THE ARCHAEOLOGY OF HOUSES AND HOUSEHOLDS IN THE

THE ARCHAEOLOGY OF HOUSES AND HOUSEHOLDS IN THE

THE ARCHAEOLOGY OF HOUSES AND HOUSEHOLDS IN THE NATIVE SOUTHEAST by BENJAMIN A. STEERE (Under the direction of Stephen A. Kowalewski) ABSTRACT This re...

18MB Sizes 0 Downloads 22 Views

THE ARCHAEOLOGY OF HOUSES AND HOUSEHOLDS IN THE NATIVE SOUTHEAST by BENJAMIN A. STEERE (Under the direction of Stephen A. Kowalewski) ABSTRACT This research is a preliminary attempt to understand how houses and households changed in the Southeastern United States from the Woodland to the Historic Indian Period (ca. 200 B.C. to A.D. 1800). The primary study area is the Southern Appalachian region, with comparative cases from farther afield. Most studies of domestic architecture in the Southeast are conducted at the single-site scale. Broader patterns of variation in houses and households are not well understood. To address this problem, I use a database that catalogs the architectural features of 1258 structures from 65 sites to identify broad spatial and temporal patterns of variation in domestic architecture, including region-wide changes in the size and spacing of houses, increasing architectural investment, and the increasing partitioning of houses over time. Using a theoretical framework developed from household archaeology and anthropology, I argue that certain aspects of architectural variation can be explained by changes in household economics and household composition, symbolic behavior, status differentiation, and settlement patterning. More generally, I propose that large-scale patterns of diachronic and synchronic variation in domestic architecture are best explained by changes in social organization. INDEX WORDS:

Houses, Household Archaeology, Social Organization, Domestic Architecture, Woodland, Mississippian, Historic Indian, Database, Scale.

THE ARCHAEOLOGY OF HOUSES AND HOUSEHOLDS IN THE NATIVE SOUTHEAST

by

BENJAMIN A. STEERE

B.A., Wake Forest University, 2003

A Dissertation Submitted to the Graduate Faculty of the University of Georgia in Partial Fulfillment of the Requirements for the Degree

DOCTOR OF PHILOSOPHY

ATHENS, GEORGIA 2011

©2011 Benjamin A. Steere All Rights Reserved

THE ARCHAEOLOGY OF HOUSES AND HOUSEHOLDS IN THE NATIVE SOUTHEAST

by

BENJAMIN A. STEERE

Electronic Version Approved: Maureen Grasso Dean of the Graduate School The University of Georgia May 2011

Major Professor:

Stephen A. Kowalewski

Committee:

David J. Hally Mark Williams Bram Tucker

DEDICATION

For Elizabeth.

iv

ACKNOWLEDGEMENTS

This work benefited enormously from the guidance and assistance of my committee. From my first days in the graduate program, Steve Kowalewski challenged me to think big and tackle a bold project, and then provided the direction and support to make it possible. His insightful feedback, challenging questions, frank critiques, and encouragement kept me moving forward. I could not have asked for a better advisor and mentor. David Hally‘s work was instrumental in developing the methods for this project, and like Steve, his feedback and suggestions greatly improved the finished product. I am especially thankful to him for teaching his Household Archaeology seminar as a readings course for Emily Beahm and me. That course provided the backbone of the theoretical and methodological framework for this dissertation. Mark Williams and I had the chance to work on houses together in the field at the Copeland site, and our discussions there and in the archaeology lab kept me tuned in to the practical problems associated with structure excavation. Mark also provided me with support in the form of hardware, software, and technical training at the Laboratory of Archaeology. He may not remember, but I was sort of a Luddite when I entered the program, and I would never have attempted a complex, database-driven project without his help. Bram Tucker encouraged me to think about the project in broad anthropological terms, and his comments helped improve the organization of the dissertation. I left his classes convinced that anthropological research could still be empirical without reducing people to

v

automatons. His Economic Anthropology seminar introduced me to new ways of thinking about houses and households. I would also like to thank Betsy Reitz. Her Environmental Archaeology course introduced me to method and theory that helped this project. As graduate coordinator, she helped me and many of my peers progress through the program in a timely manner and secure funding for our research. Likewise, the staff of the Department of Anthropology enabled me to meet deadlines and secure funding and awards. Margie Floyd helped me through every stage of the program, and offered kind encouragement every step of the way. LaBau Bryan, Lisa Norris, Deb Chasteen, Jill Morris, Curtis Combs, and Cabe Mottley assisted with administrative, technical, and financial hitches. Our department is fortunate to have such a competent and congenial staff. Good friends and colleagues in the Department of Anthropology provided camaraderie and inspiration, and prevented me from developing tunnel vision. I would especially like to thank Emily Beahm, Sarah Bergh, Dan Bigman, Steffan Brannan, Carol Colaninno-Meeks, Vicki Dekle, Amber and Pat Huff, Ellen and John Turck, Yanxi Wang, and Jared Wood for making my time in Athens more productive and fun. I am so glad that John Chamblee came back to Georgia. He provided invaluable instruction and insight for the construction of the database. I got into the house business as field technician with TRC, Inc. while working on the Ravensford project in Cherokee, North Carolina. Tasha Benyshek, our field director, encouraged me to pursue this research. So did our manager, Paul Webb. Paul and Tasha have both provided unpublished data and feedback that helped this research a great deal. I would also like to acknowledge my good friends and colleagues from the Ravensford project. John Kessler, Mandy Terkhorn, Hannah Guidry, Mike Hayden, Michael ―Chief‖ Griffin, Phoebe Gilbert, Chris

vi

Ciancibelli, Jacob Turner, Nicole Coomer, Bill Duckworth, Mike Fisher, Bryan Jackson, Sterling Howard, and Heather McAllister taught me so much about archaeology, and were like family to me in Cherokee and on other projects in North Carolina. Many of these data were hard to come by. I especially thank Aaron Deter-Wolf at the Tennessee Division of Archaeology and Susan Myers at the North Carolina Office of State Archaeology for providing access to excavation records from Tennessee and North Carolina. The interlibrary loan staff at the University of Georgia helped me track down several sources. In Athens, Joel Jones, Michael ―Chief‖ Griffin, Jerald Ledbetter, and Scott Jones of Southeastern Archeological Services provided insight and ideas on houses during informal talks at the archaeology lab and on Chief‘s front porch. Thanks especially to Jerald for supplying unpublished data. My family has always provided loving support, and especially during the graduate school years. Thanks to Jonathan, Sarah, Matt, Allen, Richard, Joanna, and Kate Steere for the encouraging calls and fun reunions on holidays. Last but not least, I would like to thank my wife, Elizabeth Lee Steere, for her loving kindness and support. She is the best partner, friend, directeur sportif, and copyeditor I could ever want. Elizabeth took time to read and comment on much of my work, even as she was working on her own dissertation. I could not have done this without her.

vii

TABLE OF CONTENTS Page DEDICATION .............................................................................................................................. iv ACKNOWLEDGEMENTS........................................................................................................... v LISTS OF TABLES ...................................................................................................................... x LISTS OF FIGURES .................................................................................................................... xii CHAPTER 1

INTRODUCTION ................................................................................................. 1

2

RESEARCH FRAMEWORK ................................................................................ 7

3

METHODS ............................................................................................................ 36

4

RESULTS: ARCHITECTURAL VARIABILITY .................................................. 64

5

ENVIRONMENTAL FACTORS ........................................................................... 153

6

HOUSEHOLD ECONOMICS AND HOUSEHOLD COMPOSITION .................. 180

7

HOUSES AS SYMBOLS ...................................................................................... 218

8

HOUSES AND STATUS ....................................................................................... 252

9

HOUSES AND SETTLEMENT FUNCTION ........................................................ 281

10

CONCLUSION...................................................................................................... 310

REFERENCES CITED ................................................................................................................. 326 APPENDICES A. SITE DATA TABLE .................................................................................................... 351 B. COMPONENT DATA TABLE..................................................................................... 356

viii

C. STRUCTURE DATA TABLE ...................................................................................... 360 D. PERIOD DATA TABLE .............................................................................................. 452 E. REFERENCES DATA TABLE..................................................................................... 457 F. SITE PLANS ................................................................................................................. 465

ix

LIST OF TABLES Page Table 3.1: Sites Recorded in the Database ..................................................................................... 50 Table 3.2: Components Recorded in the Database ......................................................................... 52 Table 4.1: Summary Statistics for All Structures ........................................................................... 117 Table 4.2: Summary Statistics for Middle Woodland Domestic Period Structures ......................... 117 Table 4.3: Summary Statistics for Late Woodland Domestic Period Structures.............................. 118 Table 4.4: Summary Statistics for Early Mississippian Period Domestic Structures ....................... 118 Table 4.5: Summary Statistics for Early Mississippian Period Non-Domestic Structures ............... 119 Table 4.6: Summary Statistics for Middle Mississippian Period Domestic Structures .................... 119 Table 4.7: Summary Statistics for Middle Mississippian Period Non-Domestic Structures ............ 120 Table 4.8: Summary Statistics for Late Mississippian Period Domestic Structures......................... 120 Table 4.9: Summary Statistics for Late Mississippian Period Non-Domestic Structures................. 121 Table 4.10: Summary Statistics for Late Mississippian Period Storage Structures.......................... 121 Table 4.11: Summary Statistics for Historic Indian Period Domestic Structures ............................ 122 Table 4.12: Summary Statistics for Historic Indian Period Non-Domestic Structures .................... 122 Table 4.13: Summary Statistics for Historic Indian Period Storage Structures ............................... 123 Table 4.14: Structure Shape, Sorted by Time Period and Functional Class .................................... 124 Table 4.15: Interior Support Post Patterns ..................................................................................... 125 Table 4.16: Structures with Hearths ............................................................................................... 128 Table 4.17: Hearth Types .............................................................................................................. 128 Table 4.18: Mississippian Period Rebuilding Episodes .................................................................. 129 x

Table 4.19: Average Post Density for Structures by Period ............................................................ 130 Table 4.20: Results of Multivariate Regression with Chronological Dummy Variables for Domestic Structures ..................................................................................................................... 131 Table 4.21: Results of Multivariate Regression with Chronological Dummy Variables for NonDomestic Structures ...................................................................................................................... 131 Table 5.1: Environmental Data for Sites ........................................................................................ 173 Table 5.2: Historic Accounts of Building Materials ....................................................................... 175 Table 5.3: Archaeological Evidence for Building Materials ........................................................... 176 Table 5.4: Summary of Environmental Variables for Sites with and without Structures with Basins ........................................................................................................................................... 177 Table 5.5: Results of Multivariate Regression for Post Density and Environmental Variables ....... 177 Table 6.1: Results of Regression with Dummy Variables for Interior Partitioning ......................... 211 Table 6.2: Results of Multivariate Regression with Interaction Variables for Maize Reliance ........ 211 Table 8.1: Summary Statistics for Mississippian Mound and Village Structures ............................ 272 Table 8.2: Results of Wilcoxon Rank-Sum Tests for Differences between Mound and Village Structure Variables ....................................................................................................................... 273 Table 9.1: Results of Regression for Domestic Structure Area ....................................................... 298 Table 9.2: Results of Multivariate Regression for Domestic Structure Area ................................... 299 Table 9.3: Summary Table for Architectural Diversity by Component........................................... 300 Table 9.4: Results of Multivariate Regression for Architectural Diversity ..................................... 304 Table 10.1: Best-Fit Model for Domestic Structure Area ............................................................... 324 Table 10.2: Best Fit Model for Partitions in Domestic Structures................................................... 325

xi

LIST OF FIGURES Page Figure 2.1: Division of Space in a Navajo Hogan .......................................................................... 35 Figure 3.1: Diagram of Database Relationships ............................................................................. 57 Figure 3.2: Map of Sites Recorded in the Database ........................................................................ 58 Figure 3.3: Map of Middle Woodland Components ....................................................................... 59 Figure 3.4: Map of Late Woodland Components ........................................................................... 60 Figure 3.5: Map of Early Mississippian Components..................................................................... 61 Figure 3.6: Map of Middle and Late Mississippian Components ................................................... 62 Figure 3.7: Map of Historic Indian Components ............................................................................ 63 Figure 4.1: Histogram of Middle Woodland Period Domestic Structure Area ................................ 132 Figure 4.2: Box Plots of Domestic Structure Floor Area for Middle Woodland Period Components .................................................................................................................................. 132 Figure 4.3: Histogram of Late Woodland Period Domestic Structure Area .................................... 133 Figure 4.4: Box Plots of Domestic Structure Floor Area for Late Woodland Period Components .. 133 Figure 4.5: Histogram of Early Mississippian Period Domestic Structure Area .............................. 134 Figure 4.6: Box Plots of Domestic Structure Floor Area for Early Mississippian Period Components .................................................................................................................................. 134 Figure 4.7: Histogram of Early Mississippian Period Non-Domestic Structure Area ...................... 135 Figure 4.8: Box Plots of Non-Domestic Structure Floor Area for Early Mississippian Period Components .................................................................................................................................. 135 Figure 4.9: Histogram of Middle Mississippian Period Domestic Structure Area ........................... 136 xii

Figure 4.10: Box Plots of Domestic Structure Floor Area for Middle Mississippian Period Components .................................................................................................................................. 136 Figure 4.11: Histogram of Middle Mississippian Period Non-Domestic Structure Area ................. 137 Figure 4.12: Box Plots of Non-Domestic Structure Floor Area for Middle Mississippian Period Components .................................................................................................................................. 137 Figure 4.13: Histogram of Late Mississippian Period Domestic Structure Area ............................. 138 Figure 4.14: Box Plots of Domestic Structure Floor Area for Late Mississippian Period Components .................................................................................................................................. 138 Figure 4.15: Histogram of Late Mississippian Period Non-Domestic Structure Area ..................... 139 Figure 4.16: Box Plots of Domestic Structure Floor Area for Late Mississippian Period Components .................................................................................................................................. 139 Figure 4.17: Histogram of Late Mississippian Storage Structure Area ........................................... 140 Figure 4.18: Box Plots of Storage Structure Floor Area for Late Mississippian Period Components .................................................................................................................................. 140 Figure 4.19: Histogram of Historic Indian Period Domestic Structure Area ................................... 141 Figure 4.20: Box Plots of Domestic Structure Floor Area for Historic Indian Period Components . 141 Figure 4.21: Bar Chart of Average and Median Floor Area for Domestic Structures ...................... 142 Figure 4.22: Histogram of Structure Orientation for Late Woodland Period Domestic Structures .. 142 Figure 4.23: Histogram of Structure Orientation for Early Mississippian Period Domestic Structures ...................................................................................................................................... 143 Figure 4.24: Histogram of Structure Orientation for Early Mississippian Period Non-Domestic Structures ...................................................................................................................................... 143

xiii

Figure 4.25: Histogram of Structure Orientation for Middle Mississippian Period Non-Domestic Structures ...................................................................................................................................... 144 Figure 4.26: Histogram of Structure Orientation for Late Mississippian Period Domestic Structures ...................................................................................................................................... 144 Figure 4.27: Histogram of Structure Orientation for Late Mississippian Period Non-Domestic Structures ...................................................................................................................................... 145 Figure 4.28: Histogram of Structure Orientation for Late Mississippian Period Storage Structures 145 Figure 4.29: Histogram of Structure Orientation for Historic Indian Period Domestic Structures ... 146 Figure 4.30: Histogram of Structure Orientation for Historic Indian Period Non-Domestic Structures ...................................................................................................................................... 146 Figure 4.31: Bar Chart of Average and Median Number of Wall Posts for Domestic Structures .... 147 Figure 4.32: Box Plots of Domestic Structure Wall Posts for Middle Woodland Period Components .................................................................................................................................. 147 Figure 4.33: Box Plots of Domestic Structure Wall Posts for Late Woodland Period Components. 148 Figure 4.34: Box Plots of Domestic Structure Wall Posts for Early Mississippian Period Components .................................................................................................................................. 148 Figure 4.35: Box Plots of Domestic Structure Wall Posts for Middle Mississippian Period Components .................................................................................................................................. 149 Figure 4.36: Box Plots of Domestic Structure Wall Posts for Late Mississippian Period Components .................................................................................................................................. 149 Figure 4.37: Box Plots of Domestic Structure Wall Posts for Historic Indian Period Components . 150 Figure 4.38: Bar Chart of Average and Median Wall Post Spacing for All Time Periods ............... 150 Figure 4.39: Bar Chart of Average and Median Wall Post Diameter for All Time Periods ............. 151

xiv

Figure 4.40: Bar Chart of Average and Median Number of Partitions in Domestic Structures ........ 151 Figure 4.41: Bar Chart of Average Number of Interior Burials for All Time Periods ..................... 152 Figure 5.1: Major Physiographic Regions of Eastern North America ............................................. 178 Figure 5.2: Braun‘s Forest Regions for Eastern North America ..................................................... 179 Figure 5.3: Location of Sites with Structure Basins ....................................................................... 179 Figure 6.1: Model of Household Clusters for Each Chronological Period ...................................... 212 Figure 6.2: Examples of Middle Woodland Period Domestic Structures ........................................ 213 Figure 6.3: Examples of Late Woodland Period Domestic Structures ............................................ 214 Figure 6.4: Examples of Early Mississippian Period Domestic Structures ...................................... 215 Figure 6.5: Examples of Middle Mississippian Period Domestic Structures ................................... 216 Figure 6.6: Examples of Late Mississippian Period Domestic Structures ....................................... 217 Figure 6.7: Examples of Historic Indian Period Domestic Structures ............................................. 217 Figure 7.1: Middle Woodland Circular Structures from the McFarland Site................................... 249 Figure 7.2: Rectangular Late Woodland Structure from 1PI61....................................................... 249 Figure 7.3: Circular Late Woodland Structures from Brasstown Valley ......................................... 250 Figure 7.4: Late Mississippian Winter House Layout .................................................................... 251 Figure 7.5: Historic Indian Winter House Layout .......................................................................... 251 Figure 8.1: Box Plots of Area for Mound and Village Structures ................................................... 274 Figure 8.2: Bar Chart of Average and Median Area for Mound and Village Structures .................. 274 Figure 8.3: Box Plots of Area for Mound and Village Structures at Bessemer ............................... 275 Figure 8.4: Box Plots of Area for Mound and Village Structures at Dallas ..................................... 275 Figure 8.5: Box Plots of Area for Mound and Village Structures at Dyar Mound ........................... 276

xv

Figure 8.6: Box Plots of Area for Mound and Village Structures at Hiwassee Island, Hiwassee Island Phase .................................................................................................................................. 276 Figure 8.7: Box Plots of Area for Mound and Village Structures at Jewell ..................................... 277 Figure 8.8: Box Plots of Area for Mound and Village Structures at Kincaid, Middle Kincaid Phase ............................................................................................................................................ 277 Figure 8.9: Box Plots of Area for Mound and Village Structures at Little Egypt ............................ 278 Figure 8.10: Box Plots of Area for Mound and Village Structures at Toqua, Hiwassee Island Phase ............................................................................................................................................ 278 Figure 8.11: Box Plots of Area for Mound and Village Structures at Toqua, Dallas Phase ............. 279 Figure 8.12: Box Plots of Area for Mound and Village Structures at Town Creek, Early Town Creek Phase .................................................................................................................................. 279 Figure 8.13: Box Plots of Area for Mound and Village Structures at Town Creek, Late Town Creek to Early Leak Phase ............................................................................................................ 280 Figure 8.14: Eighteenth-Century Cherokee House Pairs at Ravensford .......................................... 280 Figure 9.1: Structure Area at Middle Woodland Components, Sorted by Number of Structures ..... 305 Figure 9.2: Structure Area at Late Woodland Components, Sorted by Number of Structures ......... 305 Figure 9.3: Structure Area at Early Mississippian Components, Sorted by Number of Structures... 306 Figure 9.4: Structure Area at Middle Mississippian Components, Sorted by Number of Structures ...................................................................................................................................... 306 Figure 9.5: Structure Area at Late Mississippian Components, Sorted by Number of Structures .... 307 Figure 9.6: Structure Area at Historic Indian Components, Sorted by Number of Structures .......... 307 Figure 9.7: Scatterplot of House Area and Component Area .......................................................... 308 Figure 9.8: Scatterplot of House Area and Number of Structures ................................................... 308

xvi

Figure 9.9: Scatterplot of House Area and Number of Structures ................................................... 309

xvii

CHAPTER 1 INTRODUCTION

On a windy morning in December, 2004, I stood around the excavated post pattern of an early eighteenth-century Cherokee winter house with a dozen students from the Cherokee High School‘s Advanced Placement history class. At the time, I was working as a field technician for TRC, Inc. at the Ravensford site on the Qualla Boundary in Cherokee, North Carolina. Our field director had asked me to give a site tour to visiting students. By that point I had learned that houses were the place to start a tour. Many of the social phenomena we examine as archaeologists—subsistence strategies, settlement patterns, social organization, or political complexity—can be abstract and remote to the average site visitor, but everyone can relate to houses. The map of an excavated post pattern can be read like an architect‘s floor plan. Walking across the surface of an excavated structure is not unlike crossing the threshold of a small cabin. It is an intimate experience. The students were quiet, so I tried to get them talking by asking leading questions. What materials do you think people used to build this house? How could you tell? What sort of activities went on inside the house? After a while, the students began to ask questions of their own. They wondered how many people could comfortably sleep, eat, and work in the house. One student suggested that the quarters would have been cramped, but also warm in the winter. With minimal prompting, these students were making good connections between archaeological data and human behavior.

1

As a field technician, I helped excavate several of the 110 well-preserved Woodland, Mississippian, and Cherokee structures at Ravensford. My colleagues and I spent days trying to identify, delineate, map, and excavate complex post patterns. I spent hours removing roof fall from the surface of burned beams. This immersion in structure excavations made me want to learn more about the connections between the fragile remains of these houses and the people who built them. My experience with house excavation and the realization that anthropological studies of houses and households were accessible to people outside of archaeological circles led me to study prehistoric domestic architecture. As a graduate student searching for a way to make a contribution to the field, I found there were plenty of data for answering questions about people and houses in the Southeast. Houses have long been a focus of major excavations, and researchers here have made significant contributions to household archaeology for decades. What I found wanting were comparative studies of prehistoric houses at a broad spatial and temporal scale. Beyond the scale of a single site or small, well-studied region, it was hard to tell if houses at a particular site were relatively large or small, simple or complex, and whether or not diachronic changes in house form were typical or extraordinary. My focus expanded from wanting to understand architectural variability at single site to understanding architectural variability at a broader scale. This research will be the first attempt to understand how prehistoric houses and households have changed across the Southern Appalachian region of the Southeastern United States from the Woodland to the Mississippian to the Historic period. Most studies of prehistoric houses in the region have been conducted at the single-site scale. Archaeologists in the Southeast have an in-depth understanding of domestic architecture at many individual sites, but wider

2

patterns of variation are not well understood. How did prehistoric houses in the Southeast change in shape, size, layout, and permanency? How did houses at small, dispersed sites differ from houses at large, regional centers? What are the relationships between changes in domestic architecture and the social groups—the households—that made and lived in houses? What social causes underlie differences in house form across space and time? The goal of this dissertation is to address these questions by comparing the architectural features of prehistoric houses in the Southeast from the Woodland, Mississippian, and Historic Indian periods (ca. 1000 B.C. to A.D. 1750). The study is informed by a theoretical framework drawn from research in household anthropology and archaeology, which show that houses reflect the activities, composition, status, and symbolic systems of social groups. The house is a behaviorally meaningful unit of analysis that can be used to answer questions about household activities and economics, household composition and social organization, status differentiation, and symbolism and ideology. The data I use for this research are published and unpublished maps, drawings, and field notes describing well-excavated houses. These records are currently scattered in state archaeology offices, museums, and universities. Using these documents, I have measured and recorded the size, shape, degree of interior segmentation or partitioning, alignment, and spacing of houses. I recorded the size, shape, and spacing of wall posts to allow for fine-scale comparisons of house construction. I have also recorded associated features, including burials, hearths, and storage pits to determine what sorts of activities occurred in and around the structures. Site-level variables such as site size, dates of occupation, mounds, earthworks, and plazas, shed light on the relationships between the organization and function of sites and house form.

3

I have organized these data, many of which were buried in unpublished manuscripts and technical reports with limited circulation, into a Microsoft Access database. I have prepared high-quality scale drawings of nearly every recorded structure (for a few houses only written descriptions exist). These drawings have been included as image files in the database, facilitating visual comparisons of the houses. Geographic coverage is broad to allow for comparisons of changing house form through time in communities, regions, and across the Southeast. The Southern Appalachians are the heart of the study area, and cases from farther afield are used for comparison. The database contains detailed descriptions of over 1200 houses from over 60 well-documented sites in North Carolina, Georgia, Alabama, Tennessee, Kentucky, Missouri, Indiana, and Illinois. With these data, it is possible to address questions about prehistoric houses and households at a scale that has rarely been attempted in the region. In the chapters that follow, I develop a context for this research. In Chapter 2 I argue that while our understanding of houses and households has improved dramatically over the past few decades, it is still primarily limited to a single-site or regional perspective. I then outline a theoretical framework for a broad-scale comparative study of households. Most studies of houses and households build theory from the bottom up, developing models from small case studies and applying them at larger scales. I argue that we also need top-down method and theory to understand architectural variability. Studies of houses and households show that there are no simple, univariate explanations for architectural variability, so we also need theory that considers multiple social reasons for variation in house form. In Chapter 3, I describe the methods used to construct the architectural database and define the variables. Chapter 4 is comparative. There I describe the major patterns of

4

architectural variation that emerge from comparing houses at multiple spatial and temporal scales. Perhaps the most striking finding of this research is the spatial scale of the patterns of architectural variability. At various times from the Woodland to the Historic period houses are remarkably similar across geographic regions, and widespread changes in domestic architecture happen fairly quickly. Archaeological patterns such as the reduction in house size from the Middle Woodland to the Late Woodland period, the occurrence of wall trench houses all over the Southeast during the Early Mississippian period, and the increasing segmentation of houses during the Mississippian period, cannot be explained by local and regional processes alone. After establishing the major synchronic and diachronic patterns of variation in house form, I pose a series of questions about the causes of this architectural variability, and attempt to answer them with archaeological data and insights from ethnohistoric accounts, ethnoarchaeology, and cross-cultural research. There are many environmental and social reasons for variation in houses, but for each question, I build an argument for why it is more one thing than another. In Chapters 5 through 9, I ask different questions about variation in houses and answer them with appropriate variables, using the theoretical approaches outlined in Chapter 2. I hold some things equal so I can focus on a manageable number of causal factors at a time. In Chapter 5, I examine the role of environmental factors. In Chapter 6, I ask to what extent variation in houses and household organization is explained by changes in domestic production and consumption. In Chapter 7, I suggest that conceptualizing houses as symbols used to express cosmological beliefs helps explain variation in domestic architecture. In Chapter 8, I consider the effect of status differentiation. I Chapter 9, I ask to what degree variation in houses and households can be explained by a settlement‘s place in a regional settlement system or polity.

5

Finally, in Chapter 10, I return to the results of Chapters 5 through 9 and synthesize the findings of this study. I argue that certain social factors are better than others for explaining particular facets of architectural variability. I then propose that addressing questions about domestic architecture at a broad spatial and temporal scale leads to a general theoretical statement about the relationship between domestic life and architecture. I posit that broad-scale changes in architecture are best explained by broad-scale changes in social organization. Considering the multiple social factors outlined above provides theoretical expectations for the nature of these changes. The spatial and temporal scale of architectural variability will depend on the size, permanence, and degree of integration of social groups larger than households, i.e. the lineages, clans, chiefdoms, and other corporate groups to which households belong. Households in the native Southeast were part of a large, macroregional system that existed since Paleoindian times. In order to understand architectural variability in the region we need an approach that considers the society-wide processes that explain variation in houses and households.

6

CHAPTER 2 RESEARCH FRAMEWORK

Of all the classes of material culture archaeologists study, architecture is one of the most useful for understanding how the broad social world affects domestic life. Houses are both utilitarian and symbolically charged objects. People use houses as domiciles and workshops, yet they often build them according to cosmological principles. Individuals and small social groups make conscious and creative decisions about house design, but their choices are constrained by top-down societal forces. Understanding architectural variability requires theory that takes into account different social reasons for variation in house form and considers spatial and temporal scale. Most studies of houses and households have been conducted at the single-site scale. Archaeologists have an in-depth understanding of architectural variability at the scale of an individual site or region, but wider patterns of variation are often not understood. Large-scale changes in domestic architecture, such as the shift from round to rectangular houses during the Woodland to Mississippian transition, need to be explained by social processes operating at a macroregional scale. Household archaeology is not a unified theory. It is a subfield defined by its focus on the household as the primary unit of analysis (Wilk and Rathje 1982). Research in household anthropology and archaeology has mostly been carried out at small spatial and temporal scales, and has not been related to a coherent theory that delineates and weighs the different social factors that explain architectural variation. 7

However, these studies have identified the most important causes of architectural variability, the social behaviors which at different scales account for most of the variation in house form. Decades of research in household anthropology and archaeology, reviewed here, demonstrate that most of the variation we observe in houses is explained by variation in household economics and household composition, ritual and symbolic behavior, status differentiation, and settlement patterning. Household studies tell us which questions to ask about people and houses, and give us methods for linking the material remains of houses to human behavior. We can use methods and insight from household archaeology to build a framework for a broad-scale, comparative study of domestic architecture. The first step is to compile a large sample of houses, one big enough to identify meaningful patterns of synchronic and diachronic variation over a large geographic area. The next step is to consider the effect of the environment on house form. What are the major environmental and ecological constraints on house building? Finally, we ask to what extent the patterns can be explained by each of the most important social factors. Momentarily holding other factors equal, we ask how much of the variation in houses can be explained by economic behavior, by symbolism, by status differentiation, and by settlement patterning. In each case, some architectural variation may be explained by local, small-scale causes, but some patterns will require explanation in terms of broader processes. This approach could be used anywhere, but the Southeastern United States is an especially good place to start. Houses have been the focus of archaeological studies since the early twentieth century, and the architectural features of thousands of Woodland, Mississippian, and Historic Indian houses have been recorded. They have not been compiled and compared at a broad scale. In a recent discussion of architectural variability in the Southeast, Knight (2007:187)

8

critiques the study of domestic architecture in the region, stating, ―in general, our colleagues have gotten away from the broad-scale comparative studies of material culture of the kind that might reveal the geography of housing traditions. This trend, I think, is to our detriment.‖ Decades ago compiling an architectural database with broad geographic coverage would have been an extremely difficult task, but thanks to technological improvements, increasingly accessible data, and the recent publication of large architectural data sets, this kind of research is less daunting. In this chapter I develop a theoretical framework for a broad-scale comparative study of domestic architecture. I demonstrate that decades of research in household archaeology and anthropology point to four social factors that can explain architectural variability. The review of ethnographic literature is not systematic, but draws from heavily from cases that provide possible analogs for the social and economic conditions of the prehistoric Southeast. I then review the history of household research in the Southeast, and argue that this approach can shed new light on our understanding of domestic life in the region. Household Archaeology: Identifying Major Causes of Architectural Variability Research in household anthropology and archaeology has improved our understanding of the social processes that explain architectural variability in households and communities, but has done less to describe and explain broader patterns of variation in domestic architecture. Largescale changes in domestic architecture must be explained by large-scale social processes. As Kowalewski has argued, archaeologists often ―tend to favor bottom-up, locality-to-macroregion method and theory, but top-down, macroregion-to-locality method and theory provide their own insights‖ (1995:149).

9

Anthropological studies of houses and households should consider causal factors operating at different spatial and temporal scales. The call for attention to scale is not a new one, with proponents in various disciplines, including history (Braudel 1979), sociology (Giddens 1979), geography (Meentemeyer and Box 1987), ecology (Wu and Loucks 1995), and anthropology (Blanton et al. 1993; Smith 1984; Kowalewski 1995). We can think of households, communities, polities, and regional systems operating on a continuum of increasingly large socioeconomic and political and temporal scales. We might anticipate different explanations for the behaviors and processes we see at each level (Kowalewski 1995:155). Explanations of architectural variability must also account for the interplay of structure and agency (Giddens 1979; Weissner 2002). Many recent studies of houses and households use practice theory (Wilson 2008) or refer to Bordieu‘s (1977) concept of habitus to understand this relationship (Blanton 1994; Emerson 1997). On the one hand, people consciously use houses to communicate their social standing, and in a community or region subtle differences in the size and exterior decoration of houses are meaningful. Yet in communities and regions houses generally conform to a similar architectural style, and historical and archaeological studies of architecture demonstrate that coherent building styles change over time. This sort of conformity is hard to explain with strictly bottom-up, agent-based explanations. In House Form and Culture, Rapoport (1969), undertook a world-wide comparative study of architecture using a cross-cultural sample of houses. Rapoport‘s goal was to understand the relationship between culture and architectural style, and to develop a general theory to explain architectural variability. His approach was pioneering, but not very systematic, and his data were not sufficient for empirical analyses. However, in his theoretical framework he makes the important observation that ―house form is not simply the result of physical forces or any

10

single causal factor, but is the consequence of a whole range of socio-cultural factors seen in their broadest terms‖ (Rapoport 1969:47). Rapoport‘s (1969) contribution lies in his recognition that multiple social forces account for variation in house form. They are many different social reasons for variation in houses, and a theory aimed at explaining architectural variability at a broad scale cannot hinge on a single social variable. Strictly materialist explanations of architectural variability will omit important ritual and symbolic meaning embodied in houses, while ideological approaches might ignore intimate, flesh and blood realities that affect house form, such as the expansion and contraction of household size with the domestic cycle. I propose instead to consider the multiple social factors that affect house form one at a time, momentarily setting other factors equal. Research in household anthropology and archaeology, carried out primarily at small scales, shows that four big social factors are the most important for understanding variation in house form. Variation in houses is best explained by: (1), variation in household economics and household composition; (2), ritual and symbolic behavior; (3), status differentiation; and (4), settlement patterning. By tackling each of these factors independently, and then synthesizing those findings with attention to the scale at which different social processes operate, we can generate more nuanced explanations for variation in houses. Household Composition and Household Economics. A key contribution of the anthropological and archaeological studies of households in the 1980s was the movement to define households based on the activities that household members perform (Wilk and Rathje 1982; Wilk and Netting 1984). The household, as defined by Wilk and Netting (1984:2) is the most basic unit of consumption, production, distribution, transmission, reproduction, and

11

coresidence in human society. Twenty-five years later this definition is still useful for archaeologists, as it provides a framework for identifying meaningful social groups based on the material remains of their daily activities. Defining households based on what household members do allows for such complications as the presence of non-kin members in households (Bender 1967; Gonzalez 1969; Yang 1945; Yanagiasko 1979), and the division of kin into different households (Crocker 1969). The economic activities performed by households and the composition of households affect the size, shape, and interior layout of houses and their spatial arrangement in communities. Household composition can be flexible (Gonzalez 1969; Cliggett 2005), and households do not always act as cohesive decision-making units (Wilk and Cliggett 2007:180). Households grow and contract with the domestic cycle and houses are modified as household membership changes (Goody 1962). This can result in a confusing picture for archaeologists. In a single community some architectural variability will reflect households at different stages in the domestic cycle. However, despite these complications, many aspects of household composition and social organization, such as household size and post-marital residence patterns, can be inferred from domestic architecture (Blanton 1994; Kent 1990; Kramer 1982; Oswald 1987). This is especially significant for research in the Southeast, where changes in household composition from the Woodland to the Mississippian to the Historic Indian period are still poorly understood. Archaeologists have had some success in reconstructing prehistoric household composition. Although Hill (1968) was criticized for relying on Murdock‘s (1949) assumptions regarding correlations between residence and descent (Allen and Richardson 1971), he compared the size and shape of room blocks and the distribution of domestic and ritual artifacts to define

12

activity areas and possible household units at Broken K Pueblo. At Tikal, Haviland (1988) not only identified household units, but also traced the developmental cycle of a group of households in a Classic Mayan courtyard group, demonstrating that changes in household composition over time could be traced through changes in domestic architecture, such as rebuilding and renovation. Kapches's (1990) comparative, diachronic study of Iroquois longhouses links changes in the occupation span and internal segmentation of structures to the development of the Iroquois matrilocal residence system. Identifying the absolute size of households from archaeological remains is a difficult task, but comparative studies exploring social causes for variation in house and household size have been productive. Formulas for generating absolute headcounts for households from floor area and artifact counts have been developed, but have rarely held up to scrutiny (LeBlanc 1971; Naroll 1962). Studies that compare the sizes of houses and households within communities and regions have been more successful (Wilk 1983; Wilk and Netting 1984). In traditional societies, larger houses are generally correlated with larger households, which are often indicative of comparatively greater wealth or social standing (Wilk 1983). In the Southeast, recent studies have improved our understanding of Mississippian household composition and social organization through the study of architecture. At the King site (Hally 2008; Hally and Kelly 1998), a recurring pattern of two or three square domestic structures and rectangular storage facilities around small, open work areas has been interpreted as a household cluster. These household clusters were most likely ―occupied by extended families with each square structure being occupied by a separate ‗nuclear family‘‖ (Hally and Kelly 1998:57). Based on ethnohistoric accounts of matrilineal and matrilocal residence groups, Hally and Kelly (1998:61) argue that these extended households would have formed ―the core of

13

corporate matrilineal descent groups.‖ At the Warren Wilson site, Moore (2002) has identified similar Mississippian household clusters. At large sites such as Town Creek (Boudreaux 2005) and Coweeta Creek (Rodning 2004), researchers have examined change and continuity in architecture to understand changes in household size and community interaction. Wilson (2008:75) argues that multihousehold residential groups of 10 to 20 domestic structures were ―the basal social and residential units at Early Moundville.‖ Studies like these give us insight into how household composition varies, but they are limited by their single-site focus. At a broader spatial scale, Wilson (2008) argues for regional and site-level variation in household composition during the Mississippian period. In a review of Mississippian household archaeology, Wilson ( 2008:14-15) suggests that in eastern Tennessee, a household cluster composed of a winter house, summer house, small storage buildings, and a clear activity area changed little from A.D. 1300 to after A.D. 1600, despite changes in political organization. In contrast, during the Early Mississippian (A.D. 1050 – 1200) in the American Bottom, there were at least three different patterns of household organization (Wilson 2008:15). Households at rural farmsteads were represented by clusters of one to three structures with a few associated storage pits and a clear activity area. At upland villages, people lived in clusters of domestic structures and storage pits arranged around central courtyards. At mound centers, households were organized into large, multihousehold residential groups (Wilson 2008:15-16). Peregrine (1992) examined changes in houses construction, community layout, and storage patterns in the American Bottom during the Woodland to Mississippian transition. He argues that the change from large, Middle Woodland circular structures with exterior storage pits to smaller, rectangular Late Woodland structures with shared storage pits in common courtyards represents a shift from villages composed of ―cooperating extended families‖ to ―compounds of

14

individuals who formed a single social and economic unit, likely coordinated through a lineage structure‖ (Peregrine (1992:140). The larger, more distantly spaced rectangular houses of the Early Mississippian may represent villages in which individual nuclear or extended families are the basic economic unit. According to Peregrine, ―if there is a single evolutionary trajectory in this pattern, it is that the basic economic unit became attenuated during the Woodland to Mississippian transition. It began as a group of cooperating families, changed into a lineage compound group, and ended up as the individual family unit‖ (1992:141). He argues that changes in household composition are causally related to large-scale shift to intensive maize agriculture its new requirements for the organization of labor (Peregrine 1992:141-142). Certain regular patterns of variation in architecture are related to domestic subsistence production. In an ethnoarchaeological study of 31 Basarwa (―San‖) and Bakgalagadi (Bantu speakers) camps in the Kalahari desert of Botswana, Kent (1992:639-642) found that the anticipated length of site occupation was a stronger predictor of site size, house size, and the frequency of formal storage facilities than the actual length of occupation, ethnicity of the group, subsistence strategy, and the season of occupation. Rather than identifying a simple correlation between architectural investment and subsistence strategy (i.e. foragers build ephemeral houses while farmers build durable houses), Kent (1992) argues that houses with a relatively high degree of architectural investment reflect a longer intended use life. This observation, which is supported by similar observations among Mayan farming communities (Hayden and Cannon 1984, Wilk 1983), may help explain some of the unexpectedly high levels of architectural investment seen in the Middle Woodland period and the outlying cases of low architectural investment in the Mississippian period identified in Chapter 4.

15

In a more recent ethnoarchaeological study of Mikea settlements in Madagascar, Kelly et al. (2005) built on Kent‘s (1992) research by comparing the architectural features of structures from foraging camps, seasonal hamlets, forest hamlets, and villages. Mikea communities practiced a mix of horticulture, pastoralism, foraging, wage labor, and craft production (Kelly et al. 2005:403). The different settlements were associated with particular economic activities, and the structures had varying intended use lives, ranging from very short at the foraging camps to quite long in the villages. Permanent domestic structures with longer intended use lives had higher post densities, more uniform posts, and a more uniform shape than structures intended for shorter occupations (Kelly et al. 2005:411-413). This study has several important implications for Southeastern archaeologists. On the one hand, we can expect that as prehistoric communities invested more heavily in sedentary subsistence strategies, such as maize agriculture, their structures would have become more permanent and uniform. However, even after the adoption of intensive maize agriculture, Southeastern Indians practiced a mixed economy that also included hunting, gathering, and fishing (Hudson 1976; Scarry and Scarry 2005). As a result, we should expect to see a general increase in uniform, well-constructed houses over time from the Woodland to the Mississippian period, but throughout the chronological sequence we might also expect to find more lightly built structures associated with shorter intended use lives. In the Southwest, such a trend has been identified in the pithouse to pueblo transition. Diehl (1997) compared 184 Upland Mogollon pithouses from 15 sites dating between AD 200 and 1000, and argued that an increasing density of posts and investment in more durable architectural material was related to increasing sedentism and dependence on agriculture after AD 550. This finding is supported by an earlier cross-cultural study suggesting that intended

16

occupation duration predicts architectural investment in houses, as indicated by the materials used in construction (Diehl 1992). Ethnographic studies of households point to strong relationships between domestic production and consumption and household size. Wilk and Rathje (1982:631-632) argue that large households are most effective in situations that require ―task simultaneity,‖ such as agricultural systems that requires large, short-term inputs of labor, like field clearing for extensive shifting cultivation. On the other hand, small households are better for production systems that require a ―linear scheduling of labor,‖ such as small-scale intensive agriculture. In some cases, large households may be better equipped to adapt to changing economic conditions because they can quickly marshal more labor. Small households may be better equipped to transmit property between generations (Wilk and Rathje 1982:631-632). Among Kofyar communities in central Nigeria, Netting (1965) noted a strong correlation between small household size and intensive agricultural production versus a tendency for larger, extended households to engage in extensive shifting cultivation. Among Kekchi Maya farmers in Belize, Wilk (1984) discovered that household size and labor strategies (i.e. small, independent households working in small groups versus large, extended households working in communal teams) were determined by land availability. Variation in house size is often, but not always, a reflection of variation in the size of the household. These studies suggest that at least some of the synchronic variation and diachronic change we see in the size of Southeastern houses and household clusters was related to the organization of labor for subsistence production. House form can also reflect economic specialization. In a cross-cultural study of house floor plans and the division of space among 73 societies, Kent (1990) found a general trend for structures to become more internally segmented or partitioned as political complexity and social

17

inequality increased. This segmentation represented the increasing functional specialization of rooms and buildings and in many cases a greater emphasis on separating male and female work and living areas. Taken as a whole, these studies show that certain architectural characteristics of houses and settlements (e.g. the spatial organization of structures in a community) are heavily influenced by economic factors. These include house size and interior partitioning, architectural investment, regularity and uniformity of wall construction, the nature of storage, and the arrangement of houses in a settlement. Household size and household composition can sometimes be inferred by the size and arrangement of houses, and by the spatial relationship between houses, open work areas and plazas, and burials. Houses and Symbolism. Houses may also be thought of as icons representing cultural beliefs that are expressed in mythologies, cosmologies, and ritual (Blanton 1994; Hodder 1984). For example, among Atoni communities in Indonesia, the orientation of the house along cardinal directions, and the division of space into male/female, east/west, and upper/lower areas reflects and reinforces political and religious systems (Cunningham 1973). Among Tukanonan-speaking communities in Amazonia, the design of the house bolsters a social order rooted in local mythology, but the architectural symbolism is polyvalent, and can change dramatically depending on the rituals being performed in the house (Hugh-Jones 1995). In the absence of written records or historic accounts it may be difficult if not impossible for archaeologists to reconstruct such nuanced symbolic associations, but it is crucial that archaeologists consider the possibility that almost any domestic structure will be imbued with complex social, cultural, and ritual meaning.

18

Blanton‘s cross-cultural, comparative study of households reveals that houses can be used for both ―indexical‖ and ―canonical‖ communication (1994:8). Houses are highly visible symbols that communicate social identity and status differences among households, but they also reinforce social orders rooted in cosmologies and ideologies that integrate communities (Blanton 1994:8-13). Several archaeological studies have shown a positive correlation between widely accepted markers of high social status (e.g. fancy ceramics, rare or expensive trade good, and/or ritual paraphernalia) and unusually elaborate domestic structures (Gonlin 1989; Hirth 1993; Smith 1985; Wilk 1983). On the other hand, the outward appearance of houses can also be used to emphasize ―social linkages‖ between households, rather than status differences (Blanton 1994:13). Thinking about houses as symbols that both communicate individuality and reinforce group cohesion may help explain why houses shift from being more similar to more variable at the community and regional level at different times and places in the Southeast. There is strong comparative ethnographic evidence to suggest that seemingly undifferentiated structures are divided conceptually by their occupants. Among Navajo groups from Utah described by Kent (1983, 1984), hogans, circular winter dwellings, are not physically partitioned, but they are ―conceptually divided into men‘s and women‘s areas,‖ while ramadas, rectangular summer structures, are usually not partitioned architecturally or in practice (Kent 1990:133). While living with a Navajo family in a traditional hogan, Kent (1983:83) observed a clear division of the structure into female space in the north and male space in the south (Figure 2.1). While similar domestic activities took place in the ramada, she did not observe a similar conceptual partitioning of space. According to Kent (1983:83-84, 1990:133), the difference in the division of space in hogans and ramadas is ―related to the sacred circular shape of the hogan

19

and the cosmos. Both the hogan and the cosmos are sacred and both are separated into male and female areas, in contrast to the nonsacred rectangular ramada.‖ Several groups in the Amazon basin live in structures that are conceptually divided into gendered or ritually specialized areas (Kent 1990:135). In mid-twentieth-century Jívaro communities located in the foothills of the Andes near the border of Peru and Ecuador, groups of patrilineal, polygynous nuclear families lived in large, oval-shaped, non-partitioned houses measuring approximately 8 by 13 m (Harner 1973:42). On average, nine people occupied a single house (Harner 1973:46). These buildings traditionally have a male end and a female end, with an entrance on each side (Harner 1973:42-44). Male household heads have a bed on the male end of the house for daytime naps, but usually sleep with their wife or wives on the female end of the house (Harner 1973:46). Male visitors are entertained in the male end of the structure, and single men sleep at this end of the house during overnight stays. Visiting married couples, on the other hand, will tend to sleep on the female side of the house (Harner 1973:46). In addition to the daily gendered division of space, the large, open floor plans of Jívaro structures allow them to serve other functions. Harner (1973:45-46) notes that ordinary domestic houses are used for ceremonial feasts and dances that involve many households, and that several families may congregate in a single house to either stage or defend against an attack by enemies. Among mid-twentieth-century Tukanoan communities in the Vaupés region of the Amazon basin, near the border of Columbia and Brazil, groups of 4 to 30 people lived in rectangular longhouses measuring roughly 10 by 30 m (Hugh-Jones 1979:44-48). The inhabitants of a longhouse included a core descent group comprising a headman and his family, and then wives of descent group members and additional kin (Hugh-Jones 1979:40-41). Like the Jívaro structures, these longhouses had distinct male and a female ends and entrances (Hugh-

20

Jones 1979:44-46). In general, the male end of the house would face the river, while the female end of the house would face household gardens (Hugh-Jones 1979:44). Tukanoan longhouses were physically partitioned by screens, with individual nuclear families occupying single compartments along the walls of the longhouse. While family compartments line the exterior walls of the longhouse, the interior is fairly open, allowing space for communal meals and more formal ceremonial dances and gatherings (Hugh-Jones 1979:47-48). Men congregate in the male end of the house for formal nightly meetings, and the space is considered to be public (Hugh-Jones 1979:48-89). Many of the productive tasks for the household take place in the women‘s end of the house, which is separated from the men‘s side of the house by a screen, and considered to be a private area (Hugh-Jones 1979:49). The male-female division of the structure often breaks down during everyday activities, but the division is strictly enforced in the context of large gatherings and ceremonies (Hugh-Jones 1979:49). In the 1970s, Hodder (1987) performed ethnoarchaeological research among Ilchamus communities south of Lake Baringo, Kenya. At that time, these groups practiced pastoralism and agriculture, and people lived in dispersed family compounds (Hodder 1987:426). Ilchamus houses are round huts with a single interior hearth, beds along the walls, and a storage area for grain (Hodder 1987:436). There are no physical partitions, but the houses are conceptually divided into male and female halves. As in the Tukanonan case, the female half of the house is considered to be a private space where productive activities take place (Hodder 1987:436-437). In general, archaeological and anthropological case studies show a strong tendency for houses to be divided into male and female space (Cunningham 1973; Gougeon 2002; Hanson 2003; Lyons 1989). Sometimes this spatial divide is visible, as in the case of distinct

21

distributions of artifacts generally associated with males and females (e.g. ―male‖ flintknapping debris and ―female‖ ceramic debris) at the archaeological sites of Little Egypt in northwest Georgia (Gougeon 2002) and Broken K Pueblo (Hill 1968) and Carter Ranch (Longacre 1968) in Arizona. Recently progress has been made in understanding household symbolism in the Southeast. Sullivan (1987, 1995) suggests that in eastern Tennessee the layout of Mississippian towns mirrored the organization of household clusters. Domestic architecture may have reinforced the same Mississippian cosmologies and principles of social order communicated by the public architecture. Gougeon (2007:137) posits that Late Mississippian structures in northwestern Georgia shared an underlying architectural grammar, accounting for their similarity across the region. Hally (2002:108) argues that sixteenth-century square, semi-subterranean winter houses in the Southern Appalachian region were used ―to symbolically express a number of cosmological and mythological beliefs.‖ The square floor plan of the structures may correspond to the shape of the earth, the four walls and four interior support posts may correspond to the cardinal directions and the sacred number four, seven posts along each wall may correspond to the sacred number seven and the number of clans in Cherokee society, and there are similarities between the houses and prehistoric platform mounds (Hally 2002:108-109). Several scholars have explored the meaning of subfloor burials (Hally 2008; Rodning 2007) and house rebuilding (Boudreaux 2007; Hally 2008; Hally and Kelly 1998; Moore 2002; Rodning 2004, 2007; Wilson 2008) in terms of household identity. These studies provide a structure for connecting architectural data from houses to questions about symbolism and ideology. Several architectural variables may be determined by cosmological principles. These include house size and shape, the layout of houses in

22

communities, and the orientation of houses. Finer architectural variation, such as numbers of wall posts, the placement of interior posts and features, and segmentation within houses may also be explained by symbolic behavior. Status Differentiation. Understanding the relationships between the material remains of houses and status is a major focus of household archaeology. As Pluckhahn (2010:338) notes in his recent review of household archaeology in the Southeast, ―economic and status differentiation constituted the major areas of interest‖ in household studies of the 1980s and early to mid 1990s. Much of the interest in houses and status was part of a broader attempt to document variation in social and political organization, with understanding the development of social ranking and institutionalized inequality as a primary research goal (Emerson 1997; Rogers and Smith 1995). One of the most cogent and oft-cited statements on household status appears in Hirth‘s (1993) discussion of the identification of rank and status from domestic contexts in central Mexico. Hirth makes a clear argument for where and how to look for status differentiation in the archaeological record. He argues that status differences are best understood by looking for patterns of variation within identifiable residential groups (Hirth 1993:122). Second, based largely on cross-cultural and ethnographic evidence, he claims that, ―the three most common archaeological indicators of rank are household size, architectural design, and the content of its associated tool assemblages‖ (Hirth 1993:122). Hirth focuses especially on comparisons of houses, arguing ―residential architecture is perhaps the strongest and most consistent expression of wealth and rank in agrarian societies and it is the product of both a household‘s social functions and energetic capabilities,‖ and that ―the greatest source of household architectural variation derives from the political, religious, and/or economic functions which elites perform

23

within their perspective societies‖ (1993:123). Archaeological studies show a positive correlation between widely accepted markers of high social status (e.g. fancy ceramics, rare or expensive trade good, and/or ritual paraphernalia) and unusually elaborate domestic structures (Gonlin 1989; Hirth 1993; Smith 1985; Wilk 1983). The location of houses within settlements is also an indicator of special status. Accounts from the De Soto expedition describe chiefs living near or on mounds (Biedma quoted in Clayton 1993:239; Elvas quoted in Clayton 1993:73). Members of the DeSoto expedition also note special display goods in these houses (Clayton et al. 1993), suggesting special social, political, and religious activities. Following Hirth (1993), archaeologists working in the Southeast have attempted to identify corporate groups within sites and compare their domestic architecture and household artifact assemblages to find patterns of variation that indicate differences in social status and rank. Gougeon (2002, 2006) examined status variation by comparing three completely excavated Late Mississippian structures at Little Egypt, the center of the paramount chiefdom of Coosa described in the chronicles of the De Soto expedition. Structure 1 is located atop a low mound and is larger than the other two excavated structures. Given its size and location, it is considered to be an elite residence. Structure 1 also contains pigment minerals, while Structures 2 and 3 do not, and faunal evidence suggests residents of Structure 1 may have eaten slightly better cuts of deer and bear than the households represented by Structures 2 and 3 (Gougeon 2006:188). However, the method of construction, the spatial layout of interior activity areas, and botanical remains are similar across all three structures (Gougeon 2006:185-188). It appears the residents of Structure 1 had a higher status position, indicated by the special placement and larger size of their house, and their access to certain items (higher quality meat and pigment for body paint or

24

rituals). Yet from an architectural standpoint, their house was similar to the ―commoner‖ houses, Structure 2 and 3 (2006:190). Gougeon (2006) argues that the similarity in domestic architecture masked status differences within the community, promoting an ideology of community solidarity. Hammerstedt (2005:19-21) recorded the floor area of 728 houses from 21 Mississippian sites across the Southeast and compared the size of mound summit structures (mostly elite residences) to structures in villages (mostly commoner residences). Surprisingly, some smaller sites have mound summit architecture that is similar to or larger than mound summit architecture at bigger sites (Hammerstedt 2005:19-21). While the largest village buildings are generally at the largest sites, the median size of village structures is fairly evenly distributed, and some of the smallest domestic structures come from the sites with three or more mounds (Hammerstedt 2005:21, Figure 9). This study suggests that there is not a simple correlation between status, house size, and site size in the Mississippian period. At Moundville in the Roadway excavation area, Wilson (2008:87-92) documents multihousehold residential groups represented by clusters of 10-20 structures during the Early Moundville I and Late Moundville I-Early Moundville II phases. Within these clusters there are subtle differences in house size, suggesting possible status differences. However, the distribution of house sizes between the residential areas is fairly uniform, suggesting that there are not major status differences between the residential groups. Wilson (2008:129-130) argues that status differences were downplayed in the residential architecture. Studies of houses and social status have also been carried out at Mississippian sites without mounds. At Snodgrass, 93 burned structures were uncovered and mapped. Most of the houses were excavated (O‘Brien and Perttula 2001; Price and Griffin 1979). Price and Griffin

25

(1979) divided the houses into three segments. Segment 1 contains 38 structures located inside an interior wall. Segments 2 and 3 comprise the remaining structures (Price and Griffin 1979:44). The houses in Segment 1 are larger than the houses in Segments 2 and 3, and they contain more earplugs and fancy ceramics. Based on these criteria, Price and Griffin (1979:142) argued that the households in Segment 1 had higher status. In recent years, this straightforward interpretation of status differentiation at Snodgrass has been challenged. In a reanalysis of the houses and artifacts from Snodgrass, Cogswell et al. (2001) argue that the occupational history of the site was more complex than previously thought, and that while there are more artifacts in the houses in Segment 1, there is no clear evidence of status differentiation at the site. Instead, Cogswell et al. (2001) interpret the higher quantity of artifacts within the walled portion of the village to a longer occupational duration. At the King site, a Late Mississippian town in northwestern Georgia, Hally (2008) was able to infer differential household status using multiple lines of evidence, including architecture, evidence of rebuilding, artifacts, and burials. Hally‘s reconstruction of household status is more nuanced than most. Rather than postulating a simple relationship between house size and status, he demonstrates that house size and the spatial arrangement of houses changed as households moved through the domestic cycle (Hally and Kelly 1998; Hally 2008:271-279). He also finds that houses with more rebuilding episodes and interior burials are larger than houses with single building stages, and that some of the single-stage houses seemed to have been ―squeezed into small spaces‖ between existing houses and the palisade wall (Hally 2008:278). The larger, rebuilt houses may represent the founding households at King. Their size may be a function of larger household size and special status in the community, in addition to a lack of spatial constraints on building size when the houses were constructed.

26

These studies provide a framework for explaining architectural variation in terms of status differentiation. Differences in house size, the interior organization of houses, and the location of houses may be determined by the social status of household members and the functions they perform in their communities. Relative differences in house size can often be used to infer status differences, but complicating factors like occupational duration and the domestic cycle must be considered. Settlement Patterning. Archaeologists have long understood that domestic architecture varies in predictable ways within a settlement system. Important central places generally contain elite residences and special buildings for civic and ceremonial functions (Hirth 1993:124). Thus, we might expect to find a few larger and fancier buildings at major centers and less evidence of such buildings at smaller, peripheral sites. Similarly, the physical constraints on house building and expansion will be different at large, nucleated sites versus smaller, dispersed ones. Available space for building will be more limited in a densely occupied settlement as it grows, especially if it is surrounded by earthworks or a wall. These patterns have been well established in archaeological studies of stratified societies. Classic to Terminal Classic period Maya sites provide a clear example. There is not a rigid dichotomy of architecture, with palaces in major centers surrounded by hovels in rural hinterlands, but houses at urban centers are often larger and fancier than houses in surrounding rural zones, and at large sites the distribution of house sizes is positively skewed, with many small houses and only a few large ones. This is the case at Copán, where houses in the urban core are generally larger and more elaborate than houses in the rural region (Gonlin 1991:195), and at Sayil, a Terminal Classic period city Yucatán (Carmean 1991). At Aguatecas, a Late Classic period city in the Pasión region, Inomata (2008:167-169) describes 178 residential units

27

made up of houses surrounding a patio. There is a large palace in the city center with 2307 m2 of floor area, and then a continuous distribution of house sizes ranging from 5 to 617m2. Inomata (2008:169) also identified a tendency for the floor area of residential units to decrease with increasing distance from the center of the city. In the Mississippian Southeast, a site‘s place or function in a settlement system is often reflected by variation in houses. In the American Bottom near Cahokia, Emerson (1997:77-79) proposes that small farmsteads composed of one to three houses were organized in a dispersed village pattern. Dispersed farmsteads were integrated by ―nodal‖ settlements that contained sweathouses for ritual activity in addition to ordinary domestic structures. Sweathouses are also found near elite residences at Cahokia, and these buildings, in addition to providing the place for special ritual activities, may have provided a physical architectural link between leaders at the small ―nodal‖ centers and leaders at Cahokia (Emerson 1997:78). We might expect the average floor area of domestic structures to be greater at smaller, more dispersed sites and smaller at larger, more nucleated sites due to crowding during the course of settlement expansion. Hally (2008:271-279; Hally and Kelly 1998) and Moore (2002) have shown that as Late Mississippian villages expand, houses built later in the occupation are smaller as people build in less available space. The same process seems to take place at Snodgrass, where the structures along the interior edge of the fortification ditch, which have later average radiocarbon dates than structures in the center of the site, are smaller than structures in the center of the village (O‘Brien and Perttula 2001). These studies suggest that house size and the diversity of houses in a community may be determined by the place of a community in a settlement system. Major centers and smaller

28

settlements that served to integrate more scattered communities may have larger buildings and a more diverse architectural profile than smaller, more dispersed settlements. Household Research in the Southeast Long before household archaeology emerged as a formal research framework, Southeastern archaeologists looked to houses to identify social groups and reconstruct social, economic, political, and ritual activities. Beginning with the federal relief programs of the 1930s, excavations of mounds and villages uncovered hundreds of houses, mostly from the Mississippian period. At sites such as Hiwassee Island (Lewis and Kneberg 1946), Mouse Creek, Ledford Island, Hixon, Dallas, Rymer (Lewis and Lewis 1995), Jonathan Creek (Webb 1952, Schroeder 2005), and Town Creek (Boudreaux 2005) prehistoric structures were uncovered, recorded, and sometimes excavated. While record-keeping was highly variable on these projects, houses were recognized as important data sources, and many of these early excavations produced carefully drawn plan-view maps of houses and their associated features. For the most part, the houses from these early excavations were used to develop culture histories and map prehistoric culture areas. As Pluckhahn (2010:334) notes, ―house patterns appear prominently in the descriptions of various cultural-historical foci, aspects, phases, and complexes [e.g., contributors to Griffin 1952]. However, there was little consideration of the social groups that might have lived within these structures.‖ Without the benefit of radiocarbon dates, chronological associations in these early reports were understandably problematic. Architectural variation was often attributed to the norms held by different cultural groups. The Hiwassee Island report (Lewis and Kneberg 1946) and the more recently published Chickamauga Basin (Lewis and Lewis 1995) report may be the best examples of studies from this era. They continue to influence architectural studies today (see Brennan 2007; Lacquement

29

2007b). Lewis and Kneberg operated in a culture-history framework strongly influenced by McKern‘s Midwest Taxonomic Method. They carefully described individual architectural traits of buildings (e.g. walls, hearths, clay benches) and used these traits to assign houses to either the Hiwassee Island or Dallas component (Lewis and Kneberg 1946:48-79). The same strategy was used to classify Hiwassee Island, Dallas, and Mouse Creek phase houses in the Chickamauga Basin (Lewis and Kneberg 1995:54-78). With the processualist turn in the 1960s, the house became an important line of evidence for developing and testing ecological, adaptationist, and evolutionary models. As regional chronologies were developed, cultural determinist explanations for variation in house form became less common. Activity areas, seasonality of occupation, and occupational duration became staples of archaeological investigations (see Hally 1970, 1978; Smith 1979; Kline et al. 1982). In the 1970s and 1980s Southeastern archaeologists used the household as unit of analysis for understanding settlement patterns (Smith 1978) and for reconstructing ceramic assemblages (Hally 1983, 1984, 1986; Shapiro 1984). By the late 1980s there was a more formal attempt to understand Mississippian households (Polhemus 1987; Sullivan 1987), and by the time of the publication of Mississippian Households and Communities (Rogers and Lewis 1995), the household approach was a fairly coherent research framework centered on five themes: spatial analysis, social dynamics, population dynamics, subsistence, and economic activities (Rogers 1995). Coincident to the emergence of household archaeology as a coherent subfield of Southeastern research, cultural resource management projects, including some with broad horizontal excavations, produced data sets that provided new opportunities for household studies.

30

A small sample of such projects from the Southern Appalachians would include data recovery projects at Toqua (Polhemus 1987), Dog River (Poplin 1990), Rucker‘s Bottom (Anderson et al. 1995), and Brasstown Valley (Cable et al. 1997). By the 1990s, the ―building block‖ model emerged as explanation of Native American settlement and social organization. Developed first from settlement data in the Black Bottom around Kincaid by Muller (1978) and others, the model posits that self-sufficient farmsteads represented by one to three structures are the basal unit of Mississippian societies. Several of these farmsteads in a cluster make hamlet, and major centers are groups of hamlets with a mound and plaza (see discussion in Emerson 1997:67-69). This model has been used widely for the Mississippian period (Polhemus 1990) and also for the Woodland period (see Smith‘s [1992] Hopewell Household Unit). Recently scholars have criticized this model of Mississippian society as overly reductive and failing to account for both human agency and variation in household organization (Cobb 2000; Pauketat 1997, 2000, 2007), and for decoupling economic activities from political contexts (Pluckhahn 2010:338). Recent household research in the Southeast has been focused on six major themes: household production and consumption; household status; household agency, power, and resistance; households and gender; household ritual and symbolism; and households and the construction of identity (Pluckhahn 2010). Along with these new interests, in the last decade several large architectural data sets and detailed household studies have been published. These include Gougeon‘s (2002, 2006) study of three households at Little Egypt, and the reanalysis of early-to-mid-twentieth architectural data from Town Creek (Boudreaux 2005), Coweeta Creek (Rodning 2004), and Moundville (Wilson 2008). Hally‘s (2008) analysis of the households at the King site provides a rich, detailed interpretation of household and community. He uses domestic

31

and public architecture, artifacts, and burial data to reconstruct the life history of the community and infer status differences between individuals and households at the site. There has also been a renewed interest in developing regional typologies of house form and addressing technical architectural questions about houses (see contributors to Lacquement 2007). What is generally missing in the Southeast is an attempt to identify and explain the major broad-scale diachronic and synchronic changes in domestic architecture. Hally and Mainfort (2004) provide an excellent summary of changes in house form from the Woodland to the Historic Indian period, but this work is primarily descriptive, and built up from a small number of case studies compared to the total number of well-excavated houses known for the region. Peregrine (1992) compares Woodland and Mississippian period houses in the American Bottom and argues that broad changes in architecture are explained by the shift from large corporate kin groups to smaller nuclear family groups as the basal unit of social organization. Pauketat (1998) makes a similar argument using architectural data from Cahokia Tract 15A. Both of these studies link broad changes in house form to multiple social variables, but these models have not been applied beyond the American Bottom. By comparing over 1200 houses from over 60 sites across the Southeast, identifying major patterns of architectural variability, and then attempting to sort out and weigh the different social factors that explain this variation, this study can improve our understanding of domestic life in the region. It offers a chance to capture large-scale patterns of variation that require large-scale explanations, and to develop theories about houses and households using a top-down, macroregion-to-locality approach (Kowalewski 1995). In the chapters that follow I attempt determine which factors are more causally related to variation in house form than others by focusing on one set of architectural variables and explanatory socio-cultural factors at a time. After identifying patterns of change and continuity in

32

houses and their layout within communities, I develop hypotheses about the observed patterns of variation and test them with appropriate variables, at an appropriate scale, using the theoretical approaches outlined above. The questions and hypotheses are as follows: Chapter 5. How much of the variation in house form can be explained by environmental factors? I hypothesize that the distribution of houses with depressed floors or basins may be determined by variation in precipitation, elevation, temperature, and latitude. Similarly, I predict that post density may be predicted by environmental variation. Chapter 6. To what extent do variation in household economic activities and household compositions explain architectural variability? I hypothesize that changes in the size and spacing of domestic structures will track with major changes in the subsistence economy, and that the increasing partitioning of domestic structures will track with the widespread adoption of intensive maize agriculture and the establishment of larger, more permanent settlements around A.D. 900. Finally, I hypothesize that changes in architectural investment can be explained by broad-scale changes in economic behavior. Chapter 7. To what extent does architectural symbolism impact variation in house form? I predict that certain aspects of architectural variability (i.e. structure shape, the orientation of structures, the interior layout of structures, and number of posts used in walls) may be best understood as references to religious and cosmological symbolism of Southeastern Indian groups recorded during the historic period Chapter 8. How much of the variation in house form can be explained by status differentiation? I hypothesize that differences in the size of domestic structures and the frequency and diversity of interior features can be attributed to status differences in communities.

33

Chapter 9. To what degree do settlement pattern and a site‘s place in the settlement system affect house form? I hypothesize that domestic structures will be smaller at larger, nucleated sites and larger at more dispersed sites. I also predict that architecture will be more diverse at important central places. Many of these processes operate at a site or regional level, but are also structured by social phenomena operating at a broader scale. In the conclusion I synthesize the findings from each question and develop a general theory to explain the broad-scale, concordant changes in house form that occur from the Middle Woodland to the Historic Indian period.

34

Figure 2.1 Division of Space in a Navajo Hogan (Redrawn from Kent 1991:133)

35

CHAPTER 3 METHODS

There are excellent comparative data for the study of prehistoric architecture in the Southeast, but there have been few systematic attempts to compile them. Accurate drawings and descriptions of thousands of well-excavated houses are scattered across published and unpublished sources. In this chapter I describe the methods I used create the architectural database used for this research. I began my research by gathering site maps and structure drawings from welldocumented sites across the Southeast. I gathered site- and component-level data to test the effect of factors such as climate, community size, and site function on house form. Here the units of analysis are individual sites and their archaeological components. In the description of the database that follows, I describe in detail the site-level variables and how they are recorded. Second, I collected house-level data, quantitative and qualitative information on the architectural features of houses. Here the units of analysis are individual structures and their associated features. Cases were selected on three broad criteria, based heavily on Blanton‘s (1994:39) method for the cross-cultural comparison of houses and Hally‘s (2008:50-120) method for describing and analyzing architecture at the King site. The first criterion is that site maps and house drawings had to be of reasonable quality and drawn to scale. In the best cases all the associated features, including postholes and post molds, had been drawn to scale and could be measured. Most recent archaeological site reports in the Southeast meet these standards, and many WPA-era excavation records also meet these 36

criteria. Unpublished records were included provided they met these standards. The sites plans are reproduced in Appendix F. I also included some maps that only showed the outlines of house basins and wall trenches, or represented posts as ―dot‖ symbols rather than drawn-to-scale polygons. In these cases I did not have detailed data on posts, but I was still able to make reasonable size estimates, make counts of interior features, and examine the distribution of houses within sites. Second, I gave priority to site maps with broad, horizontal exposure, where household clusters could be identified. Unfortunately, some of the best site maps in the Southeast for studying houses and households are nearly indecipherable palimpsests of posts (The Rymer and Ledford Island sites in eastern Tennessee are good examples [see Lewis and Lewis 1995; Sullivan 1987]). As Knight (2007) and Benyshek et al. (2010) argue, without the aid of highquality field notes, attempting to pull houses out of dense post pattern maps long after fieldwork is completed can be a risky venture. Rather than beginning with rich but overwhelming data sets, I started with sites where houses were more readily identifiable and rebuilding episodes were clearer. Isolated houses from small archaeological excavations were also recorded. These lone structures are useful for identifying patterns of architectural variability at the regional scale, but were not used to test propositions about changing community layout over time. Only after tackling many small, manageable data sets did I move on to more complex sites. Finally, cases were selected with the goal of efficiently capturing variation across space and time. I searched for Woodland, Mississippian, and Historic Indian houses, and looked for a variety of site types, from major mound centers and large villages to small farmsteads. Geographic coverage was broad to allow for comparisons of changing house form through time at the community, regional, and macroregional scale. I initially selected architectural data sets

37

from sites in the Southern Appalachian region, and then expanded into areas with a history of research in prehistoric structures, including west-central Alabama, western Kentucky, the Eastern Highland Rim in Tennessee, and the American Bottom in southern Illinois. My sample of houses allows for comparisons of houses in single regions through time and for comparisons of houses at similar sites at a regional scale. The sample of sites is not strictly representative in a statistical sense, and it is biased toward well-documented sites. For future research the database should be expanded. On the positive side, geographic coverage is broad, temporal coverage is deep, and the data are reliable. Site maps and structure drawings were scanned at a minimum of 800 dpi using an Epson GT 20000 scanner for documents measuring 11 by 17 inches or smaller, and a Graphtec CS600 scanner for larger documents. These maps and figures were converted into vector drawings and reproduced at the same scale (1:100) in Adobe Illustrator CS4 to allow for accurate measuring of architectural features and for visual comparison of the layout and organization of settlements and structures. I measured and recorded the size, shape, orientation, and spacing of the houses, as well as the size and spacing of their postholes. These attributes were then entered into the database. In the section below, I describe each variable and how it was recorded, but in general followed recently published methods for describing prehistoric houses (Hally 2008:50-120; Lacquement 2007:54-55). In some cases, due to the quality of records and drawings, there are missing data. Observations with missing data points are dropped on a case-by-case basis in the analyses that follow. This means that in some cases the sample size is reduced. However, given the size of the data set, these smaller samples still provide a good representation for the architectural variability of the study area.

38

The Database The database has five tables: Sites, Components, Structures, Periods, and References. I describe each of the tables below. Figure 3.1 shows how the first four tables are related. Each structure is linked to the Sites table and the Components table. This allows for comparisons of houses on a site-to-site and component-to-component basis. Dates are assigned to houses by linking them to the Periods table. This reduces the tendency for data-entry errors when entering dates and allows for houses to be sorted by chronological periods. The References table stands alone, providing a searchable bibliography for the primary source material. An electronic version of the database as a Microsoft Access file is available from the author upon request. The complete database tables appear in this document as Appendices A – E. These tables are available as Microsoft Excel files from the author upon request. The site location data have been removed to protect the integrity of the archaeological resources. The ―Comments‖ fields described below have been removed in the print version, but are available in the electronic tables and the database file. In the database, each table can be viewed as a table or a form. The table view presents the data in a spreadsheet and is ideal for sorting observations by particular variables. The form view is best for examining individual observations in detail, as it provides an easy-to-read interface. When available, a thumbnail image of each structure appears in the form view. The small thumbnails are helpful for visual comparisons. Sites The Sites table contains data describing each archaeological site. I provide basic geographic and ecological data for each site as well as information about the major architectural

39

features, such as mounds and palisades. Table 3.1 is a simplified version of the Sites table. Figure 3.2 shows the location of the sites. For each site I recorded the following variables: Site ID – This is an identifier for each site. These arbitrary numbers are assigned automatically to avoid duplication, and are used to link observations in the Structures table to observations in the Sites table. State Site Number – The official state site number. Name – The site name commonly used in the archaeological literature. UTM Zone – The UTM Zone, using the WGS84 coordinate system UTM East – The UTM Easting, using the WGS84 coordinate system. UTM North – The UTM Northing, using the WGS84 coordinate system. Elevation – The elevation measured at the approximate center of the site, in meters above sea level. Data from USGS. Winter Temperature – The mean winter temperature for the site in degrees Celsius. Data from the National Climatic Data Center for the period 1961 – 1990. Mean Annual Temperature – The mean annual temperature for the site in degrees Celsius. Data from the National Climatic Data Center for the period 1961 – 1990. Physiographic Province – The physiographic province in which the site lies, as defined by the USGS. Components – This lists the numeric codes for the components present at the site, which can be referenced in the Components table. Site Size – The approximate size of the site in hectares during its most extensive occupation. Number of Structures – The total number of structures. Public – The presence or absence of public architecture, as noted by the original analyst.

40

Paired – The presence or absence of clearly paired buildings. Palisade or Other Fortification – The presence of absence of palisades, ditches or other fortifications. Mounds – A count of the total number of mounds. Plaza – The presence or absence of an obvious plaza or other clear, central, open area. Comments – In the Comments field I provide a summary description of the site and other pertinent information not captured by the variables listed above. Components Many sites in the database were occupied for centuries and have clearly distinct temporal occupations. In this table, I describe the features of components. This allows the user to group houses by more or less contemporary occupations, rather than simply by site. Using components rather than sites as an analytical unit helps control for radical changes in settlement layout over time. Table 3.2 is a simplified version of the components table. Figures 3.3 through 3.7 show the location of the components. Component ID – This is an identifier for each component. These arbitrary numbers are assigned automatically to avoid duplication, and are used to link observations in the Components table to observations in the Structures and the Sites tables. Period ID – A numeric code for the archaeological period assigned to the component. See the description of the Periods table for more detail. Site ID – This variable is used to link the component to its site. Site Name – The commonly used name of the site. Component Size – The size of the component in hectares, as defined by the original analyst.

41

Number of Structures –The total number of structures assigned to the component. I divide the structures in the database into one of four functional classes: domestic, non-domestic, storage, and unknown. This simple typology is similar to one used by Byrd (1994) to describe Neolithic structures in southwestern Asia and parallels descriptions of buildings at many well-documented sites in the Southeast (Hally 2008; Polhemus 1987; Schroedl 1986). This classification scheme has obvious limitations, but it is useful for isolating most of the houses that were primary dwellings from other kinds of buildings. In most cases the original excavators and analysts had assigned houses to similar categories, and I followed their classification. In a few cases I assigned the structures to these categories using my own judgment. Number of Domestic – The number of domestic structures assigned to the component. These are structures that appear to have served primarily as dwellings in domestic contexts. A clear example would be Primary Domestic Structures at the King site (Hally 2008:50-54), which are clearly different in function from the rectangular storage structures and public structures. Number of Non-Domestic – The number of non-domestic structures, such as large public buildings and mound summit buildings. This category is problematic, as some mound summit buildings had restricted access but were essentially residential, and some ordinary-looking structures may have been used for public functions. However, this category is useful for making broad comparisons. Problems are discussed in the comments field. Number of Storage – The number of probable storage buildings assigned to each component. Storage buildings in the Southeast are generally very small, and are associated with one or more domestic structures. Some buildings identified as storage facilities may simply be very small houses, or may have been special-purpose buildings, such as sweat lodges. These cases are discussed in the comments field.

42

Number of Unknown – The number of buildings whose function could not be determined. Paired – The presence or absence of paired structures. Mounds – The total number of mounds assigned to the components. It would be more meaningful to record the number of mound stages assigned to each component, but this is not possible for many sites. Plaza – The presence or absence of a plaza for the component. Palisade or Other Fortification – The presence of absence of palisades, ditches or other fortifications. Comments – I use the comments field to note additional information or discuss problems with the data. Structures In the Structures table, each observation is a single structure. Like other analysts in the Southeast (Hally 2008; Lacquement 2007; Rodning 2004), I define structures as roofed buildings used for dwelling, cooking, storage, ritual activities, and public gatherings. Buildings such as rectangular Early Mississippian wall trench structures, Cherokee town houses, square or circular corn cribs, and possible lean-tos are included, but palisades, fences, and outdoor screens are not. There are two types of variables in the Structures table: those that provide provenience and chronological context and those that describe house form and contents. I describe these variables below. For quantitative variables, I used the measurements provided by the original analyst after checking those measurements for accuracy and consistency. If the original analysts did not provide this information, I recorded them using measurement tools in Adobe Illustrator and recorded the data in spreadsheets prior to entering them into the database.

43

Posts and interior features were easy to associate with isolated, single-stage structures, but were often difficult to record and delineate in the rebuilt or superimposed houses. Some analysts carefully sorted out the posts and features associated with each construction stage of houses, while others, often working with older data sets, could only record houses as clusters of posts and features with an estimated number of building episodes. In the comments section for each site I indicate the level of detail for structure data. I attempt to give as detailed a description as possible for each structure. When construction stages were sorted out and listed as separate structures, I maintained this categorization in the database. For some comparisons between welldefined and poorly defined houses, I combined the construction stages of the well-defined, multistage houses and treated them as a single unit. These changes are stated explicitly in the chapters that follow. Structure ID – This is an identifier for each house. These arbitrary numbers are assigned automatically in the table to avoid duplication. Site ID – This is an identifier for each site. Assigning a Site ID to each structure allows the user to link the Sites table with the Structures table and group houses by site. Occupation – This is a convenient variable for making rough chronological comparisons. Houses are either coded as Early Woodland (EW), Middle Woodland (MW), Late Woodland (LW), Early Mississippian (EM), Middle Mississippian (MM), Late Mississippian (LM) and Historic Indian (HI). These categories are not meant to be reified, but provide an easy tool for sorting houses into broad chronological periods. Phase – If one is provided, I give the local phase name assigned to the structure by the original analyst.

44

Locus – In many cases the original analyst will divide a site into different locations, such as ―village‖ versus ―mound summit.‖ I record these for intrasite comparisons. Structure Number – The structure number assigned by the original analyst. Class – Each structure is assigned to one of the four functional classes described above: domestic, non-domestic, storage, and unknown. Shape – The shape of the house (circular, rectangular, square, etc.). Wall Type – The construction method used for the exterior wall: single post, wall trench, or both. Complete – If the house has been completely excavated and most of the post pattern is intact, I classify it as complete (Y). If the house is not completely excavated or has been badly disturbed, I classify it as not complete (N). This is subjective and somewhat problematic, but it provides a way to remove houses that should not be considered for certain quantitative comparisons, and quickly explains why a house is missing certain features. Posts – The total number of posts, interior and exterior, associated with the structure. Wall Posts – The total number of wall posts. Average Wall Post Diameter – The average diameter of the wall posts, in meters. Wall Post Spacing – The average spacing of the exterior wall posts, in meters, measured from the center of each post. For rectangular structures, this measurement was taken by measuring the length of the wall and dividing by n-1 wall posts. For circular structures, this measurement was taken by dividing the circumference of the structure by the number of posts. Interior Posts – The total number of interior posts. Interior Post Diameter – The average diameter of the interior posts, in meters.

45

Diameter – The overall diameter of the structure, in meters, if it is circular. The diameter was measured from the center of a well-defined wall post (chosen at random) to the center of a second well-defined post on the opposite side of the structure. Length – The length of the house in meters, measured along the long axis of the house, from the center of well-defined post on one wall to the center of a well-defined post on the opposite wall. Width – The width of the house in meters, measured along the short axis of the house, from the center of well-defined post on one wall to the center of a well-defined post on the opposite wall. Area – The floor area of the house in square meters, calculated using the length and width or diameter of the house. Partitions/Walls – A conservative and approximate count of the number of walls or partitions inside a structure. Interior walls and partitions may be represented by either the remains of clay or daub walls, or by closely spaced lines of posts. Lines of closely set interior posts (30 cm apart or less) were only counted as walls or partitions if they ran perpendicular to an exterior wall. Using this strategy I will miss some poorly preserved walls, and may count a few non-walls as such, but in general this method produces a conservative count of divisions within houses. Interior Hearth – I note the presence of absence of an interior hearth. In some cases (e.g. at sites with poor preservation conditions or severe post-depositional disturbance) this cannot be determined with certainty. Hearth Type – I classify most hearths as prepared clay hearths, simple fire basins, or surface fired areas. Special terms (e.g. double earth oven or undefined) are used for a few cases. Interior Features – The number of interior features, including the hearth. I define interior features as pits or depressions used for storage or cooking and molded clay seats and benches. Burials, midden concentrations, and large interior posts are noted but not counted as interior features.

46

Interior Burials – The number of interior burials. Burials must be completely within the walls of the structure and clearly contemporary with the house. Intrusive burials from later occupations are not counted. Floor Debris – I describe floor debris discovered on intact house floors. Entry Trenches – The presence or absence of parallel entry trenches. Orientation – The compass orientation in degrees of the long axis of the house, measured in compass degrees east of magnetic north. All values fall between 0 and 180 degrees. Evidence of Basin – The presence or absence of a house basin. Cases in which a house basin was likely present but not preserved are noted in the comments section. Burned – This indicates whether or not a house was destroyed by fire. Build Episodes – A count of the number of building episodes. Some analysts describe each inplace building stage of a structure as a single structure while others will describe a rebuilt structure as a single structure with multiple building stages. In the original Structures table I record as much detail for each house as possible, and treat individual building episodes as single structures, following Hally‘s (2008) convention for the King site structures. For comparisons of building episodes between sites, I created a second Structures table in which I classify all structures with multiple construction stages as single structures with multiple stages. This resulted in a table with fewer overall structures. I only used this table for appropriate intersite comparisons, and I explicitly state when I use this table in the analysis chapters. Comments – I use this section to note important architectural features of houses not captured by the variables above and to point out possible problems, discrepancies, and alternative interpretations.

