A BIOCULTURAL PERSPECTIVE ON SEX AND GENDER IN LATE PREHISTORIC
WEST CENTRAL ILLINOIS: GROWTH PATTERNS, MISSISSIPPIANIZATION, AND
INTRACEMETERY SOCIAL DIFFERENTIATION
By
Jennifer Dyhan Bengtson

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Anthropology
2012

ABSTRACT
A BIOCULTURAL PERSPECTIVE ON SEX AND GENDER IN LATE PREHISTOFIC
WEST-CENTRAL ILLINOIS: GROWTH PATTERNS, MISSISSIPPIANIZATION, AND
INTRACEMETARY SOCIAL DIFFERENTIATION
By
Jennifer Dyhan Bengtson

Uncontextualized sex-based analyses of stress and health provide simplistic, essentialized
perspectives on trends among males and females in general, propagating the idea of a singular
experience of “maleness” or “femaleness.” A research design that incorporates multiple social
units of analysis avoids essentialization of the experiences of men and women by providing a
means of discovering the ways stress and growth might vary within each sex, related to social
factors like status, kinship, and/or community/corporate membership. The research outlined here
incorporates a traditional, sex-based perspective on growth patterns over the Late WoodlandMississippian transition in West Central Illinois. However, detailed artifactual and spatial
analyses of the Mississippian component of the Schild cemetery allows for further division of
this portion of the sample into finer scale, socially meaningful units of analysis, which leads to a
more engendered interpretation of biological patterns and contributes to reconstructions of
gendered social identity.
Linear enamel hypoplasia (LEH) is used to isolate early childhood stress experiences,
while femur length is used to assess cumulative growth, which continues into early adulthood. A
less traditional method of growth assessment based on vertebral dimensions is also incorporated.
Vertebral arch growth is complete by early childhood, and it has been hypothesized that neural
canal diameter (NCD) can be used as indicator of early childhood growth corresponding roughly
to the same period as LEH formation. Vertebral body height (VBH) increases until early

adulthood, corresponding closely with the timing of femoral growth. Comparison between these
four indicators provides a detailed account of growth experience as well as a new test of the
vertebral method. Diachronic analysis assesses differences in male and female growth trends
over the Late Woodland-Mississippian transition, while more detailed analysis of the Schild
Mississippian materials allows for an assessment of the ways that male and female growth
patterns vary by burial treatment based on artifact assemblages, body positioning, burial areas,
and charnel association.
Analysis of NCD suggest that the conditions of early childhood growth improved over
the Late Woodland-Mississippian transition for females, while male growth patterns remained
stable. Mississippian males experienced fewer LEH than their Late Woodland counterparts, with
no corresponding change among females. Femur and VBH data indicate no change in cumulative
growth patterns for either males or females over time. However, a closer examination of the
Schild Mississippian component indicates intracemetery variation in these patterns among
females, with those afforded certain burial treatments, those buried in particular areas, and those
associated with charnel structures apparently experiencing better cumulative growth conditions
than other females at the site. It is suggested that some women in the broader Schild
Mississippian community experienced better biosocial circumstances of growth based on their
association with certain status, kinship/community networks, and/or the timing of acculturation.

Copyright by
JENNIFER DYHAN BENGTSON
2012

For BoomerYou were not a good dog, but you were a nice dog.

v

ACKNOWLEDGMENTS

It takes a village to raise a child, as well as to see a dissertation through to its completion. First
and foremost, I would like to thank my son, Gabriel Bengtson. Having been born during the
early stages of my undergraduate education, he grew up being dragged to libraries, conferences,
labs, classes, and department functions, and he took it all in stride. His patience and maturity are
admirable and he is, and always will be, the accomplishment of which I am most proud. I would
also like to thank my mother, Esparanza Villalobos, and my sisters, Marisa and Jessica Bengtson.
Without their assistance in raising Gabe and their regular reminders about what really matters in
life, I would have lost my mind long ago. A special thank you is in order for my dog, Violet
Bengtson, for keeping my feet warm as I typed for hours. Thanks also to the late Boomer
Bengtson, who was always ready to distract me from my work with sudden, unjustifiable barking
and leaping about. Rest in peace, my most loyal friend.
Of course, I want to thank the members of my Guidance Committee. Dr. Norm Sauer was
a source of both academic inspiration and comic relief, and I am also grateful for his assistance
in the construction of Pinewood Derby cars. Dr. Jodie O’Gorman is a great mentor and a great
friend who has challenged me to challenge my own notions of gender in archaeological analysis.
Dr. Lynne Goldstein has been an invaluable source of expertise and advice regarding
Mississippian archaeology in general and the Schild site in particular, and (for many reasons)
there is truly no way I could have finished my program without her. Dr. Gillian Bice is the best
editor a dissertation writer could ask for, and her thoughtful comments compelled me to tighten
both my writing and my logic. I could not imagine a better team of mentors.
Professors at other universities were also invaluable to the completion of this dissertation.
Dr. Della Cook at Indiana University shared her vast expertise on the biocultural prehistory of
the midcontinent. She and her students made me feel at home during my data collection.
Bloomington and Dr. Cook’s lab will always hold a very special place in my heart. Dr. Rinita
Dalan at Minnesota State University Moorhead was my undergraduate advisor. I got hooked on
anthropology in one of her introductory courses, and although I was not the most promising
student at first, for some reason she never gave up on me. Rinita provided me with research and
vi

learning opportunities that shaped me as a scientist and a teacher. I could write a whole chapter
about my gratitude toward her.
Moral and academic support from like-minded students was an important part of my
graduate experience. In particular, I want to thank Amy Michael at Michigan State University for
her friendship, her conversation (academic and otherwise), for being as hungry and vegan as I,
and for seeing me through some of the most difficult moments of my life. Susan Spencer at
Indiana University helped me navigate the Schild and Yokem collections, eagerly shared her
knowledge, and listened to me vent about data collection frustrations. Both of these women are
great friends and great scholars, and I am honored to have had the opportunity to work with
them.
Paul Curran at the MSU Center for Statistical Training and Consulting was patient and
helpful, and I would especially like to thank him for promptly replying to my numerous,
panicked, late night emails. Nancy Smith, Graduate Secretary in the MSU Anthropology
Department, helped me with technicalities, formalities, forms, and protocols that I never could
have dealt with on my own. Also, she was great fun to visit with when I had to take breaks from
SPSS.
I would like to express my extreme gratitude to the Gender, Justice, and Environmental
Change program at Michigan State University’s Center for Gender in Global Context. They
funded this research from data collection to completion, and I could not have finished in a timely
manner without their generous support.
Finally, a very special thank you to Paul Berninger for more things than I could possibly
list here.

vii

TABLE OF CONTENTS

LIST OF TABLES

x

LIST OF FIGURES

xv

CHAPTER 1
INTRODUCTION
The Research Problem
Theoretical Framework
Synthesizing Biological and Cultural Perspectives
Engendering Bioarchaeology
Research Questions
Organization

1
1
5
14
15
17
24

CHAPTER 2
REGIONAL AND BIOCULTURAL BACKGROUND
Geographic Setting
The Late Woodland Tradition
The Mississippian Tradition
Sex and Gender in the Late Woodland and Mississippian Worlds
Situating the Current Study within the Existing Literature

25
25
25
30
46
53

CHAPTER 3
GROWTH AND STRESS IN BIOANTHROPOLOGICAL ANALYSIS
Understanding Stress
Growth and Heritability in Living Populations and Historic Data
Stress and Growth in Prehistoric Populations
Social Meaning of Intragroup Variation in Growth Experience
Osteological Parameters Used in the Current Study

56
57
57
59
60
61

CHAPTER 4
SITES AND SAMPLE DESCRIPTIONS
West Central Illinois- Cultural Setting
The Sites
Late Woodland Components
Mississippian Components
Previous Bioarchaeological Research

70
70
71
73
78
86

CHAPTER 5
ANALYTICAL METHODS, MATERIALS, AND HYPOTHESES
Data Collection Methods
Schild Mortuary Data

89
90
94

viii

Statistical Methods for Incorporating Osteological and Archaeological Data
Hypotheses
Sample Sizes
CHAPTER 6
RESULTS
Hypothesis One- Internal Consistency for Segmental Mean Analysis
Hypothesis Two- Utility of the Vertebral Method of Growth Assessment
Hypothesis Three- Temporal Trends in LEH Frequency
Hypothesis Four- Temporal Trends in Femur Length
Hypothesis Five- Temporal Trends in Vertebral Growth
Temporal Trends in Growth Disruption: Individual Vertebrae
Hypothesis Six- Growth Patterns and Social Differentiation among the Schild
Mississippians
Intracemetery Trends in Growth Disruption: Individual Vertebrae
Hypothesis Seven- Multifactor Approaches
CHAPTER 7
INTERPRETATIONS AND DISCUSSION
Intrasegment Reliability
Utility of the Vertebral Method
Interpretation of Growth Data within Temporal Framework
Growth Patterns and Social Differentiation among Schild Mississippians
Engendered Interpretations
Regional Perspectives on Mississippianization, Health, Growth, and Social
Identity

100
104
111

114
114
115
118
119
120
123
125
138
143

144
145
146
155
158
164
172

CHAPTER 8
CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS
Summary
Future Research Directions

181
181
187

APPENDIX

195

WORKS CITED

221

ix

LIST OF TABLES

Table 1- Radiocarbon dates and phases for study sites

71

Table 2- Mortuary category descriptions

96

Table 3- Overall sample sizes

112

Table 4- Sample sizes for Schild burial areas

112

Table 5- Sample Sizes for Schild Knolls

112

Table 6- Sample sizes for Schild charnel association

112

Table 7- Sample sizes for Schild artifact only clusters

113

Table 8- Sample sizes for artifact and positioning clusters

113

Table 9- Thoracic Intrasegment Reliability

114

Table 10- Lumbar Intrasegment Reliability

115

Table 11- Linear Regression for Schild Female Lumbar Vertebral Measurements and Femur
Length
115
Table 12- Linear Regression for Schild Male Lumbar Vertebral Measurements and Femur
Length

116

Table 13- Lumbar NCD T-Tests between Late Woodland Males and Late Woodland
Females

117

Table 14- Lumbar NCD T-Tests between Mississippian Males and Mississippian Females

117

Table 15- Male LEH Frequency Chi-square test between Late Woodland and Mississippian 118
Table 16- Female LEH Frequency Chi-square test between Late Woodland and
Mississippian

118

Table 17- Male Right Femur Length T-Test between Late Woodland and Mississippian

119

Table 18- Female Right Femur Length T-Test between Late Woodland and Mississippian

119

Table 19- Male Thoracic Dimensions Means T-Tests between Late Woodland and
Mississippian

121

x

Table 20- Male Lumbar Dimension Means T-Tests between Late Woodland and
Mississippian

121

Table 21- Female Thoracic Dimensions Means T-Tests between Late Woodland and
Mississippian

122

Table 22- Female Lumbar Dimensions Means T-Tests between Late Woodland and
Mississippian

122

Table 23- Male Individual Vertebra T-Tests between Late Woodland and Mississippian

123

Table 24- Female Individual Vertebra T-Tests between Late Woodland and Mississippian

125

Table 25- Schild Male Thoracic Vertebral Dimensions T-test by Knoll

126

Table 26- Schild Male Lumbar Vertebral Dimensions T-test by Knoll

126

Table 27- Schild Male Right Femur Lengths T-test by Knoll

126

Table 28- Schild Male LEH frequency Chi-square by Knoll

127

Table 29- Schild Female Thoracic Dimensions T-tests by Knoll

127

Table 30- Schild Female Lumbar Dimensions T-tests by Knoll

127

Table 31- Schild Female Right Femur Length T-tests by Knoll

128

Table 32- Schild Female LEH Frequency Chi-square test by Knoll

128

Table 33- P-values for Schild Male Lumbar Dimensions by Burial Area ANOVAs

128

Table 34-Schild Male Right Femur Length ANOVA by Burial Area

128

Table 35- Schild Male LEH frequency Chi-square by Burial Areas

129

Table 36- P-values for Schild Female Thoracic Dimensions by Burial Area ANOVAs

130

Table 37- P-values for Schild Female Lumbar Dimensions by Burial Area ANOVAs

130

Table 38- Schild Female Lumbar AVBH Descriptive Statistics by Burial Area

130

Table 39- Schild Female Right Femur Length ANOVA by Burial Areas

130

Table 40- Schild Right Femur Length T-test by Burial areas (4 versus all others)

130

Table 41- Schild Female Mean AVBH T-test by Burial Areas (4 versus all others)

130

Table 42- Schild Females LEH frequency Chi-square by Burial Area

130

xi

Table 43- Schild Male Lumbar Vertebral Dimensions T-tests by Charnel Association

131

Table 44- Schild Male Right Femur Length T-test by Charnel Association

132

Table 45- Schild Male LEH Frequency Chi-square by Charnel Association

132

Table 46- Schild Female Lumbar Vertebral Dimensions ANOVA by Charnel Association

132

Table 47- Schild Female Right Femur Length T-test by Charnel Association

133

Table 48- Schild Female LEH Frequency Chi-square test by Charnel Association

133

Table 49- Schild Male Thoracic Dimensions T-tests by Artifact-Only Clusters

133

Table 50- Schild Male Lumbar Dimensions T-tests by Artifact-Only Clusters

134

Table 51- Schild Male Right Femur Length T-test by Artifact-Only Clusters

134

Table 52- Schild Male LEH Frequency Chi-square by Artifact Only Clusters

134

Table 53-Schild Female Thoracic Dimensions T-Tests by Artifact-Only Clusters

135

Table 54- Schild Female Lumbar Dimensions T-tests by Artifact-Only Clusters.

135

Table 55- Schild Female Right Femur Length T-test by Artifact Only Clusters

135

Table 56- Schild Female LEH Frequency Chi-square Test by Artifact-Only Clusters

135

Table 57- Schild Male Thoracic Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

136

Table 58- Schild Male Lumbar Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

136

Table 59- Schild Male Right Femur Length T-test by Artifact and Positioning Clusters

136

Table 60- Schild Male LEH Frequency Chi-square test by Artifact and Positioning Clusters 136
Table 61- Schild Female Thoracic Vertebral Dimensions T-tests by Artifact and Positioning
Clusters
137
Table 62- Schild Female Lumbar Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

138

Table 63- Schild Female Right Femur Length T-test by Artifact and Positioning Clusters

138

xii

Table 64- Schild Female LEH Frequency Chi-square by Artifact and Positioning Clusters

138

Table 65- Schild Female Individual Vertebral T-tests by Knoll

139

Table 66- P-values from Schild Males Individual Vertebrae ANOVAs by Burial Area

140

Table 67- P-values from Schild Females Individual Vertebrae ANOVAs by Burial Area

140

Table 68- Schild Male Individual Vertebral T-tests by Artifact Only Cluster

141

Table 69- Schild Male Individual Vertebral T-tests by Artifact and Positioning Clusters

141

Table 70- Schild Female Individual Vertebral T-tests by Artifact Only Clusters

142

Table 71- Schild Female Individual Vertebral T-tests by Artifact and Positioning Clusters

142

Table 72- P-values from Two-way ANOVAs

143

Table 73- Descriptive statistics for significant Schild male individual vertebral dimensions by
burial area
208
Table 74- Descriptive statistics for significant Schild female individual vertebral dimensions by
burial
209
Table 75- Descriptive statistics for Schild male mean lumbar vertebral dimensions by burial
area
209
Table 76- Descriptive statistics for Schild female mean lumbar vertebral dimensions by burial
area
210
Table 77- Sample sizes for multiway ANOVA- mean female lumbar APNCD by charnel
association and artifact/position clusters

211

Table 78- Multiway ANOVA table for mean female lumbar APNCD by charnel association and
artifact/position clusters
211
Table 79- Sample sizes for multiway ANOVA- mean female lumbar APNCD by mound and
artifact/position clusters
213
Table 80- Multiway ANOVA table for mean female lumbar APNCD by mound and
artifact/position clusters

213

Table 81- Sample sizes for multiway ANOVA- Schild male femur length by charnel association
and artifact only clusters
215

xiii

Table 82- Multiway ANOVA table for male femur length by charnel association and artifact
only clusters
215
Table 83- Sample sizes for multiway ANOVA- mean female lumbar AVBH by charnel
association and artifact only clusters

217

Table 84- Multiway ANOVA table for mean female lumbar AVBH by charnel association and
artifact only clusters
217
Table 85- Sample sizes for multiway ANOVA- mean male lumbar AVBH by mound and artifact
only clusters
219
Table 86- Multiway ANOVA table for mean male lumbar AVBH by mound and artifact only
clusters
219

xiv

LIST OF FIGURES

Figure 1- The American Bottom

26

Figure 2- Growth Curves

62

Figure 3- Vertebral neural canal size in children and adults

63

Figure 4- Site locations

72

Figure 5- Map of Schild Site

74

Figure 6- Schild Late Woodland Mound Nine, showing spatial relationship to Mississippian
Knoll A
75
Figure 7- Map of Yokem Site

77

Figure 8- Schild Mississippian Component

78

Figure 9- Schild Knoll A rows

80

Figure 10- Charnel structure and hypothesized burial divisions in Schild Knoll A

82

Figure 11- Charnel structure and hypothesized burial divisions in Schild Knoll B

83

Figure 12- Yokem Mississippian Mound 2 showing charnel structure

85

Figure 13- Vertebral dimensions

91

Figure 14- Schild Burial Areas 3 and 4

98

Figure 15- Schild Burial Areas 1 and 2

99

Figure 16- Spatial distribution of Schild Knoll A females used in the current study

196

Figure 17- Spatial distribution of Schild Knoll B females used in the current study

197

Figure 18- Spatial distribution of Schild Knoll A males used in the current study

198

Figure 19- Spatial distribution of Schild Knoll B males used in the current study

199

Figure 20- SPSS scatterplot of Schild female lumbar AVBH and femur length

200

Figure 21- SPSS scatterplot of Schild female lumbar PVBH and femur length

201

Figure 22- SPSS scatterplot of Schild female lumbar TNCD and femur length

202

xv

Figure 23- SPSS scatterplot of Schild female lumbar APNCD and femur length

203

Figure 24- SPSS scatterplot of Schild male lumbar AVBH and femur length

204

Figure 25- SPSS scatterplots of Schild male lumbar PVBH and femur length

205

Figure 26- SPSS scatterplot of Schild male lumbar TNCD and femur length

206

Figure 27- SPSS scatterplot of Schild male lumbar APNCD and femur length

207

Figure 28- Multiway ANOVA plot for mean female lumbar APNCD by charnel association
and artifact/position clusters
212
Figure 29- Multiway ANOVA plot for mean female lumbar APNCD by mound and
artifact/position clusters

214

Figure 30- Multiway ANOVA plot for male femur length by charnel association and artifact
only clusters
216
Figure 31- Multiway ANOVA plot for mean female lumbar AVBH by charnel association
and artifact only clusters
218
Figure 32- Multiway ANOVA plot for mean male lumbar AVBH by mound and artifact
only clusters
220

xvi

CHAPTER ONE- INTRODUCTION

The Research Problem
The relationship between health status and social status has been noted by researchers studying
both modern (Nguyen and Peschard 2003, Leatherman 1998) and past populations (Larsen 1997,
Goodman 1998). Decreasing levels of community health and increasing status-related health
differentials have been associated with transitions to agriculture and increased social complexity,
in many cases involving concomitant increases in gender related health disparities (Ambrose and
Buikstra 2003, Cohen and Bennett 1993, Cucina and Tiesler 2001, Lukacs 1992, White 2005).
However, the particulars of the complex relationships between social complexity and heath
disparities should not be simplified or generalized, especially when community or
intracommunity perspectives are desired. For those with biocultural interests in the late
prehistory of the Eastern United States, the spread of Mississippian culture/influence throughout
the Midwest and its potential effects on social relations and health disparities in communities
outside of the American Bottom provide a particularly fascinating case study.
This dissertation investigates the interrelationships between growth patterns, social
identity, and social complexity through time on the Mississippian periphery using skeletal
remains from the Late Woodland and Mississippian components of the Schild and Yokem sites
in West Central Illinois. Changes in age-sensitive patterns of stress and growth between males
and females are investigated through an analysis of enamel defects, long bone growth, and
vertebral growth. Data is analyzed within a framework of existing mortuary data and interpreted
with reference to contemporary gender scholarship and ideas regarding the specific nature of
Mississippianization in West Central Illinois. This chapter will introduce the problems to be

1

investigated, make an argument for the suitability of a biocultural perspective for approaching
these questions, outline general hypotheses, and briefly summarize results and interpretations.

Conceptual Framework
Late prehistoric West Central Illinois was occupied by a Late Woodland population that
eventually became Mississippianized via acculturation and limited colonization (Delaney-Rivera
2004, 2007, Schroeder 2004). Consequently, by Mississippian times, these communities began to
display mortuary patterns and material culture in keeping with the Mississippian phenomenon
elsewhere in the region. By analyzing the skeletal series from the Schild and Yokem sites, the
first consideration of this dissertation is a further investigation of the extent to which, in
becoming Mississippianized, traditional social relations were transformed in these communities.
It is clear that the nature of mortuary ritual was altered, but did these changes reflect concomitant
changes in the lived experiences of people? Did changes in intracommunity social relations also
occur or was the basic structure of daily life part of some body of social relations that was either
maintained from the previous period or otherwise differentially incorporated? Can changing
gender relations be detected osteologically as changes in relative stress, particularly insofar as
differential occurrence of stress in males and females, as interpreted within the context of
diachronic and intracommunity archaeological research? Is this type of phenomenon expressed
similarly across the Mississippian world?
This study can be broken down into the following four areas:
1. Diachronic Change- Between the Late Woodland and Mississippian skeletal series at
Schild and Yokem, are there changes in the overall frequency of osteological indicators
of biological stress and growth? Do differences in these patterns correspond with sex and,
2

if so, how are these patterns similar or different between the Late Woodland and
Mississippian components?
2. Relating observed osteological patterns to gender transformations- Can the observed
stress and growth patterns in the Late Woodland and Mississippian series be interpreted
in regard to possible gender-related health disparities and in terms of changing social
structure and life ways at this time and place? Because gender, as a cultural construct,
cannot be observed directly from skeletal remains, how does the explicit incorporation of
mortuary analysis and age-sensitive osteological data clarify the sex-specific patterns of
growth and stress observed in these skeletal series?
3. Regional comparisons- How do the observed patterns of gender-related health
disparities and changes over time differ from what we see at other Mississippian villages
and centers (i.e. Dickson mounds and American Bottom sites)? What does this mean as
far as differential affects of Mississippianization between peripherally versus centrally
located sites?
4. Broader applications- Given the results of this investigation, what can this study
contribute to the broader body of knowledge regarding the relationship between health,
gender/social identity, and cultural transitions? To what extent do outlying communities
who are culturally/ethnically affiliated with core sites continue to exhibit traditional
gender relations and patterns of community health in spite of having become part of a
larger, more complex system? How can the answers to these questions contribute to a
discussion of the extent of influence of centers over peripheral peoples or of the ability of
peripheral peoples to selectively incorporate or resist certain aspects of the prevailing
cultural pattern?

3

A combined osteological and archaeological approach allows for a biocultural
exploration of these questions. Multiple indicators of early childhood and late adolescent growth
were evaluated within a temporal and cultural framework and compared to published reports
regarding local and regional health patterns. Linear enamel hypoplasia (LEH) and long bone
length are frequently used measures of growth patterns and disruptions in bioarchaeological
studies, but although several authors have noted the potential sensitivity of vertebral growth to
stress related disruptions (Clark et al 1986, Larsen 1998, Hoppa and Fitzgerald 1999, Tatarek
1999, Porter 1985), few have applied this knowledge to bioarchaeological situations. Clark’s
(1985, 1988) vertebral morphometric method for growth assessment is intended to go beyond
traditional methods of evaluating sexual dimorphism by distinguishing between chronic and
acute growth disruptions, the recognition of which has culturally significant implications. The
method is based on the observation that vertebral arch growth ceases before vertebral body
growth (during childhood versus during late adolescence/early adulthood) (Clark 1988, Baker et
al 2005). Clark suggests that analysis of the relationship between measurements of these two
features may be useful in determining whether growth disruptions are chronic, resulting from
generally continuous stress and deprivation throughout the growth period, or whether catch up
growth occurred when stressful situations were eventually ameliorated later in childhood.

A Note on Terminology
Throughout this dissertation, the terms sex, gender, male, female, man, and woman are used
frequently. Following Walker and Cook (1997), the term sex is used specifically to refer to the
biological assignment of individuals into one of two categories: male or female. Gender refers to
the variable and dynamic cultural values, ideas, expectations, and social roles collectively

4

recognized by a culture, based loosely on perceptions of biological sex and sexuality. Man and
woman are gender terms used in Western societies to differentiate people within a binary gender
framework. In this dissertation, these gender terms are used only when sex-based biological data
has been interpreted within a cultural context provided by mortuary and archaeological studies.
In general, then, sex terminology is employed when discussing uncontextualized data and when
reviewing the results of statistical analyses, while gender terminology is used when discussing
interpretations based on considerations of archaeological perspectives. It is known that many
traditional Native American societies recognized three or more genders (Hollimon 2009), and it
should be noted that some of the variation seen within male and female categories may, in part,
reflect this.

Theoretical Framework
The Case for a Bioarchaeological Approach to Evaluating Prehistoric Social Change
Bioarchaeology refers to the integration of human biological remains and archaeological analysis
in order to approach and integrate such topics as burial programs and social organizations,
division of labor, paleodemography, diet, and disease. Although the term is used in Europe to
refer to the analysis of any biological materials (i.e. faunal and botanical) from archaeological
sites, American usage of the term refers specifically to the study of human remains (Buikstra
2006). Larsen (1997) contends that bioarchaeology, as a modern discipline, is inherently
interpretive (as opposed to descriptive) and that bioarchaeological analysis takes place at the
population level. Although the creation of descriptive osteological reports has long been standard
practice in archaeological investigations, the tendency for bioarchaeologists to rigorously

5

incorporate social theory, analytical techniques, and statistical methods as a significant
component of their work is a relatively recent phenomenon.
Bioarchaeological research is uniquely positioned on the cusp of biological and cultural
studies of the human past. As such, this research is critical for investigating the interaction
between health and the natural/social environment of past populations. The primary goal of the
modern bioarchaeological approach is the development of a history of the human condition on
local, regional, and global levels, and the methodologies available for accomplishing this goal
are diverse and evolving (Larsen 1997, 2002; Knudson and Stowjanowski 2008). However,
some have pointed out that progress in anthropological skeletal analysis has been severely
hindered by the persistence of historical-typological approaches originating in racialized studies
of the late eighteenth and early nineteenth centuries (Larsen 2002, Armelagos 2003, Woploff and
Caspari 2000). Armelagos (2003) points out that even researchers such as E.A. Hooton, who are
seen as pioneers in the biocultural and populational approaches, were not immune to tendencies
toward static, typological constructions and failed to focus on heterogeneity, variability, and
change as legitimate bases for osteological research. However, others suggest that, when
evaluated within the context of their own historical situations, even the earliest bioarchaeological
studies implicitly demonstrate the dynamic and interpretative attributes characteristic of the
discipline in its modern iteration (Buikstra 2006, Cook 2006).
Although the modern American approach to bioarchaeological analysis emerged in the
wake of the “New Archaeology,” the advent of processualism (with its concomitant movement
away from the culture-historical categorizations of the early twentieth century) and the
availability of advanced statistical software may have done little more than superficially change
the typological and static inter-populational comparisons that had previously dominated the field.

6

For example, Blakely (1998) shows that the continued application of a typological, racialized
model of human biology resulted from (and reified) the institutionalized marginalization of
minority ethnic groups. Furthermore, reliance on a purely biological model served to remove
historic documents and artifactual analysis from the process, essentially hindering the
construction of a socially meaningful research agenda and making it impossible to properly
reconstruct ethnic identity (Blakely 1998, Goldstein 2006).
Larsen (2002) points to more recent processual studies that have moved away from
typologies to address and explain change in human populations over time. He points out that,
although critiques of the approach outlined above were increasingly common from the early 20th
century onward, it is not until the last 20 years that these criticisms seem to make a real
difference in the types of questions being asked by bioarchaeologists. Larsen suggests that the
shift is due to 1) changing theoretical perspectives regarding the plasticity (as opposed to the
fixity) of the human body, 2) a greater understanding of the biology of the skeleton and its
relationship to the non-biological aspects of human existence, and 3) new methodologies for
gleaning information from human bones. Furthermore, Knudson and Stowjanowski (2008)
demonstrate that postmodern theoretical orientations in the social sciences have recently been
melded with the bioevolutionary focus on change and variation to create a new bioarchaeological
synthesis that focuses on social identities in the past. They argue that bioarchaeologists are in a
unique position to contribute novel ideas regarding a range of social identities in the past, but
that they can only do so through more rigorous application of social theory, archaeology, and
ethnohistory. They also suggest that it is because of recent methodological advances in
excavation, aging, sexing, and chemical and statistical analyses that appropriate data sets can

7

now be produced and interpreted through the lens of contemporary social science theory on
identity (ethnic, gender, age, etc.), embodiment, and disability.
Postmodern social theory is not, however, the only means by which theoretically minded
bioarchaeologists might uphold and expand upon the biocultural approach. For example, Smith
and Thomas (1998) discuss the ways that biological anthropologists interested in modern and
archaeological societies have explicitly incorporated political and economic theory as way to
explain the relationship between class and health in modern populations. Explicitly Marxist
approaches have led to deeper understandings of the relationship between social situation
(including gender) and the body through the application of such concepts as dialectical
adaptation, the political economy of human biology, and the development of a critical biological
anthropology (Smith and Thomas 1998, Mascia-Lees and Black 2000). This approach allows for
the incorporation of a sophisticated social theory while wholly preserving the materialist
approach focusing on class and status differentials that many bioarchaeologists find appealing
and useful.
One rapidly expanding area of archaeological interest is the analysis of gender in past
populations (Hill 1997, Stig Sorenson 2000, Milledge Nelson and Rosen-Ayalon 2002). Walker
and Cook (1998) and Armelagos (1998) have both decried the tendency of bioarchaeologists to
use the terms sex and gender interchangeably and to treat sex as a direct proxy for gender in the
archaeological record. They both insist that more care must be taken in the application of
terminology, making it clear that sex refers to the biological differences between males and
females, while gender refers to culturally-specific differentiation of social roles, relations, and
ideologies, often based loosely on biological sex. Walker and Cook (1998) particularly

8

emphasize that sex-based bioarchaeological analyses must be nested securely in gender-based
archaeology in order to appropriately engender our views of the past.
Gellar (2008) agrees with the above arguments on a general level, but criticizes the
generalized and persistent ignorance of these issues in the bioarchaeological literature. Most
bioarchaeologists have taken an essentialist view of differences between males and females and
have failed to incorporate more than overly simplistic social theory when making gendered
interpretations of the past. Her criticisms do not spare Walker and Cook (1998) and Armelagos
(1998), who rely on decades-old gender literature rather than incorporating relevant
contemporary social theory, resulting in superficial interpretations that serve to justify and
naturalize modern western ideas about gender roles and relations. This has resulted at least in
part from our reliance on biomedical approaches to sex and gender, which do not incorporate
anthropological concepts of sex, gender, and bodies. Biomedical studies generally lead to the
production of normative statements that reify assumptions about what men and women do, and
how they are constrained by their biologies (Gellar 2005, Sofaer 2006).
Gellar also (2008) laments that many biological anthropologists have not been trained to
think in terms of social theory and have been excluded from the interpretive process. She sees
this as a result of both their own apprehensions and of the assumptions of theory-makers who
operate under the post-modern notion that materialist studies (such as those of the body) have
nothing to offer the social sciences on an interpretive level. Identity is often considered
conceptually, but it is expressed materially and corporeally and is therefore amenable to both
archaeological and bioarchaeological investigation (Gellar 2005, 2008; Sofaer 2006). Gellar
insists that the first steps in fixing this problem must be to critically reevaluate our sexing
methodologies, our concept of the relationship between sex and gender, and the ways we

9

implicitly and explicitly make gendered interpretations from incomplete and/or uncontextualized
biological data.
Gender is a social construct that does not relate directly to biological sex and, as such,
must be approached from an inherently archaeological perspective. It is also important to
recognize that social institutions such as kinship, age, and ethnicity intersect with gender to
create one’s social identity (Gellar 2005, Sofaer 2006). In other words, although gender is one of
the primary foci of this dissertation, it cannot be investigated in isolation from intimately linked
cultural institutions. I argue that explicit incorporation of multiple social variables will result in a
more appropriately gendered interpretation. For example, bioarchaeological analyses that attempt
to account for the ways that experiences vary within groups based on biological sex based on the
inclusion of social institutions units of analysis like status (Martin 2000, Cucina and Tiesler
2003, White 2005) and ethnicity (Schurr 1992) have greater potential for clearly assessing
gender relations in the past. The research proposed here seeks to incorporate data for three other
social factors, which provide further cultural context for interpreting osteological data. These
factors are social status, kinship, and age.
That subadult skeletons cannot be sexed with a high degree of confidence has proven to
be a serious hindrance to the development of an engendered bioarchaeology of childhood (Baker
et al 2005). However, the skeletal remains of adults do retain information on certain aspects of
childhood growth, stress, and health, and analysis of indicators that capture conditions at known
points in an individual’s life may help to alleviate this problem by providing observable records
of childhood growth conditions in sexed adult remains. Specific types of skeletal data which
indicate the conditions of early childhood growth were collected and analyzed as part of this
study, in addition to cumulative indicators of growth over the life course (these indicators will be

10

discussed in more detail in the methods chapter). The combination of age sensitive indicators of
stress with spatial/artifactual data pertaining to kinship and social differentiation will allow for a
more nuanced interpretation of the relationship between gendered social relations, social
complexity, and growth/health in late prehistoric West-Central Illinois.

The Case for Mortuary Analysis and the Saxe-Goldstein Approach
Goldstein (2006) examines recent literature to specifically evaluate the level of cooperation and
collaboration between physical anthropologists and archeologists in the field of mortuary
analysis. Like Gellar (2005, 2008), she finds that there is significant disconnect between skeletal
analysis and the interpretation of archaeological data. Physical anthropologists studying remains
from archaeological contexts focus on topics such as bone chemistry, pathology, stress, growth
disruption, and paleodemography, which require laboratory-oriented approaches.

Intense

involvement in this type of analysis comes with the risk of further distancing oneself from the
archaeological context of the remains. This issue becomes increasingly important as modern
mortuary analysis continues to incorporate nuanced social, economic, political, and ideological
theory in the interpretation of living people and their treatment of the dead.

Literal and

figurative disconnect between the two fields can result in interpretive problems, as demonstrated
in the issue of sex and gender, as well as more specific critiques (Bice 2003 and Jurmain 1997,
for example) regarding the failure of physical anthropologists to rigorously test methods
borrowed from the natural sciences before applying them to human remains.
Binford (1971) provides an important treatise on the history of mortuary analysis, and
called for a new, processual approach to mortuary studies. He contends that descriptive accounts
abound in archaeology, but points to a lack of literature attempting to use burials as a "distinct

11

class of variable phenomena" (Binford 1971, 58). Early studies of mortuary analysis were
attached to general studies of religion and employed an “idealist-rationalist” approach that
correlate practices with postulated or observed forms of belief. Binford points out, however, that
as early as Hertz (1960), it was recognized that burial rites were related to, and vary with, social
institutions other than religious ritual, such as the status of dead within living community and the
status of souls in relation to the society. He also summarizes early 20th century functionalist
perspectives on mortuary rites as a collective (social) reaction against death as an attack on
social solidarity and goes on to argue that, while these approaches attempt to link mortuary
practice to social circumstances, they focus on abstract notions rather than studying the
distribution of variation within or among sociocultural units.
Binford calls for the development of theory to provide a context for explanation of
differences and similarities in mortuary practices, based particularly on status and group
affiliation. He used the Human Relation Area Files to argue that, 1) a high degree of
isomorphism exists between complexity of status structure and complexity of mortuary
ceremonialism, 2) mortuary differentiation is based on personal qualities (age, sex, differential
personal capacity) in “minimal complexity” societies, while in complex societies, differentiation
is based on abstract notions of social divisions, and 3) a correlation exists between quantitative
and qualitative dimensions of social persona (as recognized via mortuary custom) and general
level of sociocultural complexity. All of this is in keeping with the neoevolutionary perspective
that defined his views of culture and processualist archaeology.
Saxe’s (1970) perspective on anthropology as an explanatory tool for addressing
similarities and differences within and between sociocultural systems complements Binford’s
generalization that the form and structure, which characterize mortuary practice in a given

12

society, are conditioned by form and organizational complexity of the society itself. In an
influential dissertation, Saxe proposes eight hypotheses with the goal of explaining “the formal
disposal types found in any sociocultural system and, by doing so, shows how the domain
formed by these types can elucidate the social organization that produced them (1970, 3).” His
original hypotheses have notably been modified and applied to archaeological situations by
Goldstein (1976, 1980, 1981) in her work at the Schild site in West Central Illinois (one of the
sites studied in this dissertation). The complementary nature of Binford’s, Saxe’s, and
Goldstein’s work have led to their ideas being referred to together as the “Saxe-Binford”(Brown
1995) or “Saxe-Goldstein” (Morris 1991) hypotheses.
Of particular interest is Saxe’s eighth hypothesis, which states that “to the degree
corporate group rights to use and/or control crucial but restricted resources are attained and/or
legitimized by means of lineal descent from the dead (i.e. lineal ties to the ancestors), such
groups will maintain formal disposal areas for the exclusive disposal of their dead, and
conversely” (Saxe 1970). This approach has been criticized by Hodder (1981, 1982) and Shanks
and Tilley (1982) as ahistorical, simplistic, and passive. These criticisms have themselves been
duly criticized by Brown (1995) who points out that Goldstein’s (1980) reformulation of
hypothesis eight, linking mortuary ritual, formal structure of cemetery spaces, and corporate
rights of lineal groups, has improved the original formulation. This ties directly into the idea of
the dead being used/manipulated to serve a specific end within the world of the living. Morris
(1991) also defends the approach as useful in both historical and empirical/analytical
frameworks, as it is inherently concerned with the interaction between cemeteries/mortuary ritual
and socially significant institutions such as the household, class, and kinship. It is argued here
that 1) the Saxe-Binford-Goldstein approach to mortuary analysis remains relevant and is

13

adaptable and amenable to application within various theoretical frameworks, whether more
processually or post-processually oriented, and 2) Goldstein’s (1980) spatial analysis of the
Schild cemetery provides an appropriate sociocultural framework upon which to build a
bioarchaeological analysis of intracommunity social differentiation.

