INFORMATION TO USERS While the most advanced technology has been used to photograph and reproduce this manuscript, the quality of the reproduction is heavily dependent upon the quality of the material submitted. For example: • Manuscript pages may have indistinct print. In such cases, the best available copy has been filmed. • Manuscripts may not always be complete. In such cases, a note will indicate that it is not possible to obtain missing pages. • Copyrighted material may have been removed from the manuscript. In such cases, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, and charts) are photographed by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each oversize page is also filmed as one exposure and is available, for an additional charge, as a standard 35mm slide or as a 17”x 23” black and white photographic print. Most photographs reproduce acceptably on positive microfilm or microfiche but lack the clarity on xerographic copies made from the microfilm. For an additional charge, 35mm slides of 6”x 9” black and white photographic prints are available for any photographs or illustrations that cannot be reproduced satisfactorily by xerography. 8709987
Church, Flora
AN INQUIRY INTO THE TRANSITION FROM LATE WOODLAND TO LATE PREHISTORIC CULTURES IN THE CENTRAL SCIOTO VALLEY, OHIO, CIRCA A.D. 500 TO A.D. 1250
The Ohio State University Ph.D. 1987
University Microfilms International300 N. Zeeb Road, Ann Arbor, Ml 48106 PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a checkV mark .
1. Glossy photographs or pages_____
2. Colored illustrations, paper or print______
3. Photographs with dark background____
4. Illustrations are poor copy______
5. Pages with black marks, not original copy______
6. Print shows through as there is text on both sides of ______p ag e
7. Indistinct, broken or small print on several pages ^
8. Print exceeds margin requirements_____
9. Tightly bound copy with print lost in ______spine
10. Computer printout pages with indistinct print______
11. Page(s) ______lacking when material received, and not available from school or author.
12. Page(s) ______seem to be missing in numbering only as text follows.
13. Two pages numbered . Text follows.
14. Curling and wrinkled pages______
15. Dissertation contains pages with print at a slant, filmed as received______
16. Other______I
University Microfilms International AN INQUIRY INTO THE TRANSITION
FROM LATE WOODLAND TO LATE PREHISTORIC CULTURES
IN THE CENTRAL SCIOTO VALLEY, OHIO
CIRCA A.D. 500 TO A.D. 1250
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Flora Church, A.A., B.A., M.A
The Ohio State University
1987
Dissertation Committee: Approved by
W.S. Dancey
P.W. Sciulli
R.W. Yerkes To
Nancy Emmaline Thompson
And
Wadsworth Church ACKNOWLEDGMENTS
This dissertation was partially funded by a grant from
Sigma Xi and a Graduate Student Alumni Research Award from the Ohio State University. Martha P. Otto and Bradley
Baker of the Ohio Historical Center were immeasurably helpful in placing their collections at my disposal and in providing work space and equipment. A.G.E. Latimer created the site catchment drawings in chapter IX. I want to express my appreciation for the unwavering support-- financial and otherwise-- received from Chung-min Chen and the Department of Anthropology in general.
Professor William S. Dancey, my adviser, has provided me with a challenging model of intellectual curiosity and scientific rigor. He has shown me endless patience and encouragement, and for all of this, I thank him. I also want to thank Professor Paul W. Sciulli for the tir-e and care he gave in guiding me through quantitative analysis and for the refuge his osteology lab provided, Professor
Richard W. Yerkes for his cheerful willingness to help, and
Professor William M. Sumner, whose door was always open to me.
Thanks go as well to my colleagues Owen Hagovsky,
Annette E. Latimer, Margaret MacMinn-Barton, Dee Anne Wymer, and especially to Carol Piacentini and Professor Arthur S.
Keene of UMass/Amherst, for their friendship and support during my graduate career.
Finally, words cannot express the thanks I owe to my parents, Ann and Homer Church, to my brothers and sisters
(Joe and Lori; Marsha, Steve, Jason and Nick; Mitchell;
Douglas; Teressa and Mitchell Lee; Barbara; and Mark), nor to those friends who travelled this road with me in spirit— Robin, Grey, Michele, Priscilla, Madelyn, Tom, and
Larry. I
VITA
9 January 1954 ...... Born - Sandusky, Ohio
1975 ...... A.A., with Honors, Bowling Green State University, Firelands, Huron, Ohio
1977 ...... B.A., Cum laude. The Ohio State University, Columbus, Ohio
1 9 8 1 ...... M.A., The Ohio State University, Columbus, Ohio
PUBLICATIONS
1983 An analysis of textile fragments from three Ohio Hopewell mound groups. Ohio Archaeologist 33(1):10—16.
1984 Textiles as markers of Ohio Hopewell social identities. Midcontinental Journal of Archaeology 9(1):1 — 25.
1986 Current research on Ohio's prehistoric Hopewell textiles. Proceedings of the 1985 CHACM00L Archaeological Conference, Calgary, Canada, in press.
FIELDS OF STUDY
Major Field: North American Archaeology
Studies in Eastern Woodlands Prehistory and
Lithic Analysis .... W.S. Dancey
Style in Archaeology . . W.M. Sumner
Quantitative Methods . . P.W. Sciulli
v TABLE OF CONTENTS
ACKNOWLEDGMENTS ...... iii
VITA ...... v
LIST OF TA B L E S ...... ix
LIST OF FI G U R E S ...... xi
CHAPTER PAGE
I. INTRODUCTION TO THE PROBLEM: RESEARCH GOALS 1
II. THE PROBLEM: LITERATURE REVIEW ...... 8
The Fort Ancient Tradition ...... 8 The Baum P h a s e ...... 15 The Late Woodland in Central Ohio .... 21 Theories on the Origins of Fort Ancient . 27 Summary and Conclusions: Directions for New Research ...... 38
III. ENVIRONMENTAL DESCRIPTION OF THE CENTRAL SCIOTO DRAINAGE ...... AO
Climate ...... AO Climatic Change ...... A2 Hydrology...... A6 Bedrock Geology ...... A7 Physiography ...... A8 Present Topography ...... A9 Soils ...... 51 V e g e t a t i o n ...... 56 Environmental Summary ...... 59
IV. CHARACTERISTICS OF THE ARCHAEOLOGICAL ECORD: LATE WOODLAND AND LATE PREHISTORIC SIi ,S . . 61
The History of Archaeological Research in the Central Scioto Drainage ...... 62 The Nature of the Site Sample...... 66 General Characteristics of the Sample . . 73
vi Major Site Descriptions...... 78 Late Woodland S i t e s ...... 78 Late Prehistoric Sites ...... 84 Summary and Evaluation of the Sample . . 93
V. CHRONOLOGY AND THE QUESTION OF POPULATION. INTRUSION ...... 95
Ordering the Baum Phase Sites: An Illustration of the Problem...... 96 Methodology...... 102 Temporal Homogeneity...... 103 Ceramic Attributes in Chronology .... 108 Triangular Points as Chronological Markers 117 Absolute Chronological Data for the Central Scioto Valley ...... 123 New Chronological Order ...... 127 Evaluating the Prufer and Shane Model of Cultural Change ...... 130
VI. SETTLEMENT PATTERN COMPARISON ...... 136
General Site Distribution ...... 138 Environmental Variables ...... 140 Archaeological Variables ...... 145 Comparing Late Woodland and Late Prehistoric Site Distributions...... 147 Summary of Site Distribution Characteristics ...... 168 Conclusions ...... 170
VII. EDGE-WEAR ANALYSIS OF ARTIFACTS ...... 173
Theoretical Approach to Macro-Wear Analysis 174 Methodology: The Macro-Wear Approach . . 177 Results of Application: A Functional Classification ...... 188 Testing the Classes: Use-Wear or Non-Use Damage? ...... 194 Summary and Conclusions ...... 196
VIII. A COMPARISON OF SITE STRUCTURE...... 200
A Late Woodland Example: Water Plant . . 202 Examples of Late Prehistoric Sites: Baum and Gartner ...... 213 A Comparison of Site S t r u c t u r e ...... 218 Conclusions ...... 222
vii IX. LATE WOODLAND AND LATE PREHISTORIC: SITE CATCHMENTS IN THE CENTRAL SCIOTO VALLEY . . 226
Review of Site Catchment Analysis . . . 226 Methodology ...... 233 Results and Summary of the Site Catchment A n a l y s i s ...... 234
X. SUMMARY AND CONCLUSIONS...... 251
Recent Theories on Nucleation and D i s p e r s i o n ...... 252 Recent Theories on Sedentariness .... 254 Chronology and Site F u n c t i o n ...... 261 Settlement Patterning ...... 266 Seasonal or Sedentary Communities? . . . 270 Conclusions...... 279
BIBLIOGRAPHY ...... 284
APPENDICES
A. Statistical Tests: SE and R. X C Test of Independence...... 301
B. Master List of Functional Tool Classes . . 302
C. Statistical Tests: Kind of W e a r ...... 312
D. Statistical Tests: Location of Wear .... 313
E. Statistical Tests: Orientation of Wear . . 314
F. Statistical Tests: Shape of Worn Area and Edge Angle ...... 315
viii LIST OF TABLES
TABLE PAGE
1. Baum Focus Trait L i s t ...... 16
2. Radiocarbon DAtes for Central Scioto Drainage S i t e s ...... 77
3. Seriation of Sites by Temper Frequency . . 112
4. Ordering Sites by Percentage of Guilloche . 113
5. Ordering Sites by Percentage of Rim Form . . 115
6 . Ordering Sites by Frequency of Lip Shape . . 116
7. Comparison of Triangular Point Measurements: Late Woodland/Late Prehistoric ...... 120
8 . Comparison of Triangular Point Measurements Grouped by Base-Shape ...... 121
9. Seriation of Sites by Triangular Point Bases 124
10. Composite Chronology of Sites ...... 128
11. Stream Order Ranking for Central Scioto Drainage ...... 142
12. Late Woodland/Late Prehistoric Sites by Physiographic Province ...... 148
13. Late Woodland/Late Prehistoric Sites by Stream Order ...... 148
14. Landform Locations of Sites ...... 150
15. Landform Location of Sites by Stream Order . 154
16. Frequency of Sites With Vegetation Association 155
17. Site Distribution by Soil T y p e s ...... 156
ix 18. Sites by Size C l a s s e s ...... 156
19. Site Size Compared with Landform Location . 160
20. Site Size Compared with Soil T y p e ...... 161
21. Percentage of Temporal Components Per Site . 163
22. Assemblage Summaries for Water Plant, Baum and Gartner S i t e s ...... 179
23. Morphological/Technological Attributes of Tools ...... 183
24. Bone Scraper Measurements: Baum and Gartner 197
25. Functional Tool Classes Represented at the Water Plant Site ...... 211
26. Surface Clusters of Tools: Water Plant Site 212
27. Characteristics of Late Woodland and Late Prehistoric Sites ...... 219
28. Late Prehistoric Mound Characteristics . . . 223
29. Total Percentage of Catchment Strata Per Site in Hectares ...... 240
30. Percentages of Environmental Strat Per Site Per 1 Km Catchment Rings ...... 241
31. Vegetation Types by Environmental Strata . . 244
32. Soil Types Associated with Environmental Strata 247
33. Site Location Correlated with Artifact Categories ...... 249
34. Archaeological Correlates of Sedentarism for the Late Woodland/Late Prehistoric Period . 274
x
.U, LIST OF FIGURES
FIGURE PAGE
1. Copy of Ohio Archaeological Inventory Form . 68
2. Major Sites in the Central Scioto Valley . . 72
3. Triangular Point Measurements ...... 118
4. Late Woodland and Late Prehistoric Site Distribution in the Central Scioto Valley . 139
5. Comparison of Late Woodland and Late Prehistoric Sites by Stream Order and Topographic Landform Locations ...... 152
6. Comparison of Late Woodland and Late Prehistoric Sites by Size Class and Topographic Landform Locations ...... 158
7. Late Woodland/Late Prehistoric Sites Compared by Size Class and Soil T y p e ...... 162
8. Late Woodland Site Distribution...... 165
9. Late Prehistoric Site Distribution .... 166
10. Distribution of Late Woodland/Late Prehistoric Sites ...... 167
11. Example of a Data S h e e t ...... 182
12. Modes of Functional Dimensions of Use-Wear Analysis ...... 185
13. Dimension K: Kind of Wear: Frequency of Modes 190
14. Dimension L: Location of Wear: Frequency of Modes ...... 190
15. Dimension 0: Orientation of Wear: Frequency of Modes ...... 191
xi 16. Dimension S: Shape of Worn Area: Frequency of Modes ...... 191
17. Dimension A: Angle of Edge: Frequency of Modes ...... 191
18. Excavated Localities at Water Plant . . . 203
19. Distribution of Features at Water Plant . . 204
20. Water Plant Features: Histogram of Plan Shape 206
21. Distribution of Potsherds by Weight, Water Plant ...... 207
22. Distribution of Lithics, All Samples, Water Plant ...... 209
23. Distribution of Functional Classes and Subsurface Features, Water Plant ...... 210
24. Feature Clusters at Baum ...... 221
25. Water Plant Catchment ...... 235
26. Baum Catchment ...... 236
27. Gartner Catchment ...... 237
28. Voss Catchment ...... 238
29. Ety Catchment ...... 239
xii / CHAPTER I
INTRODUCTION TO THE PROBLEM: RESEARCH GOALS
In the field of Ohio archaeology, the Late Woodland
period (A.D. 500-900) has long been regarded as a ’Dark
Age,' falling between the demise of the elaborate
Hopewellian burial mound and earthwork complexes and the
growth of the Fort Ancient culture (A.D. 900-1750) with its
large, often fortified villages and dependence upon maize
agriculture. Stoltman (1978) shows clearly how both the
McKern and Willey classifications failed to recognize a
formal period between the end of Hopewell and the rise of
Middle Mississippian. Late Woodland has largely been
defined by what it is not— not Hopewell, not Fort Ancient,
rather than by what it is. Excavation and survey of Late
Woodland sites has resulted in the description of a kaleidoscopic array of Late Woodland cultures for central
Ohio, including the Cole Complex (Baby, Potter, and
Saurborn 1967), the Peters and Chesser Phases (Prufer and
McKenzie 1966; Prufer 1967), Intrusive Mound (Mills 1922;
Morgan 1952), and the Newtown Focus (Griffin 1952). A
series of unpublished papers have addressed some of the ambiguities and confusions inherent in the cultural morass
collectively designated as Late Woodland in central Ohio
1 (Seeman 1980, Barkes 1982, and Rigg 1977), yet the picture remains clouded. What is known about the relationship of Late Woodland to Late Prehistoric cultures like Fort Ancient? What was the nature of the post-
Hopewellian groups which inhabited the central Scioto
Valley? How do they compare to the later cultural developments of the Late Prehistoric period?
With the recent excavation of new sites which fall into this period (Water Plant: Dancey, Fricke, Church 1987;
Zencor: Otto 1982), it becomes increasingly clear that a great deal of information exists which could be used to shed some light upon this ’Dark Age.’ However, this information has never been properly synthesized nor coherently analyzed. The archaeological remains from this period are widespread throughout the southern half of Ohio and adjacent portions of Indiana, Kentucky, and West
Virginia. A bewildering array of projectile point types and pottery types have been defined for Late Woodland and
Fort Ancient sites in this area.
Attempts to explain this variability and describe the transition from Late Woodland to the Fort Ancient Tradition have resulted in two basic hypotheses: population intrusion and local derivation. Based largely on early twentieth century excavations (Mills 1904, 1906) and work done in
1965 (Prufer and Shane 1970), a seemingly abrupt cultural transition has been hypothesized with the intrusion of a 3 new culture into the region, bearing a new technology
(maize agriculture), a new way of life (sedentary villages), and new ideas. The result of this intrusion was the restriction of local Late Woodland hunting and gathering groups to the hilly hinterlands, where they remained as a hostile presence to the newcomers (Prufer and
Shane 1970). This position has been echoed in later work
(Barber 1974, Ullman 1985).
Alternately, relatively recent work in northwestern
Kentucky (Rafferty 1974), in a section of West Virginia adjacent to the Ohio River (Graybill 1981), and in the
Miami drainage (Essenpreis 1978) suggests that the roots of
Fort Ancient in these areas are to be found in local, indigenous Late Woodland populations. Recent discussions of Ohio archaeology (Murphy 1975, Griffin 1978) also express this idea. The central Scioto Valley, home of the earliest Fort Ancient sites, figures prominently in such writings. In the case of Prufer and Shane (1970), their hypothesis was developed based on data drawn from a central
Scioto Valley site. However, available data from the drainage has never been examined specifically to test the hypothesis of intrusion versus that of local development in the region.
The first objective of this research is to resolve the culture historical problem by critically evaluating the intrusion argument through a detailed study of chronometric
.u (radiocarbon) and relative (seriation) dating information, resulting in a refined chronology for the sequence of cultural events in this area. This allows a decision to be made on the intrusion hypothesis and provides a background for the second two objectives of the research. Given the results of recent research, there is reason to believe that the local derivation hypothesis may hold in the central
Scioto Valley. Seriation of sites is possible between the
Late Woodland and Late Prehistoric periods, using cultural traits commonly employed as chronological markers-- ceramic tempering, stylistic motifs, and projectile points. In addition, chronometric dates from within the Scioto and adjacent drainages were plotted to determine whether a full-scale cultural invasion can be supported.
A second goal of this research is to compare sites in order to determine more precisely the differences and/or similarities which exist between the Late Woodland and early Late Prehistoric periods in this region. Based on preliminary analyses of the Water Plant site (a recently identified Late Woodland site which has been intensively surveyed and partially excavated), it appears that the magnitude of difference between Late Woodland and early
Late Prehistoric groups may not be as great as previously supposed. The economy of both may have possessed an agricultural component, the earlier one growing local crops
(sunflower, sumpweed, tobacco) with the later one emphasizing tropical cultigens (maize, beans, squash) while retaining some local crops as well. Living sites in both periods may have been fortified— Late Woodland populations choosing easily defensible locations such as bluff edges and adding earthwork walls, and Late Prehistoric groups erecting palisades around floodplain sites. The most obvious change over time appears to occur in the artifact assemblages, which perhaps may be understood in the context of requirements for processing, storing, and preparing maize.
A third and final goal of this research is to reach an understanding of the process of cultural change which occurred in the central Scioto drainage between A.D. 500 and A.D. 1250, from the beginning of the Late Woodland period to the end of the early Late Prehistoric period.
Given a revised picture of the culture history of the central Scioto Valley and information on various aspects of settlement pattern, it is possible to re-examine these
’Dark Ages' and gain some understanding of the how and why of post-Hopewellian changes in the region. Ruling out the intrusion of a brand-new culture and accepting the local derivation hypothesis for the origins of Late Prehistoric groups in the central Scioto drainage does not, after all, constitute an explanation of how or why culture changes occurred in one direction and not others. The final objective, then, is to explore alternative explanations for the changes which have been -documented for this region.
Two points regarding terminology used throughout the dissertation should be noted. First, the terms focus and phase are not used interchangeably. The term focus is used in an historical context— that is, in a discussion of the formulation of a culture history for the central Scioto
Valley prior to the work of Prufer and Shane (1970). The term phase is used when referring to modern research on the constituent parts of the archaeological record in the study area, beginning with Prufer and Shane (1970). Second, the descriptive term ’Fort Ancient’ is used only in an historical context to refer to post-Woodland cultural developments in the study area. The term 'Late
Prehistoric’ is preferred as it simply indicates temporal placement, referring to any number of similar cultures in the Eastern Woodlands which develop certain characteristics over time such as patterned villages and maize agriculture.
It does not presume cultural relationships across space nor time.
The present study examines possible relationships among Late Woodland and Late Prehistoric groups—-though analyses of properties of the archaeological record which bear on the nature of these groups. Chapter II addresses ths: nature of the Fort Ancient and Baum phase concepts and the idea of a Late Woodland period in central Ohio through an extensive literature review. Properties of the archaeological record and the cultural patterns of human groups are of little value unless placed in a larger environmental context. This context is described in chapter III. The following chapter, chapter IV, explores the history of archaeological research in central Ohio— specifically in terms of the Late Woodland and Late
Prehistoric periods, and summarizes general properties of the archaeological record such as number of sites, site size, and site location by physiographic province, stream order, and topographic feature.
The culture history of the central Scioto Valley is examined in detail in chapter V and the population- intrusion hypothesis, the first objective of this research, is tested against the data. The second research objective is then addressed in a comparison of data for Late Woodland and Late Prehistoric sites in chapter VI, with site summaries provided for some of the major excavated sites in the region. Chapters VII through IX expand this comparison through a consideration of site assemblages in a functional context, intra-site structure, and site catchment.
The final objective is treated in chapter X. Having examined the temporal relationship of Late Woodland to the
Late Prehistoric and explored the traditional elements of a settlement pattern analysis, this last chapter attempts to synthesize this material and to explain those similarities and differences noted in previous chapters. CHAPTER II
THE PROBLEM: LITERATURE REVIEW
The Fort Ancient Tradition
Fort Ancient came to be recognized as a distinct prehistoric culture in the central Scioto drainage late in the 1800s. Putnam (1886) distinguished the Fort Ancient culture from the cultures responsible for most of the mounds and earthworks, while Moorehead (1899) wrote that
Baum was impoverished compared to Hopewell and was the remains of some nonagricultural tribe since no pestles were found. Two years earlier Moorehead had suggested that Baum village
presents somewhat of a lower culture than others connected with great earthworks.... The character of the relics and the lack of evidence of high aboriginal art at this place are taken as evidence of the primitive character of the villagers. I do not think that they were the same people who erected the earthwork.... I am convinced that it antedates the construction of the works (1897:151).
Mills echoed this conclusion:
In the same valleys occupied by the Fort Ancient culture we find evidence of a higher culture, and for my convenience I have designated this culture the Hopewell Culture, taking the articles secured by Professor Moorehead from the Hopewell Mounds as the type. The Fort Ancient culture occupied portions of the valleys which were later occupied by the Hopewell culture, as evidenced by the results of the explorations of the Harness Mound group during the past summer [1905] (1906:136).
8 The excavation of the Robert Harness Mound in 1905 had yielded a burial near the top of the mound which contained obvious Hopewell artifacts such as copper earspools (Mills
1907). Mills felt that this burial, in comparison with others of the mound, was intrusive; thus, the rest of the mound population (designated as Fort Ancient) was placed stratigraphically earlier than Hopewell. This being the case, Fort Ancient was relatively dated as younger than
Hopewell.
Likewise, both the Adena Mound (Mills 1906; Holmes
1907; Smith 1910) and the Tremper Mound group initially were identified as Fort Ancient (Mills 1917), which appeared to support Mills' chronological placement of these cultures. Only when extensive work was underway at Seip
Mound group (Shetrone and Greenman 1931) was it realized that the archaeological record actually encompassed three distinct cultural manifestations and the current temporal sequence of Adena, Hopewell, and Fort Ancient was accepted.
In the meantime, the chronological position of the Fort
Ancient component of Serpent Mound was assigned a later date than that of the Hopewell culture (Shetrone 1920), while the Fort Ancient site was determined to be Hopewell in origin (Mills 1920; Shetrone and Greenman 1931;
Moorehead 1908). Only the nearby Anderson Village site was felt to be undeniably of Fort Ancient cultural affiliation. 10
Thus, by the 1930s Fort Ancient was firmly established as a late prehistoric culture extending across southern
Ohio, southeast Indiana, northern Kentucky, and possibly into West Virginia. James B. Griffin (1943) was the first to attempt a synthesis of available information on Fort
Ancient sites. Specifically, he was interested in how it compared with other, contemporaneous prehistoric cultures.
Basically, Griffin analyzed ceramics from the major Fort
Ancient sites known at the time— Madisonville (Hooten and
Willoughby 1920), Feurt (Mills 1917), Anderson and Baum
(Mills 1906), and Gartner (Mills 1904), supplemented by all of the smaller, more recently identified sites (including
Campbell Island on the Great Miami River, Mill Grove and
Taylor village on the Little Miami River, the Stokes site on Paint Creek, Hahn's Field and Sand Ridge on the Little
Miami near Turpin, and Proctorville on the Ohio River upstream from Portsmouth). He then compared his description of Fort Ancient with similar prehistoric manifestations, such as the Oliver Focus in central Indiana, the Middle
Mississippian, the Oneota in Wisconsin, and the Fisher
Focus in Illinois.
From his results, Griffin formulated the existence of four separate foci within the over-all Fort Ancient Aspect:
Baum, Feurt, Anderson, and Madisonville. These foci were described by the presence or absence of aspect traits
(ceramic types, point and knife types, shell artifacts, and 11 the bone industry). Griffin then compared his list of aspect traits with the McKern classification and master list of traits for previously described prehistoric midwestern cultures, finding its best 'fit1 with the Upper
Mississippian where it rated the designation of an aspect.
The Midwestern Taxonomic System (sometimes called the
McKern classification after one of its originators) was developed in the 1930s because 1) the Midwest at this time had a paucity of excavated, stratified sites, and 2) many archaeological collections were available in museum and private collections as an alternate source of information, if they could be used to construct a cultural-historical framework (Willey and Sabloff 1980).
Thus, as originally developed, the Midwestern
Taxonomic System was used to build cultural relationships based on shared similarities. Formal traits could be identified at the level of the component (generally this equaled a site or some level within a site). Components were then grouped together taxonomically, by virtue of sharing a high percentage of trait similarities, into foci.
The next step was the creation of an aspect, where substantial trait-sharing by foci existed. The phase,
pattern, and base represented the highest levels of the system's hierarchy, where common traits successively decreased until the base was defined by sites having horticulture and pottery versus those lacking these 12 traits. Once classified in this system, archaeological cultures could be placed into spatiotemporal perspectives by specific studies.
Although its original intent was not to be a system of spatial or temporal classification, units in the Midwestern
Taxonomic System gradually took on more than purely formal meaning. In the case of Fort Ancient, Griffin felt that
Baum was the earliest of the foci, with Anderson and Feurt developing concurrently towards the end of Baum.
Madisonville was designated the latest in the prehistoric sequence because its components contained European trade goods such as blue glass beads, an iron cross, and an iron bead embedded in a leather fragment (Hooten and Willoughby
1920). Basically, the foci represented regional phenomena,
with temporal differences within foci due to the degree of
Mississippian influence upon them.
With the advent in the 1950s of a new system of classification described by Willey and Phillips (1958), the
Fort Ancient Aspect became known as the Fort Ancient
Tradition and its foci became referred to as phases.
Formally, a tradition differed from the McKernian aspect
because by definition it emphasized the spatiotemporal
qualities of components. The idea of the phase replaced
the use of the focus, with the phase designating an
archaeological unit limited spatially and temporally and
fitting into a larger, local chronological sequence. As 13 we have seen, the terminological shift engendered few problems because these qualities had already been incorporated to some extent in the definition of Fort
Ancient by Griffin.
Temporal refinement of the phases occurred with the work of Prufer and Shane (1970). Starting with additional information from Blain village (33-R0-128), a newly excavated Fort Ancient village in the central Scioto
Valley, Prufer and Shane developed a model of chronology for the Fort Ancient tradition. This chronology divided the tradition into three periods: Early (A.D. 950-1250),
Middle (A.D. 1250-1450), and Late (A.D. 1450-1700). Early
Fort Ancient was represented by three phases: Baum, consisting of the sites of Baum, Gartner, and Blain in central Ohio; Baldwin, represented by Baldwin village and
Graham village in the Hocking Valley to the east; and Brush
Creek to the west, represented largely by the Fort Ancient component of Serpent Mound on Ohio Brush Creek. Middle
Fort Ancient consisted of the Feurt and Anderson Phases, located southeast and southwest of the Scioto drainage, respectively. In this temporal scheme, Feurt was thought to be the continuation of the Baum Phase and Anderson the outgrowth of the Brush Creek Phase. The Late Fort Ancient period culminates with a single phase, the Madisonville
Phase which encompassed the entire Fort Ancient region from
A.D. 1450 to proto-historic times. By the 1980s the culture history of the Fort Ancient
tradition had been spatially (Hanson 1966, 1975; Dunnell et
al., 1971; Rafferty 1974) and temporally delineated,
Internally partitioned, and a certain amount of
sociocultural reconstruction had been attempted (Essenpreis
1978; Graybill 1980; Barber 1974; Brose 1982). The Fort
Ancient tradition was summarized as consisting of village agriculturalists with an economy based on maize, beans, and
squash with some supplemental hunting (largely deer and
turkey) and foraging (nuts and berries) (Prufer and Shane
1970; Graybill 1981). It could be recognized archaeologically because of a distinctive settlement type and a distinctive material culture. Fort Ancient villages contained a central plaza and circular zones of habitation, refuse pits, midden, and burials. In some cases burial
mounds were found within the settlement. After A.D. 1250 villages were usually palisaded. Their pottery exhibited attributes such as shell-tempering, curvilinear and
rectilinear guilloches, line-filled triangles, punctates, various loop and strap handles, cordmarked or smoothed exteriors with a smooth interior, and a wide variety of
shapes and sizes (Griffin 1943).
Also distinctive are triangular points which have been
found in varying percentages at Fort Ancient sites, as were
stemmed points, triangulold knives generally twice the size of triangular points, groundstone celts, mortars, and 15 pestles. Bone tools were ubiquitous, especially turkey bone awls and fish hooks. Often deer and elk scapula were made into hoes. The presence of shell artifacts also characterized Fort Ancient sites, in the form of large quantities of local river mussels and ocean shell, both worked into disc and tubular beads, gorgets, hoes, and spoons.
The Baum Phase
Because this dissertation is concerned exclusively with the Baum Phase, it is useful to summarize the properties which distinguish it from other Fort Ancient phases. First of all, geographically it was confined to the central Scioto area. Secondly, radiocarbon dates of
A.D. 970, 1040, 1225, 1303, and 1440 from Blain village
(Prufer and Shane 1970:232) suggest that this phase may be the earliest of the tradition. Third, its material culture was judged over-all to be less well developed than that which distinguished later sites.
Griffin (1943) described the Baum focus (now phase) on the basis of four sites: two communities, one incompletely represented community, and a component at the Serpent Mound
which consisted solely of pottery. Table 1 represents the
trait list for the Baum focus, some of the traits unique to
Baum and others shared with one or more of the remaining
foci. Also defined for jBaum were two pottery types: Baum
Incised and Baum Cordmarked. Griffin noted that the Baum 16
Table 1. Baum Focus Trait List*
Location Traits
Unique to Baum: Cremation, crescent-shaped shell gorgets, perforated epiphyseal discs, turkey head rattles, arrow points from deer phalanges, perforated deer phalanges, beveled edge points
Shared w/1 Focus: Burials in or around structures, pearl beads, effigy head bone awl, bird digit pendants, cut lower animal jaws, beaded cylindrical awls, circular or oval structures
Shared w/2 Foci: Mound on village site, mound for burials, decorated pipes, platform pipes, rectangular shell pendants, perforated shell spoons, cut animal jaw ornaments,bone or antler pendants, shoulder blade knives and scrapers, antler points (hole in side), turtle shell spoons, decorated awls, slate celts, arrow shaft smoothers
Shared w/3 Foci: 98 traits ranging from types of burials and associated grave goods to types of pits, shell artifacts, bone artifacts, points, drills, mortars, pestles, and other lithic implements.
* Table taken from Griffin 1943:368-369, Table 7. and Gartner sites shared 80% of 123 traits. He also stated that the "significant fact that eighty-seven per cent of the focus traits are also aspect traits indicates a close correspondence of traits between the cultural divisions"
(Griffin 1943:69). Given that aspect traits were defined by those traits represented in three of the four foci and occurring in one or more sites of the fourth focus, this close correspondence seems circular in nature. Within each focus, each component was separately described (for example, the Baum, Gartner, Baldwin, and Serpent Mound components for the Baum focus). This same approach was utilized to differentiate the remaining foci. Griffin summed up the Baum focus as being most closely related to the Feurt focus, less so to the Madisonville focus, and least of all to the Anderson focus.
In his summary of pottery traits, Griffin noted the high percentages of grit-tempering (a trait frequently associated with Woodland pottery), and noted that at
Gartner a rim sherd projection was very similar to sherds from Woodland sites in Wisconsin and Illinois. At Baldwin, certain Fort Ancient traits of rim and decorative treatment served to distinguish it from Woodland pottery, but it was more than 90% grit-tempered. Serpent Mound pottery was mostly grit-tempered and Griffin suggested that by the cordmarked surface finish and tempering alone it was hard to separate this pottery from Woodland types. Other Woodland traits noted at Baum by Griffin
Included triangular points with a wide base, cremations, and the presence of a structure in the Baum mound. The
Baum focus was, therefore, recognized as having a particularly strong Woodland influence, although Griffin stated that the precise explanation for this influence was unknown. He assumed a Woodland background for the early, northern Fort Ancient sites, but questioned the justification for such an assumption.
Prufer and Shane (1970) elaborated and modified the
Baum phase as a result of their excavation of Blain village and examination of other sites in the middle Scioto Valley, as well as the Graham site excavation and regional survey in the Hocking River drainage. One important contribution of the Blain excavation was the acquisition of radiocarbon dates, making Blain the first Baum phase site to be dated chronometrically. The series of five radiocarbon dates obtained from the site ranged from A.D. 970 in Feature 2
(charcoal from a pit with a Woodland vessel) to A.D. 1440 in Feature 6 (charcoal) (Prufer and Shane 1970:232).
Prufer and Shane (1970:233) rejected the most recent date because of the proximity of the sample to the site surface
(six inches below the plow zone) suggested possible root growth contamination. A second late date, A.D. 1303, was rejected because the sample (charred corn) could have been biased through carbon-13 enrichment. 19
Prufer and Shane found the most acceptable dates to be
A.D. 970-1225, but felt that this represented an overly long timespan. Noting the lack of observable change In the material culture and the fact that some of the pit features were contemporaneous, they concluded that the earliest date for the site was most representative.
Prufer and Shane's evaluation of the radiocarbon data was correlated with their estimate of the chronological position of Blain in relation to other Fort Ancient sites based on percentage of shell-tempering. Assuming that the percentage of shell-temper can be used as a chronological marker, increasing over time in Fort Ancient sites, they ordered the sites included by Griffin in the Baum focus as follows: Blain, Graham, and Brush Creek are the oldest, based on 3.3%, 2.7%, and 0% of shell-temper, respectively.
Baldwin, with greater than 10% shell-temper, came next, followed by Baum and Gartner, both of which have approximately 20% shell-temper (Prufer and Shane 1970:234).
Another result of Prufer and Shane's analysis was the division of the Baum focus into three distinct units. In so doing, they adopted Willey and Phillips' (1958) classificatory system. As they understood it,
a prehistoric Tradition is a cultural pattern, traceable through time within determinable geographic boundaries.... The subunits of a Tradition are Phases. They are characterized by internal homogeneity through space and time, but the range of variation remains within the 20
framework of the over-all Tradition to which they belong (Prufer and Shane 1970:239).
They assumed that these theoretical divisions of Fort
Ancient corresponded with "some sort of effectively interacting [emphasis original] socio-cultural units of varying degrees of magnitude" (1970:240).
Armed with this concept of the phase, they subdivided
Griffin's Baum focus into the Baldwin, Baum, and Brush
Creek phases. Approximately contemporaneous, each phase was distinguished primarily on the basis of ceramics, as well as traits like settlement pattern and tool assemblage composition. The revised Baum phase included five sites:
Baum, Gartner, Blain, McGraw's Garden site, and Kramer A site (the last two recorded during the Scioto Valley
Archaeological Survey).
Blain Village was described in detail as representative of the typical Baum phase cultural pattern, including settlement pattern, resource utilization, land use, and social structure. Based on a total of forty-seven features (including a central burial mound, a plaza, three house patterns, a 'ceremonial' pit, refuse pits, and sheet middens), Prufer and Shane concluded that Blain Village consisted of "single-family dwelling areas... surrounding a plaza-like open space in which the burial mound was located" (1970:246), based on an economy of agriculture and hunting and gathering. The settlement pattern was 21 described as a nucleated village surrounded by farmland, open grassland, and second growth cover. No outlying hamlets or camps existed and no Fort Ancient settlements were located in the hills. The social structure was inferred to center on a "unilinear descent system with ranked clans and lineages" (1970:255). This inference derived from an examination of burials, burial locations, house size and shape, and the fact that no quantitative or qualitative differentiation was manifested in the distribution of grave goods (1970:255). Kin groups were suggested by the presence in the mound of individuals of both sexes and all ages, plus the possibility of patterning in the location of non-mound burials next to houses.
The Late Woodland in Central Ohio
As one of the goals of this research is to examine the origins of the Baum phase, the current state of knowledge of Late Woodland will be summarized in this section.
Unlike Fort Ancient or the Baum phase, this has not been studied as intensively nor does a recent synthesis exist in print. Currently defined Late Woodland cultural manifestations will be summarized for central Ohio, and recent critical assessments of these will be synthesized as well.
The Cole Complex was first described by Baby and Potter
(1965) and Potter (1966) on the basis of ceramic similarities among the sites of Cole, Lichliter, Voss, and Zencor (known
.x 22 today as the Scioto Trails School site). The basis for their description began with the recognition by Morgan
(1947) of the W.S. Cole site as a Late Woodland site. The purpose of the research by Baby and Potter (1965) was to do a preliminary study of Late Woodland pottery and to describe its relation to Hopewell ceramics. The complex itself rests largely upon the definition of Cole Cordmarked and Cole Plain pottery types. Utilizing ceramics from the aforementioned sites, Baby and Potter limited their study to rim sherds and shoulder sherds. Their sample consisted of twenty-three rim sherds from the Cole site (most of them cordmarked and all grit-tempered with angular and rounded shoulders). Zencor provided 196 rims, while Voss Mound yielded over 100 sherds, three of them shell-tempered
(Baby, Potter, and Mays 1966). Among these sherds were found guilloche designs incised over partially obliterated cordmarking. Lichliter had thirteen rim sherds which were grit-tempered with cordmarking on the neck, rim, and body nearly obliterated (Allman 1957).
Based on this sample, two Cole pottery types were defined. Baby and Potter (1965) distinguished Cole
Cordmarked by form, surface, and lip treatment from previously described Late Woodland pottery types like
Newtown Cordmarked. Comparing their results to Hopewell ceramics, they felt that some sherds from Seip and also from McGraw (Prufer 1965) were very similar to Cole. 23
They concluded from this that Cole Cordmarked pottery represented the Cole Complex, a post-Hopewell, basic
Woodland or Scioto Tradition present in the Ohio Valley from Late Adena to Fort Ancient. This suggestion was amplified by Potter (1966).
With the publication of excavation results from the
Voss mound, Baby, Potter, and Mays (1966) expanded the definition of the Cole Complex to include evidence for the ceremonial life of "Cole Indians". Voss mound yielded triangular points and knives, marine shell, turkey bone awls, a buried structure under the mound (radiocarbon dated to about A.D. 966) and two burial pits dug into the mounds.
The presence of triangular points and shell-tempered pottery was interpreted to mean contact had occurred with
Fort Ancient groups to the south. Voss village, excavated in 1966 (Baby, Potter, and Sawyer 1967), yielded radiocarbon dates of A.D. 910 to A.D. 1500, which accorded well with this conclusion.
Cole, then, was seen as a lineal Late Woodland descendant of Ohio Hopewell, and in 1966 Baby and Potter added the sites of Erp, Hudson, Shipley, Fishinger Park, the Wolf Rockshelter, and Swinehart Village to the Cole
Complex, expanding the diagnostic trait list to include
Cole points, chipped slate discs, chipped stone celts, circular houses, and small villages on second terraces. Prufer and McKenzie (1966) described a new Late
Woodland phase based on the excavation of Peters Cave, Ross
County, Ohio. The Peters phase was a Late Woodland culture located in the unglaciated portion of southeastern Ohio.
They defined it on the basis of two pottery types (Peters
Plain and Peters Cordmarked) which have more rounded shoulders than other Late Woodland pottery types, no interior cordmarking, as well as grit-, limestone- or chert-tempering, fine vertical cordmarking, and flat lips.
Peters Plain has by definition a smooth exterior and grit- or flint-tempering. Triangular points, Chesser-Notched points, and ’fishspear' points comprise the diagnostic lithic assemblage. Prufer and McKenzie (1966) summarized the Peters phase as a system of riverine villages and seasonal upland hunting camps which frequently occurred in rockshelters such as Peters Cave.
The Chesser phase, described in the Chesser Cave
(located in Athens County, Ohio) excavation report, is much like the Peters phase (Prufer 1967). With a radiocarbon date of A.D. 1070 the Chesser phase is seen as temporally later than Peters phase. Pottery is predominantly limestone-tempered, but a small percentage of other pottery types are present, including sherds described as Peters
Plain and shell-tempered plainware. Prufer (1967) summarized the Chesser phase as belonging to an indigenous 25
Scioto Tradition which included the Peters phase, but
excluded Fort Ancient as intrusive to the valley.
A recent attempt to address questions concerning the nature of these Late Woodland cultures was made by Seeman
(1980), who compared and contrasted the Late Woodland phases of Cole, Peters, Chesser, Newtown, and Intrusive
Mound in the middle Ohio valley temporally, spatially, and in terms of material traits. Of interest here are his possible interpretations of the Cole Complex. He suggests that 1) its sites immediately precede Fort Ancient, or, 2) it constitutes a separate, less complex culture that was contemporaneous with Fort Ancient. Favoring the latter alternative (without explaining why or what the nature of
that culture might have been), Seeman reassigned Voss and Lichliter to Fort Ancient, but kept the sites of W.S.
Cole, Fryman, Green Camp, and Ufferman as a separate Late
Woodland complex at least partially contemporaneous with
Fort Ancient to the south.
Barkes (1982) reconsidered the nature of the Cole
Complex, based on her analysis of Cole ceramics. She saw a
degree of homogeneity among the ceramics from the W.S. Cole
site, DECCO, and Ufferman, based on crushed rock temper,
cordmarking, and the general similarity of vessel
morphology. While typical Fort Ancient characteristics
such as punctates, incised designs, shell-tempering, and
strap handles were absent, she noted that the Cole Complex 26 is contemporaneous with Fort Ancient if the radiocarbon dates of A.D. 1130 and A.D. 1160 from DECCO are correct.