47

Image – Whenever possible, I provide a medium-resolution line drawing or photograph of the excavated structure from the site report. This can only be viewed when using the Structure table in form view. Component ID – This is a numeric code that links the structure to the Components table. It is used to assign houses at sites to distinct components. This variable allows users to quickly sort structures by component rather than site. Period ID – This is a numeric code that links the structure to the Period table, providing a date range for each house. This variable allows the user to quickly sort structures by time period. Period I use the Period table to provide a date for each component and structure. The table contains archaeological periods and phases drawn from site reports and published sources. Each entry has an identifier, a general location, a start date, an end date, and a bibliographic reference. In most cases I simply recorded the chronology used by the original analyst and assigned the appropriate period code to each structure. Creating a table for the periods confers several advantages over manually entering a start and end date for each house. It reduces the chances for data entry errors and provides an easy way to group components and structures by date. In the event that a structure has an especially fine-grained date provided by C14 dating or some other means (e.g. a narrowly-dated European trade good) this information is entered in the comments field for the individual structure. References Unlike the other tables, which are all related, the references table stands alone. However, a Site ID is assigned to each reference, allowing the user to quickly find the original data source

48

and supplementary references for each site. This provides a list of all sources used for gathering site-level, component-level, and architectural data. Methods of Analysis and Advantages of the Database Approach Relational databases do more than provide a structure for information storage; they can enhance the analysis process by making it easier to recognize patterns in extremely large data sets. The analyses in this research are multiscalar. I examine variation in houses within and between sites, within and between regions, and over time. The ability to instantly sort houses into different meaningful groups (e.g. all houses from the Middle Woodland period, all houses from eastern Tennessee, all non-domestic houses, etc.) facilitates inductive pattern recognition at multiple scales and then allows for the rapid creation of tables for statistical analyses and hypothesis testing. To understand how house form varied and changed, I examined variation in each of the architectural variables listed above at multiple spatial and temporal scales. I used descriptive statistics, histograms, and box plots to identify patterns of variation during major chronological time periods. All quantitative analyses were carried out using Stata 11C. The results of these analyses are presented in chapter 4. In Chapters 5 through 9, when appropriate, I used inferential statistics to test hypotheses about causes for architectural variation. In the concluding chapter I discuss future avenues of research with the current data set, and also plans for expanding the database in a web-based environment. With this framework already in place, there are possibilities for creating a dynamic, growing architectural database that could be used for broader macroregional comparisons of houses and households.

49

Table 3.1 Sites Recorded in the Database Site 1GR1X1 1PI61 9GE1754 9GE1760 9GE1776 9GE1781 9GE333 9GE901 9GE903 9GE922 9WH120 Alarka Banks III Banks V BBB Motor Site Bessemer Brasstown Valley Cahokia Chota-Tanase Cold Springs Coweeta Creek Dallas Dog River Duncan Tract Dyar Ela Eoff I Fernvale Fredricks Garden Creek Go Kart Site Hickory Log Higgs Hiwassee Island Hiwassee Old Town Hoecake Jenrette Jewell Kellog Kimberly-Clark Kincaid King Kolomoki Ledford Island Little Egypt Lower Saratown Loy Macon Plateau

State AL AL GA GA GA GA GA GA GA GA GA NC TN TN IL AL GA IL TN GA NC TN GA TN GA NC TN TN NC NC GA GA TN TN TN MO NC KY GA TN IL GA GA TN GA NC TN GA

Physiographic Province Coastal Plain Coastal Plain Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Blue Ridge Blue Ridge Interior Low Plateau Interior Low Plateau Central Lowlands Ridge and Valley Blue Ridge Central Lowlands Ridge and Valley Piedmont Blue Ridge Ridge and Valley Piedmont Interior Low Plateau Piedmont Blue Ridge Interior Low Plateau Interior Low Plateau Piedmont Blue Ridge Piedmont Piedmont Ridge and Valley Ridge and Valley Blue Ridge Mississippi Alluvial Plain Piedmont Interior Low Plateau Piedmont Ridge and Valley Coastal Plain Ridge and Valley Coastal Plain Ridge and Valley Ridge and Valley Piedmont Ridge and Valley Piedmont

50

Site Size (ha) 0.2 0.4 1.1 0.9 1.0 1.1 8.1 1.8 10.2 0.0 0.6 0.3 0.1 0.1 0.5 10.0 2.4 890.0 2.0 4.3 1.2 1.0 2.0 8.6 2.1 0.6 13.9 0.1 0.1 5.0 2.4 0.3 0.1 4.6 140.0 81.0 0.2 3.5 0.3 0.0 50.0 2.3 100.0 5.0 2.0 0.0 2.5 150.0

No. Structures 1 4 2 4 3 2 2 5 15 3 2 2 3 5 21 28 52 253 34 8 16 29 1 7 6 15 4 2 10 4 1 16 2 78 6 11 3 22 1 3 127 65 1 20 3 2 20 12

Site Martin Farm McFarland Mialoquo Mitchum Morris Moundville Mouse Creek Napoleon Hollow Potts' Tract Ravensford Rivermoore Rucker's Bottom Rymer Snodgrass Summerour Toqua Town Creek Townsend Tuckasegee Tukabatchee Two Run Creek Warren Wilson Yearwood Yuchi Town

State TN TN TN NC KY AL TN IL GA NC GA GA TN MO GA TN NC TN NC AL GA NC TN AL

Physiographic Province Ridge and Valley Interior Low Plateau Ridge and Valley Piedmont Interior Low Plateau Coastal Plain Ridge and Valley Central Lowlands Ridge and Valley Blue Ridge Piedmont Piedmont Ridge and Valley Mississippi Alluvial Plain Piedmont Ridge and Valley Piedmont Blue Ridge Blue Ridge Coastal Plain Ridge and Valley Blue Ridge Interior Low Plateau Coastal Plain

51

Site Size (ha) 1.1 2.7 10.0 0.5 0.6 70.0 1.0 15.0 0.8 13.8 0.1 0.8 1.9 0.8 1.0 1.9 0.2 10.0 0.5 1.2 0.9 1.3 1.0 2.0

No. Structures 12 5 8 1 12 152 17 1 3 111 3 12 23 91 1 133 42 10 1 2 6 17 14 2

Table 3.2 Components Recorded in the Database

Component 1GR1X1 1PI61 9GE1754 9GE1760 9GE1776 9GE1781 9GE222 9GE333 9GE901 9GE903 9WH120 Alarka Banks III Banks V EM Banks V MW BBB Motor Site EM BBB Motor Site LW Bessemer Brasstown Valley LW 1 Brasstown Valley MW Brasstown Valley LW 2 Brasstown Valley HI Cahokia S Cahokia M Cahokia EM

Comp. Size (ha) 0.17 0.38 1.08 0.85 1.04 1.14 0.04 8.1 1.75 10.2 0.6 0.25 0.1 0.03 0.1

Structures (n) 1 4 2 4 3 2 3 1 5 15 2 2 3 1 4

Dom. (n) 1 4 2 4 3 2 3 1 5 14 2 2 3 1 4

Nondom. (n) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Storage (n) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Unk (n) 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0

Mds (n) 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0

EM

0.13

5

4

1

0

0

LW EM

0.16 10

16 25

16 4

0 21

0 0

LW

2.4

2

2

0

MW

2.4

8

8

LW

2.4

13

HI EM MM EM

2.4 890 890 890

20 7 24 41

Period MW LW LM LM LM LM LM MW LM LM LM HI MW EM MW

Plaza N N N N N N N N N N N N N N N

Fortified N N N N N N N N N N N N N N N

Start Date 400 1000 1580 1520 1450 1450 1450 400 1520 1450 1450 1600 200 880 200

AD AD AD AD AD AD AD AD AD AD AD AD AD AD AD

End Date 600 1100 1670 1580 1600 1600 1600 750 1580 1600 1600 1700 600 1045 600

AD AD AD AD AD AD AD AD AD AD AD AD AD AD AD

0

N

N

1050

AD

1150

AD

0 0

0 3

N Y

N N

900 1000

AD AD

1050 1200

AD AD

0

0

0

Y

Y

500

AD

1000

AD

0

0

0

0

Y

N

150

BC

500

AD

13

0

0

0

0

N

Y

500

BC

500

AD

10 5 24 35

0 2 0 0

10 0 0 0

0 0 0 6

0 100 100 100

N Y Y Y

N Y Y N

1450 1100 1200 1000

AD AD AD AD

1600 1200 1300 1200

AD AD AD AD

52

Component Cahokia LW Cahokia L Chota-Tanase EM Chota-Tanase HI Coweeta Creek HI Coweeta Creek LM1 Coweeta Creek LM2 Dallas Dog River Duncan Tract MW Dyar LM 1 Dyar LM2 Ela HI Ela MW Fernvale EM Fernvale MW Garden Creek Go Kart Site Hickory Log MM Hickory Log MW Hiwassee Island LM Hiwassee Island EM

Comp. Size (ha) 890 890

Structures (n) 62 119

Dom. (n) 53 106

Nondom. (n) 0 0

Storage (n) 1 0

Unk (n) 8 13

Mds (n) 0 100

0.1

1

1

0

0

0

HI

2

33

29

1

2

HI

1.2

1

1

0

LM

1.2

5

5

LM LM LM

1.2 1 2

10 29 1

MW LM LM HI MW EM

0.4 2.13 2.13 0.6 0.6 0.1

MW

Plaza Y Y

Fortified N Y

AD AD

End Date 1050 1100

AD AD

0

N

1000

AD

1200

AD

1

1

N

1625

AD

1800

AD

0

0

N

N

1600

AD

1700

AD

0

0

0

Y

N

1300

AD

1500

AD

7 26 1

2 3 0

0 0 0

0 1 0

Y Y N

N Y N

1500 1300 1450

AD AD AD

1600 1525 1600

AD AD AD

5 2 4 5 10 1

5 2 1 5 10 1

0 0 0 0 0 0

0 0 0 0 0 0

0 1 1 0 0 0

N Y Y N N N

N N N N N N

200 1375 1520 1600 200 1000

BC AD AD AD AD AD

200 1450 1580 1700 600 1200

AD AD AD AD AD AD

0.1

1

0

0

0

0

N

N

500

BC

500

AD

MW LM

5 2.4

0 1

0 0

0 0

1 0

2 0

N N

N N

200 1450

AD AD

600 1600

AD AD

MM

3

3

0

0

0

0

N

Y

1300

AD

1350

AD

MW

4

13

13

0

0

0

0

N

N

150

BC

500

AD

LM

4.6

1

1

0

0

0

2

Y

Y

1300

AD

1450

AD

EM

4.6

37

13

34

0

0

2

Y

Y

975

AD

1200

AD

Period LW EM EM

Start Date 900 1050

N

Y

0

0

1 0 0

0 0 3 0 0 0

1

1 1

4

53

Component Hiwassee Old Town Hoecake Jenrette Jewell Kellog KimberlyClark Kincaid LM Kincaid MM Kincaid EM Kincaid UNID King Kolomoki Little Egypt Lower Saratown Loy Martin Farm M1 Martin Farm M4 Martin Farm M2 Martin Farm M3 McFarland Mialoquo Mitchum Morris Moundville EM

Comp. Size (ha)

Structures (n)

Dom. (n)

HI LW HI EM MW

140 81 0.15 3.5 0.26

6 11 3 22 1

4 11 3 13 1

EM LM MM EM

0.3 32 32 32

3 2 6 7

MM LM MW LM

32 2.3 100 2

HI LM

Nondom. (n)

Storage (n)

Unk (n)

Mds (n)

0 0 0 9 0

1 0 0 0 0

1 0 0 0 0

3 1 0 1

0 1 6 6

0 0 0 0

12 67 1 3

4 44 1 3

6 2 0 0

0.04 2.5

2 2

2 2

EM

1.1

1

LM

1.1

EM LM MW HI HI EM

Period

EM

Plaza

Fortified

0 2 0 1 0

N Y N N N

N N Y N N

1625 500 1600 1000 150

AD AD AD AD BC

1800 1100 1680 1200 500

AD AD AD AD AD

0 0 0 0

0 27 27 27

N Y Y Y

N Y Y Y

975 1300 1200 1000

AD AD AD AD

1200 1450 1300 1100

AD AD AD AD

0 21 0 0

2 0 0 0

27 0 8 2

Y Y Y Y

Y Y Y N

1000 1450 500 1350

AD AD AD AD

1450 1625 750 1575

AD AD AD AD

0 0

0 0

0 0

0 1

N Y

Y Y

1620 1300

AD AD

1670 1450

AD AD

1

0

0

0

0

N

N

900

AD

1000

AD

1

0

0

1

0

0

N

N

1600

AD

1819

AD

1.1

5

3

1

0

1

2

N

Y

1000

AD

1300

AD

1.1 2.65 10 0.5 0.6

5 5 8 1 12

0 5 7 1 12

3 0 1 0 0

0 0 0 0 0

2 0 0 0 0

2 0 0 0 0

N N Y N Y

N N N Y Y

1300 200 1760 1600 1000

AD BC AD AD AD

1600 200 1776 1670 1300

AD AD AD AD AD

75

14

14

0

0

0

29

Y

Y

1120

AD

1260

AD

54

Start Date

End Date

Component Moundville EM 1 Moundville M1-M2 Napoleon Hollow Pott's Tract Rivermoore Rucker's Bottom LM Rucker's Bottom MM Rymer Snodgrass Summerour Toqua HI Toqua EM Toqua LM Town Creek EM Town Creek LW Town Creek LM Town Creek MM 2 Town Creek MM1 Townsend Tuckasegee Tukabatchee Two Run Creek

Comp. Size (ha)

Structures (n)

Dom. (n)

EM

75

35

35

EM

75

102

MW LM LW

15 0.8 0.1

LM

Nondom. (n)

Storage (n)

Unk (n)

Mds (n)

Plaza

Fortified

0

0

0

2

N

N

1120

AD

1200

AD

98

4

0

0

29

Y

Y

1200

AD

1300

AD

1 3 3

0 3 3

0 0 0

0 0 0

1 0 0

0 0 0

N N N

N N N

50 1450 850

BC AD AD

140 1625 1000

AD AD AD

0.8

5

3

1

1

0

0

Y

Y

1350

AD

1450

AD

MM LM MM LW HI EM LM

0.8 1.94 0.82 1 1.9 1 1.9

7 23 91 1 14 25 79

3 21 91 1 12 15 41

2 2 0 0 2 9 23

2 0 0 0 0 0 14

0 0 0 0 0 1 1

0 0 0 1 2 2 2

Y Y Y N Y Y Y

N Y Y N Y Y Y

1100 1450 1275 850 1625 1200 1300

AD AD AD AD AD AD AD

1350 1575 1350 1000 1800 1300 1620

AD AD AD AD AD AD AD

EM

0.15

1

1

0

0

0

0

Y

Y

1000

AD

1150

AD

LW

0.15

1

0

1

0

0

0

Y

Y

800

AD

1000

AD

LM

0.15

4

4

0

0

0

1

N

N

1350

AD

1450

AD

MM

0.15

19

0

16

0

3

1

Y

Y

1250

AD

1350

AD

MM HI HI LM

0.15 10 0.5 1.2

20 10 1 2

13 9 1 2

5 0 0 0

0 0 0 0

2 1 0 0

1 0 0 0

Y N N N

Y N N N

1150 1625 1700 1600

AD AD AD AD

1250 1800 1800 1715

AD AD AD AD

MW

0.9

1

1

0

0

0

1

N

N

400

AD

750

AD

Period

55

Start Date

End Date

Component Warren Wilson Yearwood

Period LM MW

Comp. Size (ha)

Structures (n)

Dom. (n)

1 1.01

12 14

12 12

Nondom. (n) 0 0

Storage (n)

Unk (n)

Mds (n)

0 0

0 2

0 0

56

Plaza

Fortified

Y N

Y N

Start Date 1250 200

End Date AD BC

1450 500

AD AD

Figure 3.1 Diagram of Database Relationships 57

Figure 3.2 Map of Sites Recorded in the Database

58

Figure 3.3: Map of Middle Woodland Components

59

Figure 3.4: Map of Late Woodland Components

60

Figure 3.5 Map of Early Mississippian Components

61

Figure 3.6 Map of Middle and Late Mississippian Components

62

Figure 3.7: Map of Historic Indian Components

63

CHAPTER 4 RESULTS: ARCHITECTURAL VARIABILITY

The quantitative and qualitative architectural variables from this study can be summarized at multiple scales. Most architectural studies begin at a fine spatial and temporal scale, usually at the level of a single house or site, and then move out for comparisons with other areas (see Hally 2008, Polhemus 1987). I start out at a larger scale and zoom in. I examine variation in structures by broad, chronological periods (following Anderson and Mainfort 2002; Bense 1994, Hally and Mainfort 2004): Middle Woodland (ca. 200 B.C. to A.D. 400), Late Woodland (ca. A. D. 400 to 1000), Early Mississippian (ca. A.D. 1000 to 1200), Middle Mississippian (ca. A.D 1200 to 1350), Late Mississippian (ca. A.D. 1350 to 1550), and Historic Indian (ca. A.D. 1550 to 1800). This strategy is effective for identifying major patterns of architectural variation over time and gives additional context for analyses and comparisons at finer scales. Some of these trends and patterns have been identified in previous research and syntheses, but others are not as well documented or understood. In this chapter and the chapters that follow, I use the functional classes of structures outlined in Chapter 3 to group and compare structures. The term structure refers to all types of buildings, regardless of function. Domestic structures are those that appear to have served primarily as dwellings in domestic contexts. These are the structures we usually define as ―houses.‖ Any use of the term ―house‖ in the text refers to a domestic structure. Non-domestic structures are mostly large public structures used for gatherings and ceremonies, like town houses and earth lodges, but also some smaller special-purpose buildings, like sweat houses. 64

They can be identified by their size and shape, and also their special locations in settlements. Storage structures were primarily used for storing maize. They are generally small and clearly associated with one or more domestic structures in a household cluster. In early Spanish accounts these are called ―barbacoas.‖ The remaining structures fall into the other category, and only make up about 4 percent of the total. Table 4.1 provides summary statistics for all the structures in the database. This gives a general sense of the range of values for each architectural variable, regardless of time period or functional class. Tables 4.2 – 4.13 provide summary statistics for the structures when they are grouped by broad chronological periods and functional classes. For example, Table 4.2 lists summary statistics for all the Middle Woodland period domestic structures, and Table 4.8 describes the Late Mississippian non-domestic structures. I do not provide summary tables for structures that fall into the other category, or for classes with fewer than five structures, but these structures are discussed in the text. In this chapter I present evidence for four major architectural trends. First, architectural investment—the quantity of material and labor embodied in structures—generally increases over time, and peaks in the Late Mississippian period. This is not especially surprising, given our general understanding of increasing sedentism in the Southeast after the adoption of intensive maize cultivation around A.D. 800 (Anderson and Mainfort 2002:18). There are, however, interesting deviations from this general trend. For example, there are Late Mississippian domestic structures in the uplands of the Oconee River valley that seem far more ephemeral than contemporary domestic structures elsewhere (Hatch 1995), and Middle Woodland domestic structures in middle Tennessee that seem very robust and permanent for purportedly semisedentary households (Faulkner 1988, 2002).

65

Second, given the radical changes in settlement patterns and community organization from the Woodland to the Mississippian to the Historic Indian period, there is more continuity in building traditions than might be expected. The average size of domestic structures changes over time, but the size difference between non-domestic structures and domestic structures remains similar throughout the chronological sequence. Certain architectural traits, such as the use of four central corner posts and semi-subterranean basins, have a deep history. In terms of broad spatial patterns of variation, changes in structure shape and size during the Late Woodland to Early Mississippian transition take a similar direction in the American Bottom and west-central Alabama. Something different happens to the east, in the Southern Appalachians. By the Early Mississippian period, structures look very similar in both areas. This regional variation has important implications for differences in household organization. Finally, across the entire study area, changes in domestic structure size, burials practices, storage practices, and the interior segmentation of domestic structures from the Woodland to the Mississippian period suggest that households become increasingly autonomous, and that relatively small extended or nuclear family households emerge as the basal unit of social organization. The timing and nature of this change plays out differently across the study area during the Mississippian period. This chapter is primarily descriptive. I focus on a single variable (or one small set of related variables) at a time, exploring synchronic variation and diachronic change. In the conclusion, I outline the major trends in architectural variability. This synthesis lays the groundwork for the chapters that follow, where I attempt to determine the social reasons behind this variation.

66

This first set of variables includes prominent architectural traits that would have been highly visible from the exterior of a structure: shape, size, orientation, and style of entryway. Given their high visibility, much of the variation observed in these traits may relate to their role in symbolic communication (see Blanton 1994:11-12). The size and shape of structures may reflect the size, activities, social position, and cosmologies of the social groups that inhabit them. The orientation of structures and their style of entryway may be determined by very prosaic causes (e.g. domestic structures in a linear settlement arranged parallel to a ridge), but they can also reflect shared norms of house building rooted in broader belief systems (Cunningham 1974; Hodder 1984). Shape I grouped structures in the database into five major shape classes: circular, ovular, rectangular, square, and irregular. A small minority of houses were t-shaped or keyhole shaped. This categorization obscures some finer details of structure form, but captures most of the variation. From a broad view, changes in shape track with breaks in the major chronological periods. There is a general shift from circular domestic structures in the Middle and Late Woodland to rectangular domestic structures in the Late Woodland and Early Mississippian, to square domestic structures and rectangular storage structures in the Late Mississippian, to circular and rectangular domestic structures in the Historic Indian period. There are deviations from this pattern at smaller scales. Table 4.14 records the shape of all the structures in the database, sorted by time period and functional class. Circular structures occur in all time periods. There are some circular domestic structures in all periods except the Early Mississippian, but there are circular non-domestic structures in Early Mississippian times, including possible earthlodges and sweat houses. This shape is most

67

common during the Middle Woodland period. Over 65 percent of Middle Woodland domestic structures in the database are circular. Only about 7 percent of all Early, Middle, and Late Mississippian structures are circular. They make up a low proportion of structures, mostly nondomestic, in the Mississippian components at Hiwassee Island, Toqua, Macon Plateau, Jewell, Bessemer, Martin Farm and Cahokia, but are common at Town Creek, Rucker‘s Bottom, several sites in the Oconee River valley in piedmont Georgia, and possibly at Brasstown Valley, where the Late Woodland to Early Mississippian chronology is not entirely clear. During the Historic Indian period at Cherokee sites, circular domestic structures, often called ―winter houses,‖ are usually paired with rectangular domestic structures called ―summer houses.‖ Twenty-three ovular structures (2 percent) were recorded at 13 sites dating to the Woodland, Early Mississippian, and Historic Indian period. Over half of these houses are from Owl Hollow phase sites (n=6) in Tennessee and the Late Woodland occupation at Brasstown Valley in northeast Georgia (n=8). Rectangular structures (n=595, 47 percent) were the most common of all, appearing in all periods, but most frequently in the Late Woodland, Mississippian, and Historic Indian periods. Approximately 80 percent of the Late Woodland and 60 percent of all the Early Mississippian structures are rectangular. Nearly half of the Historic Indian structures in the study are rectangular. Rectangular domestic structures are not as common during the Middle and Late Mississippian, but rectangular storage buildings commonly occur at Late Mississippian sites. Only four Middle Woodland structures are rectangular, and these all come from the Yearwood site, which may have been a special ceremonial center (Butler 1979). Square structures (n=358, 28 percent) occur in all periods but are especially common in the Middle and Late Mississippian periods. Domestic structures in villages and domestic and

68

non-domestic structures on mound summits are often square with rounded corners. Only seven square structures were recorded for the whole of the Woodland period; one square structure of unknown function with rounded corners in the pre-mound midden at Garden Creek, three square domestic structures at Rivermoore, and three roughly square domestic structures at the Yearwood site. Several square Connestee-phase structures with rounded corners and a single support post, nearly identical to square house at Garden Creek, have recently been uncovered at the Iotla site at the Macon County Airport near Franklin, North Carolina (Tasha Benyshek personal communication 2010). Only four of the Historic Indian period structures were square; two large non-domestic public structures at Toqua, and single domestic structures at Yuchi Town and Tukabatchee. Three Cherokee domestic structures at the Townsend site described as ―octagonal‖ closely resemble Late Mississippian domestic structures that are square with rounded or truncated corners. The remaining structures (n=119, 9 percent) did not fit into these four broad shape categories. Most of these structures (n=97) have an undefined shape due to poor preservation or incomplete excavation. However, 22 structures had complex or unusual shapes, such as a single keyhole-shaped domestic structure at Kolomoki, two t-shaped, non-domestic structures at Cahokia Tract 15A, and semi-circular structures at Yearwood. These structures do not share a common function (i.e. they are not all ―ceremonial structures‖), and have to be understood within the context of individual sites. In an early and influential attempt to link variation in house shape with variation in social and economic organization, Flannery (1972:22-44) used archaeological data from early villages in the Near East and Mesoamerica and cross-cultural data (especially from African herders and

69

horticulturists) to argue that circular dwellings correlate with partially nomadic groups with extended, possibly patrilocal and polygamous families and shared storage. Rectangular structures, on the other hand, correlate with more sedentary societies composed of nuclear, possibly monogamous families with private storage facilities. It is tempting to argue that a similar process plays out in the Southeast. There is a shift from round to rectangular structures, and increasing sedentism with the adoption of intensive maize agriculture after A.D. 800 (Anderson and Mainfort 2002:18). However, this model has been criticized as overly simple, and a recent study of early sedentary communities on the Anatolian plateau shows a poor fit between house shape and degree of sedentism (Steadman 2004:519-520). Flannery himself recently revised this model, placing less emphasis on the importance of house shape as a correlate of social and economic behavior, and more on the importance on a broader set of changes in house form that are related to the privatization of storage (2002: 421). Similarly for the Southeast, structure shape is better understood as part of a suite of architectural changes related to changes in social organization and domestic production and consumption. I return to these ideas in Chapter 6. Area Structure size has probably received more attention in archaeological studies of households than any other architectural variable in the database, for practical and theoretical reasons. Variation in house size in a community is highly visible, broadcasting strong signals about individual and household status (Wilk 1983; Blanton 1994). House size is also easily recorded, and even the earliest and most cursory excavation records in the Southeast include plan view maps of houses drawn to scale. I was able to record the floor area of 1054 of the 1258 structures in the database. Only structures with incomplete floor plans were not measured.

70

Ethnographic studies show that there is no single explanation for variation in house size. Among Kekchi Maya, the social position of a household within a household cluster is often the best predictor of household size (Wilk 1983). House size depends on the intended use life of the structure (Kelly et al. 2005), but also changes with fluctuations in the domestic cycle (Goody 1962). The reasons for variation in house size are complex, but archaeologists should not jettison house size as a useful variable for understanding households. A few recent studies point to the value in comparing the size of domestic structures within settlements. In his study of households at Tract 15A at Cahokia, Pauketat (1998:135-136) uses domestic structure size to argue for a shift from larger courtyard groups to smaller households as the basal unit of social organization. At Moundville during the Early Moundville I and Late Moundville I-Early Moundville II phases, Wilson (2008:87-92) compares house size within multihousehold residential groups. The distribution of domestic structures sizes are quite similar, suggesting that there are not major status differences between the residential groups. Hally‘s examination of Barnett-phase domestic structures at the King site reveals that size is impacted the domestic cycle (Hally and Kelly 1998; Hally 2008:271-279). Large, rebuilt domestic structures may represent the founding households at King, and their size may be a function of larger household size and special status in the community. Despite the obvious differences in Woodland, Mississippian, and Historic Indian societies, there is value in treating all the structures in the database as a single population and examining the distribution of structures size. This broad view provides a background for comparison of domestic and non-domestic structures at finer scales, and is especially informative in terms of the upper and lower size limits. The median size of all structures in the database is

71

28.4 m2 (n=1054). The smallest is a 1.15 m2 circular Qualla-Lamar-phase storage building at Brasstown Valley and the largest is a 299.8 m2 non-domestic Early to Middle Mississippian mound summit structure at Kincaid. The distribution of structure size is continuous and positively skewed; 75 percent of the structures measure 46.2 m2 or less and 95 percent fell below 103.7 m2. Middle Woodland period domestic structures range in size from 7.1 m2 to 146.6 m2, with a median floor area of 40 m2 (n=66) (Figures 4.1 and 4.2). All but 7 domestic structures fall into a normal distribution centered on 40 m2 and ranging from 7 to 80 m2. It is notable that the seven largest domestic structures are all either McFarland or Owl Hollow phase structures from the Yearwood, Duncan Tract, Banks III, and Banks V sites. The largest of these is a well-defined structure from the Banks V site with two large interior hearths and four central support posts, but several others in this larger size category are large circular domestic structures with few interior posts or features. Overall, Late Woodland domestic structures range from 3 to 94.7 m2 with a median size of 8.4 m2 (n=101) (Figures 4.3 and 4.4). Unlike the Middle Woodland period, the distribution of house size is positively skewed. All but one of the structures falls into a range of 3 to 48.9 m2, with the single outlier being a large, non-domestic Teal-phase building at Town Creek. An unusual keyhole-shaped domestic structure with a basin excavated at Kolomoki measures 7.5 m2. There are notable regional differences in structure size during the Late Woodland period. In the American Bottom during the Late Woodland/Emergent Mississippian period, at Late Woodland sites in west-central Alabama, and in the Cairo Lowlands of Missouri, settlements are organized into clusters of closely-spaced, small, rectangular domestic structures arranged around a central open area. These domestic structures range in size from 3.0 to 16.4 m2. They are so

72

small that it is hard to imagine them inhabited by any group larger than a small nuclear household. In northern Georgia, Late Woodland domestic structures at Brasstown Valley, Summerour, and Rivermoore are larger and more widely spaced, ranging in size from 17.8 to 48.9 m2. The spatial arrangement of these structures is similar to the arrangement of some Middle Woodland domestic structures at Hickory Log and McFarland. They are large compared to the small rectangular structures from sites to the west, and are spaced farther apart in their settlements. Keeping in mind the warning that house size is not always a direct reflection of household size, the architectural data seem to indicate two different forms of household organization. Late Woodland groups in the western part of the study area may be organized into courtyard groups made up of small, nuclear households. These courtyard groups may be the basic unit of social organization. In the eastern part of the study area, larger household groups, perhaps extended families, may occupy single structures, and these groups may be the basal social units. The size distribution of domestic structures during the Early Mississippian period is positively skewed, with many more small houses than large ones (Figure 4.5 and 4.6). Domestic structures range in size from a 3 to 118 m2, and the median size is 21 m2 (n=268). Most of the domestic structures are small; 50 percent measure 21 m2 or below and 75 percent measure 41 m2 or below. Non-domestic Early Mississippian structures are generally larger, with a range of 1.2 (a very small mortuary structure at BBB Motor site) to 300 m2 (a mound-summit structure at Kincaid), and a median size of 85 m2 (Figures 4.7 and 4.8). Most of the large non-domestic

73

structures occur on or near mounds and plazas. The few non-domestic structures with floor area less than 30 m2 represent special-purpose buildings, such as possible sweathouses and t-shaped buildings at Cahokia Tract 15A, and a very small mortuary structure at the BBB Motor site. The smaller non-domestic structures are generally located either on top of mounds or in some other special location set off from domestic structures. During the Middle Mississippian period the average size of domestic structures increases only slightly (Figures 4.9 and 4.10). Domestic structures range from 1.9 to 86 m2, with a median of 23 m2 (n=115). The size distribution is positively skewed, but less so than in the Early Mississippian period. All but six of the domestic structures fall between 1.9 and 55 m2. As in the Early Mississippian period, Middle Mississippian non-domestic structures are larger, with a range of 42 to 281 m2, and a median size of 129 m2 (n= 20) (Figures 4.11 and 4.12). These structures come in variety of shapes: square, circular, and rectangular. The most obvious shared feature of these buildings is their location. They are all set off from domestic structures, either on a mound summit or some other special location. In the Late Mississippian period there is a common pattern of settlements with small storage buildings, larger domestic dwellings, and a few very large public buildings. Late Mississippian domestic structures are larger than those of the Early and Middle Mississippian periods. The median size of domestic structures is 41 m2, with a minimum size of 6.6 and a maximum of 126 m2 (n=203) (Figures 4.13 and 4.14). Most Late Mississippian domestic structures range from roughly 20 to 126 m2 with a peak around 40 m2 (n=203). Houses in this size class are generally ordinary domestic structures, but some, especially those greater than 90 m2, appear to be non-domestic in function. Most of these structures are what Hally (2002) and others describe as winter houses. They are square,

74

often with rounded corners. The best-preserved examples have shallow basins, central hearths, and four interior roof supports. Additional interior posts formed walls and partitions that radiated out from the center of the house to the exterior walls. Hally (2002) provides a description and comparison of these structures from sites across the Southern Appalachians. Fine-scale analyses of floor debris distributions in houses from Loy (Polhemus 1998) and Little Egypt (Gougeon 2002) show a consistent pattern of separate activity areas within these houses. Out of 295 Late Mississippian houses in the database, 162 fit the winter house pattern. Out of this subset, only 60 have clear evidence for a basin, but in many cases the evidence for this has most likely been destroyed by deep plowing, erosion, and other post-depositional processes. Late Mississippian non-domestic structures range in size from 27 to 226 m2, with a median size of 79 m2 (Figures 4.15 and 4.16). The largest of these non-domestic structures are public buildings, such as the townhouse at Coweeta Creek and Structure 17 at King, which each measure approximately 15 m across. Both of these are essentially ―scaled-up‖ versions of the more common and smaller winter houses. A similar Late Mississippian townhouse at Ledford Island measures 13 m across, and was rebuilt up to four times (Lewis and Lewis 1995:529-530). Based on historic accounts of Cherokee and Creek townhouses, Hally (2008:138) argues that Structure 17 at King, ―functioned as a community meeting house where men and perhaps women congregated and participated in social, ceremonial, and political activities.‖ Large sixteenth- to seventeenth-century public structures which may have served a similar function have also been recorded outside the Southern Appalachians. An Atasi-phase structure at Fusitachee in south-central Alabama, probably analogous to the Creek rotunda, was square with rounded corners, measured up to 15 m on a side, and was rebuilt three times (Sheldon 1990). The late seventeenth-century Apalachee council house at Mission San Luis de

75

Talimali had a round floor plan and measured approximately 36.5 m in diameters (McEwan and Hann 2000). Structures classified as storage buildings are small rectangular or circular structures that appear to be barbacoas or corn cribs (Figures 4.17 and 4.18). These are especially well documented at King, Toqua, and Coweeta Creek. At these sites, small rectangular and round barbacoas not only stand out for their difference in size, but also their location in clusters of houses. They do not appear to be primary residences, but rather smaller buildings that complement more obvious domestic dwellings. This offers further support for their function as storage facilities. In the Oconee River valley in Georgia‘s northern piedmont, rectangular buildings measuring roughly 3 x 4 m are consistently found on upland Lamar phase sites (Hatch 1995; Ledbetter n.d.). In many cases a cluster of three or more of these rectangular structures will be located just south of a larger, circular building, but in some cases (e.g. 9GE901, 9GE1754, and 9GE901) they appear to be the only structures associated with the occupation. Hatch argues that these groups of buildings represent ―small, dispersed homesteads occupied year-round by families‖ (1995:136). He also suggests that the rectangular houses are domestic structures and not storage buildings, perhaps more analogous to historically recorded summer houses than barbacoas (Hatch 1995:146-147). Many basic characteristics of Late Mississippian houses continue into the Historic Indian period. Domestic structures range in size from 12 to 86 m2, and the median size of domestic structures decreases slightly from the previous period to 35 m2 (n=87) (Figure 4.19 and 4.20). Only 13 structures, ranging in size from 1.15 to 9 m2 are identified as possible corn cribs. This may be due to a shift toward sweet potato cultivation in some parts of the Southern Appalachians

76

(Wilson 2008:15). Sweet potatoes are better stored in slightly damp, below-ground cellars rather than dry, above-ground storage facilities (Riggs 1999). It is also possible that more corn was stored indoors, perhaps in the rectangular domestic structures described below. At most of the Historic Indian sites there are two types of domestic structures: circular, square, or octagonal ―winter‖ houses and rectangular ―summer‖ houses. These types of houses are often paired, they are recorded in several late ethnohistoric accounts in the Southern Appalachians, and have received considerable treatment by archaeologists (see Hally 2002; Schroedl 1986). The two classes of buildings, while very different in terms of architectural construction and investment, are remarkably close in size. The winter houses have an average floor area 36.0 m2 (s.d. 8.3 m2, n=40), while the summer houses average 35 m2 ( s.d. 16.1 m2, n=41). Winter and summer house pairs occur at Brasstown Valley, Chota-Tanase, Mialoquo, Townsend, and Tukabatchee. Paired summer and winter houses account for 81 of the 87 domestic structures. The remaining six houses come from the North Carolina piedmont, where a different pattern is found during the Historic Indian period. Domestic structures at the Jenrette, Lower Saratown, and Mitchum sites are rectangular or oval in shape, and range from 12 to 32 m2. Non-domestic structures include four very large circular or square townhouses from Chota-Tanase, Toqua, and Mialoquo. Floor area in these buildings ranges from 227 to 263 m2. Historic period townhouses are well documented in ethnohistoric accounts, and like some public structures from the Late Mississippian period, appear to be larger versions of domestic structures, with obvious architectural differences reflecting their function as meeting places. Note that Historic Indian townhouses, with an average size of 242 m2 (n=4) are over twice the

77

average size of non-domestic structures from the Late Woodland, Early Mississippian, Middle Mississippian, and Late Mississippian periods. Several important trends emerge from this broad analysis of house size. Compared to domestic structures from other periods, Middle Woodland domestic structures are relatively large. The average size of domestic structures decreases during the Late Woodland period (Figure 4.21; Tables 4.2 and 4.3). There is, of course, regional variation in this trend, as Late Woodland houses and households in the Southern Appalachians do not change as much, while there may be a radical change in household organization in the western part of the study area. Broadly speaking, domestic structure size increases from the Late Woodland to the Early Mississippian period (Figure 4.21; Tables 4.3 and 4.4). In the western part of the study area, Early Mississippian wall trench houses are much larger Late Woodland domestic structures. Specifically, domestic structures from the Late Woodland occupations at 1PI61, BBB Motor site, Cahokia, and Hoecake average only 7.7 m2 in floor area (n=82). In contrast, Early Mississippian domestic structures from BBB Motor, Bessemer, Cahokia, Jewell, Kincaid, and Moundville have an average floor area of 23 m2 (n=218), while non-domestic structures from the same sites average 84 m2 in area (n=34). Early Mississippian domestic structures at these sites are also spaced farther apart. Based on the changes in house size and the arrangement of houses at the sites in this database, the pattern that Pauketat (1998:135-136) identified at Cahokia Tract 15A— a shift from larger courtyard groups to smaller, nuclear family households as the primary unit of social organization—may be more widespread across the American Bottom and further south into the Cairo Lowlands and Alabama. In the Southern Appalachians, Early Mississippian domestic structures from ChotaTanase, Hiwassee Island, Kimberly-Clark, Macon Plateau, Martin Farm, Toqua, and Town

78

Creek average 39 m2 (n=36), while non-domestic structure on mound summits and in plazas average 85 m2 (n=17). Late Woodland domestic structures from Brasstown Valley, Raccoon Ridge, and Rivermoore (n=26) average 28.8 m2, while a single structure from the summit of the Summerour Mound measures 27.5 m2. Thus, in the eastern part of the study area, domestic structures increase in size from the Late Woodland to the Early Mississippian period, but the increase is not as dramatic as in the west. Because there are so few excavated Late Woodland sites with broad horizontal exposure, it is hard to identify major changes in the spacing of houses from the Late Woodland to the Early Mississippian period. Changes in the makeup of social groups that accompany these more subtle changes in structure size are less clear, but it does seem that as in the American Bottom and surrounds, nuclear households may be the basal social unit of the Early Mississippian period. By the Early Mississippian period, domestic structures and settlement plans look fairly similar across the entire study area. From this point forward, domestic structure size continues to increase, and plateaus during the Late Mississippian and Historic Indian periods (Figure 4.21; Tables 4.6, 4.8, 4.11). In the Southern Appalachians, Late Mississippian winter houses are generally larger than Early Mississippian wall trench houses, and tend to be arranged in clusters of two or three domestic structures and associated storage buildings that may represent extended, perhaps matrilineal, families (see Hally 2002). At this broad scale, changes in house size seem to be reflecting major, society-wide changes in the organization of households and communities. After the Late Woodland period houses generally become larger and more widely spaced, suggesting that small nuclear or extended family households emerge as a basal of social organization. These patterns have been documented at the scale of the single site and region (see Pauketat 1998; Peregrine 1992;

79

Pluckhahn 2010; Wilson 2008), and there are certainly exceptions, but in terms of house size and the layout of houses in communities, the data suggest these patterns may be widespread. Orientation I measured the compass orientation of the long axis of each structure. Circular structures were only assigned an orientation if an entranceway was present. To maintain consistency and facilitate comparisons, measurements were recorded in compass degrees east of magnetic north. All values fall between 0 and 180 degrees. This is the same strategy Polhemus (1987) used at Toqua. I originally set out to record the direction in which each building faced, but this was surprisingly difficult. Many structures lacked clear entrances. Sometimes this was due to poor preservation, but in many cases there are multiple gaps in exterior walls that are equally likely candidates for entryways. In cases where household clusters can be identified one can infer the direction of possible entryways with reasonable certainty, but this was not possible for all sites. In all, I was able to record the orientation of 551 houses. I plotted the orientation of each functional class of structure (with at least five examples) for each time period (Figures 4.22—4.30). Since nearly all Middle Woodland domestic structures are round and lack clear entrances, I do not discuss their orientation. For Late Woodland domestic structures, there is a roughly bimodal distribution around 0 and 90 degrees (Figure 4.22). There is a similar bimodal distribution for Early Mississippian domestic structures, but more houses are oriented close to 90 degrees (Figure 4.23). Most non-domestic Early Mississippian structures are oriented at roughly 120 degrees, but many structures are also oriented at 40 and 150 degrees (Figure 4.24). The quality of excavation records for the Middle Mississippian structures limited my ability to record the orientation of domestic structures, but the non-domestic structures are primarily oriented at approximately 60 degrees and 160 degrees.

80

There is variation in the orientation of Late Mississippian domestic structures, but the long axis of many Late Mississippian domestic structures is oriented toward the southeast, at approximately 130 degrees. This reflects the tendency for the four corners of the houses to be oriented toward the cardinal directions. This demonstrates a change from the Early Mississippian period, when the more common objective seems to have been aligning the long axis of the building with the cardinal directions. At the Lamar occupation at Little River, Williams (2002:91) notes that one residential compound is oriented 23.5 degrees north of east, the angle of the summer solstice. If this pattern is more widespread in Late Mississippian times, it would also help explain the movement away from orienting the long axis of structures with the cardinal directions. There are fewer Late Mississippian non-domestic structures than domestic structures, but the non-domestic structures are much more clearly oriented toward the southeast (Figure 4.27). Late Mississippian storage buildings generally seem to be oriented around 80 and 130 degrees (Figure 4.28). The orientation of historic period domestic structures has a trimodal distribution, with most houses oriented toward 45 degrees, 90-120 degrees, and 160-180 degrees (Figure 4.29). Many domestic structures are oriented with their corners to the cardinal directions, but there is also some tendency to orient the long axis of structures to the cardinal directions. The four large townhouses are all oriented with their entrances toward the southeast (Figure 4.30). In sum, there is wide variation in the orientation of structures, but there seems to be a general tendency to orient either the long axis or the corners of domestic and non-domestic structures with the cardinal directions. Some of this variation may be explained by shared norms of house alignment, i.e. a tradition of orienting houses with the cardinal directions.

81

The orientation of structures is also strongly influenced by the overall layout of the settlement. Contemporaneous structures are often oriented in the same direction, or perpendicular to one another. This effect is especially visible at large, densely occupied sites. At Snodgrass, for example, the houses are arranged along a tight grid, which Price and Griffin (1979) originally described as street-like. However, even at smaller sites where spatial constraints would have been less onerous, houses tend to have either similar or perpendicular orientations. This can be seen especially well at the BBB Motor site, where the Edelhart and Stirling phase houses are generally aligned parallel to the low linear rise on which the village rests. Orientation is quite useful for intrasite analyses. At the King site, the orientation of primary domestic structures, rectangular structures, and burials was a vital clue for grouping buildings and burials into household clusters (Hally 2008:269-330). Interior burials were aligned parallel to the walls of primary domestic structures, and rectangular structures were generally oriented perpendicular or parallel to the primary dwellings. In the case of rebuilding, rectangular structures shifted orientation along with the primary dwellings. I explore the symbolic reasons behind house orientation in more detail in Chapter 7. Entrances Many entryways simply appear as gaps in the exterior wall post pattern, and do not have much additional elaboration. Entrances are especially difficult to identify in circular structures from the Middle and Late Woodland period. Early Mississippian wall trench structures often have open corners which appear to have served as entrances. Experimental reconstructions of tensioned wall trench houses support this interpretation (Blanton and Gresham 2007).

82

Parallel entry trenches commonly occur in Late Mississippian domestic and non-domestic structures in the Southern Appalachians. These features appear as a parallel pair of trenches running perpendicular to the exterior wall of houses. At Toqua, upright boards were placed between these trenches, creating a low, sheltered entrance (Polhemus 1987:220). It is possible that saplings may also have been bent between the trenches to make a similar entrance passage. Entry trenches are useful for determining the orientation of structures. Unfortunately, they are generally shallow, and are probably some of the first features to disappear with heavy plowing and erosion. Only 61 structures in the database have clear evidence of entry trenches. Entry trenches appear to be a defining feature of Late Mississippian domestic structures. Of the 61 structures with entry trenches, 54 are Late Mississippian domestic and non-domestic structures in village and mound-top settings at 9WH210, Brasstown Valley, Dog River, King, Loy, Rucker‘s Bottom, Rymer, Toqua, and the Warren Wilson site. There are earlier examples from the Early Town Creek and Late Town Creek/Early Leak phases at Town Creek, and there are also entry trenches at Lodge 1 at Macon Plateau. Entry trenches seem to be absent from late eighteenth century Historic Indian domestic structures, but early eighteenth-century winter houses at Ela and Ravensford have well preserved entry trenches. In addition to keeping out rain and wind, trench entryways may be a symbolic feature. These small, dark entrances help explain sixteenth-century Spanish accounts describing houses as cave-like (Clayton et al. 1993; see also Williams and Evans 1993:60-62). Hally (2002) has pointed out that these entrances, along with earth-embanked walls, may have given Late Mississippian houses a similar appearance to earth lodges and even platform mounds, imbuing ordinary domestic structures with the trappings of older, ceremonial architecture. Walls and Posts

83

Wooden poles and posts formed the framework of the Native Southeastern houses, and were the primary building materials along with wattle and daub, reed, and thatch. There are clear diachronic changes and regional variations in wall construction and in the alignment of exterior and interior posts. Patterns of variation in walls and posts seem to be society-wide and may be related to changes in social organization, subsistence, and settlement. It is also possible that architectural decisions that appear to be strictly mechanical, such as the total number of wall posts, may have symbolic meaning (Hally 2002:108). Southeastern archaeologists have identified three kinds of wall construction that capture most of the variation in house construction. First, walls can be made with single set posts. These walls fall into two types – those with closely spaced, small posts and those with more widely spaced, larger posts. Second, walls made of small, closely spaced posts are set in long, narrow trenches. In some cases the individual posts in the wall trenches can be identified, while in others they cannot. Third, some buildings are made with a combination of wall trenches and single set posts. This classification is based on the one first developed at Hiwassee Island and the Chickamauga Basin (Lewis and Kneberg 1946, Lewis and Lewis 1995). It is also similar to Pauketat‘s (1998) typology of wall types at Cahokia Tract 15A and Wilson‘s (2008:44-45) for early Moundville house walls. Walls were made with single set posts from the Middle Woodland through the Historic Indian period. Approximately 65 percent of all the structures in the database are built with single post walls, and every site in the database has at least one structure with this kind of wall. There are temporally sensitive changes in wall post diameter and spacing, discussed below. Wall trenches are an almost strictly Early Mississippian construction pattern, although some houses with wall trenches appear in later periods, and there is at least one possible Late Woodland wall

84

trench structure at the Summerour mound (Pluckhahn 1996). Out of 397 structures with wall trenches, 365 are from the Early Mississippian period. The remaining 32 occur during the Middle Mississippian period at Snodgrass and Kincaid and during the Historic Indian period at Yuchi Town and Jenrette. Sixty-one structures with both wall trenches and single set posts have been identified at 13 sites in the database, mostly during the Early Mississippian period, but also during the Late Woodland period at 1PI61, the Middle to Late Mississippian period at Martin Farm, and the Historic Indian period at Chota-Tanase, Ravensford, and Jenrette. The total number of posts (interior and exterior) associated with structures ranges widely, from a poorly preserved Early Mississippian wall trench structure at the Morris site in Kentucky with a single interior support post, to the Cherokee townhouse at Chota-Tanase, which had two construction stages and 1215 posts. Factors other than the original intended architectural pattern affect the total number of posts. Each rebuilding or repair episodes adds additional posts, while post-depositional processes like erosion and deep plowing remove posts. In cases of superimposed structures, analysts have to make judgments about post assignments, and walk a tightrope between overestimating and underestimating the number associated with each structure. Only in the most ideal cases, when structures have single building episodes and are not superimposed, can this figure be considered totally reliable. Out of all the structures in the database, the total count of posts could only be recorded for 464 structures. Not taking building episodes into account, the average number of posts per structure is 64, with a median of 42 (Table 4.1). A more meaningful figure is the average number of posts for all structures with a single building episode and a completely excavated structure pattern. In this case, the sample size drops to 318 structures. The average drops to 55 posts per structure with a median of 39, a minimum of 4, and a maximum of 329. In both cases, the

85

distribution of the number of posts is positively skewed, with the post count for most structures falling below the mean. The average number of posts in domestic structures increases from the Middle Woodland to the Early and Middle Mississippian periods, and then declines during the Late Mississippian and Historic Indian period (Tables 4.2—4.13). From a strictly architectural standpoint, this reflects a change in construction technique. Houses in the Middle Woodland period are built with widely spaced single posts. When wall posts are identified in Early Mississippian walls and wall trenches they are small and very closely spaced. During the Late Mississippian period and Historic Indian period there is return to more widely spaced, single set posts. Structure size also matters. Late Woodland houses generally have closely spaced posts, but are so small that the average number of posts is still close to the figure for the Middle Woodland period, despite the change in construction technique. Wall posts are often easier to identify than interior posts, and can be more clearly associated with particular structures. I was able to record the total number of wall posts for 502 structures. In most cases and in all the time periods, wall posts were regularly spaced, making it possible to infer the location of posts in the event of relatively straightforward overlapping structure patterns. The median number of wall posts per structure for all time periods is 29, and the distribution is positively skewed, with most of the 502 houses falling at or below the mean of 39 wall posts per structure (Table 4.1). As with the total number of posts, rebuilding and site formation processes influence the number of identified wall posts. If only complete structures with single building episodes are considered (n=365), the average number of wall posts changes only slightly to 38 (s.d. 37), with a maximum of 283 at the Early Mississippian Bessemer site.

86

The average number of wall posts per domestic structure increases steadily from the Middle Woodland to the Middle Mississippian period, and then declines during the Late Mississippian and Historic Indian period (Figure 4.31). Without considering rebuilding, this trend reflects general diachronic changes in wall construction. Middle Woodland, Middle Mississippian, Late Mississippian, and Historic Indian buildings, with larger, widely spaced posts have fewer wall posts than Late Woodland and Early Mississippian structures, with their closely spaced, smaller posts. Figures 4.32 – 4.37 show box plots for the number of wall posts per structure for each component. If only complete structures with single buildings episodes are considered, the trend is the same. Wall Post Size and Spacing Unlike many of the other variables, which are positively skewed, the average wall post diameter for all the structures in the database is normally distributed. I was able to record the average diameter of wall posts for 363 structures (Table 4.1). Across all time periods, wall posts range in size from 5 to 62 cm with an average of 17.3 cm (s.d. 5.9 cm). Figure 4.38 shows the mean and median wall post diameter for structures of all types in each major chronological period. Post spacing ranges from 6 cm to 4 m, with an average of 79 (s.d. 51 cm). Wall post diameter and spacing are the most commonly used lines of evidence for interpreting house construction methods. In his analysis of the architecture in the Hiwassee Island and Chickamauga Basin reports, Lewis identified two types of wall construction (Lewis and Kneberg 1946; Lewis and Lewis 1995). Walls were either composed of small (9-15 cm diameter), closely spaced, poles set in wall trenches, or larger (18-24 cm diameter) posts set farther apart. Webb (1938) developed a similar typology of flexed ―small log‖ versus rigid ―large log‖ construction. Both Lewis and Webb suggested that small diameter, closely spaced posts

87

represent flexed pole houses while structures with larger, more distantly spaced posts probably have rigid poles with a separate attached roof. This dichotomy of small pole versus large post architecture still has traction in current research (cf. Lacquement 2007a). These typologies are based on both archaeological and ethnohistoric evidence. Structures at Hiwassee Island and other Chickamauga Basin sites sometimes contained evidence of small horizontal wooden poles wedged in the wall trenches, presumably used for bracing upright bent poles (Lewis and Lewis 1995: 56). At the Hixon site, where horizontal wedges were apparently not employed, posts in the wall trenches leaned toward the inside of the structure, providing further evidence for a bent-pole technique (Lewis and Lewis 1995:56-58). The walls of a burned house at Hixon appear to be a latticework of long, small poles. Lewis and Lewis (1995:58-59) also cite accounts of house building by Le Page du Pratz among the Natchez and John Lawson in the Carolinas that indicate small diameter poles were used to build flexed structure. Recent experimental studies basically confirm these typologies, and provide practical insight on the constraints of house building. Lacquement‘s (2004) experimental house construction indicates that hickory and white oak poles larger than 8 cm in diameter are very hard to bend by hand, and that poles between 5 and 7 cm were ideal in terms of the tradeoff between flexibility and strength. Blanton and Gresham‘s (2007) experimental reconstruction of a wall trench house at Etowah showed that horizontal wedges were extremely important for bracing prefabricated, flexed pole walls, and that interior roof support posts were not needed for structural integrity. At well-studied, multicomponent sites in eastern Tennessee and western Alabama, diachronic changes in post diameter follow major changes in the cultural sequence. Lewis and Kneberg (1946) and Lewis and Lewis (1995) demonstrate a shift from structures with small,

88

closely spaced posts to structures with larger, more widely spaced posts from the Early Mississippian Hiwassee Island phase to the later Dallas phase, and Polhemus (1987) documents a similar architectural sequence at Toqua. Lacquement (2007b:66-67) identifies the same trend from the Moundville II to the Moundville III phase in west-central Alabama. In the Southern Appalachians, Early Mississippian structures with small, closely spaced posts have been identified at Etowah (Blanton and Gresham 2007), while Late Mississippian and early Historic Indian house are made strictly with large, widely spaced posts (see Hally 2002 for an overview). At an even broader scale, variation in wall post size and spacing coincides with major changes in the culture history sequence. Middle Woodland structures made of large, widely spaced posts are replaced by structures with smaller, more closely spaced single posts in the Late Woodland. During the Mississippian period there is a clear change in wall construction over time. Early Mississippian structures, both domestic and nondomestic, have walls with trenches and small, closely set posts. This building practice is abandoned by the Late Mississippian and replaced by walls made with larger, more distantly spaced, and presumably rigid posts. This practice continues in the Historic Indian period (see Figures 4.38 and 4.39). There is important regional variation within this broader pattern. In the eastern part of the study area the changes in house construction are different from the western part of the study area. In the American Bottom and west-central Alabama, Late Woodland domestic structures are small, rectangular, sometimes set in basins, and have walls made with closely spaced small posts. In the Early Mississippian period this construction style is generally replaced by wall trench structures with closely spaced small posts. This architectural transition is especially clear at the BBB Motor site and at Tract 15A at Cahokia.

89

In Georgia and western North Carolina, where Late Woodland houses are not as well understood, there is a different, more ambiguous change in architecture. Possible Napier phase domestic structures at Brasstown Valley are roughly 7 m diameter circular structures with widely spaced single posts. A single wall trench structure associated with Napier pottery was found on the top of Summerour mound (Pluckhahn 1996), and Markin (2007:78-79) describes two square, single post structures with Woodstock ceramics at the Rivermoore site. This small sample suggests there was more variation in architecture during the Late Woodland period in and around northern Georgia than in the American Bottom. Middle Woodland period structures are sometimes characterized as ephemeral, flexed pole buildings, but in many cases that description does not fit the archaeological data. For example, at the McFarland site, the five circular domestic structures are described as ―single wall post, tensioned pole structures‖ with ―a single row of relatively large (averaging 15 cm in diameter), evenly but widely spaced postholes‖ (Kline et al. 1982:22). The authors suggest that the walls are flexed because there are few clear interior support posts. However, with 15 cm posts spaced between 61 and 79 cm apart, the walls may have been rigid. Based on Lacquement‘s (2007b) experimental study, if these postholes are a close reflection of post diameter, the wall poles may have been too large to bend. While there is not a pattern of central support posts like those seen in the Late Mississippian period, there are some interior posts in each structure. These may or may not have been roof supports. Middle Woodland houses at the Duncan Tract site, the Ela site, Hickory Log, and Two Run Creek show a similar pattern of wall construction: posts are large (generally greater than 15 cm) and widely spaced (over 50 cm). There are some interior posts, although few are clearly roof supports. The Owl Hollow phase structures at the Banks III and Banks V sites have large, widely

90

spaced posts and a pattern of four interior support structures around their two hearths. These buildings almost certainly have rigid walls and a separate, attached roof. Based on our understanding of the relationship between post size and spacing and wall construction, Middle Woodland structures in northern Georgia, western North Carolina, and middle Tennessee may have been rigid post, rather than flexed-pole buildings, and in some cases they may have had separate, attached roofs. Next I examine architectural characteristics that structure the organization of interior space. These include house basins, interior posts, interior partitions and walls, interior features, interior hearths, and interior burials. Rebuilding and repair episodes, important architectural processes that might not have been visible from the outside, are also discussed in this section. Evidence of Basin House basins, also referred to as semi-subterranean or depressed floors, were identified for 259 structures from the Late Woodland, Mississippian, and Historic Indian periods. Rectangular structures with shallow basins are found in Late Woodland and Emergent Mississippian components in the American Bottom at Cahokia 15A and BBB Motor, in the Cairo Lowlands at the Hoecake site, and at 1PI61 in west-central Alabama. A single, keyhole-shaped structure at Kolomoki is the only Middle Woodland structure in the database with a basin. During the Early and Mississippian period, wall trench structures with basins appear widely over the study area at BBB Motor, Bessemer, Cahokia 15 A, Kimberly Clark, Macon Plateau, Martin Farm, Moundville, and Toqua. In contrast to the Late Woodland period, structures with depressed floors are much less common, and make up a minority of the structures at the sites. Cahokia is an exception. Many of the Lohmann, Stirling, and Moorehead-phase domestic structures have basins. The Middle Mississippian Powers phase structures at Snodgrass

91

all appear to have shallow basins, as does a single mound-summit structure from the Early Town Creek phase at Town Creek. The semi-subterranean floor is a defining feature of Late Mississippian winter houses in the Southern Appalachians. Most well-preserved winter houses have evidence of a basin. There are examples from Late Mississippian components at 9WH120, Alarka Farmstead, Brasstown Valley, Coweeta Creek, Dallas, Dog River, Dyar, King, Little Egypt, Pott‘s Tract, Rucker‘s Bottom, Rymer, and Toqua. As Hally (2002) points out, sixteenth-century Spanish accounts of native structures from the de Soto, Luna, and Pardo expeditions highlight this feature. Biedma notes, ―[h]ere we found a difference in the houses of the Indians; we found them as caves beneath the ground‖ (quoted in Clayton et al. 1993), and Martinez writes, ―the Indians took shelter in the huts… which were under the ground, from which they came out to skirmish with the Spanish‖ (quoted in Hudson 1990:332). Semi-subterranean domestic structures become less common during the Historic Indian period, and this type of construction seems to be abandoned by the late eighteenth century. Hally (2002) notes this trend in his comparison of sixteenth century and eighteenth century domestic structures in the Southern Appalachians, and I observe the same trend in the database. Cherokee domestic structures at Mialoquo, Townsend, and Toqua, which mostly date to the latter half of the eighteenth century, do not appear to have basins. Nor do Creek houses from the 1730s occupation at Hoithlewalli or the post-1750s occupation at Fusihathcee (Hally 2002:103). With one exception, the domestic structures at Chota-Tanase (see Schroedl 1986:238-240) do not have basins. In some cases this may be due to plow damage, especially at Chota-Tanase, but in general the pattern seems robust. Five Cherokee winter houses at Ravensford do have shallow basins, but these houses likely date to the 1730s or 1740s.

92

Interior Posts Interior posts are difficult to interpret and assign to individual structures. When structures have clear, distinct post patterns, interior posts can be understood to represent roof supports, upright posts for partitions, and supports for benches and other furniture. However, with the exception of interior roof supports, which tend to be large and evenly spaced, interior posts are not as regularly spaced as wall posts, even in isolated, well-preserved structures. It is nearly impossible to assign interior posts to individual structures in the event of overlapping or rebuilding. Southeastern archaeologists have struggled with this problem for decades. The description of interior posts at the McFarland site could stand in for many site reports: ―Multiple posts were present in the interior of most of the structures, but whether these represent roof supports or posts required for internal facilities such as benches or racks is unknown‖ (Kline et al. 1982:22). Despite these difficulties, there are discernible patterns of change in the arrangement of interior posts over time. To begin with, interior roof supports can often be picked out from palimpsests of interior posts based on their larger size and central and/or linear placement. Using these criteria—post size and central placement—I have identified general diachronic trends in the layout of interior posts. I was able to describe basic patterns of interior support posts for 38 components (Table 4.15). As with other patterns of construction, there are changes in the layout of interior posts that span much of the study area and track with major changes in the cultural historical sequence. During the Middle Woodland period there are three distinct patterns of interior roof support layout. At Hickory Log, Ela, McFarland, and Duncan Tract, circular buildings contain some interior posts, but not obvious interior roof supports. In contrast, at Banks III and Banks V

93

the structures that Faulkner (1988) identifies as ―double earth oven structures‖ have a consistent pattern of four large, deep interior posts set around the two hearths. A single Miller II-phase structure from 1GRX1 in west central Alabama has a very similar layout. At Garden Creek a Connestee phase structure beneath the mound has a square plan and a single center support post. Benyshek has recently uncovered several structures with a nearly identical floor plan at the Macon County Airport site. They appear to be associated with Middle Woodland period Connestee ceramics (Tasha Benyshek personal communication 2010). There are few clear patterns of interior support posts during the Late Woodland period (Table 4.15). At 1PI61 some of the small rectangular structures have interior posts, but their size and location do not indicate they are roof supports. The same is true of the small rectangular Edelhardt phase structures at BBB Motor. Round structures that may be associated with Napier pottery at Brasstown Valley lack clear roof supports. Early Mississippian wall trench domestic structures generally seem to lack interior roof supports. Their flexed pole design may have rendered heavy roof support posts unnecessary, especially in the case of small buildings. There are a few exceptions. Polhemus (1987) identifies a common pattern of paired roof supports spaced equally apart from the short walls along the long axis of some rectangular domestic and non-domestic structures at Toqua. There is also a single circular wall trench structure at Toqua with four possible roof supports. At Bessemer, two square-to-rectangular wall trench houses contain a circular array of six interior posts that could be roofs supports (see Table 4.9). Interior posts that are not roof supports may represent benches or other facilities, and in some cases may have formed interior partitions. The only components dating to the Middle Mississippian period with reliable data for identifying roof supports are the Early Town Creek phase at Town Creek, the Beaverdam

94

occupation at Rucker‘s Bottom, and the Wilbanks phase occupation at Hickory Log (Table 4.15). At Rucker‘s Bottom and Town Creek there are very large round buildings with inner rings of posts. These may be roof supports, but the posts are not especially large. At Town Creek there is also a square, earth-embanked, mound-summit structure with four central roof supports. The Wilbanks-phase domestic structures at Hickory Log also have four central support posts around a central hearth. This pattern continues into the Late Mississippian (Table 4.15). In the Southern Appalachians, Late Mississippian winter houses exhibit a consistent pattern of four large, deep, interior roofs supports placed in a square around the central hearth (Polhemus 1987). These roof supports are easily distinguished from smaller posts for benches that line the walls, posts for partitions radiating out from the central support posts to the exterior walls, and other posts of undetermined function. This pattern has been identified with varying degrees of confidence and precision at 9WH120, Brasstown Valley, Coweeta Creek, Dallas, Dyar, King, Little Egypt, Loy, Rucker‘s Bottom, Rymer, Toqua, Town Creek, and Warren Wilson. There is somewhat more variation in Late Mississippian roof supports when storage buildings and non-domestic structures are considered. Rectangular barbacoas generally seem to lack interior posts, although possible Rembert phase circular corn cribs at Rucker‘s Bottom may contain a single central roof support. Structure 17, a large public structure at the King site has eight interior roof supports around the hearth, and a large Dallas-phase structure at Toqua has six roof supports arranged in a square around the hearth. Circular domestic structures at upland Lamar phase farmsteads in the Oconee River valley may have three to six roof supports in the exact center of the building (Table 4.15).

95

The pattern of four interior roof supports surrounding a central hearth continues in Historic Indian period winter houses in the Southern Appalachians at the Alarka Farmstead, Coweeta Creek, Toqua, Chota-Tanase, Townsend, Tuckasegee, and Mialoquo (Table 4.9). Other, smaller interior posts appear to represent supports for benches along walls, interior partitions, and other undetermined facilities. Rectangular summer houses generally lack a clear pattern of interior supports, with the exception of possible paired roof supports along the long axis at Toqua and Townsend. The average number of interior posts per structure basically increases over time. The figure for the Middle Woodland period may be artificially inflated by the very high counts of interior posts from the Ela site. Mississippian period structures have more interior posts on average than houses from earlier and later periods. This not only reflects not only the architectural shift toward houses with interior roof supports, but also reflects an increasing occurrence of interior benches, partitions, and other interior facilities. The diameter of interior posts is generally similar to the diameter of wall posts. When structures are grouped by time period and functional class, the average sizes of interior and exterior posts are quite similar, and track with the overall shifts from small to large post wall construction. At the scale of individual structures, interior posts identified as central roof supports are usually larger than wall posts and other interior posts. During the Late Mississippian and Historic Indian periods, posts that appear to be supports for benches and partitions have a smaller average diameter than wall posts. Interior Walls and the Partitioning of Interior Space Like floor area, measures of the segmentation of interior space have received attention from archaeologists and other researchers looking for meaningful connections between

96

architectural remains and human behavior. Ethnoarchaeologists, cultural anthropologists, archaeologists, and other researchers working from a cross-cultural perspective have examined the division of space within buildings to make inferences about changes in storage (Kelly et al. 2005), sedentism (Flannery 2002), division of labor (Flannery 2002; Gougeon 2002), and even political complexity (Kent 1990). Some of the best data for understanding the relationships between segmentation and household organization come from Neolithic sites in the Near East, where millennia-old buildings are well preserved. For example, Byrd (1994) compares domestic architecture from three successive occupations at Beidha, a seventh millennia B.C. Pre-Pottery Neolithic B farming village in the Southern Levant. He finds that interior features become more common and interior space more segmented over time, showing ―an elaboration in the spatial organization of production activities and storage within domestic dwellings, suggesting that household autonomy increased over time‖ (1994:640). Flannery (2002) makes a similar argument using comparative architectural data from Pre-Pottery Neolithic B villages in the Near East and Early Formative village in Mesoamerica. In both regions he notes a shift from settlements with round, mostly open structures to villages with rectangular, segmented structures. Like Byrd (1994), Flannery (2002) argues that these changes in architecture reflect the privatization of storage and an increasing tendency to perform production and consumption activities inside. Similar processes may unfold in the Southeast during the transition from the Woodland to the Mississippian period. In the Southeast, lines of interior posts and the remains of interior walls provide evidence for the segmentation of houses. In general, only well-preserved daub walls can unequivocally be classified as partitions. Lines of interior posts may indicate a wall, but they may also represent supports for benches or other interior furniture. Keeping these limitations in mind, I adopted a

97

conservative approach for counting the number of walls or partitions in houses. First, if interior posts and partition walls could not be clearly associated with a structure, partitions were not counted and the value was recorded as null. Second, obvious daub or cane walls were counted. These only occurred in late prehistoric and early historic structures. Third, lines of closely set (less than or equal to 30 cm apart) interior posts were only counted as walls or partitions if they ran perpendicular to an exterior wall. Ideally this eliminated the error of counting benches and other furniture lining the walls of houses as partitions. I will have missed some interior walls, and may count a few non-walls as such, but in general this method should produce a consistent and conservative count. Hearths and interior features also effect the division of space in a structure. Interior hearths create a de facto separation of space within houses. The area around a hearth will primarily be used for tasks associated with heating and cooking. Occupants will gather near the hearth for warmth, but they will also avoid placing flammable objects too close, which will limit the activities that take place in this area. Interior features can create similar divisions of space. Open pit features will be avoided, and may also be associated with particular activities that divide houses into discrete activity areas. Unfortunately, while archaeologists can probably assume that hearths were active during the life of a house, pit features may have been dug and immediately filled back in. In this case, an interior feature would only have a temporary impact on interior space. I was able to record the number of walls and partitions in 502 houses. Examining variation in the number of partitions per house over time is enlightening. Figure 4.40 shows the mean and median number of interior partitions per domestic structure for each of the major time periods. In general, the number of partitions and walls seems to increase over time. This

98

observation fits with expectations from cross-cultural and ethnoarchaeological research, and with trends seen archaeologically in other parts of the world, especially the Near East. The average number of partitions in structures gradually increases from practically zero during the Middle Woodland period to nearly two per house during the Late Mississippian period. The maximum number of divisions found within a particular house during each period also increases, and jumps dramatically between the Middle Mississippian and Late Mississippian period. Several factors have to be considered when examining this trend in construction. More recent houses will be better preserved. Some of the increase in the number of observed walls may simply be the result of better preservation. Along the same lines, Woodland period houses may have been divided with perishable partitions that required few or no posts and therefore do not leave a substantial footprint. It is also important to note that cross-culturally, houses with open floor plans are still divided conceptually by the occupants (Kent 1990). Some Navajo hogans (circular or octagonal winter houses), for example, may have an open floor plan, but residents describe the dwellings as having male and female sides (Kent 1984; 1990:133). Similarly, according to Harner (1973:42-44), oval-shaped Jívaro houses in the Amazon are not partitioned, but are divided into a male end, a female end, and a central area for cooking and eating. With these caveats in mind, the increase in the prevalence of interior walls over time is compelling, and there are discernible patterns of change in the way interiors are divided. During the Middle Woodland period almost none of the structures have visible interior divisions. I was able to record the presence or absence of interior partitions in 34 Middle Woodland houses. Out of this subset, only one structure, a McFarland phase house at the Duncan Tract site, seems to have a possible interior wall, and its identification is tentative. In general, Middle Woodland houses have round, open floor plans.

99

Hearths would have created some division of space in Middle Woodland houses. Only 14 percent of the 66 Middle Woodland domestic structures in the database have a recognizable hearth. Three of these houses are Owl Hollow phase structures from the Banks III and Banks V sites on the Eastern Highland Rim in middle Tennessee. These houses had substantial interior hearths composed of two large rock-lined pits surrounded by four interior support posts. These houses would probably have been divided into at least two zones: the central area around the hearth and the remaining floor space outside the hearth and four support posts. Other Middle Woodland houses had much less substantial hearths, and while there would have been a similar delineation of space, it would not have been as pronounced as the Owl Hollow cases. It is unclear to what degree interior features would have created divisions in Middle Woodland structures. Of the 66 Middle Woodland domestic structures in the database, 25 have at least one interior feature that is not a hearth. These are pit features of varying depths and sizes, ranging from very small and shallow basins to deep cylindrical storage pits. From comparative ethnographic evidence, it seems very likely that Middle Woodland houses would have been conceptually divided by their occupants, but these divisions do not appear to have been made manifest with wood and daub walls. A central/peripheral division of space would have been present in structures with interior hearths, but this only accounts for about 14 percent of the Middle Woodland structures in the database. During the Late Woodland period, when structures are generally small and rectangular, there is also relatively little evidence for interior partitioning. I was able to record the presence or absence of interior partitions in 54 of the structures. Out of this subset, three structures had two possible interior walls, five had one possible interior wall, and the remaining houses had none. Three of the six houses with partitions are small, rectangular, Gainesville-phase houses from

100

1PI61. These structures have shallow basins and are fairly well preserved. The remaining structures with interior walls include an Emergent Mississippian structure from Tract 15A at Cahokia, and Edelhardt-phase house from the BBB Motor site, two houses from Rivermoore, one from Summerour, and one from Brasstown Valley. I was able to identify the presence or absence of a hearth in all 113 Late Woodland structures in the database. Only 11 structures show evidence of an interior hearth, and in all cases these were relatively simple fire basins. In these structures the hearth may have created a central/peripheral division of space. Floor area must be considered when examining interior segmentation. In contrast to many Middle Woodland and Middle to Late Mississippian houses, most of the Late Woodland structures are so small that dividing the structures into two or three sections may have been impractical, severely limiting movement and space for activities. Out of the eight structures with interior walls, five range between only 7 to 16.4 m2 in floor area. The remaining structure, a Late Woodland period house from Brasstown Valley, is round and measures 47.15 m2. Both the chronological placement and function of this house are uncertain. There is some continuity in the segmentation of interior space from the Late Woodland to the Early Mississippian period. Many of these structures are small and rectangular with open floor plans. I was able to record the presence or absence of interior partitions in 164 of the 524 Early Mississippian structures in the database. Out of this subset, only 22 structures have 1 or 2 interior walls. These particular structures vary a great deal in terms of size and function, from small domestic structures to large non-domestic buildings. As in the Late Woodland period, subdividing some of these structures may have been impractical. However, many of the structures that appear to have open floor plans are quite large, and size would not have been a

101

limiting factor. The structures without interior partitions (n=142) range in size from 1.17 to 184 m2, with an average size of 31.4 m2 (s.d. 29.8 m2). As in earlier periods, interior hearths and features probably contributed to some division of interior space. Interior hearths were identified in 81 Early Mississippian structures, and interior features were present in 228 structures. Middle Mississippian period structures show more signs of interior segmentation. The presence or absence of interior partitions could only be confidently identified for 16 structures from Town Creek, Ruckers Bottom, and Hickory Log, and many of these structures are not especially well preserved. However, from this small group, 9 out of 16 structures had 1 to 4 interior walls. Along with this change is an increase in structure size, making the division of interior space more tenable. During the Late Mississippian period there is a substantial increase in the segmentation of interior space. By this time, the winter house—a square house with rounded corners, a central hearth, four interior support posts, and interior daub walls—becomes increasingly common (see Gougeon 2007; Hally 2002; Polhemus 1987). I was able to record the presence of absence of interior partitions in 132 of the 295 Late Mississippian structures. Out of this subset, 35 structures had a least one interior partition. Among these structures the number of interior walls ranged from 1 to 12, with an average of approximately 4. Partitions were mostly found in the square domestic structures described above, but they also appeared in large square, circular, and rectangular non-domestic structures. Rectangular barbacoas, ―summer‖ houses, and small circular storage buildings were generally not subdivided into smaller rooms. There are probably many reasons for this difference, but it seems likely that the small size of the structures and differences in the activities that took place in them were major factors.

102

The interior divisions in Late Mississippian winter houses have been well documented. There are especially detailed descriptions of interior partition walls from Late Mississippian winter houses at King (Hally 2008), Little Egypt (Gougeon 2002:17-18; 27-68), Toqua (Polhemus 1987) and Loy (Polhemus 1998). These walls were sometimes made of split posts, rather than whole ones, interwoven with cane, and covered with daub. These wattle-and-daub partitions radiated out at 90 degree angles from the four central support posts, dividing the winter houses into as many as eight or nine discrete areas for sleeping, storage, and domestic activities (see Gougeon 2006:185-188). This change in the division of interior space, from basically open floor plans with a likely central/peripheral division of space focused on the hearth, to houses divided up into as many as eight or nine small rooms, is one of the most significant changes in house form. Prior to the Middle and Late Mississippian periods, occupants almost certainly divided their houses conceptually. It is also likely that during the Woodland and Early Mississippian period, occupants performed different types of activities in different parts of houses, but archaeological evidence for this is not as strong as in later periods. The partitions in Late Mississippian houses demonstrate a formal division of interior space, and studies of floor debris show that different types of activities took place in these different rooms (see Gougeon 2002). On average, Historic Indian structures have just as many interior partitions as Late Mississippian period structures. I was able to record the presence or absence of interior partitions in 57 of 110 Historic Indian structures. Out of this subset 31 structures had at least one interior wall or partition. Among these structures the number of interior partitions ranged from 1 to 16, with an average of approximately 3 per house.

103

In contrast to the Late Mississippian Period, partitions or walls were found in Historic Indian rectangular summer houses as well as in the circular, octagonal, or square winter houses. Out of 21 rectangular summer houses in which the presence or absence of partitions could be identified, 12 had one or two interior partitions or walls. In most cases these partitions run perpendicular to the long walls of the building, dividing the structure into two or three segments. Bartram described this pattern in Cherokee houses in 1775, noting ―Their private Houses of Habitations consist of one Large Oblong- Square Log Building, divided transversely into several apartments, and a round Hot-House stands a little distance off for a Winter Lodging House‖ (Bartram quoted in Waslekov and Braund 2002:183-184). There are partitions in the Historic period winter houses, but they are not as common or as clearly delineated as the wattle-and-daub partitions in Late Mississippian winter houses. Out of 23 winter houses in which the presence or absence of partitions could be identified, 10 had 1 to 3 interior partitions or walls. In general, Historic Indian winter houses are more lightly built than Late Mississippian winter houses and lack the semi-subterranean basin that contributes to good preservation (although there are some exceptions to this). These two factors may account for the observed reduction in interior partitions. Cherokee townhouses are divided into segments by lines of posts radiating out from central support posts. These are more likely supports for rows of benches as opposed to interior walls. The townhouse at Chota-Tanase was rebuilt at least once, and has a minimum of 18 lines of interior posts. Likewise, the townhouse at Toqua has 16 lines of radiating posts. Interior Hearths In general, interior hearths become increasingly common and better defined over time. Preservation bias plays a major part in this trend, although hearth have been found in Late

104

Archaic structures in Georgia (Ledbetter et al. 2009). Interior hearths are rarest in the Late Woodland period, when the ratio of structures with hearths to structures without hearths is .07, and most common during the Late Mississippian period, when there are 1.38 structures with hearths to every structure without (Table 4.16). There are comparatively fewer structures with hearths during the Historic Indian period, which may be the result of both poor preservation and the tendency for summer houses not to have hearths. Hearths change from fire basins and earth ovens in the Woodland period to more clearly defined prepared clay hearths during the Mississippian period (Table 4.17). Again, younger hearths will be better preserved, but the pattern is robust. There are no built up, prepared clay hearths during the Woodland period, and this form is dominant during the Mississippian and Historic Indian periods. It is not surprising that hearths become increasingly common during the Mississippian period given the broad trend of increasing sedentism in the Southeast. Taken at face value, the virtual absence of hearths in Middle Woodland structures located in places with cold winters (e.g. northern Georgia and middle Tennessee) suggests these structures were occupied only seasonally. It is possible, however, that these early hearths may not have survived. In northern Georgia, Epenshade (2008:136) notes that the ethnobotanical remains from Cartersville-phase structures at Hickory Log and possible Middle-Late Woodland structures at Brasstown Valley provide evidence for year-round occupation. In contrast, the near ubiquity of hearths in wellpreserved Late Mississippian structures suggests year-round occupation of the structures. There are, however, noteworthy exceptions to this big trend. The Middle Woodland Owl Hollow phase ―double earth oven‖ structures described by Faulkner (1988) at the Banks III and Banks V sites have substantial interior hearths. Moreover, these structures have large interior

105

support posts and regularly spaced wall posts. These strucures look like permanent, year-round dwellings. Similarly, a subset of Lamar phase structures uncovered by the Reynolds Plantation excavations in the Oconee River valley stand out as the only group of Late Mississippian structures that seem to lack hearths. Neither the large circular buildings nor the smaller rectangular structures at the Reynolds plantation sites contain hearths (Ledbetter n.d.). This is almost certainly due to effects of heavy plowing, as Hatch (1995) did identify hearths in circular structures at the Sweetgum site, also an upland Lamar farmstead. Interior Features Structures in the Southeast tend to have few interior features, defined here as pits or depressions used for storage and cooking, and molded clay seats and benches. The average number of interior features per structure, including hearths, is never greater than two during all the major chronological periods. The modal value for all periods except Historic Indian is zero. However, there is an increase in the frequency of interior features during the Mississippian period. Most Middle Woodland structures have no interior features at all, and except for two outliers, none of the structures have more than five. During the Late Woodland period the distribution is slightly less skewed, but there are still few structures with more than one interior feature. This positively skewed distribution persists throughout all periods, but beginning in the Early Mississippian period there are more structures with one or more features than there are structures with none. In the Late Mississippian many of the well-preserved winter houses have two or more interior features, and the same is true for the Historic Indian period.

106

Taken at face value, these figures show a greater tendency for the placement of features inside houses during the Mississippian period than the Woodland period. This matches the trend of increasing segmentation over time, and would offer further support for a general pattern of greater architectural investment in houses during the Mississippian period. It is important to note, however, that some of the largest, deepest interior features occur in early structures. At the McFarland site, deep cylindrical storage pits were found in most of the houses, and the double earth ovens in the Owl Hollow phase structures are massive. In contrast, Mississippian interior features tend to be smaller. They seem to include small storage pits, refuse pits, and perhaps the footprint of food processing and tool making areas. This may point to a general increase in interior storage and production and consumption activities over time. However, there may have been far more interior feature in structures than these figures represent. Except in the case of structures with basins, many shallow features in houses were probably lost to plowing and erosion. As a result, the figures for Late Mississippian winter houses are probably the most accurate, and the counts for other types of structures are probably low. It is also worth considering that throughout the sequence in the Southeast there were probably all sorts of domestic activities that did not leave pits or soil stains behind. Storage containers would have been hung from rafters, cordage may have been strung between support posts, shelves may have been built along walls, and lightly constructed wooden frameworks for weaving or drying could have been moved and out of houses. Among the Chickasaw, Adair observed vegetables stored under benches in winter houses to prevent them from freezing (Hudson 1976:216). The lack of interior features does not strictly mean that there were few interior storage, production, or consumption activities. Interior Burials

107

Broad changes in burial practices in the Southeast are fairly well understood, and other researchers have noted that people are not commonly buried inside ordinary domestic structures until the Late Mississippian period (see Hally 2002). This pattern emerges clearly in the database. From the Middle Woodland to the Early Mississippian period there are practically no domestic structures with clearly associated interior burials or cremations, but by the Late Mississippian period interior burials are commonplace (Figures 4.41). The only Middle Woodland structure with an interior burial is the from the Two Run Creek site, recorded by Wauchope (1966). The lack of interior burials in the Middle Woodland houses is not surprising given the number of cremations and burials that appear in contemporary earthworks and mounds. Only three Late Woodland structures contain interior burials: the poorly understood Tealphase structure at Town Creek, a single domestic structure at 1PI61, a small site in the Gainesville Lake area in Alabama, and one structure at Brasstown Valley. The house at 1PI61 appears to have been a primary residential structure, whereas the function of the early structure at Town Creek is uncertain. One possible Late Woodland structure Brasstown Valley site contains two interior burials, but the chronological placement of the burials is not certain. During the Early Mississippian period, only 11 structures contain burials. One is a tiny oval building from the Stirling-phase occupation of the BBB Motor site. The structure is clearly not domestic, and was almost certainly built in association with the burial. The ten remaining structures are all from the Hiwassee Island component at Toqua. Polhemus (1987) classifies all these structures as domestic. They range in size from 27.0 to 95.1 m2, and 6 of the structures also include interior hearths. Based on these features and their placement in the village they appear to be ordinary houses, rather than special-purpose, non-domestic dwellings.