Synthesizing Biological and Cultural Perspectives
The above summaries of theoretically-oriented bioarchaeological and mortuary literature are far
from comprehensive. However, they provide a general foundation for the study undertaken for
this dissertation. Despite the apprehension of many bioarchaeologists to venture into the realm of
social theory, modern gender/social identity scholarship can productively inform a scientifically
rigorous biological analysis. It is further argued that, not only is social theory amenable to
bioarchaeological research, it is an absolutely necessary component.
Furthermore, an in-depth understanding of mortuary analytical theory is necessary to
move from stagnant description in bioarchaeology to culturally informed interpretation of the
relationship between health and social processes in the past. The Saxe-Binford-Goldstein
approach was developed decades ago, yet it has been shown that its basic tenet- the relationship
between treatment of the dead and social factors/concerns of the living- provides a sound
framework for a range of archaeological studies and is adaptable to diverse cultural, theoretical,
temporal, and regional situations and perspectives. The explicit addition of a biological
perspective can only serve to complement the strength of this basic model. One goal of this
dissertation is to be rooted firmly within a foundation of theoretically rigorous biocultural
analysis and a time-tested, adaptable mortuary paradigm. By focusing on the theoretical
complexities of social identity and treating archaeological data as primary rather than accessory,

14

this study attempts to contribute to a dynamic picture of what it meant to be a Mississippian
person in West-Central Illinois

Engendering Bioarchaeology
It is the express purpose of this dissertation to contribute to archaeological knowledge of gender
as lived experience in Mississippian communities, yet the preceding discussion emphasizes that
applying gender theory to interpretations of the archaeological past is inherently complicated.
Seeing gender through bioarchaeology is context-dependent and, therefore, not predisposed to
formulaic treatment. In other words, there is no singular “correct” way to construct a properly
engendered bioarchaeological analysis. Practitioners interested in exploring some facet of gender
in the past must consider the available archaeological, historical, and ethnographic data related to
their collections in order to design a research strategy that meaningfully engenders their sex
based osteological data.
One possible approach (and the approach that is taken in the current study) is to dispose
of the essentialist view that the experiences of all men and all women are unitary and invariable.
This is a generally accepted perspective when examining gender via cross-cultural comparison,
but it should be recognized that there is also a range of variation in the experiences of each
gender within communities, based on other aspects of social identity like kinship, status, and age
(Gellar 2008, Sofaer 2006, Stig Sorenson 2000). By considering variability in osteological
parameters within each sex, bioarchaeological notions of biological maleness and biological
femaleness is problematized, de-essentialized, and become prepared to be carefully
contextualized.

15

This approach necessitates that skeletal samples meet two important criteria, both of
which are inherently problematic for bioarchaeological studies. First, culturally meaningful units
of analysis are necessary to provide a justifiable basis for dividing male and female samples into
subsamples for within-sex comparison. In the current study, intensive mortuary analysis, based
on spatial and artifactual considerations, is available to provide the necessary sociocultural units
(Goldstein 1980). Because these mortuary groupings have been interpreted as representing
kinship and status differentiation, bioarchaeologists can explore the ways that variation within
each sex each is or is not related to membership in these social categories. In this way, we can
see how the biocultural experiences of men and women are variably related to other aspects of
their social identity. Second, sample sizes must be adequate to test for significant differences in
the osteological data. Sample size is a problem faced by most bioarchaeologists, and the issue
becomes even more pressing when the sample needs to be further subdivided as described above.
The current study benefited from the relatively large size of the Schild collection, and although
sample sizes were still small for many tests, some statistically significant results were indicated.
Based on the above discussion, it is proposed that the current study represents a
theoretically and methodologically sound attempt at a gendered perspective on biocultural
change and variability because the research design includes multiple social units of analysis for
comparison of growth trends within each sex. Division of a skeletal sample into “male” and
“female” subsets is standard practice in bioarchaeological studies, and one can tack a gendered
interpretation onto the results of their binary data analysis. The current study, however, treats the
possibility of within-sex variability as an integral part of the research design and an important
way to detect variation in the ways that people experience gender, thereby avoiding, to some
extent, the trap of essentialism.

16

Research Questions
The theoretical arguments provided in this chapter provide a framework for developing several
general hypotheses, which are presented here along with summarized results and interpretations.
Statistical treatment of the Schild and Yokem long bone, dental, and vertebral data evaluated
temporally and by sex provides information about whether or not a change in sex-specific
patterns occurred, as well as the direction and the significance of the change. As age is an
important factor in gender identity, special attention to the age-sensitive nature of the indicators
allowed investigation of changes in growth patterns by sex over the life course. Furthermore,
when mortuary data is integrated into the analysis, the nature of the social relations underlying
the appearance of the biological data became clearer.

Research Question 1Does Clark’s (1985, 1988) vertebral approach provide a sound, useful method to growth
assessment? One of the primary concerns of this research is a determination of the utility of the
vertebral method of growth assessment. This was tested through an assessment of the
relationship of vertebral dimensions to widely accepted growth indicators, as well as by making
comparisons of the results achieved through vertebral growth assessment with expectation based
on other lines of archaeological evidence. Taken together, the various lines of research presented
in this dissertation indicate that, although more detailed investigation remains to be done, the
vertebral method is, indeed, a potentially useful, and particularly sensitive, indicator of growth
conditions. In particular, neural canal diameters (NCD) may reflect the growth conditions
experienced during early childhood in a way that is fundamentally different from, and

17

complementary to, the episodic stressors detected by analyses of LEH. Vertebral body heights
(VBH) seem to be related to the same kind of cumulative growth experienced through early
adulthood that is indicated by femur length.

Research Question 2Are there changes in growth patterns between males and females over the Late WoodlandMississippian transition in West Central Illinois? If peripheral communities experienced changes
not only to their traditional gender ideologies, but also to their traditional gender relations under
Cahokian influence, changes in the patterns of differential growth, stress, and health between
males and females from the Late Woodland to the Mississippian period at Schild and Yokem
may be observable as quantitative and qualitative changes in biological parameters. Based on
trends in linear enamel hypoplasias over the Late-Woodland/Mississippian transition recognized
by Cohen and Bennett (1993), one might expect that females in this study would experience
decreasing levels of relative early childhood stress with Mississippianization, concomitant with
the increasing importance of maize agriculture. According to observations by Cook (2007),
previous analyses have shown a surprising lack of significant changes in long bone length with
Mississippianization. This could potentially be clarified by the incorporation of Clark’s (1988)
vertebral method, perhaps indicating that long bone length, as a cumulative indicator, might
mask age-specific growth differentials. Age-specificity in the osteological data may show that
the relative stress experiences of male and female children changed at some point over the course
of childhood, thereby suggesting that changes occurred in the social perceptions of male and
female children. This finding could provide clues to the timing of the social construction of
gender in Late Woodland versus Mississippian times.

18

Conversely, sex-based patterns of growth and stress might appear either unchanged or
only minimally changed in the transition from the Late Woodland to the Mississippian period at
Schild or Yokem. Lack of change would support the position that Late Woodland communities
in West Central Illinois adopted Mississippian burial patterns and material culture while resisting
some social aspects of Cahokian change, essentially preserving most features of their traditional
gendered social relations. An alternative interpretation might be that Mississippianization did not
bring significant changes to the lived experiences of men and women, regardless of the extent to
which normative ideas about men and women were transformed under Cahokian influence.
This dissertation research indicates that, on a temporal level, and indicated by
significantly increased neural canal diameters, female early childhood growth experiences
generally improved over the Late Woodland/Mississippian transition in West Central Illinois,.
There were no significant differences in vertebral body heights among these females, indicating
that improved biosocial conditions of growth in the adolescent period may have allowed Late
Woodland females to achieve a similar level of cumulative growth as their Mississippian
counterparts, masking early childhood growth differentials. Male growth conditions remained
relatively stable over the life course and over the cultural transition, in both early childhood and
adolescent growth.
The results of the vertebral analysis is generally consistent with trends observed in
comparable measures of skeletal growth at different life stages, namely previously published
LEH data (as an indicator of early growth) and femur length data from this analysis (as an
indicator of cumulative/late growth), lending support to the vertebral method as a reasonable
measure of prehistoric growth patterns. Furthermore, this general trend toward improved early
childhood growth conditions for females over the Late Woodland-Mississippian transition in

19

West Central Illinois suggests that, not only did Mississippianization introduce novel forms of
material culture and mortuary patterning to the region, but also brought with it real and
significant changes in social perceptions of male and female children in these communities.

Research Question 3Is variation in intracemetery growth patterning amonf in Mississippian males and females at
Schild related to other social parameters such as kinship, community membership, and status?
Explicit incorporation of biological and archaeological data is useful for exploring the way that
social identity, as observed as the interaction of sex, kinship, and social differentiation, is related
to stress, growth, and dietary experiences. Incorporation of Mississippian mortuary data allowed
an evaluation of the ways that social factors (e.g. kinship and status differentiation) interact with
sex to affect general health in the Schild community. Taken together, these data may indicate
that the changes leading to the adoption of Mississippian mortuary ritual and material culture
might also have brought concomitant changes in gendered social relationships affecting diet and
health, with the extent and the direction of the change perhaps offering clues to the degree and
scope of gendered social transformations. Gender relations therefore are more clearly observable
through a combined bioarchaeological and mortuary approach.
The current analysis indicates that a research orientation that emphasizes combinations of
social factors provides a more culturally sensitive account of the variation in growth data than
can be accomplished with sex alone. Although cumulative growth as indicated by femur length
and vertebral body height remained stable over the Late Woodland-Mississippian transition for
females and males in general, there is some evidence that relatively favorable cumulative growth
conditions were not enjoyed equally by all segments of the Schild Mississippian community.

20

Specifically, vertebral body height and femoral length analyses indicated that cumulative growth
conditions were significantly better for specific spatial subsets of the Schild female burial sample
compared to other spatial subsets. There was, however, no concomitant differentiation in growth
patterning between spatial groups among Schild males. Based on Goldstein’s proposition that
these spatial groupings may be indicative of kinship or corporate descent groups, this research
suggests that Schild women may have experienced more variability in adolescent life/growth
experiences than Schild men, based not only on biological sex, but also on their roles as part of
socially meaningful kinship or community-based networks. This may be related to particularly
strong productive or managerial skills by certain women within the larger Schild community,
leading to insipient social inequality and improved biosocial conditions of growth for themselves
and their daughters, with whom they likely to work closely. Alternately, the spatial subset of the
Schild Mississippian community in which the smallest females were interred may have
represented the earliest people at the site who could be said to have been living a Mississippian
lifestyle. In this case, growth variability among Schild females may reflect the gendered,
biocultural effects of rapid acculturation and/or insipient ethnogenesis.

Research Question 4How do the results of the current study compare with other studies of the relationship between
Mississippianization, community health, growth patterns, and gender? Comparison with regional
reports of the transition to a Mississippian lifestyle generally seem to indicate that the biosocial
conditions of female early childhood growth tended to improve relative to males over time
(Goodman et al. 1980, Cook 1984, Cohen and Bennett 1993). This evidence suggests that male
children were afforded relatively less “protection” (or female children were afforded relatively

21

more protection) from nutritional and pathological insult in the Mississippian period compared to
the earlier periods, possibly reflecting changes in the perceived value of women’s work with the
rise of agriculture (based on ethnographic and iconographic evidence linking females to
agriculture, i.e. Watson and Kennedy 1991). Cook (2007) also points out, however, that the lack
of change in long bone lengths over this transition is surprising given the assumption that the
transition to a maize heavy diet would likely contribute to poorer cumulative growth.
The results of the current analysis are consistent with interpretations suggesting regional
trends of improving conditions of female early childhood growth with Mississippianization, as
indicated by significant increases in female lumbar anterior-posterior neural canal diameter
(APNCD) without parallel relative increases among males. These results are further reflected and
supported when lumbar APNCD and transverse neural canal diameter (TNCD) are compared
between males and females within each cultural period, indicating that changes in sexual
dimorphism, within each time period, favor statistically significant improvement in female early
childhood growth. Furthermore, my analysis confirms Cook’s findings regarding lack of
significant regional change in femur length over the Late Woodland-Mississippian transition.
While Cook points out that this is interesting from a nutritional perspective, the lack of temporal
change in cumulative growth is also interesting in that it 1) appears to affect males and females
equally on a community level, 2) it appears to mask sex-based early childhood growth
differentials shown to occur via other skeletal methods, and 3) may be masking intracommunity
differentiation in growth patterning.

22

Research Question 5What do the results of the current investigation contribute to a broader understanding of the
nature of Mississippianization? Several archaeological studies of the concepts of tradition,
resistance, emulation, acculturation, and ethnogenesis as they pertain to the late prehistoric
Midwest provide a critical backdrop for one of the goals of this research, which is to explore the
extent to which the adoption of archaeologically visible Mississippian cultural practices,
particularly markers of social differentiation indicated by mortuary patterning, is related to the
lived biosocial experiences of late prehistoric Lower Illinois River Valley inhabitants. Several
archaeological studies have provided evidence for the active maintenance of some aspects of
traditional (i.e., rooted in Late Woodland) social relations at otherwise Mississippian peripheral
sites (Oteleer 1993, Alt 2002), while others have argued for either active resistance to
Mississippianization (Emerson 1997 in the Upper Illinois River Valley) or limited adoption of
only certain Mississippian traits (Farnsworth et al. 1991 in the Lower Illinois River Valley).
This dissertation contributes a biocultural perspective to a discussion of the extent of
influence of centers over peripheral peoples and the ability of peripheral peoples to selectively
incorporate or resist certain aspects of the prevailing cultural. The biological evidence presented
in the following chapters, along with previous bioarchaeological analyses, indicates important
health transitions associated with Mississippianization in the region. This research supports this
interpretation and goes a step further by contextualizing biological change on a finer social scale,
showing that certain patterns of health and growth are strongly correlated with specific types of
Mississippian intracommunity social differentiation (as indicated by mortuary analysis). Results
of this analysis support the interpretation that the adoption of Mississippian sociocultural traits in
the Lower Illinois River Valley was not an act of simple, superficial emulation, but were

23

reflective of demonstrable biosocial effects of Mississippianization on the lived experiences of
the late prehistoric inhabitants of this region.

Organization
The chapters that follow will situate and elaborate on the arguments made above in more detail.
Chapter two provides background information regarding the Late Woodland and Mississippian
traditions, and reviews the literature on sex and gender in the late prehistoric Midwest, while
chapter three presents the methods of growth assessment used in this research. Chapter four
describes in detail the Schild and Yokem sites and their associated skeletal samples in more
detail, and reviews previous bioarchaeological research based on these sample populations.
Chapter five describes the osteological measurement techniques, mortuary data, statistical
methods, comparative data sets, and hypotheses employed in this research. Chapter six presents
the results of the statistical analyses, while chapter seven situates and interprets these results
within the proposed theoretical and cultural framework. This dissertation concludes with a
summary of the current study and proposes future research questions and directions.

24

CHAPTER TWO- REGIONAL AND BIOCULTURAL BACKGROUND

The bulk of this dissertation focuses on two distinct but related components. The first is an
assessment of community wide changes in human growth patterns between the Late Woodland
and Mississippian periods in West Central Illinois. The second is a biosocial investigation of
intracommunity differentiation in the Mississippian component of the Schild cemetery. This
chapter provides the archaeological and bioarchaeological background necessary for a
contextualized interpretation of the results of this study.

Geographic Setting
The vast majority of the previous studies discussed in this dissertation analyzed populations from
what is now the Midwest United States, with a particular focus on Illinois. West Central Illinois
and the American Bottom are important regions for the investigation of the development of
Mississippian culture and its relationship to the preceding Late Woodland period. The American
Bottom is the wide floodplain at the confluence of the Mississippi and Missouri Rivers located
near modern day St. Louis, Missouri (Figure 1).

The Late Woodland Tradition
Late Woodland is a term referring to both a time period and a cultural tradition identified
throughout the Eastern United States, dating generally from AD 400 to 1000. Late Woodland
societies were tribal level and semi-sedentary They practiced a mixed economy, and settled near
rivers and lake margins (Schroeder 2004, McElrath et al 2000). Archaeological data suggests
that, although many Late Woodland groups eventually transitioned to Middle and Upper

25

Mississippian ways of life, others practiced a Late Woodland lifestyle until the time of European
contact.

Figure 1- The American Bottom (From Milner 1983)
26

Late Woodland culture in the midcontinent has been generally characterized by several
important transformations, including increased regional variability in settlement patterns,
incorporation of bow and arrow technology, and intensification of horticultural practices
(MacElrath et al 2000). McElrath et al (2000) suggest that the early Late Woodland is marked by
increasing regionalism when compared to the far-reaching, long distance relationships that
characterized Hopewell Interaction Sphere. Similarities in ceramic typologies between sites are
explained by rapid resettlement into new upland areas (out of river valleys) by segments of
established communities rather than extensive trade and interaction between groups. In the
American Bottom, the development of bow and arrow technology around AD 700 coincided with
the reoccupation of the floodplain and the appearance of patterned, nucleated villages.
Rather than being investigated as an expression of a unique cultural tradition, Late
Woodland social organization and regional integration have often been characterized based on
the absence of highly visible attributes (such as elaborate mortuary ritual) that were a hallmark of
their Hopewell predecessors and Mississippian successors (McElrath et al 2000; Charles 1992).
Increasing regionalism is interpreted to have resulted from the breakdown of elite controlled long
distance exchange. Decreasing hierarchical social differentiation is evidenced through the
simplification of burial ceremonialism. These observations are often cited as evidence for a lack
of social complexity and interaction in the Late Woodland Midwest and Southeast (Cobb and
Nassaney 1995, Braun and Plog 1983). However, Charles (1992) points out that, during the Late
Woodland, new settlement and mortuary patterns indicate a stabilization and localization of the
widespread and rapidly changing social and demographic processes that occurred during the
Middle Woodland. This set the stage for the eventual development of maize agriculture and,
hence, Mississippian social complexity in the American Bottom and West Central Illinois.

27

Some researchers have suggested that our perceptions of decreasing social complexity
and interaction during the Late Woodland are due to application of analytical techniques and
interpretative frameworks that are not appropriate for the investigation of a society in which
hierarchical relationships are not emphasized. Braun and Plog (1982) suggest that decreasing
exchange of supralocal goods, often indicated as evidence for decreasing intercommunity
interaction, might actually indicate shift over time in the nature of the interaction from more to
less social distance between participating communities. Cobb and Nassany (1995) argue that the
lack of hierarchical, vertically-differentiated sociopolitical relations within and between Late
Woodland communities does not preclude the existence of complex heterarchical, or
horizontally-differentiated, interactions. This should be studied as a unique cultural phenomenon,
rather than solely in comparison to other cultural periods. They cite evidence for the movement
of utilitarian goods and the rapid diffusion of shell-tempering, maize agriculture, and bow and
arrow technology during this period to indicate that regional integration and social complexity
existed despite outward appearances of the archaeological record. Although there may have been
no institutionalized positions of power to facilitate such interactions, the potential of lineages,
clans, age groups, and gender groups to have played important sociopolitical roles should not be
ignored. It is important to note that it may be more difficult to recognize these types of
relationships without application of a more nuanced mortuary analytical framework (Braun and
Plog 1982).
McElrath et al (2000) provide a limited and generalized review of Late Woodland
mortuary practices in the midcontinent, suggesting that the homogenous, localized expression of
Late Woodland social structure is indicated by variable mortuary practices. The traditional
Middle Woodland burial mound, with a centrally located log tomb, was replaced by new Late

28

Woodland interment styles such as stone lined/paved graves and burials in stone piles within
earthen mounds. Aboveground stone mortuary facilities, crematories, and accretional mounds
were also built, but patterns varied both regionally and temporally throughout the region.
Late Woodland mortuary practices are relatively well documented (but perhaps underanalyzed) for the Lower Illinois River Valley. Although accretional mounds with flexed burials
are common, characterization all Late Woodland mortuary patterning in this matter masks
intraregional variability. Previous studies of biological distance have demonstrated significant
intercommunity gene flow throughout the Late Woodland Lower Illinois River Valley compared
to Mississippian times (Steadman 2001), suggesting that people were moving between
communities, which is perhaps related to the ubiquity of some features of the burial program in
the area (Conner 1984). Charles (1992) postulates that population saturation in the Lower Illinois
Valley in the Early Late Woodland period would have provided a fertile environment for the
proliferation of kinship and marriage alliances, affording new opportunities for exchange
partners without management, manipulation, or interference on the part of a pan-local elite.
Charles contends that this could account for the variability of burial practices during the Late
Woodland period, as funerary programs became local affairs with local meanings. Following the
critiques of Braun and Plog (1982) and Cobb and Nassany (1995) discussed above, further
investigation of the scenario described by Charles (1992) would appear to require an
individualized analytical and interpretive framework for deciphering Late Woodland mortuary
practices and social relationships on a local or subregional level before making meaningful
regional characterizations.
As in the rest of the region, Late Woodland society in the American Bottom and West
Central Illinois appears to be horticultural, village-oriented, and egalitarian (Schroeder 2004,

29

MacElrath et al 2000). In the early Late Woodland, there is little evidence to indicate significant
social differentiation within villages, nor does there appear to be hierarchical political
relationships between villages (Mehrer 2000). It is during the terminal Late Woodland in the
American Bottom region that archaeologists first detect important cultural trends that would later
be intensified, forming the basis for Mississippian emergence. Mississippian characteristics and
precursors such as courtyard groups, plaza-mound site plans, population nucleation, insipient
social differentiation, and maize intensification appear in the archaeological record between AD
700 and 900 (MacElrath et al 2000, Mehrer 2000, Schroeder 2004).

The Mississippian Tradition
Late Woodland populations in the American Bottom appear to have undergone dramatic cultural
changes around AD 900, eventually leading to the development of the Mississippian tradition.
Although Mississippianization is seen early and most intensely in the American Bottom region,
the suite of associated cultural changes were likely the result of widespread interactions and in
situ development among different communities following the same basic cultural trajectory
(Schroeder 2004). The transition is marked by major changes in 1) ceramic temper and
decorative motifs, 2) development of a ranked social structure and hierarchically political
organization with ascribed status differentiation, 3) the ubiquity of cleared-field maize
agriculture, 4) and a set of common religious institutions and iconography (Schroeder 2004,
Scarry 1996). Throughout the region, rural occupations began to decrease in size as people
congregated in larger villages and small nodal centers and, by the eleventh century AD, villages
began to rely more heavily on maize agriculture to support growing populations. Middle

30

Mississippian chiefdoms developed internal social differentiation and hierarchical relationships
with surrounding villages (Emerson 2002, Schroeder 2004).
Mississippian villages have been identified throughout the area spanning from northern
Florida to Illinois and from the Atlantic plain to Eastern Oklahoma. Most notably for this
research, Cahokia emerged, peaked, and declined as a major regional center in the Midwest
between AD 900-1400, undoubtedly having a considerable impact within the American Bottom
and beyond (Fowler 1991, Milner 1998, Pauketat 1997). The Cahokia site covered almost 13
square miles and comprised 104 mounds, a palisaded downtown, and multiple mound precincts
and neighborhoods (Fowler 1989). Mounds and structures were clearly layed out in a planned,
organized way, which clearly reflected Mississippian social, poiltical, and cosmological
worldviews (Fowler 1978).
The spread of Mississippianism out of the American Bottom was instigated by diffusion,
migration, and economic/political activity or some combination of these factors, depending on
the specific time period and location in question (Stoltman 1991, Farnsworth et al 1991). The
nature of Cahokia’s influence in West Central Illinois, in particular, has important implications
for this study.

Mississippian Social Structure and Site Hierarchy
The establishment of sites of varying size and complexity, ranging from large centers, to lesser
centers, to rural farmsteads was a hallmark of Mississippian culture (Emerson 1997, Goldstein
1980). Middle Mississippian social complexity is evident through the integration of permanent
towns, hamlets, and farmsteads into a site hierarchy, and Goldstein (1980) shows that variation
in mortuary programs is patterned and closely associated with this hierarchy. Specifically, she

31

proposed several hypotheses for investigating Mississippian social organization through
mortuary analysis. First, she hypothesized that social organization at outlying sites should be
relatively egalitarian and that bounded cemetery areas are likely indicative of corporate group
structure with lineal descent. Second, she suggests that spatial organization within cemeteries
should reflect the corporate structure of the society. Her third hypothesis is that rural
Mississippian mortuary sites should exhibit a communal, as opposed to an individualized,
emphasis. All of these are supported by her analysis of the Mississippian component of the
Schild site (described in detail below).
Goldstein also hypothesizes that Mississippian mortuary sites should generally reflect a
common set of organizational principles. Although mortuary treatments at Schild and Moss are
certainly different from those observed at Cahokia and other large centers, Goldstein shows that
the level of social complexity within these small, outlying communities is in keeping with
general expectations of a ranked and kin-based Mississippian social system. Larger sites tend to
incorporate the basic elements observed at smaller sites, particularly rows/spatial clustering,
communal emphasis and charnel features, albeit with a measure of elaboration. The model that
she proposes has been instrumental in organizing subsequent analyses (both synthetic and
localized) and structuring ideas regarding Mississippian perceptions of mortuary space and social
organization.

Southeast Ceremonial Complex Through Time and Space
Although the focus of this dissertation is the lived, gendered experience of Mississippianism and
Mississippianization in West Central Illinois, a discussion of analogous processes and
phenomena in other parts of the Midwest and Southeast is warranted as a means for preliminary

32

consideration of the ways that the topics of the current research might be similarly or
differentially expressed on a regional level. Mississippian as a concept is more diverse than any
sweeping, cursory definition (including the one provided above) can properly convey, and this
requires that a critical perspective is employed when speaking of such things as “Mississippian
gender relations.” But although regional variation in the expression of Mississippian gender
relations is high, it can also be argued that regional expressions are rooted in common
ideological themes that have their basis in a generalized Mississippian worldview. One way that
this worldview has been conceptualized in the Mississippian literature is through studies of the
Southeastern Ceremonial Complex (SECC).
The SECC is a diverse set of ritually and spiritually charged themes and images
recognized and practiced throughout the Mississippian world. An important attribute of
Mississippian societies is the use of rituals and symbols associated with the Southeastern
Ceremonial Complex, presumably indicating participation in a characteristic Mississippian belief
system that expressed a particular social, political, and religious ideology. Although many
researchers have insisted that the SECC is a late Mississippian phenomenon restricted to the
Southeastern United States, Brown and Kelly (2000) show that many of the most enduring
symbols and items (i.e. long nosed god masks, copper plates, birdman motifs, chunkey player
motifs) have their origins in early Cahokia. They conclude that what has classically been
perceived as a change in the type of ceremonial complex being practiced in earlier versus later
Mississippian times is actually an evolution in the practices associated with a single set of
beliefs. They also point out the tripartite division of the complex into distinct complexes,
centering on warfare/cosmogony, mounds, and temple statuary. Awareness of these distinctions,
in addition to regional and temporal variability, is crucial.

33

Although primarily investigated for its significance in religious life, Brown (1989) points
out that SECC studies can also demonstrate how other aspects of Mississippian social life such
as economy and social organization, are simultaneously expressed and unified through the
complex. It represents an important unifying theme for a cultural tradition that is not only
spatially but also temporally expansive, and this provides a justification for the types of regional
comparisons that will be made in the interpretation of the data presented in this dissertation. The
evolution of the various schools or horizons of SECC symbolism can be seen on a broad regional
level as developing along parallel trajectories out of Late Woodland roots (Brown 2004, Muller
1989). In these ways, the SECC concept is recognized as 1) connecting Mississippian peoples
across space and time on both socioeconomic and ideological levels, and 2) linking them to Late
Woodland roots.

The Nature of Center-Periphery Interactions and Influence
The nature of interaction between Cahokia and its periphery is not fully understood or agreed
upon (Anderson 1997, Emerson 1997, Milner 1998). Farnsworth et al (1991) and Stoltman
(1991) summarize the range of theoretical perspectives regarding the extent of Cahokia’s
ideological, social and economic influences. Despite Cahokia’s importance on a regional level,
however, generalized statements regarding broad Cahokian influences can only superficially
approach the intricacies of social relations on a local level within peripheral villages. To better
explain variation in social relations on a local level, Mississippianization of the periphery must
be conceptualized as an active process of negotiation between traditional practices and new
Cahokian ideals (Pauketat 2002).

34

Alt (2002) comments on the importance of investigating Mississippianization within the
context of community identity, social interaction, and political change. Through an analysis of
site layout, ceramic production, and house construction, she suggests that, although upland
American Bottom villages often adopted Cahokian styles (as in ceramic technology), lack of
change in practices more indicative of intra-village social relations (i.e., site layout and house
construction), point to more traditional social structures being actively maintained. This suggests
that some nearby Mississippianized communities resisted certain Cahokian social ideologies into
the period of perceived Cahokian dominance.
Harn (1991) investigates Middle Mississippian emigration to the Central Illinois River
Valley region from the American Bottom, proposing that Mississippians essentially replaced and
acculturated existing Late Woodland communities, as evidenced through sudden changes in
social organization, indicated by mortuary analysis, and the sudden appearance of new
religiously- and politically-charged motifs. Based on their review of wall trench structures in
Lower Illinois River Valley, Farnsworth et al (1991) and Farnsworth and Atwell (1999) suggest
that Late Woodland communities in the Lower Illinois River Valley incorporated Mississippian
mortuary ritual while maintaining a predominantly Late Woodland lifestyle.
In contrast, others make strong arguments for true Mississippianization of the region.
Goldstein (1980) contends that the Lower Illinois Valley was occupied by dispersed
Mississippian farmsteads, as indicated by the presence of Mississippian pottery, mortuary ritual,
and architectural styles. Delaney-Rivera (2004) evaluates ceramics, excavation notes, and maps
to suggest that 1) Late Woodland (Jersey Bluff phase) groups from the Lower Illinois Valley
interacted intensely with early Mississippian groups in the American Bottom and 2) proposes a
colony-acculturation model for the spread of Mississippianism into this region. Although

35

investigating somewhat different phenomena (i.e. in situ Mississippianization of Late Woodland
communities versus intrusive Mississippian replacement of Late Woodland communities versus
a combination of acculturation and colonization), a key issue uniting all of the studies discussed
above is a concern with the extent of the economic, social, and ideological influence between
Cahokia and peripheral communities. The research presented in this dissertation will build upon
these investigations by gauging the biocultural affects of Mississippianization in West Central
Illinois, particularly with regard to building a baseline set of data for assessing its affect on
traditional gender relations.

Late Woodland/Mississippian Interactions and the Mississippianization of the Periphery
Models for the spread of Mississippian cultural practices out of the American Bottom, following
the “Cahokian Big Bang” around AD 1050 (Pauketat 1998), are numerous and theoretically
diverse. The initial development of Mississippianism is portrayed by some as having resulted
from widespread interactions and evolution among different communities that generally followed
the same basic cultural trajectory (Schroeder 2004). Kelly (1991) characterizes this transition
more specifically, as having resulted from interactions and exchanges between northern and
central American Bottom people with others in the surrounding region. In this conception, the
American Bottom occupies a prime central location, but is lacking in raw materials. Population
growth during the Late Woodland and Emergent Mississippian periods increased the demand for
the importation of such goods, contributing to the rise of Cahokia as a social, political, and
religious center (Kelly 1991).
The economic undertones in Kelly’s model of the emergence of Mississippian culture is
also seen in models for the nature of the relationship between Cahokia and the outlying regions.

36

Peregrine (1995) provides portrays Cahokia as the center of a Mississippian “World System.”
However, other researchers have criticized application of this model to situations other than the
capitalist and market-oriented economies it was originally developed to explain. For example,
Milner (1991) criticizes all exchange-oriented models for Cahokia’s regional importance and
external relations. He points out that pre-Mississippian long-distance exchange occurs in the
absence of social complexity through much of the prehistory of the midcontinent. These
exchanges certainly continued to be important into the Mississippian period, but there is little
evidence that Cahokia was the center of a regional exchange network, especially considering that
most Cahokian goods appear to have been manufactured locally, using local raw materials. Even
the few examples of seemingly abundant extralocal raw materials in the American Bottom, such
as Mill Creek chert hoes from southern Illinois, are not particularly abundant at Cahokia
compared to smaller Mississippian sites when their prevalence is corrected for population. Both
Milner (1991) and Mehrer (2000) suggest that political decisions at outlying sites were made by
local kin-based elite, based on local concerns rather than with reference to elites at larger centers.
Mehrer (2000) however, takes a somewhat more nuanced view of this relationship through an
evaluation of site layout, suggesting that community-planning decisions reflected heterarchical
relationships between regional leaders, local elites, and individual households.
Outside of the American Bottom, Milner (1998) notes the widespread distribution of
Cahokian and Cahokian-style pottery. Aside from distinct forms like Ramey Incised and Powell
Plain, he sees no need to consider every instance of cordmarked or shell tempered pottery as
evidence for Cahokian exchange. He also notes that the decline of Cahokia during Moorhead and
Sand Prairie times did not adversely affect the ability of most communities to continue to access
exotic goods, suggesting that Cahokia’s role in trade networks was minimal. Similarly, Mehrer

37

(2000) notes that the general plan of rural Mississippian sites, which certainly developed under
the influence of Cahokia, persisted after its collapse.
Like Milner and Mehrer, Pauketat and Emerson (1997) and Pauketat (1998) argue against
state-like portrayals of the Mississippian world, and they do not place the American Bottom at
the center of a regional trade network. Citing lack of evidence for large-scale trade, they suggest
that people in outlying areas strived to rapidly become associated with the Cahokia Mississippian
phenomenon via one-way acquisition of Cahokian goods and emulation of Cahokian styles, as
Mississippianism was appealing on a supernatural or ideological level. They support this
assertion with the observation that, although Cahokian pottery and other indicators of social
influence were abundant beyond a 150 kilometer radius around the polity, the most current
population estimates for the prehistoric Midwest indicate that population density was simply
insufficient for the support of a true mercantilist economy.
Aside from characterizing the low level of economic influence between Cahokia and
outlying regions, Pauketat (1998) also investigates matters of social and political influence. He
criticizes what he refers to as “minimalist” interpretations of Cahokia’s regional importance as
denying human agency and internal causality and as ignoring abundant evidence for Cahokia’s
rapid political centralization and widespread social influence. Pauketat insists that the spread of
Cahokian ideas be conceptualized as an active process of negotiation between traditional
practices and new Cahokian ideals. He suggests that these phenomena are visible
archaeologically, particularly through analyses of household and craft production refuse,
symbolic representations of a communal fertility cult, architectural regimes, and site layout. He
interprets the available lines of evidence as indicating that Mississippian political consolidation

38

in the American Bottom occurred rapidly around AD 1050 and that communal symbolism was a
medium of political influence that transmitted the “new Cahokian order” to surrounding regions.

Regional Models of Mississippianization
Although the specific archaeological appearance of Mississippian sites is certainly dependent on
local idiosyncrasies, Stoltman (1991) and Farnsworth et al (1991) provide useful models for
characterizing these local expressions of Mississippianism by reference to the ideological, social,
and/or economic relationships between Cahokia and its periphery. Stoltman focuses specifically
on the nature of the interactions between Cahokia and outlying Mississippian and nonMississippian sites, distinguishing between the archaeological correlates of site-unit intrusion,
trait-unit intrusion, and emulation and whether these resulted from direct or indirect contact. In
their study of the nature of Mississippianization of the Lower Illinois River Valley, Farnsworth
et al (1991) propose a similar framework for characterizing the relationship, although they
address the issue of ideological versus economic interests in somewhat greater detail. Both
indicate that the reasons for site-unit intrusion (purposeful “colonization” as opposed to necessity
resulting from sociopolitical fissioning or population growth) are important, but unclear,
distinctions. These ideas have obvious implications for interpretations of the archaeological
record outside of Cahokia, particularly for understanding the specific nature and archaeological
expression of Mississippian culture in individual communities or regions.

Central Illinois River Valley
The Central Illinois River Valley Mississippian groups provide a particularly interesting
comparative case study for modeling the spread of Mississippian culture into other regions,

39

including the Lower Illinois River Valley and broader West Central Illinois. Harn (1991) reviews
evidence from the Eveland site and suggests that the Mississippianization of this region resulted
from the migration of people from the American Bottom during the Lohman and Stirling phases.
This is evident through the sudden appearance of Mississippian pottery, architecture, and
mortuary practices in the region, with little evidence for transitional forms. The purpose of the
migration is, of course, less clear. Harn proposes several scenarios, including colonization and
sociopolitical fissioning. Steadman (2001) detects the introduction of new alleles into the
established local community in her analysis of biodistance among individuals interred at Dickson
Mounds. Although she cannot definitively track the origins of these alleles to the American
Bottom, her interpretation is consistent with Harn’s proposition of an influx of Mississippian
migrants into the region. Furthermore, her analysis detected only minor perturbations in the
population structure of the central valley with the decline of Cahokia.
Armelagos and Hill’s (1990) evaluation of Mississippian period skeletal remains from
Dickson Mounds reveals a decline in health over the course of the transition from a Late
Woodland to Mississippian lifestyle, despite the availability of abundant wild foodstuffs and
fertile soils for food production. Although other studies (i.e. Goodman et al 1984) yielded similar
results, the thrust of their argument lies in their interpretation of the social meaning of the data.
Specifically, they suggest that sociopolitical factors such as trading local subsistence goods for
items of social and ritual important could have played a role in declining community health.
Although they do, at least superficially, appear to support a trade-oriented model for the nature
of interactions between Cahokia and outlying regions, they also point out that it is difficult to
make a single unifying statement regarding the health effects of Mississippianization, as analyses
at different regional or social scales yield disparate results, a point also emphasized by Cook

40

(2007). Harn (1991) also entertains this possibility, but notes the lack of archaeological evidence
for production of a tradable surplus. In addition, he points out that social stress, rather than
drastic nutritional insufficiency, may have contributed to the appearance of the skeletal remains
from this site.