She noted also that the ceramics from these sites contrast with the Newtown Late Woodland complex, which dates between
A.D. 450 and A.D. 900 (Seeman 1980).
Barkes does find a few similarities between Cole and early Fort Ancient ceramics based on the presence in both of collared rims and rims with nodes, lugs, and flanges.
However, Baldwin, Graham, Blain, Voss, and Erp all exhibit
Fort Ancient ceramic characteristics such as punctates, strap handles, and incised line designs which are absent from Cole ceramics. Thus, Barkes (1982) summed up the Cole
Complex as culturally distinct but temporally contemporaneous with early Fort Ancient and possibly the
Peters phase in southern Ohio. The Cole ceramic complex, then, was viewed as a regional Late Woodland manifestation confined to the upper reaches of the Scioto and Olentangy
River drainages.
Thus, several cultural phases appear to exist within the Late Woodland period in the middle Scioto Valley.
These phases have been defined on the analysis of a handful of sites, with analysis largely centered on ceramic attributes. Clearly, further research on a broader scale of analysis is dictated. In the following chapters the traditional elements of settlement pattern analysis will be examined to determine more precisely the nature of Late 27
Woodland cultural groups in this region. Once a better understanding of Late Woodland is reached, a comparison can then be made with Late Prehistoric adaptations.
Theories on the Origins of Fort Ancient
Since the Baum Phase is the earliest manifestation of
Fort Ancient, literature on the origins of Fort Ancient as a whole is directly relevant here. The following section presents an historical overview of theories which attempt to explain the origins of Fort Ancient. Research from outside the central Ohio area which bears on this topic will also be considered (Rafferty 1974, Graybill 1981).
The first explanation for the origins of Fort Ancient was put forth by Griffin (1943), who suggested that early
Fort Ancient was the result of stimulus diffusion from the
Mississippian heartland after approximately A.D. 900 -1000.
This was followed by actual migration during the later
Madisonville focus (Griffin 1943). These ideas were based on his interpretation of the Baum trait list (drawn from its archaeological components), which he felt indicated the process of acculturation in which local Late Woodland groups acquired certain selected Mississippian characteristics. The suggestion that actual migration occurred later seems to be based in part on the wide geographical distribution of Madisonville sites, compared to the earlier, more restricted distributions of the Baum,
Feurt, and Andersonville foci. McKenzie (1967) excavated Graham Village in the
Hocking River Valley and offered further support for the
possibility of migration. As part of the Baldwin phase,
McKenzie felt this site, with its impoverished artifact
Inventory (compared to other early sites such as Blain,
Baum, and Gartner), represented an Incipient Fort Ancient
site that originated with an actual influx of Mississippian
population into this area, even though the site appears to
be geographically and culturally marginal to most Fort
Ancient developments. Based on a database of about
fourteen pits (which together yielded one corncob), one poorly preserved burial, no structures, no palisade, and
very little artifactual material (a few triangular points and some Baum-like pottery), McKenzie suggests that a
typical Fort Ancient site consisted of small, scattered hamlets with a limited material culture, and that larger sites are anomalous.
Given that sites such as Baum are the rule and not the exception, this particular idea has not received much support. However, McKenzie did offer the alternative view that Graham could represent a recently acculturated
Woodland population, or Fort Ancient with strong Woodland influence.
One of the most forceful statements on the origins of
the Fort Ancient tradition in Ohio can be found in the
Blain Village excavation report by Prufer and Shane (1970). They conclude that the Baum phase results from population
intrusion into the area, based upon the "massive and sudden appearance of too many new traits around 950 A.D.," (Prufer and Shane 1970:259), acknowledging regional differences in the degree of Mlssissippification of the various early Fort
Ancient phases— Baum, Baldwin, and Brush Creek. Prufer and
Shane suggest that early Fort Ancient cannot be explained as slow acculturation as a result of gradual direct diffusion or stimulus diffusion of new traits and ideas from outside sources, postulating instead "an actual invasion of southern Ohio by Mississippians, resulting in the physical and/or cultural annihilation of the older
Woodland cultures1' (1970:258).
Their evidence for such a view is based on radiocarbon dates and "the occurrence of occasional shell-tempered sherds in otherwise pure Late Woodland contexts [emphasis added]" (1970:258). They infer from this fact that local
Late Woodland populations were displaced into the hills as a result of the arrival of "strong Mississippian population units in the major valleys" (1970:258).
They examined archaeological evidence from rockshelters in southern Ohio to determine the degree to
which these Late Woodland populations were actually displaced valley dwellers (excavating five rockshelter
systems and reviewing published reports of others). The
conclusion of this research was that minimal reflection of 30 valley cultures occurred in the cave deposits (1970:261).
At one such site, Chesser Cave, Prufer (1967) noted the presence of an unknown series of untyped, shell-tempered, plainware sherds from Shelter A. He described the location of these sherds as follows: ”[t]he horizontal and vertical distribution of these sherds within the excavation units would seem to indicate that they were not intrusive into the deposits" (1967:15), being distributed horizontally over an area of 175 square feet and vertically through four levels of undisturbed deposits. He continued: "Two additional sherds were found... in the rear-central area of the shelter, in association with the usual Late Woodland materials ubiquitously encountered throughout the site"
(1967:15). From this evidence, he concluded:
At present, the best guess is that the plain shell- tempered sherds from Chesser Cave are affiliated with the Mississippian Fort Ancient Aspect. Certainly, these shell-tempered sherds strike the only discordant ceramic note in the Chesser Cave assemblage, which is in all other respects clearly of Late Woodland affiliation. Furthermore, if the sherds in question do indeed represent a Mississippian element at the site, they are quite atypical of Fort Ancient ceramics. This fact is the more obvious, since classic Fort Ancient ceramics are entirely absent from the deposits. (Prufer 1967:16).
In the same report he also stated that given the radiocarbon dates for Chesser Cave, the presence of the shell-tempered sherds "suggests very early Mississippian influences upon a basic Late Woodland group living in the 31
Appalachian hinterland, far from the main routes of cultural diffusion” (Prufer 1967:17).
It is instructive at this point to compare Prufer's
1967 conclusions with those made in the Blain Village report (Prufer and Shane 1970), as demonstrated in the following quotes.
"We conclude from these data that some of the roots of early Fort Ancient are to be sought for in local Late
Woodland cultures....” (1970:238).
”... older and possibly hostile Woodland peoples occupy the hilly uplands, while Fort Ancient activities are restricted to the lowland areas” (1970:253).
Compare these statements to his explanation of activities at Chesser Cave, where he reports that the faunal analysis indicates annual occupation for seasonal hunting "by groups of specialized hunters whose more permanent settlements were located elsewhere, perhaps in the valleys of the Hocking River or Raccoon Creek"
(1967:5). What, then, is the explanation? Does the mingling of Late Woodland/Fort Ancient characteristics represent evolutionary development, acculturation, invasion, diffusion, or some combination of all of these?
While Prufer and Shane generally support population intrusion, it is clear from a careful perusal of their writing that even they vacillate from chapter to chapter and article to article. Outside central Ohio, migration, or population
intrusion, was also used to explain the origins of the Mayo
Focus or Woodside Phase in eastern Kentucky (Dunnell, et
al., 1971), based on the excavation of the Slone site (and
additional Fort Ancient sites identified through survey).
The Slone site is palisaded and has a circular to
elliptical plan with four distinct zones of activity: a
plaza circled first by a cooking area defined by the
presence of storage pits, cooking fires, and porticos or
posthole configurations that did not represent a closed
geometric figure, then by a ring of structures (rectangular
with rounded corners and interior hearths, about 4.3 m x
5.5 m), followed by a sheet midden ring against the
palisade. Dunnell et al. (1971) concluded that the traits
represented by the Slone site— both artifactual and
settlement pattern— were so different from previous archaeological manifestations in the area that they must
represent actual migration of new populations into the
region.
In other work on Fort Ancient, Rafferty (1974)
examined the development of the Fort Ancient tradition in
northern Kentucky. She analyzed data from the Cleek-McCabe
site, testing for stylistic and functional continuity
and/or change from Woodland to Fort Ancient. Reasoning
that the demonstration of continuity in one or both areas
would rule out the migration hypothesis, then the reverse 33 would also hold. If no continuities were found, local development and the acculturational hypotheses would be discounted and migration would remain a potential explanation. She seriated sites based primarily on the attributes of temper and surface finish of ceramic types, but secondarily relied on burial types and projectile points to measure stylistic continuity. Also, settlement pattern and subsistence bases were analyzed for evidence of functional continuity. She concludes that evidence of stylistic continuity over time does exist.
She then investigates the earliest appearance of each stylistic marker in time and space of the so-called Fort
Ancient style that supposedly had been derived from Middle
Mississippian— curvilinear and rectilinear guilloches, strap handles, triangular points, ground stone disks, elbow pipes, and shell-tempering; 1974:210). Her evaluation of the members of this traditionally defined Fort Ancient constellation of traits leads her to conclude that, taken as a set, these styles do not originate in one area and thus do not support the acculturation or migration hypotheses. Furthermore, she states that these traits do not occur together even in Middle Mississippian until after they are seen in Fort Ancient.
Less successful with her functional types, Rafferty finds that the evidence is insufficient to discriminate between local development or acculturation as the best 34
explanatory hypothesis. She does, however, establish that
maize is present in the area before the advent of Fort
Ancient and that a sedentary settlement pattern was present
as well. For both functional classes— settlement pattern
and subsistence— she relies largely on evidence from
central Ohio, since it is here that the earliest Fort
Ancient sites are located and here where Late Woodland
sites are also well-documented (for example, Lichliter,
Peters and Chesser Caves).
Her conclusions are that local development is the best
hypothesis because of stylistic continuities between Late
Woodland and Fort Ancient and because no group of styles
appears in early Fort Ancient that has "a coherent origin
elsewhere" (1974:221). Those Fort Ancient stylistic attributes which do not have precursors in local Woodland
phases may have originated outside of the region or within
the Fort Ancient tradition itself. She suggests that while no strong evidence for functional continuity is available,
likewise no evidence exists to support the idea that the
settlement pattern and subsistence base coincidentally arose in the area with the appearance of Fort Ancient.
Finally, Rafferty concludes that agriculture developed independently in each area, and because an economy was already established in the Fort Ancient area equally as efficient as Middle Mississippian given the local environment, one can rule out expansion of Mississippian either by migration or acculturation. Rafferty predicts that a gradual transition would have occurred geographically from one area to the next if Fort Ancient groups were descendants of Middle Mississippian groups, but notes that this is not the case. Thus, the agricultural, nucleated Fort Ancient villages were built upon the introduction of maize into the region several centuries earlier and upon the process of nucleation which was already established in local Late Woodland settlements.
Local development of Fort Ancient populations from Hopewell in northern Kentucky was also supported by the seriation of
Hopewell and Fort Ancient sites based on pottery.
In the following year, Hanson (1975), based on the excavation of the Buffalo Village site, West Virginia, concluded that Fort Ancient was a series of phases within the general Woodland Tradition which were set apart largely by the degree to which they were influenced by various phases of the Mississippian tradition to the south. That is, acculturation, not migration, was responsible for the origins and development of Fort Ancient.
Later, Essenpreis (1978) built upon Griffin's hypothesis to explain the differences between the Anderson and Madisonville phases. According to her, the earlier
Anderson phase was the result of stimulus diffusion of ideas from Mississippian populations in terms of subsistence, stylistic attributes, and settlement pattern, with the subsequent acculturation and development of local regional populations. The more complex Madlsonville phase, for which she described a hierarchical settlement pattern based on the sites of Campbell Island, Madlsonville, and
Marietta, was a result of direct migration (cf Graybill
1980). This direct migration also resulted in the rapid expansion of Fort Ancient into the previously unoccupied areas of eastern Kentucky and West Virginia. Heilman and
Schwab (1980) also support this view of initial stimulus diffusion followed by actual migration, based on conclusions that discernible physical differences exist within the burial population at the Incinerator site (an
Andersonville phase site).
Local derivation figures prominently in the work of
Graybill (1980, 1981, 1984) for the eastern periphery of
Fort Ancient in West Virginia. Graybill orders a series of habitation sites over time through the seriation of ten independent measures, radiocarbon dates, and typological cross-dating. He then correlates this chronological summary with observations on site form and location to find the best explanation for observed changes. He concludes that the amalgamation of population is responsible for change over time in the Feurt-Clover phase settlement pattern, with nucleated villages resulting from climatic deterioration, extreme competition, and intense warfare
(1981:173). He also suggests that "the origins of the 37
Baum/Anderson tradition fall squarely within the preceding
Woodland era” (1981:33) and that the ceramic assemblage, circular village plan, and maize agriculture can all be derived from local antecedent cultures.
The most recent work to address the question of the origin of Fort Ancient has been done by Ullman (1985), who undertook a study of the Kramer site ceramics 1) to describe the ceramic assemblage from the 1967 excavations and determine spatiotemporal relationships, and 2) to use the ceramic data "to ascertain which of the current views of Fort Ancient is best supported" (Ullman 1985:6). Ullman concludes that Kramer is a late Baum site based on perceived site homogeneity in tool types, feature classes, subsistence, and the fact that the percentage of shell- tempering is higher than for other Scioto or Paint Creek sites. According to Ullman, a regional, temporally-based stylistic trend may be noted, whereby the amount of curvilinear incising decreases through time. An acculturational-derivation of Baum is put forth in this work, based upon the types of sherds present (some Feurt and Anderson types) which makes the situation at Kramer more 'dynamic' with an increased level of inter-regional interaction, communication, and/or influence, while the presence of Hocking Valley-like ceramics may indicate an
"indigenously-inclined sense of conservatism among certain
Kramer potters" (Ullman 1985:122). 38
Summary and Conclusions: Directions for New Research
An analysis of the literature has revealed that the concepts of Fort Ancient, the Baum Phase, and Late Woodland are based squarely in the culture-historical tradition.
Each has been described and differentiated from other archaeological cultures based on formal properties of their material cultures. Does Fort Ancient have concave-based triangular points as opposed to side-notched points? Do
Baum phase sites have Baum Cordmarked pottery? Do Late
Woodland sites lack Hopewell-like artifacts?
The question of the origins of Fort Ancient has been batted about like a tireless kitten at play with a ball that is never stilled. As recently as last year (Ullman
1985), yet another attempt was made to refine the description of Fort Ancient and the Baum phase material culture in order to evaluate theories on their origins. It is left for some new researcher to attempt an even finer, more precise definition of Baum Cordmarked, var. X ceramics, to keep the ball in motion. The problem remains unresolved.
What is needed is a fresh approach to the archaeological record. Without taking up the banner of the
'new archaeology' and speaking loftily of discovering universal laws of human behavior, it is possible that other avenues of inquiry might prove more profitable. Recently, 39 a new style of research (Roper 1979, Smith 1984, Dancey,
Wymer, and Waterworth 1986) has turned to a study of the history of the settlement of prehistoric cultures, rather than simply studying and re-studying their formal properties. This study, with its emphasis on the analysis of elements of the settlement patterns of Late Woodland and
Late Prehistoric groups, follows such an approach. It should be possible to build upon and expand previous research efforts, providing a more secure resolution of the problem. CHAPTER III
ENVIRONMENTAL DESCRIPTION OF THE CENTRAL SCIOTO DRAINAGE
Introduction
The Late Woodland and Late Prehistoric, groups who inhabited the central Scioto drainage did not do so in a vacuum. Living in this region meant adjusting to environmental conditions. Obviously, knowledge of the environment is an important component of any consideration of economic and subsistence behavior. The environment is also a basic factor in settlement patterning. Just as it can be a decisive influence upon elementary aspects of human behavior, the reverse may also hold. Human behavior can impact the environment. For example, prehistoric irrigation in the Near East and American Southwest made fertile soils useless with the build up of salts. Thus, it is impossible to study prehistoric cultures of the Scioto
Valley without exploring their environmental setting.
This chapter summarizes such basic environmental variables for the region as climate, hydrology, bedrock geology, physiography, present topography, soils, and vegetation. It also examines the controversial issue of the extent to which climatic variability may explain cultural variability over time for this area.
40 41
Climate
Based on thirty-year records from a weather station at
Chillicothe in Ross County (Wise and Rayner 1966), the climate of the study area is summarized. Central Ohio has a continental climate that is characterized by large annual, diurnal, and day-to-day temperature ranges. The region lies below a circulation system known as the
Westerlies and thus has a high percentage of westerly winds. It also lies in the zone known as the Polar Front, which subjects it to humid air from the Gulf of Mexico and cooler polar air. As a result, the weather changes every few days.
Summers in this region are warm and humid, with an average daily maximum temperature of 84°-88°F in July, while, conversely, winters are cold and cloudy with an average minimum daily temperature of 19°- 25°F in January.
Given this large annual diurnal and day-to-day variability in temperature, the latest date for the beginning of the growing season is 26 May in the northern part of central
Ohio to 18 May in the southern part. Likewise, the earliest end of the growing season may occur from 15
September in the north to 23 September in the south.
Precipitation is normally abundant and well distributed, with minimum amounts received in the fall.
The annual mean precipitation varies between 90.6 cm to
106.1 cm, and generally increases in amount from north to 42 south. This area is subject to a high evaporation rate in the summer, when it also receives approximately 67-71% of possible sunshine in July, comapred to 32-38% in December.
Snowfall averages 75.1 cm in the extreme northeast portion of central Ohio to less than 38.8 cm in the southwest. In the Columbus area, mean precipitation is about 94 cm with mean evaporation at about 66 cm, thus leaving about 28 cm of runoff available.
Climatic Change
The above information describes the modern climate of the study area (Vise and Rayner 1966). Considerable discussion has appeared on the question of the prehistoric applicability of such a description to this area.
Griffin (1961) hypothesized that the decline of Ohio
Hopewell could be correlated with a colder-than-present period which lasted from A.D. 200 to A.D. 700. Baerreis and Bryson (1965) and Baerreis, Bryson, and Kutzbach (1976) developed a detailed description of climatic variability for the Midwest in general and suggested (1965) a series of regional climatic episodes. The Sub-Atlantic period from about 600 B.C. to A.D. 400, for example, was defined as a time which saw some areas of the Midwest cooler and wetter than at present. The period from about A.D. 1350 to A.D.
1850 was called a "Little Ice Age," originally based on
European data which were assumed to correlate with North
America. In their 1976 article, based on varved lake sediments
from Hell's Kitchen Lake in northcentral Wisconsin, they
suggest the midwestern climate was cooler and wetter at
about 100 B.C. to A.D. 350, followed by a warmer or drier period. Thus, they consider it likely that at the time of
Hopewell decline, the climate was cooler and drier. For comparison, they utilize a pollen record from the White
Mountains, which indicates cooler temperatures beginning about A.D. 250 and ending at A.D. 700 (corresponding with
Middle Woodland to Early Late Woodland), followed by a
cool-dry period from A.D. 700 to A.D. 1000 (corresponding
with Early Late Woodland to Early Late Prehistoric). This
sequence was preceded in 250 B.C. to A.D. 250 (Middle
Woodland) by a warmer period. Subsequent periods, e.g.,
from A.D. 1400 to A.D. 1850 (Middle Late Prehistoric to
Historic), indicate a cooler climate, changing from moist
to drier at Hell's Kitchen Lake. This does, of course, correspond well with the "Little Ice Age."
Wood (1976) deals with climatic episodes in terms of evidence from vegetational reconstructions. Government
Land Office (GL0) survey reconstructions, he feels, were
done at the end of the Little Ice Age when the climate was cooler and moister than at present. Such reconstructions
in his opinion produce a vegetation which would best correlate with paleoclimatic episodes corresponding with
Late Middle Woodland and Late Woodland times, as "much of 44 the Woodland occupation of eastern and midwestern United
States took place during the very cool Scandic climatic episode (ca. A.D. 400-700)" (Wood 1976:207). This is similar to climatic conditions suggested by Baerreis,
Bryson, and Kutzbach (1976).
Graves (1977), however, directly applies data from
Ohio to the corroboration of such climatic episodes.
Utilizing the Silver Lake (Logan County, OH) pollen sequence (Ogden 1966), he compared it to the Pretty Lake pollen core from northeastern Indiana and to the Soden Lake sequence from southern Michigan to reconstruct the paleoclimate of Lake Erie's western basin region.
The first two sequences had the added advantage of good C-14 dates. Specifically, the Silver Lake sequence suggests a drier, warmer climate from 900 B.C. to A.D. 640
(Early Woodland to Early Late Woodland), followed by cooler, moister conditions from A.D. 640 to A.D. 1690
(Early Late Woodland through the Late Prehistoric). This latter period was characterized by a mesophytic forest of oak, hickory, beech, ash, elm, maple, walnut, sweet sumac, cherry, and basswood. This does not correlate with
Griffin's suggested beginning of a cooler phase at A.D. 200 nor other reconstructed paleoclimatic episodes pre-A.D. 700
(Wood 1976, Baerreis, Bryson, and Kutzbach 1976). The latter portions of the White Mountains sequence and the
Silver Lake data do, however, rather closely correspond. Thus, it seems likely that over time considerable variability existed within micro-paleoclimatic zones within the Midwest and that these micro-zones or regional areas cannot readily be correlated with other such zones within the Midwest nor beyond. Baerreis and Bryson (1965) point out several times that the episodes suggested by them may not be universally applicable. As Vickery (1970) points out, 1) the use of local data is very important in paleoclimatic reconstructions, and, 2) pollen studies more readily and accurately reflect climatic changes than does the movement of glaciers.
The Silver Lake sequence is the closest local pollen study done near the area of the present study. This sequence has been shown to be internally consistent with pollen sequences from southern Michigan and northeastern
Indiana and is radiocarbon-dated. Therefore, it is proposed here that it accurately reflects a local, micro- climatic zone which shifted from a warm-dry climate to a cooler-moister one that corresponded with the beginning of
Late Woodland times through the Late Prehistoric period.
Graybill (1981) suggests a similar scenario with his statement that a deteriorating climate in the Great Lakes region affected economic productivity and strained inter village relationships between Whittlesey and Late
Prehistoric populations in southern Ohio. 46
Hydrology
Four main aquifer groups are present in central Ohio.
Two are associated with bedrock formations: 1) Silurian and
Devonian dolomites and limestones, and 2) Mlssissippian
Berea sandstone. Two are associated with Quaternary glacial features: 1) superficial glacial deposits, and 2) valley fills and outwash terraces.
According to Ahmad and Blackie (1966),
Unconsolidated glacial deposits, especially sands and gravels, are perhaps the principal ground water yielders in the Central Ohio region. They have high porosity and permeability, and therefore usually transmit water readily. Since they are also found at or near the surface, they are susceptible to rapid recharge" (Ahmad and Blackie 1966:5-6).
Outwash terraces of well-sorted sand and gravel deposits laid by stream are the best unconsolidated deposits for obtaining ground water, such as those found in the
Olentangy, Paint Creek, and Scioto valleys.
Surface water availability within the Scioto drainage ranges from the following figures taken on each major tributary which is represented by a prehistoric site in the present study: 1) 3278 cfs on the Scioto at Chillicothe with a drainage area of 9969 km^, 2) 788 cfs at Paint Creek near Bourneville with a drainage area of 2090 km , 3) 178 cfs on the Big Walnut near Central College with a drainage area of 492 km^» and 4) 431 cfs on the Big Darby at
Darbyville with a drainage area of 1383 km* (Ahmad and Wood
1966). 47
Bedrock Geology
In a generalized stratigraphy, it may be said that In central Ohio the surface exposed rocks are Paleozoic, predominantly marine limestone and shale. The Mesozoic and most of the Cenozoic are represented in Ohio by a major period of erosion. Glacial deposits of the Pleistocene are found directly on Paleozoic rocks north of the glacial border.
The Devonian System "comprise the surface rocks over most of the Scioto River basin" (Smith 1966:6.11). This system has a thickness ranging from 83.8 m to 243.8 m. Of marine origin, the basal formations are predominantly dolomite, dolomitic limestone, and limestone while the upper formations are shale and sandstone. Four of the eight formations of this system which are important to central Ohio are the Columbus, Delaware, Olentangy, and
Ohio.
The Columbus Limestone exists in a belt trending north-south at about 40.2 km wide through the Scioto River basin. It has a thickness of 24.4-38.1 m. Its basal portions are porous, massive brown limestone with often bituminous matter, while the upper two-thirds are predominantly evenly bedded light gray limestone. Locally, it contains white or light gray fossiliferous chert and in the top 15-20 cm contains abundant remains of fossil fish. 48
The Delaware Limestone extends above the Columbus
Limestone with a thickness ranging from 9.1-21.3 m, having the same distribution. Its composition ranges from massive, bluish-gray, impure limestone to interbedded shale between 5- 0.6 m thick. The Olentangy Shale is bluish-gray to greenish-gray clay shale with black fissile shale beds in its upper portions. It is about 8.5 m thick.
The Misslssippian System has a width of 32.2-64.4 km in central Ohio and a thickness of 91-305 m, being closer to this last figure for central Ohio. It is all clastic material, dominantly shale with lesser amounts of sandstone and conglomerates. Six of its formations and several members of each are all exposed at the surface in central
Ohio.
Physiography
Physiographic development and present topography in central Ohio are largely the result of glaciation of the
Teays River system. In preglacial times central Ohio was drained by the Teays River. Originating in the Piedmont of
North Carolina and Virginia, the Teays flowed across West
Virginia to the present Ohio River at Wheelersburg where it turned north to Waverly, followed along the course of the present-day Scioto River to Chillicothe, and then crossed
Ross County into Pickaway, Fayette, and Madison Counties.
Three to four glacial advances and subsequent retreats
(Nebraskan, Kansan, Illlnoisian, and Wisconsin) modified 49 this preglacial drainage, eroding the former landscape in places, and depositing sediments in others. These activities resulted in the present Till Plains section and the Glaciated Allegheny Plateau. Within central Ohio,
Illinoisian and Wisconsin deposits are exposed at the surface. These consist of ground moraine, ice-contact sediments, and glacial outwash from the Illinoisian, along with surface material over most of the central Ohio region north of the border of the Wisconsin glaciation. In this area, flat ground moraine is intermittently broken by belts of hummocky end moraine. Near the southern margin of the
Wisconsin deposits, rolling bedrock hills are overlain by a thin mantle of ground moraine. Several levels of glacial outwash terrace (predominantly sand and gravel) are found in the major valleys of central Ohio. Lacustrine silts are found occasionally in valleys of secondary streams or as thin veneers behind end moraines.
Present Topography
Central Ohio is divided by a west-facing Allegheny
Escarpment which extends northeast-southwest. The
Appalachian Plateau Province lies east of this escarpment, the Central Lowlands Province to the west. The
Appalachian Plateaus (which includes the Allegheny divisions discussed here) are highly dissected uplands with well-drained valleys that have terraced bottoms and narrow streams. The Glaciated Allegheny Plateau differs
.1. 50 from the Unglaclated Allegheny Plateau in that Its topography has been nsubdued by glacial erosion and deposition" (Smith 1966:7.12).
The Central Lowlands or till plains differ in elevation, topography, and drainage from the Plateaus, having a little-dissected surface with the major streams following the slope of the terrain southward to the Scioto
River. Valleys here are not as wide nor as deep as those of the Plateaus and tend to have narrow bottoms almost entirely occupied by stream beds. In general, these till plains have poorly developed drainage with some of the till plains drift nearly impermeable.
The lowlands are referred to as the till plains because they are the product of glacial deposition during the Pleistocene Epoch. Both the Illinoisian and
Wisconsin deposits described above are located within this region, with the Wisconsin deposits the most extensive.
Till is defined as a mixture of boulders, sands, silts, and clay that is neither sorted nor stratified and is thinnest in proximity to the glacial margin and thickest in some of the preglacial valleys where sand, gravel, silt, and clay layers are to be found. South of the glacial border, one may find outwash deposits and lake silts along the major present-day valleys.
The major watershed of central Ohio is that of the
Scioto River and its tributaries. The Scioto River flows through the heart of the region and drains an area of
16,883 kmbeing 378 km long. Two-thirds of this drainage is made up of till plains. The northern Scioto Valley consists of nearly level plain to gently rolling terrain with tributary streams in wide valleys. Part of this section is swampy. The easterly portion of the drainage flows through the Glaciated Allegheny Plateau where the topography is more rolling. To the south and southeast in the Unglaciated Allegheny Plateau, the drainage is comprised of steep slopes and rugged topography. South of present-day Columbus, the river flows in wide, preglacial valleys where valley width can reach as much as 2.4 km.
Soils
Major soil types will be provided for those areas of
Franklin and Ross Counties (McLoda and Parkinson 1980;
Petro et al. 1976) in which the sites of this study are located.
The following soil types are important for Franklin
County: Miamian silty clay loam, 6-12% slopes, eroded;
Eldean silt loam, 2-6% slopes; Celina silt loam, 2-6% slopes; Ockley silt loam 2-6% slopes, and Ockley silt loam
6-12% slopes, eroded. Miamian silty clay loam, 6-12% slopes, eroded, is found on sloping, well-drained areas on convex ridgetops, on side slopes above steeper areas, and along well-drained waterways on uplands. Some may be found on long, narrow areas on hillsides. The subsoil material 52
has a high clay content and more coarse fragments than the
surface layer, but due to erosion and cultivation the
subsoil and surface layers are now mixed. Most areas of
this soil type range from 1.6- 24.3 ha. It is
characterized by moderate available water capacity,
moderately low organic matter content, and medium natural
fertility.
Eldean silt loam of 2-6% slopes is a gently sloping,
deep, well-drained soil found on undulating areas on long,
narrow stream terraces and broad outwash plains. Some of
it may be found on low knolls. Slopes are predominantly 3-
5%, and most areas with this soil type range from 1.6-30.4
ha in size. Available water capacity is low or moderate,
while potential frost action, organic matter content, and
natural fertility are also moderate.
Celina silt loam with 2-6% slopes is a gently sloping,
deep, moderately well-drained soil in the uplands. Located on convex ridgetops, it may also be found on side slopes above steeper areas and along well-defined waterways. It ranges in size from areas of 0.8-20.2 ha. Available water capacity and organic matter are moderate, and the soil has
medium natural fertility.
Ockley silt loam of 2-6% slopes is a gently sloping, deep, well-drained soil in irregularly shaped areas on low knolls and in long narrow strips on stream terraces and
broad outwash plains. Occurrences of this soil type range in size from 1.6-20.2 ha. Available water capacity ranges from moderate to high, and organic matter content is moderate. It has medium natural fertility. Ockley silt loam of 6-12% slopes, eroded, is a sloping, deep, well- drained soil found in long and narrow areas on short slopes between terraces and floodplains or between different terrace levels. Most areas with this soil range in size from 1.2-10.1 ha. Available water capacity is moderate or high, while organic matter content is moderately low. This soil has medium natural fertility.
Soils of immediate concern in Ross County include those from the Ross, Genesee, Fox, Ockley, Warsaw, Cana,
Parke, and Eel series. Those not described above are summarized in the following paragraphs.
The Ross series in general consists of nearly level, deep, dark-colored, medium-textured and well-drained soils on second bottoms along major streams. The surface layer is silt loam, loam, fine sandy loam or light silty clay loam. These soils are stratified in the subsoil and are slightly acid to mildly alkaline with moderate permeability, high available water capacity, and high productivity. Flooding is an occasional hazard. Ross silt loam, of particular interest here, is found in large, broad uniformly shaped areas on floodplains along larger creeks and the Scioto River. Some areas of this soil are dissected by large sloughs and flood channels. 54
The Genesee series consist of deep, well-drained soils of medium texture in the surface layer and subsoil. They are found on nearly level bottomland. The surface layer is silt loam, loam, fine sandy loam, or silty clay loam.
These soils are subject to flooding, are moderately permeable, and high in available moisture capacity.
Genesee silt loam occurs in nearly level floodplains, generally in areas about 152 m wide and range from 10.1-
60.7 ha in size with a stream on one side. The surface layer contains few or no pebbles or stones. On the Scioto floodplain south of Chillicothe some areas are covered with gravelly outwash downstream of the larger meanders.
Genesee fine sandy loam is sandier than the general profile and is found on floodplains in nearly level areas about 76 m wide and 4-40 ha in size. Some areas are gently undulating and are dissected by many floodwater channels and sloughs. This soil tends to be droughty and is subject to flooding, especially in the dissected areas in winter or early spring.
Warsaw loam is a nearly level soil on small, irregularly shaped terrace areas. The series in general may be described as follows: deep, dark-colored, well- drained soils with medium textured surface layer and a moderately fine-textured subsoil underlain by stratified gravel and sand outwash. These soils are found on nearly level terraces along major streams. 55
The Fox series is represented by Fox gravelly loam of
0-2% slopes and Fox gravelly loam of 2-6% slopes. The series in general consists of nearly level to moderately steep, well-drained soils that are moderately deep to gravel and sand. They are mainly located on terraces along major streams.
Cana series soils, represented by Cana silt loam of
25-45% slopes, are deep and moderately well-drained. They have a medium-textured surface layer underlain by moderately fine and fine textured, very strongly acidic to extremely acidic subsoil derived from shale bedrock. They occur on sloping to gently sloping ridgetops and moderately sloping hillsides in central and southeastern parts of Ross
County. They are slowly permeable and have moderate available moisture capacity if not severely eroded. Cana silt loam of 25-45% slopes presently is about 50% wooded areas with the rest in pasture or reverting to woodland.
Parke silt loam of 12-18% slopes belongs to the Parke soil series, which is characterized as gently sloping to steep, well-drained soils with a silty surface layer and a moderately fine-textured subsoil. These soils are found on terraces in the southeast and extreme southwest parts of the county. Specifically, Parke silt loam of 12-18% slopes is a moderately steep soil found in irregularly shaped areas on hillsides and in fairly narrow bands on terraces adjacent to small streams and branches. The final soil type to be considered is an Eel silt loam of the Eel series. Such soils are described as deep, nearly level, medium-textured and moderately well-drained, located on bottomland in northwest Ross County and along the Scioto. These soils are moderately permeable, high in available moisture capacity, and subject to flooding. Eel silt loam is a nearly level soil found in long strips about
61 m wide on bottomland and in narrower bands along smaller streams.
Vegetation
Vegetation zones identified for the study area include
Bottomland or Hardwood Forest, Elm/Ash Swamp Forest, Beech
Forest, Nixed Oak Forest, Oak/Sugar Maple Forest, Mixed
Mesophytic Forest, and Prairie Grasslands. Following
Anderson and Ring (1976), the following are brief descriptions of these vegetational zones.
Beech Forest: This association Includes beech, sugar maple, red oak, white ash, white oak with scattered individuals of basswood, shagbark hickory, black cherry, and rarely cucumbertree (Magnolia acuminata). The most familiar types are beech-sugar maple and 'wet beech' on poorly drained flatlands. In the dissected Allegheny
Plateau one sees tuliptree, red maple and/or sugar maple associated with beech in the valleys.
Mixed Oak: This covers a wide variety of forest types.
The most widespread are the white oak-black oak-hickory 57 type and a white oak type. Anderson and King note that the term 'black oak' often included black oak, red oak, and scarlet oak as used by the first surveyors of the Allegheny
Plateau. 'Hickory' likewise often meant shagbark, bitternut, pignut, and mockernut (all species in the genus
Carya). Oak/Sugar Maple: These types include zero- mesophytic forests lacking beech, chestnut, red maple, and tuliptree. The dominants include white oak, red oak, black walnut, black maple or sugar maple, white ash, black cherry, red elm, basswood, and shagbark hickory. Local components might include such trees as bitternut, hackberry, blue ash, chinquapin oak, honey locust, and the
Kentucky coffeetree. Stony soils in high lime regions also supported Eastern red cedar, redbud, and Ohio buckeye.
Elm/Ash Swamp: The dominant trees of the canopy are found among white elm, black ash and/or white ash, and silver maple and/or red maple. Extremely wet areas supported cottonwood and/or sycamore. Better drained areas had bur oak/big shellbark hickory and red oak/basswood. 'Swamp oak/hickory' communities might include, locally, swamp white oak, pin oak, white oak, black walnut, and tuliptree.
Contiguous areas had 'wet beech' forests, wet prairies, sedge swamps, and fens.
Mixed Mesophytic: These types were dominated by broad leaved and deciduous species, but not exclusively so. No single species dominated. Within this type, one might see segregates such as oak-chestnut-tuliptreey oak-hickory-
tuliptree, or white oak-beech-maple. Southwestern Ohio
generally had more beech, white basswood, and tuliptree.
Braun (1964) suggests that this is the oldest and most
complex forest association of the Eastern Deciduous Forest.
Bottomland Hardwood; These forests are composed of
vegetation types of variable composition in the older
valleys and terraces of major streams, plus recent
alluvium. Three of the types appear to be climax
associations: l)beech/white oak, 2)beech/maple, and
3)beech/elm/ash/yellow buckeye. Other types include
elm/sycamore/river birch/red maple and sweetgum/river
birch. In a typical valley profile, black walnut/elm is
found on terraces, soft maple on terrace slopes,
sycamore/willow on the floodplain, water willow near the
watercourse, and oak/hickory and beech/sugar maple on till
plain blufftops.
Prairie Grasslands: Most of these are on wetlands and are
dominated by tall grasses like giant reedgrass, cord, or
slough grass, bluejoint, and/or big bluestem. Drier
prairies and borders are dominated by big and little
bluestem, switchgrass, and/or Indian grass.
Such vegetational reconstructions as those detailed here were frequently made with the use of GLO survey
records. As Wood (1976) notes, these surveys were made at a time when the climate was generally cooler than today. 59
However, the Silver Lake pollen sequence from Ohio (Graves
1977) indicates that the climate during the period in question may also have been cooler than normal. Thus, the vegetational associations and types described above will be accepted as reasonable facsimiles for the prehistoric environment in question.
Environmental Summary
Environmentally, the central Scioto drainage offered a rich variety of resources to prehistoric populations. The climate supported a varied seasonal mix of flora and fauna and could also support local and non-local cultigens such as sunflower and maize during the growing season. Soils in many areas were naturally fertile and would have supported a simple agricultural system as has been postulated for some prehistoric groups in the region.
Water resources were abundant in the form of rivers, streams, and springs. Bedrock geology, physiography, and topography combined to create a number of micro environmental zones (supporting a variety of vegetational complexes and therefore of wildlife as well) which could be easily exploited from a single site. Non-food resources such as firewood and raw materials for tools (wood, flints, cherts, bone, and shell) were also plentiful. Topographic conditions also provided relatively sheltered areas for cold-weather occupations. In short, in spite of sharp seasonal contrasts, food, shelter, and other raw materials were readily available within the region. The story of prehistoric human utilization of the central Scioto drainage is not one of a hand-to-mouth existence, but a more complex one tied to such intangible cultural factors as social relations within and between groups, processes of nucleation and sedentism, and a change in the human relationship to the natural ecosystem, where humans gradually created and maintained an artificial, agricultural ecosystem. The relationship of the human settlement pattern to the environment will be more closely examined in a chapter which deals with site catchment. Other questions which concern changes in the human cultural adaptation to the region will be addressed in the final chapter. CHAPTER IV
CHARACTERISTICS OF THE ARCHAEOLOGICAL RECORD: LATE WOODLAND AND LATE PREHISTORIC SITES
Introduction
The basic characteristics of the archaeological record
for Late Woodland and Late Prehistoric sites in the middle
Scioto drainage is summarized below. This discussion is
prefaced by a short synopsis of the history of archaeological research in this region, limited to Late
Woodland and Late Prehistoric sites. A general explanation of the nature of the site sample— how sites were chosen
for inclusion— follows. The specific locational,
environmental, and archaeological characteristics described
in the succeeding section include physiographic province, stream order, topography, soils, vegetation, site size, and number of temporal components per site.
The remaining portions of the chapter are devoted to more detailed summaries of sites which have undergone extensive survey and/or excavation. These include the Late
Woodland sites of Zencor, Water Plant, and Walter S. Cole, located in Franklin County, and the Late Prehistoric sites of Voss, also located in Franklin County, and Baum, Blain,
Gartner, and Kramer, found in Ross County.
61 62
The History of Archaeological Research in the Central Scioto Drainage
The history of archaeological research on Late
Woodland and Late Prehistoric sites in the central Scioto drainage reveals roughly five major periods of activity in the past one-hundred fifty years: Early Exploration (1847)-
1895t Ohio State Historical and Archaeological Society
(1895-1930), Griffin's Synthesis (1930-1945), Post-war
Salvage (1945-1975), and Ohio Historic Preservation Office
Surveys and CRM (1975-present). These periods are not based on formal periods of research, but on broadly similar approaches to the archaeological record revealed from an historical perspective.
Thus, the first period is characterized by the initial discovery and speculation on archaeological monuments in the area. The second period records the initiation of formal archaeological research by local and other museums in the region. The third period centers around the World
War II era with a retrenchment of research in the area and synthesis of previous data, dominated by the Ohio
Historical Society and James B. Griffin of the University of Michigan. The post-war era of archaeological research was also carried out by personnel from the Ohio Historical
Society and consists largely of salvage efforts as a series of developmental projects were undertaken for dams, reservoirs, and housing projects. The final period 63 encompasses a wider scope of work by institutions throughout Ohio, accompanied by efforts to bring Ohio archaeology back into the mainstream of modern archaeological methods and theory. Each of these periods is discussed in more detail in the following paragraphs.
The first period opened with a limited excavation of the Cramer Mound (33-R0-33) by Squier and Davis in 1846
(Squier and Davis 1847:64). The next work in the area was initiated by H.L. Reynolds for the Smithsonian Institution, consisting of the exploration of the square, truncated mound located at the Baum site (33-R0-4) from 1881-1890
(Mills 1906). Middleton continued this work for the
Smithsonian in 1891 (Mills 1906). From 1889-1891, surveys were conducted in the central Scioto drainage by W.K.