108

Interior burials appear in 22 of the Middle Mississippian structures in the database. The bulk of these come from the large, non-domestic mortuary structures dated to the Late Town Creek/Leak phase occupation at Town Creek. Two burials appear in a large circular structure from the outer village area at Rucker‘s Bottom. At Hickory Log, two burials appear in a domestic Wilbanks-phase structure. During the Late Mississippian period there is a major shift toward the placement of burials beneath structures. Interior burials appear in 92 structures from 11 different sites. Most houses with burials (n=65) are the square, semi-subterranean winter houses. Seventeen rectangular barbacoas from Toqua and King contain burials. Only eight of the buildings with burials are non-domestic structures on mound summits and in plazas. There is one possible Late Mississippian structure of unknown function at Martin Farm with four interior burials. Interior burials continue in the Historic Indian period. There are burials in 17 structures from the Cherokee occupations at Chota-Tanase, Ela, Hiwassee Old Town, and Toqua. With the exception of one burial in the rectangular public structure at Chota-Tanase, the burials occur in domestic structures, 13 of which are rectangular summer houses and 3 of which are circular. The shift to interior burials is one of the strongest indicators of the emergence of the household, rather than a larger corporate kin group, as an important unit of social organization. Archaeologists have long argued that social groups use burials to claim territory and define group membership (see Charles and Buikstra 1983; Hodder 1984). In the place of large cemetery burials, which may place an emphasis on membership in a large corporate group, the practice of burying the deceased in house floors seems identifies houses and households as the more important locus of group identity. Building Episodes

109

Domestic and non-domestic structures and storage buildings were repaired in place, completely rebuilt in place, or rebuilt after being moved or shifted slightly from the original position. Repair and rebuilding creates palimpsests of superimposed posts, making it difficult for archaeologists to understand sequences of construction and repair. On the positive side, rebuilding and repair provides strong lines of evidence for understanding architectural investment, occupational duration, the domestic cycle, and architectural symbolism. Archaeologists in the Southeast generally report rebuilding episodes in three ways. In cases of obvious, complete rebuilding, some archaeologists will record a single structure with n number of rebuilding episodes. Others will assign each building episode a single structure number. Some archaeologists adopt a hybrid technique, and identify the superimposed building episodes as a single structure, but separately describe each building episode. To understand the variation in repair and rebuilding, I adopted the first strategy. For the analyses of rebuilding that follow, I record each structure as a single building with n number of repair or building episodes. This allows for the comparison of structures that were built once and never repaired or rebuilt with those that were repaired and rebuilt several times. I consider a repair or rebuilding event to constitute the repair or replacement of at least one entire wall, indicated by (at a minimum) a double line of posts or two parallel wall trenches. A limitation of this strategy is that I give equal weight to the replacement of a single wall and the complete reconstruction of a house. Rebuilding is virtually non-existent during the Middle Woodland period. There is no convincing evidence for wall repair or the rebuilding of structures in place. In some cases the walls of circular structures overlap slightly. Analysts generally interpret this as the seasonal

110

abandonment of a structure and perhaps a site, and the construction of a new house in the same area after a hiatus of some time. There is limited evidence of repair and rebuilding during the Late Woodland period. A single house at 1PI61 shows evidence of wall repair, and 14 percent Emergent Mississippian structures Cahokia 15A were repaired or rebuilt. In contrast, rebuilding becomes commonplace during the Mississippian period (Table 4.18). Twenty-one percent of the 459 Early Mississippian structures show evidence of rebuilding, and 55 percent of Early Mississippian components have at least one repaired or rebuilt structure. In some cases structures are repaired or rebuilt as many as five times in a single place. The rebuilt structures include 73 domestic structures, 23 non-domestic structures, and one structure of unknown function at Martin Farm. Late Mississippian houses are often repaired or rebuilt. Twenty-two percent of the 287 Late Mississippian structures have multiple building stages or repairs, and 25 of these structures were repaired or rebuilt three or more times. Rebuilt structures were identified in 41 percent of Late Mississippian components. Domestic structures accounted for most of the structures with multiple building stages (n=43), but non-domestic structures (n=12), storage buildings (n=7), and one building of undetermined function were also repaired and rebuilt. Rebuilding is much less frequent at Historic Indian sites. Only seven percent of structures show evidence of rebuilding or repair, and all of these structures were only repaired or rebuilt once. Out of this group of structures, five were domestic structures from Ela, Chota-Tanase, Tukabatchee, and the Jenrette site, and two were townhouses from Chota-Tanase and Toqua. Broadly speaking, Southeastern domestic structures were not repaired or rebuilt very often. These figures demonstrate that houses with clear evidence of rebuilding and repair are in

111

the minority at most well-documented sites. However, it is also clear that rebuilding and repair are more common during the Mississippian period than in earlier and later periods. The complete rebuilding of structures in place is more common during the Late Mississippian period. When combined with other lines of architectural evidence presented here (interior partitions, features, hearths, robust roof supports, and raw numbers of interior posts), the rebuilding data suggest that architectural investment reached a peak during the Late Mississippian period. Increasing permanency in structures may be related to broad changes in social and political organization. Blanton et al. (1993:75) argue that transition from wattle and daub structures to more permanent and durable adobe houses in the Valley of Oaxaca during Monte Albán Early I times (500 – 350 B.C.) was ―related to the increased security of life in the valley after the establishment of Monte Albán made boundary maintenance more effective. Administrators also may have encouraged the building of more permanent residences to facilitate control over the population.‖ Discussion: Major Themes in Architectural Variability At least four major themes about architectural variability emerge from this comparison. There is a general increase in architectural investment, some surprising continuity in building techniques, notable regional variation in the Late Woodland to Early Mississippian transition, and a trend toward increasing household autonomy during the Mississippian period. Several lines of evidence point to an increase in architectural investment over time, with a peak during the Late Mississippian period and a falloff during the Historic Indian period. In a study of the pithouse to pueblo transition in the upland Mogollon region of the Southwest, Diehl (1997:179) identified five lines of evidence for increasing architectural investment: more posts

112

per square meter, more expensive materials used for wall construction (in terms of labor cost), more formality of hearth construction, deeper house pits, and more evidence of rebuilding. Some of the same architectural changes occur in the Southeast from the Middle Woodland to the Late Mississippian period. Interior hearths become more defined, uniform, and common over time. There is more evidence of rebuilding and repair. I was unable to record the depth of hearth basins due to the variability of excavation records, but the frequency of house basins increases from the Woodland to the Late Mississippian period. The other lines of evidence are somewhat equivocal. I calculated the ratio of all posts (interior and exterior) to floor space for all the complete, single-stage structures for which reliable post data were available (Table 4.19). The average number of posts per square meter is highest during the Late Woodland period, due in part to the very small size of domestic structures. If the Late Woodland period is excluded, there is a gradual increase in post density over time, from 0.96 posts per square meter in the Middle Woodland period to 2.23 in the Late Mississippian. Post density drops to 1.63 posts per square meter in the Historic Indian period. Compared to other regions, building materials change little over time in the Southeast. There is a not a major shift from perishable to more permanent materials for construction, as in the Southwest. Wooden posts and poles are the primary materials for framing houses throughout the sequence, and cane, thatch, and reed seem to be used throughout for walls and partitions. Wattle and daub walls seem to be more common during the Mississippian period, but this may be due to better preservation. Knight has recently raised the important point that good evidence of daub from Mississippian houses is less abundant than commonly thought (2007:187). It is not surprising that Mississippian structures seem to be permanent, year-round dwellings, given the intensive maize cultivation. What is more remarkable is the degree of

113

architectural investment in some Middle Woodland structures, before the widespread adoption of maize. The large, widely spaced posts typical of Middle Woodland structures at McFarland, Hickory Log, Ela, and perhaps even Yearwood suggest the use of large, rigid wall posts and a separate, attached roof. In terms of broad spatial patterns of variation, Middle Woodland houses look fairly similar across most of the study area, with the exception of the unusual Owl Hollow phase structures (see Faulkner 1988, 2002). This is not the case with Late Woodland architecture, where there are substantial differences in the shape and size of structures in the eastern and western parts of the study area. By the Early Mississippian period, structures look very similar in both areas. It is possible that household organization was quite different in both areas during the important Late Woodland to Early Mississippian transition, when intensive maize cultivation was established. I continue this line of inquiry in more detail in Chapter 6. From a long view, there is more continuity in building traditions than we might expect. For all the major time periods, there is a break in structure size distribution between roughly 80 and 100 m2, and for each period this represents the break between ordinary domestic structures and larger non-domestic structures. Certain architectural elements appear early, and then seem to drop out of use and remerge later. Floor plans with four corner posts and a central hearth appear early in middle Tennessee and west central Alabama, then disappear, only to reappear in Middle to Late Mississippian period. Shallow structure basins or depressed floors are abundant during the Late Woodland, and deeper basins are a prominent feature of Late Mississippian winter house. Wall trenches first appear during the Late Woodland, spread across the Southeast during the Early Mississippian, fall out of use during the Late Mississippian, and reemerge on some Historic Indian sites.

114

Across the entire study area, changes in domestic structure size and spacing, burial practices, storage practices, and the interior partitioning of structures suggest that small nuclear or extended family households emerge as the basal important unit of social organization during the Mississippian period. The timing and nature of this change plays out differently in different parts of the study area, especially at the western versus the eastern sites during the Late Woodland to Early Mississippian transition. In general, major changes in structure form seem to track with the major cultural historical periods. One method for determining the statistical significance of these changes is regression with dummy variables. In this case, the major time periods can be used to assign dummy variables to each structure and compare the Middle Woodland, Late Woodland, Early Mississippian, Middle Mississippian, Late Mississippian, and Historic Indian structures. I performed linear regression with each of the architectural variables discussed above, using the Late Mississippian period as a reference category. I created a separate model for domestic and non-domestic structures. The results for domestic structures are shown in Table 4.20. Note that for every architectural variable except for interior features, membership in a certain time period accounts for a significant amount of variation. In some cases the R-squared values are quite low, but note that over 20 percent of the difference in wall post diameter, wall post spacing, and interior partitions may be explained by chronological variation. For non-domestic structures (Table 4.21), the results are slightly different. In this case, wall post spacing does not seem to vary significantly with time period, reflecting the general tendency for wall posts to be spaced more closely in non-domestic structures, even as wall construction techniques changed. The R-squared for partitions is extremely high, reflecting the transition from non-domestic structures with

115

essentially open floor plans early in the sequence to more heavily segmented structures by the Late Mississippian and historic period. These models drive home an important point: the changes we see in structures over time are significant. In the chapters that follow I look for the social factors behind these changes. I explore the role of classic prime movers, such as climate and material and technological constraints. Then I examine the effect of domestic production and consumption, household composition, architectural symbolism, status differentiation, and settlement pattern on house form.

116

Table 4.1 Summary Statistics for All Structures Variable Posts (n) Wall Posts (n) Wall Post Dia. (m) Wall Post Spacing (m) Int. Posts (n) Int. Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Paritions (n) Int. Burials (n) Int. Features (n) Orientation (deg)

Structures St. (n) Min. Max. Median Mean Dev. 467 1 1215 42 64.1 81.3 502 0 283 29 38.8 36.6 363 0.05 0.62 0.18 0.2 0.1 400 0.06 4 0.74 0.8 0.5 388 0 408 10 25.2 41.7 243 0.07 0.67 0.18 0.2 0.1 149 1 18.9 7 7.2 3.5 866 1.2 25 5.6 6.1 2.6 879 0.78 17.4 4.6 5.1 2.6 1054 1.15 299.8 28.4 38.9 38.9 457 0 18 0 0.7 2.0 667 0 40 0 0.9 3.0 709 0 30 1 1.2 2.1 566 0 180 91.5 83.8 49.6

Table 4.2 Summary Statistics for Middle Woodland Period Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 45 8 150 32 42 34 57 7 65 21 24 9.9 45 0.12 0.31 0.2 0.21 0.044 51 0.36 3.1 0.9 1.1 0.55 43 0 126 7 18 30 31 0.12 0.34 0.21 0.21 0.051 26 3 12 7 7.4 2.3 27 3 14 7.5 7.8 2.3 40 2.5 11 6.6 6.6 1.8 66 7.1 147 40 47 26 31 0 1 0 0.032 0.18 57 0 1 0 0.018 0.13 57 0 7 0 0.98 1.4 17 0 180 90 104 51

117

Table 4.3 Summary Statistics for Late Woodland Period Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 39 10 98 40 44 17 45 10 60 38 38 11 30 0.05 0.18 0.085 0.093 0.03 43 0.18 0.83 0.3 0.32 0.12 40 0 54 3 6.4 11 10 0.09 0.18 0.13 0.13 0.03 15 5 8 6 6.2 0.86 87 1.9 5 3.7 3.6 0.69 87 1.4 5.5 2.1 2.2 0.55 101 3 47 8.4 11 8.5 50 0 2 0 0.22 0.55 62 0 2 0 0.048 0.28 71 0 5 1 1.1 1.3 59 0 130 69 62 40

Table 4.4 Summary Statistics for Early Mississippian Period Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 56 1 205 40 49 37 59 0 171 31 35 29 47 0.08 0.21 0.14 0.14 0.032 21 0.15 4 0.39 0.58 0.81 43 0 140 8 15 23 34 0.11 0.26 0.15 0.16 0.035 0 . . . . . 266 1.9 12 5.2 5.6 1.7 270 1.3 10 4 4.1 1.6 268 3 118 21 26 18 116 0 2 0 0.086 0.31 134 0 4 0 0.16 0.64 161 0 8 1 1 1.6 143 0 175 94 74 50

118

Table 4.5 Summary Statistics for Early Mississippian Period Non-Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 34 15 329 105 121 69 34 0 283 91 110 66 35 0.07 0.25 0.15 0.14 0.037 34 0.08 2 0.18 0.39 0.46 32 0 181 6.5 18 33 22 0.08 0.67 0.19 0.2 0.13 16 5.8 15 9.9 9.6 2.7 67 1.5 20 9.1 9.6 3.4 65 0.78 17 8.8 9.2 3.3 82 1.2 300 85 96 56 36 0 2 0 0.44 0.65 41 0 1 0 0.024 0.16 48 0 7 1 1.4 1.8 35 2 180 116 96 49

Table 4.6 Summary Statistics for Middle Mississippian Period Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 12 17 92 71 65 26 15 17 71 38 39 15 6 0.15 0.23 0.2 0.2 0.029 16 0.4 0.9 0.66 0.68 0.14 4 7 30 21 20 9.5 1 0.2 0.2 0.2 0.2 . 15 5.2 11 8.2 8.2 1.4 95 1.2 6.6 4.6 4.6 1 95 1.5 7.3 4.6 4.7 0.97 115 1.9 86 23 27 15 6 0 4 1 1.3 1.5 16 0 13 0 3 4.3 15 0 2 0 0.33 0.62 4 44 170 57 82 59

119

Table 4.7 Summary Statistics for Middle Mississippian Period Non-Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 18 57 258 118 127 63 15 19 167 71 78 43 2 0.22 0.24 0.23 0.23 0.014 18 0.3 2.5 0.63 0.73 0.47 11 4 145 39 54 44 0 . . . . . 8 9.3 19 14 15 2.9 12 6.4 15 9.4 9.5 2.4 11 6.5 13 9.2 9.2 2.1 20 42 281 129 131 69 4 1 3 1.5 1.8 0.96 18 0 40 2.5 6.3 10 16 0 4 1 1.1 1.1 13 5 180 140 111 55

Table 4.8 Summary Statistics for Late Mississippian Period Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 95 2 499 51 79 86 135 2 231 28 36 32 103 0.09 0.59 0.19 0.2 0.052 112 0.06 2.1 0.82 0.88 0.38 82 0 408 23 42 63 45 0.07 0.28 0.19 0.19 0.037 9 3 10 8.6 8 2.4 171 2.6 9.9 6.5 6.4 1.5 169 2.2 10 6.1 6 1.8 203 6.6 126 41 43 21 69 0 8 0 1.7 2.4 149 0 18 0 1.4 2.5 144 0 30 1 1.5 3.4 143 0 174 89 88 47

120

Table 4.9 Summary Statistics for Late Mississippian Period Non-Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 15 6 272 55 89 96 13 3 157 40 54 42 14 0.13 0.62 0.18 0.22 0.12 6 0.14 1.1 0.45 0.57 0.36 14 3 232 18 54 76 12 0.13 0.29 0.2 0.2 0.051 1 14 14 14 14 . 23 5.2 15 9.4 9.6 2.7 24 3.7 15 7.9 8.5 2.7 24 27 226 79 90 49 26 0 12 0 1.5 3.3 26 0 35 0 2.7 7.7 27 0 9 2 2.4 2 20 66 128 118 114 15

Table 4.10 Summary Statistics for Late Mississippian Period Storage Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 37 6 51 11 17 13 23 6 23 10 10 3.6 2 0.12 0.23 0.18 0.18 0.078 15 0.3 1.4 0.9 0.87 0.29 16 0 51 26 26 16 15 0.16 0.26 0.22 0.21 0.031 2 1.7 3 2.4 2.4 0.92 36 2.4 6.6 4.2 4.3 0.94 36 1.6 4.3 2.5 2.6 0.61 38 2.3 26 9.9 11 4.2 36 0 0 0 0 0 37 0 8 0 1.5 2.3 37 0 4 0 0.62 1.2 36 0 145 84 87 43

121

Table 4.11 Summary Statistics for Historic Indian Period Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 78 14 180 48 51 29 72 1 88 24 29 16 57 0.12 0.22 0.16 0.17 0.024 56 0.23 2.4 1.1 1.1 0.48 72 0 126 17 24 23 56 0.13 0.24 0.16 0.17 0.027 26 5.8 9.5 7 6.7 0.76 61 4 12 7.6 7.4 1.8 62 2.3 9.5 4.6 5 1.6 87 12 86 35 35 13 45 0 3 1 1.1 1 78 0 7 0 0.64 1.5 78 0 11 1 1.6 2 65 0 180 90 88 59

Table 4.12 Summary Statistics for Historic Indian Period Non-Domestic Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures St. (n) Min. Max. Median Mean Dev. 4 222 1215 240 479 491 2 84 88 86 86 2.8 4 0.16 0.18 0.16 0.17 0.0096 2 0.3 1.3 0.8 0.8 0.71 3 29 160 147 112 72 5 0.16 0.42 0.16 0.21 0.12 2 17 18 18 18 0.92 3 13 16 15 15 1.5 3 6.1 16 15 12 5.3 5 79 263 231 210 74 3 16 18 16 17 1.2 5 0 1 0 0.2 0.45 5 0 2 1 1 0.71 4 100 145 122 122 23

122

Table 4.13 Summary Statistics for Historic Indian Period Storage Structures

Variable Posts (n) Wall Posts (n) Wall Posts Dia. (m) Wall Posts Spacing (m) Interior Posts (n) Interior Post Dia. (m) Diameter (m) Length (m) Width (m) Area (m2) Partitions (n) Interior Burials Interior Features Orientation (deg.)

Structures (n) Min. Max. Median Mean 9 6 34 16 17 9 6 29 16 16 2 0.18 0.18 0.18 0.18 3 0.79 1.6 0.87 1.1 9 0 5 0 1.2 2 0.2 0.21 0.2 0.2 12 1 3 2 2.1 1 3 3 3 3 1 3 3 3 3 13 1.1 9 3.6 4.2 8 0 0 0 0 8 0 0 0 0 8 0 1 0 0.13 1 150 150 150 150

123

St. Dev. 9.2 8.2 0 0.48 1.8 0.0071 0.46 . . 2 0 0 0.35 .

Table 4.14 Structure Shape, Sorted by Time Period and Functional Class Class

Middle Woodland (n=70)

Domestic

Circular

39

Rectangular

85

Rectangular

Oval

19

Circular

15

Square

57

Irregular

7

Square

3

Irregular

78

Keyhole

1

Total Non-domestic

66 N/A

Total Storage

0

N/A

0

Circular

3

Circular

1

1

Square

1

Rectangular

58

Circular

16

Square

14

Irregular

5

T-shaped

2 95

N/A

1 Irregular

8

Circular Rectangular

4

Class

Middle Mississippian (n=146)

Late Mississippian (n=295)

Domestic

Square Rectangular

74 26

Square (rounded corners) Rectangular

Circular

15

Total Non-domestic

Total Storage

Total Other

Total

Rectangular Circular Square Circular

Rectangular Irregular

115 9 8 7 24 2

2 4 1

8

Circular Irregular Square Rectangular Circular Rectangular Circular Square Rectangular Circular Irregular

5

124

169 44 6 3 222 19 11 1 31 35 2 1 38 2 1 1 4

0 0

Irregular

Total

270

405

1

0 Square

Early Mississippian (n=405)

103

0

Total Other

Late Woodland (n=103)

11 2 11 24

Historic Indian (n=110)

Rectangular Circular Square (rounded corners) Oval Circular Square Rectangular Circular Rectangular

Circular Rectangular Irregular

46 28 11 4 89 2 2 1 5 12 1 13 1 1 1 3

Table 4.15 Interior Support Post Patterns Site

Phase

Start Date

Unit

End Date

Unit

Hickory Log

Cartersville

150

BC

500

AD

Ela

200

AD

600

AD

Yearwood Garden Creek Banks III Banks V

Connestee McFarland-Owl Hollow Connestee Owl Hollow Owl Hollow

200 200 200 200

AD AD AD AD

500 600 600 600

AD AD AD AD

Duncan Tract

McFarland

200

BC

200

AD

McFarland 1GRX1

McFarland Miller II

200 400

BC AD

200 600

AD AD

Hiwassee Island

Hiwassee Island

975

AD

1200

AD

1PI61

Gainesville

1000

AD

1100

AD

Bessemer

Early Mississippian

1000

AD

1200

AD

Rucker's Bottom

Beaverdam

1100

AD

1350

AD

Town Creek

Early Town Creek

1150

AD

1250

AD

Toqua

Hiwassee Island

1200

AD

1300

AD

Town Creek

Late Town Creek/Leak

1250

AD

1350

AD

Warren Wilson

Late Pisgah

1250

AD

1450

AD

Hickory Log

Wilbanks

1300

AD

1350

AD

125

Interior Support Post Patterns Circular structures with possible interior supports, but no clear patterns. Circular structures with possible interior supports, but no clear patterns. Two large, roughly square structures with possible interior supports on center of long axis. Square single post house with single support post. Four central roof supports around double earth oven. Four central roof supports around double earth oven. Circular structures with possible interior supports, but no clear patterns. Circular structures with possible interior supports, but no clear patterns. Four central roof supports around large earth oven. Rectangular wall trench house with possible interior roof supports, not evenly spaced or well-defined Small rectangular structure with possible roof support in center. Possible array of six or seven supports in a circle in the center of a square wall trench structure. Large circular structure with 3 possible supports in center. May also be ring of supports in the big ones. Earth-embanked structures on mound with four central roof supports. Circular wall trench structure with four possible supports. Wall trench and small pole structure with two supports on long axis Four central roof supports around hearth; similar to winter house. Rectangular single post structure with 2 large posts on long axis. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern.

Site

Phase

Start Date

Unit

End Date

Unit

Dallas

Yarnell I/II

1300

AD

1525

AD

Toqua

Dallas

1300

AD

1620

AD

Coweeta Creek

Early Qualla

1300

AD

1500

AD

Rucker's Bottom

Rembert

1350

AD

1450

AD

Town Creek Brasstown Valley

Late Leak

1350

AD

1450

AD

Qualla-Lamar

1450

AD

1600

AD

9WH120

Lamar

1450

AD

1600

AD

9GE903 9GE1760

Lamar Lamar

1450 1450

AD AD

1600 1600

AD AD

King

Barnett

1450

AD

1625

AD

Coweeta Creek

Middle Qualla

1500

AD

1600

AD

Dyar Alarka Farmstead

Dyar

1520

AD

1580

AD

Late Qualla

1600

AD

1700

AD

Coweeta Creek

Late Qualla

1600

AD

1700

AD

Tukabatchee

Atasi

1600

AD

1715

AD

Toqua

Cherokee

1625

AD

1800

AD

126

Interior Support Post Patterns Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Four central roof supports around hearth; winter house. Similar pattern, but with six posts around hearth. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Four central roof supports around hearth; winter house. Circular structure (barbacoa?) with central support. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Large circular structure with 6 supports (4 in center, 2 equally spaced from center) Large circular structure with 3 to 5 supports in center. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Eight central roof supports in the public structure 17. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. On Dyar Mound: four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Rectangular house with two large supports on the long axis. Four central roof supports around hearth; winter house. In summer house, line of 3 support posts along long axis. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. One circular house with two roof supports in center, one square house with a roof support in center. Four central roof supports around hearth; winter house. In summer house, possible support posts on long axis, or lines of posts which may be partitions and/or supports.

Site

Phase

Start Date

Unit

End Date

Unit

Chota-Tanase

Cherokee

1625

AD

1800

AD

Townsend

Cherokee

1670

AD

1740

AD

Tuckasegee

Cherokee

1700

AD

1800

AD

Mialoquo

Cherokee

1760

AD

1776

AD

127

Interior Support Post Patterns Four central roof supports around hearth; winter house. Supports unclear in rectangular summer structures. Four and eight very large roof supports in Townhouse 1 and 2. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Rectangular house with two large supports on the long axis. Four central roof supports around hearth: typical Late Mississippian/Cherokee winter house pattern. Four clusters of post around hearth are likely roof supports; winter house. Possible lines of support post in rectangular structures.

Table 4.16 Structures with Hearths

Period MW LW EM MM LM HI

Structures Hearth Hearth Possible (n) (Y) (N) Hearth No Data Y/N Ratio 63 9 54 0.17 116 8 108 0.07 492 60 363 19 50 0.17 146 12 41 2 91 0.29 316 174 126 4 12 1.38 88 29 58 1 0.50

% Structures with Hearths 0.14 0.07 0.14 0.23 0.58 0.33

Table 4.17 Hearth Types

Period MW LW EM MM LM HI

Structures (n) 9 8 60 12 174 29

Fire Double Prepared Surface Fired Undefined Basin Oven clay Area 2 4 3 1 7 3 20 29 8 12 1 1 153 18 5 17 7

128

Table 4.18 Mississippian Period Rebuilding Episodes

Site Name 9GE1754 9GE1760 9GE1776 9GE1781 9GE222 9GE901 9GE903 9WH120 Banks V BBB Motor Site BBB Motor Site Bessemer

Period/Phase Bell Dyar Middle Lamar Middle Lamar Middle Lamar Dyar Middle Lamar Middle Lamar Banks Stirling Edelhardt Early Mississippian

Brasstown Valley Cahokia Cahokia

Middle Lamar Lohmann Moorehead

Structures with rebuilding (n)

Structures (n)

Ratio rebuilt structures: all structures

2 4 3 2 3 5 15 2 1 5 16

0 0 0 0 0 0 0 0 0 1 0

.. .. .. .. .. .. .. .. .. 0.20 ..

25

0

..

20 119 24

0 12 5

.. 0.10 0.21

41 7 10 5 29 1 4 2

0 0 3 1 2 0 4 0

.. .. 0.30 0.20 0.07

1 1 3

0 0 1

.. .. 0.33

14 1

6 0

0.43 ..

Jewell

Hiwassee Island Dallas Early Mississippian

22

2

0.09

Kimberly-Clark Kincaid

Hiwassee Island Early Kincaid

3 7

2 3

0.67 0.43

Kincaid Kincaid King

Kincaid Unknown Phase Late Kincaid Barnett

12 2 67

3 0 12

0.25 .. 0.18

Cahokia Cahokia Coweeta Creek Coweeta Creek Dallas Dog River Dyar Dyar Fernvale Go Kart Site Hickory Log Hiwassee Island Hiwassee Island

Cahokia Unknown Phase Stirling Middle Qualla Early Qualla Yarnell I/II Middle Lamar Dyar Duvall Early Mississippian Middle Lamar Late Wilbanks

129

1.00 ..

Site Name Little Egypt Loy Martin Farm Martin Farm Morris

Period/Phase Little Egypt and Barnett Dallas Mississippian II Mississippian I (Martin Farm) Early Mississippian

Structures with rebuilding (n)

Structures (n)

Ratio rebuilt structures: all structures

3 2 5

2 2 0

0.67 1.00 ..

1

0

..

12

2

0.17

102

46

0.45

35 14 3

3 2 2

0.09 0.14 0.67

5

2

0.40

Moundville Moundville Pott's Tract

Late Moundville I - Early Moundville II Early Moundville I Moundville I Barnett

Rucker's Bottom

Rembert

Rucker's Bottom Rymer Snodgrass Toqua

7 23 91 79

0 0 1 20

.. .. 0.01 0.25

Toqua

Beaverdam Mouse Creek Powers Dallas Late Hiwassee Island

25

3

0.12

Town Creek Town Creek Town Creek Warren Wilson

Early Town Creek Late Leak Teal Late Pisgah

20 4 1 12

1 0 0 3

0.05 .. .. 0.25

Moundville

Table 4.19 Average Post Density for Structures by Period

Period MW LW EM MM LM HI

Mean post density Structures (n) (posts/m2) 47 0.96 36 4.79 78 1.94 24 1.93 124 2.23 66 1.63

130

Table 4.20 Results of Multivariate Regression with Chronological Dummy Variables for Domestic Structures Model Area = ƒ (dummy vars for time) Posts = ƒ (dummy vars for time) Wall posts = ƒ (dummy vars for time) Wall post diameter = ƒ (dummy vars for time) Wall post spacing = ƒ (dummy vars for time) Partitions = ƒ (dummy vars for time) Interior features = ƒ (dummy vars for time) Interior burials = ƒ (dummy vars for time)

F 57.83 4.90 3.31 47.49 20.51 17.62 1.65 17.36

Prob > F R2 df n 0.000 0.253 5 840 0.000 0.071 5 325 0.006 0.042 5 383 0.000 0.457 5 288 0.000 0.259 5 299 0.000 0.221 5 317 0.146 0.016 5 526 0.000 0.151 5 496

Table 4.21 Results of Multivariate Regression with Chronological Dummy Variables for NonDomestic Structures Model Area = ƒ (dummy vars for time) Posts = ƒ (dummy vars for time) Wall posts = ƒ (dummy vars for time) Wall post diameter = ƒ (dummy vars for time) Wall post spacing = ƒ (dummy vars for time) Partitions = ƒ (dummy vars for time) Interior features = ƒ (dummy vars for time) Interior burials = ƒ (dummy vars for time)

131

F 6.10 8.13 2.76 5.99 1.98 53.27 2.19 3.50

Prob > F R2 df n 0.002 0.161 4 132 0.000 0.327 5 72 0.036 0.156 4 65 0.001 0.261 3 55 0.111 0.128 4 61 0.000 0.711 3 69 0.076 0.087 4 97 0.011 0.140 4 91

Figure 4.1 Histogram of Middle Woodland Period Domestic Structure Area

Figure 4.2 Box Plots of Domestic Structure Floor Area for Middle Woodland Period Components

132

Figure 4.3 Histogram of Late Woodland Period Domestic Structure Area

Figure 4.4 Box Plots of Domestic Structure Floor Area for Late Woodland Period Components

133

Figure 4.5 Histogram of Early Mississippian Period Domestic Structure Area

Figure 4.6 Box Plots of Domestic Structure Floor Area for Early Mississippian Period Components

134

Figure 4.7 Histogram of Early Mississippian Period Non-Domestic Structure Area

Figure 4.8 Box Plots of Non-Domestic Structure Floor Area for Early Mississippian Period Components

135

Figure 4.9 Histogram of Middle Mississippian Period Domestic Structure Area

Figure 4.10 Box Plots of Domestic Structure Floor Area for Middle Mississippian Period Components

136

Figure 4.11 Histogram of Middle Mississippian Period Non-Domestic Structure Area

Figure 4.12 Box Plots of Non-Domestic Structure Floor Area for Middle Mississippian Period Components

137

Figure 4.13 Histogram of Late Mississippian Period Domestic Structure Area

Figure 4.14 Box Plots of Domestic Structure Floor Area for Late Mississippian Period Components

138

Figure 4.15 Histogram of Late Mississippian Period Non-Domestic Structure Area

Figure 4.16 Box Plots of Domestic Structure Floor Area for Late Mississippian Period Components

139

Figure 4.17 Histogram of Late Mississippian Storage Structure Area

Figure 4.18 Box Plots of Storage Structure Floor Area for Late Mississippian Period Components

140

Figure 4.19 Histogram of Historic Indian Period Domestic Structure Area

Figure 4.20 Box Plots of Domestic Structure Floor Area for Historic Indian Period Components

141

Figure 4.21 Bar Chart of Average and Median Floor Area for Domestic Structures

Figure 4.22 Histogram of Structure Orientation for Late Woodland Period Domestic Structures

142

Figure 4.23 Histogram of Structure Orientation for Early Mississippian Period Domestic Structures

Figure 4.24 Histogram of Structure Orientation for Early Mississippian Period Non-Domestic Structures

143

Figure 4.25 Histogram of Structure Orientation for Middle Mississippian Period Non-Domestic Structures

Figure 4.26 Histogram of Structure Orientation for Late Mississippian Period Domestic Structures

144

Figure 4.27 Histogram of Structure Orientation for Late Mississippian Period Non-Domestic Structures

Figure 4.28 Histogram of Structure Orientation for Late Mississippian Period Storage Structures

145

Figure 4.29 Histogram of Structure Orientation for Historic Indian Period Domestic Structures

Figure 4.30 Histogram of Structure Orientation for Historic Indian Period Non-Domestic Structures

146

Figure 4.31 Bar Chart of Average and Median Number of Wall Posts for Domestic Structures

Figure 4.32 Box Plots of Domestic Structure Wall Posts for Middle Woodland Period Components

147

Figure 4.33 Box Plots of Domestic Structure Wall Posts for Late Woodland Period Components

Figure 4.34 Box Plots of Domestic Structure Wall Posts for Early Mississippian Period Components

148

Figure 4.35 Box Plots of Domestic Structure Wall Posts for Middle Mississippian Period Components

Figure 4.36 Box Plots of Domestic Structure Wall Posts for Late Mississippian Period Components

149

Figure 4.37 Box Plots of Domestic Structure Wall Posts for Historic Indian Period Components

Figure 4.38 Bar Chart of Average and Median Wall Post Spacing for All Time Periods

150

Figure 4.39 Bar Chart of Average and Median Wall Post Diameter for All Time Periods

Figure 4.40 Bar Chart of Average and Median Number of Partitions in Domestic Structures

151

Figure 4.41 Bar Chart of Average Number of Interior Burials for All Time Periods

152

CHAPTER 5 ENVIRONMENTAL FACTORS

How much of the variation in native architecture can be explained by variation in the late Holocene Southeastern environment? Some researchers have suggested that environmental variation has little effect on house form. In his brief cross-cultural analysis of the impact of climate on vernacular architecture, Rapoport (1969:18-24) argues that there are no simple deterministic relationships between climate and house form. Houses in similar environments worldwide vary remarkably in shape and construction technique, and similar house designs are used in diverse (and sometimes counterintuitive) ecological and environmental settings. Similarly, Susan Kent‘s (1990) cross-cultural study of housing and political complexity demonstrates that similar architectural designs cross-cut a wide range of environmental zones. Others suggest that environmental variation may explain the geographic distribution of certain house styles. In her study of the pithouse-to-pueblo transition in the American Southwest, Gilman argues that pit structures are more common in locations with a biseasonal climate and cold winters. Gilman (1987:540-548) used Murdock‘s (1967) Ethnographic Atlas to perform a cross-cultural comparison of the social and environmental context in which pit structures are used. Out of 862 groups, she found 84 that used pit structures as a primary or secondary residence, and determined that three factors always correlated with pit structures: ―non-tropical climate during the season of pit structure use, minimally a biseasonal settlement pattern, and the reliance on stored food while the pit structure is in use‖ (Gilman 1987:541). Specifically in terms of climate, she found that all but six cases of pit structure use occurred above 32 degrees latitude, 153

with most of the exceptions in high altitude areas. In all but one case, the pit structures were used as winter houses, suggesting that there is a correlation between cold weather and the use of pit structures. Gilman argues that underground houses can be more thermally efficient than aboveground houses, since the soil temperature below ground generally varies less than air temperature (Gilman 1987:542). An experimental study of heat loss by transmission in subterranean versus above-ground houses by the Underground Space Center at the University of Minnesota (1979) and a comparison of the heat loss in pithouses versus pueblos by Farewell (1981) both show underground houses to be more thermally efficient than above-ground houses. Given this physical advantage, it seems possible that differences in temperature might determine the distribution of semi-subterranean houses. While the basins of the semi-subterranean houses of the Southeast are not as deep as underground houses tested in the experimental studies cited above, they were still probably more thermally efficient than similar houses without basins. In the Southeast, there have been few quantitative attempts to evaluate the effect of environmental variables on house form, but some have suggested that synchronic or diachronic variation in temperature may explain some differences in house construction. The global warming associated with the Medieval Warm period (ca. A.D. 800 to 1300) and the global cooling associated with the Little Ice Age from the fifteenth to the eighteenth centuries are sometimes invoked as a prime movers for large-scale demographic and cultural changes in the Woodland and Mississippian periods (Anderson 2001; Little 2001; and cf. Cobb and Butler 2002:236-237). Lacquement (2007b:69-70) notes that the first examples of heavily built Late Mississippian structures in the Southern Appalachian region coincide roughly with this cooling event. However, he also notes that in other parts of the Eastern Woodlands, perhaps most notably

154

in the much colder Northeast, people continue to build flexed pole houses without daub walls during the Little Ice Age (Lacquement 2007b:70). In his comparison of sixteenth-century and eighteenth-century domestic structures in the Southern Appalachians, Hally (2002:91) notes that sixteenth-century Spanish accounts from the De Soto, Luna, and Pardo expeditions describe Late Mississippian domestic structures as wellinsulated and ―cave-like.‖ Especially relevant is Biedma‘s observation that domestic structures changed as the De Soto expedition moved northward from Florida into southern Georgia in 1540. Biedma stated, ―here we found a difference in the houses of the Indians; we found them as caves below the ground‖ (quoted in Clayton et al. 1993:228). This historic observation, combined with the archaeological observation that most (but not all) of the Late Mississippian sites with semi-subterranean domestic structures are located north of the Fall Line, suggests that climate may have had some influence on the distribution of domestic structures with basins. In this chapter I test some of these conflicting hypotheses regarding the effect of climate on house form. First I describe the environmental context of house building in the Southeast, especially as it relates to variation in temperature, rainfall, elevation, and natural resources for building materials. I review historic and archaeological data which indicate that native house builders used a relatively narrow range of plant and wood materials for house construction. Next, I perform multivariate regression to determine the effect of several environmental variables on different aspects of house form, such as floor area and post density. In general, my findings concur with Rapoport‘s (1969) statement that there is not a simple predictive relationship between environment and house form. However, certain aspects of house construction, in particular post density, may vary predictably with temperature and rainfall. Based on Gilman‘s (1987) research on pit structures and observations about the distribution of Late Mississippian

155

domestic structures with basins (Hally 2002:91, Hally 2008:68; Laquement 2007), I hypothesized that the distribution of houses with basins or sunken floors would track with latitude, elevation, and temperature. I find that the use of house basins may be better explained by social variables. Environmental Context The late Holocene Southeast had physical constraints that affected house building in at least two ways. First, Southeastern Indians built their houses to accommodate and withstand a range of climatic conditions particular to the region. Second, Southeastern Indians constructed their homes using locally available building materials, primarily wood, cane, grasses, mosses, and clay. Many of the similarities and some of the differences in Southeastern houses can be explained by the physical properties of these particular natural resources. Paleoenvironmental reconstructions suggest that Woodland, Mississippian, and Historic Indians communities in the Southeast were situated in biomes with essentially modern environmental conditions. The Holocene is broadly characterized as a period of warming and increasing seasonality, with modern climate conditions established by 3000 B.C. (Delcourt and Delcourt 1983, 1985). Recent studies show that at closer resolution, there were many local exceptions to these general trends (Goman and Leigh 2004; Lamoreaux 1999; Leigh 2006; Lowery 2001; Saenger et al. 2006). Using pollen data, Delcourt and Delcourt (1983, 1985) and Watts (1980) argue that by 8000 B.C. across much of the Southeast, mesic forests of oak, hickory, beech, elm, and ironwood replaced the spruce forests of the Late Glacial Period, and that a gradual warming trend began. The Early Archaic period appears to have been a time increasing seasonality, and may have been a wetter period than previously thought. Studies of fluvial geomorphology in North Carolina

156

(Goman and Leigh 2004; Leigh and Webb 2006) suggest that the Early and Middle Holocene were marked by relatively heavy seasonal rainfall. The Middle Archaic period in the Southeast has conventionally been described as a time of increasing warming and aridity associated with the Hypsithermal episode (Watts 1980). According to Delcourt and Delcourt (1983, 1985), climax hardwood forests were established in lowlands, and pine forests were established in uplands. In general, pine forests seem to have expanded at the expense of oak in response to the warmer, drier conditions (Delcourt and Delcourt 1985; Watts 1980). In the Southern Appalachian Mountains, species that currently live in the Coastal Plain dispersed into poorly drained environments and sheltered coves, and remain there today (Delcourt and Declourt 1985:20). This accounts for some of the remarkable modernday species diversity in pockets of the Ridge and Valley and Southern Appalachian Mountains. In recent years, this general picture of Middle Archaic aridity has been reconsidered. Harrison et al. (2003) indicate that the widespread aridity of the Hypsithermal episode may not have affected the Southeast, largely because of the influence of the Gulf of Mexico and the Atlantic Ocean. Numerous paleoenvironmental studies indicate that the Middle Holocene was a time of increased seasonality and monsoonal rainfall patterns (Goman and Leigh 2004; Jones 2005; Lamoreuax 1999, Leigh and Feeney 1995; Leigh et al. 2004). By 3000 B.C., roughly the start of the Late Archaic period, modern climatic conditions were generally established (Alvarez Zarikian et al. 2005; Delcourt and Delcourt 1983, 1985; Lamoreaux 1999). The modern southern and eastern boundaries of the deciduous, mesic forests appear to have been relatively stable by 3000 B.C., and the Prairie Peninsula, which covers most of Illinois, reached its modern extent (Delcourt and Delcourt 1987:99). Likewise, the boundaries of the southern evergreen forests of the Atlantic Coastal Plain changed little after this time

157

(Delcourt and Declourt 1987:99). In general, the biomes of the Mid-to-Late Holocene were similar to those of today, but newer paleoclimate research shows that climate conditions and perhaps plant communities were not as stable as once thought (Dunbar 2000). In an interdisciplinary literature review of climate change data, Little (2001:25) concludes:

In southeastern North America, late Holocene climate changes do not coincide with environmental changes as dramatic as those that occurred in the arctic regions. There are no major shifts in vegetation distributions in the Southeast. Instead, there were probably only perturbations in the relative abundance of certain resources accompanying changes in temperature (i.e., length of growing season) and precipitation‖ (2001:25).

The paleoenvironmental studies cited here suggest that is appropriate to use modern climate data to describe the Woodland, Mississippian, and Historic Indian environmental context, given the caveat that their environments were probably richer and more diverse than those of today. The study area encompasses most of the major physiographic regions in the southern half of the eastern woodlands (Figure 5.1). Moving from east to west, sites are located on the Atlantic and Gulf Coastal Plain, the Piedmont, the Blue Ridge, the Ridge and Valley, the Appalachian Plateau, the Interior Low Plateau, the Central Lowlands, and the Mississippi Alluvial Plain. The westernmost site, Snodgrass, lies just below the Ozark Mountains. Table 5.1 lists the sites in the database by major geographic province and provides summary environmental data.

158

Elevations range from a minimum of 33 m above mean sea level at 1GR1X1, a Middle Woodland site located near the confluence of the Tombigbee River and a small stream in Greene County, Alabama, to 787 m above mean sea level at the Garden Creek site, a Middle Woodland and Mississippian mound site in the Blue Ridge Mountains in Haywood County, North Carolina. Understanding the composition of plant communities is vital for understanding both the broader environmental context of the region and the materials available for house construction. Sixty years after its publication, Braun‘s Deciduous Forests of Eastern North America continues to be the standard reference for describing the forest communities of the Eastern Woodlands prior to major European settlement (Dyer 2006). Braun (1950:8) used published literature and her own samples from old growth stands to ―present information on the original forest pattern of eastern North American and on the composition of virgin forests‖ (original italics). The study area encompasses six of Braun‘s (1950) nine forest regions for the eastern North America (Figure 5.2). Moving from east to west, they are: the Southeastern Evergreen region, the Oak-pine region, the Oak-chestnut region, the Mixed Mesophytic region, the Western Mesopythic region, and the Oak-hickory region. Braun‘s classification provides an idealized, synchronic view of the composition of climax forests prior to European colonization. While these regions should not be reified, they provide a general picture of the tree species that would have been available to indigenous house builders. In the Southeastern Evergreen region, upland and lowland communities are dominated by coniferous and broad-leaved evergreen trees and shrubs (Braun 1950:280). Longleaf pines and live oaks (often draped with Spanish moss) dominate much of the landscape (Braun 1950:280). Braun notes that the long-leaf pine communities are actually an ―edaphic climax so modified and stabilized by recurring fires that it is considered a fire subclimax‖ (1950:280). According to

159

Braun the forest communities in this region are quite diverse, and include: longleaf pine forests, pine flats, savannas, and bays, loblolly and pine-hardwoods forests, slashpine forests, bottomland forests, swamp forests (cypress and tupelo), hardwood bottoms, ridge bottoms, oak hickory forests, mixed mesophytic hardwoods, beech-magnolia forests, and evergreen oak forests. Braun considers the Oak-Pine forest region a transition zone between northern deciduous forests and the Southern evergreen region (1950:259). There are oak-hickory climax forests, with white oak as the dominant species, and widespread communities of pine in the secondary forests. The most common hardwood species are oaks (white, black, post, red, and southern red) and hickories (white and pignut), which are usually accompanied by sourwood and sweet gum (Braun 1950:259). Braun (1950:260) divides this region into two main sections, the Atlantic Slope and the Gulf Slope. Within the study area, forest communities in the Atlantic Slope include mixed oak-hickory and pine forests and more mesophytic forests in the Piedmont uplands, and Piedmont bottomland forests with riverbirch, black willow, cottonwood, and sweet gum (Braun 1950:267). Forest communities in the Gulf Slope section vary by physiographic region, with different combinations of hardwood and pine species depending on the elevation, slope, and underlying soils. The Oak-Chestnut region covers much of western North Carolina and eastern Tennessee and a small part of northern Georgia. In general, chestnut, red oak, chestnut oak, and tuliptree are the most frequent species, with oak-chestnut forests especially common on the slopes (Braun 1950:37). Braun divides the Oak-Chestnut region into several smaller physiographic provinces, two of which, the Southern Appalachians and the Ridge and Valley, fall in the study area. The Southern Appalachians include the mountains south of Roanoke Gap (the southern portion of the Blue Ridge). Braun identifies three main forest types in this area: a northern forest,

160

mostly at high altitudes, comprising spruce-fir, hemlock, yellow birch, and northern hardwood; moist slope and cove forest, including buckeye-basswood, northern red oak, cove hardwood, yellow poplar, cove hemlock, white pine, chestnut, and white oak; and dry slope and ridge forest, comprising chestnut oak, black oak-scarlet oak, and pitch pine-mountain pine (1950:197). She also notes the occurrence of balds, which are densely covered with grass and shrubs (Braun 1950:212-213). In the Ridge and Valley province, most of the ridges are home to oak-chestnut forests. Mixed mesophytic communities are more common on the ridges of the southern portion of the Ridge and Valley than further north. In the valley floors, oak forests are dominant, with white oak as the dominant species. Only the narrow southern end of Braun‘s Mixed Mesophytic region runs through the study area, along the Cumberland Mountains and the Cumberland Plateau. These forest communities only develop on moist, well-drained soils, and are dominated by a mixture of species, including beech, tuliptree, basswood, sugar maple, sweet buckeye, chestnut, red oak, white oak, and hemlock. In the Cumberland Mountains there are patches of pine mixed in with species-rich deciduous forests, and in the Cumberland Plateau there is a higher frequency of oaks and pines relative to the rest of the region (Braun 1950:35, 51). All of the archaeological sites in middle and western Tennessee and Kentucky fall within Braun‘s Western Mesophytic Forest region. According to Braun (1950:122), the boundaries of this region essentially follow those of the Interior Low Plateau, bordered on the east by the Cumberland Plateau and on the west by the Mississippian River. This is a transition zone, with a mosaic pattern of vegetation that shifts from mixed mesophytic communities in east to oak and

161

oak-hickory communities in the west (1950:36, 122-123). In addition to upland forests, there are alluvial swamps, most notably on the Green River and in the Lower Wabash River Basin. Within the Western Mesophytic Forest region, the Bluegrass region, the Nashville Basin, the Mississippian Plateau, and the Mississippi Embayment region fall in the study area. Blue ash and bur oak are the most common tree species in the Bluegrass region, and in the Nashville Basin cedars are dominant, along with a mixture of deciduous trees and shrubs. The Mississippian Plateau includes the Eastern Highland Rim, where some of the best examples of Middle Woodland houses are found (Faulkner 1988). In this region, mixed mesophytic forests of beech, oak, oak-hickory, and oak-chestnut are dominant. There are also some areas of prairie, cedar barrens, and swamp forests. In the Mississippi Embayment region, there is a pattern of mosaic vegetation including prairies, oak-hickory forests in the uplands, swamp forests, and mixed mesophytic communities (Braun 1950:125-161). The westernmost part of the study area falls within Braun‘s Oak-Hickory Forest region, where oaks and hickories are the dominant species. This region includes parts of the upper Mississippi Valley, the Ozarks, and the American Bottom, and extends northward through the Prairie Peninsula into Michigan and Wisconsin. Cahokia and surrounding sites fall within this forest region. In bottomland forests, white oak are abundant, and silver maple, elm, willow, ash, pin oak, and river birch are common (Braun 1950:162, 181). In the Ozark uplands, there are stands of pine in the oak-hickory forests, and Pauketat (2004:34) suggests that some of the pine used for construction at Cahokia may have been brought in from the St. Francois Mountains, over 100 km to the southwest. Data collected from 1960 to 1990 by the National Climatic Data Center provide an overview of modern environmental conditions in the study area (National Climatic Data Center

162

2010). Table 5.1 shows the environmental variables recorded for each site. Temperatures across the region are relatively mild, varying with latitude and elevation. Average annual temperatures range from 45 to 50 degrees F in the Blue Ridge Mountains to 65 to 70 degrees F in the coastal plain of southwest Georgia. The median number of frost-free days ranges from 121 to 180 in the Blue Ridge Mountains to 240 to 270 in the southern Coastal Plain. In terms of extreme cold, the average daily minimum temperature in January ranges from 15 to 25 degrees F in the coldest, northernmost parts of the study area to 32 to 40 degrees F in the warmest, southernmost parts. The annual mean extreme minimum temperatures ranged from -10 to 0 degrees F in the Blue Ridge Mountains to 10 to 20 degrees F in the southern Coastal Plain. Building Materials Structures in the native Southeast were constructed with locally available materials. As Braun‘s (1950) study of forest distribution shows, builders in the Southeast had access to a wide array of tree and plant species for the frames, roofs, walls, and furnishings of houses. However, ethnohistoric and archaeological evidence indicates that a relatively narrow set of species were used as building materials. Early historic accounts of native structures provide insight into the choice and use of building materials. These accounts undoubtedly have some biases and inaccuracies, as European and Euroamerican observers filtered their observations through a mental template of European house building. By the time of the later accounts, i.e. those from the eighteenth century and beyond, many Southeastern Indians had access to iron woodworking tools, which may have changed the range of wood and plant resources they chose to exploit in terms of species and size. Table 5.4 provides a list of building materials from detailed ethnohistoric accounts of indigenous architecture. Despite spanning a wide range geographically, from Mississippi to the

163

North Carolina coast, the materials used for particular aspects of house construction are quite similar (Adair 1968; duPratz 1972[1725]; Swanton 1946; Speck 1909). Small, bendable poles cut from hickory, pine, and cedar were used to build the framework for flexed-pole buildings, while pine, oak, locust, and sassafras were used for the support posts and rafters in rigid post structures. Posts, poles, and rafters were bound with splints made of cane, splinters of white oak and hickory, split saplings (species undefined), bark, and mosses. Builders covered house walls with daub, bundles of grass, pine or cypress boards, and bark from cypress, red and white cedar, and pine trees. Roofing materials included cane mats, bundles of grass, poplar boards, long poles covered with daub and grass thatch, and bark from pine and cypress trees. Additional building materials mentioned by Adair (1968:420) include a door made of poplar planks and bound with buffalo hide. Burned structures generally offer the best archaeological evidence of building materials. While it is often difficult to identify the species of burned structural elements, it is sometimes possible to make genus-level identifications. In general, archaeologically recovered building materials are similar to those recorded in ethnohistoric accounts. Table 5.5 shows some plant and tree genera identified from archaeological contexts. In a study of burned structural remains from several sites in the American Bottom, Simon (2002) identifies white oak and hickory as primary building materials, and reports the possible use of red oak, ash, cedar, and willow. At the Rench site in Illinois, McConaughy (2007) identifies different genera used for different parts of a Mississippian ―wigwam‖ style domestic structure. Curved oak and hickory poles were bent to form the long walls, while upright hickory and walnut poles formed the end walls. Crossbeams appear to be have been hickory, basswood fibers were used for cordage binding wall members

164

with cross beams, and bulrush thatching covered the roof and walls (McConaughy 2007:109). At Loy, Polhemus (1998) records burned thatch and grass used for roof and wall covering. An analysis of 3288 charcoal samples from 26 of the 85 excavated structures at Snodgrass provides one of the best archaeological records of construction material (O‘Brien and Cogswell 2001:150-157). In all, 22 taxa were represented. The most common was red oak, which comprised 41.0 percent of the sample, followed by hickory (19.2 percent), white oak (12.3 percent), ash (8.8 percent), bald cypress (7.3 percent), and persimmon (4.5 percent). The remaining taxa, with percentages at 2 percent or below, were walnut/butternut, elm, locum, cottonwood/poplar/willow, maple, pine, sweet gum, tupelo, hackberry, sassafras, holly, buckeye, catalpa, hornbeam, and birch (O‘Brien and Cogswell 2001:154). O‘Brien and Cogswell (2001) note that while red oak and hickory were the most commonly exploited taxa, they are not the most abundant in the local area, according to General Land Office (GLO) survey notes (2001:156). Sweet gum, the highest ranked taxon according to GLO records (19.3 percent of the local trees) only accounted for 0.2 percent of the charcoal. This provides further evidence that tree species were selected based on their properties for construction, rather than their relative abundance. Considering the diversity of plant and wood species available to native house builders, ethnohistoric and archaeological evidence suggests that people used a relatively narrow range of materials. Oak and hickory were most commonly used for bent poles. Pine, cedar, and locust were more commonly used for upright posts and roof supports. Bindings for wall and roof elements were often made with cane and split saplings. Wall and roof coverings sometimes incorporated daub, and often included grass thatching, cane mats, and bark. The preservation conditions of the houses in the database are too variable to do a systematic study of roof and wall

165

coverings. However, when the 166 burned houses in the database are examined, it appears that roof and wall materials are roughly similar across a range of environmental conditions. Burned cane and thatch, which appear to have been used for wall and roof insulation, appear as far north as Cahokia in Illinois and as far south as Dog River in Georgia. In general, native Southeastern architecture was heavily dependent on saplings and young trees. This trend emerges not only from the material record of burned houses and written accounts, but from the observation that for all the houses in the database, only three structures have exterior wall posts that exceeded 30 cm in diameter. While a few large upright supports were needed for many houses, most of the wood in both flexed pole and rigid post buildings would have come from straight young saplings. Not surprisingly, studies of the physical properties of wood indicate that Southeastern Indians used some of the best available genera for house building. Oak and hickory provide a good balance of strength and flexibility, making them ideal for flexed-pole construction. Pine and cedar have a great deal of crushing strength, meaning they are better used as load-bearing members in construction (Lacquement 2004:70-76). In their experimental construction of a flexed-pole wall trench house at Etowah, Blanton and Gresham (2007:43) found that pine worked poorly for bent poles: the wood dried quickly, becoming too rigid to bend without breaking. They completely abandoned their initial pine framework and built a new one using mostly hickory, but also some oak, maple, and hornbeam. Most archaeologists estimate a use-life of 10-20 years for houses in the Eastern Woodlands (Hally 2008; Warrick 1988; Wilson 2008:79-80). It seems unlikely that exploitation of wood for house construction alone would place a burden on local ecosystems in the prehistoric Southeast, except in very large settlements. By the time houses needed to be repaired

166

or rebuilt, new saplings of appropriate size may have been in available. Firewood for fuel demands would probably have been a larger drain on local tree communities. In historic northern Iroquoian settlements in the Ontario, Jennifer Birch has identified a possible switch from smaller to larger wall poles as a result of exhausting local saplings (personal communication 2010). In that case, however, structures were much larger, and rebuilding was more frequent due to the wholesale relocation of communities. Environmental Explanations If we assume that the Woodland through Historic period environment was roughly modern, does any of the variation in environmental conditions across the study area account for architectural variability? Does variation in temperature, elevation, or latitude explain any of the observed variation in house form? Protection against rain and storms would have been a major concern for house builders. Across most of the study area, annual precipitation averages 50 to 60 inches, and the mean number of days with measurable precipitation ranges from 90 to 135. In parts of the Blue Ridge Mountains in southwestern North Carolina, the wettest part of the study area, average annual precipitation can reach 100-120 inches. In this part of the study area especially, mildew, rot, and storm damage caused by heavy rains were major limiting factors on the lifespan of native structures. Ethnohistoric records and archaeological evidence suggests that cane matting and grass thatch were commonly used to cover walls and roofs (Adair 1968; Swanton 1946; Polhemus 1998; Speck 1909). Sixteenth- and eighteenth-century accounts of native houses from different environmental settings in the Southeast suggest that they were well insulated against the cold. In his 1700 account of Siouan houses in the Carolinas, Lawson claims, ―these dwellings are as hot at stoves,

167

where the Indians sleep and sweat all night” (Swanton 1946:410-411). These were flexed pole structures covered with bark, analogs for Early Mississippian wall trench structures. Adair (1968:420), describing what seems to be a semi-subterranean winter house in northeast Mississippi between 1735 and 1768 notes, “the clothing of the Indians being very light, they provide themselves for the winter with hot-houses, whose properties are to retain, and reflect the heat, after the manner of the Dutch stoves.” However, despite the fairly strong cross-cultural correlation between climate and pit structure use, variation in climate does not seem to be a strong predictor of basin construction in the Southeast (Table 5.4). In line with Gilman‘s findings, nearly all of the sites in the study area with house basins are above 32 degrees north latitude. However, there is not a clear increase in the prevalence of house basins with increasing latitude. Figure 5.3 shows that across all time periods, houses with basins appear as far south as Kolomoki in extreme southwestern Georgia. The median latitude of sites with structures with basins is 35.269 decimal degrees, and the median latitude of sites with structures without basins is actually slightly farther north, at 35.418 decimal degrees. A Wilcoxon rank-sum test indicates this difference is not statistically significant (Z = 0.253, p = 0.800). Most semi-subterranean Late Mississippian domestic structures are found in the Southern Appalachian region, but examples of these structures have been found at sites as far south as Irene near Savannah, Georgia (Caldwell and McCann 1941). Domestic structures with basins that also appear to have been covered in earth have been recorded Bullard Landing in Twiggs County, Georgia (Williams and Evans 1993). Several measures of annual temperature do not seem to explain the occurrence of semisubterranean houses. The average and median number of frost-free days, average January minimum temperature, and average annual temperature is nearly identical for sites with and

168

without structures with basins (Table 5.4). In contrast to Gilman‘s (1987) worldwide study, structures with basins occur over a broad range of elevations; 31 percent of the components in the top quartile for elevation (263-787 m above mean sea level) have structures with basins, but 45 percent of the components in the lowest quartile for elevation (34 – 128 m above mean sea level) also have semi-subterranean houses. The median elevation of sites with structures with basins is 36 meters higher than the median elevation of sites with structures without basins (Table 5.4). However, a Wilcoxon rank-sum test indicates this difference is not statistically significant (Z = 0.417, p = 0.677). Another facet of architectural variation that might be explained by climatic variation is architectural investment. In general, houses with more posts per square meter may have had thicker exterior walls, providing a sturdier framework for hanging insulating wall and roof coverings and supporting snowfalls. This assumption is problematic, and it would be preferable to measure the thickness of wall and roof coverings. However, using posts per square meter as a proxy for insulation provides another way to test the impact of elevation, latitude, and temperature on house construction. For this analysis, only completely excavated structures with a single building episode are used. This reduces the sample size down to 326 domestic and non-domestic structures from all time periods. This sample is basically representative of the database as a whole, as it contains mostly domestic structures from all time periods, but it is biased toward well-preserved structures. The average post density (measured in posts per square meter) was calculated for each house, with values ranging from 0.16 to 14.84. I tested several models to determine the effect of environmental variables on post density, and then used the Akaike Information Criterion to determine which model fit the data best. I

169

began with a model that used all the environmental variables listed in Table 5.1 and the dummy variables for time period, and then performed stepwise regression to find the best-fit model. The best-fit model, whose partial slope coefficients, t-statistics, and p-values are shown in Table 5.5, suggests that roughly 33 percent of the variation average post density can be explained as the result of variation in average annual days of precipitation, average annual temperature, and average number of frost free days, taking into consideration the impact of time period. The Rsquared value is only moderately predictive by most standards, but the model is significant (F = 19.72, Prob>F = 0, R2 = 33.2, 8 df, n = 326). The model suggests that when it is warmer and wetter, post density declines. This provides some support for the hypothesis that houses in colder, drier climes may have been more heavily built, although many analysts would consider this R-square value too low to be explanatory. Discussion Environmental factors clearly played a part in shaping the houses of the prehistoric Southeast, but not in a simple, determinate way. Ethnohistoric and archaeological evidence indicate that native house builders used a relatively narrow set of the many available wood and plant genera for house construction. The most heavily used genera—oak, hickory, and pine—are some of the most common in the region, but they were not always the most abundant. People carefully selected wood for particular physical properties (e.g. rigid wood for upright posts and flexible wood for bendable poles), and often used different genera for different parts of houses. There was a heavy emphasis on saplings and young trees for house building, but given the estimated use life of 10-20 years for houses, it seems that local tree populations would not have been majorly taxed by the material demands for houses.

170

In a broad sense, the natural resources available for building may have been one of the stronger determinants of house form in the Southeast. Unlike the Southwest, there was never a transition to mud brick building techniques, and the limitations on the size and use-life of structures in the Southeast would have been relatively consistent from the Middle Woodland to the Historic Indian period. This helps explain some of the continuity in building traditions, but also suggest that social causes are ultimately more important than environmental causes as we attempt to understand variation in structure form. Currently available paleoclimate data do not suggest that climate change easily explains major diachronic shifts in building styles, such as the first appearance of wall trench houses or the shift from wall trench houses to rigid post houses. If we assume that modern climate data is a suitable analog for the Woodland through Historic periods, it appears that variation in environmental factors in the Southern Appalachians can only explain spatial variation in house form in a limited way. As the Spanish chronicles of the sixteenth century note, there is a definite clustering of houses with depressed floors and basins in the Southern Appalachian region north of the Fall Line, but this house form also occurs in warmer settings and sites at lower elevations. A limitation of this analysis is the geographic scale of the sample. There may not be a wide enough range of environmental diversity to fully understand the impact of environmental factors on house form when most of the structures come from the Southern Appalachian region. For future research it would be advantageous to perform these same analyses with a larger sample of houses from the coast and coastal plain. In a subset of the database, the combined effects of average precipitation, average temperatures, and frost free days may explain some of the variation in observed in post density; a logical next step would be to test this same hypothesis at a broader scale. More generally, comparing structures from Florida and the coastal Carolinas

171

with structures from the Southern Appalachians may reveal stronger empirical relationships between environmental factors and structure form.

172

Table 5.1 Environmental Data for Sites

Site Name 1GR1X1 1PI61 9GE1754 9GE1760 9GE1776 9GE1781 9GE222 9GE333 9GE901 9GE903 9WH120 Alarka Banks III Banks V BBB Motor Site Bessemer Brasstown Valley Cahokia Chota-Tanase Coweeta Creek Dallas Dog River Duncan Tract Dyar Ela Fernvale Garden Creek Go Kart Site Hickory Log Hiwassee Island Hiwassee Old Town Hoecake Jenrette Jewell Kimberly-Clark Kincaid King Kolomoki Little Egypt Lower Saratown Loy

Physiographic Province Coastal Plain Coastal Plain Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Blue Ridge Blue Ridge Interior Low Plateau Interior Low Plateau Central Lowlands Ridge and Valley Blue Ridge Central Lowlands Ridge and Valley Blue Ridge Ridge and Valley Piedmont Interior Low Plateau Piedmont Blue Ridge Interior Low Plateau Blue Ridge Piedmont Piedmont Ridge and Valley Blue Ridge Mississippi Alluvial Plain Piedmont Interior Low Plateau Ridge and Valley Coastal Plain Ridge and Valley Coastal Plain Ridge and Valley Piedmont Ridge and Valley

Frost Free Days 210 210 210 210 210 210 210 210 210 210 210 150 210 210 210 210 150 210 210 150 210 210 210 210 150 210 150 210 210 210 210

Jan. Min. Temp (F) 28 28 28 28 28 28 28 28 28 28 28 19 28 28 19 28 19 19 28 19 28 28 19 28 19 19 19 28 28 28 28

Avg. Annual Temp. (F) 63 63 63 63 63 63 63 63 63 63 57 54 57 57 57 63 54 57 57 54 57 63 57 63 54 57 54 63 57 57 57

Latitude 32.8685 33.0439 33.4406 33.4333 33.443 33.4425 33.4197 33.4524 33.439 33.439 34.6098 35.3695 35.4699 35.4732 38.6756 33.401 34.9443 38.6581 35.5544 35.0903 35.1882 33.6224 36.3485 33.6022 35.4523 35.9923 35.5399 32.4414 34.2554 35.4175 35.2433

Elev. (m AMSL) 33 34 157 151 139 149 142 139 137 143 421 530 249 250 127 127 556 128 275 617 208 230 140 132 548 175 787 73 274 208 219

210 210 210 210 210 210 255 210 210 210

19 28 19 19 19 28 36 28 28 28

57 57 57 57 57 63 63 57 57 57

36.7563 36.0715 36.8551 35.7674 37.0812 34.1982 31.4676 34.5978 36.4879 36.1071

92 152 196 228 102 172 88 212 154 268

173

House Basins N Y N N N N N N N N Y Y N N Y Y N Y N N N Y N Y N N N N N N N Y N N Y N Y Y Y N N

Site Name Martin Farm McFarland Mialoquo Mitchum Morris Moundville Napoleon Hollow Pott's Tract Rivermoore Rucker's Bottom Rymer Snodgrass Summerour Toqua Town Creek Townsend Tuckasegee Tukabatchee Two Run Creek Warren Wilson Yearwood

Physiographic Province Ridge and Valley Interior Low Plateau Ridge and Valley Piedmont Interior Low Plateau Coastal Plain Central Lowlands Ridge and Valley Piedmont Piedmont Ridge and Valley Mississippi Alluvial Plain Piedmont Ridge and Valley Piedmont Blue Ridge Blue Ridge Coastal Plain Ridge and Valley Blue Ridge Interior Low Plateau

Frost Free Days 210 210 210 210 210 210 210 210 210 210 210

Jan. Min. Temp (F) 28 28 28 28 19 28 28 28 28 28 28

Avg. Annual Temp. (F) 57 57 57 57 57 63 54 57 57 63 57

Latitude 35.5734 35.4787 35.6081 35.8318 37.2879 33.0052 39.6873 34.5974 35.5118 35.5118 34.1844

Elev. (m AMSL) 247 297 264 115 149 47 132 212 610 610 128

210 210 210 210 210 150 210 210 150 210

19 28 28 28 28 19 36 28 19 28

57 57 57 63 57 54 63 57 54 57

35.3162 36.5546 35.558 35.1826 35.6781 35.2694 32.4807 34.243 35.6084 35.1282

209 89 248 69 309 654 63 212 646 263

174

House Basins N N N N N Y N Y N Y N Y Y Y N N Y N N Y N

Table 5.2 Historic Accounts of Building Materials

Group

Natchez

Chickasaw

Various

Carolina Siouan

Yuchi

Date(s)

1725

1735-1768

1735-1768

1700

1909

Account

Construction

Framework (posts and poles)

Le Page du Pratz

Flexed pole

Hickory (young, 4" diameter, 18 20' long)

Adair

Rigid post; summer house

Pitch-pine, dried locust and sassafras

Rigid post, winter house

Flexed pole

Forked wall posts (species unidentfied), white oak rafters, four pine roof supports Bendable poles of pine, cedar, or hickory, stripped of bark

Flexed pole

Bendable poles, species not identified

Adair

Lawson

Speck

175

Binding material for framework

Cane splints, cord Splinters of white oak and hickory; cane; split saplings (species unidentified)

Split saplings Bark, especially elm, tree moss (about a yard long)

Wall covering Leaves and stalks of corn; mud plaster with some moss; bundles of grass fastened with cane splints

Roof

Other

Bundles of grass fastened with cane splints; cane mats

Clay and cut straw (for wall plaster); cane mats

Pine or cypress clap-boards; mud daub in some cases

Poplar eave boards; pine or cypress bark Long poles (species unidentified) woven with split saplings. This is covered with daub and grass thach Bark: cypress, red or white cedar; pine (not preferred) Matting weighed down with halved logs

Daub mixed with withered grass; covered with grass thatch Bark: cypress, red or white cedar; pine (not preferred)

Cypress bark; matting

Poplar doors, planks bound with straps of buffalo hide

Table 5.3 Archaeological Evidence for Building Materials

Site

Period

Reference

Construction

Flexed and rigid

Framework (posts and poles)

Hiwassee Island

Late Mississippian

Lewis and Kneberg 1946

American Bottom, multiple sites

Late Woodland, Mississippian

Simon 2002

Oak White Oak, Hickory (possibly ash, cedar, willow, red oak)

Rench Site

Late Woodland, Mississippian

McCounaughy 2007

"Wigwam;" flexed poles for long walls, straight poles for end walls

Curved hickory and oak poles in long walls; straight hickory and walnut poles; hickory cross beams

Loy

Late Mississippian

Polhemus 1998

Rigid post

176

Binding material for framework

Wall covering

Roof

River cane, grass, thatch

Bulrush thatching River cane, grass, thatch

Strips of cane; twisted cord from grass, bark fibers, wild hemp

Cordage woven from basswood fibers

Other Cane for wattle and lathing; puddled clay and dried grass in hearths

Table 5.4 Summary of Environmental Variables for Sites with and without Structures with Basins Sites with structures without basins (n = 41) Environmental variable Mean Frost Free Days 204.15 Jan. Min. Temp (F) 26.22 Avg. Annual Temp. (F) 58.39 Latitude (decimal degrees) 35.055 Elev. (m AMSL) 232

Median 210 28 57 35.4175 175

Sites with structures with basins (n=21) Environmental variable Frost Free Days Jan. Min. Temp (F) Avg. Annual Temp. (F) Latitude (decimal degrees) Elev. (m AMSL)

Median 210 28 57 35.2694 209

Mean 203.57 24.95 58.86 35.009 249

Table 5.5 Results of Multivariate Regression for Post Density and Environmental Variables Post density = ƒ (avg. days precip., avg. annual temp., avg. frost free days, dummy variables for time periods) Structures (n) 326 F( 8, 317) 19.72 Prob > F 0 R-squared 0.3323 Post density Coef. t P>|t| MW -1.18108 -3.78 0 LW 2.349376 6.35 0 EM 0.328953 1.27 0.207 MM 0.489403 1.3 0.196 HI -0.14422 -0.48 0.629 Avg. days precip. -0.0376 -2.2 0.029 Avg. annual temp. -0.18012 -3.1 0.002 Avg. frost free days -0.02101 -2.94 0.004 Constant 20.90312 3.7 0 AIC = 1207.529, lowest of all environmental models tested

177

Figure 5.1 Major Physiographic Regions of Eastern North America (Pauketat 2004:19)

178

Figure 5.2 Braun‘s Forest Regions for Eastern North America (Dyer 2006:342)

Figure 5.3 Location of Sites with Structure Basins

179

CHAPTER 6 HOUSEHOLD ECONOMICS AND HOUSEHOLD COMPOSTION

Ethnographic and archaeological research, discussed in Chapter 2, have shown that variation in a particular set of observable features of houses and communities may be caused by variation in economic behavior and household composition. These features include: structure size and the arrangement of structures in a settlement, the nature of storage, interior partitioning in domestic structures, and architectural investment in domestic structures. In this chapter I attempt to determine which of these lines of architectural variation can be explained by changes in household-level economic activities and household composition in the Southern Appalachian region of the Southeast. I begin this chapter with a brief overview of the Southeastern subsistence economy. This provides important context for the discussions that follow. First, I hypothesized that changes in the size and spacing of domestic structures would track with major changes in the subsistence economy. Peregrine (1992) and Pauketat (1998:135136), using data from the American Bottom, have argued for a diachronic shift from large courtyard groups in the Late Woodland period to smaller, perhaps nuclear family households in the Early Mississippian period as the basal unit of production and consumption. They propose this change was driven largely by the adoption of intensive maize agriculture. I use the architectural database to test this model at a broader scale.

180

Similarly, I hypothesized that the increasing partitioning of domestic structures would track with the widespread adoption of intensive maize agriculture and the establishment of larger, more permanent settlements around A.D. 900. As settlements grew larger and more permanent, I hypothesized that houses would grow larger and be would divided into smaller areas to create more space (and especially more private space) for storage, production, and consumption. Finally, I hypothesized that changes in architectural investment described in Chapter 4 can be explained by broad-scale changes in economic behavior. I especially focus on changes in architecture during the Late Mississippian to Historic Indian period, when architectural investment in domestic structures appears to decline, reversing a centuries-long trend in increasing architectural investment. The Southeastern Subsistence Economy The subsistence economy in the Southeast from the Middle Woodland to the Historic Indian period is fairly well understood. By the Middle Woodland period native groups practiced a mixed economic strategy of hunting, gathering, and fishing along with the cultivation of small, starchy seed crops (including chenopodium, goosefoot, marshelder, and sumpweed) known as the Eastern Agricultural Complex (Smith 1992). The Eastern Agricultural Complex plants were domesticated and cultivated by the Early Woodland period, and their use may have been intensified during the Middle Woodland period. Maize appears in very small quantities as early as ca. A.D. 175 at Icehouse Bottom, and ca. A.D. 220 at the Edwin Harness site, but is not used widely or intensively until the Late Woodland period (Smith 1992). By approximately A.D. 800-900 intensive maize agriculture becomes widespread across the Southeast, and by the Early Mississippian period (ca. A.D. 1000) there is clear

181

archaeological, ethnobotanical, and skeletal evidence that maize is the major staple crop in the region (Muller 1997). There are several explanations for the shift to maize agriculture, including the Boserupian notion that the increasing productivity of the Eastern Agricultural Complex and improved deer hunting with the bow and arrow raised population levels and forced people to move to a more productive agricultural system (Anderson and Mainfort 2002:17-18, Emerson et al. 2000). Agent-based models suggest that elite actors saw the highly productive potential of maize as a way to build up surplus wealth to control social reproduction (Pauketat 1994). While the shift to maize agriculture was a transformative event, hunting and gathering were never abandoned (Hudson 1976). Faunal and botanical assemblages from the Mississippian period indicate that wild foods continued to be important, and historic accounts of Indian societies provide abundant evidence of a mixed economy of hunting, gathering, fishing, and farming (Scarry and Scarry 2005). Using archaeological data and ethnohistory, Scarry and Scarry (2005) provide a reconstruction of Southeastern agriculture from roughly the eleventh century to the eighteenth century. Maize was the principal crop, and squash, beans, and sunflower were grown along with it. Men provided assistance in clearing fields, planting, and harvesting, but women performed most of the agricultural labor, and were responsible for preparing fields, planting, weeding, and harvesting. Maize was planted in small hills spaced roughly one meter apart, and beans and squash were planted in between. For the historic period, there is good evidence that agricultural production was carried out by both households and larger groups. Scarry and Scarry (2005:262) argue that the late prehistoric and historic Apalachee, Creek, Chickasaw, Powhatan, and Tunica practiced a dual strategy of farming. ―In-field gardens‖ were planted close to houses and maintained by

182

households, and ―out-field bottomland fields‖ were farmed communally by larger, suprahousehold groups. Small nuclear or extended families tended household gardens and stored food in household granaries while also contributing labor to communal fields, whose produce was stored in community granaries. Hudson‘s (1976:259) reconstruction of the native subsistence economy stresses the gendered and seasonal nature of the system: ―the men made their primary contribution to subsistence in the cold season, when they hunted, while the women made theirs in the warm season, when they cultivated crops.‖ This gendered division of labor also played out spatially within settlements, as women tended to work around their household with other women, while men ―resorted to their town house or square ground‖ (Hudson 1976:260). Especially relevant to this study is Hudson‘s (1976) discussion of gender as it related to house construction and ownership. He claims, ―men constructed all the buildings, both domestic and public, and cleared the land that was used for building and cultivation‖ (Hudson 1976:267). In Creek society a young man had to build a house and raise a crop with the help of his lineage before he could get married (Hudson 1967:198). Given that Creek society was matrilineal and matrilocal, this meant that house construction would have involved supra-household male labor, since older married men from the boy‘s lineage would have been spread out, living with their wives‘ households. After the pair was married, the house became he property of the young woman‘s lineage (Hudson 1976:198). Hudson implies that this pattern of house building and property ownership was relatively common, and that in addition to owning houses, women possibly owned agricultural fields (1976:267). While Hudson (1976) and Scarry and Scarry (2005) present good evidence for continuity in the agricultural system from Mississippian to the historic period, we also have to consider the

183

impact of contact with Europeans on the native economic system. Beyond the changes wrought by disease, slaving, and warfare (see Marcoux 2008), native involvement in the deerskin trade may have had major impacts on the subsistence economy that led directly to changes in domestic architecture. Waselkov (1994:194) argues that the abandonment of the semi-subterranean winter house by Creek communities was the result of heavy investment in deer hunting, which led to longer winter hunting trips that made solidly built winter domestic structures obsolete. This understanding of the Southeastern subsistence economy and the general relationships between houses and economic practices provides a framework for understanding how much of the variation in Southeastern architecture is driven by change domestic subsistence production and consumption. Changes in Domestic Structure Size and Arrangement Based on Peregrine (1992) and Pauketat‘s (1998:135-136) research on domestic structures and economic and social organization in the American Bottom, and Flannery‘s (2002) model for changes in domestic architecture with increasing sedentism in the Near East and Mesoamerica, I hypothesized that the changes in the size and spacing of domestic structures would track with major changes in the subsistence economy. I anticipated that with the widespread adoption of intensive maize agriculture after A.D. 900 there would be a shift away from closely-spaced domestic structures with communal storage, to more distantly-spaced domestic structures with private storage. This archaeologically observable pattern would represent a transition from larger, corporate social groups that cooperated a great deal for daily subsistence tasks to smaller, more autonomous households that invested more of their labor into household-level production. I then expected to see another transition in the size and spacing of domestic structures during the eighteenth century, when many native communities‘ subsistence

184

economies changed due to the onslaught of European diseases and warfare, and new markets like the deerskin trade. I tested these hypotheses by comparing the size and arrangement of domestic structure before and after the adoption of maize agriculture in the Southeast and before and after the contact period. Most of the Middle Woodland structures in the study area are circular or oval shaped, single post structures with few interior features and open floor plans. Middle Woodland domestic structures are the largest of any time period. Large round domestic structures are found at Brasstown Valley, Hickory Log, McFarland, Two Run Creek, Ela, 9GE333, and Duncan Tract. After accounting for superimposed buildings, the maximum number of contemporary occupied structures ranges from 3 to 10. At the sites with multiple domestic structures, there is variation in the layout of community plans. At Hickory Log, 12 of the 13 possible Cartersville phase structures are arranged in a roughly 75 m long arc in the southern locus of the site. One additional structure is located 50 m north of the larger group of houses. Two groups of houses overlap, suggesting they were not contemporary. An additional pair of houses overlaps slightly, and may or may not have been contemporary. At most, there may have been 10 contemporary structures. Some of the houses are spaced closely, with only 10 m from center to center, and the farthest two houses would have been spaced roughly 110 m apart. If these structures were contemporary, they may be grouped into three clusters of one to four houses each, but given the lack of a clear courtyard area or plaza, this interpretation is only tentative. In contrast, at Brasstown Valley, 7 possible Cartersville-phase houses in locus F are arranged in a rough horseshoe pattern measuring 50 by 60 m. They may form an arc around a clear area, although given the imprecise dating of the houses this is unclear. Four of the houses

185

are spaced between 20 and 40 m apart, suggesting that they housed separate, individual households, but three of the structures in the northeast corner may have formed a tight cluster, if they were contemporary. At Duncan Tract six circular structures are arranged in a linear pattern 130 m long by 40 m wide. Four of the structures are arranged in a tight linear pattern with some overlap in the eastern half of the habitation area, while two structures roughly 40 m to the west are spaced 15 m apart. The degree of overlap in the eastern structures suggests that no more than three of the four houses could have been contemporary. Not all of the Middle Woodland structures have round, open floor plans. These include a single, small, keyhole shaped structure at Kolomoki, the large, oval shaped Owl Hollow phase structures in Middle Tennessee and a similar structure at 9GRX1 in Alabama, a square, Connestee phase building at Garden Creek, and the large, somewhat irregularly shaped buildings at the Yearwood site. However, most Middle Woodland communities seem to be represented by a few large, closely spaced domestic structures. It seems likely that each large domestic structure represents the dwelling of a large household, and given the close spacing of the houses, it seems likely that these households could have frequently cooperated in daily subsistence activities. There is limited evidence of interior storage in domestic structures. Instead, much of the storage may have been in shared facilities. During the Late Woodland period there are major transformations in house construction and new patterns of regional variation in house form. In contrast to the Middle Woodland period, it is possible to identify multi-structure household clusters, signaling a change in the organization of domestic production and consumption.

