams (1989) recognize the importance of folded and pinched rims as a marker for the
Late Mississippian period, Williams (personal communication, 2006) has observed
hemispherical bowls with direct rim pinching among Late Savannah components
throughout Georgia and suggests a mid-14th century date.
98
Taking these two observations into account, I classify Ingram Plain as either
thick-walled, outleaned-wall bowls with appliqué strip decorations or hemispherical
bowls with pinching or notching executed directly into the clay body of the rim (e.g.,
Schnell et al. 1981:Plates 14-16). It may ultimately prove most useful to discard the
Ingram Plain and Lamar Plain designations entirely and simply use vessel form data and
rim modifications to make any possible temporal separations among plain wares.
However, given the small sample sizes involved, I would not advocate this step without
analyzing other collections.
It is possible that the Late Mississippian period can actually be split into two subperiods, although additional data are needed to test this hypothesis. According to Blitz
and Lorenz (2006), decorated ceramic types from the north (e.g. the Lamar types) and
those from the south (e.g. Fort Walton Incised and some appliqué strip forms of rim
decoration) spread through the interior coastal plain at differential rates. At various
northerly sites, Lamar diagnostics appear earlier, while Fort Walton diagnostics appeared
earlier among southerly sites. Given the project area’s location near the center of the
frontier between these ceramic traditions and low ceramic frequencies, however, there are
few opportunities to capitalize on this observation, except to note it as a possibility.
Formation Processes
Research into site formation processes is crucial if one is to understand the
transformations imposed on the archaeological record after materials are initially
deposited (Schiffer 1983 and 1996). Because this project took place in an area lacking
prior systematic excavations, considerations of site formation processes are even more
99
important. Primary concerns revolve around post depositional processes caused by
modern disturbance and the relationships between these processes and measures of
ceramic variation.
Ceramic analyses directed toward understanding formation processes attempt to
answer three primary questions:
1) In what ways has the stratigraphic integrity of archaeological sites in the
project area been affected by well-documented and wide-spread modern
agricultural and silvicultural practices?
2) How does the disturbance present at many sites affect the comparability of
ceramic assemblages across sites?
3) What insights into intra-site variation and occupation intensity can we gain by
investigating the specific branch of formation process research directed towards
understanding ceramic accumulation rates and settlement pattern change?
As noted above, some questions are addressed through the appropriate application of
field methods discussed above. This section briefly addresses the issue as related to
ceramic analysis.
The primary ceramic variable brought to bear on stratigraphic integrity and
intrasite variation is sherd size. Nielsen (1993:151) used the sherd size variable as a
means to separate primary or de facto refuse from refuse removed from its primary
context through disturbance or systemic context cleaning, that is “secondary refuse.”
Recent research suggests that formation processes that affect midden development are
even more complex than once thought (cf. Beck 2006). Because this project emphasizes
100
regional patterns of site and community development, this study simply makes the
assumption that, in all cases save one (9BX4, XU 2) all refuse is secondary and has been
disturbed.
Comparisons of sherd size are then used to determine whether or not postdepositional disturbance has been significant enough to limit comparability between
excavated and surface collected samples, as well as between surface samples collected
under different conditions. This approach deliberately emphasizes an entropy oriented
conception of formation processes (cf. Schiffer 1983:676), instead of the transformational
approach favored in later studies (Schiffer 1996:13). Field observations and other reports
of site destruction (Elliott 2004; Kowalewski 2005; Smith and Harris 2001; Wood and
Lucas 2005) indicate that it is necessary to ask whether some sites must simply be
“written off” because of the wide variety of historic disturbance activities associated with
agriculture and silviculture. As shown in Chapter 4, examining sherd size distributions
across data collection contexts provides the means to set appropriate limits on
comparative analysis.
Artifact Accumulations and Occupational Intensity
Artifact accumulations research one focus of studying formation processes in
archaeology that provides insights into questions regarding occupation duration and site
abandonment (Mills 1989; Varien and Mills 1997), as well as the nature of regional
economic integration (Pauketat 1989).These studies began to develop in earnest in
response to data from a variety of ethnoarchaeological studies (Varien and Mills
1997:143-145; Wallace 1995:78).
101
As Wallace notes (1995:78), many studies “do not gather the data necessary to
model ceramic accumulation rates.” Cases open to such modeling possibilities often
involve specific conditions, such as the systematic accumulation of spatially discrete
trash middens (Wallace 1995) or regular filling of abandoned masonry rooms with trash
(cf. Varien and Mills 1997). Data on vessel use-life and formulas for determining a
minimum number of vessels in midden deposits are often combined with various
modifications of Schiffer’s (1996) discard equation to provide an estimate of site
occupation spans (Clark 2004:188-189; Varien and Mills 1997:151-160; Wallace 1995:
102-105).
These approaches, developed most robustly in the southwestern United States,
may be appropriate for some parts of the southeastern U.S., especially using the middens
found at the base of many platform mounds (Smith and Williams 1984). No contexts
from this project provide an opportunity for such complex analysis. However, one
universally agreed upon measure of site duration may serve as a proxy for occupational
intensity in this area, where ceramic assemblages are too small to permit a more
sophisticated approach. It is widely recognized that longer site occupation spans correlate
not only with greater ceramic frequencies, but also with increased typological diversity
(Kent 1992: 646-647; Nielsen 1993:153-154). According to Varien and Mills (1997:144145), this phenomenon fueled research into vessel use life and accumulations research.
While such specificity is preferred, a recent study of ceramic diversity by
Schroeder (1997) suggests that in, in the absence of appropriate data, regional
assessments of relative occupational intensity may be obtained by measuring ceramic
102
abundance and diversity (cf. Kent 1992:646; Nielsen 1993:168). As shown in Chapters 5
and 6, this approach supports additional lines of evidence that shed light not only the
project area’s settlement hierarchy, but also its place in macroregional interaction
spheres.
Ceramic Production and Exchange
As is the case with accumulations research, most techniques for studying ceramic
production and exchange are dependent on large sample sizes and pottery taken from
primary depositional contexts (Pauketat 1997; Steponaitis 1983; Van Keuren 2001;
Welch 1981). These studies often focus on the micro-economic aspects of ceramic
production and focus on measures of standardization, local organization of production,
and learning frameworks (Costin 1991). However, some models of production and
exchange, specifically those developed for survey contexts in Highland Mesoamerica,
have proven extremely useful in understand ceramic exchange and production (cf.
Kowalewski 1989 and 2003).
Perhaps the most useful model for this study is Kowalewski’s (2003) suite of
measures for studying integration into macroregional systems of production and
exchange. Focusing on the Peñoles area, northwest of the Valley of Oaxaca, Mexico,
Kowalewski (2003) develops criteria useful to determining whether or not a specific
region is a relative “backcountry” in terms of the ceramic production practices. The
criteria (Kowalewski 2003:71-73) include standardization and production quality, vessel
form diversity, variation in decorative design, execution of decorative techniques, and
frequencies of plain vessels to
103
This model is designed for market economies with specialist producers and
assumes core/periphery relationships (Kowalewski’s 2003:67-68). Neither the former nor
the latter are appropriate for the eastern Woodlands. Arguments against the existence of
markets in the Mississippian world are long-standing (Steponaitis 1978). In addition,
Welch (1981) and Steponaitis (1983) both convincingly argue that most ceramic
production is a part-time household enterprise. However, in light of Anderson and
Mainfort’s (2002) assertion that we know almost nothing about Woodland period
economic activity and Welch’s (1981:200) crucial, but oft-overlooked observation that
there is “no reason to believe that the economy of any other polity was structured in the
same way as Moundville’s,” it seems imperative that we try to understand regional
variation in ceramic production using available data.
Following Kowalewski’s approach, I apply his criteria to the entire regional
ceramic assemblage. Some variables are not appropriate and are omitted. For instance,
there are very few recorded instances of ceramic production sites recorded in the
southeastern United States (e.g., Welch 1981:146-147), so an absence of ceramic
production areas is not necessarily meaningful. Household production is assumed. In
addition, the paddle-stamping on Late Mississippian period, Lamar Complicated Stamped
pottery is almost always “carelessly executed” (Wauchope 1966:80), so some cases of
poor design execution may not tell us much either.
At a regional scale, variables that might point to a more marginal ceramic
production economy are high plain ware frequencies, a narrow range of vessel forms, and
104
a wide variety of production anomalies, such as irregular surfaces, unusual vessel forms,
and atypical paste recipes (Kowalewski 2003:71).
At a macroregional scale, I assess the ceramic economy comparatively by looking
at the overall frequency of ceramics at mound sites through comparative measures of
decorated ceramic diversity. I combine the frequency of ceramics and their typological
diversity with site size and mound size data. Together these data provide the means to
discuss a wide variety of issues. Due to their uniform size and distribution across
depositional contexts, shovel test data are ideal for macroregional comparisons.
Ceramic Classification and Categorization
The above discussions raise a variety of theoretical issues that discourage
researchers from assuming direct relationships between variation in sherd-based ceramic
data and variation in prehistoric vessel assemblages. Such problems can be particularly
acute in areas like the southeastern United States where typological classification
schemes depend heavily on sherd-based research (cf. Wauchope 1966; Willey
1998[1949]). The classification scheme presented here seeks to avoid these pitfalls by
including variables that speak directly to formation processes and the differences between
sherd variation and variation in the vessel assemblage.
As Rice (1987:222) notes, vessel form data depend on rim sherds, a highly underrepresented class in this study. To augment this deficiency, I consider conceptual
relationships between variation in vessel form and traditional type definitions. Data
relevant to formation processes need not come from such a limited sample and are readily
available.
105
Relational Databases and Ceramic Classification
The coding of ceramic variation is typically accomplished through large and
cumbersome spreadsheets in which multiple variables are encoded in a single column.
For example, the category, “Savannah Complicated Stamped Jar Rim,” encodes not only
the decorative design and its temporal assignment, but also the vessel form and vessel
part. The categorization scheme employed here substitutes a relational database model for
such large and often unwieldy flat tables (cf. Codd 1990).
The specific relational data model for archaeological ceramics used was
developed by John Murphy (2002) and tested successfully at the Maya site of La Milpa,
Belize (Sagebiel 2005) and in by myself in the central Mixteca Alta (Kowalewski et al.
n.d.). Following Murphy (2002), I argue that a single table cannot efficiently describe all
analyzed ceramics while simultaneously cataloging the range of variation, or domain, for
all ceramic attributes in a study area. I have already described the conceptual difficulties
of using type designations as shorthand for a combination of attributes. However, in
addition to the analytical problems, the maintenance of ceramic data using computerized
data management tools means that there are technical problems, as well. Combining
multiple attributes in single column heading impedes comparisons based on individual
attributes. For example, when using columns entitled “Lamar Incised Bowls” and “Lamar
Incised Jars,” a computer will only see the columns and fail to recognize any
commonalities in the text strings comprising the names. Text headings have no meaning
for the computer, so it is the user who must see that both describe vessels from the Lamar
period that were decorated with incised lines.
106
As noted above, a single table is devoted to the actual ceramic materials at issue –
individual piles of sherds as sorted by the analyst. This table describes all the attributes
shared by each sherd pile. By coding attributes in separate columns, we break
classification down into basic elements, encode them in a way the computer can
recognize, and provide a means to separate out similarities and differences
programmatically. The use of the relational approach also allows more flexibility for
defining the range of known variation. Every column correlates with a given attribute, so
we can use Relational Database Management Systems (RDBMS) to cross-reference each
column in the main data table with columns in another table that contain legal lists of
possible values, or the domains, for each attribute. This use of “lookup tables” more
completely describes variation and decreases the possibility of coding errors.
The attributes used for coding individual piles of sherds are lot number, ware and
temper, vessel form, vessel part, decorative type, and sherd count. A separate table
records sherd weight and counts of sherd size classes, aggregating these data by lot
number. Finally, a third table presents analysis of ceramic rims and records variables
relating to decoration, rim thickness, rim fragment size, and where possible, approximate
vessel size. This section describes each of these variables. As lot numbers are described
elsewhere, they are omitted from the present description. Throughout the section, tables
present the legal values acceptable for each category.
Decorative Type. Table 3.1 presents a complete list of decorative types used in
classification, along with their temporal associations. In keeping with Murphy’s (2002)
database model, pottery classification for this project discards the implications explicitly
107
related to ware, vessel form, and rim modification – instead using type names to connote
decorative design. These names are useful in distinguishing decorative techniques that
produce different stylistic motifs. For example, Swift Creek Complicated, Stamped
Savannah Complicated Stamped, and Lamar Complicated Stamped all date to distinct
periods and look different from one another. Unless otherwise noted in the table with a
superscript modifier, all type definitions and temporal correlations follow Williams and
Thompson’s (1998) A Guide to Georgia Indian Pottery Types.
Seven of the eight categories correlated with the “Unknown Native
American/Ceramic” period apply to decorated sherds so damaged by mechanical wear or
soil acidity that they cannot be assigned to a specific temporal category (cf. Nielsen
1993). These sherds account for about 10% of the assemblage. The eighth designation in
this temporal category, UID Plain, refers to sherds that have no visible decoration, but
still exhibit enough intact surface present to determine that they are in fact undecorated.
Sherds that may have been plain, but were subject to severe erosion were marked
“eroded.”
Plain sherds account for the majority of this assemblage (77%). In cases where the
entire ceramic assemblage from a site is either UID plain or too eroded for a specific type
designation, I follow the terminology of the Georgia Archaeological Site File (n.d.) by
using the temporal designation Unknown Native American Ceramic. The “other”
category applies to very rare, but classifiable sherds. Complete descriptions for these
sherds are provided in a comments field.
108
TABLE 3.1,
Ceramic decorative types and temporal associations. Types marked with
superscripts are based on the following references: 1) White 1985 2) Pluckhahn
2003 3) Schnell et al. 1981 4) Blitz and Lorenz 2006.
Period
Late Archaic
General Woodland
Middle Woodland
Late Woodland
General Mississippian
Middle Mississippian
Late Mississippian
Protohistoric
Unknown
Indian/Ceramic
Ceramic Type
UID Fiber Tempered
Ocmulgee/Fairchild Cord Marked
Swift Creek Complicated Stamped
Wakulla Check Stamped1
Alligator Bayou Rocker Stamped
Deptford Linear Check Stamped
Deptford Simple Stamped
UID Simple Stamped
Carabelle Incised
Carrabelle Punctated
Dunlap Fabric Marked
Indian Pass Incised
Keith Incised2
Weeden Island Incised2
Lake Jackson Decorated
Lake Jackson Incised
Lake Jackson Plain
Andrews Decorated
Columbia Incised
Cool Branch Incised
Ingram Plain3
Savannah Complicated Stamped
Fort Walton Incised4
Lamar Bold Incised
Lamar Complicated Stamped
Lamar Plain
Point Washington Incised
Alachua Cob Marked
Chattahoochee Brushed
Leon Check Stamped
Leon Jefferson Complicated
Stamped
Eroded
Other
UID Complicated Stamped
UID Decorated
UID Incised
UID Plain
UID Punctated
UID Stamped
109
Temper In the southeastern United States, broad temper distinctions, such as
limestone or shell vs. “sand and grit,” are often correlated with varying decorative type
names. For this project, I used a separate category. This decision is justified through
reference to Murphy’s (2002) relational database model and by the fact a few sand and
grit tempered sherds are so dominated by either sand or grit that separating them is
occasionally warranted. I later learned that the category was of little use, as less than fifty
sherds fell outside of a large sand/grit category that cannot be subdivided without
archaeometric analysis. We recovered no shell or limestone tempered sherds from the
project area.
Vessel Form This variable is actually comprised of two dimensions – the overall
shape of the vessel and vessel size. Hally’s (1986) analysis of formal and functional
variation among Mississippian vessel assemblages demonstrates that vessels with the
same shape but different sizes may not serve the same function. As I have no whole
vessels with which to work, I use general vessel form categories and make estimates of
vessel size, where possible, as part of the rim analysis (discussed below). An additional
problem is the well-known contradiction between the somewhat arbitrary nature with
which bowls and jars are classified, and the very clear functional distinctions that are
indicated by the terms (Rice 1987:211). Thanks in part to Hally’s (1986) work, these
associations are more explicit for Mississippian period vessel assemblages – providing a
model for vessel form classification. Woodland period vessel form classes are
problematic, because similar research is lacking.
110
Table 3.2 presents the vessel form classes I use for my analysis. Broadly, my
vessel form classification parallels that of Blitz and Lorenz (2006:229). There are a few
divergences. I call their “handled jar” a “collared jar,” in recognition of the fact that this
form sometimes lacks handles. Their “open bowl” I call an “outleaned-wall bowl,” their
“rounded bowl,” a “hemispherical bowl,” and their “shallow flaring rim bowl,” a “plate.”
These categories are broadly comparable. I prefer mine simply because they are more
consistent with those used by Rice (1987), and thus may eventually aid in broader
comparisons. I also include sherd disks and pipes in my categorization.
TABLE 3.2,
Vessel form classes
Vessel Forms
Bowl
Jar
Bottle
Beaker
Unknown
Plate
Carinated Bowl
Undifferentiated
Bowl
Available vessel form categories are adequate to address what little is known
about vessel form variation during the Woodland period. However, a review of vessel
form descriptions in the type definitions by Wauchope (1966:46-50) and Willey
(1949:356-425 passim) suggest that there are few formal differences between the periods.
During the Woodland period, there seem to be more round bottomed jars. Also present
only during the Woodland period are what Rice (1987:219) might term neckless, oblong,
ellipsoid jars – a form referred to in Mesoamerica as the tecomate, or seed jar (MacNeish
et al. 1970:29).
111
Vessel Part. The vast majority of the sherds in the assemblage are unidentifiable
body sherds. The next most common vessel element is the rim. On some jars, bottles, and
bowls, it is also possible only to recognize necks, collars, and shoulders (Rice 1987:213).
A few very thick, round bottomed sherds were also recognizable as jar bases. Finally,
some conical jars without straight and outflaring rims from the earliest segments of the
Woodland period – the “Deptford …’culture’..,”according to Stephenson et al.
(2002:350) – have small feet at their base, referred to as “tetrapods” (Williams and
Thompson 1998). These are noted in the comments field CAUs, when present.
Rim Modification. Table 3.3 presents the categories of rim modification used for
this analysis and their temporal associations. Temporal associations marked with a
superscript number come from sources other than Williams and Thompson (1998). The
only modification technique that stands out as under-documented outside the project area
is the direct-rim pinching discussed in the previous section (documented in the coding
scheme as “pinched.”
TABLE 3.3,
Rim modifications and temporal associations. Superscripts designate the
references: 1) White 1985 2) Schnell et al. 1981; 3) Blitz and Lorenz 2006.
Period
General Woodland
Late Woodland
General
Mississippian
Middle Mississippian
Late Mississippian
Rim Modification
Scalloped
Wakulla Incised1
(rolled) Noded Handle
Folded and Noded
Noded
Strap Handle
Pinched2
Psuedo/Lug Handle3
Applique/Fillet Strip
Applique/Fillet Strip and Pinched
Applique/Fillet Strip and Tool
Carved
Folded and Pinched
112
Estimated Vessel Size. I use Rice’s (1987:222-223) procedures for estimating
vessel size – including the use of the template she provides in Figure 7.9 (Rice
1987:223). These techniques closely correlate with those used by Hally (1986:276). To
ensure accurate estimates of vessel size, I only measured sherds whose rim edges were
greater than 30 mm long. This reduced the sample size for sherds with vessel size
estimates to 22.
Vessel Wall Thickness. Braun (1983:119-123) and Wauchope (1966) both note
that some vessel forms exhibit long-term changes wall thickness over time. This variable
is very complex however – relating to variation in form, overall vessel size, and even
decorative design. Large samples of similar vessels are typically needed to track changes
through time (cf. Braun 1983; Chilton 1998). Knowing that I would be able to estimate
vessel size for only a small sample of sherds, I use Braun’s (1983:Figure 5.2) observed
correlation between vessel size and wall thickness, provide a preliminary view of vessel
size variation by using thickness as a supplemental proxy measure to my smaller sample.
Rice (1987:227-228) notes that wall thickness varies greatly within a single
vessel, depending on which part of the vessel is measured. To compensate for this
variation, thickness measures were only taken on rim sherds. For consistency, I measured
in such a way as to ensure the thinnest measure possible. For direct or straight rims, with
no thickening of the vessel lip, I measured at a 90 degree angle to the lip edge. For flaring
rim vessels or those with thickened lips, I measured parallel to the lip at the point at
which the sherd was thinnest.
113
Weight. Nielsen (1993:152) uses sherd weight as a proxy for sherd size. However,
in the southeastern United States, a dearth of mechanical studies limit the utility of this
variable for individual sherds because we lack a good understanding of the range of
variation regarding vessel wall thickness, porosity, density of temper etc. (Hally
1986:276). Still, aggregate sherd weight values have sometimes yielded intrasite patterns
suggestive of differential refuse disposal patterns and trampling (Nielsen 1993;
Pluckhahn 2003), as well as the existence of trails (Williams 2004). In aggregate, a
count/weight ratio may also yield some insights into the patterns of post-depositional
damage that were among Nielsen’s (1993:153-154) primary concerns.
Sherd Size Categories. Given the limitations of weight as a measure of sherd size,
I turned instead to a more refined measurement. Following Heidke (1995) and Beck
(2006), a volunteer assistant, Cherie Freeman classified every individual sherd into the
four size proposed by Beck (2006:44) and use the following four categories: 1) < 5 cm2,
2) 5 – 16 cm2, 3) 16-49 cm2, and 4) > 49 cm2 Heidke’s (1995:270) five tier scheme was
unnecessary, as only three sherds in the entire assemblage were larger than 100 cm2.
Sherd Count. Among proveniences from the 2003 and 2004 field seasons, sherd
counts were conducted twice. Cherie Freeman made the first count when she sorted
pottery on the basis of size. I made the second while conducting detailed typological
analysis. In any ceramic classification project, discrepancies can occur, even over a
variable as simple as raw count (Fish 1978). If a discrepancy greater than one sherd was
encountered in a ceramic lot, we both re-analyzed the lot. The final absolute value of the
114
discrepancy between Cherie’s count and mine was 46 sherds – an error of approximately
1.5%.
Except in a few excavated contexts, I made very little effort to look for sherd
cross-mends. In the rare instances that the sherds did mend and it was clear that their
breakage was recent (e.g., the result of excavation or screening) mended sherds were
counted as one item. Differential efforts at mending between Cherie and me may account
for the count discrepancies in the 46 cases where it occurred.
Ceramic Analysis Summary
The sample size from the project area limits analytical prospects for questions
surrounding chronology, site formation processes, ceramic accumulation and discard
rates, component occupation duration, and ceramic production and exchange.
Nevertheless, some productive and substantial insights are still possible. Analytical scales
restricted to entire sites or the whole region and presence/absence measures still describe
important settlement variation within the project area and beyond. In addition, by
including variables related to sherd size and overall vessel size and form with typical,
typological attributes I am able to investigate formation processes and the regional
ceramic economy.
Chapter 3 Summary
Apart from the soil analysis developed by Vonarx and Chamblee (2005), none of
the methods presented in this chapter are new in and of themselves. All field methods
employed for the survey follow standard practices for southeastern archaeologists. Some
of the insights I drew upon to shape field procedures were drawn from projects in
115
Mesoamerica and the southwestern United States. Likewise, some ceramic classification
and analysis methods are seldom employed in the southeastern United States (and
especially in the interior coastal plain). Nevertheless, these methods are commonplace
elsewhere.
Methodological novelty may be found in the fact that, while long-term research
projects, such as that of the Oconee Valley, integrate investigations at multiple spatial
scales, this project is among the first in the southeastern U.S. to use an explicitly multiscalar research design to investigate an area not previously subject to intensive
investigations. This bottom-up approach to local, regional, and macroregional variation
follows recommendations by Anderson (1999), Kowalewski (2004), and Muller (1997)
and yields significant results.
116
CHAPTER 4: UNDERSTANDING REGIONAL AND INTRASITE SITE
FORMATION PROCESSES
This study argues for broader variation in the scale of Woodland and
Mississippian period societies. Evidence for such variation rests strongly upon settlement
pattern data and comparisons between multiple regions. Given the extreme diversity in
the ecological, geologic and land use histories of neighboring regions, however, such
comparisons run the risk of mistaking differences resulting from post-depositional site
formation processes or archaeological methods with those representative of true scalar
differences in political economy. To avoid these problems, interpretations of intrasite,
regional, or macroregional settlement pattern data should be contextualized by a careful
consideration and analysis of site formation processes and their impact on the
archaeological record (Schiffer 1996).
Southwestern Georgia has been and continues to be the setting of intensive
archaeological site disturbance and destruction resulting from the land management
techniques connected with industrial agriculture and silviculture (Elliott 2004; Smith and
Harris 2001; Wood and Lucas 2005). Archaeologists often depend on such disturbance to
create open ground ideal for site survey. At the regional scale, such dependencies
introduce biases into overall survey coverage. At the intra-site scale, stratigraphic
sequences in the loose shallow soils of sites in the interior coastal can be removed
entirely, rather than simply having their upper portions damaged (cf. Ledbetter 1995).
While such biases cannot be avoided, their impacts on archaeological
interpretation can be mitigated when the sources of bias are quantitatively described with
117
the aim of impact assessment. This chapter considers sources of bias as they are
recognized at the regional, site, and intrasite scale. The overall goal of these assessments
is to delineate post-depositional site formation processes operating in the project area and
use this understanding to set appropriate limits on the interpretation of available data.
Results suggest that the data will not support analyses based on relative ceramic
frequencies from using surface collected assemblages, but that excavated samples are
appropriate for such inter-site and inter-regional comparison. Moreover, viable
comparisons are possible using surface assemblages using presence/absence data for
ceramic types. In addition, overall ceramic frequencies at sites subjected to high intensity
survey should be appropriate for comparisons with ceramic frequencies from surface
assemblages in other regions. Finally, intra-site data from one case suggest that many
surface collected sites may have been smaller prehistorically than their modern signatures
in plowed fields indicate.
Site Formation Processes and Survey Coverage Bias
Incomplete survey coverage and site formation processes are the most likely
sources of bias in variation. The disturbance activity necessary prior to high intensity and
low intensity survey preceded this project by decades. The surface view obtained in 2004
is thus a mere snapshot created by the most recent disturbance. As this was an intensive
reconnaissance project in which survey areas were determined opportunistically, but with
an eye towards an unbiased sample, we must assess the success of this “systematically
opportunistic” approach and set appropriate limits on data interpretation.
118
Survey Coverage
Schiffer (1996:259) recognizes that, in the southeastern United States,
“regions…contain a mix of vegetation types, each posing particular visibility and
accessibility problems. As a result, the variability in reported sites is partly determined by
variability in visibility and accessibility.” He (Schiffer 1996:259) also laments that “many
regional settlement studies have not attempted to assess the effects of these factors on the
known archaeological records, thereby adversely affecting their reconstructions to an
unknown extent.” While this was a common complaint 20 years ago (sensu White 1985)
ubiquitous shovel testing and stratified sampling programs among southeastern
archaeologists working on CRM projects have given much clearer pictures of settlement
variation across physiographic zones (see, for example Fish and Gresham 1990; Benson
2003).
In addition, the development of GIS technology and the wide availability of
remotely sensed vegetation data now allow prediction of high probability survey areas
when complete survey coverage is impracticable (e.g. Johnson et al. 1988; Kvamme
1990). When GIS methods are applied to survey coverage zones, instead of site locations,
researchers can also quantitatively measure bias. Measures of bias for and high- and lowintensity survey are presented below.
High Intensity Survey
In reporting on surveys around Cahokia, Schroeder (1997: 130 and 253) notes that
archaeological surveys in agricultural areas often under-represent certain kinds of
landforms (particularly uplands) and that these gaps in survey coverage often created
119
perceived gaps in prehispanic settlement that did not exist when these settlements were
occupied. This kind of landform under-representation was a factor for this project,
especially in areas of high-intensity survey, where project limits coincided with those of
existing fields. Since all high intensity survey took place on farmland, remotely sensed
data on land use provides no insight on the relationship between coverage areas and
ecology. However, Johnson et al. (1988) and Schroeder (1997) note correlations between
landforms and ecosystems, so a digital elevation model (DEM) serves as an adequate
proxy.
Figure 4.1 shows the full extent of high intensity survey zones in the project area.
The project area’s 30 m DEM serves as a backdrop for these areas and the red dots across
the figure are 9,615 30 m2 sample grids comprising an 8.65 km2 random sample roughly
equivalent to the project area’s 8.77 km2 of high-intensity survey. Visual examination of
the figure indicates that high-intensity survey under-represented upland areas – both in
the northern portion of the project area, and to the south, where southeast-trending ridges
on either side of Chickasawhatchee Creek were omitted. A Kolmogorov-Smirnov (KS) Z
score of 30.9 and accompanying cumulative frequency graphs comparing elevation
values from the random sample and the survey areas confirm this observation (Figure
4.2). The cumulative frequency graph confirms the KS Z-test results, showing a clear
survey bias toward lower elevations, with roughly 80% of coverage occupying areas
below 65 m above sea level.
120
FIGURE 4.1,
High intensity survey areas and random test point distributions
121
1.00
0.90
0.80
0.70
0.60
0.50
Sites
NonSites
0.40
0.30
0.20
0.10
KS Z: 30.909, p < .001
0.00
45
55
65
75
85
95
105
115
distance (m )
FIGURE 4.2,
Cumulative frequency distributions for elevation in high intensity areas
Low Intensity Survey
One hundred thirty five of 168 sites located through low-intensity survey are
located on the present-day Chickasawhatchee WMA. While low intensity survey
precludes any realistic use of artifact frequency data beyond presence/absence measures
of site duration based on artifact type, the soil data and an ecological restoration model
presented in Chapter 3 makes this section of the project area the best area for assessing
long-term relationships between regional settlement patterns and ecological variation.
Such analysis is suitable only if survey coverage is sufficiently unbiased. This section
evaluates survey coverage bias as related to soils and vegetation. Visual inspection of
transects suggests that, apart from an under-representation of regularly flooded areas,
there is little bias with regard to elevation. Moreover, since vegetation and soil regimes
122
often correlate strongly with elevation (Gremillion 2003; Johnson et al. 1988; Wharton
1978), these two variables may be more effective measures.
Sampling bias for survey transects is measured using the soil and vegetation GIS
layers provided by scientists from the Joseph Jones Ecological Research Center and fully
described in Chapter 3. Taking into account road width and surrounding open areas
created by edge effects, I assume transect widths of 10 m. By buffering transect lines into
polygons 10 m in width, I use a chi-square test to compare observed vs. expected transect
areas (in hectares) to check for bias.
At a 95% significance level, tests confirm that the survey under-represented areas
with bottomland soils and vegetation regimes (Table 4.1). However, since Jones Center
hydro-geologist Woody Hicks (D.W. Hicks, personal communication 2004) suggests that
these bottomland areas have been seasonally or permanently submerged for most of the
area’s occupation history, this large bias is negligible. Nevertheless, the minor overrepresentation of clays and under-representation of sands is noteworthy. By contrast, chi
square tests presented in Table 4.1b show an extremely even representation of upland
ecosystem types, and therefore no measurable bias.
123
TABLE 4.1,
Observed and expected distributions of soil classes and vegetation categories in
low intensity survey transects, as well as chi square test results
SOILS
Upland Soils
Transect Area
(ha)
Total Area Site Area Total
(ha)
%
Area %
Expected
Observed
X^2
Clays
145.65
1593.42
0.48
0.42
256.35
291.31
Sands
52.30
827.29
0.17
0.22
133.10
104.61
4.19
7.76
Sandy Clay Loams
104.96
1344.91
0.35
0.36
216.37
209.91
0.20
Totals
302.91
3765.62
1.00
1.00
605.83
605.83 12.15
Submerged Soils
Transect Area
(ha)
Total Area Site Area Total
(ha)
%
Area %
Expected
Observed
X^2
Bottomland
52.79
2937.09
0.60
0.71
124.10
105.58
3.25
Poorly Drained Transitions
34.34
1111.01
0.38
0.27
46.94
66.95
5.98
0.44
81.41
0.00
0.02
3.44
0.87
7.54
87.57
4145.01
1.00
1.00
175.14
Ponded
Totals
175.14 17.43
VEGETATION
Upland Vegetation
Transect Area
(ha)
Total Area Site Area Total
(ha)
%
Area %
Expected
Observed
X^2
Longleaf Pine
74.10
841.35
0.26
0.26
148.44
148.21
Fire Intolerant Upland
31.49
399.18
0.11
0.12
70.43
62.98
0.00
0.88
Upland Pine/Hardwood Mix
184.75
2050.76
0.64
0.62
361.81
369.49
0.16
Totals
290.34
3291.29
1.00
1.00
580.67
580.67
1.04
Bottomland Vegetation
Transect Area
(ha)
Hardwood Bottomland
82.93
Total Area Site Area Total
(ha)
%
Area %
2726.92
0.81
0.58
Expected
119.58
Observed
X^2
165.85 12.91
Herbacious/Open Ponds
2.87
164.01
0.03
0.03
7.19
Tupelo/Cypress Swamp
17.10
1802.05
0.17
0.38
79.03
34.20 58.75
102.90
4692.99
1.00
1.00
205.80
205.80 72.02
Totals
5.75
0.36
critical value = 5.99 at a 95% C.I.
A final potential source of survey coverage is the use of the roads themselves in
the low intensity survey area. Since uplands are under-represented in some areas, while
bottomlands are under-represented in others, it seems likely that the roads follow
boundary areas between upland and bottomland terrain. This could mean that the use of
roads themselves represent part of the bias against upland survey coverage.
124
Christopherson (2000) was able to demonstrate that, during the Madaba Plains Project,
Jordan, roads did not introduce significant biases into survey coverage.
A modified version of Christopherson’s (2000:124) measures bias in the high
intensity methods alone and in the overall sample. Kolmogorov-Smirnov Z tests compare
Euclidean distances of sites and random sample points to roads and fire breaks located in
the project area. One test compares only the sites located in and around with the high
intensity survey area, while the other compares all sites together. In order avoid biases
introduced by the swampy areas roads are artificially raised above saturated soils, only
roads running across upland landforms were considered.
Cumulative frequency graphs and KS Z test statistics describe differences
between the distribution of random points and site locations in areas of both high
intensity (Figure 4.3a) and low intensity (Figure 4.3b) survey. In the case of the high
intensity survey, the graph shows that sites are actually located further from the roads
than the random sample points, however, given the high probability value (p = .067, the
results are not significant. Given that the low intensity survey followed the roads, it is not
surprising that the site locations in the overall project area are bias toward the roads.
Taken together, high and low intensity survey zones do introduce measurable
biases in survey coverage. North of the present-day WMA boundary, high intensity
surveys omit large sections of uplands. At the same time, survey coverage in the low
intensity area is biased against non-bottomland areas away from the roads.
125
FIGURE 4.3a, Cumulative frequency distributions for distance to roads in high intensity areas
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
Sites
0.20
nonsite
0.10
KS Z: 1.203, p = .067
0.00
0
500
1000
1500
2000
2500
Distance (m)
FIGURE 4.3b, Cumulative frequency distributions for distance to roads across the project area
1.00
0.90
0.80
0.70
0.60
0.50
0.40
site
0.30
nonsite
0.20
0.10
KS Z: 5.484, p < .001
0.00
0
500
1000
1500
Distance (m)
2000
2500
126
The absence of open ground in both areas accounts for these patterns of underrepresentation. While these biases should be corrected with additional survey, the
excellent variation achieved by low-intensity survey on the WMA and the broad dispersal
of high-intensity transects along the eastern bank of Chickasawhatchee Creek sample
sufficient for investigating how Woodland and Mississippian societies confronted local
ecological variation. Results from these measures also highlight the strengths and
weakness of localized full coverage and road-based approaches to survey.
Mechanized Farming and Artifact Variation
Biases introduced by survey coverage notwithstanding, data from surface
collections present their own unique set of difficulties. In Formation Processes of the
Archaeological Record, Schiffer (1996:123) posits an example in which a Woodland
mound is bulldozed by a farmer and its contents are spread atop the surface of a
surrounding Mississippian occupation. Schiffer (1996:260) goes on to point out that,
since such disturbance is often part of a regional scale pattern of landscape change, an
effort to reconstruct the sequence of such changes in an advisable step in regional
settlement pattern analysis.
According to local informants, mechanized farming in the Chickasawhatchee
Drainage accelerated in the early 1970s with widespread adoption of center pivot
agriculture (Hicks et al. 1987) (Figure 4.4). Among the site formation phenomena
documented during the survey were a wide variety of disturbances related to the
installation of center pivot systems. These include the removal of hilltop chert outcrops,
the draining of small wetlands, the grading of terrain, and the deep (50 cm or more)
127
plowing of sites that may have had intact strata. Given the shallowness of most sites in
the project area, it is a certainty that these activities destroyed most subsurface contexts.
FIGURE 4.4,
Row plowed field planted in corn
In discussions with local farmers, it also became clear that fields are plowed using
a variety of techniques. Depending on the crop, fields are “row-tilled” (corn) “harrowed”
(post-harvest), “strip tilled” (cotton), or “bottom plowed” (peanuts). Field observations
suggest that different plowing techniques, along with the frequency of rain or irrigation,
result in differential surface visibility. Most farmers practice regular crop rotation, so,
over the course of a few years, every field will be subjected to all methods. Given
aforementioned concerns related to “background noise” and artifact visibility
(Wandsnider and Camilli 1992:182) it is important to quantitatively assess the impact of
this differential visibility. Moreover, because high-intensity survey areas have all been
farmed for at least fifty years, and more or less continuously since the installation of pivot
irrigation systems, we must also consider the some sites have been completely destroyed.
128
Impact assessment of mechanized farming focuses on the reduction of visibility in
areas recently subjected to different plowing techniques. Figure 4.5 shows the 35 sites
located by high intensity survey in the central Chickasawhatchee drainage. These sites
are mapped thematically according to the plowing techniques most recently employed
prior to site discovery. Of these techniques, bottom plowing creates the most subsurface
disturbance, as ca. 60 cm blades turn up soil from 40 cm below the surface. Strip tilling is
the least invasive and creates trenches that are 5 -10 cm deep, 20-30 cm wide, and about a
meter apart. Harrowing, or “discing,” churns up soil to a depth of about 20 cm and
mulches post harvest waste into the soil. In addition to plowing techniques, artifacts in
disturbed contexts are most likely to be exposed after hard or repeated rains.
Table 4.2 displays summary data relating variables of site frequency, site size,
sherd frequency, and sherd size according to field preparation techniques and general
assessments of rain or irrigation frequency. Strip tilled and planted fields have poor
visibility. In the case of recently bottom plowed sites, surveyors literally followed behind
the plow looking for artifacts. Such a strategy can yield fantastic results on a previously
undisturbed site (Frankie Snow, personal communication), but this technique is not
recommended for fields that have been regularly plowed for many years.
Taken together, these data suggest that differential field preparation techniques
affect surface visibility and can impede not only site discovery, but several other
analytical goals as well. Since the designation of temporal components is directly tied to
the availability of diagnostic artifacts, poor plowing or rain conditions also lead to a
reduction in the number of components. Low artifact counts also preclude meaningful
129
quantitative measures as simple as artifact frequency or the presence or absence of given
artifact types. Results suggest not only that, prior to survey, fields should be allowed to
“season” through regular exposure to irrigation water or rain, but also that sites recorded
in recently bottom plowed, planted, or strip tilled fields should be excluded from
comparative analyses focusing on artifact quantities and site size.
A final regional scale concern related to site formation processes is the outright
destruction of pottery through prolonged employment of mechanized farming. The
destruction of sherds would reduce the total number of sherds and the number of
available diagnostics. These losses would result in under-representations of temporal
variation, would negate comparisons of sherd diversity within and across regions, and
would prevent even general estimates of occupational intensity.
TABLE 4.2,
Sherd size distribution data distributed according to field preparation techniques
Field Preparation
Size 1
Size 2
Size 3
Bottom Plow - Recent
Bottom Plow +
Rain/Irrigation
2
1
133
Harrow + Rain
204
Row Plow + Planted
Row Plow + Rain
Strip Till + Irrigated
Field Preparation
Size 4
Sherd
Frequency
g/sherd
Comp.
Freq.
3
0.57
1
0
0
128
5
0
266
0.73
9
282
12
0
498
0.67
17
1
0
1
0
0
1
0.15
136
211
15
0
362
0.61
5
3
2
0
0
5
0.36
3
Comp.
Freq.
Avg Site
Size (ha)
Avg
Comp
Size (ha)
Sherds/ha
(Sites)
Sherds/ha
Comp.
0.17
0.17
Bottom Plow - Recent
Bottom Plow +
Rain/Irrigation
1
17.84
17.84
12
40.78
8.62
7.49
9.00
Harrow + Rain
24
30.38
6.36
32.01
34.04
Row Plow + Planted
1
3.14
0.69
0.32
1.46
Row Plow + Rain
6
21.20
14.58
4.16
6.89
Strip Till + Irrigated
3
4.25
1.04
0.41
9.80
130
FIGURE 4.5,
Site distributions according to plowing methods
Sherd size data from surface collections suggest no significant differences across
observed plowing techniques (Table 4.2), but since each field in the project area was
likely subjected to each technique at one time or another, this homogeneity means little.
131
Adequate assessments of sherd destruction require reference to non-surface, excavated
collections from undisturbed contexts. Discussions below present results from analysis to
demonstrate the presence of such contexts using excavation data. A follow-up
comparison of surface and excavated data addresses the issue of sherd destruction.
Site Formation Processes at the Intrasite Scale
As seen elsewhere in the coastal plain, sites in the Chickasawhatchee Creek
drainage are located in sandy soils that have little-to-no visible stratigraphy. This lack of
visual stratigraphy often leads archaeologists to the erroneous conclusion that observed
stratigraphy in coastal plain sites is solely the result of human trampling, bioturbation, or
disturbance (Brooks and Sassaman 1990:183). While some studies support the idea that
trampling can lead to artifact displacement and breakage (e.g. Nielsen 1993; Stephenson
1990; Williams 2004), several studies have shown that geomorphological strata are
discernible among coastal plain sites when soil profiles are subjected to
geoarchaeological analysis (Brooks 1992; Brooks and Sassaman 1990; Sassaman 1992).
Such analysis can distinguish intact strata from disturbed stratigraphic contexts, thereby
providing a means to ensure valid comparisons of stratigraphic data and ceramic
distribution data with data from sites excavated across the project area and beyond.
This section presents results from sedimentological and soil-geochemical studies
undertaken on two samples from excavation units at sites located in contrasting settings
within the project area. In each of the test excavation units presented below, soil columns
were taken from the both the northwestern and southeastern portions of the excavation
units. However, at this time, phosphate analysis and particle size data are available only
132
from the columns taken from the northwestern corner of each unit. Results seem to
correlate broader depositional patterns, but these sequences have been interrupted by
micro-stratigraphic events such as the tree root or rodent burrow intrusions.
Interpretations are therefore preliminary and await confirmation.
Figure 4.6 shows the location of the test sites, CAS 245 (Hay Fever Farm) and
9DU6 (Windmill Plantation). Data include descriptions of geographic settings, land use
history, soil context, as well as results from analyses focused on particle size, available
phosphate, general relative ceramic frequencies, and relative frequencies of ceramic size
classes. Apart from general comments about estimated occupation duration, life use
histories for each site are left for later discussions.