Moorehead for the Ohio State Archaeological and Historical
Society (Moorehead 1897).
Historical society personnel dominated the next 35 years of research in this area. In 1896, Moorehead excavated at the Higby Mound and Village site (33-R0-23)
(Moorehead 1897). The following year Clarence Loveberry
(under the direction of Moorehead) engaged in fieldwork at the Baum site, where two amateurs— L.M. Bean and A.B.
Coover— were involved in further exploration of the square, truncated Baum mound (Moorehead 1899). Loveberry apparently dug four mounds within an enclosure at the site, also. Two years later, in 1899, W.C. Mills continued work 64 on the Baum mound. Also associated with the Ohio State
Archaeological and Historical Society, Mills excavated the mound and the associated Baum Village until 1901 (Mills
1906). Gartner Mound (33-R0-19) was excavated in 1902 by
Mills, who subsequently excavated Gartner Village in 1903
(Mills 1904). This second period also saw Mills' identification of a post-Hopewell, pre-Fort Ancient
Intrusive Mound culture at Mound City (Mills 1922).
Continued work in the middle Scioto Valley largely focused upon Middle Woodland sites.
The third period, or Griffin Synthesis, is based on the efforts of a single individual— James B. Griffin, who undertook a synthesis of material known about the Fort
Ancient culture in general. This work began in the early
1930s and concluded with the publication of Griffin's monumental volume, The Fort Ancient Aspect, in 1943. New fieldwork was also begun during this period by the amateur crew of Hunter, Hunter, and Counts (1940) at Cramer Mound.
H. Holmes Ellis also conducted surveys in Ross County at this time for the Ohio State Archaeological and Historical
Society (Ullman 1985).
The Post-war Salvage period was ushered in with the excavation of the Walter S. Cole site by Raymond S. Baby of the Ohio Historical Society in 1947, under the direction of
Morgan (Potter 1966). Other major Late Woodland and Late
Prehistoric sites were also excavated during this period, 65 beginning with the Merion Village site (also called the
Zencor Village site and most recently the Scioto Trails
School site) by Baby and Mays in 1957 and 1958 (Potter
1966). The Voss mound was dug by Baby, Potter, and Mays in
1963 Baby, Potter, and Mays 1966), while the Enos Holmes mound was dug the same year by Baby, Potter, and Saurborn
(Baby, Potter, and Saurborn 1968). Voss Village was excavated in 1966 by the team of Baby, Potter, and Sawyer
(Baby, Potter, and Sawyer 1967). While Baby and his co workers appear to have dominated the scene during this post-war era, significant contributions were made by two other groups. In 1963, Prufer and Shane instituted the
Scioto Valley Archaeological Survey for the Case Institute of Technology (Prufer 1967). Their excavations at Blain
Village (33-R0-128) in 1966 and 1968 led to a revision of
Griffin's Fort Ancient foci and to the development of a new hypothesis for the origins of the Baum phase (Prufer and
Shane 1970). In addition, the Kramer mound and village site was subjected to limited excavation by a crew under the direction of Oriol Pi-Sunyer for the Case Institute of
Technology in 1967 (Ullman 1985).
The final period has been dominated by the Ohio
Historic Preservation Office, which instituted a series of grants to encourage regional survey programs, and by various CRM projects. From 1975 to the present such surveys and projects have resulted in additional knowledge of the portion of the archaeological record with which this study is concerned (Deaver 1980, Piotrowski 1980, and
MacMinn 1980 for 0HP0-6A in the late 1970s, Seeman 1980,
Clarke and Addington 1983, etc.). With the funding of regional surveys by the Ohio Historic Preservation Office in 1975, a different research approach to the archaeological record was adopted. These surveys are responsible for a large portion of the site distributional data utilized in this study. Several important Late
Woodland sites in Delaware County were excavated by Phagan during this period. These Include the Ufferman site (33 DL
12) in 1976 and 1978, the DECC0 site (33 DL 28) in 1977, and the periphery of the Walter S. Cole site also in 1978
(Phagan 1977, Barkes 1982). The most recent research, for example that of Dancey, Fricke, and Church at Water Plant in 1980, Otto at Zencor in 1981, and the analysis of ceramics from Kramer (Ullman 1985), is making significant new contributions to our understanding of the Late Woodland and Late Prehistoric periods of Scioto Valley archaeology.
The Nature of the Site Sample
The basic data for this research consists of a sample of 189 sites located in the central Scioto drainage
(portions of Delaware, Franklin, Pickaway, Ross, and
Highland Counties) which have a Late Woodland or Late
Prehistoric designation. These sites were culled from a statewide inventory of archaeological sites maintained by the State Historic Preservation Office in the Department of
Archaeology at the Ohio Historical Society. Each site is recorded in the inventory on a standard Ohio Archaeological
Inventory form (an OAI, as they are popularly known; Fig.
1).
The inventory grew out of response to the passage of historic and archaeological preservation legislation in the mid-1960s and 1 9 7 0 s ( building upon earlier attempts by the
Ohio Archaeological and Historical Society (now the Ohio
Historical Society) to catalogue all known archaeological sites in the state (Mills 1914; Dancey 1984, 1985). As noted in Dancey (1985), the OAI emphasizes locational and environmental information about sites, with only four of 33 items specifically concerned with archaeological characteristics of a site: site type, site dimensions, cultural assignment, and collected artifacts. No standardized definitions guide field descriptions of these dimensions; they are qualitative, intuitive designations made by field personnel. Locational information is included in the form of a xeroxed portion of a 7.5” topographic quadrangle on which the site is located.
Generally, maps of site plan, location of artifacts, and other such categories of data are lacking.
In some cases, sites identified before 1975 have been added to the inventory, but many of these early sites have not. Thus, no guarantee exists that the inventory is 68
Ml* Mm h I h ImI C— iicW OMo Hittoric Pioaom U oo M o o OHIO ARCHAEOLOGICAL INVENTORY C«Nm *M,OM c 4311I I. County I. Othot Nome* f t
I. Crty t Town Form
17. Floral Cover
20. Location of
II. U T M >1. Oroinapo tyaiam
n cvttwoi ctoMinuuon ot Tima Period
to copy portion at
Fig. 1. Copy of Ohio Archaeological Inventory Form. complete and that all known Late Woodland or Late
Prehistoric sites are Included in the sample. In addition
to noting that some Late Woodland and Late Prehistoric
sites from the region may not be included, it should be
emphasized that the set of sites used in the present study
was not randomly chosen and were not systematically
identified as part of a single, purposive survey. Many of
the recent additions to the OAI were recorded in the course
of CRM projects. However, the Ohio Historic Preservation
Office, as noted previously, administers a Survey and
Planning grant program which has allowed systematic surveys
along the major stream and river channels and adjacent
uplands (see Deaver 1980; MacMinn 1980; Piotrowski 1980;
Dancey, Wymer, and Waterworth 1986).
While lamenting the lack of robust archaeological
information contained in the OAIs and realizing that the
quality and quantity of recorded information varies with
the field archaeologist, it is important to note that the
inventory contains basic information that is useful for
certain kinds of analyses. The preponderance of geographic
and environmental data is extremely useful for settlement
pattern analysis (see chapter VI). The form routinely
includes information on topographic landform, soils,
vegetation, and stream order, along with a unified system
of general site location (7.5" topographic maps). Thus, a 70 great deal can be understood about a site's location In relation to other sites and environmental factors.
Sites Included In the sample are those indicated in the cultural period box on the OAIs as Late Woodland, Fort
Ancient, or Mississippian. Also included are sites which were not given a cultural designation, but which listed diagnostic Late Woodland and/or Late Prehistoric artifacts.
Additional OAI sites were included in the sample if the field archaeologist had documented the existence of diagnostic artifacts in private collections. Finally, some sites were added to the sample from site lists in archaeological reports, such as Prufer (1967), which have never been added to the OAI and from the older site card inventory.
Sites were recorded in the sample as single-component
Late Woodland, multi-component Late Woodland (with other components noted if available), single-component Late
Prehistoric, multi-component Late Prehistoric (as above), or as Late Woodland-Late Prehistoric. The last of these categories is necessary because some sites were assigned to a temporal period by the sole criteria of triangular
projectile points. As is discussed later (chapter V), some
controversy exists concerning the diagnostic precision of
triangular points (Seeman and Munson 1981; Baby, Potter, and Saurborn 1967). 71
The sites in the sample came from three types of investigation: amateur reports (n«3, 2%), intentional survey (n«155, 82%), and survey and excavation (n*31, 16%).
OAI site types— variously labelled as lithic clusters or scatters, camps, villages, workshops, or habitation sites— are based on recognition of a site on the basis of simple observable factors. Thus, 109 sites (58%) were recorded on the basis of a scatter of lithic debitage and/or tools, 54
(29%) were identified by the presence of both lithics and sherds on the surface, while 26 (13%) are known by the presence of lithics, sherds, possibly other types of artifacts, and/or features (mounds, hearths, or pits).
Thus, based on the OAIs and published and unpublished site reports, certain categories of information were recorded (where available) for each site included in the sample. First, each site was plotted on a map of the central Scioto drainage (Fig. 2). Next, the following characteristics were recorded: topographic landform location, soil type, vegetation, drainage system, site size, cultural classification, who did the work and when, descriptions of artifacts recovered, features identified, and any other information or pertinent comments.
Physiographic province was determined from mapped locations of sites. Coverage varied for each characteristic, ranging from 100% for physiographic province to 81% for site size.
More extensive information was gathered for those sites 72
1 W. S. Cole j— Utftrm>.R
DECCD
Zencor
i Water Plant
L__
Gartner
Baum lain
Enos Holmes I Harness-28 20 km Howard Baum
Fig. 2. Major Sites in the Central Scioto Valley. 73
which have been intensively surveyed and/or excavated, such as the Water Plant site (Dancey, Fricke, and Church 1987),
Blain (Prufer and Shane 1970), Voss (Baby, Potter, and
Sawyer 1967), Baum (Mills 1906), Gartner (Mills 1904), and
Zencor (Potter 1966, Otto 1982).
Based on the information gathered in this way, a
general characterization can be made of the archaeological
record for the period A.D. 500- A.D. 1250 in terms of
locational and environmental features. This general
picture is summarized in the following paragraphs. A more detailed comparison is made between Late Woodland and Late
Prehistoric sites in chapter VI for the settlement pattern analysis. Further, the greater detail of the better known sites allows for comparative analyses of artifacts, site structure, and site catchment to be done in later chapters.
General Characteristics of the Sample
As noted in the previous chapter, central Ohio is divided into three major physiographic provinces. Sites are nearly equally divided between the Glaciated Allegheny
Plateau (48%) and the Central Lowlands or Till Plains
(44%), with less than 10% found in the Unglaciated
Allegheny Plateau. By stream order or rank within the drainage, 57% of all sites are located directly on the
Scioto River, with 15% found along a minor tributary (small creek or intermittent stream) of the Scioto. An additional
18% of sites are found on a major tributary of the Scioto 74
(large creek or river), with 5% located on a creek or run branching from one of these major tributaries. Finally, 4% of all sites can be found on the smallest stream order, a small creek or intermittent stream feeder that runs into a tributary of a major creek or river other than the Scioto.
General topographic distinctions of valley and uplands were made. Valley encompasses any locational designation such as 'valley,' 'valley floor,' bottom, floodplain, floodplain terrace, or terrace. The term uplands includes slopes, 'blufftops,' 'bluff-edges,' hilltops, ridgetops, and 'uplands.' Specifically, floodplain is used here to indicate land immediately adjacent to a stream or river and terrace indicates a landform of higher elevation lying between the floodplain and uplands. A bluff location refers to a landform set off from the valley by a relatively abrupt or steep embankment. That is, a gentle slope does not equal a bluff. General uplands refers to all other possible upland locations, excluding bluffs as defined.
Based on this two-part topographic division of the region, 78% of all sites are located in a valley position and 20% are in the uplands (2% unknown). Within valleys, nearly half of all sites are found on a terrace with about one-third on the floodplain. The uplands show a tidy distribution of one-third bluff locations and two-thirds general uplands locations.
.i. 75
Other environmental features assessed include soils and vegetational zone. While 18 soil types were recorded
over-all, slightly more than half (51%) of the sample falls
on soils of the Fox-Ockley-Genesee-Ross association and approximately 20% on the Eldean-Warsaw-Genesee-Ross
(including Ockleys) association. The remaining 30% of the
sample was divided among the 16 other soil types, ranging
most frequently from 1-5 sites per type (one type had nine
sites). Again, nearly half (45%) of the sample was
associated with a Bottomlands Hardwood forest, 23% with an
Elm-Ash Swamp forest (not surprising, given 78% of all
sites are valley sites), while Beech forest and Mixed Oak
forests each account for 10% of the sample, Mixed
Mesophytic forest for 5%, and Prairie Boundaries for only
1% of all sites.
Only two archaeological characteristics were
summarized for the entire sample: site size and the nature
of and number of temporal components. Site size categories
were standardized in terms of hectares. Fully 47% of all
sites are Class 1 (< 1 hr ) in size, while 22% are Class 2
(>. 1 <. 2 ha). Classes 3, 4, and 5 (by 1 ha increments)
together account for less than 20% of all sites (about 6%
each), while Class 6 sites (>. 5 10 ha) account for 10% of
the sample. Only three sites (2%) are Class 7 (> 10 ha) in
size. In terms of temporal components per site, the largest percentage are those sites in the ambiguous Late Woodland-
Late Prehistoric category. That is, 30% of the sample consists of sites which are at least Late Woodland, possibly Late Prehistoric, or both. Nearly as many (29%) are single-component Late Woodland or Late Prehistoric sites. Two-component sites make up 18% of the sample; three-component ones equal 12%. Four- and five-component sites make up 5% each of all sites (all of these except one contain both a Late Woodland and a Late Prehistoric component).
Temporal placement of sites is augmented, sparingly, by C-14 dates available for a few sites (Table 2). Except for the dates from the Water Plant site, which were acquired through funding for this dissertation research, all were taken from published and unpublished reports.
Thus, while only nine sites within the drainage have been dated, the range of dates clearly spans the time period in question, with the Watet Plant dates an especially important addition, because they provide information about the early end of the temporal continuum.
All of the available dates have been presented, notwithstanding interpretations of their significance that may have been made in the context of individual site reports such as Blain (Prufer and Shane 1970) and Voss
(Baby, Potter, and Sawyer 1967). Many of the sites with 7 7
Tabic 2. Radiocarbon Datca for Central Scioto Drainage Sltea
Site Calendar Date Radiocarbon Date (BP) HASCA-Corrected
Blaln A.D. 970 1035+ 155 (OVl! 24S)1 A.D. 719/1049 A.D. 1040 970+ 220 (OVU 247b)* A.D. 721/1)8) A.D. 1225 760+ 100 ( M—1911 )jj A.D. 1057/1277 A.D. 1303 490+ 100 (H-1910) , A.D. 1335/1555 A.D. 1440 4057 150 (OVU 2479)* A.D. 1373/1693
DECCO A.D. 240 1710+ 50® A.D. 129/249 A.D. 250 17007 50® A.D. 139/259 A.D. 270 16807 45® A.D. 165/275 A.D. 370 15807 50® A.D. 263/383 A.D. 1130 620+ 45® A.D. 1050/1160 A.D. 1160 790+ 55' A.D. 1071/1201
Enos Holaes A.D. 1135 815+ 95 (OVl 27b)1 A.D. 1006/1216
Harness-26 A.D. 1175 775+ 90* A.D. 1052/1252 A.D. 1290 660+ 75 A.D. 1185/1355
Howard Baun A.D. 1240 710+ 50® A.D. 1159/1279
Morrison Village A.D. 1550 400+115 (OVl 1611® A.D. 1413/1663 A.D. 1660 2997110 (M-l760)2 A.I. 1541/1761 A.D. 1690 2667110 (H-1761)2 A.D. 1373/160C
Paint Creek Lake #7 A.D. 1390l A.D. 1610'
Voss K d . A.D. 970 970+79 (OVT 92A)® A.D. 862/1041 Voss Village A.D. 910 1040+215 (OVl 229)* A.D. 654/1104 A.D. 920 10307)20 (K 1675)*): A.D. 1070 860+100 (Y. 1862)“ A.D. 1120 8307100 (M 1861)*° A.D. 1135 8157220 (OVl' 2431* A.D. 881/1341 A.D. 1170 7807l00 Vater Plant A.D. 500 1450+80 (B-15506) A.D. 440/620 A.D. 620 1330+70 (B-15507) A.D. 570-590/700-72i A.D. 500 1450+70 (B-15506) A.D. 450/6)0 Zencor A.D. 620 1330+60’ A.D. 510/650 A.D. 650 1200+60’ A.D. 644/764 A.D. 780 1170+50’ A.D. 685/805 A.D. 1000 9507509 A.D. 911/1031 *Ogden and Hay 1960 2Crane and Griffin 1968 fBarkes 1982 tSeeaan 1981 'Skinner et al. 1981 °Ogden and Hay 1967 'Undated OAI for Highland County, on file Departaent of Archaeology, the Ohio Historical Society. 'Ogden and Hay 1965 lQRt 78 C-14 dates are among those sites which have been recorded in site reports. Some of these— Water Plant, Blain, and Voss— will be summarized in the following section, which also includes sites lacking C-14 dates such as Baum and Gartner. These sites have been identified by name on the site distribution map (Fig. 2). Ma 1or Site Descriptions In this section a general description is given of several excavated Late Woodland and Late Prehistoric sites. The summary of the Water Plant site is drawn directly from survey and excavation data and notes from 1980 fieldwork. Baum and Gartner are described on the basis of two published reports (Mills 1904, 1906) and the author's study of the collections from these sites. No field notes are available for these sites. Voss Village has been reported in an unpublished site report (Baby, Potter, and Sawyer 1967), while Blain is the subject of a published report (Prufer and Shane 1970). Zencor and the Walter S. Cole site are described in an unpublished master's thesis (Potter 1966) and Kramer in a recently published master's thesis (Ullman 1985). Late Woodland Sites The Water Plant site (33-FR-155) was investigated as part of the Big Walnut Survey in the spring, summer, and early autumn of 1980. The site is located on the bluff edge of the west side of Big Walnut Creek, a major tributary of the Scioto River. It is characterized by a high density of surface artifacts, the suggestion of a mound (33-FR-156), and the possible existence of an associated ditch/earthwork feature. This ditch/earthwork circumscribes an area of approximately 3 ha. Topographically, the site is located in the Glaciated Allegheny Plateau region, in an area where esker and moraine features remain from the Illinoisian and Wisconsin glacial advances. Elevation is 730 ft. above sea level. The site has been subject to annual cultivation for at least ten years prior to the 1980 investigation and was subject to amateur collecting over this same ten-year period. The Big Walnut Survey was the result of a Survey and Planning Grant- from 0HP0 for the survey of four sections on either side of the Big Walnut Creek near the Hamilton Meadows housing subdivision in southern Franklin County, W.S. Dancey of the Department of Anthropology, The Ohio State University, was project director; Mary Lou Fricke was field director. A considerable portion of field time was devoted to survey and limited excavation of the Water Plant site. This site had been reputed to contain artifacts from nearly every cultural period in Ohio prehistory and was threatened by ongoing construction of new facilities for the Parson's Water Treatment Plant. Thus, an attempt was made to 80 collect as much data as possible from this portion of the survey area. The Water Plant site was subjected to two Intensive surface collections within the area enclosed by traces of an arc-shaped ditch, the ends of which terminated at the bluff edge. Dual surface collections were used to control for the effects of collector activity, to test different sampling strategies, and to test the effects of differential exposure conditions. A 10 m^ grid system was developed for field provenience, with each 10 m identified by its location in meters north and south from a 0.0 point. Each of these 10 m^ grid squares was composed of sixteen 2.5 m^ sampling units numbered consecutively 1-16. A 25% sample fraction of each grid square was chosen for maximum coverage of the sampling universe, given limited resources of time, crew size, and money. Systematic coverage of 145 grid squares (71%) was achieved over a period of four weeks during the first surface collection. The second systematic surface collection was made on five successive days late in June following plowing contracted by the project in early June. In order to avoid collector bias between the date of the plowing and the onset of systematic collecting, three separate, controlled collections of diagnostic artifacts were made over the sampling universe. The result of the 81 systematic surface collection was an evenly spaced, 25% sample of the entire site. Supplementing the surface collection, limited excavation vas undertaken at the Water Plant site. One objective of the excavation was to verify the existence and extent of a prehistoric earthwork/ditch feature believed to be present at the site on the basis of aerial photographic anomalies and the exposure of a ditch profile revealed in an overflow trench excavated by the construction company at the north end of the site. Additionally, a series of eight test pits were excavated from east to west along the southern margins of the site to obtain information on plow zone artifact content and density as well as the subplow character of the archaeological deposit. These test pits were placed along the center line of a 20 m wide strip of land designated in the construction design as the location of a v-shaped emergency overflow ditch for the plant facility. Additional excavation was performed within this strip when the construction company was persuaded to remove the plow zone with heavy machinery (road grader and bulldozer) before completing the overflow ditch. Trowelling and shovel skimming of the exposed undisturbed surface in an area of approximately 20 x 70 m produced evidence of 109 cultural features, all of which were mapped and many of 82 which were excavated. While not part of the original Big Walnut Survey research design, this excavation added important information to our knowledge of the Water Plant site. Considerable lithic and ceramic collections were recovered at the Water Plant site, although very little bone was found. Also, no pipes, shell, or ornaments were found. Ceramics were recovered from both surface collection and feature contexts. In general, the ceramics reflect a Late Woodland date, being largely cordmarked with granitic temper. No shell tempering nor Late Prehistoric decorative motifs were present. Total lithics from all site samples equal 21,303 pieces of material: groundstone, N-109; projectile points, N-454; bifaces, N-93; and flakes, N-20,647. Most of the projectile points have been identified as belonging to the Chesser Notched type (Dancey, Fricke, and Church 1987). Contrary to initial reports by local artifact collectors, the surface collection and feature contexts show the Water Plant site to consist largely of a single-component, early Late Woodland settlement rather than a multi-component site representing all periods of central Ohio prehistory. The Walter S. Cole site (33-DL-ll) was the first site officially recognized as Late Woodland, or something new between Hopewell and Fort Ancient, in central Ohio. This site is known only from unpublished field notes and 83 descriptions given in an unpublished thesis (Potter 1966). The site is located on the south end of an esker on a terrace of the Olentangy river. The northern portion of the site was excavated by means of N-S and E-W trenches across the site. Typical midden cultural materials such as animal bones, few human bones, shell, projectile points, debitage, and potsherds were recovered. Nine refuse/storage pits which were oval in shape, with an irregular horizontal outline and an irregular, roughly U- shaped vertical profile, were discovered. Eight of these were grouped in a depression in the center of the excavated portion of the site, while the ninth was located on the eastern slope of the esker. The features contained dark fill which mostly yielded animal bones, shell, and potsherds. One adult female burial with a fetus was removed from a pit by the landowner prior to excavation of the site. Further test excavations were made to determine the extent of the site. Nothing was found to warrant extension of the original excavations. No structures were located at the site; nor were C-14 dates obtained. Pottery is basically grit-tempered Late Woodland— no shell temper, no guilloche or other Late Prehistoric decorative motifs, and no appendages of any kind. The ceramic assemblage has been analyzed in Barkes (1982) and summarized in Seeman (1980). Zencor (33-FR-8), another Late Woodland site, was first described by Potter (1966). More recent research at the site has been made public as well (Otto 1982). The site is located on a bluff east of the Scioto River. It is partially enclosed on three sides by a rough, semi-circular embankment approximately 274 m in length. Postholes forming at least three structures were found during the first excavations (1957-1958). These are circular, 7.6- 10.9 m in diameter with an overlapping wall serving as an entrance. One structure consists of two superimposed patterns. Numerous oval and circular pits were scattered in and near the structures. Most were filled with midden, although one contained four extended burials. Midden contained animal bones, mussel shells, debitage, broken and complete projectile points, and approximately 10,000 sherds. While this material was scattered over the entire site, denser concentrations occurred in the pits and near the structures. Further human burials have been recovered in more recent work at the site (Otto 1982). Late Prehistoric Sites Late Prehistoric sites that will be summarized here include Baum, Gartner, Blain, and Kramer in Ross County and Voss, which is located in Franklin County. Baum, located on the first gravel terrace of the South Fork of Paint Creek, was discovered in the 1800s by W.R. Keran and A.W. Stretcher, two local collectors. Paint 85 Creek Valley Is about 3.22 km vide at this point and is surrounded on the east and west by hills. The site was originally cleared for cultivation around 1830 by the grandfather of J.E. Baum (the property owner in 1906) and had been under continuous cultivation when Mills excavated there in 1899-1903. The original vegetation was noted to include such trees as black walnut, oak, sycamore, and ash. In 1897 the mound associated with the village was explored by Coover and Bean, two amateurs from Roxabell, Ohio, while a small portion of the village Itself was examined by Dr. Loveberry, under the direction of Warren K. Moorehead, then curator of archaeology at the Ohio Archaeological and Historical Society. During the summer of 1899, William C. Mills, who succeeded Moorehead as curator of archaeology, excavated those portions of the village lying south and north of the mound. His excavations continued the following summer in that portion of the village east of the mound, producing the "greatest finds" (Mills 1902:79). During the summer of 1902, Mills excavated the village area northeast of the mound, extending work along the edge of a gravel terrace southeast of the mound. The summer of 1903 was spent on that portion of the village to the east of the mound. The site is located 457.2 m south of Paint Creek and covers approximately 12.14 ha, with the greatest concentration of cultural material found within an area of 86 0.8-1.2 ha. Less than 0.81 ha or 6.6% of the site was excavated. Excavations proceeded through a series of trenches across the site. These indicated 30-90 cm of leafmold and alluvial deposits over a thin stratum of hardpan; directly beneath the hardpan was gravel. Mills (1902:78)), in quoting from Middleton's report in the Twelfth Annual Report of the Smithsonian Institution, noted that the site was subjected to annual floods, leaving only the mound summit above water, hence the alluvial deposits. Trenching revealed refuse to a depth of 76 cm beneath which was found a yellow clay of natural origins. Features identified at Baum Village include hearths, refuse, storage and burial pits, and structures. A total of 127 burials were recovered from the village, excluding the mound. Except for four adults recovered from pits and 17 infants in pits, all were extended burials grouped in close proximity to structures. The majority of the village burials contained some grave goods in the form of ordinary tools and/or ornaments. In general a great deal of ceramics, lithics, bone, and shell tools and ornaments were recovered. Pottery has been defined by Griffin (1943) as forming two types: Baum Cordmarked and Baum Plain. A variety of incised motifs are present, including curvilinear and rectilinear guilloches, and both shell and grit tempers are represented. Shell is present in less than 20% of the sample, making grit temper 87 the predominant form. Triangular points predominate in the lithic assemblage. Much floral and faunal material was also recovered, with pits yielding charred masses of corn, beans, hickory nuts, and walnuts. Faunal material included large and small herbivores, carnivores, numerous small mammals, as well as birds, fish, mussels, and turtles. Baum Mound is located almost in the center of the village, near the edge of the terrace to the west. It was first described by Squier and Davis (1847), who reported the following dimensions: 36.58 m at the base, 4.57 m in height, and 232.6 m^ at the top. Middleton (1890-91) wrote that it was composed of clay mottled with black loam and patches of a grayish, plastic lime. E-W and N-S trenches, 1.5-1.8 m wide, were excavated through the mound, then widened at the center of the mound where they intersected to create an area nearly 4 m diameter. Lateral digging was done from these trenches. The mound features included two stratified series of upright postholes, thin seams of fine sand, and seventeen burials. Of these, all were enclosed within the postholes, none were determined to be intrusive, and all except three lay on the sand seams. An oblong crematory pit was also located within the mound, as was a bed of ashes covered with large pebbles. Collections from the site consist in part of approximately 1517 bone tools, 152 pieces of animal jaws 88 and canines, 507 other animal bones, 652 pieces of worked and unworked shell, 164 miscellaneous stone artifacts, 504 pieces of worked flint, 4 pieces of mica, and 428 sherds. Boxes of charred corncobs, hickory nuts, walnuts, chestnuts, plum seeds, hazel nuts, papaw seeds, and butternuts are also curated at the Ohio Historical Museum. Gartner was excavated by W.C. Mills in the summer of 1903 (Mills 1904). It is about 8.4 ha in size and is located on the south side of the Scioto River about 90 m from the river and approximately 21 m above it. The total area excavated was about 0.24 ha or 15-20% of the site in a series of trenches to the north, south, east, and west of the mound located in the center of the village. The village midden ranged from a depth of 30.5-50.8 cm. Features recovered included refuse pits, structures, hearths, burials, and the mound. Like Baum, the refuse pits were distributed throughout the village, surrounding the structures and had similar contents. Hearths were located interior to and exterior to structures. Unlike the Baum report, no exact count of structures is given for Gartner; however, since Mills refers to structures in the plural, it is safe to assume that more than one was identified. One could probably assume that the number is close to that for Baum. According to the artifact inventory, approximately 2164 artifacts, excluding animal bone, were recovered from the 1902-1903 excavations. 89 These included bone and stone tools, sherds, and shell artifacts. Also, no indication is given of how many village burials were recovered, although groups were interspersed among the pits (visible in photographs). A crematory was located directly south of the mound. Grave goods were plentiful and on the order of those recovered with Baum burials. Artifact assemblages are similar in nature to those recovered at Baum. A total of 42 burials were found within the mound, which was located in the center of the village. The mound, when excavated, was 2.3 m in height and 23 m in diameter. It was constructed from soil from the village and pits (containing refuse and implements with sand and gravel mixed in). This mound was built in three distinct, overlapping sections (Mills 1904:130; Fig. 1). The first section contained cremations on a prepared floor and several extended burials, while the second and third sections contained no cremations. Most, including the cremations in section 1, contained grave goods. Voss Village was discovered during the excavation of the O.C. Voss mound in 1963 (Baby, Potter, and Mays 1966) and was excavated in the summer of 1966 (Baby, Potter, and Sawyer 1967). The site is located along the east side of Big Darby Creek, just south of the confluence of the Big and Little Darby Creeks. It was excavated in a series of 50 3-m^ blocks placed north of the mound. A total of 19 90 refuse pits, two structures, and two burials as well as numerous stone, shell, ceramic, and bone artifacts were recovered. Voss mound, dug by Baby, Potter, and Mays in 1963, contained four features, nine postholes, and seven burials (six contained in Feature 1 and one in Feature 3). Numerous artifacts were recovered. Like Baum and Gartner, the artifacts recovered in the mound were identical to those found in the village. Voss was described as a Late Woodland, Cole Complex habitation-ceremonial site (Baby and Potter 1965; Baby, Potter, and Sawyer 1967). A series of subrectangular structures (based on the two uncovered at the site) were described as surrounding a plaza which Was located beneath Voss mound. A C-14 date of A.D. 970 had been obtained from material in the mound. An absence of hearths interior or exterior to the structures was taken as further indication of the ceremonial nature of the structures, as this was similar to Hopewell charnel houses and dissimilar to structures from other Cole sites (Baby, Potter, and Sawyer 1967). The series of C-14 dates obtained from both village and mound (A.D. 910- A.D. 1563) were used to place the Cole Complex firmly in the Late Woodland period (the most recent dates were discarded as inappropriate). Blain was excavated by Prufer and Shane in 1968 (Prufer and Shane 1970). The site, which is about 3.2 ha in size, is located on the west bank of the Scioto River within the confines of the 660' contour, bounded on three sides by the river and the Paint Creek, their confluence forming the southernmost corner of the area. Forty-seven features were discovered, including a central burial mound (0.46 m high and 21 m diameter), plaza, three structures, pits, and small sheet middens. From this, the site structure is described as a ring of structures surrounding the central mound and plaza. Seven burials were recovered from the mounds. The artifact assemblage consists of 13,317 sherds, 7,738 pieces of chipped flint, 184 stone artifacts, 4 pipes, 257 bone artifacts, and 392 shell artifacts. The pottery is defined as Baum Cordmarked Incised, var. Blain. However, some Peters Cordmarked and Peters Plain sherds were part of the assemblage as well as some Woodland point types. Based in part on the pottery and in part on a series of C-14 dates which ranged from A.D. 970 to A.D. 1440, the site was designated as the earliest Baum phase site by Prufer and Shane (1970). Kramer was partially excavated in 1967 by Oriol Pi- Sunyer of the University of Massachusetts (Ullman 1985), surveyed by an Ohio Department of Transportation crew in 1982 (Addington and Clarke 1983), and has been subjected to two seasons of excavation in 1985 and 1986 by Seeman. The site is located on a terrace west of the Scioto River and is 8.8 ha in size. It has an abrupt terrace edge to the 92 east and a swale north of the truncated mound associated with the site. This mound Is located In the center of the village and was originally recorded as being 4.6 m in height and 30.5 m in diameter (Ullman 1985). Several skeletons were removed from the mound, while the artifacts recovered from it included grit and shell tempered ceramics, some with cordmarking, incised guilloches, and dentated collars. The 1967 excavations of the village by Oriol Pi-Sunyer opened seven 10' x 10' excavation units which yielded 25 features, among which were 14 deep bell-shaped, oval, circular, and irregularly shaped pits, 5 shallow, basin shaped pits, and 3 ashpits. Artifacts recovered from the village included triangular projectile points, worked bone, cut shell, bone and shell beads, shell hoes, bone fish hooks, faunal remains, some 11,000 potsherds, and some carbonized corn kernels, nuts, and seeds. Addington and Clarke (1983) described primary and secondary zones of cultural debris on the site surface. No evidence of structures was found during the 1967 excavations, nor were any C-14 dates obtained. Based on similarities between artifact assemblages, especially among ceramics, the site was assigned to the Baum phase (Ullman 1985). At present, no information is available from recent work at the site as to new feature discoveries (such as 93 possible structures), although it is known that C-14 samples have been submitted for dating (Otto 1986, pers. comm ■) • Summary and Evaluation of the Sample In this chapter the total data set of 189 sites has been characterized in terms of basic variables like site size, whether the sites are single-component or multi- component, location by landform, physiographic province, and drainage, and associated soil and vegetation types. Some general comparisons were made for the sample as a whole. While information concerning the archaeological record is not as complete as one would wish, the sample of sites has yielded enough information to allow more detailed comparisons of settlement patterns in the middle Scioto Valley to be made between the Late Woodland and Late Prehistoric periods (see chapter VI). However, the present study also points to a clear need for standardized recording of field information, an issue which has been addressed by the creation of a new, ten-page OAI form. While this form obviously exceeds the former in sheer volume, whether it will redress disparities in the description of characteristics of the archaeological record and paleoenvironment remains to be seen. Finally, site distributional data summarized from the OAI forms has been augmented by descriptions of some of the major sites of this data set, including the Water Plant 94 site, the Walter S. Cole site, Zencor, and Baum, Gartner, Blain, Kramer, and Voss villages. In the next chapters these sites will be compared in greater detail in terms of analyses of artifacts, features, site structure, and site catchment. Other sites will also be drawn from the sample in order to make fuller comparisons between sites of the Late Woodland and Late Prehistoric periods. Prior to a discussion of these analyses, however, the issues of general chronology and the validity of a culture historical explanation for the origins of the Late Prehistoric sites in the middle Scioto Valley will be addressed. v CHAPTER V CHRONOLOGY AND THE POPULATION INTRUSION HYPOTHESIS Introduction The previous chapters have discussed the two basic hypotheses put forth to explain the transition between Late Woodland and Late Prehistoric populations in the central Scioto Valley. The question of cultural intrusion versus local development will be addressed after a refined chronological ordering of Late Woodland and Late Prehistoric sites within the drainage is produced. This order will also be of help in dealing with the other questions addressed in later chapters. Underlying the basic opposition between population intrusion and local derivation as an explanation for culture change is a more pervasive chronological problem growing out of the history of archaeological work in this area. Chronological sequences were first developed based on seriation of sites using stylistic attributes of lithics and ceramics. Advances in dating techniques provided a basis for new chronological orders using radiocarbon dates. Such sequences conflicted with earlier ones. Evaluation of the relative and absolute chronologies has been complicated by several factors. Primary among 95 96 these has been a tendency for explanations of cultural change to be developed and put forth without testing against the archaeological data. Objective evaluation of the chronology of Late Woodland and Late Prehistoric development has been prevented because of set ideas on the sequence and causes of cultural change through time and across space in the central Scioto Valley. To address this problem, the ordering of the Baum phase sites will be discussed next, followed by a section on the methodology employed in the re-ordering of sites, and a discussion of several potentially useful chronological measures: ceramic temper, design elements, and rim and lip morphology; projectile point base shape; and radiocarbon dates. Ordering the Baum Phase Sites: An Illustration of the Chronological Problem Problems inherent in the chronology developed for the Late Prehistoric Baum phase, particularly as illustrated in the most recent chronological refinements of this phase by Prufer and Shane (1970), are the focus of the following discussion. As noted in chapter I, originally the Late Prehistoric sites in the central Scioto drainage were felt to occur earlier in time than Hopewell, which was perceived as more culturally advanced (Putnam 1886; Moorehead 1897; Mills 1906). The issue of chronology within Fort Ancient was, however, first addressed by Griffin (1943). 97 Griffin distinguished various Fort Ancient foci (now phases), based on similarity in traits to other, known Late Prehistoric groups. Baum and Anderson foci, for example, seemed to represent the fusion of a cultural tradition which had many elements also occurring in Woodland sites with early diffused Middle Mississippi traits. Early stages of this might be represented by the Baldwin Component or the Brush Creek Component at Serpent Mound (Griffin 1943:307-8). Other foci, such as Madisonville, can be placed later in time because "many of the elements which entered into the Madisonville focus came or seemed to come from cultural groups that either entered the historic period or represented the last known prehistoric group in the various areas" (Griffin 1943:307). Thus, the Baum focus (containing Baum and Gartner villages) is placed in time through comparison with other foci. Griffin's statement that the Baldwin Component and the Brush Creek Components may represent early stages of the Baum Focus, then, implies an order for Baum Focus sites from early to later as follows: Baldwin Component sites, Brush Creek Component, Baum Component sites. Tet, he notes at the site of Baum the following Late Woodland traits: triangular point with wide base, cremation, and the presence of a structure in the Baum mound (1943:65-7). Clearly, Griffin recognizes that even in the latest sites of the focus local Late Woodland traits persisted. 