186

At sites from the western part of the study area, specifically in the American Bottom, southeastern Missouri, and west-central Alabama, there is a shift toward small, uniform, rectangular domestic structures, often arranged in groups around a small courtyard. Wall posts are small and closely spaced, reflecting a bent pole construction style. In many cases these houses are set in shallow basins. In terms of shape and construction technique, these houses and community plans signal a clear break from the built environment of the Middle Woodland period, and also a shift in the nature of the primary unit of domestic production and consumption. During the Edelhardt phase at Cahokia 15A, Pauketat identifies four possible courtyard groups. These are clusters of several small rectangular structures flanking obvious courtyards, two of which contain four central pits (1998:85, Figure 6.18). There is little evidence of interior features or partitioning in the houses, and storage seems to be predominantly in pits located outside of buildings. Pauketat (1998:135) argues that ―given this storage location and the courtyard arrangements, the courtyard group is inferred to have been the fundamental unit of Edelhardt phase social organization.‖ By extension, it may also have been the primary unit of domestic production and consumption. At the BBB Motor site, Emerson and Jackson (1984) argue that 16 small, rectangular Edelhardt-phase houses are arranged into 3 clusters of 4 to 6 structures each. The houses are closely spaced, approximately 5 m apart from center to center. In contrast to Cahokia, the houses are arranged in a linear pattern, rather than in courtyard groups. The majority of the storage appears to be outside of houses, but there is at least one small feature, possibly a storage pit, in most houses. Two excavation areas at the Hoecake site revealed two clusters of 3 to 4 small rectangular domestic structures that appear to be centered around small open areas. Within the

187

clusters houses are spaced only about 5 m apart from center to center. The possible ―courtyards‖ in this case are much smaller than the ones identified at Cahokia Tract 15A, but are more obvious than possible shared open spaces at BBB Motor. There are few interior features in the Hoecake structures, suggesting storage was primarily external. Four domestic structures were identified at 1PI61. While they are similar in size and form to the houses at BBB Motor, Cahokia Tract 15A, and Hoecake, they are more widely spaced. The four structures lie in dense palimpsest of posts and features surrounding a relatively clear central courtyard or plaza. There are some interior features, but most of the storage appears to have been in exterior pits. In the Southern Appalachians, the form and arrangement of Late Woodland domestic structures is more variable. At Brasstown Valley, possible Late Woodland domestic structures are round, with few to no interior features, and no evidence of interior partitioning. Storage appears to have been extramural. The three houses at Rivermoore are small and square. The houses each have at least one hearth and one to three interior pits, but most of the storage appears to have been extramural, in several large exterior pits. These structures may be arranged around a small plaza or courtyard. By the Early Mississippian period (ca. A.D. 1000), a subsistence economy based on intensive maize agriculture was established across the Southeast. In a broad sense, settlements shared a common architectural pattern. According to Lewis and Stout (1998:5), ―the main architectural elements include plazas, platform mounds and other earthworks, various means of segregating space and activities, defensive works, and natural terrain features.‖ Across most of the study area, the rectangular wall trench house form replaced single post buildings. In general, the large courtyard groups of the Late Woodland period seem to be gradually replaced by smaller

188

social units. Smaller household clusters and perhaps even individual nuclear family households may have become the primary units of domestic production and consumption. Domestic structure size increases noticeably from the Late Woodland to the Early Mississippian period (see Tables 4.3 and 4.4).While the increase in domestic structure size is a strong trend across the Southeast, there is variability in the degree of the size increase. Figure 4.6 shows box plots for domestic structure size at each Early Mississippian component. Note that at four of the components, Lohmann phase Cahokia, Fernvale, Stirling phase BBB Motor, and the general Early Mississippian houses at Moundville, median domestic structure size is still under 25 m2. Meanwhile, at Jewell, Hiwassee Island, Toqua, Morris, and Chota-Tanase, the median is close to or above 40 m2, representing a four-fold increase in size compared to most Late Woodland houses. The difference does not appear to be strictly regional. Many Early Mississippian domestic structures from the American Bottom fall on the small end of the size continuum while houses from the Southern Appalachians tend to fall on the large end, but overall, components with smaller and larger domestic structures are spread out across the study area. Even as the general pattern of mound and plaza community plans applies to many of the Early Mississippian sites, trying to identify household clusters reveals variation in the arrangement of domestic structures. Household clusters in the Early Mississippian period range from single, isolated structures on small farmsteads, such as Fernvale and Banks V, to large, multi-structure groups at large sites like Moundville. Single Early Mississippian domestic structures were identified Banks V, Chota-Tanase, Fernvale, Town Creek, and Martin Farm. These houses range in size from only 14.4 m2 at Fernvale to 46.7 m2 at Chota-Tanase. It is hard to imagine a group much larger than a nuclear

189

family living in the small house at Fernvale, while a larger, possibly extended family may have inhabited the house at Chota-Tanase. At BBB Motor, Bessemer, Cahokia Tract 15A, Jewell, Kimberly-Clark, Morris, Moundville, and Toqua, relatively broad horizontal excavations provide a better opportunity to understand overall community layout and identify possible households. During the Stirling phase at BBB Motor, there are two possible clusters of structures. The two groups are spaced approximately 80 m apart, and within the clusters the houses are spaced roughly 20 m apart from center to center. Compared to the Edelhardt phase at BBB Motor, houses are larger and more widely spaced. There is a small, clear area in both clusters that may have served as a small work area or courtyard. Much of the storage appears to have been in exterior pits, but there are more interior features in the Stirling phase houses than in the Edelhardt phase houses. Welch (1994) argues that most of the structures in the village at Bessemer are nondomestic, while the mound summit buildings were elite residences. Whether or not this interpretation is accurate, it is clear that there were several large structures in the village area at Bessemer, and that there was some rebuilding and shifting of structuress over time. There is certainly no evidence of small, Late Woodland style courtyard groups. Rather, the large buildings seem to be oriented toward the central plaza and the large ―domiciliary mound‖ (Welch 1994). Pauketat summarizes the Late Woodland to Early Mississippian transition at Cahokia Tract 15A, suggesting that, ―Emergent Mississippian courtyard groups by and large disappeared, to be replaced by a series of rectangular and circular buildings built using a new construction technique and arranged around a possible plaza‖ (Pauketat 1998:135). By the end of the Lohmann phase, there is evidence for more intramural storage. During the Stirling phase Tract

190

15A was converted into a ceremonial, rather than a residential space, with large posts and a few large buildings replacing domestic structures. By Moorehead times, domestic structures are larger and spaced farther apart than in the Lohmann phase, and Pauketat suggests that there may have been more interior storage. Pauketat (1998:135) argues that ―by the Moorehead phase, households had probably replaced larger kin units (such as the Emergent Mississippian courtyard groups) as economic and social units.‖ At the Jewell site in Kentucky, there are no clear multistructure household clusters. A mound summit structure shows rebuilding episodes, and several overlapping structures were found in two large excavation areas. Two structures in the village area appear to have been rebuilt in place two or three times. If these houses represent individual households, they would appear to have been widely spaced on the site, and they may have been occupied for more than a generation. Schroedl (1998:69) classifies the Kimberly-Clark site as a small ―hamlet‖ that represents the occupation of no more than one or two households. There are two superimposed domestic structures, one possible structure represented by a shallow basin, and two pits. It appears that the larger domestic structure was built over the earlier, smaller house, suggesting that household membership expanded between the construction episodes (Chapman 1994). Given the isolated nature of the settlement, it seems that the household or households at Kimberly-Clark would have been fairly autonomous in terms of daily subsistence activities. Rolingson and Schwartz (1966) identified 12 Early Mississippian structures at the Morris site in Kentucky. There is a plaza at Morris, and all the domestic structures except one are arranged around this clear area. It is hard to tell if there are distinct household clusters at Morris. If the structures were all contemporary, there may be two clusters of four structures each in the

191

northern half of the site, and a possible, widely spaced cluster of four houses in the northern half of the site. It is also possible, especially in the southern half of the site, that individual structures represent individual households. At least three of the wall trench structures were rebuilt or repaired in place. There are some interior features, but much of the storage may have taken place in small pits near the houses. At Moundville in the Roadway excavation area, and the PA and ECB tracts, Wilson (2008:87-92) documents 12 residential groups represented by clusters of 10-20 structures during the Early Moundville I and Late Moundville I-Early Moundville II phases. During the Early Moundville I phase, it appears that a few households built single post domestic structures in each residential area. In the subsequent Late Moundville I-Early Moundville II phase, there was a population increase and a shift to wall trench architecture. More domestic structures were built in each residential area and there was more in-place rebuilding. Due to the overlapping building episodes, it is difficult to identify clear household clusters in the Late Moundville I-Early Moundville II phase. Wilson argues that the houses in each residential area represent multihousehold groups that would have acted as a coherent, basal unit of production and consumption. However, there may also be several smaller households within each residential area. After A.D. 1200, as intensive maize agriculture became increasingly common in the Southeast, there is evidence for both change and continuity in the size and arrangement of domestic structures. On the surface, there is a secular trend toward larger, more distantly spaced houses, suggesting that the shift toward nuclear family households as the basal economic unit is a widespread pattern. The median size of domestic structures rises from 21 m2 (n=268) in the Early Mississippian to 23 m2 (n=115) for the Middle Mississippian period, then rises sharply to 41 m2

192

(n=203) during the Late Mississippian period. The low figure for Middle Mississippian domestic structures is driven by the 93 houses at Snodgrass, which have an average area of 22.1 m2. Middle Mississippian domestic structures at other sites are much larger. Structures at Hickory Log, Kincaid, Rucker‘s Bottom, and Town Creek have an average size of 45.0 m2 (n=24). Thus, during the Middle and Late Mississippian period, most domestic structures are nearly twice as large as analogous structures during Early Mississippian times. Examining community layouts after A.D. 1200 reveals variation in the arrangement of possible household clusters. While some occupations appear to be made up of individual, distantly spaced houses, at some Late Mississippian sites there are clusters of closely spaced houses that share small, open work areas. This suggests that a range of household types, from small, single families to larger, extended family groups shared in daily production and consumption tasks in the Middle and Late Mississippian period. By the Late Mississippian period, it is possible to identify storage buildings in household clusters. These small rectangular or circular structures, referred to as barbacoas in Spanish accounts, were probably elevated corn cribs. These structures may appear as early as the Middle Mississippian period at Rucker‘s Bottom, and they are common at Late Mississippian sites in western North Carolina, eastern Tennessee, and northern Georgia (see Hally 2002). They continued to be used into the historic period, when at some settlements storage also occurs in underground cellars. In contrast to earlier times, the ability to recognize individual household storage buildings makes it somewhat easier to identify coherent household groups, and provides clear evidence for common storage at the level of a household cluster. Hally (2008:272) used the distribution of primary domestic structures, rectangular storage buildings, open work areas, and human burials to identify at least six possible households at

193

King. Most of the households contain more than one domestic structure, and therefore may have been multi-family households (Hally 2008:272-273). Five of the six household clusters contain at least two primary domestic structures, one or more rectangular storage structures, and an open work area (Hally 2008:279-290). Household 8, composed of a single primary domestic structure and two rectangular buildings, may represent the dwelling of a single conjugal family, but it may also be part of a larger household that is not as well defined due to preservation issues (Hally 2008:282). At its peak population, the King site appears to have been composed of several multifamily households arranged around a central plaza and townhouse. These households would have been fairly autonomous in terms of day-to-day subsistence, and had granaries for householdlevel food stores, but could have worked as a larger productive unit for community-level projects. At Hickory Log, three Wilbanks phase domestic structures measure approximately 30 to 40 m2 and are spaced 40 m apart. The houses are located in an area with heavy erosion, and little remains aside from the exterior wall posts and central support posts. However, given the distance between the houses, these structures probably represent three distinct households. At Little Egypt, three Barnett phase domestic structures also appear to represent the winter houses of three distinct households. Structure 1, the largest of the three houses, is located on Mound A, and is thought to be an elite residence (Gougeon 2002, 2006:183-184). Structures 2 and 3, located roughly 60 to 90 m from Mound A and 30 m from one another, are interpreted as the winter houses of commoner households. Gougeon (2002; 2006:185-187) finds that the same types of subsistence activities took place in all three houses. The architecture of the upland Lamar farmsteads of the Oconee River Valley of the Georgia piedmont points to a different pattern of household organization from the Late

194

Mississippian settlements closer to core of the Southern Appalachians. At Sweetgum, Sugar Creek, Carroll (Hatch 1995) and several sites excavated in advance of development at Reynolds Plantation on Lake Oconee (Ledbetter n.d.), there is a consistent architectural pattern of small settlements with one large circular domestic structure, several smaller rectangular structures, large pits (presumably excavated for daub then used as refuse pits), and smaller non-architectural features (Hatch 1995:144-148; Ledbetter n.d.). The circular structures contain interior hearths and evidence for interior storage (Hatch 1995:146). The rectangular structures, like those found at King and Coweeta Creek, were more lightly built and may have been used for storage and for domestic activities in warm weather. If the circular houses represent the primary dwellings of single families, as Hally (2008:272) posits for King, then many of these sites seem to represent dispersed, single-family households. In terms of daily economic production and consumption, these groups would have been more independent than households in larger town such as King and Toqua. In contrast to the settlements in the Southern Appalachians, the 93 houses at the Snodgrass site are arranged in a tight grid surrounded by an outer ditch. Price and Griffin (1979) originally argued that these houses were organized into three segments: a central segment surrounded by an interior wall and centered around a small plaza, and two additional segments to the north and east of the central segment, each with their own small plazas. In all three segments, houses were very closely spaced. On average, the center-to-center distance between houses in the segments is 7 meters. The houses are also smaller than most Middle or Late Mississippian houses, with an average size of 22.1 m2. In the wake of European diseases, warfare, and a changing political and domestic economy after A.D. 1540, there are noteworthy changes in the size and spacing of domestic

195

structures. During the Historic Indian period, household clusters, especially among the Cherokee, tend to be very uniform and widely spaced, even at large settlements. There is a slight decrease in the median size of domestic structures after the sixteenth period across the study area, from 41 m2 (n=203) in the Late Mississippian period to 35 m2 (n=87) in the Historic period. Individual household storage facilities can still be identified, and in some places cellar storage augments storage in domestic structures and corn cribs. At sites occupied by historically known groups, such as the Cherokee of the Southern Appalachians, the material remains of households are fairly consistent from site to site. The winter house and summer house pattern described in several eighteenth-century accounts of Cherokee communities can be clearly identified at several sites. At Alarka Farmstead, Brasstown Valley, Chota-Tanase, Ravensford, Toqua, and Townsend, household clusters consist of a round winter house, a rectangular summer house, an open work area, and burials. In some cases, small circular or rectangular storage buildings can be identified, but these are less common than in the Late Mississippian period. At Mialoquo, a single round winter house is part of a cluster with rectangular buildings, but there is not a clear pairing. At Tuckasegee, only a winter house was uncovered while at Hiwassee Old Town, only rectangular structures were identified. In general, eighteenth-century Cherokee household clusters are more widely spaced than those of the Late Mississippian period. While some Middle and Late Mississippian domestic structures are isolated and may represent single family households (e.g. the widely spaced houses at Hickory Log and Town Creek), many people lived in what are probably multi-family households represented by two or more closely spaced domestic structures and their associated storage buildings (e.g. the households identified at Toqua and King). At the Ravensford and Townsend sites, small Cherokee settlements with no townhouse, clearly-defined household

196

clusters with a winter and summer structure are spaced over one hundred meters from one another in a linear arrangement roughly parallel to a river. Paired houses are also widely spaced at Brasstown Valley, and a single house pair at Alarka Farmstead may represent an isolated single family household. At Chota, a larger town with a townhouse, pavilion, and plaza, house pairs are more closely spaced, but are still separated by at least ten meters, and most are spaced farther apart than the household clusters in most Late Mississippian settlements. The redundancy of domestic architecture and the wide spacing of household clusters suggest that households were economically independent on a daily basis. This fits Scarry and Scarry‘s (2005) reconstruction of historic period agricultural practices, in which individual households managed their own garden and food stores while contributing labor to community fields located farther away. In general, these broad-scale diachronic patterns of change fit with the models for changing house size and arrangement proposed by Peregrine (1992) and Pauketat (1998:135136) for the American Bottom and by Flannery (2002) for the Near East and Mesomerica. This schematic diagram shows a simplified model for household clusters for each time period (Figure 6.1). This model glosses over regional variation, but it is based on dozens of site plans and captures the general trend in the Southern Appalachian region and surrounds. The visual model shows a trend toward the increasing privatization of storage within a household cluster, and increasing distance between household clusters. With the widespread adoption of intensive maize agriculture after A.D. 900 there seems to be a shift away from closely-spaced domestic structures with communal storage, to more distantly-spaced domestic structures with private storage. This may represent a transition from larger, corporate social groups that cooperated a great deal for daily subsistence tasks to smaller,

197

more autonomous households that invested more of their labor into household-level production. In the eighteenth century, domestic structures continue to be distantly spaced, with their own storage, but they also appear to be less permanent. Household Economics and Interior Partitioning I hypothesized that the increasing partitioning of domestic structures would increase after the adoption of intensive maize agriculture and the establishment of larger, more permanent settlements around A.D. 900. As settlements grew larger and more permanent, I hypothesized that houses would grow larger and be would divided into smaller areas to create more private space for storage, production, and consumption. This appears to be the case. One of the most pronounced diachronic changes in house form is the transition from open to divided floor plans. During the Middle Woodland period, domestic structures are primarily round with open floor plans (Figure 6.2). Only one of the domestic structures in the database has clear evidence of an interior wall. Numerous ethnographic studies would suggest that the occupants of these houses mentally divided the space into different activity areas, but there is little physical evidence for the division of interior space. During the Late Woodland period houses are less uniform across the study area, but there is still little evidence of interior partitioning (Figure 6.3). In northern Georgia, for example, some domestic structures are large, round, and open, much like the Middle Woodland period houses, while others are smaller and square. In both cases there is little evidence for partitioning. Across the entire study area, only 14 percent of the Late Woodland domestic structures have partitions, and the maximum number of interior walls is any house is two. In some parts of the study area, the American Bottom especially, Late Woodland houses are extremely small, only 8-10 square meters. There was hardly any space to divide in these houses.

198

During the Early Mississippian period, domestic structures are much more uniform at a broad geographic scale. Floor plans are still essentially open (Figure 6.4). Only 9 percent of domestic structures have any evidence of partitioning, and the maximum number of interior walls is two. Interior hearths become better formed and more common. Domestic structures are generally spaced farther from one another than during Late Woodland times, and there is increasing evidence for interior storage: directly in the form of small pits, and indirectly in an increase in floor area compared to the Late Woodland period. Beginning in the Middle Mississippian period, after about A.D. 1200, there is a notable increase in the frequency of interior partitions. Over 50 percent of the well-documented domestic structures have interior partitions, and the maximum number of interior partitions rises from two to four. The size of domestic structures changes relatively little from the Early to the Middle Mississippian period, so approximately the same amount of interior floor space is divided into one or more compartments (Figure 6.5). During the Late Mississippian period interior partitions become even more common in domestic structures (Figure 6.6). In several burned Late Mississippian domestic structures, collapsed daub partition walls have been identified. Forty-one percent of the Late Mississippian domestic structures have between one and seven interior partitions. Moreover, these houses have a clearly defined prepared clay hearth and four large interior support posts. These features effectively separate the center of the structure from the periphery, and contribute to the division of interior space. As a schematic diagram of a Late Mississippian house shows, many of these structures were divided into as many as six to eight smaller areas (see Figure 6.1). Late Mississippian domestic structures are also larger than Early and Middle Mississippian houses.

199

During the Historic period in the Southern Appalachians, partition walls are common in both types of domestic structures. Both the round winter houses and the rectangular summer houses of the Cherokee are divided by interior partition walls. Figure 6.7 shows examples of winter and summer houses with interior partitions. In sum, there is a secular trend of increasing interior partitioning after the Late Woodland period. The mean number of interior partitions spikes after the Early Mississippian period, and tapers off slightly in the Historic Indian period. These temporal differences are statistically significant. Multivariate regression with dummy variables for time periods shows that the difference in the average number of partitions per domestic structure is significant between all periods except the Middle and Late Mississippian periods, when the average values are quite close (Table 6.1). In general, the architectural data confirm the hypotheses about subsistence and settlement and the organization of interior space and storage developed from smaller-scale studies of houses and households. After the widespread adoption of maize agriculture in the Southeast, households construct larger houses with more interior partitions, which would have created more private space, for storage, production, and consumption. The placement of exterior storage facilities also fits with this general model. During the Middle and Late Woodland period, before the onset of intensive maize agriculture, it appears that large corporate groups may have shared single storage facilities, whereas by the Late Mississippian period, some 500 years later, individual maize storage buildings can be identified with distinct single- and multi-family households, smaller social groups than in earlier times. There are, however, some important deviations from the expectations of small-scale studies of the relationship between house form and domestic economy. Early Mississippian

200

domestic structures show signs of additional interior storage compared to earlier periods, but there are few interior partitions. Moreover, few good examples of separate maize storage buildings have been found at Early Mississippian sites, and some of these sites were just as large if not larger than Late Mississippian towns where storage buildings can be identified. We may simply have failed to identify these structures, but it seems like the pressure to privatize storage at these sites would have been just as great as in later times. The Late Mississippian domestic structures of the Oconee River valley are rarely partitioned. This deviation from the usual Late Mississippian pattern provides further support for the argument that partitioning in domestic structures is largely explained by economic drivers. Out of all the regions samples for this study, the Late Mississippian settlement of the Oconee River valley offers the best example of a dispersed settlement pattern. At these upland Lamar settlements, the demands for sharing and the impetus to privatize storage may have been weaker than at nucleated Late Mississippian settlements such as Toqua and King. Architectural Investment Based on ethnographic and archaeological studies discussed in Chapter 2, I hypothesized that changes in architectural investment described in Chapter 4 can be explained by broad-scale changes in economic behavior. Specifically, I hypothesized that increasing architectural investment would track with increasing investment in maize agriculture, and that it might be seen to decline during the Historic Indian period, when some native groups adopted a more mobile subsistence strategy to cope with changes wrought by contact with Europeans. As demonstrated in Chapter 4, the in-place rebuilding of domestic structures is rare prior to the widespread adoption of maize agriculture around A.D. 900, and then becomes increasingly common from the Early Mississippian to the Late Mississippian period (Table 4.18). Shallow

201

house basins, an indication of additional labor investment that only appeared in the western part of the study area during the Late Woodland period, are widespread across the region during the Early Mississippian period. They appear at BBB Motor, Bessemer, Cahokia, Kimberly-Clark, Macon Plateau, Martin Farm, Moundville, and Toqua. The walls of Early Mississippian houses were built with using one of three techniques: small, closely spaced, single set posts; small, closely spaced, posts placed in a wall trench; or, some combination of wall trenches and single set posts. The majority of Early Mississippian houses have wall trenches. In some cases there appears to be a diachronic shift from single post houses to wall trench structures from the early to the late part of the Early Mississippian period (e.g. at Cahokia and Moundville), but in parts of the Southern Appalachians, single post construction persists throughout the period. Some researchers (Polhemus 1987; Peregrine 1992:137-138) suggest that the shift to wall trench construction is an indication of increasing sedentism and architectural investment. The excavation of wall trenches certainly represents a new step in house construction that could have required more labor than setting individual posts by hand. However, wall trenches may have been an innovation that made house building more efficient and less labor intensive. The wall trenches of Early Mississippian houses were probably excavated with stone hoes, and may have allowed for the insertion of prefabricated walls (Lewis and Kneberg 1946:61; Pauketat and Alt 2005:225). Pauketat and Alt (2005:220) argue that ―digging wall trenches meant that housefoundation-excavation need not have remained the same sort of family effort that it had been. That is, by digging wall trenches, many house foundations could have been dug in a single day by a solitary work crew‖ (Pauketat and Alt 2005:225). From this perspective, the shift to wall

202

trenches may actually represent a reduction, or perhaps a streamlining, of the labor required for house construction. Architectural investment in domestic structures seems to reach a peak during the Late Mississippian period. Compared to the Early Mississippian period, house walls were built with larger, more distantly spaced, single set posts, signaling a shift away from bent-pole construction. While there is regional variation in house form during the Late Mississippian period, the general shift to larger posts and the increasing frequency of in-place rebuilding is indicative of greater architectural investment than in previous periods. The winter houses of the Southern Appalachians represent the greatest labor and material investment of any domestic structure type recorded in this study. With their four large interior supports posts, deep basins (estimated to be 30-60 cm by Gougeon [2007:141]), prepared clay hearths, daubed interior partitions, earth-embanked sides, and entryway trenches, these houses were more architecturally complex than earlier structure types. They were larger than earlier domestic houses, and they were frequently rebuilt in place. Although this characterization is somewhat oversimplified, there is a secular trend of increasing architectural investment in domestic structures that reaches a peak in the fifteenth century. Historic Indian structures, though retaining many characteristics of Late Mississippian architecture, seem to have been built with a reduced investment of labor and materials. The deep basins of the Late Mississippian period semi-subterranean houses seem to be abandoned by the late eighteenth century. Hally (2002) notes this trend in his comparison of sixteenth century and eighteenth century houses in the Southern Appalachians. Cherokee domestic structures at Mialoquo, Townsend, and Toqua, which date to the latter half of the eighteenth century, do not appear to have basins. Nor do Creek houses from the 1730s

203

occupation at Hoithlewalli or the post-1750s occupation at Fusihathcee (Hally 2002:103). With one exception, the houses at Chota-Tanase (see Schroedl 1986:238-240) do not have basins. In some cases this may be due to plow damage, especially at Chota-Tanase, but in general the pattern seems to reflect an architectural reality. Five Cherokee winter houses at Ravensford do have shallow basins, but these houses likely date to the 1730s or 1740s. Walls were still built with rigid, single set posts, but the posts are slightly smaller during the Historic Indian period than in Middle and Late Mississippian period (approximately 17 cm in diameter versus 20 to 23 cm). Compared to the Late Mississippian period, wall posts are more widely and irregularly spaced in Historic period houses, and the overall post density for all house drops to 1.63 posts per square meter, the lowest post density in the region since the Early Mississippian period (Table 4.19). The in-place rebuilding of domestic structures, a hallmark of the Late Mississippian period, drops off steeply in the Historic period. Only 5 of the 110 Historic period domestic structure in the database show evidence of rebuilding or repair, and all of these houses were only repaired or rebuilt once. Out of this group of structures, five were domestic structures from Ela, Chota-Tanase, Tukabatchee, and the Jenrette site. Waselkov (1994:194) argues that increased participation in deer hunting and deerskin trade led to the abandonment of the semi-subterranean winter houses, as longer winter hunting trips made sedentary winter structures obsolete. Marcoux (2008) suggests that the variability in wall post spacing and the reduction in rebuilding and architectural investment in Cherokee houses at Townsend was a conscious adaptation to the chaotic and violent social landscape of the mid-eighteenth century. Less energy was invested in house construction because occupational duration was uncertain, and the quick abandonment of sites was a crucial survival strategy. I

204

agree with this interpretation, and argue that it extends beyond Townsend, to all the Cherokee sites in the database. Only at Ravensford, which dates to the first half of the eighteenth century, is there much evidence for rebuilding and architectural investment in domestic structures. Modeling Maize Dependence The theoretical models tested above are based on well-documented ethnographic and archaeological cases, but it would obviously be preferable to look for statistical associations between maize dependence and change in architectural features. Unfortunately, the ethnobotanical data from site reports are not detailed and consistent enough to rank the components in the dataset in order by maize dependence. I was only able to record the presence or absence of maize for each site, and not surprisingly, only the Middle Woodland components lack evidence of maize production or consumption. More broadly, there are still debates regarding the timing and scale of maize dependence in the Southeast beyond the widely acknowledged fact that maize agriculture is generally widespread after A.D. 900 (see for example Hutchinson et al. 1998). However, one possible avenue for modeling maize dependence with this dataset is to create interaction variables representing the potential for maize agriculture at a particular component. Maize is cold intolerant and sensitive to drought. Within the study area, the potential for maize production will be probably be higher at sites with more rainfall, higher annual average temperature, and more frost-free days. Two variables, one measuring the interaction of precipitation and annual average temperature, and another measuring the interaction of precipitation and frost-free days, are used to represent the differential productivity of each site. A dummy variable for the presence or absence of maize completes the model (Architectural variation = ƒ[presence of maize, interaction of precipitation and average annual temperature,

205

interaction of precipitation and frost-free days]). This model does not account for differences in soil type, which is a major issue, and this data should be included in a future analysis. I used these interaction variables to test the hypothesis that interior partitioning in domestic structures increases with increasing maize dependence. The results of this model are shown in Table 6.2. The model suggests that a small but significant amount of variation in partitioning in domestic structures can be explained by the combined effects of the presence of maize, the interaction of precipitation and average annual temperature, and the interaction of precipitation and frost-free days (F = 13.13, Prob>F = 0.000, R2 = 0.112, 3 df, n = 317). The partial slope coefficients indicate that there is a difference in partitioning between domestic structures on located on sites with or without maize, but this difference is not significant at the 0.05 level. The partial slope coefficients for the interaction variables are significant, but produce mixed results. Interior partitioning increases by a very small degree as precipitation and average annual temperature increase, but also declines by a very small degree as precipitation and frostfree days increase. This particular model operates on several problematic assumptions, but it does provide a useful counterpoint to the ethnographic models discussed above. The relationship between interior partitioning and maize reliance may not be as straightforward as my more qualitative analyses suggest. Future analyses with better environmental and ethnobotanical data would no doubt improve the hypothesis testing presented in this chapter. Discussion and Conclusions In a general sense, broad changes in certain features of prehistoric and early historic Southeastern architecture—the size and arrangement of domestic structures, interior partitioning, and architectural investment—seem to track predictably with major changes in the Southeastern

206

subsistence economy. From the Woodland to the Mississippian period, based on the size and arrangement of houses and the location of storage feature and hearths, there appears to be a shift from households composed of large corporate groups that frequently cooperated in daily subsistence activities to smaller, more independent nuclear or extended family households that were more autonomous in terms of production and consumption. This suggests that the model of Woodland to Mississippian period household development posited for the American Bottom by Peregrine (1992) is also applicable to the Southern Appalachian region and surrounding areas. There is a general trend toward the increasing partitioning of domestic structures, representing a privatization of storage and a movement of productive activities inside the house. These trends fit with our understanding of the relationships between houses and economic behavior gleaned from ethnographic and cross-cultural research. Large households would have been well suited to the economic system of the Middle Woodland period that relied on a mixed strategy of hunting, gathering, fishing, and the cultivation and collection of small seed crops. This strategy would have required what Wilk and Rathje (1982:631) refer to as ―task simultaneity.‖ Large households would have been able to quickly muster labor for different and simultaneous economic tasks, and could have adapted to changing economic conditions, such as unexpected shortages or windfalls of particular wild foods. By the Early Mississippian period, when maize agriculture was widespread, smaller households may have been a more efficient unit of production and consumption. With the subsistence economy more heavily centered on the annual planting and harvesting of maize, households would have been geared toward the ―linear scheduling of labor‖ (Wilk and Rathje 1982:632); i.e. warm season planting followed by cold season hunting (Hudson 1976; Scarry and Scarry 2005). As households became more heavily invested in sedentary agriculture, the inheritance of land would have become more important,

207

and the transmission of land and property is more efficient among smaller households with fewer sons and daughters (Wilk and Rathje 1982:631-632). There are, however, exceptions to this general trend in household organization, some of which may represent regional variation in economic strategies. In the western and eastern part of the study area there are differences in the architectural transitions from the Middle to the Late Woodland period, the time when maize agriculture and a more sedentary subsistence strategy became increasingly important. In the American Bottom, southeastern Missouri, and west-central Alabama, there is a shift from large, circular structures that may have housed extended families in the Middle Woodland period to smaller, rectangular structures arranged in multi-structure household clusters by the Late Woodland period. The Late Woodland houses could probably hold no more than a single nuclear family. A courtyard group composed of several nuclear families appears to have been the primary unit of domestic production and consumption. The architectural pattern from the Woodland to the Mississippian period in the west shows a clear change in house form at the Middle to Late Woodland transition and then some continuity in house form (i.e. a continuation of the small, rectangular house) during the Late Woodland/Emergent Mississippian to Early Mississippian transition. In the Southern Appalachians, the architecture tells a different story. The form and arrangement of Late Woodland domestic structures at Brasstown Valley seems to indicate continuity with Middle Woodland style architecture and community layout. People continue to build large round houses, and there is less evidence for an increase in architectural investment. These buildings may have housed large extended families, operating individually for most subsistence activities and joining together for occasional communal labor. These types of houses

208

are abruptly replaced by the rectangular structures of the Early Mississippian period. Perhaps the reorganization of labor accompanying the transition from the mixed economic strategy of the Woodland period to the maize agriculture economy of the Mississippian period was more gradual in the western part of the study area and happened later, and perhaps more abruptly, in the east. Settlements take on a more formal arrangement and houses are more solidly built as intensive maize agriculture becomes widespread during the Early Mississippian period. There is a reduction in architectural investment in the eighteenth century as native groups become heavily involved in the more mobile enterprise of deerskin hunting and trading. These trends fit with the expectation, based on ethnoarchaeological research, that architectural investment and the uniformity of house construction will be positively correlated with anticipated use life (Kelly et al. 2005; Kent 1992). This relationship may also explain some forms of domestic architecture that Southeastern archaeologists tend to view as atypical, such as the heavily constructed structures of the Owl Hollow phase of the Eastern Highland Rim in Tennessee and the more lightly built Late Mississippian domestic structures of the upland Lamar sites in the Oconee River valley in the Georgia Piedmont. These exceptions from the general secular trend of increasing architectural investment and uniformity may be the result of regional variation in economic strategies. Faulkner (2002) suggests the Owl Hollow phase structures represent households practicing an unusually sedentary form of hunting, gathering, and collecting for the Middle Woodland period. The dispersed Lamar farmsteads of the Oconee River valley, which seem to represent occupations of only one generation, may represent households engaging in a less intensive land use strategy than roughly contemporary Late Mississippian communities located in floodplains,

209

such as King, Toqua, Dyar, and Coweeta Creek. The households at the Lamar farmsteads may have also faced fewer pressures for sharing in their smaller settlements, resulting in less emphasis on private storage. In this chapter I have examined diachronic change and synchronic variation in observable features of Woodland, Mississippian, and Historic Indian houses and communities and found that they seem to be explained by diachronic change and synchronic variation in the subsistence economy. Along with other researchers in the Southeast, I argue that there is an important, economically driven change in household organization during the Mississippian period, what Pluckhahn (2010:347) calls ―a fundamental change in the organization of domestic production… perhaps including the development of nuclear family households as a basic economic unit.‖ However, it is also clear that there is more diversity in house form and household organization than can be accounted for by strictly economic factors. Economic drivers seem to explain the general size and spacing of domestic structures, certain aspects of the organization of interior space, and differences in the investment in domestic structures. These factors tell us less about why domestic structures are built in particular shapes, or are aligned in certain directions, or why certain floor plans emerge and disappear over time. In the chapters that follow I fill in many of the gaps left by a strict economic approach.

210

Table 6.1 Results of Regression with Dummy Variables for Interior Partitioning Model: Interior Partitioning = ƒ(dummy variables for time periods) Structures (n) 317 F( 5, 311) 17.62 Prob > F 0 R-squared 0.2207 Partitions MW LW EM MM HI Constant

Coef. t P>t -1.619916 -6 0 -1.432174 -6.18 0 -1.565967 -8.25 0 -0.3188406 -0.6 0.549 -0.5855072 -2.45 0.015 1.652174 11 0

Table 6.2 Results of Multivariate Regression with Interaction Variables for Maize Reliance Model: Interior Partitioning = ƒ(presence of maize, interaction of precipitation and average annual temperatures, interaction of precipitation and frost-free days) Structures (n) 317 F( 3, 313) 13.19 Prob > F 0 R-squared 0.1122 Partitions Maize present PrecipXavgtemp PrecipXfrostfree Constant

Coef. t P>t 0.4501445 1.78 0.077 0.0012553 5.69 0 -0.0003012 -4.84 0 -0.825934 -0.83 0.407

211

Figure 6.1 Model of Household Clusters for Each Chronological Period

212

Figure 6.2 Examples of Middle Woodland Period Domestic Structures (Sites, clockwise from left: McFarland [Kline et al. 1982], Garden Creek [Keel 1976], 9GE333 [Ledbetter 2009], Fernvale [Steere and Deter-Wolf 2011], Duncan Tract [Faulkner 1988])

213

Figure 6.3 Examples of Late Woodland Period Domestic Structures (Sites, left to right: Brasstown Valley [Cable et al. 1997], Rivermoore [Markin 2007])

214

Figure 6.4 Examples of Early Mississippian Period Domestic Structures (Sites, left to right: Hiwassee Island [Lewis and Kneberg 1946], Kimberly-Clark [Chapman 1994], Banks V [Faulkner and McCollough 1978])

215

Figure 6.5 Examples of Middle Mississippian Period Domestic Structures (Sites, left to right: Town Creek [Boudreaux 2005], Hickory Log [Paul Webb personal communication 2006])

216

Figure 6.6 Examples of Late Mississippian Period Domestic Structures (Sites, left to right: Dog River [Poplin 1991], Little Egypt [Gougeon 2002], King [Hally 2008])

Figure 6.7 Examples of Historic Indian Period Domestic Structures (Sites, left to right: Alarka Farmstead [Shumate et al. 2005], Brasstown Valley [Cable et al. 1997])

217

CHAPTER 7 HOUSES AS SYMBOLS

The ethnographic, architectural, and archaeological studies discussed in Chapter 2 indicate that ritual beliefs and symbolism are powerful determining factors in the design and construction of domestic and non-domestic architecture. Yet until recently, research on household ritual and symbolism in the Southeast has been limited. This may be due to a general tendency for archaeologists to consider ritual and domestic contexts as mutually exclusive. More broadly, even in places with excellent preservation (e.g. Mesoamerica, the Near East, and the Southwest), teasing out ritual behavior and symbolism from the material remains of everyday household activities is no simple task. In recent years, Southeastern archaeologists have linked certain features of native architecture to some of the better understood religious and cosmological symbolism of Southeastern Indian groups recorded during the historic period. Some of the more widely recognized religious and cosmological symbolism includes: the separation of opposite categories (e.g. summer/winter, male/female), a concept of the quartering of the earth and a below-world and above-world, the symbolic importance of the cardinal directions, and the symbolic importance of the numbers four and seven (see Hally 2002; Hudson 1976; Mooney 1900). Some researchers (Hally 2002, 2008; Sullivan 1995) suggest that many of the patterns of variation in Late Mississippian and Historic period domestic and non-domestic structures may be explained in terms of these symbols. This includes variation in house size and shape, numbers of wall posts, segmentation within houses, and the orientation and layout of houses in communities. 218

In this chapter I attempt to determine which of these lines of architectural variation seem to be best explained by symbolic behavior. I search for evidence of these symbolic associations in Woodland, Mississippian, and Historic period structures. In some cases, I argue that aspects of known historic Southeastern Indian cosmology, such as an emphasis on the separation of opposites through paired buildings, may stretch as far back as the Woodland period. I suggest that symbolic behavior can explain much of the variation in architecture that cannot be explained in strictly mechanical or economic terms. Household Symbolism in the Middle Woodland Period Despite recent attempts to correct a longstanding bias toward large, impressive Middle Woodland sites with monumental architecture, very little research has been directed toward understanding ritual and symbolism at the household level. In his recent review of household archaeology in the Southeast, Pluckhahn (2010) only cites one publication that addresses household ritual and symbolism in the Woodland period: his own work on feasting at Kolomoki (Pluckhahn et al. 2006). There is little discussion of ritual and symbolism in domestic contexts in The Woodland Southeast (Anderson and Mainfort 2002), and the few papers that discuss Middle Woodland domestic architecture (Clay 2002; Faulkner 2002; Wetmore 2002) do little to examine possible symbolic or ritual factors influencing house form. There is a general consensus that the complex earthworks, burial mounds, and broadly circulated exotic artifacts of the Middle Woodland period point to a widely shared symbolic and ritual system (Anderson and Mainfort 2002; Carr and Case 2006). Domestic architecture during the Middle Woodland period is still poorly understood, but most scholars argue that beneath the broad, integrative system associated with the earthworks and artifacts of the Hopewell Interaction Sphere, there was much cultural diversity in domestic contexts (Anderson and

219

Mainfort 2002:9-10). Unlike the Mississippian period, when site plans and house form are similar across broad geographic regions, there are fewer formal similarities in Middle Woodland domestic sites and structures. With a few exceptions, which I address below, Middle Woodland domestic structures are generally circular or oval shaped, single post houses with no interior partitions. Interior hearths and features are rare, as are interior burials. The most common Middle Woodland domestic floor plan is round and open, with few archaeologically visible divisions of interior space. Structures 2, 3, and 4 at McFarland provide an excellent example of this kind of floor plan (Figure 7.1). It is tempting to suggest that there would have been few ideological or cosmological associations with such simple buildings. However, the ethnographic studies discussed in Chapter 2, especially the Navajo and Amazon examples, provide useful insight for interpreting these deceptively simple buildings. First, it is clear that architecturally undifferentiated structures may be conceptually divided into male and female areas, and that in some cases the female areas are also private areas where household production takes place. In some cases, male spaces are more commonly used for public activities or gathering places. Second, these conceptual divisions can be fluid. In the Amazon cases, the same houses used for daily living are transformed into important ritual places when they serve as the setting for public ceremonies with large gatherings, feasts, and dancing. Given the strength of the ethnographic data, it seems likely that occupants of the round, non-segmented houses at Brasstown Valley, Hickory Log, McFarland, Two Run Creek, Ela, 9GE333, Banks III, Banks V, and Duncan Tract would have divided these buildings into male/female and public/private spaces. This proposition could be tested by comparing the distribution of artifacts associated with predominantly male and female production activities.

220

Gougeon (2002) performed this analysis on the Late Mississippian structures at the Little Egypt site, and successfully identified male and female activity areas. Similarly, the gender distribution of subfloor burials could also be used to test this hypothesis. Due to the lack of house basins and generally poor preservation, floor debris is rarely recovered from these structures. Subfloor burials are rare in domestic Middle Woodland structures (out of the entire sample, only one occurs, a single flexed burial on the eastern side of House 6 at Two Run Creek). While a gendered division of space seems likely but is hard to test archaeologically, there may be better evidence to support the proposition that the spaces in some Middle Woodland domestic structures were conceptually transformed during certain community or household level rituals, as in the Amazonian cases described above. There is a tendency in much of the Middle Woodland literature to separate the domestic and ritual spheres, which creates the perception that most ritual and ceremonial behavior took place at large sites with earthworks or platform mounds, while little more than subsistence and productive activities took place at domestic sites. This may largely be true. However, at sites like Brasstown Valley, Hickory Log, Ela, and McFarland, where there are at least three or more probable contemporary structures, there may have been meetings, feasts, or dances that helped foster community integration and decision making among discrete households. The open floor plans of houses at these sites, like the Jívaro and Tukanoan structures, would have allowed for such events (see Harner 1972; Hugh-Jones 1979). The average floor area of the circular and oval Middle Woodland structures with open floor plans is approximately 43 m2. This figure is relatively high: it is much larger than most Late Woodland and Early Mississippian domestic structures, and is on par with the area of Middle and Late Mississippian and Historic Indian domestic structures, which are subdivided into smaller rooms, but were not used for meeting and dancing. While the relatively large size of Middle

221

Woodland structures may simply reflect larger household size, it could be that the large, open floor plans reflect an intentionally fluid design that allowed the structures to serve both domestic and ritual functions, depending on the context. Cross-cultural research shows that even the simplest house designs can reflect cosmological symbolism. In his comparison of traditional agrarian housing in China, Java, Japan, Thailand, Vietnam, South Asia, Southwest Asia, and Mesoamerica, Blanton (1994:77-110) noted a wide range in the degree to which houses clearly communicated principles of cosmological or ritual order, but in almost all cases there was some evidence of such an influence on the placement or design of the house. At a finer scale, Kent‘s (1983:83-84) discussion of Navajo hogans and ramadas demonstrates that the shape of a structure alone may be symbolically charged. It seems reasonable to suggest that the round shape of many Middle Woodland structures reflected cosmological beliefs. Understanding the nature of the symbolism embodied in the structures is a more difficult enterprise. One avenue for research is to examine other material culture for insight into the broader symbolic system that may have influenced house construction. For example, recent studies of Middle Woodland iconography suggest that Swift Creek pottery designs reflect cosmological and religious beliefs (Anderson 1998). Swift Creek design motifs are primarily curvilinear geometric forms, and according to Frankie Snow (1998:63), ―less abstract depictions… recognized as flowers, serpents, birds, insects, and wolflike and other animal heads, plus human-mask like designs.‖ Some Swift Creek researchers argue that these designs were important religious symbols and they may have parallels with known religious symbols during the historic Indian period (Anderson 1998:291; Saunders 1998:156-158; Snow 1998:62-63). They may invoke such cosmological themes as the quartering

222

of the world (Snow 1998). The suggestion that Swift Creek pottery designs and other Middle Woodland iconography on copper ornaments, carved stone pipes, effigies, etc, offer a bridge to better understood historic Native American belief systems is intriguing, although it is still something of a leap to make the kind of associations between Woodland period house form and mythological and cosmological beliefs that Hally (2002, 2008) establishes in a Late Mississippian context. Not all of the Middle Woodland domestic structures in the database have round, open floor plans, and these houses merit special attention. These include a single, small, keyholeshaped structure at Kolomoki, the large, oval-shaped Owl Hollow phase structures in Middle Tennessee, and a similar structure at 9GRX1 in Alabama, a square, Connestee phase building at Garden Creek, and the large, somewhat irregularly shaped buildings at the Yearwood Site. Taken as a group, these structures represent variability in Middle Woodland house form that contrasts with the Mississippian period, when houses are quite similar over much of the Southeast. They may also point to regional variation in architectural symbolism. The large Owl-Hollow phase structures of the Eastern Highland Rim in Middle Tennessee are more robust than most Woodland structures, containing four substantial interior support posts and large interior hearths. These structures were sometimes paired with more lightly built rectangular structures. Faulkner (2002:196-197) considers this arrangement to represent paired winter and summer dwellings, not unlike those of the eighteenth-century Cherokee. Moreover, he suggests that, ―these structures were sometimes arranged on a circular midden ring around a debris-free open area or ‗plaza‘‖ (Faulkner 2002:196). This site plan is comparatively formal for the Middle Woodland period, and the interiors of the large winter dwellings would have been segmented into rooms by the large posts and hearths. These buildings

223

may also have been divided into male/female and public/private spaces, but the additional segmentation of the built environment, including a more obvious plaza for gatherings and ritual events, seems to reflect a different kind of household symbolism when compared to the round, open structures at other Middle Woodland sites. The large, square, rectangular, and irregularly shaped structures at Yearwood may have been primarily ceremonial or public buildings based on their size, placement, and associated artifacts (Butler 1979). The lack of regular patterning in the architecture at Yearwood is puzzling. Given the fancy exchange goods and cremations, which suggest the site may have been a symbolically charged place, one might expect the buildings to be especially well constructed, but this is not the case. The single Connestee-phase building at Garden Creek (Keel 1976) has very regularlyspaced wall posts and a single central post, and recent excavations at the Macon County Airport site have unearthed several structures with a nearly identical form (Tasha Benyshek personal communication 2010). The Garden Creek and Macon County Airport houses may have been primarily domestic. Like most other Middle Woodland houses, they would have had fairly open floor plans, and may have been conceptually divided into male/female and public/private space. As there are no clear non-domestic structures at either site, they may have occasionally served as gathering places during special events, with the interior space changing depending on the social context. Keyhole-shaped, semi-subterranean structures like the one at Kolomoki are unusual in the deep South. There are similar structures in the Midwest (Binford et al. 1970; Kelly et al. 1987). At only 7.5 m2, the house at Kolomoki could have served as little more than a nuclear family residence. In addition to improving the thermal efficiency of the structure, the house basin may

224

have had symbolic associations with cosmological concepts of emergence or underworld commonly recorded among historically known native groups in the Southeast, but this is speculative. Following Hally‘s (2002; 2008) examination of wall posts at King, I considered the possibility of cosmological symbolism in Middle Woodland wall posts. In contrast to the King site, there is less evidence for an important, canonical number of wall posts at the single-site or broader scales. For all Middle Woodland sites, the average number of wall posts ranges from only 15 at Fernvale to 37 at 1GRXI, with most falling between 20 and 30 wall posts per house. Sites with the multiple domestic structures, Duncan Tract (n=5), Ela (n=10), Hickory Log (n=13), and McFarland (n=5), offer the best opportunity to find meaningful patterns in wall posts counts. At Duncan Tract there are an average of 21 posts per structure, but the range is quite wide, and the number of posts co-varies with house size. At Ela and Hickory Log, where there are an average of 22 and 25 wall posts per house, respectively, but the ranges are large, and the distributions are too widely dispersed to suggest a clear attempt at using a pre-determined number of wall posts. At McFarland, where houses are quite uniform, there are an average of 33 wall posts per house, which falls at the high end for the Middle Woodland period. Wall posts counts range from 28 to 41 per structure. The spacing of posts is consistent within each house, but not across houses. Given the variability in wall posts counts at the intra- and intersite scale, there is little good empirical evidence for a special number of wall posts in Middle Woodland structures. Household Symbolism in the Late Woodland Period During the Late Woodland period there are wide-scale transformations in house construction and pronounced regional variation in house form that may be related to changes in

225

ritual or symbolic practices. At sites from the western part of the study area, specifically in the American Bottom, southeastern Missouri, and west-central Alabama, there is a shift toward small, uniform, rectangular structures, often arranged in small groups around a small courtyard or plaza. In many cases these houses are set in shallow basins. Structure 1 from 1PI61 provides a good example (Figure 7.2). In terms of shape and construction technique, these houses signal a clear break from the architectural style of the Middle Woodland period. In the Southern Appalachians, house form is not as well understood, but seems to be more variable. At roughly contemporaneous sites in northern Georgia and piedmont North Carolina there are round and square single post houses in a wide range of sizes, and one wall trench house on a low mound. In contrast to the western sites, in the Southern Appalachians there is some evidence of continuity with the building traditions of the Middle Woodland period, especially in the large, circular, Late Woodland period structures at Brasstown Valley and Town Creek. Figure 7.3 shows some of the Late Woodland houses in Locus B of Brasstown Valley. Researchers from various theoretical camps argue that the widespread adoption of maize agriculture in the Late Woodland period would have been correlated with significant changes in ritual and symbolic systems (see for example Cobb and Garrow 1996; Emerson et al. 2000; Muller 1997; Pauketat 1994). Anderson and Mainfort (2002:18-19) offer a fairly representative interpretation, arguing that changes in iconography from the Middle to Late Woodland period suggest a shift toward:

… a more complete collective, elite-directed communal ceremonialism centered on agricultural productivity and reinforcing the sacred and hereditary position of the leadership. Individualistic expression appears to be channeled from more

226

traditional emphases on hunting to success in warfare and militarism… In contrast, changes in the underlying cosmology of the southeastern Indians, such as the quartering or layered nature of the world, appear to have been much less pronounced.

Domestic architecture in the western part of the study area may reflect this shift toward agrarian symbolism while retaining broader, more resilient cosmological themes. The small rectangular houses of the Late Woodland or Emergent Mississippian period at sites like BBB Motor, 1PI61, Hoecake, and Cahokia represent a dramatic departure from most architectural forms of the Middle Woodland period. The uniformity of the structures would certainly seem to reflect and reinforce a communal or corporate orientation (sensu Blanton et al. 1996). The lack of interior space, along with the emergence of readily identifiable plazas suggests a possible movement of ritual and ceremonial practices out of domestic structures and into public places. We can imagine several households gathering within the open, 6 to 8 m diameter structures at the Middle Woodland period houses at Brasstown Valley and McFarland, but this would have been practically impossible in the roughly 8 to 16 m2 rectangular houses at the western Late Woodland sites. There is little evidence of interior partitioning in the Late Woodland houses in the western part of the study area. Kent‘s (1990) cross-cultural research would suggest conceptual male/female and public/private divisions of space, but given the miniscule floor areas of these houses, the divisions would have to be subtle. On one hand, it is hard to imagine that these houses provided space for little more than basic domestic activities for a small nuclear family, such as sleeping, cooking, and some storage. However, as Blanton (1994) shows in his cross-

227

cultural study of traditional agrarian houses, special ritual or private spaces are sometimes marked in small dwellings with small altars or shrines, which may or may not be detected archaeologically. Zinacantan Maya houses observed by Vogt (1969) are very small by modern standards and have few to no interior partitions, yet ―these houses illustrate, in addition to the usual Catholic altars, an orientational preference, gender-specific use of space, a cosmological basis in layout and space use, and physical evidence of lineage ideology in the form of crosses placed near the entrance to each house‖ (Blanton 1994:102). There is a clearer picture of architectural symbolism at the site level. In contrast to the Middle Woodland settlements and the Late Woodland sites in the Southern Appalachians, the Late Woodland communities in the western part of the area demonstrate a common layout that may be rooted in cosmological principles. During the George Reeves phase at the Range site (Kelly 1990) and during the Emergent Mississippian phase at Cahokia Tract 15A (Pauketat 1998:135), small rectangular domestic structures are oriented to the cardinal directions around a central plaza containing four large pits and a central pole. BBB Motor, Hoecake, and 1PI61 lack the central pits and pole seen at Cahokia and Range, but the structures at these sites are oriented either perpendicular or parallel to one another, and there is evidence at each site for a small plaza or courtyard area. A histogram of the angle of orientation of all the Late Woodland domestic structures shows quite a bit of variation, but there is a tendency toward orientation with the cardinal directions (Figure 4.22). At all of these sites the consistent pattern of houses oriented at right angles to one another around a common plaza clearly reflects planning, perhaps in accord with cosmological principles. This common architectural pattern of houses oriented to the cardinal directions around an open space suggests a reference to an axis mundi, a shared concept of a world‘s center where the four cardinal directions meet and intersect with the boundary of the

228

upper and lower worlds. This pattern is not seen in domestic contexts during the Middle Woodland period, but it does continue into the Mississippian period with further elaboration. In the Southern Appalachians, there is less conclusive evidence for broadly shared underlying symbolism in Late Woodland domestic architecture. At Brasstown Valley the form of structures changes very little from the Middle Woodland period, suggesting continuity with older building principles. The lone structure at Town Creek that may date to the Late Woodland period is non-domestic, and it is large and circular. At Rivermoore, the three square structures in Domestic Area 1 may have been arranged around a relatively clear courtyard or plaza, but the houses are not oriented at obvious right angles to one another. The layout of the settlement has more in common with 1GRX1 than to the much closer Brasstown Valley. If not for the Woodstock ceramics and an early radiocarbon date, the wall trench structure at Summerour would almost certainly be identified as an Early Mississippian building. In contrast to the Middle Woodland period, there is slightly better evidence to suggest a reflection of cosmological symbolism in the number of wall posts used in houses. For all the Late Woodland sites, the range of the average number of wall posts is fairly tight, from a minimum of 34 at Hoecake to a maximum of 46 at 1PI61. At all the sites with multiple structures, the distribution of wall post counts is close to normal. At both Hoecake and BBB Motor, there may be a wall construction pattern similar to the one identified by Hally (2002; 2008) at King. At Hoecake, there are an average of 34.2 (s.d. 7.5) wall posts, and at BBB Motor during the Edelhardt phase there are 35.1 (s.d. 11.2). In both cases the wall post count does not strictly vary with house size, and at both sites there is pattern of approximately 7 posts in the short walls and 10-12 posts in the long walls. It may be that the previously discussed uniformity in size is the driver behind this uniformity in wall posts, but this

229

pattern may point to a concerted effort to place seven posts on the end walls of houses. While it is somewhat risky to apply cosmological principles from the Historic Indian period to Late Woodland societies, the widespread belief among Southeastern Indians that seven was a sacred number (Hudson 1976) could stretch back to the Late Woodland period. Similar patterns are harder to identify at the other Late Woodland sites, but the overall consistency in the number of wall posts in both the eastern and western parts of the study area is striking. Unlike the Middle Woodland period houses, there is not a close correlation between house size and number of wall posts. Household Symbolism in the Early Mississippian Period The formalization of community plans seen in the Late Woodland period continues in the Early Mississippian period. The orientation and alignment of houses reflects this trend. Houses are often oriented to the cardinal directions and either parallel or perpendicular to one another. A histogram of the orientation of Early Mississippian domestic structures shows a bimodal distribution centered roughly around 0 and 90 degrees, although there is also a tendency for houses to be oriented toward the southeast (Figure 4.24). Household clusters composed of domestic structures and small open areas can be identified at many Early Mississippian sites. In some cases the individual household clusters will conform to a broader, sitewide orientation. This is the case during the Lohmann phase occupation at Cahokia 15A, and also during the Early Mississippian occupations at Jewell and Morris. In other cases, such as Moundville, the household clusters are less clearly aligned to a broader site plan. At some sites there is good evidence that the orientation of household clusters changed in lockstep with major changes in the organization of the settlement, creating a visual symbol of the new social or political order. For example, at the ICT-II tract at Cahokia during the Lohmann

230

phase, the long axis of almost all houses were aligned within a few degrees of the cardinal directions, providing evidence ―that the planning of the new Lohmann community at the ICT-II tract was tied to a pan-Cahokian pattern (Collins 1997:126). During the subsequent Stirling phase, the orientation and placement of structures shifts, reflecting a new alignment with a local mound and plaza complex (Collins 1997:129-131). Collins (1997:128) argues that the change in layout and orientation of houses reflects a change from households tied to ―sitewide central authority‖ to a ―local social hierarchy.‖ Compared to the Middle and Late Woodland periods, domestic structures across the Southeast look more similar over a broad geographic area. Simultaneously, there is more architectural variability within sites. Rectangular and square floor plans become increasingly common at a macroregional scale, accounting for 399 of the 524 Early Mississippian structures in the database. The rectangular shape is even more dominant among strictly domestic structures: 352 of the 444 domestic structures are rectangular or square. This trend holds in both the eastern and western halves of the study area, signaling a reduction in regional variability that lasted for hundreds of years. However, there is also increasing evidence for special purpose and high-status structures, buildings whose shape and size convey different visual messages than ordinary dwellings. There are frequently a limited number of circular, t-shaped, or irregularly shaped buildings at Early Mississippian sites. This pattern is especially well documented at Cahokia Tract 15A, Hiwassee Island, Jewell, and Toqua. There is also increasing variation in structure size during the Early Mississippian period. As shown in Chapter 4, there is often a bimodal distribution of structure sizes, with many small domestic structures and a few large public buildings and chief‘s houses, often placed on or near mounds.

231

Wall trench construction is perhaps the most notable shift in building style from the Late Woodland to the Early Mississippian period, especially in the American Bottom, where the overall shape and size of domestic structures changes relatively little. The wall trenches of Early Mississippian houses were probably excavated with stone hoes, and may have allowed for the insertion of prefabricated walls (Lewis and Kneberg 1946:61; Pauketat and Alt 2005:225). In some cases, individual postholes can still be identified archaeologically when wall trenches are excavated, and in other cases they cannot. From a broad perspective, the architectural database confirms a long recognized pattern: between A.D. 1000 and 1100 rectangular wall trench houses dominate the domestic landscape in the Southeast. On the other hand, at finer scales there are important exceptions to this trend. The shift to wall trench construction probably reflects changes in the organization of labor for house construction, but the exact nature of this change is unclear. Pauketat and Alt (2005:220) argue that the variability in post depths from Emergent Mississippian houses in the American Bottom point to house construction by families or households. The more standardized wall trenches, on the other hand, could have been dug by a single person with a hoe. They claim, ―digging wall trenches meant that house-foundation-excavation need not have remained the same sort of family effort that it had been. That is, by digging wall trenches, many house foundations could have been dug in a single day by a solitary work crew‖ (Pauketat and Alt 2005:225). Several aspects of wall trench house form and construction may carry important religious or cosmological symbolism. The four trenches may correspond to the four cardinal directions, or refer to the number four, which was a symbolically charged number for historic Southeastern groups (Hudson 1976). By the Early Mississippian period, intensive maize agriculture was the productive base for most communities. The act of constructing a house with an increasingly

232

important agricultural tool, the stone hoe, may have been a symbolic act that linked the house and the process of house construction to the subsistence base, but this is speculative. Given the difficulty of identifying individual posts in wall trench structures, I was unable to identify a consistent pattern in the number of wall posts that might reflect cosmological symbolism. For the Early Mississippian structures with single set posts, the number of posts generally varies with house size. The emphasis seems to have been on regular, tight post spacing, rather than achieving a predetermined number of posts in each wall. Wall trenches are the most common wall type at Early Mississippian components, accounting for over 50 percent of all structures. However, people continued to build single post structures in the Early Mississippian period. These account for 86 structures at BBB Motor, Bessemer, Cahokia, Hiwassee Island, Jewell, Kimberly-Clark, Kincaid, Macon Plateau, Martin Farm, Morris, Moundville, Toqua, and Town Creek. In most cases, they make up a minority of the structures, but they are quite common at Hiwassee Island, Toqua, and Martin Farm. Fifty-five structures from Banks V, Bessemer, Cahokia, Chota-Tanase, Hiwassee Island, Kincaid, Macon Plateau, Martin Farm, Morris, Moundville, and Toqua have a hybrid wall style, with both single set posts and wall trenches. This provides some support for the recent suggestion that single-post structures persist in the Southern Appalachians during the Early Mississippian while the transition to wall-trench houses in the Central Mississippi valley is more complete (Cobb and King 2005:169). At a regional scale, Pauketat (2003) and Pauketat and Alt (2005) have argued that the single post or hybrid wall houses at Early Mississippian sites near Cahokia represent a kind of resistance against a Mississippian architectural template imposed from the top down by rulers at Cahokia. Similar processes may be at play in other places, and perhaps within sites. At Hiwassee

233

Island, Jewell, Kincaid, Macon Plateau, Martin Farm, Morris, Moundville, and Toqua, there is some evidence that structures with different wall types were built and occupied at the same time. This variability at the single-site scale may represent individual households electing to maintain an older architectural style that carried symbolic ties to a different and/or older social order. Early Mississippian floor plans appear to be fairly open. Only 22 of the 524 Early Mississippian period houses have clearly visible partitions, and the only site with an especially high frequency of partitioned houses is Hiwassee Island, where 6 structures, 5 of which are mound-summit buildings, have well-defined interior walls. Only 14.2 percent of the Early Mississippian houses in the database have interior hearths, which would have created a de facto division of central and peripheral interior space. On the other hand, there is an increase in the number of interior posts from the Late Woodland to the Early Mississippian period. This may point to an increasing prevalence of interior furnishings, such as benches. Most Early Mississippian domestic structures range from 20 to 45 m2 in floor area, and some public and mound-top structures measure over 100 m2. Compared to the much smaller domestic structures of the Late Woodland period, in these larger houses there is more room for defining gendered space and specialized activity areas. In domestic structures located off mounds, there may have been a central-public versus peripheral-private division of space, with a relatively open area in the center of the structure for cooking, eating, and receiving visitors, while the walls of houses may have been lined with sleeping benches that were considered private household space. Compared to the Late Mississippian period, there are fewer symbolic markers of permanency and continuity in Early Mississippian houses. With a few notable exceptions, such as the Late Moundville I-Early Moundville II stage domestic structures at Moundville and the

234

Hiwassee Island phase domestic structures at Hiwassee Island, there is relatively little in-place rebuilding during the Early Mississippian. Mound summit structures are often rebuilt, but domestic structures in villages most often have a single building stage. This is in keeping with the buildings patterns of the Late Woodland period, but offers a sharp contrast to the Late Mississippian period, when domestic structures are often rebuilt in place. Sub-floor burials in Early Mississippian houses are extremely rare. This is also a sharp contrast with the Late Mississippian period, when sub-floor burials are common, suggesting that domestic structures were imbued with mortuary symbolism. In sum, certain features of Early Mississippian structures were no doubt guided by mythological or cosmological principles. It seems plausible that the wall trench house form was one symbol of a new economic, political, and social order. Yet domestic structures themselves seem somewhat unadorned and anonymous, especially compared to the more elaborate domestic structures of the Late Mississippian period. Instead, architectural symbolism at the site level— the orientation of houses and the orderly, planned layout of mound and plaza complexes—seems to carry more weight. This would seem to communicate a more corporate or communal kind of ideology at Early Mississippian sites, one in which individual differences between households in villages were played down, and in which the permanency and continuity of individual households not located on top of mounds was less important. The relatively small size of domestic structures, the ubiquity of plazas, and the new pattern of a few special structures (e.g. the circular wall trench buildings at Cahokia, Toqua, and Hiwassee Island) and public buildings suggest that most of the major ceremonial and public activities took place outside of individual houses. Some ritual behavior almost certainly took place in domestic structures, but evidence for this is less clear.

235

Household Symbolism in the Late Mississippian Period There are important changes in domestic architecture from the Early to the Late Mississippian period. In some places, one could argue that the architectural transitions from the Early to the Late Mississippian are more dramatic than those that mark the transition from the Late Woodland to the Early Mississippian period. For example, by the Late Mississippian period, small, bent pole construction is almost completely abandoned in favor of large, widely set, rigid posts. This is just as dramatic a shift as the change from small single set poles to wall trenches in the American Bottom during the Late Woodland to Early Mississippian transition. In that case, the same size poles were still cut, and the construction technique—weaving a latticework of bent poles—may have changed little after the walls were set in the ground. Wall trench construction appears to be abandoned after the Early Mississippian period in the Southern Appalachians, although it continues to the west at Snodgrass during the Powers phase and possibly in late occupations at Kincaid during the Middle Mississippian period. If wall trench construction was a ritually or cosmologically loaded act, then its absence may speak to changes in household symbolism after the Early Mississippian period. In recent years researchers have made major gains in the understanding architectural symbolism in the Late Mississippian period. Compared to houses from earlier and later periods, Late Mississippian houses show a high degree of architectural investment and rebuilding, regular patterns of interior segmentation, regular patterns of subfloor burials, and a clearer division of public and private space. The architectural database presents the opportunity to test some recently proposed models and identify new patterns of variation that may be explained by underlying ritual or cosmological symbolism. There is strong evidence to suggest that household identity is strongly expressed in domestic architecture.

236

Compared to some Early Mississippian communities, in which the site plan is quite orderly while there is more variation in the layout of households, in many Late Mississippian communities the layout of the individual household cluster seems to be just as important and formal as the spatial organization of the site as a whole. Sullivan (1987, 1995) offers support for this pattern, arguing that the overall layout of Late Mississippian, Mouse Creek-phase settlements in eastern Tennessee mirrors the spatial organization of individual households. She notes that at Ledford Island:

A large structure measuring 15.25 m on a side was situated on the north side of the plaza. Except for its size and location, this structure was architecturally identical to the household winter structures. To the northeast and southwest corners of the plaza were burial groups not associated with household units. The arrangement of the central or ‗public‘ portion of the Ledford Island site, including the large structure and plaza cemeteries mimics that of the ‗private‘ household (Sullivan 1995:107).

This suggests that household-level architecture may have reinforced the same Mississippian cosmologies and principles of social order made manifest at the community level. At King, Warren Wilson, and Toqua, the entranceways of houses are arranged to provide access to the small open areas or courtyards of the household clusters, rather than uniformly facing the larger plaza. This is obviously a practical arrangement for daily living, but it may also point to the symbolic importance of the small, open area in the household cluster. Figures 4.26 shows a histogram for the orientation of Late Mississippian structures. Note that there is more variability than in the Early Mississippian period, but a general tendency toward a southeastern orientation,

237

and also a tendency to orient the corners of the structures with the cardinal directions. The variability in orientation may reflect this tendency to prioritize the orientation of houses to the household cluster rather than to a sitewide plan. Note that Late Mississippian non-domestic structures, larger townhouses and public buildings, are almost strictly oriented toward the Southeast, with their corners facing the cardinal directions (Figure 4.27). This finding bolsters Sullivan‘s (1995) argument for parallels in architectural symbolism in domestic and nondomestic Late Mississippian structures. Late Mississippian domestic structures in the Southern Appalachians are quite uniform across sites. Gougeon (2007:137) argues that Late Mississippian structures in northwestern Georgia shared an underlying architectural grammar, which accounts for their similarity across the region. Hally (2002:108, 2008:85-86) has recently offered an interpretation of the cosmological symbolism of Late Mississippian houses in the Southern Appalachians. He argues that sixteenth-century, square, semi-subterranean winter houses were used ―to symbolically express a number of cosmological and mythological beliefs‖ (Hally 2002:108). The square floor plan of the structures may correspond to the shape of the earth, the four walls and four interior support posts may correspond to the cardinal directions and the sacred number four. The pattern of four interior support posts around an interior hearth and the partitioning of the house into as many as eight or nine small areas (an entrance, a central area around the hearth, and six or seven smaller rooms) occurs at King, Little Egypt, Dog River, Coweeta Creek, Dallas, Dyar, Loy, Pott‘s Tract, Rucker‘s Bottom, Rymer, Toqua, and Warren Wilson. Figure 7.4 is a schematic drawing of the Late Mississippian winter house (see also Gougeon 2002:190, Figure 10.2) In addition to the symbolic associations that Hally (2002, 2008) identifies, I would argue that the division of space within the house created by the four interior posts and the interior

238

partitions may be a reference to the quartering of the earth, another underlying cosmological principle shared by many Southeastern Indian groups (Hudson 1976). As these houses were constructed, they would have resembled the circle and cross motif commonly interpreted as a reference to the sacred fire. Hally (2008:85) also found that most houses at King have 28 or 32 wall posts, with either 7 or 8 along each wall. The seven posts along each wall may correspond to the sacred number seven and the number of clans in Cherokee society. Hudson (1976) explains that in Cherokee society the number seven represents the number of levels in the sky dome and the seven directions, which are the four cardinal directions plus up, down, and center. It is especially noteworthy that the number of posts in each wall does not seem to vary with structure size. There are also physical similarities between the houses and prehistoric platform mounds, suggesting symbolic linkages between public and domestic ritual (Hally 2002:108-109; 2008:85-86). Few sites have the fine-scale wall post data necessary to test for the exterior wall post pattern Hally identified at King, but it does appear that the domestic structures at Little Egypt, Dyar, and Dallas have approximately 28 or 32 wall posts. As Hally (2008:87) indicates, domestic structures at Toqua do not exhibit this pattern. This is partly due to a high degree of rebuilding at Toqua, but it may also represent a genuinely different pattern. Structure 2 at Rucker‘s Bottom has the floor plan typical of houses in the Southern Appalachians. It was rebuilt once, and I interpret it as having 27 wall posts in the first stage and 36 in the second, and four central supports in both stages, falling roughly in line with Hally‘s (2002, 2008) pattern. Away from the Southern Appalachian core, there is notable variation in Late Mississippian house form that may have symbolic or cosmological significance. For example, upland Lamar sites in the Oconee River valley have a very different architectural pattern. There

239

is often a single, large, circular structure with an open floor plan and several smaller rectangular structures. The large circular structures are very uniform in shape and size, but the number of wall posts ranges from 15 to 31, and spacing is variable. The smaller rectangular structures are all very similar, and usually have double posts at the corner. There is no evidence for a predetermined number of posts being used in either structure type. The prevalence of rebuilding and subfloor burials in Late Mississippian houses is one of the most notable architectural changes from the Early Mississippian period. There have been some attempts to examine ancestor veneration through subfloor burials and house rebuilding at King (Hally 2008) and Coweeta Creek (Rodning 2007). It seems clear that the widespread practice of burying deceased kin within the physical space of the household would emphasize connections between lineages and domestic structures, and express a sense of continuity and permanence in households that is not seen in previous times in the region. Several researchers have suggested that the rebuilding of domestic structures may have been a ―symbolic expression of household identity‖ (Pluckhahn 2010:360) at Late Mississippian sites, including Coweeta Creek (Rodning 2004, 2007), Warren Wilson (Moore 2002), and King (Hally 2008; Hally and Kelly 1998). There is good ethnohistoric and archaeological evidence for a division of space along gender lines within Late Mississippian domestic structures. There was a strong division of gender roles between men and women in Creek and Cherokee society (Hudson 1976:317-319). Separating male and female space had a structural significance that fit into a broader ideology of purity and pollution that required keeping binary opposites (e.g. upper world versus lower world) apart. Polhemus (1987; 1998) identified male and female work areas in Dallas-phase houses from Toqua and Loy, as did Gougeon (2002) in his analysis of activity areas at three structures

240

from Little Egypt. These findings were based on the distribution of artifacts that are considered to be associated with male and female activities according to ethnographic and cross-cultural studies. There was some variation in the exact location of these gendered spaces, but in general the organization of space was similar. These studies require the careful examination of floor debris and are time-consuming, but they could be carried out at other Late Mississippian sites in the Southern Appalachians and beyond. The similarity in overall layout of Southern Appalachian Late Mississippian houses suggests that this pattern of gendered space would be widespread. In terms of architectural symbolism, the most striking features of Late Mississippian communities are developments that point to increasing an emphasis on individual households: increasing architectural investment and rebuilding, sub-floor burials, orientation toward household courtyards, and houses that on an individual level may have reflected cosmological symbolism. Much of the important ritual and ceremonial behavior clearly takes place in large plazas and public structures, but the domestic architecture suggests that there may have also been more room, spatially and metaphorically, for household-level ritual and symbolism than in the Early Mississippian period. Household Symbolism in the Historic Indian Period Historic accounts from the eighteenth century provide important details on architectural symbolism in Creek, Cherokee, Chickasaw, and Choctaw communities. These are summarized by Hudson (1976:213-218), and have been instrumental in interpretations of Late Mississippian architecture (see for example Hally 2002, 2008; Rodning 2004, 2007). In addition to the rich historical detail these accounts provide, archaeological data point to important patterns of house construction (distinct from those of the Late Mississippian period) with possible symbolic and cosmological underpinnings.

241

Cherokee households clusters are relatively easy to identify and are fairly uniform in and across sites. At Alarka Farmstead, Brasstown Valley, Chota-Tanase, Ravensford, Toqua, and Townsend, clusters consist of a round winter house, a rectangular summer house, and in some cases, small circular or rectangular storage buildings. At Mialoquo, a single round winter structure is part of a cluster with rectangular buildings, but there is not a clear pairing. At Tuckasegee, only a winter structure was uncovered, while at Hiwassee Old Town, only rectangular structures were identified. There is some tendency to orient domestic structures with the cardinal directions. As the histogram for domestic structure orientation shows, there is a trimodal distribution. Many structures are oriented near 90 and 180 degrees, as in previous periods, but there is also a peak near 45 degrees (Figure 4.29). This may reflect an attempt to orient the corners of domestic structures, rather than the long axis, to the cardinal directions. Compared to the Late Mississippian period, there seems to be less of a conscious effort toward a southeastern orientation. In general, household clusters are more widely spaced in the eighteenth-century than in the Late Mississippian period. At Ravensford and Townsend (small Cherokee settlements with no townhouse) clearly-defined household clusters with a winter and summer structure are spaced over 100 m from one another in a linear arrangement roughly parallel to a river. At Chota, a larger town with a townhouse, pavilion, and plaza, similar house pairs are more closely spaced, but are still separated by at least ten meters, and most are spaced more distantly, farther apart than the household clusters in most Late Mississippian settlements with several identifiable household clusters. We know from historic accounts that Cherokee communities were socially and politically integrated over a broad geographic area through matrilineal clans and political

242

alliances (Hudson 1976; Marcoux 2008). Viewed together, the site-level variation in structure orientation, the redundancy of domestic architecture, and the wide spacing of household clusters suggests a symbolic emphasis on household autonomy and identity. As Hally (2002) notes in his comparison of sixteenth- and eighteenth-century native housing traditions, historic period winter houses seem to lack semi-subterranean basins by the second half of the eighteenth century, and they are round rather than square or eight-sided. The basic floor plan of the winter house changes little. There is still a prepared clay central hearth and four interior support posts, and in the better preserved houses, partition walls radiating out from the four central support posts can be identified. Figure 7.5 provides a fairly typical example of a late eighteenth-century Cherokee winter house from the Townsend site. Note that the four interior support posts and central hearth are easily identified, but partition walls are less clear. The interior floor plan of these winter houses may have been a reference to the cosmological quartering of the earth. House size is similar across sites. The diameter of winter houses at Chota-Tanase, Mialoquo, Ravensford, Toqua, Townsend, and Tuckasegee falls between 5.8 to 8.2 m. While the four central posts in Cherokee winter houses probably had symbolic cosmological associations, there is less evidence for numerical symbolism in the wall posts. Compared to the Late Mississippian winter houses, there is more variability in wall post counts. Out of all the Cherokee winter houses in the database, only four houses from Townsend, three from Chota-Tanase, and two from Mialoquo fall close having seven or eight posts in each walls. Wall post spacing is sometimes more irregular than in many Late Mississippian structures. Rectangular summer houses were more lightly insulated than the winter houses, and are sometimes divided into two or three sections. The number and spacing of posts in these

243

structures varies widely. In some cases (e.g. Ravensford) the post patterns of the rectangular structures are very well defined, whereas at other sites (e.g. Mialoquo and Hiwassee Old Town) the post spacing is highly irregular. Differential preservation no doubt plays a part in this phenomenon, but there is no clear evidence for a symbolically important number of posts being set aside for each house. In contrast to the variation in wall construction, the regional similarity in the basic household plan in Cherokee settlements is remarkable. This shared plan, the round winter house and rectangular summer house, may reflect aspects of a belief system in which the ritual separation of binary opposites was important (Hudson 1976:317). The domestic household plan may also be a symbolic reference to both Late Mississippian and historic period public architecture. At Coweeta Creek (Rodning 2004) and perhaps at King (Hally 2008), a rectangular ramada was placed next to the townhouse. There is a similar ramada adjacent to the townhouse at Chota-Tanase. Just as Sullivan (1987, 1995) argues for the Mouse Creek phase in eastern Tennessee, household architecture may have symbolized and reinforced the same cosmologies and principles of social order symbolized at the community level by townhouses and ramadas. The similarity in domestic structures across a broad geographic area may be the result of refugee groups forming new communities in Cherokee territories, following Kowalewski‘s (2006) model for coalescent societies. Sharing architectural patterns at the household level, patterns that possibly refer to older, Late Mississippian public architecture, may have been an important mechanism for integrating newly formed communities. Based on the domestic architecture at the Townsend sites, Marcoux (2008) argues that the variability in wall post spacing and the reduction in rebuilding and architectural investment in Cherokee houses (compared to the Late Mississippian period) was a conscious adaptation to the

244

chaotic and violent social landscape of the mid-eighteenth century. Less energy was invested in house construction because occupational duration was uncertain, and the quick abandonment of sites was a crucial survival strategy. I agree with this interpretation, and argue that it extends beyond Townsend, to all the Cherokee sites in the database. Only at Ravensford, which dates to the first half of the eighteenth century, is there much evidence for rebuilding and architectural investment in domestic structures. On the other hand, historic period council houses or townhouses have an enormous number of posts and represent a substantial labor investment. Eighteenth-century Cherokee townhouses are just as well built as the townhouses of the Late Mississippian period, and in some cases they are larger. Despite the uncertain and dangerous social, political, and biological context of the late eighteenth century, the heavily built townhouse survived. This provides further evidence for coalescence and the integration of newcomers during this period. The townhouse may have been an even more important locus of social integration than in earlier times, absorbing labor that had previously been invested in domestic structures. There is strong historical evidence for importance for gendered divisions of space in Historic Indian communities. For example, in Creek society, women and children were sometimes excluded from entering the townhouse, while men could enter freely (Hudson 1976:219). Historic accounts also indicate that Creek women lived apart from men, sat on different mats, and ate with different dishes during menstruation (Bell 1990). In Cherokee society women had more political power and influence, but there were still strong norms of gender separation (Braund 1993:15; Hudson 1976). We would expect for there to be division of male and female space within domestic structures such as those posited by Polhemus (1987; 1998) and Gougeon (2002) in Late Mississippian contexts. Ongoing analyses of the floor debris

245

distribution in the Cherokee winter houses from Ravensford may reveal gender-specific activity areas (Tasha Benyshek personal communication 2010). Discussion and Conclusions In this chapter I conceptualize houses as icons representing cultural beliefs rooted in mythologies, cosmologies, and other religious and symbolic systems. Based on archaeological and ethnographic studies from beyond the Southeast, I argue that seemingly undifferentiated structures are often conceptually divided by residents. Further, I suggest that variation in house size and shape, numbers of wall posts, interior partitioning within houses, and the orientation and layout of houses in communities can in some cases be understood as references to well-known aspects of historically recorded Southeastern Indian religious and cosmological concepts. Several key trends emerge when we think of houses as symbols and consider diachronic changes in architecture. The reduction in domestic structure size from the Middle Woodland to the Late Woodland period and the shift from open floor plans to increasingly segmented buildings may signal a movement of some household-level ritual out of domestic structures and into plazas and public spaces. This an admittedly speculative hypothesis that needs further testing, but it offers an alternative to simply thinking of Middle Woodland domestic structures and habitation sites as devoid of architectural symbolism and ritual. From the Late Woodland to the Early Mississippian period domestic structures become more similar across broad geographic areas, in concert with the rise of the expansive early Mississippian polities such as Cahokia, Etowah, and Moundville (see Cobb and King 2005). In the western part of the study area there is some tendency for houses to be aligned to sitewide layouts, as though signaling membership with the broader community was more important than signaling membership to the household cluster. This changes in the Late Mississippian period,

246

when there is increasing archaeological evidence that households became more autonomous, and architectural symbolism was directed toward expressing household identity. This trend becomes even more pronounced in the Historic Indian period. There is clear evidence from the Early Mississippian period onward that key aspects of Southeastern Indian cosmology are reflected in domestic and public architecture. Specifically, references to the four cardinal directions, the seven layers of the sky dome, the above world and underworld, and the separation of opposite categories may all be expressed in Mississippian domestic and non-domestic structures. Some of these concepts may also help explain architectural patterns in the Woodland period, such as possible early examples of paired buildings, house basins, and structures with four central support posts. In the Late Mississippian and Historic Indian periods, visual parallels between public and domestic structures suggest a similar set of underlying cosmological beliefs and rules of social order. In a landscape fraught with disease, violence, and uncertainty, Historic Indian domestic structures show less evidence of architectural investment, but highly similar floor plans that may invoke older, Mississippian notions of house building. This common template may have helped integrate refugees and migrants into new, diverse communities. There is strong historical evidence for a gendered division of space in Historic Indian communities and strong cross-cultural evidence to suggest that houses would have been divided into male and female activity areas as far back as the Middle Woodland period. This hypothesis has only been tested in a few locations. It would be productive to perform similar analyses in structures from the Early and Late Mississippian and Historic Indian periods. Understanding the impact of symbolic and ritual behavior on architecture is a difficult enterprise. There is always the possibility that features that seem to be heavily imbued with

247

cosmological references, such as the four central supports of Late Mississippian and Cherokee winter houses, could be more about structural integrity than underlying cosmological structures. However, cross-cultural studies of houses and symbolism and the historical literature from the Southeast provide strong evidence suggesting house building and community planning are guided by ritual, mythological, and cosmological beliefs.

248

Figure 7.1 Middle Woodland Circular Structures from the McFarland Site (Redrawn from Kline et al. 1982:27)

Figure 7.2 Rectangular Late Woodland Structure from 1PI61 (Redrawn from Jenkins et al. 1981) 249

Figure 7.3 Circular Late Woodland Structures from Brasstown Valley (Redrawn from Cable et al. 1997)

250

Figure 7.4 Late Mississippian Winter House Layout (Adapted from Gougeon 2002)

7.5 Historic Indian Winter House Layout (Redrawn from Marcoux 2008:281)

251

CHAPTER 8 HOUSES AND STATUS

Understanding the relationships between the material remains of houses and status is a major focus of household archaeology. Based on anthropological and archaeological studies of houses and household status discussed in Chapter 2, I hypothesized that differential household status would account for variation in house size and interior features. During the Middle Woodland, Late Woodland, and Historic Indian periods, when there is little evidence for status differentiation beyond ranking and achieved status positions, I anticipated that domestic structures in a defined settlement would be similar in terms of size and number of interior features. During the Mississippian period, archaeological and early historical evidence points to at least a two-part division of society into chiefs and commoners. Chiefs often lived on top of or near mounds, while commoners lived in structures surrounding the mound and plaza. I hypothesized that Mississippian mound-top domestic structures would be larger and have more interior features and partitions than village-level domestic structures. In this chapter I test these hypotheses. During the Middle Woodland, Late Woodland, and Historic Indian periods there is little evidence to suggest that status differentiation can explain variation in the size and interior organization of domestic structures. In contrast, during the Mississippian period, differential household status may account for much of the variation in the size of domestic structures, and more broadly, helps explain intrasite variation in domestic structures. Houses and Status in the Middle and Late Woodland Periods 252

While there is a large body of research exploring the relationship between house form and status during the Mississippian period, few such studies have been applied to Middle Woodland-period houses. Based on evidence from earthworks, burials, artifacts, and iconography, Middle Woodland societies were not strictly egalitarian, but they were clearly not as stratified as Mississippian societies (Anderson and Mainfort 2002:10). Archaeological data and cross-cultural comparisons suggest that Middle Woodland period households may have been organized into clans or other descent groups with more or less equal status (Anderson and Mainfort 2002:10). The ceremonialism and long-distance exchange of the Middle Woodland period could have provided arenas for status competition and the development of social ranking. Smith (1986:48) suggests that charismatic individuals could have risen to leadership positions in the manner of Melanesian ―Big Men.‖ Mainfort (1988) uses the distribution of burials goods and other fancy artifacts from Pinson to argue that certain clans or lineages had more wealth than others, and that these ranked descent groups had differential power over long-distance exchange and public ceremonies. Based on the distribution of faunal and botanical remains at Kolomoki, Pluckhahn et al. (2006) claim that households used small-scale feasting for status competition. However, none of the archaeological data from Kolomoki, Pinson, or other well-documented Middle Woodland sites point to the kind of hierarchical status divisions apparent in Mississippian societies. If we assume that there was relatively little status differentiation in Middle Woodland societies, then we might expect houses within communities to be similar in size, appearance, and construction. Given the regional variation in artifacts not associated with the ceremonialism and exchange of the Hopewell Interaction Sphere (see Caldwell 1964), we might also expect there to be geographic variation in domestic structures.

253

To begin, I compared the size of domestic structures from 16 Middle Woodland components from 16 different sites in the study area. Figure 4.2 shows a box plot for the area of domestic structures at each component. In general, domestic structure size at each component is fairly similar. The coefficient of variation in floor area ranges from only 0.09 at McFarland to 1.31 at Banks V. At McFarland, Brasstown Valley, Duncan Tract, and Ela, the houses within each settlement are nearly identical in size. In each case the size distribution is normal and fairly tightly clustered around the mean. The coefficient of variation for domestic structure area at these four sites is less than 0.25. At Banks III and Hickory Log, domestic structures are also similar in floor area, although at Hickory Log the distribution is skewed toward smaller houses. Only at Banks V and Yearwood are a few structures remarkably larger than others. At Banks V, three circular to oval-shaped domestic structures are very close in size, ranging from 15.9 to 19.6 m2, while one large structure with four interior supports and two hearths measures 146.6 m2. It seems unlikely that this pattern represents an elite versus commoner dichotomy in housing. Rather, based on the features and ecofacts found in the large structure, the difference in structure types appears to be seasonal. Faulkner (2002:196-198) argues that the large structures with hearths are winter dwellings while the other structures are summer dwellings. There is a similar pattern of possible winter and summer dwellings at Banks III. Butler (1977) argues that the Owl Hollow phase structures at Yearwood are not domestic, but were instead related to gatherings for exchange and ceremony, based on the presence of cremation burials and exotic materials including copper, mica, galena, Flint Ridge chert, quartz crystals, and rocker-stamped ceramics. The unusual floor patterns of the structures at Yearwood do not resemble other domestic structures from the Middle Woodland period.