Lower Illinois River Valley
Further south, in the Lower Illinois River Valley, local manifestations of Mississippianization
appear to have resulted from different sociocultural processes. The late prehistoric Lower Illinois
River Valley was occupied by a Late Woodland population that eventually became
Mississippianized via diffusion and acculturation with only limited colonization (Delaney-Rivera
2004, 2007; Schroeder 2004). Consequently, these communities began to display mortuary
patterns and material culture in keeping with the Mississippian phenomenon elsewhere in the
region. However, unlike the Central Illinois River Valley and other outlying regions, the Lower
Illinois River Valley seems to be largely devoid of large sites with temple mounds. Instead, small
farmstead sites, occupied by one or two families, seems to be more common (Conner 1984,
Goldstein 1981).
Connor (1984) reports on Hill Creek, a dual structure Mississippian farmstead in the
Lower Illinois River Valley. Although Hill Creek dates to Moorhead or Sand Prairie times in the
Cahokian chronology, Connor revisits a previous study of Lower Illinois Valley farmstead
settlement by Goldstein and suggests that, while analyses of ceramics from early farmsteads in
the region reveal affinities with Cahokian wares, those from later sites like Hill Creek are
stylistically related to the Central Illinois Valley, but manufactured using local materials.
Farnsworth et al (1991) criticize this work, suggesting that the lack of substantial Mississippian

41

sites in the region indicates that the Lower Illinois Valley was populated by groups who
essentially retained their Late Woodland identities while incorporating some aspects of
Mississippian mortuary ritual. However, the presence of Mississippian style pottery, houses,
social relations, and subsistence regimes among Mississippian period inhabitants of the region
(Connor 1984, Goldstein 1981, Rose 2007) seems to indicate otherwise

Mississippianization in Other Regions
The research presented here focuses on the Mississippianization of West Central Illinois, and as
such much of the foregoing discussion centered on that region. A brief exploration of a few other
key regions further demonstrates the variety and complexity in the process of
Mississippianization on a local level. Sites like Hiawassee Island and Toqua demonstrate a
unique expression of the Mississippian phenomenon in Eastern Tennessee. Although first
thought to have resulted from immigration and replacement, more recent models for the
Mississippianization of this region suggest local development from Late Woodland roots, with
movement of ideas into the region having a stronger effect than any movement of actual people
(Schroedel et al 1990). However, Sullivan (1995) shows that there existed significant variation in
Mississippian households and communities throughout Eastern Tennessee, and used some
mortuary based gender studies to demonstrate some of her ideas regarding similarities and
difference between sites within the region. This is taken here as an important caveat for
considering scale of analysis in studies of the expression of gender ideology.
The Mississippian center of Moundville in West Central Alabama emerged around the
same time as Cahokia, and appears to have resulted from social reorganization and cultural
change among local Late Woodland populations (Welch 1990). There is some evidence that

42

Mississippianization of this region occurred relatively abruptly in a Late Woodland environment
of increasing population density, intergroup violence, and insipient status differentiation and
competition (Knight and Steponaitis 1998). There is little evidence for substantial Late
Woodland occupation at the Moundville site, but production of local Late Woodland pottery
persists well into the Mississippian period here (Steponaitis 1998), further suggesting an
essentially in situ Mississippianization of a local Late Woodland population. Steponaitis (1998)
also points out, however, that the recognition of Late Woodland burials at the site is hindered by
the preference for interring the dead without diagnostic artifacts. This might make
bioarchaeological analyses of the Late Woodland Mississippian transition, such as that presented
in the current study, difficult to accomplish for Moundville.
Goldstein (1991) provides an overview of research on Mississippianization at the Aztalan
site in southeast Wisconsin in order to propose an organized approach to explaining its
relationship to Cahokia and situating it within the larger Mississippian settlement system. Mound
construction, subsistence data, and many aspects of the artifact assemblage indicate clear Middle
Mississippian influence, but other observations are more difficult to explain. For example, many
researchers have noted the physiographic and environmental differences between the American
Bottom and the Aztalan landscape, and Goldstein suggests that people from northern Illinois
would have been more familiar with, and comfortable settling into, such environments. This
scenario might also help to explain the ceramic assemblage, which is dominated by grittempered, collared vessels (similar to Northern Illinois varieties) rather than the typical shell
tempered varieties associated with most Mississippian sites. Goldstein suggests that an initial
influx of northern Illinois Late Woodland peoples around AD 800-900 was followed by a later
migration of Middle Mississippians, explaining both the location and the ceramics. Furthermore,

43

a relatively low population or threat of violence from nearby Oneota populations may have
prevented the establishment of smaller, outlying hamlets and farmsteads. She suggests that the
abandonment of Aztalan might have been related to the collapse of Cahokia.
At the Mississippian site of Angel, the nature of the Late Woodland-Mississippian
transition is less clear. Marshall (2011) suggests three possible scenarios: Mississippian
immigration to Angel, continuity between local Late Woodland/Emergent Mississippian peoples
and Angel phase inhabitants of the site, and a combination of local and non-local people who
came to be united under a common Mississippian ideology. Muller (2009) states that Terminal
Late Woodland Yankeetown phase settlements persist into early Mississippian times in the
Angel area, but he suggests that this should not be taken as contemporaneity of Late Woodland
and Mississippian peoples in the region. Instead, he implies a more complex transitional scenario
in which cannot be understood simply by dating the transition from grog to shell temper in
pottery. Expanding on this notion and applying it on a broader regional level, it is argued in this
dissertation that a complete understanding Mississippian and Mississippianization at any site is
contingent on synthesizing diverse archaeological data (i.e. moving beyond artifact typologies)
with problem-oriented bioarchaeological analysis.

Tradition and Acculturation on the Mississippian Periphery
Closer to the American Bottom, at the Bridges site in South Central Illinois, Oetelaar (1993)
reconstructs community life and situates rural settlements within the larger regional
Mississippian framework. He delineates family and communal areas within the site, suggesting
that public space was necessary for the development and maintenance of a Mississippian lifestyle
by providing venues for ritual, subsistence, and tribute activities. He also points to the
44

contemporaneous use of wall trench houses and Late Woodland pit features as potential evidence
for the maintenance of some traditional practices in spite of overall Mississippianization. Alt
(2002) takes a similar approach in her evaluation of Mississippian Richland Complex sites in the
uplands surrounding the American Bottom, evaluating ethnicity from the perspective of
community identity, social interaction, and political change. She shows that traits such as house
construction and ceramic production techniques might indicate acceptance of new Mississippian
ideas on the one hand, while other features such as site layout, which may be more indicative of
intercommunity social relations, point to the active maintenance of traditional social relations.
This might indicate some level of resistance to complete Mississippianization well into the
period of perceived Cahokian dominance.
In another take on the issue of resistance, tradition, and emulation, Emerson (1997)
suggests that significant cultural changes (ethnogenesis) can indicate a response to cultural
intrusion from other groups, a mechanism for defining “otherness,” and the act of competing for
cultural survival. Framing the development of the Upper Mississippian Langford tradition as a
cultural response to Middle Mississippian intrusion in what is now northern Illinois, he suggests
that ethnogenesis can be visible archaeologically through things like population aggregations,
centralization of leadership, increased violence, limited incorporation of outside cultural
practices, and increased territorial boundedness. Through an investigation of subsistence,
settlement patterns, and mortuary practices, Emerson shows that each of these things, although
clearly rooted in the terminal Late Woodland culture that preceded it, are amplified in the
Langford Tradition. These changes are sudden and contemporaneous with the rise of the nearby
Illinois River chiefdoms, lending support to the idea that Langford development was in response
and resistance to the new social situation.

45

Some of these ideas might be applied to the situation further to the south, in the Lower
Illinois River Valley. Unlike the Central Valley and other outlying regions, the Lower Valley
seems to be largely devoid of large sites with temple mounds. Instead, smaller Mississippian
farmstead sites appear to be more common. Farnsworth et al (1991) suggest that the lack of
typical, substantial Mississippian village sites in the region indicate that the Lower Illinois
Valley was populated by groups who essentially retained their Late Woodland identities while
incorporating some aspects of Mississippian mortuary ritual. However, the presence of
Mississippian style pottery, houses, social relations, and subsistence regimes among
Mississippian period inhabitants of the region (Connor 1984, Goldstein 1981, Rose 2008)
indicate deep involvement in a Mississippian lifestyle. A more nuanced approach to identity and
ethnicity and its relationship to material culture in this region, coupled with discussion of the
meaning of these concepts in individual archaeological interpretations, could potentially clarify
these issues. This dissertation is intended to contribute a small piece of this puzzle by exploring
the extent to which the adoption of archaeologically visible Mississippian cultural practices (i.e.
mortuary patterning) in the Lower Illinois River Valley is reflective of real social differentiation
and lived biosocial experiences of Lower Illinois River Valley inhabitants.

Sex and Gender in the Late Woodland and Mississippian Worlds
Archaeological Research as a Baseline for an Engendered Bioarchaeology
Gender is a fundamental cultural institution that organizes behavior and social interaction in
complex ways, even within relatively egalitarian societies (Hill 1998, Macias-Lees and Black
2000). Specific investigations of changing gender ideologies resulting from Mississippianization
are rare, but a few available studies provide a starting point for interpreting the results of this

46

research. Fritz (1999) and Watson and Kennedy (1991) both report on the origins of cultivation
in the Eastern United States, proposing that women were responsible for the technological
advances that lead to the development of agriculture in this region. Benn (1995) suggests that
changing subsistence regimes had important implications for gender relations in the past, as a
change in the importance of collected or cultivated plant foods would be directly related to
women’s economic power. Pauketat (2004) points to frequent iconographic depictions of
Mississippian females engaged in agricultural activities and males engaged in priestly and gamerelated activities as evidence for the polarization and politicization of gender ideology
concomitant with increasing social complexity and regional integration. Alt and Pauketat (2007)
elaborate further and propose that the “theatrical” nature of Cahokian political ritual
disseminated changing gender ideologies throughout the Mississippian world. As an example of
this, Pauketat and Emerson (1991) suggest that decorative motifs on ubiquitous Ramey-incised
pottery may have had sexualized and gendered meanings. Furthermore, Emerson et al (2003)
observe that Mississippian figurines depicting males are more widely dispersed throughout the
periphery than female representations, posing important questions regarding the nature of
receptivity and resistance to Cahokian gender ideologies in outlying communities.
Another perspective on Cahokian influence can be seen through an evaluation of
ideology and gendered social relations in the Mississippian world. Alt and Pauketat (2007)
reevaluate excavation data from Wilson Mound in the American Bottom and propose that the
unique burial assemblage might offer clues regarding a gendered component to Mississippian
political activities. The central burial feature contained primary interments of several women,
infants, and children. One of the females was carrying a full term fetus, which they interpret as a
woman who died in childbirth. Bundle burials were also interred in the mound. The authors

47

suggest that this may represent a transition of power from one powerful lineage to another, with
the new leaders ceremonially ending the reign of the old lineage by publically killing and
interring their reproductive base. The importance of women in both the productive (i.e.
agricultural) and reproductive roles in the Mississippian world certainly had implications for
many aspects of culture, but it is unclear how this may have been absorbed by outlying
Mississippian communities. In particular, Alt and Pauketat point to evidence for a wider
dispersal of Cahokian male warrior/ritual iconography outside of the American Bottom
compared to only limited distribution of female iconography. They suggest that peripheral
Mississippian communities chose to emphasize some aspects of the Mississippian ideology while
downplaying others. This is an interesting interpretation, and one that might have important
implications for studies of gender relations in the American Bottom and outlying areas.
Nevertheless, several of these same authors also point out that there is no a priori reason
to believe that people in outlying sites, with their own local histories, would have passively
absorbed and reflected new Cahokian ideas, nor would they have necessarily accepted all new
Cahokian social ideologies and practices as a total package (Alt and Pauketat 2007). This
observation may be particularly astute when one considers that the exact nature of the
relationship between centers like Cahokia and sites located in the periphery, particularly
regarding the extent of social and political influence, is not fully understood and likely varies by
locality (Anderson 1997, Emerson 1997, Milner 1998, Farnsworth et al 1991, Stoltman 1991).
Another complicating issue is that, other than Benn’s (1995) suggestion that changing
subsistence regimes had important implications for gender relations in the past, there is little
gender-oriented research regarding Late Woodland social dynamics. This makes it difficult to
evaluate the strength of models that propose revolutionary changes in gender constructs with

48

Mississippianization or the way that any changes may have affected the experiences of real
people, as we are lacking a baseline against which to gauge observed Mississippian patterns.
Furthermore, although the studies cited above build strong arguments for a normative ideological
shift with Mississippianization, none has adequately approached the way that this shift affected
the lived experiences of ordinary Mississippian people. Gender may well have acquired new
political and ritual meanings at Cahokia, but the extent to which gender relations in outlying
communities were reflective of the Cahokian pattern in not known.

Bioarchaeological Perspectives- Criticisms, Shortcomings, and Sex-based Research
The disconnect between studies of physical bodies and studies of culture and social life has
sometimes led to interpretive problems, as demonstrated in the occasional conflation of the terms
sex and gender in bioarchaeology (Goldstein 2006, Gellar 2008, Walker and Cook 1998).
Although osteological determinations of sex are useful and necessary in mortuary analyses, in
isolation, osteological data are insufficient for exploring gender as a cultural phenomenon
(Gellar 2008, Walker and Cook 1998). Gellar (2008) points out that bioarchaeology has the
unique role of investigating the intersection of body and culture, but she also recognizes that
bioarchaeologists have taken an essentialist view of differences between males and females and
have failed to incorporate more than overly simplistic social theory when making gendered
interpretations of the past. While bioarchaeologists are skilled at discerning biological
differences, they are often challenged to interpret these differences within an archaeological
framework to produce more nuanced observations about gender in the past.
Several important studies have focused on patterns of health and stress within the context
of the Late Woodland/Mississippian transition, although few have done so from an

49

unequivocally gendered perspective. Some studies have indicated changes in the division of
labor by sex with the advent of agriculture based on occupational stress markers and crosssectional geometry (Bridges 1991, 2000), but studies of this nature have been plagued by both
practical and interpretative issues (Jurmain 1998, Bice 2003). Furthermore, when taken on their
own, they offer little in the way of social implications of the patterns they are detecting. For this,
we must investigate not only to the type of work being done, but also to the changing gender
relations indicated by the observed patterns. These data must be collected with careful attention
to sex, but interpreted within a larger framework of appropriate archaeological investigations.
A few bioarchaeological studies from the Eastern United States and beyond provide a
backdrop for the research questions investigated in this dissertation. Goodman et al (1980)
investigated changes in the frequency of linear enamel hypoplasia in males and females interred
at Dickson Mounds in West Central Illinois, concluding that during the Late Woodland, males
displayed fewer hypoplasias than females, but that these proportions were essentially equal by
Mississippian times. Cook (1984) reported that, for a set of Illinois skeletal samples (including
Schild), the frequencies of microscopic enamel defects in the first permanent molar were the
same in males and females in the Middle Woodland, but that the frequency in males increased in
the Mississippian period. No data for the intervening Late Woodland period were available. In
reviewing the implications of these and other studies, Cohen and Bennett (1993) suggest that
male children were afforded less “protection” from nutritional and pathological insult in the
Mississippian than in the Late Woodland period, possibly reflecting changes in the perceived
value of women’s work with the rise of agriculture. However, other studies have resulted in
fundamentally different conclusions. In his analysis of growth and stress in the Dickson Mounds
Late Woodland and Mississippian series, Clark (1985) concludes that, while both sexes

50

experienced early growth disruption in Mississippian times, female growth disruption was
chronic while males underwent a period of catch-up growth during adolescence. He interprets
this as an increase in male status with Mississippianization, achieved primarily during
adolescence. The reason for the discrepancy in the results of these Dickson Mounds studies is
unclear, but may be related to the disparate nature of the biological phenomena they are
measuring (i.e. episodic versus cumulative growth disruptions). This idea will be explored
further in chapter six.
The ways in which Mississippian ideas regarding social organization and status affected
the proportion of maize in ones diet is of considerable interest to many archaeologists, including
bioarchaeologists. One promising and increasingly popular source of paleodietary information is
isotopic studies. Despite some interpretative and methodological problems, stable isotopic
analyses have contributed to a wide range of bioarchaeological studies, particularly in light of
recent advances in technology and understanding of metabolic and taphonomic issues (Cook
2007, Ambrose et al 2003, Schoeninger and Moore 1992). In the Eastern Woodlands, many
isotopic studies have centered on clarifying the timing and impact of maize agriculture,
suggesting that maize played only a minor dietary role well into the Late Woodland period
(Cook 2007). Intracommunity social relations are investigated by Rose (2007) through an
evaluation of variability in diet over time during the adoption of maize agriculture. Stable isotope
analysis of bone collagen allowing for estimation of the relative proportion of maize and meat in
the diet of Middle Woodland, Late Woodland, and Mississippian individuals from five sites in
the Lower Illinois Valley and Central Mississippi Valley in West Central Illinois. Although
variable maize intensification is evident at some late Late Woodland sites, the general trends are

51

nearly identical to those seen in the early Late Woodland, with no correlation between diet and
age, sex, or status.
During Mississippian times, maize constituted a major component of the diet of most
individuals, regardless of social differentiation. Schurr (1992) used stable isotope analysis to
compare dietary variability to mortuary treatment at the Angel site in Southern Indiana. He
suggests that social differentiation, as evidenced through mortuary differentiation, is not
correlated with the amount of maize in ones diet at this site. Rose (2007) found that only Schild
Knoll A showed significant sex differences in carbon isotope ratios and nitrogen values. Females
had enriched carbon isotope values and lower nitrogen values, apparently indicating a diet with
more maize and less meat than males. Females in mass graves in Mound 72 at Cahokia also
appeared to have had a high maize diet, but the nature of this particular deposit, especially the
possibly disparate and extraordinary life histories of these women, makes interpretation and
relation to the results to the rest of the Mississippian world somewhat problematic (Ambrose and
Buikstra 2003). Other authors have also attempted to analyze dietary data from a gender
perspective by incorporating status into their interpretation of sex related health differentials in
complex societies (Ambrose et al 2003, White et al 2001). An important study by Hedman et al
(2002) addresses dietary variability within the American Bottom, indicating possible sex and
status-related differences within and between both floodplain and upland sites. As an added
component, they also compare their data with previous isotopic studies from Schild Knoll A
remains. Both male and female Schild individuals showed dietary similarities with males and
females from the American Bottom uplands, but the cultural meanings of these similarities are
difficult to ascertain because of small sample sizes, the effects of individualized diets, and the
possibility of interregional migration (Hedman et al 2002).

52

Situating the Current Study Within the Existing Literature
This chapter has provided a basic overview of Late Woodland and Mississippian cultural
traditions of the late prehistoric midcontinent, and has specifically emphasized issues of
particular relevance to this dissertation. An inherently complicating issue for scholars of gender
transformations with Mississippianization is the general lack of gender-oriented research
regarding Late Woodland social dynamics. This makes it difficult to evaluate the strength of
models that propose revolutionary changes in gender constructs with Mississippianization or the
way that any changes may have affected the experiences of real people, as we are lacking a Late
Woodland baseline against which to measure proposed changes that lead to Mississippian
patterns. This dissertation begins to broach this issue by contributing another component (i.e.
changes in growth patterns) to the sex based bioarchaeological literature regarding the Late
Woodland-Mississippian transition. However, a truly engendered interpretation of this
phenomenon awaits an appropriate rigorous mortuary analytical perspective on Late Woodland
intracommunity social dynamics, which is currently lacking.
Ideas about the role of tradition in acculturation are important because they remind us
that, although broad regional trends in Mississippianization are of obvious importance, local
variation should not be discounted. Pauketat’s framework, particularly his emphasis on the role
of negotiation between traditional practices and new Cahokian ideals, provides a good starting
point for the exploration of Mississippianization on a more local level. Local perspectives are
particularly important for exploring the relationship between Cahokia and peripherally located
Mississippian sites. Issues of power and social organization at the community level are clearly
tied into the types of large-scale, regionally oriented models of political and economic interaction
described above. However, it would be an error to assume that cultural change was

53

homogenously accepted and incorporated by existing communities or that the spread of
Mississippian culture was always accomplished through similar mechanisms. Evidence
suggesting either maintenance or change of traditional intracommunity social relations, including
those associated with gender, might suggest that the Mississippianization of outlying
communities was a locally variable phenomena, complicating conceptions of the nature and
definition of Mississippian ethnic identity. This dissertation will contribute a West Central
Illinois/ Lower Illinois River Valley perspective to this problem.
“Cahokia-centric” notions of Mississippianization may be predisposed to ignoring local
trends in the Mississippian periphery that, instead of being somehow less authentically
Mississippian, might be more fruitfully considered as equally legitimate variants on a broader
cultural theme. Furthermore, publically and ritually focused investigations of gender in religious
and political life, which seem to be the norm in Mississippian gender scholarship, are interesting
in their own right, but they inherently result in normative statements regarding Mississippian
gender ideologies. For example, Emerson et al (2003) observe that Mississippian figurines
depicting men and masculine activities are more widely dispersed throughout the periphery than
female representations, and this observation might be interpreted as suggesting that
Mississippianization may have had a more profoundly transformative effect on the lives of men
(as compared to women) in outlying communities. This is something that can and should be
tested in the archaeological record, but which must be done on a local level. Community-focused
archaeological analyses are better positioned to determine the ways in which gender ideologies
affected social relations for the average Mississippian person, and this research aims to ascertain
the lived experiences of gendered Mississippian peoples through a socioculturally sensitive
biological analysis of growth pattern differentials.

54

Based on iconographic evidence from the archaeological record in addition to
ethnohistoric accounts, the ideological association between women, maize agriculture, and
fertility is clear, but the way the relationship to traditional gender ideologies (historically rooted
in the Late Woodland) or to the lived lives of Mississippian women is not entirely clear from the
available evidence. Inter- and intracommunity social relations in the Mississippian world had
important implications regarding food production and consumption, and studies such as those
provided by Rose (2007) and Hedman et al (2002) will certainly continue to come out of isotopic
approaches to these problems. However, non-chemical studies of skeletal growth can also
contribute to this issue. Agriculture, as the primary means of subsistence in Mississippian
societies, may have provided an avenue through which women could gain power and prestige,
which can potentially be reflected in various health and growth patterns. This is a basic premise
of bioarchaeological research, one that this dissertation is built upon, and one that can
complement a wide range of approaches to gender dynamic in the past.

55

CHAPTER THREE- GROWTH AND STRESS IN BIOANTHROPOLOGICAL
ANALYSIS

A basic premise of this dissertation research is that human growth patterns vary in relation to the
biological, social, psychological, and environmental circumstances of life, and that such
circumstances affect individuals differentially according to various social parameters (i.e. class,
gender, etc). Although no single variable provides a perfect proxy for general health (Arora
2001), exploring the relationships between multiple indicators contributes to this study in
particular, and to the development of bioanthropological method and theory, in several ways.
Testing for relationships between certain indicators (between linear enamel hypoplasia and
neural canal diameter, and between femur length and vertebral body height in the current study)
can provide a framework for evaluating the general utility of poorly understood methods of
growth assessment. Building upon this, interpretation of significant relationships (or lack of
relationships) between various growth indices provide a basis for understanding the ways that
various measures are reflective of similar or different growth phenomena.
This chapter provides an overview of relevant clinical and bioanthropological literature in
order to make a case for studies of human skeletal growth as a valid means of exploring the
biological implications of social differentiation in prehistoric communities. The specifics of the
growth variables chosen for this study are reviewed and assumptions regarding the relationships
between them explained.

56

Understanding Stress
A basic tenet underlying much bioanthropological research holds that unmitigated physiological
and psychosocial stress is related to growth stunting, compromised immune response, and
relatively reduced levels of health in affected individuals. Goodman et al (1988) incorporate a
biomedical concept of stress and associated psychosocial and physiological processes into an
anthropological understanding of biological and cultural adaptations. The Selyean approach to
stress, which forms the foundation of their model, emphasizes two key issues: 1) A wide variety
of stressors can manifest themselves via a small number of indicators, and 2) Perceived stressors
have an important impact by providing new sources of psychosocial stress and exacerbating
stress resulting from biotic sources (e.g. pathogens, conditions in the natural environment).
These concepts allow for the integration of biobehavioral responses to stress into anthropological
models, accounting for stress-mitigating and stress-exacerbating affects of cultural behaviors.

Growth and Heritability in Living Populations and Historic Data
Notions of plasticity of the human phenotype in relation to social and environmental conditions
can be seen in the anthropological literature as far back as Boas’s (1912) research on growth
among American born children of immigrants. That growth/body dimensions are a reflection of
the health and well-being of living individuals and populations is well supported by research
focused on living communities (WHO 1995, Himes 2004). The clinical literature on growth
typically focuses on stature and/or body mass in general, rather than emphasizing skeletal growth
per se. Still, the establishment of a solid relationship between body growth and generalized
health status is fundamental to the current study.

57

Stature is one of the most heritable of all anthropometric growth indices, with heritability
reaching 0.92 in some populations (Czerwinski and Towne 2004). In a literature review,
Silventoinen (2003) shows that in modern Western societies, about 20% of total adult stature
attainment can be attributed to environmental (rather than genetic) factors, the most important of
which are childhood nutrition and disease status. She also suggests that the average proportion of
body height under environmental control in developing countries may be even higher, with lower
general heritability and greater discrepancies between status groups.
Beard and Blaser (2002) suggest a direct relationship between pathogenic growth
disruptions and decreased adult stature across time and space based on both historic and
experimental analyses, and Steckel (1995) makes a particularly strong historically,
contextualized argument for a general and significant correlation between economic standard of
living and adult stature in modern developed nations. In an evaluation of historic penitentiary
records, Tatarek (2006) finds that geographic origin and occupation better explain variation in
height data for male inmates than do other factors such as race and birth cohort, although she
suggests that unevaluated factors (such as education level) may also have had a strong effect.
Because occupation and geographic are proxies for natural and social (i.e. economic) conditions,
her analysis supports the hypothesis of a strong relationship between environment (both social
and natural) and attained adult stature. Steckel’s (1987) analysis of African slave height records
documents the extraordinary ability of catch up growth, during the adolescent growth period, to
virtually obliterate evidence of significant early childhood growth disruption related to disease
load, poor nutrition, and poor prenatal health.
The above discussion emphasizes that the relationship between generalized cultural and
environmental conditions and growth patterns can be ascertained for past populations using an

58

approach that combines historic records with modern biosocial growth models developed from
living communities. These studies provide a theoretical and methodological bridge between
studies of living groups (as discussed above) and studies of skeletal remains from archaeological
sites lacking written records. Prehistoric bioarchaeological analyses are clearly hindered by lack
of documentation. However, incorporation of appropriate archaeological data can largely
compensate for the lack of written sociocultural records.

Stress and Growth Patterns in Prehistoric Populations
Prehistoric studies of the effects of stress on skeletal biology yield insights regarding
generalized, long-term patterns of stress throughout space and time, despite the fact that they are
not well suited to identifying the proximate cause(s) of stress in individuals and, in many cases,
have difficulty incorporating the concept of perceived stress (Armelagos et al 1988). Larsen
(1997) emphasizes that the skeletal expression of stress and deprivation must be understood as
the result of the synergistic effects of multiple factors, including genes, hormones, physiological
stress, environment, nutrition, sanitation, hygiene, disease, and degree of sedentism. Goodman
et al (1984) summarize the uses of skeletal stress indicators in determining health patterns in
prehistoric populations, emphasizing a culturally-oriented, multiple indicator approach. Skeletal
stress indicators are discussed within three categories: 1) general, cumulative indicators, 2)
general, episodic indicators, and 3) specific indicators. The particular qualities and utility of each
of these categories have important implications for the interpretation of ancient health patterns,
and will be discussed in this dissertation where appropriate.

59

Social Meaning of Intragroup Variation in Growth Experience
Patterned variability in growth experiences among living humans, as indicated by various types
of anthropomorphic measures including skeletal growth and height, can be seen not only across
populations, but also within communities and even within households (WHO 1995). The nature
of contemporary populations allow for the development of directly testable hypotheses regarding
the specific biological, social, and environmental correlates of intracommunity health
discrepancies. In general, the nature of bioarchaeological samples precludes such detailed levels
of analysis, although it should be noted that advances in the recognition of certain specific
diseases in the bioarchaeological record, detailed mortuary analyses, and the historic
documentation pertaining to social and biological status of the deceased can help to address this
problem.
It is not the express intention of the current research to identify specific causes or
explanations for the growth patterns detected. However, a general correlation between growth
differentials and socioeconomic status differentials has been firmly established in
bioanthropological literature regarding living peoples (Bogin 1988, Leatherman 1998, Bogin et
al 2007). Goodman (1998) contends that ideology and economy should inform studies of
intercommunity studies of growth in the past, and, building on this, it is suggested here that
bioarchaeologists should also be concerned with 1) the ways that ideology and economy inform
intracommunity social relations and, 2) that those relationships might become more visible in
studies of growth. As such, intracommunity social parameters like status and kinship should be
given as much credence in bioarchaeological studies as the biological categories of sex and age.
It is argued here that this can be accomplished through the integration of the results of
osteological analyses with the results of other archaeological and bioarchaeological analyses (i.e.

60

mortuary patterning, nutritional analyses, epidemiology and paleopathology, trauma analyses,
etc.). This approach can contribute to a more complete understanding of biosocial processes and,
hence, quasi-proximate causality for changes in health patterns. In any case, exploration of the
nature of the relationship between multiple stress indicators and multiple biocultural categories,
combined with a recognition of the strengths and weaknesses of various methods and levels of
analysis in bioanthropological studies of growth, will contribute to more complete interpretations
of the concept of stress within the scope of a broader understanding of human adaptability.

Osteological Parameters Used in the Current Study
Vertebral Dimensions in Anthropological Analyses
Several authors have noted potential sensitivity of the vertebral column to growth disruptions
(Clark et al 1986, Larson 1998, Hoppa and Fitzgerald 1999), but few have applied this
knowledge to bioarchaeological situations. Clark (1988) developed a vertebral morphometric
method for growth assessment and used it to investigate the relationship between social
complexity, sex, and growth in Mississippian and Pre-Mississippian skeletal remains from
Dickson Mounds. A detailed review of his vertebral growth model and his method are provided
in his dissertation (Clark 1985).
Clark’s vertebral method goes beyond traditional methods of evaluating stress (e.g. long
bone length) by distinguishing between chronic and acute growth disruptions, the recognition of
which has culturally significant implications. Individualized and generalized curves for four
broad, physiologically based categories of bodily growth first described by Scrammon (1930) are
still cited as reliable qualitative depictions of human growth trends (Molinari and Gasser 2004).
Of particularly interest for the current study are 1) the general type, which comprises most

61

skeletal growth, including the femur, and 2) the neural type, which comprises the brain, spinal
cord, and associated structures. Clark et al (1986) cite vertebral growth data and suggest that,
just as skull growth occurs based on the neural curve in order to accommodates rapid brain
development during early childhood, vertebral arch growth progresses according to a similar
growth trajectory due to its association with the spinal cord. General growth increases gradually
until puberty, when more rapid, exponential growth begins to occur. Neural growth, on the other
hand, progresses exponentially through early childhood, tapering off prior to puberty (Figure 2) .
This means that the ultimate size of the neural canal is achieved during early childhood, long
before the completion of general body growth (Figure 3).

Figure 2- Growth Curves (from Clark 1986)
62

Figure 3- Vertebral neural canal size in children and adults (from Clark 1986)

Because vertebral arch growth ceases before vertebral body growth (during childhood
versus during late adolescence/early adulthood), Clark suggests that analysis of the relationship
between measurements of these two features may be useful in determining whether growth
disruptions are chronic, resulting from generally continuous stress and deprivation throughout
the growth period, or whether catch up growth occurred when stressful situations were
eventually ameliorated during the later adolescent period. Clark’s interpreted vertebral growth
patterning in the Dickson Mounds skeletal series as indicating that Mississippian females were
smaller than their pre-Mississippian counterparts, but that Mississippian females experienced
63

chronic growth disruption. Mississippian males, on the other hand, experienced early growth
disruption followed by periods of catch up growth. This has important implications regarding
differential treatment of male and female children in Mississippian society, and Clark interprets
his results as indicating an increase in male status with Mississippianization.
The current study provides new data on the relationship between generalized health/stress
and vertebral growth patterns, and contributes to the limited body of anthropological research on
the subject. Clark’s (1985) analysis, upon which the current study is largely based, is discussed
throughout this dissertation, but Tatarek (1999) offers another important analysis that warrants
detailed discussion. She used the Hammond Todd collection to analyze the relationship between
spinal stenosis in the cervical spine and various health parameters, many of which are only
available by virtue of documentation and records that are inherently lacking in a prehistoric
analysis. Tatarek provides a detailed account of morphological variability throughout the spinal
column and offers biomechanical, genetic, and ancestry-related explanations for the variation.
The current study did not test the same hypotheses as Tatarek, and therefore comparison with
many of her results is not possible at this time. Additionally, she dedicated much attention to
vertebra-by vertebra analysis. In contrast, this research focused more on the means based
analysis, as this research is modeled largely on Clark’s study. However, future research will
focus on reanalyzing the data collected for the current study so that it might be more directly
comparable to Tatarek’s analysis.
Tatarek (1999, 2005) found that sex was more highly correlated with cervical neural
canal diameter than age or ancestry, although interactions between sex and ancestry were also
found to be significant. These associations were not observed in the thoracic or lumbar segments.
She suggests that transverse diameters in the cervical spine may continue to grow until the age of

64

10, and are thereby more subject to environmental perturbation. Tatarek does not find a
correlation between LEH occurrence and decreased canal diameters (contra Porter 1986), but she
recognizes that vertebral canal diameter is likely dependent on more than genetics and sexual
dimorphism, concluding that more research is necessary before the relationship between neural
canal diameter and environmental factors can be understood. Further investigation of the
relationship between vertebral neural canal growth and early childhood growth disruption is
required.

Timing of Vertebral Growth
Successful application/testing of Clark’s method requires a detailed understanding of vertebral
growth. Unfortunately, the clinical literature is sparse, and what is available focuses heavily on
reviews of abnormal development. The bioanthropological literature provides the basic
framework for the timing of bone growth used in this study, and complementary information
from the clinical literature is cited when necessary and available.
At birth, vertebrae typically consist of three primary centers of ossification: the body and
two (left and right) arch components. The posterior aspects of the arches generally fuse at the
spinous process during the first and second year of life in the thoracic region, progressing from
the lower to the upper elements. In the lumbar segment, posterior fusion starts in the upper
elements at approximately one year of age, progressing downward to L5, which fuses by age
five. Fusion of the neural arch to the body occurs earliest in the lower thoracic elements, starting
at approximately three to four years of age and progressing superiorly through the sixth year. In
the lumbar region, fusion of the arch to the body occurs in the upper elements around two to
three years of age, and culminates with the fusion of L5 around age five (Baker et al 2005).

65

Supplementary information from clinical literature complements the growth model
presented above. Gepstein et al (1991) find that the anterior-posterior canal diameter is
significantly correlated with overall neural canal area (p=.05) in the second through fifth lumbar
vertebrae, while neither transverse diameter nor vertebral body height measurements are
significantly correlated with canal size. Ursu et al (1996) shows that, in the lumbar region of the
spinal column, transverse neural arch diameter in L1 through L4 reach 70% of adult size by the
time of birth, but L5 has generally attained only 50% of its adult size by then. This suggests that
growth disruption in infancy might have a more stenotic effect on L5 compared to the rest of the
lumbar spine.
While the vertebral arch growth data cited above pertains specifically to the conditions of
early childhood growth, vertebral body growth continues throughout childhood and into early
adulthood, making it a potential indicator of cumulative growth conditions over the entire growth
period. Sarwark and Aubin (2007) calculate that, on average, each vertebral body grows by about
one millimeter (0.5 mm per epiphysis) in height per year, with accelerated growth during the
early (0-2 years old) and prepubescent period. Song and Little (2000) note that growth in height
had ceased in 61% of their study sample by the time participants reached age 17, although they
did not distinguish between differential growth between various body segments (i.e. limbs versus
trunk). Baker et al (2005) note that thoracic vertebral body height growth ceases with epiphyseal
fusion by early adulthood, while the same process occurs somewhat earlier in the lumbar region.
Beyond his original analysis, Clark’s vertebral method has not been retested in a
bioarchaeological context. Explicit comparisons between vertebral growth patterns and patterns
of growth observed using methods that have been more widely applied will help in assessing its
validity in bioarchaeological contexts. Comparative methods that record growth occurring at

66

approximately the same time as those observable in vertebral canal and body height (specifically
early childhood and late adolescence) will be useful for assessing the validity of the vertebral
method. Two traditional methods of bioarchaeological growth assessment, LEH and long bone
length, were chosen for evaluation.

Long Bone Length
Human growth models are typically developed from living populations with both skeletal and
soft tissue intact. Long bone proportionality, as proxy for adult stature, is an established and
standard practice in bioarchaeology. Long bone length can be used as a cumulative, non-specific
indicator of stress and growth (Goodman et al 1984), although it must be recognized that final
adult length is the net result of a number of factors, including genetics and nutrition (Larson
1998).
The femur comprises five centers of ossification, but for the purposes of recording data
pertaining to adult stature, only the fusion of the proximal and distal epiphyses to the femoral
diaphysis is of concern. The proximal epiphysis first appears within a year of birth, and does not
fuse to the diaphysis until between 12 and 16 years of age in females and 14 to 19 years of age in
males. The distal epiphysis is present before birth and fuses between the ages of 14 to 18 in
females and 16 to 20 in males. Of particular interest in the current study is the relationship
between femoral length and vertebral body height, both of which are hypothesized to record
general, cumulative growth through early adulthood.

67

Linear Enamel Hypoplasia (LEH)
Enamel hypoplasia is an episodic stress indicator commonly cited in bioarchaeological literature
as the premier marker of early childhood stress events, with particular utility in discerning the
timing of stress within the early childhood growth period. However, like long bone growth, LEH
is a non-specific stress indicator and cannot, on its own, be used to determine proximate
causality. Nonetheless, the non-adaptive nature of LEH is suggested by the general correlation
between the exhibition of these characteristics and reduced life expectancy throughout time and
space. This observation suggests that these factors are indeed useful for making inferences
regarding quality of life in prehistoric populations (Larsen 1997, Cook 1981). LEH etiology,
factors contributing to their appearance and distribution, and implications for archaeological
studies have been thoroughly reported by Goodman and Rose (1990) and Goodman and
Armelagos (1985).
Developmental enamel defects represent disruptions to normal amelogenesis and result
from a wide variety of biological insults, including dietary insufficiency, trauma, and infections.
Goodman and Rose (1990) provide a review of amelogenesis and specific enamel structures,
such as lines of retzius, Wilson bands, and enamel prisms.

Reference to normal dental

development, including differential development and chronology, can assist in gauging the
timing and severity of these defects. Because all dental development occurs during childhood,
bioarchaeologists can use both microscopic and macroscopic enamel defects to explore the
connection between poor early childhood health and broader community health.
Several sources recommend collecting LEH data for the central maxillary incisors and
mandibular canines in the permanent dentition (Goodman and Rose 1990, Ubelaker and Buikstra
1994). Maxillary incisor crown development generally occurs between ages two months and four

68

years, while mandibular canine crown development typically occurs between ages four months
and six years (White and Folkens 2005). It is suggested here that the established relationship
between LEH and early childhood stress make this indicator comparable to neural canal
measurements collected using Clark’s method, as both are hypothesized to reflect conditions of
early childhood growth.

69

CHAPTER FOUR- SITES AND SAMPLE DESCRIPTIONS

This study evaluates human skeletal remains from two West Central Illinois mortuary sites:
Schild and Yokem. These collections are part of a large, temporally and culturally expansive set
of mortuary and skeletal series from the Lower Illinois Valley, and their importance for
investigations of the prehistory of the midcontinent is clearly demonstrated by their prominence
in the regional archaeological and bioarchaeological literature. This chapter provides an
overview of the mortuary contexts of the skeletal remains and briefly review pertinent previous
research focusing on these collections.

West Central Illinois- Cultural Setting
The Lower Illinois River Valley is commonly demarcated as the area surrounding the
southernmost 110 kilometers of the Illinois River (Delaney-Rivera 2007). It is widely recognized
in the literature as possessing distinct regional manifestations of both Late Woodland and
Mississippian cultural traditions, recognized primarily on the basis of ceramic sequences (Perino
1971, 1973), and is sometimes conceived of as a cultural extension of the American Bottom
(Brown et al 2007). Regional similarities in cultural practice reach beyond the boundaries of the
Lower Illinois River Valley proper to encompass a somewhat broader area that includes the
Yokem site, which is located outside of the Lower Illinois River Valley proper, on a bluff
overlooking the Mississippi River floodplain. Intercommunity gene flow has also been
demonstrated within this more inclusively defined region of West Central Illinois (Droessler
1981, Conner 1984, Steadman 2002). Based on these cultural and biological relationships, some
archaeological and bioarchaeological analyses have considered Yokem and Schild to be part of

70

the same regional expressions of Late Woodland and Mississippian traditions. As such skeletal
samples from these sites were combined in the current study when necessary for sample size.