Hay Fever Farm
Hay Fever Farm (Figure 4.7) is located on an upland ridge overlooking the
intersection between Chickasawhatchee Creek and Brantley Creek. There is no historic
evidence for disturbance prior to the 1950s, when the ridge-top was cleared of its
predominantly hardwood land cover and replanted as a peach orchard. In the 1970s, the
orchard was removed and the site was used as a deeply plowed, center-pivot irrigated,
peanut field, with the edges of the field planted in pine trees. After a few years, land use
changed again, and the ridge-top has since been utilized as a hay field, subjected to only
periodic harrowing. Prehistorically, Hay Fever Farm was occupied from the Early
Archaic period through the Late Mississippian period.
133
FIGURE 4.6,
Sites used for sedimentological analysis
134
FIGURE 4.7,
Topographic context of Hay Fever Farm (CAS 245)
The Lee and Terrell County USCS Soil Survey indicates that the site was
composed of Americus Sand, a friable, porous reddish-brown soil with few nutrients,
with a possible overlay of an Americus Sandy Loam characterized by organics and fine
sediments (USDASCS 1978) Apart from the plowzone, it is difficult to discern soil strata
through visual inspection (Figure 4.8). The site is located along a ridge-top that is flooded
rarely, if it all. Since the excavation unit is located down slope from the ridge summit,
most soil deposition likely occurred as a result erosion from the wearing-down of the
ridge.
135
FIGURE 4.8,
Hay Fever Farm, XU 1, Final excavation level
While colluvial deposition processes are not well understood in the coastal plain,
Sassaman (1992) hypothesizes that the upward coarsening observed at 38AK157 might
indicate a depositional signature for colluvium accumulated through erosional episodes
tied to periodic soil exposures related to human occupation, anthropogenic or natural
fires, or historic activities. According to Sassaman’s (1992) hypothesis, a series of
discrete depositional episodes would be indicated by individual upward coarsening events
through the soil column.
We would have expected correlated spikes in the in available phosphate (reported
in ppm) and in large artifact and overall artifact frequencies of artifacts, as well (Vonarx
and Chamblee 2005). Failure to identify these trends, or a series of contradictory trends,
lends support to the hypothesis, consistent with the site’s land use history, that the area
136
around XU 1 was subjected to deep disturbance and mixing of deposits by agriculture,
silviculture, and rootwork (despite efforts to locate the XU 1 in the area least likely to
have been subjected to disturbance).
Table 4.3 and Figures 4.9 and 4.10 present the results of phosphate and particle
size analysis, respectively. If examined in isolation, the phosphate analysis suggests only
one intensively utilized/occupied surface, located between 30 cm and 40 cm (Figure 4.9).
While this significant spike in phosphate levels correlates with relative sherd frequencies
by depth (Figure 4.11 and the last column in Table 4.4), particle size data (Figure 4.10)
and diagnostic sherds and sherd sizes data (Figure 4.12) tell a different story.
TABLE 4.3,
Soil Particle size and available phosphate data, according to stratigraphic level.
CAS 245, XU 1
Site/Sample ID
CAS245 XU1 0-5
CAS245 XU1 5-10
CAS245 XU1 10-15
CAS245 XU1 15-20
CAS245 XU1 20-25
CAS245 XU1 25-30
CAS245 XU1 30-35
CAS245 XU1 35-40
CAS245 XU1 40-45
CAS245 XU1 45-50
CAS245 XU1 50-55
CAS245 XU1 55-60
CAS245 XU1 60-65
CAS245 XU1 65-70
CAS245 XU1 70-75
CAS245 XU1 75-80
CAS245 XU1 80-85
CAS245 XU1 85-90
CAS245 XU1 90-95
CAS245 XU195100
VCOS
1.51
1.21
1.53
2.12
1.90
2.84
1.40
1.41
1.49
1.43
1.95
1.51
2.25
0.24
2.20
2.29
1.59
1.72
1.41
COS
17.13
18.37
19.21
20.02
17.13
20.43
17.03
17.71
18.84
17.82
10.55
16.67
12.67
20.73
12.42
16.43
13.59
15.44
14.27
MS
43.84
43.69
43.94
44.70
33.22
37.45
41.67
41.25
41.20
40.14
24.16
38.96
25.00
31.44
24.51
39.41
40.22
37.09
35.36
FS
11.75
11.42
10.38
13.37
7.68
11.29
10.77
10.36
9.44
9.90
25.05
9.90
22.71
11.18
22.27
10.50
10.24
10.11
9.74
VFS
4.25
4.30
3.87
2.56
2.51
3.83
3.50
3.49
3.26
3.39
11.07
3.26
9.47
5.03
9.29
3.59
3.38
3.32
3.32
%
SILT
6.88
7.77
9.19
3.14
21.60
4.30
4.66
5.58
4.63
5.35
6.20
13.08
7.21
6.31
6.42
6.13
6.68
5.67
5.66
%
CLAY
14.64
13.24
11.87
14.09
15.97
19.86
20.97
20.20
21.14
21.98
21.02
16.63
20.68
25.07
22.89
21.65
24.31
26.65
30.25
ppm
1500
1460
1320
1580
1610
1740
1800
1700
1100
940
890
450
670
800
520
630
660
580
620
1.42
14.74
36.85
10.15
3.38
5.86
27.61
420
cm
-5
-10
-15
-20
-25
-30
-35
-40
-45
-50
-55
-60
-65
-70
-75
-80
-85
-90
-95
100
137
Available Phosphate at CAS 245 (ppm vs. depth)
0
-10 0
500
1000
1500
2000
-20
-30
-40
-50
CAS 245
-60
-70
-80
-90
-100
FIGURE 4.9,
Stratigraphic distribution of available phosphate, CAS 245
Particle Size Distributions by Percent and Depth (CAS 245)
0
-10 0
-20
10
20
30
40
50
% CLAY
-30
% SILT
-40
VFS
-50
FS
-60
MS
-70
COS
-80
VCOS
-90
-100
FIGURE 4.10, Stratigraphic distribution of soil particle size distributions, CAS 245
138
TABLE 4.4,
Level
1
2
3
4
5
6
7
8
9
Average
Sherd size data, by stratigraphic level. CAS 245, XU 1
Size 1
(5cm^2)
0.75
0.66
0.73
0.77
0.56
0.47
0.29
0.00
0.00
0.47
Size 2
(16cm^2)
0.21
0.30
0.23
0.19
0.38
0.32
0.71
0.00
0.00
0.26
Size 3
(49cm^2)
0.04
0.03
0.03
0.04
0.05
0.21
0.00
0.00
0.00
0.05
Size 4
(100cm^2)
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Depth Total
-10
0.11
-20
0.21
-30
0.33
-40
0.24
-50
0.06
-60
0.03
-70
0.01
-80
0.00
-90
0.00
n = 628
FIGURE 4.11, Stratigraphic distribution of ceramic frequencies, CAS 245
0
0.00
-10
-20
-30
Depth
-40
-50
-60
-70
-80
-90
-100
0.05
0.10
0.15
0.20
0.25
0.30
0.35
139
FIGURE 4.12, Stratigraphic distribution of ceramic frequencies by sherd size, CAS 245
0
0.00
-10
0.20
0.40
0.60
0.80
1.00
-20
-30
-40
-50
-60
Size 4
Size 3
Size 2
Size 1
-70
-80
-90
-100
Relative frequencies of very fine sand (VFS), fine sand (FS), and medium sand
(MS), indicate a single upward coarsening sequence between 100 cm and 80 cm,
followed by alternating episodes of upward fining and coarsening every five centimeters
throughout the unit from 80 cm to 50 cm. From 50 cm to the ground surface, there is a
general upward coarsening interrupted by a single episode in which silt proportions
increase dramatically (Figure 4.10). These multiple episodes of short term deposition,
followed by one or two large depositional episodes could correlate with an area subjected
to deposition from uphill deposits from logging, seasonal plowing and harvesting, or
replanting in pine trees.
A sequence consistent with heavy disturbance is corroborated by sherd size data.
The largest sherds are located in the deepest levels of the unit (Figure 4.12) and represent
140
a very small proportion of the ceramic assemblage. Temporal data on ceramic types
support the hypothesis of widespread disturbance as well. Archaic period fiber tempered
sherds were recovered between 10 and 30 cm below surface as well as at 70 cm below
surface (where they belong). Additionally, a Mississippian sherd disk was also found near
bottom of the unit. Seen in this light, the spike in available phosphate may be the result of
run-off from agro-chemicals and the coincident spike in overall sherd frequency probably
marks the bottom of a recent plowzone.
Finally, shovel tests provide additional insights into the disturbance at Hay Fever
Farm. While this may still have been one of the most intensively occupied, long-term
sites in the project area, shovel test data call into question whether the site was actually
15 hectares prior to agricultural disturbance. Figure 4.13 combines an inverse distance
weighting model of ceramic frequency with a topographic map of the site.
The abundance of artifacts in the east central portion of the site, along a microtopographic drainage, suggests that the site may have expanded in size due to artifact
redeposition from erosion and agricultural disturbance. Such circumstances, combined
with aforementioned soil and ceramic profiles, not only suggest that ceramics from XU 1
may have been redeposited, but also indicate that Hay Fever Farm may have been a
smaller and more intensively occupied hilltop site that has been redeposited downhill by
modern disturbance and runoff.
141
FIGURE 4.13, Topography and Inverse Distance Weighting surface of ceramic density, CAS
245
142
Windmill Plantation
Windmill Plantation 9DU6 is a small civic-ceremonial center with a 2.5 m conical
mound. Artifacts suggest Early Middle Woodland period and Late Woodland period
occupations. As shown in Figure 4.14, the site is located on a low hammock between two
large, open water ponds. These ponds are connected directly to the water table through
limestone sinks and act as groundwater recharge agents (D.W. Hicks, personal
communication, 2004). The site was subjected to seasonal flooding, as indicated by the
prevalence all around the site of cypress knees, or small, stalactite-like protrusions of
wood that are part of the root systems of cypress tree in regularly flood areas. Given this
topographic situation, it is unlikely that Windmill Plantation ever supported a resident
population at the mound site itself.
The Dougherty County Soil Survey (USDACS 1968), and the soil survey
conducted by researchers from the Jones Ecological Research Center (Cammack et al.
2002) characterize the soil around the mound as a Grady Clay Loam. The profile of this
soil is characterized by high sand content yielding to mottled yellow, gray and orange
clays with depth. The soil is generally characterized as shallow, poorly drained, and often
inundated (Cammack et al. 2002; USDASCS 1968). Excavation Unit 2 (XU 2) was
excavated on high ground in an area that shovel tests suggested were artifact rich.
Mapping and land use histories also suggested minimal disturbance in the XU 2 vicinity.
The soil profile from XU 1 fits well with the overall characterization of the area (Figure
4.15).
143
FIGURE 4.14, One meter resolution aerial photograph of Windmill Plantation (9DU6)
FIGURE 4.15, 9DU6, XU 2, Final excavation level
The depositional situation, then, is one exhibiting little post-occupation
disturbance, but in which particles are laid down repeatedly each season through low
144
energy rises and falls in standing water. If this is the case, particle size data taken in 5
cm3 samples shed little light on this fine-grained process. A sampling strategy comparing
sedimentation with centimeter or sub-centimeter precision would be more appropriate.
However, phosphate data (Table 4.5 and Figure 4.16) draw out two or three
potential use surfaces, with visible in spikes at 15 cm, 25 cm, and 40 cm. These spikes
correlate nicely with total artifact count, suggesting potential geomorphological and
activity surfaces. Given the impossibility of agriculture in the area and the stability of
available phosphate in the absence of mechanical disturbance, spikes in available
phosphate present a viable route to the identification of ancient surfaces in deep
swampland settings.
TABLE 4.5,
Available phosphate data, according to stratigraphic level. 9DU6, XU 2
depth
-5
-10
-15
-20
-25
-30
-35
-40
-45
-50
-55
-60
-65
-70
-75
-80
-85
-90
ppm
980
640
1300
720
1120
940
670
1460
470
520
420
370
400
200
390
420
340
400
145
Available Phosphate at 9DU6 (ppm vs. depth)
0
-10 0
200
400
600
800
1000
1200
1400
1600
-20
-30
-40
-50
9DU6
-60
-70
-80
-90
-100
FIGURE 4.16, Stratigraphic distribution of available phosphate, 9DU6
TABLE 4.6,
Level
1
2
3
4
5
Average
Size 1
(5cm^2)
0.39
0.45
0.28
0.00
0.20
0.26
Sherd size, by stratigraphic level. 9DU6, XU 2
Size 2
(16cm^2)
0.61
0.35
0.56
0.50
0.60
0.52
Size 3
(49cm^2)
0.00
0.20
0.12
0.50
0.20
0.20
Size 4
(100cm^2)
0.00
0.00
0.04
0.00
0.00
0.01
%
Depth Total
-10
0.23
-20
0.26
-30
0.32
-40
0.13
-50
0.06
n = 78
146
Relative Ceramic Frequencies by Size Category (9DU6)
0
-10
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
-20
-30
Size 4
-40
Size 3
-50
Size 2
-60
Size 1
-70
-80
-90
-100
FIGURE 4.17, Stratigraphic distribution of ceramic frequencies by sherd size, 9DU6
XU2 Relative Ceramic Frequency by Depth (9DU6)
0
-10
0
5
10
15
20
25
30
35
-20
-30
-40
-50
-60
-70
-80
-90
-100
FIGURE 4.18, Stratigraphic distribution of ceramic frequencies, 9DU6
Relative ceramic frequencies by depth and ceramic proportions sorted by size
suggest similar views of artifact surfaces (Table 4.6 and Figures 4.17 and 4.18).
Differences in ceramic vessel size and form notwithstanding, large, horizontally oriented
sherds can indicate potential surfaces. Though their exact orientation was not
documented, many large sherds were found between 30 and 40 cm below surface. Total
147
counts and sherd size data suggest up to three surfaces between 20 and 30 cm, between
30 and 40 cm, and between 40 and 50 cm.
As noted in Chapter 4, correlations are limited by the fact that the excavations
were carried out in 10 cm levels, while the soil samples were taken in 5 cm levels.
Moreover, without particle size data, adequate reconstructions of depositional history
cannot be certain. Over and above the phosphate data, several ceramic distribution trends
argue for undisturbed strata. Unlike Hay Fever Farm, where small sherds represent up to
80% of the sample, the smallest size class (Class 1) is never dominant. Instead, larger
sherds (Classes 3 and 4) consistently represent between 20 and 40 percent of the
assemblage. Finally, as shown in Chapter 6, temporal data likewise suggest stratigraphic
integrity and separate components tied to the Middle Woodland and Late Woodland
periods.
Additional Excavated Sites
Soil samples were also taken from excavations at Magnolia Plantation (9DU1)
and the Red Bluff Earthlodge Site (9BX4), but have not been analyzed. The central
occupation zone at Magnolia Plantation was bisected by a railroad sometime between
1840 and 1870, severely damaging the site and destroying half of one platform mound.
The Red Bluff Earthlodge site may also have been partially destroyed by logging, but the
presence of surface architecture suggests that portions of the site are still intact. At
present, intact areas comprise Red Bluff Earthlodge’s defined site boundaries. At the
least, both sites have areas of intact strata, although neither is undisturbed. The most
148
secure strata at these sites are thus located in the remains of intact above-ground
architecture.
Intrasite Pattern Summary
Hay Fever Farm and Windmill Plantation present contrasting cases in terms of
geographic setting, natural depositional history, prehistoric occupation history, and
historic land-use history. As such, these cases are instructive as to the kinds of signatures
one may look for when attempting to determine whether or not a site has been subjected
to heavy post-depositional stratigraphic disturbance.
Shallow spikes of available phosphate in soils otherwise lacking nutrients, an
abundance of small sherds in shallow excavation unit levels, and contradictory results
from different measures of stratigraphic integrity all represent indicators of disturbed soil
contexts. While a true understanding of depositional history would require the kinds of
broad excavations advocated by Brooks and Sassaman (1990), one may use the
techniques above to determine whether or not sites are worth the investment to establish
large horizontal exposures and piece plot materials (Vonarx and Chamblee 2005). These
results, while promising, must be viewed as preliminary. If the additional soil columns
from the opposite sides of each excavation units do not produce comparable results, it
will be necessary to reconsider interpretations presented here and perhaps to discard
them.
Comparing Surface and Subsurface Assemblages
Having established a general understanding of site formation processes in both
surface and subsurface contexts within the project area, it is now possible to compare
149
these assemblages and thereby gain some idea of what questions are appropriate for each
data set. The next chapter will focus on the regional settlement history and political
economy of the project area. Inferences concerning the distribution, density, duration,
and intensity of settlement will rest heavily on ceramic data. By understanding the site
formation processes described above, it is possible to set appropriate limits on
interpretation and develop plausible inferences.
Tables 4.7a and 4.7b present data concerning sherd size distributions, counts,
average weights, and an average measure of grams per sherd – all from surface contexts.
Table 4.7a presents data from the eight largest surface assemblages in the project area.
These are the only surface collections with 15 or more sherds per collection area. Table
4.7b presents the same data for four stratigraphically intact excavation contexts.
Comparisons of surface and subsurface assemblages reveal significant differences
in the relative proportions of sherds according to size class. In subsurface context, larger
sherds comprise 20% more of the assemblage, overall. These differences are reflected in
observable differences in overall measures of weight and in the aggregate sherd weight
measures.
150
TABLE 4.7a,
Sherd size data from surface collections with more than 15 sherds
Field #
CAS
9CAL10(SGC)
CAS 89
CAS
9CAL11(SGC)
CAS 99
CAS 122
CAS
9CAL5(SGC)
CAS 162
CAS 124
CAS 184
CAS 163
CAS 64
CAS
9CAL7(SGC)
CAS 126
CAS 139
CAS 78
CAS 187
CAS 173
CAS 183
CAS 68
CAS 115
CAS 181
CAS 213
CAS 5
Average
Median
TABLE 4.7b,
Field #
9DU6
CAS 245
CAS 10
CAS 4
Average
Median
Shovel
Test
Count
123
117
53
91
96.00
104.00
Area (ha)
Count
sherds/ha
0.17
9.37
59
2448
355.8
261.2
1.22
2.63
4.60
139
170
290
113.9
64.7
63.0
2.85
6.75
0.94
4.10
0.92
0.57
170
379
41
149
31
17
59.6
56.2
43.6
36.4
33.6
29.8
2.58
0.69
1.80
1.07
1.66
2.72
2.03
4.40
4.53
3.65
6.26
45.04
4.81
2.68
74
16
41
24
36
55
39
66
38
29
17
18
188.957
41.00
28.6
23.1
22.8
22.5
21.7
20.2
19.2
15.0
8.4
7.9
2.7
0.4
57.0
25.87
Sherd size data from shovel tests at four test sites
Positive
Tests
28
56
46
32
40.50
39.00
%
positive
23
48
87
35
48.15
41.51
Sherd
Count
47
169
186
72
118.50
120.50
sherds
/test
0.4
1.4
3.5
0.8
1.23
1.12
sherds/
positive
1.7
3.0
4.0
2.3
2.93
2.63
151
An additional difference between the assemblages is the strongly bipolar
distribution of sherd frequencies among surface assemblages. Even among the largest
ceramic collections, surface assemblages with more than 100 sherds are in a qualitatively
separate class. In spite of the very different intrasite contexts where excavations were
carried out, the distribution of sherd frequencies among the excavated samples is more
even.
While these differences indicate greater breakage among surface collected sherds,
the case for sherds in surface contexts being destroyed beyond recognition is not strong.
Were this the case, expected differences between relative frequencies of class sizes
should be much greater, and small sherds would be better represented. Instead, the three
largest assemblages among the surface collected sites are represented by a majority of
Class 2 and Class 3 sherds. The largest of these is CAS 162, a bulldozed mound site that
has been plowed consistently for around 40 years. This site has a proportion of Class 1
sherds matching the Class 1 assemblage average among undisturbed contexts.
Is Time on Our Side? The Effect of Repeated Disturbances on Artifact Variation
A final consideration regarding site formation processes revolves around the issue
of the duration of disturbance at a given site. Throughout this chapter, I have emphasized
the fact that subsurface disturbance is an ongoing process in the interior coastal plain, not
simply a one-time occurrence. However, with the exception of habitation zones that are
never abandoned, every site is subjected to a first-time, post-occupational disturbance.
Conversations with local collectors and data from the Ocmulgee Big Bend Survey
suggest that when interior coastal plain sites are first disturbed, or are disturbed after
152
more than fifty years of lying dormant, observed artifact frequencies can be orders of
magnitude higher than in collections taken from long-disturbed contexts (Snow 1975 and
1990; Snow and Stephenson 1998). This section considers the veracity of this
phenomenon and offers some ideas about possible interpretive limitations.
Surface data provide the best basis for time series comparisons. As noted in
Chapter 3, a subset of the sites recorded in the Chickasawhatchee WMA was originally
recorded in the late 1970s by Frankie Snow, Chris Trowell, and others. In most cases
Snow and Trowell collected these sites following the first recorded disturbance each
received.
Tables 4.8a and 4.8b present ceramic frequency comparisons between the sites
collected by Snow and Trowell and those collected on Bermuda Plantation, a harrowed
field that presented the best surface collection conditions at the time of the 2004 survey.
Note that Chickasawhatchee Knoll (CAS 89) is reported separately. Although this site
had seen little disturbance prior to being recorded in 1976, it was collected much more
intensively than the other sites recorded by Snow et al. and must be considered
separately.
A Mann/Whitney U test and a Kolmogorov/Smirnov Z (KS) test comparing the
ceramic frequencies of the newly disturbed and long disturbed component sets present p
values of 0.28 and 0.58, respectively. These low probability values prevent rejection of
the null hypothesis that there are statistically significant differences between ceramic
frequency distributions.
153
TABLE 4.8a,
Sherd counts, site areas, and sherds/ha estimates from recently disturbed sites
Field Site #
CAS 9CAL5(SGC)
CAS 9CAL11(SGC)
CAS 9CAL7(SGC)
CAS 9CAL10(SGC)
CAS 9CAL13(SGC)
CAS 9CAL18(SGC)
CAS 9CAL12(SGC)
CAS 9CAL8(SGC)
CAS 9CAL4(SGC)
CAS 9CAL9(SGC)
CAS 9CAL17(SGC)
Average
Median
Standard Deviation
CAS 89
TABLE 4.8b,
Count
170
139
74
59
15
13
12
10
6
5
1
46
13
56
Site Area
(ha)
2.85
1.22
2.58
0.17
0.17
1.35
0.16
0.27
0.07
0.42
1.30
0.96
0.42
0.95
Sherds/ha
59.60
113.94
28.63
355.84
85.75
9.61
76.96
37.57
88.64
11.90
0.77
79.02
59.60
94.42
2448
9.37
261.20
Sherd counts, site areas, and sherds/ha estimates from repeatedly disturbed sites
Field Site #
CAS 122
CAS 124
CAS 139
CAS 115
CAS 126
CAS 123
CAS 121
CAS 117
CAS 137
CAS 120
CAS 116
CAS 149
Average
Median
Standard Deviation
Count
290
41
41
38
16
9
5
3
2
2
2
1
38
7
78
Site Area
(ha)
4.60
0.94
1.80
4.53
0.69
0.34
0.14
0.24
0.03
0.34
1.60
0.90
1.35
0.79
1.53
Sherds/ha
63.04
43.62
22.78
8.39
23.10
26.72
35.35
12.44
70.53
5.81
1.25
1.12
26.18
22.94
22.18
Interestingly, although differences in component area and sherds/ha likewise lack
statistically significant differences, there are large differences between the significance
levels themselves across the two variables. Significance levels for differences in site area
154
are p= 0.70 for Mann/Whitney and p = 0.97 for Kolmogorov Smirnov. For sherds/ha, the
Mann/Whitney significance level is p = 0.079 and the KS significance level is p=0.187.
Chickasawhatchee Knoll is a special case. In their initial collections, Trowell and
Snow collected 1344 sherds in a single day (Snow n.d.). Follow-up collections yielded
another 1100 sherds. If Snow’s initial collection is considered alone, sherd density per
hectare is reduced to about 143. Such a figure ranks second highest among sites that were
newly disturbed at the time of collection and is extraordinary for a site so large. While it
is possible that surface conditions immediately after the site was logged contributed to
the site’s tremendous artifact assemblage, it is also necessary to point out that the site was
being looted and that some of the largest and most impressive decorated ceramics had
probably already been removed. As such, the artifact frequencies at Chickasawhatchee
Knoll are, at least in part, reflective of a long and intense occupation.
Taken together, tabular data and statistical tests indicate that, although there are
clear and observable differences in ceramic frequencies between recently disturbed sites
and those recorded after many years of continual disturbance, these differences are not
statistically significant. However, there do seem to be differences in ceramic density. It is
possible that these differences are the result of different survey methods, but it is more
likely that they stem from the blurring of site boundaries that would occur in regularly
plowed or logged contexts. Artifact frequencies at Chickasawhatchee Knoll are partially a
result of more intensive survey and may also be due in part to surface conditions.
Nevertheless, the greater number of artifacts at this key site reflects systemic behavior to
some degree.
155
Regional and Intrasite Site Formation Processes: Summary and Conclusions
In contrast to highland Mesoamerica, or the southwestern United States, surface
survey in the southeastern U.S. is always dependent on patterns of disturbance. These
patterns introduce biases, from overall survey coverage relative to local ecology, to the
reduction of recorded artifact variation, to the destruction of entire stratigraphic
sequences. However, once such biases are understood, appropriate data subsets and the
limitation of analytical approaches will still yield excellent results.
In the Chickasawhatchee Swamp, a number of cautions apply. While the shape of
the survey areas affect site location samples (Schroeder 1997), it does seem that pockets
of full-coverage survey mitigate biases introduced by survey transects limited to roads
and trails. Excavated sites should be tested for stratigraphic integrity. Comparisons
dependent on relative frequencies of diagnostics should be limited to excavated sites, or
to the entire project area assemblage. Low artifact densities at sites recorded along roads
or in unplowed fields should be regarded skeptically. Sites recorded shortly after their
initial periods of disturbance will have observably higher artifact frequencies, but
differences between these components and those recorded after years of repeated
disturbance are not statistically significant.
Presence/absence comparisons of artifact types are appropriate using any surface
data. Interpretations of occupational intensity should depend either on high intensity
survey or excavation. Conclusions using comparisons that respect these limits adequately
represent prehistoric empirical realities and will not produce interpretations that are
skewed by site formation processes or methods.
156
CHAPTER 5: THE REGIONAL POLITICAL ECONOMY
OF THE CHICKASAWHATCHEE SWAMP
This chapter uses evidence from fieldwork and collections analysis to present a
reconstruction of settlement history from the Middle Woodland through Late
Mississippian periods. Period-by-period descriptions of settlement pattern data focus on
site location, site size, duration of occupation across periods, as well as assessments of
occupational intensity based on an intensive mapping and testing program at four sites.
Finally, an analysis of the project area’s entire ceramic assemblage focuses on
production-related dimensions of ceramic variation. Several trends emerge from these
discussions:
1) The variables of size, duration of occupation, and occupational intensity
support a division of settlements into four spatially distinct settlement zones. These zones
are interesting because, although they are spatially discrete, their separation is based on
the non-spatial criteria of occupational duration and intensity and are therefore less
subject to bias resulting from gaps in survey coverage.
2) Low intensity survey data indicate strong correlations between site location and
landscape patches shaped by anthropogenic disturbance. The exact nature of the
disturbance is unclear, but the correlation highlights the probable importance of fire in
Woodland and Mississippian subsistence strategies and raises questions about the exact
nature of these strategies.
157
3) Intensive mapping and testing at four sites in the project area suggest that civicceremonial sites in the project area have shorter occupation durations than several key
sites that lack preserved, above-ground civic-ceremonial architecture.
4) Data from regional survey and intensive site mapping suggest a bimodal
distribution of site size and intensity of occupation. Large sites with high sherd
frequencies are a minority.
5) Low ceramic frequencies at most sites, low percentages of diagnostic ceramics,
a general lack of variation in vessel thickness, and the persistence of limited ceramic
variation all suggest an economy that, in the realm of ceramic production, typifies a
“backcountry” to broader spheres of interaction.
Concluding discussions tie these trends together in a model of Woodland and
Mississippian period political economies based on the hypothesis that, although
Chickasawhatchee Creek polities existed at a crossroads between prehispanic societies
located in present-day Florida and Georgia, contemporary polities varied significantly
from their neighbors in terms of scale. Although short-term shifts in settlement patterns
may be responses to any number of agent-driven processes, longer term shifts in political
economy reflect the unintended consequences of short-term responses to dual structural
constraints imposed by a) macroregional shifts in the geopolitical landscape and b) the
opportunities and constraints imposed by local ecological variation.
Because the Chickasawhatchee Swamp is characterized by small, relatively
evenly distributed landscape patches and abundant wetlands, incentives for settlement
aggregation are low. This disincentive would affect the sectors of Woodland and
158
Mississippian period political economies that are most dependent on village life,
including mound building and the production of non-utilitarian craft items. Supporting
arguments for these hypothetical dependencies and shifts require links to archaeological
data at multiple spatial scales. This chapter presents data at regional and intrasite scales.
The nature, extent, and sources macroregional scale variation are explored in the next
chapter.
Regional Settlement History in the Chickasawhatchee Swamp
Settlement pattern data describe the shifts and continuities that comprise the
settlement history of the Chickasawhatchee Swamp. For each period, I outline settlement
patterns, discuss trends in component and site size and frequency, and discuss key
occupations. Period-by-period narratives are supplemented by settlement pattern maps,
distribution maps for diagnostic ceramic types, and tables presenting data on site size,
component size, and continuity of settlement.
General Native American Ceramic Components
Fifty-six out of 135 ceramic components are classified as General Native
American Ceramic. As noted in Chapter 3, this means that such components were
occupied sometime after the Late Archaic period, and before the region was abandoned.
These temporally undesignated components, cataloged in Table 5.1, range between
approximately 0.1 and 11 hectares in size. The majority of these sites are less than 2
hectares, and only eight sites had ceramic components that can be spatially distinguished
from surrounding lithic scatters.
159
Unknown Indian/Ceramic
Late Mississippian
Middle Mississippian
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
41
100
13
20
100
100
19
100
100
100
100
100
100
1
100
100
100
General Mississippian
0.12
2.72
0.90
0.16
0.57
0.41
3.22
0.16
0.10
7.05
1.69
0.48
0.73
0.11
0.79
0.15
1.40
3.90
0.34
15.35
3.52
0.03
0.90
5.10
1.12
4.45
0.03
1.76
0.03
8.92
5.55
1.91
2.11
0.08
Late Woodland
% of Site Area
0.12
2.72
0.90
0.16
0.57
0.41
3.22
0.16
0.10
7.05
1.69
0.48
0.73
0.11
0.79
0.15
1.40
1.60
0.34
1.96
0.69
0.03
0.90
0.98
1.12
4.45
0.03
1.76
0.03
8.92
0.04
1.91
2.11
0.08
Middle Woodland
Site Area
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
General Woodland
Component Area
CAS 13
CAS 24
CAS 37
CAS 47
CAS 64
CAS 70
CAS 72
CAS 73
CAS 77
CAS 81
CAS 85
CAS 88
CAS 93
CAS 98
CAS 104
CAS 108
CAS 109
CAS 116
CAS 120
CAS 122:2
CAS 126
CAS 137
CAS 149
CAS 155
CAS 160
CAS 161
CAS 167
CAS 169
CAS 172
CAS 174
CAS 175
CAS 176
CAS 177
CAS 179
Component Count
Site area and component data, Unknown Native American Ceramic period
components
Field Site #
TABLE 5.1,
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
160
CAS 183:1
CAS 183:2
CAS 186
CAS 187
CAS 190
CAS 191
CAS 192
CAS 197
CAS 199
CAS 209
CAS 215
CAS 220
CAS 225
CAS 229
CAS 238
CAS 240
CAS 9CAL9(SGC)
CAS9CAL20(SGC)B
CAS 9CAL2(SGC)
CAS 9CAL17(SGC)
CAS 9CAL4(SGC)
CAS 9CAL8(SGC)
Totals (ceramic
n=135)
Totals (all n =259)
Average
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.28
1.76
0.05
1.66
6.64
3.56
4.30
0.09
10.88
0.11
1.12
1.37
1.72
0.12
1.07
0.33
0.42
0.17
0.39
1.30
0.07
0.27
18.74
18.74
0.05
12.60
6.64
3.56
4.30
0.09
10.88
0.11
1.12
1.37
1.72
0.12
1.07
0.33
0.42
0.17
0.39
1.30
0.07
0.27
1
9
100
13
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
56
56
n/a
85.35
85.35
1.52
159.89
522.72
2.86
53
16
88
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
When General Native American Ceramic components are plotted on all settlement
pattern maps, visual associations suggest, but cannot firmly establish, relationships
between the general components and more temporally specific component clusters.
Smaller non-diagnostic occupations were located in upland areas and were probably
short-term encampments or special use zones. However, a few components, such as CAS
183, were probably large occupations for which diagnostics were either overlooked or not
present on the surface at the time of the survey. A complete catalog of ceramics
recovered from these and all other sites is included as part of Appendix C. Generally,
only one or two sherds were present in General Native American Ceramic components.
161
The Woodland Period
Forty-two out of 135 ceramic components date to the Middle, Late, or General
Woodland periods (Table 5.2). As noted in Chapter 3, Early Woodland diagnostics
typical of those found on the Georgia coast area absent from southwestern Georgia. It
appears that the Chickasawhatchee drainage was largely vacant during this period (Wood
1995). Woodland period components comprise 71% of the total site area recorded for
sites on which they are found. These components represent 36% of the area represented
by all sites, irrespective of period.
Twenty-four components cannot be dated to a Middle or Late subdivision of the
Woodland period. As with General Native American Ceramic components, evidence for
these occupations typically consists of fewer than 10 artifacts, and only one diagnostic
sherd (Figure 5.1). In only four cases can component boundaries be distinguished from
site boundaries. In these cases, the components are less than a third of the total site
occupation. General Woodland Period sites are smaller than sites with more specific
temporal designations. More than half of these sites are less than two hectares and all but
three are less than four hectares.
45.04
9.37
100
100
x
x
x
x
General Woodland
% of Site Area
45.04
9.37
Late Woodland
Site Area
2
2
Middle Woodland
Component Area
CAS 5
CAS 89
Woodland Components
Site area and component data, Woodland period components
Field Site #
TABLE 5.2,
162
CAS 122:1
CAS 9DU6
CAS 245
CAS 68
CAS 99
CAS 115:1
CAS 115:2
CAS 124
CAS 162
CAS 182
CAS 184:2
CAS 4
CAS 6
CAS 7
CAS 9
CAS 49
CAS 63
CAS 66
CAS 69
CAS 71
CAS 94
CAS 117
CAS 123
CAS 139:2
CAS 153
CAS 173
CAS 194
CAS 222
CAS 235
CAS 242
CAS 9CAL18(SGC)
CAS 9CAL10(SGC)
CAS 9CAL11(SGC)
CAS 9CAL5(SGC)
CAS 9CAL7(SGC)
Total Areas
(ceramic)
Total Areas (all)
Average
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2.64
1.31
13.62
4.40
2.63
0.60
3.93
0.94
6.75
0.42
2.48
0.37
36.68
0.90
19.36
2.47
0.70
2.39
1.17
1.75
0.20
0.24
0.34
1.74
0.69
2.72
0.11
0.46
6.39
8.99
1.35
0.17
1.22
2.85
2.58
15.35
1.31
13.62
4.40
2.63
15.41
15.41
0.94
8.53
0.42
18.21
0.37
36.68
0.90
19.36
2.47
0.70
2.39
1.17
1.75
0.20
6.95
0.34
9.07
3.14
7.58
0.11
0.46
6.39
8.99
1.35
0.17
1.22
2.85
2.58
17
100
100
100
100
4
26
100
79
100
14
100
100
100
100
100
100
100
100
100
100
3
100
19
22
36
100
100
100
100
100
100
100
100
100
42
42
n/a
189.95
189.95
5.13
267.81
522.72
7.24
71
36
82
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
163
FIGURE 5.1,
General Woodland period ceramic diagnostic distributions
The Middle Woodland Period
Archaic period settlement patterns are beyond the scope of this study, but
comparisons between Middle Woodland period settlement patterns (Figure 5.2) and the
overall pattern represented by the distribution of lithic-only sites in the project area
164
suggests a significant departure from the Archaic period, as settlement consolidates
considerably. Only four large sites and one small ceremonial center can be firmly dated
to the Middle Woodland period (Figure 5.3). These components form three distantly
spaced clusters near areas of major wetland resource concentration. The Mountain (CAS
5), Chickasawhatchee Knoll (CAS 89), Blue Hole Hill (CAS 122), and Hay Fever Farm
(CAS 245) are all consolidated settlements – perhaps functioning as small, sedentary
villages. Each site is located near a key resource.
Given their proximity, it is likely that The Mountain and Chickasawhatchee Knoll
sites were at one time part of related settlements. The Mountain is a one-kilometer-long,
25 m high ridge extending into Hellgate Swamp from the northwest. Although its
geologic origin is unknown, this anomalous feature likely represents the southernmost
extension of the Fall Line Hills into the Dougherty Plain (Beck and Arden 1984).
Ceramics were found in erosional exposures and road cuts along the summit and
base of the hill, indicating that the entire geographic feature was in use to some degree or
another throughout the Middle Woodland. The heaviest artifact concentration was located
on a flat promontory just northeast of the hill summit and facing Chickasawhatchee
Knoll. This high concentration area may be a site discovery bias, rather than being related
to occupational intensity. However, the proximity of this area to Chickasawhatchee Knoll
and The Mountain’s unusually large size (45 ha) argues for the idea that this area in
particular was home to long-term occupation(s).
165
FIGURE 5.2,
Middle Woodland period site distributions
166
Chickasawhatchee Knoll is a 15 m, flat-topped hill with a conical base. It rests on
the same overall landform as the Mountain. Both sites straddle the slough connecting the
main channel of Chickasawhatchee Creek with Hellgate Swamp. Hay Fever Farm is
located on a knoll just above the convergence of Chickasawhatchee Creek and Brantley
Creek. Blue Hole Hill is situated on a large conical rise within 50 m of one of the many
groundwater-fed seeps that give Spring Creek its name. These four sites range in size
from 45 ha (or, if The Mountain was not fully utilized, perhaps a third of that), to 2.5 ha.
Chickasawhatchee Knoll and The Mountain sit at the northern end of a string of hills that
extend an additional 2.6 km to the southeast. These southerly hills are outside the survey
area and it is likely that they were occupied prehistorically, as well.
In contrast to the larger sites, Windmill Plantation (9DU6) is a small isolated
conical mound located on a low hammock between two large open-water ponds.
Although it is conceivable that a resident population could have been supported on a 1.3
ha site, the high probability of flooding makes this unlikely. Instead, Windmill Plantation
appears to be a vacant civic-ceremonial center that was deliberately located in the exact
center of an oval depression (shown in Figure 4.13), in which the shallowness of
underlying limestone creates an abundance of large, open, freshwater ponds that we now
know to be connected to the Florida aquifer (Hicks et al. 1987). More details about
Windmill Plantation are provided below.
Although diagnostics from the Middle Woodland Period are few, it is worth
noting that only two sites, Windmill Plantation and Chickasawhatchee Knoll, exhibit
more than one diagnostic type for this period (Figure 5.3). Of these two sites, only
167
Chickasawhatchee Knoll exhibits all four diagnostic types. As I usually require two
diagnostics for specific period designation, I have doubts about the Middle Woodland
period designations for The Mountain, Blue Hole, and Hay Fever Farm. But, given the
large size and prominence of these sites, their possible occupation during this period
merits emphasis.
FIGURE 5.3,
Middle Woodland period ceramic diagnostic distributions
168
The Late Woodland Period
The Late Woodland period was a time of settlement expansion and dispersal. The
number of securely dated Late Woodland period sites jumped to 13 (Figure 5.4). This
expansion of settlement included not only the expansion of existing settlements, but also
the establishment of new settlement zones. Chickasawhatchee Knoll, Blue Hole Hill, Hay
Fever Farm, and Windmill Plantation were in use, but evidence for a Late Woodland
occupation at the Mountain is equivocal (one eroded sherd that might be Weeden Island
Incised). The absence of significant numbers of Swift Creek Complicated Stamped
pottery in the project area suggests a short regional abandonment between occupations
across the project area, but it is equally possible that people living along
Chickasawhatchee Creek simply did not participate in the macroregional interactions
discussed in Chapter 3 and involving the nearby sites of Kolomoki and Mandeville.
To the southwest of The Mountain, four general Woodland components, each
with a single check stamped sherd, extend in a line – each spaced about 1 km apart.
Chickasawhatchee Knoll and The Mountain were the closest components to a cluster of
General Woodland Components along the west bank of Spring Creek. These components
could be Late Woodland Sites or Early Woodland sites, but a likely scenario is that they
represent a variety of short-term occupations from both periods.
169
FIGURE 5.4,
Late Woodland period site distributions
North-Plowed Hill (CAS 115) is just to the north of Blue Hill Hole – and is
located in a similar in setting. However, instead of being located next to a flowing spring,
this large hilltop site is adjacent to a cypress wetland. Given the clustering in the area, it
170
is likely that the General Woodland sites between Blue Hole Hill and North Plowed Hill
are Late Woodland in date.
Check stamped sherds with incised lines immediately below the vessel lip and a
few Weeden Island sherds are the primary indicators of Windmill Plantation’s continued
use during the Late Woodland period (Figure 5.5). Hay Fever Farm appears to persist as a
Late Woodland settlement, though Late Woodland period diagnostics are scarce at this
site.
FIGURE 5.5,
Late Woodland period ceramic diagnostic distributions
171
Two new settlement clusters between the existing groups indicate expansion of
settlement into previously uninhabited areas – a pattern that coincides with White’s
(1981:666) observed settlement expansion in what is now the Lake Seminole basin. The
Long Walk Site (CAS 162) is primarily a Mississippian mound center. However, a
Weeden Island sherd, a Wakulla Check Stamped sherd, and a variety of Late Woodland
period folded and smoothed rims indicate Late Woodland period settlement (Figure 5.5).
Another new settlement cluster is established between Chickasawhatchee Creek and
Willow Branch, about 8 km north of the Long Walk Site. On the opposite side of
Chickasawhatchee Creek stand three large, flat topped, but very steep mounds with
narrow summits. I was unable to investigate these mounds, known as the Tallassee
Plantation mounds (9DU22), but reports from John Worth and Marvin Smith (personal
communication 2006) suggest mound configurations similar to those found at several
Weeden Island sites in northwestern Florida, including the Letchworth site (Wheeler
2004). Tallassee Plantation is located about one kilometer north of the Hartford Road, a
historic period trail that may have been used during prehistoric periods (Goff 1975).
Aerial photos of the site area dating to 1938 provide no insights regarding mound
dimensions and location. Unlike at Plant Hammond (Chamblee et al. 1998:Figure 3),
where aerial photos show the mound with relative clarity, no such features are discernible
in the area where the Tallassee Plantation are thought to be. More research is needed
concerning this potentially important site.
Late Woodland Period settlement expansion leaves us with several questions.