254

With the exception of Yearwood, the Middle Woodland domestic structures tend to have similar floor plans at the level of a single settlement. At Hickory Log, McFarland, and Brasstown Valley, domestic structures have round-to-oval shaped post patterns with regularly spaced single posts. The house patterns at Ela are poorly defined, but they also appear to be circular, and are quite similar in size. At Banks III and Banks V, there is little size difference when the winter double earth-oven structures are considered separately from single post summer houses. At Duncan Tract, the domestic structures are large compared to other Middle Woodland sites, but they are uniform in size at the single-site scale. Comparing the interiors of Middle Woodland domestic structures provides little evidence for status differences. Most houses have very few interior features, and the distribution of features is fairly even (Figure 4.31). At McFarland one domestic structure has far more interior pits than the other four, but it is not certain that all the pits were contemporaneous with the house. The similarly low number of features in each domestic structure does not suggest that any one household in a community had access to special activities or goods. With the exception of the Owl Hollow phase houses of middle Tennessee, with their large hearths and interior roof supports, Middle Woodland structures have open floor plans. The interior space is sometimes divided by hearths, and was almost certainly conceptually divided by the occupants, but there are not clearly defined rooms. Only one structure, the largest building at Duncan Tract, has a clear interior partition. This is another indicator of architectural similarity, but perhaps more significantly it points to a lack of areas of restricted access in dwellings. Based on cross-cultural data compiled by Kent (1990), open floor plans are generally associated with egalitarian societies with frequent sharing between households.

255

In recent years our perspective on social organization and complexity during Late Woodland period has changed. Once conceptualized as a dark age populated by ―good gray cultures‖ (S. Williams 1967:297) between the more archaeologically impressive Hopewell and Mississippian periods, the Late Woodland period is now seen as a time of important cultural and demographic changes that set the stage for the rise of Mississippian chiefdoms (Anderson and Mainfort 2002:15-19; Cobb and Nassaney 2002; Muller 1997:117-140). Given the lack of fancy exchange and display goods associated with many Late Woodland sites, it is generally thought that there was little status differentiation in Late Woodland societies (Kidder 2002; Muller 1997:135), with households organized into egalitarian clans and lineages that may or may not have been ranked (Anderson and Mainfort 2002:17). On the other hand, the first formal civicceremonial complexes in the Southeast (sites that were political centers as well as ceremonial ones) emerge during these years at places like Toltec in Arkansas and at Coles Creek sites in the Lower Mississippi Valley (Anderson and Mainfort 2002:16), suggesting the development of a more hierarchical social order (see Rolingson 2002). However, at the Late Woodland sites in the study area, there is less evidence from artifacts, burials, or mound to suggest this sort of emerging social inequality. As a result, we might expect houses to be relatively similar within communities. As in Chapter 4, I consider the Late Woodland components from the western and eastern halves of the study area separately. In the American Bottom, southeastern Missouri, and west-central Alabama, Late Woodland domestic structures are uniform in size. Figure 4.4 shows box plots for the floor area of Late Woodland domestic structures. At BBB Motor, Cahokia Tract 15A, Hoecake, and 1PI61, the size distributions are very tightly clustered around the mean, and the coefficient of variation of floor area for the components ranges from 0.19 to 0.42. Moreover, these domestic structures

256

are all similar in form, both from an inter- and intrasite perspective. Almost all of these structures are small, rectangular, single post houses set in shallow basins. Barring differences in perishable exterior decoration that cannot be detected archaeologically, these domestic structures would have looked nearly identical to viewers. These are not elaborate or individualistic-looking structures. Domestic structures at the western Late Woodland sites consistently have few interior features or partitions. At BBB Motor and 1PI61, most houses have two interior features: one hearth and one pit. At Hoecake and Cahokia there are more domestic structures with only one interior feature. There are no instances of a single domestic structure having far more interior features than surrounding houses. From a broad perspective, this suggests there was little difference in the sorts of activities that took place within these structures. Of all the western Late Woodland domestic structures, only four had interior partitions: two of the houses at 1PI61, one house at Cahokia Tract 15A, and one house at BBB Motor. These structures are not in special locations in their communities, and they do not appear to serve special functions. In the east there are fewer good examples of Late Woodland domestic structures. At Town Creek and Summerour, single structures were identified, and their function is unclear. The house at Summerour is a wall trench structure on top of a low mound. Ceramics at Summerour are predominantly Woodstock phase, and a radiocarbon date from the structure places the occupation squarely in the Late Woodland period, despite the diagnostic Early Mississippian architectural style (Pluckhahn 1996). The village at Summerour was never excavated, so it is impossible to compare the mound-top structure to village houses. At Town Creek, the lone possible Late Woodland structure is large non-domestic structure in the plaza.

257

At Brasstown Valley, possible Late Woodland domestic structures are round, single-post houses. The spread of house sizes is fairly tight, and the coefficient of variation for area, 0.27, falls within the range of the western sites. The three domestic structures at Rivermoore are small, square, single post houses. The coefficient of variation for their floor area is 0.23, similar to domestic structures from the western part of the study area. Unfortunately, there are few good data on the interior features in the eastern Late Woodland domestic structures. At Rivermoore, the domestic structures each had at least one hearth and one interior pit. One domestic structure had three pit features in addition to the hearth, but this house does not seem to be located in a special place or have a non-domestic function. The single structures at Town Creek and Summerour have a relatively high number of interior features for Woodland period structures, but we lack other contemporaneous structures at those sites for comparison. Much like the structures in the Middle Woodland period, Late Woodland structures exhibit regional diversity, but from an intrasite perspective, do not seem to be signaling major status differences between the residents. The homogeneity of structure size and form at 1PI61, BBB Motor, Cahokia 15A, and Hoecake is remarkable. These houses seem practically interchangeable, not only at a community, but also at a regional scale. At Rivermoore and Brasstown Valley, the Late Woodland structures are also quite similar. Given its mound summit placement and relatively complicated interior, it seems highly likely that the structure at the Summerour Mound is a high status residence, but this is speculative. House Size and Status During the Mississippian Period In a review of Mississippian research, Schroeder (2004) identified three different positions with regard to the interpretation of Mississippian social and political organization:

258

―minimalist,‖ ―maximalist,‖ and ―idealist‖ (see also Cobb 2003; Muller 1997; Hally 2009; Hammerstedt 2005). The minimalist position tends to downplay the degree of status differentiation in Mississippian society (e.g. Muller 1997), while the maximalist position argues that Mississippian society was significantly more hierarchical and stratified than eighteenthcentury Indian societies (e.g. O‘Brien 1972, 1989, 1991). The idealist position focuses on ideological manifestations of status differentiation and social inequality, such as the development of cultural hegemony through the use of widely used symbols on quotidian material culture, such as ceramic serving ware (e.g. Emerson 1997; Pauketat 2004; Pauketat and Emerson 1991). These different positions produce different expectations for the relationship between social status and houses. A ―maximalist‖ might expect that houses on top of mounds would be significantly larger than village houses, and that they would have more interior facilities, following Hirth‘s (1993) argument that elite residences should have more evidence for special ritual and political activities. These differences should be especially pronounced at major political centers, where presumably, paramount chiefs live. A ―minimalist‖ might expect the differences between mound-top and village houses to be less dramatic in terms of size and interior facilities. An ―idealist‖ might expect to find symbolically charged differences and similarities between mound and village architecture, perhaps in terms of size or appearance (interpretable via floor plans), that communicated both differentiation and community solidarity between elite and commoners (see Gougeon 2006; Wilson 2008). While ethnohistoric evidence from the sixteenth century clearly points to a correlation between high status and mound-top residence, there is debate among Mississippianists over the degree of status differentiation between households living on mounds and off mounds. Like Hammerstedt (2005), I compare the size of structures on top of mounds to structures in villages

259

using a large sample. I also compare additional architectural characteristics, including post size and spacing, and the number of interior partitions, features, and burials. To refine the parameters of this analysis I compare structures by class, i.e. domestic versus non-domestic structures. This allows for a more meaningful comparison of primary residences, and avoids comparing large public structures in villages to mound-top residences. The results of this comparison are shown in Tables 8.1, and graphically in Figures 8.1 and 8.2. To determine if the differences in the architectural features are significant, I use nonparametric statistics. For example, when comparing the average area of different groups of structures, I use a Wilcoxon rank-sum test. This test is appropriate because the distributions of structure size are sometimes non-normal and their means have high standard deviations, but they are similarly shaped (i.e., positively skewed). The results of these tests are shown in Table 8.2. The sample includes 584 structures from 27 components at 14 sites: Banks V, Bessemer, Cahokia, Dallas, Dyar, Hiwassee Island, Jewell, Kincaid, Little Egypt, Loy, Martin Farm, Moundville, Toqua, and Town Creek. There are far fewer mound-summit domestic structures than village-level domestic structures, and in general, the mound-summit domestic structures are larger. For the Early, Middle, and Late Mississippian period, the median size of domestic structures on mounds is greater than domestic structures in villages, and Wilcoxon rank-sum tests indicate these differences are statistically significant (Tables 8.1 and 8.2). The structures I have identified as non-domestic in both mound and village settings are structures that seem to have served as public buildings, mortuary structures, and other specialpurpose structures. Non-domestic structures in villages, not on mound summits, are some of the largest buildings identified in this analysis. These large structures include the circular mortuary buildings from the Late Town Creek/Leak phase at Town Creek, large Stirling-phase buildings

260

from Cahokia Tract 15A, and the large off-mound structures at Bessemer. Judging by their floor plans and associated features, these large structures do not appear to be elite residences, but rather structures for public gatherings or special rituals. Access to these buildings may or may not have been restricted. Visually, these large off-mound buildings may have been just as impressive as large mound-top structures. The differences in area between non-domestic structures located on and off mounds are not statistically significant (Table 8.2). It is noteworthy that domestic structures on mounds are often more than twice the size of village domestic structures. If we only considered size, this would suggest that mound summit architecture reflected substantial differences in household size, household status, special functions, or some combination of all these factors. This would support the ―maximalist‖ assertion that there were major status differences between chiefs and commoners. However, examining variation in other architectural traits complicates this relationship. While the size of mound and village structures differs significantly, there are similarities in construction. In general, mound-top domestic structures have the same shape and wall type as domestic structures in villages. Gougeon (2006) notes this trend at Little Egypt, and it is also the case at Bessemer, Dallas, Dyar, Hiwassee Island, Jewell, Kincaid, Martin Farm, and Toqua. In these cases, houses that are presumably chiefly residences appear to be ―scaled up‖ versions of ordinary domestic structures. These houses may have been decorated on the outside with pigments or other perishable forms of decoration, but in terms of their basic shape and method of construction, they would have looked much like village-level houses. Comparing the number, size, and spacing of posts reveals additional similarities between mound and village structures. The number of posts in all structure types is closely correlated with structure size, so the largest structures have the most posts. Wall post size and spacing does not

261

differ greatly when mound and village structures are compared on a component-by-component basis. Thus, while more materials and labor were needed for structures on top of mounds and for large non-domestic structures in villages, the basic construction methods were the same. Throughout the Mississippian period, the number of interior partition walls in domestic structures on mounds and in villages does not differ significantly (Tables 8.1 and 8.2). Early Mississippian non-domestic structures in villages have significantly fewer partition walls than mound-top non-domestic structures, while the opposite is true during the Late Mississippian period, when mound-top non-domestic structures have more interior partitions. The higher frequency of interior partitions in Late Mississippian non-domestic moundsummit structures may reflect special social and/or economic activity areas that were not present in village-level houses. This would follow Hirth‘s (1993) expectations for architecture associated with special activities restricted by elite, and would fit with ethnohistoric accounts that described special rooms for holding display goods in chiefly, mound summit houses (Clayton et al. 1993). The number and distribution of interior features and burials in mound and village structures sheds light on possible status differences. In both mound and village contexts, most structures have only a few interior features and burials, but there are a few high outliers. In the cases of both domestic and non-domestic structures, mound-top buildings generally have more interior features, but only during the Late Mississippian period do domestic structures on mounds have significantly more interior features than domestic structures in villages. During the Early and Middle Mississippian periods, non-domestic structures on mounds have significantly more interior features than village-level structures (Tables 8.1 and 8.2). Domestic mound-summit structures in this sample contain the same sorts of interior features as village buildings: hearths and pits for storage, refuse, or processing. However, they

262

also have a small set of features that most village structures lack, such as raised clay seats and platforms. Following Hirth‘s (1993) criteria for detecting status, this could be an indicator of elite control over special political or ceremonial activities. However, it also appears that presumably elite households carried out many of the same mundane, ―village-level‖ tasks represented by hearths and pits. As others have argued with finer-scale data such as ratios of ceramic cooking ware to serving ware (Boudreuax 2005) and food remains (Gougeon 2006), this suggests that Mississippian chiefs and their families had special status in their communities, but were not exempt from quotidian subsistence tasks. Burials are indicators of social status, but the presence or absence of subfloor burials seems to be more related to changing burial practices over time than the social status of a particular household. When mound and village structures are compared, there is a not a statistically significant difference in the frequency of interior burials. As the broad-scale diachronic analysis in Chapter 4 indicates, interior burials are rare during the Early Mississippian and then quite common by the Late Mississippian period. Comparing mound and village architecture at the single-site scale generally confirms these trends, but also points to noteworthy exceptions. Figures 8.3 through 8.13 show box plots for mound and village structure area at Bessemer, Dallas, Dyar, Hiwassee Island, Jewell, Kincaid, Little Egypt, Toqua, and Town Creek. At all the sites except for Bessemer, mound summit structures have a larger average size than village structures. Welch (1994) argues that most of the buildings at Bessemer are non-domestic, which could explain the large village-level buildings. At Cahokia, Dallas, Dyar, Kincaid, Little Egypt, Martin Farm, and Moundville, there is a clear difference in the size of mound and village structures. In some cases the size distributions for the two samples do not overlap at all. In contrast, at Hiwassee Island, Jewell,

263

Toqua, and at Town Creek in the Early and Late Town Creek phases, mound structures are larger, but there is quite a bit of overlap in the size of mound and village buildings. At these sites, some village structures are larger than mound structures. In some cases this is due to functional differences in buildings. The large circular structures at the late occupation at Town Creek seem to have served as mortuary buildings rather than residences, and at Hiwassee Island phase Toqua, some of the smaller buildings on the mound summit are lightly built rectangular ―porches‖ near chiefs‘ houses. However, there are also cases of overlap in the size of domestic buildings on and off mounds at Jewell, Hiwassee Island and Dallas phase Toqua. Following this line of inquiry, examining the size distributions of village structures is revealing. For this analysis, I examine only the domestic structures located off mounds. In general, these domestic structures all have the same construction style and basic floor plan. At most of the 14 Mississippian sites considered here, the range of floor area in domestic structures in villages is fairly small. The coefficient of variation for floor area falls between 0.10 at Little Egypt (where there are only two structures) and 1.03 at Early Kincaid, where the sample is biased toward large buildings. For the remaining sites, the coefficient of variation falls between 0.30 and 0.46. This figure is more meaningful when compared to values for the Woodland and Historic Indian period. The coefficient of variation for the floor area of domestic structures at the Late Woodland components ranges from 0.19 to 0.43, with 75 percent of components falling at or below 0.30. For the Historic Indian period, the figure ranges from 0.08 to 0.40, with 72 percent of components falling at or below 0.30. While Mississippian domestic structures in the shadow of mounds are generally similar, from a long-term diachronic perspective, they exhibit more variation in size than domestic

264

structures during other time periods. Given the other lines of evidence for status differentiation in the Mississippian period, this is not surprising, but it does point to possible status differences expressed within villages, and not simply along the lines of mound versus village occupation. This finding confirms research by others at Mississippian sites with mounds (Pauketat 1998; Wilson 2008) and without (Price and Griffin 1979; Hally 2008) pointing to evidence of subtle status differences between household clusters in Mississippian villages. If mound summit residence is an indicator of higher social status, then status differentiation seems to explain much of the variation in house size and some of the variation in interior features and partitions at the single-site scale during the Mississippian period. Domestic and non-domestic structures on mounds are significantly larger than their counterparts in villages. In some cases mound summit domestic structures tend to have more interior features and partitions, but not always at a significantly higher rate. This suggests that households on the summits of mounds were larger, and had regular access to facilities for special social and economic activities, such as raised clay seats and platforms, and additional rooms which may have held symbolically charged display goods. On the other hand, mound-summit structures were made with the same materials and construction techniques as houses in villages. From a broader archaeological perspective, this is not a trivial similarity. In much of Mesomerica, elite houses are made with finely cut stone while the houses of commoners are made of wattle and daub or coarse stone (Gonlin 1991; Hirth 1993; Smith 1987). These structures would look radically different from the outside, immediately communicating differential status to viewers. Moreover, the labor inputs for constructing houses from finely cut stone and wattle and daub are remarkably different, resulting in a much greater difference in the labor value of elite and commoner housing. There may have been important,

265

archaeologically undetectable status markers on mound-summit structures, such as perishable, high cost decoration (perhaps especially fine wood carving or ornamentation with plant materials), but from a construction standpoint, the differences between presumably elite and commoner houses are not as pronounced. The findings from these comparisons confirm certain aspects of ―maximalist,‖ ―minimalist,‖ and ―idealist‖ views of Mississippian status differentiation. If, as a ―maximalist‖ would argue, Mississippian society was highly stratified and households on mounds had significantly higher status positions than households in villages, then mound summit houses should be larger, fancier, and contain evidence of restricted social and economic activities. On the other hand, in support of the ―minimalist‖ view of Mississippian society, houses on mounds and in villages were built with the same materials and in the same style, and while differences in the number of features and interior partitions are statistically significant, they are not great. From an ―idealist‖ viewpoint, the similarities in mound and village structures may represent an attempt by chiefs or commoners obscure or play down some of the status differences within their villages. While chiefs had larger houses, possibly reflecting larger household size, and may have performed certain special social and economic functions (represented by the unique features in their houses), minimizing visible differences in the exterior of their houses may have been a strategy to obscure growing social inequality in communities. Houses and Status in the Historic Indian Period Ethnohistoric records indicate that social and political organization during the eighteenth century was less hierarchical and stratified than in the Mississippian period. Among the Cherokee, councils made important political decisions, and leadership does not appear to have been centralized or inherited (Gearing 1942; Gilbert 1943). In both Creek and Cherokee society,

266

status was achieved rather than ascribed (Hudson 1976). In this more egalitarian context, we would expect domestic structures to look similar within communities. Compared to the Mississippian period, there is much less variation in the size and interior layout of historic period native houses. Focusing on the size distribution of historic period domestic structures reveals similarities across household clusters at individual sites (see Figure 4.20). Among the Historic Indian components with multiple domestic structures, Brasstown Valley, Chota-Tanase, Ela, Hiwassee Old Town, Jenrette, Lower Saratown, Mialoquo, Ravensford, Toqua, and Townsend, the size distribution of domestic structures is tightly clustered around the mean and close to normal. The coefficient of variation for the average size of domestic structures in settlements ranges from 0.08 at Alarka Farmstead to only 0.40 at Jenrette. At the Cherokee components in particular—Alarka Farmstead, Brasstown Valley, Mialoquo, Ravensford, Toqua, and Townsend—household clusters look remarkably similar. The well-preserved early eighteenth century house pairs at Ravensford provide an especially good visual representation of this uniformity (Figure 8.14). There is minor variation in the arrangement of winter and summer structures, but in general these household clusters are more uniform than those from well understood Late Mississippian sites like King and Toqua, where there is more variation in the size and number of structures in household clusters. Historic Indian domestic structures are not only similar within sites, they are similar across sites. This pattern is especially clear when Historic Indian winter houses are compared to Late Mississippian winter houses. When all the historic period winter houses are grouped together, they are normally distributed and have an average size of 35.1 m2 (s.d. 8.5 m2) and a coefficient of variation of 0.24 (n=43). Sizes range from 18.6 to 70.9 m2. The Late Mississippian winter houses have an average size of 47.4 (s.d. 19.1 m2, n=160). The coefficient of variation,

267

0.40 is nearly twice that of the historic sample, and the range is wider, with a minimum of 15.6 and a maximum of 126.3 m2. The size distribution of historic period domestic structures is tightly clustered around the mean with a single high outlier. In contrast, in Late Mississippian times the size distribution is positively skewed, with many houses falling close to or slightly below the mean (Figure 4.13). These size distributions are what we would expect when comparing houses from a more egalitarian society to a more stratified one. To the extent that houses serve as a visual signal of household status, it seems likely that Historic Indian structures would have emphasized similarities between households while Late Mississippian structures may have emphasized status differences. Interior spaces in Historic Indian domestic structures are also uniform across sites. Most Historic Indian structures have three or fewer interior features, usually a hearth and one or two interior pits (Table 4.11). Single high outliers at Mialoquo, Toqua, and Chota-Tanase represent large town houses, which not surprisingly have more interior features related to their special functions as public buildings. Ela has an unusually high number of interior features (approximately six per house), but the distribution is normal, and Wetmore (1991) states that feature assignment at Ela was difficult due to superimposed Woodland and Cherokee occupations. Generally speaking, there is little evidence from the Historic Indian components indicating that any one household in a settlement had an unusually high or low number of interior features, which suggests uniformity in the kinds of production and consumption activities that took place within these households. Discussion and Conclusions I rely on decades of archaeological and ethnohistoric research on status differentiation to make several guiding assumptions in this chapter: first, status relations during the Middle

268

Woodland and Late Woodland period were primarily egalitarian, with some possible status competition; second, Mississippian societies were stratified, at least into elite and commoner groups, but there was regional and chronological variation in the degree of status differentiation; and third, Historic Indian societies were fairly egalitarian, with achieved rather than ascribed status. Comparing houses at multiple spatial and temporal scales suggests that status differentiation may account for differences in house size and variation in interior spaces. On the other hand, status differences seem to have had less of an effect on outward appearance and basic architectural form. Middle and Late Woodland domestic structures generally conform to expectations for egalitarian societies. Domestic structures are similar within sites in terms of size, style, and interior features, and would have communicated minimal status differences between household clusters. The similarity in Late Woodland structures in the western part of the study area is especially impressive, not just at the level of a single site, but at a broader regional scale. The differences in houses in mound and village contexts at Mississippian sites fit with expectations for societies with some degree of social stratification (Hirth 1993). Mound summit domestic structures are generally larger than village structures, and have special features that village houses lack. However, as others (Gougeon 2006; Hammerstedt 2005) have pointed out, differences in house form do not seem to be adequately explained by an overly-simple ―elite on mound versus commoner in village‖ model of social differentiation. In terms of overall appearance and interior features, domestic structures on mounds and in villages are often similar, and in a few isolated cases, domestic structures in villages rival mound summit buildings in terms of size. Compared to the Woodland and Historic Indian periods, there is more variation in the size of domestic structures in ordinary village contexts

269

during the Mississippian period. This may be the result of status differences within ―commoner‖ domestic contexts that we typically think of as homogenous in our models of Mississippian social organization and political complexity. The similarities between mound and village domestic structures raise important questions. Did the physical similarities between mound and village domestic structures help ameliorate tensions between mound summit and village households in an atmosphere of social inequality, as Gougeon (2006) and other ―idealists‖ have suggested, or do they represent a limit on the amount of labor chiefs could muster for house construction? Domestic structures in the Historic Indian period are remarkably similar in form. Cherokee household clusters are uniform both within and across sites. This uniformity fits with our understanding of house from in egalitarian societies, but it may also be related to the historical forces of the eighteenth century in the Southern Appalachians. The similarity in eighteenth century Cherokee domestic structures may have been a strategy to integrate newcomers in a social landscape of disease, warfare, and frequent movement, following Kowalewski‘s (2006) model for coalescent societies. Marcoux (2008) argues that this accounts for some of the similarities in structures and subtle differences in ceramics at the Townsend sites, and I would argue that this is the case across much of the Southern Appalachians during the eighteenth century. More broadly, this analysis shows the value of using a comparative, diachronic, and multiscalar approach to understanding the relationship between houses and status. Comparing the differences in house size at Woodland, and Historic Indian components provides a new basis of comparison for heavily studied Mississippian status differences. For the Woodland period, there are few published comparisons of architectural data, and while this analysis generally confirms

270

the notion that these societies were roughly egalitarian and lived in similar structures, it points to noteworthy similarities and differences in architecture at a broad spatial scale that have received little attention.

271

Table 8.1 Summary Statistics for Mississippian Mound and Village Structures Early Mississippian Period Domestic, Village Variable n Min Area (m2) 257 3 Int. Features 154 0 Int. Burials 127 0 Partitions 112 0

Max 106.5 8 4 2

Median 20.7 1 0 0

Non-domestic, Village Variable n Min Area (m2) 35 1.17 Int. Features 28 0 Int. Burials 29 0 Partitions 23 0

Max 228.3 7 1 2

Median 76.9 0 0 0

Domestic, Mound Variable n Area (m2) 11 Int. Features 7 Int. Burials 7 Partitions 4

Max 117.5 2 0 0

Median 27 0 0 0

Non-domestic, Mound Variable n Min Area (m2) 47 29.9 Int. Features 20 1 Int. Burials 12 0 Partitions 13 0

Max 299.8 6 0 2

Median 89.9 2 0 1

Middle Mississippian Period Domestic, Village Variable n Min Max Area (m2) 113 1.9 85.9 Int. Features 13 0 2 Int. Burials 14 0 13 Partitions 6 0 4

Median 22.3 0 0 1

Non-domestic, Village Variable n Min Area (m2) 12 41.5 Int. Features 6 0 Int. Burials 8 0 Partitions 2 2

Max 280.5 4 40 3

Median 139.2 0 4.5 2.5

Domestic, Mound Variable n Area (m2) Int. Features Int. Burials Partitions

Min 18.5 0 0 0

Min 48.7 0 6 .

Max 65.7 1 10 .

Median 57.2 0.5 8 .

Non-domestic, Mound Variable n Min Area (m2) 8 49.1 Int. Features 10 0 Int. Burials 10 0 Partitions 2 1

Max 195.1 2 24 1

Median 86.9 1 1.5 1

Late Mississippian Period Domestic, Village Variable n Min Area (m2) 194 6.6 Int. Features 133 0 Int. Burials 138 0 Partitions 59 0

Max 113.8 30 18 8

Median 39.4 1 0 0

Non-domestic, Village Variable n Min Area (m2) 6 37.2 Int. Features 5 1 Int. Burials 4 0 Partitions 4 0

Max 226.1 5 35 11

Median 121.45 2 5 3.5

Domestic, Mound Variable n Area (m2) 9 Int. Features 11 Int. Burials 11 Partitions 10

Max 126.3 6 3 7

Median 63.4 1 0 2

Non-domestic, Mound Variable n Min Area (m2) 18 27 Int. Features 22 0 Int. Burials 22 0 Partitions 22 0

Max 134.6 9 18 12

Median 73.5 2 0 0

2 2 2 0

Min 37.2 0 0 0

272

Table 8.2 Results of Wilcoxon Rank-Sum Tests for Differences between Mound and Village Structure Variables Early Mississippian Period Domestic, mound vs. village Non-domestic, mound vs. village

Area z = -2.471, p = 0.014 z = -1.453, p = 0.146

Interior features z = 1.137, p = 0.255 z = -4.560, p = 0.000

Interior burials z = 0.768, p = 0.442 z = 0.643, p = 0.520

Partitions z = -0.588, p = 0.557 z = -2.099, p = 0.036

Middle Mississippian Period Domestic, mound vs. village Non-domestic, mound vs. village

Area z = -2.183, p = 0.029 z = 1.119, p = 0.263

Interior features z = -0.658, p = 0.517 z = -1.996, p = 0.046

Interior burials z = -1.751, p = 0.079 z = 1.076, p = 0.282

Partitions .. ..

Late Mississippian Period Domestic, mound vs. village Non-domestic, mound vs. village

Area z = -3.091, p = 0.002 z = 1.167, p = 0.243

Interior features z = -2.295, p = 0.022 z = 0.649, p = 0.516

Interior burials z = 0.858, p = .3911 z = 1.456, p = 0.145

Partitions z = -0.835, p = .4036 z = 2.366, p = 0.018

273

Figure 8.1 Box Plots of Area for Mound and Village Structures

Figure 8.2 Bar Chart of Average and Median Area for Mound and Village Structures

274

Figure 8.3 Box Plots of Area for Mound and Village Structures at Bessemer

Figure 8.4 Box Plots of Area for Mound and Village Structures at Dallas

275

Figure 8.5 Box Plots of Area for Mound and Village Structures at Dyar Mound

Figure 8.6: Box Plots of Area for Mound and Village Structures at Hiwassee Island, Hiwassee Island Phase

276

Figure 8.7: Box Plots of Area for Mound and Village Structures at Jewell

Figure 8.8 Box Plots of Area for Mound and Village Structures at Kincaid, Middle Kincaid Phase

277

Figure 8.9 Box Plots of Area for Mound and Village Structures at Little Egypt

Figure 8.10 Box Plots of Area for Mound and Village Structures at Toqua, Hiwassee Island Phase

278

Figure 8.11 Box Plots of Area for Mound and Village Structures at Toqua, Dallas Phase

Figure 8.12 Box Plots of Area for Mound and Village Structures at Town Creek, Early Town Creek Phase

279

Figure 8.13 Box Plots of Area for Mound and Village Structures at Town Creek, Late Town Creek to Early Leak Phase

Figure 8.14 Eighteenth-Century Cherokee House Pairs at Ravensford

280

CHAPTER 9 HOUSES AND SETTLEMENT FUNCTION In this chapter I examine the effect of a site‘s place and settlement system function on domestic architecture. How do domestic structures at large, nucleated settlements differ from houses at small, dispersed settlements? Can measurable characteristics of sites that represent their function in a settlement system predict variation in architecture? Based on previous research discussed in Chapter 2, I developed three sets of hypotheses about the relationship between settlement function and architecture. First, I hypothesized that the average size of domestic structures would decline as the overall number of structures, site size, and number of mounds increased. Second, I hypothesized that there would be greater diversity in structure form at large, central sites than at small, dispersed settlements. Third, I hypothesized that domestic structures at sites with palisades might be smaller due to the impact of crowding. I test these hypotheses by comparing the architectural characteristics of domestic structures at different kinds of settlements and use multivariate regression to examine the effect of site size and function on domestic architecture. The first step in this analysis is to define the terms for different kinds of settlements. As Williams (1995:124-125) has indicated, there is some diversity in Mississippian settlement systems, and sometimes the terminology used by archaeologists to describe these settlements is confusing. Williams proposes a two-tiered model for Mississippian settlement, in which there are systems with either many dispersed farmsteads and small chiefly compounds, or systems with large towns and few farmsteads (1995:133). The dispersed settlement system of the Late 281

Mississippian Oconee River valley would be an example of the former, while the more nucleated settlement system of the Late Mississippian paramount chiefdom of Coosa would be an example of the latter (Williams 1995:127). Finally, he argues that ―higher levels or tiers of social and political complexity are likely evidenced by such sites as Moundville or Cahokia or Etowah, but these sites are the exceptions rather than the rule‖ (Williams 1995:133). Taking this model as a starting point, I propose that three criteria are especially useful for identifying differences between settlements: the areal size, the total number of roughly contemporaneous structures, and the number of mounds. All of these variables are problematic in a different way, but together they clearly separate large nucleated settlements and central places from small, dispersed settlements. Site size can be used to distinguish large, nucleated settlements from small, dispersed settlements. Site size is often an estimate, and is usually impacted by site destruction or a lack of complete survey and excavation. For this study I use an estimate of the total areal extent for each site or component provided by the original analyst. The total number of structures present at a site can also be used to distinguish large, nucleated settlements from small, dispersed settlements. Sites are rarely completely excavated, so the total number of structures found is often lower than the total number that may have been occupied at any given time. I do not attempt to estimate the total number of structures at partially excavated sites, but I take the scale of excavation into consideration in the discussion that follows. In the Southeast, settlements with mounds, whether in dispersed or nucleated settlement systems, are generally important central places. I use the term central place to refer to large settlements that provide the location for important social, political, and ceremonial services for

282

the surrounding area (Christaller 1966, Smith 1978:61). It seems likely that there may be significant differences between the architecture at settlements with many mounds and settlements with no mounds. During the Mississippian period, societies were organized into politically centralized chiefdoms, and sites with mound were the political and ceremonial centers of these polities (Anderson 1994; Blitz and Livingood 2004; Hally 2006). For the Historic Indian period, a large townhouse is a better indicator for the centrality of a site. Middle Woodland ceremonial centers may have been largely devoid of domestic structures, making the analysis here less pertinent. During the Late Woodland period, on the other hand, there are large population centers with multiple mounds (Rolingson and Mainfort 2002:35). Keeping these problems in mind, for this analysis I assume that sites with mounds are more important central places than sites with no mounds. Domestic Structure Size and Site Size Previous research on Mississippian structures discussed in Chapter 2 suggests that domestic structure size may vary with the size and function of a site in its settlement system. We might expect the average floor area of domestic structures to be greater at smaller, more dispersed sites and smaller at larger, more nucleated sites due to crowding during the course of settlement expansion. On average, sites with one or more mounds that served as political, social, and ceremonial centers may have smaller domestic structures than dispersed sites, where more space may have been available for house construction. To visualize the differences in the size of domestic structures at different kinds of settlements, I generated a box plot for domestic structure area at each component, and then sorted them in ascending order, based on site size, then by total number of structures, then by

283

number of mounds. By way of example, Figures 9.1 – 9.6 show the results when the components are sorted by total number of structures. For each time period there does not seem to be a strong linear relationship between domestic structure size and site size. During the Late Woodland and Historic Indian periods, domestic structure size is uniform across spatially large and small settlements, both in terms of median size and the degree of dispersion. During the Middle Woodland, Middle Mississippian, and Late Mississippian periods the size of domestic structures varies more from site to site, but does not clearly vary with the spatial size of the settlement. The Early Mississippian period presents a different pattern. Domestic structures tend to be small at the low and high ends of the continuum, i.e. at very large and very small sites. At Cahokia and Moundville, which have some of the most expansive Early Mississippian components in the database, domestic structures are small (with average sizes under 25m2), and distributions are clustered tightly around the mean. Domestic structures that fall in the middle of the site-size continuum are comparatively large, and their size distributions more dispersed. A different pattern emerges when the components are sorted in ascending order based on the number of structures (Figures 9.1 – 9.6). During the Middle and Late Woodland period, the overall number of structures seems to have little effect on domestic structure size. During the Early Mississippian period, there is some evidence that domestic structure size decreases as the number of structures increases. At sites with fewer than five domestic structures, house size is generally low. However, for sites with more than five domestic structures, there is something of a downward trend in house size as settlements become more densely occupied. The components with the most houses, the Emergent Mississippian and Lohmann phases at Cahokia, and the Early Moundville I and Late Moundville I-Early Moundville II phases at Moundville, have some

284

of the smallest Early Mississippian domestic structures in the sample. Variation in domestic structures size seems less pronounced at the extremes; the broadest ranges of domestic structure size occur in the middle of the continuum, with components that contain between 5 and 35 structures. The total number of structures seems to have less influence on domestic structure size during the Middle Mississippian, Late Mississippian, and Historic Indian periods. The relationship between mounds and domestic structure size seems to be different during the Early, Middle, and Late Mississippian periods. In Early Mississippian times, domestic structure size does not clearly increase or decrease as the number of mounds increases. The largest domestic structures appear at sites with two or more mounds, as Hammerstedt (2005) found in his research, but large numbers of small domestic structures occur at Moundville and Cahokia. Very small domestic structures appear on sites with either no mounds or many mounds. Many of the components with the largest average domestic structure size fall in the middle of the continuum, with one to three mounds. Likewise, the relationship between domestic structure size and the frequency of mounds during the Middle Mississippian period is unclear. In contrast, during the Late Mississippian period, there appears to be a slight increase in the average size of domestic structures as the frequency of mounds increases. At sites with one or more mounds, average domestic structure size is generally close to or above 45 m2, which is above average for the period. These observations provide an important complement to Hammerstedt‘s (2005) earlier comparison of mound and non-mound houses, which did not break sites down into finer-scale components to examine chronological differences between Early, Middle, and Late Mississippian occupations. A side-by-side comparison of the domestic structure size distributions for the Early Mississippian components suggests that variability in domestic

285

structure size is higher at sites with two or more mounds than at sites with one or no mounds. In contrast, for the Late Mississippian period, the components with mounds do not have much more variation in house size than the components without. Based on the patterns identified above, I developed three testable hypotheses with domestic structures size as the dependent variable. I then performed linear regression analyses to determine the degree to which the site-level variables accounted for variation in domestic structure size. First, I hypothesized that there would be a negative empirical relationship between site size and domestic structure size. At large, nucleated sites, I expected that people would build smaller domestic structures as a result of increasing restraints on space. Second, I hypothesized that the average size of domestic structures would decline as the total number of structures increased, again reflecting the constraints of crowding. Third, I hypothesized there would be a negative empirical relationship between domestic structure size and the number of mounds at a site. This stems from the observation that multi-mound sites tend to have a few very large domestic structures, but most of the village-level domestic structures are quite small. This would also fit the pattern of more stratified ancient societies in which core urban centers typically contained a few large residential structures surrounded by many smaller ones (Hirth 1993). I tested these hypotheses using all the domestic structures in the database, regardless of location and dates of occupation. While this approach ignores important differences in Woodland, Mississippian, and Historic Indian societies, it provides a useful baseline for comparison for the multivariate analysis that follows. The results of these tests are displayed in Table 9.1. Figures 9.7 – 9.9 show scatterplots with a line of best fit for the three pairs of variables.

286

There is a significant but weak negative empirical relationship between domestic structure area and the number of structures present in a settlement (Figure 9.7, Table 9.1). According to the model, only 9.7 percent of the variation in domestic structure area can be accounted for by variation in the total number of structures, with average floor area decreasing by approximately 0.16 m2 for each additional structure on a site (F = 89.31, Prob>F = 0.000, R2 = 0.097, 1 df, n = 838). The empirical relationship between domestic structure area and site size is also significant and negative, with a low magnitude (Figure 9.8, Table 9.1). According to the model, only 15.4 percent of the variation in domestic structure area can be accounted for by variation in the total number of structures, with average floor area decreasing by approximately 0.02 m2 for each additional hectare of settlement area (F = 151.67, Prob>F = 0.000, R2 = 0.154, 1 df, n = 838). Finally, there is a significant but weak negative empirical relationship between domestic structure area and the number of mounds present in a settlement (Figure 9.9, Table 9.1). According to the model, 10.5 percent of the variation in domestic structure area can be accounted for by variation in the number of mounds, with average floor area decreasing by approximately 0.21 m2 for each additional mound on a site (F = 87.11, Prob>F = 0.000, R2 = 0.105, 1 df, n = 741). These initial tests suggest that variation in settlement size and centrality may influence domestic structure size, but at a very small rate. Each model is significant, but the componentlevel variables account for little variation in domestic structure size, and the magnitude of change in domestic structure size with increasing site size, number of mounds, or number of structures is small.

287

A multivariate regression model that takes time periods into consideration reveals that the combined effects of increasing site size and centrality have a greater effect on domestic structure size than any single variable. Table 9.2 shows the results of this model (domestic structure size = ƒ[component size, mounds, number of structures, dummy variables for time periods]). There is a significant, moderate empirical relationship between domestic structure area and component area, number of mounds, and number of structures. According to the model, 25.7 percent of the variation in domestic structure area can be accounted for by variation in the three measures of site size and centrality, controlling for time (F = 31.58, Prob>F = 0.000, R2 = .257, 8 df, n = 741). However, of these three independent variables, only the number of structures on a site has a significant effect on domestic structure area; with every additional structure on a site, domestic structure area decreases by 0.06 m2, net of the combined effects of additional mounds, component size, and time period. The results of the univariate and multivariate regression analyses suggest that the three measures of site size and centrality account for a significant but very small amount of variation in domestic structure size. The total number of structures present in a settlement has the most significant influence on the size of domestic structures when all of the variables are considered at once, but this influence is small. House Form and Site Size As previous chapters have shown, architectural diversity can vary a great deal from one site to another. For example, the Late Woodland occupations at Hoecake and 1PI61 are remarkably homogenous; all the structures at both sites can be categorized as small, rectangular, single post dwellings. In contrast, during the Early Mississippian, Hiwassee-Island phase at Toqua, Polhemus (1987:231) identifies seven different types of structures based on their shape,

288

wall style, and interior posts patterns. Based on the assumption that large, nucleated settlements and central places should have a higher frequency of special-purpose buildings for suprahousehold civic and ceremonial functions, I hypothesized that diversity in structure form would be higher at settlements with more structures, greater area, and more mounds. Here I am considering all types of structures: domestic, non-domestics, storage, and structures of unknown function. Following Polhemus‘s (1987) strategy for developing architectural typologies, I sorted the structures at each component into types based on their functional class, shape, and wall construction technique. For example, the early Cherokee occupation at Brasstown Valley has three types of structures: domestic structures with rounded corners (winter houses), rectangular domestic structures (summer houses), and small circular storage buildings. Table 9.3 shows the number of types of structures for each component. The number of types per component ranges from one to seven, and the average is approximately two. Unfortunately, among many of the components with only a single structure type, excavations were limited, and the lack of diversity reflects our incomplete understanding of the settlement, rather than true homogeneity in the architecture. This is likely the case at Pott‘s Tract and Little Egypt, where time constraints only allowed for the excavation of a few primary domestic structures. There are probably small, rectangular storage buildings at these sites, given their common occurrence in the Southern Appalachians. In other cases, the lack of architectural diversity may reflect chronological changes in the use of a particular space. During the Stirling phase at Cahokia, Tract 15A had very little residential use and was instead the site of large public monuments, including circles of large posts, or ―wood henges.‖ Similarly, at Town Creek during

289

the Late Mississippian phase, a handful of small, square domestic structures replace the villagelevel mortuary structures of the Middle Mississippian occupation. In some cases the low number of structure types represents genuine uniformity in the architecture. At many of the better understood Middle and Late Woodland settlements, such as the Cartersville phase occupation at Hickory Log and the Edelhardt phase component at the BBB Motor site, all of the structures appear to conform to a single architectural style. The components that received fairly extensive excavations and have only one structure type share common features. They have relatively few contemporaneous houses (no more than 16), and they usually do not have mounds. As expected, many of the components with the most structure types were important civicceremonial centers (Table 9.3). The Yearwood site, though poorly understood, is thought to be a Middle Woodland period ceremonial center based on the high frequency of fancy goods and the presence of cremations (Butler 1979; Faulkner 2002). For the Early Mississippian period, major mound centers including Cahokia, Moundville, and Kincaid fall into this category. So do sites with fewer mounds, including Hiwassee Island, and Toqua. Kincaid and Town Creek have a high number of house types during their Middle Mississippian period occupations. During the Late Mississippian period in the Southern Appalachians, most well documented sites have at least two types of structures, semi-subterranean winter houses and smaller, rectangular storage buildings, or barbacoas. In addition to these two types, larger settlements, such as King, Coweeta Creek, and Toqua, also have large townhouses, which were used for community gatherings and ceremonies. The Late Mississippian occupations at 9GE903 and Rucker‘s Bottom, both on the edge of the Southern Appalachians, do not fit this pattern, but have three and four house types, respectively. At 9GE903, there is a small, circular building in

290

addition to the large round and small rectangular houses typical of the Oconee River Valley. Both the Middle and Late Mississippian occupations at Rucker‘s Bottom have an unusually high degree of architectural variability. In addition to a square structure that closely resembles Southern Appalachian winter houses, there are several large and small circular buildings whose precise function is unknown. The nature and function of this settlement is not entirely understood. Most Cherokee settlements have two to three types of structures: summer houses, winter houses, and in some cases, storage buildings. As in the Late Mississippian period, Historic period settlements with large townhouses stand out when the components are ranked by number of house types. Chota-Tanase, Mialoquo, and Toqua have large townhouses that distinguish them from smaller, more peripheral settlements like Townsend and Ravensford. When the total number of structure types is used as an indicator of architectural variability, the most diverse components share certain characteristics. They are nucleated settlements that also had non-domestic structures for important civic-ceremonial centers, such as King and Coweeta Creek. While the patterns of variation outlined above are fairly clear, comparing the number of structure types per component fails to account for differences in sample size. Because the total number of structures at each component varies widely, simply comparing the total number of structure types per component can be misleading. The Simpson‘s Diversity Index provides a simple way to determine the variation in categorical data. By using proportions of observations, the index accounts for the impact of sample size. Commonly used in demographic and sociological research to describe the homogeneity or heterogeneity of a population, the Simpson‘s Diversity Index (D) is defined as: N

291

D = 1 - ∑ pi2 i=1 Where p equals the proportion of individuals or objects in a category and N equals the number of categories. A perfectly homogenous population has a diversity index of 0, while a perfectly heterogeneous population has an index of 1. The diversity indices for the components in the database range from 0 to 0.73 (Table 9.3). The average value for D is 0.22. Fifty of the sites have a diversity index of zero. This group is comprised mainly of components with ten or fewer structures, but there are several components with multiple structures that are all the same style. The more diverse sites span a wide range in terms of size and number of structures. The Late Town Creek/Early Leak-phase occupation at Town Creek and the Beaverdam and Rembert phase components at Rucker‘s Bottom appear to be the most architecturally diverse of the components. It is noteworthy that these settlements are just outside of the Southern Appalachian core, and they both have unusual architecture compared to most settlements in the Southern Appalachian region. The most diverse components are generally the same important central places and nucleated settlements with public architecture identified by the comparison of the number of structure types. However, calculating the diversity index also draws attention to less impressive sites with a great deal of architectural diversity. Two sites in particular, Rucker‘s Bottom and Jenrette, are not major civic-ceremonial centers, but do have high diversity indices. Rucker‘s Bottom is a roughly 1 ha site with a palisade that Anderson (1994:174) classifies as an agriculture village. It appears to have an unusually high degree of architectural diversity given its small size, lack of mounds, and relatively low number of houses. Based on survey data from the surrounding area, it is possible that Rucker‘s Bottom was the only gathering point for surrounding communities over a fairly large area (Anderson 1994:201). 292

Preservation bias can also explain the unusually high architectural diversity. The Jenrette site has three structures, and each one is different, giving it a high diversity index. However, the site is badly eroded and not completely excavated (Ward and Davis 1999:237-239). There are almost certainly additional houses that have not been identified. Sites with a very low diversity index are primarily those with only a few structures, and which do not appear to be central places in settlement systems. In some cases, the limited extent of the excavations is clearly an issue. On the other hand, some components with low diversity indices are found at sites with broad horizontal excavations. These include the Middle Woodland occupation at Hickory Log, the Edelhardt phase occupation at BBB Motor, and the Stirling phase occupation at Cahokia Tract 15A. In these cases the houses are genuinely very similar. The diversity indices also reveal that some sites with a high number of structure types (three or more) are relatively homogenous due to the distribution of structures in each of their categories. The Late Mississippian occupation at Toqua, for example, has 4 different kinds of houses and 79 total structures, but their distribution results in a more homogenous profile than the Historic Indian component at Brasstown Valley, when only 21 structures are divided among 3 categories. Despite having 5 structure types, the Lohmann phase occupation of Tract 15A at Cahokia has a diversity index of only 0.30 because 83 percent of the houses fall into a single category: small, rectangular wall trench houses. The same is true of the houses in the Roadway area at Moundville. From the perspective of a resident or visitor, the residential areas of these large sites may have seemed more homogenous than the residential areas with the same number of house types but a smaller overall population. These observations suggest that important central places and large, nucleated sites will have more architectural variability than small, dispersed settlements, but a more empirical

293

analysis casts doubt on this correlation. I performed bivariate correlation analyses to examine the effect of site-level variation on house diversity. There is a weak, positive correlation between the diversity index and the number of structures, but it is not significant at the 0.05 level most analysts would accept (Pearson‘s Correlation Coefficient=0.19, R2=0.060, N=95, 1 df, p>0.06). There is a weak, negative correlation between the diversity index and the number of mounds, but it is not statistically significant (Pearson‘s Correlation Coefficient=-0.02, R2=0.001, N=95, 1 df, p>0.812). This is also true of component size (Pearson‘s Correlation Coefficient=-0.078, R2=0.006, N=95, 1 df, p>0.499). A multivariate regression model examining the relationship between the diversity index, all three indices of settlement function, and controlling for time with dummy variables suggests that the combined effects of site size, number of structures, and mounds may have more influence than any single variable. Table 9.4 shows the results of this model (diversity = ƒ[component size, mounds, number of structures, dummy variables for time periods]). There is a significant, moderate empirical relationship between diversity and component area, number of mounds, and number of structures (F = 2.94, Prob>F = 0.006, R2 = .215, 8 df, n = 95). According to the model, 21.5 percent of the variation in domestic structure area can be accounted for by variation in the three measures of site size and centrality, controlling for time. However, of these three independent variables, only the number of structures on a site comes close to having a significant effect on diversity index (t = 1.98, p = 0.051), and this impact is very small. In sum, comparisons of the number of structures types and the diversity indices suggest that important central places and large, nucleated sites will have more architectural variability, but this hypothesis was not confirmed by testing with univariate and multivariate regression

294

models. Of the three indicators of site size and centrality, the total number of structures may be the best predictor or architectural diversity. Fortifications and Domestic Structures Palisades, ditches, and other fortifications constrain the amount of building space available, which may result in smaller houses, especially as villages expand. Fortifications may also draw in full- or part-time residents during times of conflict, increasing the number of domestic structures and reducing house size due to spatial constraints. But fortified walls occur more often at important central places with large public buildings, elite residences, and other special purpose buildings. I hypothesized that there would not be simple empirical relationships between measures of domestic structure size and diversity and fortifications, but that domestic structures at sites with palisades might be smaller due to the impact of crowding. The results were mixed. Out of 95 total components in the database, 35 have palisades and 6 either do not have palisades, or their palisades have not been located (Table 9.3). From a broad perspective, in contrast to my expectations, structures at sites with fortifications are larger than structures at sites without such features. The median size of domestic structures at sites without palisades is 18.2 m2 (n=393) while the median size of domestic structures at sites with palisades is 31 m2 (n=431). A Wilcoxon rank-sum test indicates that the difference in area is statistically significant (Z=-8.62, p=0.000). This pattern holds when the domestic structures are grouped by time period. At a grand scale, there is strong tendency for domestic structures at sites with palisades to be larger. This suggests that the occurrence of large domestic structures at sites with palisades and ditches outweighs the impact of crowding that fortifications may have caused. At the single-site scale, palisades and ditches can be seen to reduce to size of domestic structures, primarily on the edges of settlements. At King, Morris, Snodgrass, and Warren

295

Wilson, there is clear evidence that houses were built smaller to fit into small spaces between existing structures and the edges of palisade walls or ditches. However, these site-level cases are not frequent enough to dramatically lower the average size of houses. There is also a significant empirical relationship between fortifications and the architectural diversity of a site. The 35 components with palisades or other fortifications have an average diversity index of 0.29 and 2.5 structure types, while the 60 components without recorded fortifications have an average diversity index of 0.19 and only 1.8 structure types. Wilcoxon rank-sum tests for the difference in the diversity index (Z = -2.055, p = 0.039) and the number of structure types (Z = -2.629, p = 0.009) reveal that these differences are statistically significant. In some cases this simply reflects the tendency for major central places to be fortified. However, it also helps account for sites that appear to be unusually diverse outliers when site size, total number of structures, and mounds are used as predictors of architectural variability. Discussion and Conclusions In this chapter I have moved beyond the customary scale of household studies, examining the effect of site‘s function in a settlement system on domestic architecture. This strategy provides an important complement to the household-level analyses of architectural variation employed in other chapters, and produces some surprising results. There do not appear to be simple, linear relationships between site-level variables and house form. The combined effects of settlement size, the number of structures present at a settlement, and the presence of mounds seem to have a significant but small effect on variation in domestic structure size and architectural diversity. Of these three predictor variables, the overall number of structures, a good proxy for nucleated versus dispersed settlements, may be the most

296

important. Differences in house size and form seem to be better explained by social factors operating at multiple spatial scales, such as synchronic variation in domestic production and consumption, household symbolism, and status differentiation. Palisades and fortifications are a surprisingly good predictor of architectural variability. At fortified sites there are usually more house forms compared to unfortified sites, and a higher degree of architectural diversity. This does point to meaningful differences in domestic structures at large, nucleated sites and smaller, dispersed sites. Crowding and diminishing space in a settlement accounts for reduced size in domestic structures, but since it mostly affects the edges of habitation of areas, it does not drive down average house size as much as we might imagine.

297

Table 9.1 Results of Regression for Domestic Structure Area Domestic structure area = ƒ (number of structures) Structures (n) F( 1, 836) Prob > F R-squared Area No. structures Constant

838 89.31 0 0.0965 t Coef. -0.1617172 38.28501

P>|t| -9.45 36.29

0.000 0.000

Domestic structure area = ƒ (component size) Structures (n) F( 1, 836) Prob > F R-squared Area Comp. size Constant

838 151.67 0 0.1536 t Coef. -0.0233057 35.12623

P>|t| -12.32 46.48

0.000 0.000

Domestic structure area = ƒ (number of mounds) Structures (n) F( 1, 739) Prob > F R-squared Area No. mounds Constant

741 87.11 0 0.1054 t Coef. -0.2133701 35.3907

P>|t| -9.33 44.21

298

0.000 0.000

Table 9.2 Results of Multivariate Regression for Domestic Structure Area Domestic structure area = ƒ (no. mounds, component size, no. structures, dummy variables for time periods) Structures (n) 741 F( 8, 732) 31.58 Prob > F 0 R-squared 0.257 Area No. mounds Comp. size No. structures MW LW EM MM HI Constant

Coef. t 0.0187254 -0.0177983 -0.0641368 2.879639 -28.42443 -9.223896 -12.84617 -8.408538 44.521

P>|t| 0.15 -1.44 -2.48 1.08 -9.64 -4.05 -5.48 -3.48 27.96

0.879 0.152 0.013 0.282 0 0 0 0.001 0

299

Table 9.3 Summary Table for Architectural Diversity by Component

Component 1GR1X1 1PI61 9CK(DOT)7 9GE1754 9GE1760 9GE1776 9GE1781 9GE222 9GE333 9GE901 9GE903 9WH120 Alarka Banks III Banks V EM Banks V MW BBB Motor Site EM BBB Motor Site LW Bessemer Brasstown Valley HI Brasstown Valley LW Brasstown Valley LW Brasstown Valley MW Cahokia EM Cahokia L Cahokia LW Cahokia M Cahokia S ChotaTanase EM ChotaTanase HI Coweeta Creek HI Coweeta Creek LM1

D 0.00 0.38 0.00 0.00 0.38 0.00 0.50 0.44 0.00 0.00 0.42 0.50 0.50 0.44

Occ. MW LW EW LM LM LM LM LM MW LM LM LM HI MW

Structures (n) 1 4 1 2 4 3 2 3 1 5 15 2 2 3

Types (n) 1 2 1 1 2 1 2 2 1 1 3 2 2 2

Dom. (n) 1 4 1 2 4 3 2 3 1 5 14 2 2 3

Nondom. (n) 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Stor. (n) 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Unk. (n) 0 0 0 0 0 0 0 0 0 0 1 0 0 0

Comp. size (ha) 0.17 0.38 0.8 1.08 0.85 1.04 1.14 0.04 8.1 1.75 10.2 0.6 0.25 0.1

Mounds (n) 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Plaza N N N N N N N N N N N N N N

Fortified N N N N N N N N N N N N N N

0.00

EM

1

1

1

0

0

0

0.03

1

N

N

0.38

MW

4

2

4

0

0

0

0.1

0

N

N

0.64

EM

5

3

4

1

0

0

0.13

0

N

N

0.00 0.22

LW EM

16 25

1 2

16 4

0 21

0 0

0 0

0.16 10

0 3

N Y

N N

0.63

HI

20

3

10

0

10

0

2.4

0

N

N

0.00

LW

2

1

2

0

0

0

2.4

0

Y

Y

0.00

LW

13

1

13

0

0

0

2.4

0

N

Y

0.00 0.29 0.30

MW EM EM

8 41 119

1 3 5

8 35 106

0 0 0

0 0 0

0 6 13

2.4 890 890

0 100 100

Y Y Y

N N Y

0.08 0.08 0.00

LW MM EM

62 24 7

2 2 1

53 24 5

0 0 2

1 0 0

8 0 0

890 890 890

0 100 100

Y Y Y

N Y Y

0.00

EM

1

1

1

0

0

0

0.1

0

N

N

0.64

HI

33

5

29

1

2

1

2

1

Y

N

0.00

HI

1

1

1

0

0

0

1.2

0

N

N

0.00

LM

5

1

5

0

0

0

1.2

0

Y

N

300

Component Coweeta Creek LM2 Dallas Dog River Duncan Tract EW Duncan Tract MW

D

Occ.

Structures (n)

Types (n)

Dom. (n)

Nondom. (n)

Stor. (n)

Unk. (n)

Comp. size (ha)

Mounds (n)

Plaza

Fortified

0.48 0.18 0.00

LM LM LM

10 29 1

5 2 1

7 26 1

2 3 0

1 0 0

0 0 0

1.2 1 2

0 1 0

Y Y N

N Y N

0.00

EW

2

1

2

0

0

0

0.4

0

N

N

0.00

MW

5

1

5

0

0

0

0.4

0

N

N

Dyar LM 1 Dyar LM2 Ela HI Ela MW Fernvale EM Fernvale MW Garden Creek

0.00 0.38 0.00 0.00

LM LM HI MW

2 4 5 10

2 2 1 1

2 1 5 10

0 3 0 0

0 0 0 0

0 0 0 0

2.13 2.13 0.6 0.6

1 1 0 0

Y Y N N

N N N N

0.00

EM

1

1

1

0

0

0

0.1

0

N

N

0.00

MW

1

1

1

0

0

0

0.1

0

N

N

0.00

MW

1

1

0

0

0

1

5

2

N

N

Go Kart Site Hickory Log MM Hickory Log MW Hiwassee Island EM Hiwassee Island LM Hiwassee Old Town

0.00

LM

1

1

1

0

0

0

2.4

0

N

N

0.00

MM

3

1

3

0

0

0

4

0

N

Y

0.00

MW

13

1

13

0

0

0

4

0

N

N

0.45

EM

47

4

13

34

0

0

4.6

2

Y

Y

0.00

LM

1

1

1

0

0

0

4.6

2

Y

Y

0.28

HI

6

2

4

0

1

1

140

0

N

N

Hoecake Jenrette Jewell Kellog KimberlyClark

0.00 0.67 0.31 0.00

LW HI EM MW

11 3 22 1

1 3 3 1

11 3 13 1

0 0 9 0

0 0 0 0

0 0 0 0

81 0.15 3.5 0.26

2 0 1 0

Y N N N

N Y N N

0.44

EM

3

2

3

0

0

0

0.3

0

N

N

Kincaid EM Kincaid LM Kincaid MM Kincaid UNID

0.44 0.50

EM LM

7 2

3 2

1 1

6 1

0 0

0 0

32 32

27 27

Y Y

Y Y

0.51

MM

6

3

0

6

0

0

32

27

Y

Y

0.28

MM

12

3

4

6

0

2

32

27

Y

Y

King Kolomoki Little Egypt

0.45 0.00 0.00

LM MW LM

67 1 3

4 1 1

44 1 3

2 0 0

21 0 0

0 0 0

2.3 100 2

0 8 2

Y Y Y

Y Y N

301

Component Lower Saratown

D

Occ.

Structures (n)

Types (n)

Dom. (n)

Nondom. (n)

Stor. (n)

Unk. (n)

Comp. size (ha)

Mounds (n)

Plaza

Fortified

0.00

HI

2

1

2

0

0

0

0.04

0

N

Y

Loy Martin Farm M1 Martin Farm M2 Martin Farm M3 Martin Farm M4

0.00

LM

2

1

2

0

0

0

2.5

1

Y

Y

0.00

EM

1

1

1

0

0

0

1.1

0

N

N

0.48

EM

5

2

3

1

0

1

1.1

2

N

Y

0.48

LM

5

2

0

3

0

2

1.1

2

N

N

0.00

LM

1

1

0

0

1

0

1.1

0

N

N

McFarland

0.00

MW

5

1

5

0

0

0

2.65

0

N

N

Mialoquo Mitchum Morris Moundville EM Moundville EM 1 Moundville M1-M2 Napoleon Hollow

0.42 0.00 0.40

HI HI EM

8 1 12

3 1 3

7 1 12

1 0 0

0 0 0

0 0 0

10 0.5 0.6

0 0 0

Y N Y

N Y Y

0.00

EM

14

1

14

0

0

0

75

29

Y

Y

0.43

EM

35

2

35

0

0

0

75

2

N

N

0.08

EM

102

2

98

4

0

0

75

29

Y

Y

0.00

MW

1

1

0

0

0

1

15

0

N

N

Pott's Tract

0.00

LM

3

1

3

0

0

0

0.8

0

N

N

Rivermoore Rucker's Bottom LM Rucker's Bottom MM

0.00

LW

3

1

3

0

0

0

0.1

0

N

N

0.72

LM

5

4

3

1

1

0

0.8

0

Y

Y

0.73

MM

7

4

3

2

2

0

0.8

0

Y

Y

Rymer

0.16

LM

23

2

21

2

0

0

1.94

0

Y

Y

Snodgrass Summerour Toqua EM Toqua HI Toqua LM Town Creek EM Town Creek LM Town Creek LW

0.35 0.00 0.70 0.60 0.49

MM LW EM HI LM

91 1 25 14 79

2 1 7 4 4

91 1 15 12 41

0 0 9 2 23

0 0 0 0 14

0 0 1 0 1

0.82 1 1 1.9 1.9

0 1 2 2 2

Y N Y Y Y

Y N Y Y Y

0.00

EM

1

1

1

0

0

0

0.15

0

Y

Y

0.00

LM

4

1

4

0

0

0

0.15

1

N

N

0.00

LW

1

1

0

1

0

0

0.15

0

Y

Y

302

Component Town Creek MM 2 Town Creek MM1

D

Occ.

Structures (n)

Types (n)

Dom. (n)

Nondom. (n)

Stor. (n)

Unk. (n)

Comp. size (ha)

Mounds (n)

Plaza

Fortified

0.72

MM

19

4

0

16

0

3

0.15

1

Y

Y

0.53

MM

20

4

13

5

0

2

0.15

1

Y

Y

Townsend

0.50

HI

10

2

9

0

0

1

10

0

N

N

Tuckasegee Tukabatchee Two Run Creek Warren Wilson

0.00 0.50

HI LM

1 2

1 2

1 2

0 0

0 0

0 0

0.5 1.2

0 0

N N

N N

0.00

MW

1

1

1

0

0

0

0.9

1

N

N

0.50

LM

12

2

12

0

0

0

1

0

Y

Y

Yearwood

0.60

MW

14

4

12

0

0

2

1.01

0

N

N

303

Table 9.4 Results of Multivariate Regression for Architectural Diversity Diversity = ƒ (no. mounds, component size, no. structures, dummy variables for time periods) Components (n) 95 F( 8, 732) 2.94 Prob > F 0.0059 R-squared 0.215 Diversity No. structures Comp. size No. mounds MW LW EM MM HI Constant

Coef. t 0.0024805 -0.0001454 -0.0010705 -0.1407094 -0.1845001 0.0264434 0.1612732 0.0874493 0.2206232

P>|t| 1.98 -0.58 -0.44 -1.93 -1.99 0.37 1.67 1.13 4.9

0.051 0.563 0.66 0.056 0.05 0.714 0.098 0.263 0

304

Figure 9.1 Domestic Structure Area at Middle Woodland Components, Sorted by Number of Structures

Figure 9.2 Domestic Structure Area at Late Woodland Components, Sorted by Number of Structures

305

Figure 9.3 Domestic Structure Area at Early Mississippian Components, Sorted by Number of Structures

Figure 9.4 Domestic Structure Area at Middle Mississippian Components, Sorted by Number of Structures

306

Figure 9.5 Domestic Structure Area at Late Mississippian Components, Sorted by Number of Structures

Figure 9.6 Structure Area at Historic Indian Components, Sorted by Number of Structures

307

Figure 9.7 Scatterplot of Domestic Structure Area and Component Area

Figure 9.8 Scatterplot of Domestic Structure Area and Number of Mounds

308

Figure 9.9 Scatterplot of Domestic Structure Area and Number of Structures

309

CHAPTER 10 CONCLUSION

Household archaeology is a research approach that focuses on the household as the primary unit of analysis, but it is not a unified theory. However, small-scale household studies provide the tools for building a coherent theory that delineates the different social factors that explain variation in houses at larger scales. In this chapter I synthesize the findings of this study and use these architectural data, along with insight from household archaeology and anthropology, to propose such a theory. The first step is to understand the degree to which environmental variation determines architectural variation. In the case of the Southern Appalachians, I found that the greatest environmental influence was related to the use of natural resources for building materials. A multivariate regression analysis suggests the combined effects of rainfall, temperature, and elevation may account for some variation in how solidly domestic structures are built, but this finding is based on a subset of the structures in the database. At the scale of this study—the Southern Appalachian region and surrounds—environmental variation does not seem to be a key explanatory factor for variation in domestic structures. The social reasons for architectural variation seem to be more important. At much larger spatial scales, however, where differences between plant communities and climate are more pronounced, it seems likely that environmental variables will play a stronger role in explaining architectural variation (e.g. a comparison of domestic structures in the Southern Appalachians and the North Carolina coast). The next step is to sort out and weigh the different social factors that explain architectural 310

variation. The social causes for variation in house form are complex, but the household studies reviewed in Chapter 2 tell us which ones are generally the most important. Architectural variability is often the result of four social factors: household economics and household composition, symbolic behavior, status differentiation, and settlement patterning. In Chapters 6 through 9 I tested a set of related hypotheses to determine how each of these factors impacted synchronic variation and diachronic changes in Southeastern architecture. My findings suggest that certain social factors are better than others at explaining certain lines of architectural variation. I have used architectural data from the Southern Appalachian region as a case study, but I argue that this approach is broadly applicable. Finally, I argue that addressing these questions about domestic architecture at a broad spatial and temporal scale, with a large data set, leads to another, even simpler theoretical statement about the relationship between domestic life and architecture. I posit that broad-scale changes in architecture are best explained by broad-scale changes in social organization. Considering the social factors outlined above provides theoretical expectations for the nature of these changes. The spatial and temporal scale of architectural variability will depend on the size, permanence, and degree of integration of social groups larger than households, i.e. the lineages, clans, chiefdoms, and other corporate groups to which households belong. The Social Reasons for Architectural Variability Household Economics and Household Composition. In the Southern Appalachian region, it appears that major changes in the subsistence economy account for significant changes in the size, spacing, interior partitioning, and the degree of architectural investment in domestic structures at a broad spatial and temporal scale. With the adoption of intensive maize agriculture there appears to be a shift toward more autonomous households with more private storage. This

311

is consistent with observations from ethnographic and archaeological studies in other parts of the world (Byrd 1994; Kent 1990; Flannery 2002). We can expect variation in domestic structures to reflect variation in subsistence patterns. In the case of the Southeast, with the adoption of a more intensive and sedentary subsistence strategy, we see a shift from larger, less permanent structures housing larger households that cooperate frequently in daily subsistence activities to smaller, more widely spaced structures with more autonomous households. After a sedentary, intensive subsistence strategy has been established for some time, domestic structures become larger, but more partitioned, probably reflecting an increasing tendency to move storage and subsistence activities indoors. This pattern of development is widespread, but it is not strictly unilinear. For example, Late Mississippian and Cherokee winter houses share many formal similarities, but the Cherokee domestic structures are not rebuilt as often. This is likely the result of the shorter intended use lives of the Cherokee structures, due to the participation of households in the more mobile deerskin trade and the uncertain political, biological, and social milieu of the times. The domestic structures of the upland Lamar farmsteads of the Oconee River valley are more lightly built and less partitioned than contemporaneous houses in the floodplain below, and this may be explained by differences in household economics and household composition within the same time period. Houses and Symbolism. Considering the well-documented tendencies for houses to be imbued with important cosmological and religious symbolism helps explain many of the particulars of architectural variation left unexplained by economic changes. There are numerous ways to arrange partitions within a domestic structure, or position houses in a settlement, or erect the walls of a house, but in the Southern Appalachian region of the Southeast it seems that major

312

cosmological and religious themes recorded in the historic period may explain these specific lines of architectural variability. There is some danger in extending these historically documented cosmologies as far back as the Mississippian and Woodland period. However, there is good evidence from other kinds of material culture (e.g. surface treatments on ceramics and iconography on gorgets and pipes) to suggest that concepts including the separation of opposite categories (e.g. summer/winter, male/female), the quartering of the earth and a below-world and above-world, the symbolic importance of the cardinal directions, and the symbolic importance of the numbers four and seven, have deep roots in the region (Anderson 1998:291, Saunders 1998:156-158). In places like the Southeast, where we have knowledge of these concepts, we can predict that certain aspects of house construction including orientation, the spatial layout of interior posts and features, the number of posts used for construction, and the overall shape of houses may be explained with reference to important cosmological and religious symbols. Status Differentiation. Archaeologists have long understood that status differentiation explains some differences in the size and appearance of domestic structures, the frequency and diversity of interior features, and the location of structures in settlements (Hirth 1993). In general, we can expect that in more egalitarian societies, domestic structures will be more similar along these lines of variation, while in more stratified societies, the differences will be more pronounced. In the Southern Appalachian region this is the case, as differences in the size and location of structures within settlements are most pronounced in the chiefdom societies of the Mississippian period, while domestic structures seem more uniform at large and small spatial scales in the more egalitarian contexts of the Woodland and Historic periods. In general, considering status differentiation is especially useful for explaining intrasite differences in domestic structures that might otherwise go unexplained by variation in household economics

313

and household composition, or by symbolic associations. Settlement Patterning. The function of a site in a larger settlement system may explain certain aspects of architectural variation not explained by the other three big social factors. I had predicted that there might be fairly simple relationships between site-level measures of the size and centrality of settlements and the size of domestic structures, but this is not the case. Univariate and multivariate regression analyses suggest that combined measures of site size and centrality (component size, number of structures, and number of mounds) may account for some of the variation in domestic structure size and architectural diversity. Of these three site-level variables, the total number of structures, a good indicator of nucleated versus dispersed settlement, may be especially important in determining the size of domestic structures. There is some indication that fortified and nucleated sites may have more diverse architecture than dispersed settlements. Architecture and Social Organization These insights from small-scale household studies help explain much of the architectural variation observed in the Southern Appalachian region of the Southeast, and these ideas could be applied to other regions to generate testable hypotheses about broad-scale synchronic and diachronic variation in domestic architecture. However, this theoretical framework, which considers multiple reasons for architectural variation, needs to be linked to a more general theory that can account for the broad spatial and temporal scale at which some patterns of architectural variability are observed. I argue that broad-scale changes in architecture, which can be delineated using the fourpart framework outlined above, are best explained by broad-scale changes in social organization. House construction is not an individual activity. Houses in non-state societies are built using

314

community and even supra-community labor, often organized along kin lines (see for example Horne 1994; Kramer 1982). Houses are thus usually constructed by social groups larger than households, and it is the interaction between these larger social groups that we need to consider in our explanations of architectural variability. House building in historic Creek society provides an example. Before getting married, a young Creek man had to raise a crop and build a house with the help of his male relatives (Hudson 1976:198). Given that Creek society was matrilineal and matrilocal, this meant that house construction involved supra-household male labor, since older married men from the boy‘s lineage would have been spread out, probably living with their wife‘s household. Within this tradition of house construction, domestic structures would look more similar when societies were integrated, with high rates of marriage between groups living in different settlements. Houses would look dissimilar at a regional and macroregional scale when society was less integrated and marriage alliances were more localized. Houses are also a major investment of materials and labor, much more so than pottery or other classes of material culture. Thus the nature of house construction, both in terms of material investment and the social processes, makes houses especially good indicators of broad-scale changes in social organization. When there are changes the size, organization, and connectedness of important social groups, we should expect to see related changes in architecture, even more than in pottery and projectile point styles, which can be more easily manipulated by individuals and households at small scales. We should expect to see change in domestic architecture track with major changes in social organization identified through other lines of evidence (settlement patterns, subsistence strategies, burial patterns, etc.), and we should expect the scale of these architectural changes to

315

vary depending on the size and scale of integration of societies. Households, even isolated ones on dispersed farmsteads, are always part of larger social groups, but those larger corporate groups (lineages, clans, chiefdoms, polities, regional cultures) are not static. They grow and contract over time in response to large-scale political, economic, and social forces. When the larger social groups that households belong to are more integrated, we might expect houses to look more similar at a broad spatial scale, as the movement of people across the landscape would be less constrained by political and social boundaries. Several findings from this study support this general theory. During the Middle Woodland period, the pattern of round houses is more widespread and regular than we might have thought. This calls into question the conventional wisdom that Middle Woodland domestic activity beneath the level of the integration of the Hopewell Interaction Sphere varies widely across the Southeast (see Anderson and Mainfort 2002). Instead, there are signs of social interaction and integration in the domestic architecture. The round houses that look similar in middle Tennessee, northern Georgia, and even the American Bottom might be an indicator of pan-regional social integration in the domestic sphere. During the Late Woodland and Emergent Mississippian period, houses become more dissimilar at a macroregional scale, but more similar at the scale of the single site and region. In the western part of the study area, domestic structure size and form is remarkably uniform in sites and across a broad area from the American Bottom southward to west-central Alabama. This is a distinct pattern from the Southern Appalachian region, where there seems to be more architectural variability. This could point to a more localized level of social integration; marriage alliances and the social process of house construction would have been more circumscribed, resulting in a pattern of regional similarity and macroregional differences in houses. Cobb and

316

Nassaney come to a similar conclusion by examining the distribution of Late Woodland small triangle projectile points, maize, shell-tempered pottery, and shell ornaments: We suggest that from the macroperspective the scale of social reproduction was altered during the Late Woodland period. There seems to have been a contraction in the spatial extent of the alliance networks that forged and replicated social relations. Groups became more dependent on close-range ties to obtain goods central to effecting daily social transactions… Imported items found during the Late Woodland period were likely used in the transactions of daily ritual occasions, such as puberty rites, bride-wealth payments, and religious fees (1995:217). In the Early Mississippian period, the common occurrence of small rectangular houses, and more specifically, wall trench houses, points again to widespread integration between social groups from the American Bottom to Georgia. There is some regional variation in Early Mississippian house form. Cobb and King (2005:169) state that wall trench houses were more common in the Central Mississippi valley while single post structures were more common in the Southern Appalachians. However, Early Mississippian wall trench houses occur in eastern Tennessee and northern Georgia, and single post structures persist throughout the sequence in small numbers on both sides of the Southeast. From a broad view, houses are more similar at a macroregional scale during the Early Mississippian period than during any other time in the Southeast. Early Mississippian polities may have been extensive; we can infer this from a political landscape with a few large mound centers, such as Cahokia, Etowah, and Moundville, that may have dominated large geographic areas. Exchange, competition, and emulation between these polities, Renfew‘s (1986) concept of peer-polity interaction, would have created broad avenues for exchanges of people, material, and ideas. Households could have used wall trench

317

houses could as a symbol of group membership in these big new polities, or they may have been imposed on households from the top down through coercion. The general disappearance of wall trench houses beginning after A.D. 1200 is also noteworthy, and tracks with a general transition from Early to Late Mississippian political organization; a shift from few polities with large centers (Cahokia, Moundville, Etowah) to more polities with smaller centers (e.g. Barnett phase polities like Coosa) (Cobb and King 2005). If wall trench houses did have a strong association with the Early Mississippian social order, the large-scale abandonment of the form may be related to the waning strength of the early polities. This change could be the result of bottom-up or top-down processes operating on a large scale. As the political clout of leaders at Cahokia and Moundville declined, corporate groups may have gradually abandoned the old house form, creating symbolic distance between themselves and the old elite. The change in house style could also have been mandated from the top down by new elite filling the power vacuum left by the older, larger polities. Houses are more similar in the Southern Appalachians during Late Mississippian times than during Early Mississippian times, suggesting a high degree of integration and exchange between the polities in eastern Tennessee, western North Carolina, and northern Georgia. Compared to the Early Mississippian period, there is something of a contraction in the spatial extent of similar houses, but the complex Late Mississippian winter house form is remarkably consistent in and between sites. This points to the existence of a large regional culture, spread over much of the Southern Appalachians. During Historic Indian times, against a backdrop of disease, violence, and political upheaval, there is less investment in domestic houses, but among historically known groups house form is remarkably uniform. The similarity in domestic structures across a broad

318

geographic area may be the result of refugee groups forming new communities in the Southern Appalachians, following Kowalewski‘s (2006) model for coalescent societies. Sharing architectural patterns at the household level, patterns that possibly refer to older, Late Mississippian public architecture, may have been an important mechanism for integrating newly formed communities. In a broad geographic area that once had similar Late Mississippian winter houses, we see divergence in architectural forms and architectural similarity at a smaller spatial scale as these historic groups become archaeologically and historically visible (Hally 2002). Moreover, the architectural investment in some Cherokee townhouses is remarkable. Few structures in the database look as permanent as the townhouse at Chota-Tanase, with its two building episodes and 1215 posts. This heavy investment in public structures rather than domestic structures may indicate a shift toward the importance of social groups larger than households, such as matrilineal clans, as the more important unit of social organization. Warfare was common in the Southeast from the Late Woodland period onward, as evidenced by the widespread occurrence of settlements with palisades and violent deaths (Milner 1999). The in-place rebuilding of houses is extremely rare in the Late Woodland period, but then becomes more common in the Early Mississippian period, and more so in Late Mississippian times. During the Historic Indian period, in the wake of the biological, social, political, and economic upheaval of contact with Europeans, the rebuilding of domestic structure diminishes. During Early Mississippian times, large polities like Cahokia and Moundville may have contributed to increased boundary maintenance and social stability (as in Pauketat‘s [2004] Pax Cahokiana), resulting in an uptick in the rebuilding of houses, even in the face of endemic Southeastern warfare. In the Late Mississippian period in the Southern Appalachians, the inplace rebuilding of houses is even more common. Hally (2006) has argued that these Late

319

Mississippian polities were quite stable at a regional level, even as they were prone to cycling and collapse at smaller scales. Widespread changes in Southeastern domestic architecture happened concordantly. The broad spatial and temporal extent of these changes requires large-scale driving forces, which I argue are major changes in social organization, and the degree of interaction and interdependence of Southeastern societies. The major transitions in house form in the Southeast happen at times of system-wide political and economic change. These include: the possible intensification of small seed crop plants and the establishment of the Hopewell Interaction Sphere; the Middle Woodland to Late Woodland/Emergent Mississippian transition that coincides with the adoption of maize agriculture and the emergence of chiefdom political organization; the breakdown of large, Early Mississippian polities and the establishment of smaller polities and paramount chiefdoms; and the emergence of Historic Indian groups of the eighteenth century. The expansion and contraction of alliance and exchange networks associated with these major changes would have mediated the smaller-scale processes of house building. The approach that I propose in this research offers an alternative to either strictly materialist, historical, or ideological explanations of architectural variability. Grounded in household studies, it explains architectural variation in terms of definable social processes. Operating at a multiple scales, it allows for the identification of patterns of behavior that require top-down, as well as bottom-up, explanations, and leaves room for considerations of both structure and agency. Future Avenues of Research This study is not intended to be the final say on architectural variability in the Southeast. The houses recorded in this database represent a large fraction, but only a fraction, of the well-

320

excavated houses available for study in the Southeast. A logical next step for this project would be to expand and improve the regional coverage, especially around Cahokia and Moundville, and test the validity of these patterns and explanations with a more diverse sample of houses. It would also be advantageous to attempt to record all the structures in a known polity and make comparisons among and between sites. At a broader scale, it would be informative to compare domestic architecture in the Southeast with the Northeast. Large, high-quality architectural data sets exist over much of the Northeast. Why did long houses never emerge as a house form in the Southeast, despite the existence of fairly similar environmental parameters? The emergence of the more hierarchical chiefdom societies in the Mississippian Southeast, as compared to the more egalitarian societies of the Late Woodland Northeast may explain this difference. Due to constraints of time and the quality of record-keeping, I was unable to record finescale data, such as post diameter and spacing, for many of the structures in this study. This points to a broader problem in structure excavation. Too often structures are not identified and delineated in the field, and interpreting post maps years after structure excavation is complete is a risky endeavor (see Knight 2007; Benyshek et al. 2010). However, as GIS software and scanning hardware improves, it may become more efficient to tackle and revive some of the older, more complicated architectural data sets. In addition to expanding the geographic coverage, another step for this project would be to return to some of the more complex site maps and look for finer patterns of variation. Along similar lines, as this data set expands and missing data are filled in, there is great potential for testing hypotheses about architectural variability using more nuanced statistical analyses. In chapter 5, I used stepwise regression and the Akaike information criterion (AIC) to

321

identify the environmental variables that best explained variation in post density. This approach could be applied to all of the dependent and independent variables to empirically determine which causal variables best explain differences in certain architectural features. Given the problems in this dataset with missing data points, I am hesitant to apply this strategy to all the architectural variables in the database, but two case studies demonstrate the potential of this approach. Using stepwise regression and the AIC, I identified the best-fit models for variation in the area of domestic structures and the number of interior partitions in domestic structures. For each model, I began with a model that included all the independent variables in the database, and removed variables until I identified the model with the lowest AIC. The results are shown in Tables 10.1 and 10.2. Note that several environmental variables, the interaction variables for maize productivity, and the number of structures present in a component appear to have a significant effect on variation in domestic structure area, but the number of mounds at a site and the size of the component do not (Table 10.1). In the case of interior partitions, more of the environmental variables are removed from the model, and both mounds and component size do not appear to make affect variation in partitioning (Table 10.2). In both models the scale of the effects are often small, but are still significant. These models should be interpreted with some caution, but they do provide a useful complement to some of the more traditional archaeological analyses applied in this research. In both cases environmental variables may explain some variation in domestic architecture, and of all the indicators of settlement size and function, the number of structures present in a settlement seems to consistently have an effect on domestic structure form.

322

The architectural database constructed here can grow. There have been recent calls in the discipline to make archaeological data buried in gray literature more accessible. Databases like these, which focus on a particular class of archaeological data, are a good way to start. A webbased database led by a working group of archaeologists interested in architecture would be a logical structure. As Knight (2007:187) suggests, there is perhaps no better time than the present to revisit broad comparisons of domestic architecture. This is fertile ground for theory building, and there are still multitudes of neglected houses to be pulled from dusty map drawers and explored.

323

Table 10.1 Best-Fit Model for Domestic Structure Area Domestic structure area = ƒ (Jan. temp., days precip., avg. temp., frost free days, interaction of precip. and avg. temp., interaction of precip. and frost free days, elevation, no. structures, dummy variables for time periods) Structures (n) 737 F( 13, 723) 32.59 Prob > F 0 R-squared 0.3695 Area Coef. t Jan. temp. -1.270547 Days precip. -32.55008 Avg temp. -64.11655 Frost free days 3.089296 Precipxavgtemp 0.6901901 Precipxfrostfree -0.0292727 MW -2.389214 LW -33.19712 EM -15.57982 MM -9.747398 HI -15.70169 Elevation 0.0415273 No. structures -0.1459047 Constant 3009.574 AIC = 6260.32, lowest of all models tested.