The Sites
Use of the Schild and Yokem cemeteries spanned the Late Woodland-Mississippian transition,
providing researchers with a good skeletal series for investigating the biological effects of social
change over time. Schild and Yokem are located approximately 105 and 240 kilometers
(respectively) north of Cahokia. The Schild site is located near the town of Eldred in Greene
County, Illinois, while the Yokem site is near Quincy in Pike County (Goldstein 1980, Perino
1971 and 1973). Both the Late Woodland and Mississippian components of Schild and Yokem
were excavated by Gregory Perino in the early and mid-1960s (see Table 1 for radiocarbon dates
and Figure 4 for site locations). Skeletal remains from both sites are curated at Indiana
University.

Site
Schild

Component
Date
Late Woodland
AD 1200-1300
Mississippian
AD 1100-1200
Yokem
Late Woodland
AD 900-1100
Mississippian
AD 800-900
Table 1- Radiocarbon dates and phases for study sites (from Steadman 2002)

71

Figure 4- Site locations (From Perino 1971a)

72

Late Woodland Components
Schild (Perino 1973)
Nine Late Woodland mounds were excavated at the Schild site, with approximately 230
individuals represented in the series (Figure 5). Schild Late Woodland mound nine directly abuts
knoll A of the Schild Mississippian Cemetery (Figure 6). The majority of the burials were single
interments in circular graves with few or no grave goods. Although flexed burials were the most
frequently encountered, semiflexed and bundle burials were also common. The graves were
generally excavated into the ground surface prior to mound construction, with only few added
after the mound was completed. The Schild graves were subsequently refilled with the excavated
soil, in contrast to Late Woodland graves at sites like Koster, which appeared to be refilled with
clean loess. Stone lined graves and individuals covered with limestone slabs are also found at
Schild, although this type of burial is more common at the Yokem site.
Rectangular pits associated with the Late Woodland mounds were interpreted as potential
charnel structures, from which processed bodies were periodically removed and interred in the
immediate vicinity. The subfloor tombs likely served a similar purpose, a proposition that is
supported by the scatter of isolated human bones around the immediate vicinity. Perino discusses
the “crematories” located in several of the Schild mounds,. These 7x9 foot rectangular structures
were similar in size and shape to stone tombs found at Yokem, but at Schild they were
presumably constructed of some perishable material. As was the case at Yokem, these structures
were clearly burnt, but whether the burning event was accidental or purposeful is unclear.

73

Figure 5- Map of Schild Site (From Perino 1971)

74

Figure 6- Schild Late Woodland Mound Nine, showing spatial relationship to Mississippian
Knoll A (From Perino 1971)

Yokem (Perino 2006)
The Late Woodland component of the Yokem site consists of five burial mounds (numbered 4
through nine) constructed upon naturally occurring knolls (Figure 7). Several of the mounds
contained tombs made of limestone slabs and logs. The above ground tombs had doorways and
the floors were prepared so that the surface was about six inches deeper than the outside ground
75

level. Others were constructed by excavation of a subsurface crypt, which was subsequently
lined with stone slabs and roofed with logs. At nearby, contemporaneous sites (such as Koster),
remains were periodically removed from wooden charnel structures and reinterred in the vicinity.
At Yokem, however, it appears that remains were allowed to accumulate. Some of the tombs
show evidence for burning, although it is unclear whether this burning was intentional or
accidental. Regardless, , the tombs were filled with soil after the burning event. At the nearby Joe
Gay site, two similar structures are interpreted as charnel houses, and there is evidence that
bodies were cremated within this structure and subsequently interred in surrounding mounds
(Cook and Palcovich 2006).
Other Late Woodland mounds at Yokem did not contain stone tombs and may date to a
slightly later period. Stone remained an important part of the burial program, as evidenced by the
presence of limestone and shale slabs over many burials. Individual and multiple burials were
excavated into the natural ground surface, while others were incorporated into the mound fill.
Like many Late Woodland mortuary sites in the region, few contained any durable grave
accompaniments. The most common inclusions were Anculosa shell beads, presumably as
jewelry or clothing embellishments. Several of the mounds contained charnel pits, each of which
contained disarticulated remains. Perino interprets the charnel pit associated with Yokem Mound
four as the initial mortuary event for the mound. He also mentions the presence of charred bone
fragments that appeared to have been scattered over the central region of two mounds at the site,
a phenomenon he observed at other Late Woodland mounds (he specifically mentions Koster).
Also notable in mound four at Yokem is the interment of an older male, accompanied by a pipe,
a bone tool, and covered with a series of deer antlers. This individual may have had a unique
social role, or his interment may have been ritually meaningful in some other way. Citing the

76

progression of smaller (younger deer) to larger (older deer) antler racks as one moved from head
to feet, Perino suggests that deer ceremonialism may have played a role in this arrangement.

Figure 7- Map of Yokem Site (From Perino 1971)
Mississippian Components
Schild

77

The Schild Mississippian component was originally excavated by Perino in the 1960’s (along
with the Late Woodland component) and likely dates to Stirling and Moorhead phases (Perino
1971a, Goldstein 1980). Over 300 burials were recovered from two distinct burial areas (Figure
8). Perino’s field observations indicate his suspicion that apparent spatial grouping and artifact
associations were related to social roles and that fire and/or cremation may have played
important part of mortuary ritual at this site.

Figure 8- Schild Mississippian Component (From Goldstein 1980)

78

Goldstein (1980), in her study of Mississippian mortuary practice at Schild and the
nearby Moss Cemetery (not part of the current analysis), hypothesized that bounded cemetery
areas are likely indicative of corporate group structure with lineal descent. Initial artifact analysis
indicated some differentiation in burial treatment, based loosely on sex and age, with the most
restrictive associations limited to a few individuals. Distinct rows of burials, as indicated by
spatial analysis, further defined more restrictive versus more inclusive burial treatments within
the cemeteries, while also allowing for investigations of group membership (i.e. kinship) beyond
that which is indicated by the artifact analysis alone (Figure 9).

79

Figure 9- Schild Knoll A rows (from Goldstein 1980)

80

Although traces of a structure were not clearly visible during excavations, the discrete
arrangement of certain burials and the presence of charred plant materials and limestone set two
groups of individuals off from the prevailing row-based arrangement of the site. Based on this
evidence, Goldstein suggests that a charnel facility is indicated (Figure 10). For Knoll A, it
appears that the charnel structure served as a focus around which a group of burials were
arranged. It is unclear which bodies were processed through the charnel house before interment,
although Goldstein points out that articulated burials that were afforded the most exclusive burial
treatments were often interred with parts of individuals who may have been processed through
the facility. The structure was eventually burnt and scattered, with some bundled individuals and
charred plant remains deposited on top before the area was covered with a low mound of earth.
The Knoll B structure appeared to go through the same general sequence, although it did not
appear to have been capped with a mound upon burning (Figure 11). It is important to note that
not all of the burials in either knoll seem to be directly related to the charnel facilities, and
Goldstein suggests that this may be an important clue to social divisions within the communities
that used the site.

81

Figure 10- Charnel structure and hypothesized burial divisions in Schild Knoll A (from
Goldstein 1980)

82

Figure 11- Charnel structure and hypothesized burial divisions in Schild Knoll B (from
Goldstein 1980)

Mortuary treatments at Schild are certainly different from those observed at Cahokia and
other large centers, but Goldstein’s analysis indicates that the level of social complexity within
these small, outlying communities is in keeping with general expectations of a ranked and kinbased Mississippian social system. Larger sites tend to incorporate the basic elements observed
at smaller sites. Goldstein’s model for Mississippian mortuary practices emphasizes spatial
83

arrangement as an important indicator of social differentiation and constitutes a critical
component of the research outlined in this dissertation.

Yokem
Like Schild, the Yokem site has both Late Woodland and Mississippian components (Perino
1971b). Over 100 burials at this site were attributed to the Mississippian occupation of the area,
which likely dates to slightly later than the Mississippian component at Schild. The
Mississippian mounds at Yokem are numbered 1, 2, and 3 (Figures 7 and 12)
Although Yokem has not been subject to the same type of rigorous spatial analysis as
Schild, Goldstein (1980) notes several key features of this site in her model of Mississippian
mortuary practice. Natural knolls and an existing Late Woodland mound at the site were
accentuated with additional earth to construct three low mounds. She suggests that each of the
three mounds may represent a corporate or descent group and that the Yokem mounds seem to be
used by a more localized community compared to Schild, which likely represents the cooperative
mortuary activities of several dispersed communities.

84

Figure 12- Yokem Mississippian Mound 2 showing charnel structure (from Perino
1971b)

Unlike Schild, Yokem has preserved structural evidence for charnel facilities, specifically
post molds and delineated floors (Perino 1971b). Charnel features were built upon original
surfaces that had been prepared prior to the main mound construction events. Two of these
buildings showed evidence of burning. Rows of graves containing both primary and bundled
interments were placed around these charnel structures as well as throughout the rest of the
mounds. One of the mounds contained a row of articulated and bundled individuals, while the
other two contained scattered human bones. One of the structures contained the partially
disarticulated remains of an adult, which Perino interprets as a burial that had been disturbed by
subsequent charnel activity (Perino 1971b).

85

Previous Bioarchaeological Research
Several important bioarchaeological studies have incorporated data from the Schild and Yokem
series and will be considered when interpreting the results of this study. Cook, Buikstra, and
their colleagues have extensively studied the Lower Illinois Valley series from the periods in
question to better understand the origins of diseases such as tuberculosis and syphilis in the New
World (Braun et al 1998; Buikstra 1992; Buikstra and Cook 1978, 1981; Cook 1976, 1994,
2005). These studies have obvious implications for evaluating the general health of the
populations that are part of this study, but these specific pathologies have not been directly
incorporated into the current analysis. It is important to note, however, that disease processes
associated with tuberculosis can affect the vertebral column; such affected individuals were
excluded from the current analysis.
Other studies have more direct implications for this dissertation, particularly those that
explicitly address sex and status based differences in the expression of various biological
indicators of stress, growth, health, and activity. Bridges et al (2000) suggest that Schild
Mississippian females had decreased arm strength as compared to their Late Woodland
predecessors, perhaps indicating advances in harvesting and food processing technologies.
Although these types of cross sectional geometry studies have been criticized from both
methodological and interpretive perspectives (Jurmain 1992, Bice 2003), the basic association
between women and agricultural pursuits in the prehistoric Eastern Woodlands is generally
agreed upon. Also related to agricultural perspectives is Buikstra et al’s (1986) use of the Schild
sample to show increasing fertility rates over time, suggesting that Mississippian women were
consciously making decisions regarding birth-interval based in part on access to adequate
weaning foods.

86

Rose (2007) contributes an isotopic perspective to this literature in her broad temporal
analysis of dietary trends in West Central Illinois. She notes that members of the Schild Late
Woodland community seems to have consumed somewhat less maize than their contemporaries
throughout the region, including those individuals interred at Yokem. Regarding the
Mississippian components, only Schild Knoll A showed significant sex differences in carbon
isotope ratios and nitrogen values (females having enriched carbon isotope values and lower
nitrogen values), apparently indicating a diet with more maize and less meat than their male
counterparts.
Hamilton (1982) shows that sexual dimorophisms in several skeletal features decreases
from the Late Woodland to the Mississippian at Schild and other West Central Illinois sites,
suggesting a relative increase in female nutritional or health status. Cook (1984) evaluates femur
length relative to dental age in a sample that includes juveniles from both Schild components,
and concludes that there is a general trend toward improving growth conditions for children with
Mississippianization. However, because juveniles cannot be sexed, it is impossible to determine
whether male and female children were differentially affected. She also notes a relative decrease
in the prevalence of microscopic enamel defects among Schild Mississippian females compared
to males when Schild is compared to earlier Middle Woodland samples in the broader region.
Furthermore, LEH frequency among Schild Mississippians does not appear to be related to social
status. Cook also points out, however, that there appears to be a bias toward females in the Schild
Mississippian cemetery, perhaps indicating that higher status males were interred elsewhere.
Genetic analyses are also important for this research; the general consensus on temporal
and regional biological continuity adds an important level of genetic control for the comparisons
made in this study. Conner’s (1990) analysis of non-metric cranial traits indicated significant

87

intercommunity gene flow in his West Central Illinois sample, which included both Schild and
Yokem. Steadman’s (2001) craniometric data, interpreted within a more sophisticated population
genetics framework, suggests that, while there was significant gene flow among Late Woodland
communities in the Lower and Central Illinois Valleys, the Mississippian pattern indicated that
gene flow was more restricted within each region. Droessler’s (1981) investigation of biological
distances, based on Lower Illinois Valley cranial measurements, also indicated genetic continuity
between Late Woodland and Mississippian populations in this region.
Recent mtDNA analysis suggests that there may have been limited immigration into the
Lower Illinois Valley, possibly from the American Bottom, with Mississippianization (Raff
2008). In a related analysis, Raff (2008) shows that rows and spatial groups at Schild do not
seem to be based on biological matrilineal descent. However, Raff also notes that notions of
kinship or other ways of relating oneself to (or differentiating oneself from) other segments of a
community (i.e. clans, moieties, sodalities, etc.) are not purely biological, and DNA analyses
alone cannot preclude kinship as an organizing principle in the use of mortuary spaces.

88

CHAPTER FIVE- ANALYTICAL METHODS, MATERIALS, AND HYPOTHESES

In order to address the research questions proposed in chapter one, a set of hypotheses must be
developed in a way that provides a framework for simultaneously testing the vertebral method
and addressing the bioarchaeological concerns of this research given the available skeletal
samples. Data on four specific indicators of stress and growth are used in this analysis: vertebral
body height, vertebral canal diameter, linear enamel hypoplasia, and long bone length. The
nature of this study requires limiting the sample to only those adult remains for whom sex could
be accurately determined. The usable sample size will therefore vary between the different
components of the analysis; general summary tables are included at the end of this chapter.
Sample sizes for specific statistical tests are provided in chapter six and in the appendix.
A diachronic analysis is used to determine whether changes in growth patterns within
either sex occurred over the Late Woodland-Mississippian transition. The osteological data are
further analyzed with reference to mortuary data to determine whether sex- and age-related
biological growth patterns vary with burial treatments and spatial patterning, allowing for a more
specific assessment of Mississippian gender phenomena at Schild. For this portion of the
analysis, of particular interest is the extent to which variation in the osteological data is
correlated with sex, status, and kinship, or community membership. Results are interpreted
within the framework of contemporary gender scholarship and existing archaeological and
bioarchaeological reports from the region. Temporal trends are compared to published reports
from the American Bottom (Milner 1982) and Dickson Mounds (Clark 1985, Goodman et al

89

1984) in order to identify any temporal (Late Woodland versus Mississippian) or regional (center
versus periphery) trends in stress/growth discrepancies by sex.

Data Collection Methods
Sexing Methods
Accurate sexing methods are critical for making appropriate interpretations of biological and
cultural data. Sex is a basic unit of bioarchaeological analysis and is of particular importance
when attempting a properly engendered bioarchaeological analysis. There are no reliable
techniques for sexing subadult skeletal remains, as the most commonly employed qualitative
methods for sexing skeletal remains are based on secondary sexual characteristics of the pelvis
and cranium. This is a significant limitation, resulting in a notable lack of engendered
bioarchaeological studies of childhood, as well as a disregard for subadult analysis in general
(Baker et al 2005). One limitation inherent to ascertaining sex from adult skeletons is that the
extent of sexual dimorphism varies by population, which limits the applicability of methods
developed on modern, documented samples and introduces a significant source of error. Despite
these concerns, osteologists routinely assign sex to prehistoric skeletonized remains with a high
level of reported accuracy (Ubelaker and Buikstra 1994, Phenice 1969).
For the samples involved in this study, sex determinations were previously made by Dr.
Della Cook and her colleagues at Indiana University. Genuinely indeterminate cases were not
included in the current analyses, however those individuals categorized as “probable females”
and “probable males” were included.

90

Vertebral Measurements
To evaluate vertebral growth, measurements of vertebral body heights and neural canal
diameters were taken on all available thoracic and lumbar vertebrae. Measurement standards are
provided by Clark (1985). Clearly pathological vertebrae (i.e., collapsed or particularly arthritic)
were excluded from analysis. Body height measurements were taken on the anterior and
posterior aspects, and both anterior-posterior and transverse neural canal diameters were
recorded (Figure 13). A more detailed explanation for each dimensional measure is provided
below.

Figure 13- Vertebral dimensions (from Clark 1986) (Note that nerve root tunnel measurements
were not part of the present study)

Anterior vertebral body height (AVBH): AVBH is defined as the maximum height at the anterior
border of the vertebral body, as taken on the superior and inferior annular rings. Using the
outside jaws of the vernier caliper, measurements were taken by placing the body of one jaw

91

along the anterior aspect of the superior annular ring and moving the other jaw to meet the most
anterior aspect of the most inferior annular ring. Osteophytic growths were avoided.

Posterior vertebral body height (PVBH): PVBH is the maximum height at the posterior border of
the vertebral body, as taken on the superior and inferior annular rings. Using the outside jaws of
the vernier caliper, measurements were taken by placing the body of one jaw along the posterior
aspect of the superior annular ring and moving the other jaw to meet the posterior aspect of the
inferior annular ring. Osteophytic growths were avoided.

Transverse Neural Canal Diameter (TNCD): TNCD is the maximum transverse diameter at the
widest point of the neural canal as taken from the superior aspect, measured using the inside jaws
of the vernier calipers. The locus at which this measurement is taken varies anteriorly-posteriorly
based on the specific location of the particular vertebra in the spinal column (i.e. lumbar
vertebral canal diameters are widest more anteriorly than are those of thoracic vertebrae).
Calipers were gently “wiggled” around so that maximum diameter could be recorded.

Anterior-Posterior Neural Canal Diameter (APNCD): APNCD is the maximum anteriorposterior diameter of the neural canal as taken from the superior aspect, measured using the
inside jaws of the vernier calipers. This distance is typically at the midline, between the center
point of the posterior vertebral body and the point where the right and left laminae fuse to form
the base of the spinous process. Care must be taken to hold the calipers so that the sliding
horizontal portion is parallel with the superior aspect of the vertebral body, as irregular
angulations produce inconsistent measurements.

92

Linear Enamel Hypoplasia (LEH)
LEH were observed under a 100 watt incandescent desk lamp. LEH was recorded as present
when a clearly observable line was apparent on the labial surface. Non-linear abnormalities (i.e.
pits) were not recorded. Additionally, the number of hypoplasias per tooth was recorded, as was
distance of each LEH from the cemento-enamel junction. This distance was measured in
millimeters using a graduated hand lens.

Femur Length
Maximum femur length was measured to the nearest half centimeter for all left and right femora
using a standard osteometric board according to standard osteological data collection methods
(Ubelaker and Buikstra 1994).

Controlling for Error
All continuous measurements were taken at least twice. For femur length, measurements were
taken to the nearest half centimeter. A random sample of 20 femora were remeasured to test for
intraobserver error, and no inconsisitencies were detected. Vertebral measurements proved to be
more problematic than femoral measurements. Clark (1984) provided methods for all vertebral
measurements, which were taken to the nearest tenth of a millimeter (see descriptions above for
measurement details). To develop proficiency and intraobserver consistency, unrecorded practice
measurements were taken. During actual data collection, two measurements were initially taken.
If the first two measurements were not in agreement, the skeletal element in question was set
aside and remeasured at a later time. In cases of continued measurement problems where three

93

measurements did not yield consistent results, the average of three measurements was recorded.
At the start of each day, the last specimens from the previous day were remeasured to achieve
consistency in placement and pressure of the calipers.
To test for error in LEH assessment, a random sample of 25 individuals was evaluated,
and then reevaluated. This test revealed one case in which an LEH was originally marked as
present, but was subsequently marked as absent.

Schild Mortuary Data
As one of the key methods by which archaeologists study social differentiation, mortuary
analysis is an important line of evidence for interpreting sex-based biological data from a
gendered perspective. Incorporation of mortuary data was more readily accomplished for the
Mississippian component of the analysis. Mortuary data from the Late Woodland component is
presently limited to Perino’s field observations. Comparison of Late Woodland and
Mississippian mortuary samples allow for investigation of change over time in patterns of
stress/growth by sex and age and will be useful from a general interpretive perspective, but the
more socially in-depth portion of this analysis of limited to the Schild Mississippian component.
Goldstein’s (1980) spatial analysis of mortuary practices provides critical information
regarding the level of social differentiation in the Mississippian component of the Schild site.
She showed that the Schild Mississippian burials displayed a generally egalitarian, grouporiented (rather than individualized) mortuary pattern. Although the overall Schild Mississippian
pattern can be characterized as egalitarian, both Goldstein (1980) and Rothschild (1990) point to
a pattern of restricted access to certain burial treatments, often based on age and sex, in
Mississippian mortuary practices throughout the broader region. Some of the burial treatments

94

detected in Goldstein’s cluster analysis were more restrictive than others in terms of the
proportion of the population afforded them. Goldstein’s (1980) work on the Schild Mississippian
component also detected spatially discrete units (rows), which she interpreted as representing kin
groups. Although a similar analysis is not available for the Yokem Mississippian sample,
Goldstein (1980) pointed out that row patterning is also evident at that site and that each of the
three mounds may represent a corporate group.
Analysis of the relationship between biological data and archaeologically-derived spatial
data is complicated by sample size concerns. To maximize cell counts and increase statistical
amenability, the original rows and clusters proposed by Goldstein (1976, 1980, 1981) as
indicated in Table 2 and the descriptions below.

95

Mortuary category
Knoll

Subcategories
A

Description
Burials in northern knoll at Schild Mississipian
cemetery as defined by Perino (1971) and
Goldstein (1980)- See figure

B

0

Burials assigned to the four clusters separated
by the primary division in Goldstein’s (1980)
cluster analysis, based on arm extension.
Burials assigned to the four clusters separated
by the primary division in Goldstein’s (1980)
cluster analysis, based on arm non-extension.

1

Non-charnel associated burial area defined by
Goldstein’s (1980) hypothesized division of
knoll B burials- see figure
Charnel associated burial area defined by
Goldstein’s (1980) hypothesized division of
knoll B burials- see figure

3

Charnel associated burial area defined by
Goldstein’s (1980) hypothesized division of
knoll A burials- see figure

4

Charnel Association

1

2

Burial Area

Burials assigned to cluster 6, the most exclusive
of the Schild Mississippian artifact only clusters

2

Non-charnel associated burial area defined by
Goldstein’s (1980) hypothesized division of
knoll A burials- see figure

Yes

Charnel associated burial areas (burial areas 2
and 3)

No

Artifact and Positioning cluster

Burials assigned to the least exclusive clusters in
Goldstein’s (1980) analysis. This category
includes all clusters except for cluster 6.

6

Artifact only clusters

Burials in southern knoll at Schild Mississipian
cemetery as defined by Perino (1971) and
Goldstein (1980)- See figure

Non-charnel associated burial areas (burial areas
1 and 2)

Table 2- Mortuary category descriptions

96

Artifact-Only Clusters- Goldstein’s analysis indicated eight clusters based on similarities in
grave accompaniments. Sample size issues precluded an analysis based directly on these clusters,
therefore they were collapsed into two groups- cluster six was maintained, but all others were
collapsed into a single group. Most burials were assigned to cluster six, and can therefore be
considered the least exclusive group. The other, more exclusive clusters comprised the other
group. These groups are referred to as six (6) and zero (0), respectively.

Artifact and positioning clusters- Goldstein’s analysis by artifact and body position indicated 10
clusters. The primary division was based on extension versus non-extension of arms, with
subsequent divisions based on other positional and artifact parameters. For the current analysis,
the results of this cluster analysis were collapsed into two groups of roughly equal size, based on
the primary positional division. These groups are referred to as group one (1) (Goldstein’s
original cluster numbers 1,8,3,5, and 7) and group two (2) (Goldstein’s original cluster numbers
2,4,6,9, and 10) in tables and statistical output.

Areas and Charnel Associations- The spatial portion of Goldstein’s mortuary analyses indicated
16 rows, but sample size issues precluded a by row analysis of this data. However, Goldstein
recognized other spatial divisions within the site (see map), and these “burial areas” correspond
well with both the burial areas recognized in Perino’s excavations, as well as clusters of the
original rows from Goldstein’s analysis.
For Knoll A, Goldstein’s hypothesized division of the burials essentially separated one
curvilinear row from the other burials, most of which were associated with the charnel structure
and capping mound. The curvilinear row is referred to as group four (4) in this analysis, while

97

the charnel/mounds associated burials are referred to as group three (3). The few burials that
were not associated with either the curvilinear row or the charnel structure/mound were
considered to be part of group four (4) (Figure 14).

Figure 14- Schild Burial Areas 3 and 4 (from Goldstein 1980)

For Knoll B, Goldstein’s hypothesized division also delineated a separation between
charnel and non-charnel areas (Figure 15). The charnel associated area (group 2) is not capped
98

with a mound as is the case in Knoll A. However, the burials not directly associated with the
charnel area in Knoll B are strongly aligned with the structure, and therefore considered to be
part of the same group. In both knolls, burial areas delineated by Goldstein inherently correspond
to groups of rows, presumed charnel structure, and Perino’s original groupings, and it is
suggested here that the spatial groupings used in this study account for all of these spatial
consideration in a culturally meaningful way.

Figure 15- Schild Burial Areas 1 and 2 (From Goldstein 1980)

99

Statistical Methods for Incorporating Osteological and Archaeological Data
This study involves both continuous and categorical data. Investigation of the interrelationships
between growth patterns drawn from the osteological data and mortuary data drawn from the
archaeological literature involves both types of data to be analyzed simultaneously. All tests are
performed using the SPSS 19 statistical package.

Cronbach’s Alpha
In his original analysis, Clark (1985,1988) used a coefficient alpha test for intrasegment
relaibaility to show that vertebral dimensions within each segment were related to each other in a
regular pattern (i.e. increase in size predictably as one moves downward), justifying use of the
intrasegmental mean for evaluating changes in size over time (Clark 1988). SPSS allows for
Cronbach’s Alpha testing to provide a measure of intrasegment reliability. A high alpha score
indicates that measurements of any one vertebral dimension (thoracic AVBH, for example) in
any one vertebra (The first thoracic vertebra (T1), for example) is predictably related to all other
thoracic AVBHs (T2 through T12). An alpha score is obtained by comparing intrasegmental
patterns in individuals to intrasegmental patterns in the sample as a whole. In standard practice,
an alpha score over 0.7 is acceptable, over 0.8 is good, and over 0.9 is excellent, on a scale from
0 to 1 (George and Mallery 2003).
Use of this technique requires exclusion of individuals with missing thoracic or lumbar
vertebrae, i.e., fewer than 12 or five, respectively. Throughout this dissertation, the term
“segment-based” refers to tests in which measurements for each vertebrae in a segment were
averaged for each individual, and a single mean value used for statistical comparison between
time periods and mortuary groups. For hypothesis testing, using segmental means based analysis

100

limits sample sizes. However, the benefit is that each individual is represented by one variate
rather than many, and that may reduce overall noise in the data resulting from individual
idiosyncrasies or error. Alpha scores were also useful for finding and correcting data input
errors.

Linear Regression
Linear regression analyses are used to gauge the relationship between vertebral dimensions and
femur length. One of the premises of the vertebral method is that vertebral body height is a
measure of cumulative growth in the same way femur length has been traditionally used in
bioarchaeological studies, so there should exist a significant, positive linear relationsip between
these measures. Because neural canal diameter (NCD) measurements are hypothesized to be
indicative of early childhood (as opposed to cumulative) growth conditions, there may or may
not be a linear relationship between femur length and neural canal diameters. If NCD is not
significantly correlated with femur lengths, these two measures may be measuring growth
conditions experienced during different time periods (i.e. early childhood-only versus
cumulative). If NCD is positively and significantly correlated with femur length, this does not
necessarily prove that the two measures are not measuring growth conditions covering different
time periods, as it may alternatively indicate that there was not a general trend toward differential
growth trajectories between the early childhood and late adolescence periods.

T-tests
Standard t-tests are used for the majority of the vertebral and femur length comparisons, both
over time and between Mississippian mortuary groupings. Femur tests were relatively simple and

101

straightforward, as each individual was represented by only one measurement. However, the
multiplicity of vertebrae and the more complex nature of statistical investigation of the vertebral
data made for a more complicated scenario. For each vertebral segment, the mean of each
dimension (AVBH, PVBH, TNCD, APNCD) was calculated for each sex within each cultural
period and t-tests were used to evaluate change over time between the Late Woodland and
Mississippian periods. Low sample sizes for the Late Woodland materials preclude separate
temporally based analysis of the two sites, so Yokem and Schild were considered together for
this portion of the study. Comparisons of segmental means are also performed to compare
individuals based on Goldstein’s mortuary parameters where there were only two categories
(knolls, clusters, charnel association).
After segment-based testing, tests were performed for each individual vertebral
dimension (e.g. comparing T1 AVBH’s only to other T1 AVBH’s, L2 TNCD’s only to other L2
TNCD’s, etc). This allowed me to isolate the combinations of vertebral elements that were
responsible for significant results yielded by the more general average comparisons described
above. Both of these analytical techniques are useful and necessary for investigating the utility of
vertebral measurements as growth indicators because 1) they address both generalized and
specific patterns of spinal development, 2) comparison between the two might allow for isolation
of vertebrae that are more or less sensitive to growth disruption, complementing the available
clinical literature on the timing of vertebral growth. This may also be useful for developing an
analytical technique that can be applied in situations where taphonomic and cultural processes
may result in a skewed or incomplete vertebral sample.
As the other biological parameters were less complicated from an analytical perspective
(one LEH measurements and one femoral length measurement for each individual as opposed to

102

up to 68 vertebral measurements per individual) they did not require the same level of
reorganization and retesting. For these measures, each test (for change over time and the various
differences between Schild Mississippian mortuary categories) was performed only once for
males and once for females.
Results were considered significant at the α=.05 level, but because of small sample sizes,
results that approached significance are also noted, briefly discussed, and cautiously interpreted.

ANOVA
Two-way ANOVA was used to explore the nature of differences in growth indicators between
different burial areas. Multiway ANOVA tests are used to evaluate interactions between multiple
categorical mortuary parameters (factors) and continuous growth indicators, and significant
combinations were plotted.
Results were considered significant at the α=.05 level, but because of small sample sizes,
results that approached significance are also noted, briefly discussed, and cautiously interpreted.

Pearson’s Chi-square
Because LEH is considered only as a categorical/nominal measure (present or absent) in this
analysis, contingency tables and chi-square analyses are used to identify differences in frequency
by sex over time and between mortuary groupings. As with tests described above, results were
considered significant at the α=.05 level, but because of small sample sizes, results that
approached significance are also noted, briefly discussed, and cautiously interpreted.

103

Hypotheses
Seven specific hypotheses were tested in this study. These were based in part on much of the
West Central Illinois bioarchaeological literature already discussed in preceding chapters.
Hypotheses one through four are related to change over the Late Woodland-Mississippian
transition, while hypotheses five addresses expectations of differentiation in growth patterns
based on social categories among the Schild Mississippians.

Intrasegmental Reliability to Justify Segmental Means Testing
Hypothesis One- As a preliminary test, it must be determined whether the vertebral
measurements within either segment (thoracic or lumbar) are reliabably and predictably related
to one another to a statistically significant extent (see discussion of Cronbach’s alpha above).
The hypothesis is stated as follows:
Hypothesis 1:

H0: Intrasegmental reliability is not significant at the α=.7 level
H1: Intrasegmental reliability is significant at the α=.7 level.

Testing the Vertebral Method
Hypothesis Two- In order to test whether the vertebral method is a useful indicator of early
childhood and vertebral growth, a series of hypotheses will be tested. These are based on an
assumption that, if VBH can be used as an indicator of cumulative growth, there should be a
significant linear relationship between femur length and VBH. Furthermore, if NCD is a growth
indicator that reflects a growth period restricted to early childhood only, there will not
necessarily be a linear relationship between femur length and NCD. Also, if NCD is a growth
indicator that reflects a growth period restricted to early childhood growth before sexually

104

dimorphic characteristics develop, one would not necessarily observe a typical pattern of sexual
dimorphism whereby males will be larger than females, as this feature develops prior to the onset
of genetically controlled, sex specific growth differentials. As such, the level of sexual
dimorphism in neural canal diameter is expected to change to reflect improving female growth
conditions as we move from Late Woodland to Mississippian times. The hypotheses can be
stated as followed:
Hypothesis 2a:

H0- A linear relationship does not exist between VBH and femur
length.
H1- A linear relationship exists between VBH and femur length.

Hypothesis 2b:

H0- There is no linear relationship between NCD and femur length.
H1- There is a linear relationship between NCD and femur length.

Hypothesis 2c:

H0- There is no difference between male and female NCD
measurements within cultural periods.
H1- Male NCD is consistently larger than female NCD within
cultural periods.

Trends in LEH Frequency Over Time
Hypothesis Three- Hypothesis two is designed to test whether there is temporal change in LEH
frequencies over time for either males or females over the Late Woodland- Mississippian
transition. Based on literature that points increased reliance on maize agriculture facilitated
primarily by women, which may have lead to female children being more valued in society and
hence afforded better care and improved growth conditions (Cook 1984, Cohen and Bennet
1993, Goodman et al 1980), there may be a temporal change in LEH frequencies among females
in the study sample, with frequencies decreasing from the Late Woodland to the Mississippian
period. Lack of corresponding changes in LEH frequency over time among the males has been
105

interpreted as indicating that increased reliance on maize agriculture disproportionally affected
women in Mississippian society. The hypotheses to be tested to investigate this phenomenon are
stated below.
Hypothesis 3a:

H0- Male Late Woodland LEH frequencies are equal to male
Mississippian LEH frequencies.
H1- Male Late Woodland LEH frequencies are unequal to male
Mississippian LEH frequencies.

Hypothesis 3b:

H0- Female Late Woodland LEH frequencies are equal to female
Mississippian LEH frequencies.
H1- Female Late Woodland LEH frequencies are unequal to female
Mississippian LEH frequencies.

Trends in Femur Length Over Time
Hypothesis Four- Based on Cook’s (1984, 2007) observations, there was no statistically
significant difference in femur length among either males or females between the Late Woodland
Mississippian periods. Statistical analyses will test the following set of hypotheses to address this
issue.
Hypothesis 4a:

H0- Male Late Woodland femur length measurements are equal to
male Mississippian femur length measurements.
H1- Male Late Woodland femur length measurements are unequal
to male Mississippian femur length measurements.

Hypothesis 4b:

H0- Female Late Woodland femur length measurements are equal
to female Mississippian femur length measurements.
H1- Female Late Woodland femur length measurements are
unequal to female Mississippian femur length measurements.

106

Vertebral Growth Trends Over Time
Hypothesis Five- Clark (1985, 1988) suggests that neural canal diameter dimensions capture
early childhood growth conditions. If increased status of women associated with the transition to
Mississippian lifestyle may result in changes in growth patterning for men and women. Like
LEH (discussed above) improved early childhood growth conditions for women over the Late
Woodland Mississippian transition might become visible as a change in female neural canal
dimensions over time, with relative female neural canal diameter measurements increasing from
the Late Woodland to the Mississippian period. There may or may not be corresponding changes
in neural canal dimensions among the males in this study, depending on whether they were
differentially affected by the cultural changes associated with this transition. Clark (1985, 1988)
also suggests that VBH measures cumulative growth, a parameter that is more traditionally
measured using long bone lengths (see hypothesis four above). As such, VBH and femur length
may be measuring the same cumulative growth phenomenon.
These ideas will be tested using the following set of hypotheses:
Hypothesis 5a:

H0- Male Late Woodland VBH measurements are equal to male
Mississippian VBH measurements.
H1- Male Late Woodland VBH measurements are unequal to male
Mississippian VBH measurements.

Hypothesis 5b:

H0- Male Late Woodland NCD measurements are equal to male
Mississippian NCD measurements.
H1- Male Late Woodland NCD measurements are unequal to male
Mississippian NCD measurements.

Hypothesis 5c:

H0- Female Late Woodland VBH measurements are equal to
female Mississippian VBH measurements.

107

H1- Female Late Woodland VBH measurements are unequal to
female Mississippian VBH measurements.

Hypothesis 5d:

H0- Female Late Woodland NCD measurements are equal to
female Mississippian NCD measurements.
H1- Female Late Woodland NCD measurements are unequal to
female Mississippian NCD measurements.

Intracommunity Variation in Growth Patterns in the Schild Mississippian Component
Hypothesis six- Mississippianization of West Central Illinois brought with it clear changes in
mortuary patterning (Goldstein 1980) and although community/household life has proven
difficult to investigate archaeologically due to the nature of the Mississippian presence in this
region (Farnsworth et al 1991), the introduction of new Mississippian ideals must have had a
significant effect on intracommunity social relations. New types of social differentiation can be
inferred from mortuary analysis, and the research outlined in this dissertation can help to detect
the ways that new ideas about social differentiation affected biosocial life experiences within the
community. Specifically, it is hypothesized that that analysis of skeletal data according to
archaeologically defined mortuary groupings (artifact clusters, artifact-positioning clusters, and
burial areas) will new yield insights regarding biological growth patterning, indicating whether
or not certain subsets of the Schild Mississippian community enjoyed higher status and, thus,
better growth conditions.
This set of hypotheses specifically addresses the ways in which those individuals of
either biological sex may have had different life experiences based on cultural perceptions of the
interaction between sex and other (non-sexual) facets of social identity. This is important
because it removes the unfounded a priori assumption of a simplistic, unitary gender phenomena

108

and forces a consideration of the range of experiences of gendered social identity within a
community. When the intersection of sex, kinship, and status are investigated with regard to
growth experiences, the ways in which various social factors combine to contribute to gender
identity at this place and time will become more clear.

By KnollHypothesis 6a1:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between males in knoll A and males in knoll B.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between males in knoll A and males in knoll
B.

Hypothesis 6a2:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between females between females in knoll A and
females in knoll B.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between females in knoll A and females in
knoll B.

By Burial AreaHypothesis 6b1:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between males based on burial area.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between males based on burial area.

Hypothesis 6b2:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between females based on burial area.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between females based on burial area.

By Charnel Association-

109

Hypothesis 6c1:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between males based on charnel association.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between males based on charnel association.

Hypothesis 6c2:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between females based on charnel association.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between females based on charnel association.

By Artifact only ClusterHypothesis 6d1:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between males in artifact only cluster 0 and males in
artifact only cluster 6.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between males in artifact only cluster 0 and
males in artifact only cluster 6.

Hypothesis 6d2:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between females in artifact only cluster 0 and females
in artifact only cluster 6.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between females in artifact only cluster 0 and
females in artifact only cluster 6.

By Artifact and Positioning ClusterHypothesis 6e1:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between males in artifact and positioning cluster 1
and males in artifact and positioning cluster 2.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between males in artifact and positioning
cluster 1 and males in artifact and positioning cluster 2.

110

Hypothesis 6e2:

H0- There will be no difference in VBH, NCD, LEH frequency, or
femur length between females in artifact and positioning cluster 1
and females in artifact and positioning cluster 2.
H1- There will be a difference in VBH, NCD LEH frequency,
and/or femur length between females in artifact and positioning
cluster 1 and females in artifact and positioning cluster 2.