Even if additional surveys discover Late Woodland sites on the shores of the ponds near
172
Windmill Plantation, the mound site would have been isolated compared to Tallassee
Plantation. Tallassee Plantation fits well into the model of Woodland period mound
locations near major transport routes, as discussed in Chapter 3. However, the location of
Windmill Plantation in the exact center of a complex patchwork of ponds is curious, from
the perspective of land transportation. In contrast, in the likely event that boats were a
major avenue for transport, Windmill Plantation’s location is both prominent and central.
Recent research on Native American boat use makes the latter hypothesis more
plausible. Curci’s (2006) dissertation compiles a variety of ethnohistoric and
archaeological examples for the use of logboats in the southeastern United States. These
examples date from the Archaic period through American Colonialism and include an
excerpt from accounts of the de Soto entrada (Curci 2006:4-5). Archaeological data for
logboat use point to their importance for both transport and wider spheres of wetland
resources exploitation. The deep ponds surrounding the Windmill Plantation mound
would have been exploited via boats – as would the larger channels of Chickasawhatchee,
Kiokee, and Spring Creeks.
The Mississippian Period
Mississippian period settlements are smaller than Woodland period sites overall.
In a survey such as this one, the 39 components recorded for the Mississippian period is
roughly equivalent to the Woodland period’s 42. In spite of the equivalence in frequency,
Mississippian period sites account for only 49% of the total site area occupied where
these components were found (Table 5.3). In terms of total site area, the contrast is even
173
more striking, as Mississippian components only occupy about 18% of the total site area
– a 50% decline from the Woodland Period.
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Unknown Indian/Ceramic
General Mississippian
Late Mississippian
100
100
100
100
100
100
100
100
79
13
9
14
48
52
100
100
100
100
100
100
100
100
100
100
4
26
17
1
5
100
100
Middle Mississippian
4.16
9.37
13.62
0.37
0.12
1.21
2.63
0.14
8.53
7.09
18.21
18.21
6.27
6.27
8.99
0.17
1.22
2.85
0.17
0.44
0.35
4.79
1.07
2.69
15.41
15.41
15.35
9.07
5.99
0.07
0.37
General Woodland
% of Site Area
4.16
9.37
13.62
0.37
0.12
1.21
2.63
0.14
6.75
0.92
1.62
2.48
3.01
3.26
8.99
0.17
1.22
2.85
0.17
0.44
0.35
4.79
1.07
2.69
0.60
3.93
2.64
0.06
0.30
0.07
0.37
Late Woodland
Site Area
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Middle Woodland
Component Area
CAS 10
CAS 89
CAS 245
CAS 4
CAS 27
CAS 74
CAS 99
CAS 121
CAS 162
CAS 163
CAS 184:1
CAS 184:2
CAS 213:1
CAS 213:2
CAS 242
CAS 9CAL10(SGC)
CAS 9CAL11(SGC)
CAS 9CAL5(SGC)
CAS 9CAL13(SGC)
CAS 30
CAS 35
CAS 51
CAS 78
CAS 86
CAS 115:1
CAS 115:2
CAS 122:1
CAS 139:1
CAS 156
CAS 157
CAS 170
Mississippian Components
Site area and component data, Mississippian period components
Field Site #
TABLE 5.3,
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
174
CAS 181
CAS 235
CAS 9DU11(SGC)
CAS 9CAL18(SGC)
CAS 9CAL12(SGC)
CAS 9CAL7(SGC)
Totals (ceramic
n=135)
Totals (all n =259)
Average
1
1
1
1
1
1
3.65
6.39
1.09
1.35
0.16
2.58
3.65
6.39
1.09
1.35
0.16
2.58
100
100
100
100
100
100
39
39
n/a
95.58
95.58
2.58
195.84
522.72
5.29
49
18
77
x
x
x
x
x
x
x
x
x
Although only a third of all Mississippian sites were occupied during the
Woodland period, most settlement shifts were local and did not represent the occupation
of entirely new areas. Unlike General Woodland period components, which could be
represented by a number of diagnostic artifacts, Mississippian components are much
more specific. General Mississippian components are only those represented by Lake
Jackson Jars or Mississippian triangular projectile points, and no other diagnostics.
The Middle Mississippian Period
As noted in Chapter 4, no Early Mississippian period diagnostic ceramics were
recorded in the project area. This observation meshes well with widespread evidence that
interior coastal plain groups continued practicing material culture traditions associated
with the Woodland period well into eras more typically thought of as Mississippian (Blitz
and Lorenz 2006; Schnell 1981; Schnell et al. 1981; Schnell and Wright 1993;
Stephenson 1990; Stephenson et al. 2002). It is even possible that some Late Woodland
period components were contemporary with the earliest Middle Mississippian period
components in neighboring regions.
Unless we accept a scenario in which many Late Woodland settlements were
occupied into the Middle Mississippian period with no noticeable shift in political
175
economy, there was a dramatic consolidation of settlement at this time. Only three sites
are clearly datable to the Middle Mississippian period (Figure 5.6). Interestingly,
however, this consolidation resulted primarily in abandonment of smaller sites. With the
exception of The Mountain, two of the largest settlements remained in use –
Chickasawhatchee Knoll and Hay Fever Farm. At least one of the sites around
Chickasawhatchee Knoll (CAS 9Cal18(SGC)) was likewise established during the
Middle Mississippian period.
Settlement shift involves civic-ceremonial activity centers. Magnolia Plantation
(9DU1) is a three mound Middle Mississippian period civic-ceremonial center located on
the eastern bank of Chickasawhatchee Creek’s easternmost channel. This site was tested
intensively and contrasts with Windmill Plantation in that it is large enough and well
situated enough for permanent inhabitants (Figure 5.7). Magnolia Plantation is the only
site with more than one diagnostic ceramic type from the Middle Mississippian period
(Figure 5.8). As discussed below, this site was densely inhabited, but was abandoned
after no more than a century. A lack of surface visibility prevented identification of sites
surrounding Magnolia Plantation. Given what we know about biased survey coverage,
the probability of nearby contemporary sites is high.
176
FIGURE 5.6,
Middle Mississippian period site distributions
177
FIGURE 5.7,
Topographic context of Magnolia Plantation (9DU1)
Nevertheless, the continuity of occupation around Chickasawhatchee Knoll and at
Hay Fever Farm is a striking contrast to the newly established and short-lived mound
center. Continuity of occupation at large sites located more than 8 km from the nearest
mound center is contrary to the concentrated village layouts that typify many models of
Middle Mississippian period settlement (e.g. Blitz and Lorenz 2006; Williams 1994a).
However, if we consider the possibility that Chickasawhatchee Knoll was itself a site of
civic-ceremonial importance due to its prominence on the landscape (cf. Cobb and Butler
2006; Williams 1999), the continuity is more explicable.
178
FIGURE 5.8,
Middle Mississippian period ceramic diagnostic distributions
179
The Late Mississippian Period
In cyclic fashion, the Late Mississippian period follows the Late Woodland
pattern as another period of expansion and settlement shift (Figure 5.9). The number of
sites dating to the Late Mississippian period increases to 13, but component size declines
– suggesting settlement dispersal. Except for the Red Bluff and the Long Walk Sites, both
discussed below, no component is larger than 3 ha.
As noted in Chapter 4, further ceramic analysis may eventually lead to defining
two distinct sub-periods in the project area. In this scenario, sites exhibiting only Fort
Walton Incised vessels or bowls with carved appliqué strip designs might occupy a
period between the abandonment of Magnolia Plantation and the establishment of
settlements where occupants produced ceramics associated with the Lamar decorative
designs of what is now northern Georgia. If this proves true, Hay Fever Farm and the area
around Chickasawhatchee Knoll would again represent zones of occupational continuity
(Figure 5.10).
In any case, a combined pattern of decreasing site size and settlement expansion
coincides with another round of settlement shift. Only 60% of Late Mississippian sites
are established in locations with prior occupations. Of those ten sites, only
Chickasawhatchee Knoll and Hay Fever Farm have any evidence for Middle
Mississippian period occupations. When comparing the northern two-thirds of the project
area, with the southern third, there is a noticeable difference in how these settlement
shifts occur between the Middle and Late Mississippian periods. In the area within about
180
5 km of Chickasawhatchee Knoll, settlements shift very slightly, with differences
between earlier and later sites amounting to less than a kilometer.
FIGURE 5.9,
Late Mississippian period site distributions
181
FIGURE 5.10, Late Mississippian period ceramic diagnostic distributions
The Red Bluff Site (CAS 242) is a large, concentrated occupation less than onehalf of one kilometer southwest of Chickasawhatchee Knoll. This site suggests that
occupational continuity was accompanied by local settlement expansion. The Red Bluff
182
Earthlodge Site may simply be a series of small houses connected to Red Bluff, but it
could also have been a small ceremonial area – deliberately separated from the main
settlement (cf. Nass and Yerkes 1995). Apart from this small potential ritual site,
discussed in greater detail in the next section, and potentially Chickasawhatchee Knoll
itself, this area of highest settlement concentration had no artificial Mississippian civicceremonial architecture that remains visible about the ground surface. To the north, civicceremonial architecture was present, but settlement continuity was reduced, and
settlement shifts involved greater distances.
Magnolia Plantation was abandoned after the Middle Mississippian period, but
the Chickasawhatchee drainage was not without a Late Mississippian period mound
center. Late Mississippian diagnostics from the Long Walk Site (CAS 162) include Fort
Walton Incised and Lamar Complicated Stamped pottery. Two land managers say that a
flat-topped mound was bulldozed on the site prior to 1980s. One manager remembers the
date as 1954 and the other as 1978. Aerial photographs suggest an actual date between
1964 and 1969. According to the both managers, the Long Walk Mound was destroyed
when the area was cleared for agriculture. In the 1964 photo, the area around the mound
was forested. The 1969 photo shows a cleared agricultural field and no evidence of a
mound. Independent reports of the mound’s existence are supported by high ceramic
frequencies – the second highest among surface collected sites in the project area.
There is a Late Woodland period component at Long Walk, so we must consider
Schiffer’s (1996:260) hypothetical scenario in which a Late Woodland period mound is
bulldozed onto a Mississippian village. The relative under-representation of Late
183
Woodland pottery and the large number of Mississippian period diagnostics argue against
this idea. North of the Long Walk site, across from Tallassee Plantation, Late Woodland
sites between Willow Branch and Chickasawhatchee Creek were re-occupied. The reoccupations may suggest re-use or continued use of the Tallassee Plantation mounds as
well.
By comparing Figures 5.9 and 5.10, some general Mississippian period
components can be provisionally assigned to Early or Late subperiods on the basis of
spatial association. Sites near Long Walk (CAS 162) and Willow Branch (CAS 184) are
spatially associated with the Late Mississippian period. Sites within about 4 km of
Magnolia Plantation (9DU1) on the east side of the creek are probably Middle
Mississippian period components. Sites within the Chickasawhatchee Swamp itself and
along Spring Creek could be associated with short-term occupations dating to either or
both subperiods.
Summary of Regional Settlement Patterns
Settlement pattern trends spanning the occupation history of the project area are
better understood when combined with data from intensive, site-level testing. These data
are taken up in the next section. At the regional scale, three trends stand out through the
entire occupation sequence. These trends relate to the duration and intensity of
occupation around mound centers, the contrast between these centers and other prominent
sites in the project area, and the overall distribution of settlement.
From period to period, civic-ceremonial centers in the project area shift in
location and prominence. In spite of biased survey coverage, available evidence suggests
184
that settlements did cluster around the mound centers at Windmill Plantation, Tallassee
Plantation, Magnolia Plantation, the Long Walk Site, and the Chickasawhatchee
Knoll/Red Bluff Earthlodge cluster. Among these prominent sites, only the settlements
clustered around Chickasawhatchee Knoll escaped cycles of subregional abandonment.
Assuming that the entire region was not abandoned from A.D. 350 – 650 during
the height of Kolomoki’s occupation, only two settlements were continuously occupied
throughout the project area’s settlement history: Hay Fever Farm and Chickasawhatchee
Knoll. Hay Fever Farm’s role in the settlement system cannot be fully understood without
additional survey, but Chickasawhatchee Knoll is clearly an anchor in the regional
settlement system. These settlements both stand out in terms of size and ceramic
diversity, despite having no artificial above-ground ceremonial architecture.
During the Mississippian period, the cycle of abandonment strongly resembles the
cycles of abandonment characteristic of paired towns in the Oconee and Chattahoochee
Valleys (Blitz and Lorenz 2006; Williams and Shapiro 1990). As noted in Chapter 2,
these researchers disagree about the cause of the paired town phenomenon. Williams and
Shapiro (1990) favor a model built around local ecological degradation, while Blitz and
Lorenz (2006) favor a model driven by the internal politics of chiefdoms. Given available
data, their arguments cannot be resolved here.
Overall, the project area’s component size distribution is bi-modal for each
period. Three to five sites located near key wetland resources are consistently much
larger than their counterparts. Chickasawhatchee Knoll and Hay Fever Farm are always
among these sites. Additional sites include Blue Hole Hill and North Plowed Hill for the
185
Woodland period, as well as Magnolia Plantation, Long Walk, and Red Bluff for the
Mississippian period. From the Woodland to Mississippian period, settlements also
became smaller and more dispersed. Implications of these trends are better understood
when complemented by data from intensive mapping, testing, and excavations.
Intrasite Settlement Patterns: Four Examples
According to regional data, Windmill Plantation, Magnolia Plantation, the Red
Bluff Earthlodge, Chickasawhatchee Knoll, and Hay Fever Farm are five of the most
prominent sites in the project area. As mound centers, Windmill Plantation, Magnolia
Plantation, and possibly the Red Bluff Earthlodge, represent short-term focal points for
civic-ceremonial activity and the management of power relationships. Chickasawhatchee
Knoll and Hay Fever Farm draw their prominence from continuity of occupation. Despite
their importance, these sites vary greatly in their size, apparent intensity and duration of
occupation, and overall settlement structure. This section reports on mapping and testing
projects at these sites, highlights the contrasts between the sites, and raises broader
comparisons useful for synthesizing intrasite and regional data.
Windmill Plantation
Windmill Plantation is a Woodland period civic-ceremonial site located between
several large open ponds in near the eastern edge of the project area, close to
Cooleewahee Creek (Figure 5.11). The overall site size is about 0.4 ha The mound itself
is conical, about 35 m in diameter at its base and about 2.5 m high (Figure 5.12).
According to Mark Williams (personal communication) and the records on the artifact
bags, the mound was first excavated in 1962 by Don and Betty Smith, but these
186
excavations were never reported. Figure 5.11 shows a shallow trench on the mound’s
eastern flank – a likely result of the Smith excavations.
FIGURE 5.11, Topography and Inverse Distance Weighting surface of ceramic density, 9DU6
Intrasite Patterns
Mapping and shovel testing at Windmill Plantation was executed in a modified
radial pattern (cf. Williams 1999). Crews excavated 108 shovel tests on the site, first in a
radial pattern away from the mound, then following a grid along the mound’s eastern
skirt, where the first sequence of tests revealed the highest artifact frequency.
Eighty-five percent of the site perimeter is bounded by open water. The only
entrance by land is a narrow strip of land extending to the northeast. Apart from this
187
easterly approach, the site was most accessible via boat crossings from settlements on
opposite sides of the site’s flanking ponds. There is a wide trough dividing the landform
to the east of the site’s zone of highest artifact concentration. Such an area would likely
have served well as a boat landing.
FIGURE 5.12, The Windmill Plantation mound
Discussions with land owners and land managers suggest that the eastern and
southwestern portions of the site were partially cleared in preparation for logging, but that
disturbance was limited because the clearing was never completed. Apart from Don
Smith’s backdirt piles on the eastern flank of the mound, and a small looter’s trench on
the northern side of the mound, no disturbance seems to have come within about 10 m of
the mound itself.
188
Artifact frequencies across the site are generally low (Table 5.4). Shovel tests are
dominated by plain and check stamped sherds, with just a few Middle and Late
Woodland period diagnostics within about 20 m of the mound. In some cases Middle and
Late Woodland diagnostics occur in the same shovel test, suggesting that overall use
patterns at the site changed little from period to period. Shovel tests revealed two areas of
high artifact concentration. As the concentration east of the mound was the largest, it was
selected for test excavations in non-mound areas.
0
0
0
0
0
0
0
0
0
0
0
0
1962 Excavations
0
8
1
0
6 36
0
1 144
Alachua Cob Marked
Weeden Island Punctated
0
1
1
0
0
0
0
0
2
0
1
3
0
0
0
0
0
0
0
0
0
0
1
1
0 2
0 7
0 1
0 0
0 0
0 0
0 0
0 0
0 10
0 4
1 3
1 17
0
5
4
1
1
0
0
0
11
0
3
14
3 18
0 20
0 25
0 10
0
5
0
1
0
2
0
1
3 82
0 17
0 47
3 146
0
2 34
0
0 232
0
Totals
2 5
0 3
3 5
1 1
1 1
0 0
2 0
0 0
9 15
6 0
13 8
28 23
UID Plain
UID Stamped
Weeden Island Incised
0 4
0 4
1 8
0 4
0 1
0 0
0 0
0 1
1 22
1 5
0 10
2 37
UID Decorated
Wakulla Check Stamped
2
0
0
1
0
0
0
0
3
1
2
6
UID Complicated Stamped
Ocmulgee/Fairchild Cord Marked
0
0
0
0
0
0
0
0
0
0
0
0
Other
Deptford Simple Stamped
0
0
1
1
1
1
0
0
4
0
2
6
Eroded
Deptford Linear Check Stamped
0
0
1
1
0
0
0
0
2
0
3
5
Level #
2 1 Level 1
2 2 Level 2
2 3 Level 3
2 4 Level 4
2 5 Level 5
2
E Wall Profile
2
Floor Cleaning
2
N
XU 2 Totals
1
All Levels
Shovel Tests
9DU6 Totals
Unit #
Alligator Bayou Rocker Stamped
Swift Creek Complicated Stamped
Ceramic type distribution from excavation units and shovel tests, 9DU6
Level Name
TABLE 5.4,
Crews excavated two test trenches on the site. Excavation Unit 1 (XU 1) profiled
a small looter’s trench excavated into the lower skirt of the mound’s northern edge. Only
189
17 sherds were collected from this trench. Mound fill in this location consisted of
undifferentiated brown sand with no visible depositional stratigraphy.
Excavation Unit 2 (XU 2) was a two-by-two meter text pit excavated in 10 cm
arbitrary levels in the zone of highest artifact concentration located east of the mound.
The unit was excavated to a depth of 60 cm below the surface and terminated after the
last five centimeters of Level 5 revealed no artifacts. Because the mass of embedded and
organic and root matter near the ground surface was so deep, Level 1 was excavated to
depth of 20 cm. This excavation unit’s soil profile is described in detail in Chapter 4.
Below the organic yellow, yellow and brown sands graded into a yellow, orange, and
gray clay. This transition from sand to clay was complete at 45 cm below the surface.
Artifact concentrations in XU 2 are low, with only 82 sherds recovered from all
levels (Table 5.4). Artifact distributions through the unit from XU 2 support the idea,
presented in Chapter 4, of two distinct occupations dating the Middle and Late Woodland
periods respectively. Middle Woodland period diagnostics are most concentrated in Level
4, but are found in Levels 3 – 5. Conversely, Late Woodland period artifacts are most
common in Levels 1 and 2, but are also found in Level 3. Diagnostics for each period
were discussed in Chapter 3, but the paucity of Swift Creek Complicated Stamped pottery
is noteworthy and discussed below. Bioturbation and trampling probably accounts for
some of the mixing of artifacts between levels (e.g. Brooks and Sassaman 1990;
Stephenson 1990).
190
Windmill Plantation in Regional Context
Data regarding the site plan, artifact distribution, and stratigraphy support the
hypothesis that Windmill Plantation was a vacant civic-ceremonial site that was the locus
of occasional ritual use during the Middle and Late Woodland periods. A lack of Swift
Creek Complicated Stamped pottery may indicate a possible gap in settlement coincident
with the ascendancy of Kolomoki, but there are other possibilities to consider. At first
glance, the argument for non-participation in broader exchange networks at a mound site
seems less plausible than a possible period of disuse. However, there are also other, more
mundane explanations.
Milanich et al. (1997:87) argue that so-called “elite ceramics” were items of
restricted use and subject to control by a small and select group. They stand by this
argument in spite of a clear pattern showing wide intra-site distribution of these vessel
forms (Milanich et al. 1997:84-87; Pluckhahn 2003). Jeffries (1976:48 and 1979:170)
also argues that “Hopewellian” items are restricted to burial contexts. As Smith’s
excavations were quite shallow, and our excavations were not on the mound, it is possible
that some of the ceramic variation at Windmill Plantation has not been properly
documented. The arguments by Milanich et al. (1997) are not as convincing those of
Jeffries (1976), but even so, a complete view of the site’s occupation history awaits future
and more extensive excavations.
The exact life-history of the site notwithstanding, Windmill Plantation is a bit
unusual in terms of its setting. Other Woodland period mounds, potentially including
Tallassee Plantation, are typically located on somewhat higher ground or along what may
191
be major transport routes (e.g. Jeffries 1976; Milanich et al. 1997). Milanich et al.
(1997:203) note the Carter Complex is adjacent to a similar pond but this site, too, differs
from Windmill Plantation in that there is a large associated village complex adjacent to
the mound.
In terms of its out-of-the-way location, Windmill Plantation resembles Kolomoki.
Pluckhahn (2002:46) argues that Kolomoki served as a territorial marker between two
widely scattered groups. Given Windmill Plantation’s location in the center of the pond
network southwest of Kiokee Creek, this site may have served a similar function for
those making use of the resources in the ponds.
Magnolia Plantation
Magnolia Plantation is a 4.16 ha. Mississippian Period mound and village site
located about two kilometers east of Chickasawhatchee Creek, at the base of fall line
hills. The site is dominated by a large, well preserved platform mound (Figure 5.13).
Broad swamps that are part of the Chickasawhatchee Creek floodplain surround the site
on three sides. Magnolia Plantation was originally visited by Robert Wauchope (1939) in
the late 1930s. He recorded the site as “Pine Island,” a misnomer according to locals who
apply that name to sand hammock located about five miles to the south. Wauchope
(1939) also noted the possibility that the site was composed of three mounds.
During investigations related to the southern portion of the revised Hernando De
Soto route, John Worth and Marvin Smith visited the mound with a local avocational
archaeologist, Eugene C. Black, who had been to the site many years before (Black 1977;
Hudson et al. 1984, Worth 1989). Frankie Snow (n.d.) also visited the site in 1986.
192
However, none of these archaeologists could confirm Wauchope’s observations
concerning the number of mounds on the site.
FIGURE 5.13, Mound A at Magnolia Plantation, facing SSE
The landowners of Magnolia Plantation have been protective of the main mound.
Prior requests to conduct fieldwork have been denied. In this case, land managers
allowed us to complete a map, conduct limited testing to determine site boundaries, and
excavate two test units. This work established the site’s boundaries and confirmed that
there are three mounds at the site (Figure 5.14).
Mapping
The largest mound is designated Mound A. It is three meters high, square and flattopped, with 39 m wide base and 27 m wide on the summit. Except for a few places on
193
the mound slopes where old hardwood trees have collapsed, the mound is undamaged and
there is no evidence of looting. To the south of Mound A is Mound C, a low, eroded rise
similar to many of the smaller mounds on the perimeter of Macon Plateau (cf. Williams
1994c). Mound C is just over one meter high and is 20 m in diameter.
FIGURE 5.14, Topographic map and shovel test distributions of Magnolia Plantation (9DU1)
Mound B is located 270 km West/Northwest of Mound A and is located just on
the edge of a railroad track that is been operative since before 1870 (Figure 5.15). Most
of this mound was has been destroyed, presumably when the railroad was built. The
remains of Mound B are also three meters high, but are only 27 m wide along the base of
the long axis and only 15 m wide along the short axis due to the impact of the railroad.
194
The mound summit is 9 m x 20 m. If this mound were as undisturbed as Mound A, it still
would have been slightly smaller.
FIGURE 5.15, Mound B at Magnolia Plantation, Facing NE
Although the apparent two mound layout of this center is similar to others in the
lower Appalachian southeastern United States, mapping reveals some important
differences. First, the mounds are separated by two ribbons of wetland. Railroad
construction may have made these wetlands larger than they were during the site’s
occupation, but it is unlikely the entire area between the mounds was high ground prior to
the site’s disturbance. Additionally, the mounds are not oriented in line with each other.
This characteristic is uncommon. Other differences pertaining to artifact distributions are
summarized below.
195
Shovel Testing
Landowner reluctance to allow extensive work restricted the shovel test program
at Magnolia Plantation to 53 tests. Many tests were excavated immediately around
Mound B in an effort to locate deep deposits of high artifact frequencies suitable for a test
excavation. Additional tests were excavated in radial lines away from the mounds as a
means to determine site boundaries.
Across the site, buried artifacts are found at shallow depths and mixed sandy loam
soils suggest heavy disturbance over time by earth moving equipment. Artifact
frequencies in many shovel tests were much higher than at other sites, but this result may
again be due to the large number of tests excavated around Mound B. Shovel tests near
the edge of a small channel that now divides the site showed evidence of deep, high
frequency deposits.
This concentration may have resulted from historic period disturbance, but a
cause is difficult to determine without sedimentological and geochemical analysis.
Interestingly, shovel tests excavated between the mounds all contained some artifacts.
This pattern raises the possibility that Magnolia Plantation deviated from typical twomound Mississippian centers through the absence of an open, clean plaza between the
mounds (cf. King 1996; King and Stephenson 2001; Williams 1999). It is entirely
possible, however, that the plaza was destroyed by the railroad. A systematic grid of
shovel tests might shed more light on this issue.
196
Excavations and Mound Cores
Crews excavated two stratigraphic test units at the site – one on Mound B and one
in the area of highest artifact concentration near Shovel Test 38. Excavation Unit 1 (XU
1) was a one-by-two meter stratigraphic test trench excavated into western skirt of
Mound B. Two deep mound cores were also excavated two meters inward from the
northwestern and northeastern corners of Mound B (Figure 5.16). Excavation Unit 2 (XU
2) was a one-by-one meter test pit excavated in a zone of high artifact concentration 10 m
from the edge of the channel bisecting the site (Figure 5.14). This unit yielded little
information in terms of visually discernible strata. Artifact data from XU 2 are addressed
below.
FIGURE 5.16, Topographic map of Mound B
197
Observations from shovel testing and mapping around Mound B documented
extensive disturbance and provided challenging conditions for the excavation of a test pit
to record mound stratigraphy. There was a deep looter’s trench on the eastern skirt of the
mound, but I decided that cleaning it would have destabilized the mound more than an
excavation unit I could backfill effectively. The high flat prominence extending westward
on the lower portion of the mound’s western edge seemed likely to be displaced mound
fill. The southern flank of the mound was destroyed by the railroad and the northern flank
has long-since collapsed. In the end, XU 1 was excavated high up the mound’s western
skirt because of the likelihood of encountering intact stratigraphy.
Figures 5.17 and 5.18 illustrate stratigraphic differentiation in the mound fill
along the eastern wall of the excavation unit profile. This profile consists of eight distinct
stratigraphic layers, along with a number of inclusions that may represent features. The
lowest layers of the unit consist of thin horizontal lenses of grey, yellow, and brown
sands. These sands are capped by a 1 cm vein of fine white clay that is probably refined
or crushed kaolin.
The white kaolin vein is topped by a thick yellow clay cap. The yellow clay is in
turn overlain by dark red clay layer of roughly equal thickness that also exhibits small,
mottled yellow clay inclusions. The dark red clay layer intrudes into the yellow clay
layer. The nature of this intrusion is not clear. If this had been a post mold, one might
expect some indication of decayed wood or charcoal. However, the feature is consistent
with a post that had been deliberately removed and filled in. Two thin layers of yellow
198
and orange clay intrude the profile from the west and underlay a dark brown sandy clay
rich in organic matter that caps the dark red clay.
FIGURE 5.17, East Profile Drawing, Mound B, XU 1, 9DU1
199
FIGURE 5.18, East Profile Photo, Mound B, XU 1, 9DU1
Figures 5.19 and 5.20 illustrate stratigraphic differentiation in the mound fill
along the southern wall of the excavation unit. Strata from this lateral profile of the
mound flank reflect those observed in the eastern profile, but with added complexity that
might reflect mound construction technology. The mixed lensed sands at the bottom of
200
the unit are intruded by a layer of kaolinitic white clay and sand. A layer of maroon sand
is underneath the yellow clay layer, and intermixed orange and red clays occur in solid
units; these likely represent distinct sequences of fill dumped by the basket load as the
mound was built.
FIGURE 5.19, South Profile Photo, Mound B, XU 1, 9DU1
The two mound cores present an additional view of the variation in the mound fill
(Shovel Tests 23 and 30; Figure 5.21). Both cores were excavated from the summit in an
area of roughly equivalent elevation. Each was excavated to depth of approximately four
meters, thus penetrating the mound completely and extending through the mound subsoil
and to the water table or bedrock.
201
FIGURE 5.20, South Profile Drawing, Mound B, XU 1, 9DU
Shovel Test 23, on the northwestern corner of the mound, extends to 4.25 m
below the summit. Possible occupational debris, including daub, charcoal and pottery,
occur continuously between approximately 70 cm and 240 cm. These midden layers of
orange sandy clay and yellow sandy loam overlay a thick, homogeneous clay red clay
mantle. Below the red clay mantle are pre-mound soils similar to non-mound shovel tests.
Shovel Test 30, on the northeastern corner of the mound contrasts with Shovel
Test 23. Ceramics are less frequent and more dispersed through the test, beginning at 30
cm below the summit. There is an apparent gap between these ceramics and the test’s
202
greatest concentrations of possible occupational debris between 110 cm and 170 cm.
Below this midden layer are alternating layers or yellow and tan sands, separated by a
white sand and kaolin clay layer similar to that observed in the mound profiles. Beneath
an additional layer of white sand are premound soil layers that extend to a large rock at
380 cm. There are a variety of confounding factors when considering the occupational
sequence of any Mississippian period platform mound. Many mounds begin with sand fill
overtopped by a clay cap (e.g. Coe 1995; Rudolph and Hally 1986). Construction stages
that do not represent actual occupation surfaces are also common (cf. Schnell et al. 1981;
Smith 1995). In addition, (2001) recent work by Anderson and Cornelison (2001) on the
importance of color symbolism in prepared clay mound surfaces and the reports from
Vogel et al. (2005) concerning ritual pit excavation on mounds indicate that mound
alteration may have occurred more regularly than if construction stages were only added
after an episode of chiefly succession (cf. Hally 1996).
These potential complications are reflected in the contrast between the adjacent
eastern and southern profiles and the mound cores. As such, they are illustrative of the
difficulties associated with interpreting mound construction on the basis of such small
snapshots of mound stratigraphy. In the profiles, the most likely candidate for an
occupation surface is the 20 cm layer of dark brown sandy loam at the top of the eastern
profile. The mound cores show clear evidence of midden layers between 30 cm and 210
cm below the summit. However, given that Mound B is located near the edge of the site
and that Shovel Test 23 faces away from the probable village center, it is possible that the
203
broad midden in this test represents part of a dump on an earlier summit (cf. Williams
and Smith 1994).
FIGURE 5.21, Soil profile schematics from posthole tests of Mound B, 9DU1
204
The frequent appearance of thin, clean, white clay and sand layers, suggests ritual
surface preparations (cf. Chamblee 1997; Rudolph and Hally 1986). Whether these
surfaces were cleaned occupation layers or ritual preparations that were then immediately
covered cannot be determined without additional excavation. The intrusive red clay
feature in the eastern profile suggests the possibility of an occupation surface on the
yellow clay cap, but the lack of occupational debris between these layers means that this
interpretation is by no means certain. Taken together, stratigraphic data provide clear
evidence for only one occupational stratum, but could support interpretations involving
up to four stages. The thick midden near the mound summit suggests multiple remodeling
episodes on the uppermost stage. Even if these remodeling episodes are not, as Anderson
and Cornelison (2001) suggest, part of ongoing maintenance, the overall occupation of
the mound could still be short – no more than 50 or 100 years.
Ceramic Data
Ceramic data are presented in Table 5.5. The lack of typological variation
between different excavation zones is striking. Ceramic frequencies from XU 1 suggest
that, even if Shovel Test 23 represents a mound flank midden, the test unit excavation
unit did not sample it. Frequencies from XU 2 suggest the possibility of excavation into a
village midden or occupation layer, but analysis of sediment samples should be
performed to verify stratigraphic integrity.
205
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
0
0
1
0
0
0
2
1
5
4
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
1
1
1
0
0
0
3
1
5
Savannah Complicated Stamped
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Lake Jackson Plain
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
Lake Jackson Incised
Ingram Plain
XU 1, Level 1
XU 1, Level 2
XU 1, Level 2, Profile Cleaning
XU 1, Level 3
XU 1, Level 3, Profile Cleaning
XU 1, Level 4
XU 1, Level 5
XU 1, Level 6
XU 1, Level 7
XU 1, Level 8
XU 1, Level 9
XU 1, Level 10
XU 1, Level 11
XU 1, Level 12
XU 1, Level 13
XU 1, Level 14
XU 1 Totals
CAS 10 XU2, Level 1
CAS 10 XU 2, Level 2
XU 2, Level 3
XU 2, Level 4
XU2, Level 5
XU 2, Level 6
XU 2, Level 7
XU 2 Totals
Shovel Tests
CAS 10 Totals
Lake Jackson Decorated
Cool Branch Incised
1
2
3
4
5
6
7
Columbia Incised
2
2
2
2
2
2
2
Carabelle Incised
1
2
2
3
3
4
5
6
7
8
9
10
11
12
13
14
Andrews Decorated
Level #
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Ceramic type distribution from excavation units and shovel tests, 9DU1
Level Name
Unit #
TABLE 5.5,
0
0
0
0
0
0
0
0
0
0
1
3
0
0
0
0
4
0
0
0
0
0
0
0
0
0
4
206
Totals
1
2
3
4
5
6
7
Other
2
2
2
2
2
2
2
Eroded
3
4
5
6
7
8
9
10
11
12
13
14
UID Stamped
1
1
1
1
1
1
1
1
1
1
1
1
UID Incised
2
3
13
11
0
0
0
0
0
1
3
4
0
0
1
2
2
15
0
0
0
0
0
1
0
0
0
0
0
0
1
4
6
7
6
6
5
1
1
1
0
0
79
41
65
25
12
12
5
0
160
163
402
0
0
0
0
1
0
0
0
0
0
0
0
1
0
2
0
0
0
0
0
2
3
6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
4
5
0
0
0
1
0
0
0
0
1
0
0
0
4
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
0
0
7
2
6
0
0
0
1
0
9
2
18
0
0
0
3
0
0
1
0
0
0
0
0
4
0
0
0
0
3
0
0
3
6
13
0
0
0
1
0
0
2
3
0
0
0
0
9
1
1
1
2
0
1
0
6
6
21
UID Decorated
1
1
XU 1, Level 1
XU 1, Level 2
XU 1, Level 2, Profile
Cleaning
XU 1, Level 3
XU 1, Level 3, Profile
Cleaning
XU 1, Level 4
XU 1, Level 5
XU 1, Level 6
XU 1, Level 7
XU 1, Level 8
XU 1, Level 9
XU 1, Level 10
XU 1, Level 11
XU 1, Level 12
XU 1, Level 13
XU 1, Level 14
XU 1 Totals
CAS 10 XU2, Level 1
CAS 10 XU 2, Level 2
XU 2, Level 3
XU 2, Level 4
XU2, Level 5
XU 2, Level 6
XU 2, Level 7
XU 2 Totals
Shovel Tests
CAS 10 Totals
UID Complicated Stamped
1
2
UID Plain
Level #
1
1
Level Name
Unit #
TABLE 5.5 (continued)
2
2
The distribution of ceramic types from all XUs is consistent regardless of the
stratigraphic context. Data from shovel testing and excavation units indicate a Middle
207
Mississippian period occupation consistent with general descriptions of the Late Rood
Phase (A.D. 1100 – 1200) (Blitz and Lorenz 2006; Schnell et al. 1981). A small sample
size limits considerations of ceramic variation in terms of decorative design and vessel
form. It is still worth noting, however, that Andrews Decorated is present only at
Magnolia Plantation. Andrews Decorated vessels are beakers or cups with incised
exteriors that have been found in contexts at Cemochechobee suggesting ritual disposal
patterns (Schnell et al. 1981:69-72; cf. Walker 1999). Savannah Complicated Stamped, a
decorative design usually found on large storage jars throughout the piedmont and coastal
zones (Williams and Thompson 1999:108), likewise only appears at Magnolia Plantation.
Magnolia Plantation Regional Context
At 4.16 hectares, Magnolia Plantation is comparable to a number of other Middle
Mississippian period mound centers located along the Georgia Fall Line (see Chapter 6).
Although the site has been severely disturbed, evidence may suggest an unusual
circumstance in which this multi-mound center lacks a central plaza. The situation would
be even stranger if all three mounds were not in simultaneous use – an unlikely scenario.
Stratigraphic evidence from shovel tests, XU 1, and mound excavations all suggest a
relatively short-term occupation characterized by a narrow subsurface occupation zone.
Ceramic evidence corroborates the notion of a short-term occupation, as all diagnostics
coincide with the Chattahoochee Valley’s Middle Mississippian period Rood Phase.
Magnolia Plantation’s location over eight kilometers from the largest known,
continuously occupied population centers is noteworthy. This fact, along with the site’s
small size, slightly greater ceramic diversity, and short-term occupation all suggest that
208
this center of Mississippian ideology, power, and economy was deliberately set apart
from the project area’s greater settlement system. Possible reasons for such separation
require a broader, more comparative view, and are taken up later.
Red Bluff Earthlodge
The Red Bluff Earthlodge (9BX4) site is a small series of low mounds on a
narrow hammock near the eastern bank of Chickasawhatchee Creek (Figure 5.22). The
site is located to the south of small slough feeding the creek – 1.5 km above the
confluence with Spring Creek. The Red Bluff Earthlodge was first recorded by Eugene C.
Black (1977). Black and John Worth visited the site in the 1980s and recorded a linear
arrangement of three defined landscape prominences, including one doughnut-shaped rise
at the southern end of the site (Worth 1989). Black and Worth interpreted these features
as mounds, though, at the time, no artifacts were recorded and true site boundaries were
not established. The goals of this project’s investigations at 9BX4 were to understand the
nature of any prehistoric above ground architecture, determine chronological and spatial
site boundaries, and develop a preliminary view of intrasite patterns.
Investigations show that at least one and possibly two of the three above ground
features are the remains prehistoric architecture. Testing determined that the site is 0.53
ha in size, with General Woodland period and Late Mississippian period components
together. Artifact densities are very low, especially away from the surface features
(Figure 5.23). Red Bluff Earthlodge is located among the largest cluster of Late
Woodland period and Late Mississippian period components in the project area. Red
Bluff Earthlodge is than less than 1 km from the major components at Red Bluff,
209
Chickasawhatchee Knoll, The Mountain, and Hoek’s Hole (CAS 1), among others. As
such, Red Bluff Earthlodge components are best viewed as integrated parts of a larger
community.
Mapping
The Red Bluff Earthlodge is bounded on the north, west, and east sides by cypress
and tupelo wetlands. The southern site boundary is a clear line delineated by a pine
plantation estimated to be less than 15 years old. The site may have extended further to
the south, but any traces of it have been destroyed by logging. Pedestrian survey of an
open food plot 150 m to the south located no artifacts. As the 10 cm contours show, there
is very little relief. In order to document surface features, 286 elevation readings were
taken with a total station using a radial pattern centered on the approximate center of the
small landform supporting the site. The maps (Figure 5.23) suggest three low mounds,
including the large doughnut shaped feature documented by Black (1977). Looter’s pits
intrude upon two of the mounds (Structures 2 and 3).
Shovel Testing
In response to low artifact densities and the site’s small size, shovel tests were
executed at 10 m intervals, rather than the usual 30 m intervals. One hundred and
seventeen shovel tests were excavated to depths between 50 and 70 cm. Some ceramics
were recovered. A deeply deposited Archaic period component and associated lithics
were also documented. Interestingly, the site’s landform is comprised of shallow brown
sand underlain by a thick red clay layer that was observed in disturbed areas on several
other sites deep in the swamp.
210
FIGURE 5.22, Topographic and site distribution context, Red Bluff Earthlodge (9BX4)
Artifact concentrations were moderate around the potential mounds on the site.
The highest artifact concentration is on the site’s northwestern edge. The steeper
elevation and proximity to the swamp of this concentration suggests a possible dump, but
211
further testing would be required to confirm this supposition. Shovel tests in Structure 1
were equivocal as to whether this feature is man-made. There were several artifacts in
one test on the western side of the feature, but the feature’s soil profile strongly
resembled that of the surrounding terrain. Shovel tests in Structure 3 revealed high
artifact concentrations and reddish sands and clays intermingled with small quantities of
charcoal – all good indicators of human construction. Because Structure 2 had large
looter’s holes in its summit, a large excavation unit to profile one hole was substituted for
shovel tests.
Excavations
Five excavation units were carried out on the site. The cleaning of looters pit on
Structure 2 (XU 1) revealed little. A solid red clay matrix beneath 10 cm of brown sand
and organic materials is not visually distinct from the surrounding subsoil. A few quartz
flakes and plain sherds were interspersed throughout this clay layer. Quartz does not
occur naturally in southwestern Georgia and there are records of quartz materials being
used in Woodland period ritual activity (cf. Jeffries 1976 and Pluckhahn 2003). Given the
paucity of quartz across the project area, these flakes represent the best support for the
hypothesis that Structure 2 is a Woodland period mound.
Excavation units in Structure 3 were considerably more instructive. The looter’s
pit in the southern third of the feature provided a means to create a quick profile. The
trench into this profile (XU 2) revealed a clear midden layer, flat-laying sherds, and
charcoal. This initial trench was extended as a one-by-two meter stratigraphic test pit that
212
extended from the south wall of XU 2 to the edge of the feature (XU3). Figure 5.24
illustrates the results from the XU 3 excavation.
FIGURE 5.23, Topography and Inverse Distance Weighting surface of ceramic density, 9BX4
Less than 10 cm below the ground surface lays a layer of dark brown cultural fill
that extends about 25 cm southwards into the excavation unit. The fill rests at an angle
213
against a uniform layer of red, part of which was likely the original ground surface and
part of which formed an earthen structural support berm. Beneath the cultural fill lay
burned daub and charred timbers identified by Michael Diehl (personal communication
2005) as Slash Pine (P. elliottii).
FIGURE 5.24, Profile Drawing, Level 4, XU 3, Structure 3, 9BX4
A plan view of XU3 (Figure 5.25) shows the lateral configuration of daub, fill,
and ceramics. The northernmost section of the unit is likely internal fill from the
structure, now lightly disturbed by looting activities. In between this fill and the outer
layers of red clay are the remains of a collapsed wall. The large burned timber intruding
through the west wall of the unit exhibited visible wood grain alignment pointing to the
southeast, suggesting that the structure may have collapsed outward in this area.
214
FIGURE 5.25, Plan View Drawing, Level 4, XU 3, Structure 3, 9BX4
Two of the large sherds near the wall fill were two diagnostic of the Late
Mississippian period (Figure 5.26). Interestingly, the Ft. Walton incised bowl suggests
affiliations with material culture traditions to the south of Chickasawhatchee Creek, in
the Florida panhandle. By contrast, the Lamar Complicated Stamped sherd suggests
interactions with local groups slightly to the north. Frankie Snow (personal
communication, 2005) has observed the same stamping motif along Kinchafoonee Creek
– some 60 km from the site and only a couple of drainages over from Chickasawhatchee
Creek.