P>|t| -3.92 -8.9 -8.15 3.63 8.53 -4.07 -0.86 -9.5 -6.98 -3.35 -6.34 2.2 -7.11 8.36

324

0 0 0 0 0 0 0.388 0 0 0.001 0 0.028 0 0

Table 10.2 Best-Fit Model for Partitions in Domestic Structures Partitions = ƒ (Jan. temp., days precip., interaction of precip. and avg. temp., elevation, no. structures, dummy variables for time periods) Structures (n) 266 F( 10, 255) 15.6 Prob > F 0 R-squared 0.3796 Partitions Coef. t Jan temp. -0.1158294 Days precip. -0.2386754 MW -2.03758 LW -1.961748 EM -1.401075 MM -1.452667 HI -0.4708793 Precipxavgtemp 0.0031511 No structures -0.0087668 Elevation 0.0091037 Constant 9.213241 AIC = 864.97, lowest of all models tested

P>|t| -4.12 -7.12 -6.54 -4.62 -5.71 -2.67 -1.71 5.46 -3.43 5.37 6.15

325

0 0 0 0 0 0.008 0.088 0 0.001 0 0

REFERENCES CITED

Adair, James A. 1968 (1775) History of American Indians: Particularly Those Nations Adjoining to the Mississippi, East and West Florida, Georgia, South and North Carolina, and Virginia. Johnson Reprint Corporation, New York. Allen, William L., and James B. Richardson III 1971 The Reconstruction of Kinship from Archaeological Data: The Concepts, the Methods, and Feasability. American Antiquity 36:41-53. Alvarez Zarikian, Carlos A., Peter K. Swart, John A. Gifford, and Patricia L. Blackwelder 2005 Holocene Paleohydrology of Little Salt Spring, Florida, Based on Ostracod Assemblages and Stable Isotopes. Palaeogeography, Palaeoclimatology, Palaeoecology 225(1-4):134156. Anderson, David G. 1994 The Savannah River Chiefdoms: Political Change in the Late Prehistoric Southeast. University of Alabama Press, Tuscaloosa. 1998 Swift Creek in a Regional Perspective. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams, and D. T. Elliot, pp. 274-300. University of Alabama Press, Tuscaloosa. 2001 Climate Change and Culture Change in Prehistoric and Early Historic Eastern North America. Archaeology of Eastern North America 29:143-186. Anderson, David G. and Robert C. Mainfort, Jr. 2002 An Introduction to Woodland Archaeology in the Southeast. In The Woodland Southeast, edited by D. G. Anderson and R. C. Mainfort, Jr., pp. 1-19. University of Alabama Press, Tuscaloosa. Anderson, David G., and Joseph Schuldrein (editors) 1985 Prehistoric Human Ecology along the Upper Savannah River: Excavations at Rucker's Bottom, Abbeville and Bullard Site Groups. Prepared for the United States Department of Interior National Park Service, Archaeological Services Branch, Atlanta, Georgia, by Commonwealth Associates, Inc. The Branch, Atlanta. Bell, Amelia R. 326

1990 Separate People: Speaking of Creek Men and Women. American Anthropologist 92:332345. Bender, Donald R. 1967 A Refinement of the Concept of Household: Families, Co-residence, and Domestic Functions. American Anthropologist 69:495-504. Bense, Judith A. 1994 Archaeology of the Southeastern United States: Paleoindian to World War I. Academic Press, New York. Benyshek, Tasha, Benjamin A. Steere, Paul Webb, Joel Jones, and Hannah Guidry 2010 Finding, Delineating, and Recording Structure Patterns: Recent Excavations in Western North Carolina. Paper Presented at the 67th Annual Meeting of the Southeastern Archaeological Conference, Lexington, Kentucky. Binford, Lewis R., Sally R. Binford, Robert Whallon and Margaret Ann Hardin 1970 Archaeology at Hatchery West. Memoirs of the Society for American Archaeology No. 24. U.S. Department of the Interior, Northeast Region. Blanton, Dennis B., and Thomas H. Gresham 2007 An Experimental Perspective on Mississippian Small Pole Structures. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 32-47. University of Alabama Press, Tuscaloosa. Blanton, Richard E. 1994 Houses and Households: A Comparative Study. Plenum Press, New York. Blanton, Richard E., Stephen A. Kowalewski, Gary M. Feinman, and Laura M. Finsten 1993 Ancient Mesoamerica: A Comparison of Change in Three Regions. Second Edition. Cambridge University Press, Cambridge. Blitz, John 1993 Ancient Chiefdoms of the Tombigbee. The University of Alabama Press, Tuscaloosa. 2010 New Perspectives in Mississippian Archaeology. Journal of Archaeological Research 18:1-39. Blitz, John H., and Patrick Livingood 2004 Sociopolitical Implications of Mississippian Mound Production. American Antiquity 69(2): 291-301. Bourdieu, Pierre 1977 Outline of a Theory of Practice. Cambridge, Cambridge University Press. Boudreaux, Edmond A. 327

2005 The Archaeology of Town Creek: Chronology, Community, Patterns, and Leadership at a Mississippian Town. Unpublished Ph.D. dissertation, Department of Anthropology, University of North Carolina, Chapel Hill. Braudel, Fernand 1979 The Mediterranean World and the Mediterranean World in the Age of Philip II. Translated by Sian Reynolds. Harper and Row, New York. Braun, E. Lucy 1950 Deciduous Forests of Eastern North America. Blakiston, Philadelphia. Braund, Kathryn E. Holland 1993 Deerskins and Duffels: The Creek Indian Trade with Anglo-America, 1685-1815. University of Nebraska Press, Lincoln. Brennan, Tamira K. 2007 In-Ground Evidence of Above-Ground Architecture at Kincaid Mounds. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 73-100. University of Alabama Press, Tuscaloosa. Byrd, Brian F. 1994 Public and Private, Domestic and Corporate: The Emergence of the Southwest Asian Village. American Antiquity 59(4): 639-666. Butler, Brian M. 1977 The Yearwood Site: A Specialized Middle Woodland Occupation on the Elk River. Tennessee Anthropologist 1(2):1-15. 1979 Hopewell Contacts in Southern Middle Tennessee. In Hopewell Archaeology: The Chillicothe Conference, edited by D. Brose and N. Greber, pp. 150-156. Kent State University Press, Kent, Ohio. Cable, John S., Lisa O'Steen, Leslie E. Raymer, Johannes H. N. Loubser, David S. Leigh, Dr. J.W. Joseph, Mary Beth Reed, Lotta Danielsson Murphy, Undine McEvoy, Thaddeus Murphy, Mary Theresa Bonhage-Freund, and Deborah Wallsmith 1997 "A Picture Unsurpassed": Prehistoric and Historic Indian Settlement and Landscape, Brasstown Valley, Towns County, Georgia. Technical Report Number 457. New South Associates, Inc., Stone Mountain, Georgia. Submitted to The Georgia Department of Natural Resources Maintenance and Construction. Parks, Recreation, and Historic Sites Division, Atlanta, Georgia. Copies available from the University of Georgia Laboratory of Archaeology, Athens. Caldwell, Joseph R. 1964 Interaction Spheres in Prehistory. In Hopewellian Studies, edited by J. R. Caldwell and R. L. Hall, pp. 133-143. Scientific Papers 12, Illinois State Museum, Springfield.

328

Caldwell, Joseph R., and Catherine McCann 1941 Irene Mound Site, Chatham County, Georgia. University of Georgia Press, Athens. Carmean, Kelli 1991 Architectural Labor Investment and Social Stratification at Sayil, Yucatán, Mexico. Latin American Antiquity 2:151-165. Carr, Christopher, and D. Troy Case (editors) 2006 Gathering Hopewell: Society, Ritual, and Ritual Interaction. Springer, New York. Chapman, Jefferson 1994 Tellico Archaeology: 12,000 Years of Native American History. The University of Tennessee, Department of Anthropology, Report of Investigations Number 43. Frank McClung Museum, Occasional Paper No 5. The Tennessee Valley Authority Publications in Anthropology Number 41. University of Tennessee Press, Knoxville. Charles, Douglas K., and Jane E. Buikstra 1983 Archaic Mortuary Sites in the Central Mississippi Drainage: Distribution, Structure, and Behavioral Implications. In Archaic Hunters and Gatherers in the American Midwest, edited by J. L. Phillips and J. A. Brown, pp. 117-145. Academic Press, New York. Christaller, Walter 1966 Central Places in Southern Germany. Englewood Cliffs, Prentice Hall. Clark, Jeffery J. 2001 Tracking Prehistoric Migration: Pueblo Settlers among the Tonto Basin Hohokam. Anthropological Papers of the University of Arizona No. 65. The University of Arizona Press, Tuscon. Clay, R. Berle 2002 Deconstructing the Woodland Sequence from the Heartland: A Review of Recent Research Directions in the Upper Ohio Valley. In The Woodland Southeast, edited by D. D. Anderson, and R. C. Mainfort, Jr., pp.162-184. University of Alabama Press, Tuscaloosa. Clayton, Lawrence A., Vernon J. Knight, Jr., and Edward C. Moore 1993 The DeSoto Chronicles: The Expeditions of Hernando de Soto to North America in 15391543. 2 vols. University Press of Alabama, Tuscaloosa. Cliggett, Lisa 2005 Grains from Grass: Aging, Gender, and Famine in Rural Africa. Cornell University Press, Ithaca, New York. Cobb, Charles R. 2000 From Quarry to Cornfield: The Political Economy of Mississippian Hoe Production. University of Alabama Press, Tuscaloosa. 329

2003 Mississippian Chiefdoms: How Complex? Annual Review of Anthropology 32: 63–84. Cobb, Charles R., and Brian Butler 2002 The Vacant Quarter Revisited: Late Mississippian Abandonment of the Lower Ohio Valley. American Antiquity 67:625-641. Cobb, Charles R., and Patrick H. Garrow 1996 Woodstock Culture and the Question of Mississippian Emergence. American Antiquity 61(1):21-37. Cobb, Charles R., and Adam King 2005 Re-Inventing Mississippian Tradition at Etowah, Georgia. Journal of Archaeological Method and Theory 12(3):167-192. Cobb, Charles R. and Michael S. Nassaney 1995 Interaction and Integration in the Late Woodland Southeast. In Native American Interactions: Multiscalar Analyses and Interpretations of the Eastern Woodlands, edited by M. S. Nassaney and K. E. Sassaman, pp. 205-226. The University of Tennessee Press, Knoxville. 2002 Domesticating Self and Society in the Woodland Southeast. In The Woodland Southeast, edited by D. G. Anderson, and R. C. Mainfort, Jr., pp. 525-539. University of Alabama Press, Tuscaloosa. Coe, Joffre L. 1964 The Formative Cultures of the Carolina Piedmont. Transactions of the American Philosophical Society, n.s., 54, pt. 5. American Philosophical Society, Philadelphia. Cogswell, James W., Michael J. O‘Brien, and Daniel S. Glover 2001 The Artifactual Content of Selected House Floors at Turner and Snodgrass. In Mississippian Community Organization: The Powers Phase in Southeastern Missouri, edited by M. J. O‘Brien, pp. 181-230. Kluwer Academic/Plenum Publishers, New York. Collins, James M. 1997 Cahokia Settlement and Social Structures as Viewed from the ICT-II. In Cahokia: Domination and Ideology in the Mississippian World, edited by T. R. Pauketat, and T. E. Emerson, pp. 124-140. University of Nebraska Press, Lincoln. Crocker, Christopher 1969 Men's House Associates among the Eastern Bororo. Southwestern Journal of Anthropology 25(3):236-260. Crumley, Carol L. 1978 Toward a Locational Definition of State Systems of Settlement. American Anthropologist 78(1):59-73.

330

Cunningham, Clark E. 1973 Order in the Atoni House. In Right and Left: Essays on Dual Symbolic Classification, edited by R. Needham, pp. 204-238. University of Chicago Press, Chicago. DeJarnette, David L., and Steve B. Wimberly 1941 The Bessemer Site: Excavation of Three Mounds and Surrouding Areas near Bessemer, Alabama. Museum Paper 17, Geological Survey of Alabama, University of Alabama, Tuscaloosa. Delcourt, Hazel R., and Paul A. Delcourt 1994 Postglacial Rise and Decline of Ostrya Virginiana (Mill.) K. Koch and Carpinus caroliniana Walt. In Eastern North America: Predictable Responses of Forest Species to Cyclic Changes in Seasonality of Climates. Journal of Biogeography 21(2):137-150. 1985 Quaternary Palynology and Vegetational History of the Southeastern United States. In Pollen Records of Late-Quaternary North American Sediments, edited by V. M. Bryant, Jr., and R. G. Holloway, pp. 1–37. American Association of Stratigraphic Palynologists Foundation. Delcourt, Paul A., and Hazel R. Delcourt 1987 Long-term Forest Dynamics of the Temperate Zone. Springer, New York. Dickens, Roy S., Jr. 1976 Cherokee Prehistory: The Pisgah Phase in the Appalachian Summit Region. The University of Tennessee Press, Knoxville. Diehl, Michael W. 1992 Architecture as a Material Correlate of Mobility Strategies: Some Implications for Archaeological Interpretation. Behavior Science Research 26:1-35. 1997 Changes in Architecture and Land Use Strategies in the American Southwest: Upland Mogollon Pithouse Dwellers, A. C. 200-1000. Journal of Field Archaeology 24(2):179194. Dunbar, Robert B. 2000 Climate Variability During the Holocene: An Update. In The Way the Wind Blows: Cimate, History, and Human Action, edited by R. J. McIntosh, and S. K. McIntosh, pp. 45-88. Columbia University Press, New York. DuPratz, Le Page 1972 [1725] The History of Louisiana. Glaitor's Publishing Division, Baton Rouge. Dyer, James M. 2006 Revisiting the Deciduous Forests of Eastern North America. BioScience 56(4):341-352. Emerson, Thomas E., and Douglas K. Jackson 331

1984 The BBB Motor Site (11-Ms-595). American Bottom Archaeology FAI-270 Site Reports, Volume 6. University of Illinois Press, Urbana and Chicago. Emerson, Thomas E. 1997 Cahokia and the Archaeology of Power. University of Alabama Press, Tuscaloosa. Emerson, Thomas E., Dale L. McElrath, and Andrew C. Fortier 2000 Late Woodland Tradition and Transformations across the Midcontinent. University of Nebraska Press, Lincoln. Epenshade, Christopher 2008 Woodland Period Archaeology of Northern Georgia: 2008 Update. New South Associates, Inc., Stone Mountain, Georgia. Submitted to The Georgia Department of Transportation, Office of Environment/Location, Atlanta, Georgia. Copies available from New South Associates, Inc., Technical Report Number 1593. Farwell, Robin Y. 1981 Pit Houses: Prehistoric Energy Conservation? El Palacio 87(3):43-47. Faulkner, Charles H. 1988 Middle Woodland Community and Settlement Patterns on the Eastern Highland Rim, Tennessee. In Middle Woodland Settlement and Ceremonialism in the Mid-South and Lower Mississippi Valley, edited by R. C. Mainfort, Jr., pp. 76-98. Proceedings of the 1984 Mid-South Archaeological Conference, Pinson Mounds Tennessee, June 1984. Mississippi Department of Archives and History, Jackson. 2002 Woodland Cultures of the Elk and Duck River Valleys, Tennessee: Continuity and Change. In The Woodland Southeast, edited by D. G. Anderson, and R. C. Mainfort, Jr., pp. 185-203. University of Alabama Press, Tuscaloosa. Flannery, Kent V. 1972 The Origin of the Village as a Settlement Type in Mesoamerica and the Near East: A Comparative Study. In Man, Settlement, and Urbanism, edited by P. Ucko, R. Tringham, and G. Dimbleby, pp.25-53. Duckworth, London. 2002 The Origins of the Village Revisited: From Nuclear to Extended Households. American Antiquity 67(3):417-433. Gearing, Fred 1962 Priests and Warriors: Structures for Cherokee Politics in the Eighteenth Century. American Anthropological Association Memoir 93, Washington, D. C. Giddens, Anthony 1979 Central Problems in Social Theory: Action, Structure, and Contradiction in Social Analysis. University of California Press, Berkeley and Los Angeles. Gilbert, William H., Jr. 332

1943 The Eastern Cherokee. Smithsonian Institution, Bureau of American Ethnology Bulletin 133, Washington, D. C. Gilman, Patricia 1987 Architecture as Artifact: Pit Structures and Pueblos in the American Southwest. American Antiquity 52(3):538-564. Gonlin, Nancy 1994 Rural Household Diversity in Late Classic Copan, Honduras. In Archaeological Views from the Countryside: Village Communities in Early Complex Societies, edited by G. M. Schwartz, and S.E. Falconer, pp. 173-197. Smithsonian Institution Press, Washington, D.C. Gonzalez, Nancy L. Solien 1969 Black Carib Household Structure: A Study of Migration and Modernization. University of Washington Press, Seattle and London. Goody, Jack (editor) 1962 The Developmental Cycle in Domestic Groups. Cambridge University Press, Cambridge. Goman, Michelle and David S. Leigh 2004 Wet Early to Middle Holocene Conditions on the Upper Coastal Plain of North Carolina, USA. Quaternary Research 61(3):256-264. Gougeon, Ramie Alphonse 2002 Household Research at the Late Mississippian Little Egypt Site (9MU102). Unpublished Ph.D. dissertation, Department of Anthropology, University of Georgia, Athens. 2006 Different but the Same: Social Integration of Households in Mississippian Chiefdoms. In Leadership and Polity in Mississippian Society, edited by B. M. Butler, and P. D. Welch, pp. 178-194. Occasional Paper No. 33, Center for Archaeological Investigations, Southern Illinois University, Carbondale. 2007 An Architectural Grammar of Late Mississippian Houses in Northwest Georgia. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 49-72. University of Alabama Press, Tuscaloosa. Griffin, James (editor) 1952 Archeology of Eastern United States. University of Chicago Press, Chicago. Hally, David J. 1970 Archaeological Investigations of the Pott's Tract (9Mu103), Carter's Dam, Murray County, Georgia. Laboratory of Archaeology Series Report No. 6. University of Georgia Department of Anthropology, Athens.

333

1978 Archaeological Investigations of the Little Egypt Site (9Mu102), Murray County, Georgia. Laboratory of Archaeology Series Report No. 18. University of Georgia Department of Anthropology, Athens. 1983 The Interpretive Potential of Pottery from Domestic Contexts. Midcontinental Journal of Archaeology 8:163-196. 1984 Vessel Assemblages and Food Habits: A Comparison of Two Aboriginal Southeastern Vessel Assemblages. Southeastern Archaeology 3: 46–64. 1986 The Identification of Vessel Function: A Case Study from Northwest Georgia. American Antiquity 51:267-295. 2002 As Caves Beneath the Ground: Protohistoric Houses of the Southeastern United States. In Between Contact and Colonies: Archaeological Perspectives on the Protohistoric Southeast, edited by M. Rees and C. B. Rodning, pp. 99-109. University of Alabama Press, Tuscaloosa. 2006 The Nature of Mississippian Regional Systems. In Light on the Path: The Anthropology and History of the Southeastern Indians, edited by T. J. Pluckhahn, and R. Ethridge, pp. 26-42. University of Alabama Press, Tuscaloosa. 2008 King: The Social Archaeology of a Late Mississippian Town in Northwestern Georgia. University of Alabama Press, Tuscaloosa. Hally, David J., and Hypatia Kelly 1998 The Nature of Mississippian Towns in Georgia: The King Site Example. In Mississippian Towns and Sacred Spaces: Searching for an Architectural Grammar, edited by R. B. Lewis, and C. Stout, pp. 49-63. University of Alabama Press, Tuscaloosa. Hally, David J., and Robert C. Mainfort 2004 Prehistory of the Eastern Interior After 500 B.C. In Southeast, edited by R. D. Fogelson, pp. 265285. Handbook of North American Indians. vol. 14, W. C. Sturevant, general editor. Smithsonian Institution, Washington. Hammerstedt, Scott W. 2005 Mississippian Status in Western Kentucky: Evidence from the Annis Mound. Southeastern Archaeology 24(1):11-27. Hanson, Lee H., Jr 1970 The Jewell Site, Bn21 Barren County Kentucky. Tennessee Archaeological Society Miscellaneous Paper No. 8. University of Tennessee, Knoxville. Hanson, Julienne 2003 Decoding Homes and Houses. Cambridge University Press, Cambridge.

334

Hillier, Bill, and Julienne Hanson 1984 The Social Logic of Space. Cambridge University Press, Cambridge. Hargrave, Michael 1998 The Yuchi Town Site (1RU63), Russell County, Alabama: An Assessment of the Impacts of Looting. USA-CERL Special Report No. 98/48. US Army Corps of Engineers Construction Engineering Research Laboratories, Cultural Resources Research Center, Champaign, Illinois. Harner, Michael 1973 The Jivaro: People of the Sacred Waterfalls. Anchor Press, Garden City, New York. Harrison, S.P., J. E. Kutzbach, Z. Liu, P. J. Bartlein, B. Otto-Bliesner, D. Muhs, I. C. Prentice, and R. S. Thompson 2003 Mid-Holocene Climates of the Americas: A Dynamical Response to Changed Seasonality. Climate Dynamics 20:663-688. Hatch, James W. 1995 Lamar Period Farmsteads of the Oconee River Valley, Georgia. In Mississippian Communities and Households, edited by J. D. Rodgers, and B. D. Smith, pp. 135-155. University of Alabama Press, Tuscaloosa. Haviland, William A. 1988 Musical Hammocks at Tikal: Problems with Reconstructing Household Composition. In Household and Community in the Mesoamerican Past, edited by R. R. Wilk, and W. Ashmore, pp. 121-134. University of New Mexico Press, Albuquerque. Hayden, Brian, and Aubrey Cannon 1983 Where the Garbage Goes: Refuse Disposal in the Maya Highlands. Journal of Anthropological Archaeology 2:117–163. Hendon, Julia 1996 Archaeological Approaches to the Organization of Domestic Labor: Household Practice and Domestic Relations. Annual Review of Anthropology 25: 45–61. Hill, James K. 1968 Broken K Pueblo: Patterns of Form and Function. In New Perspectives in Archaeology, edited by S. R. Binford, and L. R. Binford, pp. 103-142. Aldine Publishing Company, Chicago. Hirth, Kenneth G. 1993 Identifying Rank and Socioeconomic Status in Domestic Contexts: An Example from Central Mexico. In Prehistoric Domestic Units in Western Mesoamerica: Studies of the Household, Compound, and Residence, edited by R. S. Santley, and K. G. Hirth, pp. 121146. CRC Press, Boca Raton.

335

Hodder, Ian 1984 Burials, Houses, Women and Men in the European Neolithic. In Ideology, Power, and Prehistory, pp. 51-68. Cambridge University Press, Cambridge. 1987 The Meaning of Discard: Ash and Domestic Space in Baringo. In Method and Theory for Activity Area Research: an Ethnoarchaeological Approach, edited by S. Kent, pp. 424448. Cambridge University Press, Cambridge. Horne, Lee 1994 Village Spaces: Settlement and Society in Northeastern Iran. Smithsonian Institution Press, Washington and London. Hugh-Jones, Stephen 1979 The Palm and the Pleiades: Initiation and Cosmology in Northwest Amazonia. Cambridge University Press, Cambridge. 1995 Inside-Out and Back-To-Front. In About the House, edited by J. Carsten, and S. Hugh Jones, pp. 226-252. Cambridge University Press, Cambridge. Hudson, Charles 1976 The Southeastern Indians. University of Tennessee Press, Knoxville. 1990 The Juan Pardo Expeditions: Exploration of the Carolinas and Tennessee, 1566-1568. Smithsonian Institution Press, Washington D.C. Hutchinson, Dale L., Clark Spencer Larsen, Margaret J. Schoeninger, and Lynette Norr 1998 Regional Variation in the Pattern of Maize Adoption and Use in Florida and Georgia. American Antiquity 63(3):397-416. Ingmanson, J. Earl 1964 Archaeology of the South Plateau, Ocmulgee National Monument. United States Department of the Interior, National Park Service Southeast Region. Richmond, Virginia. Manuscript on file at the Georgia Archaeological Site File, Athens. Inomata, Takeshi 2008 Warfare and the Fall of a Fortified Center: Archaeological Investigations at Aguateca. Vanderbilt University Press, Nashville, Tennessee. Jenkins, Ned J., and H. Blaine Ensor 1981 The Gainesville Lake Area Excavations. Archaeological Investigations in the Gainesville Lake Area of the Tennessee-Tombigbee Waterway, Vol. 1. Report of Investigations No. 11, Office of Archaeological Research, University of Alabama, Tuscaloosa. Kapches, Mima 1990 The Spatial Dynamics of Ontario Iroquoian Longhouses. American Antiquity 55(1):49-67. Keel, Bennie C. 336

1976 Cherokee Archaeology: A Study of the Appalachian Summit. University of Tennessee Press, Knoxville. 2007 The Ravensford Tract Archeological Project. National Park Service Oversight Report. Southeastern Archaeological Center, Tallahassee. Kelly, John E. 1990 Range Site Community Patterns and the Mississippian Emergence. In The Mississippian Emergence, edited by B. D. Smith, pp. 67-112, Smithsonian Institution Press, Washington, D.C. Kelly, John E., Andrew C. Fortier, Steven J. Ozuck, and Joyce A. Williams 1987 The Range Site: Archaic through Late Woodland Occupations. University of Illinois Press, Urbana. Kelly, Robert L., Lin Poyer, and Bram Tucker 2005 An Ethnoarchaeological Study of Mobility, Architectural Investment, and Food Sharing among Madagascar's Mikea. American Anthropologist 107:403-416. Kent, Susan 1983 The Differentiation of Navajo Culture, Behavior, and Material Culture: A Comparative Study in Culture Change. Ethnology 22(1):81-91. 1984 Analyzing Activity Areas: An Ethnoarchaeological Study of the Use of Space. University of New Mexico Press, Albuqurque. 1990 A Cross-Cultural Study of Segmentation, Architecture, and the Use of Space. In Domestic Architecture and the Use of Space: An Interdisciplinary Cross-Cultural Study, edited by S. Kent, pp. 127-152. Cambridge University Press, Cambridge. 1992 Studying Variability in the Archaeological Record: An Ethnoarchaeological Model for Distinguishing Mobility Patterns. American Antiquity 57:635–659. Kidder, Tristam R. 2002 Woodland Period Archaeology of the Lower Mississippi Valley. In The Woodland Southeast, edited by D. G. Anderson and Robert C. Mainfort, Jr., pp. 66-90. University of Alabama Press, Tuscaloosa. Kline, Gerald W., Gary D. Crites, and Charles H. Faulkner 1982 The McFarland Project: Early Middle Woodland Settlement and Subsistence in the Upper Duck River Valley in Tennessee. Miscellaneous Paper Number 8. Tennessee Anthropological Association, Knoxville. Knight, Vernon J., Jr.

337

1985 Tukabatchee : Archaeological Investigations at an Historic Creek Town, Elmore County, Alabama. Office of Archaeological Research, Alabama State Museum of Natural History, University of Alabama, Tuscaloosa. 2007 Conclusions: Taking Architecture Seriously. In Architectural Variability in the Southeast, edited by C. H. Lacquemont, pp. 186-192. University of Alabama Press, Tuscaloosa. Knight, Vernon James, Jr., and Vincas P. Steponaitis (editors) 1998 A New History of the Moundville Chiefdom. Smithsonian Institution Press, Washington D.C. Kowalewski, Stephen A. 1995 Large-Scale Ecology in Aboriginal Eastern North America. In Native American Interactions: Multiscalar Analyses and Interpretations of the Eastern Woodlands, edited by M. S. Nassaney and K. E. Sassaman, pp. 147-173. The University of Tennessee Press, Knoxville. 2006 Coalescent Societies. In Light on the Path: The Anthropology and History of the Southeastern Indians, edited by T. J. Pluckhahn and R. Ethridge, pp. 95-122. University of Alabama Press, Tuscaloosa. Kramer, Carol 1982 Village Ethnoarchaeology: Rural Iran in Archaeological Perspective. Academic Press, New York. Lacquement, Cameron H. 2004 How to Build a Mississippian House: A Study of Domestic Architecture in West-Central Alabama. Unpublished Masters Thesis, University of Alabama. 2007a Introduction to Architectural Variability in the Southeast. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 1-11. University of Alabama Press, Tuscaloosa. 2007b Typology, Chronology, and Technological Changes of Mississippian Domestic Architecture in West-Central Alabama. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 49-72. University of Alabama Press, Tuscaloosa. Lamoreaux, Heidi Jo Kieffer 1999 Human-Environmental Relationships in the Coastal Plain of Georgia Based on High Resolution Paleoenvironmental Records from Three Peat Deposits. Unpublished Ph.D. dissertation, Department of Geography, University of Georgia, Athens. LeBlanc, Steven 1971 An Addition to Naroll's Suggested Floor Area and Settlement Population Relationship. American Antiquity 36:210-211.

338

Ledbetter, Jerald 1997 The Bull Creek Site, 9ME1, Muscogee County, Georgia. Occasional papers in Cultural Resource Management, No. 9. Georgia Department of Transportation, Office of Environment/Location, Atlanta. Ledbetter, Jerald R., Lisa D. O'Steen, and Scott Jones 2009 The Late Archaic to Early Woodland Transition in Northwest Georgia: Evidence for Terminal Archaic (ca. 1100 - 600 B.C.) Period Occupation in the Region. Occasional Papers in Cultural Resource Management No. 14. Georgia Department of Transportation Office of Environment/Location, Atlanta. Leigh, David S. 2006 Terminal Pleistocene Braided to Meandering Transition in Rivers of the Southeastern USA. CATENA 66(1-2):155-160. Leigh, David S., and T. P. Feeney 1995 Paleochannels Indicating Wet Climate and Lack of Response to Lower Sea Level, Southeast Georgia. Geology 23:687-690. Leigh, David S. and Paul A. Webb 2006 Holocene Erosion, Sedimentation, and Stratigraphy at Raven Fork, Southern Blue Ridge Mountains, USA. Geomorphology 78(1-2):161-177. Lewis, R. Barry 1996 Mississippian Farmers. In Kentucky Archaeology, edited by R. B.Lewis, pp. 127-160. The University Press of Kentucky, Lexington. Lewis, R. Barry, and Charles Stout (editors) 1998 Mississippian Towns and Sacred Spaces: Searching for an Architectural Grammar. University of Alabama Press, Tuscaloosa. Lewis, Thomas M. N., and Madeline Kneberg 1946 Hiwassee Island: An Archaeological Account of Four Tennessee Indian People. University of Tennessee Press, Knoxville. Lewis, Thomas N., and Madeline D. Kneberg Lewis (editors) 1995 The Prehistory of the Chickamauga Basin in Tennessee. The University of Tennessee Press, Knoxville. Little, Keith J. 2001 Late Holocene Climate Fluctuations and Culture Change in Southeastern North America. Southeastern Archaeology 22(3):9-32. Longacre, William A. 1968 Some Aspects of Prehistoric Society in East-Central Arizona. In New Perspectives in Archaeology, edited by S. Binford, and L. R. Binford, pp. 89-102. Aldine Publishing Company, Chicago. 339

Lowery, Darrin 2001 Archaeological Survey of the Chesapeake Bay Shorelines associated with Accomack County and Northampton County, Virginia. Virginia Department of Historic Resources Survey and Planning Report Series 6. Virginia Department of Historic Resources, Richmond. Lyons, Diana 1989 Men's Houses, Women's Spaces: The Spatial Ordering of Households in Duolo, North Cameroon. In Households and Communities: Proceedings of the 21st Annual Chacmool Conference, edited by S. MacEachern, D. J. W. Archer, and R. D. Garvin, pp. 28-34. University of Calgary Archaeology Association, Calgary. Mainfort, Robert C., Jr. 1988 Middle Woodland Ceremonialism at Pinson Mounds. American Antiquity 53(1):158-173. Marcoux, Jon Bernard 2007 On Reconsidering Display Goods Production and Circulation in the Moundville Chiefdom. Southeastern Archaeology 26:232-245. 2008 Cherokee Households and Communities in the English Contact Period, A.D. 1670-1740. Unpublished Ph.D. dissertation, Department of Anthropology, University of North Carolina, Chapel Hill. Markin, Julie Gayle 2007 Woodstock: The Rise of Political Complexity in Northern Georgia. Unpublished Ph.D. dissertation, Department of Anthropology, University of Georgia, Athens. McConaughy, Mark A. 2007 A Comparison of Burned Mississippian Houses from Illinois. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 101-116. University of Alabama Press, Tuscaloosa. McEwan, Bonnie G., and John H. Hann 2000 Reconstructing a Spanish Mission: San Luis de Talimali. OAH Magazine of History 14(4):16-19. McGimsey, Charles R., and Michael D. Wiant 1986 Middle Woodland Features. In Woodland Period Occupations of the Napoleon Hollow Site in the Lower Illinois Valley, edited by M. D. Wiant and C. R. McGimsey, pp. 114-170. Center for American Archaeology Kampsville Archaeological Center, Kampsville, Illinois. Meentemeyer, Vernon, and Elgene O. Box 1987 Scale Effects in Landscape Studies. In Landscape Heterogeneity and Disturbance, edited by M. G. Turner, pp. 15-34. Springer-Verlag, New York.

340

Milner, George R. 1999 Warfare in Prehistoric and Early Historic North America. Journal of Archaeological Research 7(2):105-151. Mooney, James 1900 Myths of the Cherokee. Annual Report of the Bureau of American Ethnology, 1897-98. Government Printing Office, Washington, D.C. Moore, David G. 2002 Pisgah Phase Village Evolution at the Warren Wilson Site. In The Archaeology of Native North Carolina: Papers in Honor of H. Trawick Ward, edited by J. M. Eastman, C. D. Rodning, and E. A. Boudreaux, III, pp. 76-83. SEAC Special Publications 7. Southeastern Archaeological Conference, Biloxi, Mississippi. Muller, John 1978 The Kincaid System: Mississippian Settlement in the Environs of a Large Site. In Mississippian Settlement Patterns, edited by B. D. Smith, pp. 269-292. Academic Press, New York. 1997 Mississippian Political Economy. Plenum Press, New York. Murdock, George P. 1949 Social Structure. MacMillan Company, New York. 1967 Ethnographic Atlas. University of Pittsburgh Press, Pittsburgh. Naroll, Raoul 1962 Floor Area and Settlement Population. American Antiquity 27:587-89. Netting, Robert McC 1965 Household Organization and Intensive Agriculture: The Kofyar Case. Africa 35:422-429. O‘Brien, Michael J., and James W. Cogswell 2001 The Construction and Abandonment of Powers Phase Structures. In Mississippian Community Organization: The Powers Phase in Southeastern Missouri, edited by M. J. O‘Brien, pp. 141-180. Kluwer Academic/Plenum Publishers, New York. O‘Brien, Michael J. and Timothy K. Perttula 2001 Community Organization and Dates of Occupation. In Mississippian Community Organization: The Powers Phase in Southeastern Missouri, edited by M. J. O‘Brien, pp. 99-140. Kluwer Academic/Plenum Publishers, New York. O'Brien, Patricia J. 1972 Urbanism, Cahokia, and Middle Mississippian. Archaeology 25:189-197. 1989 Cahokia: The Political Capital of the "Ramey" State? North American Archaeologist 10:275-292. 341

1991 Early State Economics: Cahokia, Capital of the Ramey State. In Early State Economics, edited by H. J. M. Glaessen and P. van de Velde, pp. 143-175. Political and Legal Anthropology 8, Transaction Publishers, New Brunswick, New Jersey. Oswald, Dana 1987 The Organization of Space in Residential Buildings: A Cross-Cultural Perspective. In Method and Theory for Activity Area Research: An Ethnoarchaeological Approach, edited by S. Kent, pp. 295-344. Columbia University Press, New York. Pauketat, Timothy R. 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 16:1-15. 1998 The Archaeology of Downtown Cahokia: The Tract 15a and Dunham Tract Excavations. Illinois Transportation Archaeological Research Program Studies in Archaeology No. 1. Univerisity of Illinois, Urbana-Champaign. 2000 Politicization and Community in the Pre-Columbian Mississippi Valley. In The Archaeology of Communities: A New World Perspective, edited by M. A. Canuto, and J. Yaeger, pp.16-43. Routledge, London. 2003 Resettled Farmers and the Making of a Mississippian Polity. American Antiquity 68(1):3966. 2004 Ancient Cahokia and the Mississippians. Cambridge University Press, Cambridge. 2007 Chiefdoms and Other Archaeological Delusions. AltaMira, Lanham, Maryland. Pauketat, Timothy R. and Susan M. Alt 2005 Agency in a Postmold? Physicality and the Archaeology of Culture Making. Journal of Archaeological Method and Theory 12: 213–236 Pauketat, Timothy R. and Thomas E. Emerson 1991 The Ideology of Authority and the Power of the Pot. American Anthropologist 93(4):919941. Peregrine, Peter N. 1992 Social Change in the Woodland-Mississippian Transition: A Case Study of Household and Community Patterns in the American Bottom. North American Archaeologist 13: 131– 147. Pluckhahn, Thomas J.

342

1996 Joseph Caldwell‘s Summerour Mound (9FO16) and Woodland Platform Mounds in the Southeastern United States. Southeastern Archaeology 15(2):191-211. 2003 Kolomoki: Settlement, Ceremony, and Status in the Deep South, A.D. 350 to 750. University of Alabama Press, Tuscaloosa. 2010 Household Archaeology in the Southeastern United States: History, Trends, and Challenges. Journal of Archaeological Research 18(4):331-385. Pluckhahn, Thomas J., Matthew J. Compton, and Mary T. Bonhage-Freund 2006 Archaeological Correlates of Small-Scale Feasting: Evidence from the Woodland Period Site of Kolomoki in Georgia. Journal of Field Archaeology 31: 263–284. Polhemus, Richard R. 1987 The Toqua Site: A Late Mississippian Dallas Phase Town. University of Tennessee Department of Anthropology Report of Investigations No. 41, Tennessee Valley Authority Publications in Anthropology No. 44. Tennessee Valley Authority, Knoxville. 1990 Dallas Phase Architecture and Sociopolitical Structure. In Lamar Archaeology: Mississippian Chiefdoms of the Deep South, edited by M. Williams and G. Shapiro, pp. 125-138. University of Alabama Press, Tuscaloosa 1998 Activity Organization in Mississippian Households: A Case Study from the Loy Site in East Tennessee. Unpublished Ph.D dissertation, Department of Anthropology, University of Tennessee, Knoxville. Poplin, Eric C. 1990 Prehistoric Settlement in the Dog River Valley: Archaeological Data Recovery at 9DO34, 9DO39, and 9DO45, Douglas County, Georgia. Brockington and Associates Technical Report. Copies available from UGA Laboratory of Archaeology, Athens, GA. Prentice, Guy 1983 Cottage Industries: Concepts and Implications. Midcontinental Journal of Archaeology 8:1–16. 1985 Economic Differentiation among Mississippian Farmsteads. Midcontinental Journal of Archaeology 10:77-122. Price, James E., and James B. Griffin 1979 The Snodgrass Site of the Powers Phase of Southeast Missouri. Reports on Anthropology, Vol. 8, no. 1. University of Kentucky, Lexington. Prokopetz, A. Wayne 1974 An Analysis of Post Houses, Site 1Bi4, Macon, Georgia. Unpublished manuscript on file, University of Georgia, Laboratory of Archaeology, Athens. Rapoport, Amos 343

1969 House Form and Culture. Prentice Hall, Englewood Cliffs, New Jersey. Renfrew, Colin 1986 Introduction: Peer Polity Interaction and Socio-Political Change. In Peer Polity Interaction and Socio-Political Change, edited by C. Renfrew and J. Cherry, pp. 1-18. Cambridge University Press, Cambridge. Riggs, Brett H. 1999 Removal Period Cherokee Households in Southwestern North Carolina: Material Perspectives on Ethnicity and Cultural Differentiation. Unpublished Ph.D. dissertation, Department of Anthropology, University of Tennessee, Knoxville. 2009 A Synthesis of Documentary and Archaeological Evidence for Early 18th Century Cherokee Villages and Structures: Data for the Reconstruction of the Tsa-La-Gi Ancient Village, Cherokee Heritage Center, Park Hill, Oklahoma. Research Laboratories of Archaeology, University of North Carolina at Chapel Hill. Riggs, Brett H., Norman D. Jefferson, and George M. Crothers 1998 Hiwassee Old Town Draft Manuscript. Tennessee Department of Conservation, Division of Archaeology, Knoxville. Rodning, Christopher B. 2004 The Cherokee Town at Coweeta Creek. Unpublished Ph.D. dissertation, Department of Anthropology, University of North Carolina, Chapel Hill. 2007 Building and Rebuilding Cherokee Houses and Townhouses in Southwestern North Carolina. In The Durable House: Architecture, Ancestors, and Origins, edited by R. A. Beck, Jr., pp. 464-484. Southern Illinois University, Center for Archaeological Investigations Occasional Paper 35, Carbondale. Rogers, J. Daniel 1995 The Archaeological Analysis of Domestic Organization. In Mississippian Communities and Households, edited by J. D. Rogers and B. D. Smith, pp. 1-31. University of Alabama Press, Tuscaloosa. Rogers, J. Daniel, and Bruce D. Smith (editors) 1995 Mississippian Communities and Households. The University of Alabama Press, Tuscaloosa. Rolingson, Martha Ann 2002 Plum Bayou Culture of the Arkansas-White River Basin. In The Woodland Southeast, edited by D. G. Anderson and R. C. Mainfort, Jr., pp. 20-43. University of Alabama Press, Tuscaloosa. Rolingson, Martha Ann, and Douglas W. Schwartz

344

1966 Late Paleo-Indian and Early Archaic Manifestations in Western Kentucky. Studies in Anthropology No. 3. University of Kentucky Press, Lexington. Russ, Kurt C., and Jefferson Chapman 1983 Archaeological Investigations at the Eighteenth Century Overhill Cherokee Town of Mialoquo. The University of Tennessee, Department of Anthropology Report of Investigations Number 37. The Tennessee Valley Authority Publications in Anthropology Number 36. University of Tennessee Press, Knoxville. Saenger, Casey, Thomas Cronin, Robert Thunell, and Cheryl Vann 2006 Modelling River Discharge and Precipitation from Estuarine Salinity in the Northern Chesapeake Bay: Application to Holocene Paleoclimate. The Holocene 16(4):467-477. Saunders, Rebecca 1998 Swift Creek Phase Design Assemblages from Two Sites on the Georgia Coast. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams and D. T. Elliot, pp 154-180. University of Alabama Press, Tuscaloosa. Scarry and Scarry 2005 Native American ‗Garden Agriculture‘ in Southeastern North America. World Archaeology 37(2):259-274. Schroeder, Sissel 2004 Current Research on Late Precontact Societies of the Midcontinental United States. Journal of Archaeological Research 12: 311–372. 2005 Reclaiming New Deal Era Civic Archaeology: Exploring the Legacy of William S. Webb and the Jonathan Creek Site. CRM: The Journal of Heritage Stewardship 2:53-71. Schroedl, Gerald F. (editor) 1986 Archaeology at Chota-Tanase. Univeristy of Tennessee Department of Anthropology Report of Investigations No. 38. University of Tennessee, Knoxville. 1998 Mississippian Towns in the Eastern Tennessee Valley. In Mississippian Towns and Sacred Spaces: Searching for an Architectural Grammar, edited by R. B. Lewis and C. Stout, pp. 64-92. University of Alabama Press, Tuscaloosa. Schroedl, Gerald F., R. P. Stephen Davis, Jr., and C. Clifford Boyd, Jr. 1985 Archaeological Contexts and Assemblages at Martin Farm. University of Tennesee, Department of Anthropology, Report of Investigations Number 39. Tennessee Valley Authority Publications in Anthropology Number 37. University of Tennessee Press, Knoxville. Scott, Robert J.

345

2007 Interpreting Changes in Historic Creek Household Architecture at the Turn of the Nineteenth Century. In Architectural Variability in the Southeast, edited by C. H. Lacquement, pp. 166-185. University of Alabama Press, Tuscaloosa. Shapiro, Gary 1984 Ceramic Vessels, Site permanence, and Group Size: A Mississippian Example. American Antiquity 49: 696–712. Sheets, Payson 2000 Provisioning the Ceren Household: The Vertical Economy, Village Economy, and Household Economy in the Southeastern Maya Periphery. Ancient Mesoamerica 11:217230. Sheldon, Craig 1990 The Council House at Fusihatchee. In Archaeological Investigations at the Early Historic Creek Town of Fusihatchee (Phase I, 1988-89), edited by G. Waselkov, J. Cottier, and C. Sheldon, pp 45-57. Report submitted to the National Science Foundation, Washington, D.C. Shumate, M. Scott, Brett H. Riggs, and Larry P. Kimball 2005 The Alarka Farmstead Site: Archaeological Investigations at a Mid-Seventeenth-Century Cherokee Winter House/Summer House Complex, Swain County, North Carolina. Appalachian State University Laboratories of Archaeological Science and Research Laboratories of Archaeology, Boone and Chapel Hill, North Carolina. Submitted to Rodney J. Snedeker, Forest Archaeologist, National Forests in North Carolina. Simon, Mary L. 2002 Red Cedar, White Oak, and Bluestem Grass: The Colors of Mississippian Construction. Midcontinental Journal of Archaeology 27(2):273-308. Smith, Bruce D. 1978 Prehistoric Patterns of Human Behavior: A Case Study in the Mississippi Valley. Academic Press, New York. 1986 The Archaeology of the Southeastern United States: From Dalton to DeSoto, 10,500-500 B.P. In Advances in World Prehistory, Vol. 5, edited by F. Wendorf and A. Close, pp. 192. Academic Press, Orlando. 1992 Hopewellian Farmers of Eastern North America. In Rivers of Change: Essays on Early Agriculture in Eastern North America, edited by B. D. Smith, pp. 201-248. Smithsonian Institution Press, Washington, D.C. Smith, Bruce D. (editor) 1978 Mississippian Settlement Patterns. Academic Press, New York. Smith, Carol 346

1984 Local History in Global Context: Social and Economic Transitions in Western Guatemala. Comparative Studies in Society and History 26(2):193-228. Smith, Marvin T. 1994 Archaeological Investigations at the Dyar Site, 9GE5. Laboratory of Archaeology Series, Report No. 32, University of Georgia, Athens. Smith, Marvin T., and Mark Williams 1994 Mississippian Mound Refuse and Disposal Patterns and Implications for Archaeological Research. Southeastern Archaeology 13(1):27-35. Smith, Michael E. 1987 Household Possessions and Wealth in Agrarian States: Implications for Archaeology. Journal of Anthropological Archaeology 6:297-335. Snow, Frankie 1998 Swift Creek Design Investigations: The Hartford Case. In A World Engraved: Archaeology of the Swift Creek Culture, edited by M. Williams and D. T. Elliot, pp 61-98. University of Alabama Press, Tuscaloosa. Speck, Frank G. 1909 Ethnology of the Yuchi Indians. University of Pennsylvania Museum Paper 1(1). Philadelphia. Steadman, Sharon R. 2004 Heading Home: The Architecture of Family and Society in Early Sedentary Communities on the Anatolian Plateau. Journal of Anthropological Research 60(4):515-558. Steere, Benjamin A., and Aaron Deter-Wolf 2009 Postholes and Structures. In The Fernvale Site (40WM51): Late Archaic and Multicomponent Occupations Along the South Harpeth River in Williamson County, Tennessee, edited by A. Deter-Wolf and M. C. Moore. Tennessee Department of Environment and Conservation, Division of Archaeology, Nashville, in preparation. Sullivan, Lynne P. 1987 The Mouse Creek Household. Southeastern Archaeology 6(1):16-29. 1995 Mississippian Community and Household Organization in Eastern Tennessee. In Mississippian Communities and Households, edited by J. D. Rogers and B. D. Smith, pp. 99-123. University of Alabama Press, Tuscaloosa. Swanton, John R. 1946 The Indians of the Southeastern United States. Bulletin 137. Smithsonian Institution, Washington, D.C. Vogt, Evan 347

1969 Zinacantán: A Maya Community in the Highlands of Chiapas. Belknap Press of Harvard University Press, Cambridge. Ward, H. Trawick, and R. P. Stephen Davis 1993 Indian Communities on the North Carolina Piedmont, A.D. 1000 to 1700. Research Laboratories of Anthropology, Monograph No. 2. University of North Carolina, Chapel Hill. 1999 Time Before History: The Archaeology of North Carolina. University of North Carolina Press, Chapel Hill. Warrick, Gary A. 1988 Estimating Ontario Iroquoian Village Duration. Man in the Northeast 36: 21-60. Waselkov, Gregory A. 1994 The Macon Trading House and Early European-Indian Contact in the Colonial Southeast. In Ocmulgee Archaeology, 1936-1986, edited by D. J. Hally, pp. 190- 196. University of Georgia Press, Athens. Waselkov, Gregory A., and Kathleen E. Holland Braund (editors) 2002 William Bartram on the Southeastern Indians. University of Nebraska Press, Lincoln. Watts, W. A. 1980 The Late Quaternary Vegetation History of the Southeastern United States. Annual Review of Ecology and Systematics 11:387-409. Wauchope, Robert 1966 Archaeological Survey of Northern Georgia with a Test of some Cultural Hypotheses. Memoirs of the Society for American Archaeology No. 21. Society for American Archaeology, Salt Lake City. Webb, William S. 1952 The Jonathan Creek Village, Site 4, Marshall County, Kentucky. Reports on Anthropology, Vol. 8, no. 1. University of Kentucky, Lexington. Wiessner, Polly 2002 The Vines of Complexity: Egalitarian Structures and the Institutionalization of Inequality among the Enga. Current Anthropology 43(2):233-269. Welch, Paul 1991 Moundville’s Economy. University of Alabama Press, Tuscaloosa. 1994 The Occupational History of the Bessemer Site. Southeastern Archaeology 13(1):1-26. Wetmore, Ruth Y.

348

1990 The Ela Site (31SW5): Archaeological Data Recovery of Connestee and Qualla Phase Occupations at the East Elementary School Site, Swain County, North Carolina. Western Carolina Archaeology Laboratory, Cullowhee, North Carolina. 2002 The Woodland Period in the Southern Appalachian Summit of Western North Carolina and the Ridge and Valley Province of Eastern Tennessee. In The Woodland Southeast, edited by D. G. Anderson and R. C. Mainfort, Jr., pp. 249-270. University of Alabama Press, Tuscaloosa. Wilk, Richard R. 1983 Little House in the Jungle: The Causes of Variation in House Size among Modern Kekchi Maya. Journal of Anthropological Archaeology 2(2):99-116. 1984 Households in Process: Agricultural Change and Domestic Transformation Among the Kekchi Maya of Belize. In Households: Comparative and Historical Studies of Domestic Groups, edited by R. M. Netting, R. R. Wilk, and E. J. Arnould, pp. 217-244. University of California Press, Berkeley. 1990 The Built Environment and Consumer Decisions. In Domestic Architecture and the Use of Space: An Interdisciplinary Cross-Cultural Study, edited by S. Kent, pp. 43-72. Cambridge University Press, Cambridge. Wilk, Richard R., and Lisa C. Cliggett 2007 Economies and Cultures: Foundations of Economic Anthropology. Westview Press, Boulder. Wilk, Richard R., and Robert McC. Netting 1984 Households: Changing Forms and Functions. In Households: Comparative and Historical Studies of the Domestic Group, edited by R. Netting, R. R. Wilk, and E. Arnould, pp. 128. University of California Press, Berkeley. Wilk, Richard R., and William L. Rathje 1982 Household Archaeology. American Behavioral Scientist 25(6):617-640. Williams, J. Raymond 1974 The Baytown Phases in the Cairo Lowland of Southeast Missouri. The Missouri Archaeologist 36:1-109. Williams, Mark 1995 Chiefly Compounds. In Mississippian Households and Communities, edited by J. D. Rogers and B. D. Smith, pp. 124-134. University of Alabama Press, Tuscaloosa. 2002 Archaeological Excavations at Little River: The 1998-2000 Seasons. LAMAR Institute Publication No. 49. LAMAR Institute, Athens. Williams, Mark, and Don Evans 349

1993 Archaeological Excavations at the Bullard Landing Site (9TW1). LAMAR Institute Publication No. 24. LAMAR Institute, Athens. Williams, Stephen 1963 The Eastern United States. In Early Indian Farmers and Villages and Communities, edited by W. G. Haag, pp. 267-325. U.S. Department of the Interior, National Park Service, Washington, D.C. Wilson, Gregory D. 2008 The Archaeology of Everyday Life at Early Moundville. University of Alabama Press, Tuscaloosa. Wu, Jianguo, and Orie L. Loucks 1995 From Balance of Nature to Hierarchical Patch Dynamics: A Paradigm Shift in Ecology. The Quarterly Review of Biology 70:439-466. Yanagiakso, Sylvia 1979 Family and Household: An Analysis of Domestic Groups. Annual Review of Anthropology 8:161-295. Yang, Martin C. 1945 A Chinese Village: Taitou, Shantung Province. Columbia University Press, New York.

350

54

1GR1X1

1GR1X1

AL

33

55 48 49 50 51 44 45 46 47 28

1PI61 9GE1754 9GE1760 9GE1776 9GE1781 9GE333 9GE901 9GE903 9GE922 9WH120

1PI61 9GE1754 9GE1760 9GE1776 9GE1781 9GE333 9GE901 9GE903 9GE922 9WH120

AL GA GA GA GA GA GA GA GA GA

34 157 151 139 149 139 137 143 142 421

17

31SW173

Alarka

NC

530

68

40CF108

Banks III

TN

249

69

40CF111

TN

250

8

11MS595

Banks V BBB Motor Site

IL

127

Coastal Plain Coastal Plain Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Piedmont Blue Ridge Blue Ridge Interior Low Plateau Interior Low Plateau Central Lowlands

Plaza

Mounds (n)

Palisade

Paired

Public

Structures (n)

Site Size (ha)

Avg. Ann. Temp. (F)

Avg. Win. Temp. (F)

Phys. Prov.

Elev (m AMSL)

State

Site Name

State Site No

SiteID

APPENDIX A. SITE DATA TABLE

28

63

0.17

1

N

N

N

0

N

28 28 28 28 28 28 28 28 28 28

63 63 63 63 63 63 63 63 63 57

0.38 1.08 0.85 1.04 1.14 8.10 1.75 10.20 0.04 0.60

4 2 4 3 2 2 5 15 3 2

N N N N N N N N N N

N N N N N N N N N Y

N N N N N N N N N N

0 0 0 0 0 0 0 0 0 0

N N N N N N N N N N

19

54

0.25

2

N

Y

N

0

N

28

57

0.10

3

N

Y

N

0

N

28

57

0.13

5

N

N

1

N

19

57

0.50

21

Y

N

0

N

351

N

1JE12

Bessemer

AL

127

9

9TO45, 9TO48, 9TO49

Brasstown Valley

GA

556

IL

128

TN

275

NC

617

27

Ridge and Valley

Plaza

Mounds (n)

Palisade

Paired

Public

Structures (n)

Site Size (ha)

Avg. Ann. Temp. (F)

Avg. Win. Temp. (F)

Phys. Prov.

Elev (m AMSL)

State

Site Name

State Site No

SiteID 61

28

63

10.00

28

Y

N

N

3

Y

Blue Ridge Central Lowlands Ridge and Valley

19

54

2.40

52

Y

Y

Y

0

Y

19

57

890.00

253

Y

Y

Y

100

Y

28

57

2.00

34

Y

Y

Y

1

Y

19

54

1.20

16

Y

Y

N

0

Y

28 28

57 63

1.00 2.00

29 1

Y N

N N

Y N

1 0

Y N

19 28 19

57 63 54

8.60 2.13 0.60

7 6 15

0 Y N

N Y N

N N N

0 1 0

N Y N

19

57

0.10

2

N

N

N

0

N

35

11MS2 40MR2, 40MR62

21

31MA34

Cahokia ChotaTanase Coweeta Creek

10 6

9DO39

Dallas Dog River

TN GA

208 230

70 71 37

40TR27 9GE5 31SW5

Duncan Tract Dyar Ela

TN GA NC

140 132 548

36

40WM51

TN

175

64 33

31HW2 9ME50

NC GA

787 73

Blue Ridge Piedmont

19 28

54 63

5.00 2.40

4 1

Y N

N N

N N

3 0

N N

24

9CK9

Fernvale Garden Creek Go Kart Site Hickory Log

Blue Ridge Ridge and Valley Piedmont Interior Low Plateau Piedmont Blue Ridge Interior Low Plateau

GA

274

28

57

0.27

16

N

N

Y

0

N

18

40LO45

TN

226

28

57

0.14

2

N

N

N

0

N

4

40MG31

TN

208

28

57

4.60

78

Y

Y

Y

7

Y

19

40PK3

Higgs Hiwassee Island Hiwassee Old Town

Piedmont Ridge and Valley Ridge and Valley

TN

219

Blue Ridge

28

57

140.00

6

N

N

N

0

N

352

23MI8

Hoecake

MO

92

42

31OR231a

Jenrette

NC

152

59 58

15BN21 9CK62

KY GA

196 282

39

40LD208

Jewell Kellog KimberlyClark

TN

228

Kincaid

IL

102

73 1

9FL5

King

GA

172

29

9ER1

Kolomoki

GA

88

7

9MU102

GA

212

41

31RK1

Little Egypt Lower Saratown

NC

154

53

40JE10

TN

268

40

9BI1

Loy Macon Plateau

GA

101

16

40MR20

Martin Farm

TN

247

14

40CF48

McFarland

TN

297

15 38

40MR3 31CH452

Mialoquo Mitchum

TN NC

264 115

Plaza

Mounds (n)

Palisade

Paired

Public

Structures (n)

Site Size (ha)

Avg. Ann. Temp. (F)

Avg. Win. Temp. (F)

Phys. Prov.

Elev (m AMSL)

State

Site Name

State Site No

SiteID 72

Mississippi Alluvial Plain

19

57

81.00

11

N

N

N

2

Y

Piedmont Interior Low Plateau Piedmont Ridge and Valley Coastal Plain Ridge and Valley Coastal Plain Ridge and Valley

28

57

0.15

3

N

N

Y

0

N

19 28

57 57

3.50 0.26

22 1

Y N

N N

N N

1 0

N N

19

57

0.03

3

N

N

N

0

N

19

57

50.00

127

Y

U

Y

27

Y

28

63

2.30

65

Y

N

Y

0

Y

36

63

100.00

U

N

N

8

Y

28

57

2.00

3

N

N

N

2

Y

Piedmont Ridge and Valley

28

57

0.04

2

N

N

Y

0

N

28

57

2.50

20

Y

N

Y

1

Y

Piedmont Ridge and Valley Interior Low Plateau Ridge and Valley Piedmont

28

63

60.00

12

Y

N

Y

8

Y

28

57

1.10

12

Y

N

Y

2

N

28

57

2.65

5

N

N

N

0

N

28 28

57 57

10.00 0.50

8 1

Y N

Y N

N Y

0 0

Y N

353

15HK49

Morris

KY

149

65

1TU500

Moundville Napoleon Hollow

AL

47

IL

132

Potts' Tract

GA

212

Ravensford Rivermoore Rucker's Bottom

NC GA

610 295

GA

128

Rymer

TN

209

56 34 2 75

9MU103 31SW78, 9SW136 9GW70

20

9EB91

12

Plaza

Mounds (n)

Palisade

Paired

Public

Structures (n)

Site Size (ha)

Avg. Ann. Temp. (F)

Avg. Win. Temp. (F)

Phys. Prov.

Elev (m AMSL)

State

Site Name

State Site No

SiteID 60

Interior Low Plateau Coastal Plain Central Lowlands Ridge and Valley

19

57

0.60

12

U

N

Y

0

Y

28

63

70.00

152

Y

N

Y

29

Y

28

54

15.00

1

U

N

N

0

N

28

57

0.80

3

N

N

N

0

N

Blue Ridge Piedmont

19 28

54 57

13.80 0.10

111 3

Y N

Y N

N N

0 0

Y N

Piedmont Ridge and Valley Mississippi Alluvial Plain Piedmont Ridge and Valley Piedmont

28

63

0.80

12

Y

N

Y

0

Y

28

57

1.94

23

Y

Y

Y

0

Y

19 28

57 57

0.82 1.00

91 1

Y U

N N

Y N

0 1

Y N

28 28

57 63

1.90 0.15

133 42

Y Y

Y Y

Y Y

2 1

Y Y

28 19

57 54

10.00 0.50

10 1

N N

Y N

N N

0 0

N N

36

63

1.20

2

N

N

N

0

N

28

57

0.90

6

Y

N

N

1

N

19

54

1.30

17

N

N

Y

0

Y

62 74

23BU21b 9FO16

Snodgrass Summerour

MO GA

89 326

5 23

40MR6 31MG2

Toqua Town Creek

TN NC

248 69

66 63

40BT89, 40BT90, 40BT91 31JK12

Townsend Tuckasegee

TN NC

309 654

52

1EE32

AL

63

31

9BR3

GA

212

3

31BN29

Tukabatchee Two Run Creek Warren Wilson

Blue Ridge Blue Ridge Coastal Plain Ridge and Valley

NC

646

Blue Ridge

354

43 40LN16 Yearwood TN 263

32 1RU63 Yuchi Town AL 63

Interior Low Plateau Coastal Plain 28 57 1.01 14 Y Y N 0 N

28 63 2.00 2 Y Y Y 0 Y

355

Plaza

Mounds (n)

Palisade

Paired

Public

Structures (n)

Site Size (ha)

Avg. Ann. Temp. (F)

Avg. Win. Temp. (F)

Phys. Prov.

Elev (m AMSL)

State

Site Name

State Site No

SiteID

1 6 7 8 9 10 11 12 13 14

18 56 6 12 7 81 11 38 82 114

1 3 4 4 5 5 5 6 7 8

15

73

9

16

74

9

17

29

9

18 19 22 24

38 1 13 3

9 10 12 14

King Warren Wilson Hiwassee Island Hiwassee Island Toqua Toqua Toqua Dog River Little Egypt BBB Motor Site Brasstown Valley Brasstown Valley Brasstown Valley Brasstown Valley Dallas Rymer McFarland

Palisade

Plaza

Mounds (n)

Paired

Unk. (n)

Stor. (n)

Non-dom. (n)

Dom. (n)

Types (n)

Structures (n)

Comp Size (ha)

Occ.

SiteName

Site ID

Pd. ID

Comp. ID

APPENDIX B. COMPONENT DATA TABLE

LM LM EM LM EM LM HI LM LM LW

2.30 1.00 4.60 4.60 1.90 1.90 1.90 2.00 2.00 0.16

67 12 14 1 25 79 14 1 3 16

4 1 4 1 7 4 4 1 1 1

44 12 14 1 15 41 12 1 3 16

2 0 0 0 9 23 2 0 0 0

21 0 0 0 0 14 0 0 0 0

0 0 0 0 1 1 0 0 0 0

N N Y N Y Y Y N N N

0 0 2 2 2 2 2 0 2 0

Y Y Y Y Y Y Y N Y N

Y Y Y Y Y Y Y N N N

LW

2.40

2

1

2

0

0

0

N

0

N

Y

LW

2.40

13

1

13

0

0

0

N

0

Y

Y

MW

2.40

8

1

8

0

0

0

N

0

Y

Y

HI LM LM MW

2.40 1.00 1.94 2.65

20 29 23 5

3 1 1 1

10 26 21 5

0 3 2 0

10 0 0 0

0 0 0 0

Y N Y N

0 1 0 0

N Y Y N

N Y Y N

356

83 25 26 27 28 59 72

15 16 16 16 16 17 18

32

11

19

33

68

20

34 35 36 37 39 40 41 42 43 44 45 49 50 51 52 53 54 55 59

69 57 58 59 64 60 61 62 63 29 23 97 101 105 110 75 38 52 30

20 21 21 21 23 23 23 23 23 24 24 27 27 27 27 27 28 29 31

Mialoquo Martin Farm Martin Farm Martin Farm Martin Farm Alarka Higgs Hiwassee Old Town Rucker's Bottom Rucker's Bottom Coweeta Creek Coweeta Creek Coweeta Creek Town Creek Town Creek Town Creek Town Creek Town Creek Hickory Log Hickory Log Cahokia Cahokia Cahokia Cahokia Cahokia 9WH120 Kolomoki Two Run Creek

Palisade

Plaza

Mounds (n)

Paired

Unk. (n)

Stor. (n)

Non-dom. (n)

Dom. (n)

Types (n)

Structures (n)

Comp Size (ha)

Occ.

SiteName

Site ID

Pd. ID

Comp. ID 25 26 27 28 29 30 31

HI EM EM LM LM HI EW

10.00 1.10 1.10 1.10 1.10 0.25 0.14

8 1 5 5 1 2 1

3 1 2 2 1 2 1

7 1 3 0 0 2 1

1 0 1 3 0 0 0

0 0 0 0 1 0 0

0 0 1 2 0 0 0

Y N N N N Y N

0 0 2 2 0 0 0

Y N N N N N N

N N Y N N N N

HI

140.00

6

2

4

0

1

1

N

0

N

N

MM

0.80

7

4

3

2

2

0

N

0

Y

N

LM LM LM HI LW EM MM MM LM MW MM LW EM EM MM EM LM MW MW

0.80 1.20 1.20 1.20 0.15 0.15 0.15 0.15 0.15 4.00 4.00 890.00 890.00 890.00 890.00 890.00 0.60 100.00 0.90

5 5 10 1 1 1 20 19 4 13 3 62 119 7 24 41 2 1 1

4 1 5 1 1 1 4 4 1 1 1 3 5 1 2 3 2 1 1

3 5 7 1 0 1 13 0 4 13 3 53 106 5 24 35 2 1 1

1 0 2 0 1 0 5 16 0 0 0 0 0 2 0 0 0 0 0

1 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0

0 0 0 0 0 0 2 3 0 0 0 8 13 0 0 6 0 0 0

N N Y N N Y Y Y N N N N N N N N Y N N

0 0 0 0 0 0 1 1 1 0 0 0 100 100 100 100 0 8 1

Y Y Y N Y Y Y Y N N N Y Y Y Y Y N Y N

Y Y Y N Y Y Y Y N Y Y N Y Y Y Y N Y N

357

78 38 18 6 11 73 75 54 59 40 6 75

32 33 34 35 35 36 36 37 37 38 39 40

72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87

42 41 2 30 38 48 38 38 49 48 38 38 66 12 70 71

41 42 43 44 44 45 46 47 48 49 50 51 52 53 54 55

88

84

56

Yuchi Town Go Kart Site Pott's Tract Chota-Tanase Chota-Tanase Fernvale Fernvale Ela Ela Mitchum Kimberly-Clark Macon Plateau Lower Saratown Jenrette Yearwood 9GE333 9GE333 9GE901 9GE903 9GE222 9GE1754 9GE1760 9GE1776 9GE1781 Tukabatchee Loy 1GR1X1 1PI61 Napoleon Hollow

Palisade

Plaza

Mounds (n)

Paired

Unk. (n)

Stor. (n)

Non-dom. (n)

Dom. (n)

Types (n)

Structures (n)

Comp Size (ha)

Occ.

SiteName

Site ID

Pd. ID

Comp. ID 60 61 62 63 64 65 66 67 68 69 70 71

HI LM LM EM HI MW EM MW HI HI EM EM

2.00 2.40 0.80 0.10 2.00 0.10 0.10 0.60 0.60 0.50 0.30 60.00

1 1 3 1 33 1 1 10 5 1 3 5

2 1 1 1 5 1 1 1 1 1 2 4

2 1 3 1 29 1 1 10 5 1 3 5

0 0 0 0 1 0 0 0 0 0 0 0

0 0 0 0 2 0 0 0 0 0 0 0

0 0 0 0 1 0 0 0 0 0 0 0

Y N N N Y N N N N N N N

0 0 0 0 1 0 0 0 0 0 0 8

Y N N N Y N N N N N N Y

Y N N N N N N N N Y N Y

HI HI MW MW LM LM LM LM LM LM LM LM LM LM MW LW

0.04 0.15 1.01 8.10 8.10 1.75 10.20 0.04 1.08 0.85 1.04 1.14 1.20 2.50 0.17 0.38

2 3 14 1 1 5 15 3 2 4 3 2 2 2 1 4

1 3 4 1 1 1 3 2 1 2 1 2 2 1 1 2

2 3 12 1 1 5 14 3 2 4 3 2 2 2 1 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 0 0 0 0 0

0 0 2 0 0 0 1 0 0 0 0 0 0 0 0 0

N N Y N N N N N N N N N N N N N

0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0

N N N N N N N N N N N N N Y N N

Y Y N N N N N N N N N N N Y N N

MW

15.00

1

1

0

0

0

1

N

0

N

N

358

31 29 75 85 75 86 87 54 93 94 11 4 4 96 113 3 115 46 48 116 117 118 119

57 58 59 60 61 62 63 64 65 65 66 68 69 69 70 70 8 71 71 72 73 73 73

9CK(DOT)7 Kellog Jewell Morris Bessemer Snodgrass Tuckasegee Garden Creek Moundville Moundville Townsend Banks III Banks V Banks V Duncan Tract Duncan Tract BBB Motor Site Dyar Dyar Hoecake Kincaid Kincaid Kincaid

127 128 129 130

120 89 19 19

73 65 74 75

Kincaid Moundville Summerour Rivermoore

EW MW EM EM EM MM HI MW EM EM HI MW MW EM EW MW EM LM LM LW EM MM LM EM, LM EM LW LW

Palisade

Plaza

Mounds (n)

Paired

Unk. (n)

Stor. (n)

Non-dom. (n)

Dom. (n)

Types (n)

Structures (n)

Comp Size (ha)

Occ.

SiteName

Site ID

Pd. ID

Comp. ID 89 92 93 94 95 96 97 98 99 100 101 103 104 105 118 119 120 121 122 123 124 125 126

0.80 0.26 3.50 0.60 10.00 0.82 0.50 5.00 75.00 75.00 10.00 0.10 0.10 0.03 0.40 0.40 0.13 2.13 2.13 81.00 32.00 32.00 32.00

1 1 22 12 25 91 1 1 35 102 10 3 4 1 2 5 5 2 4 11 7 6 2

1 1 3 3 3 1 1 1 2 2 2 2 2 1 1 1 3 2 2 1 3 3 2

1 1 13 12 4 91 1 0 35 98 9 3 4 1 2 5 4 2 1 11 1 0 1

0 0 9 0 21 0 0 0 0 4 0 0 0 0 0 0 1 0 3 0 6 6 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 0 0 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0

N N N N N N N N N N Y Y N N N N Y N Y N N N N

0 0 1 0 3 0 0 2 2 29 0 0 0 1 0 0 0 1 1 2 27 27 27

N N N Y Y Y N N N Y N N N N N N N Y Y Y Y Y Y

N N N Y N Y N N N Y N N N N N N N N N N Y Y Y

32.00 75.00 1.00 0.10

12 14 1 3

3 1 1 1

4 14 0 3

6 0 1 0

0 0 0 0

2 0 0 0

N N N N

27 29 1 0

Y Y N N

Y Y N N

359

656

1PI61

LW

3

N

Dom estic

660

9CK62

MW

Gain esvill e Carte rsvill e

1

N

Dom estic

548

9GE1754

LM

Lama r, Bell

2

N

Dom estic

547

9GE1754

LM

Lama r, Bell

1

N

Dom estic

561

9GE1760

LM

4

N

Dom estic

560

9GE1760

LM

Lama r, Dyar Lama r, Dyar

3

N

Dom estic

559

9GE1760

LM

Lama r, Dyar

2

N

Dom estic

558

9GE1760

LM

Lama r, Dyar

1

N

Dom estic

563

9GE1776

LM

Lama

2

N

Dom

Re

Posts

Pd ID

Dom estic

Comp ID

N

Site ID

1

Builds

Gain esvill e

Burned

LW

Basin

1PI61

Y

N

1

5 4

8 6

7 0

2

Y

N

1

5 5

8 7

7 1

N

130

Y

N

1

5 5

8 7

7 1

0

N

99

Y

N

1

5 5

8 7

7 1

2

1

N

105

Y

N

2

5 5

8 7

7 1

Fire basin

3

0

N

130

N

N

1

5 8

9 2

2 9

N

N/A

0

0

N

N

N

1

4 8

8 0

4 9

7.7

N

N/A

0

0

N

18

N

N

1

4 8

8 0

4 9

12.8

N

N/A

0

0

N

140

N

N

1

4 9

8 1

4 8

55.14

N

N/A

0

0

N

N

N

1

4 9

8 1

4 8

Y

Fire basin

2

0

N

Y

Fire basin

1

0

N

N

N/A

0

0

2

Y

Fire basin

2

2

Y

Fire basin

8.5

Y

2.2

6.6

3.5

2.2

4

3.2

37

0.24

0.67

31

0.24

10.1

6.1

51.2

30

0.07

0.21

6

0.18

3.1

2.2

6.9

55

47

0.18

0.18

8

0.18

2.9

2.4

7

Y

88

59

0.09

0.18

29

0.09

4.6

3.4

16.4

Both

Y

70

49

0.14

0.18

21

0.14

4

3

12

Posts

Y

27

27

0.13

0.36

0

Posts

Y

8

8

0.29

1.3

0

3

Posts

Y

16

16

0.24

1.1

0

Posts

Y

13

11

0.2

1.9

2

0.27

Posts

Y

15

12

0.26

2.1

3

0.24

Posts

Y

19

19

0.21

1

0

Posts

Y

18

17

0.21

1.3

1

Posts

Y

18

18

0.23

1.3

0

360

3.3

0.2

8.4

Orient. (deg E of N)

654

Posts

Entry Trenches

Dom estic

Int. Bur. (n)

N

Int. Feat. (n)

2

Hearth Type

Gain esvill e

Int. Hearth

LW

Int. Walls (n)

1PI61

Y

Area (m 2)

655

Posts

N

W (m)

Dom estic

L (m)

N

Dia. (m)

4

Int. Post Dia. (m)

Gain esvill e

68

Int. Posts (n)

LW

Y

Wall Post Spac. (m)

1PI61

Posts

Wall Post Dia. (m)

657

Ov al Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Re cta ng ula r Re cta ng ula r

Wall Posts (n)

Dom estic

Posts (n)

N

Complete

1

Wall Type

Shape

Mille r II

Locus

MW

Phase

1GRX1

Occ.

653

Site

Class

Mound Top

Str. No.

Str. ID

APPENDIX C. STRUCTURE DATA TABLE

0

4

2.7

10.8

N

N/A

0

0

N

22

N

N

1

4 9

8 1

4 8

5

2.8

14

N

N/A

0

0

N

40

N

N

1

4 9

8 1

4 8

4.4

3.2

14.1

N

N/A

0

0

N

31

N

N

1

5

8

3

562

9GE1776

LM

Lama r

1

N

Dom estic

564

9GE1776

LM

Lama r

3

N

Dom estic

565

9GE1781

LM

Lama r

1

N

Dom estic

566

9GE1781

LM

Lama r

2

N

Dom estic

546

9GE222

LM

Lama r

3

N

Dom estic

545

9GE222

LM

Lama r

2

N

Dom estic

544

9GE222

LM

Lama r

1

N

Dom estic

538

9GE333

LM

2

N

Dom estic

532

9GE333

MW

Lama r Swift Cree k

1

N

Dom estic

539

9GE901

LM

Lama r, Dyar

1

N

Dom estic

543

9GE901

LM

Lama r, Dyar

5

N

Dom estic

542

9GE901

LM

Lama r, Dyar

4

N

Dom estic

541

9GE901

LM

Lama r, Dyar

3

N

Dom estic

540

9GE901

LM

Lama r, Dyar

2

N

Dom estic

11

N

Dom estic

535

9GE903

LM

Lama r

Bl oc k1

Pd ID

Builds

Burned

Basin

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Comp ID

cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Re cta ng ula r Sq uar e Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng

Site ID

estic

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

r

0

2

8

Posts

Y

21

21

0.23

1.6

0

6.6

3.7

24.4

N

N/A

0

0

N

45

N

N

1

5 0

8 2

3 8

Posts

Y

20

20

0.24

1.6

0

5.8

4

23.2

N

N/A

0

0

N

117

N

N

1

5 0

8 2

3 8

Posts

Y

13

13

0.19

0.9

0

7.1

N

N/A

0

0

N

N

N

1

5 1

8 3

3 8

Posts

N

14

14

0.23

0.8

0

N

N/A

0

0

N

5

N

N

1

5 1

8 3

3 8

Posts

Y

8

8

0.21

1.73

0

2.6

2.8

7.3

N

N/A

0

0

N

5

N

N

1

4 7

7 9

3 8

Posts

Y

16

16

0.19

0.82

0

3.2

2.4

7.7

N

N/A

0

0

N

0

N

N

1

4 7

7 9

3 8

Posts

Y

19

19

0.2

1

0

4.8

2.5

12

N

N/A

0

0

N

75

N

N

1

4 7

7 9

3 8

Posts

N

5

5

0.26

1.2

0

N

N/A

0

0

N

N

N

1

4 4

7 6

3 8

Posts

Y

19

16

0.2

0.75

3

12.6

Y

Fire basin

1

0

N

N

N

1

4 4

7 5

3 0

Posts

Y

26

26

0.18

0.76

0

4

2.6

10.4

N

N/A

0

0

N

65

N

N

1

4 5

7 7

4 8

Posts

Y

17

17

0.2

1.43

0

4

2.7

10.8

N

N/A

0

0

N

152

N

N

1

4 5

7 7

4 8

Posts

Y

13

13

0.24

1.1

0

4

2.9

11.6

N

N/A

0

0

N

27

N

N

1

4 5

7 7

4 8

Posts

Y

25

25

0.21

0.77

0

4

2.6

10.4

N

N/A

0

0

N

45

N

N

1

4 5

7 7

4 8

Posts

Y

13

13

0.21

1.3

0

4

3.2

12.8

N

N/A

0

0

N

65

N

N

1

4 5

7 7

4 8

Posts

Y

15

15

0.21

1.2

0

3.2

2.4

7.68

N

N/A

0

0

N

112

N

N

1

4 6

7 8

3 8

361

3

2.6

4

Lama r

550

9GE903

LM

Lama r

537

9GE903

LM

Lama r

536

9GE903

LM

Lama r

552

9GE903

LM

Lama r

551

9GE903

LM

Lama r

Bl oc k3 Bl oc k3

557

9GE903

LM

Lama r

556

9GE903

LM

534

9GE903

553

N

2

N

Dom estic Unkn own/ other

Bl oc k1

14

N

Dom estic

Bl oc k1

12

N

Dom estic

4

N

Dom estic

3

N

Dom estic

Bl oc k4

1

N

Dom estic

Lama r

Bl oc k3

10

N

Dom estic

LM

Lama r

Bl oc k1

13

N

Dom estic

9GE903

LM

Lama r

Bl oc k3

7

N

Dom estic

549

9GE903

LM

Lama r, Dyar

1

N

Dom estic

531

9GE903

LM

Lama r

2

N

Dom estic

554

9GE903

LM

Lama r

8

N

Dom estic

312

9WH120

LM

Lama r

Bl oc k3 Bl oc k1

1

N

Dom estic

313

9WH120

LM

Lama r

Bl oc k1

2

N

Dom estic

Bl oc k2 Bl oc k1

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

N/A

0

0

N

150

N

N

1

4 6

7 8

3 8

8.6

58.1

N

N/A

0

0

N

90

N

N

1

4 6

7 8

3 8

1.5

1.8

N

N/A

0

0

N

N

N

1

4 6

7 8

3 8

2.8

12.6

N

N/A

0

0

N

120

N

N

1

4 6

7 8

3 8

5.7

3.2

18.2

N

N/A

1

0

N

153

N

N

1

4 6

7 8

3 8

5.7

3.4

19.4

N

N/A

0

0

N

22

N

N

1

4 6

7 8

3 8

83.3

N

N/A

0

0

N

N

N

1

4 6

7 8

3 8

2.3

6.9

N

N/A

0

0

N

110

N

N

1

4 6

7 8

3 8

3.6

2.9

10.4

N

N/A

0

0

N

117

N

N

1

4 6

7 8

3 8

0

3.7

3.3

12.2

N

N/A

0

0

N

13

N

N

1

4 6

7 8

3 8

1.1

0

3.3

2.3

7.6

N

N/A

0

0

N

60

N

N

1

4 6

7 8

3 8

0.19

0.7

0

4.6

2.5

11.5

N

N/A

0

0

N

30

N

N

1

4 6

7 8

4 8

17

0.28

1.9

0

78.5

N

N/A

0

6

N

N

N

1

4 6

7 8

3 8

21

21

0.2

0.7

0

3.2

2.4

7.7

N

N/A

0

0

N

40

N

N

1

4 6

7 8

3 8

Y

62

36

26

7.2

7.5

52

Y

Prepare d clay

2

0

Y

97

Y

N

1

2 8

5 4

3 8

Y

58

35

24

6.7

4.6

29

N

N/A

N

88

N

N

1

2 8

5 4

3 8

L (m)

Posts

Y

21

21

0.2

0.63

0

Posts

Y

56

31

0.22

0.8

25

0.22

Posts

Y

11

8

0.18

0.58

3

0.18

Posts

Y

21

21

0.22

0.85

0

4.5

Posts

Y

22

22

0.24

1

0

Posts

Y

17

17

0.25

1.9

0

Posts

Y

25

18

0.28

2

7

Posts

Y

17

17

0.23

1

0

3

Posts

Y

27

27

0.23

0.73

0

Posts

Y

15

15

0.26

1.8

Posts

Y

16

16

0.27

Posts

Y

23

23

Posts

Y

17

Posts

Y

Posts

Posts

362

2.7

0.28

W (m)

N

Dia. (m)

11.3

Posts (n)

Wall Type

Shape

Class

9

Pd ID

LM

N

Comp ID

9GE903

Dom estic

5

Site ID

555

Bl oc k1 Bl oc k3 Bl oc k2

Builds

Lama r

Burned

LM

Basin

9GE903

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID 533

ula r Re cta ng ula r Cir cul ar Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Re cta ng ula r Sq uar e Re cta ng ula

4.2

10.3

10

1

Orient. (deg E of N)

Burned

Builds

Site ID

Comp ID

Pd ID

34

0.16

10

4.5

45

2

Y

Fire basin

3

0

N

160

N

N

1

1 7

3 0

5 9

N

Dom estic

Sq uar e

Posts

Y

98

29

0.22

0.75

69

0.18

7

7

40

3

Y

Prepare d clay

2

0

N

160

Y

N

1

1 7

5 9

N

Dom estic

Ov al

Posts

Y

25

21

9.1

7.9

56.5

0

Y

Double oven

2

0

N

0

N

N

1

6 8

N

Dom estic

Ov al

1

6 8

III

N

Dom estic

II

N

Dom estic

Ov al Re cta ng ula r

3 0 1 0 3 1 0 3 1 0 3

V

N

Dom estic

N

Dom estic

Ov al Cir cul ar

N

Dom estic

Ov al

N

Dom estic

Ov al

Posts

Y

79

65

0.14

0.6

14

Dom estic

Re cta ng ula r

Posts

Y

7

7

0.1

0.75

0

Dom estic

Re cta ng ula r

Posts

Y

33

32

0.15

0.52

1

Dom estic

Re cta ng ula r

WT

Y

16

3

0.12

13

Dom estic

Re cta ng ula r

WT

Y

8

8

0.11

0

L (m)

W (m)

Area (m2)

0.73

Dia. (m)

Basin

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

0.22

Posts (n)

42

Complete

Wall Type

76

Class

Shape

Mound Top

Y

Str. No.

Posts

Locus

Phase

Occ.