Hypothesis 7- This hypothesis is designed to test for a relationship between each growth
indicator and all possible combinations of one spatial variable and one cluster variable for each
sex using two way ANOVAs. A significant relationship would suggest that spatial and cluster
variables interact in a way that is related to sex-based intracemetery growth patterns. In order to
maximize cell size, only binary spatial categories (knoll and charnel association) will be used.
The main hypotheses guiding this portion of the research can be stated as follows:
Hypothesis 7:

H0- There is no significant interactions between spatial categories,
cluster categories, or growth parameters for Schild Mississippian
males or females
H1- There are significant interactions between spatial categories,
cluster categories, and growth parameters for Schild Mississppian
males and/or females.

Sample Sizes
For the Schild site, a total of 138 Late Woodland and 175 Mississippian numbered and sexed
adult individuals are available for this analysis. The Yokem sample consists of 115 Late
Woodland and 72 sexed Mississippian adults. Several factors limit the number of usable
individuals for any particular test, and sample sizes for specific tests are therefore highly
variable. Not all observations (left and right femur lengths, four teeth used for LEH, and 68 total
vertebral measurements) were observable in every individual, therefore, sample sizes varied for
each of the statistical tests presented in the results chapter. General sample size tables are
111

provided in Tables 3 through 8 below, and sample sizes for specific tests can be found in the
following chapter as well as in the appendix.

Site

Period

N

Schild

Late Woodland
Mississippian

46
130

Late Woodland
Mississippian

37
42

Yokem

Table 3- Overall sample sizes. N=number of individuals with at least one
measurement/observation in the current study

Burial Area
1
2
3
4

Sex
M
F
M
F
M
F
M
F

N
18
16
13
11
9
18
13
14

Table 4- Sample sizes for Schild burial areas. N=number of individuals with at least one
measurable/observable indicator

Knoll
A
B

Sex
M
F
M
F

N
28
35
32
37

Table 5- Sample Sizes for Schild Knolls. N=number of individuals with at least one
measurable/observable indicator

Charnel Association
Yes
No

Sex
M
F
M
F

N
27
29
31
30

112

Table 6- Sample sizes for Schild charnel association. N=number of individuals with at least one
measurable/observable indicator

Cluster
0
6

Sex
M
F
M
F

N
31
42
31
30

Table 7- Sample sizes for Schild artifact only clusters. N=number of individuals with at least one
measurable/observable indicator

Cluster
1
2

Sex
M
F
M
F

N
28
25
32
47

Table 8- Sample sizes for artifact and positioning clusters. N=number of individuals with at least
one measurable/observable indicator

Preservation was generally good for both series, and this is particularly true for the Schild
Mississippian component. However, some remains were too taphonomically damaged for
confident measurement, and those were excluded. Some of the youngest adults could be sexed
with a reasonable degree of confidence, but were excluded if the epiphyseal rings of their
vertebrae were not completely fused. Likewise, a few of the oldest individuals who experienced
the most extensive arthritic and other pathological changes were also excluded. In general, mild
to moderate osteophystic lipping did not interfere with taking vertebral measurements.

113

CHAPTER SIX- RESULTS

In this chapter, results of the data analysis are presented and organized by hypothesis. Each
hypothesis is referenced, and summary data is provided. The results of the statistical analysis are
presented concisely, and hypotheses are explicitly accepted or rejected. Interpretation and
contextualization is presented in the following chapter.

Hypothesis One- Internal Consistency for Segmental Mean Analysis
Cronbach’s Alpha scores indicate that there is a relatively constant relationship between the
various measurements within both the thoracic and lumbar vertebral segments (Tables 9 and 10).
This is not surprising, as even cursory macroscopic observation clearly indicates that vertebral
body height (VBH) and neural canal diameter (NCD) measurements do not vary randomly but
rather increase as one moves down the spinal column. The null hypothesis is accepted, and
subsequent vertebral tests are performed using both segmental means and individual vertebral
comparisons. .

Sex
Males

Females

Dimension
N
AVBH
45
PVBH
59
TNCD
53
APNCD
41
AVBH
96
PVBH
122
TNCD
108
APNCD
96
Table 9- Thoracic Intrasegment Reliability

114

Cronbach’s Alpha
.921
.942
.950
.953
.943
.952
.966
.967

Sex

Males

Females

Cronbach’s Alpha
(with L5)
AVBH
56
.905
PVBH
65
.893
TNCD
57
.917
APNCD
46
.730
AVBH
115
.876
PVBH
129
.897
TNCD
135
.781
APNCD
105
.876
Table 10- Lumbar Intrasegment Reliability
Dimension

N

Cronbach’s Alpha
(without L5)
.884
.903
.939
.794
.921
.902
.787
.907

Hypothesis Two- Utility of the Vertebral Method of Growth Assessment
Hypotheses 2a and 2b
The results of the linear regression analysis indicate a statistically significant (or near statistically
significant) relationship between vertebral body height measurements and femur length. The low
R-square values, however, indicate that the relationship between the two measures is not
exceptionally strong. The statistically insignificant p-values for TNCDs and APNCDs indicate
that neural canal diameter measurements are independent of femur lengths (Tables 11 and 12).

Vertebral Dimension
P
R-square
AVBH
.035
.124
PVBH
.003
.208
TNCD
.274
.030
APNCD
.473
.017
Table 11- Linear Regression for Schild Female Lumbar Vertebral Measurements and Femur
Length

115

Vertebral Dimension
P
R-square
AVBH
.070
.105
PVBH
.051
.117
TNCD
.486
.019
APNCD
.881
.003
Table 12- Linear Regression for Schild Male Lumbar Vertebral Measurements and Femur
Length

The null hypothesis 2a is rejected, because there is a significant linear relationship
between femur length and VBH. This indicates a weak association between increased vertebral
body height and increased femur length.
The null hypothesis 2b is accepted, because there is no significant relationship between
NCD and femur length. This suggests that the two measures are reflecting fundamentally
different growth phenomena, which may be related to differentiation between growth restricted
to early childhood (as indicated by NCD) and cumulative growth (as indicated by femur length).

Hypothesis 2c
Male and female TNCDs and APNCDs were compared directly to one another as an additional
test of the hypothesis that NCDs can be used as indicators of early childhood growth experiences
(Tables 13 and 14). The underlying assumption behind this test is that significant sexual
dimorphism in the size of most skeletal elements generally does not occur until later in the
growth period, and that it is therefore possible that males and females could lack sexual
dimorphism in neural canal diameter if this measure does, indeed, reflect early childhood growth
as opposed to cumulative growth.

116

Vertebral
Dimension

Sex

N

Mean

SD

P

Male
17
23.18
1.68
.006
Female
31
21.85
1.45
Male
12
15.21
1.10
APNCD
.514
Female
23
14.93
1.19
Table 13- Lumbar NCD T-Tests between Late Woodland Males and Late Woodland
Females
TNCD

Vertebral
Dimension

Sex

N

Mean

SD

P

Male
55
22.87
1.44
.154
Female
77
22.48
1.61
Male
51
15.18
1.06
APNCD
.0502
Female
69
15.58
1.12
Table 14- Lumbar NCD T-Tests between Mississippian Males and Mississippian Females
TNCD

The null hypothesis 2c is accepted. The results indicate that there was no sexual
dimorphism in APNCD measurements during the Late Woodland. Male APNCD was greater,
but the difference was not statistically significant. However, during the Mississippian period,
female APNCD was greater than male APNCD, and the result was statistically significant. For
TNCD, there was a statistically significant difference between males and females in the Late
Woodland, with males being larger. However, by the Mississippian period, the difference was no
longer significant. Although the results for TNCD and APNCD do not exactly parallel each
other, both cases indicate an increase in female NCD over time.

117

Hypothesis Three- Temporal Trends in LEH Frequency
Hypothesis 3a- Temporal Trends in Male LEH Frequency
Chi-square analysis of the LEH frequencies collected for this study indicated that Mississippian
males were significantly less likely to experience episodic stressors in early childhood growth
when compared to their Late Woodland counterparts (Table 15).

Period
N
LEH present
Fisher’s Exact
Late Woodland
40
27
.006
Mississippian
51
19
Table 15- Male LEH Frequency Chi-square test between Late Woodland and Mississippian

Based on the results of this analysis, the null hypothesis 3a is rejected. There is a
statistically significant difference between LEH frequencies in males between the Late
Woodland and Mississippian periods. Late Woodland males were more likely to have at least
one LEH than their Mississippian counterparts

Hypothesis 3b- Temporal Trends in Female LEH Frequency
A similar analysis of the female component of the sample did not indicate any significant
difference in the experiences of episodic early childhood growth disruptions (Table 16).

Period
N
LEH present
Fisher’s Exact
Late Woodland
50
21
.699
Mississippian
61
23
Table 16- Female LEH Frequency Chi-square test between Late Woodland and Mississippian

118

Based on these results, hypothesis 3b is accepted. There is no significant difference
between the number of Late Woodland females with at least one LEH compared with the number
of Mississippian females experiencing at least one LEH.

Hypothesis Four- Temporal Trends in Femur Length
Hypothesis 4a- Temporal Trends in Male Femur Length
A T-test between Late Woodland male femur length and Mississippian male femur length
indicated no statistically significant difference (Table 17).
Period
N
Mean
SD
P
Late Woodland
26
45.31
1.85
.942
Mississippian
59
45.27
2.24
Table 17- Male Right Femur Length T-Test between Late Woodland and Mississippian

Based on this result, the null hypothesis 4a is accepted. There is no evidence of a change
in male femur length over the Late Woodland- Mississippian transition.

Hypothesis 4b-Temporal Trends in Female Femur Length
A T-test between Late Woodland female femur length and Mississippian female femur length
indicated no statistically significant difference (Table 18).
Period
N
Mean
SD
P
Late Woodland
34
42.559
2.06
.607
Mississippian
78
42.353
1.90
Table 18- Female Right Femur Length T-Test between Late Woodland and Mississippian

119

Based on this result, the null hypothesis 4b is accepted. There is no evidence of a change
in female femur length over the Late Woodland-Mississippian transition.

Hypothesis Five- Temporal Trends in Vertebral Growth
The results of the vertebral growth analysis are presented here. Tables are divided by sex and
vertebral segment, and individual hypotheses are addressed in the text.
Hypothesis 5a- Temporal Trends in Male VBH
There is no statistically significant difference in AVBH or PVBH for either the lumbar or the
thoracic segment in the male portion of the sample (Tables 19 and 20). Based on these
observations, the null hypothesis 5a is accepted. There is no detectable change in male VBH over
the Late Woodland-Mississippian transition.
Hypothesis 5b- Temporal Trends in Male NCD
There is no statistically significant difference in TNCD or APNCD for either the lumbar or the
thoracic segment in the male portion of the sample (Tables 19 and 20). Based on these
observations, the null hypothesis 5b is accepted. There is no detectable change in male NCD
over the Late Woodland-Mississippian transition.

120

Dimension

Period
N
Mean
SD
P
Late Woodland 8
20.55
1.13
AVBH
.812
Mississippian
41
20.47
.831
Late Woodland 9
23.713
0.65
PVBH
.470
Mississippian
50
23.147
2.31
Late Woodland 7
18.011
1.52
TNCD
.351
Mississippian
44
17.609
1.00
Late Woodland 4
15.310
0.79
APNCD
.845
Mississippian
37
15.220
0.88
Table 19- Male Thoracic Dimensions Means T-Tests between Late Woodland and Mississippian

Dimension

Period
N
Mean
SD
P
Late Woodland 12
26.600
1.36
AVBH
.585
Mississippian
46
26.816
1.17
Late Woodland 18
29.199
1.07
PVBH
.198
Mississippian
27
28.796
1.17
Late Woodland 17
23.175
1.67
TNCD
.471
Mississippian
55
22.874
1.44
Late Woodland 12
15.208
1.10
APNCD
.935
Mississippian
51
15.180
1.06
Table 20- Male Lumbar Dimension Means T-Tests between Late Woodland and Mississippian

Hypothesis 5c- Temporal Trends in Female NCD
There is no statistically significant difference in AVBH or PVBH for either the lumbar or the
thoracic segment in the female portion of the sample (Tables 21 and 22). Based on these
observations, the null hypothesis 5c is accepted. There is no detectable change in female VBH
over the Late Woodland-Mississippian transition.

Hypothesis 5d- Temporal Trends in Female NCD
Among females in this portion of the analysis, significant or near significant results were
indicated for the two neural canal dimensions within the lumbar segment, with the lumbar
APNCD making the strongest case for an increase in neural canal growth in the Mississippian
121

period when compared to the Late Woodland (Tables 21 and 22). The homologous
measurements in the thoracic segment were not statistically significant. Based on this
observation, the null hypothesis 5d is partially rejected, as there is some evidence suggesting a
change in female NCD over the Late Woodland-Mississippian transition, with APNCD
increasing significantly with Mississippianization, and TNCD increasing nearly significantly
with Mississippianization.

Dimension

Period
N
Mean
SD
P
Late Woodland 14
19.132
0.80
AVBH
.504
Mississippian
51
18.965
0.83
Late Woodland 15
21.176
0.80
PVBH
.812
Mississippian
61
21.293
1.85
Late Woodland 14
16.252
0.82
TNCD
.110
Mississippian
52
16.737
1.04
Late Woodland 13
14.828
0.55
APNCD
.449
Mississippian
47
15.042
1.00
Table 21- Female Thoracic Dimensions Means T-Tests between Late Woodland and
Mississippian

Dimension

Period
N
Mean
SD
P
Late Woodland 20
26.202
1.24
AVBH
.902
Mississippian
54
26.160
1.35
Late Woodland 29
27.353
1.37
PVBH
.770
Mississippian
77
27.434
1.23
Late Woodland 31
21.848
1.46
TNCD
.059
Mississippian
77
22.483
1.61
Late Woodland 23
14.935
1.19
APNCD
.021
Mississippian
69
15.579
1.12
Table 22- Female Lumbar Dimensions Means T-Tests between Late Woodland and
Mississippian

122

Temporal Trends in Growth Disruption: Individual Vertebrae
The following analyses regarding individually-based vertebral tests are provided as
supplementary information, intended to assist in interpreting the results presented in the
segmental-based hypotheses testing presented in the above temporal analyses. This is particularly
important for providing a starting point for a recognizing potentially useful patterns in vertebral
growth.
Temporal Trends in Males
For the male sample, the only significant differences were in some PVBHs, with Late Woodland
male measurements being larger than their Mississippian counterparts (Table 23). This trend held
true for all vertebrae with few exceptions, although only those listed yielded statistically
significant results. Only two out of 12 (17%) of thoracic vertebrae showed significant change in
PVBH over time (T1 and T3), while 3 out of 5 lumbar PVBHs (60%) showed significance
changes (L1, L2, and L5). These differences in PVBH were apparently not strong enough to
affect the means-based PVBH comparison (discussed above).

Dimension

Period
N
Mean
SD
P
Late Woodland 20
21.350
1.19
T1 PVBH
.005
Mississippian
60
20.395
1.30
Late Woodland 20
21.605
1.04
T3 PVBH
.011
Mississippian
61
20.852
1.14
Late Woodland 23
29.843
1.42
L1 PVBH
.035
Mississippian
64
29.131
1.35
Late Woodland 26
29.604
1.18
L2 PVBH
.053
Mississippian
65
28.897
1.68
Late Woodland 22
26.655
1.92
L5 PVBH
.003
Mississippian
59
25.319
1.67
Table 23- Male Individual Vertebra T-Tests between Late Woodland and Mississippian
(Unlisted results were not significant at the α=.05 level)
123

Temporal Trends in Females
Significant differences in APNCDs over time were indicated in a number of female vertebrae, all
of which were in the direction of increased diameter among Mississippians over their Late
Woodland counterparts (Table 24). The trend generally held true for most vertebrae with few
exceptions, although only those listed on the table yielded statistically significant results. Out of
12 thoracic vertebrae, 4 (33%) showed significant changes in APNCD over time (T1, T2, T11,
and T12), but this effect was apparently not strong enough to cause significant results in the
analysis of thoracic APNCD means discussed above. However, of the five lumbar APNCD
measurements that were part of the means-based analysis, four (or 80%) were significantly
different, with the Mississippian females having larger measurements in all cases. This correlates
well with the significant results in lumbar APNCD means over time. Two out of 5 (40%) female
lumbar TNCDs (were also significantly different between the Late Woodland and the
Mississippian, and this dimension approached significance in the means-based analysis.

124

Dimension

Period
N
Mean
SD
P
Late Woodland 25
13.692
1.20
T1 APNCD
.004
Mississippian
85
14.389
1.00
Late Woodland 20
13.765
0.84
T2 APNCD
.001
Mississippian
80
14.701
1.08
Late Woodland 26
15.135
0.92
T11 APNCD
.004
Mississippian
79
15.956
1.31
Late Woodland 24
16.408
1.13
T12 APNCD
.057
Mississippian
75
16.915
1.11
Late Woodland 32
16.384
1.18
L1 APNCD
.046
Mississippian
80
16.901
1.24
Late Woodland 30
15.273
1.40
L2 APNCD
.015
Mississippian
83
15.959
1.27
Late Woodland 35
14.171
1.65
L4 APNCD
.016
Mississippian
80
14.895
1.36
Late Woodland 32
14.608
2.28
L5 APNCD
.010
Mississippian
66
15.781
1.97
Late Woodland 35
21.511
1.58
L2 TNCD
.035
Mississippian
85
22.194
1.60
Late Woodland 38
21.805
2.15
L3 TNCD
.012
Mississippian
82
22.685
1.55
Table 24- Female Individual Vertebra T-Tests between Late Woodland and MississippianUnlisted results were not significant at the α=.05 level

Hypothesis Six- Growth Patterns and Social Differentiation among the Schild
Mississippians
Because of relatively high sample size and availability of previous mortuary studies (Goldstein
1980), the Mississippian component of the Schild site can be subjected to more specific and
culturally sensitive types of analyses. By analyzing growth patterns among Schild Mississippians
on a finer level, subtler patterns of growth differentials related to intracommunity social
differentiation can be detected. The spatial groupings and their hypothesized social significance
are described in detail in chapter four, and the results of the archaeologically-based growth
analysis are presented in this section.

125

Hypotheses 6a1 –Male Growth Patterns by Knoll
There are no significant differences in the mean thoracic or lumbar dimensions between males in
knoll A compared to knoll B, nor were there significant differences in femur length or LEH
between knoll A and knoll B males (Tables 25, 26, 27, and 28). Based on these observations, the
null hypothesis 6a1 is accepted. There are no observable differences in Schild Mississippian male
growth and stress indicators based on knoll.

Dimension
AVBH
PVBH
TNCD
APNCD

Dimension
AVBH
PVBH
TNCD
APNCD

Knoll
A
B

Knoll
N
Mean
SD
P
A
13
20.59
1.01
.531
B
17
20.39
0.69
A
15
22.31
0.96
.560
B
21
22.80
3.17
A
16
17.48
0.94
.551
B
20
17.69
1.12
A
14
15.20
0.71
.454
B
17
15.43
0.93
Table 25- Schild Male Thoracic Vertebral Dimensions T-test by Knoll

Knoll
N
Mean
SD
P
A
16
27.19
1.44
.904
B
22
27.24
1.02
A
16
28.24
1.37
.679
B
26
28.08
1.05
A
17
23.41
1.12
.305
B
18
23.92
1.74
A
12
15.12
0.99
.540
B
19
15.37
0.74
Table 26- Schild Male Lumbar Vertebral Dimensions T-test by Knoll

N
Mean
SD
P
20
45.18
2.00
.833
26
45.31
2.18
Table 27- Schild Male Right Femur Lengths T-test by Knoll

126

Knoll
A
B

N
LEH Present
Fisher’s Exact
15
5
.732
21
9
Table 28- Schild Male LEH frequency Chi-square test by Knoll

Hypotheses 6a2 -Female Growth Patterns by Knoll
There are no significant differences in the mean thoracic or lumbar dimensions between females
in knoll A compared to knoll B, nor were there significant differences in femur length or LEH
between knoll A and knoll B females (Tables 29, 30, 31, and 32). Based on these observations,
the null hypothesis 6a2 is accepted. There are no observable differences in Schild Mississippian
male growth and stress indicators based on knoll.

Dimension
AVBH
PVBH
TNCD
APNCD

Dimension
AVBH
PVBH
TNCD
APNCD

Knoll
N
Mean
SD
A
19
19.03
0.90
B
15
19.16
0.55
A
24
21.05
1.04
B
17
22.15
3.06
A
20
16.54
1.18
B
16
17.20
1.13
A
19
15.08
1.09
B
13
15.05
0.84
Table 29- Schild Female Thoracic Dimensions T-tests by Knoll

P

Knoll
N
Mean
SD
A
20
26.53
1.33
B
20
26.62
0.82
A
23
26.92
1.15
B
24
26.87
0.96
A
26
22.96
1.51
B
24
23.35
1.37
A
22
15.54
1.10
B
16
15.55
1.05
Table 30- Schild Female Lumbar Dimensions T-tests by Knoll

P

127

.611
.112
.096
.935

.796
.878
.349
.982

Knoll
A
B

Knoll
A
B

N
Mean
SD
P
28
42.46
2.20
.545
27
42.80
1.82
Table 31- Schild Female Right Femur Length T-tests by Knoll
N
LEH Present
Fisher’s Exact
19
8
.533
22
7
Table 32- Schild Female LEH Frequency Chi-square test by Knoll

Hypotheses 6b1- Male Growth Patterns by Burial Area
There were no significant differences in the mean vertebral dimensions between males based on
which burial area they were interred in, nor were there significant differences in femur length or
LEH between burial areas (Tables 33, 34, and 35). Based on these observations, the null
hypothesis 6b1 is accepted for both vertebral segments and all growth indicators.

Dimension
P
AVBH
.949
PVBH
.640
TNCD
.931
APNCD
.179
Table 33- P-values for Schild Male Lumbar Dimensions by Burial Area ANOVAs

Burial Area
1
2
3
4

N
Mean
SD
P
15
45.83
2.38
11
44.59
1.72
.384
5
45.90
2.67
15
45.93
1.75
Table 34-Schild Male Right Femur Length ANOVA by Burial Area

128

Burial Area
1
2
3
4

N
LEH Present
Fisher’s Exact
13
7
8
2
.254
7
1
8
4
Table 35- Schild Male LEH frequency Chi-square by Burial Areas

Hypotheses 6b2- Female Growth Patterns by Burial Area
P-values for the burial area ANOVAs for each vertebral dimension are presented in Tables 36
and 37. These results indicate a strong significant result for lumbar AVBH between females by
burial area at Schild, with group four females had the smallest measurements, while group three
were the largest (Table 38). Although there was no overall significant difference in femur length
by burial among Schild females in the one way ANOVA, further testing revealed that females in
burial area four had significantly smaller femur measurements when compared to the rest of the
sample (Tables 39 and 40). This pattern held true for AVBH (Table 41) No significant
differences were detected for LEH, but sample sizes were very small (Table 42). The results of
the burial area analysis indicated that 1) females in group four had significantly smaller AVBH
and femur lengths than Mississippian females in other burial areas, and 2) females in group
three may have had somewhat better growth experiences than other females in the community.
Based on these observations, the null hypothesis 6b2 is only accepted in part. For both LEH and
NCD, the null hypothesis 6b2 is accepted. For VBH, the null hypothesis 6b2 is rejected, as
ANOVA detected a statistically significant difference in lumbar AVBH by burial area. For femur
length the null hypothesis 6b2 is rejected in part- the ANOVA did not detect statistically
significant differences by burial area, but a t-test revealed that burial area four femur lengths
were smaller than all others to a statistically significant extent. The possibility that this pattern in
129

femur length and lumbar AVBH was related to burial area associations with charnel structures is
explored in the next section.

Dimension
P
AVBH
.141
PVBH
.380
TNCD
.429
APNCD
.780
Table 36- P-values for Schild Female Thoracic Dimensions by Burial Area ANOVAs

Dimension
P
AVBH
.028
PVBH
.182
TNCD
.743
APNCD
.537
Table 37- P-values for Schild Female Lumbar Dimensions by Burial Area ANOVAs

Burial Area
N
Mean
SD
P
1
8
26.73
0.75
2
12
26.55
0.90
.028
3
11
27.17
1.21
4
9
25.75
1.05
Table 38- Schild Female Lumbar AVBH Descriptive Statistics by Burial Area

Burial Area
1
2
3
4

N
Mean
SD
P
11
42.50
1.90
16
43.00
1.80
.146
17
43.11
2.18
11
41.45
1.90
Table 39- Schild Female Right Femur Length ANOVA by Burial Areas

Burial Area
N
Mean
SD
P
1,2,and 3 combined 44
42.90
1.95
.029
4
11
41.45
1.90
Table 40- Schild Right Femur Length T-test by Burial areas (4 versus all others)
130

Burial Area
N
Mean
SD
P
1, 2, and 3 combined 31
26.82
1.00
.008
4
9
25.75
1.05
Table 41- Schild Female Mean AVBH T-test by Burial Areas (4 versus all others)

Burial Area
1
2
3
4

N
LEH Present
Fisher’s Exact
9
3
13
4
.859
16
7
2
1
Table 42- Schild Females LEH frequency Chi-square by Burial Area

Hypotheses 6c1 - Male Growth Patterns by Charnel Association
There were no significant differences between charnel associated and non-charnel associated
males for AVBH or femur length (Tables 45 and 46). LEH frequencies were not quite
statistically significant between males based on charnel association (Table 47), but given the
small sample size, the trend toward charnel associated males having a lower incidence of LEH
should be noted. Therefore, the null hypotheses 6c2 is cautiously rejected as it pertains to LEH
frequencies, and accepted as it pertains to femur length and lumbar vertebral dimensions.

Dimension

Charnel
N
Mean
SD
P
Yes
15
27.38
1.15
AVBH
.516
No
23
27.11
1.24
Yes
19
28.30
1.36
PVBH
.458
No
23
28.02
1.01
Yes
16
23.77
1.77
TNCD
.731
No
19
23.59
1.22
Yes
11
15.52
0.60
APNCD
.283
No
20
15.18
0.93
Table 43- Schild Male Lumbar Vertebral Dimensions T-tests by Charnel Association
131

Charnal
Yes
No

N
Mean
SD
P
16
45.00
2.08
.558
30
45.38
2.10
Table 44- Schild Male Right Femur Length T-test by Charnel Association

Charnal
Yes
No

N
LEH present
Fisher’s Exact
15
3
.083
21
11
Table 45- Schild Male LEH Frequency Chi-square by Charnel Association

Hypotheses 6c2- Female Growth Patterns by Charnel Association
Tests to compare females from groups associated with charnel structures to females from those
groups not associated with charnel structures revealed that charnel associated females had nearly
significantly larger lumbar AVBHs, as well as significantly larger femur lengths (Tab;es 48 and
49). There was not a statistically significant difference in LEH frequencies between charnel and
non-charnel associated females (Table 50). Therefore, the null hypotheses 6c2 is cautiously
rejected as it pertains to lumbar vertebral dimensions, rejected as it pertains to femur length, and
accepted as it relates to LEH frequencies among Schild Mississippian females as analyzed by
charnel association.

Dimension

Charnel
N
Mean
SD
P
Yes
23
26.85
1.08
AVBH
.066
No
17
26.20
1.02
Yes
27
27.07
1.08
PVBH
.202
No
20
26.68
0.98
Yes
29
23.23
1.33
TNCD
.804
No
20
23.12
1.61
Yes
23
15.34
1.02
APNCD
.137
No
15
12.87
1.10
Table 46- Schild Female Lumbar Vertebral Dimensions ANOVA by Charnel Association

132

Charnal
Yes
No

N
Mean
SD
P
33
43.06
1.97
.049
22
41.98
1.93
Table 47- Schild Female Right Femur Length T-test by Charnel Association

Charnal
N
LEH present
Fisher’s exact
Yes
29
11
.188
No
12
4
Table 48- Schild Female LEH Frequency Chi-square test by Charnel Association

Hypotheses 6d1- Male Growth Patterns by Artifact Only Cluster
There were no significant differences in the mean vertebral dimensions between males in
artifact-only cluster 6 compared to males who were in cluster 0 (Tables 51 and 52), nor were
there significant differences in femur length or LEH between cluster 0 and cluster 6 males
(Tables 53 and 54). Based on these observations, the null hypothesis 6d1 is accepted, as there are
no observable differences in male growth patterns when analyzed by artifact-only clusters.

Dimension

Cluster
N
Mean
SD
P
0
19
20.62
0.95
AVBH
.227
6
11
20.23
0.53
0
23
22.19
0.96
PVBH
.197
6
13
23.31
3.93
0
22
17.65
1.12
TNCD
.659
6
14
17.50
.090
0
19
15.36
0.94
APNCD
.785
6
12
15.30
0.67
Table 49- Schild Male Thoracic Dimensions T-tests by Artifact-Only Clusters

133

Dimension

Cluster
N
Mean
SD
P
0
22
27.28
1.18
AVBH
.705
6
18
27.13
1.28
0
24
28.06
1.24
PVBH
.896
6
18
28.26
1.11
0
26
23.04
1.35
TNCD
.647
6
19
22.84
1.51
0
22
15.13
0.88
APNCD
.151
6
19
15.54
1.08
Table 50- Schild Male Lumbar Dimensions T-tests by Artifact-Only Clusters

Cluster
0
6

N
Mean
SD
P
25
45.72
1.56
.095
21
44.69
2.49
Table 51- Schild Male Right Femur Length T-test by Artifact-Only Clusters

Cluster
0
6

N
LEH Present
Fisher’s Exact
20
9
.501
16
5
Table 52- Schild Male LEH Frequency Chi-square by Artifact Only Clusters

Hypotheses 6d2- Female Growth Patterns by Artifact Only Cluster
There were no significant differences in the mean vertebral dimensions between females in
artifact-only cluster 6 compared to females who were in cluster 0 (Tables 55 and 56), nor were
there significant differences in femur length or LEH between cluster 0 and cluster 6 males
(Tables 57 and 58). Based on these observations, the null hypothesis 6d2 is accepted, as there are
no observable differences in female growth patterns when analyzed by artifact-only clusters.

134

Dimension

Cluster
N
Mean
SD
P
0
24
19.16
0.84
AVBH
.450
6
10
18.94
0.52
0
27
21.77
2.61
PVBH
.291
6
14
21.00
0.64
0
22
16.98
0.99
TNCD
.373
6
14
16.61
1.46
0
20
15.31
0.95
APNCD
.071
6
12
14.67
0.93
Table 53-Schild Female Thoracic Dimensions T-Tests by Artifact-Only Clusters

Dimension

Cluster
N
Mean
P
0
26
26.72
AVBH
.254
6
14
26.03
0
35
27.67
PVBH
.253
6
18
27.28
0
33
23.30
TNCD
.316
6
17
22.86
0
25
15.60
APNCD
.704
6
13
15.45
Table 54- Schild Female Lumbar Dimensions T-tests by Artifact-Only Clusters

Cluster
0
6

N
Mean
SD
P
36
42.67
1.80
.844
19
42.55
2.40
Table 55- Schild Female Right Femur Length T-test by Artifact Only Clusters

Cluster
N
LEH Present
Fisher’s Exact
0
25
8
.517
6
16
7
Table 56- Schild Female LEH Frequency Chi-square Test by Artifact-Only Clusters

Hypotheses 6e1- Male Growth Patterns by Artifact and Positioning Clusters
There were no significant differences in the mean thoracic or lumbar dimensions between males
in artifact and positioning cluster 1 compared to males who were in cluster 2 (Tables 59 and 60),

135

nor were there significant differences in femur length or LEH between cluster 1 and cluster 2
males (Tables 61 and 62). Based on these observations, the null hypothesis 6e1 is accepted, as
there are no observable differences in male growth patterns when analyzed by artifact and
positioning clusters.

Dimension

Cluster
N
Mean
SD
P
1
16
20.70
0.89
AVBH
.319
2
14
20.31
0.76
1
19
23.13
3.26
PVBH
.173
2
17
22.00
0.92
1
18
17.70
1.01
TNCD
.569
2
18
17.49
1.06
1
17
15.46
0.68
APNCD
.323
2
14
15.16
0.99
Table 57- Schild Male Thoracic Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

Dimension

Cluster
N
Mean
SD
P
1
18
27.35
1.26
AVBH
.528
2
20
27.10
1.16
1
21
28.15
1.42
PVBH
.981
2
21
28.14
0.90
1
19
23.82
1.52
TNCD
.529
2
16
23.49
1.44
1
15
15.42
0.96
APNCD
.424
2
16
15.19
0.74
Table 58- Schild Male Lumbar Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

Cluster
N
Mean
SD
P
1
24
45.52
2.19
.576
2
22
45.66
2.06
Table 59- Schild Male Right Femur Length T-test by Artifact and Positioning Clusters

136

Cluster
N
LEH Present
Fisher’s Exact
1
15
5
.731
2
21
9
Table 60-Schild Male LEH Frequency Chi-square test by Artifact and Positioning Clusters

Hypotheses 6e2- Female Growth Patterns by Artifact and Positioning Clusters
Vertebral results are listed on Tables 63 and 64 below. The only significant results of the female
artifact and positioning cluster based analysis occurred in the lumbar AVBHs, with cluster two
females being significantly larger. Female lumbar PVBH results approached statistical
significance. There were no significant differences in femur length or LEH between cluster 1 and
cluster 2 females (Tables 65 and 66). Based on these observations, the null hypothesis 6d1 is
cautiously rejected as it applies to female VBH. It is rejected for all other growth indicators, as
there are no observable differences in female NCD, LEH or femur length patterns when analyzed
by artifact and positioning clusters.

Dimension

Cluster
N
Mean
SD
P
1
12
19.05
0.70
AVBH
.825
2
22
19.11
0.81
1
14
20.91
0.92
PVBH
.206
2
27
21.82
2.54
1
12
16.43
1.16
TNCD
.153
2
24
17.03
1.17
1
11
14.89
0.91
APNCD
.455
2
21
15.17
1.03
Table 61- Schild Female Thoracic Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

137

Dimension

Cluster
N
Mean
SD
P
1
15
26.05
1.00
AVBH
.017
2
25
26.89
1.04
1
16
26.51
1.00
PVBH
.067
2
31
27.10
1.03
1
18
22.84
1.07
TNCD
.269
2
32
23.32
1.60
1
14
15.57
1.15
APNCD
.927
2
24
15.54
1.04
Table 62- Schild Female Lumbar Vertebral Dimensions T-tests by Artifact and Positioning
Clusters

Cluster
N
Mean
SD
P
1
21
42.55
1.84
.820
2
34
42.68
2.16
Table 63- Schild Female Right Femur Length T-test by Artifact and Positioning Clusters

Cluster
N
LEH Present
Fisher’s Exact
1
17
8
.328
2
24
7
Table 64- Schild Female LEH Frequency Chi-square by Artifact and Positioning Clusters

Intracemetery Trends in Growth Disruption: Individual Vertebrae
As was provided for the temporal analysis, the following analyses regarding individually-based
vertebral tests is provided as supplementary information. It is intended to assist in interpreting
the results presented in the segmental-based hypotheses testing in the intracemetery analyses
above, particularly insofar as providing a starting point for a recognizing potentially useful
patterns in vertebral column growth.

138

By Knoll
There were no significant differences between vertebral dimensions for males in knoll A versus
knoll B. However, analysis of females revealed statistically significant results in two thoracic
TNCDs (16% of TNCDs- T5 and T6), with knoll B females being larger in all significant cases
(Table 67). This pattern generally held true for other thoracic vertebrae, approaching significance
in several cases. However, these results taken together were not strong enough to result in a
significant difference in the means based analysis, as discussed above.

Vertebral
Dimension

Knoll

N

Mean

SD

P

A
29
15.42
1.27
.037
B
24
16.15
1.19
A
28
15.51
1.25
T6 TNCD
.038
B
24
16.24
1.20
Table 65- Schild Female Individual Vertebral T-tests by Knoll (Unlisted results were not
significant at the α=.05 level)
T5 TNCD

By Burial Areas
Significant results for the Schild male individual vertebrae tests by row are listed in Table 68. In
general, group three had the largest VBH measurements in all significant dimensions (see
appendix for more specific results, including means for all groups). When AVBHs are
considered, only five out of 12 (41.6%) thoracic vertebrae yielded significant results (T1, T3, T5,
T6, and T8), while only one out of 12 thoracic PVBHs (8.3%) yielded significant results (T2). In
significant NCD dimensions, group two tended to be the largest, with three out of 12 (25%)
thoracic APNCDs (T8, T9, and T10), one out of five (20%) lumbar APNCDs (L3), and one out
of five (20%) lumbar TNCDs (L4) significant at the α= .05 level. None of these results were
strong enough to affect the means based analysis, for which no significant results were obtained
for the male portion of the Schild sample.
139

Vertebral Dimension
P
T1 AVBH
.011
T2 PVBH
.032
T3 AVBH
.051
T5 AVBH
.034
T6 AVBH
.005
T8 AVBH
.017
T8 APNCD
.023
T9 APNCD
.028
T10 APNCD
.027
L3 APNCD
.035
L4 TNCD
.009
Table 66- P-values from Schild Males Individual Vertebrae ANOVAs by Burial Area (Unlisted
results were not significant at the α=.05 level)

A different pattern is evident when Schild females are considered. Only three significant results
were obtained in the row based individual vertebral analysis, and all of these were lumbar
AVBHs (Table 69). Overall, three out of five (60%) of lumbar AVBHs were significant (L1, L2,
and L4). In all of these cases, group three females were the largest, and in two of the three cases
group two females were the second largest.

Vertebral Dimension
P
L1 AVBH
.014
L2 AVBH
.011
L4 AVBH
.057
Table 67- P-values from Schild Females Individual Vertebrae ANOVAs by Burial Area
(Unlisted results were not significant at the α=.05 level)

By clusters
No strong patterns emerged when Schild male vertebral measurements were analyzed by artifactonly clusters. One significant result (T12) indicated that APNCD was larger among cluster six
males when compared to males in all other clusters (Table 70), but this discrepancy only
140

constituted one out of 12 (8.3%) vertebral dimensions and was not indicative of a broader
pattern.

Vertebral
Dimension

Artifact Only N
Mean
SD
P
Cluster
0
28
17.12
0.99
T12 APNCD
.005
6
19
17.96
0.94
Table 68- Schild Male Individual Vertebral T-tests by Artifact Only Cluster (Unlisted results
were not significant at the α=.05 level)

When Schild male vertebral measurements were analyzed by artifact and positioning
clusters, cluster one males had larger VBH and NCD measurements in all three significant cases
(Table 71).

Vertebral
Dimension

Artifact and
N
Mean
SD
P
Positioning
Cluster
1
22
18.81
1.04
T4 AVBH
.027
2
23
18.11
0.99
1
23
21.61
0.93
T5 PVBH
.013
2
21
20.79
1.35
1
25
24.18
1.91
L4 TNCD
.030
2
27
23.00
1.93
Table 69- Schild Male Individual Vertebral T-tests by Artifact and Positioning Clusters (Unlisted
results were not significant at the α=.05 level)

Statistically significant VBH and NCD measurements were detected at several vertebral
levels among Schild females, but there was no clear patterning to the results (Table 72).
However, in all of these cases, cluster six females were significantly smaller than those who
were not in cluster six.