215
FIGURE 5.26, Fort Walton Incised (Lot 34) and Lamar Comp. Stamped (Lot 43) pottery, XU 3,
Str. 3, 9BX4
Ceramic Data
Ceramic data from XU 1, XU 3 and all other excavation units and shovel tests
across the site point to a very ephemeral occupation at the Red Bluff Earthlodge (Table
5.6). Only 93 sherds were recovered from 4.5 m3 of stratigraphic test pits and trenches.
This low frequency could be due to the placement of the units in areas that saw little use
or were regularly cleaned. However, only 72 sherds were recovered from 117 shovel
tests, as compared to more than twice then number of sherds from less than half the
number of tests at Magnolia Plantation Together, XU and shovel test data point to a very
216
ephemeral Woodland occupation around a possible low mound and a slightly more
intense Mississippian occupation centered around a single house or earthlodge.
UID Decorated
UID Incised
UID Stamped
2
1
0
0
0
0
0
0
0
0
0
0
0
0
3
1
0
0
0
2
6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
3
4
Totals
UID Complicated Stamped
0
1
0
3
0
0
0
6
0
2
2 17
0
0
0
2
0
1
0 15
0 10
0
2
0
1
0
0
2 60
0
8
0
2
0
6
0
2
0 47
2 125
Chattahoochee Brushed
Other
0
0
0
0
0
1
1
0
1
3
0
0
0
0
6
0
0
0
0
1
7
Eroded
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
UID Plain
Lamar Complicated Stamped
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 1
0 0
1 9
1 10
Point Washington Incised
Fort Walton Incised
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
Wakulla Check Stamped
1
2
4
5
Level 1 (0-10cm)
Level 2 (10-20cm)
Level (Floor Removal)
Level 3 (20-30cm)
Balk Removal (40-50cm)
Level 4 (30-40cm)
Level 4 (Floor sherd)
Center, Level 5 (40-50cm)
North, Level 5 (pp)
North, Level 5 (40-50cm)
South, Level 5 (40-50cm)
Center, Level 6 (50-60cm)
South, Level 6 (50-60cm)
South, Level 7 (60-70cm)
XU 3 Totals
All Levels
All Levels
All Levels
All Levels
Shovel Tests
CAS 4 Totals
Swift Creek Complicated Stamped
1
2
3
3
4
4
4
5
5
5
5
6
6
7
Ocmulgee/Fairchild Cord Marked
Level #
3
3
3
3
3
3
3
3
3
3
3
3
3
3
Ceramic type distribution from excavation units and shovel tests, 9BX4
Level Name
Unit #
TABLE 5.6,
0
3
0
4
0
1
0
6
0
2
0 20
0
1
0
2
0
2
0 18
0 10
0
2
0
1
0
0
0 72
0
9
0
2
0
8
0
2
1 72
1 165
Red Bluff Mississippian Architecture: House or Earthlodge
The prehistoric use of the structure documented in part by XU 3 is difficult to
determine without a more complete picture of the feature’s stratigraphic profile and
artifact distributions. The overall configuration of this structure matches that of either a
217
large Mississippian house or a small earthlodge or other ceremonial structure. Structures
with earthen berms and single set post supports are a common form of residential
architecture among many lower Appalachian Mississippian societies, and Hally
(1994:154) cites several excavated examples with features that broadly resemble those at
9BX4. Although few of these structures have survived with above-ground portions intact,
there are examples in central Georgia (Williams 1993; Williams and Evans 1993) and on
the Georgia coast (Keene 2004).
The Red Bluff Earthlodge’s configuration may resemble a domestic structure, but
its size suggests a more publicly oriented purpose. Hally (1994:155) suggests an average
domestic structure size of six or seven meters in diameter. Published house schematics
from the Apalachee area suggest a similar size profile (Scarry 1995). Polhemus’ (1987)
recorded a wide variety of structure forms in the course of his extensive analysis of
Dallas Phase (A.D. 1300 – 1650) architecture at the Toqua site. At Toqua, structures that
are morphologically similar to that of 9BX4 and similar in size are typically located in
public contexts, such as the mound summit. Structures closer to Hally’s average size for
domestic structures are often located in settings away mounds. The only large structures
(8-10 m in diameter) that are similar to the one at Red Bluff and yet argued to have been
the site of domestic activities were the circular structures Hatch (1995) recorded on
isolated farmsteads in the Oconee Valley.
The Red Bluff Earthlodge Site in Regional Context
Sites located adjacent to large wetlands can be the locus of ritual activities
(Kikuchi 1977; Pauketat 1994). Additionally, Nass and Yerkes (1995) and Emerson
218
(1997) both argue that in the absence of mound ceremonialism, small ritual loci may exist
in the form occupations that resemble domestic areas, but are spatially distinct and
exhibit some level of architectural variation.
These lines of evidence, along with the possibility that the Red Bluff Earthlodge
Site was previously the locus for Woodland period ceremonialism, provide preliminary
support for the hypothesis that this Late Mississippian period component served some
civic-ceremonial purpose for the local community. However, the extensive modern
disturbance around the site prevents verification of the site’s true spatial separation from
nearby settlements. In addition, more fieldwork would be necessary to understand the
site’s function using intra-site or architectural data. The site may be simply an extension
of local settlements and not a ceremonial center. In either case, occupations during both
the Woodland period and Late Mississippian period were small, short-lived, ephemeral,
and separated by a long period of disuse.
Chickasawhatchee Knoll
Chickasawhatchee Knoll may very well be the largest, most intensively occupied
site in the project area (Figure 5.22). As noted in Chapter 3, the site was investigated
twice in the early months of 1976. General surface collections and small 100% pickup
sample units were carried out at the site. Figure 5.27 shows site layout and the
distribution of sample points.
219
FIGURE 5.27, Site map of Chickasawhatchee Knoll, showing dog-leash collection unit locations
220
Temporally diagnostic sherds densities are low and exhibit no discernible spatial
patterns (Table 5.8). Diagnostic ceramics from the Middle Woodland through Late
Mississippian periods are evenly distributed across the site. There are, however, some
spatial patterns in terms of overall artifact frequency. The highest ceramic densities are at
the edges of the plateau atop the knoll. In contrast, very few sherds were recovered from
Unit 8, the collection area nearest to the site’s center. Units 2 – 4, on slight rise and also
located closer the site’s center, likewise had low artifact frequencies. This very general
pattern should be tested by more intensive more systematic testing, but there is
preliminary evidence for a formal site structure, perhaps including a central plaza, at
Chickasawhatchee Knoll.
The ceramic data are impressive in terms of the sheer frequency of sherds present
on the site compared to other sites (Table 5.7). The 2,448 sherds at Chickasawhatchee
Knoll are 38% of the total assemblage from the project area. As noted in Chapter 4, at
least half of these total results from surface conditions and the much greater intensity
with which the site was investigated. However, even if ceramic counts from this site were
25% of the total collected, there would still have been more sherds recovered from
Chickasawhatchee Knoll than from any site in the project area except for Hay Fever
Farm. The site was occupied or in use from the Paleoindian period forward (Snow n.d.),
and seems to have served as anchor for regional settlement patterns throughout the
Woodland and Mississippian periods.
Other
0 0
0 2
0 5
0 0 0 0 0 0 29
0 0 0 0 0 0 4
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 1
0 0 0 0 0 0 2
1 2 0 1 1 0 212 1 14 0 16 0 5 0 1 1 1 703
1 2 1 1 6 2 342 2 42 2 30 1 8 3 1 2 1 1713 52 12 108 1 21 94 2448
WGC Unit # 12
WGC Unit # 13
WGC Unit # 14
WGC Unit # 15
WGC Unit # 16
WGC Unit # Unknown
WGC Eastern Edge Recon 0 0 0 0 0 0 3
0 0 0 0 1 2 13
WGC Unit # 11
WGC General Surface
SCG Collection
Totals
0 1
0 0
0 0
0 0
0 2
1 0
0 3
0 3
0 4
0 0
0 1
0 0
0 0
1 0
0 0
0 1
0 0
0 2 1 0 0 0 69
0 0 1 0 0 0 61
0 0 0 0 0 0 31
0 0 0 0 0 0 100
0 0 0 0 0 0 8
0 0 0 0 0 0 2
0 0 0 0 0 0 20
1 0 0 0 0 0 114
0 0 1 0 0 0 108
0 0 0 0 0 0 134
0 1 0 0 0 0 82
0
0
0
1
2
2
1
0
0
8
0
1
0
0
2
10 0
1
1
0
3
0
4
6
4
0
0
5
5
4
3
0
0
0
8
12
17
19
0
6
144
11
69
78
80
47
108
9
3
25
139
162
10 111
0
0
1
0
1
0
6
3
20 198
4
0
0
1
1 13 17 998
0 0
0 1
0 0
0 0
0 0
0 0
0 0
0 2
0 1
0 0
0 3
0 0
0 1
114
24 93
0
0 0 0 0 2 0 15
0 0
0 0 0 0 0 0 6
0
0 0
0 0
0 0
WGC Unit # 10
0 0
0 0
0
1
3
25
0 0 1 0 0 0 30
0 2
0 0 0 0 0 0 50
0
2
2
WGC Unit # 8
59
0 0 0 0 0 0 3
0 1
6
2
5
7
WGC Unit # 7
0 6
0 0 0 0 0 0 65
0 0 0 0 0 0 52
0 0 0 0 1 0 63
0 0
0 0 0 0 1 0 4
1 0
0 2
0 2
0
WGC Unit # 6
0 4
0 0
0 2
0
0 0 0 0 0 0 0
Provenience
WGC Unit # 5
Wakulla Check
Swift Creek Comp.
Dunlap Fabric Mkd.
Keith Inc.
Carrabelle Punc.
Deptford Simple
Deptford Linear
0 0 0 0 0 0 8
Lake Jackson Deco.
Cool Branch Inc.
WGC Unit # 3
Fort Walton Inc.
Lake Jackson Inc.
0 0 0 0 1 0 13
UID Plain
Leon Jeff. Comp.
Leon Check
Alachua Cob Mkd.
Point Washington
Lamar Comp.
Lamar Bold Inc.
WGC Unit # 2
1
UID Comp. Stamped
0 0 0 0 0 0 45
UID Decorated
0 0
UID Incised
UID Fiber Tempered
0 0 0 0 0 0 5
UID Stamped
0 1
Totals
WGC Unit # 1
221
TABLE 5.7,Ceramic type distributions, CAS 89 surface collections
222
Hay Fever Farm
Chapters 3 and 4 presented a detailed consideration of the procedures undertaken
to document Hay Fever Farm as well as most of the results. Along with
Chickasawhatchee Knoll, the site is one of the two most durable sites in the project area.
Excavation data and landowner collections reveal occupations from the Early Archaic
period forward. At 15 hectares the site is also one of the largest in the project area.
However, as noted in the figures and discussions from Chapter 4, some of the site’s size
may be attributed to artifact spread resulting from historic period disturbance and erosion.
Cool Branch Incised
Fort Walton Incised
Ingram Plain
Lake Jackson Plain
UID Plain
Eroded
Other
UID Complicated Stamped
UID Decorated
0
0
0
3
0
0
0
0
0
3
0
3
0
0
0
1
0
0
0
0
0
1
0
1
0
1
1
0
0
0
0
0
0
2
0
2
0
1
0
0
0
0
0
0
0
1
2
3
1
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
0
1
0
1
61
128
190
142
38
17
3
0
0
579
162
741
0
0
0
0
0
0
0
0
0
0
1
1
0
1
2
0
1
1
3
0
0
8
0
8
0
0
0
0
0
0
0
0
0
0
1
1
2
0
5
0
0
0
0
0
0
7
0
7
Unit #
Totals
Wakulla Check Stamped
0
0
5
2
0
0
0
0
0
7
0
7
UID Stamped
Ocmulgee/Fairchild Cord Marked
0
2
0
0
0
0
0
0
0
2
1
3
UID Incised
Carrabelle Punctated
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
UID Fiber Tempered
Carabelle Incised
1 Level 1
2 Level 2
3 Level 3
4 Level 4
5 Level 5
6 Level 6
7 Level 7
8 Level 8
9 Level 9
XU 1 Totals
Shovel Tests
CAS 245 Totals
Level #
Deptford Simple Stamped
Ceramic type distribution from excavation units and shovel tests, CAS 245
Level Name
TABLE 5.8,
0 2
1 0
0 7
0 2
0 0
0 0
1 0
0 0
0 0
2 11
0 1
2 12
0
2
0
1
0
0
0
0
0
3
0
3
67
136
210
151
39
18
7
0
0
628
169
797
223
Table 5.8 presents data from the shovel tests and a two-by-two meter excavation
unit. In terms of overall frequencies, data are most comparable to those of
Chickasawhatchee Knoll, with some important differences. Comparatively, there are
fewer Woodland period diagnostics at Hay Fever Farm, perhaps indicating that the
majority of the settlement activity at the site took place during the Mississippian period.
A lack of regional settlement around the site precludes any conclusions about Hay Fever
Farm’s role in the regional settlement system, but future survey in the surrounding
uplands may provide evidence of a large community centered on this site, as is the case at
Chickasawhatchee Knoll.
Settlement Zones and Cycles of Growth and Contraction
Settlement patterns in the Chickasawhatchee watershed had alternating cycles of
growth and contraction. These cycles did not affect the region uniformly. Instead, the
area can be divided into four distinct zones of settlement (Figures 5.28a and 5.28b), each
of which displayed differential responses to changing circumstances in the watershed. In
general, Zones 1 and 2 were areas of settlement continuity and local shift, Zones 3 and 4
exhibited cyclic subregional abandonment.
Zone 1 is essentially the Hay Fever Farm site. There is strong evidence for
occupational continuity here from the Middle Woodland period forward, and perhaps
earlier as well. However, a lack of survey around this site raises questions as to the
possibility of occupational continuity in the northern portion of the project area as a
whole.
224
FIGURE 5.28a, Settlement zones during the Woodland periods
Zone 2 is the area within five kilometers of Chickasawhatchee Knoll. Because
this area was surveyed by Snow and colleagues when the area was first logged, survey
coverage in this area is as good as, if not better than, high intensity survey coverage
throughout the project area. This area also represents strong evidence for occupational
continuity. Chickasawhatchee Knoll exhibits artifact densities unmatched by other sites.
This difference may be the partial result of long-term site disturbance that did not affect
Chickasawhatchee Knoll when it was first collected. However, lower sherd densities at
neighboring sites collected at the same time combined with the higher artifact frequencies
from the extensively disturbed Hay Fever Farm suggest the importance of this site from
the Middle Woodland period forward into the 16th century. Away from
225
Chickasawhatchee Knoll, Zone 2 is characterized by small sites that, when abandoned,
are replaced by neighbors that are generally one or two kilometers away.
FIGURE 5.28b, Settlement zones during the Mississippian periods
Zone 3 is the area of periodic abandonment around Windmill Plantation. Survey
coverage in this area is also spotty, but the General Woodland period and non-diagnostic
ceramic sites in the area are most likely associated with this mound center. Overall,
ceramic data support a pattern of abandonment and reoccupation at Windmill Plantation.
As shown in Chapter 6, these cycles may coincide with the rise and decline of Kolomoki.
Zone 4 is the area of alternating occupation around Magnolia Plantation. It is
difficult to say how long the mound at Long Walk was occupied, but stratigraphic
226
evidence clearly demonstrates that the Magnolia Plantation (9DU1) site was occupied for
no more than four successive mound construction stages (75 – 100 years). As noted
above, this pattern is consistent with other paired town situations in which closely spaced
sites are occupied sequentially (e.g. Blitz and Lorenz 2006; Williams and Shapiro 1990).
These spatially separated areas of abandonment and continuity are unusual in that
the strongest evidence for settlement continuity and density are in the heart of the
Chickasawhatchee Swamp, in Zone 2. During both the Woodland and Mississippian
periods, civic-ceremonial sites are isolated from what appears to be the most intensely
occupied parts of the settlement system. Spatial separations around Windmill Plantation
and Magnolia Plantation may the result of spotty survey coverage, so the pattern of
separation may be somewhat exaggerated. However, the lack of settlement continuity
outside from the cluster around Chickasawhatchee Knoll is still intriguing and does
suggest deliberate attempts to keep mound ceremonialism at a distance.
Fully understanding the source of this distancing behavior requires not only an
understanding of the macroregional context of the Chickasawhatchee Swamp, but also an
explanation as to why the swamp would prove the more attractive area for settlement.
The former question is taken up in Chapter 6 and the latter in the section below.
Landscape Variation and Regional Settlement
Broad-scale models of Mississippian subsistence are not likely to be well adapted
to low-lying environments such as the Chickasawhatchee Swamp. Models of eastern
Woodlands land use model typically center on the creation of open spaces through
anthropogenic fire in regularly flooded areas (Smith 1989) and the control of small,
227
spatially restrictred wetlands (Pauketat 1994; Schroeder 1997). Data from areas outside
so-called “core” regions in the Eastern Woodlands suggest that these models are
incomplete.
Wagner (2003) notes a high correlation between archaeological site location and
certain fire resistant landscape mosaics, including oak/hickory/pine forests and longleaf
pine forests. There is a general consensus concerning the importance of oak and hickory
masts as part of subsistence strategies in the southeast (cf. Gremillion 2003; Scarry
2003). Wagner (2003) in particular emphasizes the abundance of understory plant and
game resources in open pine forests.
In addition to these observations, Pluckhahn and McKivergan (2002:157) note
correlations between wetland abundance and patterns of high settlement dispersal,
suggesting that in regions where wetlands are ubiquitous, such resources cannot be easily
used as leverages of political and economic power, as happens elsewhere (cf. Pauketat
1994: Schroeder 1997). Given these observations, it seems likely that a region
characterized by ecological heterogeneity, fire tolerant hardwood and coniferous species,
and ubiquitous wetlands might exhibit Woodland and Mississippian settlement patterns
that differ from the general model.
Using the same methods described above for testing bias in low-intensity survey,
I focus on the sites located on the present-day Chickasawhatchee Wildlife Management
area (Figure 5.29) in order to understand relationships between local ecological variation
and settlement. These comparisons will primarily shed light on the settlements in Zone 2,
but the insights provided below also lead to questions regarding why the strategies in
228
Zone 2 were not exported to other parts of the project area, or at least, why similar
patterns of occupational continuity did not exist.
FIGURE 5.29, Site distributions and topographic context on the Chickasawhatchee WMA
Visual inspection of soil data in Figure 5.30 suggests preferences for settling on
clays and sandy clay loams. However, chi square results do not fully bear out this
conclusion (Table 5.9). There is no statistical preference for sandy clay loams and the
229
tendency towards clays is weak at best. Instead, what we see is an overwhelming
disinclination towards sands. From a behavioral point of view, this could relate to the
slightly higher elevation gradients between bottomlands and uplands in areas where clays
and loams dominate – or perhaps to the need for clay in home construction, ceramic
manufacture, or similar activities. However, the under-representation of sands raise the
possibility of sampling bias. Bias seems especially likely given that, in south-central
Georgia, Kirkland (1994:92) notes heavy settlement concentrations around well-drained
upland sand ridges.
FIGURE 5.30, Site and soil distributions on the Chickasawhatchee WMA
230
TABLE 5.9,
Observed and expected distributions of soil classes in low intensity survey
transects, as well as chi square test results
Upland Soils
Clays
Sands
Sandy Clay Loams
Totals
TRANSECT DATA
Transect
Total
Site
Total
Area (ha) Area (ha) Area % Area % Expected Observed X^2
145.65 1593.42
0.48
0.42
256.35
291.31 4.19
52.30
827.29
0.17
0.22
133.10
104.61 7.76
104.96 1344.91
0.35
0.36
216.37
209.91 0.20
302.91 3765.62
1.00
1.00
605.83
605.83 12.15
Transect
Total
Site
Total
Area (ha) Area (ha) Area % Area % Expected Observed X^2
Submerged Soils
Bottomland
52.79 2937.09
0.60
0.71
124.10
105.58 3.25
Poorly Drained Transitions
34.34 1111.01
0.38
0.27
46.94
66.95 5.98
Ponded
0.44
81.41
0.00
0.02
3.44
0.87 7.54
Totals
87.57 4145.01
1.00
1.00
175.14
175.14 17.43
SITE DATA
Site Area
Total
Site
Total
(ha)
Area (ha) Area % Area % Expected Observed X^2
Upland Soils
Clays
43.04 1593.42
0.58
0.42
31.60
43.04 3.04
Sands
3.98
827.29
0.05
0.22
16.41
3.98 38.79
Sandy Clay Loams
27.66 1344.91
0.37
0.36
26.67
27.66 0.04
Totals
74.67 3765.62
1.00
1.00
74.67
74.67 41.86
Site Area
Total
Site
Total
(ha)
Area (ha) Area % Area % Expected Observed X^2
Submerged Soils
Bottomland
3.33 2937.09
0.33
0.72
7.31
3.33 4.75
Poorly Drained Transitions
6.78 1111.01
0.50
0.27
2.76
5.05 1.03
Ponded
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Totals
10.11 4063.60
1.00
1.00
10.11
10.11 7.44
critical value = 5.99 at a 95% C.I.
Visual inspection of Figure 5.31 suggests preferences for long-leaf pine and fire
intolerant hardwoods. In contrast to the equivocal soil results, chi square results for
landscape patches are robust and match subjective expectations. Individual chi square
scores exhibit strong preferences toward long-leaf pine ecosystems and lesser use of fire
tolerant mixed pine and hardwood patches (Table 5.10). In both of these cases, individual
chi square scores exceed critical values by a comfortable margin. Preferences for fire
231
intolerant hardwood ecosystems are weaker, but are nevertheless worthy of mention due
to the low proportion of total area that this ecosystem type represents.
FIGURE 5.31, Sites and vegetation zone reconstruction on the Chickasawhatchee WMA
Survey data from narrow transects like these should be viewed cautiously, but if
we provisionally assume that the data and statistical results are valid, a few trends
emerge. Upland fire intolerant hardwoods and long-leaf pine ecosystems are located
closest to the confluence of Chickasawhatchee and Kiokee Creeks. In this area, maps
show boundaries between uplands and open water wetlands at their sharpest, suggesting
232
that the distribution of waterine resources would be at its most abundant. This trend,
combined with the swamp’s dispersed settlement pattern, and the lack of civicceremonial sites, suggests that ubiquitous wetlands do encourage settlement dispersal and
close off certain avenues for leveraging political and economic power.
TABLE 5.10,
Observed and expected distributions of reconstructed vegetation classes in low
intensity survey transects and among sites, as well as chi square test results
TRANSECT DATA
Transect Total Area Site
Total
Upland Vegetation
Area (ha)
(ha)
Area % Area % Expected Observed
Longleaf Pine
74.10
841.35
0.26
0.26
148.44
148.21
Fire Intolerant Upland
31.49
399.18
0.11
0.12
70.43
62.98
Upland Pine/Hardwood Mix
184.75 2050.76
0.64
0.62
361.81
369.49
Totals
290.34 3291.29
1.00
1.00
580.67
580.67
X^2
0.00
0.88
0.16
1.04
Transect Total Area Site
Total
Area (ha)
(ha)
Area % Area % Expected Observed X^2
82.93 2726.92
0.81
0.58
119.58
165.85 12.91
2.87
164.01
0.03
0.03
7.19
5.75
0.36
17.10 1802.05
0.17
0.38
79.03
34.20 58.75
102.90 4692.99
1.00
1.00
205.80
205.80 72.02
SITE DATA
Site Area Total Area Site
Total
Upland Vegetation
(ha)
(ha)
Area % Area % Expected Observed X^2
Longleaf Pine
33.90
841.35
0.47
0.26
18.36
33.90
7.13
Fire Intolerant Upland
15.78
399.18
0.22
0.12
8.71
15.78
3.17
Upland Pine/Hardwood Mix
22.13 2050.76
0.31
0.62
44.75
22.13 23.12
Totals
71.81 3291.29
1.00
1.00
71.81
71.81 33.42
Bottomland Vegetation
Hardwood Bottomland
Herbacious/Open Ponds
Tupelo/Cypress Swamp
Totals
Site Area Total Area Site
Total
Bottomland Vegetation
(ha)
(ha)
Area % Area % Expected Observed
Hardwood Bottomland
9.56 2726.92
0.85
0.58
6.53
9.56
Herbacious/Open Ponds
0.36
164.01
0.03
0.03
0.39
0.36
Tupelo/Cypress Swamp
1.32 1802.05
0.12
0.38
4.31
1.32
Totals
11.24 4692.99
1.00
1.00
11.24
11.24
critical value = 5.99 at a 95% C.I.
X^2
0.96
0.00
6.79
7.76
With regard to landscape measures of regional ecological variation, upland fire
intolerant hardwood and longleaf pine ecosystems tend correlate with red clays and red
sandy clay loams, respectively. Given the correlations between soils, ecosystem types,
233
and site location, we have evidence for the hypothesis that these landscape patches are
most likely the direct or indirect result of anthropogenic fire management. Wagner (2003)
notes that high frequency, high temperature burning is conducive the spread of long-leaf
pine habitats. In the case of upland fire intolerant hardwoods, human settlement activity
would have created cleared spaces for housing. Once settlements were established, fire
suppression would have allowed bottomland hardwoods to invade upland habitats
normally occupied by long-leaf pine. In modern southwestern Georgia, the correlation
between human settlements and the upland invasion of bottomland oak species is
apparent to even casual observers.
There are a number of wild resources that would have been available among
upland hardwoods and long-leaf pine ecosystems. However, the likelihood of sixteenth
century abandonment begs the question as to whether fire intolerant hardwood and longleaf upland habitats resulted from the deliberate, if casual, use of fire by Woodland and
Mississippian peoples, or whether these ecosystem types resulted from post-occupation
ecological succession in areas previously cleared for more typical eastern Woodlands
agriculture. Either way, the tendency for sites to be located next to transition zones
between uplands and lowlands suggests a mixed strategy with a heavy emphasis on
wetland resources.
Project-Wide Ceramic Variation and Density
Throughout this chapter ceramic variation has been discussed in generalized
terms. At the regional and intra-site scale, ceramics provide temporal control and are the
building blocks of settlement histories. Taking into account the methods and limits
234
discussed in Chapters 3 and 4, ceramic frequencies serve as proxy measures of
occupational intensity.
One of the major goals of this dissertation is to demonstrate significant
differences in the political economy of the region’s Woodland period and Mississippian
period polities using multiple lines of evidence. I argue that the low ceramic frequencies
observed throughout the project area are not the result of site formation processes or
methodological bias, but instead reflect these socio-political conditions. A complete
assessment of this argument requires reference to broader patterns in neighboring regions.
This section describes ceramic variation in terms of decorative design, vessel
form, and vessel standardization at a regional scale. These descriptions form the basis for
the preliminary hypothesis, addressed in Chapter 6, that aggregate ceramic data from the
Chickasawhatchee Creek watershed are consistent with ceramic production and exchange
patterns characteristic of a “backcountry” region (cf. Kowalewski 2003).
A full discussion of Kowalewski’s model and its limited applications to
Woodland and Mississippian period societies is provided in Chapter 3. To review, the
critical criteria are reduced ceramic production overall, a more narrow range of vessel
forms, greater frequencies of undecorated vessels, less variation in the type of decorative
designs present, and less skillful execution of decorative motifs and overall vessel
formation. The figures and tables presented below address these issues while describing
the ceramic variation in the project area.
One remarkable result is that ceramic temper among Woodland period and
Mississippian period pottery is almost exclusively a continuous blend of sand and ground
235
quartz or grit. Sherds with temper consisting almost exclusively of one or the other
material cannot be reliably identified without petrographic analysis. There are exceptions,
including 40 sherds tempered with pure sand, similar to sand-tempered sherds in the
Florida Panhandle. There are also 31 examples of hard-fired Woodland period sherds
with glittering micaceous paste and four sherds tempered with grog, or ground-up
ceramics. Together, these sherds represent just over one percent of the total assemblage. I
collected no instances of the shell-tempered pottery common at Early Mississippian
period sites in the adjacent Chattahoochee Valley (Blitz and Lorenz 2006).
Table 5.11 is a complete catalog of sherd counts from the project area according
to period, decorative design, and vessel form. The most salient feature of this table is the
overwhelming number of plain sherds. Wakulla Check Stamped pottery is the most
common decorated type, but as noted in Chapter 3, this decorative technique is so
ubiquitous, so temporally un-diagnostic, and found in so many socio-economic contexts
that it is virtually a plain ware (White 1985).
Excluding check stamped pottery, only 313 out of 6,556 sherds, or five percent,
can be identified to a specific decorative design. Even if all the unidentified decorated
sherds in the project area were added to this figure (minus the UID Stamped sherds, most
of which are also probably check stamped) decorated pottery still only accounts for 13%
of the total assemblage. Only two other decorative designs, Lake Jackson Decorated and
Fort Walton Incised are represented by more than 50 sherds.
236
TABLE 5.11,
Period
Late Archaic
Sherd counts by decorative type, temporal designation, and vessel form
Type
Bowl Jar Other Pipe Unknown Totals
UID Fiber Tempered
0
0
0
0
1
1
Ocmulgee/Fairchild Cord Marked
0
1
0
0
11
12
General
Swift Creek Comp. Stamped
0
0
0
0
23
23
Woodland
Wakulla Check Stamped
2
3
0
0
463
468
Alligator Bayou Rocker Stamped
0
1
0
0
5
6
Middle
Deptford Linear Check Stamped
0
0
0
0
16
16
Woodland
Deptford Simple Stamped
0
0
0
0
4
4
UID Simple Stamped
0
0
0
0
1
1
Carabelle Incised
0
0
0
0
3
3
Carrabelle Punctated
0
0
0
0
13
13
Late
Dunlap Fabric Marked
1
0
0
0
5
6
Woodland
Indian Pass Incised
0
0
0
0
2
2
Keith Incised
0
0
0
0
1
1
Weeden Island Incised
0
0
0
0
9
9
Lake Jackson Decorated
3 21
1
0
53
78
General
Lake Jackson Incised
0
1
0
0
2
3
Mississippian Lake Jackson Plain
0
7
0
0
0
7
Andrews Decorated
0
0
1
0
0
1
Columbia Incised
1
0
0
0
0
1
Middle
Cool Branch Incised
0
3
0
0
2
5
Mississippian Ingram Plain
11
0
0
0
1
12
Savannah Complicated Stamped
0
0
0
0
4
4
Fort Walton Incised
10
1
3
0
38
52
Lamar Bold Incised
1
0
0
0
1
2
Late
Lamar Complicated Stamped
0
2
0
0
20
22
Mississippian Lamar Plain
0
1
0
0
2
3
Point Washington Incised
1
3
0
0
12
16
Alachua Cob Marked
0
0
0
0
4
4
Protohistoric
Chattahoochee Brushed
0
0
0
0
1
1
Leon Check Stamped
1
0
0
0
1
2
Leon Jefferson Comp. Stamped
0
0
0
0
3
3
Eroded
0
1
0
0
84
85
Other
1
0
0
0
84
85
UID Complicated Stamped
0
0
0
0
114
114
Unknown
UID Decorated
0
1
0
0
214
215
Indian/Ceramic UID Incised
6
0
0
0
81
87
UID Plain
9 36
7
8
4967
5027
UID Punctated
0
0
0
0
10
10
UID Stamped
0
0
0
0
152
152
Totals
47 82
12
8
6407
6556
237
In terms of vessel form, a small sample size (n=149) limits inferences that can be
drawn from the available data, but a few observations are possible. First, jars outnumber
bowls by a ratio of nearly two to one. Non-standard forms are uncommon, represented by
only Fort Walton, plain, and UID incised plates and a single Andrews Decorated beaker
or bottle sherd. Pipes are also rare. Though pipes are common elsewhere in the South
Appalachian area, their under-representation here may be an artifact of surface collecting,
as pipe fragments are easily recognizable and more likely to be collected by those looking
for arrowheads.
In his study that associates changes in ceramic form to functional variation
through time, Braun (1983) proposes a temporal sequence in which potters abruptly begin
making thicker vessels and including lower densities of temper in order increase
resistance to thermal shock (at the expense resistance to mechanical shock); this is due to
changes in domesticated corn and cooking technology. Chilton (1998) demonstrates a
similar pattern, contrasting vessel assemblages between two historic period Iroquoian and
Algonquian potters. Chilton (1998:151-152) builds on Braun’s argument by linking
vessel form variation to sedentary and non-sedentary subsistence strategies. A sedentary
strategy dependent on cooked foods requires thinner vessel walls while thicker vessel
walls will be more resistant to the mechanical shocks entailed in a mobile lifestyle.
Archaeologists working in the South Appalachian area have long believed that
there are quantitative differences in vessel thickness between Woodland period and
Mississippian period ceramic vessels (Wauchope 1966; Williams, personal
communication), but this has not been systematically demonstrated. If such a difference
238
were observed, it might also imply a possible shift in subsistence strategies – probably
accompanied by a settlement shift as well.
Tables 5.12a and 5.12b present data on vessel thickness according to decorative
design, temporal placement, and vessel form. As Braun (1983) notes, differences in
vessel thickness or even temper density require an accounting of morphological variation
in vessel size and vessel form. Vessel size data are not available for this sample, but two
trends stand out among the temporal and formal data.
TABLE 5.12,
Vessel thickness measures according to A) decorative type and temporal
designation and B) vessel form
Form
Thickness(mm) n
Bottle
5.00
1
Carinated Bowl
7.79
1
Collared Jar
5.77
7
Jar
6.17 22
Outleaned Wall Bowl
7.48 15
Plate
2.53
3
Undifferentiated Bowl
5.50
7
Unknown
4.33 81
5.57 137
Form
Bowl
Jar
Other
Unknown
Thickness(mm) n
6.89 23
6.08 29
3.15
4
4.33 81
5.11 137
The range of variation in thickness across periods exhibits considerable overlap.
In addition, the extreme ends of the range of variation are represented by a specific class
of vessels. Very thin sherds include Weeden Island Incised and Fort Walton Incised
vessels, which are often fine effigy vessels in the former case and delicate plates in the
latter (Milanich et al. 1997; Lazarus and Hawkins 1976). The thickest vessels are the
Ingram Plain and Point Washington Incised bowls. Although Point Washington Incised
239
vessels vary greatly in size (Willey 1998[1949]), Ingram Plain bowls are well known for
their thick walls and somewhat “clunky” character (Schnell et al. 1981).
When these extreme cases of vessel form variation are omitted, the range of
variation among vessels is less than 2 mm across periods. Data do not support a temporal
shift in vessel thickness similar to those recorded by Braun (1983) and Chilton (1998).
Although this is not necessarily evidence for continuity in settlement strategies in and of
itself, it does lend support to other lines evidence in an argument for long-term continuity
in approaches to subsistence and settlement. Vessel thickness data also support the idea
of somewhat limited vessel forms.
TABLE 5.13,
Rim modification counts
Rim Treatment
Flat Plain
Folded
Folded and Decorated
Large Tics
Notched
Other
Rolled
Rounded Plain
Scalloped
Small Tics
% (n=222)
11.7
11.3
8.1
12.6
6.3
5.0
18.9
15.3
0.5
10.4
100.0
Table 5.13 presents data on vessel lip modifications for the entire project area.
Kowalewski (2003: 72) suggests that flattened, reinforced, rounded, and rolled rims are
shortcuts in vessel production aimed at cutting down the effort needed to complete a rim.
He makes a similar argument for folded rims, but many of the observed folded rims in
this project area are either have a very finished and squared off appearance (in the
Woodland period) or are decorated (in the Mississippian period). In this assemblage, 46.9
percent of the rim lips have some form of finish that fit a model of lesser production
240
value. Evidence for expedient ceramic production also comes from several cases of
idiosyncratic vessel forms. Figure 5.32 illustrates several of these unusual forms. Figure
5.32a illustrates a sherd with unusual amounts of coarse grit temper. Figures 5.32b, c, and
d illustrate sherds in which vessel rims and shoulders are irregularly smoothed. Figure
5.32e is a plain rim that looks as if the lip was deformed when the vessel was flipped over
to dry or be fired. Decorative anomalies even existed at mound sites. Figures 5.32b and
and 5.32c are from Windmill Plantation. Figure 5.32e illustrates an attempted pseudonode on a Lake Jackson Decorated Jar from Long Walk mound site that is crudely
executed and oriented to the horizontal, rather than vertical axis of the jar. Figure 5.32f,
likewise from Long Walk is another example of an irregularly smoothed vessel surface.
FIGURE 5.32a, large quantities of heavy grit temper
241
FIGURE 5.32b, Irregularly smoothed vessel rim
FIGURE 5.32c, Irregularly smoothed vessel rim
242
FIGURE 5.32d, Irregularly smoothed vessel rim
FIGURE 5.32f, Horizontal pseudo node on low fired potter
243
FIGURE 5.32g, Irregular vessel surface
Table 5.14 presents frequencies of decorative design exclusive from the
overwhelming numbers of plain and check stamped sherds. One contrast that appears is
between the overall distributions of decorative types in the Woodland and Mississippian
periods. During the Woodland period, Swift Creek Complicated Stamped, cord marked,
and Carrabelle Punctated sherds are most common, but none comprise more than six
percent of the total decorated assemblage.
In contrast, the Mississippian period is dominated by Lake Jackson wares and
Fort Walton Incised pottery. Apart from Ingram Plain, Point Washington Incised, and
Lamar Complicated Stamped sherds, no decorative type exceeds two percent of the total
decorated assemblage. Even accounting for the fact that the majority of some vessel
forms, such as Lake Jackson jars, are undecorated, these data support the idea that vessel
assemblages in the Chickasawhatchee Swamp were dominated by plain wares.
244
TABLE 5.14,
Decorated ceramic relative frequencies
Period (n =313)
Late Archaic
General Woodland
Middle Woodland
Late Woodland
General Mississippian
Middle Mississippian
Late Mississippian
Protohistoric
% Total
Type
UID Fiber Tempered
Ocmulgee/Fairchild Cord Marked
Swift Creek Complicated Stamped
Alligator Bayou Rocker Stamped
Deptford Linear Check Stamped
Deptford Simple Stamped
UID Simple Stamped
Carabelle Incised
Carrabelle Punctated
Dunlap Fabric Marked
Indian Pass Incised
Keith Incised
Weeden Island Incised
Lake Jackson Decorated
Lake Jackson Incised
Lake Jackson Plain
Andrews Decorated
Columbia Incised
Cool Branch Incised
Ingram Plain
Savannah Complicated Stamped
Fort Walton Incised
Lamar Bold Incised
Lamar Complicated Stamped
Lamar Plain
Point Washington Incised
Alachua Cob Marked
Chattahoochee Brushed
Leon Check Stamped
Leon Jefferson Complicated Stamped
Total %
0.32
3.83
7.35
1.92
5.11
1.28
0.32
0.96
4.15
1.92
0.64
0.32
2.88
24.92
0.96
2.24
0.32
0.32
1.60
3.83
1.28
16.61
0.64
7.03
0.96
5.11
1.28
0.32
0.64
0.96
100.00
Finally, it is worth re-emphasizing that the overall ceramic collection from the
project area most closely approximates the assemblage Blitz and Lorenz (2006) outline
for the Chattahoochee Valley, and differs from what Hally (1986) defines for piedmont
and montane areas. Blitz and Lorenz (2006) argue that the vicinity Chattahoochee and
Apalachicola Valleys represent ceramic production and exchange networks that have
relatively little contact for most of the Mississippian period. In the Chickasawhatchee
245
Swamp, we have evidence of ceramic forms and decorative techniques from both areas.
Nevertheless, vessel forms that differ from the Chattahoochee Valley assemblage are
rare. This general lack of diversity supports arguments for more restricted vessel forms in
the project area.
Data from the project area support the idea that ceramic production in the
Chickasawhatchee Creek drainage was that of a marginal system not intensively involved
in long-distance exchange. Plain ware vessels dominate the assemblage. In terms of
vessel form, the significantly greater frequency of jars and the paucity of specialized
forms suggest that cooking was the overarching use for pottery. There were a few serving
bowls, bottles, and plates for public rituals and feasting, but these vessel forms are
exceedingly rare.
The range of variation in vessel thickness, combined with the fact that outliers in
terms of thickness are connected to a few rare types, likewise suggests a limited number
of forms overall and restricted use of ceramics. The lack of change through time in vessel
thickness supports other lines of evidence for settlement pattern continuity. Finally,
irregularly executed vessels and vessel decorative designs supports the hypothesis that
trends in vessel construction and decoration were being initiated elsewhere and copied at
the household scale.
Uncommonly friable sherds with large temper and inclusions and irregularly
smoothed and finished vessel shoulders and rims demonstrate tendencies toward less
standardization and less careful vessel execution. These trends may reflect less skill
potters. Reduced decorative design variation also suggests less intense participation in
246
broader spheres of interaction were common. Low ceramic densities across the project
area also indicate that people simply did not make as many pots as in other areas – a
possible contributing factor in reduced standardization and decorative variation.
Together, these trends raise the possibility of either differential subsistence practices, or
product substitution, or both.
Summary and Conclusions
This chapter presented results of regional and intrasite surveys in the
Chickasawhatchee Swamp. Settlements in the heart of the swamp represent a
continuously occupied community in which individual settlements cluster around
boundaries between wetland and upland ecosystems. Beyond the swamp, subregions
exhibit cycles of intensive occupation and abandonment. Intrasite data from civicceremonial centers indicate either a lack of intensive occupation or much shorter term
occupations than the two most intensively and continuously occupied sites in the project
area – both of which lack above ground architectural remains. Ceramic variation suggests
a production regime in which household producers of utilitarian wares selectively mimic
external trends in the production of ritual and serving wares.
Spatial Settlement variation contrasts with temporal continuity. Through the
occupation sequence, the Chickasawhatchee Swamp exhibits a bi-modal settlement
system with a few settlements located near major wetland resources and a broader
distribution of small, short-term occupations. In terms of occupational intensity and
ceramic production, the Chickasawhatchee Swamp seems marginal to its neighbors.
These hypotheses are tested with macroregional data in the next chapter.
247
CHAPTER 6: THE CHICKASAWHATCHEE SWAMP
IN MACROREGIONAL CONTEXT
In the preceding chapters, I have used multi-scalar analysis to make the case for
long term variance between the settlement strategies of Chickasawhatchee Swamp
inhabitants and those of its neighbors. By tacking between regional and intra-site scales, I
demonstrated a spatial disconnect between civic-ceremonial activity and areas of
continuous occupation, differential patterns of subregional abandonment, and, possibly, a
less intensive ceramic production economy than in other areas. However, in the absence
of a quantitative view of macroregional variation, these suggestions lack support.
Additionally, as noted by Meentenmeyer and Box (1987:21), patterns occurring over long
time periods and requiring more overall energy are often visible at much broader scales.
This chapter places the Chickasawhatchee Swamp in macroregional, long-term
context using two primary lines evidence. First, corrected data from the Georgia
Archaeological Site File to present a period-by-period narrative of macroregional
settlement pattern change. This narrative demonstrates concordant shifts between
macroregional settlement strategies and those observed locally in the Chickasawhatchee
Swamp – lending support to arguments favoring the long-term interconnectedness of
eastern Woodlands societies.
In addition, I provide insights into the relationship between physiographic
variation and settlement structure during a single time period. As physiographic variation
is a major determinant of landscape structure, and by extension, constrains land use
strategies, physiography-related settlement pattern variation supports hypotheses linking
248
anthropogenic landscapes to differential participation in macroregional interaction
spheres.