Site

Str. ID

Dom estic

Re cta ng ula r

r

106

Alarka Farmstea d

HI

105

Alarka Farmstea d

HI

983

Banks III

MW

982

Banks III

MW

984

Banks III

MW

989

Banks V

EM

988

Banks V

MW

986

987

985

117 4

117 5

117 7

117 6

Banks V

Banks V

Banks V

BBB Motor

BBB Motor

BBB Motor

BBB Motor

MW

MW

MW

EM

EM

EM

EM

Quall a

Quall a Owl Hollo w Owl Hollo w Owl Hollo w

Stirli ng

Stirli ng

Stirli ng

N

II

I

Bank s Owl Hollo w Owl Hollo w Owl Hollo w Owl Hollo w

Stirli ng

Su m me r Ho use Wi nte r Ho use

III

IV

I Sti rlin g No rth ern Un it Sti rlin g No rth ern Un it Sti rlin g So uth ern Un it Sti rlin g So uth ern Un it

52

87

16

17 6

N

N

N

N

Posts

Y

20

4

19

1

Posts

Y

23

19

Both

Y

4

50

0.11

0.24

9

Posts

Y

19

19

0.18

0.88

0

Posts

Posts

Y

Y

7.6

4

0.11

37

14

6.7

2

0

363

0

N

N/A

2

0

N

90

N

N

2

0

N

75

N

N

1

6 8

11.5

9.5

85.8

0

Y

Double oven

5.5

3.6

19.8

1

N

N/A

0

0

N

17

N

N

1

6 9

4.7

4.3

15.9

0

N

N/A

0

0

N

100

N

N

1

6 9

1

6 9

1

6 9

1 0 5 1 0 4 1 0 4 1 0 4 1 0 4

4.5

14

40

15.9

5

0

19.6

N

N/A

0

0

N

N

N

N

N

N

N/A

N

2

0

N

58

N

N

1

6 9

4

4

4

9 6

4

4

4

13.7

10.7

146.6

0

Y

Double oven

5.54

3.26

18.1

0

Y

Fire basin

1

0

N

63

Y

Y

1

8

1 2 0

1 1 5

0.26

5.2

3.17

16.5

0

N

N/A

8

0

N

61

N

N

1

8

1 2 0

1 1 5

0.15

5

3.16

15.8

2

N

N/A

3

0

N

73

N

N

1

8

1 2 0

1 1 5

4.24

2.5

10.6

0

N

N/A

2

0

N

60

N

N

2

8

1 2 0

1 1 5

0.32

4

Edelh ardt

10 8

18 7

N

N

N

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

0.78

1.17

0

N

N/A

0

1

N

63

N

N

1

8

1 2 0

1 1 5

3.49

1.84

6.42

0

N

N/A

2

0

N

69

Y

N

1

8

1 4

1 1 4

0

2.65

1.51

4

0

N

N/A

0

0

N

Y

N

1

8

1 4

1 1 4

5

3.94

2.14

8.43

0

Y

Fire basin

4

0

N

75

Y

N

1

8

1 4

1 1 4

3.9

2.5

9.75

0

N

N/A

4

0

N

58

Y

N

1

8

1 4

1 1 4

2.78

1.7

4.73

0

Y

Fire basin

3

0

N

7

Y

N

1

8

1 4

1 1 4

4.51

2.41

10.87

0

N

N/A

1

0

N

70

Y

N

1

8

1 4

1 1 4

WT

Y

15

15

0.1

0.08

9

Dom estic

Re cta ng ula r

Posts

Y

38

38

0.1

0.28

1

Dom estic

Re cta ng ula r

Posts

Y

10

10

0.08

0.83

Dom estic

Re cta ng ula r

Posts

Y

50

45

0.09

0.27

Dom estic

Re cta ng ula r

Posts

Y

40

40

0.08

0.32

Dom estic

Re cta ng ula r

Posts

Y

35

35

0.13

0.26

0

Dom estic

Re cta ng ula r

Posts

Y

37

28

0.11

0.49

9

364

0.1

0.14

L (m)

Ov al

W (m)

1.5

Nondome stic

Dia. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Mound Top

Str. No.

Locus

Phase

Occ. LW

Edelh ardt

23

N

Pd ID

BBB Motor

LW

Edelh ardt

95

N

Comp ID

BBB Motor

LW

Edelh ardt

26 2

N

Site ID

118 3

BBB Motor

LW

Edelh ardt

16 1

Builds

118 1

BBB Motor

LW

Edelh ardt

N

Burned

117 8

BBB Motor

LW

Stirli ng

Basin

118 4

BBB Motor

EM

Shape

118 9

BBB Motor

Bu r Fe a 25 0

Class

118 7

Site

Str. ID 119 4

Sti rlin g No rth ern Un it Ed elh ard t Ce ntr al Cl ust er Ed elh ard t Ce ntr al Cl ust er Ed elh ard t We ste rn Cl ust er Ed elh ard t Ea ste rn Cl ust er Ed elh ard t We ste rn Cl ust er Ed elh ard t Ce ntr al Cl ust er

98

N

N

0.26

2

Dom estic

Re cta ng ula r

Posts

Y

55

55

0.11

0.23

Dom estic

Re cta ng ula r

Posts

Y

23

23

0.11

Dom estic

Re cta ng ula r

Posts

Y

35

35

Dom estic

Re cta ng ula r

Posts

Y

40

Dom estic

Re cta ng ula r

Posts

Y

39

Orient. (deg E of N)

0.13

Entry Trenches

49

Int. Bur. (n)

51

Int. Feat. (n)

Y

Hearth Type

Posts

Int. Hearth

Dom estic

Re cta ng ula r

Int. Walls (n)

0

Area (m2)

0.28

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n) 0.1

1.73

4.97

0

N

N/A

2

0

N

86

Y

N

1

8

1 4

1 1 4

4.2

2.18

9.16

0

Y

Fire basin

1

0

N

92

Y

N

1

8

1 4

1 1 4

0

4.1

2.26

9.27

0

N

N/A

3

0

N

75

Y

N

1

8

1 4

1 1 4

0.36

0

2.58

1.46

3.77

0

N

N/A

0

0

N

87

Y

N

1

8

1 4

1 1 4

0.13

0.33

0

3.44

2.43

8.36

0

N

N/A

4

0

N

78

Y

N

1

8

1 4

1 1 4

39

0.09

0.32

1

0.14

3.44

2.36

8.12

0

N

N/A

2

0

N

62

N

N

1

8

1 4

1 1 4

36

0.12

0.36

3

0.12

4.24

2.22

9.41

1

N

N/A

3

0

N

65

Y

N

1

8

1 4

1 1 4

365

0.13

L (m)

W (m)

2.87

Dia. (m)

Posts (n)

33

Complete

Mound Top

Str. No. 61

N

33

Pd ID

Edelh ardt

Locus

Phase

Occ. LW

Edelh ardt

10 0

N

Y

Comp ID

BBB Motor

LW

Edelh ardt

17 5

N

Posts

Site ID

BBB Motor

LW

Edelh ardt

29

N

Dom estic

Builds

118 0

BBB Motor

LW

Edelh ardt

94

N

Burned

119 3

BBB Motor

LW

Edelh ardt

23 5

Re cta ng ula r

Basin

118 6

BBB Motor

LW

Edelh ardt

Wall Type

118 2

BBB Motor

LW

Shape

119 2

BBB Motor

Class

117 9

Site

Str. ID 119 1

Ed elh ard t Ce ntr al Cl ust er Ed elh ard t We ste rn Cl ust er Ed elh ard t Ea ste rn Cl ust er Ed elh ard t Ea ste rn Cl ust er Ed elh ard t We ste rn Cl ust er Ed elh ard t Ea ste rn Cl ust er Ed elh ard t Ce ntr al Cl ust er

EM

717

Bessemer

EM

EM

Early Missi ssippi an

EM

Early Missi ssippi an

EM

Early Missi ssippi an

Vil lag e Do m. Mo un d

Early Missi ssippi an Early Missi ssippi an

Vil lag e Ce re mo nia

704

719

700

Bessemer

Bessemer

Bessemer

707

Bessemer

EM

695

Bessemer

EM

Vil lag e Do m. Mo un d

0.11

0.18

5

Re cta ng ula r

Posts

Y

30

30

0.08

0.4

WT

Y

Both

Y

Posts

Y

Both

N

Both

Y

13 0

12 5

0.15

0.15

5

0.33

7.5

5.8

43.5

Both

N

89

68

0.09

0.18

21

0.2

6.6

6.4

Both

Y

18 1

18 1

0.15

0.15

0

10.8

Both

Y

60

47

0.15

0.18

13

5.5

21 8

N

Dom estic

4

Y

Dom estic

3

Y

Nondome stic

12

N

Nondome stic

6

Y

Nondome stic

14

N

Nondome stic

2

Y

Nondome stic

2

N

Nondome stic

1

Y

Dom estic

Sq uar e Re cta ng ula r

Ov al Re cta ng ula r Re cta ng ula r

Sq uar e Re cta ng ula r Re cta ng ula

0

25 3

25 1

0.09

0.18

Orient. (deg E of N)

45

Entry Trenches

50

Int. Bur. (n)

Y

Int. Feat. (n)

Posts

Hearth Type

Dom estic

Re cta ng ula r

Int. Hearth

0

Area (m2)

0.6

Int. Walls (n)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

0.08

2.18

9.16

0

N

N/A

0

0

N

40

Y

N

1

8

1 4

1 1 4

2.26

1.49

3.37

0

N

N/A

1

0

N

78

Y

N

1

8

1 4

1 1 4

0

3.88

2.22

8.61

0

N

N/A

1

0

N

69

Y

N

1

8

1 4

1 1 4

8

4.3

4.3

18.5

N

N/A

N

0

N

N

1

6 1

9 5

7 5

15.2

12.2

185.4

Y

Fire basin

N

150

N

N

1

6 1

9 5

7 5

N

N/A

N

N

N

1

6 1

9 5

7 5

N

N/A

N

N

N

1

6 1

9 5

7 5

Y

Prepare d clay

1

0

N

40

Y

N

1

6 1

9 5

7 5

42.2

Y

Fire basin

1

0

N

70

N

N

1

6 1

9 5

7 5

7.8

83.9

Y

Prepare d clay

4

0

N

165

N

N

1

6 1

9 5

7 5

4.9

27

Y

Fire basin

2

0

N

0

N

Y

1

6 1

9 5

7 5

2

0.11

0.67

L (m)

W (m)

4.2

Dia. (m)

Posts (n)

21

Complete

Mound Top

Str. No.

21

Pd ID

Bessemer

Y

Comp ID

701

Early Missi ssippi an Early Missi ssippi an

Posts

Site ID

EM

Early Missi ssippi an

N

Dom estic

Builds

Edelh ardt

Locus

Phase

Occ. LW

96

N

Burned

Bessemer

Edelh ardt

24

Re cta ng ula r

Basin

BBB Motor

LW

Edelh ardt

Wall Type

698

BBB Motor

LW

Shape

118 8

BBB Motor

Class

118 5

Site

Str. ID 119 0

Ed elh ard t Ce ntr al Cl ust er Ed elh ard t We ste rn Cl ust er Ed elh ard t Ce ntr al Cl ust er Ce re mo nia l Mo un d Do m. Mo un d

0.18

6.1

366

0.15

0

0

2

0

10

N

Nondome stic

1

Y

Nondome stic

Vil lag e

13

N

Nondome stic

Cir cul ar Re cta ng ula r Re cta ng ula r

Bessemer

EM

718

Bessemer

EM

EM

Early Missi ssippi an

Vil lag e

6

N

Nondome stic

EM

Early Missi ssippi an

Vil lag e

1

N

Nondome stic

EM

Early Missi ssippi an

11

N

Nondome stic

EM

Early Missi ssippi an

Vil lag e Do m. Mo un d

4

Y

Nondome stic

EM

Early Missi ssippi an

Vil lag e

7

N

Nondome stic

EM

Early Missi ssippi an

4

N

Nondome stic

Sq uar e Re cta ng ula r Re cta ng ula r

EM

Early Missi ssippi an

2

Y

Dom estic

Sq uar e

3

Y

5

Y

Dom estic Nondome stic

Sq uar e Re cta ng

706

716

702

712

709

696

Bessemer

Bessemer

Bessemer

Bessemer

Bessemer

Bessemer

697

Bessemer

EM

703

Bessemer

EM

Early Missi ssippi an Early Missi ssippi

Vil lag e Ce re mo nia l Mo un d Ce re mo nia l Mo un d Do m. Mo

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

0

N

N/A

0

0

N

120

N

N

1

6 1

9 5

7 5

0

N

N/A

0

0

N

35

N

N

1

6 1

9 5

7 5

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Pd ID

Nondome stic

699

Bessemer

Comp ID

N

Early Missi ssippi an Early Missi ssippi an

711

51.5

r

8

Vil lag e Do m. Mo un d

Site ID

EM

Early Missi ssippi an

Vil lag e

Builds

Bessemer

EM

Burned

715

Bessemer

Basin

713

Early Missi ssippi an

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

l Mo un d

Sq uar e Re cta ng ula r Re cta ng ula r Re cta ng ula r

Posts

Y

72

71

0.18

0.3

1

0.21

Both

Y

14 1

13 4

0.15

0.15

7

0.15

Both

Y

16 5

16 5

0.09

0.18

Both

Y

12 5

11 8

0.15

0.15

7

Both

Y

18 5

0.15

0.15

Both

Y

32 9

28 3

0.15

Posts

Y

16 9

16 5

Both

Y

83

Both

Y

Both

8.1

8.8

7.5

66

10.1

8.5

85.9

Y

Prepare d clay

1

0

N

65

N

N

1

6 1

9 5

7 5

0.21

5.5

5.3

29.1

N

N/A

0

0

N

30

N

N

1

6 1

9 5

7 5

13

0.21

11.9

9.1

108.3

N

N/A

0

0

N

115

N

N

1

6 1

9 5

7 5

0.15

46

0.21

18.5

11.9

220.2

N

N/A

0

0

N

N

N

1

6 1

9 5

7 5

0.15

0.15

4

0.18

12.8

9

115.2

N

N/A

0

0

N

48

N

N

1

6 1

9 5

7 5

83

0.09

0.18

0

7.6

7

53.2

Y

Fire basin

2

N

150

N

N

1

6 1

9 5

7 5

10 5

10 3

0.15

0.15

2

0.21

7.2

6.1

43.9

0

N

N/A

0

0

N

150

N

N

1

6 1

9 5

7 5

Y

26 9

22 8

0.15

0.15

41

0.15

16.4

10.7

175.5

2

N

N/A

0

0

N

120

N

N

1

6 1

9 5

7 5

Posts

Y

38

32

0.15

0.45

6

0.15

4.3

4.3

18.5

N

N/A

0

0

N

40

N

N

1

6 1

9 5

7 5

Both

Y

87

81

0.15

0.18

6

0.15

4.6

4.6

21.2

N

N/A

0

0

N

63

N

N

1

6 1

9 5

7 5

Both

N

14 2

14 5

0.09

0.18

3

0.39

19.5

6.4

124.8

Y

Fire basin

1

0

N

155

N

N

1

6 1

9 5

7 5

367

0

0

708

Bessemer

EM

705

Bessemer

EM

63

Brasstow n Valley

HI

68

Brasstow n Valley

HI

61

Brasstow n Valley

HI

58

Brasstow n Valley

HI

60

Brasstow n Valley

HI

62

Brasstow n Valley

HI

71

Brasstow n Valley

HI

65

Brasstow n Valley

HI

54

Brasstow n Valley

HI

73

Brasstow n Valley

HI

69

Brasstow n Valley

HI

66

Brasstow n Valley

HI

Early Missi ssippi an Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

N

N

1

6 1

9 5

7 5

N

N

1

6 1

9 5

7 5

Y

N

1

6 1

9 5

7 5

ula r

N

Nondome stic

Re cta ng ula r

Both

Y

14 0

14 0

0.15

0.15

9

N

Nondome stic

Cir cul ar

Posts

Y

81

81

0.18

0.4

3

N

Nondome stic

Both

Y

15 7

15 7

0.15

0.15

7

Y

Nondome stic

Sq uar e Re cta ng ula r

Both

N

48 F

C11

N

Stora ge

Posts

Y

45 L

42

N

Dom estic

Cir cul ar Re cta ng ula r

Posts

Y

20

18

2

45 L

C8

N

Stora ge

Cir cul ar

Posts

Y

21

21

0

49 B

C5

N

Stora ge

Cir cul ar

Posts

Y

19

19

0

49 B

C7

N

Stora ge

Cir cul ar

Posts

48 F

C10

N

Stora ge

Cir cul ar

48 H/I

5

N

Dom estic

45 L

31

N

Dom estic

Sq uar e Re cta ng ula r

48 F

C1

N

Stora ge

45 L

30

N

Dom estic

45 L

1

N

Dom estic

48 F

40

N

Dom estic

Vil lag e

5

Vil lag e Vil lag e Do m. Mo un d

Pd ID

EM

Comp ID

Bessemer

Site ID

714

Builds

EM

Burned

Bessemer

un d

Basin

710

Early Missi ssippi an Early Missi ssippi an Early Missi ssippi an

Locus

Phase

Occ.

Site

Str. ID

an

0

8.7

7.2

10.2

0

N

N/A

0

0

N

81.7

0

N

N/A

0

0

N

0

N

N/A

0

0

N

142

9

8.7

78

11.2

5.3

59.4

U

N/A

N

N

N

1

6 1

9 5

7 5

4.52

N

N/A

N

N

N

1

9

1 8

3 8

21.32

N

N/A

N

N

1

9

1 8

3 8

2

2.95

N

N/A

N

N

N

1

9

1 8

3 8

2

3.3

N

N/A

N

N

N

1

9

1 8

3 8

Y

1

1.15

N

N/A

N

N

N

1

9

1 8

3 8

Posts

Y

3

6.83

N

N/A

N

N

N

1

9

1 8

3 8

Posts

Y

Y

Prepare d clay

N

N

1

9

1 8

3 8

Posts

Y

65

36

29

N

N

1

9

1 8

3 8

Cir cul ar

Posts

Y

27

25

2

N

N

1

9

1 8

3 8

Sq uar e

Posts

Y

32

27

5

6.48

6.27

40.63

N

N

1

9

1 8

3 8

Posts

Y

11 3

49

64

6.32

6.01

37.98

Posts

Y

51

26

25

5.48

3.37

18.47

Sq uar e Re cta ng ula

0

62.6

2

368

6.27

3.4

0

0

0

110

4.72

4.55

21.47

5.25

3.75

19.69

1

N

N/A

0

0

N

4.52

0

N

N/A

0

0

N

Y

Prepare d clay

3

Y

Prepare d clay

1

0

Y

140

Y

N

1

9

1 8

3 8

0

N

N/A

0

0

N

138

N

N

1

9

1 8

3 8

2

105

Orient. (deg E of N)

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

N

N

1

9

1 8

3 8

N

N

1

9

1 8

3 8

N

N

1

9

1 8

3 8

N

N

1

9

1 8

3 8

N

N

N

1

9

1 8

3 8

N

N

N

1

9

1 8

3 8

N

N

1

9

1 8

3 8

r

HI

Quall aLama r

48 F

C2

N

Stora ge

HI

Quall aLama r

48 J

32

N

Dom estic

45 L

41

N

Dom estic

Cir cul ar Re cta ng ula r Re cta ng ula r

48 F

7

N

Dom estic

49 B

C6

N

48 J

C4

48 H/I

55

Brasstow n Valley

64

Brasstow n Valley

67

Brasstow n Valley

HI

72

Brasstow n Valley

HI

59

Brasstow n Valley

HI

57

Brasstow n Valley

HI

70

Brasstow n Valley

HI

56

Brasstow n Valley

HI

39

Brasstow n Valley

LW

49

Brasstow n Valley

LW

42

Brasstow n Valley

LW

44

Brasstow n Valley

LW

47

Brasstow n Valley

LW

40

Brasstow n Valley

LW

Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Quall aLama r Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/

Posts

Y

8

8

0

Posts

Y

30

20

10

Posts

Sq uar e

Stora ge

N

4

48 F

2

2.4

0

N

N/A

0

0

N

2

N

N/A

0

0

N

9.03

2.27

20.5

Y

5.32

2.88

15.32

N

N/A

Posts

Y

6.82

6.36

43.37

Y

Prepare d clay

Cir cul ar

Posts

Y

N

N/A

Stora ge

Cir cul ar

Posts

Y

N

Dom estic

Sq uar e

Posts

Y

C3

N

Stora ge

Cir cul ar

Posts

Y

34

29

5

2

4.6

49 B

12

N

Dom estic

Cir cul ar

Posts

Y

30

25

5

5

49 B

27

N

Dom estic

Cir cul ar

Posts

Y

98

44

54

49 B

16

N

Dom estic

Cir cul ar

Posts

49 B

18

N

Dom estic

Cir cul ar

49 B

24

N

Dom estic

49 B

13

N

Dom estic

16

16

0

2.54

2

3.53

N

N/A

25.98

Y

Prepare d clay

0

N

N/A

0

0

N

N

N

1

9

1 8

3 8

17.79

0

N

N/A

0

0

N

N

N

1

9

1 6

7 4

8

47.15

2

N

N/A

0

2

N

N

N

1

9

1 6

7 4

Y

6

32.35

N

N/A

N

N

N

1

9

1 6

7 4

Posts

Y

7

34.51

N

N/A

N

N

N

1

9

1 6

7 4

Cir cul ar

Posts

Y

7

35.66

N

N/A

N

N

N

1

9

1 6

7 4

Cir cul ar

Posts

Y

6

25.95

N

N/A

N

N

N

1

9

1 6

7 4

369

4.94

0

0

140

2

5.26

0

Y

37

Burned

Builds

Site ID

Comp ID

Pd ID

29.88

N

N/A

N

N

N

1

9

1 6

7 4

49 B

23

N

Dom estic

Cir cul ar

Posts

Y

6

26.96

N

N/A

N

N

N

1

9

1 6

7 4

49 B

34

N

Dom estic

Cir cul ar

Posts

Y

6

24.79

N

N/A

N

N

N

1

9

1 6

7 4

49 B

21

N

Dom estic

Cir cul ar

Posts

Y

6

24.88

N

N/A

N

N

N

1

9

1 6

7 4

48 H/I

3

N

Dom estic

Cir cul ar

Posts

Y

5

21.8

N

N/A

N

N

N

1

9

1 5

7 3

48 F

10

N

Dom estic

Cir cul ar

Posts

Y

5

17.64

N

N/A

N

N

N

1

9

1 5

7 3

49 B

26

N

Dom estic

Cir cul ar

Posts

Y

7

33.78

N

N/A

N

N

N

1

9

1 6

7 4

49 B

14

N

Dom estic

Cir cul ar

Posts

Y

6

23.75

N

N/A

N

N

N

1

9

1 6

7 4

49 B

35

N

Dom estic

Cir cul ar

Posts

Y

7

33.37

N

N/A

N

N

N

1

9

1 6

7 4

48 F

33

N

Dom estic

Ov al

Posts

Y

7.33

6.67

48.89

N

N/A

N

N

N

1

9

1 7

2 9

48 F

11

N

Dom estic

Ov al

Posts

Y

7.33

5.36

39.29

N

N/A

N

N

N

1

9

1 7

2 9

Area (m2)

6

W (m)

Y

L (m)

Posts

Dia. (m)

Basin

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Wall Type

Cir cul ar

Posts (n)

Shape

Dom estic

Complete

Class

Mound Top N

Str. No. 17

Locus

Phase

Occ.

Site

Str. ID

49 B

Napi er

43

Brasstow n Valley

LW

46

Brasstow n Valley

LW

50

Brasstow n Valley

LW

45

Brasstow n Valley

LW

52

Brasstow n Valley

LW

53

Brasstow n Valley

LW

48

Brasstow n Valley

LW

41

Brasstow n Valley

LW

51

Brasstow n Valley

LW

35

Brasstow n Valley

MW

34

Brasstow n Valley

MW

Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Woo dstoc k/Na pier/ Swift Cree k Woo dstoc k/Na pier/ Swift Cree k Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Etow ah/W oodst ock/ Napi er Late Woo dland /Cart ersvil le Late Woo dland /Cart ersvil le

370

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

38

N

Dom estic

Ov al

Posts

Y

6.98

5.64

39.37

N

N/A

N

N

N

1

9

1 7

2 9

48 H/I

2

N

Dom estic

Ov al

Posts

Y

5.49

4.65

25.53

N

N/A

N

N

N

1

9

1 7

2 9

48 F

37

N

Dom estic

Ov al

Posts

Y

6.69

6.42

42.95

N

N/A

N

N

N

1

9

1 7

2 9

48 F

39

N

Dom estic

Ov al

Posts

Y

5.3

5.23

27.72

N

N/A

N

N

N

1

9

1 7

2 9

48 F

8

N

Dom estic

Posts

Y

7.65

5.05

38.63

N

N/A

N

N

N

1

9

1 7

2 9

EM

Unkn own

Tra ct 15 A

un no. Str C

N

Dom estic

WT

N

N

N/A

N

N

N

1

2 7

5 3

7 9

EM

Unkn own

Tra ct 15 A

Fe a 32 1

N

Dom estic

WT

N

N

N/A

N

N

N

1

2 7

5 3

7 9

EM

Unkn own

Tra ct 15 A

7

N

Dom estic

WT

Y

6.3

3.7

27.4

N

N/A

0

N

N

1

2 7

5 3

7 9

EM

Unkn own

Tra ct 15 A

15

N

Dom estic

WT

Y

3.8

2

7.6

N

N/A

1

N

N

1

2 7

5 3

7 9

EM

Unkn own

Tra ct 15 A

14 6

N

Dom estic

WT

N

N

N/A

Y

N

1

2 7

5 3

7 9

EM

Unkn own

Tra ct 15 A

50

N

Dom estic

WT

Y

N

N/A

N

N

1

2 7

5 3

7 9

EM

Unkn own

Tra ct 15 A

23 8

N

Dom estic

Ov al Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

WT

N

N

N/A

N

N

1

2 7

5 3

7 9

MW

32

Brasstow n Valley

MW

115 7

Cahokia

Orient. (deg E of N)

48 F

Brasstow n Valley

Cahokia

Entry Trenches

2 9

38

114 0

Int. Bur. (n)

1 7

MW

Cahokia

Int. Feat. (n)

9

Brasstow n Valley

114 4

Hearth Type

1

36

Cahokia

Int. Hearth

N

MW

113 5

Int. Walls (n)

N

W (m)

L (m)

Locus

Area (m2)

N

Dia. (m)

N/A

Brasstow n Valley

Cahokia

Int. Post Dia. (m)

N

31

112 9

Int. Posts (n)

37.15

MW

Cahokia

Wall Post Spac. (m)

5.75

Brasstow n Valley

116 6

Wall Post Dia. (m)

6.46

37

Cahokia

Wall Posts (n)

Wall Type

Y

Posts (n)

Shape

Posts

Complete

Class

Ov al

MW

116 4

Mound Top

Dom estic

Brasstow n Valley

Str. No.

N

Phase

9

Occ.

Site

Str. ID

48 F

33

Late Woo dland /Cart ersvil le Late Woo dland /Cart ersvil le Late Woo dland /Cart ersvil le Late Woo dland /Cart ersvil le Late Woo dland /Cart ersvil le Late Woo dland /Cart ersvil le

2.8

3.7

371

1.9

7

0

0

N

96

N

N

2

90

90

N

N

5

Cahokia

Cahokia

N

Dom estic

EM

Unkn own

Tra ct 15 A

15 9

N

Dom estic

EM

Unkn own

Tra ct 15 A

23 7

N

Dom estic

EM

Unkn own

Tra ct 15 A

8

N

Dom estic

24 0

N

Dom estic

22 6

N

Unkn own/ other

115 9

Cahokia

EM

Unkn own

115 1

Cahokia

EM

Unkn own

Tra ct 15 A Tra ct 15 A

EM

Unkn own

Tra ct 15 A

6

N

Dom estic

EM

Unkn own

Tra ct 15 A

5

N

Dom estic

20 0

N

Dom estic

30

N

Dom estic

112 8

112 7

Cahokia

Cahokia

114 7

Cahokia

EM

Unkn own

113 9

Cahokia

EM

Unkn own

Tra ct 15 A Tra ct 15

Orient. (deg E of N)

21 2

Entry Trenches

Tra ct 15 A

Int. Bur. (n)

EM

Unkn own

Int. Feat. (n)

Dom estic

Hearth Type

N

Int. Hearth

57

Int. Walls (n)

Tra ct 15 A

Area (m2)

EM

Unkn own

N/A

N

10

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

100

N

N

1

2 7

5 3

7 9

Posts

N

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

Y

5.6

3.4

19

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

N

6.1

4.1

25

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

N

2.6

N

N/A

N

90

N

N

1

2 7

5 3

7 9

WT

N

3.6

N

N/A

N

92

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

20

N

N

1

2 7

5 3

7 9

WT

Y

N

N/A

N

5

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

97

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

95

N

N

1

2 7

5 3

7 9

WT

Y

11.2

5.8

65

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

Y

5.2

2.8

14.6

N

N/A

N

Y

N

1

2 7

5 3

7 9

WT

N

3.7

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

N

4.7

N

N/A

N

N

N

1

2 7

5 3

7 9

W (m)

Dom estic

N

L (m)

N

N

Dia. (m)

17

Int. Post Dia. (m)

Tra ct 15 A

Int. Posts (n)

EM

Unkn own

Wall Post Spac. (m)

Dom estic

Wall Post Dia. (m)

N

Wall Posts (n)

13

WT

Posts (n)

Tra ct 15 A

Complete

EM

Unkn own

Wall Type

Dom estic

Class

N

Shape

Mound Top

Str. No. 17 6

Occ.

Tra ct 15 A

Pd ID

113 0

Cahokia

EM

Unkn own

Un def ine d Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Un def ine d Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng

Comp ID

115 6

Cahokia

N

Unkn own/ other

Site ID

114 5

Cahokia

21 4

Builds

114 8

Cahokia

Tra ct 15 A

Burned

114 1

Cahokia

EM

Unkn own

Basin

113 6

Cahokia

Locus

113 3

Cahokia

Phase

114 6

Site

Str. ID 114 9

2.5

9.7

372

5

2.5

0

48.5

11.7

0

1

2

0

5

0

0

95

Dom estic

un no. Str H

N

Dom estic

31 0

N

Unkn own/ other

31 2

N

Unkn own/ other

Cahokia

EM

116 2

Cahokia

EM

Unkn own

116 3

Cahokia

EM

Unkn own

EM

Unkn own

Tra ct 15 A

4

N

Dom estic

EM

Unkn own

Tra ct 15 A

22

N

Dom estic

EM

Unkn own

Tra ct 15 A

24

N

Dom estic

23 9

N

Dom estic

22 7

N

Unkn own/ other

30 9

N

Dom estic

23 3

N

Unkn own/ other

113 8

Cahokia

Cahokia

115 8

Cahokia

EM

Unkn own

115 2

Cahokia

EM

Unkn own

Tra ct 15 A Tra ct 15 A

116 1

Cahokia

EM

Unkn own

115 4

Cahokia

EM

Unkn own

Tra ct 15 A Tra ct 15 A

EM

Unkn own

Tra ct 15 A

22 1

N

Dom estic

EM

Unkn own

Tra ct 15 A

23 5

N

Dom estic

EM

Unkn own

Tra ct 15 A

14

N

Dom estic

115 0

115 5

113 4

Cahokia

Cahokia

Cahokia

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

N

N

1

2 7

5 3

7 9

N

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

5

N

N

1

2 7

5 3

7 9

Unid

N

N

N/A

0

N

90

Y

N

1

2 7

5 3

7 9

WT

Y

6.5

4.4

28.6

N

N/A

2

N

N

Y

1

2 7

5 3

7 9

WT

Y

6.6

5.4

35

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

5

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

2

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

8

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

103

N

N

1

2 7

5 3

7 9

WT

N

N

N/A

N

96

N

N

1

2 7

5 3

7 9

WT

N

3.3

N

N/A

N

95

Y

N

1

2 7

5 3

7 9

WT

Y

4

N

N/A

N

N

N

1

2 7

5 3

7 9

L (m)

W (m)

N

Dia. (m)

N/A

Posts (n)

Pd ID

N

Tra ct 15 A Tra ct 15 A Tra ct 15 A

113 7

Comp ID

12

Unkn own

Cahokia

N

ula r

116 5

112 6

Site ID

Unkn own

Builds

EM

Burned

Cahokia

Basin

113 2

Tra ct 15 A

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

A

Re cta ng ula r Re cta ng ula r Un def ine d Un def ine d Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Un def ine d Re cta ng ula r Un def ine d Re cta ng ula r Re cta ng ula r Re cta ng ula r

WT

N

WT

12

0

4.6

0

373

2.2

9

0

0

0

994

105 6

Cahokia

Cahokia

Cahokia

Cahokia

22 4A

N

Dom estic

Tra ct 15 A

85

N

Dom estic

Tra ct 15 A

D1 9B

N

Dom estic

EM

Loh mann 1 Loh mann Stirli ng

EM

Loh mann 2

Tra ct 15 A

21 1

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

22 2

N

Dom estic

23 0

N

Dom estic

17 0

N

Unkn own/ other

14 7

N

Dom estic

EM

Cahokia

EM

104 1

Cahokia

EM

Other Loh mann

Tra ct 15 A Tra ct 15 A

EM

Loh mann 3

Tra ct 15 A

Cir cul ar Re cta ng ula r

N

WT

N

WT

Y

5.4

2.6

14

WT

Y

4.7

2.4

WT

Y

4.5

WT

Y

WT

Y

WT

N

WT

N

WT

Y

WT

Y

N

N

1

2 7

5 3

7 9

N

N

N

2

2 7

5 3

7 9

N

N

N

1

2 7

5 3

7 9

N

N

1

2 7

5 3

7 9

N

N

1

2 7

5 3

7 9

91

N

N

1

2 7

5 0

1 0 1

N

100

N

N

2

2 7

5 0

1 1 2

N

10

Y

N

1

2 7

5 0

1 0 1

N

N

1

2 7

5 0

1 0 4

104

N

N

1

2 7

5 0

1 0 2

N

95

N

N

1

2 7

5 0

1 1 2

N

3

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 0 3

N/A

N

N/A

N

N/A

N

N/A

N

N/A

0

N

N/A

1

0

N

11.3

0

N

N/A

1

0

2.5

11.3

0

N

N/A

1

0

5.1

2.5

12.8

N

N/A

4.1

2.4

9.8

0

N

N/A

0

0

N

0

N

N/A

0

0

N

N/A

5

4.7

23.5

2.9

0

4.2

2.6

4.7

374

N

N

11.7

2

1

0

0

Orient. (deg E of N)

Tra ct 15 A

WT

Entry Trenches

EM

Other Loh mann

N

Int. Bur. (n)

Dom estic

WT

Int. Feat. (n)

N

N

Hearth Type

20 8

WT

Int. Hearth

Tra ct 15 A

N

Int. Walls (n)

EM

Loh mann 1

Area (m2)

Dom estic

W (m)

N

L (m)

30 7

Dia. (m)

Tra ct 15 A

Int. Post Dia. (m)

EM

Unkn own

Int. Posts (n)

Dom estic

Wall Post Spac. (m)

N

Wall Post Dia. (m)

12 8

Wall Posts (n)

Tra ct 15 A

Posts (n)

EM

Unkn own

Complete

Dom estic

Wall Type

N

Shape

58

Class

Tra ct 15 A

Other Loh mann

Cahokia

Mound Top

EM

Unkn own

106 1

100 0

Str. No.

Dom estic

Occ.

N

WT

Pd ID

109 0

Cahokia

11

Dom estic

Comp ID

206

Cahokia

Tra ct 15 A

N

Site ID

105 8

Cahokia

EM

Unkn own

22 8

Builds

217

Cahokia

Unkn own

Burned

116 0

Cahokia

EM

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Basin

114 3

Cahokia

Locus

114 2

Cahokia

Phase

113 1

Site

Str. ID 115 3

Tra ct 15 A

N

102

N

N

5.3

0

N

N/A

1

0

N

9.4

0

N

N/A

1

0

N

94

36

N

Dom estic

EM

Loh mann 3

Tra ct 15 A

11 0

N

Dom estic

EM

Loh mann 1

Tra ct 15 A

16 5

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

55

N

Dom estic

Tra ct 15 A

21 7

N

Dom estic

Tra ct 15 A

49

N

Dom estic

Tra ct 15 A

13 8

N

Dom estic

Tra ct 15 A

D3 8

N

Dom estic

998

214

101 4

105 3

107 5

103 4

109 4

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

EM

EM

EM

EM

Other Loh mann Loh mann Stirli ng

Other Loh mann Loh mann Stirli ng

Orient. (deg E of N)

Pd ID

Tra ct 15 A

Comp ID

EM

Other Loh mann

101 0

Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Site ID

Other Loh mann

Entry Trenches

Unkn own/ other

EM

Int. Bur. (n)

N

Cahokia

Int. Feat. (n)

16 1

103 8

Hearth Type

Dom estic

EM

Tra ct 15 A Tra ct 15 A

Int. Hearth

N

Cahokia

Other Loh mann

Int. Walls (n)

52

101 3

Area (m2)

Dom estic

2.4

9.1

0

N

N/A

0

0

N

95

Y

N

1

2 7

5 0

1 0 2

WT

Y

4.5

2.4

10.8

0

N

N/A

1

0

N

3

N

N

2

2 7

5 0

1 1 2

WT

N

N

N/A

N

Y

N

2 7

5 0

1 0 3

WT

Y

N

N/A

N

N

N

2

2 7

5 0

1 0 4

WT

Y

N

N/A

1

N

N

N

1

2 7

5 0

1 1 2

WT

Y

N

N/A

1

N

N

N

1

2 7

5 0

1 1 2

WT

Y

3.5

0

N

N/A

1

N

N

1

2 7

5 0

1 1 2

WT

N

4.3

1

N

N/A

1

N

N

1

2 7

5 0

1 0 3

WT

Y

6.3

3.9

24.6

0

N

N/A

0

Y

N

2

2 7

5 0

1 0 1

WT

Y

5.8

3.2

18.5

N

N/A

N

N

N

1

2 7

5 0

1 1 2

WT

N

N

N/A

N

N

N

1

2 7

5 0

1 1 2

WT

Y

5.9

2.8

16.4

N

N/A

N

N

N

1

2 7

5 0

1 0 4

WT

N

4.6

2.6

12

N

N/A

N

N

N

1

2 7

5 0

1 1 2

WT

N

6.1

3.4

20.7

N

N/A

N

N

N

1

2 7

5 0

1 0 4

W (m)

N

3.8

L (m)

D1 9A

EM

N

Dia. (m)

Tra ct 15 A

Int. Post Dia. (m)

Dom estic

Int. Posts (n)

N

Wall Post Spac. (m)

17 6

EM

Wall Post Dia. (m)

Tra ct 15 A

Loh mann 3 Loh mann Stirli ng

Wall Posts (n)

Dom estic

WT

Posts (n)

N

Complete

20 4A

Wall Type

Tra ct 15 A

Class

EM

Other Loh mann

Shape

Mound Top

Str. No.

Dom estic

Occ.

N

Builds

Cahokia

69

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Burned

Cahokia

EM

Tra ct 15 A

Basin

108 9

Cahokia

Locus

100 2

Cahokia

Phase

104 7

Site

Str. ID 991

Loh mann 2

4.9

2.3

3

375

11.3

7.1

2.2

7.7

0

N

23

N

0

N

10

20 1

N

Dom estic

104 5

Cahokia

EM

Other Loh mann

EM

Other Loh mann

Tra ct 15 A

un no. Str E

N

Dom estic

Tra ct 15 A

20 3

N

Dom estic

Tra ct 15 A

D3 0

N

Dom estic

106 8

104 6

109 2

997

101 6

105 5

207

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

EM

Other Loh mann Loh mann Stirli ng

EM

Loh mann 3

Tra ct 15 A

10 8

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

60

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

22 0

N

Dom estic

EM

Loh mann 1

Tra ct 15 A

10 1

N

Dom estic

EM

15.4

WT

Y

4.4

2.3

10.1

WT

Y

3.2

1.4

WT

N

4

WT

N

WT

Y

WT

N

WT

Y

Unid

N

WT

Y

WT

N

WT

N

5.9

2.7

15.9

Posts

Y

4.9

2.5

12.3

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

97

N

N

1

2 7

5 0

1 1 2

8

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 1 2

Y

N

2 7

5 0

1 0 4

N

N/A

0

N

N/A

2

N

N/A

0

5.6

N

N/A

2.7

10.8

N

N/A

4.8

1.5

7.2

0

N

N/A

0

0

N

6.5

2.9

18.9

0

N

N/A

0

0

N

N

N/A

N

N/A

N

N/A

N

N/A

N

N/A

N

N/A

N

N/A

6.1

3.4

3.2

1.8

19.5

6.1

0

0

0

10

376

Pd ID

Dom estic

EM

Tra ct 15 A Tra ct 15 A

2.7

Comp ID

N

Cahokia

Other Loh mann

5.7

0

Orient. (deg E of N)

22 9

106 0

Y

0

Entry Trenches

Dom estic

WT

Int. Bur. (n)

N

0

Int. Feat. (n)

un no. Str J

3.1

Hearth Type

Tra ct 15 A

N

Int. Hearth

EM

Other Loh mann

WT

Int. Walls (n)

Dom estic

Area (m2)

N

W (m)

73

L (m)

Tra ct 15 A

Dia. (m)

EM

Other Loh mann

Int. Post Dia. (m)

Dom estic

Int. Posts (n)

N

Wall Post Spac. (m)

15 8

Wall Post Dia. (m)

Tra ct 15 A

Wall Posts (n)

EM

Other Loh mann

Posts (n)

Dom estic

Complete

N

Wall Type

87

Class

Tra ct 15 A

Shape

Mound Top

Str. No.

EM

Other Loh mann

Occ.

Dom estic

Site ID

Cahokia

N

Builds

Cahokia

30 4

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Tsha pe d Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Burned

107 1

Cahokia

EM

Tra ct 15 A

Basin

101 9

Cahokia

Locus

103 7

Cahokia

Phase

102 2

Site

Str. ID 106 5

Other Loh mann

N

94

N

0

N

100

N

0

0

N

83

N

0

0

0

0

N

6

Y

N

1

2 7

5 0

1 0 3

N

96

N

N

1

2 7

5 0

1 1 2

N

10

N

N

1

2 7

5 0

1 1 2

Y

N

1

2 7

5 0

1 0 1

N

EM

102 6

Cahokia

EM

Other Loh mann

100 5

108 1

104 3

102 8

106 9

108 0

992

101 7

212

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

D1 3A

N

Dom estic

Tra ct 15 A

20 6B

N

Dom estic

20 4B

N

Dom estic

11 2

N

Unkn own/ other

Tra ct 15 A

D1 8

N

Dom estic

Tra ct 15 A

16 4

N

Dom estic

Tra ct 15 A Tra ct 15 A

EM

Loh mann 3 Loh mann Stirli ng

EM

Other Loh mann

Tra ct 15 A

17 7

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

12 2

N

Dom estic

Tra ct 15 A

un no. Str F

N

Dom estic

Tra ct 15 A

16 3

N

Dom estic

EM

EM

Other Loh mann Loh mann Stirli ng

EM

Loh mann 2

Tra ct 15 A

20 6A

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

68

N

Dom estic

EM

Loh mann 1

Tra ct 15 A

15 6

N

Dom estic

EM

Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

4.4

2.5

11

N

N/A

N

N

N

1

2 7

5 0

1 1 2

WT

Y

5

3

15

N

N/A

N

N

N

3

2 7

5 0

1 0 4

WT

N

4.5

2.2

9.9

0

N

N/A

0

0

N

15

N

N

1

2 7

5 0

1 0 2

WT

Y

5.2

2.7

14

0

N

N/A

1

0

N

3

N

N

1

2 7

5 0

1 1 2

WT

Y

18.9

N

N/A

1

N

N

N

1

2 7

5 0

1 1 2

WT

N

26.3

N

N/A

N

N

N

1

2 7

5 0

1 0 3

WT

N

N

N/A

N

N

N

1

2 7

5 0

1 0 4

WT

N

N

N/A

N

N

N

1

2 7

5 0

1 1 2

WT

Y

N

N/A

N

N

N

1

2 7

5 0

1 1 2

WT

N

N

N/A

N

Y

N

1

2 7

5 0

1 1 2

WT

Y

5.2

3.3

17.2

0

N

N/A

0

0

N

2

N

N

1

2 7

5 0

1 0 4

WT

N

4.5

2.5

11.3

0

N

N/A

0

0

N

15

N

N

2

2 7

5 0

1 0 2

WT

Y

5.3

2.3

12.2

0

N

N/A

0

0

N

100

N

N

1

2 7

5 0

1 1 2

WT

Y

4.8

2.1

10.1

0

N

N/A

0

0

N

14

N

N

1

2 7

5 0

1 0 1

L (m)

W (m)

Y

Dia. (m)

WT

Posts (n)

Shape

Class

Mound Top

Str. No.

Tra ct 15 A

Pd ID

Cahokia

Other Loh mann

Dom estic

Comp ID

104 8

Locus

Phase

Occ. EM

Loh mann 2

N

Site ID

Cahokia

EM

26

Builds

Cahokia

EM

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Burned

993

Cahokia

Tra ct 15 A

Basin

108 3

Site

Str. ID 100 8

Other Loh mann Loh mann Stirli ng

4.9

7.1

3.7

3.4

4.1

377

2

8.2

1

94

996

103 1

Cahokia

Cahokia

Cahokia

Unkn own/ other

Tra ct 15 A

D1 3B

N

Dom estic

EM

EM

Loh mann 3

Tra ct 15 A

61

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

12 6

N

Dom estic

12 5

N

Dom estic

21 6

N

Unkn own/ other

103 0

Cahokia

EM

Other Loh mann

105 2

Cahokia

EM

Other Loh mann

Tra ct 15 A Tra ct 15 A

EM

Other Loh mann

Tra ct 15 A

88

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

20 5

N

Dom estic

EM

Loh mann 3

Tra ct 15 A

20 2

N

Dom estic

Loh mann 3

21

N

Dom estic

21 0

N

Unkn own/ other

Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Irr eg ula r

37

N

Unkn own/ other

Co mp lex

21 5

N

Unkn own/ other

61

N

Dom estic

Cir cul ar Re cta ng ula r

102 3

104 9

100 3

Cahokia

Cahokia

Cahokia

995

Cahokia

EM

218

Cahokia

EM

107 4

Cahokia

EM

105 1

Cahokia

EM

107 7

Cahokia

EM

Loh mann 1 Loh mann Stirli ng Other Loh mann Loh mann Stirli ng

Tra ct 15 A Tra ct 15 A Tra ct 15 A Tra ct 15 A Tra ct 15 A

N

N

N

1

2 7

5 0

1 0 4

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 0 4

N

Y

N

1

2 7

5 0

1 0 3

N

N

N

1

2 7

5 0

1 1 2

0

N

N

N

1

2 7

5 0

1 1 2

0

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

100

N

N

1

2 7

5 0

1 1 2

12

N

N

1

2 7

5 0

1 0 3

N

N

1

2 7

5 0

1 0 3

Y

N

1

2 7

5 0

1 0 1

N

N

N

1

2 7

5 0

1 0 4

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 0 4

WT

N

6.5

4.2

27.3

N

N/A

WT

N

4.1

2.1

8.6

N

N/A

WT

Y

5.6

3.2

17.9

N

N/A

WT

Y

5.2

3.4

17.7

N

N/A

WT

Y

5.3

2.6

13.8

N

N/A

WT

Y

4.6

2.7

12.4

0

N

N/A

1

WT

N

0

N

N/A

0

WT

Y

3.5

1.5

5.3

N

N/A

WT

Y

3.9

2.2

8.6

0

N

N/A

0

0

N

WT

Y

5.5

2.9

16

0

N

N/A

1

0

N

WT

Y

7.2

3.7

26.6

N

N/A

Unid

Y

N

N/A

0

WT

Y

N

N/A

1

WT

Y

N

N/A

1

WT

Y

N

N/A

1

0

25

3

7.1

5.2

378

3.4

17.7

0

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

N

Pd ID

108 4

12 3

Other Loh mann Loh mann Stirli ng

Comp ID

EM

Dom estic

Site ID

Cahokia

N

Builds

102 9

D1 7

Re cta ng ula r Irr eg ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Burned

EM

Tra ct 15 A Tra ct 15 A

Basin

Cahokia

Phase

Occ.

Site

Str. ID 108 8

Loh mann Stirli ng

1

1

N

0

0

N

10

103 2

102 4

Cahokia

Cahokia

Cahokia

EM

Other Loh mann

Tra ct 15 A

13 6

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

91

N

Dom estic

10 9

N

Dom estic

20 7

N

Dom estic

16 7

N

Unkn own/ other

23 4

N

Dom estic

17 2

N

Unkn own/ other

210

Cahokia

EM

Loh mann 1

105 0

Cahokia

EM

Other Loh mann

100 1

Cahokia

EM

Loh mann 3

Tra ct 15 A Tra ct 15 A Tra ct 15 A

106 4

Cahokia

EM

Other Loh mann

216

Cahokia

EM

Loh mann 1

Tra ct 15 A Tra ct 15 A

EM

Other Loh mann

Tra ct 15 A

un no. Str A

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

78

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

97

N

Dom estic

106 6

102 0

102 5

Cahokia

Cahokia

Cahokia

Cir cul ar Re cta ng ula r Un def ine d Re cta ng ula r Re cta ng ula r Re cta ng ula r

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Dom estic

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 0 3

N/A

N

Y

N

1

2 7

5 0

1 0 4

N

N/A

N

N

N

1

2 7

5 0

1 1 2

9.9

N

N/A

N

Y

N

1

2 7

5 0

1 0 4

3.2

18.9

N

N/A

2

N

N

N

1

2 7

5 0

1 1 2

3.9

2.1

8.2

0

N

N/A

3

0

N

13

N

N

1

2 7

5 0

1 1 2

Y

2.7

2.7

7.3

0

N

N/A

0

0

N

18

N

N

1

2 7

5 0

1 0 1

WT

N

6

2.6

15.6

0

N

N/A

0

0

N

11

N

N

1

2 7

5 0

1 1 2

WT

Y

N

N/A

1

N

N

1

2 7

5 0

1 0 3

WT

N

0

N

N/A

0

0

N

12

N

N

1

2 7

5 0

1 1 2

Unid

N

0

N

N/A

0

0

N

94

Y

N

1

2 7

5 0

1 0 1

WT

Y

5.4

2.5

13.5

0

N

N/A

0

0

N

94

N

N

1

2 7

5 0

1 1 2

WT

Y

4

2.2

10.3

N

N/A

1

N

N

N

1

2 7

5 0

1 1 2

WT

Y

5.6

3

16.8

N

N/A

1

N

N

N

1

2 7

5 0

1 1 2

W (m)

N

1

L (m)

D1 4

N

Dia. (m)

Tra ct 15 A

Int. Post Dia. (m)

Dom estic

Int. Posts (n)

N

Wall Post Spac. (m)

16 6A

EM

EM

Wall Post Dia. (m)

Tra ct 15 A

Other Loh mann Loh mann Stirli ng

Wall Posts (n)

Dom estic

N

Posts (n)

N

Complete

D1 5

Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Tsha pe d

Wall Type

Mound Top

Str. No.

Tra ct 15 A

Locus

Unkn own/ other

Pd ID

108 5

Cahokia

EM

N

Comp ID

104 0

Cahokia

11 1

Site ID

108 6

Loh mann 3 Loh mann Stirli ng

Dom estic

Builds

EM

N

Burned

Cahokia

14 2

Re cta ng ula r

Tra ct 15 A Tra ct 15 A

Basin

999

Shape

EM

Class

Cahokia

Phase

Occ.

Site

Str. ID 103 5

Other Loh mann

WT

N

4.2

2

WT

Y

WT

Y

WT

N

WT

Y

4.3

2.3

WT

Y

5.9

WT

Y

Posts

2.7

6.8

3.5

8.4

0

N

N/A

0

0

N

5.7

0

N

N/A

1

0

N

23.8

3.1

3.4

9.1

1.9

1.6

3

6

379

111

N

215

107 8

102 7

105 4

107 9

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

EM

Other Loh mann

Tra ct 15 A

56

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

86

N

Dom estic

EM

Loh mann 3

Tra ct 15 A

D1 2

N

Dom estic

Tra ct 15 A

17 3

N

Dom estic

Tra ct 15 A

62

N

Dom estic

Tra ct 15 A

11 8

N

Dom estic

Tra ct 15 A

21 9

N

Dom estic

Tra ct 15 A

16 2

N

Dom estic

EM

Loh mann 1 Loh mann Stirli ng

EM

Other Loh mann

EM

EM

EM

Other Loh mann Loh mann Stirli ng

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

96

N

N

1

2 7

5 0

1 1 2

96

Y

N

1

2 7

5 0

1 0 1

N

N

N

1

2 7

5 0

1 0 4

N/A

N

N

N

1

2 7

5 0

1 1 2

N

N/A

N

N

N

1

2 7

5 0

1 1 2

N

N/A

N

Y

N

1

2 7

5 0

1 0 3

N

N/A

0

Y

N

1

2 7

5 0

1 0 1

N

N/A

1

N

N

1

2 7

5 0

1 0 4

0

N

N/A

0

0

N

3

Y

N

1

2 7

5 0

1 1 2

0

N

N/A

0

0

N

103

N

N

1

2 7

5 0

1 1 2

N

N/A

N

94

N

N

2

2 7

5 0

1 0 4

WT

Y

3.4

1.3

4.4

N

N/A

WT

Y

4.7

2.8

13.2

N

N/A

WT

Y

4.6

2.4

11

N

N/A

WT

Y

6.2

2.9

18

1

N

N/A

0

0

N

WT

Y

4.7

2.5

13.2

1

N

N/A

6

0

N

WT

Y

3.8

3.3

12.6

N

N/A

1

WT

N

7.1

4.1

29.1

N

WT

Y

4.9

2.2

10.8

WT

N

7.4

3.5

25.9

WT

Y

6.5

4.3

27.4

WT

Y

3.7

1.9

7.2

WT

Y

3.6

1.9

6.8

WT

Y

4.9

2.2

10.8

WT

N

3.4

380

Orient. (deg E of N)

Dom estic

Entry Trenches

N

Int. Bur. (n)

28

EM

Int. Feat. (n)

Tra ct 15 A

Hearth Type

Dom estic

Int. Hearth

N

Int. Walls (n)

15 7

EM

Area (m2)

Tra ct 15 A

Loh mann 1 Loh mann Stirli ng

W (m)

Dom estic

L (m)

N

Dia. (m)

38

Int. Post Dia. (m)

Tra ct 15 A

Int. Posts (n)

EM

Other Loh mann

Wall Post Spac. (m)

Dom estic

Wall Post Dia. (m)

N

Wall Posts (n)

13 7

Posts (n)

Tra ct 15 A

Complete

EM

Other Loh mann

Wall Type

Dom estic

Class

N

Shape

Mound Top

Str. No. 25

Occ.

Tra ct 15 A

Pd ID

100 4

Cahokia

EM

Other Loh mann

Comp ID

102 1

Cahokia

Dom estic

Site ID

101 5

Cahokia

N

Builds

107 2

Cahokia

14 9

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Burned

213

Cahokia

EM

Tra ct 15 A

Basin

101 1

Cahokia

Locus

103 3

Cahokia

Phase

100 7

Site

Str. ID 103 6

Other Loh mann

0

1

0

N

10

N

EM

104 2

Cahokia

EM

Other Loh mann

100 6

108 2

104 4

107 0

106 2

205

109 1

209

106 3

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

54

N

Dom estic

Tra ct 15 A

34

N

Dom estic

70

N

Dom estic

17 1

N

Unkn own/ other

Tra ct 15 A

23

N

Dom estic

Tra ct 15 A

17 41 75

N

Dom estic

Tra ct 15 A Tra ct 15 A

EM

Other Loh mann Loh mann Stirli ng

EM

Other Loh mann

Tra ct 15 A

17 8

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

un no. Str I

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

23 1

N

Dom estic

Tra ct 15 A

66

N

Dom estic

Tra ct 15 A

D2 3

N

Dom estic

EM

EM

Loh mann 1 Loh mann Stirli ng

EM

Loh mann 1

Tra ct 15 A

10 3

N

Dom estic

EM

Other Loh mann

Tra ct 15 A

23 2

N

Dom estic

EM

Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Y

5.2

2.4

12.5

WT

Y

6.8

3.9

Both

Y

5.3

3.6

WT

N

WT

Y

WT

Y

4.2

2.9

12.2

WT

Y

6.7

3.8

25.5

WT

N

WT

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 0 4

N

N

N

1

2 7

5 0

1 1 2

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

N

N

N

1

2 7

5 0

1 1 2

1

26.4

N

N/A

1

19.1

N

N/A

5

0

N

N/A

0

0

N

0

N

N/A

0

0

N

N/A

N

N/A

3.2

N

N

2.4

WT

N

2.2

WT

Y

5

2.6

WT

N

5.8

WT

Y

4

WT

N

7.1

N

100

100

N

94

N

N

1

2 7

5 0

1 0 4

N/A

N

94

N

N

1

2 7

5 0

1 1 2

N

N/A

N

93

N

N

1

2 7

5 0

1 1 2

N

N/A

N

97

N

N

1

2 7

5 0

1 1 2

13

N

N/A

N

Y

N

1

2 7

5 0

1 0 1

2.9

16.8

N

N/A

N

N

N

2

2 7

5 0

1 0 4

2.1

8.4

0

N

N/A

0

0

N

100

N

N

1

2 7

5 0

1 0 1

0

N

N/A

0

0

N

102

N

N

1

2 7

5 0

1 1 2

2.7

381

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

N

N/A

3

0

N

N

2.7

0

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Tra ct 15 A

WT

Pd ID

Cahokia

Other Loh mann

Dom estic

Comp ID

101 8

Locus

Phase

Occ. EM

Other Loh mann

N

Site ID

Cahokia

EM

22 4B

Builds

Cahokia

EM

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Burned

100 9

Cahokia

Tra ct 15 A

Basin

107 6

Site

Str. ID 105 9

Other Loh mann Loh mann Stirli ng

0

1

0

Cahokia

1 1 2

N

N/A

N

Y

Y

1

2 7

5 0

1 0 4

N

N/A

0

N

N

1

2 7

5 0

1 1 2

13.2

N

N/A

4

N

Y

N

1

2 7

5 0

1 0 2

2.8

14.8

N

N/A

N

N

N

1

2 7

5 0

1 0 4

6.5

3.8

24.7

N

N/A

1

N

N

N

1

2 7

5 0

1 1 2

Y

4.6

2.4

11

N

N/A

0

N

N

1

2 7

5 0

1 0 1

WT

N

5.4

3

16.2

N

N/A

N

N

N

1

2 7

5 0

1 0 4

Dom estic

Re cta ng ula r

Posts

Y

3.8

1.9

7.2

N

N/A

N

N

N

1

2 7

4 9

1 0 0

N

Unkn own/ other

Un def ine d

Posts

N

N

N/A

N

N

N

1

2 7

4 9

1 0 0

30 8

N

Unkn own/ other

Posts

N

N

N/A

N

N

N

1

2 7

4 9

1 0 0

22 5

N

Unkn own/ other

Posts

N

N

N/A

Y

N

1

2 7

4 9

1 0 0

D3 3

N

Dom estic

Tra ct 15 A

51

N

Dom estic

Tra ct 15 A

14 5

N

Dom estic

Tra ct 15 A

D1 6

N

Dom estic

N

203

Cahokia

LW

201

Cahokia

LW

Tra ct 15 A

11 4

Tra ct 15 A

14 1

Un def ine d Un def ine d

L (m)

Tra ct 15 A

Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Class Unkn own/ other

Occ.

N

Tra ct 15 A Tra ct 15 A

Orient. (deg E of N)

5 0

41

LW

Entry Trenches

2 7

Dom estic

LW

Int. Bur. (n)

2

N

EM

Int. Feat. (n)

N

22 3

EM

Hearth Type

N

Dom estic

Loh mann 1 Loh mann Stirli ng Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi

Int. Hearth

110

N

EM

Int. Walls (n)

N

33

Other Loh mann

Area (m2)

N/A

W (m)

N

Dia. (m)

1 1 2

Tra ct 15 A

EM

Int. Post Dia. (m)

5 0

Dom estic

Loh mann 2 Loh mann Stirli ng

Int. Posts (n)

2 7

N

Tra ct 15 A Tra ct 15 A

Wall Post Spac. (m)

1

16 6

Other Loh mann

Wall Post Dia. (m)

N

Tra ct 15 A

EM

Wall Posts (n)

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Posts (n)

3

N

EM

Complete

N

Shape

N/A

EM

un no. Str D

Other Loh mann Loh mann Stirli ng

Wall Type

Mound Top

1 0 1

Str. No.

Pd ID

195

Cahokia

Comp ID

186

Cahokia

Site ID

108 7

5 0

Tra ct 15 A

EM

Cahokia

2 7

Other Loh mann

Cahokia

211

1

Dom estic

990

Cahokia

N

N

EM

101 2

N

10 2

Cahokia

Cahokia

85

EM

Tra ct 15 A

105 7

109 3

Builds

Cahokia

Burned

Cahokia

Basin

107 3

Cahokia

N

Locus

103 9

Cahokia

Phase

106 7

Site

Str. ID 208

Loh mann 1

WT

Y

3.2

2.2

WT

N

3

3.4

WT

N

6.1

3

WT

Y

6.9

WT

Y

3

WT

Y

N

N/A

N

18.3

4.3

29.8

1.9

5.7

WT

N

5.3

WT

Y

WT

4.1

7

0

0

0

0

382

0

0

0

0

0

0

N

N

N

103

95

60

Orient. (deg E of N)

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

N

N

1

2 7

4 9

9 9

N

N

1

2 7

4 9

9 9

N

N

N

1

2 7

4 9

9 9

N

Y

N

1

2 7

4 9

9 9

Y

N

1

2 7

4 9

1 0 0

N

N

2

2 7

4 9

9 8

0

N

N

1

2 7

4 9

1 0 0

7

N

N

1

2 7

4 9

1 0 0

N

N

1

2 7

4 9

1 0 0

N

N

1

2 7

4 9

9 9

N

N

N

1

2 7

4 9

1 0 0

N

N

N

2

2 7

4 9

1 0 0

N

N

2

2 7

4 9

1 0

ssippi an

180

Cahokia

LW

Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 3 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Emer gent Missi ssippi an 3 Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an

183

Cahokia

LW

Other Emer

165

159

164

171

187

148

194

198

184

160

193

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

LW

LW

LW

LW

LW

LW

LW

LW

LW

LW

LW

Tra ct 15 A

79

N

Unkn own/ other

Tra ct 15 A

43

N

Dom estic

Tra ct 15 A

76

N

Dom estic

Tra ct 15 A

21 3

N

Unkn own/ other

Irr eg ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Posts

N

Both

N

Posts

Y

Unid

N

Posts

Y

2.3

1.5

3.5

Posts

Y

3.7

1.9

7

63

60

0.23

3

4.9

2.5

12.2

3.2

1.9

6.1

Tra ct 15 A

12 0

N

Dom estic

Tra ct 15 A

81

N

Dom estic

Re cta ng ula r Re cta ng ula r

Dom estic

Re cta ng ula r

Posts

N

3

Dom estic

Re cta ng ula r

Posts

Y

1.9

Posts

N

Posts

Y

Tra ct 15 A

Tra ct 15 A

14 0

15 5

N

N

Tra ct 15 A

10 4

N

Dom estic

Tra ct 15 A

44

N

Dom estic

Re cta ng ula r Re cta ng ula r

Dom estic

Re cta ng ula r

Posts

Tra ct 15 A

13 4

N

Tra ct 15 A

71

N

Dom estic

Re cta ng ula r

Tra ct

84

N

Dom estic

Re cta

1.6

3

N

N/A

N

N/A

N

N/A

N

N/A

N

N/A

N

N/A

0

N

N/A

0

0

N

0

N

N/A

0

0

N

N

N/A

N

N/A

0

0

1.4

44

44

0.24

0

3.3

1.9

6.3

Y

2.3

2.1

4.8

N

N/A

Posts

Y

4.1

2

8.2

N

N/A

Posts

Y

2.5

2

5

N

N/A

383

0

N

0

0

0

0

N

N

18

90

N

N

1

3

0

N

N

18

35

168

204

157

188

144

169

151

202

172

158

156

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

LW

LW

LW

LW

LW

LW

LW

LW

LW

LW

LW

LW

15 A

ng ula r

Tra ct 15 A

16 0

N

Dom estic

Tra ct 15 A

10 0

N

Dom estic

Tra ct 15 A

31 1

N

Dom estic

Tra ct 15 A

19

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Tra ct 15 A

12 7

N

Dom estic

Tra ct 15 A

47

N

Dom estic

Tra ct 15 A

12 1

N

Dom estic

Tra ct 15 A

90

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Tra ct 15 A

23 6

N

Unkn own/ other

Un def ine d

Tra ct 15 A

27

N

Dom estic

Tra ct 15 A

20

N

Dom estic

Tra ct 15 A

18

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r

Pd ID

Comp ID

Site ID

Builds

Burned

Basin

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID 199

gent Missi ssippi an Other Emer gent Missi ssippi an Emer gent Missi ssippi an 3 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 3 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 2 Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 3

0

Posts

Y

3.9

1.9

7.2

Posts

Y

4

2.2

8.8

Posts

N

2.8

2.1

5.9

Posts

Y

4.5

2.2

10

Posts

N

3.3

2

Posts

Y

2.7

Posts

N

Posts

Y

Posts

N

Posts

Y

3.4

1.9

Posts

Y

4

Posts

Y

4.1

60

49

52

49

0.26

8

0.22

0

49

43

0.3

6

384

1

2 7

4 9

1 0 0

N

N

2

2 7

4 9

9 9

90

Y

N

1

2 7

4 9

1 0 0

97

N

N

1

2 7

4 9

9 9

N

Y

N

1

2 7

4 9

1 0 0

N

N

N

1

2 7

4 9

9 8

Y

N

1

2 7

4 9

9 9

N

N

1

2 7

4 9

9 8

Y

N

1

2 7

4 9

1 0 0

N

N

N

1

2 7

4 9

1 0 0

N

N

N

1

2 7

4 9

9 9

Y

N

1

2 7

4 9

9 9

N

N/A

0

N

N/A

0

0

N

1

N

N/A

1

0

N

6.6

N

N/A

1.8

4.9

N

N/A

3.2

2.3

7.4

N

N/A

0

3.7

2.1

7.8

N

N/A

2

N

N/A

0

6.5

N

N/A

1

2.2

8.8

N

N/A

1

2.4

9.8

N

N/A

0

0

0

0

N

N/A

0

0

N

N

N

2

N

0

N

90

N

0

0

N

N

2

104

178

190

Cahokia

Cahokia

Cahokia

LW

LW

LW

192

Cahokia

LW

153

Cahokia

LW

N

Posts

Y

2.6

2.2

5.7

Posts

Y

3.7

1.9

Posts

Y

3.7

Posts

Y

Posts

Tra ct 15 A

83

N

Dom estic

Tra ct 15 A

82

N

Dom estic

Tra ct 15 A

92

N

Dom estic

Tra ct 15 A

74

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

W (m)

L (m)

Orient. (deg E of N)

Posts

Entry Trenches

Un def ine d

Int. Bur. (n)

N

Unkn own/ other

Int. Feat. (n)

8.6

Hearth Type

2.1

Int. Hearth

4.1

Area (m2)

N

Dia. (m)

Posts

Posts (n)

Dom estic

Complete

N

Re cta ng ula r

Int. Walls (n)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Mound Top

Str. No.

Locus

Phase

Occ. LW

4.3

Pd ID

Cahokia

LW

1.6

Comp ID

150

Cahokia

LW

21 8

2.7

Site ID

181

Cahokia

LW

Tra ct 15 A

Y

Builds

163

Cahokia

LW

15 3

Posts

N

Burned

152

Cahokia

LW

Tra ct 15 A

Dom estic

Re cta ng ula r

Basin

149

Cahokia

LW

75

Wall Type

182

Cahokia

LW

Shape

200

Cahokia

Tra ct 15 A

Class

196

Site

Str. ID 146

Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Emer gent Missi ssippi an 2 Emer gent Missi ssippi an 2 Emer gent Missi ssippi an 3 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Emer gent Missi ssippi

N

N

1

2 7

4 9

9 8

100

Y

N

1

2 7

4 9

1 0 0

N

100

Y

N

1

2 7

4 9

1 0 0

N

35

N

N

2

2 7

4 9

1 0 0

N

N

N

2

2 7

4 9

9 8

N

N

N

2

2 7

4 9

9 8

N

N

N

1

2 7

4 9

9 9

N

N

N

1

2 7

4 9

1 0 0

N

N

N

1

2 7

4 9

9 8

N

N/A

N

0

N

N/A

0

0

N

0

N

N/A

0

0

0

N

N/A

0

0

7

N

N/A

2.1

7.8

N

N/A

2.9

1.6

4.6

N

N/A

Y

2.9

1.5

4.4

N

N/A

Posts

Y

2.5

2

5

N

N/A

2

Tra ct 15 A

72

N

Dom estic

Tra ct 15 A

89

N

Dom estic

Re cta ng ula r Re cta ng ula r

Dom estic

Re cta ng ula r

Posts

Y

3.7

1.9

7

0

N

N/A

0

0

N

98

N

N

1

2 7

4 9

1 0 0

Dom estic

Re cta ng ula r

Posts

Y

4.1

2.5

10.3

0

N

N/A

2

0

N

10

N

N

3

2 7

4 9

1 0 0

Posts

Y

2.2

1.8

4

N

N/A

N

N

N

1

2 7

4 9

1 0 0

Posts

Y

3.9

2

7.8

N

N/A

N

N

N

2

2 7

4 9

9 8

Tra ct 15 A

Tra ct 15 A

Tra ct 15 A Tra ct 15 A

59

13 1

N

N

13 3

N

Dom estic

98

N

Dom estic

Re cta ng ula r Re cta ng ula

385

1

197

175

185

173

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

LW

LW

LW

LW

LW

LW

N

4.4

2.1

9.2

Posts

Y

4

2.3

9.2

Posts

Y

3.2

2.5

8

Posts

Y

4.5

3.1

Posts

Y

3.4

W (m)

L (m)

Orient. (deg E of N)

Posts

Entry Trenches

7.8

Int. Bur. (n)

1.7

Int. Feat. (n)

4.6

Hearth Type

Y

Int. Hearth

Posts

Area (m2) 6.6

Dia. (m)

1.7

Int. Walls (n)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

3.8

Posts (n)

Y

Complete

Posts

Pd ID

147

Cahokia

LW

7.2

Comp ID

166

Cahokia

LW

2

Site ID

191

Cahokia

LW

3.6

Builds

155

Cahokia

LW

Y

Burned

174

Cahokia

LW

Posts

Basin

145

Cahokia

LW

Wall Type

189

Cahokia

Shape

176

Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 2 Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an Other Emer gent Missi ssippi an

Re cta ng ula r

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

an 2

N

N

2

2 7

4 9

1 0 0

N

N

1

2 7

4 9

1 0 0

87

N

Y

1

2 7

4 9

9 8

5

N

N

1

2 7

4 9

1 0 0

N

N

2

2 7

4 9

9 8

N

N

3

2 7

4 9

1 0 0

N

N

N

1

2 7

4 9

9 9

N

N

N

1

2 7

4 9

9 8

0

N

N

1

2 7

4 9

1 0 0

N

15

N

N

1

2 7

4 9

1 0 0

0

N

96

Y

N

1

2 7

4 9

1 0 0

0

N

102

N

N

1

2 7

4 9

1 0 0

r

Tra ct 15 A

45

N

Dom estic

Tra ct 15 A

12 9

N

Dom estic

Tra ct 15 A

63

N

Dom estic

Tra ct 15 A

39

N

Dom estic

Tra ct 15 A

11 511 7

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Tra ct 15 A

13 2

N

Dom estic

Tra ct 15 A

93

N

Dom estic

Tra ct 15 A

80

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r

Dom estic

Re cta ng ula r

Posts

Y

Dom estic

Re cta ng ula r

Posts

Y

Dom estic

Re cta ng ula r

Posts

Y

Dom estic

Re cta ng ula r

Posts

Y

Tra ct 15 A

Tra ct 15 A

Tra ct 15 A

Tra ct 15 A

15 4

40

10 7

29

N

N

N

N

2

50

32

42

42

0.22

0.3

3

0.29

386

0

N

N/A

N

N/A

0

N

N/A

0

0

N

0

N

N/A

1

0

N

N

N/A

N

N/A

2

14

N

N/A

1

2.1

7.1

N

N/A

3

2.8

1.8

5.3

0

N

N/A

0

0

N

3.6

1.9

6.9

0

N

N/A

1

0

3.9

2.3

9

0

N

N/A

0

4.1

1.9

7.8

0

N

N/A

1

0

1

0

N

101

N

N

0

N

10

154

162

179

161

167

143

177

111 8

110 3

111 7

111 2

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

EM

Moor ehead 1

Tra ct 15 A

9

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

11 9

N

Dom estic

EM

Moor ehead 2

Tra ct 15 A

30 2

N

Dom estic

13 5

N

Dom estic

30 7

N

Dom estic

LW

111 9

Cahokia

EM

Other Moor ehead

112 5

Cahokia

EM

Other Moor ehead

Dom estic

Tra ct 15 A

64

N

Dom estic

Tra ct 15 A

65

N

Dom estic

Tra ct 15 A

48

N

Dom estic

Tra ct 15 A

96

N

Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r

Tra ct 15 A

42

N

Stora ge

Sq uar e

Tra ct 15 A

46

N

Dom estic

Tra ct 15 A Tra ct 15 A

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula

Y

Posts

Y

Posts

N/A

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type Posts

N

Pd ID

Dom estic

LW

N

N

Comp ID

N

LW

10 5

Posts

Site ID

13 0

LW

Tra ct 15 A

Un ide ntif ied Re cta ng ula r Re cta ng ula r

Builds

Tra ct 15 A

LW

N

Unkn own/ other

Burned

EM

Other Moor ehead

LW

13 9

Shape

Class

Mound Top

Str. No.

Locus

Phase

LW

LW

Tra ct 15 A

Basin

Cahokia

Occ.

Site

Str. ID 170

Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 2 Emer gent Missi ssippi an 3 Other Emer gent Missi ssippi an Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 3 Emer gent Missi ssippi an 1 Other Emer gent Missi ssippi an

N

14

Y

N

1

2 7

4 9

9 9

4.1

2.1

8.6

0

N

N/A

0

0

N

93

Y

N

1

2 7

4 9

9 8

3.7

2.3

8.5

0

N

N/A

0

0

N

104

Y

N

1

2 7

4 9

9 9

Y

2.5

1.7

4.3

N

N/A

N

Y

N

1

2 7

4 9

1 0 0

Posts

Y

4.2

2.2

9.2

N

N/A

N

Y

N

1

2 7

4 9

9 9

Posts

Y

4

1.8

7.2

N

N/A

1

N

N

N

1

2 7

4 9

9 9

Posts

Y

2.4

2.3

5.5

0

N

N/A

1

0

N

3

N

N

1

2 7

4 9

9 7

Posts

Y

4.4

2.6

11.4

0

N

N/A

0

0

N

15

N

N

1

2 7

4 9

1 0 0

WT

Y

6.3

3.7

23.3

N

N/A

N

N

N

1

2 7

5 2

1 1 1

WT

Y

8

4.4

35.2

N

N/A

4

N

Y

N

1

2 7

5 2

1 0 9

WT

N

6.6

4.7

31

0

N

N/A

4

0

N

12

Y

N

1

2 7

5 2

1 1 1

WT

Y

5.3

3.9

20.7

0

N

N/A

2

0

N

90

Y

N

1

2 7

5 2

1 1 0

WT

N

4.2

3.3

13.9

N

N/A

N

N

N

2

2 7

5 2

1 1 1

WT

N

N

N/A

N

N

N

1

2 7

5 2

1 1 1

47

47

0.27

34

0.35

0

2

387

Orient. (deg E of N)

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

Y

N

1

2 7

5 2

1 1 0

N

Y

N

1

2 7

5 2

1 1 0

N

Y

N

1

2 7

5 2

1 1 0

10

N

N

1

2 7

5 2

1 1 1

N

10

Y

N

1

2 7

5 2

1 0 9

0

N

2

N

N

1

2 7

5 2

1 1 1

0

N

2

N

N

1

2 7

5 2

1 1 1

N

N

1

2 7

5 2

1 1 1

97

N

N

1

2 7

5 2

1 0 9

N

96

Y

N

1

2 7

5 2

1 0 9

0

N

96

Y

Y

1

2 7

5 2

1 0 9

0

0

N

94

N

N

1

2 7

5 2

1 1 0

N/A

2

0

N

90

N

N

1

2 7

5 2

N/A

0

0

N

23

N

N

1

2 7

5 2

r

111 4

110 9

111 0

112 4

110 6

112 2

112 1

111 5

110 5

110 7

110 4

111 1

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

EM

Moor ehead 2

Tra ct 15 A

30 5

N

Dom estic

EM

Moor ehead 2

Tra ct 15 A

10 B

N

Dom estic

EM

Moor ehead 2

Tra ct 15 A

35

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

30 1

N

Dom estic

EM

Moor ehead 1

Tra ct 15 A

77

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

15 1

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

15 0

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

31

N

Dom estic

EM

Moor ehead 1

Tra ct 15 A

67

N

Dom estic

EM

Moor ehead 1

Tra ct 15 A

95/ 99

N

Dom estic

EM

Moor ehead 1

Tra ct 15 A

32

N

Dom estic

EM

Moor ehead 2

Tra ct 15 A

14 3

N

Dom estic

Moor ehead 2 Moor ehead 1

Tra ct 15 A Tra ct 15

30 3

N

Dom estic

2

N

Dom estic

111 3

Cahokia

EM

110 2

Cahokia

EM

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng

WT

Y

8.1

6.6

53.5

WT

Y

10.3

4.9

WT

Y

5.7

WT

Y

WT

N

N/A

3

50.5

N

N/A

1

4.2

23.9

N

N/A

5

7.5

5.5

41.3

0

N

N/A

5

0

N

Y

7

4.3

30.1

0

N

N/A

8

0

WT

N

5

0

N

N/A

0

WT

N

5.7

0

N

N/A

0

WT

Y

6.3

5.2

32.8

N

N/A

6

WT

Y

6.9

4

27.6

0

N

N/A

0

0

N

Posts

N

0

N

N/A

0

0

WT

Y

9.3

5.2

48.4

0

Y

Fire basin

1

WT

Y

4.1

3.9

16

0

N

N/A

WT

N

5.7

4.4

25.1

0

N

WT

Y

5.9

4.9

28.9

1

N

388

1

0

N

10

N

1 1 0 1 0 9

110 1

109 5

109 6

109 7

Cahokia

Cahokia

Cahokia

Cahokia

Cahokia

109 8

Cahokia

528

ChotaTanase

397

ChotaTanase

396 393

ChotaTanase ChotaTanase

EM

Tra ct 15 A

14 4

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

11 3

N

Dom estic

EM

Moor ehead 2

Tra ct 15 A

10 A

N

Dom estic

EM

Other Moor ehead

Tra ct 15 A

15 2

N

Dom estic

EM

Stirli ngMoor ehead

Tra ct 15 A

12 4

N

Dom estic

EM

Stirli ngMoor ehead

Tra ct 15 A

1

N

Dom estic

EM

Stirli ngMoor ehead

Tra ct 15 A

30 6

N

Dom estic

EM

Stirli ng 1

Tra ct 15 A

20 9

N

Dom estic

EM

Stirli ng 2

Tra ct 15 A

D1 0

N

Dom estic

EM

Other Stirli ng

Tra ct 15 A

3

N

Nondome stic

EM

Other Stirli ng

Tra ct 15 A

14 8

N

Nondome stic

EM

Hiwa ssee Islan d

M R2

Fe atu re 1

N

Dom estic

HI

Cher okee

M R2

25

N

Dom estic

Cher okee Cher okee

M R2 M R2

20

N

10

N

HI

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

4.8

23

0

N

N/A

0

0

N

2

N

N

5

2 7

5 2

1 1 1

WT

Y

6.3

3.6

22.7

0

N

N/A

0

0

N

10

N

N

1

2 7

5 2

1 1 1

WT

Y

8.4

4.9

41.2

N

N/A

1

Y

N

2

2 7

5 2

1 1 0

WT

N

4.2

N

N/A

0

N

N

1

2 7

5 2

1 1 1

WT

Y

8.3

N

N/A

2

N

N

1

2 7

5 1

1 0 8

WT

N

N

N/A

0

WT

N

N

N/A

WT

Y

N

N/A

WT

N

N

N/A

WT

Y

18.6

12.3

228.3

1

N

N/A

7

0

N

WT

N

9.9

9.5

94.1

0

N

N/A

2

0

Both

Y

7.6

6.2

46.7

N

N/A

0

0

L (m)

W (m)

4.8

Dia. (m)

Y

Posts (n)

WT

ula r

Other Moor ehead

HI

Pd ID

109 9

Cahokia

Comp ID

110 0

Cahokia

Site ID

112 3

Cahokia

Builds

110 8

Cahokia

Burned

111 6

Cahokia

Basin

112 0

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

A

Dom estic Dom estic

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Cir cul ar Cir cul

0

4.6

38.2

0

3.9

2.2

8.6

0

3.6

10 0

80

0.1

0.19

20

0.15

0

N

0

N

N

0

0

N

3

N

Y

1

2 7

5 1

1 0 8

N

4

N

N

1

2 7

5 1

1 0 8

N

100

N

N

1

2 7

5 1

1 0 5

Y

N

1

2 7

5 1

1 0 6

95

N

N

1

2 7

5 1

1 0 7

N

2

N

N

1

2 7

5 1

1 0 7

N

0

N

N

1

3 5

6 3

6

3 5

6 4

1 1

3 5 3 5

6 4 6 4

1 1 1 1

N

Posts

N

36

0.16

0.16

7

38.6

N

N/A

Posts

N

22

0.16

0.16

6

29.2

N

N/A

0

0

N

Posts

Y

57

0.16

7

38.6

N

N/A

0

2

N

20

0.16

1.1

37

389

2

2

N

15

22

N

N

1

N

N

1

N

N

1

Orient. (deg E of N)

Burned

Builds

Site ID

Comp ID

Pd ID

N/A

3

0

N

180

N

N

1

3 5

6 4

1 1

6.1

61

N

N/A

3

6

N

3

N

N

1

3 5

6 4

1 1

6.2

4

24.4

N

N/A

1

0

N

100

N

N

1

3 5

6 4

1 1

0.16

13

6.1

79.3

N

N/A

0

1

N

105

N

N

2

3 5

6 4

1 1

0.18

9.8

5.5

53.5

N

N/A

0

3

N

15

N

N

1

3 5

6 4

1 1

4.6

4

18.1

N

N/A

0

0

N

10

N

N

1

3 5

6 4

1 1

38.6

N

1

0

N

115

N

N

1

3 5

6 4

1 1

38.6

Y

N/A Surface fired area

3

0

N

N

Y

1

3 5

6 4

1 1

50.5

Y

Fire basin

1

7

N

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

L (m)

W (m)

N

Dia. (m)

Basin

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

35.3

ar

398

ChotaTanase

HI

Cher okee

M R2

27

N

Dom estic

530

ChotaTanase

HI

Cher okee

M R6 2

6

N

Dom estic

416

ChotaTanase

HI

Cher okee

M R2

N

Dom estic

N

Nondome stic

527

ChotaTanase

HI

Cher okee

M R2

2 Su m me r Pa vili on

413

ChotaTanase

HI

Cher okee

M R2

26

N

Dom estic

417

ChotaTanase

HI

Cher okee

M R2

3

N

Dom estic

394

ChotaTanase

HI

Cher okee

M R2

12

N

Dom estic

401

ChotaTanase

HI

Cher okee

M R2

5

N

Dom estic

410

ChotaTanase

HI

Cher okee

M R2

13

N

Dom estic

403

ChotaTanase

HI

Cher okee

M R2

15

N

Dom estic

411

ChotaTanase

HI

Cher okee

M R2

19

N

Dom estic

392

ChotaTanase

HI

Cher okee

M R2

6

N

Dom estic

391

ChotaTanase

HI

Cher okee

M R2

1

N

Dom estic

529

ChotaTanase

HI

Cher okee

M R6 2

4

N

Dom estic

405

ChotaTanase

HI

Cher okee

M R2

22

N

Stora ge

409

ChotaTanase

HI

Cher okee

M R2

11

N

Dom estic

Cir cul ar Re cta ng ula r Re cta ng ula r

Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Cir cul ar Re cta ng ula r Cir cul ar Re cta ng ula r Cir cul ar Cir cul ar Re cta ng ula r Cir cul ar Re cta ng ula r

Posts

Y

35

17

0.13

1.4

18

0.13

Posts

Y

51

20

0.18

1.8

31

0.18

10

Posts

Y

46

22

0.17

2

24

0.15

Posts

Y

22 2

0.16

1.3

Posts

Y

51

0.18

Posts

Y

14

14

0.16

1.6

0

Posts

Y

48

17

0.19

1.3

31

0.19

7

Posts

Y

33

14

0.15

1.6

19

0.2

7

Posts

Y

18

1

0.19

2.4

7

0.22

Posts

Y

45

16

0.16

1.2

29

0.16

Posts

Y

53

46

0.18

0.91

7

0.2

Posts

Y

57

38

0.19

0.82

19

0.24

6

Posts

Y

40

13

0.16

1.7

27

0.16

7

Posts

Y

61

17

0.17

1.9

44

0.17

Posts

Y

10

9

0.18

0.79

1

0.21

Posts

Y

10 3

24

0.16

1.2

79

0.16

390

6.7

9.8

5.2

6.7

8.8

8.5

5.8

5.6

2.1

10.7

5.8

0

115

35.3

3

N

N/A

0

0

N

51.2

1

N

N/A

0

3

N

32.1

Y

Prepare d clay

1

0

N

36.9

Y

Prepare d clay

5

0

N

90

N

N

1

3 5

6 4

1 1

47.6

N

N/A

3

6

N

5

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

3.6

0

N

N/A

0

0

N

61.8

2

N

N/A

3

2

N

30

22

M R2

16

N

Dom estic

395

ChotaTanase

HI

Cher okee

18

N

Dom estic

399

ChotaTanase

HI

Cher okee

M R2 M R6 2

N

Dom estic

526

ChotaTanase

HI

Cher okee

N

Nondome stic

400

ChotaTanase

HI

Cher okee

M R2 M R6 2

3 To wn ho use 1-2

5

N

Dom estic

415

ChotaTanase

HI

Cher okee

M R2

8

N

Dom estic

402

ChotaTanase

HI

Cher okee

M R2

14

N

Dom estic

418

ChotaTanase

HI

Cher okee

M R2

9

N

Dom estic

412

ChotaTanase

HI

Cher okee

M R2

21

N

Dom estic

408

ChotaTanase

HI

Cher okee

M R2

4

N

Dom estic

414

ChotaTanase

HI

Cher okee

M R2

28

N

Dom estic

355

Coweeta Creek

HI

14

N

Dom estic

350

Coweeta Creek

LM

9

N

Dom estic

11

N

12

N

352 353

Site Coweeta Creek Coweeta Creek

LM LM

Late Quall a Early Quall a

Early Quall a Early Quall

Dom estic Dom estic

Cir cul ar Cir cul ar Re cta ng ula r Cir cul ar Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Sq uar e Re cta ng ula r Sq uar

Pd ID

Cher okee

Comp ID

HI

Site ID

ChotaTanase

Builds

404

Burned

Dom estic

Basin

N

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

N

N

1

3 5

6 4

1 1

30

N

N

1

3 5

6 4

1 1

N

5

N

N

1

3 5

6 4

1 1

0

N

100

N

N

2

3 5

6 4

1 1

3

1

N

0

N

N

2

3 5

6 4

1 1

N/A

1

0

N

88

N

N

1

3 5

6 4

1 1

N

N/A

0

0

N

N

N

1

3 5

6 4

1 1

23.7

Y

Prepare d clay

1

2

N

30

N

N

1

3 5

6 4

1 1

4.9

44.6

N

N/A

0

3

N

35

N

N

1

3 5

6 4

1 1

9.3

4.1

38.4

Y

Surface fired area

1

3

N

90

N

N

1

3 5

6 4

1 1

7.9

4.9

38.6

N

N/A

1

2

N

180

N

N

1

3 5

6 4

1 1

7

38.4

Y

Prepare d clay hearth

N

N

1

2 1

3 7

5 9

10

74.7

Y

Prepare d clay

N

N

1

2 1

3 5

5 7

5.8

37.1

N

N/A

N

N

1

9.1

82.8

U

N/A

N

N

1

2 1 2 1

3 5 3 5

5 7 5 7

Posts

Y

40

30

0.13

0.4

10

0.13

3.3

8.6

N

N/A

0

0

N

Posts

Y

13

10

0.18

0.87

3

0.2

2.6

5.3

N

N/A

1

0

N

Posts

Y

55

50

0.15

0.63

5

0.15

34.7

N

N/A

1

0

N

Posts

Y

69

22

0.16

0.95

47

0.16

7

35.3

Y

Prepare d clay

1

0

N

Posts

Y

77

26

0.16

1.3

51

0.16

7.2

40.7

2

N

N/A

0

0

N

Posts

Y

80

28

0.18

0.95

42

0.18

7

38.6

2

Y

Prepare d clay

1

0

Posts

Y

12 15

0.17

0.3

0.17

18.3

263

18

Y

Prepare d clay

2

Posts

Y

42

22

0.2

1.4

20

0.2

7

35.3

1

Y

Prepare d clay

Posts

Y

82

42

0.16

0.72

40

0.16

30.7

2

N

Posts

Y

26

16

0.16

1.4

10

0.2

19.6

Both

Y

35

20

0.16

1

15

0.16

5.2

4.6

Posts

N

18

0.16

9.1

Posts

Y

42

21

0.16

1.3

21

0.16

Posts

N

21

13

0.14

2

7

0.14

Posts

Y

Posts

N

Posts

Y

6.4

Posts

Y

9.1

0.16

391

Orient. (deg E of N)

2

Entry Trenches

M R6 2

Int. Bur. (n)

Cher okee

Int. Feat. (n)

HI

Hearth Type

ChotaTanase

Int. Hearth

419

Int. Walls (n)

Stora ge

Area (m2)

N

1

W (m)

23

N

L (m)

M R2

N

Dia. (m)

Cher okee

Int. Post Dia. (m)

HI

Int. Posts (n)

ChotaTanase

Wall Post Spac. (m)

406

Cir cul ar Cir cul ar Re cta ng ula r Cir cul ar Cir cul ar Cir cul ar

Wall Post Dia. (m)

Unkn own/ other

Wall Posts (n)

N

Posts (n)

1

Complete

M R6 2

Wall Type

Shape

Cher okee

Locus

HI

Phase

ChotaTanase

Occ.