141

Vertebral
Dimension

Artifact Only N
Mean
SD
P
Cluster
0
34
17.24
0.93
T3 AVBH
.032
6
16
16.64
1.59
0
30
16.06
1.12
T7 TNCD
.038
6
19
15.25
1.54
0
37
20.90
1.08
T8 PVBH
.050
6
21
20.35
0.83
0
36
15.35
1.09
T8 APNCD
.032
6
19
14.65
1.18
0
34
20.31
1.01
T9 AVBH
.043
6
24
19.76
0.98
0
36
21.20
1.11
T9 PVBH
.022
6
24
20.58
0.82
0
36
21.20
1.08
T9 APNCD
.037
6
24
20.58
1.15
0
36
25.05
1.43
T11 PVBH
.039
6
22
24.29
1.15
0
34
16.27
1.31
T11 APNCD
.031
6
23
15.48
1.37
0
33
20.57
0.90
T12 APNCD
.027
6
20
20.14
1.13
0
35
17.23
0.94
L1 APNCD
.003
6
21
16.34
1.22
0
34
26.25
1.58
L2 AVBH
.043
6
18
25.40
1.00
Table 70- Schild Female Individual Vertebral T-tests by Artifact Only Clusters (Unlisted results
were not significant at the α=.05 level)

When analyzed by artifact and positioning clusters, female vertebral measurement
comparisons yielded only one significant result, with T1 PVBHs being significantly larger
among cluster 2 females (Table 73).

Vertebral
Dimension

Artifact and
N
Mean
SD
P
Position
Cluster
1
22
18.98
1.30
T1 PVBH
.035
2
38
19.77
1.41
Table 71- Schild Female Individual Vertebral T-tests by Artifact and Positioning Clusters
(Unlisted results were not significant at the α=.05 level)
142

Hypothesis Seven- Multifactor Approaches
Multifactor approaches (comparing biological indicators to multiple social indicators
simultaneously) are inherently more complicated than the single factor approaches above, as the
addition of more factors decreases cell counts. Using the broadest categories available (i.e.
binary categories only, which excludes the burial area category) allowed for maximization of cell
counts, but the availability of other binary spatial factors (knolls and charnel associations)
allowed for an investigation of the interactions between spatial organization and
artifact/positioning attributes. Despite the issue of sample size, several significant interactions
were discovered (Table 74).
Measurement
Female lumbar APNCD

Factors
P
Artifact and Positioning clusters, Knoll
.028
Artifact and Positioning clusters, charnel
Female lumbar APNCD
.053
association
Female lumbar AVBH
Artifact only clusters, Charnel association
.048
Male lumbar AVBH
Artifact only clusters, Knoll
.012
Male Femur Length
Artifact Only Clusters, Charnel Association
.038
Table 72- P-values from Two-way ANOVAs (Unlisted results were not significant at the α=.05
level)

Except for the significant relationship between charnel association and female AVBH, the one
way ANOVA’s between the osteological measurements and individual factors listed above did
not yield statistically significant results. When a spatial and an artifactual factor were considered
together, however, some interesting patterns emerge. Notably, the multifactor results for male
lumbar AVBH and male femur length were the only significant results for males in the entire
analysis. Because statistically significant patterns were found, hypothesis seven is rejected.

143

CHAPTER SEVEN- DISCUSSION AND INTERPRETATION

In previous chapters, it was suggested that 1) traditionally employed bioarchaeological indicators
of early childhood stress (LEH) and cumulative growth (femur length) would provide a
framework for testing the vertebral method of growth assessment, and 2) a nuanced perspective
on temporal changes and intracommunity differentiation in growth patterns could be achieved
through application of nontraditional osteological methods and the use of finer scale, mortuary
based units of analysis. A set of specific hypotheses were provided in chapter five, upon which
the statistical tests presented in chapter six were based.
In the first part of the current chapter, results of the statistical testing of each hypothesis
will be interpreted. It will be argued that, although the vertebral methods tested here do not
correspond perfectly with the LEH and femur length results, there is evidence that vertebral
metrics may provide meaningful insights into growth patterns in the past, and that the vertebral
method may provide unique information that complements traditional methods. Statistical tests
of the temporal and intracemetery growth trends for males and females will be interpreted within
the archaeological framework presented earlier in this dissertation. The overall results of this
analysis suggest that growth differentials between individuals are related to archaeologicallyderived mortuary groupings at the Schild Mississippian site, and that this is particularly true for
the female portion of the sample. This chapter will also summarize comparable data from other
Late Prehistoric sites, suggesting that that there may be regional variation in temporal growth
trends and intrasite patterns of differential growth.
It is proposed that the Late Woodland- Mississippian transition in West Central Illinois
was characterized by improving conditions of in generalized/cumulative early childhood growth

144

for females, decreasing incidence of episodic growth disruption for male children, and no change
in cumulative growth through early adulthood for either males or females. Within the Schild
Mississippian community, women may have experienced more variability in biocultural growth
experiences than men, based not only on biological sex, but also on the roles they played as part
of socially meaningful kinship or village networks, or based on finer scale temporal variation
within the Mississippian component. In either case, variability in the growth experiences of
Mississippian women related to cultural parameters is indicated. It will further be argued that
correlations between Mississippian mortuary patterning and biological growth experiences at
Schild suggest that the adoption of Mississippian mortuary attributes in the Lower Illinois River
Valley were not acts of simple, superficial emulation, but were reflective of the demonstrable
biosocial effects of Mississippianization on the lived experiences of the late prehistoric
inhabitants of this region.

Intrasegment Reliability
Hypothesis one tested for intrasegment reliability in vertebral measurmnts in order to justify a
segment-based approach to subsequent tests. All Cronbach’s Alpha tests yielded results that were
sufficiently high (greater than .7) to justify segment-based testing of vertebral data. This is not
surprising, as visual observation of vertebral dimensions reveals that they do not vary randomly
in relation to one another- there is a clear progression in size/shape variation in a sequence of
vertebrae.
The results for the fifth lumbar vertebra NCDs are noteworthy and warrant more detailed
discussion, as the alpha scores indicate that L5 measurements were (except in male AVBH)
consistently the most variable and, subsequently, the least statistically “reliable” of all vertebrae.

145

This may be because the L5 neural arch grows for a longer period of time than the other lumbar
vertebrae (Ursu et al 1996), and often does not complete its growth until age five (Baker et al
2005). In the analyses presented here, all lumbar vertebral dimensions (L1-L5) were used to
compute means. In future research where sample size is perhaps more problematic, omitting L5
from averages may serve to increase both sample sizes and intrasegment reliability scores.
However, this comes with the trade off of limiting averages to the first four lumbar vertebrae,
which are restricted to an earlier period of postnatal growth. Alternately, because L5 seems to be
indicative of childhood growth for a longer postnatal period than all of the other vertebrae, trends
in L5 neural canal diameters as analyzed on their own may also be informative.

Utility of Vertebral Method
This section specifically interprets the results of statistical testing of hypothesis two. It also
draws on observations of the relationship between NCD and LEH (as indicators of early
childhood growth) and VBH and femur length (as indicators of cumulative growth through early
adulthood) to bolster an argument for the vertebral method as a potentially sensitive indicator of
growth trends.

Relationship Between NCD and Femur Length
One of the basic premises of this dissertation was that traditionally utilized indicators of growth
and stress could be used to test the efficacy of the vertebral method for studies of growth in the
past, as has been suggested by Porter (1986) and Tatarek (1999). Linear regression analysis
revealed that neither TNCD or APNCD were significantly correlated with femur length. This
suggests that NCD is independent of cumulative growth as indicated by femur length and may

146

lend credence to an assertion that NCD are reflective of early childhood rather than cumulative
growth experience. However, this finding is in contrast to a similar analysis in Tatarek’s (1999)
study of the Hammond Todd collection, and the reason for this discrepancy must be explored in
future research. Possible explanations may be attributable to between-population variation in
growth patterning, or differences in the natures of the two skeletal collections (i.e. one represents
a genetically-related community, while the other is a collection of genetically dissimilar
individuals).
Another issue that must be addressed is the fact that, in the current analysis, more
strongly significant results (among females over the Late Woodland-Mississippian transition)
were obtained with APNCD than with TNCD. This phenomenon was also recognized in Clark’s
(1988) original study. He cites several previous studies in addition to providing new evidence
(using an age-graded sample of the Dickson Mounds collection) that APNCD and TNCD were
not measuring the same period of growth, and that TNCD likely corresponds to growth
continuing into a later period of childhood, while the APNCD is limited specifically to early
childhood (prior to age 4). Furthermore, and as discussed in chapter four, Gepstein et al (1991)
found that APNCD was the only vertebral dimension to be significantly correlated with total
vertebral canal size, so it is not entirely surprising that it was this dimension, rather than TNCD,
that yielded the strongest statistically significant results and seemed most useful for
differentiating between groups.
Sexual Dimorphism in NCD Over Time
The NCD data merits further discussion here, as one of the basic assumptions underlying this
study- that NCD’s reflect early childhood growth experiences- has implications for a facet of
sexual dimorphism that is less well understood. Sexual dimorphism in skeletal features typically
147

develops after puberty, which is the primary reason that subadult remains cannot be sexed.
Therefore, it would be reasonable to hypothesize that any skeletal element completing growth
during the early childhood years (i.e. neural canal) would not inherently display sexual
dimorphism, and that any significant difference between males and females would therefore be
more fully explained with a consideration of social or environmental factors.
In statistical testing of hypothesis 2c, lumbar APNCD showed no sexual dimorphism
during the Late Woodland period (α=.514). However, during the Mississippian period, female
APNCD were significantly larger than males (α=.051). Lumbar TNCD’s were significantly
larger in males compared to females in the Late Woodland period (α=.006), but the difference
between male and female TNCD was no longer statistically significant in the Mississippian
(α=.924).
Interpretation of these results is facilitated by keeping in mind two key issues. First,
although the specific nature of change in the two dimensions is different (i.e. females surpassing
males in one measure, and females catching up to males in the other measure), it should be noted
that both scenarios indicate an improvement in the conditions of female early childhood growth
over time. Second, APNCD may be indicative of an earlier period of early childhood growth
than TNCD, and so the differences may indicate a more complex pattern of differential treatment
of male and female children in which 1) female Mississippian children had better growth
conditions early on as indicated by APNCD, and 2) a slight later childhood advantage among
Late Woodland male children had disappeared by Mississippian times, as indicated by TNCD.
All of these issues considered together lend support to a hypothesis that neural canal
diameter (APNCD in particular) is a potentially sensitive indicator of early childhood growth,
although further investigation is necessary before this can be said with certainty.

148

Comparing Neural Canal Diameter and LEH Results
The relationship between APNCD, LEH, and early childhood growth disruption can be further
explored by comparing results for changes in LEH frequency to changes in canal diameters over
the Late Woodland-Mississippian transition. Temporal analysis revealed that female APNCD
increased significantly with Mississippianization (see next section for further discussion) which,
according to the basic assumptions of the vertebral method, indicates improving early childhood
growth conditions for females over this cultural transition. However, this analysis did not
indicate a concomitant decrease in LEH frequency over time, which would have offered an
excellent corroboration of APNCD as a sensitive indicator of early childhood growth conditions.
Exacerbating this issue is the fact that there was a statistically significant difference in LEH
frequencies for males over the Late Woodland-Mississippian transition, with Mississippian
males experiencing fewer early childhood growth disruption episodes than their Late Woodland
counterparts. However, there was no concomitant change in neural canal diameter.
A possible explanation for the observed discrepancies between vertebral canal data and
LEH data may be related to the nature of stress being recorded by each measure. Although LEH
is traditionally relied upon as an indicator of early childhood stress and growth disruption, it
represents episodic stress while the vertebral method represents more generalized, cumulative
growth over the entire early childhood period (Goodman et al 1984). Occasional, episodic
perturbations would not necessarily have an effect on total, cumulative growth, so there is no
reason that the two different analyses would necessarily yield the same results. It may be,
therefore, that LEH and NCD are measuring inherently different phenomena, albeit during the
same growth period.

149

Previous bioarchaeological studies of macroscopic and microscopic enamel defects with
Mississippianization in Illinois have revealed patterns different from those reported here.
Goodman et al (1980) investigated changes in frequencies of linear enamel hypoplasia in males
and females interred at Dickson Mounds in West Central Illinois, concluding that during the Late
Woodland, males displayed fewer hypoplasias than females, but that these proportions were
essentially equal by Mississippian times. The discrepancy between the results of the current
study and the results of the Goodman et al (1980) study may be due to regional variation in sexbased patterns of growth disruption. More locally, Cook’s (1984) analysis of microscopic enamel
defects in in Lower Illinois River Valley samples (including Schild) indicated a trend toward
decreasing frequencies of LEH in females relative to males between the Middle Woodland and
the Mississippian. However, Cook’s (1984) results are not directly comparable to the LEH
results presented here because her study used 1) different teeth (she used first molars), 2)
different methodologies (she used dental microdefects) and 3) a different temporal framework
(she compared Middle Woodland frequencies to Mississippian frequencies).

Scale of Difference
It should also be noted that the discrepancies in segment-based neural canal diameters in this
study are very small (on the scale of .5 millimeter), even when compared to the seemingly small
discrepancies demonstrated by Clark (1985) using the Dickson Mounds collection. The
biological and cultural meaning of this discrepancy is not known, and warrants further research.
One possible explanation is that, despite their apparent phenotypic sensitivity, vertebral
size is still under a significant amount of genetic control and is very population specific. If this is
the case, one might expect the scale of differences to not be directly comparable across

150

populations, particularly when one population seems to have experienced an influx of new alleles
from a foreign source over the Late Woodland-Mississippian transition (as was the case in the
Dickson Mounds sample) while the other population was relatively genetically stable over time,
with no significant biological discontinuity (as was the case at Schild and Yokem).
Another explanation for the may lie in slight differences in measurement techniques (i.e.
one observer holding the calipers slightly differently, or squeezing the jaws together with
somewhat different amounts of pressure). These types of interobserver error are difficult to
control for, and as such it is argued here that while statistical trends may be comparable between
studies, direct comparison of measurements taken by different observers is not.
A special case is also presented by L5. As pointed out earlier, the L5 vertebral canal
measurements may eventually prove to be a better indicator of early childhood growth than the
other vertebral arches by virtue of its relatively extended growth period (although further
investigation is required). The segment-based analysis averaged the results for the entire lumbar
segment, but the individually-based analysis revealed that the change in APNCD over the Late
Woodland-Mississippian transition in L5 this measure was nearly 60% larger than the change in
the other lumbar vertebrae that yielded significant results in the individual-based analysis.
Despite the small scale of the observed differences in vertebral dimensions, several were
strong enough to achieve statistical significance and their potential biocultural significance is
considered.

Comparing Vertebral Body Height and Femur Length Results
It was hypothesized that the reliability of vertebral body heights as an indicator of cumulative
growth could be evaluated by looking at its relationship with femur length, a more traditionally

151

employed indicator. Linear regression analysis indicated that both AVBH and PVBH were
significantly correlated with femur length, lending support to the hypothesis that vertebral body
height is a good measure for cumulative growth experiences. This is not in agreement with
Tatarek’s (1999) results based on cervical body heights. R-square values for the strength of the
association were relatively low, suggesting that although the measures were positively
correlated, there are some other factors determining differential cumulative growth between
these elements.
It is surprising that, in this study, all significant results indicating change in vertebral
body heights were in lumbar AVBH, despite the facts that both AVBH and PVBH were
significantly correlated with femur length and that those correlations were of equal strength. It
may be that AVBH is a more reliable measure of cumulative growth in the lumbar region than
PVBH because the lordodic lumbar curvature would put more compressive force on the posterior
aspect of the lumbar vertebrae, resulting in enough subtle collapse over time to obliterate the
difference that are preserved in AVBH. However, it also must be considered that intervertebral
disc morphology (specifically the position of the nucleus pulposus) contributes more to the
development of secondary spinal curvatures than do vertebral bodies themselves (Taylor 1975).
In any case, this phenomenon needs to be explored more thoroughly.
The relationship between vertebral body height, femur length, and cumulative growth
patterns is further complicated when one evaluates the pattern of significant results in the
mortuary based analysis if the Schild Mississippian cemetery. Cook (2007) points out that there
is no evidence for a decrease in cumulative growth patterns over the late WoodlandMississippian transition as indicated by femur length. Likewise, the vertebral body height data
presented here does not indicate a statistically significant difference in VBH over this cultural

152

transition for either sex. In several tests involving Schild females by charnel association, nearly
statistically significant lumbar AVBH data seemed to correlate well with statistically significant
femur length data in several instances, indicating that certain segments of the Schild
Mississippian female sample experienced better cumulative growth. However, in other cases, the
relationship between the two indicators was less clear. For example, in one case (Schild females
by burial area), lumbar AVBH yielded significant results while femur length did not, although
retesting the smallest group (group four) against all others did yield significant results in both
femur length and AVBH. There were no cases where femur length was statistically significant
without concomitant significant (or very near significant) difference in lumbar AVBH. This may
suggest that vertebral body heights are somewhat more sensitive to growth disruption than femur
length, but this cannot be known for certain without further research.
All of this, considered with the weakly significant correlation between VBH and femur
length, suggests that vertebral body height should cautiously be considered as an indicator of
cumulative growth. However, the intricacies of the involved relationships between the two
measures requires further investigation, and it may be that femur length provides the same
growth information in a more accessible manner.

Thoracic versus Lumbar Sensitivity to Growth Disruption
Although not formulated as a specific hypothesis, it is important to recognize that significant
results in the means-based analyses were limited to the lumbar vertebrae, despite the fact that
significant results were obtained for several thoracic vertebrae in the individual vertebral
analyses. Several factors are suggested to explain this discrepancy.

153

First, the thoracic vertebrae are, quite obviously, smaller than their lumbar homologues.
Larger vertebrae simply “grow more” than smaller vertebrae, perhaps giving them more
opportunity to be affected by growth disruptions. This issue was also addressed by Clark (1985)
who suggests that lumbar NCDs in particular have a longer period of postnatal growth than do
thoracic vertebrae, thereby providing a larger window during which cultural factors can affect
growth differentials. Second, larger vertebrae might also simply be easier to measure without
error. Because there are more thoracic vertebrae (n=12) than lumbar vertebrae (n=5) significant
differences in, for example, three thoracic vertebrae might get lost in a segment-based analysis,
while significant differences in three lumbar vertebrae might be strong enough to affect the
segment-based analysis. This would be particularly true if thoracic growth trajectories are
significantly different between the upper and lower portions, indicating that subdivision of the
thoracic segment may be called for.
Individual-based vertebral testing indicated some support for the proposed explanations.
Patterning of the significant results sometimes indicated that sequences of thoracic vertebrae
yielded significant results, or that a strong pattern of sequential significant results in the lumbar
vertebrae were sometimes continued into the lowermost thoracics. However, in both cases, nonsignificant results in the rest of the thoracic segment negated their significance in the segmentbased testing. Future research will involve subdividing the thoracic segment into anatomically
appropriate subsets for segment-based testing.
All of these propositions should be investigated more fully, but based on preliminary
observations, it is suggested that the lumbar segment of the vertebral column may be the better
choice for growth studies of the kind attempted in this dissertation. This is in agreement with

154

Tatarek (1999), who also suggests that thoracic segment displays much more variability in
growth patterning than either the cervical or lumbar segments.

Interpretation of Growth Data within Temporal Framework
In this section, results of statistical testing of hypotheses four and five are integrated and
discussed. A generalized picture of growth trends for males and females is interpreted within a
diachronic archaeological context, with particular reference to the increasingly important role of
maize agriculture over the Late Woodland-Mississippian transition.

Late Woodland-Mississippian Transition
The analysis of femur length data collected for this study corroborates Cook’s (2007)
observation that there did not seem to be a change in stature with Mississippianization. This was
true for both the males and the females in this study. Furthermore, there were no statistically
significant differences in vertebral body height data for either males or females between the Late
Woodland and Mississippian periods, which complement the femoral analysis.
Analysis of vertebral canal dimensions yielded some significant differences. There were
no significant changes in vertebral canal growth among males, but females APNCD
measurements were found to increase significantly with Mississippianization (TNCD’s were also
very close to significantly different, with a p-value of α= .059). This would seem to suggest that
early childhood growth experiences (as indicated by APNCD) were ameliorated for females with
the transition to a Mississippian lifestyle, and that this advantage may have lasted into later
childhood (as indicated by TNCD). The preponderance of ethnographic, archaeological, and
iconographic evidence link agriculture in this region to women. As maize agriculture became

155

increasingly important with the advent of Mississippianization, it makes sense that women could
gain social status through their productive and reproductive potentials (Fritz 1999, Watson and
Kennedy 1990, Alt and Pauketat 2007, Thomas 2000), and that female children would thereby
become more socially valued. The lack of concomitant changes in APNCD among males with
Mississippianization suggests that the biosocial circumstances of early childhood growth for
male children were not different between the two periods, or at least were not different enough to
affect change in neural canal growth in a statistically detectable way given the limitations of the
sample.
However, accounting for the LEH trends complicates this interpretation. The data
collected for this study showed a significant decrease in the frequency of LEH in males with
Mississippianization, but no temporal difference in female LEH frequencies. This suggests that
Mississippian male children experienced fewer episodes of early childhood stress and illness
than their Late Woodland counterparts, and also that the female early childhood stress/illness
experience did not change with Mississippianization. As discussed above, several explanations
can be offered and evaluated to account for this apparent contradiction with the vertebral
analysis. The first is that LEH is a measure of episodic stress, while neural canal diameter is a
measure of cumulative growth during early childhood, so the very nature of the data may
preclude them from being directly comparable. Instead of one being used to validate the other,
perhaps the information they provide should be interpreted as representing different and
complementary perspectives on early childhood growth conditions. Second, previous studies
(Cook 1984, Goodman et al 1980) suggest trends toward improved female early childhood
conditions in Mississippian samples compared to earlier samples from elsewhere in the regions,
without a concomitant similar trend in male frequencies.

156

Overall, there is much to consider when interpreting the sex-based growth data collected
for the temporal portion of this study. Because of the evidence discussed in the previous section
(Utility of the Vertebral Method), it is suggested that the lack of change in VBH, as well as the
significant changes in female APNCD toward larger canal size with Mississippianization, reflect
meaningful changes in the biosocial conditions of growth during this transition. Females
experienced improved early childhood growth patterns compared to their Late Woodland
counterparts, while the male early childhood growth experience remained largely unchanged
with Mississippianization. This pattern may be related to ethnographic and iconographic
evidence for the importance of females in the development and intensification of maize
agriculture, which played an important socioeconomic and ideological role in Mississippian life.

Further Contextualizing Temporal Trends
There is a clear tendency to refer to the female association with agriculture and increased
importance of agricultural pursuits with Mississippianization as an explanation for the
improvement in female early childhood growth conditions suggested above. However, to
continue to do this relegates the Late Woodland results to simple “baselines” against which to
measure Mississippian growth differentials rather than treating the nature of Late Woodland
sexual dimorphism as a phenomenon worthy of interpretation in and of itself. The detailed
intracommunity interpretations that follow for the Schild Mississippian component are simply
not possible for the Late Woodland component at this time, as we are lacking adequate mortuary
data to use as units of analysis for the osteological data. Until such analyses are available, an
exploration of the literature on Late Woodland social dynamics may offer cursory explanations
for the apparent equity of Late Woodland male and female early growth experiences, as

157

indicated by comparison of APNCD between Late Woodland males and females when testing
hypothesis two.
As discussed in chapter two, the Late Woodland tradition is characterized by
heterarchical relationships between communities (Cobb and Nassany 1995), and this may have
been the prevailing theme in intracommunity social relations (including gender relations) as it
was for intercommunity interactions. The social distance, then, between Late Woodland men and
women may have been somewhat more horizontal than vertical in nature. Authors like Pauketat
and Alt (2007) seem to implicitly suggest this type of gender relations in the Late Woodland
when they argue for increased polarization and politicization of gender roles with
Mississippianization. If early childhood growth experiences were equitable between males and
females in the Late Woodland, it may be because there were fewer sociocultural forces (like
agricultural intensification, for example) to compel inequitable treatment during this period.
However, if we take TNCD to represent a somewhat later period of childhood growth (following
Clark 1985 and Tatarek 1999), we need to account for an apparent late childhood improvement
of male growth relative to females, which also suggested by testing of hypothesis 2. This may
have been related to the young boys becoming more involved in hunting with their adult male
kin as they got older. However, none of these ideas can be tested until more rigorous Late
Woodland mortuary and gender studies begin to come out of archaeology.

Growth Patterns and Social Differentiation among Schild Mississippians
The second part of this study focused specifically on the Mississippian component of the Schild
site, as substantial mortuary data is available for incorporation into the osteological analysis.
This section integrates, interprets, and contextualizes test results for hypotheses five and 6.

158

No significant associations were found between male growth data and any of the
mortuary parameters in single factor analyses. No significant associations were found for either
the knoll-based or the artifact only clusters, nor were any significant differences in NCD
detected. However, several statistically significant associations between female cumulative
growth patterns (AVBH and femur length) and other Schild mortuary parameters (artifact and
positioning clusters, burial area, and charnel association) were discovered and are discussed
below.

Artifact and Positioning Clusters
T-tests based on artifact and positioning clusters revealed that females in cluster two had
significantly larger lumbar AVBH compared to their cluster one counterparts. There was no
concomitant difference in femur length.
Interpreting the meaning of the statistically significant difference between females in
these two clusters is somewhat complicated. Goldstein (1980) points out that burial positioning
(extension versus non-extension, in particular) seems to have been an important attribute in
differentiating the Schild burials, although variation in artifact assemblages is also significant.
Because the primary division in Goldstein’s cluster analysis was one of burial positioning
(extension versus non-extension), there is significant variability in artifact assemblages within
each collapsed cluster. The significant results in the vertebral body height data may be
highlighting subtle differences in cumulative growth experiences of females related to the type of
social differentiation indicated by burial positioning, which may be a status related (Goldstein
1980).

159

Burial Areas
Interesting patterns of differentiation in growth experience among females seem to be indicated
by the tests based on burial areas. ANOVA indicated statistically significant differences in
AVBH, with group four females having the smallest measurements and group three having the
largest. For femur length, group four was once again the smallest while group three was once
again the largest, but this difference was not reflected in the femur length ANOVA. However,
upon further investigation, t-tests revealed that group four females had significantly smaller
measurements in both AVBH and femur length when compared to the rest of the female sample
(groups 1, 2, and 3 combined).
As with the cluster-based analysis, it would have been ideal to perform ANOVAs on the
growth data based on the original rows provided by Goldstein (1980). However, sample size
issues prevented this. The groupings referred to as “burial areas” that were used in this analysis
roughly correspond to sets of rows in Goldstein’s analysis, and correspond almost directly to
spatial divisions recognized by both Goldstein (1980) and Perino (1972). Goldstein (1980)
provides a strong argument for a relationship between kinship groups and row-based mortuary
patterning, and although the rows do not perfectly correspond to the burial areas defined by
Goldstein (1980, pp. 107 and 109) and used for this portion of the current analysis, it is
cautiously suggested here that these groupings may represent kinship groups, community based
corporate descent groups, or some similar manner of intracommunity social differentiation.
However, Goldstein also raises two other important issues that should be discussed here.
First, she suggests that the individuals considered to be part of group four in this analysis may be
the earliest burials at the site. Therefore, the fact that females in this group have significantly
smaller AVBHs and femur lengths may reflect their experiences under a very early version of

160

Mississippianism at the site. Furthermore, the issue of community membership must also be
considered. Specifically, Goldstein (1980) points out that there is no Mississippian village site in
the region with a population large enough to be solely responsible for the number of burials at
Schild. It is likely that multiple Mississippian smaller communities interred their dead at Schild,
and therefore each of these spatial groups may represent separate communities. However,
enough social unity was recognized between these groups to compel them to inter their dead in at
the same site. In this scenario, females from the community or communities represented by
group four seem to have less favorable cumulative growth experiences than females from other
communities, while group three females may have been somewhat better off. In any case, the
results of this analysis indicate that growth experiences among Schild females were
differentiated and patterned in a way that is related to socially-significant spatial clustering,
which may represent kinship, corporate descent groups, community membership, cultural change
over time, or some combination of these factors. This suggests that Mississippian women in this
region may have experienced more variability in cumulative growth experiences than Schild
men, based not only on biological sex, but also on their cultural experiences over time and/or the
roles they played as part of socially meaningful kinship or community networks.

Charnel Association
The descriptive statistics from the analysis discussed above indicate not only that group four
females were significantly smaller in both femoral and vertebral body height measures, but also
suggested that there may be some discrepancies in growth patterns based on charnel structure
association. Groups two and three generally correspond with charnel structures while groups one
and four generally do not. For the charnel association based analysis, the results for the male

161

subsample were not statistically significant. For the females, however, both AVBH
measurements and femur lengths were significantly greater for those associated with a charnel
facility when compared to those who were not.
Many of the same caveats regarding the cultural meaning of mortuary patterning
discussed in the previous section also apply here, specifically regarding the relationship between
time, community membership, and kinship in regards to access to a charnel related burial.
Goldstein (1980) points out, however, that the knoll B charnel and non-charnel areas were likely
used simultaneously, while the knoll A charnel and non-charnel areas may have been used either
sequentially or simultaneously. If the sequential scenario were correct, then one might make an
argument for excluding group four from this analysis because, although it is indeed a noncharnel group, these individuals should not be considered to have been “excluded” from charnel
association in the same way that group one individuals were, and it might therefore be
inappropriate to lump the two together. However, assuming that charnel associated interment
was available in some way to group four individuals is not completely unfounded, as charnel
ceremonialism was not a Mississippian innovation that simply had not been invented yet.
In this analysis, the biological patterns observed are meaningful regardless of the timing
of the introduction of charnel ceremonialism, and the charnel/noncharnel groupings were
retained for component of this analysis. The two non-charnel areas had lower mean AVBH and
femur length then the two charnel areas, but it should also be recognized that the non-charnel
female results may have been disproportionately skewed by group four measurements. The
results of the growth pattern analysis suggest that Schild women may have experienced
variability in cumulative growth patterns (as indicated by vertebral and femoral measurements)
based not only on their biological sex, but also on their membership in social groups (kinship or

162

community associated) that either entitled or excluded them from charnel associated burial
treatment.

Multifactor Analyses
The two way ANOVAs simultaneously comparing osteological measurements with two
categorical factors (one spatial, one artifactual) yielded five significant results. The meaning of
any one of these results is difficult to interpret specifically, but what seems particularly apparent
is that seemingly insignificant relationships between a biological dimension and any individual
factor may be masking more complicated multifactor interactions between artifact assemblages,
space, and biology. This is not surprising, as mortuary programs are complicated cultural
phenomena, and human decisions regarding these programs and their implementation are
concerned with multiple levels of sociocultural meaning. It is likely that a suite of mortuary
characteristics is more meaningful than any one parameter on its own. However, sample size
problems will often preclude this level of investigation of skeletal remains from archaeological
mortuary contexts.
Nevertheless, the results of the two way ANOVAs presented in this research further
support Goldstein’s contention that both spatial and artifactual attributes were important
components of the Mississippian burial program and reflect meaningful sociocultural divisions,
as (at least to some extent) the interactions of these traits are significantly associated with the life
experiences of the interred individuals (as indicated by growth trends). More research is needed
in order to more meaningfully interpret the significant results presented in this study, as well as
to fine tune the units of analysis and statistical tests used in order to detect hidden trends in the
biocultural data.

163

Engendered Interpretations
In the above sections, it is suggested that apparent improvement in female early childhood
growth experiences over the Late Woodland/Mississippian transition is likely related to women’s
roles in agricultural production. Furthermore, some Mississippian women enjoyed better overall
biosocially significant health and growth conditions through their membership in certain status,
kin, or community groups. An alternate interpretation is that the biosocial experiences of
Mississippian women changed over time, improving as Mississippianization of the region
progressed and women became more adept at agricultural practices. These suggestions
(particularly the latter) require further contextualization within the limited literature on late
prehistoric gender relations in the Midwest.
Late Prehistoric Women, Domestic Economies, and Social Power
O’Gorman (1995, 2001) investigates gender, domestic economies, and insipient social inequality
at Tremaine, an Oneota site in southwest Wisconsin. Through an investigation of storage feature
distributions, mortuary ritual, and artifact assemblages among Tremaine longhouses, she
suggests that certain women at the site were able to gain, maintain, and exercise social power
through the production and management of surpluses. She explains,
“Using Johnson's (1982) proposed mechanisms for dealing with scalar stress, I suggest
that social inequalities emerged at the household level as simultaneous or vertical
hierarchies were needed to coordinate and regulate the utilization of some resources. Less
permanent, short-term sequential hierarchies probably occurred whereby short-term
decisions were made by consensus. However, as some households became more
successful at acquiring resources (subsistence or labor), it is possible that social
inequalities between households and between women also developed. Surpluses of food
and other items may have been used by women to gain and demonstrate social position
by providing the necessary provisions and wealth to undertake socially enhancing
ventures, such as warfare, hunting, or religious ceremonies.” (O’Gorman 2001, 26)

164

This model of insipient social inequality among Oneota women provides a potentially
helpful perspective for interpreting the results of the current analysis, which indicate socially
patterned variability in growth experiences among females at Schild. There are, of course, certain
caveats. The Oneota tradition is an expression of Upper Mississippian culture that is distinct
from the Middle Mississippian cultural tradition practiced by the communities who interred their
dead at Schild. However, Oneota and Middle Mississippian cultural traditions are known to have
shared several key characteristics, including the practice of maize agriculture. Iconography
clearly demonstrates an ideological link between Mississippian women and agriculture.
O’Gorman cites ethnographic literature to support her assertions, which she can do because of
the established archaeological links between the prehistoric Oneota and historically known
tribes. Although direct archaeological links have yet to be made between prehistoric
Mississippian groups in Illinois to historically known tribes, ethnohistoric interpretations of early
Euroamerican encounters with native peoples of the Southeast are helpful, as several of these
tribes continued to practice a Mississippian lifestyle into historic times. Nearly all of the
accumulated

evidence,

archaeological,

ethnohistoric,

or

iconographic,

suggests

that

Mississippian women were responsible for agricultural pursuits.
Furthermore, although men may have been more active in political and public life in a
normative sense, ethnohistoric accounts of native women in the southeast and beyond
demonstrate that there were definite avenues through with industrious women could influence
political decisions. Furthermore, women and/or their matrilines were said to own the fields that
they worked in and the houses that they lived in (Hudson 1976). It is not a particularly risky leap
of faith to suggest that some Middle Mississippian women may have been able to manipulate

165

their productive and reproductive power to the social benefit of themselves and their female kin,
whom they presumably worked closely with and passed on practical skill sets, both subsistence
and social in nature. Thomas (2000) suggests that Mississippian women in Illinois may have
gained social power through their role in salt production and exchange and, based on
O’Gorman’s model, it is suggested here that they may have done the same at Schild by
manipulating their roles as producers, reproducers, and managers of domestic economies.

Acculturation, Ethnogenesis, and Gender Transformations
Another explanation offered for the variation in female growth patterns between burial areas
considers time and acculturation. Goldstein points out that Knoll A is earlier, and that the
individuals interred in what has been categorized as burial area four in the current study may
represent the earliest burials at the site. Delaney-Rivera (2007) evaluates hybrid vessel forms
from the Schild Mississippian cemetery, and suggests that:
“The changing nature of the ceramic assemblage, with a lower percentage of hybrid
vessels recovered in Knoll B (40% compared to 51% from Knoll A), may be evidence of
the acculturation of the local Late Woodland population to a Mississippian way of life.
Through time, fewer Late Woodland individuals from a Late Woodland culture were
brought into the Mississippian one because the local population from Lower Illinois
River valley had already acculturated, thus fewer hybrid vessels. The percentages of
vessels may also support the relative use of the cemeteries, with Knoll A assumed to have
been older. The hybrid vessels, therefore, represent attempts by culturally Late Woodland
individuals to demonstrate their new affiliation. Through time fewer Late Woodland
individuals were present in the surrounding region, hence the lower quantity of hybrid
vessels from Knoll B. (Delaney-Rivera 2007, 320)
This line of reasoning supports Goldstein’s (1980) contention of an earlier use-period for Knoll
A, while indirectly bolstering a hypothesis that the group of individuals from burial area four
may have been the earliest at the site. A future study may look specifically at the ceramics
166

interred with the group four females to determine whether a disproportionate number of the
hybrid pots were interred with them. In any case, Delaney-Rivera’s (2007) contention of
archaeologically visible, active acculturation in Schild Knoll A provides an interesting platform
from which to explore the meaning of growth discrepancies among females observed in the
current study. Based on this, it is suggested that the growth patterning among females at Schild
may be related to changing gender roles through time with insipient Mississippianization and
continued ethnogenesis at this site. Further on this point, and related to the ideas presented by
O’Gorman (2001), Delaney Rivera continues:
“Although Late Woodland women were familiar with maize (the archaeological data
support an increase in the use of maize among Late Woodland populations through time),
Late Woodland women marrying into Mississippian families had to learn to manage
crops differently, as well as engage in new types of community activities and other social
obligations associated with the Mississippian worldview.” (Delaney-Rivera, 2007)
As Mississippianization occurred, local women adapted their life skills to accommodate and
thrive in a new cultural context. Growth differentiation between group four females and other
females interred at the site may represent the biological expression of this temporally-based
cultural phenomenon. What we may be seeing then, with improved growth conditions in other
burial areas compared with those observed in group four, is the way that active process of
Mississippianization is expressed through women gradually becoming pre adept at an
agricultural lifestyle through the time period represented at Schild, which Delaney-Rivera (2007)
suggests may be as little as 100 years.

Women and Power in Mississippian Society
The unique nature of particular female burials in the Schild Mississippian might be interpreted as
emphasising the special and powerful social roles that certain women played within this
167

community. For example, burial 66 from area three in knoll A was initially assumed to be male
based on grave inclusions traditionally interpreted as symbolizing masculine power, such as
knives, an antler staff, bird wings, and the disarticulated remains of other individuals.
Osteological analysis later revealed, however, that the remains were that of a female (Cook
1980). Furthermore, the data presented in this dissertation reveals that the female represented by
burial 66 may have experienced somewhat better growth conditions than her Mississippian
female counterparts at Schild, as burial 66 values were higher than average in femur length and
all vertebral dimensions. This case generally holds true for burial 96, who was also interred with
and also mistakenly identified as a male based on these artifacts. These examples underscore the
fact that women did, in fact, obtain roles of significant social power in Mississippian societies,
while also highlighting traditional archaeological biases toward associationing men with political
power.
Explicitly engendered Mississippian mortuary analysis with which we might
contextualize osteological data presented here are rare, but Sullivan’s (2001, 2006) work with
Mississippian data from the Toqua site in Tennessee offers a reasonable starting point for
interpreting that patterns seen at Schild. The mortuary situation at Toqua is inherently different
from that seen at Schild, as there is a clear spatial differentiation- between mound burials and
village burials- which appears to be related in some manner to gender differentiation.
Specifically, Sullivan points out that traditional interpretations rely on assumptions that men
dominate the public realm and are therefore more powerful than women, and that this indicated
by the preponderance of males interred in mound contexts and females interred in village
contexts. Sullivan’s analysis demonstrates, however, that a multifactoral mortuary perspective
that incorporates age and grave goods as variables results in a more nuanced perspective on

168

Mississippian gender relations. The level of variation indicated by burial treatments among
males and females when multiple lines of mortuary evidence are incorporated shows that
gendered Mississippian social identity may have varied with age over the lifecourse, and that
individual achievement may have played a larger role in social identity than most archaeologists
have previously considered.
In light of such a perspective on Mississippian gender/sociopolitical dynamics, it might
be suggested that the females identified here as Schild burials 66 and 96 may have achieved
positions of considerable sociopolitical power within the Schild community. As these are only
two individuals, the osteological evidence cannot be presented with statements of statistical
significance. However, disparate age-sensitive growth patterning between these two females, as
indicated by the current analysis suggests, that positions of female social power are not
necessarily related to relatively improved growth conditions throughout life, as burial 66 appears
to have experienced improved conditions throughout her life as indicated by higher than average
NCD and femur length measurements, while burial 96 may have experienced relative
improvement (i.e. catch up growth as indicated by higher than average femur length without
concomitant high NCD values) during the adolescent years of her life. Perhaps these
osteobiographies are indicating a bioarchaeological method for differentiation ascribed (burial
66) versus achieved (burial 96) female power.