Through time, settlement pattern shifts begin to reflect differential participation in
the Mississippian political economy, as evidenced through the expansion and contraction
of mound building below the fall line. The nature of this expansion and contraction is
explored through macroregional comparison of intra-site variation at mound sites which
shows that sites established below the fall line hills are generally less intensely occupied
than to the north. Macroregional comparisons also show long-term differences in mound
building intensity and settlement density that are most pronounced when comparing the
interior coastal plain with the fall line hills and the piedmont. The coastal zone differs yet
again from all these areas. Through time, there is continuity in terms of settlement
structure within individual physiographic zones.
Methods
Data for this chapter come from the Georgia Archaeological Site File, as well as
published and unpublished data concerning reports of excavations at mound sites along
the fall line (Pluckhahn 1996 and 2002; Williams 1996, 1999a, and 1999b). Using site
file data, I present settlement pattern maps showing the locations of all mound and nonmound centers that can be dated to specific time periods, as well as a table detailing
temporal shifts in the frequency of mound and non-mound centers. To compare
macroregional trends in intra-site patterning I present cartographic and tabular data
detailing ceramic frequencies and typological diversity from shovel tests at key mound
and non-mound sites.
249
Site file data are limited by the relative completeness of the data set and the
widely varying methods involved archaeological investigations in CRM and academic
settings (Williams 2000). For this analysis, I focus on the relative locations and
frequencies of mound and non-mound centers from the Middle Woodland through Late
Mississippian periods. Despite the importance of variables such as mound form, volume,
quantity and arrangement, I overlook these variables and instead focus narrowly on the
potential relationship between mound ceremonialism and inter-regional exchange, as
discussed in Chapter 2.
Most of the mound centers in Georgia have been recorded – though some were
certainly destroyed and will remain unknown (Elliott 2004; Smith and Harris 2001). With
this relatively unbiased sample, we can understand some macroregional shifts in the
political economies of eastern Woodlands societies that emerge as temporal changes in
mound center density, arrangement, and distribution relative to natural features. By
contrast, the spatial patterning of non-mound sites is closely tied to survey coverage
(Schroeder 1997; Williams 2004). Nevertheless, non-mound sites provide contextual data
and suggest hypotheses to be investigated later.
While mound center and site distribution maps reveal macroregional trends,
relationships between these trends and changes in local political economies require more
refined data. Ideally, we should compare multiple regions in terms of long-term shifts in
component frequency, component size, settlement pattern hierarchies, relationships
between settlement locations and the environment, and intra-site patterns. These
250
comparisons require massive amounts of data and conceptual tools that account for
differences in field methods and analytical frameworks (Smith 2002).
For this analysis, I focus on variables comparing intra-site patterns in the
Chickasawhatchee Drainage with those of selected sites in the nearby fall line hills and
coastal zone. I compare relative measures of occupation intensity and duration. For this
limited study, I refer to ten sites that have been subjected to systematic shovel testing. As
noted in Chapter 3, there are several ways to measure occupational intensity – only a few
of which depend exclusively on ceramic data. A smaller subset of such measures are
useful in the absence of estimates concerning ceramic and other artifact accumulation
rates and data related to architectural and storage features. Specifically, I focus on
typological diversity based on the presence and absence of decorative types, and ceramic
density using sherds per shovel test as a unit of measure.
Macroregional Settlement Pattern Change
Table 6.1 presents data concerning changes in the relative frequencies of mound
and non-mound centers through time. Note that these counts are of mound sites occupied
during the specific periods. Some sites may be counted twice and other sites are counted
for given periods in spite of the fact the site may predate or post-date the mound’s actual
construction. As this analysis does not account for specific social contexts of mound use,
the primary impact of this bias is an overstatement of the uniformity in mound
construction from period to period.
251
TABLE 6.1,
Mound and non-mound site frequencies in Georgia by time period
Period
Middle Woodland
Late Woodland
General Woodland
Woodland Totals
Early Mississippian
Middle Mississippian
Late Mississippian
General Mississippian
Mississippian Total
Unknown
Totals
Mound Non-Mound Sites/Mound
74
2346
31.7
71
1499
21.1
9
3258 n/a
154
7103
46.1
30
574
19.1
39
614
15.7
75
2574
34.3
8
2425 n/a
152
6187
40.7
83 n/a
n/a
389
13290
34.2
For the sites of earlier periods on which mounds were later constructed, this bias
is significant. For later sites, with mounds built during an earlier period, I argue that
mound sites are re-occupied deliberately. In these cases, mounds are re-integrated into the
settlement layout. Even in the absence of new construction activity and minimal
associated material deposits, the mounds are in use – albeit in a different context (see
Williams and Shapiro 1990; Williams 1990; Smith 1979).
Mound counts from the Middle and Late Woodland periods are stable, but nonmound site counts are not and suggest a Late Woodland period decline in site frequency.
These data contradict published hypotheses of Late Woodland period expansion
(Milanich et al. 1997; Rudolph 1991; Chapter 5, this volume) – likely reflecting the
difficulties associated with recognizing Late Woodland period sites. Because these sites
are characterized by pottery decorated with cord marking, check stamping, or simple
stamping (Elliott and Wynn 1991; Stephenson 1990; White 1985), many sites that can
only be given a General Woodland period designation may date to the Late Woodland
period. Such a hypothesis is supported by significantly larger numbers of General
Woodland period sites than General Mississippian period sites.
252
The Early and Middle Mississippian periods seem to suggest dramatic reductions
in settlement, to the point of suggesting a dramatic population decline (e.g. Kowalewski
et al. 1989:499-501). However, considering the overwhelming evidence for contemporary
occupations of so-called “Woodland” and “Mississippian” settlements (Blitz and Lorenz
2002 and 2006; Elliott and Wynn 1991; Pluckhahn 1997; Schnell and Wright 1993;
Williams 1996), there is little doubt that this decline reflects the theoretical baggage of
neo-evolutionary models in which the “good grey cultures” of the Late Woodland period
(Williams 1964) are replaced by Mississippian complexity (e.g. Kowalewski 1996:29).
These apparent patterns of stability should not be exaggerated, however. Problems
with chronological and cultural classification notwithstanding, intrasite data from a
variety of sources suggest major differences in intra-site settlement strategies between the
Woodland and Mississippian periods. Available data suggests that many Mississippian
period villages were larger and more populous than their Woodland period counterparts
(compare, for example, Milanich et al. 1997:200-208 with Hatch et al. 1997). In addition,
the differential length of the Late Woodland period in some areas confounds efforts at
broad-scale comparison. More research on the Woodland period will help clarify
relationships between macroregional trends and local settlement variation.
Figures 6.1 and 6.2 are settlement pattern maps of the Middle Woodland and Late
Woodland periods. Overall the maps show remarkable continuity through time. In both
periods, the coastal zone and interior coastal plain contain slightly more mound sites than
the less expansive north. However, from the Middle Woodland period to the Late
Woodland period, the number of mound sites in the piedmont and mountain zones drops
253
from 33 to 26. The expansion of mound sites in the coastal plain is most dramatic near
the mouths of the Savannah and Ogeechee Rivers and along between confluence of the
Flint and Chattahoochee Rivers.
FIGURE 6.1,
Middle Woodland period settlement patterns
254
FIGURE 6.2,
Late Woodland period settlement patterns
These data support hypotheses positing Late Woodland period settlement
expansion into upland areas (Percy and Brose 1975 [cited in White 1981]). At this coarse
255
scale, settlement pattern data shed little light on relationships between upland expansion
and the Late Woodland reorganization of exchange networks – again highlighting the
need for more research into Woodland period settlement patterns.
FIGURE 6.3,
Early Mississippian period settlement patterns
256
Early Mississippian settlement patterns (Figure 6.3) reflect the expansion of a new
model for political economy based on platform-mounds, consolidated villages, longdistance exchange of preciosities, new ceramic vessel forms, paste recipes, and
decorative designs, and, in most areas, more intensive cultivation of corn (Blitz and
Lorenz 2002 and 2006; Williams 1994; King 2003). The Macon Plateau site, located in
the center of the state and far from most other mound centers, may illustrate how this
expansion worked in the context of existing Late Woodland period populations.
Following Fairbanks (2003), Williams (1994 and 2003) posits that at Macon
Plateau, this new economy arrived with recent migrants to the area. Similar arguments
have been posited by Blitz and Lorenz (2002), who argue that fortified Middle
Mississippian period sites such as Cool Branch indicate the marshal intentions of recent
immigrants. By contrast, Pluckhahn (1997:49-50) documents interaction between people
at the Macon Plateau and nearby Tarver sites. Shell-tempered pottery from Macon
Plateau at a site dominated by Vinings Simple Stamped pottery suggests the possibility of
close contact and perhaps intermarriage (Pluckhahn 1997:51).
Comparative cases yield some insights the question of hostility and the
establishment of Macon Plateau. Wiessner (2002:342) describes large scale migrations
among the Enga that are linked to coordinated military action. In these narratives, groups
from highland settlements invaded nearby valleys and forced existing inhabitants out of
the conquered territories. In the southwestern United States, mobility is often correlated
not with warfare, but with existing social ties, as people tend to move to places where
they have existing connections (Varien 1999:60). However, even in situations
257
characterized by amicable, or at least, non-hostile relationships, migrants moving into
areas with established settlements have no guarantees of obtaining the best land (Lyons et
al. 2005:11).
Williams’ (1996:136-137) primary argument in favor of immigration rests on
dissimilarities between the pottery at Macon Plateau and those from nearby areas and
later periods. Williams’ observations and those of Pluckhahn may be balanced through a
hypothesis that acknowledges the likelihood of pre-existing long-distance, inter-regional
contact. In this circumstance, any number of non-hostile scenarios might serve to explain
this abrupt expansion of the Mississippian political economy – including new, ritually
based exchange networks, intermarriage, and even migration. Evidence for these
activities taking place in a context of hostility is slim.
The Middle Mississippian period is one of considerable changes in macroregional
settlement dispersal, but relatively little change in overall site frequency (Figure 6.4). The
total number of mound centers in the northern third of the state decreases from 21 to 12.
At the same time, there is a dramatic increase in the number of mounds in the interior
coastal plain and along the coast.
Four new mound sites are established just below the transition between the fall
line hills and the interior coastal plain. From west to east, these sites are Lawton
(38OC11), Sawyer (9LS1), Sandy Hammock (9PU1), and Magnolia Plantation (9DU1 –
with which the reader is by now familiar). Each site is separated by at least one
watershed, suggesting possible buffer zones and is reminiscent in other ways of patterns
observed by Hally (1994:161-163), with regard to the location of later polities. In
258
discussing polities along the transition zones between the piedmont physiographic
province and the fall line hills and ridge and valley provinces, he suggests that polities
were located to maximize resource in these areas of extreme diversity (Hally 1994:163).
FIGURE 6.4,
Middle Mississippian period settlement patterns
259
By contrast, Middle Mississippian period mound sites were founded just below
the fall line hills. Figure 6.5, illustrates the relationship between these polities and smallscale data describing soil distributions. With the exception of Magnolia Plantation, all of
the mound centers are located on floodplain soils and in areas where these soils begin to
expand to into what Smith (1978:481) refers to the meander belt. The three sites located
in this belt are also associated with the Chewacla soil series, floodplain soil associates
with excellent agricultural production (TerraTech 2004). The locations of three out of
four of these sites are consistent with Smith’s (1978:48-483) model for the adaptive niche
of Mississippian period populations.
However, in addition to the relationship to local ecology, these site locations
suggest that occupants located mound centers on the basis of macroregional exchange
relationships. By being located on major rivers, each site, or cluster of sites are located on
major routes between the interior and the coast. Although waterborne travel was probably
more important, we may infer an overland route between the coast and interior from
traverses the Chickasawhatchee Creek watershed, as both Hudson (1997) and Swanton
(1985[1939]) agree that de Soto crossed the Chickasawhatchee Swamp on the way to
Ichisi.
Comparative analogues support this explanation. On the Pacific Slope of
Guatemala, Bove (1989:120-123) notes a southerly settlement expansion along routes
between the coast and the mountains. Junker (1999) notes similar trends among
chiefdoms of the Philippines. In each of these cases, researchers partially explain these
expansions as attempts to control the flow of resources between coastal and inland
260
groups. King’s (2003:124) recent research concerning trade routes between coastal
groups and the people at Etowah (9BR1) provides an additional line of support.
FIGURE 6.5,
Middle Mississippian period mound and soil distributions A) statewide in
Georgia B) near Magnolia Plantation, C) near the Sawyer and Sandy Hammock
mound sites, and D) near the Lawton Mound
261
Several Middle Mississippian period sites exist at the most southerly points of the
fall line hills – in the Chattahoochee Valley. Between 1100 and 1200 Rood’s Landing
(9SW1) and Singer Moye (9SW2) were probably contemporaries (Blitz and Lorenz
2006:83). Interaction between the occupants of Magnolia Plantation and Chattahoochee
Valley mound centers is suggested by ceramic similarities and the short distance between
Singer Moye and Magnolia Plantation (62 km). At the Middle Mississippian period’s
end, only one Chattahoochee Valley center remained in use (Blitz and Lorenz 2006:83).
During the Late Mississippian period (Figure 6.6), some areas, such as the
Oconee Valley provide clear evidence for settlement expansion and population growth
(Kowalewski and Hatch 1991; Williams 1994a). At the same time, the Middle
Mississippian period sites below the fall line are abandoned and a new set of mound
centers are established or re-occupied along the transition between the fall line and the
piedmont (Hally 1994:162; Worth 1988). Nearly the entire Savannah River Valley is
abandoned in the Late Mississippian period (Anderson 1994). As noted by Blitz and
Lorenz (2006:134-140) the Chattahoochee Valley exhibits local cycles of mound
abandonment and reoccupation beginning around A.D. 1450 and continuing through the
Protohistoric period. This cycle strongly resembles the cycles of the Savannah and
Oconee Valleys, but seems to continue longer than either of them (Anderson 1994;
Williams 1994a).
Across present-day Georgia, the abandonment of the mound centers below the fall
line was accompanied by a general southward shift in settlement (Figure 6.6). A similar
trend exists along the Ocmulgee River, where the abandonment of Sandy Hammock is
262
also accompanied by a southward settlement shift. Lamar Complicated stamped pottery is
evident, but with variations in decorative design not seen elsewhere in Georgia (Snow
1990). Southward shifts within the Ocmulgee Valley and possible new occupations in the
Satilla River Valley are not accompanied by additional mound building activities.
Instead, sites in this area seem to be located along trails (Snow 1990:92).
Despite the absence of mounds, there is persistent evidence for Late Mississippian
period interregional contacts in some parts of coastal plain. In addition, Mississippian
ceramic diagnostics from 17-Mile Creek, south of the Satilla River, suggest that such
occurred in both the Middle Mississippian and Late Mississippian periods (Kirkland
1994). There is strong evidence for continued occupation and some mound construction
along the Atlantic coast (Pearson 1977), but apart from the Middle Mississippian
platform mound at Irene Platform Mound (Caldwell and McCann 1941) there are
significant differences in coastal settlement strategies (Hally 1994:164; Thompson n.d.).
Evidence for regional continuity of settlement in the interior coastal plain is
strong, in spite of evidence for sub-regional abandonment. However, variation in the
continuity of mound construction practices is striking. Mound construction ceased along
Georgia rivers draining into Atlantic coast, while continuing along rivers that drain into
the Apalachicola River, and ultimately, the Gulf Mexico. These patterns may correspond
to differences between interregional contact strategies among groups interaction with
Atlantic and Gulf Coast populations – and indeed to the preferences of these populations
themselves.
263
FIGURE 6.6,
Late Mississippian period mound centers
In the interior coastal plain, settlement aggregation and mound construction
appears only to have been advantageous in a few places. Trade with coastal areas has
frequently been posited as an engine of macroregional exchange in the southeast (see
264
King 2003). As noted in Chapter 2, trade goods are often found in large quantities at
mound centers and it is plausible to infer that mound centers are indexical of exchange.
Exchange, like settlement aggregation or mound building, has its costs. Among
the greatest of these is transportation or travel (Skinner 1964 and 1965). More
opportunities for exchange over shorter distances provide incentives for increased trade.
If mound ceremonialism also facilitates trade, local investments in mound building may
be more likely to pay off in areas where trade goods can be transported and exchanged
over shorter distances. Figure 6.7 compares straight line distances between the Atlantic
and Gulf Coastal zones and associated regions in the interior.
Along the Gulf Coast interior, the Tallahassee Red Hills (Scarry 1996) provide a
resource rich environment similar to the fall line hills. This area is only 50 km from the
Gulf Coast. Between this environment and the southernmost extension of the fall line
hills, the overall intervening distances required for trade goods is also short, about 100
km. These distances form part of King’s (2003:121) model for an early and Middle
Mississippian period trade route.
As noted, Middle Mississippian period mound center expansion occurred
immediately below the fall line hills. Along the Atlantic coast, it appears that settlement
aggregation was not pursued further downriver. Thus, even in the Middle Mississippian
period, when the expansion of mound sites had reached its zenith, journeys to interior
mound sites would have been nearly 200 km.
265
FIGURE 6.7
Straight-line distances between Gulf and Atlantic Coast drainages and interior
settlement zones
Ceramic evidence also suggests linkages between Gulf Coast drainages and
intensified trade. In the Chattahoochee Valley, Blitz and Lorenz (2006:138-139) note that
reoccupations of mounds to the south of Singer-Moye and Rood’s Landing coincide with
266
the appearance of Fort Walton Incised bowls and other forms associated with the
Apalachee polity and the Fort Walton ceramic tradition (Scarry 1985 and 1996). In the
Chickasawhatchee Swamp, the transition between Magnolia Plantation and Long Walk
suggests a similar pattern. However, Fort Walton materials are also present at
Chickasawhatchee Knoll, suggesting that the shift between Magnolia Plantation and
Long Walk had little effect on Chickasawhatchee Knoll’s importance in the settlement
system. In addition to Fort Walton Incised pottery, Lamar Complicated Stamped sherds
are also present at sites in the Chattahoochee Valley and in the Chickasawhatchee
Swamp. Perhaps these two areas were also part of late Mississippian trade networks with
the interior as well.
Hally’s (1994:164) summary of the Georgia coast highlights the need for more
research into the settlement patterns and political economy of this area. Blitz and Lorenz
(2006:104) note the need for additional research along the Apalachicola river and in the
Tallahassee Red hills, observing that several mound sites have yet to be investigated.
Bearing these difficulties in mind, I can only speculate that lower transport costs along
Gulf Coast drainages provided incentives to link mound building, settlement aggregation
and trade in way lacking along the Atlantic Coast.
Differences in potential transportation costs may have helped to fuel differences
in the relationships between mound construction, settlement aggregation, and exchange.
While the Middle Mississippian period expansion of mounds may have been part of a
broader expansion of this new model for political economy, the greater transport costs
267
along the Atlantic coast may have worked in concert with the interior coastal plain’s
landscape structure to bring about an end to mound building activities.
Figure 6.8 presents synchronous view correlating mound occupation history
physiographic variation. The most striking feature of this map is the strong evidence for
continuity in mound construction and use at a few key locations, such as the area near
Macon Plateau, the Etowah River Valley, and the central Oconee and Upper Savannah
River Valleys. Apart from the areas, the map suggests expanding and receding waves of
mound ceremonialism – possibly tied to the expansion and contraction of regional
exchange networks.
Perhaps due to the unevenness of archaeological investigations through time and
the corresponding variation in data quality, this pattern seems more pronounced during
the Mississippian period (Figure 6.9). In an overall context of settlement and population
growth, the mound construction associated with the Mississippian political economy was
relatively stable in northern Georgia. Beginning with Macon Plateau, this system
expanded southwards. However, in all parts of the interior coastal plain, mound building
ceased after the Middle Mississippian period.
The Chickasawhatchee Swamp is an exception to this observation, in that the
Long Walk site represents a newly established Late Mississippian period mound site.
However, as noted in Chapter 5, the social context of mound use seems to have been
quite different in this area. Even though mound use persisted in the swamp, mound
centers were consistently located in sub-regions away from the main zones of
continuously occupied settlement.
268
FIGURE 6.8,
Physiographic provinces and long-term mound distribution patterns
269
FIGURE 6.9,
Physiographic provinces, Mississippian period mound distributions, and Late
Mississippian period non-mound site distributions
While the differences between Atlantic and Gulf Coastal polities offer possible
partial explanations for the cessation of mound building, the settlement pattern variation
in the Chickasawhatchee Swamp again raises the issue of local ecological variation and
270
selective participation in broader spheres of exchange. Upon close inspection, it seems
that Middle Mississippian period mound centers below the fall line hills do not conform
perfectly to Smith’s (1978) model of meander belt resource exploitation. Understanding
these differences requires a return to a finer scale analysis.
A Macroregional View of Intrasite Variation at Mound Centers
Instead comparative data concerning the size, arrangement, and ceramic richness
and diversity of selected mound centers provides preliminary evidence concerning the
relationships between local political economy, regional ecological variation, and
macroregional exchange. Given the limited nature of the sample, conclusions from this
comparison are preliminary and subject to revision.
Table 6.2 presents data from mound centers in the fall line hills and interior
coastal plain – as well as the four sites that were tested in the Chickasawhatchee Swamp.
Each site has been subjected to systematic shovel testing. To ensure data comparability
data between sites from the Chickasawhatchee Swamp and other areas, site area
estimates, component counts, and indices of ceramic richness and diversity are derived
exclusively from shovel test data. Several patterns emerge from these data, some relevant
to specific time periods, others to more long-term trends.
The lower diversity and frequency of sherds per shovel tests among
Chickasawhatchee Swamp sites lends support to the hypothesis, raised in Chapter 5, that
ceramic production in the was less intensive than in neighboring regions. Figures 6.10
and 6.11 presents these differences in graph and map form. In Figure 6.10, the X axis is
the natural logarithm (ln) of sherds per positive shove text and the Y axis is number of
271
types per component The size of each data point represents component counts. The three
sites in the bottom left hand corner of Figure 6.10 are Hay Fever Farm, Windmill
Plantation, and Red Bluff. Of the four sites in the Chickasawhatchee Swamp (in italics),
Magnolia Plantation is the only the site that exhibits similar typological diversity to sites
in neighboring regions. Given the diversity observed from the XU 1, at Hay Fever Farm,
it is likely that the history of surface collecting at this site has affected the site’s lower
type count.
Mound Count
Total Tests
Total Positive Tests
Sherd Count
% Positive
Sherds/Test
Sherds/Positive Test
Types/Component
13.0
3.6
4
2
0
0
111
91
56
31
169
72
7 50.5
12 34.1
1.5
0.8
3.0
2.3
1.8
6.0
1.3
144.0
2
2
1 103
9 1310
28
707
47
5367
11 27.2
18 54.0
0.5
4.1
1.7
7.6
5.5
9.0
4.1
6.0
1
1
2
3
1
2
53
122
215
33
117
179
186
4014
1368
9 62.3 3.5 5.6 9.0
10 95.9 32.9 34.3 10.0
16 83.3 6.4 7.6 8.0
1.4
12.0
6.3
3
2
1
2
1
1
1
2
301
146
40
150
72
82
39
126
1356
369
439
3374
14
16
10
18
Type Count
Components
Hay Fever
Non-Mound Farm*
Red Bluff*
Sites
Windmill
Woodland Plantation
Kolomoki
Magnolia
Plantation
Lawton
Middle
Mississippian Sawyer*
Sandy
Hammock*
Lewis*
Hartley Posey
Late
Mississippian Lamar
Total or
Average
Site Area (ha)**
Site
Site area, shovel test frequency, and ceramic density data for mound sites in the
Fall Line Hills and Coastal Plain physiographic provinces
Site Type
TABLE 6.2,
23.9 4.5 18.8 4.7
56.2 2.5 4.5 8.0
97.5 11.0 11.3 10.0
84.0 22.5 26.8 9.0
6.0 22 21 2642 1470 16761 12.8 60.8
8.2 11.2
7.4
Comparisons between Woodland period sites emphasize massive differences in
scale with respect to Kolomoki and other Woodland period mound centers. Data from the
272
Lewis Mound formed part of the database from which Pluckhahn and McKivergan
(2002) argue for an inverse relationship between wetland ubiquity and settlement
consolidation. In spite of the differences in political economic and environmental context
provided by the Lewis Mound’s proximity to the coast, the Lewis Mound and Windmill
Plantation are similar both in terms of mound and occupation size. Given these
similarities, data from Windmill Plantation provide further support for Pluckhahn and
McKivergan’s hypothesis.
During the Mississippian period, the tight clustering between Magnolia Plantation
and Sawyer suggest similar processes at work in the foundation, occupation, and
abandonment of these mounds. The slightly higher typological diversity and ceramic
frequency indices at Sawyer and the higher frequencies and Sandy Hammock may be
related to multiple components. In Figure 6.10 the three outlying sites in terms of ceramic
density are, from left to right Hartley Posey, Lamar, and Lawton. Given their location
above the fall line, the higher ceramic densities at Hartley Posey and Lamar are
consistent with expected differences in fall line and coastal plain mound centers. Lawton
however, is an outlier that cannot be explained at this time.
Figure 6.10 also supports the idea of differences in settlement intensity between
Woodland and Mississippian period mound centers. In spite of its remarkable scale and
400 year occupation history, Kolomoki’s ceramic density values are most comparable
with the occupations at the smaller sites of Sawyer and Magnolia Plantation. More lightly
occupied still are occupations at Windmill Plantation and the Lewis Mound. The
273
geographic distributions of ceramic frequency in Figure 6.11 highlight relationships
between physiography and occupational intensity.
12
10
Types/Component
8
Hay Fever Farm
Red Bluff
Windmill Plantation
Kolomoki
Magnolia Plantation
Lawton
Sawyer
Hartley Posey
Lamar
Lewis
Sandy Hammock
6
4
2
Data point sizes reflect number of ceramic components
0
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
ln(Sherds/Positive Shovel Test)
FIGURE 6.10, Ceramic richness and diversity at sites in the Fall Line Hills and Coastal Plain
physiographic provinces
274
FIGURE 6.11, Ceramic density data according to sherds/positive shovel test at ten study sites
within the Fall Line Hills and Coastal Plain physiographic provinces
275
Summary
This chapter presented a macroregional overview of settlement pattern variation
in the physiographic provinces of Georgia and a snapshot of intrasite variation among
mound centers in the fall line hills and the interior coastal plain. Though based on data
that require additional processing and refinement, this overview supports hypotheses,
raised in earlier chapters, concerning the Chickasawhatchee Swamp. They also highlight
several trends relevant to the larger themes in the dissertation.
At the macroregional scale, concordant changes in settlement patterns support
arguments linking mound use and interregional exchange and argue for the long-term
interconnectedness of eastern Woodlands societies. Of particular note is the Middle
Mississippian period expansion of mound centers below the fall line. While these centers
were located in an analogue of Smith’s (1978) meander belt, the relatively light
occupation at these centers and their rapid abandonment suggest that the Atlantic slope
coastal plain did not provide the adaptive advantages of similar areas in the Mississippi
Valley. Instead, the differential transport costs may have led to diverging relationships
between mound building, settlement aggregation, and exchange in the Gulf and Atlantic
drainages.
Macroregional and intrasite data together strongly support arguments for broad
differences in political economic strategies between the piedmont montane, and fall line
hills zones and the interior coastal plain. These differences have been noted before,
especially by Hally (1994:163) and Scarry (1994:23-24). This chapter’s comparison
supports the hypotheses originally presented by Pluckhahn and McKivergan (2002) and
276
developed in Chapter 5, that the structure of many interior coastal plain landscapes
discourages economic strategies that revolve around settlement aggregation. In the
Chickasawhatchee Swamp, settlement aggregation persisted throughout the Woodland
and Mississippian period, but only as part of broader, more mobile strategy involving
repeated local and sub-regional abandonment. These cycles and the abandonment of
mound building activities suggest that macroregional interactions were undertaken
strategically by coastal plain populations who maintained dispersed settlement strategies
as a means of coping with coastal plain landscapes. The full implications of these trends
and a summary of the results of this dissertation are presented in the final chapter.
277
CHAPTER 7: SUMMARY AND CONCLUSIONS
This research provides a new approach to understanding long term variation
among Woodland and Mississippian period societies. Delineating variation in the scale
and complexity of social systems is more important than determining why a given
prehistoric development sequence does or does not fit a chiefdom, or even “middle range
society” model. Relationships between local ecological variation and macroregional
exchange provide new insights into social change in the prehistoric eastern Woodlands.
Beyond the results relevant to southeastern archaeologists, this research also
contributes to broader goals in anthropological archaeology. Multiscalar analysis
provided the means to attempt a balanced investigation into agency and structure. Within
this multi-scalar view, an emphasis on macroregional, long-term perspectives provides a
more complete picture of social variation. Theories of landscape patch dynamics provide
improved comparative views of macroregional ecology. In this multi-scalar framework,
inter-regional interaction and regional ecological variation are powerful variables.
At the regional and intrasite scale, new settlement pattern data demonstrate that
the heterogeneous landscape patches of the Chickasawhatchee Swamp are not conducive
to residential and subsistence strategies involving settlement aggregation. However,
periodic mound construction at swamp edges suggests the importance of participation in
broader spheres of interaction. Macroregional patterns of concordant change in the
distribution of non-mound and mound centers support both the idea of greater temporal
continuity in settlement structure and the idea that broad scale landscape structures
shaping local settlement intensity.
278
These insights are not dependent on neo-evolutionary models or social typologies,
and indeed, may not be possible under such frameworks. Instead of discovering the
"laws” that govern human societies, researchers should seek to explain the diversity and
variation present in the archaeological record (Hegmon 2003), recognizing that any
generalizations that social scientists formulate are historically and socially constituted
(Giddens 1979; Trigger 1989). I summarize specific results below and discuss them in
terms of emerging trends in anthropological theory and the need for future research. I
continually return to the themes of macroregional comparison and a long term
perspective because these approaches yield key insights into the old and significant
question of why humans were, are, and will continue to be different from one another.
Results
New data from the Chickasawhatchee Swamp provide a bottom-up view of local
settlement pattern variation across the Woodland and Mississippian periods. Data
collection strategies reflect a multiscalar approach aimed at maximizing the
documentation of settlement diversity and providing ease of access to ceramic variation
data at virtually any scale. Formation process research strengthens the study’s empirical
base.
Careful considerations of site formation processes and evaluations of appropriate
methods mitigated biases introduced by the high degree of post-depositional disturbance
in the Chickasawhatchee Swamp. The site formation process data highlight the fragility
of coastal plain sites located in shallow, sandy soils. However, it is also clear that major
279
insights concerning past behavior are possible using measures of overall occupational
continuity, presence/absence measures of ceramic diversity, and shovel test data.
Chi square tests from low intensity survey correlate prehistoric settlement
locations with areas that were likely disturbed by anthropogenic fires. This strategy
seems unlikely to have been successful in creating the riparian agricultural habitats
favored by many Mississippian societies (e.g. Smith 1978 and 1995). However, burning
would likely have provided a variety of opportunities for wild resource exploitation in
longleaf pine forests and adjacent hardwood habitats (Cowell 1995; Wagner 2003). A bimodal settlement system of large and small sites and cyclic local and sub-regional
abandonments likely reflected a strategy for exploiting the swamp’s heterogeneous
landscape.
The deliberate spatial separation between the areas of greatest settlement
continuity and the location of civic-ceremonial centers suggest strategic participation in
inter-regional interaction. According to Hudson’s (1997:51) account of the De Soto
entrada into the Chickasawhatchee area (then known as Capachequi), the primary
strategy for dealing with De Soto and his men was simply to disappear into the swamp.
The major centers at Blue Hole Hill, Spring Hill, Tallassee Plantation, Magnolia
Plantation, and Long Walk were outside the swamp, in areas more conducive to
aggregated settlements. However, each was also quite close to a dense wetland. Apart
from Chickasawhatchee Knoll, no center was occupied for very long. These patterns
suggest that that some distancing between the Swamp’s inhabitants and outside visitors
may have been a long-term strategy.
280
The correlation between settlements and probable locations for prehistoric
anthropogenic fire also suggests a strategic approach to landscape affordances
(Heckenberger 1998). Over time, the anthropogenic landscape became an institution,
created and recreated by the community. As such, it played a role in structuring social
change. Together, strategies for locating and maintaining settlements reflect strategic
needs for regional and local autonomy and the institutional constraints imposed by the
natural and geo-political landscape.
Ceramic data from the Chickasawhatchee Swamp support the conclusion that the
area’s occupants were not as intensely engaged in pottery manufacturing and use as
neighboring regions. A more limited range of vessel forms and sizes, inconsistencies in
manufacturing techniques, the ubiquity of plainwares, and the generally lower ceramic
frequencies all point to a marginal ceramic economy. The ceramic data alone do not
allow us to determine whether these differences reflect differential subsistence strategies,
lower population densities, or shorter occupations and greater mobility between sites, but
Chickasawhatchee Swamp settlement patterns and the contrast between coastal plain and
fall line sites indicate that some combination of the three is the most likely scenario.
Without specifying the exact social context of exchange, macroregional
settlement pattern data support arguments for the importance of inter-regional contact in
driving settlement pattern change and link mound ceremonialism to exchange (e.g. Elson
1998; Jeffries 1976 and 1981; Kowalewski 1996; Williams 1996). The expansion and
abandonment of Middle Mississippian mound centers below the fall line hills is an
especially strong indicator of concordant change. The fact that Late Mississippian
281
societies respond differently to changes on the Atlantic and Gulf coasts demonstrates the
relationship between this concordant settlement shift and inter-regional interaction.
Overall, these data suggest broader stability in strategic approaches to local
landscape affordances. The long-term settlement pattern trends across the physiographic
provinces of Georgia illustrate patterns of concordant change that are highly suggestive
of long-term interaction over great distances. In addition, long-term macroregional
variation in intra-site patterns at mound centers support arguments for differing
settlement strategies in interior coastal plain, fall line hills, piedmont, and montane
physiographic provinces.
Through time, the overall number of mound centers in any of the landscapes
represented by physiographic province divisions is relatively constant. In the interior
coastal plain, the more resilient, easily disturbed landscapes have less intensive
occupations. By contrast, large concentrations of population and early shifts in social
structure consistently take place in the same areas, such as around Macon Plateau area
and in the Ridge and Valley province. Through each period, mound ceremonialism of
some form is a constant. And yet, the changing configurations of sites, mounds, and
exchange materials indicate shifting social contexts for exchange.
Discussion
The similarities and differences in the social context of mound ceremonialism are
reminiscent of variation Wiessner (2002) observes among the Enga of Papua New Guinea
– but with an important exception. The introduction of the sweet potato helped to provide
282
a production context for three Enga ritual systems associated with inter-regional
exchange. The Tee cycle and Great War ceremonies are most similar to this case.
Tee cycle and Great War ritual performances served as a backdrop for the
arrangement of marriages, the formation of alliances, and the exchange of wealth. These
rituals provided the means for organizers to augment personal power (Wiessner
2002:247-248). Over time, each ritual cycle grew in terms of the number of participants,
the formalization of leadership offices, and the amount of wealth invested. Each ritual
system also underwent a final period of collapse. In the case of the both the Tee cycle and
the Great Wars, organizers simply decided the events were too expensive – to borrow the
words Pluckhahn (2002:218) chose to describe Kolomoki, the systems “collapsed under
the weight of their own success.”
The important area of divergence between the Enga case and that of the
southeastern United States is that, although Enga cycles were highly ritualized, they were
also somewhat secular. Those who managed cycles of inter-regional exchange were
barred from holding offices linked to supernatural power and fertility (Wiessner
2002:252). The contrast between the Enga and southeastern United States is especially
striking with regard to Mississippian period leadership offices, where ceremonial and
political offices seem to have been unified in many cases.
Southeastern archaeologists necessarily miss the variation Wiessner (2002)
recorded and under-represent the dynamism of prehistoric exchange networks.
Nevertheless, the very stability of mound-building and recurring concordant changes in
settlement should lead archaeologists interested in middle range societies to heed
283
Kowalewski’s (1996:32) observations that interconnection between groups is a given –
even over large distances. Woodland and Mississippian period mound ceremonialism
should remind us of the Enga case, in which the ritual events that attracted the largest
numbers of people tended to revolve around inter-regional exchange.
Future Research Directions
The influence of research aimed at describing diversity in middle-range societies
is evident in current southeastern archaeological practice. Relevant trends include a
renewed emphasis on agency and leadership (e.g. Barker and Pauketat 1992; Beck 2004;
Blitz and Lorenz 2006; Emerson 1997; Pauketat 1994), recognition of the importance of
inter-regional interaction (King 2003; Kowalewski 1996; Muller 1997; Peregrine 1992
and 1996), and attempts to explore variation across space (King and Meyers 2002).
Southeastern archaeologists have also been collecting massive quantities of new
data and generating new regional syntheses (Anderson 1994; Blitz 1993; Blitz and
Lorenz 2006; Hally 1994; Knight and Steponaitis 1998; Milner 1998; Moore 2002;
Muller 1997; Pauketat 1994; Scarry 1996; Schroeder 1997; Williams 1994a; Williams
and Shapiro 1990). Much of this work focuses on the emergence of social inequality and
has revolves around theoretical frameworks associated with the chiefdom model. These
major contributions provide a better understanding of Mississippian societies than ever
before.
However, recent discussions suggest that there may be limits to the number of
research questions that can be answered by research into variation among chiefdoms
alone (Kowalewski 1996; King and Meyers 2002; Muller 1997). Although there are
284
many similarities between archaeological correlates of Mississippian period social
organization and the social structure laid out in the chiefdom model, we must remember
that it is, after all, only a model. The long-term, macroregional view strongly suggests
that, if we look closely at the temporal and spatial edges of the so-called “Mississippian
world,” the chiefdom model is a bit frayed around the ends.
A continued focus exclusively on chiefdoms may hamper additional research by
encouraging continuing emphasis on short-term processes and the role of agency while
overlooking the balancing role of structure and structuration (Giddens 1979). The neoevolutionary baggage of the chiefdom model has also led to a shortage of problem
oriented research and data regarding the Woodland period (Anderson and Mainfort 2002;
Cobb and Garrow 1996; McElrath et al. 2000; Pluckhahn 2003).
Instead of continuing with this approach, research from the Chickasawhatchee
Swamp demonstrates the utility of a historically grounded model and a balanced
approach to agency and structure. Two relatively straightforward variables,
macroregional exchange and landscape structure – contribute to our understanding of
variation among middle range societies without making a particularly strong appeal to or
case against the chiefdom model. Additional research will contribute the theories and
methods necessary to place the understanding of diversity at the center of research.
Within the Chickasawhatchee Swamp, additional excavation and survey data will
provide refined chronological control, a more detailed picture of the regional political
economy, and, in the northern part of the project area, an expanded view of settlement
patterns. At the macroregional scale, additional efforts to assemble a comprehensive view
285
of settlement patterns and the social context of mound use could provide insights similar
to those achieved by Smith (2002) in her macroregional study of highland Mesoamerica.
In the southeastern case, ecological context is a crucial factor. Publicly available GIS data
concerning elevation and soil distribution can provide two elements of the natural
context, but historical data from witness trees would provide a more detailed picture of
pre-colonial ecological diversity, thereby increasing our understanding of variation
(Foster et al. 2004).
These individual analytical steps will further clarify the interior coastal plain’s
place in southeastern prehistory, but a broader agenda is also necessary. Theoretical
frameworks for human behavior must continue to eschew typological classification in
favor of historically constituted models of variation. Macroregional exchange networks
and local ecological variation have proven effective in building such models. Landscape
patch dynamics and alternative stable state theories provide flexible frameworks for
broad ecological comparisons. Together, agency, structure, and landscape ecology may
provide the tools for a more integrated view of change among Eastern Woodlands
societies.
286
APPENDIX A:
TABLE LISTING SITES, SITE AREAS, AND COMPONENTS OF ALL
CERAMIC SITES
100
100 x
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100 x
100
100
100
100
100
100
100
4
x
Unknown Indian/Ceramic
General Mississippian
Late Mississippian
Middle Mississippian
General Woodland
0.37
45.04
36.68
0.90
19.36
4.16
0.12
2.72
0.12
0.44
0.35
0.90
0.16
2.47
4.79
0.70
0.57
2.39
4.40
1.17
0.41
1.75
3.22
0.16
1.21
0.10
1.07
7.05
1.69
2.69
0.48
9.37
0.73
0.20
0.11
2.63
0.79
0.15
1.40
15.41
Late Woodland
0.37
45.04
36.68
0.90
19.36
4.16
0.12
2.72
0.12
0.44
0.35
0.90
0.16
2.47
4.79
0.70
0.57
2.39
4.40
1.17
0.41
1.75
3.22
0.16
1.21
0.10
1.07
7.05
1.69
2.69
0.48
9.37
0.73
0.20
0.11
2.63
0.79
0.15
1.40
0.60
Middle Woodland
Site Area
2
2
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
4
1
1
1
2
1
1
1
2
% of Site Area
Component Area
CAS 4
CAS 5
CAS 6
CAS 7
CAS 9
CAS 10
CAS 13
CAS 24
CAS 27
CAS 30
CAS 35
CAS 37
CAS 47
CAS 49
CAS 51
CAS 63
CAS 64
CAS 66
CAS 68
CAS 69
CAS 70
CAS 71
CAS 72
CAS 73
CAS 74
CAS 77
CAS 78
CAS 81
CAS 85
CAS 86
CAS 88
CAS 89
CAS 93
CAS 94
CAS 98
CAS 99
CAS 104
CAS 108
CAS 109
CAS 115:1
Component Count
Field Site #
287
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
26
41
3
100
100
17 x
13
100
100
20
100
1
19
100
22
19
5
100
100
100
79
13
100
100
100
100
36
100
1
100
100
100
100
100
100
1
9
9
14
100
x
Unknown Indian/Ceramic
General Mississippian
Late Mississippian
Middle Mississippian
General Woodland
15.41
3.90
6.95
0.34
0.14
15.35
15.35
0.34
0.94
3.52
0.03
9.07
9.07
0.90
3.14
5.10
5.99
0.07
1.12
4.45
8.53
7.09
0.03
1.76
0.37
0.03
7.58
8.92
5.55
1.91
2.11
0.08
0.44
3.65
0.42
18.74
18.74
18.21
18.21
0.05
Late Woodland
3.93
1.60
0.24
0.34
0.14
2.64
1.96
0.34
0.94
0.69
0.03
0.06
1.74
0.90
0.69
0.98
0.30
0.07
1.12
4.45
6.75
0.92
0.03
1.76
0.37
0.03
2.72
8.92
0.04
1.91
2.11
0.08
0.44
3.65
0.42
0.28
1.76
1.62
2.48
0.05
Middle Woodland
Site Area
2
1
1
1
1
3
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
% of Site Area
Component Area
CAS 115:2
CAS 116
CAS 117
CAS 120
CAS 121
CAS 122:1
CAS 122:2
CAS 123
CAS 124
CAS 126
CAS 137
CAS 139:1
CAS 139:2
CAS 149
CAS 153
CAS 155
CAS 156
CAS 157
CAS 160
CAS 161
CAS 162
CAS 163
CAS 167
CAS 169
CAS 170
CAS 172
CAS 173
CAS 174
CAS 175
CAS 176
CAS 177
CAS 179
CAS 180
CAS 181
CAS 182
CAS 183:1
CAS 183:2
CAS 184:1
CAS 184:2
CAS 186
Component Count
Field Site #
288
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
CAS 187
CAS 190
CAS 191
CAS 192
CAS 194
CAS 197
CAS 199
CAS 209
CAS 213:1
CAS 213:2
CAS 215
CAS 220
CAS 222
CAS 225
CAS 229
CAS 235
CAS 238
CAS 240
CAS 242
CAS 9CAL9(SGC)
CAS 9DU11(SGC)
CAS9CAL20(SGC)B
CAS 9CAL2(SGC)
CAS 9DU6
CAS 9CAL18(SGC)
CAS 9CAL17(SGC)
CAS 9CAL10(SGC)
CAS 9CAL4(SGC)
CAS 9CAL8(SGC)
CAS 9CAL11(SGC)
CAS 9CAL5(SGC)
CAS 9CAL13(SGC)
CAS 9CAL12(SGC)
CAS 9CAL7(SGC)
Total Areas (ceramic)
Total Areas (all)
1
1.66 12.60
1
6.64
6.64
1
3.56
3.56
1
4.30
4.30
1
0.11
0.11
1
0.09
0.09
1 10.88 10.88
1
0.11
0.11
2
3.01
6.27
2
3.26
6.27
1
1.12
1.12
1
1.37
1.37
1
0.46
0.46
1
1.72
1.72
1
0.12
0.12
2
6.39
6.39
1
1.07
1.07
1
0.33
0.33
2
8.99
8.99
1
0.42
0.42
1
1.09
1.09
1
0.17
0.17
1
0.39
0.39
2
1.31
1.31
2
1.35
1.35
1
1.30
1.30
2
0.17
0.17
1
0.07
0.07
1
0.27
0.27
2
1.22
1.22
2
2.85
2.85
1
0.17
0.17
1
0.16
0.16
2
2.58
2.58
137 291.76 487.91
137 291.76 522.72
13
100
100
100
100
100
100
100
48
52
100
100
100
100
100
100
100
100
100
100
100
100
100
100 x
100
100
100
100
100
100
100
100
100
100
60
56
Unknown Indian/Ceramic
General Mississippian
Late Mississippian
Middle Mississippian
General Woodland
Late Woodland
Middle Woodland
% of Site Area
Site Area
Component Area
Component Count
Field Site #
289
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
290
APPENDIX B:
COMPLETE PROVENIENCE LIST FOR PROJECT SURVEY AND TESTING
291
Lot numbers higher than 1000 are from West Georgia College and South Georgia
College surface collections from the Chickasawhatchee Swamp. Lot numbers higher than
2000 are from Don Smith’s 1962 excavations at 9DU6.