Class

Mound Top

Str. No.

Str. ID 407

6.7

9.1

5.2

3.5

5.8

0

90

95

135

Coweeta Creek

LM

346

Coweeta Creek

LM

356

Coweeta Creek

LM

351

Coweeta Creek

LM

347

Coweeta Creek

LM

634

Coweeta Creek

LM

357

Coweeta Creek

LM

633

Coweeta Creek

LM

344

Coweeta Creek

LM

345

Coweeta Creek

LM

Midd le Quall a Midd le to Late Quall a Midd le Quall a Midd le Quall a

139

Dallas Site

LM

Yarn ell I

123

Dallas Site

498

Dallas Site

135

Dallas Site

134

Dallas Site

140

Dallas Site

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

5 7

135

N

N

4

2 1

3 5

5 7

Prepare d clay

90

N

N

1

2 1

3 6

5 8

Y

Prepare d clay

135

N

N

1

2 1

3 6

5 8

N

U

N/A

N

N

1

2 1

3 6

5 8

Posts

N

U

N/A

135

N

N

1

2 1

3 6

5 8

Posts

Y

6.1

6.1

37.2

Y

Prepare d clay

135

N

N

1

2 1

3 6

5 8

Posts

Y

14.9

6.7

99.8

N

N/A

N

N

2

2 1

3 6

5 8

Posts

Y

4.6

2.4

11

N

N/A

N

N

1

2 1

3 6

5 8

Sq uar e

Posts

Y

14.9

14.9

226.1

Y

Prepare d clay

Y

N

5

2 1

3 6

5 8

Sq uar e

Posts

N

6.4

6.4

41

Y

Prepare d clay

135

N

N

3

2 1

3 6

5 8

Posts

Y

5.5

5.5

30.3

Y

Prepare d clay

135

N

N

1

2 1

3 6

5 8

Posts

Y

4.3

4.3

18.2

Y

Prepare d clay

N

N

1

1 0

1 9

1

Y

Prepare d clay

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

L (m)

W (m)

3 5

Dia. (m)

2 1

Posts (n)

1

Shape

Pd ID

349

Comp ID

LM

Site ID

Coweeta Creek

Builds

348

Burned

LM

Basin

354

N

e

Early Quall a Early Quall a Midd le Quall a Midd le Quall a Midd le Quall a Midd le Quall a Midd le Quall a Midd le to Late Quall a

Coweeta Creek

N

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

a

13

N

Dom estic

7

N

Dom estic

Un def ine d Sq uar e

8

N

Dom estic

5

N

15

Posts

N

Posts

Y

Sq uar e

Posts

N

7

Dom estic

Sq uar e

Posts

N

7

N

Dom estic

Sq uar e

Posts

10

N

Dom estic

Sq uar e

6

N

Dom estic

2

N

Nondome stic

16

N

Stora ge

Sq uar e Re cta ng ula r Re cta ng ula r

1

N

Nondome stic

3

N

Dom estic

4

N

Dom estic

EU 8

258

N

Dom estic

LM

Yarn ell II

EU 7

247

LM

Yarn ell II

EU 7

317

LM

Yarn ell I

EU 7

487

LM

Yarn ell I

EU 7

437

LM

Yarn ell II

EU 8

368

N

Dom estic Nondome stic

N

Dom estic

N

Dom estic

N

Dom estic

N

Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e

Posts

Posts

Posts

Posts

Posts

N/A

60.8

Y

Prepare d clay

6.7

46.9

Y

7

49

8.8

24

Y

0.18

0.18

Y

0.18

Y

0.18

Y

Y

U

0.18

34

31

0.18

0.6

0.6

6.4

0.36

7

0.6

5.2

0.6

0.6

7

3

392

6.1

5.8

7

4.9

6.7

5.5

37.1

45

25.2

41.2

33.4

3

N

N/A

Y

Prepare d clay

Y

Prepare d clay

Y

Prepare d clay

N

135

N

4

4

N

2

N

Y

N

N

0

0

N

Y

Y

N

N

Y

LM

Yarn ell II

EU 8

138

499

Dallas Site

LM

Yarn ell II

EU 8

148

131

Dallas Site

LM

Yarn ell II

EU 7

407

126

Dallas Site

LM

Yarn ell II

EU 8

378

122

Dallas Site

LM

Yarn ell II

EU 7

177

127

Dallas Site

LM

Yarn ell II

EU 7

417

136

Dallas Site

LM

Yarn ell II

EU 7

197

129

Dallas Site

LM

Yarn ell I

EU 7

377

120

Dallas Site

LM

Yarn ell II

EU 7

121

Dallas Site

8-7

LM

Yarn ell II

EU 7

157

132

Dallas Site

LM

Yarn ell I

EU 7

107

138

Dallas Site

LM

Yarn ell II

EU 7

357

124

Dallas Site

EU 7

267

LM

Yarn ell II Gene ral Late Missi ssippi an

LM

130

Dallas Site

EU 8

128

Dallas Site

LM

Yarn ell II

137

Dallas Site

502

Dallas Site

141

Dallas Site

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

N

Dom estic

428

N

Dom estic

EU 7

137

N

Dom estic

LM

Yarn ell I

EU 7

347

N

Dom estic

LM

Yarn ell II

EU 7

297

N

Dom estic

LM

Yarn ell II

EU 8

418

N

Dom estic

N

Y

Sq uar e Sq uar e Sq uar e Un def ine d Sq uar e

Posts

Posts

Posts

Posts

Posts

Posts

Posts

Y

30

0.18

Y

Y

Y

0.25

34

24

45

Y

0.18

0.45

0.45

6.2

5.3

26.8

39

31.8

Orient. (deg E of N)

Entry Trenches

Y

Prepare d clay

Y

Prepare d clay

Y

Prepare d clay

N

N

Int. Bur. (n)

Y

Prepare d clay

Int. Feat. (n)

Int. Walls (n)

Area (m2)

W (m)

Dia. (m)

Int. Post Dia. (m)

L (m) 4.9

35.3

Y

Prepare d clay

1

1 0

1 9

1

N

N

N

N

N

N

N

1

0

N

N

1

1 0

1 9

1

2

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

Y

1

1 0

1 9

1

N

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1

1 0

1 9

1

1 9

1

N

Y

Y

N

N

Y

N

20.9

Y

0.18

0.6

6.1

5.5

33.4

Y

Prepare d clay

Y

Prepare d clay

Y

Prepare d clay

Y

Prepare d clay

Y

Prepare d clay

53.5

Y

Prepare d clay

N

N

1

1 0

17.8

Y

Prepare d clay

N

N

1

1 0

1 9

7 7

22.1

Y

Prepare d clay

N

Y

2

1 0

1 9

1

Y

Prepare d clay

1 9

1

Y

Prepare d clay

Y

Prepare d clay

0.18

0.18

0.6

7

0.6

0.18

0.6

0.19

5.2

0.6

Y

0.18

4.6

0.6

Posts

17

5.8

89.2

Y

Prepare d clay

Y

4.6

0.6

Y

9.4

105

Y

N

4.6

0.18

N

5.8

10.6

36.8

Prepare d clay

N

0.6

N

Posts

6.6

0.6

Posts

31

5.5

0.6

0.6

Y

6.1

0.6

0.18

Posts

9.4

0.6

Y

41

10.6

5.6

Prepare d clay

0.18

0.18

31

6.8

Y

Prepare d clay

0.18

Y

Y

0.18

0.6

Posts

Posts

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m) 0.18

N

Y

Posts

0.18

Y

Posts

Posts

0.18

Y

Y

Posts

Wall Posts (n)

Y

Posts

Posts

Posts (n)

Complete

Wall Type

Shape

Class

Mound Top Y

Dom estic Nondome stic Nondome stic

N

Pd ID

287

Posts

Comp ID

EU 7

Dom estic

Site ID

LM

Yarn ell II

N

Builds

6-8

Burned

EU 8

Basin

500

Dallas Site

Str. No.

Yarn ell II

Hearth Type

125

Dallas Site

Locus

LM

Int. Hearth

Dallas Site

Phase

Occ.

Site

Str. ID 501

Re cta ng ula r Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e Re cta ng ula r/O val Sq uar e Sq uar e Sq uar e Sq uar e Sq uar e

6.1

9.1

5.2

0.6

5.2

0.6

4.3

393

4.6

5.2

7

6.1

7.5

4.3

4.9

4.3

20.9

26.8

42.9

37.1

25.2

18.2

N

N

N

N

0

0

N

Y

N

N

N

N

N

1

1 0

N

N

1

1 0

1 9

1

1

1 0

1 9

1

Y

N

Dyar

LM

Duva ll

119 7

Dyar

LM

Duva ll

119 5

Dyar

LM

Dyar

Vil lag e

2

120 0

Dyar

LM

Dyar

119 9

Dyar

LM

Dyar

119 8

Dyar

LM

Dyar

Mo un d Mo un d Mo un d

Ea ste rn We ste rn No rth ern

451

Ela

HI

462

Ela

HI

Quall a Quall a

63.6

0

N

N/A

0

0

8

9.8

75.5

0

N

N/A

1

1.8

19

11.6

105.7

0

N

N/A

18

1

9

12

113.1

1

N

32

1.3

3

11

95

0

Cir cul ar

Posts

Y

15

13

2.2

2

Dom estic

Cir cul ar

Posts

Y

29

21

1

N

Dom estic

Cir cul ar

Posts

Y

38

19

I

N

Dom estic

Cir cul ar

Posts

Y

27

VII

N

Dom estic

Posts

Y

35

1

N

Dom estic

4

N

Dom estic

N

Dom estic

Y

Unkn own/ other

Y

Dom estic

Y

Dom estic

4

N

15

N

Dom estic Dom estic

Cir cul ar Sq uar e Sq uar e Re cta ng ula r Re cta ng ula r Sq uar e Sq uar e Re cta ng ula r Re cta

7

7.7

Pd ID

1

1 9

1 3 8

Y

Y

1

6

1 2

N

N

N

1

7 0

1 1 9

3

0

N

N

N

1

7 0

1 1 9

3

0

0

N

N

N

1

7 0

1 1 9

3

N/A

0

0

N

N

N

1

7 0

1 1 9

3

N

N/A

2

0

N

N

N

1

7 0

2

0

Y

174

N

N

1

7 1

157

56.3

1

N

N/A

1

0

N

36

Y

Y

1

7 1

7.2

6.6

47.5

Y

Undefi ned

0

2

N

160

N

N

2

7 1

1 2 2

Y

Prepare d clay

3

0

N

35

N

N

6

7 1

2

Y

Prepare d clay

1

0

N

45

N

Y

1 8

7 1

2

Y

Prepare d clay

1

0

N

45

N

N

6

7 1

1 2 2 1 2 2 1 2 2

N

N/A

5

4

N

160

N

N

1

U

N/A

N

N

1

3 7 3 7

6 8 6 8

Y

15 5

10 2

0.48

53

7.9

7.9

62.4

Posts

N

14 3

86

0.63

57

8.2

8.4

68.9

Posts

N

11 4

21

7

8

56

Posts

N

394

1

1 0

7.5

Posts

0.2

N

1

7.5

N

93

Y

Y

Posts

0.21

N

1 9

3

N

0.13

1

1 0

30.3

Posts

24

1 9

5.5

Y

0.5

1 0

5.5

Posts

0.15

1

Prepare d clay

Y

38

N

1 1 9 1 2 1 1 2 1

Posts

62

Comp ID

Site

119 6

9

44

Site ID

McFa rland

Y

Y

42

Builds

MW

0

6

86

Y

N

50

Y

Burned

Duncan Tract

2

0

Basin

117 3

Y

Prepare d clay

1

Orient. (deg E of N)

McFa rland

Y

Prepare d clay

Prepare d clay

Posts

Entry Trenches

MW

Prepare d clay

Int. Bur. (n)

Duncan Tract

37.1

Y

Int. Feat. (n)

116 9

III

5.9

30

Hearth Type

McFa rland

6.1

5.2

15.6

Int. Hearth

MW

0.6

3.7

Int. Walls (n)

Duncan Tract

N

5.8

Area (m2)

117 0

V

0.18

4.3

0.6

W (m)

McFa rland

Dom estic

33

0.6

L (m)

MW

N

Y

Dia. (m)

Duncan Tract

IV

0.14

Int. Post Dia. (m)

117 2

N

39 Ar eas R, S, an dT Ar eas R, S, an dT Ar eas R, S, an dT Ar eas R, S, an dT Ar eas R, S, an dT Vil lag e Vil lag e

Posts

Y

0.19

Int. Posts (n)

McFa rland

1

Dom estic

29

Wall Post Spac. (m)

MW

117 1

N

Posts

Y

Wall Post Dia. (m)

Duncan Tract

104

Dog River

Dom estic

Posts

Wall Posts (n)

LM

Lama r

133

N

Dom estic

Sq uar e Sq uar e Sq uar e Sq uar e

Posts (n)

427

Dom estic

Complete

LM

EU 7

Dallas Site

N

Wall Type

7-7

Yarn ell II

119

Shape

EU 7

Class

Str. No.

LM

Yarn ell II

Dallas Site

Mound Top

Locus

438

Dallas Site

Phase

EU 8

Occ.

Yarn ell II

Str. ID

LM

142

3 4 6 4 6

4 8

4 8 4 8 4 8

5 9 5 9

0

Y

7

38.5

Y

Undefi ned

3

0

0.19

7

38.5

N

N/A

0

25

0.2

6.6

34.2

N

N/A

1

20

0.21

8

50.3

N

0.29

1.6

12

0.24

8

50.3

28

0.19

1

11

0.23

15 0

29

0.21

0.9

111

0.21

Y

14 9

40

0.17

0.68

109

Posts

Y

61

19

0.23

1.5

Posts

Y

24

14

0.22

1.1

Sq uar e

WT

Y

5

5

0.21

Cir cul ar

Posts

Y

19

15

0.21

0.72

4

0.22

Posts

Y

29

24

0.21

0.83

4

0.18

Posts

Y

N

Dom estic

460

Ela

MW

Conn estee

13

N

Dom estic

457

Ela

MW

Conn estee

10

N

Dom estic

458

Ela

MW

Conn estee

11

N

Dom estic

456

Ela

MW

Conn estee

9

N

Dom estic

454

Ela

MW

Conn estee

7

N

Dom estic

455

Ela

MW

Conn estee

8

N

Dom estic

459

Ela

MW

Conn estee

12

N

Dom estic

448

Ela

MW

Conn estee

1

N

Dom estic

461

Ela

MW

Conn estee

14

N

Dom estic

453

Ela

MW

6

N

Dom estic

447

Fernvale

EM

2

N

Dom estic

446

Fernvale

MW

1

N

Dom estic

1

N

Unkn own/ other

1

N

Dom estic

340

Go Kart Site

LM

Sq uar e Irr eg ula

Y

2

3 7

6 8

5 9

N

N

1

3 7

6 8

5 9

N

N

2

3 7

6 8

5 9

N

N

N

1

3 7

6 7

5 4

0

N

N

N

1

3 7

6 7

5 4

2

0

N

N

N

1

3 7

6 7

5 4

N/A

2

0

N

N

N

1

3 7

6 7

5 4

N

N/A

0

0

N

N

1

3 7

6 7

5 4

60

N

N/A

0

0

N

N

N

1

3 7

6 7

5 4

8

50.3

N

N/A

1

0

N

N

N

1

3 7

6 7

5 4

0.16

8.5

56.7

N

N/A

2

0

N

N

N

1

3 7

6 7

5 4

42

0.2

8.5

56.7

N

N/A

4

0

N

N

N

1

3 7

6 7

5 4

10

0.25

7

38.5

N

N/A

0

0

N

N

N

1

3 7

6 7

5 4

14.4

N

N/A

0

0

N

N

Y

1

3 6

6 6

7 5

7.1

N

N/A

0

0

N

N

N

1

3 6

6 5

7 3

0

0

N

144

N

N

1

6 4

9 8

5 4

N

45

N

N

1

3 3

6 1

7 7

N

18 0

54

0.19

0.45

126

0.18

Posts

N

63

31

0.14

0.4

31

0.16

Posts

Y

14 7

88

0.18

0.55

59

0.17

Posts

N

14 6

20

0.15

0.79

126

0.13

Posts

Y

89

20

0.19

1.2

69

Posts

Y

64

39

0.18

Posts

Y

44

24

0.21

Posts

Y

25

13

Posts

Y

39

Posts

Y

Posts

395

L (m) 7.5

9

67.5

7

9

7.5

0

0.4

W (m)

N

Posts

Dia. (m)

Pd ID

3

Comp ID

Quall a

Site ID

HI

Builds

Ela

Burned

450

Basin

Dom estic

Orient. (deg E of N)

Int. Feat. (n) 1 1

N

Conn estee Bull Cree k

Entry Trenches

Hearth Type Surface fired area

5

Int. Bur. (n)

Int. Hearth Y

Quall a

MW

Int. Walls (n)

85.5

HI

Area (m2)

N

Ela

Garden Creek

Int. Post Dia. (m)

0

452

Int. Posts (n)

0

Dom estic

819

Wall Post Spac. (m)

N/A

N

Ol d Vil lag e Mi dd en

Wall Post Dia. (m)

N

2

Conn estee Gene ral Early Missi ssippi an Gene ral Midd le Woo dland

Wall Posts (n)

N

Quall a

Posts (n)

0

HI

Complete

6

Ela

Wall Type

Prepare d clay

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

Y

449

ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Cir cul ar Cir cul ar Cir cul ar Cir cul ar Cir cul ar Cir cul ar Cir cul ar Cir cul ar Cir cul ar Cir cul ar

3.8

9.5

8

3.8

3

2

6.6

6.4

39.5

N

N/A

7

7

49

N

N/A

160

160

30

Orient. (deg E of N)

Burned

Builds

Site ID

Comp ID

Pd ID

5.6

5.3

29.7

2

N

N/A

0

2

N

44

N

N

2

2 4

4 5

2 3

Posts

Y

55

0.18

0.4

6.6

6.3

41.6

1

N

N/A

2

0

N

59

N

N

1

2 4

4 5

2 3

Posts

Y

51

0.15

0.55

6.4

6.2

39.7

4

N

N/A

0

0

N

54

N

N

1

2 4

4 5

2 3

Posts

Y

26

0.2

1.3

8.9

62.2

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

25

0.2

1

7.1

39.6

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

28

0.2

0.8

6

28.3

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Cir cul ar

Posts

Y

16

0.2

0.8

4.8

18.1

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Cir cul ar

Posts

Y

27

0.3

1

8

50.2

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

17

0.2

0.8

5.2

21.2

0

N

N/A

0

0

N

N

N

1

2 4

4 4

7 3

Posts

Y

31

0.2

0.6

7.1

39.6

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

19

0.2

0.9

5.2

21.2

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

19

0.2

0.8

7.3

41.83

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

18

0.2

0.9

5.3

22.1

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

14

0.9

5.2

21.2

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

16

0.8

9.2

66.4

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Posts

Y

25

0.9

6.9

37.4

0

N

N/A

0

0

N

N

N

1

2 4

4 4

2 9

Cir cul ar

Posts

N

32.2

0

N

N/A

0

N

N

N

1

4

7

6

Sq uar e

WT

Y

44.9

1

Y

Prepare d clay

N

N

N

4

7

6

L (m)

W (m)

0.6

Dia. (m)

0.2

Posts (n)

Basin

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

32

Shape

Y

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

Posts

r

98

Hickory Log

MM

Wilb anks

100

Hickory Log

MM

Wilb anks

99

Hickory Log

MM

91

Hickory Log

MW

88

Hickory Log

MW

85

Hickory Log

MW

90

Hickory Log

MW

86

Hickory Log

MW

93

Hickory Log

MW

87

Hickory Log

MW

92

Hickory Log

MW

95

Hickory Log

MW

89

Hickory Log

MW

96

Hickory Log

MW

97

Hickory Log

MW

94

Hickory Log

MW

28

Hiwassee Island

EM

Wilb anks Carte rsvill e Carte rsvill e Carte rsvill e Carte rsvill e/Swi ft Cree k? Carte rsvill e/Swi ft Cree k? Midd le Woo dland Carte rsvill e Carte rsvill e Carte rsvill e Carte rsvill e/Swi ft Cree k? Carte rsvill e Carte rsvill e Carte rsvill e Hiwa ssee Islan d

126 6

Hiwassee Island

EM

Hiwa ssee Islan d

S. Lo cus S. Lo cus S. Lo cus S. Lo cus S. Lo cus N. Lo cus

2

N

Dom estic

4

N

Dom estic

3

N

Dom estic

10

N

Dom estic

7

N

Dom estic

1

N

Dom estic

Sq uar e Sq uar e Sq uar e Cir cul ar Cir cul ar Cir cul ar

S. Lo cus

9

N

Dom estic

N. Lo cus

5

N

Dom estic

12

N

Dom estic

6

N

Dom estic

11

N

Dom estic

14

N

Dom estic

8

N

Dom estic

15

N

Dom estic

16

N

Dom estic

13

N

Dom estic

Cir cul ar Cir cul ar Cir cul ar Cir cul ar

H 71

N

Dom estic

B 62

Y

Dom estic

S. Lo cus S. Lo cus S. Lo cus S. Lo cus

S. Lo cus S. Lo cus S. Lo cus S. Lo cus Un it 37 Un it 37 Md Le v G

Cir cul ar Cir cul ar Cir cul ar Cir cul ar

0.2

6

6.4

396

7

1

125 6

124 4

11

124 8

Hiwassee Island

Hiwassee Island

Hiwassee Island

Hiwassee Island

Hiwassee Island

EM

Hiwa ssee Islan d

EM

EM

EM

EM

Hiwa ssee Islan d Hiwa ssee Islan d

Hiwa ssee Islan d

Un it 38 Un it 37 Md Le v D Un it 37 Md Le vC Un it 63 Un it 37 Md Le vC

Dom estic

Re cta ng ula r

WT

Y

Dom estic

Sq uar e

Posts

Y

Dom estic

Re cta ng ula r

WT

Y

Dom estic

Sq uar e

Posts

Y

Dom estic

Re cta ng ula r

WT

Dom estic

Re cta ng ula r

B 15

H 8

B 21

Y

Y

N

Y

62

10.7

6

Fire basin

N

N

N

4

7

6

N

N/A

N

N

N

4

7

6

N

N/A

N

N

4

7

6

Y

Prepare d clay

N

N

4

7

6

Y

Prepare d clay

N

N

4

7

6

Y

Fire basin

N

N

4

7

6

N

N

4

7

6

N

N

4

7

6

4

7

6

4

7

6

4

7

6

89.4

7.6

7.6

57.76

9.1

12.8

117.1

7.9

7.9

62.41

0

Y

Fire basin

1

Y

9.4

13.7

129.6

1

Y

Fire basin

1

WT

Y

9.1

11

100.3

Y

Fire basin

Dom estic

Sq uar e

WT

Y

6.1

6.7

40.87

Y

Prepare d clay

Nondome stic

Re cta ng ula r

WT

Y

10.1

17.4

174.7

N

N/A

397

0

0

0

N

40

N

0

0

N

0

N

1

3

N

147

N

0

N

N

15

N

N

N

N

Builds

7

11.3

30

Orient. (deg E of N)

4

7.9

30

Entry Trenches

N

52.48

0

Int. Bur. (n)

N

8.2

77

0

N

6.4

77

62

Int. Walls (n)

Area (m2)

W (m)

Y

Dom estic

0.15

L (m)

Posts

N

B 35

Dia. (m)

Y

H 73

N

Int. Post Dia. (m)

Posts

Cir cul ar Re cta ng ula r/O val

H 36

Int. Posts (n)

11.3

Nondome stic

Y

Wall Post Spac. (m)

8.2

Y

B 51

Wall Post Dia. (m)

Y

B 49

N

Wall Posts (n)

WT

Dom estic

H 64

Posts (n)

Complete

Wall Type

Mound Top

Str. No.

Locus

Phase

EM

Hiwa ssee Islan d Hiwa ssee Islan d

Re cta ng ula r

1

Pd ID

3

Hiwassee Island

EM

Un it 38 Un it 37 Md Le vF

89.9

123.8

Y

N

Y

13.1

B 20

B 70

92.8

9.4

Comp ID

126 3

Hiwassee Island

EM

4

Y

Site ID

9

Hiwassee Island

Hiwa ssee Islan d Hiwa ssee Islan d

Un it 37 Vil lag e un der mo un d

Prepare d clay

WT

Burned

126 8

EM

Hiwa ssee Islan d

Y

Dom estic

Re cta ng ula r

Basin

12

Hiwassee Island

EM

Hiwa ssee Islan d

Int. Feat. (n)

Hiwassee Island

Hearth Type

127 0

EM

Hiwa ssee Islan d

Y

Int. Hearth

Hiwassee Island

B 58

Shape

124 7

EM

Class

Hiwassee Island

Occ.

Site

Str. ID 126 5

Hiwa ssee Islan d

Un it 37 Md Le v G Un it 37 Md Le vC Un it 37 Md Le vE 2

1

1

2

1

125 9

125 1

124 2

4

Hiwassee Island

Hiwassee Island

Hiwassee Island

Hiwassee Island

124 9

Hiwassee Island

125 8

Hiwassee Island

7

126 7 126 9

Hiwassee Island

Hiwassee Island Hiwassee Island

EM

EM

EM

EM

EM

EM EM

Hiwa ssee Islan d Hiwa ssee Islan d

Hiwa ssee Islan d

Hiwa ssee Islan d Hiwa ssee Islan d

Hiwa ssee Islan d Hiwa ssee

Un it 38 Un it 37 Md Le vC Un it 37 Md Le vE 1 Un it 38 Vil lag e un der mo un d Un it

11.3

120.3

N

N/A

B 46

Y

Dom estic

Sq uar e

WT

Y

9.1

10.4

94.8

N

N/A

Dom estic

Re cta ng ula r

WT

Y

8.5

11.3

96.2

Y

Prepare d clay

Dom estic

Re cta ng ula r

WT

Y

11

7.9

87

Y

Prepare d clay

Dom estic

Re cta ng ula r

WT

Y

7

8.8

61.9

Y

Fire basin

Dom estic

Sq uar e

Posts

Y

6.7

7.6

50.92

Y

Fire basin

WT

Y

9.1

12.2

111.5

N

N/A

B 43

B 25

B 10

H 37

Y

Y

Y

N

Fire basin

80

80

0

N

5

0

1

N

N

Builds

10.7

Y

Orient. (deg E of N)

Y

139.7

Entry Trenches

Posts

14.3

Int. Bur. (n)

Sq uar e

L (m)

Dom estic

9.7

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Mound Top

Str. No.

Y

Y

Pd ID

EM

Hiwa ssee Islan d

B 16

WT

Comp ID

EM

Hiwa ssee Islan d

Locus

Phase

EM

Hiwa ssee Islan d

Dom estic

Site ID

Hiwassee Island

Y

Burned

126 1

EM

Hiwa ssee Islan d

B 29

Re cta ng ula r

Basin

Hiwassee Island

Shape

124 5

EM

Class

Hiwassee Island

Occ.

Site

Str. ID 125 4

Hiwa ssee Islan d

Un it 37 Md Le v D Un it 37 Md Le vC Un it 37 Md Le vE 2 Un it 37 Md Le vE 1 Un it 37 Md Le v D Un it 37 Md Le vB

4

7

6

4

7

6

N

N

N

4

7

6

N

N

N

4

7

6

N

N

N

4

7

6

4

7

6

4

7

6

N

175

N

N

1

N

N

N

4

7

6

N

N

N

4

7

6

N

N

4

7

6

B 22

Y

Dom estic

Re cta ng ula r

B 41

Y

Dom estic

Sq uar e

WT

Y

7.9

8.2

65.2

1

Y

Prepare d clay

3

Dom estic

Sq uar e

WT

Y

4.9

5.5

26.95

0

Y

U

1

Nondome stic Dom estic

Re cta ng ula r Re cta

WT

Y

10.4

15.8

164.2

Y

Prepare d clay

N

N

N

4

7

6

WT

Y

7.3

11.3

82.5

N

N/A

N

N

N

4

7

6

H 60

N

B 69

N

B9

N

398

2

3

N

130

2

Hiwassee Island

126 4

Hiwassee Island

1

Hiwassee Island

127 3

Hiwassee Island

126 2

Hiwassee Island

125 5

Hiwassee Island

29

Hiwassee Island

6

Hiwassee Island

14

Hiwassee Island

126 0

Hiwassee Island

EM

Hiwa ssee Islan d

EM

Hiwa ssee Islan d Hiwa ssee Islan d

EM

Hiwa ssee Islan d

EM

EM

Hiwa ssee Islan d

EM

Hiwa ssee Islan d Hiwa ssee Islan d Hiwa ssee Islan d

EM

Hiwa ssee Islan d

EM

Hiwa ssee Islan d

EM

EM

Un it 37 Md Le v D Un it 37 Md Le vC Un it 37 Md Le vC Un it 37 Md Le v G Un it 38 Un it 37 Md Le vC Un it 37 Md Le vF Un it 37 Md Le v D Un it 38 Un it 38

Un it 38 Un it 37 Md Le

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

8.5

72.8

Y

Fire basin

N

N

N

4

7

6

WT

Y

7.9

9.1

72.8

Y

Fire basin

N

N

N

4

7

6

89.9

N

N/A

N

N

N

4

7

6

N

N

N

4

7

6

N

N

4

7

6

L (m)

Builds

8.5

W (m)

Y

Dia. (m)

WT

Posts (n)

Sq uar e

63

Shape

Dom estic

Class

Mound Top

Str. No.

Pd ID

127 2

EM

Hiwa ssee Islan d

Comp ID

Hiwassee Island

Site ID

125 0

EM

Hiwa ssee Islan d

Burned

Hiwassee Island

Basin

125 3

Locus

Phase

Occ.

Site

Str. ID

Islan d

ng ula r

B 28

Y

B 24

Y

Dom estic

Re cta ng ula r

B 13

Y

Nondome stic

Cir cul ar

Both

Y

Y

Nondome stic

Re cta ng ula r

WT

Y

11.6

15.2

176.5

1

N

N/A

WT

Y

5.8

6.1

35.38

0

Y

Fire basin

167.4

N

N/A

N

N

N

4

7

6

B 57

2

10.7

H 25

N

Dom estic

Sq uar e

B 14

Y

Nondome stic

Cir cul ar

Both

Y

WT

Y

9.1

14.3

131

N

N/A

N

N

N

4

7

6

7.3

7.3

53.5

Y

Fire basin

N

N

N

4

7

6

N

N

1

4

7

6

B 50

Y

Dom estic

Re cta ng ula r

B 30

Y

Dom estic

Sq uar e

WT

Y

Dom estic

Cir cul ar

WT

Y

WT

Y

Both

Y

WT

Y

H 62

N

H 58

N

Dom estic

H 26

N

Dom estic

Sq uar e Re cta ng ula r/O val

B 44

Y

Dom estic

Sq uar e

2

14.6

7

55

51

4

399

1

4

N

40.7

0

N

N/A

0

0

N

145

1

6.4

6.4

40.96

0

N

N/A

0

1

N

131

N

N

2

4

7

6

4.9

7.3

35.77

0

Y

Fire basin

2

0

N

151

N

N

1

4

7

6

9.4

9.8

92.2

Y

Prepare d clay

1

N

N

4

7

6

N

N

N

N

N

N

Pd ID

N

Comp ID

N

Site ID

Burned

N

4

7

6

4

7

6

4

7

6

2

4

7

6

1

4

7

6

4

7

6

4

7

6

4

7

6

4

7

6

Builds

Basin

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

WT

Int. Post Dia. (m)

Y

Int. Posts (n)

Posts

Wall Post Spac. (m)

Y

Wall Post Dia. (m)

WT

Wall Posts (n)

Cir cul ar

Posts (n)

Complete

Shape

Wall Type

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

Nondome stic

vE 2

127 1

Hiwassee Island

2

Hiwassee Island

125 2

Hiwassee Island

8

Hiwassee Island

124 1

Hiwassee Island

124 6

Hiwassee Island

5

124 3

Hiwassee Island

Hiwassee Island

125 7

Hiwassee Island

30

Hiwassee Island

495

Hiwassee Old Town

496

Hiwassee Old Town

493

Hiwassee Old Town

EM

EM

Hiwa ssee Islan d Hiwa ssee Islan d

EM

Hiwa ssee Islan d Hiwa ssee Islan d

EM

Hiwa ssee Islan d

EM

EM

Hiwa ssee Islan d Hiwa ssee Islan d

Un it 37 Md Le vC Un it 38 Un it 37 Md Le v D Un it 38

Un it 37 Un it 37 Md Le vC

EM

Hiwa ssee Islan d

EM

Hiwa ssee Islan d

LM

Dalla s

Un it 38 Un it 37 Md Le vB Un it 37 Md Le v D Un it 38

HI

Cher okee

EM

B 17

Y

H 34

N

Dom estic

Sq uar e

B 27

Y

Dom estic

Sq uar e

1

64

13.7

64

147.3

0

N

N/A

Y

Fire basin

Y

Prepare d clay

N

6.7

7.6

50.92

1

3

N

Y

8.2

8.2

67.7

WT

Y

5.2

5.8

30.16

0

Y

U

1

3

N

WT

Y

7.3

5.2

37.9

0

Y

Prepare d clay

1

0

N

N

N

Y

Fire basin

N

N

N

N

N/A

N

N

N

H 61

N

Dom estic

H 74

N

Dom estic

Sq uar e Re cta ng ula r

B 19

Y

Dom estic

Sq uar e

WT

Y

6.1

6.7

40.9

Dom estic

Sq uar e

WT

Y

4.9

5.5

26.95

Dom estic

Re cta ng ula r

WT

Y

7.9

10.7

84.5

Y

Fire basin

WT

Y

6.1

7.9

48.3

y

Prepare d clay

Y

Prepare d clay

0

0

N

H 57

B 11

B 36

N

Y

Y

Dom estic

H 11

N

Dom estic

Bl oc k A

5

N

Dom estic

HI

Cher okee

Bl oc k A

6

N

Dom estic

HI

Cher okee

Bl oc kC

3

N

Stora ge

Re cta ng ula r Sq uar e Re cta ng ula r Re cta ng ula r Re cta ng ula r

Posts

Y

23

23

0.18

0.6

Posts

Y

39

27

1.1

Posts

Y

29

23

Posts

Y

6

6

0

0

1

N

130

131

N

2

2

N

N

124

N

N

1

4

8

1 2

4.6

4.6

21.16

12

9.6

3.6

34.6

N

N/A

0

0

N

115

N

N

1

1 9

3 2

1 1

1.1

6

9.4

3.4

32

N

N/A

0

0

N

23

N

N

1

1 9

3 2

1 1

1.65

0

3

3

9

N

N/A

0

0

N

150

N

N

1

1 9

3 2

1 1

400

0

147

0

Cher okee

Bl oc kB

1

N

Dom estic

492

Hiwassee Old Town

HI

Cher okee

Bl oc kB

2

N

Unkn own/ other

120 5

Hoecake

LW

Hoec ake

Ar ea I

5

N

Dom estic

121 0

Hoecake

LW

Hoec ake

Ar ea IV

10

N

Dom estic

120 3

Hoecake

LW

Hoec ake

Ar ea I

3

N

Dom estic

120 8

Hoecake

LW

Hoec ake

Ar ea IV

8

N

Dom estic

120 2

Hoecake

LW

Hoec ake

Ar ea I

2

N

Dom estic

120 7

Hoecake

LW

Hoec ake

Ar ea III

7

N

Dom estic

120 6

Hoecake

LW

Hoec ake

Ar ea I

6

N

Dom estic

121 1

Hoecake

LW

Hoec ake

Ar ea IV

11

N

Dom estic

120 1

Hoecake

LW

Hoec ake

Ar ea I

1

N

Dom estic

120 9

Hoecake

LW

Hoec ake

Ar ea IV

9

N

Dom estic

120 4

Hoecake

LW

Hoec ake

Ar ea I

4

N

Dom estic

476

Jenrette

HI

Jenre

3

N

Dom

Ov

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

29

22

1.5

7

9.8

4.25

41.65

1

N

N/A

1

0

N

25

N

N

1

1 9

3 2

1 1

Posts

Y

35

20

1.4

15

7.4

3

22.2

2

N

N/A

1

1

N

155

N

N

1

1 9

3 2

1 1

Posts

Y

25

22

1

3

5.3

3

15.9

0

N

N/A

N

N

1

1 9

3 2

1 1

Posts

N

3.8

2

7.5

N

N/A

0

0

N

0

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

4

2

7.9

N

N/A

0

0

N

123

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

3.7

2.1

7.8

N

N/A

1

0

N

42

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

Posts

20

0

L (m)

W (m)

Y

Dia. (m)

Posts

Posts (n)

Pd ID

HI

Dom estic

Comp ID

Hiwassee Old Town

N

Site ID

491

4

Builds

Cher okee

Burned

HI

Basin

Hiwassee Old Town

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID 494

Bl oc kC

N

20

0.05

0.35

48

0.05

0.25

35

32

0.08

0.33

3

3.2

2.4

7.8

N

N/A

0

0

N

110

Y

N

1

7 2

1 2 3

1 1 6

Y

36

36

0.08

0.38

0

4.3

2.4

10.4

N

N/A

0

0

N

35

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

36

34

0.07

0.3

2

3.5

2.4

8.5

N

N/A

2

0

N

120

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

44

40

0.08

0.35

4

3.7

2.4

8.9

N

N/A

1

0

N

125

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

48

40

0.08

0.3

8

4.4

2.9

12.8

N

N/A

0

0

N

30

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

42

32

0.06

0.39

10

4.3

2.4

10.4

N

N/A

0

0

N

30

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

32

31

0.06

0.4

1

4.3

2.7

11.7

N

N/A

0

0

N

110

Y

N

1

7 2

1 2 3

1 1 6

Posts

Y

29

0.06

0.32

3.7

2.3

8.4

N

N/A

1

0

N

42

Y

N

1

7 2

1 2 3

1 1 6

WT

Y

5.8

4.9

28.2

N

N/A

2

0

N

33

N

N

1

4

7

4

401

475

Jenrette

HI

474

Jenrette

HI

677

676

Jewell

Jewell

EM

EM

673

Jewell

EM

662

Jewell

EM

668

675

661

670

664

665

Jewell

Jewell

Jewell

Jewell

Jewell

Jewell

EM

EM

EM

EM

EM

EM

Vil lag e

Vil lag e

2

N

Dom estic

1

N

Dom estic

Re cta ng ula r Re cta ng ula r

21

20

N

N

Comp ID

Pd ID

Builds

Burned

Basin

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Wall Type

Shape al

Site ID

estic

Jenre tte

Jenre tte Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

tte

2

3

1

4

3

12

N

N/A

1

0

N

175

N

N

1

4 2

7 3

4 1

Y

5.4

4.2

20.4

N

N/A

3

0

N

33

N

N

1

4 2

7 3

4 1

WT

Y

9.9

9

88.6

U

N/A

N

N

N

1

5 9

9 3

7 5

Dom estic

Re cta ng ula r

WT

Y

8.5

7.9

67.2

U

N/A

N

N

N

2

5 9

9 3

7 5

WT

N

6.3

U

N/A

N

N

N

1

5 9

9 3

7 5

U

N/A

N

N

N

1

5 9

9 3

7 5

Posts

Y

Both

Dom estic

Re cta ng ula r

67

51

0.15

0.3

16

0.15

Vil lag e

17

N

Dom estic

Re cta ng ula r

Vil lag e

2

N

Dom estic

Cir cul ar

WT

N

Dom estic

Re cta ng ula r

Posts

N

Dom estic

Re cta ng ula r

WT

Y

6.6

6.3

44

U

N/A

N

N

N

1

5 9

9 3

7 5

Dom estic

Re cta ng ula r

WT

Y

5.6

4.8

26.9

N

N/A

N

N

Y

1

5 9

9 3

7 5

Dom estic

Re cta ng ula r

WT

Y

9.3

8.8

80.9

N

N/A

N

N

N

1

5 9

9 3

7 5

Dom estic

Re cta ng ula r

WT

N

Dom estic

Re cta ng ula r

WT

N

9 3

7 5

Mo un d

Vil lag e

Vil lag e

Vil lag e

Mo un d

Mo un d

12

19

1

14

5

8

Y

N

N

N

Y

Y

12

113

5 9

5 9

U

402

N/A

N

N

N

1

5 9

667

682

680

666

674

672

Jewell

Jewell

Jewell

Jewell

Jewell

Jewell

EM

EM

EM

EM

EM

EM

669

Jewell

EM

681

Jewell

EM

Comp ID

Pd ID

N

1

5 9

9 3

7 5

Dom estic

Re cta ng ula r

WT

Y

8.5

7.6

64.2

N

N/A

N

N

N

2

5 9

9 3

7 5

Dom estic

Re cta ng ula r

WT

N

9.6

U

N/A

N

N

N

1

5 9

9 3

7 5

Posts

N

L (m)

Y

Dom estic

Cir cul ar

WT

N

Dom estic

Re cta ng ula r

WT

N

Dom estic

Re cta ng ula r

WT

N

5.8

Dom estic

Re cta ng ula r

WT

Y

14.5

12

Dom estic

Re cta ng ula r

WT

Y

6.5

6.4

Dom estic

Re cta ng ula r

WT

N

5.5

WT

Y

10.5

WT

N

Vil lag e

Vil lag e

Vil lag e Pre Mo un d

9

18

16

Y

N

N

Orient. (deg E of N)

N

10

Mo un dA

Entry Trenches N

Mo un dB'

N

Int. Bur. (n)

N/A

Dom estic

24

Int. Feat. (n)

U

Y

Vil lag e

Hearth Type

107.8

4

Y

Int. Hearth

9.8

Re cta ng ula r

28

Int. Walls (n)

11

W (m)

N

Dia. (m)

WT

Posts (n)

Dom estic

Re cta ng ula r

Mo un d

Mo un dB'

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Site ID

N

Wall Type

Mound Top

Str. No.

Locus

Phase

Occ. EM

23

N

Builds

Jewell

EM

Vil lag e

15

Y

Burned

Jewell

EM

Vil lag e

22

Basin

663

Jewell

EM

Shape

679

Jewell

Pre Mo un d

Class

671

Site

Str. ID 678

Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi an Gene ral Early Missi ssippi

13

N

Dom estic

25

Y

Dom estic

Re cta ng ula r Re cta ng ula r

5 9

9.6

403

U

N/A

N

N

N

1

5 9

9 3

7 5

U

N/A

N

N

N

1

5 9

9 3

7 5

U

N/A

N

N

N

1

5 9

9 3

7 5

174

U

N/A

N

N

N

1

5 9

9 3

7 5

41.6

U

N/A

N

N

N

1

5 9

9 3

7 5

U

N/A

N

N

N

1

5 9

9 3

7 5

N

N/A

N

N

y

1

5 9

9 3

7 5

U

N/A

N

N

N

1

5 9

9 3

7 5

71.9

10.1

106.5

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

Orient. (deg E of N)

Entry Trenches

Int. Feat. (n)

Prepare d clay

1

0

N

155

N

Y

2

3 9

7 0

6

Posts

Y

36

36

0.11

0.39

3.9

3.8

14.8

0

Y

Prepare d clay

1

0

N

155

Y

N

2

3 9

7 0

6

Posts

N

Y

Prepare d clay

1

0

N

Y

N

1

3 9

7 0

6

L (m)

Int. Bur. (n)

Hearth Type

Y

Area (m2)

1

W (m)

27.8

Dia. (m)

4.8

Int. Posts (n)

Int. Hearth

Int. Walls (n)

Int. Post Dia. (m)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

5.8

Posts (n)

0.56

Posts

Complete

0.1

Wall Type

29

Shape

29

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

Y

Dom estic

Re cta ng ula r

an

464

Kimberly -Clark

EM

465

Kimberly -Clark

EM

Hiwa ssee Islan d Hiwa ssee Islan d Hiwa ssee Islan d

121 9

Kincaid

EM

Early Kinc aid

Mx 7

Mx v1 D

463

Kimberly -Clark

EM

1

121 8

Kincaid

EM

Early Kinc aid

121 6

Kincaid

EM

Early Kinc aid

Mx 4

121 5

Kincaid

EM

Early Kinc aid

Mx 7

EM

Early Kinc aid

Mx 4

EM

Early Kinc aid

Mx 4

Mx 7

Mx 7

121 7

121 4

Kincaid

Kincaid

122 0

Kincaid

EM

Early Kinc aid

122 8

Kincaid

LM

Late Kinc aid

N

2

N

Dom estic

3 Mx 7, str uct ure 10 2 Mx v1 D, F.1 , E. Se cti on Mx 4, F2 2 Mx 7, str uct ure 10 1 Mx 4, W T3/ 4/5 , Ho use 2 Mx 4, F1 3, Ho use 1 Mx 7, Su bz on e 3D Mx 7, L7, Z2 a, Ho use

N

Dom estic

Sq uar e Un ide ntif ied

N

Dom estic

Sq uar e

WT

Y

64.7

N

N/A

N

N

N

1

7 3

1 2 4

1 1 7

N

Dom estic

Re cta ngl e

WT

Y

15.7

Y

Prepare d clay

N

N

N

2

7 3

1 2 4

1 1 7

N

Dom estic

Sq uar e

WT

Y

12.3

Y

Prepare d clay

N

N

N

1

7 3

1 2 4

1 1 7

N

Dom estic

Sq uar e

Both

Y

10.2

N

N

N

N

1

7 3

1 2 4

1 1 7

N

Dom estic

Re cta ngl e

WT

Y

12.9

N

N

N

N

2

7 3

1 2 4

1 1 7

N

Dom estic

Re cta ngl e

WT

Y

9.6

N

N

N

N

2

7 3

1 2 4

1 1 7

Y

Dom estic

Re cta ngl e

Posts

Y

117.5

N

N/A

N

N

N

1

7 3

1 2 4

1 1 7

Y

Dom estic

Sq uar e

WT

Y

126.3

N

N/A

N

N

N

1

7 3

1 2 6

1 1 9

404

N/A

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

Orient. (deg E of N)

Entry Trenches

N

1

7 3

1 2 6

1 1 9

N

Dom estic

Re cta ngl e

WT

Y

41.8

N

N/A

N

N

N

1

7 3

1 2 5

1 1 8

N

Dom estic

Sq uar e

Posts

Y

18.6

N

N/A

N

N

N

1

7 3

1 2 5

1 1 8

N

Dom estic

Re cta ngl e

WT

Y

33.2

Y

Prepare d clay

N

N

N

1

7 3

1 2 5

1 1 8

N

Dom estic

Sq uar e

WT

Y

11.5

Y

Prepare d clay

N

N

N

2

7 3

1 2 5

1 1 8

Y

Nondome stic

Re cta ngl e

WT

Y

195.1

N

N/A

N

N

N

1

7 3

1 2 5

1 1 8

N

Dom estic

Sq uar e

WT

Y

39

Y

Prepare d clay

N

N

N

1

7 3

1 2 5

1 1 8

N

Dom estic

Sq uar e

WT

Y

14.1

Y

Prepare d clay

N

N

N

1

7 3

1 2 7

1 2 0

N

Dom estic

Sq uar e

WT

Y

33.5

Y

Prepare d clay

N

N

N

1

7 3

1 2 7

1 2 0

N

Dom estic

Sq uar e

WT

Y

20.9

N

N/A

N

N

N

1

7 3

1 2 7

1 2 0

W (m)

Int. Bur. (n)

N

Int. Feat. (n)

Hearth Type

N

Area (m2)

Prepare d clay

L (m)

Y

Dia. (m)

35.1

Int. Posts (n)

Int. Hearth

Int. Walls (n)

Int. Post Dia. (m)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Y

Posts (n)

WT

Complete

Shape Sq uar e

Wall Type

Class Dom estic

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID

N

1

122 7

Kincaid

LM

Late Kinc aid

Mx 1C

122 5

Kincaid

MM

122 2

Kincaid

MM

Midd le Kinc aid Midd le Kinc aid

MM

Midd le Kinc aid

Mx 4

MM

Midd le Kinc aid

Mx 4

MM

Midd le Kinc aid

Mx 4

122 3

122 1

122 6

122 4

123 1

Kincaid

Kincaid

Kincaid

Kincaid

Kincaid

MM

EM

123 6

Kincaid

EM

123 2

Kincaid

EM

Midd le Kinc aid Kinc aid Unkn own Missi ssippi an

Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own

Mx 4

Mx 4

Mx 4

We st Mo un d

Mx 1A

Mx 1A

Mx 1C, F. 20 Mx 4, W T5 358, Ho use 2 Mx 4, F.3 1/8 Mx 4, F.7 , Ho use 1 Mx 4, F.1 2, Ho use 2 Mx 4, W T6 8/9 3 Mx 4, W T5 1/5 9/6 0 We st Mo un d, F.2 Mx 1A , Str uct ure 1, 19 41 Mx 1A , F.2

405

Kincaid

EM

122 9

Kincaid

EM

123 9

Kincaid

EM

123 0

123 8

Kincaid

Kincaid

EM

EM

Orient. (deg E of N)

Entry Trenches

Int. Walls (n)

Int. Post Dia. (m)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

N

N

N

1

7 3

1 2 7

1 2 0

Y

Dom estic

Re cta ngl e

WT

Y

72

Y

Prepare d clay

N

N

N

2

7 3

1 2 7

1 2 0

N

Dom estic

Re cta ngl e

WT

Y

27.9

N

N/A

N

N

N

1

7 3

1 2 7

1 2 0

N

Dom estic

Sq uar e

WT

Y

33.5

Y

Prepare d clay

N

N

N

1

7 3

1 2 7

1 2 0

Y

Nondome stic

Re cta ngl e

WT

Y

299.8

N

N/A

N

N

N

1

7 3

1 2 7

1 2 0

N

Dom estic

Re cta ngl e

WT

Y

9.8

N

N

N

N

1

7 3

1 2 7

1 2 0

Y

Nondome stic

Sq uar e

WT

Y

183.8

Y

Prepare d clay

N

N

N

1

7 3

1 2 7

1 2 0

N

Dom estic

Re cta ngl e

WT

Y

13.9

N

N/A

N

N

N

2

7 3

1 2 7

1 2 0

Y

Dom estic

Re cta ngl e

WT

Y

113

Y

Prepare d clay

N

N

N

2

7 3

1 2 7

1 2 0

W (m)

Int. Bur. (n)

Prepare d clay

Int. Feat. (n)

Y

Area (m2) 30

L (m)

Y

Dia. (m)

Posts

Int. Posts (n)

Sq uar e

Posts (n)

Dom estic

Complete

N

Mound Top

Pd ID

124 0

Comp ID

EM

Site ID

Kincaid

Builds

123 5

Burned

EM

Basin

Kincaid

Hearth Type

123 3

Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an

Int. Hearth

EM

Wall Type

Kincaid

EM

Shape

123 7

Kincaid

Class

123 4

Kinc aid Unkn own Missi ssippi an Kinc aid Unkn own Missi ssippi an

Str. No.

Locus

Phase

Occ.

Site

Str. ID

Missi ssippi an

9 inn er (19 35)

Mx 1C

Mx 9

Mx 1A

Mx 10

Mx 4

Pp 13

Mx 1D

Mx 4

Mx 1A

Mx 1C, F.1 9b

Mx 9, F.1 9 Mx 1A , F.2 9, out er (19 35)

Mx 10, F.8 Mx 4, W T7 1/7 2/7 9

Pp 13, F.7 Mx 1D , W T1/ 3/1 6

Mx 4, F.2 3 Mx 1A , F.1 /12 (19 35)

406

613

King

LM

Barn ett

RS 1.3

N

Stora ge

Sq uar e Re cta ng ula r

603

King

LM

Barn ett

25. 2

N

Dom estic

Sq uar e

594

King

LM

Barn ett

20

N

Dom estic

623

King

LM

Barn ett

RS 9.2

N

Stora ge

571

King

LM

Barn ett

2.2

N

Dom estic

627

King

LM

Barn ett

RS 11. 2

N

Stora ge

619

King

LM

Barn ett

RS 6

N

Stora ge

607

King

LM

Barn ett

28

N

Dom estic

616

King

LM

Barn ett

RS 3

N

Stora ge

Sq uar e Re cta ng ula r

605

King

LM

Barn ett

26. 2

N

Dom estic

Sq uar e

587

King

LM

Barn ett

13

N

Dom estic

629

King

LM

Barn ett

RS 13

N

Stora ge

Sq uar e Re cta ng ula r

575

King

LM

Barn ett

5.1

N

Dom estic

581

King

LM

Barn ett

7

N

Dom estic

Y

Posts

Y

Posts

Y

Posts

Y

Posts

Y

Posts

Y

Posts

Y

8

8

Posts

Y

8

Posts

Y

Posts

Y

Posts

25

5.9

33.9

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

N

N/A

0

0

N

22

N

N

1

1

1

1 8

N

N/A

0

0

N

77

N

N

1

1

1

1 8

N

N/A

0

1

N

2

N

N

1

1

1

1 8

0.9

4.1

2.1

8.7

26

0.9

8.2

8.3

68

5.6

6

33.8

0

Y

Prepare d clay

2

1

N

16

N

N

1

1

1

1 8

3.9

2.4

9.5

0

N

N/A

0

1

N

25

N

N

1

1

1

1 8

7.6

7.7

58.7

Y

Prepare d clay

1

2

N

86

N

N

1

1

1

1 8

0.7

3.7

1.6

6

N

N/A

0

0

N

85

N

N

1

1

1

1 8

8

0.9

3.7

2.2

8.4

N

N/A

0

3

N

67

N

N

1

1

1

1 8

26

0.8

6.4

6.4

41

N

N/A

0

0

N

58

N

N

1

1

1

1 8

11

1.1

4.5

3.8

10.2

N

N/A

0

0

N

72

N

N

1

1

1

1 8

Y

28

0.9

8.2

8

65.8

N

N/A

0

5

N

66

N

N

1

1

1

1 8

Posts

Y

14

5.4

5.5

29.9

Y

Prepare d clay

0

N

65

N

N

1

1

1

1 8

Posts

Y

7

3.9

1.9

7.4

N

N/A

0

0

N

76

N

N

1

1

1

1 8

Sq uar e

Posts

Y

38

0.76

7.3

7.3

53.3

Y

Prepare d clay

1

2

N

77

Y

N

3

1

1

1 8

Sq uar e

Posts

Y

36

0.82

6.3

6.5

41.1

Y

Prepare d clay

1

1

Y

81

Y

Y

1

1

1

1 8

Sq uar e Re cta ng ula r Re cta ng ula r

7

11

7

7

1.1

407

0

Int. Hearth

Int. Walls (n)

Area (m2)

W (m)

L (m)

Dia. (m) 5.7

10

Sq uar e Re cta ng ula r

10

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Posts

Pd ID

Dom estic

Comp ID

N

Site ID

18

Builds

Barn ett

Burned

LM

Basin

King

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID 592

0

0

0

6

572

King

591

Pd ID

LM

Comp ID

King

Site ID

568

Builds

Barn ett

Burned

LM

Basin

King

2

10

N

86

Y

N

1

1

1

1 8

4.1

2.4

9.9

0

N

N/A

0

1

N

0

N

N

2

1

1

1 8

7.4

7.5

55.3

Y

Prepare d clay

1

4

N

2

Y

N

1

1

1

1 8

0.9

7.4

7.5

55.6

Y

Prepare d clay

1

0

N

67

Y

Y

1

1

1

1 8

44

0.94

7.4

7.4

54.8

Y

Prepare d clay

1

3

N

3

N

N

1

1

1

1 8

Y

15

1

7.7

7.8

60.4

N

N/A

1

N

65

N

N

2

1

1

1 8

Posts

Y

46

0.82

7.3

7.3

54

Y

Prepare d clay

1

1

N

85

N

N

1

1

1

1 8

Posts

Y

7.6

7.9

60.4

Y

Prepare d clay

1

1

N

85

N

N

1

1

1

1 8

Posts

Y

8.5

8.5

72.3

N

N/A

0

1

N

5

N

N

1

1

1

1 8

Posts

Y

11

11

5.6

2.7

15.1

0

N

N/A

0

4

N

39

N

N

1

1

1

1 8

Posts

Y

12

12

0.9

5.3

2.1

11.4

0

N

N/A

0

2

N

86

N

N

1

1

1

1 8

Posts

N

1

7.4

N

N/A

0

N

78

N

N

1

1

1

1 8

Posts

Y

N

N/A

0

2

N

82

N

N

2

1

1

1 8

Posts

Y

Y

Prepare d clay

1

4

N

88

Y

Y

1

1

1

1 8

Posts

Y

Posts

Y

Posts

Y

Sq uar e

Posts

Y

Dom estic

Sq uar e

Posts

Y

40

N

Dom estic

Sq uar e

Posts

Y

25. 1

N

Dom estic

Sq uar e

Posts

Barn ett

1.2

N

Dom estic

Sq uar e

LM

Barn ett

2.3

N

Dom estic

Sq uar e

King

LM

Barn ett

15. 3

N

Dom estic

621

King

LM

Barn ett

RS 8

N

Stora ge

612

King

LM

Barn ett

RS 1.2

N

Stora ge

586

King

LM

Barn ett

12

N

Dom estic

626

King

LM

Barn ett

RS 11. 1

N

Stora ge

599

King

LM

Barn ett

23. 3

N

Dom estic

617

King

LM

Barn ett

RS 4

N

Stora ge

Sq uar e Re cta ng ula r

121 2

King

LM

Barn ett

16

N

Nondome stic

Sq uar e

Pla za

Sq uar e Re cta ng ula r Re cta ng ula r Sq uar e Re cta ng ula r

Orient. (deg E of N)

602

Entry Trenches

1.3

Prepare d clay

10

10

32

7

11

1

7

3.9

1.6

6.3

0

Int. Bur. (n)

Barn ett

Int. Feat. (n)

LM

Hearth Type

King

Int. Hearth

569

Int. Walls (n)

N

Y

1.1

Area (m2)

6.2

11

0.2

W (m)

Barn ett

211.8

65

L (m)

LM

14.6

Y

Dia. (m)

King

580

Int. Post Dia. (m)

Dom estic

King

14.5

Posts

Int. Posts (n)

N

576

Wall Post Spac. (m)

5.2

LM

Wall Post Dia. (m)

Barn ett

King

Wall Posts (n)

LM

624

Pla za

Posts (n)

Stora ge

Barn ett

Complete

N

LM

Wall Type

Shape

RS 10. 1

King

Locus

Barn ett

Sq uar e Re cta ng ula r

Phase

Nondome stic

Occ.

N

Site

Class

Mound Top

Str. No.

Str. ID

17

121 3

25

1.2

8.4

8.5

71.8

11

0.7

3.8

2.4

9.3

0

N

N/A

0

0

N

28

N

N

1

1

1

1 8

6.1

6.1

37.2

5

N

N/A

3

0

Y

66

N

N

1

1

1

1 8

37

0.62

408

620

King

LM

Barn ett

RS 7

N

Stora ge

Sq uar e Re cta ng ula r

600

King

LM

Barn ett

23. 4

N

Dom estic

Sq uar e

608

King

LM

Barn ett

29

N

Dom estic

630

King

LM

Barn ett

RS 14

N

Stora ge

622

King

LM

Barn ett

RS 9.1

N

Stora ge

631

King

LM

Barn ett

RS 15

N

Stora ge

Sq uar e Re cta ng ula r Re cta ng ula r Re cta ng ula r

588

King

LM

Barn ett

14

N

Dom estic

589

King

LM

Barn ett

15. 1

N

570

King

LM

Barn ett

2.1

573

King

LM

Barn ett

593

King

LM

567

King

574

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

0

1

N

76

N

N

1

1

1

1 8

0

N

N/A

0

0

N

81

N

N

1

1

1

1 8

5

Y

Prepare d clay

1

2

N

64

Y

Y

1

1

1

1 8

Y

Surface fired area

0

1

N

73

N

N

1

1

1

1 8

0

N

N/A

0

0

N

81

N

N

1

1

1

1 8

8

0

N

N/A

0

0

N

13

N

N

2

1

1

1 8

2.5

16.5

0

N

N/A

0

0

N

82

N

N

1

1

1

1 8

8

7.8

62.1

5

Y

Prepare d clay

2

3

N

31

Y

Y

1

1

1

1 8

1.1

9.4

9.5

88.7

Y

Prepare d clay

1

1

N

87

N

N

3

1

1

1 8

0.82

6.6

6.5

43.1

Y

Surface fired area

1

3

N

89

N

N

1

1

1

1 8

5.7

5.9

33.9

Y

Prepare d clay

1

0

N

65

Y

N

1

1

1

1 8

0.9

6.6

6.6

42.9

N

N/A

0

1

N

4

N

N

1

1

1

1 8

39

0.82

9.9

10.1

100.2

Y

Prepare d clay

1

1

Y

86

N

Y

3

1

1

1 8

30

0.76

5.3

5.3

28.4

Y

Prepare d clay

1

0

N

67

Y

Y

1

1

1

1 8

7

7.2

50.4

Y

Prepare d clay

4

N

5

Y

N

1

1

1

1 8

6

6

36

Y

Prepare d clay

1

Y

84

N

N

1

1

1

1 8

29

Posts

Y

Posts

Y

21

Posts

Y

Posts

Y

Posts

1.1

L (m)

W (m)

N/A

Y

Dia. (m)

N

Posts

9.1

9.4

86.1

6.1

2.7

16.6

0.8

6.1

6.5

39.8

25

1.1

9.1

9.1

83.1

9

9

1.1

5.2

2.5

12.9

Y

11

11

0.9

3.5

2.3

Posts

Y

7

7

1.4

6.6

Sq uar e

Posts

Y

34

0.9

Dom estic

Sq uar e

Posts

Y

37

N

Dom estic

Sq uar e

Posts

Y

26

3

N

Dom estic

Sq uar e

Posts

Y

17

Barn ett

19

N

Dom estic

Sq uar e

Posts

Y

20

LM

Barn ett

1.1

N

Dom estic

Sq uar e

Posts

Y

King

LM

Barn ett

4

N

Dom estic

Sq uar e

Posts

Y

577

King

LM

Barn ett

5.3

N

Dom estic

Sq uar e

Posts

Y

601

King

LM

Barn ett

24

N

Dom estic

Sq uar e

Posts

Y

7

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Pd ID

Dom estic

Comp ID

N

Site ID

27

Builds

Barn ett

Burned

LM

Basin

King

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID 606

7

8

409

0

4

1

Orient. (deg E of N)

Burned

Builds

Site ID

Comp ID

Pd ID

N

N/A

0

0

N

37

N

N

2

1

1

1 8

Posts

Y

12

12

0.8

5.2

2.1

11.3

0

N

N/A

0

2

N

86

N

N

3

1

1

1 8

Sq uar e

Posts

Y

39

0.82

7.3

7.4

54.6

Y

Prepare d clay

1

0

Y

32

Y

Y

1

1

1

1 8

Dom estic

Sq uar e

Posts

Y

34

0.9

7.8

8.1

63.5

Y

Prepare d clay

3

0

N

60

Y

Y

1

1

1

1 8

N

Dom estic

Sq uar e

Posts

Y

7.7

7.8

59.6

Y

Prepare d clay

4

N

6

Y

Y

1

1

1

1 8

30

N

Dom estic

Sq uar e

Posts

N

N

N/A

0

6

N

90

N

N

1

1

1

1 8

Barn ett

15. 2

N

Dom estic

Posts

Y

Y

Prepare d clay

1

2

N

90

N

N

1

1

1

1 8

LM

Barn ett

RS 5

N

Stora ge

Posts

N

N

N/A

0

0

N

78

N

N

1

1

1

1 8

King

LM

Barn ett

6.1

N

Dom estic

Posts

Y

Y

Prepare d clay

1

0

N

72

Y

Y

1

1

1

1 8

628

King

LM

Barn ett

RS 12

N

Stora ge

Sq uar e Re cta ng ula r

Posts

Y

N

N/A

0

0

N

56

N

N

1

1

1

1 8

604

King

LM

Barn ett

26. 1

N

Dom estic

Sq uar e

Posts

Y

28

598

King

LM

Barn ett

23. 2

N

Dom estic

Sq uar e

Posts

Y

584

King

LM

Barn ett

9

N

Dom estic

Posts

Y

625

King

LM

Barn ett

RS 10. 2

N

Stora ge

Sq uar e Re cta ng ula r

Posts

Y

596

King

LM

Barn ett

22

N

Dom estic

Sq uar e

Posts

Y

40

582

King

LM

Barn ett

8.1

N

Dom estic

Sq uar e

Posts

Y

37

RS 2.1

N

Stora ge

611

King

LM

Barn ett

RS 1.1

N

Stora ge

583

King

LM

Barn ett

8.2

N

Dom estic

595

King

LM

Barn ett

21

N

578

King

LM

Barn ett

5.4

609

King

LM

Barn ett

590

King

LM

618

King

579

Sq uar e Re cta ng ula r

31

7

10

6

1

L (m)

Locus

Site

Barn ett

W (m)

0

LM

Dia. (m)

9.8

King

Posts (n)

2.1

Shape

4.7

Class

Basin

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Complete

Wall Type

Mound Top

0.9

Str. No.

12

Phase

12

Occ.

Y

Str. ID

Posts

614

Re cta ng ula r Re cta ng ula r

8

8.8

9

79

7

4.9

2.6

12.8

37

6.1

6.3

38.6

10

4.8

1.9

8.9

0.8

7.6

7.8

59.3

N

N/A

0

5

N

66

N

N

2

1

1

1 8

26

0.85

7.3

7.7

56.2

Y

Prepare d clay

2

3

Y

87

Y

N

1

1

1

1 8

28

0.9

7.2

7.4

53.2

6

Y

Prepare d clay

1

2

N

77

Y

Y

1

1

1

1 8

4.3

2.5

10.7

0

N

N/A

0

1

N

0

N

N

1

1

1

1 8

0.9

7.5

7.8

58.2

7

Y

Prepare d clay

2

5

N

46

N

Y

1

1

1

1 8

0.9

7

7.3

50.8

Y

Prepare d clay

1

2

Y

12

Y

Y

2

1

1

1 8

6

410

0

0

Orient. (deg E of N)

Int. Feat. (n)

Basin

Burned

Builds

Site ID

Comp ID

Pd ID

9

1.1

5.2

1.9

10

0

N

N/A

0

0

N

38

N

N

1

1

1

1 8

Y

45

1.1

8.2

8.4

69

Y

Prepare d clay

1

5

Y

86

Y

N

4

1

1

1 8

Posts

Y

16

1.1

7.3

7.3

53.3

N

N/A

0

1

N

85

N

N

1

1

1

1 8

Ke yh ole

Posts

Y

3

2.5

7.5

0

Y

Fire basin

1

0

N

76

Y

N

1

2 9

5 5

3 0

Dom estic

Sq uar e

Posts

Y

28

1

6.6

6.2

41.3

4

Y

Prepare d clay

1

0

N

0

Y

Y

2

7

1 3

8 2

N

Dom estic

Sq uar e

Posts

Y

20

1

7.3

6.3

47.3

6

Y

Prepare d clay

2

5

N

30

Y

N

2

7

1 3

8 2

1

Y

Dom estic

Sq uar e

Posts

Y

11 5

33

1.17

9.6

9.2

89.2

7

Y

Prepare d clay

1

0

N

45

Y

Y

1

7

1 3

8 2

2

N

Dom estic

Ov al

Posts

Y

67

67

0.12

0.34

7.6

5.3

29.7

Y

Surface fired area

7

0

N

N

N

1

4 1

7 2

4 2

1

N

Dom estic

Ov al

Posts

Y

69

69

0.12

0.23

5.8

4.9

18.6

Y

Surface fired area

7

0

N

N

N

2

4 1

7 2

4 2

Dalla s

2

N

Dom estic

Sq uar e

Posts

Y

40 1

0.19

0.75

0.19

6.55

6.58

43.1

Y

Prepare d clay

5

7

Y

102

N

Y

4

5 3

8 5

1 2

Dalla s

3

N

Dom estic

Sq uar e

Posts

Y

22 3

0.19

0.75

0.19

6.86

6.1

41.8

Y

Prepare d clay

6

6

Y

14

N

N

3

5 3

8 5

1 2

Po st Ho use 1

N

Nondome stic

Re cta ng ula r

Posts

Y

77

77

0.15

0.6

0

14.5

14.2

206.5

N

N/A

0

0

N

45

N

N

2

4 0

7 1

7 5

Po st Ho use 8a

N

Nondome stic

Cir cul ar

Posts

Y

32

32

0.1

0.6

0

N

N/A

0

0

N

N

N

1

4 0

7 1

7 5

Barn ett

RS 2.2

N

Stora ge

597

King

LM

Barn ett

23. 1

N

Dom estic

610

King

LM

Barn ett

31

N

635

Kolomoki

MW

1

103

Little Egypt

LM

10 2

102

Little Egypt

LM

101

Little Egypt

LM

473

Lower Saratown

HI

472

Lower Saratown

HI

Swift Cree k Barn ett and Little Egyp t Barn ett and Little Egyp t Barn ett and Little Egyp t Midd le Sarat own Midd le Sarat own

651

Loy

LM

652

Loy

LM

Maco n Plate au Gene ral Early Missi ssippi an

Posts

Y

Sq uar e

Posts

Dom estic

Sq uar e

N

Dom estic

3

N

10 2

2

10 2

Mi ddl e Pla tea u Mi ddl e Pla tea u

9

22

22

82

411

10.1

W (m)

L (m)

Locus

LM

EM

Entry Trenches

Hearth Type

1 8

76.9

5

0

Int. Bur. (n)

Int. Hearth

1

King

Macon Plateau

Int. Walls (n)

1

Y

Area (m2)

1

Posts

Dia. (m)

N

615

471

Int. Post Dia. (m)

Y

Sq uar e Re cta ng ula r

Int. Posts (n)

85

Dom estic

EM

Wall Post Spac. (m)

Y

N

Macon Plateau

Wall Post Dia. (m)

0

11

467

Wall Posts (n)

1

Barn ett

Posts (n)

Prepare d clay

LM

Complete

Y

King

Wall Type

6

Shape

55.1

Class

7.4

Mound Top

7.4

Str. No.

1

Phase

Occ.

Site

Str. ID

28

585

N/A

0

0

1

Y

Fire basin

0

0

Y

Fire basin

1

N

N/A

Y

Fire basin

1

0

Y

Prepare d clay

5.9

0

N

65

1

184.3

55

55

0.15

0.8

0

14.5

13

191.5

Y

11 5

97

0.15

0.3

22

8.2

7

57.3

Posts

Y

13 8

12 3

0.18

0.44

15

0.18

13.7

11.9

163

Cir cul ar

Posts

N

21

20

0.11

0.23

1

0.21

4.6

N

Nondome stic

Irr eg ula r

Posts

Y

54

32

0.13

0.5

22

0.13

5.8

26.4

Lo dg e1

N

Nondome stic

Cir cul ar

Both

Y

10 5

81

0.11

0.23

24

0.19

7.9

49

Ho use B

N

Unkn own/ other

Cir cul ar

Posts

Y

35

0.1

0.2

0

2.7

Lo dg e3

N

Nondome stic

Cir cul ar

Posts

Y

10 5

82

0.11

0.25

23

9.1

Both

Y

68

40

0.12

0.3

28

Both

Y

59

59

0.09

0.3

Posts

N

9

9

0.12

0.45

Macon Plateau

EM

720

Macon Plateau

EM

722

Macon Plateau

723

Macon Plateau

EM

721

Macon Plateau

EM

724

Macon Plateau

EM

725

Macon Plateau

EM

221

Martin Farm

EM

Missi ssippi an I

3

N

Dom estic

219

Martin Farm

EM

Missi ssippi an II

1

N

Dom estic

224

Martin Farm

EM

Missi ssippi an II

6

N

Unkn own/ other

Sq uar e Re cta ng ula r Un ide ntif ied

0

412

0.15

0

Pd ID

Lo dg e2

N

11.9

Comp ID

Nondome stic

N

14.5

Site ID

N

0

0

Builds

Lo dg e 1A

0

0.5

Burned

Unkn own/ other

N/A

0.15

Basin

N

N

100.3

Orient. (deg E of N)

Posts

Ho use A

Re cta ng ula r Re cta ng ula r

N

84

Po st Ho use 2

Entry Trenches

N

Nondome stic

0

0.3

Int. Bur. (n)

Po st Ho use 8

2

0.15

Po st Ho use 5

11.3

Int. Walls (n)

Y

Prepare d clay

Area (m2)

Posts

Y

W (m)

Re cta ng ula r

L (m)

N

Nondome stic

Dia. (m)

84

Int. Post Dia. (m)

Y

Int. Posts (n)

Posts

Wall Post Spac. (m)

Re cta ng ula r

Wall Post Dia. (m)

N

Nondome stic

Wall Posts (n)

Posts (n)

Complete N

470

EM

Wall Type

Mound Top

Str. No.

WT

Int. Feat. (n)

Maco n Plate au

Cir cul ar

Locus

Phase

Occ. EM

Maco n Plate au

Nondome stic

Hearth Type

Macon Plateau

EM

Maco n Plate au

N

Lo dg e 4, 4a, 5, 6

Int. Hearth

Macon Plateau

EM

Shape

468

Macon Plateau

Class

469

Site

Str. ID 726

So uth Pla tea u Mi ddl e Pla tea u Mi ddl e Pla tea u Mi ddl e Pla tea u So uth Pla tea u So uth Pla tea u So uth Pla tea u So uth Pla tea u So uth Pla tea u So uth Pla tea u

Y

N

4

4 0

7 1

7 5

120

N

N

1

4 0

7 1

7 5

N

45

N

N

1

4 0

7 1

7 5

0

N

120

N

N

1

4 0

7 1

7 5

0

N

N

Y

2

4 0

7 1

7 5

N

N

N

1

4 0

7 1

7 5

0

N

N

N

1

4 0

7 1

7 5

1

0

Y

N

N

1

4 0

7 1

7 5

N/A

0

0

N

N

N

1

4 0

7 1

7 5

Y

Fire basin

1

0

N

180

N

N

1

4 0

7 1

7 5

52

4.9

4.9

23.8

0

Y

Prepare d clay

1

0

N

150

Y

Y

1

1 6

2 6

2 5

5.5

4.3

23.4

0

Y

Prepare d clay

7

0

N

63

N

N

1

1 6

2 7

2 6

0

N

N/A

0

0

N

N

N

1

1 6

2 7

2 6

4

N

Dom estic

223

Martin Farm

EM

Missi ssippi an II

5

N

Dom estic

227

Martin Farm

LM

Missi ssippi an III or IV

Mo un d2

9

Y

Dom estic

226

Martin Farm

LM

Missi ssippi an III or IV

Mo un d2

8

Y

Dom estic

229

Martin Farm

EM

Missi ssippi an II or III

11

N

Unkn own/ other

228

Martin Farm

LM

Missi ssippi an II or III

10

N

Unkn own/ other

225

Martin Farm

LM

Missi ssippi an III or IV

7

Y

Dom estic

220

Martin Farm

LM

Missi ssippi an IV

2

N

Stora ge

111

McFarlan d

MW

McFa rland

Ar ea L

5

N

McFa rland

Ar ea A

4

107

McFarlan d

MW

Mo un d2

1

Y

35

Posts

N

Posts

Orient. (deg E of N)

Entry Trenches

Int. Bur. (n)

Int. Feat. (n)

Hearth Type

Int. Hearth

Int. Walls (n)

Area (m2)

Int. Post Dia. (m)

0

0

N

22

N

N

1

1 6

2 7

2 6

0

Y

Fire basin

5

0

N

24

N

N

1

1 6

2 7

2 6

0

N

N/A

0

0

N

63

N

N

1

1 6

2 7

2 6

N

N/A

0

0

N

135

N

N

2

1 6

2 8

2 7

N

N/A

3

1

N

62

N

N

1

1 6

2 8

2 7

Y

Prepare d clay

2

0

N

N

N

1

1 6

2 8

2 7

0

Y

Prepare d clay

2

4

N

N

N

2

1 6

2 8

2 7

63.4

2

Y

Prepare d clay

6

0

N

N

N

1

1 6

2 8

2 7

1.7

2.3

0

N

N/A

0

0

N

N

N

1

1 6

2 9

2 8

L (m)

W (m)

N/A

Dia. (m)

Int. Posts (n)

Wall Post Spac. (m)

Wall Post Dia. (m)

Wall Posts (n)

Posts (n)

Complete

Posts

N

0.15

0.15

6.9

6.7

46.2

35

0.15

0.45

5

4.6

23

22

22

0.11

0.45

4.3

4.1

N

20 9

20 9

0.15

7.9

2.7

Both

Y

20 9

14 3

0.12

0.07

8.8

7.4

Posts

N

69

0.15

0.33

6.1

4.3

Posts

N

71

0.18

4.9

4.9

24

Both

Y

9.2

6.9

Cir cul ar

Posts

Y

Dom estic

Cir cul ar

Posts

Dom estic

Cir cul ar

N

N

N

Pd ID

Missi ssippi an II

Both

Comp ID

EM

Dom estic

Site ID

Martin Farm

Y

Builds

222

12

Burned

EM

Basin

Martin Farm

Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r Re cta ng ula r

Wall Type

Shape

Class

Mound Top

Str. No.

Locus

Phase

Occ.

Site

Str. ID 230

Missi ssippi an II

65.5

2

23 1

0.09

0.06

14

14

0.12

0.3

0

Y

32

28

0.15

0.78

4

0.15

6

31.2

0

N

N/A

4

0

N

N

N

1

1 4

2 4

3

Posts

Y

48

41

0.15

0.74

7

0.15

6

31.7

0

N

N/A

2

0

N

N

N

1

1 4

2 4

3

Dom estic

Cir cul ar

Posts

Y

41

31

0.15

0.74

10

0.15

7

37.9

0

N

N/A

3

0

N

N

N

1

1 4

2 4

3

Cir cul ar

Posts

Y

33

28

0.15

0.79

5

0.15

7

33.7

0

N

N/A

7

0

N

N

N

1

1 4

2 4

3

Posts

Y

47

36

0.15

0.61

11

0.15

7

36.8

0

N

N/A

4

0

N

N

N

1

1 4

2 4

3

Posts

Y

36

26

0.16

0.8

10

0.13

13.3

0

N

N/A

1

0

N

N

N

1

1 5

2 5

8 3

110

McFarlan d

MW

McFa rland

Ar ea B/I

108

McFarlan d

MW

McFa rland

Ar ea B/I

2

N

Dom estic

109

McFarlan d

MW

McFa rland

Ar ea J/B

3

N

Dom estic

118

Mialoquo

HI

Cher okee

Ar ea 3

8

N

Dom estic

Cir cul ar Re cta ng ula r

413

4.6

2.9

150

65

N

Dom estic

117

Mialoquo

HI

Cher okee

Ar ea 4

6

N

Dom estic

497

Mialoquo

HI

Cher okee

Ar ea 5

7

N

Nondome stic

113

Mialoquo

HI

Cher okee

Ar ea 3

2

N

Dom estic

Cir cul ar Re cta ng ula r

466

Mitchum

HI

1

N

Dom estic

686

683

693

692

690

Morris

Morris

Morris

Morris

Morris

EM

EM

EM

EM

EM

Mitc hum Kent ucky Early Missi ssippi an Kent ucky Early Missi ssippi an Kent ucky Early Missi ssippi an Kent ucky Early Missi ssippi an Kent ucky Early Missi ssippi an

4

1

13

12

10

N

N

N

N

N

Pd ID

4

Comp ID

Ar ea 4