What About the Men?
As was indicated by LEH analysis, Mississippian males seem to have experienced fewer episodic
growth disruptions than their Late Woodland counterparts. Although it appears that males in the
most exclusive Mississippian artifact only clusters (cluster 0 in the current analysis) experienced

169

fewer episodic growth disruptions as indicated by LEH, this result was not quite statistically
significant (α=.083).
Despite these examples, one of the most obvious patterns to have emerged in this analysis
is that statistically significant results were, in most cases, associated with females. As was
pointed out in several of the interpretations outlined above, Schild women seem to have had
disparate growth experiences related to their membership in one or more intra or intercommunity
social and status groups. As interesting as this fact may be, the lack of similar variability among
Schild males is also noteworthy in and of itself, and warrants further exploration.
One obvious interpretation that would constitute a natural extension of the interpretation
proposed for the female data would be that, for Schild men, the biosocial conditions of growth
were more regular across different social groups when compared to Schild females. It may be
that Mississippianization and its focus on maize agriculture brought with it opportunities for
certain Schild women to advance their social position and improve the quality of life for
themselves and their daughters. This is not to say that female children were necessarily treated
better than their male counterparts, or that these the observed patterns were necessarily the result
of conscious decisions. But this research indicates that certain female children were provided
significantly better conditions of cumulative growth, that those differences seem to be correlated
with membership in certain social subsets of the Schild community, and that the differentiation
in growth conditions may have occurred during adolescence as opposed to early childhood (as
there were not concomitant differences in neural canal diameters). It may be that very young
female children who were not yet active contributors to their communities were all viewed
equally, and on equal terms with their male siblings. As male children aged and spent more time
with their fathers, female children would have remained with their mothers and aunts, who

170

would continue to teach them the social and technical skills necessary to manage village
activities, including managing agriculture pursuits. As they came to be viewed as contributing
adult women, certain among them may have been treated somewhat preferentially, perhaps due
to their association with household, kin, or community networks who exercised particularly
noteworthy managerial or productive prowess within the broader region. The conditions of
growth among male children, either in early childhood or later adolescence, did not seem to vary
based on group membership, and this may mean that Mississippian boys and young men at
Schild were afforded similar biosocial conditions of growth in spite of status or kinship
differentiation.
However, it must also be noted that, although cell sizes in ANOVAs were also small for
females, the problem was somewhat exacerbated for males. In fact, many of the statistically
significant patterns observed for females (variation in cumulative growth indicators by burial
area, for example) were reflected in the male data, but not to a statistically significant extent.
Furthermore, Schild Mississippian sample is skewed toward females and it is possible that a
certain subset of the Mississippian community (presumably skewed toward males) was interred
elsewhere (Cook 1984). These may have been individuals of higher or lower social status, or
higher or lower ranking kin or community groups, but this cannot be said for certain unless the
other group was to be located and analyzed. Goldstein (1980) points out that some of the
variation in mortuary patterning at Schild seems to be related to sex, with many of the clusters
(artifact only as well as artifact and positioning clusters) dominated by one sex or the other. This
suggests that, although a segment of the male population seems to be missing from the Schild
sample, at least some portion of male social differentiation is represented within this skeletal
series and we should still be able to make some cursory conclusions regarding the lived

171

experiences of Schild men. Based on the analysis presented here, there seems to be less variation
in the growth experiences of Schild males when considered in relation to females at the site, and
this may be related to 1) the ability of women to exercise social mobility related to the role they
play as producers (of maize) and reproducers (of humans), and/or 2) the biocultural results of
Schild women actively acculturating and adapting to a new set of sociocultural circumstances
with Mississippianization.
The multifactor two way ANOVAs, on the other hand, did reveal some significant
interactions between space, artifact assemblages, and growth indicators for the male subsample.
In particular, male lumbar AVBH varied significant depending on artifact cluster and knoll,
while male femur length varied significantly based on artifact cluster and charnel association. As
was mentioned above, these results indicate the complexity of mortuary phenomena and
generally suggest that spatial and artifactual parameters considered together may indicate
something about the social lives of Schild individuals that is related to biocultural patterns of
growth. However, it is difficult to comment about what these patterns specifically mean
regarding men and women at this site.

Regional Perspectives on Mississippianization, Health, Growth, and Social Identity
Broader regional comparisons are helpful for more completely understanding variability in the
biocultural effects of Mississippianization on intracommunity social relations. The nontraditional methods used in this study preclude direct comparison to most of the comparative
literature. However, the primary goal here is to contextualize the interpretations provided in this
chapter within a regional framework of Mississippianization, health, and gender transformations.
Studies from the American Bottom, Dickson Mounds, and Moundville are discussed.

172

Dickson Mounds
Much of the growth data provided in Dickson Mounds reports is based on juvenile skeletal
remains, which cannot be reliably sexed. This means that some of this information cannot be
directly compared to the current analysis. However, certain aspects of these reports are useful for
contextualizing the findings of the current study within a regional framework. For example,
Goodman (1980) reports that his analysis of tibial growth patterns in subadult skeletal remains
indicates increasing childhood stress with Mississippianization, with the years between two and
five being the most stressful. This likely coincides with “weaning stress” reported by Blakely
and Armelagos (1985) and Goodman et al (1984), the effects of which are seen earlier and
expressed more strongly during Mississippian times, when nutritionally poor maize gruel could
be used as a transitional food for very young children.
It could be reasonably hypothesized that this decline in overall childhood health and
growth conditions would similarly manifest itself in vertebral growth as a decrease in neural
canal diameters. And, indeed, Clark’s (1985) analysis does indicate a statistically significant
decrease in APNCD in both males and females over the Late Woodland-Mississippian transition.
This observation seems to lend further credence to the utility of APNCD as an indicator of early
childhood growth. However, Clark (1985) also points out that TNCD decreased in females and
increased in males over time. This fact, combined with the complicated statistical results
regarding TNCD in this dissertation, indicate that the meaning of this vertebral parameter
requires further exploration. Based on this, it would appear that the biocultural circumstances of
early childhood growth in the Central Illinois River Valley seems to generally decline over the
Late Woodland-Mississippian transition

173

Overall, Clark interprets and contextualizes his vertebral data as follows: Over the Late
Woodland-Mississippian transition, both Mississippian males and Mississippian females
experienced stunted early childhood growth compared to their Late Woodland counterparts, as
evidenced by decreased APNCD (significant only in females). Later early childhood growth and
adolescent growth (as evidenced through TNCD’s and PVBH’s respectively) improved
significantly for males with Mississippianization. Clark suggests that all of this, taken together,
likely reflects an increase in social status for Mississippian males, coinciding with the adolescent
years.
Application of similar methodologies to Lower Illinois River Valley samples in this
study, however, paints a different picture of gendered growth experiences than that observed in
the Central Illinois River Valley. In the Schild and Yokem samples, male growth patterns remain
statistically unchanged over the Late Woodland-Mississippian transition. For females, however,
early childhood growth improves over the transition, as evidenced by APNCD’s and, to a lesser
extent, TNCD’s. This may indicate better treatment of female children over time as agriculture
becomes more important in this region, perhaps to the extent that female children were
(consciously or unconsciously) provided preferential treatment. However, as indicated by lack of
statistically significant change in vertebral body height and femur length data, this amelioration
of growth conditions did not last throughout the entire growth period. Perhaps this means that
female productive (i.e. agricultural) potential was emphasized and recognized in a particularly
notable manner during early childhood and that, by adolescence, both males and females were
seen as equal contributors to the community. However, the cumulative growth data from the
Schild-only analysis forces us to recognize that not all Mississippian females interred at this site
experienced this sustained, favorable cumulative growth conditions equally, as some had better

174

cumulative growth conditions based on certain intracommunity social parameters (see above
discussion). Unfortunately, we cannot know the extent to which the Mississippian pattern at
Schild is a departure from the Late Woodland situation (without further mortuary analysis) and,
furthermore, similar mortuary-based analyses among the Dickson Mounds population are not
available for direct comparison with this segment of the current study.

American Bottom
Rigorous bioarchaeological studies of American Bottom skeletal samples are sparse and tend to
be limited by poor preservation and small sample sizes. Furthermore, much of this literature
deals more with high status burials than it does with general, community level changes over the
Late Woodland-Mississippian transition, which may inherently result from the lack of large Late
Woodland mortuary sites in the region (Emerson et al 2003). This may be partially due to
preservational factors, but Carter (2003) points out that poor preservation alone cannot account
for the lack of Late Woodland interments in this region. Isolated interment and scattered human
remains in village deposits is the most widely documented context for human remains in this
area (McElrath et al 2000). Other reports have indicated that mortuary treatments at this time and
place were highly variable and complex and that the nature of these practices may require
unconventional methods in order to detect, let alone interpret, these processes.
Ambrose et al (2003) present a discussion of status and gender related dietary variation
among the Mound 72 burials at Cahokia. Individuals afforded particularly ornate burial
treatments are considered to be high status, while those in mass graves are considered to be of
lower social status. They present isotopic evidence that the higher status individuals consumed
significantly larger amounts of meat compared to the lower status individuals, yet consumed only

175

slightly more maize. However, low status females in Mound 72 (interred in a mass grave)
consumed significantly more maize than any of the other burials. These presumably low status
females also suffered from increased incidence of stress and poor health (Fowler et al 1999).
Based on this analysis, one might suggest that low status Mississippian women
experienced significantly less favorable biocultural conditions than either their low status male
counterparts or high status individuals of either sex, which is not congruent with the trends
detected in the current analysis. However, the nature of the Mound 72 sample complicates
interpretation. First, Mound 72 is likely part of a complex, theatrical mortuary ritual that does not
represent “typical” Mississippian burial treatments (Porubcan 2000, Goldstein 2000, Brown
2003). Second, it is unclear whether the individuals interred in the mound, particularly the low
status females, were from Cahokia or elsewhere in the Mississippian world, nor do we know
whether they were slaves, captives, or members of some other non-representative social group
(Ambrose et al 2003).
Although the status and gender differentials seen in the Mound 72 burial sample are
intriguing and may provide clues to a normative, ideological basis for Mississippian
intracommunity social/gender differentiation and concomitant biocultural circumstances of
health, they contribute little to an elucidation of the lived experiences of the general
Mississippian population. Non-elite Cahokian cemeteries have been excavated, but the remains
are typically poorly preserved, presumably complicating biological analyses. Fingerhut, for
example, is an early Mississippian cemetery at the Cahokia site (Klepinger 1978, 1993, Witty
1993). Bone preservation was poor, but it was clear that all age and sex groups are represented.
The few graves goods recovered did not appear to be associated with any particular
demographic. Males, females, and children were sometimes interred as bundle burials. This

176

indicates that secondary treatment was accessible to a segment of this community, likely based
on status differentials more than age or sex alone. This is similar to mortuary phenomena seen at
Schild, but the lack of detailed bioarchaeological reports preclude biocultural comparison.
Although we are unable to link these social parameters to stress/health experiences, it does
suggest that both males and females qualified for somewhat elaborate mortuary ritual within the
general Cahokian population.
Milner (1982) provides what might be considered the most comprehensive account of
Mississippian health in the American Bottom, but most of the data is presented in a way that
does not facilitate comparison to the study presented here. His analysis of LEH among Kane
Mound individuals is not explicitly presented or interpreted in regard to sex differentials. He
does discuss sexual dimorphism to some extent, but not in a way that is comparable to how it
was approached in the current study. Future research may focus on applying Milner’s analytical
methodologies to the Schild and Yokem femur data so that levels of sexual dimorphism can be
directly compared.

Moundville and the Late Prehistoric Southeast
Powell (1991) provides status-based summary data on the Mississippian skeletal series from
Moundville that is reasonably comparable to the results of the current study, as maximum femur
length is among the parameters she explores. She divides the sample into groups based on cluster
analyses by Peebles and Kus (1977) although the presumed highest status burials (from the
mounds) were unavailable. The “higher status” clusters were collapsed into two categories, based
primarily on cemetery type (cemetery near a mound or cemetery near a village), with those
interred near mounds being considered of higher status. The third group, referred to as

177

“residual,” consisted of those remains that had no grave goods, regardless of cemetery type. No
assumption regarding relative rank was made for the residual group.
Powell’s analysis of pathology, trauma, and dental wear yields no statistically significant
differences between status groups. Of particular interest is her discussion of adult body size, for
which she also determines that there were also no statistically significant differences (although
she does point out that mean male body size, based on femur length and measures of robusticity,
were slightly larger for the elite group, but not significantly so). She concludes that status
differentiation at Moundville did not seem to translate into biosocial differentiation in health,
diet, or growth experience, but cautions that exclusion of the highest status mound burials may
have biased the results.
This provides an interesting comparative case for the research presented in this
dissertation. Although she does not explicitly frame it as such, Powell’s research design
incorporates, a priori, a mechanism for detecting within-sex variation. This is one way of
addressing the problem of essentialism when characterizing the experiences of men and women
in the past. As argued in chapter one, bioarchaeological analyses that attempt to account for the
ways that experiences vary within groups based on biological sex based on the inclusion of
social institutions units of analysis like status (White 2005, Martin 2000, Cucina and Tiesler
2003, White 2005) and ethnicity (Schurr 1992) have greater potential for clearly assessing
gender relations in the past. Although variable growth experiences attributable to social
differentiation within each sex were not detected (as they were in the current study), this result
does not preclude the possibility that social differentiation might be expressed in some way other
than cemetery type. Furthermore, as Powell stresses, inclusion of high status mound burials
could have affected the result. Although she does not contextualize her results within the gender

178

literature, it should be emphasized that this study was published in 1991, when gender
archaeology was in its infancy and had not yet begun to have a measurable influence on
bioarchaeological interpretation. Overall, Powell’s analysis provides as solid example of a
research design that incorporates multiple units of analysis and has the potential for beginning to
approach the intricacies of gender in the past. Furthermore, when compared to the results of the
current study, Powell’s results indicate that females had less variable growth experiences at
Moundville than they did at Schild, although the mortuary units of analysis are not necessarily
directly comparable.

Mississippian Identity, Gender, and Health
Based on the their review of wall trench structures in the Lower Illinois River Valley,
Farnsworth et al (1991) suggest that Late Woodland communities in this region adopted
Mississippian mortuary ritual while maintaining a predominantly Late Woodland lifestyle.
However, Goldstein (1980) contends that the Lower Illinois Valley was occupied by dispersed
Mississippian farmsteads, as indicated by the presence of Mississippian pottery, mortuary ritual,
and architectural styles. Furthermore, Delaney-Rivera’s (2004) analysis of Schild ceramics
indicate that Late Woodland communities in the Lower Illinois River Valley rapidly adopted a
Mississippian

lifestyle.

Both

Goldstein

and

Delaney-Rivera

contend

that

the

“Mississippianization” of the Lower Illinois River Valley was a case of full acculturation rather
than selective emulation.
The research presented in this dissertation lends some level of further support to the
Goldstein/Delaney-Rivera position. The mortuary patterning observed at Schild was clearly
Mississippian in nature, but proponents of emulation-oriented hypotheses might be unconvinced

179

that this new mortuary program was reflective of true Mississippian-style social differentiation in
the lived experiences of these people, or that these burial groupings meant much as far as
intracommunity social relations are concerned. In the above discussions, I have asserted that, to
some extent, variation in the biological conditions of growth among Schild females is correlated
with social differentiation as indicated by mortuary patterning. Noteworthy are six key
observations/arguments: 1) There is biological evidence for sex-based differences in growth
experiences in Schild and Yokem communities with Mississippianization, 2) a particularly wellstudied Mississippian burial program existed at Schild, 3) mortuary programs likely reflect ideas
about social structure in living communities, as expressed through treatment of the dead, 4)
discernable patterns in biocultural experiences of Schild females correlated significantly with
these social categories 5) Mississippian women likely exercised new types of managerial and
social power and roles as agriculturalists 6) The overall cultural picture at Schild is that the burial
program reflects, to some extent, a social structure that had real effects on the lives of
community members, and Mississippian social structure and gender relations were established in
community life.
If Schild people were burying their dead according to a Mississippian mortuary program,
and that mortuary program reflected social differentiation within communities in a way that
becomes more apparent in analyses of health and stress, then it would seem difficult to contend
that these people were essentially Late Woodland people with a superficially Mississippian
burial program. Of course, bioarchaeological analyses alone cannot settle this question, and
much more work needs to be done regarding habitation sites in the region. But it is suggested
that that the current research contributes important, if limited, data for exploring the nature of
Mississippianization in the Lower Illinois River Valley.

180

CHAPTER EIGHT- CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS

The previous chapters present specific results and discussions of data analysis and contextualized
interpretations. In this chapter, conclusions are presented succinctly and future research
directions are proposed. The contributions of this research to investigations of the utility of the
vertebral method for growth assessment will be stated, and interpretations summarized. This
chapter concludes with future research questions that build upon the study presented here, and
are predominately centered on two major themes 1) clarifying remaining uncertainties regarding
the nature of human vertebral growth, and 2) expanding gendered perspectives on growth to new
skeletal collections.

Summary
Utility of Vertebral Method
The literature regarding the utility of vertebral dimensions as indicators of growth conditions
across the life span is limited (Clark 1985, 1988; Tatarek 1999, 2005), and the current analysis
provides supplementary data interpreted within a unique context of cultural change and
intracemetery variation. In the mortuary-based testing, significant results were limited to the
lumbar segment. The explanation for this may be anatomical in nature, as the timing of growth
cessation of certain subsets of the thoracic segment (i.e. upper thoracics versus lower thoracics)
may skew thoracic segmental mean-based results. Cursory observation of the vertebra-byvertebra analysis lends some support to this hypothesis, which should be further tested through

181

experimentally dividing the thoracic data into anatomically-based subsets and subsequently
recreating the tests performed in the current analysis.
Clark (1985) suggests that TNCD may represent later early childhood growth while
APNCD may be indicative of early early childhood growth. That significant results were
obtained for APNCD and not TNCD in the current study is, therefore, not entirely surprising, and
so it is possible that these two measures are capturing the growth conditions of somewhat
different time periods. NCD was not correlated with femur length, suggesting that these
measures are recording a kind of growth that is fundamentally different from one another. This
fundamental difference is likely an issue of timing, with NCD representing only the early
childhood period of growth, which is not inherently related to conditions of cumulative growth
through adolescence and early adulthood. Diachronic change in the nature of sexual dimorphism
in NCD in the study samples examined here indicate that some non-genetic, non-sex linked
factor affects canal size, and it is suggested that this may be due to conditions in social or natural
environments. However, the unclear relationship to the cumulative picture of early childhood
growth provided by NCD and the episodic perturbations detected through LEH requires further
exploration through application of this research design to new skeletal samples.
This research demonstrates a reasonably strong relationship between VBH and femur
length. These measures were significantly correlated with one another, suggesting that they are
both measuring a similar phenomenon. However, R-square values for the strength of the
association were low, indicating that there might be some other factors determining differential
cumulative growth between these elements. Furthermore, it is unclear why both AVBH and
PVBH were both correlated with femur length, yet only AVBH yielded any significant results in

182

the mortuary-based analyses. Anatomical explanations regarding the effects of spinal curvatures
were suggested, but this is a problem that requires further exploration.
The mortuary-based portion of this research also showed that femur length and VBH
seem to be measuring similar, but not identical, growth phenomena. In the mortuary-based
analyses, there were some instances where AVBH indicated growth differentiation between
groups that was not detected by femur length. However, there were no cases where femur length
was statistically significant without concomitant significant (or very near significant) difference
in lumbar AVBH. This may suggest that vertebral body heights are somewhat more sensitive to
growth disruption than femur length, but the specific biological and cultural meaning of this
pattern cannot be known for certain without further research.
Overall, it is suggested that the vertebral method is a potentially sensitive indicator of
both early childhood and cumulative growth conditions. Although the exact nature of the growth
and how it related to social conditions requires further investigation, the results of this study
suggest that it cannot be rejected as a useful tool for bioarchaeological analysis. The analysis
and interpretations outlined in this dissertation show that vertebral measurements are a
potentially useful tool for the assessment of stress and growth that can complement more
traditionally employed methods, such as LEH and long bone length.

Growth Trends Over the Late Woodland- Mississippian Transition
When applied to a bioarchaeological analysis of the Late Woodland-Mississippian transition at
Schild and Yokem, NCD data suggest that females experienced improvement in early childhood
growth conditions, which is in keeping with traditional archaeological and ethnohistoric
183

interpretations of women’s increasing social status as agriculturalists. Analysis found that female
lumbar APNCD increased significantly with Mississippianization. Lumbar TNCD increased
among females as well, and was very nearly significant. This supports the hypothesis that rowth
conditions for young females improved over the Late Woodland-Mississippian transition, as
women’s roles as agricultural producers became more highly valued in an increasingly foodproducing society. Although analyses of LEH frequencies did not reflect the same trend, it is
suggested that this does not necessarily negate the vertebral results, as LEH and NCD may be
measuring fundamentally different growth processes (i.e. episodic versus cumulative stress).
Male Mississippian children appear to have experienced fewer episodic growth
disruptions during early childhood than did their Late Woodland counterparts, as suggested by a
statistically significant decrease in LEH frequency. There were no significant changes in NCD
over the transition for the male sample. The reason for the incongruent results of the LEH and
NCD portions of the temporal analysis may be due to the disparate nature of the growth
phenomena being recorded, as explained above for the opposite discrepancy in the female
sample.

Gender and Growth Among Schild Mississippians
There was no statistically significant improvements in cumulative growth through early
adulthood for either males or females in the temporal analysis, as indicated by either VBH or
femur length. However, the intracemetery component of this analysis suggests that not all
members of the Schild Mississippian community enjoyed this apparent stability in cumulative
growth to the same extent. This portion of the analysis was designed specifically to engender the
184

analysis through an investigation of culturally meaningful, within sex variation in growth
patterning. Gender categories are not unitary, stagnant phenomenon, and within the broad
groupings of “men” and “women”, there exists much potential for variable biosocial experiences
based on the interaction of one’s gender with other facets of social identity, such as status,
kinship, age, ethnicity, etc. Clear mortuary groupings and clusters emerged in Goldstein’s
spatially and artifactually based analyses of the Schild Mississippian cemetery, and these
groupings were likely reflective of significant social differentiation within the community
represented by this skeletal series. Hypothesis five recognizes the potential for there to be
variation in the biosocial conditions of growth for men and women based on these socially
significant groupings, and the research design was created in a way that allowed for potential
detection of this variation.
More detailed, gender-oriented analysis of the Mississippian component of the Schild
Cemetery yielded within-sex variability in cumulative growth patterning correlated with social
differentiation, as indicated by mortuary analysis. Vertebral body height data provides limited
evidence for slight discrepancies in cumulative growth patterns in females depending on which
artifact and positioning clusters they belonged to, and this may be related to their social status
within the Schild community. Stronger evidence based on both vertebral body heights and femur
length suggest that females interred in certain burial areas, and particularly those associated with
charnel-focused activity, experienced better cumulative growth conditions than females in other,
non charnel areas. Because the spatial patterning of these areas corresponds to sets of rows
recognized in Goldstein’s (1980) analysis, it is suggested that access to specialized burial
treatment may have been related to membership in certain kin or community networks, or some
similar manner of intracemetery differentiation. Women in some of these groups may have been
185

able to exercise more or less social power through their particular social circumstances, perhaps
through managing productive and reproductive activities. As young women came of age and
worked closely with their female kin, those who were members of particularly successful
households or lineages would have benefitted from their success, perhaps becoming visible
bioarchaeologically through improved growth.
Despite increasingly small sample sizes as tests became more complex, two-way
ANOVAs simultaneously comparing osteological measurements with two categorical factors
(one spatial, one artifactual) yielded five significant results, indicating that seemingly nonsignificant relationships between a biological dimension and any individual factor may be
masking more complicated multifactoral interactions between artifact assemblages, space, and
biology. It is suggested that these results indicate that that a suite of mortuary characteristics is
may be as meaningful than any one parameter on its own.
Overall, the data suggest that women at Schild experienced variable cumulative growth
conditions based not only on their biological sex, but on their membership in certain status or
kinship, or community groups, represented archaeologically through artifactually and spatially
based mortuary differentiation. Some Mississippian women enjoyed better overall growth
conditions related to their membership in certain status, kin, or community groups. An alternate
interpretation, based on the proposition that burial area four (where females experienced the least
favorable conditions of growth as evidenced through VBH and femur length) is that the biosocial
experiences of Mississippian women changed over time, improving as Mississippianization of
the region progressed and women became more adept at agricultural practices. Preliminary
multifactoral analysis indicate that both spatial and artifactual/positional categories were
important components of the Mississippian burial program and reflect meaningful sociocultural
186

divisions, as (at least to some extent) the interactions of these traits are significantly associated
with the life experiences of the interred individuals (as indicated by growth patterns). Hypothesis
five is therefore accepted. More research is needed in order to more specifically interpret the
significant results presented in this study, as well as to fine tune the units of analysis and
statistical tests used in order to detect hidden trends in the biocultural data.

Future Research Directions
The research presented in this dissertation generated several new research directions, which will
be further explored using the data collected for this project, as well as new growth data from
skeletal collections in West Central Illinois and beyond. The collection of new data on vertebral
dimensions and other growth indicators will provide more data for the further exploration of
vertebral growth and its relationship to stress and growth disruption. More specific descriptions
of potential research projects are described below.

Further Methodological Research
NCD as Maternal Health- Ursu et al (1996) show that in the lumbar region of the spinal column,
transverse neural arch diameter in L1 through L4 reach 70% of adult size by the time of birth.
The intrauterine environment is generally thought of as offering a high level of protection to the
unborn fetus, buffering it from physiological stresses in all but the most extreme cases of
maternal illness or malnutrition (Cook and Buikstra 1979). In a study of juvenile skeletal remains
from Middle and Late Woodland Illinois contexts, Cook and Buikstra (1979) show that the age at

187

death distributions of juveniles with a prenatal enamel defect was significantly different than the
age at death distribution of juveniles without prenatal defects. The nature of differential survival
shows that those individuals who suffered the effects of physiological stress prior to birth were at
increased risk of dying earlier than those who did not experience prenatal stress.
Future research will investigate neural canal diameters in subadult remains in a similar
manner. The relationship between age at death and neural canal diameter can be tested using
linear regression, and it is hypothesized that there will be a positive correlation between age at
death and NCD. Subadult remains can typically be aged with relatively high accuracy, which
will contribute a certain level of detail to this analysis. Because adult remains cannot be aged as
precisely, broader categories such as young, middle-aged, and old adults are sometimes used.
While this does not completely preclude a similar analysis among adults (i.e. one that tests for a
relationship between neural canal diameter and age at death), it may require statistical analysis
like ANOVA rather than linear regression, as age would be treated as a categorical variable in
this instance.
Another approach may be to measure the NCDs of subadults and compare those
measurements to those individuals who survived into adulthood (i.e. those measured for this
study). Because juvenile skeletal remains cannot be sexed, a sex balanced sample of the adult
remains would likely provide the soundest comparisons, assuming that the childhood mortality
sample is not skewed toward one sex or the other (which is difficult to assess). In this scenario,
one would expect NCD in the subadult sample to be smaller than the adult sample if NCD is
indeed a good measure of early childhood growth disruption.

188

Similar analyses have been done using adult remains, and the results have been mixed.
Clark et al (1985) found a statistically significant relationship between spinal stenosis and earlier
age at death using the Dickson Mounds series, but Tatarek (1999) found no such relationship
using the Hammond Todd collection. The nature of this discrepancy is presently unclear, but
may be due to disparate nature of the two collections. In any case, such an investigation using the
Schild and Yokem data can contribute to this research area. Furthermore, and aside from
paleodemographic concerns, detailed statistical investigation of the relative proportion of early
versus later childhood death, prenatal enamel defects, and neural canal diameters may provide a
picture of change or stability in maternal health over the Late Woodland-Mississippian
transition.
Focus on L5 APNCD- Ursu et al (1996) point out that, at the time of birth, the fifth lumbar
vertebral canal has approximately 50% of its adult size, while the other lumbar vertebral canals
are already 70% complete. This means that the fifth lumbar has an extended opportunity to be
affected by environmental factors compared to the other lumbar vertebrae, as more L5 growth
occurs postnatally, when there is no protection by the intrauterine environment. Furthermore,
APNCD is perhaps focusing in more specifically on the early early childhood period (Clark
1985) and is the only vertebral measure consistently and significantly correlated with actual
canal size (Gepstein et al 1991), perhaps making it a better choice for investigating early
childhood neural canal growth when compared to TNCD.
Several results from the current study seem to reflect this phenomenon. Cronbach’s
Alpha testing indicated that, for most vertebral dimensions, internal reliability was improved
when L5 was removed from analysis, suggesting that L5 dimensions were more variable and less
predictable than the same dimension in the other lumbar vertebrae. This was particularly true for
189

APNCD. Additionally, the only significant result for APNCD in the mean-based analysis was
between Late Woodland and Mississippian females, with Mississippian females being
significantly larger. Individual vertebral analysis indicated that most of this difference came from
the L5 APNCD, which makes sense considering Ursu et al’s (1996) observation. Similar
research on new skeletal collections will isolate and focus on L5 in order to determine whether it
is a better indicator of early childhood growth disruption on its own rather than measuring all
five lumbars and combining them into a means based analysis. Not only may L5 be a more
sensitive indicator than a lumbar means-based analysis, but it is also far less time consuming to
measure APNCD in one vertebra rather than five or more.

Subdivision of Vertebral Segments- The vertebral column is traditionally conceptualized as
having three distinct segments, which are based on real functional and morphological
differences. However, visual observation of variation in the size and shape of vertebral bodies
and neural canals reveals that the change is gradual as one moves up or down the column, even
over the transition from one segment to the next. In this way, the division of the current study
sample in to thoracic and lumbar segments is somewhat arbitrary. This, coupled with patterns in
the individual, vertebra by vertebra analysis indicate that further division of the vertebral
column, and the thoracic segment in particular, might provide for a more valuable method of
investigating vertebral growth and assessing growth disruption, or that careful scrutinization of
individual vertebral results may serve to fine tune or otherwise inform the results of the
segmental mean based analysis.

190

For example, analysis revealed that APNCD was significantly different between Late
Woodland and Mississippian females, and in the segmental mean based analysis, this result was
only significant in the lumbar. Table 24 shows that APNCD was significant in the first and
second thoracic vertebrae, which may be related to the timing of NCD growth cessation in this
particular reason. Perhaps more informative, though, are the significant APNCD results for the
11th and 12th thoracic vertebrae. If we consider the entire sequence of significant APNCD results
throughout the lower vertebral column, regardless of segment, nearly the entire lower third of the
vertebral column (except L3) yielded significant results in APNCD. Similar cases are provided
when other patterns of significant results are observed, such as with male AVBH by burial area
in table 68 and female APNCD by artifact only cluster in 72. This may indicate that either 1)
further subdivision, based on growth timing rather than morphological/functional segment) may
be a better means of investigating growth disruption, or 2) the overall pattern of significant,
vertebra-by-vertebra results should be more intensively investigated and interpreted in studies
such as that presented here. The latter suggestion seems difficult to accomplish objectively
without further investigation via the former suggestion. In any case, more research regarding
patterning in the vertebra-by-vertebra analysis is necessary.

Future Applied Research Directions
Two-way ANOVA for Sexual Dimorphism- One of the more interesting details to emerge from
the current study was that the nature of sexual dimorphism in NCD is poorly understood. Tatarek
(1999) notes that cervical NCD seems to be closely correlated with ancestry and sex. The
explanation for the former is likely genetic, as genetically similar individuals are more likely to

191

be phenotypically similar, while those individuals who are less genetically similar may be more
phenotypically different from one another. This line of reasoning is essentially, one of the
explanations that was provided in the previous chapter when discussing the issue of scale of
differentiation between the current study and the original study by Clark (1985, 1988). In the
current analysis, there was a change in the level of sexual dimorphism in NCD over time.
Specifically, there was no sexual dimorphism in APNCD during the Late Woodland, but sexual
dimorphism was evident by Mississippian times, with female APNCD being larger. The case
was somewhat different for TNCD, with males being significantly larger during Late Woodland
times and being statistically equal to females by the Mississippian. It was argued that this
variability by sex offers support to the contention that NCD is fixed in early childhood, long
before sexual dimorphism develops in other skeletal elements during adolescence. If one takes
this argument a step further, it is suggested that if there is no genetic predisposition for sexual
dimorphism in this characteristic, environmental or sociocultural explanations for observed
differences is necessary. This is, of course, a basic premise of the Clark method and, as such, was
tested on the Schild and Yokem collections.
In addition to the basic exploration included as part of this study, the bioarchaeological
literature provides other means of further exploring the meaning of sexual dimorphism,
specifically ones that address temporal changes in the degree of sexual dimorphism in
populations. For example, Lallo (1973) used two-way ANOVA to assess changes in sexual
dimorphism between the Late Woodland and Mississippian series at Dickson Mounds. Sex and
culture are used as factors, while continuous measurements of pelvic and femoral dimensions
were used as the continuous variable. A significant interaction would indicate that there was a
change in the amount of sexual dimorphism over time.
192

Lallo did not detect any changes in the amount of sexual dimorphism in femur or pelvic
measures over the Late Woodland/Mississippian transition. The current study did not test for
sexual dimorphism in this way, but future reanalysis will involve analyzing femur lengths as well
as all vertebral measures following Lallo’s methods, hence quantifying change in the amount of
sexual dimorphism over time and making them more directly comparable to Lallo’s Dickson
Mounds data.
American Slave Samples- As discussed in chapter three, Steckel’s (1987) analysis of African
slave height records documents the extraordinary ability of adolescent catch up growth to
virtually obliterate evidence of significant early childhood growth disruption related to disease
load, poor nutrition, and poor prenatal health. The skeletal remains of the individuals used in
Steckel’s analysis are, of course, not available for analysis. If they were, they would provide an
excellent additional test of the vertebral method. Application of the vertebral method of growth
assessment of other collections of slaves, from American contexts or otherwise, might provide a
useful comparative assessment of the biosocial growth experiences of enslaved peoples.
Moundville- Powell’s (1991) provides an interesting intracommunity assessment of growth
patterns by sex at using skeletal remains from the Moundville site in Alabama (see chapter seven
for discussion). She found no significant differences between status groups (as determine by
cluster analysis) for either males or females. Assessment of NCD for these individuals, and
according to the same social groupings, would complement this analysis by adding a dimension
of early childhood growth that complements that already investigated by Powell using LEH.
Late Woodland Mortuary Analysis- The bulk of the engendered bioarchaeological analysis in the
study presented here was limited to the Mississippian component of the Schild site, as formal,

193

multifactor mortuary units of analysis were available. Such detail is rare in mortuary
archaeology, and much of what is available for Late Woodland mortuary sites (as well as other
Mississippian sites) is largely descriptive in nature. Although sample size is always a primary
concern, formal mortuary analyses of more sites would allow for detailed biocultural analyses of
intracommunity social differentiation. A pending dissertation by King evaluates ideology,
biological relatedness, and mortuary treatment at a number of Late Woodland mortuary sites in
West Central Illinois, and it is anticipated that such analyses will provide useful social units for
further biocultural analyses, including gender oriented ones such as that presented in this
dissertation.
Ceramic Hybridity, Growth Trends, and Gender at Schild: The gendered interpretation of the
Schild female data presented in this dissertation relied heavily on Delaney-Rivera's (2004, 2007)
analysis of ceramic trends at Schild. Future bioarchaeological research will focus on more
explicitly and completely integrating Delaney Rivera's ceramic trend data as a critical variable,
focusing on the relationship between growth trends over time and Mississippianization vis á vis
the transition from Late Woodland, to hybrid, to Mississippian ceramic forms.