Project Lot #
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Provenience Description
CAS 1, Collection Area A.
CAS 1, Collection Area B.
CAS 1, Collection Area C.
CAS 1, ShovelTest No: 1
CAS 1: ShovelTest No: 2
CAS 1: ShovelTest No: 3
CAS 1: ShovelTest No: 4
CAS 1: ShovelTest No: 5
CAS 1: ShovelTest No: 7
CAS 1: ShovelTest No: 8
CAS 1: ShovelTest No: 10
CAS 1: ShovelTest No: 12
CAS 1: ShovelTest No: 13
CAS 1: ShovelTest No: 14
CAS 1: ShovelTest No: 15
CAS 1: ShovelTest No: 16
CAS 1: ShovelTest No: 17
CAS 1: ShovelTest No: 18
CAS 1: ShovelTest No: 19
CAS 1: ShovelTest No: 20
CAS 1: ShovelTest No: 21
CAS 4, Str. 2, Xu 1, Level 1 (0-10cm)
CAS 4, Str 2., Xu 2, Level 1 (10-20cm)
CAS 4, Str 2., Xu 1, Level 3 (20-30cm)
CAS 4: ShovelTest No: 3
CAS 4, Str. 3, Xu 2, Trench
CAS 4, Str. 3, Xu 2, Trench Profile Cleaning
CAS 4, Str. 3, XU 3, Level 1 (0-10cm)
CAS 4, Str. 3, XU 3, Level 2 (10-20cm)
CAS 4, Str. 3, XU 3, Level 3 (20-30cm)
CAS 4, Str. 3, XU 3, Level 3 (Floor Material)
CAS 4, Str. 3, XU 3, Level 4 (Surface Charcoal)
CAS 4, Str. 3, XU 3, Level 4 (30-40cm)
CAS 4, Str. 3, XU 3, Level 4 (Floor sherd)
CAS 4, Str. 3, XU 3 South, Level 5 (40-50cm)
CAS 4, Str. 3, XU 3, Level 6 (50-60cm)
CAS 4, Str. 3, XU 3 South, Level 7 (60-70cm)
CAS 4, Str. 3, XU 3, Balk Removal (35-40cm)
CAS 4, Str. 3, XU 3, Balk Removal (40-50cm)
CAS 4, Str. 3, XU 3 Center, Level 5 (40-50cm)
292
41
42
43
44
45
46
47
48
49
50
53
55
56
57
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
90
91
92
93
CAS 4, Str. 3, XU 3 Center, Level 6 (50-60cm)
CAS 4, Str. 3, XU 3 Center, Level 7 (60-70cm)
CAS 4, Str. 3, XU 3 Level (Floor Removal)
CAS 4, Str. 3, XU 3 North, Level 5 (40-50cm)
CAS 4, Str. 3, XU 3 North, L. 5 (post-map removal)
Profile Cleaning
CAS 4, ShovelTest No: 4
CAS 4, ShovelTest No: 5
CAS 4, ShovelTest No: 6
CAS 4, ShovelTest No: 7
CAS 4, ShovelTest No: 10
CAS 4, ShovelTest No: 14
CAS 4, ShovelTest No: 16
CAS 4, General Surface 12 Aug 03
CAS 5, ShovelTest No: 3
CAS 5, Xu 1, Level 1 (0-10cm)
CAS 6, General Surface
CAS 7, General Surface
CAS 10, ShovelTest No: 1
CAS 10, ShovelTest No: 2
CAS 10, ShovelTest No: 3
CAS 10, ShovelTest No: 4
CAS 10, ShovelTest No: 5
CAS 10, ShovelTest No: 6
CAS 10, ShovelTest No: 7
CAS 10, ShovelTest No: 8
CAS 10, ShovelTest No: 9
CAS 10, ShovelTest No: 10
CAS 10, ShovelTest No: 11
CAS 10, ShovelTest No: 12
CAS 10, ShovelTest No: 13
CAS 10, ShovelTest No: 15
CAS 10, ShovelTest No: 17
CAS 10, ShovelTest No: 18
CAS 10, ShovelTest No: 19
CAS 10, ShovelTest No: 20
CAS 10, ShovelTest No: 21
CAS 10, ShovelTest No: 22
CAS 10, ShovelTest No: 23
CAS 10, ShovelTest No: 23
CAS 10, ShovelTest No: 23
CAS 10, ShovelTest No: 24
CAS 10, ShovelTest No: 25
CAS 10, ShovelTest No: 26
CAS 10, ShovelTest No: 27
CAS 10, ShovelTest No: 28
CAS 10, ShovelTest No: 29
CAS 10, ShovelTest No: 30
293
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
135
136
137
138
139
140
141
142
CAS 10, ShovelTest No: 31
CAS 10, ShovelTest No: 32
CAS 10, ShovelTest No: 33
CAS 10, ShovelTest No: 34
CAS 10, ShovelTest No: 36
CAS 10, ShovelTest No: 37
CAS 10, ShovelTest No: 39
CAS 10, ShovelTest No: 40
CAS 10, ShovelTest No: 41
CAS 10, ShovelTest No: 42
CAS 10, ShovelTest No: 43
CAS 10, ShovelTest No: 44
CAS 10, ShovelTest No: 45
CAS 10, ShovelTest No: 48
CAS 10, ShovelTest No: 49
CAS 10, ShovelTest No: 50
CAS 10, XU 1, Level 1
CAS 10, XU 1, Level 2
CAS 10, XU 1, Level 2, Profile Cleaning
CAS 10, XU 1, Level 3
CAS 10, XU 1, Level 3, Profile Cleaning
CAS 10, XU 1, Level 4
CAS 10, XU 1, Level 5
CAS 10, XU 1, Level 5a
CAS 10, XU 1, Level 6
CAS 10, XU 1, Level 7
CAS 10, XU 1, Level 8
CAS 10, XU 1, Level 9
CAS 10, XU 1, Level 10
CAS 10, XU 1, Level 11
CAS 10, XU 1, Level 11
CAS 10, XU 1, Level 12
CAS 10, XU 1, Level 14
Surface Collection, pre-excavation, XU 1
CAS 16 A: General Surface
CAS 18 A: General Surface
CAS 15 A: General Surface
CAS 17 A: General Surface
CAS 19 A: General Surface
CAS 24 A: General Surface
CAS 32 A: General Surface
CAS 34 A: General Surface
CAS 40 A: General Surface
CAS 42 A: General Surface
CAS 46 A: General Surface
CAS 56 A: General Surface
CAS 58 A: General Surface
CAS 64 A: General Surface
294
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
CAS 66 A: General Surface
CAS 68 W: West Collection
CAS 68 E: East Collection
CAS 70 A: General Surface
CAS 72 A: General Surface
CAS 74 A: General Surface
CAS 78 A: General Surface
CAS 86 A: Collection Area A.
CAS 86 C: Collection Area C.
CAS 88 A: General Surface
CAS 90 A: General Surface
CAS 92 A: General Surface
CAS 94 A: General Surface
CAS 96 A: General Surface
CAS 98 A: General Surface
CAS 102 A: General Surface
CAS 104 A: General Surface
CAS 106 A: General Surface
CAS 108 A: General Surface
CAS 23 A: Late Archaic Biface
CAS 27 A: General Surface
CAS 29 A: General Surface
CAS 33 A: General Surface
CAS 35 A: General Surface
CAS 37 A: General Surface
CAS 43 B, Collection Area B
CAS 43 A: MALA Base Waypoint
CAS 47 A: General Surface
CAS 49 A: General Surface
CAS 49 B: Check Stamped Pottery
CAS 49 C: Ceramics
CAS 51 General Surface
CAS 55 General Surface
CAS 63 General Surface
CAS 69 Waypoints B & C General Surface
CAS 71 A: General Surface
CAS 71 B: General Surface
CAS 71 C: General Surface
CAS 73, Collection Area: Lithics and Ceramics
General Surface (Biface)
CAS 77 General Surface
CAS 79 West Side
CAS 81 A: General Surface
CAS 83 W: West Side
CAS 85, North Side of Site, Ceramics
CAS 93 General Surface
CAS 105 General Surface
CAS 107 General Surface
295
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
CAS 74 17 Feb General Surface
CAS 74 ShovelTest No: 1
CAS 74 ShovelTest No: 2
CAS 74 ShovelTest No: 3
CAS 74 ShovelTest No: 4
CAS 74 ShovelTest No: 5
CAS 74 ShovelTest No: 6
CAS 74 ShovelTest No: 8
CAS 74 ShovelTest No: 9
CAS 74 ShovelTest No: 10
CAS 51 General Surface 18 Feb 04
CAS 51 General Surface 19 Feb 04
CAS 51 General Surface 19 Feb 04
CAS 51 General Surface 19 Feb 04
CAS 51 General Surface 19 Feb 04
CAS 51 General Surface 19 Feb 04
CAS 51 A: Collection Area A
CAS 51 B: Collection Area B
CAS 51 C: Collection Area C
CAS 51 D: Collection Area D
CAS 51 ShovelTest No: 2
CAS 51 ShovelTest No: 4
CAS 51 ShovelTest No: 5
CAS 51 ShovelTest No: 8
CAS 68 General Surface 19 Feb 04
CAS 68 ShovelTest No: 5
CAS 68 ShovelTest No: 7
CAS 109 General Surface
CAS 109 ShovelTest No: 1
CAS 109 ShovelTest No: 2
ShovelTest No: 5
CAS 109 ShovelTest No: 6
CAS 110 ShovelTest No: 1
CAS 115 General Surface
CAS 115 General Surface
CAS 115 PPK
CAS 115 General Surface
CAS 115 General Surface - North End of Site
CAS 115 General Surface
CAS 115 General Surface
CAS 115 General Surface - North End of Site
CAS 116 General Surface
CAS 117 General Surface
CAS 117 Area B (pottery)
CAS 117 Area C (PPK)
CAS 117 Area D (Large Core)
CAS 117 Area E (PPK)
CAS 117 General Surface (27 Feb 04)
296
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
CAS 118 General Surface
CAS 119 General Surface
CAS 120 General Surface
CAS 121 General Surface
CAS 122 General Surface
CAS 122 General Surface
CAS 122 General Surface
CAS 122 General Surface
CAS 122 Collection Area B
CAS 122 Collection Area D
CAS 122 Collection Area E
CAS 122 Collection Area F
CAS 122 Collection Area G
CAS 122 Collection Area H
CAS 122 Collection Area I (PPK)
CAS 122 General Surface
CAS 122 Collection Area N
CAS 122 Collection Area N
CAS 122 Collection Area O (PPK)
CAS 122 Collection Area B
CAS 122 Collection Area Q (PPK)
CAS 122 Collection Area R (PPK)
CAS 122 Collection Area S
CAS 122 Collection Area S
CAS 122 Collection Area S
CAS 122 Collection Area S
CAS 122 Collection Area U (PPK)
CAS 122 Collection Area V (PPK)
CAS 122 Collection Area W (PPK)
CAS 123 General Surface
CAS 123 General Surface
CAS 124: General Surface, Bag 1
CAS 124: General Surface, Bag 2
CAS 125: General Surface
CAS 126: General Surface
CAS 128: General Surface
CAS 129: General Surface
CAS 130: General Surface
CAS 131: General Surface
CAS 132: General Surface
CAS 133: General Surface
CAS 134: General Surface
CAS 135: General Surface
CAS 136: General Surface
CAS 137: General Surface
CAS 138: General Surface
CAS 139: General Surface
CAS 139: Areas A & B
297
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
336
CAS 139: Area D
CAS 139: Area E
CAS 139: Area E2 & E5
CAS 139: Area E4
CAS 139: Area F PPK
CAS 139: Area G PPK
CAS 139: Area H PPK
CAS 139: Area I PPK
CAS 140: General Surface
CAS 139: Area A
CAS 140: Area D PPK
CAS 140: Area E PPK
CAS 141: General Surface
CAS 142: General Surface
CAS 144: General Surface
CAS 144: PPK
CAS 120: Sherd collected 27 FEB 04 (area b)
CAS 127: General Surface
CAS 143: PPK
CAS 145: General Surface
CAS 147A: PPK & Flakes
CAS 147B: Flakes & Whiteware
CAS 148: General Surface
CAS 149A: Historic & Flakes
CAS 149B: Historic Materials
CAS 149C: Large Biface
CAS 146: PPK
CAS 150: PPK
CAS 151: Historic
CAS 152: PPK
CAS 153: General Surface
CAS 153D: PPK
CAS 154: General Surface
CAS 155: General Surface
CAS 155: General Surface
CAS 155B: PPK
CAS 156: General Surface
CAS 156F: PPK
CAS 156J: PPK
CAS 156G: PPK
CAS 156: General Surface 13 April 04
CAS 157: General Surface
CAS 158: General Surface
CAS 159: General Surface
CAS 159A: PPK
CAS 160: General Surface
CAS 161: General Surface
CAS 162 A: General Surface
298
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
CAS 162B: General Surface
CAS 162C: General Surface
CAS 162C: General Surface Waypoint 4,5,6
CAS 162C: General Surface
CAS 162C7: PPK
CAS 162C8: PPK
CAS 162C10: PPK
CAS 163: General Surface, South of Branch
CAS 163: General Surface, South of Branch
CAS 163A: PPK
CAS 163B: PPK
CAS 163B: PPK
CAS 163 D: PPK
CAS 163 E: PPK
CAS 163 F: PPK
CAS 163 G: PPK
CAS 163 H: PPK
CAS 163 : General Surface 15/16 APR 2005 Bag 1
CAS 163 : General Surface 15/16 APR 2005 Bag 2
CAS 163 L: PPK
CAS 163 M: PPK
CAS 164 : PPK
CAS 165 : PPK
CAS 166 : PPK
CAS 167 : General Surface
CAS 168 : General Surface
CAS 169 : General Surface
CAS 170 : General Surface
CAS 170 C: PPK
CAS 171 : PPK
CAS 172 : General Surface
CAS 173 1: Collection Area 1
CAS 173 1A: Collection Area 1: PPK
CAS 173 1C: Collection Area 1: PPK
CAS 173 1E: Collection Area 1E: 2 PPKs 20m apart
CAS 173 2: Collection Area 2
CAS 173 3: Collection Area 3
CAS 174 : General Surface
CAS 175 : General Surface
CAS 176 : General Surface
CAS 177 : General Surface
CAS 178 : General Surface
CAS 178 B: PPK
CAS 179 : General Surface
CAS 180 : General Surface
CAS 182 A: Elmodel Sand Pit and Surround
CAS 181 : General Surface, Locke Farm
CAS 183 1: Collection Area 1
299
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
407
408
413
421
424
425
426
438
440
442
448
450
451
459
465
466
468
475
479
480
485
489
491
497
502
503
504
507
CAS 183 B: PPK
CAS 183 C: PPK
CAS 183 D: PPK
CAS 183 F: PPK
CAS 183 G: PPK
CAS 182 II: Elmodel Sand Pit Areas I and II
CAS 183 2: Collection Area 2 Bag 1
CAS 183 2: Collection Area 2 Bag 2
CAS 183 3: Collection Area 3 Bag 1
CAS 183 3: Collection Area 3 Bag 2
CAS 183 : Collection Area 4
CAS 183 L: PPK
CAS 183 M: PPK
CAS 183 N: PPK
CAS 183 O: PPK
CAS 183 Q: PPK
CAS 183 R: PPK
CAS 183 S: PPK
CAS 183 X: PPK
CAS 183 Y: PPK
CAS 184 : Collection Area 1
CAS 184 : Collection Areas 2 & 3
CAS 184 J: Collection Area J
CAS 184 4: Collection Area 4
CAS 186 : General Surface
CAS 187 : General Surface Bag 1
CAS 187 : General Surface Bag 2
CAS 190 : General Surface
CAS 191 : General Surface
CAS 192 : General Surface
CAS 197 : General Surface
CAS 199 : General Surface
CAS 199 C: Collection Area C
CAS 209 : General Surface
CAS 213 : General Surface
CAS 213 I: Collection Area I
CAS 215 : General Surface
CAS 6 HH: Collection Area HH
CAS 9 D: Collection Area D
CAS 9 A: Collection Area A/B/C
CAS 13 : General Surface
CAS 220 : General Surface
CAS 222 : General Surface
CAS 229 : General Surface
CAS 234 : Ceramic & Lithic Scatter
CAS 235 A: PPK, ceramic, & lithic
CAS 235 B: ceramic and lithic
CAS 238 : General Surface
300
508
510
512
513
516
517
520
522
523
524
525
528
529
530
532
540
542
544
545
546
547
548
550
552
556
557
558
559
561
566
568
569
570
574
577
582
583
586
589
591
595
596
597
598
601
602
605
606
CAS 240 : General Surface
CAS 225 B: Ceamics (possibly part of 9BX5)
CAS 242 AB: 9BX5, Areas A & B (JFC collections)
CAS 242 : Road NE of 9BX5 original boundaries
CAS 242, ShovelTest No: 1
CAS 242 ShovelTest No: 2
CAS 242 ShovelTest No: 5
CAS 4, Level 2 (30-40)
CAS 4, Level 3 (40-50)
CAS 4, Level 4 (50-60)
CAS 4, Level 5 (60-70)
CAS 242 : General Surface
CAS 5 : General Surface 7 JUNE 04
CAS 242 ShovelTest No: 6
CAS 242 ShovelTest No: 8
CAS 242 ShovelTest No: 16
ShovelTest No: 21
ShovelTest No: 24
ShovelTest No: 25
ShovelTest No: 26
ShovelTest No: 27
ShovelTest No: 28
ShovelTest No: 31
ShovelTest No: 34
ShovelTest No: 43
ShovelTest No: 44
ShovelTest No: 45
ShovelTest No: 46
ShovelTest No: 48
ShovelTest No: 52
ShovelTest No: 58
ShovelTest No: 59
ShovelTest No: 60
ShovelTest No: 68
ShovelTest No: 71
ShovelTest No: 78
ShovelTest No: 83
ShovelTest No: 89
ShovelTest No: 95
ShovelTest No: 99
CAS 242: ShovelTest No: 19
CAS 242, XU 1, Level 2
CAS 242, XU 1, Level 2
CAS 242, XU 1, Level 3
CAS 9DU6, Shovel Test: 13
CAS 9DU6, Shovel Test: 14
CAS 9DU6, Shovel Test: 21
CAS 9DU6, Shovel Test: 22
301
609
611
612
614
615
619
620
622
623
625
629
630
632
633
634
635
639
640
643
649
651
653
654
655
656
657
658
659
660
662
663
664
665
666
668
669
670
674
675
676
677
679
682
683
684
685
686
687
CAS 9DU6, Shovel Test: 33
CAS 9DU6, Shovel Test: 35
CAS 9DU6, Shovel Test: 39
CAS 9DU6, Shovel Test: 43
CAS 9DU6, Shovel Test: 45
CAS 9DU6, Shovel Test: 54
CAS 9DU6, Shovel Test: 56
CAS 9DU6, Shovel Test: 61
CAS 9DU6, Shovel Test: 65
CAS 9DU6, Shovel Test: 42
CAS 9DU6, Shovel Test: 69
CAS 9DU6, Shovel Test: 71
CAS 9DU6, Shovel Test: 75
CAS 9DU6, Shovel Test: 77
CAS 9DU6, Shovel Test: 80
CAS 9DU6, Shovel Test: 83
Top of floor of looters pit
Surface
CAS 9DU6, Shovel Test: 95
ShovelTest No: 105
9DU6, ShovelTest No: 107
9DU6, ShovelTest No: 109
9DU6, ShovelTest No: 111
9DU6, ShovelTest No: 113
West Side, Looter's Trench Cleaning
South Side, Looter Trench Cleaning
East Side, Looter's Trench Cleaning
Floor Cleaning, Looter's Trench
9DU6, ShovelTest No: 110
9DU6, XU 2, Level 1
9DU6, XU 2, Level 2
9DU6, XU 2, Level 3
9DU6, XU 2, Level 4
9DU6, XU 2, Level 5
9DU6, North Wall Profile Cleaning
9DU6, XU 2, East Wall Profile Cleaning
9DU6, XU 2, Floor Cleaning
9DU6, ShovelTest No: 96
CAS 245, Shovel Test: 1
CAS 245, Shovel Test: 2
CAS 245, Shovel Test: 3
CAS 245, Shovel Test: 5
CAS 245, Shovel Test: 9
CAS 245, Shovel Test: 11
CAS 245, Shovel Test: 13
CAS 245, Shovel Test: 14
CAS 245, Shovel Test: 15
CAS 245, Shovel Test: 17
302
688
689
690
691
692
693
694
695
697
698
699
700
701
703
704
705
706
708
709
710
712
715
717
718
719
720
721
722
723
726
727
728
736
737
738
740
742
744
745
746
747
748
749
750
751
754
755
756
CAS 245, Shovel Test: 19
CAS 245, Shovel Test: 22
CAS 245, Shovel Test: 23
CAS 245, Shovel Test: 25
CAS 245, Shovel Test: 27
CAS 245, Shovel Test: 29
CAS 245, Shovel Test: 33
CAS 245, Shovel Test: 24
CAS 245, Shovel Test: 34
CAS 245, Shovel Test: 36
CAS 245, Shovel Test: 38
CAS 245, Shovel Test: 39
CAS 245, Shovel Test: 41
CAS 245, Shovel Test: 43
CAS 245, Shovel Test: 40
CAS 245, Shovel Test: 51
CAS 245, Shovel Test: 52
CAS 245, Shovel Test: 56
CAS 245, Shovel Test: 61
CAS 245, Shovel Test: 62
CAS 245, Shovel Test: 46
CAS 245, Shovel Test: 49
CAS 245, Shovel Test: 58
CAS 245, Shovel Test: 59
CAS 245, Shovel Test: 64
CAS 245, Shovel Test: 65
CAS 245, Shovel Test: 66
CAS 245, Shovel Test: 68
CAS 245, Shovel Test: 69
CAS 245, Shovel Test: 77
CAS 245, Shovel Test: 79
CAS 245, Shovel Test: 80
CAS 245, Shovel Test: 91
CAS 245, Shovel Test: 94
CAS 245, Shovel Test: 95
CAS 245, Shovel Test: 101
CAS 245, Shovel Test: 97
CAS 245, Shovel Test: 99
CAS 245, Shovel Test: 104
CAS 245, Shovel Test: 105
CAS 245, Shovel Test: 106
CAS 245, Shovel Test: 108
CAS 245, Shovel Test: 109
CAS 245, ShovelTest No: 110
CAS 245, ShovelTest No: 112
CAS 245, XU 1, Level 1
CAS 245, XU 1, Level 2
CAS 245, XU 1, Level 3
303
757
758
759
760
761
762
763
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
785
786
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
CAS 245, XU 1, Level 4
ShovelTest No: 117
CAS 245, XU 1, Level 5
CAS 245, XU 1, Level 6
CAS 245, XU 1, Level 7
CAS 245, XU 1, Level 8
CAS 245, XU 1, Level
CAS 245, General Surface
CAS XU2, Level 1
CAS XU 2, Level 2
CAS 10, XU 2, Level 3
CAS 10, XU 2, Level 4
CAS 10, XU2, Level 5
CAS 10, XU 2, Level 6
CAS 10, Xu2, Level 7
Surface Collection Adjacent to Looter's pit West of
XU 5
CAS 4, XU 5, Level 1
CAS 4, XU 5, Level 2
CAS 4, XU 5, Level 3
CAS 4, XU 5, Level 4
CAS 4, XU 5, Level 5
CAS 5, XU 5, Level 6
N510/E530 - Surface Collection Near Swamp Edge
CAS 68 ShovelTest No: 1
Bulldozed Mound Area
only collection
Chickasawhatchee Knoll, Collection 1, SGC
WGC Collection, General Surface, 28 March 1976
WGC Collection, Eastern Edge Recon
WGC Collection, Unit # 1
WGC Collection, Unit # 2
WGC Collection, Unit # Unknown
WGC Collection, Unit # 3
WGC Collection, Unit # 5
WGC Collection, Unit # 6
WGC Collection, Unit # 7
WGC Collection, Unit # 8
WGC Collection, Unit # 10
WGC Collection, Unit # 11
WGC Collection, Unit # 13
WGC Collection, Unit # 14
WGC Collection, Unit # 15
WGC Collection, Unit # 12
WGC Collection, Unit # 16
9Cal10, Only Collection
9Cal14, Only Collection
9Cal12, Only Collection
304
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1039
1040
1041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
9Cal18, Only Collection
9Cal5, General Surface
9Cal5, Lone Oak SE 100m
9Cal5, Beaver Lake
9Cal2, General Surface
9Cal3, General Surface
9Cal4, General Surface
9Cal7, General Surface
9Cal7, Mid - Mt. Pleasant Cemetary - Boat Landing
9Cal8, General Surface
9Cal9, General Surface
9Cal11, General Surface
9Cal13, General Surface
9Cal17, General Surface
9Cal15, Historic
9Cal15, South of Cedars -150 Meters
9Cal20, General Surface
9DU6, Surface
9DU6, test 1 6-12
9DU6, test 1 0-6
9DU6, test 2, 0-6
9DU6, test 2, 6-12
unknown 9DU6, test 6-12
9DU6, test 3, 0-6
9DU6, test 3, 6-12
9DU6, test 3, 12-18
9DU6, test 4, 0-6
9DU6, test 4, 6-12
9DU6, test 5, 0-6
9DU6, eu 6, 0-6
9DU6, test 7, 6-12
9DU6, exu 10, 6-12
9DU6, exu 10, 0-6
9DU6, exu 10, 12-18
9DU6, exu 10, 18-24
9DU6, exu 10, 24-30
9DU6, exu 10, 30-40
ex 11, 0-6
ex 11, 6-12
9DU6, e8, 0-6
9DU6, fx 8, 6-12
9DU6, fxu 8, 12-18
9DU6, exu 9 0-6
9DU6, eux 9, 6 - 12
9DU6, eux 9, 12-18
305
APPENDIX C:
CERAMIC CLASSIFICATION DATA, ACCORDING TO THE LOT NUMBERS
IN APPENDIX B.
306
Lot
Temper
1 Grit
1 Grit
3 Grit
3 Grit
3 Grit
4 Grit
4 Grit
5 Grit
7 Grit
7 Grit
8 Grit
8 Grit
9 Grit
10 Grit
11 Grit
12 Grit
13 Grit
16 Grit
17 Grit
17 Grit
23 Grit
24 Grit
24 Grit
25 Grit
25 Grit
26 Grit
28 Grit
28 Grit
29 Grit
29 Grit
30 Grit
33 Grit
33 Grit
33 Grit
33 Grit
33 Grit
34 Grit
35 Grit
35 Grit
36 Grit
39 Grit
39 Grit
Type
UID Plain
UID Plain
UID Plain
UID Complicated Stamped
UID Plain
UID Complicated Stamped
UID Plain
UID Incised
UID Plain
UID Plain
UID Incised
UID Decorated
UID Plain
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
Lake Jackson Decorated
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Lamar Complicated Stamped
Point Washington Incised
Point Washington Incised
UID Plain
Lamar Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Form
Part
Count
Jar
Rim
1
Unknown
Unknown
1
Unknown
Unknown
2
Jar
Rim
1
Unknown
Unknown
2
Unknown
Unknown
1
Unknown
Unknown
1
Unknown
Unknown
6
Unknown
Unknown
1
Unknown
Unknown
5
Unknown
Unknown
2
Unknown
Unknown
2
Unknown
Unknown
1
Unknown
Unknown
1
Unknown
Unknown
1
Unknown
Unknown
1
Unknown
Unknown
2
Unknown
Unknown
1
Unknown
Unknown
1
Collared Jar
Rim
1
Unknown
Unknown
3
Unknown
Unknown
5
Unknown
Unknown
1
Unknown
Other
1
Unknown
Unknown
2
Unknown
Unknown
2
Jar
Other
1
Unknown
Unknown
2
Unknown
Unknown
1
Unknown
Unknown
3
Unknown
Unknown
6
Unknown
Unknown
15
Unknown
Unknown
1
Undifferentiated Bowl Rim
1
Unknown
Unknown
1
Unknown
Unknown
2
Jar
Unknown
1
Unknown
Unknown
9
Unknown
Unknown
1
Unknown
Rim
1
Unknown
Unknown
1
Unknown
Unknown
1
307
40 Grit
41 Grit
43 Grit
44 Grit
44 Sand
44 Grit
44 Grit
45 Grit
45 Grit
48 Grit
48 Grit
48 Grit
49 Grit
49 Grit
50 Grit
53 Grit
55 Grit
56 Grit
56 Grit
57 Grit
57 Grit
57 Grit
59 Grit
60 Grit
61 Grit
62 Grit
64 Grit
66 Grit
66 Grit
66
66 Grit
67 Grit
67 Grit
68 Grit
69 Grit
70 Grit
70 Grit
70 Grit
70 Grit
70 Grit
70 Grit
71 Grit
72 Grit
73 Grit
74 Grit
UID Plain
UID Plain
Fort Walton Incised
UID Plain
UID Plain
Lamar Complicated Stamped
UID Plain
UID Plain
Lamar Complicated Stamped
Wakulla Check Stamped
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
UID Stamped
UID Plain
Swift Creek Complicated Stamped
UID Stamped
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
Other
UID Plain
UID Incised
Other
Lake Jackson Plain
UID Plain
UID Complicated Stamped
UID Plain
UID Plain
Lake Jackson Decorated
Other
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Unknown
Unknown
Outleaned Wall Bowl
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Unknown
Unknown
Unknown
Unknown
Collared Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Base
Unknown
Unknown
Unknown
Unknown
Unknown
Base
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
2
2
1
13
1
3
1
1
1
3
2
1
1
1
10
1
1
1
1
1
1
1
1
2
1
1
1
18
1
3
1
3
1
15
1
1
1
1
1
12
1
4
5
7
5
308
74 Grit
75 Grit
76 Grit
77 Grit
78 Grit
79 Grit
79 Grit
80 Grit
82 Grit
84 Grit
84 Grit
84 Grit
86
86 Grit
86 Grit
86 Grit
87 Grit
87 Grit
88 Grit
88 Grit
90 Grit
90 Grit
93 Grit
94 Grit
96 Grit
97 Grit
98 Grit
98 Grit
99 Grit
99 Grit
99 Grit
101 Grit
103 Grit
103 Grit
104 Grit
104 Grit
106 Grit
108 Grit
110 Grit
110 Grit
110 Grit
111 Grit
111 Grit
111 Grit
111 Grit
UID Incised
UID Plain
UID Incised
UID Plain
UID Plain
Lake Jackson Decorated
UID Plain
UID Plain
UID Decorated
UID Plain
UID Plain
Ingram Plain
Other
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Decorated
Lake Jackson Decorated
UID Plain
UID Plain
Lake Jackson Decorated
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Andrews Decorated
UID Stamped
Ingram Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Undifferentiated Bowl Rim
Unknown
Unknown
Outleaned Wall Bowl Rim
Unknown
Unknown
Unknown
Unknown
Pipe
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Shoulder
Unknown
Unknown
Unknown
Unknown
Collared Jar
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Undifferentiated Bowl Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Bottle
Rim
Unknown
Unknown
Outleaned Wall Bowl Rim
Unknown
Unknown
Unknown
Unknown
1
10
1
1
1
1
2
1
1
1
7
1
1
1
1
6
3
1
1
1
4
1
8
8
1
1
10
1
1
1
12
1
1
3
1
4
1
2
13
3
1
1
2
4
2
309
111 Grit
112 Grit
113 Grit
113 Grit
114 Grit
115 Grit
115 Grit
116 Grit
118 Grit
118 Grit
118 Sherds
118 Grit
119 Grit
119 Grit
119 Grit
120 Grit
121 Grit
121 Grit
121 Grit
121 Grit
122 Grit
122 Grit
123 Grit
123 Grit
125 Grit
127 Grit
133 Grit
142 Grit
142 Grit
142 Grit
143 Grit
143 Grit
144 Grit
144 Grit
145 Grit
145 Grit
145 Grit
145 Grit
145 Grit
146 Grit
147 Grit
148 Grit
148 Grit
149 Grit
149 Grit
UID Plain
UID Plain
UID Plain
UID Stamped
UID Plain
UID Plain
UID Plain
UID Plain
Lake Jackson Plain
UID Plain
Other
UID Stamped
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
Other
UID Plain
UID Plain
UID Plain
UID Stamped
UID Plain
UID Plain
UID Plain
Other
UID Plain
UID Incised
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
UID Stamped
Wakulla Check Stamped
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
UID Complicated Stamped
UID Plain
UID Incised
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Shoulder
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Undifferentiated Bowl Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
9
2
15
1
1
2
2
6
1
7
3
1
1
2
4
6
1
1
5
1
1
3
1
1
1
1
2
1
15
1
5
1
17
1
39
1
1
2
2
2
1
6
1
21
1
310
149 Grit
149 Grit
150 Grit
150 Grit
150 Grit
152 Grit
155 Grit
155 Grit
157 Grit
159 Grit
161 Grit
163 Grit
166 Grit
167 Grit
171 Grit
172 Grit
173 Sand
174 Grit
178 Grit
178 Sand
180 Grit
182 Grit
184 Grit
186 Grit
188 Grit
189 Grit
192 Grit
193 Grit
193 Grit
193 Grit
193 Grit
193 Grit
193 Grit
201 Grit
202 Grit
203 Grit
203 Grit
208 Grit
210 Grit
212 Grit
216 Grit
216 Grit
217 Grit
222 Grit
223 Grit
UID Stamped
UID Plain
UID Plain
UID Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Complicated Stamped
Point Washington Incised
Point Washington Incised
UID Plain
UID Complicated Stamped
UID Incised
UID Plain
UID Plain
Other
UID Plain
Other
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Body
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
1
1
1
1
1
2
11
1
1
2
1
1
1
4
1
2
1
1
6
1
1
3
2
2
3
1
1
5
2
1
6
1
1
1
1
1
1
1
1
2
2
1
2
1
1
311
225 Grit
225 Grit
225 Grit
226 Sand
226 Sand
228 Grit
228 Grit
229 Grit
229 Grit
229 Grit
229 Grit
230 Grit
230 Grit
231 Grit
231 Grit
232 Grit
232 Grit
233 Grit
235 Grit
235 Grit
239 Grit
242 Grit
243 Grit
243 Grit
247 Grit
247 Grit
247 Grit
249 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
256 Grit
Swift Creek Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Incised
Other
UID Plain
UID Incised
UID Plain
Lake Jackson Decorated
UID Plain
UID Plain
UID Plain
UID Plain
UID Stamped
UID Plain
UID Plain
Lamar Complicated Stamped
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Stamped
UID Incised
UID Incised
UID Incised
UID Plain
UID Decorated
UID Decorated
UID Decorated
Lake Jackson Decorated
UID Plain
UID Incised
Other
UID Complicated Stamped
UID Complicated Stamped
Swift Creek Complicated Stamped
Unknown
Unknown
Unknown
Disk
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Base
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Base
Unknown
Unknown
Unknown
Shoulder
Unknown
Unknown
Unknown
Base
Unknown
Unknown
Unknown
2
5
1
1
3
1
1
1
1
1
11
1
1
5
1
1
1
2
1
1
1
1
1
4
7
1
1
1
1
7
2
1
1
5
2
1
1
1
1
81
1
1
2
7
2
312
256 Grit
256 Grit
256 Grit
256 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
257 Grit
262 Grit
262 Grit
262 Grit
262 Grit
263 Grit
263 Grit
263 Grit
263 Grit
263 Grit
263 Grit
263 Grit
263 Grit
263 Grit
265 Sand
265 Grit
265 Grit
265 Grit
265 Grit
269 Grit
269 Grit
269 Grit
270 Grit
271 Grit
271 Grit
271 Grit
271 Grit
272 Grit
272 Grit
272 Grit
272 Grit
UID Plain
UID Plain
UID Incised
UID Complicated Stamped
Other
UID Incised
UID Decorated
UID Plain
UID Complicated Stamped
UID Decorated
Carrabelle Punctated
Wakulla Check Stamped
UID Plain
Swift Creek Complicated Stamped
UID Plain
UID Plain
UID Decorated
UID Complicated Stamped
Swift Creek Complicated Stamped
UID Plain
UID Plain
UID Decorated
UID Plain
UID Complicated Stamped
UID Incised
UID Plain
UID Decorated
UID Plain
UID Plain
UID Plain
UID Plain
Swift Creek Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Complicated Stamped
Swift Creek Complicated Stamped
UID Plain
UID Complicated Stamped
Carrabelle Punctated
Jar
Rim
Unknown
Rim
Jar
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Undifferentiated Bowl Rim
Unknown
Rim
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Pipe
Jar
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
1
1
2
1
6
1
1
1
3
1
2
1
1
47
1
16
2
1
1
1
1
1
1
4
32
1
6
1
1
6
2
14
2
5
1
1
2
3
1
5
1
24
1
1
1
313
272 Grit
272 Grit
272 Grit
272 Grit
274 Grit
284 Grit
284 Grit
287 Grit
287 Grit
289 Grit
290 Sand
290 Grit
291 Grit
304 Grit
311 Grit
318 Grit
322 Grit
324 Grit
328 Grit
328 Grit
328 Grit
329 Grit
329 Grit
333 Grit
333 Grit
334 Grit
334 Grit
334 Grit
334 Grit
336 Grit
336 Grit
336 Grit
336 Grit
336 Grit
338 Grit
339 Grit
339 Grit
339 Grit
339 Grit
339 Grit
339 Grit
339 Grit
339 Grit
339 Grit
339 Grit
Carrabelle Punctated
UID Plain
UID Plain
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Decorated
Carrabelle Punctated
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Fort Walton Incised
UID Incised
UID Plain
UID Decorated
UID Plain
UID Stamped
UID Plain
UID Incised
Fort Walton Incised
UID Plain
UID Plain
UID Plain
Lake Jackson Decorated
UID Complicated Stamped
UID Plain
UID Plain
Lake Jackson Decorated
Lake Jackson Decorated
Lake Jackson Decorated
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Pipe
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Disk
Unknown
Unknown
Unknown
Unknown
Unknown
Plate
Unknown
Unknown
Unknown
Unknown
Unknown
Outleaned Wall Bowl
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Rim
Unknown
Rim
Rim
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
1
1
1
1
16
1
1
1
1
10
1
17
11
1
1
1
3
1
1
9
1
1
1
1
1
1
1
1
3
1
1
1
1
15
2
1
2
2
3
1
1
1
1
1
109
314
339 Grit
339 Grit
339 Grit
339 Grit
340 Grit
340 Grit
340 Grit
340 Grit
340 Grit
340 Grit
340 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
341 Grit
345 Grit
345 Grit
345 Grit
351 Grit
351 Grit
355 Grit
355 Grit
356 Grit
356 Grit
356 Grit
356 Grit
357 Grit
362 Grit
364 Grit
365 Grit
368 Grit
369 Grit
373 Grit
373 Grit
UID Incised
Ingram Plain
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
Lake Jackson Decorated
Fort Walton Incised
UID Plain
Lake Jackson Decorated
Ingram Plain
UID Plain
Fort Walton Incised
Lamar Complicated Stamped
UID Decorated
Lake Jackson Decorated
UID Plain
UID Complicated Stamped
UID Plain
UID Plain
UID Decorated
UID Plain
UID Decorated
UID Decorated
UID Plain
UID Incised
UID Plain
Point Washington Incised
UID Plain
UID Plain
UID Incised
UID Plain
UID Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Outleaned Wall Bowl Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Base
Unknown
Unknown
Jar
Shoulder
Undifferentiated Bowl Rim
Jar
Rim
Collared Jar
Rim
Outleaned Wall Bowl Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
1
3
1
1
1
1
2
1
1
28
1
1
1
1
2
2
3
3
1
1
1
1
2
1
133
2
3
1
1
1
1
10
1
1
1
1
9
1
3
2
1
8
1
7
1
315
374 Grit
374 Grit
374 Grit
374 Grit
375 Grit
375 Grit
375 Grit
376 Grit
377 Grit
378 Grit
381 Grit
383 Grit
383 Grit
384 Grit
384 Grit
384 Grit
385 Grit
385 Grit
385 Grit
385 Grit
385 Grit
391 Grit
391 Grit
391 Grit
393 Grit
394 Grit
395 Grit
395 Grit
396 Grit
407 Grit
407 Grit
407 Grit
407 Grit
407 Grit
408 Grit
408 Grit
408 Grit
408 Grit
408 Grit
408 Grit
408 Grit
408 Grit
408 Grit
413 Grit
421 Grit
UID Plain
UID Decorated
UID Decorated
UID Plain
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
Simple Stamped
UID Decorated
UID Incised
Lake Jackson Plain
UID Plain
UID Plain
UID Incised
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
UID Incised
UID Plain
Ingram Plain
UID Plain
UID Plain
UID Plain
UID Stamped
UID Incised
UID Plain
UID Plain
UID Decorated
UID Incised
UID Plain
UID Incised
Lake Jackson Decorated
Lake Jackson Decorated
UID Plain
UID Incised
UID Plain
UID Incised
Disk
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Plate
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Unknown
Unknown
Unknown
Rim
Rim
Other
Unknown
Rim
Unknown
Unknown
Rim
Rim
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Rim
Rim
Other
Other
Unknown
Unknown
Unknown
Rim
Unknown
1
1
1
10
1
1
1
1
1
1
1
1
1
1
1
27
14
1
1
1
1
5
1
1
2
1
2
8
8
1
1
1
13
2
1
1
1
1
1
1
2
19
1
1
1
316
421 Grit
421 Grit
421 Grit
421 Grit
421 Grit
421 Grit
421 Grit
421 Grit
424 Grit
425 Grit
425 Grit
426 Grit
426 Grit
438 Grit
440 Grit
440 Grit
442 Grit
448 Grit
450 Grit
451 Grit
459 Grit
465 Grit
465 Grit
465 Grit
466 Grit
466 Grit
466 Grit
468 Grit
475 Grit
479 Grit
479 Grit
479 Grit
480 Grit
485 Grit
489 Grit
491 Grit
491 Grit
497 Grit
502 Grit
503 Grit
504 Grit
504 Grit
504 Grit
504 Grit
507 Grit
UID Complicated Stamped
UID Incised
UID Plain
UID Plain
Fort Walton Incised
Lamar Bold Incised
UID Plain
UID Plain
UID Incised
UID Plain
UID Incised
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Complicated Stamped
UID Plain
Lamar Complicated Stamped
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Decorated
Wakulla Check Stamped
Fort Walton Incised
UID Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Carinated Bowl
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Base
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
1
1
1
91
3
2
1
1
3
11
1
23
1
4
1
6
2
1
2
2
1
7
2
1
1
1
5
2
1
2
1
2
1
1
1
1
1
1
2
1
1
1
1
10
1
317
508 Grit
510 Grit
512 Grit
513 Grit
513 Grit
516 Grit
520 Grit
522 Grit
523 Grit
523 Grit
524 Grit
525 Grit
528 Grit
529 Grit
529 Grit
529 Grit
529 Grit
529 Grit
530 Grit
532 Grit
540 Grit
542 Grit
542 Grit
544 Grit
545 Grit
545 Grit
545 Grit
546 Grit
547 Grit
548 Grit
550 Grit
550 Grit
552 Grit
556 Grit
557 Grit
558 Grit
559 Grit
561 Grit
566 Grit
568 Grit
568 Grit
569 Grit
570 Grit
574 Grit
577 Grit
UID Plain
UID Incised
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
UID Stamped
UID Plain
UID Plain
UID Incised
UID Plain
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Stamped
UID Plain
Wakulla Check Stamped
Other
UID Plain
UID Plain
UID Decorated
UID Plain
UID Stamped
UID Decorated
UID Plain
UID Plain
UID Decorated
UID Plain
UID Plain
UID Plain
UID Incised
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Complicated Stamped
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Rim
Unknown
Unknown
Shoulder
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
2
1
2
1
2
1
1
3
1
1
2
1
1
1
2
1
1
10
2
2
1
2
1
2
1
1
4
2
1
2
2
1
1
1
1
1
1
1
2
1
1
2
1
1
1
318
577 Grit
577 Grit
577 Grit
577 Grit
582 Grit
582 Grit
583 Grit
586 Grit
589 Grit
591 Grit
595 Grit
596 Grit
597 Grit
597 Grit
597 Grit
597 Grit
597 Grit
598 Grit
598 Grit
598 Grit
601 Grit
601 Grit
602 Grit
605 Grit
605 Grit
605 Grit