98 The next attempt at chronologically ordering the sites present within the various Ohio Fort Ancient foci came in the work of Prufer and Shane (1970). With their excavation of Blain Village, Prufer and Shane made use of radiocarbon dates and seriation to construct a spatiotemporal sequence for the Late Prehistoric in this area. Using the Blain Village data as a starting point, Prufer and Shane suggested that internal evidence supports an interpretation of the village occupation as homogeneous [cf Schambach 1971]. Five radiocarbon dates ranging from A.D. 970 to A.D. 1440 were obtained from several features. Evaluating these dates in terms of contamination, stratigraphic considerations, and the explicit assumption that over time the percentage of shell-tempering increased in Late Prehistoric ceramics, they viewed the A.D. 970 date as most representative of Plain's temporal position. Using amount of shell-temper to produce a chronological sequence, Prufer and Shane seriated sites based on the following frequencies of shell-tempering for early Late Prehistoric sites in central Ohio: 1) Brush Creek, 0%, 2) Graham village, 2.7%, 3) Blain village, 3.3%, 4) Baldwin village, less than 10%, 5) Baum and Gartner, about 20%. Thus, a relative chronology was established which differed from and expanded Griffin's tentative sequence in that it explicitly placed Brush Creek before Baldwin and separated the two sites with the addition of 99 Graham Village and Blain Village. Baum and Gartner remain at the end of the sequence. Next, Prufer and Shane discuss the regional phases of the early Late Prehistoric period in terms of their chronological implications. For example, they present the Brush Creek Component, the predecessor to Anderson, as the temporal equivalent of Baum (Prufer and Shane 1970:238). Furthermore, "in the last analysis Baum and Anderson are 'genetically' related in the sense that both derive from the same ultimate source" (Prufer and Shane 1970:238). Here, too, Griffin's comments on the Woodland traits of the Brush Creek and Baldwin ceramics are noted, with the conclusion that some of the roots of early Fort Ancient are to be found in local Late Woodland cultures. They write: "those assemblages which exhibit the greatest resemblance to Late Woodland materials are probably the earliest in the Fort Ancient sequence" (ibid.). Again, the sites are ordered approximately as above with Blain seen as contemporaneous or slightly later than Voss, Graham, Baldwin, and Brush Creek based on the percentage of shell- tempering present and its occurrence with well-developed Fort Ancient rim and decorative treatments. Baum and Gartner fit at the end of the sequence. Thus, the absolute and relative chronologies appear to support one another. Regarding the origins of the Late Prehistoric culture, Prufer and Shane conclude that it is difficult to explain as .1.. a gradual modification of Late Woodland by slow acculturation resulting from gradual direct diffusion or the stimulus diffusion of new traits and ideas from outside sources. Their alternative explanation is that local Late Woodland populations were displaced by strong Mississippian population units intruding into major valleys such as the Scioto, with regional variability in Late Prehistoric phases now explained by degree of Mississlppification. Baum Phase was most directly affected and Baldwin, Voss, and Brush Creek less so. In other words, the degree of Late Woodland affinities is described as a function of geographic and environmental considerations, such that "the farther such phases are removed from the centrally located Scioto Valley, the less completely 'Mississlppified' they are" (Prufer and Shane 1970:260). The "significant numbers of more or less modified Woodland traits persisting into early Fort Ancient" (1970:261) found at Voss, Baldwin, and Brush Creek are explained by the fact "that the acculturation process of phases originated from the Scioto Valley, i.e., from the Baum P h a s e . Such dates as are presently available, support this hypothesis" (1970:262). As indicated, the presence and percentage of shell- tempering has been used by Prufer and Shane to chronologically order early Late Prehistoric sites. However, they also use this trait— which they feel indicates Mississippian contact— to infer degrees of 101 cultural contact between populations. Baldwin, Brush Creek, Blain, and the other sites with very low or no frequency of shell-temper are determined to be less influenced by Mississippian cultural contact. The higher frequencies at Baum and Gartner, on the other hand, indicate direct Mississippian influence. Low frequencies at the other sites are explained as the result of acculturation from centers of direct population intrusion such as Baum and Gartner. The implication of this analysis is that Baum and Gartner must be coeval with or earlier than Voss, Brush Creek, Blain, or Baldwin. The original chronological order of these sites produced by their seriation based on percentages of shell-tempering has been altered in addressing the questions of the origins of Fort Ancient. Two chronological orders exist in Prufer and Shane's work, both supported with available radiocarbon dates. They obviously conflict with one another. Part of the problem may result from the amount of space included by Prufer and Shane in their analysis. They have seriated sites across all of southern and central Ohio, an area which encompasses at least three separate drainages— Brush Creek, the Scioto River, and the Hocking River. Recent work (Cowan et al. 1985; Carskadden 1985; Murphy 1975) suggests that successful chronological sequences can be developed for Individual drainages. It is 102 possible that the conflict apparent in Prufer and Shane's work could be eliminated if sites in the Brush Creek and Hocking river drainages were treated independently. This is not to infer that cultural ties or interaction did not cross-cut drainages, but that our initial unit of analysis should be more restricted. Methodology The objective of the following sections is to determine whether central Scioto Valley sites can be more precisely ordered in time. Will this new order also yield a new perspective on the question of population Intrusion, as expressed by Prufer and Shane, versus local derivation? Obviously, seriation in and of itself is not sufficient to test the validity of these hypotheses. It will produce a relative chronology. Evaluation of that chronological sequence, however, must be made independently. The attempt to resolve the problem of chronology for the central Scioto Valley during the Late Woodland-Late Prehistoric transitional period will focus on data from a select group of sites drawn from the over-all sample. This group consists of the Water Plant, DECCO, Zencor, Voss, Gartner, Kramer, Blain, Baum, Howard Baum, and Enos Holmes sites. They were selected because each provided certain categories of data which could be used to seriate the sites. These categories of data include the ceramic attributes of tempering material, design element, rim form 103 and lip shape, plus base shape of triangular points. Radiocarbon dates are also discussed. The comparability of the sites as temporally homogeneous units will be addressed first. This is followed by a discussion of each chronological measure. The nature and validity of these measures will be explored individually and conclude with the application of each measure to the data base. A composite order, based on all the measures, will be produced. With the establishment of this order, it will then be possible to evaluate the population intrusion versus local derivation question for the central Scioto drainage. Temporal Homogeneity A factor which must be considered in any determination of a chronological order for the sites is the means by which they were identified as homogeneous, single component sites. The cultural homogeneity of an assemblage refers to the fact that it can be characterized as being predominantly the result of the occupation of a site by a single cultural group. Heterogeneous, or mixed assemblages would be the result of two or more cultural groups dominating the archaeological assemblage at a site. Thus, it is important to examine the determination of homogeneity for some of the better known sites of the sample. The Water Plant site yielded ceramics from features which are homogeneous stylistically and conform to 104 definitions of Late Woodland types. Surface collections from across the site also reflect this predominance of Late Woodland cultural traits. Radiocarbon dates from the features support a single component, early Late Woodland assignment. Zencor, located near Water Plant on the Scioto River in Franklin County (Fig. 2), is presently undergoing excavation by Martha Potter Otto of the Ohio Historical Society. This site, much like Water Plant, has produced evidence of pits, structures, and artifacts which are all typically Late Woodland in character (Otto 1982). No earlier Woodland or Late Prehistoric components are present. Carbon-14 dates support its assignment as a single-component Late Woodland site. The Howard Baum site is located on Paint Creek in southern Ross County. It was tested as part of a CRM project, providing pits, lithics, ceramics, and a radiocarbon date of 710 + 50 years: A.D. 1240 (Skinner et al. 1981) which confirm it as a single component site that has a number of Late Prehistoric characteristics. Baum and Gartner villages are described as homogeneous by Mills (1906, 1904), who noted that artifacts within the mounds were similar in all respects to those of the villages, from all levels within the mound and village features. Furthermore, descriptions of artifacts suggest that they are reflected village-wide in terms of 105 distribution, both in pits and structures. Mills, in his summaries of artifact classes, indicated that some classes of artifacts may be limited to representation in some feature classes and not others, but within that class they seem to have been distributed across the site. Mills also indicated that patterning existed in pottery distribution within a site, in the form of discernible stylistic differentiation between groups of burials associated with different structures. Unfortunately, this report cannot be tested as 1) specific item-by-item provenience is lacking, and 2) village structures and burials are not mapped. However, Griffin's pottery analysis indicated a relatively homogeneous style. Also, while hearths, pits, and structures show evidence of being rebuilt, artifacts remain the same from one end of the village to the other throughout the pits, structures, and burials. The large number of pits, structures, hearths, and burials examined suggests that an interpretation of homogeneity is sound for Baum and Gartner villages. Prufer and Shane (1970) consider Blain Village to be a single-component site based on over-all homogeneity of artifacts and the fact that the kind and distribution of structures and features form a cohesive settlement plan. Some Late Woodland-like ceramics were recovered, but these are explained as a result of the very early dates of Blain. Schambach (1971), however, has observed that less than 2% of the site was excavated and that Feature 2, which yielded the date of A.D. 970 (uncorrected), also yielded the Late Woodland ceramics. The relationship of this feature to the other features is obscured by its absence from published site maps. Schambach also observes that undecorated cordmarked, grit-tempered sherds are identified as Baum Cordmarked rather than as Woodland (Schambach 1971:1404). While Late Prehistoric ceramics and other artifacts indisputably predominate at the site, Schambach offers these observations to question the interpretation by Prufer and Shane (1970) that the A.D. 970 date best fits this Late Prehistoric material instead of some of the other, later dates obtained from this site. Voss Village could be subjected to the same criticisms. The portion of the site which has been excavated produced a relatively homogeneous artifact distribution; no ambiguous artifact concentrations have been identified. However, the mixture of traits present at the site has caused it to be identified as both Late Woodland (Baby, Potter, and Sawyer 1967) and Late Prehistoric (Prufer and Shane 1970). Given the consistency of artifact distribution, the site will be considered single-component. The possibility that a Late Woodland component underlies a Late Prehistoric component is one possibility which might ultimately explain the confusion 107 over its temporal placement if more of the site is excavated in the future. DECCO, located in the far north of the Scioto drainage on the Olentangy river in Delaware County, has produced artifacts and radiocarbon dates which support both a Middle Woodland and Late Woodland component at the site (Phagan 1977). It is an intriguing site based on its size, location, and radiocarbon dates which place it as late as A.D. 1160 (Phagan 1977). Enos Holmes is located at the opposite end of the drainage on Paint Creek in the extreme northeastern corner of Highland County where it borders Ross County. This mound site has been assigned, like Voss Village, to both Late Woodland (Baby, Potter, and Saurborn 1968) and to Late Prehistoric (Rigg 1977). Its temporal assignment will be tested since it has produced a quantity of ceramics and lithics which are consistent from all features in the mound and which provide the kind of data outlined above. The final site to be considered is Kramer, located almost directly west of Gartner on the Scioto River in Ross County. The portion of the site which was excavated by Oriol Pi-Sunyer (Ullman 1985) points to a single-component Late Prehistoric site that is very similar to Baum and Gartner in types of artifacts and features present. Thus, the basic temporal homogeneity of each site is accepted. The assemblages are sufficient and comparable in 108 that all provide data on lithic and ceramic attributes which will be used to construct frequency seriations. In addition, data pertinent to other analyses— functional use-wear, site structure, and site catchment— are also available. Ceramic Attributes in Chronology: Shell-Temper Ceramics are one of two major diagnostic elements (the other is triangular points) used to establish chronology between Late Woodland and Late Prehistoric sites (Rafferty 1974; Ullman 1985; Barkes 1982; Griffin 1943; Prufer and Shane 1970; etc.). Temper and attributes of surface finish, such as design motifs and smoothing or cordmarking, are the most commonly utilized factors in the determination of pottery types as Late Woodland or Late Prehistoric. In this section the attributes of shel1-temper, surface finish— specifically the guilloche design element, and rim and lip form, will be discussed. Shell-temper refers to the use of crushed shell to temper raw clay in the production of ceramic wares. It first appears in archaeological contexts in small ratios where other temper forms (grit, grog, sand, and chert) predominate. In the lower Mississippi Valley, for example, radiocarbon dates indicate that the Mississippian period began approximately A.D. 700 with the Coles Creek Phase (Morse and Morse 1983). Tempering materials such as grog, grit, and limestone precede the appearance of shell- 109 tempering, which is not present until A.D. 800-850 when the Coles Creek phase is fully developed. Another example, from the northern portion of the American Bottoms, is the Dohack Phase (A.D. 800-850) which represents a transitional phase between the Late Woodland Patrick Phase (A.D. 600-800) and the Emergent Mississippian Range phase (A.D. 850-900) (Kelly et al. 1984). Its tempering frequencies of 84% limestone, 3% grog, 1% grit, and 12% mixtures of these contrast with the predominantly grit and grog tempers of the preceding phase, although neither has shell-temper. It has been suggested that the shift to limestone temper... represents the initiation of a local ceramic tradition in which limestone is used as a tempering agent well into the initial phases of Mississippian culture. This localized pattern contrasts with the predominant use of grit and grog tempering and the eventual predominance of shell tempering by the Emergent Mississippian and the early Mississippian populations (Kelly et al. 1984:130). This is seen in the Merrell Phase, which is dated to A.D. 900-950 (Kelly et al. 1984). Tempering in Merrell Phase ceramics consists of grit, grog, or a combination of these in 93% of jars. Limestone and shell-temper are both present in very low percentages, making up the remaining 7% of temper material in ceramics. However, by the Edelhardt Phase (A.D. 950-1000), the last Emergent Mississippian phase in the same region, various temper types are present. Grit and grog predominant at AO.8% in jars in the straight, unmodified category, while shell temper now predominates (52.3%) in jars with flared rims. "Differences in tempering, along with surface treatment and stylistic attributes, form the basis for the spatial segregation of phases within the Emergent Mississippian period. Such patterns persist into the Mississippian phases for the area" (Kelly et al. 1984:157). To summarize, shell-temper first appears in the lower Mississippi Valley. There may be technological reasons, rather than purely stylistic ones, for its appearance (Morse and Morse 1983). Prior to the Coles Creek phase, shell-temper was present in very low frequencies if at all in Late Woodland phases. It appears consistently in Mississippian contexts. In the more northerly portions of the Mississippi Valley, shell-temper does not appear to be found in any Late Woodland phases. In phases transitional between Late Woodland and Mississippian the percentage of shell-temper goes from 0% in the Dohack phase to more than 50% in certain ceramic categories of the Edelhardt phase. Shell-temper predominates in the fully Mississippian phases. Thus, shell-temper does appear to be a chronologically sensitive marker, with full-blown Mississippian sites having high frequencies of the trait. Late Woodland sites or those which are transitional to Mississippian (Late Prehistoric) either have no shell- Ill temper or very small amounts (less than 7% in the Merrell phase, for example). In regards to the central Scioto Valley, shell- tempered ceramics are present at Enos Holmes, Baum, Blain, Kramer, Gartner, and Voss sites. They are not found at Water Plant, Zencor, Howard Baum, or DECCO. When sites are seriated by the frequency of temper types (Table 3), three groups result: those which lack shell-temper, those which have less than 5%, and those which have 15% or more. The presence and amount of shell-temper does appear to be a chronologically sensitive attribute for differentiating sites in the Scioto drainage. The significance of these amounts will be evaluated in conjunction with the results of the other seriations. Ceramic Attributes in Chronology: Guilloche The presence of the curvilinear guilloche (the rectilinear guilloche is present in such a minuscule proportion that it has been added to the curvilinear element; hereafter it is referred to as the guilloche) has been used as a diagnostic Late Prehistoric trait. Like shell-temper, it has also been used to support cultural ties between Fort Ancient and Mississippian groups, a point to be considered later. Ordering sites by guilloche frequencies versus those for other designs, an order is produced which again results in three groups (Table 4). The first group has no 1 1 2 Table 3. Seriation of Sites by Temper Frequency Site Shell Grit Limestone Chert Other Water Plant 0 100 0 0 0 Zencor 0 100 0 0 0 Howard Baum 0 80 12 8 0 DECCO 0 99 0 1 0 Voss 0.2 99.8 0 0 0 Enos Holmes 0.6 99.A 0 0 0 Blain 3.3 96.1 0. A 0 0.2 Baum 15 80 5 0 0 Gartner 20 80 0 0 0 Kramer 29 69 0 0 2 113 Table 4. Ordering of Sites by Percentage of Guilloche Site Guilloche Other Designs Water Plant 0 0 Zencor 0 0 DECCO 0 100 Kramer 22 78 Baum 44 56 Gartner 50 50 Howard Baum 52 48 Voss 63 37 Enos Holmes 68 32 Blain 85 15 114 guilloche. The second contains the highest frequencies of guilloche (63-85%), while the third is characterized by the lowest frequencies of guilloche (22-52%) for sites in which it is present. These groups are similar to those produced by the seriation using temper. Ceramic Attributes in Chronology; Rim and Lip Form The other ceramic attributes for which there was sufficient data for analysis are rim and lip form. Both appear to be temporally sensitive attributes (Tables 5 and 6). For rim forms, rims were categorized as either thickened (includes appliqued, collared, and folded rims) or unthickened. When sites are ordered on this basis, three groups of sites are produced: one clearly Late Woodland, one Late Prehistoric, and one differentiated between these two with consistently higher frequencies of thickened rims. Water Plant, however, can be securely placed at the beginning of any sequence because of its C-14 dates; thus, it is out of place in this order. Lip shape, on the other hand, when defined as either flattened or rounded, results in two groups, one in which flattened lips predominate (78-100%) and one in which rounded lips are dominant (60-90%). To some degree, then, these ceramic attributes seem to be temporally-sensitive categories of data. Before comparing the orders produced with the use of ceramic attributes, one final category of data must be examined for its potential use in seriation. 115 Table 5. Ordering of Sites by Percentages of Rim Form Site Thickened Unthickened Unknown Voss 78 10 12 DECCO 67 33 0 Blain 61 39 0 Howard Baum 52 44 0 Water Plant 30 57 13 Kramer 26 62 12 Baum^ <25 75 0 Enos Holmes 21 79 0 Gartner1 20 80 0 ^Figures estimated from Griffin (1943). 116 Table 6. Ordering of Sites by Frequencies of Lip Shape Site Flattened Rounded/Other Zencor 100 0 DECCO 78 22 Blain 40 60 Baum* 40 60 Howard Baum 32 68 Enos Holmes 31 69 Gartner* 20 80 Voss 10 90 ^Figures estimated from Griffin (1943). 117 Triangular Points as Chronological Markers Triangular projectiles have long been a point of controversy for those who attempt to differentiate Late Woodland from Late Prehistoric sites. Attempts to use the triangular point as a diagnostic element of Late Woodland versus Late Prehistoric sites are critiqued in this section. These attempts are basically of two kinds, one which tries to differentiate points on the basis of metric measurements and one which seriates sites based on such attributes as base shape. To test both models, data on seven metric variables (Fig. 3) for side lengths, central axis length, base width, length from base to thickest part of the point, and weight were collected for the present sample of 333 triangular points from Baum, Gartner, Enos Holmes, and several miscellaneous small sites. Base shape, categorized as concave, convex, or straight, was also recorded. As proponents of the first approach, Baby, Potter, and Saurborn (1968) give descriptions and measurements for both Late Woodland and Late Prehistoric triangular points. Out of a sample of 18 triangular points, 5 were defined as Late Prehistoric, being "long and slender, with flaring sides and straight or convex bases " (1968:11). Late Woodland triangles (Cole), n-13, were described as "short and stubby with relatively thick cross-sections" (ibid.) and straight, convex, or concave bases. Mean values were given for 118 Key: 1 side length la side length lb central axis length 2 th ick n ess 3 31 from base to thickest point base length Fig. 3. Triangular Point Measurements. 119 length, base width, and thickness for both types. It was suggested that triangles from Voss Village showed a similar division into these two types. Table 7 indicates that values for both types fall well within the range of values for the present sample, as do Griffin's (1943) small and large groups of triangles. Prufer (1967) and Prufer and Shane (1970) provide two more samples of triangular point measurements from the Late Woodland Chesser Cave site and from Blain. Prufer provides the range and mean values for length, width, and thickness of concave-, convex-, and straight-based triangles. Tables 7 and 8 illustrate that when the same categories are compared to the present sample, no bimodal distribution of values is evident between the Late Woodland and Late Prehistoric sites. Prufer's values do not differ significantly from these (Table 8), leading to the conclusion that his sample size may have suggested the proposed differences between categories. In a more recent effort, Seeman and Munson (1981) use discriminant analysis, based on seven measurements and four non-metric attributes described for a sample of 319 points, to separate Late Woodland from Late Prehistoric triangular points. Points from sites of known dates were utilized as controls, with a success rate of about 81% directly discriminated. Only 49% of points of unknown affiliation were assigned in a separate test. This paper, however, Table 7. Comparison of Triangular Point Measurements: Late Woodland/Late Prehistoric. Baby, Potter, Griffin Prufer & Shane Church & Saurborn (1968) (1943) (1970) .(1986) n=18 n=l 36 n=l 79 n=333 Range/Mean Range/Mean Range/Mean Range/Mean Fort Ancient Length: 34.3-47.1/38.7 17-47mm/25mm(Sm) 16-45/NA 17.15-91.5/38.3 Width: 14.5-19.1/17.0 NA NA 6.60-40.0/16.5 Thickness: 3.7-5.8/4.34 3.0-6.0/NA 3.0-9.0/NA 2.50-13.2/5.7 Late Woodland Length: 27.0-32.3/29.2 NA/45-47mm(Lg) (see Table 4.5) None Width: 13.0-20.6/17.3 NA None Thickness:3.3-7.3/5.1 NA None Table 8. Comparison of Triangular Point Measurements Grouped By Base-Shape Concave Convex Straight Site Range/Mean Range/Mean Range/Mean Chesser A (n-79) Length:18-49/32.3 21-59/29.4 28-46/32mm Width: 17-32/22.8 14-25/19.2 14-31/22.7 Thick: 3-7/3.95 2-7/3.6 2-8/4.4 Enos Holmes (n»18) Length:24.5-48.5/38.5 None 26.35-80.2/40.4 Width: 13.6-29.0/19.0 None 12.15-27.0/16.9 Thick: 4.8-7.55/6.2 None 3.7-10.0/5.4 Gartner (n=59) Length:24.35-71.6/41.5 21.35-52.0/32.7 27.15-80.85/38.5 Width: 8.0-25.0/17.0 11.25-23.25/15.6 8.65-32.55/14.6 Thick: 3.45-8.8/5.7 3.8-10.5/6.0 3.45-9.45/5.1 Baum (n-231) Length:24.7-82.2/42.8 19.65-72.7/33.8 17.15-91.5/34.9 Width: 6.6-40.0/16.9 10.7-22.15/15.0 8.8-33.0/15.3 Thick: 3.0-13.2/6.3 3.7-10.0/5.6 2.5-9.75/5.3 122 does not describe the specific variables or values of variables which serve to discriminate the points. Thus, it is impossible to test the results with additional data. The hypothesis that triangular points can be separated into Late Woodland and Late Prehistoric types by such measurements is not supported by the larger sample available here. One might reasonably conclude that 1) an even larger sample is necessary, with more triangles from single-component Late Woodland sites included, or 2) the sample is sufficiently large to demonstrate that Late Woodland triangles cannot be differentiated from Late Prehistoric ones on the basis of such measurements alone, or 3) a single type exists which does not discernibly change over time in terms of these measurements. In the second approach both Rafferty (1974) and Graybill (1981) utilize the triangular point as a diagnostic artifact type. Rafferty seriates sites by its presence and frequency in comparison with other point types. Graybill (1981) seriates triangular points by base shape and illustrates that convex-based points decrease over time, concave-based points increase over time, and straight-based points increase then decrease within the same time span. For the sample given here, sites were seriated by frequencies of triangular point bases to compare the results with Graybill's findings for a Late Prehistoric 123 sequence of sites in West Virginia. Seriating only sites from the central Scioto drainage produces the order shown in Table 9. Enos Holmes was not included because of its small sample size (n«9). Thus, Graybill's results are . mirrored in the central Scioto Valley. Base shape, then, appears to be one attribute of triangular points which does reflect temporal change. Absolute Chronological Data for the Central Scioto Valiev Absolute chronological data is available in the form of a series of radiocarbon dates (Table 2). These have been obtained for most of the sites discussed in this section, excluding Baum, Gartner, and Kramer. In addition, several other sites have been dated, including Harness-28, Morrison Village, and Paint Creek Lake #7. A general discussion of the dates indicates which are generally accepted and which are questionable. Where dates may be used confidently, they provide additional support for a revised chronological sequence, but it is emphasized that as of the present, radiocarbon dates from central Scioto Valley sites are not sufficient by themselves to establish a secure chronological sequence. Far from producing an unambiguous time sequence, radiocarbon dates have added to the controversial interpretation of Late Woodland versus Late Prehistoric cultures in the study area. The Late Woodland Cole Complex, for example, was originally described on the basis 124 Table 9. Seriation of Sites by Triangular Point Bases Site Convex Concave Straight Total n Z n Z n Z n Z Voss 11 58 1 5 7 37 19• 3 Blain 88 49 2.4 13 67 37 179 33 Gartner 17 29 17 29 25 42 59 11 Kramer 12 19 14 23 36 58 62 11 Baum 28 12 120 52 83 36 231 42 Total 550 100 125 of such sites as Voss, Enos Holmes, and DECCO which have yielded C-14 dates that place them within the range of time conventionally recognized as early Late Prehistoric. This has led Seeman (1980) and Baby and Potter (1965) to suggest contemporaneity between the Cole Complex to the north and Fort Ancient groups to the south within the same drainage. The two apparently existed side by side for a hundred or two hundred years, exchanging little more than a few pottery traits. Voss and Blain have produced an array of dates which span the Late Woodland- Late Prehistoric transition. While Baby and Potter selected the earliest dates from Voss as most representative and called the site Late Woodland, Prufer and Shane likewise selected the earliest dates from Blain, which are nearly identical to those from Voss, as most representative of their site's character, but called it Late Prehistoric. They also suggested that Voss was Late Prehistoric as well, which would effectively end the suggested contemporaneity of Late Woodland- Late Prehistoric cultures in the area. Two other sites with anomalous radiocarbon dates are DECCO and Harness-28. DECCO's dates clearly indicate two occupations at the site. The first, dating to A.D. 129-383 (MASCA-corrected), is not in question. The second, dated to A.D. 1050-1201 (MASCA-corrected), is of considerable interest because once again a Late Woodland assemblage is associated with a date that fits into the timespan when Late Prehistoric sites are known to have existed in the valley. Unlike Blain and Voss, excavations were not extensive at this site, and the assemblage is correspondingly small. Harness-28 has two published dates which also place it in this transitional period. Given that a site report is not available, evaluation of the dates is not possible. Radiocarbon dates from both sites remain questionable and cannot be used as clearcut evidence of absolute temporal placement within a chronological sequence for the Scioto Valley. The Water Plant and Zencor sites have produced a series of C-14 dates that place them consistently as Late Woodland, which agrees with the nature of their artifact assemblages. Of interest here is the fact that the Zencor dates are slightly later than those from Water Plant and overlap the earliest dates from Voss and Blain. The final sites which have radiocarbon dates are Howard Baum, Morrison Village, and Paint Creek Lake #7. The date from Howard Baum can be used more confidently because it was published as part of a comprehensive site report (Skinner et al. 1981). Morrison Village and Paint Creek Lake #7 are included even though their dates place them beyond the immediate range of time discussed here. Their relatively late dates are Middle Fort Ancient, 127 according to Prufer and Shane (1970). Yet( these dates overlap the upper end of the sequence of dates from several of the earlier Scioto Valley sites such as Blain, Harness- 28, and Voss. They point to an unbroken, continued local development of Late Prehistoric cultures in the central Scioto Valley. With this picture of development in mind, the following section orders the sites into a new chronological sequence. A. New Chronological Order for Late Woodland- Late Prehistoric Sites A re-ordering of the sites utilizing the attributes of shell-temper, the guilloche design element, rim and lip shape, and base shape of triangular points produced three to four chronological groups of sites. Radiocarbon dates are secure for Water Plant and Zencor, supporting their position as consistently earlier than the other sites and anchoring the early end of the sequences. In Table 10 a four-part summary chronological order has been established: Early Late Woodland, Late Late Woodland, Transitional Late Prehistoric, and Early Late Prehistoric. Characteristics of each are summarized in Table 10. Limiting the sites to those within the central Scioto drainage still encompasses approximately 100 km N-S and includes sites from Enos Holmes on the south fork of Paint Creek to DECCO on the Olentangy River. Both sites fit into an over-al'l chronology. 1 2 8 Table 10. Composite Chronology of Sites Period Sites Description Early Late Woodland Water Plant Ceramics 100% grit-teraper; no guilloche Zencor unthickened rim & flattened lip dominant; Chesser-type points major type; C-14 dates cluster A.D. 500-700. Late Late Woodland DECCO Ceramics predominantly grit-temper, no shell-temper; no guilloche; thick rims & flat lips dominant; Chesser-type points also dominant. Transitional Late Prehistoric Howard Baum Ceramics < 5% shell-temper; > 50% Voss design elements guilloche; thick rims; Enos Holmes lips flat or round; Convex-base Blain triangular points predominant. C-14 dates overlap Late Woodland-Late Prehistoric. Early Late Prehistoric Baum Ceramics v/>15% shell-temper; < 50% Gartner design elements guilloche; thick rims Kramer < 25%; Concave- and straight-base triangular points dominant. Enos Holmes consistently ranks with sites such as Howard Baum and Blain, which are relatively close geographically, and with Voss, which is in the northern portion of the study area. In three of four seriations, DECCO ranks with the Early Late Woodland sites; however, it differs from these in terms of different tempering, decorative elements, rim form, and radiocarbon dates. While it shares some characteristics with Transitional Late Prehistoric sites, such as presence of limestone temper, rim form, and radiocarbon dates, DECCO clearly lacks shell- temper and guilloche. Thus, it makes up the sole occupant of the Late Late Woodland class. Additional work in the area is needed to evaluate the position of this period. The Late Prehistoric period is divided into two groups of sites— Transitional and Early Late Prehistoric. Those sites included in the first period share specific chronological measures with the second group of sites such as shell-temper, guilloche, and a predominance of triangular projectile points. They differ considerably in the frequency of occurrence of each measure. Given this consistent difference, it was felt that a two-part division of Late Prehistoric sites could be made. The position of the Howard Baum site was a toss-up between Late Late Woodland and Transitional Late Prehistoric, based on its lack of shell-temper. However, two factors weighed in its placement into the latter 130 period. First was the fact that all other ceramic attributes placed it within this period, as did its lithic assemblage. The C-14 date from Howard Baum also supports such a position. The very high percentage of guilloche particularly supports its placement as Late Prehistoric. While no shell-tempered sherds were recovered, it should be noted that the site was tested as part of a CRM project where large-scale excavations were not possible. The lack of shell-temper was not considered sufficient to place this site as unequivocally Late Woodland. Baum, Gartner, and Kramer consistently were ordered together in all seriations. There seems to be no doubt that they rank at the most recent end of the chronological sequence. In the final chapters of this work, it will be seen whether other data such as site structure and site catchment also support this chronology. Before turning to these topics, the Prufer and Shane (1970) population intrusion hypothesis is evaluated in light of the analyses in the preceding sections of this chapter. Evaluating the Prufer and Shane Model of Culture Change This evaluation centers upon the question of whether or not the evidence supports a view of population intrusion into the central Scioto Valley as an explanation for the transition from Late Woodland to Late Prehistoric cultures. An examination of early attempts to establish a chronology for this area has illustrated the contradictory nature of 131 such efforts, even when separate attempts are made by the same individuals (Prufer and Shane 1970). The major sites have been determined to be more or less homogeneous. That is, most places have a major occupation that is either Late Woodland or Late Prehistoric, but many have a suggestion of Late Woodland when they have been claimed as pure Late Prehistoric. Accepting that such sites are predominantly Late Prehistoric, the implication is that Late Woodland groups were around and using localities like Voss and Blain prior to the establishment of Late Prehistoric groups at these sites. Nothing in the assemblages supports site-unit intrusion, or the replacement of one culture by another, as an explanation for the establishment of these latter groups in the region (Smith 1984). Some specific sites raise questions about the population intrusion model. Harness-28 is described as a Late Woodland site (Seeman 1980) with published C-14 dates of A.D. 1175 and A.D. 1270 (uncorrected). Located on the Scioto floodplain, it poses a puzzle for the Prufer and Shane hypothesis if it is indeed Late Woodland. According to them, Mississippian groups intruded into the central Scioto area by A.D. 900, displacing local Late Woodland populations to the hinterlands where their old way of life continued. How, then, can a single-component Late Woodland site on the Scioto floodplain be explained two centuries or 132 more after Late Woodland groups were supposed to have been scattered to the hills? Either the site represents an exception to the rule or Late Woodland populations were not displaced according to the Prufer-Shane scenario. Neither do ceramic attributes support the idea of population intrusion. Based on the Prufer-Shane model, the expected relationship between Mississippian and central Ohio Late Prehistoric ceramics does not hold. Shell-temper is first documented in the northern Mississippi Valley between A.D. 900-950. By A.D. 1000 shell-temper is present in up to 52.3% of certain ceramic classes in sites of this region, but in the central Scioto Valley during the same time (Transitional Late Prehistoric period), it appears in very low (less than 1%) frequencies. Even sites considered fully Late Prehistoric in this study— Baum, Gartner, and Kramer— do not approach such frequencies of shell-temper. If populations from the Mississippi Valley migrated or intruded or invaded the central Scioto drainage, why does such a disparity exist in the frequency of temper? While the use of shell-temper may have moved up the Mississippi Valley from south to north, through migration or diffusion, its appearance in the study area is coincident with, not later than, its first appearance in the middle Mississippi Valley. Parallel development can explain its presence in both areas. Further research into the technological 133 ramifications of shell-tempering for ceramics may provide additional support for this idea. Likewise, attempts to establish a link between Mississippian and Scioto Valley Late Prehistoric populations by tracing design elements from the former to the latter have failed (Rafferty 1974; Rigg 1977). Rigg found several design motifs similar to the guilloche in the lower Yazoo Basin of the Mississippi Valley (A.D. 1550), north of Memphis, Tennessee (A.D. 1625), in the lower Mississippi Valley (A.D. 1440-1675), and at Cahokia (A.D. 1050-1300, A.D. 1250-1500). None of these are identical to the guilloche found in the central Scioto Valley, and all occur later than the Transitional Late Prehistoric period in which this design is well-established. It seems clear that this is a local trait. Graybill (1981) suggests that the guilloche decreases over time as shell-tempering increases in the Late Prehistoric period in West Virginia. Such a correlation is supported by this study. Taken as a group, the ceramic traits used to establish a chronology for Late Woodland and Late Prehistoric sites of the Scioto Valley put to rest suppositions of contemporaneity for these groups postulated by Prufer and Shane (1970), Seeman (1980), Barkes (1982), and others. Recognizing the basic differences and similarities between the Late Woodland Cole Complex and Late Prehistoric ceramics, the confusion is resolved by the temporal order given here. Sites such as 134 DECCO, Voss, Blain, and Howard Baum fit between Early Late Woodland and Early Late Prehistoric, indicating that Late Prehistoric populations were derived from local antecedents. This viewpoint is sustained by the data on triangular points. . Rafferty (1974), in an effort to trace possible antecedents of the Fort Ancient complex of traits, notes that the triangular point appears in the central Ohio area in assemblages which predate both the Mississippian tradition and Fort Ancient. This has been supported by data from the Scioto Valley. Sites from this area have been successfully seriated using the base shape of triangular points. Furthermore, I suggest that the inability to seriate sites based on a description of triangular points by the range and mean values of length, width, and thickness is because this combination of attributes is not temporally sensitive. This also indicates continuity between the Late Woodland and early Late Prehistoric sites in the central Scioto drainage. Temporally sensitive categories of data based on ceramics (tempering materials, design elements, rim form, and lip shape) and on triangular points (base-shape) have been used to develop a new chronology for the central Scioto Valley that spans the time period from early Late Woodland to early Late Prehistoric. No evidence is present to support the idea that any of these elements, 135 singly or as a complex, was introduced into this region due to population intrusion from the Mississippi Valley. The best possible explanation is that Late Prehistoric populations in the central Scioto drainage were derived from local Late Woodland populations. Support for such a position is not limited to a comparison of ceramic and lithic attributes. In the following chapters the course of this local development is explored through several avenues of inquiry. This exploration is initiated with a comparison of Late Woodland and Late Prehistoric settlement pattern. Do the two periods have radically dissimilar patterns of settlement within the Scioto Valley? An analysis of functional classes of artifacts should reveal whether similar patterns of activity characterized Late Woodland and Late Prehistoric populations. Further analyses of site structure and site catchment will enrich our understanding of this crucial time period. It would have been much easier to account for the transition from Late Woodland to Late Prehistoric if indeed the population-intrusion hypothesis had been supported. In that case, the explanation would have been simple— different people, different culture. Now, the picture becomes much more complex. Is it possible to reach some understanding of how and why cultural change occurred as it did in the central Scioto Valley between A.D. 500- A.D. 1200? CHAPTER VI SETTLEMENT PATTERN COMPARISON Introduction Having established that a sound basis exists for the derivation of Scioto Valley Late Prehistoric populations from local Late Woodland groups, the second major problem of this research is addressed in this and subsequent chapters. In what ways do Late Woodland and Late Prehistoric sites compare? Do the elements of settlement patterning provide evidence of local development? Preliminary analysis of the site sample (chapter IV) has shown that sufficient data are available for several locational variables to make a settlement pattern analysis possible. These variables are systematically analyzed to obtain a fuller picture of how the distribution of Late Woodland sites compares to that of Late Prehistoric ones. Does an association exist between time period and location of sites? What variables seem to influence location? In the actual analysis a three-way comparison was necessary because all sites could not be confidently assigned to either a Late Woodland or a Late Prehistoric category. As discussed previously, an ambiguous group of sites exist which do not yield sufficient material to allow 136 137 a more precise temporal discrimination than that they could be Late Woodland and/or Late Prehistoric. These are labelled Late Woodland-Late Prehistoric (LW-LP) sites. This category of sites was included in the analysis because it was felt that an evaluation of the results might indicate whether these sites were more similar to Late Woodland, Late Prehistoric, or might in fact fit squarely between the two. To address the questions asked above, site distribution in regards to physiographic province, stream order, soils, vegetation, and topographic location was determined for Late Woodland and Late Prehistoric sites. The variables of site size and the nature of and number of temporal components were also recorded. These characteristics are presented in a series of graphs and tables. While differences in value of double or triple percentages seem to indicate a significant difference, the Standard Error of Proportion (SEp) indicates the variability of these estimates. R. X jC G-tests were performed where warranted (Appendix A). To explore which factors may have been influencing site location, a series of cross-tabulations were done between certain variables. These were limited to landform and stream order, site size and landform, and site size and soil group. It was felt that these were sufficient to establish whether any such associations existed. The 138 results were once more evaluated with the SEp and an R X £ G-test. This chapter will proceed with a description of over all site distribution drawn from compiled site maps for Late Woodland, Late Woodland-Late Prehistoric, and Late Prehistoric categories. The implications of these distributions are examined in more detail, beginning with a general description and discussion of each of the settlement pattern variables. This is followed by a comparison of these variables for each site category and ends with the summary and conclusions. General Site Distribution Not surprisingly, most sites are found along the Scioto and its major tributaries (Fig. 4). However, when the sites of a single period are examined, it is apparent that differences exist between the Late Woodland and Late Prehistoric. Clusters of Late Woodland sites occur in the mid-north area of the drainage (northern and middle Pickaway County and northern Ross County). If this distribution is compared to that of the Late Woodland-Late Prehistoric sites, more often than not, the two cluster. More of the Late Woodland-Late Prehistoric sites, however, are found in Franklin County, where they do not cluster. In Pickaway and Ross Counties they are found on the Scioto and its major tributaries. 139 Key; A lp A LW & LP • LW 0 LW &/or LP 20 Km Fig. 4. Late Woodland and Late Prehistoric Site Distribution in the Central Scioto Valley. When Late Prehistoric sites are added to the picture, in almost every case they occur in the same places as Late Woodland and Late Woodland-Late Prehistoric sites. Late Prehistoric sites also cluster with unambiguous multi- component sites which clearly have Late Woodland and Late Prehistoric components. The largest clusters occur in Ross County along the Scioto and on Paint Creek. Some clusters of these multi-component sites and Late Prehistoric sites occur where no Late Woodland sites have been recorded. The details of these distributions are examined variable by variable in the following sections. Explanations for these distributions will be explored in the final chapter. Environmental Variables The first variable to be considered is that of physiographic province, which refers to broad environmental zones. These have been differentiated into three major divisions: Unglaciated Allegheny Plateau, Glaciated Allegheny Plateau, and Till Plains. As each of these have been previously described in chapter III, a simple characterization will suffice here. Unglaciated areas occur primarily in the eastern and southern portions of the region and consist of upland, hilly areas. The glaciated plateau area refers to uplands that were subject to the effects of Pleistocene glaciations. The Till Plains encompass the western portions of the area. 141 Looking at data for all sites within the time period indicates (chapter IV) that about 90% of the sites are found in either the Till Plains or the Glaciated Allegheny Plateau with 10% in the Unglaciated Allegheny Plateau. It is informative to examine this distribution in more detail. What kind of sites make up that 10%? Are they all Late Woodland sites? Are they very small, specialized activity sites or very large, multi-purpose ones? Stream order or rank is the next variable to be scrutinized. The population-intrusion model suggests that intruding groups settled in major river valleys like the Scioto because of an agricultural subsistence base. Where are Late Prehistoric sites situated within the drainage? Are they spaced primarily along the broad Scioto Valley and its major tributaries? Does a relationship exist between stream order and landform location? Are Late Prehistoric sites found mostly on the valley floor or terraces of major streams? How does the distribution of Late Prehistoric sites compare to Late Woodland sites? If it is similar, for example, what does this say about Late Woodland and Late Prehistoric economies in terms of the population- intrusion hypothesis? An arbitrary ranking was established beginning with the the Scioto River (Table 11) as the first order stream. The second order streams are those minor creeks and intermittent streams branching directly from the Scioto. 142 Table 11. Stream Order Ranking for Central Scioto Drainage Stream Order Named Streams in Sample First Scioto River Second Dry Run Hargus Creek Lick Run Peters Run Van Meter Run Third Big Darby Creek Big Walnut Creek Deer Creek Little Walnut Creek Olentangy River Paint Creek Fourth North Fork Paint Creek Alum Creek Sugar Run Hellbranch Run Blacklick Creek Fifth Compton's Creek Herrod's Creek Mud Run (Little Walnut) Mud Run (N. Fork Paint Crk) Sulphur Lick Creek 143 Third order streams are major creeks and rivers tributary to the main trunk, and fourth order streams are those which branch from the 3rd order ones. Finally, the fifth order streams are the minor creeks, runs, and intermittent streams which branch from fourth order streams. Consideration of the over-all data set reveals that slightly more than half (57%) of all sites are found on the Scioto, with an additional one-quarter located on a second or third order stream. On first glance, this might seem to support the idea of population-intrusion into the region because the hypothesis predicts most sites will be found on the main trunk and its major tributaries. Thus, it is important to evaluate Late Woodland and Late Prehistoric sites separately for this variable. Similarly, an analysis of where sites are to be found in terms of specific landforms can suggest possible reasons for such location. A set of general landform types was utilized to describe topographic differences within the central Scioto drainage. Given the uncertainty of quality of landform assessment on the OAIs, broad differences were designated, rather than precise topographic divisions. These include valley floor (encompassing descriptions like bottom, floodplain, and terrace which combines all terrace levels), bluff (a bank rising from the edge of the valley floor), slope, and uplands. A slope is any gradient other 144 than a bluff. Uplands refer to all flat areas, hills, and ridges above the level of the valley floor and terraces. What is the pattern of occupation for Late Woodland sites compared to Late Prehistoric sites? Are most Late Prehistoric sites found in the valley? Are any Late Woodland sites found there? A cursory examination of all sites reveals that 74% are located on a terrace or valley floor. However, 26% are located away from the valley. Are these sites all Late Woodland? What kind of sites are found in the valleys compared to the uplands? Another variable which was examined was proximity of sites to vegetation zones. These zones are described in chapter III, but can be broadly summarized as those occurring on valley floors and terraces: Bottomland Hardwood forests, Elm-Ash forests; those found on slopes: Beech forests and Mixed Oaks; and those on uplands: Mixed Oaks, Oak-Sugar Maple groups, and so forth. Two-thirds of all sites are associated with vegetation associations found on valley floors or terraces, with one-third on the remaining upland types. Does this pattern equally characterize Late Woodland and Late Prehistoric sites? One final environmental variable— soil type— was considered. This variable should, of course, closely correlate with landform as the nature of landforms, bedrock geology, and other factors (weathering, vegetation) determine the type of soil which forms in an area. Soils 145 are of particular importance when the possibility of an agricultural subsistence system, based on maize, is raised. While a number of soil types have been identified for sites in the sample (chapter III), these have been lumped into four categories for comparative purposes which are broadly divided into valley and upland soil types: Fox- Ockley-Genesee-Ross soils and Eldean-Warsaw-Ockley-Ross soils as valley and Miamian and Other as upland. Evaluating over-all site distribution, about one-half are located on Fox-Ockley-Genesee-Ross soils with the remainder nearly equally divided among the other three categories. It would be expected that the proportion of sites on valley soils would dramatically increase from Late Woodland to Late Prehistoric, assuming a drastic change in subsistence to full-time maize agriculture. Archaeological Variables The final two variables analyzed are those dealing with parameters of site size and number of temporal components present at sites. Do sites exist with both Late Woodland and Late Prehistoric components? One superceding the other in time? Or, acculturation of one by the other? How does the percentage of such sites compare to the percentage of single-component Late Woodland and Late Prehistoric sites? Is the percentage of Late Woodland sites greater or lesser than Late Prehistoric sites? What does this indicate about the pattern of settlement in the 146 central Scioto area over time, especially when this variable is correlated with site size and environmental variables? As was noted earlier (chapter IV), one-third of all sites are clearly single-component sites and nearly one-third are ambiguous Late Woodland- Late Prehistoric sites, which leaves an additional one-third as multi- component containing both Late Woodland and Late Prehistoric components. Site size is an important variable to consider at this point because it too may indicate something about the nature of settlements and why they are found in one place and not another. Size is frequently assumed to be related to such factors as the kinds and numbers of activities carried out at a site, population size, and duration of occupation. For the present study seven site classes were defined, the first five in 1 ha increments up to 5 ha. Given the limited percentage of sites greater than or equal to 5 ha, a decision was made to divide such sites into two classes: greater than or equal to 5 ha and less than 10 ha and greater than or equal to 10 ha. Also, because it is not possible, barring systematic site surveys and/or excavation, to delimit size parameters for Late Woodland and Late Prehistoric components of a single site, such sites were excluded from this analysis. The possible relationship of site size to the function of a settlement suggests further lines of inquiry. Are 147 sites randomly distributed among all site size classes or are they clustered into certain ones? Does size correlate with other variables such as soil types, landform location, or stream order? Nearly half of the sites taken from the OAIs are Class 1 or less than 1 ha in size, while at the other end of the spectrum about 10% are Class 6 or 7, greater than or equal to 5 ha. Sites appear to cluster at either end of the size classes, and, therefore, are not randomly distributed. Is the distribution different for Late Woodland versus Late Prehistoric sites? These and other questions will be explored in the following section. Comparing Late Woodland and Late Prehistoric Site Distributions The results which follow characterize the nature of the change in settlement pattern within the central Scioto drainage over time. In regards to physiographic province, the location of Late Woodland sites is fairly equally divided between the Glaciated Allegheny Plateau and the Till Plains. Six percent are found in the unglaciated portions of the region. By the Late Prehistoric period, settlement has increased in the Glaciated Allegheny Plateau to 61% with Till Plains decreasing to 25% of sites. Surprisingly, sites have increased to over 10% in the Unglaciated Allegheny Plateau as well (Table 12). The category Late Woodland-Late Prehistoric is more similar to Late Woodland sites when the proportions are 148 Table 12. Late Woodland/Late Prehistoric Sites by Physiographic Province Province LW LW-LP LP Totals n % SEp n % SEp n % SE n % P Unglaciated 4 6+.054 3 5+.056 8 14+.092 15 8 Glaciated 33 46+.118 23 38+.124 35 61+.129 91 48 Till Plains 34 48+. 119 35 57+.127 14 25+. 114 83 44 Totals 71 100 61 100 57 100 189 100 G/q-15.079; X2(.01)(9)-14 .684. Table 13. Late Woodland/Late Prehistoric Sites by Stream Order Rank LW LW-LP LP Totals n % SEp n % SEp n % SEp n % 1st 40 56+.117 31 58+.137 33 59+.131 104 57 2nd 14 19+.093 6 11+.087 8 14+.094 28 15 3rd 13 18+.091 11 21+.Ill 8 14+.094 32 18 4th 2 3+.039 2 4+.052 5 9+.076 9 5 5th 3 4+.047 3 6+.063 2 4+.050 8 4 Totals 72 100 53 100 56 100 181 100 G/q®4.718; 149 compared to SEp. The only group of sites which stands out is the proportion of Late Prehistoric sites located in the Unglaciated Allegheny Plateau. When this difference is evaluated using an R X C G-test, it indicates the association is significant at the 0.01 level. Slightly more Late Prehistoric sites than expected occur in this province. Locational patterns have also been examined in terms of stream order (Table 13). Based on the scheme outlined earlier for ranking streams within the drainage, it is apparent that the proportion of sites on the Scioto increases by 3% from Late Woodland to Late Prehistoric. The percentage of sites on second and third order streams decreases by 4-5%, and the percentage of sites on fifth order streams remains the same. The only increase in percentages greater than 5% occurs on fourth order streams. The variables of stream order and time period show no association when the data are statistically compared, and Late Woodland-Late Prehistoric sites are not significantly different from either of the other categories. Meanwhile, how are sites distributed by landform when Late Woodland sites are compared to Late Prehistoric? The percentage of valley sites (valley floor and terrace) increases by 20% from Late Woodland to Late Prehistoric (Table 14), with approximately 75% of all sites found in the valley. Within the valley category, valley floor 150 Table 14. Landform Locations of Sites Landform LW LW-LP LP Totals n Z SEp n Z SEp n Z SEp n Z Valley floor 30 41+.116 25 52+.144 25 41+.126 80 44 Terrace 18 25+.102 11 23+.121 25 41+.126 54 30 Bluff 10 14+.082 3 6+.070 6 10+.076 19 10 Slope 2 3+.039 1 2+.041 0 0 3 2 Uplands 12 17+.088 8 12+.108 5 8+.070 25 14 Totals 72 100 48 100 61 100 181 100 G/q=l4.468 ; X2<#x(g)=13.362. 151 settlement remains the same proportionately with a 20% increase occurring on terraces. Upland settlements appear to decrease by about 50% from Late Woodland to Late Prehistoric. This decrease occurs in bluff and general upland areas. An analysis of the frequencies of occurrence of this variable by time periods indicates a slight association that is statistically significant only at the 0.1 level. When landform is correlated with stream order (Fig. 5), one finds that 58% of Late Woodland sites on the first order (Scioto) are on the valley floor and 25% are on the terrace. On the same stream order for Late Prehistoric 53% are on the valley floor, 45% are on terraces, and none are in the uplands. For the major tributaries of the Scioto (third order), the increase is even more striking: 75% of Late Prehistoric sites are located on terraces and 25% on valley floors, while Late Woodland sites are almost equally distributed on valley floors, terraces, and uplands. The percentage of Late Prehistoric sites decreases for fourth order streams with most located in the uplands. Late Woodland fourth order sites are found in a ratio of 2:1 on bluffs compared to valley floor. On the fifth rank streams all Late Prehistoric sites and 67% of Late Woodland sites are found on the bluff. A significant association between landform and stream order occurs for Late Woodland sites at the 0.01 level, while the association is even stronger for 152 «o1 - s• i, ■ i—i r 1 V T B S U V T B S U v T BSD V T B S U L W I. p I IS' LP a 1st Order Stream b 2nd Order Streams s o '/ s o '/,- ' so'/-- V T B S U V T B S U V T B SU VT n B S U L IV L P L W LP c 3rd Order Streams d 4th Order Streams so S o '/ VT BSU VT BSU L W L P * 5th Order Streams Key: V- Valley floor; T- Terrace; B- Bluff; S* Slope; U- Uplands. Fie. 5. Comparison of Late Woodland and Late Prehistoric Sites by Stream Order and Topographic Landform Locations. 153 Late Prehistoric sites at the 0.001 level (Table 15). In general, vegetation association (Table 16) reflects an increase in valley utilization (20% increase from Late Woodland to Late Prehistoric in Bottomland Hardwood forest association); also an increase in Beech forest with a decrease of more than 50% in all upland vegetation associations. Noticeable differences in the expected proportions of sites occurred for Late Woodland-Late Prehistoric sites and Late Prehistoric sites associated with Beech forest, Late Woodland and Late Prehistoric sites and the Mixed Mesophytic forest, and Late Prehistoric sites and Oak-Sugar Maple forest. Statistically, this apparent association between time period and vegetation zone is only significant at a 0.1 level. The percentage of sites associated with upland soil types decreases by about 20% from Late Woodland to Late Prehistoric, while the percentage found on valley soils increases by 20% (Table 17). Still, however, nearly 50% of all Late Woodland sites to begin with are found on these same soils. An evaluation of the proportions of soil types by time period indicates only one possible association where fewer Late Prehistoric sites than expected are found on upland soils. This association is significant at the 0.05 level. Considering site size, about 70% of Late Woodland sites are Class 1 with an additional 15% Class 2 (Table Table 15. Landform Location of Sites by Stream Order Stream Order 1 2 3 4 5 n % SEp n % SEp n % SEp n Z SEp n % SEp Late Voodland Valley 23 58+.156 3 27+.269 4 29+.242 1 33+.544 0 Terrace 10 25+.137 3 .27+.269 5 36+.256 0 0 Bluff 3 8+.083 0 1 7+.137 2 67+.544 2 67+.544 Slope 1 3+.049 1 9+!173 0 0 0 Uplands 3 8+.083 4 36+.290 4 29+.242 0 1 33+.544 Total 40 100 11 100 14 100 3 100 3 100 (G/q=21. 871; x2(.01)(8)==20.090) Late Prehistoric Valley 20 53+. 162 0 2 8+.192 0 0 Terrace 17 45+.161 3 50+.408 6 75+.306 0 0 Bluff 1 2+.052 1 17+.305 0 1 20+.358 2 100 Slope 0 0 0 0 0 Uplands 0 2 33+.385 0 4 80+.358 0 Total 38 100 6 100 8 100 5 100 2 100 G/q=46.95; x2(.QOl)(8)=26*124‘ 155 Table 16. Frequency of Sites with Vegetation Associations Vegetation LW LW-LP LP Totals n X SEp n X SEp n •X SEp n Z Bottomland- Hardwood 27 37+.113 21 40+.136 34 58+.129 82 45 Elm-Ash 19 26+.103 1A 27+.123 9 15+.094 42 23 Beech A 5+.053 7 13+.095 8 14+.089 19 10 Mixed Mesophytic A 5+.053 1 2+.038 4 7+.066 9 5 Mixed Oak 11 15+.08A A 8+.07A 3 5+.057 18 10 Oak-Sugar Maple 7 10+.069 A 8+.074 1 2+.034 12 7 Prairie Boundary . 1 1+.028 1 2+.038 0 0 2 1 Totals 73 100 52 100 59 100 184 100 G/q=19.011; X2( . i)(12)“18•549• 156 Table 17. Site Distribution by Soil Types Soils LW LW- LP LP Totals n Z SEp n % SEp n % : n % SEP Valley (F-O-G-R) 30 44+.120 24 48+. 141 35 64+ .130 89 51 Valley (E-W-O-R-G) 12 18+.092 12 24+. 121 7 13+ .090 31 18 Upland (Miamian) 16 24+.103 9 18+. 109 2 4+.050 27 16 Other 10 15+.086 5 10+. 085 11 20+ .108 26 15 Totals 68 100 50 100 55 100 173 100 G/q-15.55; x2(.05)(6) = 12-;592. Table 18. Sites by Size Classes Class LW LW-LP LP Totals n Z SEp in Z SEp n Z SEp n % 1 32 70+.136 23 48+.144 12 44+. 191 67 55 2 7 15+.106 13 27+.128 7 26+. 169 27 22 3 2 4+.060 3 6+.070 2 7±. 101 8 7 4 1 2+.043 3 6+.070 2 7±- 101 11 9 5 2 4+.060 2 4+.058 1 4±. 073 5 4 6 2 4+.060 3 6+.700 3 11±. 121 8 7 7 0 0 1 2+.041 0 0 1 0.8 Totals 46 100 48 100 27 100 121 100 G/q-33.617; X2v2 £'Qq j ^£}2)m 32.910. n m 157 18). Only 4% each fall in Classes 5 and 6. This pattern changes later, where 44% of Late Prehistoric sites are Class 1, a decrease of 26% from Late Woodland. Over 25% are now Class 2, representing an increase of 10% from Late Woodland. The proportion of sites also increases in all other classes except Classes 5 and 7, nearly tripling the percentage in Class 6. The SEp indicates greater than expected proportions for Late Prehistoric sites in Classes 1, 2, 3, 4, and 6. Late Woodland-Late Prehistoric category sites are clorest to the Late Woodland percentages in all cases except Class 4 (intermediate in value between Late Woodland and Late Prehistoric) and Class 7, for which there were no Late Woodland or Late Prehistoric sites. These differences are significant at the 0.001 level. Site size has been analyzed in conjunction with several environmental factors to determine if this variable is associated with other site locational attributes. An examination of size and landform location, for example, reveals that over 67% of Class 1 Late Woodland sites are located either on the valley floor or in the uplands, with less than 20% on terraces (Fig. 6). During the Late Prehistoric, 67% of all Class 1 sites are found on a terrace with an additional 25% on valley floors. Approximately 7% of all sites are located on a bluff; all of these are less than 4 ha and all but one is Late 158 »•/. HI V T B S U V T B S U S II V T B S M I. IV L P L IV I. P Class 1 b Class 2 100 so y. so V T B S U V T B S U V T B S U V T B S U L IV I P L IV L P C Class 3 d Class 4 ioo 1*0 soy. so y.- •o y . s»y.- VTBSU VTBSU VTBSU VTBSU L W I. P L IV LP e Class S I Class b Key: V* Valley floor; T* Terrace; B* Bluff; S* Slope; U* Unlands. No comnarablo data was available for Class 7 sites. Fig. 6. Comparison of Late Woodland and Late Prehistoric Sites by Size Class and Topographic Landform Locations. 159 Woodland. Of those sites in Class 4 or higher, 50% are found on terraces, greater than 33% on valley floors, and one site (Late Prehistoric) in the uplands. Statistically, the variables of site size and landform are not associated (Table 19). Correlated with soil types, 70% of all Late Woodland sites on valley soils are Class 1 compared to 47% for Late Prehistoric (Fig. 7). Almost all sites larger than Class 1 are found on valley soils (80%), regardless if they are Late Woodland or Late Prehistoric. The small sample sizes for Late Woodland and Late Prehistoric sites make it hard to evaluate these results. When compared statistically, for example, site size and soils appear to be associated for Late Woodland sites but not for Late Prehistoric sites. The last characteristic examined is the number of temporal components present at a site (Table 21). Single component sites increase by over 15% from Late Woodland to Late Prehistoric, while the proportion of multi-component sites generally decreases from Late Woodland to Late Prehistoric. An additional 35% of sites are ambiguous Late Woodland-Late Prehistoric. Fully 87% of such sites contained no other temporal component. Also, of those sites which had four to five components, 77% contained both Late Woodland and Late Prehistoric components, while only 2% of such sites contained only a Late Woodland component. Table 19. Site Size Compared with Landform Location Size Class 1 2 3 4 5 6 n X SEp n X SEp n X SEp n X SEp n X SEp n X SEp Late Woodland Valley 12 38+.03 4 57+.14 0 1 50+.50 1 50+.50 2 100 Ter.race 6 19+.02 114+.10 0 0 1 50+.50 0 Bluff 1 3+.01 1 14+.10 ‘ 1 100 1 50+.50 0 0 Slope 2 6+.01 0 0 0 0 0 Upland 11 34+.03 1 14+. 10 0 0 0 0 Total 32 100 7 100 1 100 2 100 2 100 2 100 (G/q=16. 32; *2 (.5)(20)i-:19.337) Late Prehistoric Valley 3 25+.07 2 29+.13 0 2 67+.31 0 0 Terrace 8 67+.08 2 29+.13 0 1 100 2 67+.31 Bluff 0 1 13+.09 0 0 0 0 Slope 0 0 0 0 0 0 Upland 1 8+.04 2 29+.13 0 133+.31 0 1 33+.31 Total 12 100 7 100 2 100 3 100 1 100 3 100 G/q=12.985; X2(.5)(15)=14.339. Table 20. Site Size Compared with Soil Type Soil Type Fox F.ldean Miamian Other Total Class n % SGp n % SEp n % SEp n % SEp n % 1 7 24+.029 5 17+.026 11 38+.033 6 21+.028 29 100 2 0 5 71+.129 0 2 29+.129 7 100 3 0 0 0 0 0 4 0 1 50+.-500 0 1 50+.500 2 100 5 2 100 0 0 0 2 100 6 1 50.500 1 50+.500 0 0 2 100 (G/q=21.37; X^(.5 )(i5 )*14.339) ■ Late Prehistoric 1 8 67+.078 2 17+.063 0 2 17+.063 12 100 2 2 29+.129 2 29+.129 0 3 43+.141 7 100 3 2 100 0 0 0 2 10.0 4 2 67+.310 0 0 1 33+.310 3 100 5 1 100 0 0 0 1 100 6 267+.310 0 0 1 33+.310 3 100 G/q=5.03; 5)(10)=^*^^^' 162 so# jo y n a n F E MO F E M 0 F E M 0 F E M 0 LW t P L W L P a. CLASS I b. C LASS 2 i»o so X- SO <0 l-l 30 /•' FEMO E M 0 E M 0 F EM L W L P L W L P c. CLASS 3 CLASS 4 i«e 33 190 s o y . AM soy.. so r—i *°X> FEMO FEMO FEMO FEMO L W L P L W L P e. CLASS 5 f. CLASS 6 Key: F-Fox Association; E-Eldean Association; M-Miamian Association; O-Other. Fig. 7. Late Woodland/Late Prehistoric Sites Compared by Size Class and Soil Type. 163 Table 21. Percentage of Temporal Component8 per Site Components LW LW-LP1 LP Both^ Totals n n % n Z n Z n Z n Z 1 26 49 0 0 25 66 0 0 51 29 2 14 26 8 13 8 21 2 9 32 18 3 12 23 0 0 5 13 3 14 20 12 4 0 0 0 0 0 0 9 41 9 5 5 1 2 0 0 0 0 8 36 9 5 Ambiguous 0 0 52 87 0 0 0 0 52 30 Totals 53 100 60 100 38 100 22 100 173 100 1 Sites that cannot be differentiated. Sites contain both a LW and LP component. 164 In no case did a site with four-five components contain a Late Prehistoric but not a Late Woodland component. Returning now to a consideration of the site clusters mapped in Fig. 4, and Figs. 8-10, it is possible to examine them in terms of some the variables discussed above, especially landform, stream order, site size, and the nature and number of components. Ross County, with 88 sites, has four major clusters of sites (representing 68% of the sites in the county). Late Woodland makes up 8-33% of each cluster, with Late Prehistoric ranging from 0-50% of a cluster. Late Woodland-Late Prehistoric sites closely parallel the Late Prehistoric, present at 5-56% per cluster. Sites with both Late Woodland and Late Prehistoric occur in three out of four clusters at 25-33%. Almost all of the sites in these four clusters, which are all along the Scioto, occur on the valley floor or a terrace. All size classes are represented, though Class 1 sites predominate. Only one-tenth fall in Class 6 or 7. Pickaway County contains 47 sites, with 81% found in three major clusters all of which are located on the main trunk of the drainage. Late Woodland sites make up 53% of the sample, ranging from more than one-half to two-thirds of the sites in two clusters. Late Prehistoric, on the other hand, makes up 11-21% of each cluster, and Late Woodland-Late Prehistoric sites range from 21-67% of a cluster. No multi-component sites which contain both Late 165 W Km t Fig. 8. Late Woodland Site Distribution. 166 10 Km 9. Late Prehistoric Site Distribution. O LW &/or LP A L W & LP Fig. 10. Distribution of Late Woodland/Late Prehistoric Sites. Woodland and Late Prehistoric are found in these site clusters. From 33-67% of these sites are found on the valley floor with only 25% of one cluster found on a terrace. Each cluster contains a bluff site, while two clusters have about 25% of their sites located in the uplands. In terms of site sizes, only three sites (8%) are Class 4 or 5 and fully 79% are Class 1 and 2. Interestingly enough, no comparable site clusters are found in Franklin County. Only 14 sites are located here. Given the largely metropolitan aspect of the county and its history of settlement and development, this lack may represent the effects of community growth on this portion of the archaeological record for the central Scioto drainage. On the other hand, this apparent lack of site clusters in Franklin County may represent very real differences between northern and southern portions of the drainage. This question will be explored more fully in following chapters. Summary of Site Distribution Characteristics Over-all, the variables examined here make it apparent that sites are found in all physiographic provinces from Late Woodland through Late Prehistoric, and thus there is no indication of a major population shift between these divisions with the transition from Late Woodland to Late Prehistoric. Also, no significant increase is seen in the 169 proportion of Late Prehistoric to Late Woodland sites located on the Scioto and its major tributaries. In fact, there may be a slight decrease because*, over-all, the percentage of Late Prehistoric sites decreases within the region. However, within stream orders a major difference is notable in location by landform with a significant increase in Late Prehistoric sites on terraces compared to valley floor. A difference is also seen in fourth order stream sites where the Late Prehistoric sites are mostly upland compared to mostly bluff sites for Late Woodland sites located here. Most Late Prehistoric bluff sites are on the fifth rank streams. Also, while 75% of Late Prehistoric sites are on valley soils, nearly 50% of all Late Woodland sites are on similar soils. This does not represent the dramatic difference that would be expected from the population-intrusion model. Clearly, Late Woodland activities also favored such areas. Furthermore, while more Late Woodland sites (17%) are found in the uplands, this represents only slightly more than Late Prehistoric sites in the same area (12%), again, suggesting fairly similar utilization of these areas over time. Increasing use of the valleys is also reflected in vegetation associations for sites. Comparing size, the proportion of Late Prehistoric sites is greater than expected for almost every size class. 170 Conclusions From Late Woodland to Late Prehistoric, the following changes are seen in the central Scioto Valley. While an increasing proportion of Late Prehistoric sites are found in the Glaciated and Unglaciated Allegheny Plateaus, with a corresponding decrease in the Till Plains, this represents only a slight statistical association between the variables of time and physiographic province. No statistical association exists between stream order and time period, with very little difference noted between the proportions of Late Woodland, Late Woodland-Late Prehistoric, and Late Prehistoric sites. A slight association does occur between time period and landform, where the proportion of sites located on terraces compared to valley floors increases by 20% from Late Woodland to Late Prehistoric and upland sites decrease by 50%. However, when landform is considered with stream order, the association is highly significant for Late Prehistoric sites which increase on first and third order terraces, and, contrary to Late Woodland sites, are found mostly in the uplands on fourth order streams and 100% on bluffs for fifth order streams. A slight association is again apparent between vegetational zone and time period. From Late Woodland to Late Prehistoric an increase in the proportion of 171 settlements associated with the Bottomland Hardwood and Beech forests occurs, as well as a more than 50% decrease in the proportion of sites in all upland vegetational zones. Site size is highly associated with time period. The Late Woodland has a higher percentage of Class 1 sites, but the Late Prehistoric has greater than expected proportions of every other size class except for Classes 5 and 7, with two times more Class 6. There seems to be no association between size and landform, and a slight association between size and soils for Late Woodland, but not for Late Prehistoric sites. This difference may be due to small sample sizes. Clearly, the differences between Late Woodland and Late Prehistoric settlement locations are not of a magnitude to support the complete replacement of one culture by another in the central Scioto Valley as expected under the population-intrusion model. The most striking differences between the two settlement patterns in terms of location occur when landform, stream order, and time are considered. Site size and time period is the next most significant difference. These differences can readily be explained by factors inherent in local development. Changes in economy, in social organization, and in population distribution could all account for the differences observed here. A use-wear analysis of Late Woodland and Late Prehistoric artifact assemblages and a comparison of site structure and site catchment will further explore the potential to document the kinds of changes occurring between A.D. 500-A.D. 1250 in the central Scioto Valley. Use-wear, for example, can indicate whether one set of activities was replaced by another, Site structure can point to differences in social relations and community organization through the arrangement of structures and affiliated features at sites. Site catchment will illustrate whether site location has changed in relation to the location of resources. CHAPTER VII EDGE-WEAR ANALYSIS OF ARTIFACTS Introduction This chapter compares artifact assemblages present in three of the major sites of this study — Baum, Gartner, and Water Plant— through an analysis of edge-wear damage resulting from use. These sites were chosen because they have large, varied artifact assemblages, they provided an example of a Late Woodland and two Late Prehistoric sites, and they were available for analysis. Such a comparison should illustrate any differences and similarities in activities associated with Late Woodland and Late Prehistoric populations which inhabited this region in the past. Theoretical support for such an analysis is presented, followed by a discussion of methodology which details the design and application of a system of macro- use-wear analysis. To implement this discussion, the dimensions and modes of the functional analysis are described. Then the results obtained from the application of this system to the assemblages are presented, and a test of the system against the data is described. Finally, the Late Woodland functional classes are compared to the Late Prehistoric ones and conclusions are summarized. 174 Theoretical Approach to Macro-Wear Analysis The method of analysis utilized In this research is an approach designed to locate and identify edge-wear on artifacts with low-power magnification. Such an approach is in contrast to micro-wear studies which use high-power microscopic analysis (e.g., Keeley 1974, 1980; Newcomer and Keeley 1979; Nance 1971; Yerkes 1985). Functional analysis is based on the idea that use-wear of an object can be identified and separated from damage that might occur to a tool due to natural agents such as wind and water action, the process of manufacture, human actions like plowing, excavation, and curation, or to characteristics of the raw material (Greiser and Sheets 1979). Most commonly, use-wear is identified as patterns of scars or polishes caused by chipping or abrasion during use (Cotterell and Kamminga 1979, Del Bene 1979, Hayden 1979, Kamminga 1979, Lawrence 1979, Unger-Hamilton 1984, and Witthoft 1967). It is also based on the proposition that functional attributes of an object can be distinguished from stylistic attributes (Binford 1979; Dunnell 1975, 1978; Jelinek 1976; Sackett 1977, 1982, 1985). The methodological questions which arise concern the best method for identification of use-wear and material worked. The basic criticisms of low power approaches are that they cannot distinguish 1) use-wear from edge-damage 175 resulting from other factors (Keeley 1980), 2) differentiate functional categories of abrasive forms of damage like polishes (Keeley 1980, Odell 1982), and 3) differentiate between worked materials (Odell 1982, Yerkes 1985). A host of studies suggests that low power approaches can distinguish use-wear damage from edge damage caused by other factors (Tringham, et al. 1974, Odell 1974, Odell and Odell-Vereecken 1980, Hayden 1979, Holley and Del Bene 1980, Moss 1983). Keeley (1980) uses a method of micro-wear analysis with incident light microscopy at 50-400X (high power) to differentiate functional categories of polish and striations. A series of blind tests and experimental replication of these forms of use-wear supports his conclusion that distinct polishes indicative of different methods of use and worked materials can be identified. This information is used by Keeley in the reconstruction of the range of economic activities at prehistoric sites. High power magnification is the only method to identify and discriminate classes of use-wear polishes. The two methods of analysis, in fact, are not mutually exclusive. At a grosser level of analysis, however, the presence of polishes can be observed. By noting their presence and frequency within an assemblage as compared to other classes of wear patterns, useful information can be gained. At this level of analysis low power is sufficient. 176 As for the third criticism of low-power approaches, a series of blind tests were also made by Odell and Odell- Vereecken (1980) and Newcomer and Keeley (1979). Using a low-power approach, Odell and Odell-Vereecken achieved successful identifications of 80% of the utilized portions of tools, 70% of activity, and 61-68% of relative worked material compared to 65% in this category for Newcomer and Keeley. It has been observed that "the only justifiable product of use-wear analysis is the enumeration of aggregates of attributes that correspond to wear types" (Odell 1982, summing up Bunnell's 1975 position). Odell's response is: "who cares if sites A and B share 13 wear types if we have no idea what those types represent?" (Odell 1982:27). Yet, each wear type represents the identification of a distinct action. The range of wear types describes the range of actions engaged in by the prehistoric users of the tools in an assemblage. The value of such basic information cannot be ignored. By putting those wear types into the contexts of sites, and the sites into even larger spatiotemporal contexts, further information can be gained. For, as Odell (1982) points out, such research needs to be problem-oriented towards such questions as culture change and environmental exploitation. 177 Odell and Odell-Vereecken (1980) and Holley and Del Bene (1980) have found that an even better rate of success is possible if worked material is identified in terms of relative resistance rather than specifying the exact material. Is it, in fact, always necessary for such detailed reconstruction? Depending upon the nature of the research problem such exactitude may be superfluous. As noted in this paper and elsewhere (Keeley 1980), such an analysis— at high or low-power magnification— when used as part of several lines of inquiry and in conjunction with information on spatial and temporal context, can give highly satisfactory results. High-power analysis is thus best utilized in situations where high contact material specificity is needed for small samples, and both money and time are in good supply (Odell 1982). This analysis was short on time and money. The basic goal was the identification of functional classes which could be compared for Late Woodland and Late Prehistoric sites. High contact specificity was not necessary in the context of the problem. A comparison of functional classes should shed some light on whether similar or different activities were occurring in these sites, regardless if we know precisely what material was worked by each tool. Methodology: The Macro-Wear Analysis The analysis was based upon lithic, bone, and shell artifacts from Baum and Gartner villages which are curated at the Ohio Historical Society, and lithics recovered from the Water Plant site which are curated at The Ohio State University (Table 22). A relative paucity of lithic debitage and broken tools from Baum and Gartner, compared to Water Plant, can most likely be explained in terms of the historical context of the excavations. Collecting such material when plentiful, unbroken artifacts existed apparently was not a high priority. Three major differences between the assemblages immediately assert themselves: the Water Plant site is virtually lacking triangular points and no bone or shell tools have yet been identified, while these elements constitute major components of the Baum and Gartner assemblages. The lack of triangular points at Water Plant is understandable given the temporal assignment of the site as Early Late Woodland. The lack of bone or shell tools may be the result of two factors: 1) poor preservation in acidic soils, and 2) limited nature of the excavations. Lithics, including chipped stone tools and groundstone tools, comprised the major focus of the study. Bone and shell tools were also analyzed from Baum and Gartner. Pottery was excluded from analysis because the time and energy to be expended was felt to outweigh the amount of potential information to be gained. A hand lens of lOx and a strong, fluorescent light were used as aids in an examination of each artifact. An attempt was made to use 179 Table 22. Assemblage Summaries for Water Plant, Baum, and Gartner Sites Artifact Types Water Plant Baum Gartner Points: Triangular 1 286 60 Other 382 5°2 192 76 5 Chipped Stone 406 273 18 Flakes 20647 74 18 Cores 63 46 1 Ground Stone 117 109 14 Bone Items 0 1517 176 Shell Items 0 652 878 Ceramics 1000 4283 453 Total 22615 3864 1244 ^Whole points ^Point Fragments ^Numbers from Accession Lists 180 higher magnification (a Bausch & Lomb Stereozoom 4 microscope with zoom range of 7x to 30x), but as no appreciable difference was noted in ability to distinguish wear patterns at this level, the attempt was discontinued. No attempt was made to use an SEM or similar microscope for reasons outlined above. A total of 1,128 lithic implements was examined from the Water Plant site. All lithics curated at the Ohio Historical Society for the Baum and Gartner sites were analyzed for use-wear, as were bone and shell implements (ornaments such as beads were excluded). As only 12 shell tools were present, all from Baum, they obviously cannot serve as a point of comparison between Late Woodland and early Late Prehistoric sites except to note that such were not recovered from Water Plant. Therefore, they are not discussed further. A large sample of bone tools, however, was available from Baum and was supplemented by a smaller sample from Gartner. These will be described following the discussion of lithics in order to compare the two Late Prehistoric sites. No preparation of artifacts was made for analysis other than washing if necessary. In all cases analysis proceeded with the separation of artifacts into general categories such as points, flakes, bifaces, groundstone, scrapers, awls, and drills. All items were examined at least twice— with the hand lens and without— before a 181 decision was made to include or exclude an item from further analysis. All broken items were examined for evidence of use-wear. Where breakage obscured a possible use-wear pattern, the piece was excluded as inconclusive. Each item which remained in the analysis was then sketched and weighed. Modal distinctions were determined for each of five dimensions of use-wear, then the results noted on data sheets (Fig. 11). Several morphological and technological attributes were also recorded for each item (Table 23). The dimensions and their modes are defined as follows. DIMENSION K: Kind of Wear. The nature of the , alteration of the tool after use has occurred. This defines the first step of the analysis— the differentiation of edge damage caused by use from that which results from technological, depositional, and post- depositional events in the life of the tool. Some measures of non-use events include: 1) fresh, unweathered edge damage, 2) random, limited damage (based on the assumption that purposeful wear will leave a pattern of wear), 3) evidence of manufacture and preparation of a tool for use or re-use. For example, in heating or retouch the pattern of chippage would be fairly regular, but use would crosscut the arris of such flake scars, resulting a a different pattern. Other factors also have to be considered here, 182 Site: Dimensions Stylistic \ttrihutes Comments I tern 1 7 t Fig. 11. Example of a Data Sheet. 183 Table 23. Morphological/Technological Attributes of Tools DIMENSION 1: Raw material mode 1: Flint Ridge flint mode 2: Local flint mode 3: Other DIMENSION 2: Fracture mode 1: feather mode 2: hinge mode 3: step mode A: none DIMENSION 3: Retouch mode 1: present mode 2: absent DIMENSION 4: Kind of Technique mode 1: unifacial chipping mode 2: bifacial chipping mode 3: ground stone mode 4: other such as effects of raw material, duration and intensity of use, and material worked (relatively hard or soft). The five modes, described below, cover the most basic possibilities for physical alteration of the kinds of raw materials known to have been used as tools (those that have survived in the archaeological record) by prehistoric human groups (Fig. 12). Mode _1: Chipping— small conchoidal fragments broken from the edge. Mode .2: Abrasion— striations on an edge, point, or surface. Mode J3: Polishing— gloss which cannot be resolved into striations. Mode _4: Crushing— irregular fragments broken from an edge, point, or surface, leaving pits. Mode J5: Other— anything not fitting the above modes. DIMENSION L: Location of Wear. The location of wear dimension is described by that portion of the artifact which constitutes a tool, exhibiting use-wear. The geometry of this portion of the artifact is defined broadly as some variation of an angular or curvilinear geometric representation. Angular refers to a linear configuration of lines, planes, points, and curvilinear to spherical and elliptical configurations. Mode 1_: Angular point— intersections of 3 or more 185 0 < • Kind ol K n ot.! m2 Akniit 0. 2. iocuioa *1 Won flll m 2 • 3 Anpilai Point Uiifacial Angular Edge Bi f a c in * V ip i’**- Edge «S *6 Angular l*l«* rjtnilinear Paint C unilinear Edge 0 3 OfiiAUtitii of Wear J! * P ' a 4 r a l l e l ' k ------* r — HF— --- M \ PprpcnJicular V a r i a b l e \ ! \ 'R2 C b l i q u r 0.4: tkl|i it Win A m Strai0if N o tc h Fig. 12. ^alysis FunCtional dimensions of Use-Wear 186 planes at a point. Including the point. Mode 2 : Unifacial angular edge— Intersection of 2 planes, including the line of intersection. Wear occurs on 1 plane. Mode ,3: Bifacial angular edge— same as mode 2, except wear extends onto both planes. Mode 4^ Angular plane— a single plane surface. Mode _5: Curvilinear point— a 3-dimensional parabola or hyperbola. Mode Curvilinear edge— a curved plane (2- dimensional parabola or hyperbola) bent along one axis. Mode .7: Curvilinear plane— a curved plane with spherical or elliptical distortion of large radius. Mode ,8: Nonlocalized— none of the above. DIMENSION 0: Orientation of Wear. This dimension accounts for the factor of direction of the force which forms a use-wear pattern on a tool. This is recorded by noting the central tendency of the linear orientation of the wear pattern in relation to a Y-plane established perpendicular to the X-axis of the artifact. This is a line or plane connecting use-wear to the body (X-axis) of the tool. The modes describe the possible relationships between the force and the T-plane in terms of the tool (Fig. 12). .t. 187 Mode i: Perpendicular to the T-plane Mode 2: Obliaue to the T-plane. Mode 3: Variable to the T-plane. Mode A: Parallel to the T-plane. Mode 5: Nonlinear wear. Mode 6: None. DIMENSION S: Shape of Worn Area. This dimension describes the shape of the X-plane's intersection with the tool’s surface in relation to the body of the artifact (Fig. 12 ). Mode _1: Convex— an arc away from the tool center. Mode 2: Concave— an arc toward the tool center. Mode 3^: Straight— no detectable curve to edge. Mode 4: Point— angular expression of mode 1. Mode J5: Notch— angular expression of mode 2. Mode 6z None* eliminated from analysis on the ground that if there is no shape there is no wear. Mode 7.: Flat— expression of mode 3 for a plane. DIMENSION A: Edge Angle. Edge angle approximates the angle of damage to the tool from use-wear, equivalent to the spine plane angle. Modes are defined qualitatively and quantitatively. Angles were measured against a pattern (Fig. 12). Mode 1_: Low— 0-25°. Mode 2: Medium-low— 25-45°. Mode 3,: Medium-high— 45-65°. Mode 4: High— 65-90°. In essence, the five principal dimensions of analysis were chosen to deal with specific aspects of function such as the raw material of the tool and the general nature of the material worked, the part of the artifact actually contacting worked material, the direction of the force applied during use, the effect of this on the tool, and edge angle. The data collected for each assemblage has been recorded on data sheets as noted. This information, which includes sketches of artifacts with each use-wear pattern numbered, any provenience information recorded on an artifact, and the morphological and technological attributes listed in Table 23, will not be fully reproduced in this volume due to its extensive nature. A copy of this data is on deposit with the Central Ohio Archaeological Survey at The Ohio State University. Results of Application: A, Functional Classification A total of 246 preliminary functional classes, made up of 1058 tools, was recorded in the application of the functional analysis to artifacts (Appendix B). The following sections sum up the results of the functional analysis. The frequency of occurrence of modes for each dimension are compared to determine the general nature of differences and similarities among the assemblages. 189 It Is clear that for Dimension K, Kind of Wear, mode 1 (chipping) predominates at all three sites (Fig. 13). The percentage is well over 50% for Water Plant and Gartner and less than 50% at Baum. Baum and Gartner in general share a similar pattern of occurrence of modes for this dimension, although modal values for Gartner are less than Baum for mode 2 (abrasion), mode 3 (polish), and mode 4 (crushing). The pattern of occurrence at Water Plant differs with much less of mode 2 and mode 3, but is similar to Baum in the frequency of mode 4. Plotting Dimension L, Location of Wear, by frequency of occurrence of modes (Fig. 14), indicates a completely different pattern of modal frequencies for Water Plant compared to the Late Prehistoric sites. Baum and Gartner have similar proportions of modes 1-3 (angular point, unifacial and bifacial angular edges), differ at mode 4 (angular plane: 20% Baum, less than 5% Gartner), and are similar at modes 6 and 7 (curvilinear edge and plane). Water Plant has similar frequencies of modes 1 and 2, but greater amounts of modes 3,4, and 5, with a steady decrease from modes 2-7. No increase is seen in modes 6 and 7, such as is present at the early Late Prehistoric sites. When the dimension of Orientation of Wear is graphed (Fig. 15), Baum and Gartner again share a similar pattern of modal frequencies, differing primarily in that mode 5 (nonlinear wear) is about five times more frequent at Baum 190 «.>(- « y .- ry=L 1 2 5 4 5 3 2 3 4 5 u i 1 2 3 4 5 Modes Modes Modes Baum G artner Water Plant Fig. 13. Dimension K: Kind of Wear: Frequency of Modes. L0j£L 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 3 4 5 6 7 S Modes Modes • Modes Baum Gartner Water Plant Fig. 14. Dimension L: Location of Wear: Frequency of Modes. 191 «• 1 2 3 4 5 6 1 2 3 4 S 6 1 2 3 J 5 6 Modes Modes '■odes Gartner Kstcr V ant Fig. 1 5. Dimension 0: Orientation of Wear: Frequency of Modes. «%■ Modes Baum Gartner Kater Plant Fig. 16. Dimension S: Shape of Worn Area: Frequency of Modes. ■On 1 2 3 4 1 2 3 4 1 2 3 4 Modes Modes Modes Bairn Fig. 17 Dimension A: Angle of Edge: Frequency of Modes 192 than Gartner. Water Plant has a much smaller proportion of mode 1 (perpendicular orientation), falls between Baum and Gartner with mode 2 (oblique orientation), is about the same as Baum for mode 3 (variable orientation), has much more of mode 4 (parallel orientation), and is close to the value for Gartner for mode 5 (nonlinear wear). For the next dimension, Shape of Worn Area, three different patterns of occurrence of modes distinguish each site (Fig. 16). Baum increases in value from modes 1-3 (convex, concave, straight edge), has a very low percentage of mode 4 (point), none of mode 5 (notch), but some mode 7 (flat plane). Gartner has a decrease in percentages from modes 1-2, then a dramatic increase in mode 3, similar amounts of mode 4 as Baum, but differs in that mode 5 is present and mode 7 has a lower frequency. Water Plant starts with a high percentage of mode 1 which decreases steadily through modes 2, 3, and 4. None of mode 5 nor mode 7 is present. The final dimension, Edge Angle (Fig. 17), shows similar patterns for Baum and Gartner, both of which have the highest values for mode 1, then mode 2, lower values for mode 3, and differ only by mode 4 which is higher at Baum. Water Plant has a lower percentage of mode 1, and by contrast, values increase between modes 1 and 2. It is nearly equal in mode 3 to Baum and Gartner, but has a much higher figure for mode 4 than Baum or Gartner. 193 Proportional differences in frequency of occurrence of these modes were tested with a Standard Error of Proportion Test (SEp) and an II X C 6-test for each dimension of wear. Only data from Water Plant and Baum were used because they provided comparable sample sizes. For Dimension K, Kind of Wear, greater than expected differences occurred for modes 2 (abrasion) and 3 (polish) in the Baum sample; this difference was highly significant at the 0.001 level (Appendix C). Dimension L, Location of Wear, showed greater than expected differences in proportion for modes 3 (bifacial angular edge) and 5 (curvilinear point) for the Late Woodland site and in modes 7 (curvilinear plane), and 8 (nonlocalized) for the Late Prehistoric site. The association between modes and time period again proved significant at the 0.001 level (Appendix D). Modes 1 (perpendicular) and 5 (nonlinear) of Dimension 0, Orientation of Wear, were significantly higher for the Late Prehistoric site, while mode 4 (parallel) was significantly higher for the Late Woodland site at the 0.001 level (Appendix E). Expected values did not significantly differ between the two sites for either Dimension S (Shape of Worn Area) or Dimension A (Edge Angle) (Appendix F). Modal combinations from all five dimensions created the preliminary functional classes. Water Plant shares only six such classes with both Baum and Gartner (representng 17.5% of all tools). It shares an additional four classes with each individual site. Baum and Gartner also share an additional 16 classes. In addition to lithics, bone tools were also analyzed. These are described in the following paragraphs. Bone tools, largely from Baum, clustered into five major classes. Two of these (Classes '313' and '355': identified by the modes of the first three dimensions) are limited to narrow, polished, pointed tools (awls), and one ('322') to tools made from deer or elk metatarsals (scrapers). One class ('375') cross-cut both of these categories, and one final group ('422') was associated with occurrences of class '322' and '375' tools only. Thus, the dominant modal values were Dimension K, mode 3 (polish), with Dimension L varying among modes 1, 5, and 7 for awls and 2 and 7 for scrapers. Orientation was generally mode 5 (nonlinear) for awls and either mode 2 (oblique) or mode 5 (nonlinear) for scrapers. That the patterns identified as use-wear are not the result of manufacture or post- depositional factors is supported by an examination of whole and broken bone tools, as described below. Testing the Classes: Use-Wear or Non-Use Damage? An independent test of the data was done to further evaluate the validity of the functional classes. Data was collected for both triangular points and bone tools because these were the largest groups of artifacts present in the sample. Both broken and complete examples were plentiful. 195 Three functional classes are present on triangular points, where wear Is most frequently exhibited on tips, bases, and one or both side edges. Class '122' accounts for 125 tools, with 81% located on tips or sides of points and 19% at the base. Class '132' is represented by 13 tools, about 50% at the base, 25% each at tips and sides. Class '162' (8 tools) is found at 50% on bases, 33% on sides, and 17% at tips. Basal wear is assumed to be a function of hafting. Wear at the sides and tips must be accounted for in some other manner. When broken triangular points are examined (n-111), 96% represent point tips or finished points missing only the tip. Use at the tip is borne out by the pattern of breakage. Whole midsections would not be expected if pressure vas applied to side edges during use, and, in fact, account for less than 3% of broken triangular points. Side wear is clearly secondary to that of the point. Since basal wear is assumed to be due to hafting, breakage at this part of the point also is not expected to be significant. Less than 1% of broken point fragments are bases. Thus, the identification of use-wear on triangular points is supported by patterns of breakage. A similar comparison was made for bone awls and bone scrapers. Mean thickness of whole awls is about 7.2 mm (n»174), while mean thickness of those broken in use is 6.35 mm. Polish, the kind of wear dominant on these tools, 196 may be a form of attrition (Kamminga 1979, Diamond 1979, Del Bene 1979). With use, greater attrition and breakage would be expected. Of those awls with a very sharp tip (could still prick the skin as opposed to a blunt tip), all had wear confined to the tip itself, compared with 4% of the blunt awls. Blunt awls had a greater area of polish and were broken much more frequently (68%) compared to sharp awls (32%). A total of 10 complete bone scrapers from both Baum and Gartner were compared to a sample of 63 broken ones from Baum (Table 24). Use-wear on broken scrapers was confined to that portion of the tool closest to the break and were broken in the middle of the artifact lengthwise, at the thinnest point. On the average broken pieces tend to be wider and thicker at this point than whole scrapers. However, when the amount of attrition is measured (over-all thickness of the artifact minus the measurement for the thinnest point at use, Table 24: col. t-T), it is shown that the amount of attrition is much higher for the broken scrapers than for whole ones. With breakage consistently occurring at the point of usage, the idea that it was due to function and not manufacture, deposition, or some other factor is supported. Summary and Conclusions On the surface, the early Late Prehistoric sites are more similar to one another than to the Late Woodland site 197 Table 24. Bone Scraper Measurements: Baum and Gartner Type L WT 1 t-T Total # Whole: Unused 244 16.8 10.9 —— 2 Used 252 16.3 7.7 105 3.2 3 Recon. 254 18.5 8.2 103 2.7 5 Mean 251. 4 15.95 7.7 103.5 2.95 Total 10 Broken: DMT1*2 124 17 10.3 50.2 7.3 27 PMT1 *3 115 17 12.9 68.6 16.8 36 Mean 120 17 11 .6 59 .4 12 .1 Total 63 ^Baum only. ^Distal end of metatarsals. ^Proximal end of metatarsals. 198 as might be expected. They are not, however, carbon copies of one another which would be expected given the original descriptions of the sites. In contrast, the Late Woodland site is similar in some respects to one or the other of the early Late Prehistoric sites. It was not possible to adequately compare functional classes among the sites given the biased nature of the assemblages. Flakes, for example, are grossly underrepresented at the Late Prehistoric sites, probably a result of collection methods and decisions at the turn of the century. Also, the Late Woodland bone assemblage has not been analyzed, so that no comparison was possible using bone tools. At a gross level of analysis, it seems that similar activities were occurring at the Late Woodland site and both of the early Late Prehistoric sites. The high proportion of groundstone classes at the sites suggests the importance of woodworking tools. (The smaller percentage at Gartner may be due to sampling, since this site has fewer numbers of most artifact types.) It is clear that the Late Woodland site was not a specialized activity site, for example, a meat-processing or nut-processing camp, but that a wide range of activities occurred here which necessitated a variety of tool types. The same can be said for the Late Prehistoric sites. Equally clear is that activities, while similar, were not identical at any of the sites. The presence of bone tools marks another difference between the sites, with the Late Prehistoric sites having a significant number of such tools. These apparently worked a soft material due to the nature and extent of wear upon the tool (Lawrence 1979:118). That this sheen does not result from the manufacturing process is clear from Mills' (1904) description of and my own examination of tools in production in which a splinter was removed from a metatarsal, leaving striae and sharp edges, but no gloss. As Late Woodland sites in the study area have yielded bone tools, this apparent difference between assemblages is probably the result of non-use factors like differential preservation. Thus, the Late Woodland and Late Prehistoric sites share a similar toolkit of groundstone and bifacially worked lithics. The differences that are apparent in the frequency of modal distinctions for dimensions of wear in general show an increase in proportion from Late Woodland to Late Prehistoric (or a decrease in a few cases). At no time do modes completely replace one another. The differences are hardly of a magnitude to falsify a hypothesis of local development in the drainage from Late Woodland through Late Prehistoric. The changes in economy and social interaction which were occurring in the central Scioto Valley in this period are explored further in the following chapters. CHAPTER VIII A COMPARISON OF SITE STRUCTURE Introduction A comparison of site structure is undertaken in this chapter in order to develop a better understanding of the Late Woodland-early Late Prehistoric period in the central Scioto drainage. This examination begins with a short discussion of the concept of site structurey then proceeds to a description of the structure of a Late Woodland site, Water Plant, and two Late Prehistoric sites, Baum and Gartner, through an analysis of features and other pertinent data. The analysis concentrates upon these three sites because they provide examples of both Late Woodland and Late Prehistoric sites, they have been extensively excavated and described in the literature, and they formed the basis for the functional use-wear analysis completed in the previous chapter. This information is augmented with data from other Late Woodland and early Late Prehistoric sites located in the Scioto and adjacent drainages. Site structure, as used here, refers to the location of features, artifacts, and other cultural material remains within a site. In this sense, a site could represent either a segment of or an entire community, when such is 2 0 0 201 defined as an "arrangement of artifactual materials remaining from the domestic occupations used by a group of economically, socially, and politically related people" (Fuller 1982:1). Site structure could be defined as simply as a few flint flakes and firecracked rocks on the surface or as a much more substantial concentration of subsurface features, artifacts, and ecofacts. Fuller (1982) refines this definition to include two basic community patterns: nucleated and dispersed. A nucleated community pattern occurs when all members of the community live within one contiguous settlement. When the community is separated into smaller settlement units located at some distance from one another, this is known as a dispersed community. Rafferty (1985) raises the related issue of how communities can be identified as sedentary, since dispersed or nucleated defines a pattern of habitation in space, but not necessarily in time. Sedentary is taken to mean that a site is occupied by at least some members of a community year-round (Rafferty 1985:115). This means that some members may be elsewhere at certain times of the year. These temporary activity sites would be part of the year- round settlement system of the community. At this point, however, only the details of site structure will be considered. The implications of this pattern in terms of year-round community activities will be dealt with in the following chapters. 202 A Late Woodland Example: Water Plant As Indicated previously, this site was subjected to intensive surface surveys in the spring and summer of 1980 and to test excavations also in the summer of 1980. Subsurface features are described, then compared with a distribution of artifacts across the surface of the site. Also, the distribution across the site of functional tool classes, which were defined in the previous chapter, is examined to see what this information can tell us about the community which occupied the Water Plant site. Initial trenches at the site were placed to confirm the existence of a prehistoric ditch which encloses the site on the bluff-edge of Big Walnut Creek. A series of test pits were systematically placed in the southern portion of the site. Other test pits were nonrandomly placed to investigate the nature of certain features exposed in surface stripping of the area from 0-20N, 10-60E (Fig. 18). In addition to more than 20,000 artifacts which were surface collected from the site, 80 features were o excavated. Most of these were clustered within a 20 m* area (Fig. 19), except for those which confirmed the existence of the ditch. All other features, when plotted by plan shape fall into one of four feature classes (Fig. 20). Plan shape was derived from a measure of length minus width, which avoided subjective descriptions of shapes. 203 10 m Test Trench 1 Feature 1 Test Trench 2 Feature 2 \ (Ditch) Troweled Area T.T.7 T.T.5 Test Trench 4 T.T. T.T.8 iT.T.6 ON 50W Fig. 18. Excavated Localities at Water Plant. \ 40F_^0N 60F 20N F.S6 F.4h S6A O F.4/ F.50 * F49 . F.44 q F.45 O f ’" F4J ,®l'.43 «F9B • F99 V J ' •> •-P?36-}jZ |F8f> O F . 3 8 \ n ' 34 • F96 *f:91 ^95 F95 ,ra7 oP48 •F42.r ^ . S 3 ol:92 > 1:12 1-89 O v — ' r8S F.25 Q f S S c F79C GTl:79A,B. ®F88 ® F94 »F29 1-28 «I;SI ISA FSB • I-S4 •1-27 . 1-30^ p32 Key: • Circular ® Hllintical F31 © Small Klongatc O l.nrgc F.longatc jJOm, 40R“^ N 60F ON Fig. 19. Distribution of Features at Water Plant. V02 205 These classes have been designated 1) Circular (n«47), with L-W equal to 0-3, 2) Elliptical (n-22), where L-W equals 4- 9, 3) Small Elongate (n«6), with L-W equal to 14-35, and 4) Large Elongate (n-5), where L-W equals 50-70. These classes are supported by consistent differences In other attributes. For example, Class 2 features average 5 cm more in depth than Class 1, while Class 4 features are more than twice as deep (mean depth - 53 cm) than Class 3 (mean depth - 23 cm). Class 3 features are shallow, symmetrical pits that tend to be flat-bottomed, while Class 4 features are deeper, symmetrical, and mostly round-bottomed. Classes 1 and 2 are considered to be postholes because length and width are so small, but are further distinguished by the fact that 21% of Class 1 contained sherds, fragmentary animal bone, and lithic debitage, while 27% of Class 2 features contained only lithics (a single feature had sherds and no lithics). Classes 3 and 4 were similar in contents, although about 33% of Class 3 and 20% of Class 4 yielded no cultural material. Those that did contained mostly lithic debitage, sherds, and fragmentary animal bpne. Class 3 features tended to have firecracked rock and Class 4 to have more bone. Both surface collections and excavation clearly illustrate that this Early Late Woodland site lacks the planned site structure consistently associated with Late Prehistoric sites. The distribution of potsherds (Fig. 21) Count 10 12 14 16 i. 0 Wtr ln Faue: itga o Pa Shape Plan of Histogram Features: Plant Water 20.Fig. 8 - . ZZI \ Length Width - iclr Small - j Elongate ^ j | Circular li tc l Large Elongate I H Elliptical 'ds f r n __ 4 4T“ b s t 60 65 206 70 207 ZOOjm I ^ Key: ' J -20 gm l‘.*M 21-40 gm 41-60 gm 61-80 gin Fig. 21. Distribution of Potsherds by Weight, Water Plant (Weight more indicative than count). and of lithics (Fig. 22) is ubiquitous across the site, as is that of firecracked rock. No plaza, which is defined as an open area devoid of artifactual material, exists at this site. While some areas do have higher surface clusters of sherds and lithics, these clusters are not arranged in tidy concentric activity rings. The same can be seen from the distribution of features (Fig. 19). While the postmolds do not neatly resolve themselves into the outlines of structures, it is possible to construe Features 33, 34, 41, 48, 79C, 88, and maybe 29 or 28, as part of a single structure, especially given their general regularity of spacing and the fact that they are all members of Class 2 features. Likewise, Features 49, 43, 45, and 46 are suggestive for the same reasons. Class 3 features are dispersed across the excavated area, while Class 4 features are all in relative proximity to one another. This distribution will now be compared to data on the spatial distribution of tools across the site. The portion of the site which was thoroughly surface- collected illustrates several clusters of tools (Fig. 23). In general, as indicated in the previous chapter, the predominant functional classes at this site are Classes '122', '132', (both chipped stone tool classes), and '414', and '475' (groundstone tool classes). In addition, a number of lesser groups are present, such as 'Ill's,' and some other variants of groundstone tools (Table 25). 209 \ 200N 10 m 50 w Key: n = 101-200 n =2 0 1*400 Fig. 22. Distribution of Lithics All Plant. ' A11 Samples, Water 2 1 0 \ V *>* I Key: A-L * Clusters of Tools 130 p x ■ tool o m pjtS • ■ nostholes Fig. 23. Distribution of Functional Classes and Subsurface Features, Water Plant. 211 Table 25. Functional Tool Classes Represented at the Water Plant Site. Class Definition Description N (1) ’122’ Chipped Stone 118 (2) '132’ Chipped Stone 76 (3) ’414' Groundstone 53 (4) ’475’ Groundstone 23 (5) '111' Chipped Stone 8 (6) '152' Chipped Stone 5 (7) '324' Groundstone 3 (8) ’344’ Groundstone 1 (9) '375' Groundstone 1 (10) ’544’ Groundstone 4 Total 292 These clusters are spread across the site with no indication of a planned open space, or plaza, nor obvious doughnut-shaped rings. Clusters of tools are relatively the same size and have nearly equal distributions of chipped stone tools and groundstone tools (Table 26). Clusters A, B, and C are somewhat isolated— between these two and the remaining clusters is a scattering of tools from the predominant chipped stone classes. Also, these northernmost clusters are rather limited in the functional classes represented. Greater variability exists in the other clusters, especially in those nearest the excavated portion of the site. Over-all, Classes '122' and *132' are ubiquitous across the site, while Class ’414’ is located either within Table 26. Surface Clusters of Tools: Water Plant Site Tools: 1 2 3 4 5 6 7 8 9 10 Cluster : #/% #/% #/% #/% #/% #/% #/% #/% #/% #/% A 12/10 10/13 0 0 0 0 0 0 0 0 B 13/11 7/09 3/06 0 0 0 0 0 0 0 C 8/07 7/09 6/11 3/13 0 0 0 0 0 0 D 9/08 4/05 2/04 0 1/12.5 0 0 0 0 0 E 6/05 8/11 8/15 1/04 0 0 0 0 0 0 F 10/08 4/05 0 1/04 2/25 0 0 0 0 0 6 5/04 1/01 6/11 2/09 3/37. 5 1/20 0 0 2/50 1/100 H 4/03 4/05 5/09 1/04 0 0 3/100 0 0 0 I 6/05 2/03 5/09 6/26 1/12. 5 1/20 0 0 1/25 0 J 8/07 6/08 3/06 2/09 0 1/20 0 0 1/25 0 K 6/05 4/05 0 0 1/12. 5 0 0 0 0 0 L 3/03 5/07 0 1/04 0 0 0 1/100 0 0 Total 90/76 62/82 38/72 17/74 8/100 3/60 :3/100 1/100 4/100 1/100 N= 227/78% of all tools 213 or adjacent to a cluster. Class '475' only occurs outside of a cluster in 4 cases, and Class '111' is only found in clusters D, F, G, I, and K— located in the southern and western portion of the site. The remaining five tool categories occur as relatively isolated incidences. Most of the excavated features (Fig. 23) lie beneath Surface Cluster J. If the posthole patterns represent more than one structure, then it is obvious that the tool clusters on the surface are too large to represent the interior of a single structure. However, given the concept of a domestic unit of scale (Fuller 1982; Kent 1984), it is possible that surface clusters indicate a diffuse domestic unit, one that could include more than one structure and exterior features and activity areas. While all of the clusters may not represent contemporaneous occupations (given the range of C-14 dates), their composition remains relatively similar, suggesting similar activities at this site over time. Examples of Late Prehistoric Sites; Baum and Gartner Features identified at Baum include structures, hearths, refuse and storage pits, burials, and a mound. A total of 49 structures were excavated by Mills (1906), most of which were associated with groups of hearths, pits, and burials and were located north, south, and east of the mound (west of the mound is the Scioto River). Only 10 structures had no associated burials and few pits. All 214 structures had an interior hearth, and all contained similar living debris accumulated on the occupation floors. One structure was twice as large as the other structures (6.4 m x 3.7 m compared to 3.2 x 1.8 m). This feature never changed position from its original construction while the others had evidence of rebuilding from place to place. Sometimes structures were rebuilt over burials, or were used for burials or as pits. The largest structure was oblong in shape compared to circular for the others and contained 21 interior posts (7 of which had bark and charcoal in the bottom of the posthole). It also had a large hearth in the center which was 1.2 m in diameter and 15 cm deep at the center (7.6 cm deep at edges). The hearth had been plastered with successive layers of clay (like hearths in other structures) and the earth beneath had been burned to a depth of 20 cm. The interior floor of this structure was originally smooth, but had an accumulated height of 15 cm. Scattered in this floor were unfinished and finished tools and ornaments, polishing stones, sherds, hammerstones, a large stone mortar, and many animal bones (especially deer, raccoon, bear, and wild turkey). A total of seven hearths and 19 pits were clustered north of this structure, while 30 burials were clustered just to the south. This burial group included 20 adolescents and four infants (from refuse pits) and were the only burials in the village from which 215 whole pots were recovered (two with adults, two with children). Burials were located close to structures at less than 1 m, and seldom greater than 2.7 m away at an average depth of 60-75 cm (Mills 1902:79). Other burial clusters were also distinguished such as one group of four adults and six children that contained no pottery (Mills 1906). The pits grouped around the structure associated with these burials contained elaborately decorated pottery and fragments of the largest pot found in the village. The pits were characterized as "unusually large [average pit size was 0.9-1.2 m diameter x 1.2-2.1 m deep] and the contents were very rich in animal remains and implements and ornaments of various kinds" (Mills 1906:122). Only four of 127 village burials were in pits— two in permanent pits which were very deep with bodies covered with clay to about 10 cm, then the pit filled with refuse. The other two were found in pits which had originally been dug as refuse pits since they were nearly one-quarter filled with trash before the bodies were deposited and covered with a layer of ashes about 8 cm in depth before the pits were completely filled with trash. The only other burials in pits were those of infants (n-17) which were also found in very deep pits. Non-pit infant burials occurred in the village; these did not preserve well. Cremation is the only other burial mode present at the 216 site. A cremation was discovered in an ashpit, and it was observed that "around the skeleton were many worked bones, perforated shells, cup stones, hammer stones, etc." (Mills 1906:146). Pits were extensive, their number apparently a function of the size of the associated structure. Most pits contained stratified layers of ashes, burnt animal bones, and firecracked rock. They also contained quantities of artifacts such as sherds, broken scrapers, bone awls, beads, small triangular points, hammerstones, and, according to Mills (1902:84), stone mortars. Gartner Village has a similar arrangement of structures, pits, hearths, and burials. This site, although excavated after Baum, was published prior to the Baum site report. The report makes many comparisons with Baum and is rather summary, as though to save time and effort by eliminating repetition with the Baum report which is much more thorough in its descriptions. As an example, the presence of many structures and pits is noted at Gartner; however, the exact number of each is never indicated in the report. At times, Mills seems to wander back and forth between the sites, so that one is not sure whether a descriptive passage concerns Gartner or Baum. This surely indicates, in Mills' view at least, the degree of similarity between these two sites. 217 Since the plural is used, it is assumed that more than one structure existed at Gartner. In fact, beneath the platform of Section 1 of the mound were pits, hearths, and postholes representing more than one structure. Dimensions of approximately 10.4 m x 7 m x 15 cm deep were given as average for these structures. This is larger than the largest village structure at Baum. Refuse pits and hearths, again, are similar to those at Baum and cluster around structures. These also provide evidence of repeated use of the site, since some hearths had been constructed directly over previously filled pits and pits frequently were dug over a burial, as evidenced by the placement of the disturbed portions of a burial in the bottom of the intruding pit, which were covered with dirt before the pit was filled with refuse. Mills found that about 20% of all pits at Gartner had originally been lined with plant material or bark and had 1) ears of corn in rows on the bottom of the pits, or 2) shelled corn in woven bags, or 3) shelled corn and beans in woven bags, or 4) hickory nuts, walnuts, chestnuts, and pawpaw seeds in "goodly numbers" (Mills 1904:150). All of this material was charred. Burials within the village were generally extended, with graves interspersed with pits. A crematory which contained a half-charred individual was located directly south of the mound; no other cremations were discovered in 218 the village. The first section of the mound contained cremations, but all other mound burials were noncremated. The crematory (approximately 5 m x 12 m x 36 cm) had been covered to a depth of about 15 cm with earth and midden, while a structure was built just north of it. A single, intentional pit burial was also found (a flexed inhumation, not a disturbed extended burial). Thus, the two Late Prehistoric sites are virtually identical in their feature populations. They differ in location, size, and in the over-all proportions of functional classes represented at them, although the kinds of activities are very similar. A Comparison of Site Structure The structure of the communities can perhaps best be examined by comparison with other Late Woodland and Late Prehistoric sites, both within and exterior to the Scioto drainage (Table 27). The Water Plant site is recognizably similar to other Late Woodland sites in over-all size, the presence of structures and their shape, and their decided lack of a defined pattern within the site. The Pyles site, called an early Late Woodland site based on a C-14 date of A.D. 398 (Railey 1984) differs from the other Late Woodland sites in this latter respect as it consists of a concentrically planned 'doughnut-shaped' village. However, thermoluminescent dating of materials from the site have yielded more recent dates of A.D. 1000+260 and A.D. Table 27. Characteristics of Late Voodland & Late Prehistoric Sites Site Name Size(ha) Features Structures P M H B Nd F pi # Shape Orient. Late Voodland Water Plant 3 2-3 Circ. None. + ? + 7 D Zencor 2 3 Circ. None. + + ? + - E - Harness-28 4 ? ?? + 7 ? ?- D - Ety 1 ? ? 7 ???? - D,E - Ford 0.5 i? ??- 7 + - - - ~ Lichliter 1.7 4 Circ. ? + ? + Pyles (KY) ? 1 ? Plaza 7 + ?T + - + Late Prehistoric Voss 0.2 2 Rect. Plaza + + _ + + _ + Blain 3.2 3 Oval Plaza + + + + + - + Roseberry (W.VA) 1.4 + Pit. Plaza + + + + + + Enos Holmes 1 Circ. Md. + 7 ? + + D - Gartner 7.1 >1 Circ. Plaza + + + + + - + Baum 12-14 49 Circ. Plaza + + + + + - + Kramer 11.0 ?? ? + ? 7 ? + ■** 7 KEY P=pit B=burial Pl=plaza M=raidden Md=mound H=hearth F=fortification 1100+230, which would place occupation during the late Late Woodland period (Railey 1984:114-116). Two other characteristica are almost universally shared by these Late Woodland sites— the lack of associated mounds and the presence of a ditch or earthwork or both. Most sites also have pits, but no hearths interior to structures. Also, only Zencor has yielded human burials, these located within the habitation area. The lack of burials may be more apparent than real, given that some of the sites have not been fully excavated and some are still undergoing excavation. It may also indicate that burials were occurring elsewhere, away from habitation areas. Late Prehistoric sites differ from the Late Woodland sites by kind and degree of characteristics. These sites become progressively larger when they are ordered by time, ranging from very small (Voss) to about three times the size of the Late Woodland sites (Baum and Kramer). All sites apparently exhibit a plaza and have structures oriented around this feature. The number of structures varies from two recovered at Voss to 49 documented at Baum, with the total sample having a mean diameter of about 4 m (range - 3.2-10 m). Pits, midden, hearths, intravillage burials, and a mound are ubiquitous to the early Late Prehistoric sites. These features are located in definite patterns in relation to the central plaza (the concentric ring pattern of domestic activities) (Fig. 24). None of o o <9 0 • • • no0 o O 1.0 O o O 0 0 o 0 0 • • • -+o "+ 0 Key: -H> * ^ ? Burial Fig. 24. Feature Clusters at Baum. 222 these sites has any evidence of fortifications in the form of ditches, earthworks, palisades, or topographic location. Conclusions Both early Late Woodland and early Late Prehistoric sites are nucleated communities or villages rather than dispersed communities or hamlets. The Late Prehistoric communities are larger and more complex than the earlier villages. A planned, patterned structure is evident in each Late Prehistoric community. That this pattern was not imported wholesale into the region from elsewhere is evident from the existence of a similar pattern at such Newtown Focus sites as the Pyles site (Railey 1984). Like the Late Woodland site of Zencor, the Late Prehistoric sites also have extended human burials within the confines of the village, but, unlike these earlier sites, the later ones all have associated burial mounds (Table 28) which are similar in terms of size and internal features. Most contain a structure, while the number of burials increases from earlier to later sites. Also, in spite of long popular theories of intergroup hostilities, none of the Late Prehistoric villages bear any evidence of defensive structures. On the other hand, nearly all of the Late Woodland communities have an enclosing structure such as an earthwork or a ditch. Many are also situated in what might be called strategic locations such as blufftop edges. Some other possible late 223 Table 28. Late Prehistoric Mound Characteristics Site Mound Characteristics Ht Diam Burials Structures Shape Voss ? ? 7 pits; 1 str. ? Blain 0.46m 21m 7 none Circ. Gartner 2.3m 23m 42 prepared platf; 3 sec. > 1 struc. Baum 4 . 6m 36.6m 17 dbl. struc. pyramid Kramer 4m ? ? ? pyramid Enos Holmes 1.2 24.3m 14 1 struc, ditch; subcon. pits 224 Late Woodland sites, for example, Harness-28, are actually located on the floodplain, which again does not support the suggestion of overt intergroup hostility between Late Woodland and Late Prehistoric populations. Similar information for the Late Late Woodland period is not well- documented. Sites like DECCO, which bears questionable C-14 dates, W.S. Cole and Ufferman, nevertheless, if they are Late Late Woodland, differ from earlier Late Woodland sites in that they are very much smaller (on the average less than or equal to 0.10 ha). Thus, a transition is seen in the central Scioto Valley from small, nucleated villages on bluffs to larger nucleated villages located primarily on terraces. Both sets of villages have evidence of reoccupation. Frequently, such evidence is interpreted as seasonal reoccupation. However, based on the results of a use-wear analysis, it appears that neither the Late Woodland nor Late Prehistoric sites examined were single-activity sites, but represent full-time, permanent communities. This question of sedentariness is addressed more fully in the final chapter. The most striking difference between the Late Woodland and Late Prehistoric periods is the planned community lay out of the Late Prehistoric sites. While this has in the past been used as evidence for cultural intrusion, it has been amply demonstrated- that local derivation best explains 225 the differences between the Late Woodland and Late Prehistoric up to this point. Similarly, the difference in site structure can be economically explained as resulting from some change or changes that affected the populations inhabiting this area. The community is made up of "a group of economically, socially, and politically related people” (Fuller 1982:1) whose artifactual materials reflect those relations. A change in artifactual materials over time points to change over time in one or a combination of those factors which bind community members together. This idea is explored further in the final chapter. Next, however, a final line of evidence is introduced: a comparison of Late Woodland and Late Prehistoric catchments. CHAPTER IX LATE WOODLAND AND LATE PREHISTORIC SITE CATCHMENTS IN THE CENTRAL SCIOTO VALLEY Introduction In this chapter a site catchment analysis was done for the Late Woodland and Late Prehistoric sites of Water Plant, Baum, and Gartner. For comparison, more preliminary analyses were also done for the Ety Enclosure, a Late Woodland site located on Walnut Creek in Fairfield County, Ohio, on the far eastern edge of the drainage, and for Voss, a Transitional Late Prehistoric site. The purpose of this chapter is to examine the relationship between a site's location in the environment and the resources made available by that location. Potential resource availability and evidence for resource utilization from the archaeological record are considered. A discussion of the theoretical basis for site catchment analysis is followed by a description of methodology, a summary of the results for the five sites listed above, and conclusions. Review of Site Catchment Analysis Site catchment analysis has been defined as the "study of the relationships between technology and those natural resources lying within economic range of individual sites" (Vita Finzi and Higgs 1970:5). The method was built on a 226 .t. i n foundation of assumptions concerning resources, economy, population, site location and distribution. The following assumptions of resources were made: resources utilized by human groups can be ranked by staple resources (principal elements of a diet), casual foods (delicacies, medicines, and opportunistic procurement), and emergency foods (edible non-staples not regularly utilized). Much of the recent literature revolves around ranking preferred resources on the part of a group, demonstrating their knowledge of useful resources and the actual incidence or conditions of use of each resource (for example, Lee 1967). It has been suggested that "a human group will in the long run make use of those resources within its territory that are economic for it to exploit and that are within reach of the available technology" (Vita Finzi and Higgs 1970:2), given knowledge of the distribution of resources in time and space. An assumption is also made that population rises to a level allowed by those resources which are accessible to a group based on available technology at any point in time. Sites are assumed to be located based on their potential for exploiting a set of complementary resources. An assumption is also made for site distribution that typologically similar assemblages indicate contemporaneous sites. Finally, a basic assumption is that of rational economic behavior on the part of a group, with Vita Finzi 228 and Higgs (1970) arguing that although sites may be located for decisions based on other than rational economic choices, such behavior would in the long run prove less than optimal and this would be reflected in the archaeological record. In their discussion of method, Vita Finzi and Higgs (1970) distinguish between a site exploitation territory, that area surrounding a site which is habitually exploited by a site's inhabitants, and an annual territory, which is the total area exploited by a human group in a year. The perimeter of the site exploitation territory is deemed to be that distance from the site such that energy consumed by movement within the perimeter does not exceed energy derived from resources procured within this same area. Distance is measured in terms of X amount of time equals Y km. A two-hour perimeter of 10 km is established by Vita Finzi and Higgs (1970) for non-agricultural sites, given the above assumptions. For agricultural sites, a weighting factor related to distance is figured in, and a 5 km range is used for intensive subsistence agricultural economies. They note two major problems in their application of the method. First, the site exploitation territory is divided into zones which are summarized by land-use capability classes. These classes, however, are drawn from present-day situations and may not reflect prehistoric land use patterns. Secondly, discussion of environmental 229 change within the site exploitation territory should stress local environmental factors like drainage and erosion rather than broad environmental categories like "climate." The basic assumptions and method have been criticized and expanded upon over the years. Hassan (1975) notes that in measuring resource use or needs by a human group, consumption of resources depends not only upon biological requirements, but upon social factors as well. Population is a function not only of actual or potential resource availability, but also of factors such as feeding strategy and interpersonal interactions between group members and neighboring groups. Therefore, population size is dependent upon population density, social factors, and site exploitation territory. Above all, "the size of a local group seems to be rather a function of the optimum size required for a stable organizational unit based on daily face-to-face relationship ...for cooperation in hunting, locating food, and defense" (Hassan 1975:40). One might add to this the factor of food-production in economies not solely dependent upon hunting and gathering. Roper (1979) noted in her review of site catchment analyses, however, that those analyses which followed Vita Finzi and Higgs (1970) tended to be less concerned with questions of band- spacing and population density than with determinations of site location. Other discussions have concerned the use of concentric circles as the unit of measurement for analysis of the site exploitation territory. Part of the justification for this shape of site catchment comes from ethnographic studies (Lee 1967, Chisholm 1968) which indicate the distance modern human groups habitually travel away from a settlement or camp on a daily basis. This distance is then utilized as a radius and a circle drawn to encompass the site. Hassan (1975) notes that while hunters and gatherers do not usually operate in a straight line, a circle may not best define a catchment given physical barriers and culturally-defined territories. Zarky (1976) treats a catchment circle as a sample drawn from a regional universe, calculating the percentage of land-use categories within the sample, then using the chi-square statistic to determine if prehistoric groups were selecting certain categories at higher percentages than their ranking within the total universe. Two problems that Flannery (1976) points out are 1) how to define the limits of the catchment area, and 2) how to account for changes in the past and present environment of the catchment. In answer to Flannery's first criticism, the 5-10 km concentric circles are not intended to reproduce the exact pattern of resource exploitation by a group. As in Zarky (1976), the site exploitation territory works as an arbitrarily limited sample of a regional universe based on analogy with modern hunters and gatherers or subsistence agriculturalists. Other categories of information pertaining to the site(s) under analysis can be compared to this model area to reveal a closer approximation of actual resource use. The shape of the catchment itself can be adjusted if environmental or other parameters suggest an obvious pattern of utilization. For example, a coastal riverine-uplands system might be long and narrow. Site catchment must be seen as a model, one to be compared to the environment of the region as a whole and to empirical evidence from sites. Archaeological evidence may point to environmental differences between the past and present at a site, and it may indicate social and cultural factors such as trade and social networks which could have overcome environmental barriers or deficiencies. Attempts have been made to quantify environmental diversity within a site exploitation territory by constructing an environmental diversity index (Styles 1981, Tiffany and Abbott 1982, Schermer and Tiffany 1985). Styles used four measures of dominance and diversity to categorize the faunal assemblage in her work, ranking species by abundance, by presence per excavated unit, by MNI per species, and by dominance in the total faunal assemblage. The catchment area was partitioned into habitat types based on soil, vegetation, topography, and hydrology. Resources were then rated by plant foods, 232 animal foods, and human strategies of Resource procurement. These last, like Hassan's (1975), were based not only upon technological levels, but also upon cultural and social behaviors like taste preferences, scheduling, and food tabus. Predictions of resource utilization were made and evaluated against empirical data. This concept of an environmental diversity index was evaluated by Schermer and Tiffany (1985). The presence and proximity o-f a number of environmental variables to a series of sites were measured under the assumption that sites are not randomly located. Tiffany and Abbott (1982) utilize variations in environmental characteristics to construct micro-environmental differences within the catchment area. In summary, if prehistoric human behavior is at all analogous to modern human behavior, site location is obviously not randomly chosen. Site catchment provides a method of analyzing the reasons behind selection of a site location. Selection will depend upon biological and cultural factors which affect the population. By making explicit our assumptions about resource utilization, a model can be constructed which can be tested by empirical evidence. Thus, the method will provide a useful technique for exploring the degree of difference between the Late Woodland period and the early Late Prehistoric period in the central Scioto drainage. Methodology Site catchments of 5 km radii were chosen because this distance allowed the diverse environmental zones which surrounded each site to be sampled. A larger radius would have been redundant in the kind of zones sampled. Furthermore, a 5 km radius is equivalent to walking distance common to agricultural villages (see Clark and Haswell 1970, Vita Finzi and Higgs 1970). As all early Late Prehistoric sites have been defined as agricultural (Mills 1906; Griffin 1943; Graybill 1981), this was considered a point of comparison between Late Woodland and Late Prehistoric catchments. The environment within a 5 km radius of each site was stratified according to basic topographic differences, which were then correlated with vegetation types (following Braun 1964; Wistendahl and Lewis 1976) and soil types (Petro, et al. 1976; McLoda and Parkinson 1980). Six environmental strata were differentiated as follows: I. Flat land adjacent to the major watercourses of the catchment, less than 700' elevation. II. Broad expanses of relatively flat land greater than or equal to 700-800' elevations, generally near watercourses. III. Slopes, or, land that rises in elevation greater than or equal to 700-1000'. A. North-facing slopes. B. South-facing slopes. IV. Landforms that slope on all sides with no measurable summit; dissects an area. V. Landform that slopes on all sides with a measurable, flat or relatively flat summit. VI. Broad expanses of land at greater than equal to 800' elevations (generally undissected uplands). The catchments were divided into five concentric rings at 1 km intervals with the proportion of each strata calculated for each ring (Figs. 25-29). Site locations and major watercourses were also mapped. The proportion of each strata per ring was measured using a dot-acreage counter with acres converted to hectares and recorded for each site (Tables 29 and 30). Relative estimates of strata were made for Voss Village and the Ety site. Results and Summary of the Site Catchment Analysis An examination of the tables and figures indicates that catchments are relatively dissimilar for the Water Plant site compared to Baum and Gartner. Over-all. the Water Plant catchment may be summed up as consisting of more than 50% Stratum VI. about 20% Stratum I. approximately 10% each of Strata II and IV, lessthan 1% of Stratum III, and no Stratum V. The predominant environmental zone in Baum's catchment is Stratum III 2 3 5 v a » fo'/fiooJ EKf plain ridges undissected uplands Jivers/cr^mr Fig. 25. Water Pla„t Catchment. 236 valley/flood E S plain terraces undissected unlands rivers/stream! Fig. 26. Baum Catchment. 237 Va™ y/fi°®plain terraces ridees undissected unlanns rivers/sti 27. Gartner Catchment. 238 a k m I 1 1 , 1 1 1 111' vAUE t / r„OCD- IHXfcSETC u*.atCi Fig. 28. Voss Catchment. 239 7\v valii- Flood- K.AM Fig. 29. Ety Catchment. 240 Table 29. Total Percentage of Catchment Strata Per Site in Hectares Water Plant Baum Gartner n % n % n % Stratum I. 3841 19.5 3949 20 11811 60 II. 2559 13.0 2045 10 4609 26 III. 16 0.1 11489 59 1702 9 IV. 1878 9.6 0 0 630 3 V. 0 0 2153 11 884 5 VI. 11342 57.8 0 0 0 0 Totals 19636 100 19636 100 19636 100 241 Table 30. Percentages of Environmental Strata Per Site Per 1 Km Catchment Rings Baum Gartner Water Plant Catchment Strata Ha % Ha % Ha % Ring 1 Km I. 471 60 745 95 188 24 II. 288 38 20 2.5 393 50 III. 26 3 20 2.5 16 2 IV. v 0 0 0 0 188 24 V. 0 0 0 0 0 0 VI. 0 0 0 0 0 0 2 Km I. 785 33 1547 66 118 5 II. 337 14 745 32 353 15 III. 1179 50 24 1 0 0 IV. 0 0 41 2 1179 50 V. 56 2 0 0 0 0 VI. 0 0 0 0 707 30 3 Km I. 1002 26 3534 90 982 25 II. 177 5 353 9 471 12 III. 1963 50 40 1 0 0 IV. 0 0 0 0 511 13 V. 785 20 0 0 0 0 VI. 0 0 0 0 1963 50 4 Km I. 1282 23 3629 66 1375 25 II. 916 17 1665 30 458 8 III. 2200 40 204 4 0 0 IV. 0 0 0 0 0 0 V. 1100 20 0 0 0 0 VI. 0 0 0 0 3665 67 5 Km I. 409 6 2356 33 1178 17 II. 327 5 1826 26 884 13 III. 6121 87 1414 20 0 0 IV. 0 0 589 8 0 0 V. 212 3 884 13 0 0 VI. 0 0 0 0 5007 70 242 (59%), with 20% Stratum I, and about 10% each of Strata II and V. Strata IV and VI are not included in the catchment. Stratum I accounts for the largest share of Gartner's catchment at 60%, followed by Stratum II at 23%, and less than 10% each of Strata III, IV, and V. Strata VI is also not present in this catchment. Comparing strata within catchment rings, Water Plant has more of Strata IV and VI than either Baum or Gartner except in the first ring (Table 30, Fig. 25). Unlike these sites, it lacks Stratum V in all rings and has almost no Stratum 3. It shares similar percentages of Strata I and II with Baum. Differences also exist between the two early Late Prehistoric sites (Table 30, Figs. 26 and 27). These sites are remarkably similar in structure, artifact and tool classes, but have different catchments. Stratum I steadily decreases for Baum from Ring 1 to 5, but remains extremely high for Gartner throughout all rings. Gartner exhibits a nearly constant proportion of Stratum II within the catchment rings, while the same stratum percentages are more uneven for Baum, generally decreasing from Rings 1-5. Gartner also has a much lower incidence of Stratum III which only occurs in any appreciable frequency in Ring 5, while this stratum is very high for Baum beginning with Ring 2. Very little of Stratum IV is present in the Gartner catchment; Baum has none. Neither site has Stratum 243 VI. Baum does have an appreciable amount of Stratum V present from Rings 3-4, but Gartner has only a small percentage in Ring 5 at the edges of the catchment. Voss, the Transitional Late Prehistoric site, has a low frequency of Strata I and II which occur only along the Big and Little Darby Creeks in narrow margins (Fig. 28). Zones of Stratum III ring the creeks. Elsewhere throughout the catchment circle, away from the immediate environs of the creeks, Stratum VI predominates. The Ety site is similar to Water Plant in structure and topographic location. It is located on the eastern edge of the Scioto drainage. In spite of its similarities to Water Plant, it represents a different catchment pattern (Fig. 29). Their greatest similarity lies in their immediate location on a bluff-edge. Ety's catchment is characterized by narrow bands of Stratum I and even less of Stratum II along the major streams (Walnut and Poplar Creeks). Some bands of Stratum III surround stream feeders which crosscut the remaining catchment area of Stratum IV. The catchment here is basically a dissected upland with many steep ravines with both permanent and intermittent streams, while the Water Plant site has a much greater expanse of relatively flat uplands within its catchment. Vegetation types, when correlated with environmental strata, indicate high frequencies of oak-hickory resources within the over-all catchment of the Water Plant site where 244 Strata IV and VI occur in all rings (Table 31). This association is represented in Rings 3 and 4 at Baum and in Rings 2 and 5 at very low percentages (less than 10%) for Gartner. Table 31. Vegetation Types By Environmental Strata Strata Vegetation Type I. Bottomland hardwood, sycamore/willow on floodplain II. Elm, maple, black walnut III. Beech-sugar maple, pine S- slope, hemlock N-slope IV. Yellow poplar-white oak, northern red oak, hickories, oaks, basswood, white ash, elm, black walnut, butternut, red maple V. Scarlet oak association, white oak, chestnut, black oak, hickory, blueberry VI. Upland oaks— oak-hickories, beech-sugar maple, pine on slopes Mills (1904, 1906) noted that hickories were abundant in pits at Baum and Gartner. Black walnuts (occurring in Strata II and IV) were found in pits and hearth ashes at Baum and in small quantities at Gartner. Butternuts (Stratum IV) and hazelnuts also occurred in small 245 quantities at Baum and Gartner, while small quantities of chestnut (Stratum V) were recovered at Baum. Other wild plant types found at these two sites include great quantities of pawpaw seeds, as well as some wild red plum. Grape seeds were found in a few pits in small numbers at Baum. Paleoethnobotanical evidence from the Water Plant site based on seven flotation samples from six features (all pits) reveals the presence of hickory, white oak group and red oak, and black walnut, as well as sugar maple, ash, elm, and Gvmnocladus or Kentucky coffeetree at the site. Starchy members of the Eastern Agricultural Complex like maygrass, goosefoot, and erect knotweed make up 91% of the seed assemblage with maygrass predominant. Squash, blackberry, and sumac were also present (Wymer 1986). The nearby Late Woodland site of Zencor, located on the Scioto River, also yielded carbonized seeds which included maygrass, goosefoot, knotweed, blackberry, sunflower, and pigweed (Wymer 1985), in addition to the nut types listed for the Water Plant site. Estimates of faunal productivity of the catchments were not attempted. Location and availability of local plant resources would have supported game species within and beyond each catchment. That the environment around Baum and Gartner had plentiful and varied faunal resources is clear in Mills' site reports (1904, 1906). The same situation is expected to have held for the Water Plant site. Easily cultivable, productive soils are abundant within the catchments of Baum and Gartner, decreasing from 60% to 6% in Rings 1-5 at Baum and remaining much higher throughout Rings 1-5 at Gartner. The Water Plant site has similar soils at 25% in Ring 1, 3, and 4, reduced to about 17% in Ring 5 (Table 32). All sites are located in close proximity to major watercourses that could provide more than food resources like transportation routes for trade and other social contacts. Flint Ridge flint and marine shell are at least two exotic materials noted at Baum and Gartner which were obtained from beyond a 5 km or even 10 km catchment. Mills (1904) observed that caches of prepared clay (already tempered with broken shells and crushed quartzite) were the same as the clay into which pits had been dug. Coarse grained sandstone identical to that formed into artifacts was also available from the top of nearby hills southwest of Baum (Mills 1906). A majority of all stone tools at Gartner, Baum, and the Water Plant sites were made from locally available flints, slates, and granites. Thus, the catchments were productive along a number of dimensions in terms of edible and nonedible resources and raw materials. 247 Table 32. Soil Types Associated with Environmental Strata Strata Soil Types by Sites Baum Gartner Water Plant I. Genesee Genesee Genesee Ross Ross Medway Ockley Ockley Crosby II. Genesee Genesee Eldean Fox Ross Warsaw Miamian Ockley Cana Udorthents III. Crosby Miamian Miamian Cana Latham Celina IV. Latham Eldean Ockley V. Avonburg Kendallville Rossmoyne Crosby Miamian VI. Crosby Kendallville Kokomo Celina Eldean Miamian 248 In a similar examination of Late Woodland sites in the Lower Sangamon Vallej in Illinois, a dichotomous settlement pattern with sites located either on a slope (forest) or oriented to bottomland has been described (Roper 1979). This pattern was supported by evidence from other areas of Illinois and in the Wood River Terrace area of the Mississippi River, where Early Bluff sites occur on bluff edges and Late Bluff sites have a bottomland location (Roper 1979). This suggests a temporal trend in settlement pattern similar to that postulated by Graybill (1981) for West Virginia. Roper cautiously interprets this pattern as possibly representing "seasonally interchangeable occupation of the two types of sites" (1979:140). In the present sample of sites for which locational data was available, nearly three times as many sites with ceramics and lithics present (possible habitations) from the Late Woodland-early Late Prehistoric period are located in the valley (a ratio of 2:1, terrace/bottom to floodplain) as opposed to the uplands where 100% of such sites are on a bluff-edge (Table 33). Sites with lithics only are located at an almost identical ratio between valley and uplands. However, valley sites are almost evenly split between floodplain and terrace locations, while upland sites are scattered among bluffs, general uplands, ridgetops, and slopes. 249 Table 33. Site Location Correlated with Artifact Categories Sites w/ Sites v/ Ceramics & Lithics Lithics Only Total Location: Uplands 11 (100% bluff) 12 23 Valley 10 Floodplain 15 Floodplain 25 22 Bottom & Terrace 17 Terrace 39 Totals 43 44 87 The question of whether this pattern in the central Scioto drainage represents seasonally differentiated and utilized sites as Roper (1979) suggests for her data, or more long-term temporal differences as Graybill (1981) suggests for West Virginia is taken up in the last chapter. It does point to intensive use of valleys and uplands (particularly bluffs) both for habitation and more temporary activities throughout the time period under consideration. Catchments are observably different between Late Woodland and Late Prehistoric sites. It has already been shown that Late Prehistoric sites are more likely to occur on a terrace than on the valley floor or uplands. If they are in the uplands, they are on bluffs. Yet, Late Prehistoric site catchments are not replicas of each other. 250 This again argues against a population-intrusion model of culture change. One would predict that groups coming into the area bringing an established way of life would create a set of nearly identical sites. What is seen in the central Scioto Valley is a more dynamic picture of change from early Late Woodland through early Late Prehistoric. This change can be more easily explained by a shift in economic and social foci than by an invading culture. CHAPTER X SUMMARY AND CONCLUSIONS Introduction The larger question of this research is perhaps the most difficult to deal with— how one might explain changes in the archaeological record over a period of 400-500 years which resulted in different cultural patternings across the same physical environment. Questions of origin argue that the explanation can be pinpointed as external to the system of cultural development, as internal, or as a combination of both. In this case, external factors, such as acculturation or population intrusion are self-explanatory as the ultimate causes of changes. If, however, the changes in the archaeological record reflect internal changes in cultural developments occurring in the central Scioto drainage, then further explanatory avenues must be explored. Processes which could operate in local development might produce changes in the way groups arrange themselves over the environment in space and time, for example, dispersed versus nucleated communities, and where such communities are found along a continuum of mobile to sedentary settlements. The following discussion focuses 251 252 upon recent research which addresses the definition and recognition of these factors in the archaeological record. Recent Theories on Nucleation and Dispersion In its broadest descriptive sense, the settlement pattern can be defined as the distribution of sites in a regional landscape (Willey 1953). The human activities which create a settlement pattern are generated by rules of behavior, often referred to as the settlement system (Winters 1969). In a study of settlement pattern change, the fundamental question is how have the behavioral rules altered to produce a new pattern of site distribution. A simple distinction in kinds of settlement pattern that has important meaning for the organization of human groups is that between dispersed and nucleated settlements. This distinction is used in the present study, following the lead of Fuller (1982) who examined the record of settlement pattern change from dispersed to nucleated in the Upper Ohio Valley from 500 B.C. to A.D. 1500. Central to Fuller's work is the concept of the human community, which in archaeological terms is defined as "the arrangement of artifactual materials remanent of the domestic occupation used by a group of economically, socially, and politically related people (Fuller 1982:1). It is recognizable by the presence of a cohesive stylistic assemblage from a single occupation or set of contemporaneous occupations. A community may take two forms: dispersed or nucleated. In a dispersed community, members are separated into very small settlements (hamlets or farmsteads) which are located at some distance from each other within a local area. A local area refers to a culturally homogeneous area outside an occupation easily accessible to human travel. By contrast, a nucleated community is one in which all members reside within one contiguous settlement. Fuller's purpose is to explain the observed change from dispersed to nucleated communities which occurred in the West Virginia panhandle of the Upper Ohio Valley during the Woodland and Late Prehistoric periods. The explanatory potential of two models are explored: the growth model and the nucleation model. The growth model posits that nucleation occurs as single settlements increase in size from hamlets to nucleated communities. In the nucleation model nucleated settlements originate when the members of a community who were formerly dispersed in hamlets aggregate. Lacking evidence of steady increase in site size through time in any part of his study region, Fuller rejects the growth model. The archaeological record contains evidence, at different times in different parts of the study area, of the sudden appearance of a few, large, nucleated villages where previously there had been numerous small, dispersed hamlets. Fuller argues that this kind of change can be explained best by the nucleation model. 254 Four factors are described which activate the models: defense, diffusion, population growth, and agglomeration. Arguing that nucleation "represents an adaptation towards increasing efficiency” (Fuller 1982:278), Fuller identifies two changes which must occur (separately or together) to allow nucleation: 1) a change in the environment towards greater food productivity, and 2) a change in technological adaptation which allows production of the same or more food and nonfood resources from a smaller space. Such a perspective has frequently been tied to changes in subsistence. For example, a dispersed community has been associated with extensive agriculture while nucleated ones are associated with intensive agricultural systems (cf. P.E.L. Smith 1972). The settlement system is, of course, closely tied to economic behavior: the distribution of settlements in space reflects utilization of resources by a group. The length of time a site is occupied during the year is also tied to resource use. Thus, hunting camps or quarry sites are transient, single purpose sites occupied for a short period of time. In contrast, societies dependent upon agricultural food production may occupy a site year-round, year after year. The following section discusses this aspect of settlement systems. Recent Theories on Sedentariness Beardsley et al. (1956) see human communities ranged 255 along a continuum from completely mobile to completely sedentary. A seven-part classification was utilized to describe possible settlement systems based on this concept. The patterns described which are of interest to the present research are the Central-Based Wandering (CBW), the Semi- Permanent Sedentary (SPS), and the Simple Nuclear Centered (SNC) patterns. The CBW pattern is based on a central base establishment at which a group spends part of the year. The remaining months are spent moving about, presumably within a defined territory, as the result of adjustment to one of three subsistence bases: 1) storable or preservable wild foods, 2) locally abundant resources, and 3) incipient agriculture. The SPS pattern refers to villages established in successive locations, each for a period of years, based on simple [extensive?] agriculture, while the SNC pattern indicates a permanently occupied, patterned center with or without satellite communities, dependent upon cultivated foods and having an increased population. These levels are inter-related with each growing out of a previous one and into one which follows, although development is not necessarily unilinear. Binford (1980, 1982) discusses the concept of mobility versus sedentariness, bringing up the point that mobility has less to do with simple abundance of food resources than with resource distribution. Mobility can be separated into 256 tactical and logistical forms. Tactical refers to the accomodation of a settlement system to a broader environmental geography, and logistical mobility refers to task groups which range into the environment from a residential base. Mobility patterns are not immutable, however, and can change from season to season within a group. A distinction between such mobility patterns is addressed by Eder (1984). He points out that mobility is manifested at the individual level in the context of social organization. Thus, several patterns can be present in a single human group. Sedentariness, on the other hand, occurs at the level of the social group. These are not necessarily polar concepts at the ends of a single continuum. Mobility can still be a behavior of individuals within a sedentary system. The question Eder poses is how does the amount and kinds of sedentariness change with the advent of sedentariness? Possible reasons for a shift from more to less mobile settlements are 1) change in the availability of resources and 2) a reduction of mobility due to factors such as territorial confinement, technological innovations, or the fact that sufficient material resources become available (Harris 1977). These factors may result from natural or human-made changes in the environment, from innovations in 257 subsistence technology, or from changes in the perception of acceptable food resources. A recent summary paper on sedentariness takes as its starting point a definition of sedentariness as "settlement systems ... in which at least part of the population remains at the same location throughout the entire year" (Rafferty 1985:115 from Rice 1975:97). This definition allows for individual mobility, where sites created by members of a community who are temporarily away from the major settlement would be considered part of the over-all settlement pattern. Secondly, it also allows for different degrees of residential permanence from one year to infinity. According to Rafferty, the crucial factor in the development of sedentariness is not resource abundance (cf Fuller 1982), but resource availability (Binford 1980, 1982; Harris 1977). This factor enables efficient resource procurement without constant settlement relocation. The ultimate causes of sedentariness, similar to those proposed by Harris (1977) and including demographic factors as suggested by Smith (1972), are population growth, territorial constriction, and environmental deterioration. The proximate cause is resource stress with a number of possible responses by a group: population limitation, emigration, and technological/ organizational change. Unlike Harris (1977), however, environmental change 258 operates in only one direction in this model. It is conceivable, as Salzman (1980) points out, for change to occur in any direction. Thus, logically, increased resource availability could arise from an amelioration of environmental conditions. Rafferty's model is tied to Cleland's (1976) development of focal and diffuse economic models in which the productivity and diversity of resources result in four broad settlement patterns: dispersed mobile, dispersed sedentary, nucleated mobile, and nucleated sedentary. The model is flexible in that both a diffuse or a focal economy can foster a nucleated, sedentary system based on the response of a group to resource stress. Based on these considerations, Rafferty proposes a set of material correlates by which sedentariness can be recognized in the archaeological record. Primary among these is settlement pattern change in which a change occurs in the distribution of settlements across the environment and new kinds of settlements appear. Comparable sites in such a settlement pattern should have similar storage features and residential structures, while special purpose sites should be located in a variety of microenvironments and be farther from habitation sites. Also, the appearance of a number of other changes over a short period of time can be an indication of sedentariness. These kinds of changes include house form, house shape, community 259 planning, and artifacts (for example, mounds, kivas, pottery, storage facilities, and heavy artifacts) which become more varied and numerous. Other indications are an increase in site size and depth (midden development), a change in the intra-site distribution of artifacts, year- round seasonal indications of site use, and the location of the site relative to certain environmental features, such as a year-round source of water. Longer term changes which could accompany sedentariness include population growth, an increasingly complex political organization which acts to decrease conflict and to regulate and organize trade, increasing site size, and increased ritual activity. Salzman (1980) defines sedentism as the settled, immobile location of the household (what Fuller 1982 calls the 'domestic unit') during the annual round of productive activities. This is similar to that which is used by Rafferty (1985), but he distinguishes this from sedentarization which is any change on a continuum between nomadism to sedentism. While he is specifically interested in the analysis of nomads, Salzman makes an important point when he discusses the common assumption that sedentarization is an irreversible and absolute process. Change, from his perspective, comes about in terms of an 'adaptation and response' model. A shift can occur from one available pattern to another in a society in response to changing pressures, constraints, and opportunities that are internal, external, or both. Such a shift "is guided by ends and means institutionalized in the society" (Salzman 1980:14), where the means lie in alternative behavior patterns, organizational forms, and ideologies which are already part of the society as deviant minorities, operational generalizations, and asserted ideologies. Operational generalizations refer to the existence of a variety of types and combinations of ongoing activities common to a community. Asserted ideologies imply organizational frameworks held and expressed by all members of a society, which are sometimes acted upon and sometimes not (Salzman 1980:4). In essence, the culture can adapt and change patterns of behavior (including mobility) in response to such factors as those listed by Harris (1977) through alternative courses of action which are extant components of their behavioral repertoire. The problem of concern in the present study, then, lies in the nature of the differences and similarities which existed between communities located in the central Scioto drainage from A.D. 500 to A.D. 1250. Are differences the result of external forces acting upon and changing local groups? Can similarities be traced to a common cultural pattern that has evolved and changed in situ ? Is it possible to identify those factors responsible for such changes and the direction they took? 261 Can the over-all character of the archaeological record of the central Scioto drainage during the Late Woodland/Late Prehistoric period be explained in terms of the behavior of people who inhabited this region? Chronology and Site Function The purpose of this research has been to explore the possibility that early Late Prehistoric cultural groups of the central Scioto drainage were derived from local Late Woodland populations, through a comparison of Late Woodland and Late Prehistoric sites from the region. Preliminary analyses supported such a position, contrary to Prufer and Shane's (1970) hypothesis that population-intrusion best explained the differences between Late Woodland and Late Prehistoric in this region. Graybill (1981), for example, had expressed the opinion that the origins of the Baum tradition would be found in the Late Woodland era, based on similarities in ceramics, village plan, and the presence of maize agriculture. Rafferty (1974) and Rigg (1977), in independent efforts, found that no group of ceramic styles appeared in the early Late Prehistoric period that could be traced as a complete entity elsewhere, suggesting therefore that those styles diagnostic of the early Late Prehistoric originated either in preceding Late Woodland populations, within early Late Prehistoric itself, or from a combination of these two factors. 262 Prufer (1967), for example, had noted the presence of non-Fort Ancient, untyped shell-tempered ceramics at Chesser Cave, an otherwise Late Woodland site. He interpreted their presence as a result of the acculturative effects of intrusive Mississippian populations on a basic Late Woodland group living in the hinterlands. In the same discussion he also allowed for the possibility that some of the roots of early Fort Ancient were to be found in local Late Woodland cultures. Given this statement, it seems equally plausible that the untyped shell-tempered sherds at Chesser Cave and those found elsewhere in similar contexts may represent a local Late Woodland ceramic type transitional between Late Woodland and early Late Prehistoric contexts. The sites of Blain, Voss, Howard Baum, and Enos Holmes may span the transition from Late Woodland to early Late Prehistoric in the central Scioto drainage. As Schambach (1971) has pointed out, Blain contains Late Woodland materials which are glossed over in the site report. Without some knowledge of the extent of this material, it is difficult to accept Blain as the earliest manifestation of a pure Late Prehistoric site in this region. Temporally, the C-14 dates put it early, but the earliest date from the site comes from a pit with Late Woodland ceramics. It may be that a Late Woodland component is present. The site may have been occupied by two different 263 cultural groups— Late Woodland then Late Prehistoric. Or, the Late Woodland material, coupled with shell-tempering and basic Baum stylistic ceramic traits, may represent a culture in transition. In fact, shell-tempered ceramics only predominate in one of twenty-nine total pits at the site. Again, the typical early Late Prehistoric site structure of Blain— concentric rings of domestic activities about a central plaza— has not been adequately demonstrated as parts of only two structures have been recovered (another three postmolds have been interpreted as a third structure: see Prufer and Shane 1970:33). The fact that the two structures which are documented were relatively undisturbed with no indications of rebuilding suggests, as Prufer and Shane interpret it, a short-lived occupation. Maize, while present, is not overwhelmingly represented (10% of the pits). Thus, the evidence does not contradict an alternative explanation of this site as transitional Late Woodland-Late Prehistoric. Brose and White (1983) feel that a similar transition has been documented southwest of the central Scioto drainage. Enos Holmes and Voss, also C-14 dated as very early in the Late Prehistoric sequence, also contain a mixture of Late Woodland and Late Prehistoric traits. Voss, for example, yielded 3943 sherds which include typical Late Prehistoric stylistic attributes as curvilinear and rectilinear guilloches, chevrons, oblique lines, and punctates. Tet, only nine sherds (0.2%) have shell- tempering; the remaining 3934 are granitlc-tempered (Baby, Potter, and Sawyer 1967:9-10). Late Prehistoric stylistic motifs occur, then, in sites with extremely low frequencies of shell-tempering. A similar interpretation has been made for the Pipeline site (A.D. 1070), which "was inhabited by Late Woodland groups whose ceramic assemblage had begun to adopt some attributes which later came to characterize the regional Fort Ancient complex" (Brose and White 1983:14). The frequency of shell-temper has been suggested as a temporal marker within the early Late Prehistoric period. While this does seem to hold true, seriation of sites with shell-temper suggests a gradual development of this trait in the region, rather than wholesale importation from outside the region. The percentage of shell-tempering among early Late Prehistoric sites of the central Scioto drainage increases gradually from 0-30% over a span of approximately 300 years, never approaching the percentages calculated for Emergent Mississippian sites at a time comparable to the beginning of the Late Prehistoric in central Ohio. The gradual increase in frequency better approximates independent invention or diffusion of an idea that reached this region and spread gradually. It has been argued recently that shell-temper may serve a functional purpose (creating a more cohesive paste and more even 265 heating surface) and that its gradual increase parallels changing subsistence practices in the early Late Prehistoric (Morse and Morse 1983; Bronitsky and Hamer 1986). Another trait which has served as an ambiguous chronological marker of this time period is the triangular point. While claims have been made (Seeman and Munson 1981; Baby, Potter, and Saurborn 1968) that these points can be differentiated into Late Woodland and Late Prehistoric types based on their measurements, the data from the central Scioto drainage does not support such a position. Either a better, larger sample of points is needed, a better system of analysis is needed, or a single type exists which does not change sufficiently in its metric attributes from Late Woodland to Late Prehistoric to allow fine-grained temporal discrimination. In a more positive light, this simply points to another line of continuity between Late Woodland and early Late Prehistoric groups in this area. However, successful seriation of sites in the drainage has been done based on base-shape of triangular points, such as Graybill (1981) has demonstrated for sites on the eastern periphery of this region. Three basic points were addressed concerning the magnitude of differences between cultural groups in this time period in the central Scioto drainage: 1) similarities in subsistence economy, 2) differences between site 266 locations, and 3) differences in material culture. Discussion of these points will be based on preceding analyses of settlement pattern, artifacts, site structure, and site catchment for sites within the central Scioto drainage. Two of the most obvious categories of diagnostic material— ceramics and points— have been dealt with in the above paragraphs. The results of a functional analysis of artifacts will be considered here. The Late Woodland tool assemblage at the Water Plant site differs primarily from the Late Prehistoric ones in its over-all higher percentage of chipped stone tools such as bifaces and scrapers, although Baum has nearly as many scrapers and Gartner a similar percentage of bifaces. Water Plant does have a much higher percentage of flake tools than either Late Prehistoric site, but, like the lack of bone tools at Water Plant, this may be a reflection of factors other than prehistoric activity like differential preservation and/or curation of artifacts. Water Plant also has a slightly higher proportion of ground stone tool functional classes than the Late Prehistoric sites. Over-all, the Late Prehistoric functional types seem slightly more varied with the presence of different chipped stone classes, triangular point classes, and new ground stone classes. The proportion of groundstone and triangular point types dominates all other stone tool 267 functional classes in these assemblages. Water Plant, on the other hand, is dominated by groundstone tools, flaked tools, and bifaces, in that order. Thus, clear differences exist between the functional assemblages of Late Woodland and early Late Prehistoric sites, but these differences do not argue for the complete replacement of one set of activities by another. They also indicate differences within the early Late Prehistoric, as the assemblages of Baum and Gartner by no means duplicate one another. These tools represent the results of daily activities performed at these sites by prehistoric populations. They probably best reflect economic activities and support the conclusion that these were not specialized, temporary activity sites, but were full-time functioning communities. An examination of site structure and site catchments may shed some light on economic and/or social changes mirrored in tool differences. Settlement Patterning The Water Plant site is similar to other Late Woodland sites in its size, presence and shape of structures, lack of hearths interior to structures, and lack of a defined intravillage pattern. Late Woodland sites in the central Scioto drainage lack mounds, but most have a ditch and/or earthwork enclosure. On the other hand, early Late Prehistoric sites are proportionately larger than the Late Woodland sites, increasing in size over time. Most of 268 these sites exhibit a planned site structure, having a plaza and a burial mound with occupational zones ranged about these and containing evidence of structures which have interior hearths. Pits, midden deposits, and village burials are also present. None have associated ditches or earthworks and none have palisades. Both the early Late Woodland and early Late Prehistoric sites represent nucleated communities, the Late Prehistoric sites being larger and more complex in structure. Support for nucleation of Late Woodland sites comes from their spatial distribution, size, and associated features. In a small portion of the total site, Water Plant has yielded evidence of several structures. Surface distributions of tool classes across the site indicate substantial occupational areas. Thus, this site does not appear to be a single segment of a dispersed community whose members are distributed into hamlets or individual farmsteads located at some distance from each other within the local area (Fuller 1982:2). Rafferty (1974) suggests a change from a dispersed to a nucleated settlement pattern occurs in this period, for example, at Lichliter. Furthermore, the Water Plant community, as is common with other Late Woodland sites, is spatially bounded by a ditch and/or earthwork enclosure. This feature serves to emphasize the nucleated nature of the site. Also, the site is not part of some small, clearly associated cluster of 269 sites which would be expected if a dispersed community pattern was present. This pattern is less clear for those sites which were earlier designated as possible Late Late Woodland— DECCO, Ufferman, W.S. Cole, and Harness-28. While hampered by the lack of secure information concerning these sites, the problem may be compounded by the tendency to assume that all Late Woodland sites must be part of a single, redundant pattern. For example, if early Late Woodland sites like Zencor and Water Plant are nucleated, 2-3 ha in size, located on bluffs, and are enclosed by a ditch or earthwork feature, then so must DECCO and other sites reflect this pattern. They are, in fact, much smaller— on the order of 0.10 ha in size. Ufferman and W.S. Cole cluster in Delaware County on the northern end of the Scioto drainage, looking suspiciously like elements of a dispersed community. As Salzman (1980) points out, cultures adapt and change in response to internal and external changes in their world. Knowledge of previous adaptations is not lost upon the adoption of a new pattern of behavior, such as sedentarism. If the conditions that led to change are altered, a corresponding change may be evident in a culture, one which might very well mirror previous behavior patterns. Because nucleation in the early Late Woodland appears to develop out of previously dispersed Middle 270 Woodland communities does not mean that all subsequent cultural developments must build upon this pattern and never change. Given sufficient incentive, a return to a dispersed community pattern might be entirely plausible. To think otherwise is to express a unilinear notion of cultural evolution, to suggest that inherent in change is a rigid unidirectionality. In contrast, nucleation of early Late Prehistoric settlements in the drainage has been widely accepted. The size and structure of the Early Late Prehistoric settlements indicates a face-to-face arrangement of members within these communities. While Transitional Late Prehistoric sites are smaller, their site structure parallels that of the larger sites, indicating that the difference between them is one of scale. These communities, unlike Late Woodland ones, are oriented around a common plaza and burial mound— or, in the case of Baum, around a possible communal structure as well— which indicates a corresponding difference in social structure than that which characterized Late Woodland groups. According to Fuller (1982), nucleated, sedentary communities appeared in West Virginia around A.D. 900, rapidly developing out of previously dispersed settlements. Evidence from the central Scioto drainage supports a different pattern of development. Nucleated settlements are present in this region from early in the Late Woodland 271 period. The transition from a dispersed community pattern may have occurred earlier; present evidence places it more likely between Middle and Late Woodland. Thus, in terms of the temporal sequence developed earlier for the central Scioto drainage, Early Late Woodland sites were nucleated. Late Late Woodland sites may represent a period of flux in settlement patterns, with some— such as DECCO, Ufferman, and W.S. Cole— representing components of dispersed communities, and others nucleated— possibly Harness-28. Transitional Late Prehistoric sites, on the other hand, are again all nucleated, as are Early Late Prehistoric sites. This may represent the agglomeration of populations into larger, successfully nucleated communities once again. Seasonal or Sedentary Communities? In order to understand more fully the changes which occurred over time in the Scioto Valley, the question of whether or not the Late Woodland and Late Prehistoric communities were sedentary is explored. Nucleated communities, like their dispersed counterparts, may be part of a sedentary or a seasonal settlement pattern (Fuller 1982, Rafferty 1985, Yerkes 1986). Rafferty argues that the impetus for the development of a sedentary settlement pattern is resource stress caused by factors like population growth, environmental deterioration, or territorial constriction, such that resource availability 272 and not simply resource abundance, becomes a motivating factor in human behavior (Rafferty 1985:119). Resource availability can be increased by intensifying production in one microenvironment, by expanding into other, richer zones, by linking microenvironments with -trade networks, and by locating a settlement on the boundary of several microenvironments. These actions serve to concentrate resources efficiently about a site without the need for constant population relocation. In a discussion of the direct material consequences or archaeological correlates of sedentariness, Rafferty (1985) points out that when a group settles down, a change in the distribution of sites across the environment can occur. New kinds of sites may be apparent over a short span of time, with differences in such characteristics as house form and shape, community planning, and associated features and artifacts. One would expect indications of year-round seasonal occupation of sites to occur with all sites having similar storage features. Special-purpose sites, if such exist, would be located in a variety of microenvironments and at a distance from the habitations. The above points can be considered in terms of the central Scioto drainage during the Late Woodland to Late Prehistoric periods (Table 34). The nucleated communities of the early Late Woodland period could be part of a seasonal or a sedentary settlement pattern. The same 273 options hold for sites of the other periods. Does a transition from one to the other occur through time? Information relative to tool classes, site location, and site catchment is drawn upon in consideration of these alternatives. Changes in the distribution of sites are evident from the Late Woodland to the Late Prehistoric. From Early Late Woodland to Late Late Woodland the proximity of sites to different microenvironments also occurs, with sites moving from bluffs down into the valleys— to either floodplains or terraces (cf Roper 1979). All Early Late Woodland sites are located on bluff-edges which are in close proximity to both valley and upland resources. Late Late Woodland and Transitional Late Prehistoric sites are predominately located on terraces or floodplains, while all major Early Late Prehistoric sites are located on terraces. Upland Late Prehistoric sites are restricted to bluffs, unlike Late Woodland sites. During the Late Woodland, 34% of all sites are located in the uplands. This percentage decreases by almost half for Late Prehistoric sites, with the biggest decrease (8%) occurring in undissected uplands. New kinds of sites are evident both in the restricted upland location of Late Prehistoric sites which suggests a specialized activity and in the appearance of a new site structure. Settlements illustrate a number of differences in terms of community planning. While elaborate communal 274 Table 34. Archaeological Correlates of Sedentarism for the Late Woodland-Late Prehistoric Period Correlate ELW LLW TLP ELP House-shape Circular ? Circular Circular Village plan Bluffedge None Concentric Concentric w/ditch rings rings Communal structures Ditch/ None Mound( Mound, enclosure plaza plaza Pottery Yes Yes Yes, shell- Yes, shell- temper temper Storage Pits Deeper Deeper Deeper pits pits pits Heavy artifacts Yes Yes Yes Yes > #, kinds artifacts Bone tools Bone, > Bone, > Bone triangular > tr. pts, > tr. pts. points fishhooks > fishhooks Village size Small Smaller Larger Largest 275 structures such as earthwork and mound groups are no longer constructed in the Early Late Woodland, the nucleated village is found behind a ditch and/or earthwork which was probably built with community labor. This structure is not evident in either the Late Late Woodland or Late Prehistoric. Sites appear to be much smaller and possibly dispersed during the Late Late Woodland, at least in the northern portion of the drainage. In Late Prehistoric settlements, energy is expended in the construction of habitation zones within the community which include a mound in all cases and possibly a communal structure as well (beneath the mound at Gartner and within the village at Baum). Material changes are also evident in associated features and artifacts, which are more numerous and varied. An analysis of tools at the Late Woodland site, Water Plant, indicates that it was not a specialized activity site, but that a wide variety of activities took place. Density of artifacts also suggests an intensive occupation. The differences in ground stone tools which exist between this site and the early Late Prehistoric ones probably indicates the processing of different resources. Intensified production (from Middle Woodland to Late Woodland) is indicated by the increase in starchy components of the Eastern Agricultural Complex recorded for Late Woodland sites (Wymer 1986). Comparison with other 276 Late Woodland sites such as the Rench site in central Illinois (King 1986) supports the importance of the Eastern Agricultural Complex as a component of the Late Woodland subsistence system. Water Plant overlooks an extensive bottom on the opposite side of Big Walnut Creek which has yielded evidence of extensive prehistoric utilization. No evidence of maize utilization is present from any of the Early Late Woodland sites, however. Seasonal indicators of occupation are not abundant at Water Plant. Animal bone is present in some features, but has not been analyzed. Seeds like maygrass and goosefoot, however, indicate an early spring to summer occupancy of the site. In addition, hickory and other kinds of nuts have been identified. Available in the fall, these are storable for other seasons of use, and two types of pits are present at the site. Thus, seasonality indicators from the paleoethnobotanical evidence make the strongest argument for year-round occupancy of this site. Similar data from the Early Late Woodland site of Zencor corroborates this view. Possible structures at the site do not exhibit the degree of interior occupational debris accumulated within structures at Baum and Gartner. This, coupled with the lack of interior hearths could suggest that these structures were not lived in year-round. Other Late Woodland sites, however, .1.. 277 such as rockshelters (Chesser Cave) do have interior hearths and better seasonal indicators of use. If Late Late Woodland sites were dispersed, then a return to nucleation may have provided increased resource availability with a reduction in energy costs needed to move the population for resource procurement. Features and artifacts in Late Prehistoric sites reflect a highly sedentary way of life. New artifact types like bone fishhooks and shell hoes indicate an increase in resource availability through technological innovation. New feature types like deep pits become a dominant feature of these sites. New resources like maize appear, while the proportion of starchy seeds decreases (Wagner 1983). Maize is clearly not an economic staple in Transitional Late Prehistoric sites, but it increases in Early Late Prehistoric sites and may represent an even greater intensification of production in one segment of the economy— horticulture over hunting and gathering. The concomitant shift in site location could be explained by such a change in subsistence activities. Expansion into a richer zone or an increased trade network are not evident. Another line of evidence for a potential increase in resource availability is from site catchment analysis. As indicated previously, catchments are dissimilar for the Late Woodland and Late Prehistoric sites. Water Plant has much more upland area within its catchment, no ridges, and little sloping areas with the obvious exception of the bluff upon which the site is located. Included within the catchment, however, is a segment of the Scioto River with its floodplain and terrace environments, so that this site had easy access to a variety of resources such as oak- hickory forests and soils with agricultural potential (easily worked, available moisture, and fertile) within three of five 1 km catchment rings. A comparison between this site and another Late Woodland site, that of Ety, reveals a difference in catchments. Ety's is similar to the Water Plant site in the blufftop location of the site, but otherwise consists largely of uplands dissected by numerous small streams which have narrow floodplains and steep slopes. A closer examination of catchments reveals that the three Late Prehistoric sites are not identical and do not provide a uniform contrast to the Late Woodland sites. However, the catchment for Baum encompasses all environmental strata except for generalized, undissected uplands which characterizes a large portion of the Water Plant catchment. This same stratum is also lacking at Gartner, yet predominates in the catchment of such Transitional Late Prehistoric sites such as Voss. Thus, the idea that resource availability increased from Early Late Woodland to Early Late Prehistoric is strongly supported and, over-all, the evidence supports an 279 interpretation of sedentary sites throughout this same time period. Conclusions What, then, is the nature of the transition from Late Woodland to Late Prehistoric in the central Scioto drainage? The Late Woodland has long been perceived as a period of cultural stagnation when compared to the glory of earlier Woodland cultures or to the 'progressive' Mississippianized cultures which replaced it. Yet, the evidence yields a far different picture. The Late Woodland period is far more complex than previously thought. Early Late Woodland appears as a logical development out of Middle Woodland, from dispersed to small nucleated villages consistently located on bluffs with a simple plan of households clustered behind an enclosing wall or ditch. An apparent shift occurs in the settlement pattern after about A.D. 800, during which blufftop settlements decrease while those in the valley increase. These Late Late Woodland settlements lack the enclosing structures of earlier sites and generally are much smaller in size. If the late placement of sites such as DECCO, Ufferman, Cole, and others can be accepted, this represents a decided change. If the dates for such sites are acceptable, then the archaeological record presents a gap devoid of human occupation in the region for a period of 280 several hundred years. This latter alternative Is highly unlikely, given the evidence for cultural and genetic (Robbins and Neumann 1972) continuity within this region from Late Woodland to Late Prehistoric. In certain material traits, these sites mark a transition from Early Late Woodland to Late Prehistoric. This transition is emphasized by the shift in site location. That is, Early Late Woodland settlements do not simply increase in size or move to other bluff locations by Late Late Woodland. The shift suggests the failure of the culture to maintain nucleated communities in the context of previous cultural, social, and economic systems. As Salzman (1980) points out, change usually occurs based on institutionalized behaviors that are already part of the cultural system. The fact that Late Late Woodland sites are located in different microenvironments indicates changing circumstances. Dispersion, noted above as a behavior practiced in the central Scioto drainage prior to Late Woodland, may also have been practiced on a different level during Early Late Woodland. Seasonal dispersion may have been an institutionalized behavior practiced by some members of these communities. The nature of Early Late Woodland villages argues for fluidity of social mobility among members. The unstructured village plan with its haphazard arrangement of structures indicates relative ease of movement of members into and out of the group. Such a 281 view is not inconsistent within a sedentary context, since the definition of sedentariness states only that some members of the community must remain in the settlement year-round. Late Woodland is followed in time by the Transitional Late Prehistoric, initiating the Late Prehistoric period which may be characterized in general by a settlement pattern of fewer, larger sites located overwhelmingly on terraces. Transitional Late Prehistoric sites have all of the cultural traits of the Early Late Prehistoric. Shell- temper appears, triangular points predominate, fish hooks, deep pits, and maize appear regularly, as does a structured community plan. These sites are, however, on the average smaller than their later counterparts and are more variable in their location— either floodplain or terrace on the valley floor. These Transitional Late Prehistoric sites are located in the same variable mix of microenvironmental zones as the Late Late Woodland sites and make the transition from dispersed to small nucleated sites once again. Tet, this obviously does not represent a return to the nucleated Early Late Woodland sites. A different pattern of nucleation is established, evident in the completely different village plan, and is much more successful with this same village plan enduring throughout the Late Prehistoric in the region for some 500 years. The success of nucleation is probably tied to a host of factors, some of which can be guessed at and others which might be tested given adequate archaeological data. Two important factors are most certainly changes in social order and in subsistence economy. The planned nature of the village implies greater rigidity in the composition of the community structure. Mills (1906) and Heilman and Hoeffer (1979) have observed clusters of features and artifacts (structures, pits, hearths, burials, and ceramics) that suggest distinct social groups within the village. These groups form a cohesive community bound together by activities focused upon centered communal structures. Subsistence becomes focused upon resources over which a certain amount of human control (through technological innovations and the presence of plant species adapted to the environment) is possible, ultimately leading to more consistent resource availability and more efficient resource procurement (Rafferty 1985; Reidhead 1981). Success of the system is apparent. A process of uninterrupted, local development occurs in the central Scioto Valley from A.D. 500 to A.D. 1250. An economy that uses more cultigens develops, linked to new patterns of social organization as reflected in changes in the settlement system over time. While striking differences characterize Late Woodland and Late Prehistoric, the explanation for such differences does not lie in a 283 traumatic upheaval and replacement of local groups by an intruding culture. The explanation may come, instead, from an examination of forces operating within the system to produce and sustain nucleated, sedentary groups over time. As discussed above, such an explanation will be partly grounded in subsistence changes. Given the limitations of the data used here, from early 20th century excavations and random CRM surveys, it is not possible to satisfactorily explore this problem. Why does nucleation appear after the end of Middle Woodland cultures? Why was it not sustained throughout the Late Woodland? Why does it reappear in a different form after A.D. 800? Only when systematic regional surveys are undertaken, when a cohesive program of selected (not salvage) excavation and survey projects is instituted will it be possible to directly address these issues. BIBLIOGRAPHY Ahmad, M.U. and G, Blackie 1966 Ground water. In Environmental analysis of central Ohio— an initial approximation, Volume JE, edited by D.M. Anderson and C.C. King, Chapter 5. Ohio Biological Survey, Columbus, Ohio. Ahmad, M.U. and K. Wood 1966 Surface water quantity. In Environmental analysis of central Ohio— an initial approximation, Volume edited by D.M. Anderson and C.C. King, Chapter 3. Ohio Biological Survey, Columbus, Ohio. Allman, J.C. 1957 A new Late Woodland culture for Ohio: the Lichliter village site near Dayton. Ohio Archaeologist 7(2) :59— 65. Anderson, Dennis M. and Charles C. King (editors) 1976 Environmental analysis of central Ohio— an initial approximation: map folio. United States Army Corps of Engineers, Huntington, WV. Baby, Raymond S. and M. A. Potter 1965 The Cole complex: a preliminary analysis of the Late Woodland ceramics in Ohio and their relationships to the Ohio Hopewell phase. The Ohio Historical Society, Papers in Archaeology 2. Baby, Raymond S., M.A. Potter, and A. Mays, Jr. 1966 Exploration of the O.C. Voss mound, Big Darby Reservoir Area, Franklin County, Ohio. The Ohio Historical Society, Papers in Archaeology 3. Baby, Raymond S., M.A. Potter, and B.L. Sawyer 1967 Exploration of the O.C. Voss village site, Big Darby Reservoir Area, Franklin County, Ohio. Ms. on file, Department of Archaeology, The Ohio Historical Society. Baby, Raymond S., M.A. Potter, and B.S. Saurborn 1968 Excavation of the Enos Holmes mound, Paint Creek Reservoir Area, Highland County, Ohio. Ms. on file, Department of Archaeology, The Ohio Historical Society. 284 285 Baerreis, David A. and Reid A. Bryson 1965 Climatic episodes and the dating of the Mississippian cultures. The Wisconsin Archaeologist 46:203-220. Baerreis, David A., R.A. Bryson, and J.E. Kutzbach 1976 Climate and culture in the western Great Lakes region. Midcontinental Journal of Archaeology 1:39- 57. Barber, M.B. 1974 Fort Ancient settlement patterns. Unpublished M.A. thesis, Department of Anthropology, Kent State University. Barkes, Beula M. 1982 An analysis of Late Woodland ceramics from the DECCO (33DL28), Ufferman (33DL12), and W.S. Cole (33DL11) sites: the Cole complex revisited. Ms. on file, Department of Archaeology, The Ohio Historical Society. Beardsley, Richard K., et al. 1956 Functional and evolutionary implications of community patterning. In Seminars in archaeology: 1955, edited by R. Wauchope. Memoirs of the Society for American Archaeology 11:129-157. Binford, Lewis R. 1979 Organization and formation processes: looking at curated technologies. Journal of Anthropological Research 35:255-273. Binford, Lewis R. 1980 Willow smoke and dogs' tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45(1):4 — 21. Binford, Lewis R. 1982 The archaeology of place. Journal of Anthropological Archaeology 1(1):5—31. Blank, John 1970 The Ohio Archaic: a study in culture history. Unpublished Ph.D. dissertation, Department of Anthropology, University of Massachusetts. Braun, E.L. 1964 Deciduous forests of eastern North America (second ed.). Hafna, New York. 286 Bronitsky, Gordon, and Robert Hamer 1986 Experiments in ceramic technology: the effects of various tempering materials on impact and thermal- shock resistance. American Antiquity 51(1):89-101. Brose, David S. 1982 The archaeological investigation of a Fort Ancient community near Ohio Brush Creek, Adams County, Ohio. Cleveland Museum of Natural History. Kirtlandia 34 Brose, David S. and N.M. White 1983 Recent data on Fort Ancient occupation in the Caesar Creek valley, southwestern Ohio. West Virginia Archaeologist 35(2):3-26. Carskadden, Jeff 1985 Fort Ancient in the central Muskingum Valley: 1985 perspective. Ohio Archaeologist 35(1):4-10. Chisholm, M.D.I. 1968 Rural settlement and land use: an essay in location. Hutchinson, London. Clark, C. and M. Haswell 1970 The economics of subsistence agriculture. MacMillan, London. Clarke, Wesley S. and J.E. Addington 1983 Phase I-III archaeological survey of the R0S-207- 0.00 project: demonstration of the effective use of multiphase cultural resource studies in early highway planning at an archaeologically sensitive locale. Ms. on file, Bureau of Environmental Services, Ohio Department of Transportation. Cleland, Charles 1976 The focal-diffuse model: an evolutionary perspective on the prehistoric cultural adaptations of the Eastern United States. Midcontinental Journal of Archaeology l(l):59-76. Cotterell, B. and J. Kamminga 1979 The mechaics of flaking. In Lithic use-wear analysis, edited by B. Hayden, pp. 97-113. Academic Press, New York. Cowan, C. Wesley, et al. 1985 Fort Ancient in the Miami drainages. Paper presented at the Annual Fall Meeting of the Ohio Archaeological Council. 287 Crane, H.R. and J.B. Griffin 1963 University of Michigan dates VIII. Radiocarbon 5:228-253. Crane, H.R. and J.B. Griffin 1968 University of Michigan dates XII. Radiocarbon 10: Dancey, William S. 1984 The 1914 archaeological atlas of Ohio: its history and significance. Paper presented at the 49th Annual Meeting of the Society for American Archaeology, Portland. Dancey, William S. 1985 Archaeological survey in central Ohio: the 1970s. Paper presented at the 50th Annual Meeting of the Society for American Archaeology, Denver. Dancey, William S., D.A. Wymer, and R. Waterworth 1986 Survey report for the Big Darby Creek Mounds Project. Ms. on file, Department of Anthropology, The Ohio State University. Dancey, William S., M.L. Fricke, and F. Church 1987 A siteless survey of a portion of the Big Walnut Creek drainage in southern Franklin County, Ohio, report in progress. Deaver, Ken 1980 Site distribution in the middle Scioto river valley. Ms. on file, Department of Archaeology, The Ohio Historical Society. Del Bene, T.A. 1979 Once upon a striation: current models of striation and polish formation. In Lithic use-wear analysis, edited by B. Hayden, pp. 167-179. Academic Press, New Yor k. Diamond, G. 1979 The nature of so-called polished surfaces on stone artifacts. In Lithic use-wear analysis, edited by Brian Hayden, pp.159-167. Academic Press, New York. Dunnell, R.C., L.H. Hanson, and D. Hardesty 1971 The Woodside component of the Slone site, Pike County, Kentucky. Southeastern Archaeological Conference, Bulletin 14. Dunnell, R.C. 1978 Style and function: a fundamental dichotomy. 2 8 8 American Antiquity 43:192-202. Dunnell, R.C. 1975 Archaeological potential of anthropological and scientific models of function. In Archaeological essays in honor of Irving B. Rouse, edited by R.C. Dunnell and E.S. Hall, pp.41-73. Mouton: The Hague. Eder, James E. 1984 The impact of subsistence change on mobility and settlement pattern in a tropical forest foraging economy: some implications for archaeology. American Anthropologist 86(4) : 837-853. Essenpreis, P. 1978 Fort Ancient settlements: an analysis of Anderson and Madisonville foci. In Mississippian settlement patterns, edited by Bruce D. Smith, pp. 141-168. Academic Press, New York. Flannery, Kent V. 1976 The village and its catchment area: introduction. In The ear1v Mesoamerican village, edited by K.V. Flannery, pp. 91-95. Academic Press, New York. Fuller, John W. 1982 Developmental change in prehistoric community patterns: the development of nucleated village communities in northern West Virginia. Unpublished Ph.D. dissertation, Department of Anthropology, Graves, Lawrence S. 1977 Reconstruction of the environmental conditions for human cultural development in the western Lake Erie basin during late Holocene time. Unpublished N.A. thesis, Department of Geology, Bowling Green State University. Graybill, J.R. 1980 Fort Ancient occupations in the Ohio River valley of West Virginia. Paper presented at the 1980 Society for American Archaeology Meetings, Philadelphia. Graybill, J.R. 1981 The eastern periphery of Fort Ancient (A.D. 1050- 1650): a diachronic approach to settlement variability. Unpublished Ph.D. dissertation, Department of Anthropology, University of Washington. Graybill, J.R. 1984 The eastern periphery of Fort Ancient. 289 Pennsylvania Archaeologist 54(1-2):40-50. Greiser, S.T. and P.D. Sheets 1979 Raw materials as a functional variable in use-wear studies. In Lithic use-wear analysis, edited by B. Hayden, pp.289-296. Academic Press, New York. Griffin, James B. 1943 The Fort Ancient aspect: its cultural and chronological position in Mississippi Valley archaeology. Museum of Anthropology, University of Michigan. Anthropological Papers 28. Griffin, James B. 1952 The late prehistoric cultures of the Ohio Valley. Ohio State Archaeological and Historical Quarterly 61: 186-195. Griffin, James B. 1961 Some correlations of climatic and cultural change in Eastern North America prehistory. New York Academy of Science. Annals 95:710-717. Griffin, James B. 1978 The Late Prehistoric of the Ohio Valley. In Handbook of North American Indians: the Northeast 15, edited by Bruce Trigger, pp. 547-559. Smithsonian Institution, Washington, D.C. Hanson, L .H., Jr. 1966 The Hardin village site. University of Kentucky, Studies in Anthropology 4. Hanson, L.H., Jr. 1975 The Buffalo site. West Virginia Geological Survey, Report of Investigations 5. Harris, David R. 1977 Settling down: an evolutionary model for the transformation of mobile bands into sedentary communities. In The evolution of social systems. edited by J. Friedman and M.J. Rowlands, pp. 401-418. Duckworth, London. Hassan, F.A. 1975 Determination of size, density, and growth rate of hunter-gatherer populations. In Population, ecology, and social evolution, edited by S. Polgar, pp. 27-52. Mouton, The Hague. 290 Hayden, Brian 1979 Snap, shatter, and superfractures: use-wear of stone skin scrapers. In Lithic use-wear studies, edited by B. Hayden, pp. 207-230. Academic Press, New York. Heilman, J.M. and R.R. Hoefer 1979 Astronomical alignments in a Fort Ancient settlement at the Incinerator site in Dayton, Ohio. Ms. on file, Dayton Museum of Natural History. Heilman, J.M. and A, Schwab 1980 Fort Ancient in southwestern Ohio. Ms. on file, Dayton Museum of Natural History. Holley, G.H. and T.A. Del Bene 1980 An evaluation of Keeley's microwear approach. Journal of Archaeological Science 8:337-52. Holmes, W.H. 1907 [Review of Mills’ Certain mounds and village sites, I.] American Anthropologist, N.S. 9:614-618. Hooten, E.A. and C.C. Willoughby 1920 Indian village site and cemetery near Madisonville, Ohio. Peabody Museum of American Anthropology and Ethnology 8(1). Hunter, D., Jr., C. Hunter, and H. Counts 1940 Report on Mound Number 14 (Cramer Mound) R0-33. Ms. on file, Department of Anthropology, The Ohio Historical Society. Jelinek, A.J. 1976 Form, function, and style in lithic analysis. In Cultural change and continuity, edited by C.E. Cleland, pp. 19-34. Academic Press, New York. Kamminga, J. 1979 Snap, shatter, and superfractures: use-wear of stone skin scrapers. In Lithic use-wear analysis, edited by Brian Hayden, pp. 207-231. Academic Press, New York. Keeley, Lawrence H. 1974 The methodology of microwear analysis: a comment on Nance. American Antiquity 39( 1):126-128. Keeley, Lawrence H. 1980 Experimental determination of stone tool uses. University of Chicago Press, Chicago. 291 Kelly, John E., et al. 1984 Late Woodland period. In American Bottom archaeology: a summary of the FAI-270 proiect contribution to the culture history of the Mississippi River Valley. edited by C.J. Bareis and J.W. Porter, pp. 104-127. University of Illinois Press, Chicago. Kelly, John E., et al. 1984 Emergent Mississippian period. In American Bottom archaeology: a summary of the FAI-270 pro iect contribution to the culture history of the Mississippi River Valley, edited by C.J. Bareis and J.W. Porter, pp. 128-157. University of Illinois Press, Chicago. Kent, Susan 1984 Analyzing activity areas: an ethnoarchaeological study of the use of space. University of New Mexico Press, Albuquerque. King, Frances B. 1986 Prehistoric human plant use at the multicomponent Ranch site, central Illinois. Illinois State Museum and Crop Evolution Laboratory. Lawrence, R.A. 1979 Experimental evidence for the significance of attributes used in edge-damage analysis. In Lithic use-wear analysis, edited by B. Hayden, pp. 113-123. Academic Press, New York. Lee, R.B. 1966 !Kung Bushman subsistence: an input-output analysis. In Ecological essays, edited by D. Damas. Proceedings of the Conference on Cultural Ecology, National Museum of Canada Bulletin 230. MacMinn, E.M. 1980 An archaeological survey of Scioto and adjacent townships on Paint Creek, Ross County, Ohio. Ms. on file, The Ohio Historical Society. McKenzie, D.H. 1967 The Graham village site: a Fort Ancient settlement in the Hocking valley, Ohio. In Studies in Ohio archaeology. edited by O.H. Prufer and D.H. McKenzie, pp.63-97. Kent State University Press, Kent, Ohio. McLoda, Niles A. and R.J. Parkinson 1980 Soil survey of Franklin County, Ohio. USDA, Soil Conservation Service, with the Ohio Department of Natural Resources, Division of Land and Soils, and the 292 Ohio Agricultural Research and Development Center, Columbus, Ohio. Mills, W.C. 1902 Report of the curator of the Ohio State Archaeological and Historical Society. Ohio State Archaeological and Historical Society Publications 10:78-123. Mills, W.C. 1904 Explorations of the Gartner mound and village site. Ohio State Archaeological and Historical Society Publications 13:129-191. Mills, W.C. 1906 Baum prehistoric village. Ohio State Archaeological and Historical Society Publications 15:45-136. Mills, W.C. 1907 Explorations of the Edwin Harness mound. Ohio State Archaeological and Historical Society Publications 16:113-194. Mills, W.C. 1914 Archaeological atlas of Ohio. Ohio Historical Society, Columbus, Ohio. Mills, W.C. 1917 The Feurt mounds and village site. Ohio State Archaeological and Historical Society Publications 26:305-449. Mills, W.C. 1920 Map and guide to Fort Ancient, Warren County. Ohio, the largest prehistoric fortification in the Mississippi Valley. F.J. Heer, Columbus, Ohio. Mills, W.C. 1922 Exploration of the Mound City group. Ohio State Archaeological and Historical Quarterly 31:423-584. Moorehead, Warren K. 1897 Report of fieldwork. Ohio Archaeological and Historical Society Publications 5:165-275. Moorehead, Warren K. 1899 Report of fieldwork. Ohio State Archaeological and Historical Society Publications 7:110-205. 293 Moorehead, Warren K. 1908 Fort Ancient, the great prehistoric earthwork of Warren County, Ohio. Phillips Academy. Bulletin 4. Morgan, R.G. 1947 Fort Ancient. Ohio State Archaeological and Historical Society, Columbus, Ohio. Morgan, R.G. 1952 Outline of cultures in the Ohio region. In Archaeology of the Eastern United States, edited by J.B. Griffin, pp. 83-98. University of Chicago Press, Chicago. Morse, Dan F. and Phyllis A. Morse 1983 Archaeology of the central Mississippi Valley. Academic Press, New York. Moss, Emily H. 1983 The functional analysis of flint implements: Pincevent and Ponce d’Ambou: two case studies from the French Final Paleolithic. British Archaeological Reports, International Series 177. Murphy, J. 1975 kn_ archaeological history of the Hocking Valley. Ohio University Press, Athens, Ohio. Nance, Jack D. 1971 Functional interpretations from microscopic analysis. American Antiquity 36(3): 362-366. Newcomer, M.H. and L.H. Keeley 1979 Testing a method of micro-wear analysis with experimental flint tools. In Lithic use-wear analysis, edited by Brian Hayden, pp. 195-207. Academic Press, New York. Odell, George H. 1974 Microwear in perspective: a sympathetic response to Lawrence H. Keeley. World Archaeology 7:226-240. Odell, George H. 1982 Emerging directions in the analysis of prehistoric stone tool use. Review of: Lawrence H. Keeley: Experimental determination of stone tool uses: a fflicrowear analysis. Reviews in Anthropology 9(1):17- 33. Odell, George H. and F. Odell-Vereecken 1980 Verifying the reliability of lithic use-wear 294 assessments by blind tests: the low power approach. Journal of Field Archaeology 1:171-196. Ogden, J.G., III 1966 Forest history of Ohio, Part I. Radiocarbon dates and pollen stratigraphy of Silver Lake, Logan County, Ohio. Ohio Journal of Science 66:387-400. Ogden, J.G., III and R.J. Hay 1960 Ohio Wesleyan University natural radiocarbon measurements II. Radiocarbon 7: Ogden, J.G., III and R.J. Hay 1965 Ohio Wesleyan University natural radiocarbon measurements IV. Radiocarbon 9:316-332. Ogden, J.G., III and R.J. Hay 1967 Ohio Wesleyan University natural radiocarbon measurements VI. Radiocarbon 11(1 ):137-149. Otto, Martha P. 1982 Excavations at the Scioto Trails School site. Paper presented at the Annual Fall Meeting of the Ohio Archaeological Council. Otto, Martha P. 1983 Late Woodland study unit for 58 counties of Ohio: Miami, Scioto, Hocking, Muskingum, and Ohio River drainages. Ms. on file, Ohio Historic Preservation Office. Petro, J.H., W.H. Shumate, and M.F. Tabb 1976 Soil survey of Ross County, Ohio. USDA, Soil Conservation Service. with ODNR, Division of Lands and Soil, and Ohio Agricultural Experimental Station. Phagan, Carl J. 1977 Intensive archaeological survey of the S.R. 315 Wastewater Treatment Facility location known as the DECC0-1 site (33 DL 28). Ms. on file, Ohio Historical Society. Piotrowski, L.R. 1980 The Deer Creek project (Ross County) final report. Ms. on file, The Ohio Historical Society. Potter, Martha A. 1966 Cole ceramics: ji study of Late Woodland pottery. Unpublished M.A. thesis, Department of Anthropology, The Ohio State University. 29 5 Prufer, O.H. 1965 The McGraw site: a study in Hopewellian dynamics. Cleveland Museum of Natural History. Science Publications. N.S. 4:1-144. Prufer, O.H. 1967 Chesser Cave: a Late Woodland phase in southeastern Ohio. In Studies in Ohio archaeology, edited by O.H. Prufer and D.H. McKenzie, pp.1-63. Kent State University Press, Kent, Ohio. Prufer, O.H. 1967 The Scioto valley archaeological survey. In Studies in Ohio archaeology. edited by O.H. Prufer and D.H. McKenzie, pp. 267-329. Kent State University Press, Kent, Ohio. Prufer, O.H. and D.H. McKenzie, editors. 1966 Peters Cave. Ohio Journal of Science 66(3):233- 253. Prufer, O.H. and O.C. Shane III 1970 Blain village and the Fort Ancient tradition in Ohio. Kent State University Press, Kent, Ohio. Putnam, Frederick W. 1886 Report of the curator. Peabody Museum Reports 3. Rafferty, Janet E. 1974 The development of the Fort Ancient tradition in northern Kentucky. Unpublished Ph.D. dissertation, Department of Anthropology, University of Washington. Rafferty, Janet E. 1985 The archaeological record on sedentariness: recognition, development, and implications. Advances in archaeological methods and theory 8:113-156. Railey, J.A. 1984 The Pyles site (15MS28): a Newtown village in Mason County, Kentucky. William S. Webb Archaeological Society, Occasional Paper 1. Reidhead, Van A. 1981 Optimization and food procurement at the prehistoric Leonard Haag site, southeastern Indiana: 41 linear programming approach. Ph.D. dissertation, University Microfilms, Ann Arbor. 296 Rigg, Diana 1977 Holmes mound: Cole or Fort Ancient. Ms. on file, Department of Anthropology, The Ohio State University. Robbins, Louise M. and George K. Neumann 1972 The prehistoric people of the Fort Ancient culture of the central Ohio valley. Museum of Anthropology, University of Michigan, Anthropological Paper 2. Rohlf, F.J. and R.R. Sokal 1981 Statistical tables (second edition). W.H. Freeman and Company, San Francisco. Roper, Donna C. 1979a Archaeological survey'and settlement pattern models in central Illinois. Illinois State Museum Scientific Papers, 16. Roper, Donna C. 1979b The method and theory of site catchment analysis: a review. Advances in archaeological methods and theory 2:119-140. Sackett, J.R. 1977 The meaning of style in archaeology: a general model. American Antiquity :369-380. Sackett, J.R. 1982 Approaches to style in lithic archaeology. Journal of Anthropological Archaeology 1:59-112. Sackett, J.R. 1985 Style, ethnicity, and stone tools. In Status. structure, and stratification: current archaeological reconstructions, edited by M. Thompson, M.T. Garcia, and F.J. Kense, pp.277-281. Proceedings of the 16th CHACM00L Conference. Salzman, Philip C. 1980 Introduction: processes of sedentarization as adaptation and response. In When nomads settle. edited by Philip C. Salzman, pp.1-20. Praeger, New York. Schambach, Frank F. 1971 Review of Blain village and the Fort Ancient tradition in Ohio by Olaf H. Prufer and Orrin C. Shane, III. American Anthropologist 73:1402-1404. Schermer, Shirley J. and J.A. Tiffany 1985 Environmental variables as factors in site 297 location: an example from the Upper Midwest. Midcontinental Journal of Archaeology 10(2):215-240. Seeman, Mark F. 1980 An archaeological and gepmorphological floodplain survey of the Scioto River- Paint Creek confluence. Ms. on file, 0HP0. Seeman, Mark F. 1980 A^ taxonomic review of southern Ohio Late Woodland. Paper presented at the 1980 Midwest Archaeological Conference, Chicago. < Seeman, Mark F. and C.A. Munson 1981 Determining the cultural affiliation of terminal Late Woodland- Mississippian hunting stations: a lower Ohio Valley example. North American Archaeologist 2(1):53-66. Shane, Orrin C., Ill 1975 Appendix: Ohio radiocarbon chronology. In Studies in Ohio archaeology (second edition), edited by O.H. Prufer and D.H. McKenzie, pp. 357-368. Kent State University, Kent, Ohio. Shetrone, H.C. 1920 The culture problem in Ohio archaeology. American Anthropologist, N.S., 22:144-172. Shetrone, H.C. and E.F. Greenman 1931 Explorations of the Seip group og prehistoric earthworks. Ohio State Archaeological and Historical Quarterly 50(3):349-51 0. Skinner, S.M., et al. 1981 Results of a data recovery program for portions of the Howard Baum site (33 RO 270), Ross County, Ohio. Ms. on file, The Ohio Historical Society. Smith, Bruce D. 1984 Mississippian expansion: tracing the historical development of an explanatory model. Southeastern Archaeology 3(l):13-32. Smith, Geoffrey W. 1966 Physiography. In Environmenta1 analysis of central Ohio: an initial approximation. Vol. I, edited by D.M. Anderson and C.C. King, Chapter 7. Ohio Biological Survey, Columbus, Ohio. 298 Smith, H.I. 1910 The prehistoric ethnology of a Kentucky site. American Museum of Natural History. Anthropological Papers 6(2):173-241, Smith, Philip E.L. 1972 Land-use, settlement patterns and subsistence agriculture: a demographic perspective. In Man. settlement. and urbanism, edited by P.J. Ucko, R. Tringham, and H. Dimbleby, pp.409-425. Duckworth, London. Sokal, R.R. and F.J. Rohlf 1981 Biometry (second edition). W.H. Freeman and Company, San Francisco. Squier, E.G. and E.H. Davis 1847 Ancient monuments of the Mississippi Valley, comprising the results of extensive original surveys and explorations. The Smithsonian Institution. Contributions to Knowledge, Vol. I. Stoltman, J.B. 1978 Temporal models in prehistory: an example from Eastern North America. Current Anthropology 19(4):703-731. Styles, Bonnie W. 1981 Faunal exploitation and resource selection: Early Late Woodland subsistence in the lower Illinois Valley. Northwestern University Archeological Program, Evanston, Illinois. Tiffany, J.A. and L.R. Abbott 1982 Site catchment analysis: applications to Iowa archaeology. Journal of Field Archaeology 9(3):313- 322. Tringham, Ruth, et al. 1974 Experimentation in the formation of edge damage: a new approach to lithic analysis. Journal of Field Archaeology 1:171-196. Ullman, Kyle L. 1985 The ceramics from the Kramer village site (33 R0 33), Ross County, Ohio. Kent State Research Papers in Archaeology 5. Unger-Hamilton, Romana 1984 The formation of use-wear polish on flint: beyond the "deposit versus abrasion" controversy. Journal of 299 Archaeological Science 11(1):91—98. Vickery, K.D. 1970 Evidence supporting the theory of climatic change and the decline of Hopewell. Wisconsin Archaeologist 51(2):57. Vita Finzi, C. and E.S. Higgs 1970 Prehistoric economy in the Mount Carmel area of Palestine: site catchment analysis. Proceedings of the Prehistoric Society 36:1-37. Wagner, Gail E. 1983 Fort Ancient subsistence: the botanical record. West Virginia Archeologist 35(2):27-40. Willey, G.R. 1953 Prehistoric settlement pattern in the Viru Valley, Peru. Smithsonian Institution Bulletins of American Ethnology 155:2-6. Willey, G.R. and P. Phillips 1958 Method and theory in American archaeology. University of Chicago Press, Chicago. Willey, G.R. and J.A. Sabloff 1980 A^ history of American archaeology (second edition). W.H. Freeman, San Francisco. Winters, Howard D. 1969 The Riverton culture: a second millenium occupation in the central Wabash Valley. Illinois Archaeological Survey. Monograph 1^. Wise, D.L. and J.N. Rayner 1966 Climate. In Environmental analysis of central Ohio: an initial approximation, Vol. I, edited by D.M. Anderson and C.C. King, Chapter 1. Ohio Biological Survey, Columbus, Ohio. Wistendahl, W.A. and K.P. Lewis 1976 Vegetation. In Environmental analysis of central Ohio: an initial approximation. Vol. II, edited by D.M. Anderson and C.C. King, Chapter 10. Ohio Biological Survey, Columbus, Ohio. Wittoft, J. 1967 Glazed polish on flint tools. American Antiquity 32(3):383-388. 3 0 0 Wood, W.R. 1976 Vegetational reconstruction and climatic episodes. American Antiquity Al(2):206-208. Wymer, Dee Anne 1985 Preliminary report on the paleobotanical analysis of material from the Zencor/Scioto Trails School site. Franklin County. Ohio. Paper presented at the Annual Spring Meeting of the Ohio Archaeological Council. Wymer, Dee Anne 1986 The Middle Woodland-Late Woodland interface in central Ohio: subsistence continuity amid cultural change. Paper presented at Midwest Archaeological Conference, Columbus, Ohio. Yerkes, Richard W. 1985 Stone tool function and social differentiation within the Mississippian component at the Labras Lake site, Illinois. Paper presented at the 50th Annual meetings of the Society for American Archaeology, Denver. Zarkey, A. 1976 Statistical analysis of site catchments at Ocos, Guatemala. In The early Mesoamerican village, edited by Kent V. Flannery, pp.117-128. Academic Press, New York. APPENDIX A Statistical Tests: SE and R X C Test of Independence, G-test ^ Two statistical tests were utilized for this and all subsequent appendices of statistical analyses: the Standard Error of Proportions (SEp, and, where indicated by the first test, an R X C Test of Independence, using the G-Test (Sokal and Rohlf 1981). An example of their application is given below. SEp Test: Province LW LW-LP LP U.002-.110 0-.105 .048-.232 G .347-.583 .253-.501 .485-.743 T .359-.598 .443-.697 .131-.359 This table was derived from: [r(l-r)/n]2 where r equals the percentage of sites per cell. Expected values should be nearly equivalent. Discrepancies are then tested to examine possible association of variables, when the value G/q is compared to X (chi-squared) for the degrees of freedom present in the data table. J* X C Test of Independence, using G-test: The original data table of frequencies is used to obtain four quantities based on an f In f table: 1) the sum of each datum _f In f_\ 2) the sum of f In f for each row total; 3) the sum of f In f for the sum of each column total; and 4) f In f for the data total (N). G=2[quantity 1-quantity 2-quantity 3+ quantity 4] q = l + (a - 1)(b - l)/6n(a - 1)(b - 1), where a = number of columns and b = number of rows. For Physiographic Province by Time Period: G/q=15.079 X2( .01)(9) = 14*684 301 APPENDIX B MASTER LIST OF FUNCTIONAL TOOL CLASSES Dimensions WP* EH^ BB^ Totals K-L-0-S-A6 1 1 1 4 4 1 0 0 0 0 1 1 1 2 3 2 0 1 0 0 0 1 1 1 2 3 3 0 0 1 0 0 1 1 1 2 4 1 1 0 4 0 0 5 1 1 2 4 2 0 0 3 0 0 3 1 1 2 4 3 1 0 0 0 0 1 1 1 2 4 4 2 0 0 0 0 2 1 1 3 1 2 2 2 0 0 0 4 1 1 3 4 2 0 0 1 0 0 1 1 1 4 4 4 1 0 0 0 0 1 1 2 1 1 3 1 0 0 0 0 1 1 2 1 3 1 0 0 3 0 0 3 1 2 1 3 3 0 0 0 1 0 1 1 2 1 3 4 0 3 1 0 0 4 1 2211 10 0 2 0 0 12 1 2 2 1 2 9 3 7 2 0 21 1 2 2 1 3 5 0 0 0 0 5 1 2 2 2 1 8 0 0 0 0 8 1 2 2 2 2 24 0 3 0 0 27 1 2 2 2 3 5 0 0 0 0 5 1 2 2 3 1 11 3 24 0 0 38 Appendix: B, cont. • 3 0 3 1 2 2 3 2 12 36 34 3 0 85 1 2 2 3 3 4 24 10 2 0 40 1 2 2 3 4 0 5 2 0 0 7 1 2 2 4 2 1 1 0 0 0 2 1 2 2 4 3 0 0 1 0 0 1 1 2 2 6 2 3 0 0 0 0 3 1 2 3 1 1 2 0 0 0 0 2 1 2 3 1 2 7 0 0 0 0 7 1 2 3 1 3 2 0 0 0 0 2 1 2 3 2 1 1 0 0 0 0 1 1 2 3 2 2 2 0 0 0 0 2 1 2 3 2 3 3 0 0 0 0 3 1 2 3 3 1 0 0 7 0 0 7 1 2 3 3 2 2 1 2 0 0 5 1 2 3 4 2 0 2 0 0 0 2 1 2 4 1 2 1 0 0 0 0 1 1 2 4 1 3 1 0 0 0 • 0 1 1 2 4 2 3 1 0 0 0 0 1 1 2 4 3 2 1 0 0 0 0 1 1 2 4 3 3 2 0 0 0 0 2 1 2 6 3 2 0 1 0 0 0 1 1 3 1 3 1 0 0 1 0 0 1 1 3 2 1 1 24 0 0 0 0 24 1 3 2 1 2 12 0 0 0 0 12 1 3 2 1 3 0 2 0 0 0 2 1 3 2 2 1 6 0 0 0 0 6 pendix B, cont. 304 1 3 2 2 2 3 1 0 0 0 4 1 3 2 2 3 1 0 2 0 0 3 1 3 2 3 1 11 1 8 0 0 20 1 3 2 3 2 4 2 4 0 0 10 1 3 2 3 3 0 2 1 0 0 3 1 3 2 6 1 2 0 0 0 0 2 1 3 3 1 2 4 0 0 0 0 4 1 3 3 1 4 1 0 0 0 0 1 1 3 3 2 1 4 0 0 0 0 4 1 3 3 2 2 1 0 0 0 0 1 1 3 3 2 3 1 0 0 0 0 1 1 3 3 3 2 1 0 0 0 0 1 1 3 3 4 1 0 0 1 0 0 1 1 3 4 1 2 1 0 0 0 0 1 1 4 1 1 1 0 0 1 0 0 1 1 5 2 1 1 0 0 1 0 0 1 1 5 2 1 2 0 0 1 0 0 1 1 5 2 1 4 3 0 0 0 0 1 5 2 4 2 0 0 1 0 0 1 1 5 2 4 4 1 0 0 0 0 1 1 5 3 1 3 1 0 0 0 0 1 1 6 1 1 4 0 2 0 0 0 2 1 6 1 2 4 0 0 0 1 0 1 1 6 2 1 2 0 0 2 0 0 2 1 6 2 1 4 0 3 1 0 0 4 1 6 2 2 1 0 0 3 0 0 3 Appendix EI, cont. 305 1 6 2 2 2 0 1 8 0 0 9 1 6 2 2 3 0 0 3 0 0 3 1 6 2 2 4 0 7 0 1 0 8 1 6 2 3 2 0 0 1 0 0 1 1 6 2 5 2 0 0 1 0 0 1 1 6 3 1 2 0 1 0 0 0 1 1 6 3 2 4 0 1 0 0 0 1 1 6 5 1 2 0 1 1 0 0 2 1 6 5 2 3 0 0 1 0 0 1 1 6 5 2 4 0 0 0 1 0 1 2 2 1 3 2 0 0 2 0 0 2 2 2 1 3 4 0 1 0 0 0 1 2 2 2 3 2 0 1 0 0 0 1 2 2 2 7 1 0 1 0 0 0 1 2 3 2 1 4 1 0 0 0 0 1 2 3 2 3 4 0 0 1 0 0 1 2 3 2 7 1 0 1 0 0 0 1 2 4 2 1 4 1 0 0 0 0 1 2 4 2 2 1 0 2 0 0 0 2 2 4 2 7 1 0 14 0 0 0 14 2 4 2 7 2 0 6 0 0 0 6 2 4 3 2 1 0 1 0 0 0 1 , 2 4 3 2 4 3 0 0 0 0 3 2 4 3 3 1 0 1 0 0 0 1 2 4 3 7 1 0 3 0 0 0 3 2 4 4 1 4 1 0 0 0 0 1 Appendix B, cont. 306 2 4 4 2 4 4 0 0 0 0 4 2 4 4 3 4 4 0 0 0 0 4 2 4 4 7 1 0 6 0 0 0 6 2 4 4 7 4 0 1 0 0 0 1 2 4 5 1 1 0 1 0 0 0 1 2 5 3 4 1 0 1 0 0 0 1 2 5 3 4 4 0 1 0 0 0 1 2 6 1 1 3 0 0 1 0 0 1 2 6 2 1 2 0 0 1 0 0 1 2 6 2 2 2 0 0 2 0 0 2 2 6 2 5 3 0 0 2 0 0 2 2 6 3 1 2 0 0 1 0 0 1 2 6 3 2 2 0 2 0 0 0 2 2 6 4 1 4 1 0 0 0 0 1 2 6 5 1 1 0 0 1 0 0 1 2 6 5 1 2 0 0 1 0 0 1 2 6 5 1 4 1 0 0 0 0 1 2 7 1 1 1 0 0 1 0 0 1 2 7 2 1 3 0 0 2 0 0 2 2 7 2 2 2 0 0 1 0 0 1 2 7 4 1 1 0 0 1 0 0 1 2 7 5 1 1 0 8 2 0 5 15 2 7 5 1 3 0 0 1 0 0 1 2 7 5 1 4 2 0 0 0 0 2 2 7 5 2 1 0 8 2 0 0 10 2 8 2 1 2 0 0 1 0 0 1 Appendix B, cont. 307 2 8 4 7 1 0 1 0 0 0 1 2 8 5 2 1 0 2 0 0 0 2 3 1 2 3 1 0 0 0 0 2 2 3 1 2 4 1 0 0 1 0 5 6 3 1 2 5 1 0 0 0 0 2 2 3 1 3 2 1 0 0 0 0 1 1 3 1 3 3 2 0 0 0 0 2 2 3 1 3 3 4 0 0 0 0 1 1 3 1 3 4 1 0 0 0 0 21 21 3 1 3 4 2 0 0 0 0 14 14 3 1 3 4 3 0 0 0 0 6 6 3 1 3 4 4 0 0 0 0 4 3 1 5 4 1 0 0 0 0 2 2 3 2 2 1 1 0 0 0 0 26 26 3 2 2 2 1 0 0 1 0 0 1 3 2 2 3 1 0 1 0 0 0 1 3 2 2 3 3 0 0 0 1 0 1 3 2 2 2 4 0 2 0 0 0 2 3 2 2 6 1 2 0 0 0 0 2 3 2 3 3 2 0 0 0 0 1 1 3 2 4 1 1 0 0 0 0 5 5 3 2 4 3 1 0 0 0 0 1 1 3 2 4 3 4 1 0 0 0 0 1 3 2 5 3 1 0 0 0 2 0 2 3 3 1 3 4 0 1 0 0 0 1 3 3 1 4 1 0 1 0 0 0 1 Appendix B, cont. 3 0 8 3 3 2 2 2 0 1 0 0 0 1 3 3 2 2 3 0 1 0 0 0 1 3 3 3 3 1 0 0 0 0 3 3 3 3 5 3 2 0 0 0 1 0 1 3 4 1 3 1 0 1 0 0 0 1 3 4 2 3 1 0 1 0 0 0 1 3 4 4 1 1 0 2 0 0 0 2 3 4 4 1 4 1 0 0 0 0 1 3 4 4 7 1 0 0 3 0 0 3 3 5 2 4 1 0 0 3 0 0 3 3 5 5 2 1 0 1 0 0 3 4 3 5 5 2 3 0 0 0 0 1 1 3 5 5 4 1 0 0 0 0 32 32 3 5 5 4 2 0 0 0 0 41 41 3 5 5 4 3 0 - 0 0 0 12 12 3 5 5 4 4 0 0 0 0 4 4 3 6 1 2 4 0 1 0 0 0 1 3 6 3 2 3 0 0 0 1 0 1 3 6 5 1 3 1 0 0 0 0 1 3 6 5 2 1 0 0 0 2 0 2 3 6 5 3 4 0 4 0 0 0 4 3 7 1 1 1 0 0 2 0 0 2 3 7 2 1 1 0 0 1 0 0 1 3 7 2 4 1 0 0 0 0 1 1 3 7 5 1 1 0 0 0 0 39 39 3 7 5 2 1 0 12 0 0 24 36 Appendix B1» cont. 309 3 7 5 2 2 0 0 0 0 4 4 3 7 5 2 3 0 0 0 0 1 1 3 7 5 2 4 0 0 0 0 2 2 3 7 5 3 1 0 0 0 1 0 1 3 7 5 4 1 0 0 0 0 3 3 3 7 5 7 1 0 0 0 0 1 3 7 5 7 2 0 0 0 0 1 3 8 2 4 1 0 0 0 0 1 3 8 5 4 1 0 0 0 0 1 4 1 1 4 1 0 1 0 0 0 4 1 2 4 2 0 1 0 0 0 4 2 1 2 2 0 2 0 0 0 4 2 2 2 4 0 0 0 0 11 11 4 2 2 3 1 0 1 0 0 0 4 2 2 7 2 1 0 0 0 0 4 2 3 2 1 0 1 0 0 0 4 2 3 2 2 0 2 0 0 0 4 2 3 2 3 0 1 0 0 0 4 3 1 2 2 0 1 0 0 0 4 3 2 3 1 0 2 0 0 0 4 3 2 3 2 0 1 0 0 0 4 3 2 7 2 0 1 0 0 0 4 3 3 1 4 7 0 0 0 0 4 3 3 2 2 0 2 0 0 0 2 4 3 3 2 4 1 0 0 0 0 1 4 3 3 3 2 0 2 0 0 0 2 Appendix B, cont. 310 4 4 1 7 1 0 3 0 0 0 3 4 4 1 7 4 0 1 0 0 0 1 4 4 2 3 4 0 1 0 0 0 1 4 4 3 1 1 0 3 0 0 0 3 4 4 3 7 1 0 1 0 0 0 1 4 4 3 7 2 0 1 0 0 0 1 4 4 4 1 4 4 0 0 0 0 4 4 4 4 2 4 22 0 0 0 0 22 4 4 4 3 4 5 0 0 0 0 5 4 4 4 7 1 0 1 0 0 0 1 4 5 2 1 2 1 0 0 0 0 1 4 5 2 1 4 1 0 0 0 0 1 4 5 2 4 4 2 0 0 0 0 4 5 3 1 3 1 0 0 0 0 1 4 5 3 2 4 1 0 0 0 0 1 4 5 4 1 4 1 0 0 0 0 1 4 5 4 3 4 1 0 0 0 0 1 4 5 4 4 3 1 0 0 0 0 1 4 5 5 1 4 3 0 0 0 0 4 5 5 4 3 0 0 0 0 1 1 4 6 1 1 4 0 0 1 0 0 1 4 6 3 1 1 0 0 0 0 1 1 4 6 3 1 4 1 0 0 0 0 1 4 6 4 2 4 2 0 0 0 0 2 4 6 4 3 4 1 0 0 0 0 1 4 6 5 1 4 2 0 0 0 0 2 .A 1 1 1 5 2 8 15 1058 N- 0 1 0 1 2 3 0 1 00 0 1 0 0 1 0 1 0 1 2 0 2 0 0 0 0 0 11 0 00 0 0 0 0 1 2 0 0 0 0 0 0 0 0 0 15 1 1 1 0 0 0 0 2 9 0 0 00 00 0 0 5 0 0 0 0 0 1 1 2 0 1 0 0 0 0 0 1 00 0 2 0 8 0 1 0 1 0 2 1 3 4 0 4 4 0 4 Classes 1 1 1 1 2 2 1 7 1 3 2 3 2 1 4 1 0 5 5 4 4 2 4 4 7 1 7 1 0 5 5 2 4 5 3 2 1 5 3 2 5 5 7 3 5 7 5 1 4 4 6 5 4 4 6 5 2 4 7 1 4 4 7 1 4 4 7 3 2 4 7 4 7 1 4 7 5 4 4 7 5 1 4 4 7 5 4 7 5 N= N= 246 1 wear; wear; S= Shape of worn area; A= Edge angle. Numbers below Baum. Enos Holmes. Baum: bone assemblage. K= Kind of wear; L= Location of wear; 0= Orientation of dimensions identify modes (see chapter XII). Gartner. Water Plant. Appendix B, cont. 311 cm co m so APPENDIX C Statistical Tests of Kind of Wear: Modes by Sites SEp Test of Proportions: Modes Water Plant(LW) Baum(LP) 1 .656-.762 .360-.486 2 .034-.090 .197-.307 3 .002-.032 .077-.159 4 . 152-. 246 .152-.254 5 0-.028 0-.012 II X C Test of Independence, using G-test: G/q= 77. 299 X2(.001)(4) = 18.467 312 APPENDIX D Statistical Tests of Location of Wear: Modes 1 Sites SEP Test of Proportions: Modes Water Plant(LW) Baum(LP) 1 .008-.046 .002-.038 2 .356-.472 .316-.440 3 .238-.344 .046-.116 4 . 124-.212 .152-.254 5 .031-.085 0-.026 6 .011-.051 .060-.136 7 0-.022 .144-.246 8 0 0-.026 R. X C Test of Independence, using G-test: G/q= 112.786 x2(.001)(7) = 24.322 313 APPENDIX E Statistical Tests of Orientation of Wear: Modes by Sites Test of Proportion s : SEP Mod es Water Plant:(LW) Baum(LP) 1 0- .017 .043-.106 2 .534- .650 .460-.588 3 .121- .207 .080-.164 4 .155- .249 .019-.071 5 .013- .055 .178-.286 6 0 0-.012 R X C TeSt Of Independ ence,, using G--test: G/q= 94. 701 y2 = 20.515 X (•, 001)(5) s 314 APPENDIX F Statistical Tests of Shape of Worn Area and Edge Angle SEp Test of Proportion: Shape Modes Water Plant(LW) Baum(LP) 1 .347-.461 .101-.191 2 .274-.384 .189-.299 3 .158-.252 .339-.465 4 .016-.060 .010-.056 5 0 0 6 .006-.042 0 7 0 .127-.223 Edge Angle Modes Water Plant(LW) Baum(LP) 1 .228-.334 .368-.494 2 .263-.373 .254-.372 4 .238-.344 .097-.187 315