194

APPENDIX

195

Figure 16- Spatial distribution of Schild Knoll A females used in the current study (adapted from Goldstein 1980)

196

Figure 17- Spatial distribution of Schild Knoll B females used in the current study (adapted from Goldstein 1980)

197

Figure 18- Spatial distribution of Schild Knoll A males used in the current study (adapted from Goldstein 1980)

198

Figure 19- Spatial distribution of Schild Knoll B males used in the current study (adapted from Goldstein 1980)

199

Figure 20- SPSS scatterplot of Schild female lumbar AVBH and femur length
200

Figure 21- SPSS scatterplot of Schild female lumbar PVBH and femur length
201

Figure 22- SPSS scatterplot of Schild female lumbar TNCD and femur length
202

Figure 23- SPSS scatterplot of Schild female lumbar APNCD and femur length
203

Figure 24- SPSS scatterplot of Schild male lumbar AVBH and femur length
204

Figure 25- SPSS scatterplots of Schild male lumbar PVBH and femur length
205

Figure 26- SPSS scatterplot of Schild male lumbar TNCD and femur length
206

Figure 27- SPSS scatterplot of Schild male lumbar APNCD and femur length
207

Vertebral
Dimension
T1 AVBH

T2 PVBH

T3 AVBH

T5 AVBH

T6 AVBH

T8 AVBH

T8 APNCD

T9 APNCD

T10 APNCD

L3 APNCD

Burial Area

N

Mean

SD

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

16
12
11
10
17
12
12
10
16
11
11
10
15
12
10
6
15
12
10
7
15
13
10
9
14
12
9
9
15
12
10
9
16
10
11
9
16
15
14
9

17.11
16.31
17.52
16.83
20.71
20.34
21.43
20.44
18.48
17.58
18.81
17.94
19.19
18.70
19.97
18.65
20.14
19.23
10.76
19.24
21.53
21.05
22.25
20.70
15.08
16.13
15.42
15.12
14.83
15.86
15.66
14.83
15.14
16.10
15.76
15.16
14.00
15.03
14.02
13.67

0.71
1.01
0.71
0.98
1.01
0.75
0.78
1.04
1.19
0.90
1.00
1.23
0.90
0.67
1.09
1.55
0.94
0.98
0.83
1.65
0.94
0.96
1.35
1.18
0.86
0.87
0.85
1.00
1.00
0.83
1.19
1.03
0.98
0.78
0.85
0.71
1.24
1.22
1.23
1.31

Table 73- Descriptive statistics for significant Schild male individual vertebral dimensions by burial area

208

Vertebral
Dimension
L4 TNCD

Burial Area

N

Mean

SD

1
2
3
4

15
13
14
10

23.55
24.41
23.11
23.14

1.67
2.26
2.25
1.51

Table 73 (cont'd)

Vertebral
Dimension
L1 AVBH

L2 AVBH

L4 AVBH

Burial area

N

Mean

SD

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

11
16
16
12
13
14
15
10
12
17
16
11

25.07
25.16
25.77
24.20
25.32
26.15
26.81
25.24
27.28
26.79
280.2
26.55

1.28
0.77
1.24
1.22
1.33
1.18
1.61
1.07
1.57
1.31
1.65
1.56

Table 74- Descriptive statistics for significant Schild female individual vertebral dimensions by burial
area

Vertebral
Dimension
AVBH

PVBH

TNCD

APNCD

Burial area

N

Mean

SD

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

12
90
6
10
14
12
7
9
8
10
6
11
12
7
4
8

27.36
27.07
27.84
26.80
28.25
27.90
28.98
27.67
23.82
24.01
23.37
23.42
15.32
15.47
15.62
14.97

1.07
1.00
1.29
1.44
0.98
1.14
1.50
1.00
1.29
2.10
1.09
1.19
0.79
0.70
0.45
1.14

Table 75- Descriptive statistics for Schild male mean lumbar vertebral dimensions by burial area
209

Vertebral
Dimension
AVBH

PVBH

TNCD

APNCD

Burial area

N

Mean

SD

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

8
12
11
9
9
15
12
11
9
15
14
11
6
10
13
9

26.72
26.55
27.17
25.75
26.97
26.82
27.38
26.43
23.28
23.39
23.04
23.90
15.85
15.38
15.31
15.89

0.75
0.90
1.21
1.06
0.93
1.00
1.14
1.00
1.27
1.47
1.20
1.89
1.10
1.05
1.04
1.16

Table 76- Descriptive statistics for Schild female mean lumbar vertebral dimensions by burial area

210

Factor
Charnel Association

Category
N
no
15
yes
23
Artifact/position cluster
1
14
2
24
Table 77- Sample sizes for multiway ANOVA- mean female lumbar APNCD by
charnel association and artifact/position clusters

Source

Type III Sum
of Squares

Mean
Square

df

F

Sig.

6.738a

3

8300.678

1

4.167

1

4.167

3.994

.054

Artifact/position cluster

.153

1

.153

.147

.704

Charnel Association *
Artifact/position cluster

4.193

1

4.193

4.019

.053

Error

35.471

34

1.043

Total

9232.570

38

42.209

37

Corrected Model
Intercept
Charnel Association

Corrected Total

2.246

2.153

.112

8300.678 7956.391

.000

Table 78- Multiway ANOVA table for mean female lumbar APNCD by charnel
association and artifact/position clusters (a. R Squared = .160 (Adjusted R Squared =
.085))

211

Figure 28- Multiway ANOVA plot for mean female lumbar APNCD by charnel association and
artifact/position clusters. For interpretation of the references to color in this and all other figures,
the reader is referred to the electronic version of this dissertation.

212

Factor
Mound

Category
N
1
22
2
16
Artifact/position cluster
1
14
2
24
Table 79- Sample sizes for multiway ANOVA- mean female lumbar APNCD by mound and
artifact/position clusters

Source

Type III Sum
of Squares

Mean
Square

df

F

Sig.

5.682a

3

8018.373

1

Mound

.462

1

.462

.430

.517

Artifact/position cluster

.169

1

.169

.157

.694

Mound *
Artifact/position cluster

5.671

1

5.671

5.279

.028

Error

36.527

34

1.074

Total

9232.570

38

42.209

37

Corrected Model
Intercept

Corrected Total

1.894

1.763

.173

8018.373 7463.643

.000

Table 80- Multiway ANOVA table for mean female lumbar APNCD by mound and
artifact/position clusters (a. R Squared = .135 (Adjusted R Squared = .058))

213

Figure 29- Multiway ANOVA plot for mean female lumbar APNCD by mound and
artifact/position clusters

214

Factor
Charnel Association

Category
N
0
23
1
23
Artifact only cluster
0
25
6
21
Table 81- Sample sizes for multiway ANOVA- Schild male femur length by charnel association
and artifact only clusters

Source

Type III Sum
of Squares

Mean
Square

df

F

Sig.

30.873a

3

88162.000

1

Charnel Association

.304

1

.304

.078

.782

Artifact only cluster

11.231

1

11.231

2.876

.097

Charnel Association *
Artifact only cluster

17.855

1

17.855

4.573

.038

Error

164.002

42

3.905

Total

94382.750

46

194.875

45

Corrected Model
Intercept

Corrected Total

10.291

2.635

.062

88162.000 22577.808

.000

Table 82- Multiway ANOVA table for male femur length by charnel association and
artifact only clusters (a. R Squared = .158 (Adjusted R Squared = .098))

215

Figure 30- Multiway ANOVA plot for male femur length by charnel association and artifact only
clusters

216

Factor
Charnel Association

Category
N
no
23
yes
23
Artifact only cluster
1
25
2
21
Table 83- Sample sizes for multiway ANOVA- mean female lumbar AVBH by charnel
association and artifact only clusters

Source

Type III Sum
of Squares

Mean
Square

df

F

Sig.

3.449

.027

24136.633 23983.58

.000

10.413a

3

24136.633

1

Charnel Association

1.775

1

1.775

1.763

.193

Artifact only cluster

.984

1

.984

.978

.329

4.217

1

4.217

4.191

.048

Error

36.230

36

1.006

Total

28301.183

40

46.643

39

Corrected Model
Intercept

Charnel Association *
Artifact only cluster

Corrected Total

3.471

Table 84- Multiway ANOVA table for mean female lumbar AVBH by charnel
association and artifact only clusters (a. R Squared = .223 (Adjusted R Squared =
.159))

217

Figure 31- Multiway ANOVA plot for mean female lumbar AVBH by charnel association and
artifact only clusters

218

Factor
Mound

Category
N
1
16
2
22
Artifact only cluster
0
22
6
16
Table 85- Sample sizes for multiway ANOVA- mean male lumbar AVBH by mound and artifact
only clusters

Source

Type III Sum
of Squares

Mean
Square

df

F

Sig.

9.398a

3

26796.985

1

Mound

.341

1

.341

.265

.610

Artifact only cluster

.819

1

.819

.635

.431

9.166

1

9.166

7.104

.012

Error

43.867

34

1.290

Total

28209.633

38

53.265

37

Corrected Model
Intercept

Mound * Artifact Only
Cluster

Corrected Total

3.133

2.428

.082

26796.985 20769.37
6

.000

Table 88- Multiway ANOVA table for mean male lumbar AVBH by mound and
artifact only clusters (a. R Squared = .176 (Adjusted R Squared = .104))

219

Figure 32- Multiway ANOVA plot for mean male lumbar AVBH by mound and artifact only
clusters

220

WORKS CITED

221

WORKS CITED

Alt, Susan M.
2002 Identities, Traditions and Diversity in Cahokia 's Uplands . Midcontinental
Journal of Archaeology 27:217-236.
Alt, Susan and Timothy Pauketat
2007 Sex and the Southern Cult. In Southeastern Ceremonial Complex: Chronology, Context,
and Content, edited by Adam King, pp. 332-350. University of Alabama Press,
Tuscaloosa.
Ambrose, Stanley H and Jane E. Buikstra
2003 Status and Gender Differences in Diet at Mound 72, Cahokia, Revealed by Isotopic
Analysis of Bone. Journal of Anthropological Archaeology 22: 217-226.
Anderson, David G.
1997 The Role of Cahokia in the Evolution of Southeastern Mississippian Society. In Cahokia:
Domination and Ideology in the Mississippian World, edited by Timothy R. Pauketat and
Thomas E. Emerson, pp. 248-286. University of Nebraska Press, Lincoln.
Arnold, Betinna
2006 Gender and Archaeological Mortuary Analysis. In Handbook of Gender in Archaeology,
edited by Sarah M. Nelson, pp. 137-170. Alta Mira Press, Walnut Creek.
Armelagos, George and Cassandra Hill
1990 An Analysis of the Biocultural Consequences of the Mississippian Transformation. In
Towns and Temples along the Mississippi, edited by Cheryl Anne Cox and David H. Dye,
pp. 6-37. University of Alabama Press, Tuscaloosa.
Atwell, Karen and Kenneth Farnsworth
1999 Crimson Circle Mounds: Late Woodland Ritual and Settlement Patterns at the
Confluence of the Illinois and Mississippi Rivers. Illinois Archaeology 11(1/2): 80-148.
Baker, Brenda, Tosha L. Dupras, and Matthew W. Tocheri
2005 The Osteology of Infants and Children. Texas A&M University, College Station.

Beard A. S. and M.J. Blaser
2002 The Ecology of Height: The Effect of Microbial Transmission on Human Height.
Perspectives in Biology and Medicine 45: 475-99.
Benn, David R.
1995 Woodland People and the Roots of the Oneota. In Oneota Archaeology: Past, Present,
and Future, edited by William Green, pp. 91-139. Office of the State Archaeologist of
Iowa, Iowa City.
Bice, Gillian

223

2003

Reconstructing Behavior from Archaeological Skeletal Remains: A Critical Analysis of
the Biomechanical Model. Unpublished Ph.D. Dissertation, Michigan State University.

Binford , Lewis R.
1971
Mortuary Practices: Their Study and Their Potential. In Approaches to the Social

Dimensions of Mortuary Practices, edited by J. A. Brown. Memoirs of the
Society for American Archaeology 25:58-67.
Blakey, Michael
1998 Beyond European Enlightenment: Toward a Critical and Humanistic Human Biology. In
Building a New Biocultural Synthesis: Political Economic Perspectives on Human
Biology, edited by Alan H. Goodman and Thomas L. Leatherman, pp. 379-406. The
University of Michigan Press, Ann Arbor.
Boas, Franz
1912

Bogin, Barry
1988

Changes in the Bodily Form of the Descendants of Immigrants. American Anthropologist
14(3): 530-562.

Patterns of Human Growth. Cambridge University Press, Cambridge.

Bogin, Barry, Maria Silva, and Luis Rios
2007 Life History Trade-Offs in Human Growth: Adaptation or Pathology? American Journal
of Human Biology 19(5): 631-642.
Braun, Mark, Della Collins Cook, and Susan Pfeiffer
1998 DNA from Mycobacterium Tuberculosis Complex Identified in North American, PreColumbian Human Skeletal Remains. Journal of Archaeological Science 25:271-277.
Braun, David, and Steven Plog
1982Evolution of Tribal Social Networks: Theory and Prehistoric North American Evidence. American
Antiquity 47: 504-525.
Bridges, Patricia S.
1991 Skeletal Evidence of Changes in Subsistence Activities between the Archaic and
Mississippian Time Periods in Northwestern Alabama.” In What Mean These Bones?
Studies in the Southeastern Bioarchaeology, edited by Mary L. Powell, Patricia S.
Bridges, and Anne Marie W. Mires, pp. 89-101. University of Alabama Press,
Tuscaloosa.
2000

Changes in Long Bone Diaphyseal Strength with Horticultural Intensification in West
Central Illinois. American Journal of Physical Anthropology 112:217-238.

Brown, James
1989 On Style Divisions of the Southeastern Ceremonial Complex: A Revisionist
Perspective. In The Southeastern Ceremonial Complex: Artifacts and Analysis,
edited by Patricia Galloway, pp. 183-204. University of Nebraska Press, Lincoln.
1995 On Mortuary Analysis – with Special Reference to the Saxe-Binford Research Program.

224

In Regional Approaches to Mortuary Analysis, pp. 3-26, edited by Lane Anderson Beck.
Plenum Press, New York.

2004

The Cahokian Expression: Creating Court and Cult. In Hero, Hawk, and Open
Hand: American Indian Art in the Ancient Midwest and South, edited by Richard
F. Townsend, pp. 105-124. Yale University Press, New Haven.

Brown, James A. and John Kelly
2000 Cahokia and the Southeastern Ceremonial Complex. In Mounds, Modoc, and
Mesoamerica: Papers in Honor of Melvin K. Fowler, edited by S.R. Ahler, pp.
469-510. Scientific Papers 28. Illinois State Museum, Springfield.
Buikstra, Jane E.
1992 Diet and Disease in Late Prehistory. In Disease and Demography in the Americas, edited
by John Verano and Douglas Ubelaker, pp. 87-101. Smithsonian Institution, Washington
DC.
Buikstra, Jane E and Della Collins Cook
1978 Pre-Columbian Tuberculosis: An epidemiological Approach. Medical College of Virginia
Quarterly 14:32-44.
1981

Pre-Columbian Tuberculosis in West-Central Illinois: Prehistoric Disease in Biocultural
Perspective. In Prehistoric Tuberculosis in the Americas, edited by Jane Buikstra, pp.
115-139. Northwestern University Archaeological Program, Scientific Papers 5,
Evanston Illinois.

Buikstra, Jane and Douglas Ubelaker
1994 Standards for Data Collection from Human Skeletal Remains. Arkansas Archaeological
Survey Research Series No. 44, Fayetville.

Charles, Douglas K.
1992 Woodland Demographic and Social Dynamics in the American Midwest: Analysis of a
Burial Mound Survey. World Archaeology 24 (2): 175-197.
2005

The Archaeology of Death as Anthropology. In Interacting with the Dead:
Perspectives on Mortuary Archaeology for the New Millennium, edited by Gordon F.M.
Rakita, Jane E. Buikstra, Lane A. Beck, and Sloan R. Williams , pp. 15-24. University
Press of Florida, Gainesville.

Clark, George A.
1985 Heterochrony, Allometry, and Canalization in the Human Vertebral Column: Examples
from Prehistoric Amerindian Populations. Unpublished PhD Dissertation, University of
Massachusetts Amherst.
1988

New Method for Assessing Changes in Growth and Sexual Dimorphism in
Paleoepidemiology. American Journal of Physical Anthropology 77: 105-116.

225

Clark, George A., Nicholas R. Hall, George J. Armelagos, Gary A. Borkan, Manohar M. Panjabi, and F.
Todd Wetzel
1986 Poor Growth Prior to Early Childhood: Decreased Health and Life-Span in the Adult.
American Journal of Physical Anthropology 70:145-160.
Cobb, Charles R. and Michael S. Nassany
1995 Interaction and Intensification in the Late Woodland Southeast. In Native American
Interactions: Multiscalar Analysis and Interpretation in the Eastern Woodlands. Pp. 205226. Knoxville: University of Tennessee Press.
Cohen, Matthew Nathan and Sharon Bennett
1993 Skeletal Evidence for Gender Hierarchies in Prehistory. In Sex and Gender Hierarchies,
edited by Barbara Miller, pp. 273-296. Cambridge University Press.
Conner, Michael
1985 Site Structure and Function. In The Hill Creek Homestead and the Late Mississippian
Settlement of the Lower Illinois Valley, edited by Michael D. Connor, pp. 193-220.
Center for American Archaeology, Kampsville.
1990

Population Structure and Skeletal Variation in the Late Woodland of West-Central
Illinois. American Journal of Physical Anthropology 82: 31-43.

Cook, Della Collins
1980 Schild Pathologies. In Mississippian Mortuary Practices: A Case Study of Two Cemeteries
in the Lower Illinois Valley by Lynne G. Goldstein. Northwestern University
Archaeological Program.
1981

Mortality, Age-Structure, and Status in the Interpretation of Stress indicators in
Prehistoric Skeletons. An Example from the Lower Illinois Valley. In The Archaeology of
Death, edited by R. Chapman and K. Randsborg, pp. 133-144. Cambridge University
Press, Cambridge.

1984

Subsistence and Health in the Lower Illinois Valley: Osteological Evidence. In
Paleopathology at the Origins of Agriculture, edited by Mark N. Cohen and George J.
Armelagos, pp. 237-270. Academic Press, New York.

1994

Dental Evidence for Congenital Syphillis (and its Absence) Before and After the
Conquest of the New World. In L’Origine de la syphilis en Europe: Anant ou Apres
1493, edited by Oliver Dutour, Gyorgi Palfi, Jaques Berato, and Jean-Pierre Brun, pp.
169-175. Errance, Paris.

2005

Syphillis? Not Quite. Paleoepidemiology in the Evolutionary Context of the Midwest. In
The Myth of Syphillis: A Natural History of North American Trepanematosis.

2007

Maize and Mississippians in the American Midwest: Twenty Years Later. In Ancient
Health: Skeletal Indicators of Agricultural and Economic Intensification, edited by Mark
Nathan Cohen and Gillian M.M. Crane-Kramer, pp. 10-19. University Press of Florida,
Gainesville.

226

Cook, Della Collins and Ann Palcovich
2006 Joe Gay Mound Group, Crematory A Excavations. In Illinois Hopewell and Late
Woodland Mounds: The excavations of Gregory Perino 1950-1975, edited by
Kenneth Farnsworth and Michael Wiant, pp. 498-500. Studies in Archaeology
Number 4. University of Illinois, Urbana-Champaign.
Cucina, Andrea and Vera Tiesler
2001 Dental Caries and Antemortem Tooth Loss in the Northern Peten Area, Mexico: A
Biocultural Perspective on Social Status Differences Among Classic Maya. American
Journal of Physical Anthropology 122:1-10.

Czerwinski, Stefan and Bradford Towne
2004 Methods for the Study of the Genetics of Growth and Development. In Methods in
Human Growth Research, edited by Roland C. Hauspie, Noel Cameron, and Luciano
Molinari, pp. 3-26. Cambridge University Press, Cambridge.
Delaney-Rivera, Colleen
2004 From Edge to Frontier: Early Mississippian Occupation of the Lower Illinois River
Valley. Southeastern Archaeology 23(1): 41-56.
2007

Examining Interaction and Identity in Prehistory: Mortuary Vessels from the Schild
Cemetery. North American Archaeologist 28(4) 295-331.

Droessler, Judith
1981 Craniometry and Biological Distance: Biological Continuity and Change at the Late
Woodland- Mississippian Interface. Center for North American Archaeology.
Emerson, Thomas E.
1997 Reflections from the Countryside on Cahokian Hegemony. In Cahokia: Domination and
Ideology in the Mississippian World, edited by Timothy R. Pauketat and Thomas E.
Emerson, pp.167-189. University of Nebraska Press, Lincoln.
2000

Langford Tradition and the Process of Tribalization on the Middle
Mississippian Borders. Midcontinental Journal of Archaeology 24(1): 3-56.

2002

Introduction to Cahokia 2002: Diversity, Complexity, and History. Midcontinental
Journal of Archaeology 27(2):128-148.

Emerson, T.E., R. Hughes, M. Hynes, S. Wisseman
2003 The Sourcing and Interpretation of Cahokia-Style Figures in the Trans-Mississippi South
and Southeast. American Antiquity 68(2): 287-314.
Farnsworth, Kenneth B., Thomas E. Emerson, and Rebecca Miller Glenn
1991 Patterns of Late Woodland/Mississippian Interaction in the Lower Illinois Valley
Drainage: A View from Star Village. In Cahokia and the Hinterlands: Middle
Mississippian Cultures of the Midwest, edited by T.E. Emerson and R. B. Lewis, pp. 83118. University of Illinois Press, Urbana.
Fritz, Gayle J.

227

1999

Gender and the Early Cultivation of Gourds in Eastern North America. American
Antiquity 64(3): 417-429.

Fowler, Melvin L.
1978 Cahokia and the American Bottom: Settlement Archaeology. In Mississippian Settlement
Patterns, edited by Bruce Smith, pp. 455-478. Academic Press, New York.
1989

The Cahokia Atlas: A Historical Atlas of Cahokia Archaeology. Studies in
Illinois Archaeology, edited by Melvin L. Fowler. Illinois Historic Preservation
Agency, Springfield

1991

Mound 72 and Early Mississippian at Cahokia. In New Perspectives on Cahokia: Views
from the Periphery, edited by James B. Stoltman, pp. 1-28. Monographs in World
Archaeology, Volume 2. Prehistory Press, Madison.

Gellar, Pamela
2005 Skeletal Analysis and Theoretical Complications. World Archaeology 37(4): 597–609.
2008

Conceiving Sex: Fomenting a Feminist Bioarchaeology. Journal of Social Archaeology
8(113): 113-136.

George D. and P. Mallery
2003 SPSS for Windows Step by Step: A Simple Guide and Reference. 11.0 update
(4th edition). Allyn & Bacon, Boston.
Gepstein, R, Y. Folman, P. Sagiv, Y. Ben Davis, and T. Hallel.
1991 Does the Anteroposterior Diameters of the Bony Spinal Canal Reflect its Size?
An Antaomical Study. Surgical and Radiologic Anatomy 13 : 289-291.
Goldstein, Lynne G.
1980 Mississippian Mortuary Practices: A Case Study of Two Cemeteries in the Lower Illinois
Valley. Northwestern University Archaeological Program.
1981

One Dimensional Archaeology and Multidimensional People: Spatial Organization and
Mortuary Analysis. In The Archaeology of Death, edited by R. Chapman, I. Kinnes, and
K. Randsborg, pp. 53-69. Cambridge University Press, Cambridge

1991

The Implication’s of Aztalan’s Location. In New Perspectives on Cahokia: Views
from the Periphery, edited by James B. Stoltman, pp. 209-228. Prehistory Press,
Madison.

2006

Mortuary Analysis and Bioarchaeology. In A History of American Bioarchaeology (Beck
and Buikstra, eds.). Elsevier Publishing.

Goodman, Alan H.
1998 The Biological Consequences of Inequality in Antiquity. In Building a New Biocultural
Synthesis: Political- Economic Perspectives on Human Biology, edited by Alan H.
Goodman and Thomas L. Leatherman, pp. 147-170. University of Michigan Press, Ann
Arbor.

228

Goodman, Alan H. and Jerome C. Rose
1980 Enamel Hypoplasias as Indicators of Stress in Three Prehistoric Populations from Illinois.
Human Biology 52:515-528.
1990

Assessment of Systemic Physiological Perturbations from Dental Enamel Hypoplasias
and Associated Histological Structures. Yearbook of Physical Anthropology 33:59-110.

Goodman, Alan, George Armelagos, and Jerome Rose
1984 The Chronological Distribution of Enamel Hypoplasias from Prehistoric Dickson
Mounds Populations. American Journal of Physical Anthropology 65: 259-256.

Goodman, Alan H., R. Brooke Thomas, Alan C. Swedlund and George Armelagos,
1988 Biocultural Perspectives on Stress in Prehistoric, Historical and Contemporary
Population Research. Yearbook of Physical Anthropology 31:169-202.
Graslund, Anne-Sofie
2001 The Position of Iron Age Scandinavian Women: Evidence from Graves and Runestones.
In Gender and the Archaeology of Death (Arnold, B. and Wicker, N.L., eds.). Alta Mira
Press.
Harn, Alan
1991

The Eveland Site: Inroad to Mississippian Society. In New Perspectives on Cahokia:
Views from the Periphery, edited by James B. Stoltman, pp. 129-154. Prehistory Press,
Madison.

Hamilton, Margaret E.
1982 Sexual Dimorphism in Skeletal Samples. In Sexual Dimorphism in Homo
Sapiens: A Question of Size, edited by Roberta L. Hall, pp. 107-164. Praeger,
New York.
Hedman, Kristin, Eve Hargrave, and Stanley Ambrose
2002 Late Mississippian Diet in the American Bottom: Stable Isotope Analyses of Bone
Collagen and Apatite. Midcontinental Journal of Archaeology 27(2):
Hertz, Robert
1960

Hill, Erica
1998

A Contribution to the Study of the Collective Representation of Death." Death and the
Right Hand, translated by Rodney and Claudia Needham. Free Press, Glencoe, IL.

Gender-Informed Archaeology: The Priority of Definition, the Use of Analogy, and the
Multivariate Approach. Journal of Archaeological Method and Theory 5 (1): 99-128.

Hillson, Simon
1996 Introduction to Dental Anthropology. In Dental Anthropology, pp. 1-5. Cambridge
University Press, Cambridge.
2000

Dental Pathology. In Biological Anthropology of the Human Skeleton, edited by M.Anne
Katzenberg and Shelley R. Saunders, pp.249-286. Wiley-Liss: New York.

229

Himes, John H.
2004 Why Study Child Growth and Maturation? In Methods in Human Growth Research,
edited by Roland C. Hauspie, Noel Cameron, and Luciano Molinari, pp. 3-26. Cambridge
University Press, Cambridge.
Hodder, Ian
1982 The identification and interpretation of ranking in prehistory: A contextual perspective. In
Ranking, Resource and Exchange: Aspects of the Archaeology of Early European
Society, edited by A. C. Renfrew and S. J. Shennen, pp. 150-154. Cambridge University
Press, Cambridge.
Hoppa, Robert D. and Charles M. Fitzgerald
1999 From Head to Toe: Integrating Studies of Bones and Teeth in Biological Anthropology.
In Human Growth in the Past: Studies from Bones and Teeth, edited by Robert D. Hoppa
and Charles M. Fitzgerald, pp. 1-31. Cambridge University Press, Cambridge.
Jurmain, Robert
1999 Stories from the Skeleton: Behavioral Reconstruction in Human Osteology. Canada:
Gordon and Breach.
Kelly, John E.
1991 The Evidence for Prehistoric Exchange and Its implications for the Development of
Cahokia. In New Perspectives on Cahokia: Views form the Periphery (J. Stoltman, ed.).
Madison: Prehistory Press.
Knudson, Kelly J. and Christopher Stojanowski
2008 New Directions in Bioarchaeology: Recent Contributions to the Study of Human Social
Identities. Journal of Archaeological Research 16:397-432.
Koningsberg, Lyle
1988 Migration Models of Prehistoric Postmarital Residence. American Journal of Physical
Anthropology. 77:471-482
Knight, Vernon James and Vincas Steponaitis
1998 A New History of Moundville. In Archaeology of the Moundville Chiefdom:
Chronology, Content, and Contest, edited by Vernon James Knight and Vincas
Steponaitis, pp. 1-25. University of Alabama Press, Tuscaloosa.
Larsen, Clark Spencer
1997 Bioarchaeology: Interpreting Behavior from the Human Skeleton. Cambridge University
Press, Cambridge.
2002

Bioarchaeology: The Lives and Lifestyles of Past People. Journal of Archaeological
Research 10(2): 119-166.

Leatherman, Thomas L.
1998 Illness, Social Relations, and Household Production in the Andes of Southern Peru. In
Building a New Biocultural Synthesis: Political- Economic Perspectives on Human
Biology, edited by Alan H. Goodman and Thomas L. Leatherman, pp. 245-268.
University of Michigan Press, Ann Arbor.

230

Lukacs, J.R.
1992

Dental Paleopathology and Agricultural Intensification in South Asia: New Evidence
from Bronze Age Harappa. American Journal of Physical Anthropology 87:133-150.

Marshall, Charla
2011 An Ancient DNA Perspective on Angel Mounds (12-Vg-1): A Mississippian
Archaeological Site. Unpublished PhD Dissertation. Indiana University.
Martin Debra L.
2000 Bodies and Lives: Biological Indicators of Health Differentials and Division of Labor by
Sex. In Women and Men in the Prehistoric Southwest (Crown, P.L., ed.). School of
American Research Press.
Macia–Lees, Frances and Nancy Johnson Black
2000 Gender and Anthropology. Waveland Press.
McElrath, Dale L., Thomas E Emerson, and Andrew C.Fortier
2000 Social Evolution or Social Response? A Fresh Look at the “Good Grey Cultures” after
Four Decades of Midwest Research. In Late Woodland Societies: Tradition and
Transformation across the Midcontinent, edited by Thomas E. Emerson, Dale L.
McElrath, and Andrew C. Fortier, pp. 3-21. University of Nebraska Press: Lincoln.
Mehrer, Mark W.
2000 Heterarchy and Hierarchy: The Community Plan as Institution in Cahokia’s Polity. In
The Archaeology of Communities: A New World Perspective edited by Marcello Canouto
and Jason Yaeger, pp. 44-57. Routledge, London.
Milledge Nelson, Sarah
2004 Gender in Archaeology: Analyzing Power and Prestige. Altamira Press.
Milledge Nelson, Sara and Myriam Rosen-Ayalon
2002 In Pursuit of Gender: Worldwide Archaeological Approaches. Altamira Press.

Milner, George R.
1982 Measuring Prehistoric Levels of Health: A Study of the Mississippian Period Skeletal
Remains from the American Bottom, Illinois. Unpublished PhD Dissertation,
Northwestern University.
1983

The East St. Louis Stone Quarry Site Cemetery. American Bottom Archaeology FAI-270
Site Reports. University of Illinois Press, Urbana.

1984

Social and Temporal Implications of Variation among American Bottom and
Mississippian Cemeteries. American Antiquity 49(3): 468-488.

1991

American Bottom Mississippian Cultures: Internal Developments and External Relations.
In New Perspectives on Cahokia: Views from the Periphery, edited by James Stoltman,
pp. 29-49. Prehistory Press, Madison

1998

The Cahokia Chiefdom. Smithsonian Institution Press.

231

Morris, Ian
1991 The Archaeology of Ancestors: The Saxe/Goldstein Hypothesis Revisited. Cambridge
Archaeological Journal 1(2): 147-169.

Muller, Jon
1989

2009

The Southern Cult. In The Southeastern Ceremonial Complex: Artifacts and
Analysis, edited by Patricia Galloway, pp. 11-26. University of Nebraska Press,
Lincoln.
Archaeology of the Lower Ohio River Valley. Left Coast Press, Walnut Creek.

Nguyen, Ainh-Kim and Karine Peschard
2003 Anthropology, Inequality, and Disease: A Review. Annual Review of Anthropology 32:
447-474.
O’Gorman, Jodie
2001 Life, Death, and the Longhouse: A Gendered View of Oneota Social Organization. In
Gender and the Archaeology of Death, edited by B. Arnold and N.L Wicker. Alta Mira
Press, Walnut Creek.
O’Shea, John
1981

Social Configurations and the Archaeological Study of Mortuary Practices: A Case
Study. In The Archaeology of Death, edited by R. Chapman and K. Randsborg, pp. 3952. Cambridge University Press, Cambridge.

Oetelaar, Gerald
1993 Identifying Site Structure in the Archaeological Record: An Illinois Example. American
Antiquity 58(4): 662-687.
Parker Pearson, Mike
1999 The Archaeology of Death and Burial. Texas A&M University Press, College Station.
Pauketat, Timothy
1998 Refiguring the Archaeology of Greater Cahokia. Journal of Archaeological Research
6:45-89.
1997

Refiguring the Archaeology of Greater Cahokia. Journal of Archaeological Research
6:45-89.

2004

Ancient Cahokia and the Mississippians. Cambridge University Press, Cambridge.

2002

A Fourth Generation Synthesis of Cahokia and Mississippianization. Midcontinental
Journal of Archaeology 27(2):

Pauketat, Timothy and Thomas E. Emerson
1991 The Ideology of Authority and the Power of the Pot. American Anthropologist 93: 919941.
Perino, Gregory

232

1971a The Mississippian Component of the Schild Site, Greene County, Illinois. In
Mississippian Site Archaeology in Illinois I: Site Reports from the St. Louis and Chicago
Areas, edited by James A. Brown, pp. 1-141. Illinois Archaeological Survey, Bulletin 8.
1971b The Yokem Site, Pike County, Illinois. In Mississippian Site Archaeology in Illinois I:
Site Reports from the St. Louis and Chicago Areas, edited by James A. Brown, pp. 149186. Illinois Archaeological Survey, Bulletin 8.
1973

The Late Woodland Component of the Schild Site, Greene County, Illinois. In Late
Woodland Site Archaeology in Illinois I: Investigations in South Central Illinois, edited
by James A. Brown, pp.90-140. Illinois Archaeological Survey, Bulletin 9.

2006

1967-1968 Excavation of the Yokem Site Late Woodland Mounds, Pike County, Illinois.
In Illinois Hopewell and Late Woodland Mounds: The excavations of Gregory Perino
1950-1975, edited by Kenneth Farnsworth and Michael Wiant, pp. 347-390. Studies in
Archaeology Number 4. University of Illinois, Urbana-Champaign.

Peregrine, Peter
1995 Networks of Power: The Mississippian World-System. In Native American Interactions:
Multiscalar Analyses and Interpretations in the Eastern Woodlands. University of
Knoxville Press, Knoxville.
Phenice, Terrell
1969 A Newly Developed Visual Method of Sexing the Os Pubis. American Journal of
Physical Anthropology 30(2): 297-301
Raff, Jennifer
2008

An Ancient DNA Perspective on the Prehistory of the Lower Illinois River Valley.
Unpublished PhD Dissertation. Indiana University.

Rose, Fionnuala
2007 Intra-community Variation in Diet During the Adoption of a new Staple Crop in the
Eastern Woodlands. American Antiquity 73(3): 413-439.
Rothschild, Nan
1990 Prehistoric Dimensions of Status: Gender and Age in Eastern North America. Garland
Publishing.

Sarwark, John and Carl Eric Aubin
2007 Growth Considerations for the Immature Spine. The Journal of Bone and Joint
Surgery 89(1) 8-13.
Saxe, Arthur A.
1970 Social Dimensions of Mortuary Practices. Unpublished Ph.D. dissertation,
University of Michigan.
Scarry, John F.
1996 The Nature of Mississippian Societies. In Political Structure and Change in the
Prehistoric Southeast United States, edited by John F. Scarry, pp. 12-24. University of
Florida Press, Gainsville.

233

Schober, Teresa M
1998 Reinvestigation of Maize Introduction in West-Central Illinois: A Stable Isotope Analysis
of Bone Collagen and Apatite Carbonate from Late Archaic to Mississippian Times.
Unpublished paper-in-lieu of thesis. University of Illinois, Urbana.
Schroedl, Gerald F., C. Clifford Boyd, and R.P Stephen Davis Jr.
1990 Explaining Mississippian Origins in East Tennessee. In The Mississippian Emergence,
edited by Bruce Smith, pp.175-196. Smithsonian Institution Press, Washington.
Schroeder, Sissel
2004 Current Research on Late Precontact Societies of the Midcontinental United States.
Journal of Archaeological Research 12(4): 311-372.
Schoeninger, MJ and KM Moore
1992 Bone stable isotopes in archaeology. Journal of World Prehistory 6(2):247-296.
Schurr, Mark R.
1992 Isotopic and Mortuary Variability in a Middle Mississippian Population. American
Antiquity 57(2):300-320
Shanks, Michael and Christopher Tilley
1982 Ideology, Symbolic Power, and Ritual Communication: A Reinterpretation of Neolithic
Mortuary Practices. In Symbolic and Structural Archaeology, edited by Ian Hodder, pp.
129-154. Cambridge University Press.
Silventoinen
2003

Determinants of Variation in Adult Body Height. Journal of Biosocial Science 35: 263285.

Smith, Gavin A and R. Brooke Thomas
1998 What Could Be: Biological Anthropology for the Next Generation. In Building a New
Biocultural Synthesis: Political-Economic Perspectives in Human Biology, edited by
A.H. Goodman and T.L. Leatherman, pp. 379-405. The University of Michigan Press,
Ann Arbor.
Sofaer, Joanna R.
2006 The Body as Material Culture. Cambridge University Press.

Song, Kit and David Little
2000 Peak Height Velocity as a Maturity Indicator for Males with Idiopathic Scoliosis.
Journal of Pediatric Orthopaedics 20(3): 286-288
Steadman, Dawnie Wolf
2001 Mississippian in Motion? A Population Genetic Analysis of Interregional Gene Flow in
West-Central Illinois. American Journal of Physical Anthropology 114(1):61-73

Steckel , Richard H.
1987 Growth Depression and Recovery: The Remarkable Case of American Slaves.
Annals of Human Biology 14(2): 111-132.
234

1995

Stature and the Standard of Living. Journal of Economic Literature 33: 19031940.

Steponaitis, Vincas
1998 Population Trends at Moundville. In Archaeology of the Moundville Chiefdom:
Chronology, Content, and Contest, edited by Vernon James Knight and Vincas
Steponaitis, pp. 26-43. University of Alabama Press, Tuscaloosa.

Stig-Sorenson, Marie Louise
2000 Gender Archaeology. Polity Press.
Stoltman, James B.
1991 Cahokia as seen from the Peripheries. In New Perspectives on Cahokia: Views from the
Periphery, edited by James B. Stoltman, pp. 349-354. Prehistory Press, Madison.
Sullivan, Lynne
1995 Mississippian Household and Community Organization in Eastern Tennessee. In
Mississippian Communities and Households, edited by J. Daniel Rogers and Bruce D.
Smith, pp. 99-123. The University of Alabama Press, Tuscaloosa.
2001

Those Men in the Mounds: Gender, Politics, and Mortuary Practices in Late Prehistoric
Eastern Tennessee. In Archaeological Studies of Gender in the Southeastern United
States, edited by Jane Eastman and Christopher Rodning, pp. 101-126. University Press
of Florida, Knoxville.

2006

Gendered Contexts of Mississippian Leadership in Southern Appalachia. In Leadership
and Polity in Mississippian Society, edited by Paul Welch & Brian Butler, pp. 264-285.
Southern Illinois University Press, Carbondale.

Tatarek, Nancy E.
1999 Changing Views of Health: Spinal Stenosis and its Impact on Well-being. Unpublished
PhD Dissertation. The Ohio State University.
2006

Geographical Height Variation among Ohio Caucasian Male Convicts Born 1780-1849.
Economics and Human Biology 4(2):222-236.

Ursu, Traian and Richard Porter
1996 Development of the Lumbar and Sacral Vertebral Canal in Utero. Spine 21(23):
2705-2708.
Van der Merwe, N.J. and J.C. Vogel
1978 13-C Content of Human Collagen as a Measure of Prehistoric Diet in Woodland North
America. Nature 276: 815-816.
Walker, Phillip L and Della Collins Cook
1998 Brief Communication: Gender and Sex: Vive la Difference. American Journal of Physical
Anthropology 106: 255-259.

235

Watson, Patty Jo and Mary C. Kennedy
1991 The Development of Horticulture in the Eastern Woodlands of North America: Women’s
Role. In Engendering Archaeology, edited by Joan M. Gero and Margaret W. Conkey,
pp. 255-275. Basil Blackwell, Oxford.
Welch, Paul
1990

Mississippian Emergence in West Central Alabama. In The Mississippian Emergence,
edited by Bruce Smith, pp.197-226. Smithsonian Institution Press, Washington.

White, Christine
2005 Gendered Food Behaviour among the Maya: Time, Place, Status, and Ritual. Journal of
Social Archaeology 5(3):356-382.
White, C.D., D.M. Pendergast, F.J. Longstaffe, and K.R. Law.
2001 Social Complexity and Food Systems at Altun Ha, Belize: The Isotopic Evidence. Latin
American Antiquity, 12 (4), 371-393.

World Health Organization
1995 Physical Status: the Use and Interpretation of Anthropometry. Report of WHO
Expert Committee. Technical Reports Series No. 854. WHO, Geneva.

236