606 Grit
609 Grit
611 Grit
612 Grit
612 Grit
614 Grit
615 Grit
619 Grit
619 Grit
620 Grit
622 Grit
623 Grit
625 Grit
629 Grit
629 Grit
629 Grit
629 Grit
630 Grit
632 Grit
Lamar Complicated Stamped
UID Plain
Chattahoochee Brushed
Other
Wakulla Check Stamped
UID Plain
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
Swift Creek Complicated Stamped
Lamar Complicated Stamped
UID Complicated Stamped
UID Decorated
UID Plain
UID Plain
Other
UID Complicated Stamped
UID Plain
UID Plain
Wakulla Check Stamped
UID Complicated Stamped
Deptford Linear Check Stamped
Wakulla Check Stamped
Deptford Linear Check Stamped
UID Stamped
UID Stamped
UID Plain
UID Decorated
UID Plain
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Outleaned Wall Bowl
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Unknown
1
2
1
1
1
2
2
1
1
1
2
18
1
1
2
1
6
2
1
1
1
1
5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
319
633 Grit
634 Grit
634 Grit
635 Grit
635 Grit
639 Grit
639 Grit
640 Grit
640 Grit
640 Grit
640 Grit
643 Grit
649 Grit
651 Grit
651 Grit
653 Grit
654 Grit
655 Grit
656 Grit
656 Grit
657 Grit
657 Grit
658 Sand
658 Grit
658 Sand
659 Grit
659 Grit
659 Grit
660 Grit
662 Sand
662 Grit
662 Sand
662 Sand
662 Grit
662 Grit
662 Grit
663 Grit
663 Grit
663 Grit
663 Grit
663 Grit
663 Grit
663 Grit
664 Grit
664 Grit
Wakulla Check Stamped
UID Stamped
UID Decorated
UID Decorated
Wakulla Check Stamped
UID Plain
UID Decorated
Wakulla Check Stamped
Swift Creek Complicated Stamped
Other
UID Plain
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
UID Decorated
UID Plain
UID Plain
UID Plain
UID Decorated
UID Decorated
UID Plain
UID Decorated
Wakulla Check Stamped
Wakulla Check Stamped
Alachua Cob Marked
Wakulla Check Stamped
UID Decorated
UID Stamped
UID Stamped
Wakulla Check Stamped
UID Decorated
UID Plain
Unknown
Unknown
Jar
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Rim
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Undifferentiated Bowl Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
2
1
1
1
1
5
2
2
1
2
3
3
1
1
7
3
3
2
3
1
3
320
664 Grit
664 Sand
664 Grit
664 Sand
664 Grit
664 Sand
664 Grit
664 Grit
664 Sand
665 Grit
665 Grit
665 Grit
665 Grit
665 Grit
665 Sand
665 Grit
665 Grit
666 Grit
666 Grit
666 Grit
666 Grit
666 Grit
668 Grit
669 Grit
670 Grit
674 Grit
675 Grit
675 Grit
675 Grit
676 Grit
677 Grit
679 Grit
682 Grit
683 Grit
684 Grit
685 Grit
686 Grit
687 Sand
688 Grit
689 Grit
690 Grit
691 Grit
691 Grit
691 Sand
692 Grit
Wakulla Check Stamped
Wakulla Check Stamped
Swift Creek Complicated Stamped
Deptford Linear Check Stamped
UID Stamped
UID Plain
UID Stamped
Deptford Linear Check Stamped
Wakulla Check Stamped
Wakulla Check Stamped
Deptford Linear Check Stamped
Wakulla Check Stamped
UID Stamped
UID Plain
Wakulla Check Stamped
Deptford Linear Check Stamped
UID Plain
UID Plain
Deptford Simple Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Ingram Plain
UID Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Outleaned Wall Bowl
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Base
Unknown
Unknown
Unknown
Unknown
Shoulder
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
1
7
1
1
1
4
1
1
4
1
1
1
1
1
1
1
3
1
1
1
1
1
1
1
2
2
1
10
1
1
3
2
2
2
2
2
2
1
1
5
1
3
1
2
2
321
693 Grit
694 Grit
695 Grit
697 Grit
698 Grit
699 Grit
699 Grit
700 Grit
701 Grit
703 Grit
704 Grit
705 Grit
706 Grit
708 Grit
709 Grit
710 Grit
712 Grit
715 Grit
717 Grit
718 Grit
719 Grit
720 Grit
721 Grit
722 Grit
723 Grit
723 Grit
726 Grit
726 Grit
727 Grit
728 Grit
728 Grit
736 Grit
737 Grit
737 Grit
737 Grit
738 Grit
740 Grit
742 Grit
742 Grit
744 Grit
744 Grit
745 Grit
746 Grit
747 Grit
748 Grit
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Incised
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Pipe
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Other
Unknown
Other
Unknown
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
4
1
2
2
1
2
1
1
3
1
3
1
2
1
2
2
1
1
6
3
1
4
1
1
6
2
3
1
4
12
1
1
1
3
1
1
1
2
6
1
3
8
5
2
4
322
748 Grit
749 Grit
749 Grit
749 Grit
750 Grit
751 Grit
754 Grit
754 Grit
754 Grit
754 Grit
754 Grit
754 Grit
754 Grit
754 Grit
755 Grit
755 Grit
755 Other
755 Grit
755 Grit
755 Grit
755 Grit
755 Grit
755 Grit
755 Grit
755 Grit
755 Sherds
755 Grit
756 Grit
756 Other
756 Grit
756 Grit
756 Grit
756 Grit
756 Grit
756 Grit
756 Grit
756 Grit
756 Grit
756 Grit
757 Grit
757 Grit
757 Grit
757 Grit
757 Grit
757 Grit
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Incised
UID Plain
UID Plain
UID Decorated
Ingram Plain
Lake Jackson Plain
Fort Walton Incised
Cool Branch Incised
Other
UID Plain
Other
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
UID Stamped
Other
Other
UID Plain
UID Plain
UID Plain
UID Plain
UID Plain
Carrabelle Punctated
Cool Branch Incised
UID Incised
UID Incised
UID Decorated
UID Plain
UID Incised
Carrabelle Punctated
UID Stamped
Wakulla Check Stamped
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Pipe
Unknown
Rim
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Undifferentiated Bowl Rim
Collared Jar
Rim
Plate
Rim
Collared Jar
Shoulder
Unknown
Unknown
Pipe
Unknown
Unknown
Collared Jar
Shoulder
Unknown
Rim
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Rim
Unknown
Unknown
Unknown
Neck
Jar
Rim
Unknown
Unknown
Collared Jar
Shoulder
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Pipe
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
1
1
1
3
1
1
13
1
2
1
46
2
1
1
1
1
1
1
1
1
4
2
1
119
1
1
2
1
1
2
1
185
1
1
5
1
6
1
5
1
2
2
1
1
3
323
757 Grit
757 Grit
757 Grit
758 Grit
758 Grit
759 Grit
759 Grit
759 Grit
760 Grit
760 Grit
761 Grit
761 Other
761 Grit
767 Grit
768 Grit
768 Grit
768 Grit
768 Grit
768 Grit
768 Grit
769 Grit
769 Grit
769 Grit
769 Grit
769 Grit
769 Grit
769 Grit
770 Grit
770 Grit
770 Grit
771 Grit
771 Grit
771 Grit
771 Grit
772 Grit
772 Grit
772 Grit
773 Grit
773 Grit
773 Grit
773 Grit
775 Grit
776 Grit
1001
1001
UID Plain
UID Plain
UID Plain
UID Plain
Ingram Plain
UID Plain
Other
UID Plain
Other
UID Plain
UID Plain
Other
UID Plain
UID Plain
Ingram Plain
UID Plain
UID Plain
UID Plain
UID Plain
Lake Jackson Plain
UID Incised
UID Plain
UID Complicated Stamped
UID Plain
UID Plain
UID Plain
Lake Jackson Plain
UID Plain
Lake Jackson Plain
Ingram Plain
UID Plain
UID Plain
UID Plain
Other
Other
UID Plain
UID Plain
Lake Jackson Incised
UID Plain
UID Plain
UID Plain
Lamar Complicated Stamped
Unknown
Unknown
Jar
Unknown
Outleaned Wall Bowl
Unknown
Unknown
Unknown
Unknown
Unknown
Disk
Unknown
Unknown
Unknown
Unknown
Outleaned Wall Bowl
Unknown
Unknown
Unknown
Jar
Collared Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Jar
Collared Jar
Outleaned Wall Bowl
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Collared Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Shoulder
Unknown
Rim
Unknown
Unknown
Base
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Rim
Rim
Rim
Rim
Rim
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Rim
Rim
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Rim
Unknown
Unknown
Other
Rim
135
5
1
1
1
37
1
1
1
17
1
4
2
1
2
1
38
1
1
1
1
1
1
2
2
62
6
24
1
1
1
1
3
9
12
2
1
4
1
1
1
2
2
1
1
324
1001
1001
1001
1001
1001
1001
1001
1001
1001
1001
1001
1001
1001
1001
1002
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
1003
Wakulla Check Stamped
Unknown
Rim
UID Decorated
Unknown
Body
UID Plain
Unknown
Rim
UID Plain
Unknown
Shoulder
UID Stamped
Unknown
Body
UID Plain
Unknown
Body
Fort Walton Incised
Unknown
Rim
UID Incised
Unknown
Body
UID Incised
Unknown
Rim
Point Washington Incised
Unknown
Body
Carrabelle Punctated
Unknown
Rim
Fort Walton Incised
Unknown
Body
Fort Walton Incised
Unknown
Shoulder
UID Plain
Unknown
Other
UID Plain
Unknown
Body
Wakulla Check Stamped
Unknown
Rim
Other
Unknown
Body
UID Plain
Unknown
Unknown
UID Plain
Jar
Shoulder
UID Plain
Unknown
Other
Other
Unknown
Rim
Other
Unknown
Body
Other
Unknown
Body
Other
Unknown
Body
Lake Jackson Decorated
Jar
Rim
Wakulla Check Stamped
Unknown
Body
UID Incised
Undifferentiated Bowl Rim
UID Stamped
Unknown
Unknown
Alachua Cob Marked
Unknown
Body
Deptford Simple Stamped
Unknown
Body
Deptford Linear Check Stamped
Unknown
Body
Lamar Complicated Stamped
Unknown
Body
UID Fiber Tempered
Unknown
Body
Leon Check Stamped
Unknown
Body
Leon Jefferson Complicated Stamped Unknown
Body
Deptford Simple Stamped
Unknown
Rim
Keith Incised
Unknown
Unknown
Dunlap Fabric Marked
Unknown
Body
UID Incised
Unknown
Body
UID Stamped
Unknown
Body
Lake Jackson Decorated
Unknown
Shoulder
UID Plain
Unknown
Body
UID Plain
Unknown
Body
UID Stamped
Unknown
Body
Fort Walton Incised
Undifferentiated Bowl Rim
1
1
15
5
2
4
2
1
1
1
1
1
2
1
33
14
4
9
11
1
1
1
1
1
9
198
5
12
1
1
1
5
1
1
1
1
1
1
2
4
4
404
221
1
5
325
1003
1003
1003
1003
1003
1003
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1004
1005
1005
1005
1005
1005
1005
1005
1005
1005
1005
Lake Jackson Decorated
Cool Branch Incised
UID Plain
Fort Walton Incised
Fort Walton Incised
UID Incised
Lake Jackson Decorated
UID Decorated
UID Decorated
UID Plain
Lake Jackson Decorated
Fort Walton Incised
UID Plain
UID Plain
Wakulla Check Stamped
Lamar Complicated Stamped
Dunlap Fabric Marked
Lake Jackson Decorated
UID Plain
UID Stamped
Point Washington Incised
UID Decorated
UID Plain
UID Plain
Fort Walton Incised
UID Decorated
UID Stamped
UID Plain
UID Stamped
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
Swift Creek Complicated Stamped
UID Plain
Wakulla Check Stamped
UID Plain
Other
UID Decorated
Lake Jackson Decorated
Wakulla Check Stamped
UID Complicated Stamped
UID Decorated
UID Incised
UID Plain
Undifferentiated Bowl Rim
Jar
Rim
Unknown
Rim
Unknown
Unknown
Plate
Shoulder
Unknown
Body
Unknown
Rim
Unknown
Body
Unknown
Shoulder
Unknown
Rim
Unknown
Other
Unknown
Shoulder
Unknown
Other
Unknown
Body
Unknown
Body
Unknown
Body
Undifferentiated Bowl Rim
Jar
Rim
Undifferentiated Bowl Rim
Unknown
Body
Unknown
Body
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Body
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Unknown
Body
Unknown
Base
Unknown
Body
Jar
Shoulder
Jar
Other
Unknown
Body
Unknown
Unknown
Jar
Rim
Unknown
Body
Unknown
Unknown
Unknown
Body
Undifferentiated Bowl Rim
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Body
Jar
Shoulder
1
1
57
10
1
6
3
1
1
1
1
1
1
36
5
2
1
1
3
1
1
1
21
1
2
2
1
1
8
1
8
1
1
2
2
1
58
1
1
2
2
2
2
1
1
326
1005
1005
1005
1005
1005
1006
1006
1006
1006
1006
1006
1006
1006
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1007
1008
1008
1008
1008
1009
1009
1009
1009
1009
1009
1009
1009
1009
1010
1010
Fort Walton Incised
Point Washington Incised
Fort Walton Incised
UID Plain
UID Decorated
UID Plain
Lake Jackson Decorated
UID Plain
UID Stamped
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
Other
Wakulla Check Stamped
Other
Other
Other
Lake Jackson Decorated
Leon Check Stamped
UID Plain
UID Plain
UID Decorated
UID Complicated Stamped
Other
Dunlap Fabric Marked
UID Plain
UID Decorated
Other
Fort Walton Incised
Wakulla Check Stamped
Wakulla Check Stamped
UID Decorated
Other
UID Plain
Fort Walton Incised
Wakulla Check Stamped
UID Plain
UID Complicated Stamped
Other
Other
UID Decorated
UID Stamped
UID Plain
Other
Lake Jackson Decorated
Carinated Bowl
Shoulder
Unknown
Body
Jar
Rim
Unknown
Rim
Unknown
Body
Unknown
Other
Jar
Rim
Unknown
Rim
Unknown
Body
Unknown
Rim
Unknown
Body
Unknown
Body
Unknown
Rim
Undifferentiated Bowl Rim
Unknown
Body
Undifferentiated Bowl Rim
Unknown
Body
Collared Jar
Rim
Undifferentiated Bowl Rim
Unknown
Body
Unknown
Other
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Rim
Unknown
Rim
Unknown
Rim
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Unknown
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Rim
Unknown
Body
Unknown
Shoulder
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Body
Unknown
Rim
3
1
1
2
2
2
1
4
7
1
4
39
1
1
1
1
1
2
1
58
1
24
2
1
1
4
1
1
2
12
2
4
10
31
2
8
2
2
1
1
3
24
50
4
3
327
1010
1010
1010
1010
1010
1010
1010
1010
1011
1011
1011
1011
1011
1011
1011
1011
1011
1011
1012
1012
1012
1013
1013
1013
1013
1013
1013
1013
1013
1013
1013
1013
1013
1014
1014
1014
1014
1014
1014
1014
1014
1014
1014
1014
1014
Lake Jackson Incised
UID Plain
UID Stamped
UID Decorated
Lake Jackson Decorated
Other
UID Plain
Other
Lake Jackson Decorated
UID Plain
UID Incised
UID Plain
Dunlap Fabric Marked
UID Decorated
UID Plain
Fort Walton Incised
UID Plain
Wakulla Check Stamped
Lake Jackson Decorated
Wakulla Check Stamped
UID Plain
UID Plain
Lamar Complicated Stamped
UID Incised
UID Incised
Other
UID Stamped
UID Decorated
UID Decorated
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
Carrabelle Punctated
Wakulla Check Stamped
UID Plain
UID Plain
UID Decorated
UID Stamped
Wakulla Check Stamped
UID Complicated Stamped
Other
Dunlap Fabric Marked
Lake Jackson Decorated
UID Plain
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Rim
Body
Rim
Other
Body
Body
Body
Rim
Shoulder
Body
Rim
Body
Body
Rim
Body
Body
Body
Rim
Body
Body
Rim
Body
Rim
Body
Body
Body
Rim
Body
Rim
Body
Body
Body
Rim
Rim
Body
Body
Body
Body
Body
Body
Body
Rim
Other
Shoulder
1
1
1
1
1
1
64
1
6
1
1
3
1
6
1
1
45
4
2
3
6
2
1
1
2
4
4
1
18
1
29
80
1
1
8
123
17
20
14
1
6
2
3
2
1
328
1015
1015
1015
1015
1015
1015
1015
1015
1015
1015
1015
1015
1015
1015
1016
1016
1016
1016
1016
1016
1017
1017
1017
1018
1018
1018
1019
1019
1019
1019
1019
1019
1019
1019
1019
1019
1020
1020
1020
1020
1020
1020
1020
1020
1020
UID Plain
Wakulla Check Stamped
UID Incised
UID Decorated
UID Complicated Stamped
Other
UID Stamped
UID Plain
Wakulla Check Stamped
UID Plain
Fort Walton Incised
Point Washington Incised
Cool Branch Incised
Other
Other
UID Plain
Other
UID Plain
UID Complicated Stamped
Lake Jackson Incised
UID Plain
UID Plain
UID Stamped
UID Plain
UID Plain
Other
UID Incised
UID Plain
UID Plain
UID Decorated
UID Stamped
UID Complicated Stamped
UID Plain
Lake Jackson Decorated
Lamar Bold Incised
Wakulla Check Stamped
Wakulla Check Stamped
Fort Walton Incised
UID Stamped
Other
UID Plain
UID Decorated
UID Plain
UID Plain
Lake Jackson Decorated
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Body
Body
Rim
Body
Body
Body
Body
Shoulder
Rim
Rim
Rim
Body
Rim
Rim
Body
Body
Body
Rim
Body
Rim
Body
Base
Body
Body
Rim
Body
Body
Body
Shoulder
Body
Body
Body
Rim
Rim
Body
Body
Body
Body
Body
Rim
Rim
Body
Other
Body
Rim
99
28
1
12
2
3
3
1
1
8
1
1
1
1
1
18
2
2
1
1
1
1
1
6
2
1
2
107
1
8
6
2
6
2
1
4
1
1
1
1
7
3
1
92
1
329
1021
1021
1021
1021
1021
1021
1021
1022
1022
1022
1022
1022
1022
1022
1022
1022
1022
1023
1023
1023
1024
1024
1024
1024
1024
1024
1025
1025
1025
1025
1025
1025
1025
1025
1025
1025
1025
1025
1026
1026
1026
1026
1027
1027
1027
Lake Jackson Decorated
UID Plain
Wakulla Check Stamped
UID Incised
UID Incised
Fort Walton Incised
UID Complicated Stamped
Wakulla Check Stamped
UID Plain
UID Incised
Lake Jackson Decorated
Wakulla Check Stamped
UID Stamped
UID Complicated Stamped
UID Plain
Fort Walton Incised
Lake Jackson Decorated
UID Incised
UID Plain
Lake Jackson Decorated
Cool Branch Incised
Deptford Linear Check Stamped
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
UID Decorated
Lake Jackson Decorated
Swift Creek Complicated Stamped
UID Plain
UID Plain
Fort Walton Incised
UID Plain
UID Plain
UID Complicated Stamped
UID Decorated
Wakulla Check Stamped
Fort Walton Incised
Lake Jackson Decorated
UID Plain
Fort Walton Incised
Lake Jackson Decorated
UID Plain
UID Plain
UID Complicated Stamped
Lake Jackson Decorated
Collared Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Plate
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Plate
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Collared Jar
Rim
Body
Rim
Body
Rim
Body
Body
Body
Rim
Rim
Rim
Rim
Body
Body
Body
Rim
Body
Body
Body
Rim
Rim
Body
Body
Rim
Body
Body
Rim
Body
Rim
Rim
Rim
Body
Body
Body
Body
Body
Body
Rim
Body
Body
Rim
Rim
Body
Body
Shoulder
4
42
1
1
1
1
9
9
7
1
2
5
9
10
124
1
2
1
10
1
1
1
1
1
7
2
3
1
1
3
1
101
2
3
2
2
1
1
25
2
2
2
16
1
1
330
1030
1031
1031
1031
1031
1031
1031
1031
1031
1032
1032
1032
1032
1032
1032
1032
1033
1034
1034
1034
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1035
1036
1036
1036
1037
2042
2042
2042
2042
2042
UID Plain
Unknown
Wakulla Check Stamped
Unknown
UID Plain
Unknown
UID Decorated
Unknown
UID Plain
Unknown
Swift Creek Complicated Stamped
Unknown
UID Complicated Stamped
Unknown
Swift Creek Complicated Stamped
Unknown
UID Decorated
Unknown
Wakulla Check Stamped
Unknown
Fort Walton Incised
Unknown
Lake Jackson Decorated
Unknown
Lake Jackson Decorated
Plate
UID Plain
Unknown
UID Plain
Unknown
UID Plain
Jar
UID Plain
Unknown
UID Stamped
Unknown
UID Decorated
Unknown
UID Plain
Unknown
Point Washington Incised
Unknown
UID Complicated Stamped
Unknown
Leon Jefferson Complicated Stamped Unknown
UID Plain
Unknown
UID Plain
Unknown
UID Incised
Unknown
Lake Jackson Decorated
Collared Jar
UID Decorated
Unknown
UID Decorated
Unknown
UID Decorated
Unknown
UID Decorated
Unknown
UID Complicated Stamped
Unknown
UID Stamped
Unknown
Wakulla Check Stamped
Unknown
UID Plain
Unknown
UID Incised
Unknown
Point Washington Incised
Unknown
UID Plain
Unknown
Point Washington Incised
Unknown
UID Plain
Unknown
UID Plain
Unknown
UID Plain
Jar
UID Decorated
Unknown
Other
Unknown
Other
Unknown
Body
Body
Body
Body
Rim
Rim
Body
Body
Body
Body
Body
Rim
Rim
Rim
Body
Rim
Body
Body
Rim
Body
Body
Body
Body
Other
Rim
Shoulder
Rim
Rim
Shoulder
Body
Unknown
Body
Body
Body
Body
Unknown
Rim
Body
Rim
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
6
2
28
1
1
1
2
1
1
2
1
1
1
1
30
1
10
1
1
3
1
1
2
1
3
1
2
1
1
1
1
14
7
2
100
1
2
11
2
1
9
2
12
6
2
331
2042
2042
2042
2042
2042
2042
2044
2044
2044
2044
2043
2043
2043
2043
2043
2045
2046
2046
2046
2047
2047
2047
2047
2047
2048
2048
2049
2049
2049
2049
2050
2050
2051
2051
2051
2051
2052
2052
2052
2052
2052
2052
2053
2053
2053
Deptford Linear Check Stamped
Wakulla Check Stamped
Wakulla Check Stamped
Weeden Island Zone Punctated
Other
Other
Other
Swift Creek Complicated Stamped
Wakulla Check Stamped
UID Plain
Other
UID Plain
Other
Other
UID Plain
UID Plain
UID Decorated
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
Deptford Linear Check Stamped
Swift Creek Complicated Stamped
UID Plain
Wakulla Check Stamped
UID Plain
UID Decorated
Wakulla Check Stamped
UID Plain
Wakulla Check Stamped
Deptford Linear Check Stamped
UID Plain
Other
UID Decorated
Deptford Linear Check Stamped
Wakulla Check Stamped
Wakulla Check Stamped
UID Plain
UID Plain
Other
Other
Wakulla Check Stamped
UID Decorated
UID Decorated
Unknown
Unknown
Jar
Bowl
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Rim
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Other
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
4
3
1
1
1
1
1
1
2
9
1
8
1
1
1
1
2
3
1
11
1
0
2
2
3
2
1
1
1
1
1
1
3
2
2
1
1
1
1
2
2
2
3
2
1
332
2054
2054
2055
2055
2055
2055
2056
2058
2058
2058
2058
2059
2059
2059
2059
2059
2060
2061
2061
2061
2062
2062
2063
2064
2064
2064
2065
2065
2065
2065
2065
2065
2065
2066
2066
2066
2067
2067
2067
2067
2067
2067
2068
2068
2069
Other
Swift Creek Complicated Stamped
Other
Wakulla Check Stamped
Other
UID Plain
UID Plain
UID Decorated
UID Plain
Wakulla Check Stamped
Other
UID Plain
UID Decorated
Other
Swift Creek Complicated Stamped
UID Plain
UID Plain
Wakulla Check Stamped
Deptford Simple Stamped
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
UID Plain
Wakulla Check Stamped
UID Plain
UID Plain
Wakulla Check Stamped
Wakulla Check Stamped
Other
UID Complicated Stamped
UID Decorated
UID Plain
Other
Other
UID Plain
UID Decorated
UID Complicated Stamped
Wakulla Check Stamped
UID Plain
Other
UID Plain
Wakulla Check Stamped
UID Plain
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Jar
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Rim
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Shoulder
Unknown
Unknown
Unknown
Unknown
1
2
13
5
1
1
4
10
5
1
1
5
1
1
1
1
5
2
1
3
1
1
4
1
10
2
14
5
1
0
2
1
1
11
1
1
7
1
1
2
1
1
9
2
3
333
2069
2069
UID Decorated
Wakulla Check Stamped
Unknown
Unknown
Unknown
Unknown
1
2
334
REFERENCES CITED
Anderson, David G.
1985 Middle Woodland Societies on the Lower South Atlantic Slope: A View from
Georgia and South Carolina. Early Georgia 13(1 and 2):1-14.
1994
The Savannah River Chiefdoms: Political Change in the Late Prehistoric
Southeast. University of Alabama Press, Tuscaloosa.
1995
Paleoindian Interaction Networks in the Eastern Woodlands. In Native American
Interactions: Multiscalar Analyses and Interpretations in th Eastern Woodlands,
edited by Michael S. Nassaney and Kenneth E. Sassaman, pp. 3-26. University of
Tennessee Press, Knoxville.
1998
Swift Creek in a Regional Perspective. In A World Engraved: Archaeology of the
Swift Creek Culture, edited by Mark Williams and Daniel T. Elliott, pp. 274-300.
University of Alabama Press, Tuscaloosa.
1999
Examining Chiefdoms in the Southeast: An Application of Multiscalar Analysis.
In Great Towns and Regional Polities in the Prehistoric Southwest and Southeast,
edited by Jill E. Neitzel, pp. 215-242. Amerind Foundation New World Studies
Series. vol. No. 3. University of New Mexico Press, Albuquerque.
Anderson, David G. and John E. Jr. Cornelison
2001 Excavations at Mound A, Shiloh: The 2001 Season. Southeast Archeological
Center. Memorandum Report.
Anderson, David G. and Robert C. Mainfort, Jr.
2002 Epilogue: Future Directions for Woodland Archaeology in the Southeast. In The
Woodland Southeast, edited by David G. Anderson and Robert C. Mainfort, Jr.,
pp. 540-542. University of Alabama Press, Tuscaloosa.
Anderson, David G. and Kenneth E. Sassaman
1996 Modeling Paleoindian and Early Archaic Settlement in the Southeast: A Historical
Perspective. In The Paleoindian and Early Archaic Southeast, edited by David G.
Anderson and Kenneth E. Sassaman, pp. 16-28. University of Alabama Press,
Tuscaloosa.
Beck, Barry F. and Daniel D. Arden
1984 Karst Hydrogeology and Geomorphology of the Dougherty Plain, Southwest
Georgia. Southeastern Geological Society Guidebook No. 26. Southeastern
Geological Society and the Florida Sinkhole Research Institute, Tallahassee, FL.
335
Beck, Margaret E.
2006 Midden Ceramic Assemblage Formation: A Case Study from Kalinga,
Phillippines. American Antiquity 71(1):27-52.
Beck, Robin A., Jr.
2003 Consolidation and Hierarchy: Chiefdom Variability in the Mississippian
Southeast. American Antiquity 68(4):641-664.
Beisner, B. E., D. T. Haydon and K. Cuddington
2003 Alternative stable states in ecology. Frontiers in Ecology and the Environment
1(7):376-382.
Benson, Robert W.
2003 Aerating Fort Gordon: Prehistoric Cultural Resources Data from Survey of
28,000. Early Georgia 31(1):39-52.
Black, Eugene C.
1977 An Archaeology Survey of Dougherty County, GA. Georgia Archaeological Site
File, Manuscript File # 56.
Blanton, Richard E., Gary M. Feinman, Stephen A. Kowalewski and Peter N. Peregrine
1996 A Dual-Processual Theory for the Evolution of Mesoamerican Civilization.
Current Anthropology 37(1):1-14.
Blinman, Eric
2000 The Foundations, Practice and Limitations of Ceramic Dating
in the American Southwest. In It's About Time, edited by Stephen E. Nash. The
University of Utah Press, Salt Lake City.
Blitz, John H.
1993a Ancient Chiefdoms of the Tombigbee. University of Alabama Press, Tuscaloosa.
1993b Big Pots for Big Shots: Feasting and Storage in a Mississippian Community.
American Antiquity 58(1):80-96.
Blitz, John H. and Karl G. Lorenz
2002 The Early Mississippian Frontier in the Lower Chattahoochee-Apalachicola River
Valley. Southeastern Archaeology 21(2):117-135.
2006
The Chattahoochee Chiefdoms. University of Alabama Press, Tuscaloosa.
Bourdieu, Pierre
1990 The Logic of Practice. Translated by Nice, R. Stanford University Press, Stanford.
336
Bove, Frederick J.
1989 Formative Settlement Patterns on the Pacific Coast of Guatemala. 39 ed. BAR
International Series. B.A.R., Oxford.
Braun, David P.
1983 Pots as Tools. In Archaeological Hammers and Theories, edited by James A.
Moore and Arthur S. Keene, pp. 107-134. Academic Press, New York.
Brooks, Mark J.
1992 Alluvial Contexts of Site Formation in the Coastal Plain of the Savannah and
Edisto River Valleys. Paper presented at the Conference, Site Formation
Processes: Bioturbation and Gravity as a Factor in Sandy Coastal Plain Sites.,
USC - Coastal Carolina College.
Brooks, Mark J. and Kenneth E. Sassaman
1990 Point bar geoarchaeology in the Upper Coastal Plain of the Savannah River
Valley, South Carolina; A case study. Geological Society of America Centennial
Special 4:183-197.
Brose, David S. and George W. Percy
1978 Fort Walton Settlement Patterns. In Mississippian Settlement Patterns, edited by
Bruce D. Smith, pp. 81-108. Academic Press, New York.
Brown, James A., Richard A. Kerber and Howard D. Winters
1990 Trade and the Evolution of Exchange Relations at the Beginning of the
Mississippian Period. In The Mississippian Emergence, edited by Bruce D. Smith,
pp. 251-280. Smithsonian Institution Press, Washington, D.C.
Caldwell, Joseph R.
1964 Interaction Spheres in Prehistory. In Hopewellian Studies, edited by Joseph R.
Caldwell and Robert L. Hall, pp. 133-144. vol. XII. Illinois State Museum,
Springfield.
Caldwell, Joseph R. and Catherine McCann
1941 The Irene Mound Site, Chatham County, Georgia. The University of Georgia
Press, Athens.
Cammack, Shan, Nathan Klaus, L. Katherine Kirkman, E. Leonard and D. W. Hicks
2002 Restoration Ecology in the Chickasawhatchee Swamp: Where the Future Meets
the Past. Poster presented at the Proceedings of the Fourth Longleaf Alliance
Regional Conference, Southern Pines, N.C.
337
Carniero, Robert L.
1981 The Chiefdom as Precursor of the State. In The Transition to Statehood in the
New World, edited by Grant D. Jones and Robert R. Krautz, pp. 37-79.
Cambridge University Press, Cambridge.
Chamblee, John F.
1997 The Fishing Creek Survey. Early Georgia 25(1):55-80.
2000
The Classic-Postclassic Transition in the Central Mixteca Alta, Oaxaca. M.A.
Report, Department, University of Arizona.
Chamblee, John F., Barnet Pavao and Thomas W. Neumann
1998 Archaeological Salvage of the Plant Hammond Site. Early Georgia 27(2):1-53
and 81-113.
Chamblee, John F. and Frankie Snow
2002 The Archaeology of the Chickasawhatchee Swamp and its Environs: An
Introduction. Paper presented at the 59th Annual Southeastern Archaeological
Conference, Biloxi, MS.
Chapin, F. Stuart III, Pamela A. Matson and Harold A. Mooney
2002 Principles of Terrestrial Ecosystem Ecology. Springer-Verlag, New York.
Chilton, Elizabeth S.
1998 The Cultural Origins of Technical Choice: Unraveling Algonquian and Iroquoian
Ceramic Traditions in the Northeast. In The Archaeology of Social Boundaries,
edited by Miriam T. Stark, pp. 132-160. Smithsonian Institution Press,
Washington, D.C.
Christopherson, Gary L.
2000 Pursuit of the longue duree : using a geographic information system to model
archaeological settlement patterns in the region of Tell El- Umeiri, Jordan. Ph.D.
Dissertation, Department, University of Arizona.
Cobb, Charles R. and Brian M. Butler
2006 Mississippian Migration and Emplacement in the Lower Ohio Valley. In
Leadership and Polity in Mississippian Society, edited by Brian M. Butler and
Paul D. Welch, pp. 328-347. Center for Archaeological Investigations,
Carbondale.
Cobb, Charles R. and Patrick H. Garrow
1996 Woodstock Culture and the Question of Mississippian Emergence. American
Antiquity 61(1):21-38.
338
Cobb, Charles R. and Michael S. Nassaney
1995 Interaction and Integration in the Late Woodland Southeast. In Native American
Interactions: Multiscalar Analyses and Interpretations in th Eastern Woodlands,
edited by Michael S. Nassaney and Kenneth E. Sassaman, pp. 205-226.
University of Tennessee Press, Knoxville.
Codd, E. F.
1990
The relational model for database management: version 2. Addison-Wesley
Longman Publishing Co., Inc., Boston.
Coe, Joffre Lanning
1995 Town Creek Indian Mound: A Native American Legacy. University of North
Carolina Press, Chapel Hill.
Costin, Cathy Lynn
1991 Craft Specialization: Issues in Defining, Documenting, and Explaining the
Organization of Production. In Archaeological Method and Theory, edited by
Micheal B. Schiffer, pp. 1-56. vol. 3. Univeristy of Arizona Press, Tucson.
Cowell, C. Mark
1995 Presettlement Piedmont Forests - Patterns of Composition and Disturbance in
Central Georgia. Annals of the Association of American Geographers 85(1):6583.
1998
Historical Change in Vegetation and Disturbance on the Georgia Piedmont.
American Midland Naturalist 140(1):78-89.
Curci, Jessica Lee
2006 Logboats of the Eastern United States: Investigating the Question of Form. Ph.D.
Dissertation, Department, Eastern Carolina University.
Dean, Jeffrey
1990 Intensive Archaeological Survey of Long House Valley, Northeastern Arizona. In
The Archaeology of Regions: A Case for Full-Coverage Survey, edited by
Suzanne K. Fish and Stephen A. Kowalewksi, pp. 173-189. Smithsonian Series in
Archaeological Inquiry, Robert McC. Adams and Bruce D. Smith, general editor.
Smithsonian Institution Press, Washington, D.C.
Delcourt, Paul A. and Hazel R. Delcourt
1981 Vegetation Maps for Eastern North America: 40,000 YR B.P. to The Present. In
Geobotany II, edited by Robert C. Romans, pp. 123-167. Plenum, New York.
339
Denevan, William M.
1992 The Pristine Myth: The Landscape of the Americas in 1492. Annals of the
Association of American Geographers 82(3):369-385.
DePratter, Chester B.
1994 The Chiefdom of Cofitachequi. In The Forgotten Centuries: Indians and
Europeans in The American South, 1521-1704., edited by Charles M. Hudson and
Carmen Chaves-Tesser, pp. 197-226. University of Georgia Press, Athens.
Dobres, Marcia Ann and Christopher R. Hoffman
1994 Social Agency and the Dynamics of Prehistoric Technology. Journal of
Archaeological Method and Theory 1(3):211-258.
Drew, Mark B., L. Katherine Kirkman and Jr. Angus K. Gholson
1998 The Vascular Flora of Ichauway, Baker County, Georgia: A Remnant
Longleaf/Wiregrass Ecosystem. Castanea 63(1):1-24.
Earle, Timothy K.
1987 Specialization and teh Production and Exchange of Wealth: Hawaiian Chiefdoms
and the Inka Empire. In Specialization, Exchange, and Complex Societies, edited
by Elizabeth M. Brumfiel and Timothy K. Earle, pp. 64-75. Cambridge University
Press, Cambridge.
1997
How Chiefs Come to Power: the political economy in prehistory. Stanford
University Press, Stanford.
Elliott, Daniel T.
1981 Finch's Survey. Early Georgia 9(1 and 2):14-24.
2004
Southwest Georgia Archaeological Survey. LAMAR Institute. LAMAR Institute
Publication 60.
Elliott, Daniel T. and Jack T. Wynn
1991 The Vining Revival: A Late Simple Stamped Phase in the Central Georgia
Piedmont. Early Georgia 19(1):1-18.
Elson, Mark D
1998 Expanding the view of Hohokam platform mounds : an ethnographic perspective.
University of Arizona Press, Tucson.
Emerson, Thomas E.
1997 Cahokia and the Archaeology of Power. University of Alabama Press,
Tuscaloosa.
340
Emerson, Thomas E., Dale L. McElrath and Andrew C. Fortier (editors)
2002 Late Woodland Societies: Tradition and Transformation across the Midcontinent.
University of Nebraska Press, Lincoln.
Ethridge, Robbie Franklyn
1996 A Contest for Land: The Creek Indians on the Southern Frontier, 1796-1816.
Ph.D. Dissertation, Department, University of Georgia.
Fairbanks, Charles H.
2003 Archaeology of the Funeral Mound, Ocmulgee National Monument, Georgia.
Williams, Stephens. The University of Alabama Press, Tuscaloosa.
Feinman, Gary M.
1999 The Changing Structure of Macroregional Mesoamerica: The Classic-Postclassic
Transition in the Valley of Oaxaca. In World-Systems Theory in Practice:
Leadership, Production, and Exchange, edited by P.N. Kardulias, pp. 53-62.
Rowman and Littlefield, Lanham.
Feinman, Gary M. and Jill E. Neitzel
1984 Too Many Types: An Overview of Sedentary Prestate Societies in the Americas.
In Advances in Archaeological Method and Theory 7, edited by Michael B.
Schiffer, pp. 39-102. Academic Press, New York.
Fish, Paul R.
1976 Patterns of Prehistoric Site Distribution in Effingham and Screven Counties,
Georgia. University of Georgia Laboratory of Archaeology Series Report Number
11. Department of Anthropology, University of Georgia, Athens.
1978
Consistency in Archaeological Measurement and Classification: A Pilot Study.
American Antiquity 43(1):86-89.
Fish, Paul R. and Suzanne K. Fish
1977 Prehistoric Settlement in the Dry Creek Watershed. University of Georgia
Laboratory of Archaeology Series Report Number 14. Department of
Anthropology, University of Georgia, Athens.
Fish, Paul R. and Thomas Gresham
1990 Insights from Full-Coverage Survey in the Georgia Piedmont. In The Archaeology
of Regions: A Case for Full-Coverage Survey, edited by Suzanne K. Fish and
Stephen A. Kowalewski, pp. 147-172. Smithsonian Institution Press, Washington,
D.C.
341
Fish, Paul R. and David J. Hally
1983 The Wallace Reservoir Archaeological Project: An Overview. Early Georgia 11(1
and 2):1-18.
Fish, Paul R. and William W. Mitchell
1976 Prehistoric Settlement in the Dry Creek Watershed. University of Georgia
Laboratory of Archaeology Series Report Number 13. Department of
Anthropology, University of Georgia, Athens.
Fish, Suzanne K., Paul R. Fish and John H. Madsen (editors)
1992 The Marana Community in the Hohokam World. No 56. The University of
Arizona Press, Tucson.
Fish, Suzanne K. and Richard W. Jeffries
1983 The Site Plan at Cold Springs: 9Ge10. Early Georgia 11(1 and 2):61-73.
Fish, Suzanne K. and Stephen A. Kowalewski (editors)
1990 The Archaeology of Regions: A Case for Full-Coverage Survey. Smithsonian
Institution Press, Washington, D.C.
Foster, H. Thomas, Bryan Black and Marc D. Abrams
2004 A witness tree analysis of the effects of Native American Indians on the preEuropean settlement forest in east-central Alabama. Human Ecology 32(1):27-47.
Foster, H. Thomas II.
2001 Long Term Average Rate Maximization of Creek Indian Residential Mobility: A
Test of the Marginal Value Theorem. Ph.D. Dissertation, Department, The
Pennsylvania State University.
Fried, Morton H.
1967 The Evolution of Political Society. Random House, New York.
Giddens, Anthony
1979 Central Problems in Social Theory: Action, Structure, and Contradiction in
Social Analysis. University of California Press, Berkeley.
Goebel, P. Charles, Brian J. Palik, L. Katherine Kirkman, Mark B. Drew, Larry West and Dee C.
Pederson
2001 Forest Ecosystems fo Lower Gulf Coastal Plain landscape: multifactor
classification and analysis. Journal of the Torrey Botanical Society 128(1):47-75.
342
Golladay, Stephen W. and Juliann Battle
2001 Does the Chickasawhatchee Swamp Influence Water Quality. Paper presented at
the Proceedings of the 2001 Georgia Water Resources Conference, Athens,
Georgia.
Gómez-Pompa, Arturo and Andrea Kaus
1992 Taming the Wilderness Myth: Environmental policy and education are currently
based Western beliefs about nature rather than on reality. Bioscience 42(4):271279.
Gosselain, Olivier P.
1998 Social and Technical Identity in a Clay Crystal Ball. In The Archaeology of Social
Boundaries, edited by Miriam T. Stark, pp. 78-106. Smithsonian Institution Press,
Washington, D.C.
Gougeon, Ramie, Maureen S. Meyers, Jack T. Wynn and Elizabeth Shirk
1996 Georgia's Vining Phase: A Cultural Isolate? Paper presented at the 53rd
Southeastern Archaeological Conference: Birmingham, AL.
Gremillion, Kristen J.
2003 Eastern Woodlands Overview. In People and Plants in Ancient Eastern North
America, edited by Paul E. Minnis, pp. 17-49. Smithsonian Books, Washington,
D.C.
Gunderson, Lance H.
2000 Ecological Resilience -- In Theory and Application. Annual Review of Ecological
Systems 31:425-439.
Hally, David J.
1986 The Identification of Vessel Function: A Case Study from Northwest Georgia.
American Antiquity 51(2):267-295.
1994
An Overview of Lamar Culture. In Ocmulgee Archaeology, edited by David J.
Hally, pp. 144-174. University of Georgia Press, Athens.
1996
Platform-Mound Construction and the Instability of Mississippian Chiefdoms. In
Political Structure and Change in the Prehistoric Southeastern United States,
edited by John F. Scarry, pp. 92-127. Ripley P. Bullen Series, Jerald T. Milanich,
general editor. University Press of Florida, Gainesville.
343
Hally, David J., Marvin T. Smith and James B. Langford, Jr.
1990 The Archaeological Reality of de Soto's Coosa. In Columbian Consequences:
Archaeological and Historical Perspectives on the Spanish Borderlands East,
edited by David Hurst Thomas, pp. 121-134. vol. 2. 3 vols. Smithsonian
Institution Press, Washington, D.C.
Hatch, James W.
1995 Lamar Period Upland Farmsteads of the Oconee River Valley, Georgia. In
Mississippian Households and Communities, edited by J. Daniel Rogers and
Bruce D. Smith, pp. 135-155. University of Alabama Press, Tuscaloosa.
Hatch, James W., Jerald Ledbetter, Adam King, Pter van Rossum, Thomas Foster and Ervan
Garrison
1997 Excavations at the Marshall Site (9OC25): A Preliminary Report on the 19941995 Field Seasons. Early Georgia 25(2):1-35.
Hatch, James W., Stephen L. Whittington and Bennett Dyke
1982 A Simulation Approach to the Measurement of Change in Ceramic Frequency
Seriation. North American Archaeologist 3(4):333-350.
Heckenberger, Michael J.
1998 Manioc agriculture and sedentism in Amazonia: the Upper Xingu Example.
Antiquity 72:633-648.
Heckenberger, Michael J., Afukaka Kuikuro, Urissapá Tabata Kuikuro, J. Christian Russell,
Morgan Schmidt, Carlos Fausto and Bruno Franchetto
2003 Amazonia 1492: Pristine Forest or Cultural Parkland. Science 301(19 September
2003):1710-1713.
Heckenberger, Michael J., James B. Petersen and Eduardo Goés Neve
1999 Village Size and Permanence in Amazonia: Two Archaeological Examples from
Brazil. Latin American Antiquity 10(4):353-376.
Hegmon, Michelle
1998 Technology, Style, and Social Practices: Archaeological Approaches. In The
Archaeology of Social Boundaries, edited by Miriam T. Stark, pp. 264-279.
Smithsonian Institution Press, Washington, D.C.
Heidke, James M.
2004 Utilitarian Ceramic Production and Distribution in the Prehistoric Tonto basin. In
2000 Years of Settlement in the Tonto Basin: Overview and Synthesis of the Tonto
Creek Archaeological Project, edited by Jeffrey J. Clark and James M. Vint, pp.
77-138. Anthropological Papers No. 25. Center for Desert Archaeology, Tucson.
344
Hicks, D. W., H. E. Gill and S. A. Longsworth
1987 Hydrogeology, Chemical Quality, and Availability of Ground Water in the Upper
Floridan Aquifer, Albany Area, Georgia. Water-Resources Investigations Report
87-4145. U. S. Geological Survey, Doraville, GA.
Hill, J. Brett, Jeffrey J. Clark, William H. Doelle and Patrick D. Lyons
2004 Prehistoric Demography in the Southwest: Migration, Coalescence, and Hohokam
Population Decline. American Antiquity 69(4):689-716.
Holley, George R.
1999 Late Prehistoric Towns in the Southeast. In Great Towns and Regional Polities in
the Prehistoric American Southwest and Southeast, edited by Jill E. Neitzel, pp.
23-37. Amerind Foundation New World Studies Series. vol. No. 3. University of
New Mexico Press, Albuquerque.
Holling, C.S. and Lance H. Gunderson
2002 Resilience and Adaptive Cycles. In Panarchy: Understanding Transformations in
Human and Natural Systems, edited by Lance H. Gunderson and C.S. Holling, pp.
25-62. Island Press, Washtington, D.C.
Hudson, Charles, Marvin T. Smith and Chester B. DePratter
1984 The Hernando De Soto Expedition: From Apalachee to Chiaha. Southeastern
Archaeology 3(1):65-77.
Hudson, Charles M.
1997 Knights of Spain Warriors of the Sun: Hernando de Soto and the South's Ancient
Chiefdoms. University of Georgia Press, Athens.
Janitzky, Peter
1986 Particle Size Analysis. In Field and Laboratory Procedures used in a Soil
Chronosequence Study, edited by Michael J. Singer and Peter Janitzky, pp. 11-15.
U.S. Geological Survey Bulletin 1648. U.S. Geological Survey, Washington. D.C.
Jefferies, Richard W.
1976 The Tunacunnhee Site: Evidence of Hopewell Interaction in Northwest Georgia.
Anthropological Papers of the University of Georgia Number 1. Laboratory of
Archaeology, University of Georgia, Athens,.
1979
The Tunacunnhee Site: Hopewell in Northwest Georgia. In Hopewell
Archaeology: The Chillicothe Conference, edited by David S. Brose and N'omi
Greber, pp. 162-170. The Kent State University Press, Kent.
345
Jefferies, Richard W. (cont.)
1994 The Swift Creek Site and Woodland Platform Mounds in the Southeastern United
States. In Ocmulgee Archaeology, edited by David J. Hally, pp. 71-83. University
of Georgia Press, Athens.
Johnson, Jay K., Thomas L. Sever, Scott L.H. Madry and Harry T. Hoff
1988 Remote Sensing and GIS Analysis in Large Scale Survey Design in North
Mississippi. Southeastern Archaeology 7(2):124-131.
Junker, Laura L.
1999 Raiding, Trading, and Feasting: The Political Economy of Phillipine Chiefdoms.
The University of Hawai'i Press, Honolulu.
Keene, Deborah A.
2004 Reevaluating Late Prehistoric Coast Subsistence and Settlement Strategies: New
Data from Grove's Creek Site, Skidaway Island, Georgia. American Antiquity
69(4):671-688.
Kelly, Arthur R., Richard Nonas, Betty Broyles, Clemens De Baillou, David W. Chase and Jr.
Frank T. Schnell
1962 Survey of Archaeological Sites in Clay and Quitman Counties, Georgia: 9CLA2,
9CLA7, 9CLA15, 9CLA28, 9CLA38, 9CLA51, 9QU25. University of Georgia
Laboratory of Archaeology Series Report Number 24. Department of
Anthropology, University of Georgia, Athens.
Kelly, John E.
1990 The Emergence of Mississippian Culture in the American Bottom Region. In The
Mississippian Emergence, edited by Bruce D. Smith, pp. 113-152. Smithsonian
Institution Press,, Washington, D.C. :.
Kent, Susan
1992 Study Variability in the Archaeological Record: An Ethnoarchaeological Model
for Distinguishing Mobility Pattersn. American Antiquity 57(4):635-660.
King, Adam
1996
2003
Tracing Organizational Change in Mississippian Chiefdoms of the Etowah River
Valley, Georgia. Doctoral Dissertation, Department, The Pennsylvania State
University.
Etowah: The Political History of a Chiefdom Capital. University of Alabama
Press, Tuscaloosa.
346
King, Adam and Maureen S. Meyers
2002 Exploring the Edges of the Mississippian World. Southeastern Archaeology
21(2):113-116.
Kirkland, Servey Dwight
1994 Ten millennia of human land use on the interior coastal plain of Georgia. M.A.
Thesis, Department, University of Georgia.
Kirkman, L. Katherine, P. Charles Goebel, Larry West, Mark B. Drew and Brian J. Palik
2000 Depressional Wetland Vegetation Types: A Question of Plant Community
Development. Wetlands 20(2):373-385.
Knight, Vernon James, Jr.
1990 Social Organization and the Evolution of Hierarchy in Southeastern Chiefdoms.
Journal of Anthropological Research 46(1):1-23.
Knight, Vernon James, Jr.
1997 Some Developmental Parallels between Cahokia and Moundville. In Cahokia:
Domination and Ideology in the Mississippian World, edited by Timothy R.
Pauketat and Thomas E. Emerson, pp. 229-247. University of Nebraska Press,
Lincoln.
Knight, Vernon James, Jr. and Vincas P. Steponaitis
1998 A New History of Moundville. In Archaeology of the Moundville Chiefdom,
edited by Vernon James Knight, Jr. and Vincas P. Steponaitis. Smithsonian
Institution Press, Washington, D.C.
Kowalewski, Stephen A.
1990 Merits of Full-Coverage Survey: Examples from the Valley of Oaxaca, Mexico.
In The Archaeology of Regions: A Case for Full-Coverage Survey, edited by
Suzanne K. Fish and Stephen A. Kowalewksi, pp. 33-85. Smithsonian Series in
Archaeological Inquiry, Robert McC. Adams and Bruce D. Smith, general editor.
Smithsonian Institution Press, Washington, D.C.
1995
Large-Scale Ecology in Aboriginal Eastern North America. In Native American
Interactions: Multiscalar Analyses and Interpretations in th Eastern Woodlands,
edited by Michael S. Nassaney and Kenneth E. Sassaman, pp. 147-173.
University of Tennessee Press, Knoxville.
1996 Clout, Corn, Copper, Core Periphery, Culture Area. In Pre-Columbian WorldSystems, edited by Peter N. Peregrine and Gary M. Feinman, pp. 51-64. Prehistory
Press, Madison.
347
Kowalewski, Stephen A. (cont.)
2000 Cyclical Transformations in North American Prehistory. In Aternatives of Social
Evolution, edited by Nikolay N. Kradin, Andrey V. Korotayev, Dmitri M.
Bondarenko, Victor de Munck and Paul K. Wason, pp. 175-184. Institute of
History, Archaeology, and Ethnohistory, Far Eastern Branch of the Russiona
Academy of Sciences, Vladivostok.
2003
Backcountry Pots. Ancient Mesoamerica 14(2003):67-75.
2004
The New Past: From Region to Macroregion. The Journal of Social Evolution and
History 2004.
2005
When the Past is Destroyed: Loss of Archaeological Sites due to Urbanization.
Early Georgia 33(1):1-44.
Kowalewski, Stephen A. , Andrew K. Balkansky, Laura R. Stiver, Thomas J. Pluckhahn, John
F. Chamblee, Verónica Pérez Rodríguez, Verenice Y. Heredia Espinoza and Charlotte A. Smith
n.d. The Origins of the Ñuu: Archaeology in the Mixteca Atla, Mexico. Book
Manuscript Under Review. Submitted, June 2005, University of Colorado Press,
Boulder.
Kowalewski, Stephen A., Gary M. Feinman, Laura Finsten, Richard E. Blanton and Linda M.
Nicholas
1989 Monte Alban's Hinterland, Part II: Preshispanic Settlement Patterns in Tlacolula,
Etla, and Ocotlan, The Valley of Oaxaca, Mexico. Memoirs of the Museum of
Anthropology Number 23 1. University of Michigan, Ann Arbor.
Kowalewski, Stephen A. and James W. Hatch
1991 The Sixteenth-Century of Settlement in the Upper Oconee Watershed, Georgia.
Southeastern Archaeology 10(1):1-17.
Kowalewski, Stephen A. and Mark Williams
1989 The Carroll Village Site: Analysis of 1936 Excavations at a Mississippian
Farmstead in Georgia. Southeastern Archaeology 8(1):46-66.
Kramer, Elizabeth A., Michael J. Conroy, Mathew J. Elliott, Elizabeth A. Anderson, William R.
Bumback and Jeanne Epstein
2003 A Gap Analysis of Georgia: August 2003, Final Report. Institute of Ecology and
Georgia Cooperative Fish & Wildlife Research Unit, University of Georgia. U.S.
Geological Survey, Gap Analysis Program.
Kvamme, Kenneth L.
1992 Predictive Site Location model on the High Plains: an example with an
independent test. Plains Anthropologist 37(138):19-40.
348
Kvamme, Kenneth L. (cont.)
2003 Geophysical Surveys as Landscape Archaeology. American Antiquity 68(3):435458.
Laboratory of Archaeology, University of Georgia
n.d.
Main Artifact Catalog.
Lazarus, Yulee and Carolyn B. Hawkins
1976 Pottery of the Fort Walton Period. Temple Mound Museum, Fort Walton Beach,
FL.
Ledbetter, R. Jerald
1995 Lamar Stuff from the Greenbrier Tract. Georgia Archaeological Site File,
Manuscript File # 456.
Lyons, Patrick D., J. Brett Hill and Jeffrey J. Clark
2005 Demography, Social Power, and the Dynamics of Identity among Ancient
Immigrants. Paper presented at the 70th Annual Meeting of the Society for
American Archaeology, Salt Lake City, Utah.
MacNeish, Richard S., Frederick A. Peterson and Kent V. Flannery
1970 The Prehistory of the Tehuacan Valley, Volume 3: Ceramics 3. University of
Texas Press, Austin.
McElrath, Dale L., Thomas E. Emerson and Andrew C. Fortier
2002 Social Evolution or Social Response? A Fresh Look at the "Good Grey Cultures"
After Four Decades of Midwest Research. In Late Woodland Societies: Tradition
and Transformation across the Midcontinent, edited by Thomas E. Emerson, Dale
L. McElrath and Andrew C. Fortier, pp. 3-36. University of Nebraska Press,
Lincoln.
Meentenmeyer, Vernon and Elgene O. Box
1987 Scale Effects in Landscape Studies. In Landscape Heterogeniety and
Disturbance, edited by Turner. Monica Goigel, pp. 15-34. Springer-Verlag, New
York.
Mehlic, A.
1984
Mehlic 3 Soil Test Extractant: A Modification of Mehlic 2 Extractant.
Communications in Soil Science and Plant Analysis 15(12):1409-1416.
Meyers, Maureen S.
2002 The Mississippian Frontier in Southwestern Virginia. Southeastern Archaeology
21(2):178-191.
349
Milanich, Jerald T.
1994 Archaeology of Precolombian Florida. University Press of Florida, Gainesville.
1996
Native Chiefdoms and the Exercise of Complexity in Sixteenth-Century Florida.
In Chiefdoms and Chieftaincy in the Americas, edited by Elsa M. Redmond, pp.
245-264. University Press of Florida, Gainesville.
Milanich, Jerald T., Ann S. Cordell, Vernon J. Knight, Jr., Timothy A. Kohler and Brenda J.
Sigler- Lavelle
1997 Archaeology of Northern Florida: A.D. 200-900, The McKeithen Weeden Island
Culture. Second Edition ed. University Press of Florida, Gainesville.
Mills, Barbara J.
1989 Integrating Functional Analyses of Vessels and Sherds through Models of
Ceramic Assemblage Formation. World Archaeology 21(1):133-147.
2002
Acts of Resistance: Zuni Ceramics, Social Identity, and the Pueblo Revolt. In
Archaeologies of the Pueblo Revolt, edited by Robert W. Preucel, pp. 85-98.
University of New Mexico Press, Albuquerque.
Milner, George R.
1998 The Cahokia Chiefdom: The Archaeology of a Mississippian Society.
Smithsonian Institution Press, Washington, D.C.
Muller, Jon
1997
Mississippian Political Economy. Interdisciplinary Contributions to Archaeology.
Plenum Press, New York.
Murphy, John T.
2002 Potential Contributions of Relational Database Theory to Archaeological Method
and Practice. Paper presented at the 68th Annual Society for American
Archaeology Meetings, Denver, Co.
Museum, Arizona State
2001 Field Guide and Excavation Manual. Arizona State Museum. Manuscript on File.
Nash, Roderick
1967 Wilderness and the American Mind. Yale University Press, New Haven.
Nass, John P., Jr. and Richard W. Yerkes
1995 Social Differentiation in Mississippian and Fort Ancient Societies. In
Mississippian Communities and Households, edited by J. Daniel Rogers and
Bruce D. Smith, pp. 58-80. University of Alabama Press, Tuscaloosa.
350
Nassaney, Michael S. and Charles R. Cobb
1991 Patterns and Processese of Late Woodland Development in the Greater
Southeastern United States. In Stability, Transformation, and Variation: The Late
Woodland Southeast, edited by Michael S. Nassaney and Charles R. Cobb, pp.
285-322. Plenum, New York.
Nielsen, Axel E.
1993 Formation Processes of Ceramic Assemblages: Examples from the San Juan
Basin, Upper Puerco River, and Little Colorado River Regions. In Across the
Colorado Plateau: Anthropological Studies for the Transwestern Pipeline
Expansion Project, Volume XVI: Interpretation and Ceramic Artifacts, edited by
Barbara J. Mills, Christine E. Goetze and Maria Nieves Zedeño, pp. 151-174.
Office of Contract Archaeology and Maxwell Museum of Anthropology,
University of New Mexico., Albuquerque.
Pauketat, Timothy R.
1989 Monitoring Mississippian Homestead Occupation Span Using Ceramic Refuse.
American Antiquity 54(2):288-310.
1994
The Ascent of Chiefs: Cahokia and Mississippian Politics in Native North
America. University of Alabama Press, Tuscaloosa.
1997
Specialization, Political Symbols, and the Crafty Elite of Cahokia. Southeastern
Archaeology 16(1):1-15.
2002
Cahokian ritual and the Ramey field: new insights from old collections.
Southeastern Archaeology 21(1):79-91.
Pearson, Charles E.
1977 Analysis of Late Prehistoric Settlements of Ossabaw Island, G eorgia. University
of Georgia Laboratory of Archaeology Series Report Number 12. Department of
Anthropology, University of Georgia, Athens.
Peebles, Christopher S. and Susan M. Kus
1977 Some Correlates of Ranked Societies. American Antiquity 42(3):421-448.
Percy, George W. and David S. Brose
1975 Weeden Island Ceremonialism in northwest Florida: A reappraisal, Paper
Presented at the 39th Annual Meeting of the Society for American Archaeology,
Washington, D.C.
351
Peregrine, Peter N.
1992 Mississippian Evolution: A World Systems Perspective. Monographs in World
Archaeology, No. 9. Prehistory Press, Madison.
1996 Hyperopia or Hyperbole? The Mississippian World System. In Pre-Columbian
World-Systems, edited by Peter N. Peregrine and Gary M. Feinman, pp. 39-50.
Prehistory Press, Madison.
Pluckhahn, Thomas J.
1994 The Evolution of Settlement and Land Use in Jackson and Madison Counties,
Georgia. Unpublished Masters Thesis, Department, University of Georgia.
1996
Archaeological Survey, Testing, and Damage Assessment of Lewis Mound and
Village Site (9BN39): Fort Stewart Military Reservation, Bryan County, Georgia.
LAMAR Institute. LAMAR Institute Publication 39.
1997
Rethinking Early Mississippian Chronology and Cultural Contact in Central
Georgia: The View from Tarver (9JO6). Early Georgia 25(1):21-54.
2003
Kolomoki: Settlement, Ceremony, and Status in the Deep South, A.D. 350 to 750.
Unversity of Alabama Press, Tuscaloosa.
Pluckhahn, Thomas J. and David A. McKivergan
2002 The Middle Mississippian Period in the Georgia Coastal Zone. Southeastern
Archaeology 21(2):149-161.
Polhemus, Richard R.
1987 The Toqua Site: A Late Mississippian Dallas Phase Town. Report of
Investigations No. 14. University of Tennessee, Department of Anthropology,
Knoxville.
Rangel, Rodrigo
1993 [1851] Account of the Northern Conquest and Discovery of Hernando de Soto. In
The De Soto Chronicles: The Expedition of Hernando de Soto to North America
in 1539-1543, edited by Lawrence A. Clayton and Jr. Knight, Vernon James, pp.
247-306. Translated by John E .Worth. vol. I. University of Albama Press,
Tuscaloosa.
Renfrew, Colin and John F. Cherry (editors)
1986 Peer Polity Interaction and Socio-Political Change. Cambridge University Press,
Cambridge.
Rice, Prudence M.
1987 Pottery Analysis. University of Chicago Press, Chicago.
352
Rudolph, James L. and David J. Hally
1985 Archaeological Investigations at the Beaverdam Creek Site (9EB85). Russell
Papers. Archeological Services, National Park Service, Atlanta.
Rudolph, Teresa P.
1991 The Woodland "Problem" in North Georgia. In Stability, Transformation, and
Variation: The Late Woodland Southeast, edited by Michael S. Nassaney and
Charles R. Cobb, pp. 259-248. Plenum, New York.
Russell, Emily W. B.
1983 Indian-Set Fires in the Forests of the Northeastern United States. Ecology
64(1):78-88.
Sagebiel, Kerry
2005 Shifting Allegiances at La Milpa, Belize: A Typological, Chronological, and
Formal Analysis of the Ceramics. Ph.D. Dissertation, Department, University of
Arizona.
Sahlins, Marshall D.
1958 Social Stratification in Polynesia. American Ethnological Society Monograph 29.
University of Washington Press, Seattle.
Sanders, William T., Jeffrey A. Parsons and Robert S. Santley
1979 The Basin of Mexico: Ecological Processes in the Evolution of a Civilization.
Academic Press, New York.
Sassaman, Kenneth E.
1992 Colluvial Sites in the Aiken Plateau. Paper presented at the Conference, Site
Formation Processes: Bioturbation and Gravity as a Factor in Sandy Coastal Plain
Sites., USC - Coastal Carolina College.
Scarry, C. Margaret
1983 Paleoethnobotany of the Granada Site. In Excavation at the Granada Site:
Archaeology and History of the Granada, edited by John W. Griffin, pp. 181-245.
vol. I. Florida Division Archives, History and Records Management, Miami.
2003
Patterns of Wild Plant Utilizaton in the Prehistoric Eastern Woodlands. In People
and Plants in Ancient Eastern North America, edited by Paul E. Minnis.
Smithsonian Books, Washington, D.C.
Scarry, C. Margaret (editor)
1993 Foraging and Farming in the Eastern Woodlands. University Press of Florida,
Gainesville.
353
Scarry, John F.
1985 Proposed Revision of the Fort Walton Ceramic Typology: a type variety system.
Florida Anthropologist 38(3):199-233.
1990
Mississippian Emergence in the Fort Walton Area: The Evolution of the Cayson
and Lake Jackson Phases. In The Mississippian Emergence, edited by Bruce D.
Smith, pp. 227-250. Smithsonian Institution Press,, Washington, D.C. :.
1994
The Apalachee Chiefdom: A Mississippian Society on the Fringe of the
Mississippian World. In The Forgotten Centuries: Indians and Europeans in The
American South, 1521-1704., edited by Charles M. Hudson and Carmen ChavesTesser, pp. 156-178. University of Georgia Press, Athens.
1995 Apalachee Homesteads: The Basal Social and Economic Units of a Mississippian
Chiefdom. In Mississippian Households and Communities, edited by J. Daniel
Rogers and Bruce D. Smith, pp. 201-223. University of Alabama Press,
Tuscaloosa.
1996
Looking for and at Mississippian Political Change. In Political Structure and
Change in the Prehistoric Southeastern United States, edited by John F. Scarry,
pp. 2-11. Ripley P. Bullen Series, Jerald T. Milanich, general editor. University
Press of Florida, Gainesville.
1999 How Great were Southeastern Polities? In Great Towns and Regional Polities in
the Prehistoric Southwest and Southeast, edited by Jill E. Neitzel, pp. 59-74.
Amerind Foundation New World Studies Series. vol. No. 3. University of New
Mexico Press, Albuquerque.
Schiffer, Michael B.
1983 Toward the Identification of Formation Processes. American Antiquity 48(4):675706.
1996
Formation Processes of the Archaeological Record. Second ed. University of
New Mexico Press, Albuquerque.
Schiffer, Michael B., Alan P. Sullivan and Timothy C. Klinger
1978 The Design of Archaeological Surveys. World Archaeology 10(1):1-28.
Schnell, Frank T., Vernon James Knight and Gail S. Schnell
1981 Cemochechobee: archaeology of a Mississippian ceremonial center on the
Chattahoochee River. Ripley P. Bullen Monographs in Anthropology and History.
University Press of Florida, Gainesville.
354
Schnell, Frank T. and Newell O. Wright, Jr.
1993 Mississippi Period Archaeology of the Georgia Coastal Plain. Laboratory of
Archaeology Series Report Number 26. University of Georgia, Athens.
Schortman, Edward M. and Patricia A. Urban
1992 Current Trends in Interaction Research. In Resources, Power, and Interregional
Interaction, edited by Edward M. Schortman and Patricia A. Urban, pp. 235-256.
Plenum Press, New York.
Schroeder, Sissel
1997 Place, Productivity, and Politics: The Evolution of Cultural Complexity in the
Cahokia Area. Ph.D. Disssertation, Department, The Pennsylvania State
University.
Sears, William H.
1956 Excavations at Kolomoki: Final Report. University of Georgia Series in
Anthropology Number 5. The University of Georgia Press, Athens.
Service, Elman R.
1962 Primitive Social Organization. Random House, New York.
Shaul, David Leedom and Jane H. Hill
1998 Tepimans, Yumans, and Other Hohokam. American Antiquity 63(3):417-436.
Shirk, Elizabeth Crabill
1979 Intra-site Phosphate Analysis: A Test Case at Cold Springs. University of Georgia
Laboratory of Archaeology Series Report Number 20. Department of
Anthropology, University of Georgia, Athens.
Smith, Betty A.
1979 The Hopewell Connection in Southwest Georgia. In Hopewell Archaeology: The
Chillicothe Conference, edited by David S. Brose and N'omi Greber, pp. 181-186.
The Kent State University Press, Kent.
Smith, Bruce D.
1978 Variation in Mississippian Settlement Patterns. In Mississippian Settlement
Patterns, edited by Bruce D. Smith, pp. 479-503. Academic Press, New York.
1995
Seed Plant Domestication in Eastern North America. In Last Hunters, First
Farmers, edited by T. Douglas Price and Anne Birgette Gebauer, pp. 193-213.
School of American Research, Santa Fe.
355
Smith, Bruce D. and C. Wesley Cowan
2003 Domesticated crop plants and the evolution of food production economics in
eastern North America. In People and Plants in Ancient Eastern North America,
edited by Paul E. Minnis, pp. 105-125. Smithsonian Books, Washington, D.C.
Smith, Charlotte A.
2002 Concordant Change and Core-Periphery Dynamics: A Synthesis of Highland
Mesoamerican Archaeological Survey Data. Ph.D. Dissertation, Department,
University of Georgia.
Smith, Charlotte A. and Jennifer Freer Harris
2001 Sprawl and the Destruction of Georgia's Archaeological Resources. Early
Georgia 29(1):65-77.
Smith, Marvin T.
1994 Archaeological Investigations at the Dyar Site, 9GE5. Laboratory of Archaeology
Series Report Number 32. University of Georgia, Athens.
Smith, Marvin T. and Stephen A. Kowalewski
1980 Tentative Identification of Prehistoric "Province" in Piedmont Georgia. Early
Georgia 8(1 and 2):1-13.
Smith, Marvin T. and Mark Williams
1994 Mississippian Mound Refuse Disposal Patterns and Implications for
Archaeological Research. Southeastern Archaeology 13(1):27-35.
Snow, Frankie
1990 Pine Barrens Lamar. In Lamar Archaeology: Mississippian Chiefdoms in the
Deep South, edited by Mark Williams and Gary Shapiro, pp. 82-93. University of
Alabama Press, Tuscaloosa.
n.d.
Field Notes from the Chickasawhatchee. South Georgia College. Manuscript on
File.
Snow, Frankie H and Keith Stephenson
1998 Swift Creek Designs: A Tool for Monitoring Interaction. In A World Engraved:
Archaeology of the Swift Creek Culture, edited by Mark Williams and Daniel T.
Elliott, pp. 99-111. University of Alabama Press, Tuscaloosa.
Stephenson, Donald Keith
1990 Investigations of Ocmulgee Cord-Marked Pottery Sites in the Big Bend Region of
Georgia. M.A., Department, Georgia.
356
Stephenson, Keith, Judith A. Bense and Frankie Snow
2002 Aspects of Deptford and Swift Creek of the South Atlantic and Gulf Coastal
Plains. In The Woodland Southeast, edited by David G. Anderson and Robert C.
Mainfort, Jr. University of Alabama Press, Tuscaloosa.
Stephenson, Keith, Adam King and Frankie Snow
1996 Middle Mississippi Occupation in the Ocmulgee Big Bend Region. Early Georgia
24(2):1-41.
Stephenson, Keith, Mark Williams, Tom Pluckhahn, Jerald Ledbetter, Jeff Price, John Wood,
Dot Wood, Carol McCanless, Hal Ellison and ken Carleton
1993 Archaeological Investigations at the Wildcat Creek Site (9OC169): A Late Lamar
Occupation in the Oconee River Floodplain. Lamar Institute Reports 29. Lamar
Institute, Watkinsville.
Stephenson, Keith, John E. Worth and Frankie Snow
1990 A Savannah Period Mound in the Upper Interior Coastal Plain of Georgia. Early
Georgia 18(1 and 2):41-65.
Steponaitis, Vincas P.
1978 Location Theory and Complex Chiefdoms: A Mississippian Example. In
Mississippian Settlement Patterns, edited by Bruce D. Smith, pp. 417-454.
Academic Press, New York.
Steponaitis, Vincas P.
1983 Ceramics, Chronology, and Community Patterns: An Archaeological Study of
Moundville. Academic Press, New York.
Stoltman, James and Frankie H Snow
1998 Cultural Interaction within Swift Creek Society: People, Pots and Paddles. In A
World Engraved: Archaeology of the Swift Creek Culture, edited by Mark
Williams and Daniel T. Elliott, pp. 130-153. University of Alabama Press,
Tuscaloosa.
Swanton, John R.
1985[1939] Final Report of the United States De Soto Expedition Commission. Classics
of Smithsonian Anthropology. Smithsonian Institution Press, Washington, D.C.
Tetra Tech, Inc.
2004 Upper Yadkin River Basin Detailed Assessment Report. Tetra Tech, Inc.
Trigger, Bruce G.
1989 A History of Archaeological Thought. Cambridge University Press, Cambridge.
357
USDA Soil Conservation Service
1968 Soil Survey of Dougherty County, Georgia. United States Department of
Agriculture, Washington, D.C.
1986
Soil Survey of Baker and Mitchell Counties, Georgia. United States Department
of Agriculture, Washington, D.C.
Van Keuren, Scott
2001 Ceramic Style and the Reorganization of Pueblo Communities in East Central
Arizona, A.D. 1275 - 1375. Ph.D. Dissertation, Department, University of
Arizona.
Van Voorhies, Christine and Mark Williams
1993 Archaeological Survey of a Clear-Cut Tract Off Liberty Church Road in Southern
Greene County, Georgia. Lamar Institute Reports 28. Lamar Institute,
Watkinsville.
Varien, Mark D.
1999 Sedentism and Mobility in a Social Landscape. University of Arizona Press,
Tucson.
Varien, Mark D. and Barbara J. Mills
1997 Accumulations Research: Problems and Prospects for Estimating Site Occupation
Span. Journal of Archaeological Method and Theory 4(2):141-191.
Vogel, Gregory
2002 A Handbook of Soil Description for Archeologists. Arkansas Archeological
Survey Technical Paper 11. Arkansas Archeological Survey, Fayeteville, AR.
Vogel, Gregory, Marvin Kay and Louis Vogele, Jr.
2005 A Platform Mound at the Norman Site (34WG2), Eastern Oklahoma.
Southeastern Archaeology 24(1):28-45.
Waggoner, James C., Jr.
2005 Ridgeway Road: Archaeological Survey in West-Central Georgia. Early Georgia
3(2):143-158.
Wagner, Gail E.
2003 Anthropogenic Ecology in Eastern North America. In People and Plants in
Ancient Eastern North America, edited by Paul E. Minnis, pp. 126-169.
Smithsonian Books, Washington, D.C.
358
Walker, William
1999 Ceremonial Trash? In Pottery and People: A Dynamic Interaction, edited by
James M. Skibo and Gary M. Feinman. University of Utah Press, Salt Lake City.
Wallace, Henry D.
1995 Archaeological Investigations at Los Morteros: A Prehistoric Settlement in the
Northern Tucson Basin: Part I. Center for Desert Archaeology. Anthropological
Papers. No. 17.
Wandsnider, LuAnn and Eileen L. Camilli
1992 The Character of Surface Archaeological Deposits and Its Influence on
Survey Accuracy. Journal of Field Archaeology 19(2):169-188.
Warren, Robert E.
1990 Predictive modeling in archaeology: a primer. In Interpreting Space: GIS and
Archaeology, edited by Kathleen M.S. Allen, Stanton W. Green and Ezra B. W.
Zubrow, pp. 90-111. Applications of geographic information systems. Taylor &
Francis, London.
Wauchope, Robert
1966 Archaeological Survey of Northern Georgia: with a test of some cultural
hypotheses. Memoirs of the Society for American Archaeology 31(5). Society for
American Archaeology, Washington, D.C.
1939
Welch, Paul
1981
1990
Progress of the WPA Archaeological Survey of Georgia. Georgia Archaeological
Site File, Manuscript File # 544.
Moundville's Economy. University of Alabama Press, Tuscaloosa.
Mississippian Emergence in West Central Alabama. In The Mississippian
Emergence, edited by Bruce D. Smith, pp. 197-225. Smithsonian Institution
Press,, Washington, D.C. :.
West Georgia College and State University, Antonio J. Waring Laboratory of Archaeology
n.d.
Main Artifact Catalog.
Wharton, Charles H.
1978 The Natural Environments of Georgia Bulletin 114. Department of Natural
Resources, Environmental Protection Division, Georgia Geologic Survey,
Atlanta.
359
Wheeler, Brian
2004 Florida State University Archaeological Field School at Letchworth Mounds,
Summer B 2004 (May 12-June 20, 2004). Bureau of Archaeological Research.
Research Design.
White, Nancy Marie
1981 Archaeological Survey at Lake Seminole, Jackson and Gadsden Counties,
Florida; Seminole and Decatur Counties, Georgia. Cleveland Museum of Natural
History. Archaeological Research Report. Number 29.
1985
Nomenclature and Interpretation in Borderland Chronology: A Critical Overview
of Northwest Florida Prehistory. The Florida Anthropologist 38(1-2):163-174.
Wiessner, Polly
2002 The Vines of Complexity: Egalitarian Structures and the Institutionalization of
Inequality among the Enga. Current Anthropology 43(2):233-269.
Wilcox, David R.
1999 A Peregrine View of Macroregional Systems in the North American Southwest.
In Great Towns and Regional Polities in the Prehistoric Southwest and Southeast,
edited by Jill E. Neitzel, pp. 115-142. Amerind Foundation New World Studies
Series. vol. No. 3. University of New Mexico Press, Albuquerque.
Willey, Gordon R.
1998 [1949] Archaeology of the Florida Gulf Coast. Southeastern Classics in
Archaeology, Anthropology, and History. University Press of Florida,
Gainesville.
Williams, John Mark
1983 The Joe Bell site : Seventeenth Century lifeways on the Oconee River. Ph.D.
Dissertation, Department, University of Georgia.
Williams, Mark
1990a Archaeological Excavations at Shoulderbone Mounds and Village (9HK1). Lamar
Institute Reports 3. Lamar Institute, Watkinsville.
1990b Archaeological Excavations at Shinholser (9BL1). Lamar Institute Reports 4.
Lamar Institute, Watkinsville.
1990c Archaeological Investigations at Punk Rock Shelter (9PM211). Lamar Institute
Reports 9. Lamar Institute, Watkinsville.
360
Williams, Mark (cont.)
1993a The View from Above: Archaeological Excavations at Brown's Mount. Lamar
Institute Reports 23. Lamar Institute, Watkinsville.
1994a Archaeological Site Distributions in Georgia: 1994. Early Georgia 22(1):35-76.
1994b Growth and Decline of the Oconee Province. In The Forgotten Centuries: Indians
and Europeans in The American South, 1521-1704., edited by Charles M. Hudson
and Carmen Chaves-Tesser, pp. 179-196. University of Georgia Press, Athens.
1994c The Origins of the Macon Plateau Site. In Ocmulgee Archaeology, edited by
David J. Hally, pp. 130-137. University of Georgia Press, Athens.
1996
Archaeological Investigations at the Sawyer Site (9LS1). LAMAR Institute.
LAMAR Institute Publication 32.
1999
Archaeological Testing at the Hartley Posey Site, 9TR12. LAMAR Institute.
LAMAR Institute Publication 43.
1999
Lamar Revisited: 1996 Test Excavations at the Lamar Site (9Bi2). Lamar Institute
Reports 43. Lamar Institute, Watkinsville.
2000a Archaeological Site Distributions in Georgia: 2000. Early Georgia 28(1):1-55.
2000b The Carolina-Walker Clearcut in Morgan County, Georgia. LAMAR Institute.
LAMAR Institute Publication 66.
2003a Introduction to the 2003 edition. In Archaeology of the Funeral Mound,
Ocmulgee National Monument, Georgia., pp. vii-xii. Classics in Southeastern
Archaeology, Stephens Williams, general editor. The University of Alabama
Press, Tuscaloosa.
2003b Archaeological Excavations at Little River: The 1998-2000 Seasons. LAMAR
Institute. LAMAR Institute Publication 49.
2003c Test Excavations at the Leah and Zack Site, 9PM1182. LAMAR Institute.
LAMAR Institute Publication 74.
2004
Archaeological Excavation at the Little River Site: The 2001 Season. LAMAR
Institute. LAMAR Institute Publication 66.
2005a 4,000 Years at a Glance: Patterns of Ceramic Style Distribution over Georgia.
Early Georgia 33(2):181-190.
361
Williams, Mark (cont.)
2005b Archaeological Testing and Excavations at the Bullard Bottom Site, 9PM169,
2002-2004. LAMAR Institute. LAMAR Institute Publication 69.
Williams, Mark and Daniel T. Elliott (editors)
1998 A World Engraved: Archaeology of the Swift Creek Culture. University of
Alabama Press, Tucscaloosa.
Williams, Mark and Don Evans
1993 Archaeological Excavations at the Bullard Landing Site (9TW1) Publication 24.
Lamar Institute, Watkinsville.
Williams, Mark and Gary Shapiro
1990a Paired Towns. In Lamar Archaeology: Mississippian Chiefdoms in the Deep
South, edited by Mark Williams and Gary Shapiro. University of Alabama Press,
Tuscaloosa.
1990b Archaeological Excavations at Little River (9MG46), 1984 & 1987. Lamar
Institute Reports 2. Lamar Institute, Watkinsville.
1996
Mississipian Political Dynamics in the Oconee Valley, Georgia. In Political Structure
and Change in the Prehistoric Southeastern United States, edited by John F. Scarry, pp.
128-149. University of Florida Press, Gainesville.
Williams, Mark and Gary Shapiro (editors)
1990 Lamar Archaeology: Mississippian Chiefdoms in the Deep South. University of
Alabama Press, Tuscaloosa.
Williams, Mark and Victor Thompson
1999 A Guide to Georgia Indian Pottery Types. Early Georgia 27(1):1-154.
Wood, Malcom Jared and Gregory Lucas
2005 Land-Use Change and Impact on Archaeological Sites in Georgia. Early Georgia
33(1):45-66.
Wood, W. Dean
1995 Woodland Period Archaeology of Northern Georgia. University of Georgia
Laboratory of Archaeology Series Report Number 33. Department of
Anthropology, University of Georgia, Athens.
Worth, John E.
1988 Mississippian Occupation of the Middle Flint River. Department of
Anthropology, University of Georgia. Unpublished Master Thesis.
362
Worth, John E.(cont.)
1989 Mississippian Mound Centers along the Chickasawhatchee Swamp. LAMAR
Briefs 13:7-8.
1996
Middle Mississippian Occupation of the Ocmulgee Big Bend Region. Early
Georgia 24(1):34-83.
1998
The Timucuan Chiefdoms of Spanish Florida, Volume 1: Assimilation. The Ripley
P. Bullen Series. The University Press of Florida, Gainesville.
Worth, John E. and W. Maxwell Duke
1991 Hogcrawl Creek: Early Mississippian Period Occupation in the Middle Flint River
Valley. Early Georgia 19(1):19-42.
Yoffee, Norman
1993 Too Many Chiefs? (Or, Safe Texts for the 90s). In Archaeological Theory: Who
Sets the Agenda?, edited by Norman Yoffee and Andrew Sherrat, pp. 60-79.
Cambridge University Press, Cambridge.
Yoffee, Norman, Suzanne K. Fish and George R. Milner
1999 Communidades, Ritualities, Chiefdoms: Social Evolution in the American
Southwest and Southeat. In Great Towns and Regional Polities in the Prehistoric
Southwest and Southeast, edited by Jill E. Neitzel, pp. 261-272. Amerind
Foundation New World Studies Series. vol. No. 3. University of New Mexico
Press, Albuquerque.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement