The Paleoindian and Early Archaic Southeast

By David G. Anderson, Michael F. Johnson, Lisa D. O'Steen and 14 more

The southeastern United States has one of the richest records of early human settlement of any area of North America. This book provides the first state-by-state summary of Paleoindian and Early Archaic research from the region, together with an appraisal of models developed to interpret the data. It summarizes what we know of the peoples who lived in the Southeast more than 8,000 years ago—when giant ice sheets covered the northern part of the continent, and such mammals as elephants, saber-toothed tigers, and ground sloths roamed the landscape. Extensively illustrated, this benchmark collection of essays on the state of Paleoindian and Early Archaic research in the Southeast will guide future studies on the subject of the region's first inhabitants for years to come.

Divided in three parts, the volume includes:

Part I: Modeling Paleoindian and Early Archaic Lifeways in the Southeast

Environmental and Chronological Considerations, David G. Anderson, Lisa D. O'Steen, and Kenneth E. Sassaman
Modeling Paleoindian and Early Archaic Settlement in the Southeast: A Historical Perspective, David G. Anderson and Kenneth E. Sassaman

Models of Paleoindian and Early Archaic Settlement in the Lower Southeast, David G. Anderson
Early Archaic Settlement in the South Carolina Coastal Plain, Kenneth E. Sassaman
Raw Material Availability and Early Archaic Settlement in the Southeast, I. Randolph Daniel Jr.
Paleoindian and Early Archaic Settlement along the Oconee Drainage, Lisa D. O'Steen
Haw River Revisited: Implications for Modeling Terminal Late Glacial and Early Holocene Hunter-Gatherer Settlement Systems in the Southeast, John S. Cable
Early Archiac Settlement and Technology: Lessons from Tellico, Larry R. Kimball
Paleoindians Near the Edge: A Virginia Perspective, Michael F. Johnson

Part II: The Regional Record

The Need for a Regional Perspective, Kenneth E. Sassaman and David G. Anderson
Paleoindian and Early Archaic Research in the South Carolina Area, David G. Anderson and Kenneth E. Sassaman
The Taylor Site: An Early Occupation in Central South Carolina, James L. Michie
Paleoindian and Early Archaic Research in Tennessee, John B. Boster and Mark R. Norton
A Synopsis of Paleoindian and Early Archaic Research in Alabama, Eugene M. Futato
Statified Late Pleistocene and Early Holocene Deposits at Dust Cave, Northwestern Alabama, Boyce N. Driskell
Bone and Ivory Tools from Submerged Paleoindian Sites in Florida, James S. Dunbar and S. David Webb
Paleoindian and Early Archaic Data from Mississippi, Samuel O. McGahey
Early and Middle Paleoindian Sites in the Northeastern Arkansas Region, J. Christopher Gillam

Part III: Commentary

A Framework for the Paleoindian/Early Archaic Transition, Joel Gunn
Modeling Communities and Other Thankless Tasks, Dena F. Dincauze
An Arkansas View, Dan F. Morse
Comments, Henry T. Wright

• Language: English
• Publisher: University Alabama Press
• Release date: Sep 15, 2009
• ISBN: 9780817382995

Book preview

The Paleoindian and Early Archaic Southeast

The Paleoindian and Early Archaic Southeast

Edited by

David G. Anderson and Kenneth E. Sassaman

The University of Alabama Press

Tuscaloosa and London

Copyright © 1996

The University of Alabama Press

Tuscaloosa, Alabama 35487–0380

All rights reserved

Manufactured in the United States of America

manuscript designed by Virginia Horak

The paper on which this book is printed meets the minimum requirements of American National Standard for Information Science-Permanence of Paper for Printed Library Materials, ANSI Z39.48–1984.

Library of Congress Cataloging-in-Publication Data

The Paleoindian and Early Archaic Southeast / edited by David G. Anderson and Kenneth E. Sassaman

p.    cm.

Includes bibliographical references and index.

ISBN 0–8173–0835–0 (pbk. : alk. paper)

1. Paleo-Indians—Southern States. 2. Indians of North America—Southern States—Antiquities. 3. Southern States—Antiquities.

I. Anderson, David G., 1949–. II. Sassaman, Kenneth E.

E78.S65P35    1996

975'.01—dc20                                    96–19012

British Library Cataloguing-in-Publication Data Available

ISBN-13: 978-0-8173-8299-5 (electronic)

For those who came first . . .

Contents

Figures and Tables

Preface

PART I: MODELING PALEOINDIAN AND EARLY ARCHAIC LIFEWAYS IN THE SOUTHEAST

1. Environmental and Chronological Considerations

David G. Anderson, Lisa D. O'Steen, and Kenneth E. Sassaman

2. Modeling Paleoindian and Early Archaic Settlement in the Southeast: A Historical Perspective

David G. Anderson and Kenneth E. Sassaman

3. Models of Paleoindian and Early Archaic Settlement in the Lower Southeast

David G. Anderson

4. Early Archaic Settlement in the South Carolina Coastal Plain

Kenneth E. Sassaman

5. Raw Material Availability and Early Archaic Settlement in the Southeast

I. Randolph Daniel, Jr.

6. Paleoindian and Early Archaic Settlement along the Oconee Drainage

Lisa D. O'Steen

7. Haw River Revisited: Implications for Modeling Terminal Late Glacial and Early Holocene Hunter-Gatherer Settlement Systems in the Southeast

John S. Cable

8. Early Archaic Settlement and Technology: Lessons from Tellico

Larry R. Kimball

9. Paleoindians Near the Edge: A Virginia Perspective

Michael F. Johnson

PART II: THE REGIONAL RECORD

10. The Need for a Regional Perspective

Kenneth E. Sassaman and David G. Anderson

11. Paleoindian and Early Archaic Research in the South Carolina Area

David G. Anderson and Kenneth E. Sassaman

12. The Taylor Site: An Early Occupation in Central South Carolina

James L. Michie

13. Paleoindian and Early Archaic Research in Georgia

R. Jerald Ledbetter, David G. Anderson, Lisa D. O'Steen, and Daniel T. Elliott

14. Recent Paleoindian Research in Tennessee

John B. Broster and Mark R. Norton

15. A Synopsis of Paleoindian and Early Archaic Research in Alabama

Eugene M. Futato

16. Stratified Late Pleistocene and Early Holocene Deposits at Dust Cave, Northwestern Alabama

Boyce N. Driskell

17. Bone and Ivory Tools from Submerged Paleoindian Sites in Florida

James S. Dunbar and S. David Webb

18. Paleoindian and Early Archaic Data from Mississippi

Samuel O. McGahey

19. A Stone's Throw from Kimmswick: Clovis Period Research in Kentucky

Andrea K. L. Freeman, Edward E. Smith, Jr., and Kenneth B. Tankersley

20. Early and Middle Paleoindian Sites in the Northeastern Arkansas Region

J. Christopher Gillam

PART III: COMMENTARY

21. A Framework for the Paleoindian/Early Archaic Transition

Joel Gunn

22. Modeling Communities and Other Thankless Tasks

Dena F. Dincauze

23. An Arkansas View

Dan F. Morse

24. Comments

Henry T. Wright

References

Contributors

Index

Figures and Tables

Cover/frontispiece: Closing In by Dean Quigley. Early Paleoindian hunters stalk a mastodon and her calf in the swampy terrain of Late Pleistocene Florida some 11,000 years ago. © 1996 by Dean Quigley, all rights reserved. Used with permission of Dean Quigley.

FIGURES

1.1   Reconstructed paleovegetational communities in the Southeast at 14,000 and 10,000 B.P.

1.2   Paleoindian/Early Archaic chronology and diagnostic projectile points in the Southeast

2.1   Major Paleoindian and Early Archaic sites

3.1   Occurrence of Early and Middle Paleoindian projectile points in eastern North America

3.2   Early Paleoindian initial population concentrations and colonization routes in eastern North America

3.3   Inferred Middle Paleoindian cultural traditions in eastern North America

3.4   Early Archaic band and macroband annual mobility patterns in the vicinity of the Savannah River valley

3.5   Terminal Early Archaic Bifurcate and Stanly point distributions in North Carolina and South Carolina

4.1   Contour distributions of diagnostic hafted bifaces recorded by Charles in his surveys of private collections in South Carolina

4.2   Edgefield scraper from G. S. Lewis East, 38AK228

4.3   Raw material percentages for Palmer/Kirk Corner-Notched hafted bifaces by counties along the Savannah River and the Saluda-Congaree-Santee rivers

4.4   Raw material percentages for Palmer/Kirk Corner-Notched hafted bifaces by counties along the Broad-Congaree-Santee rivers and the Catawba-Wateree-Santee rivers

4.5   Raw material percentages for Palmer/Kirk Corner-Notched hafted bifaces by counties along the Fall Line and the middle Coastal Plain

4.6   Raw material percentages for Palmer/Kirk Corner-Notched hafted bifaces by counties across the lower Coastal Plain and across the middle Coastal Plain and along the Savannah River

4.7   Model of Early Archaic settlement organization for the upper Coastal Plain of the Savannah River valley, with location of key sites on the Savannah River Site

4.8   Kirk Corner-Notched bifaces in the G. S. Lewis East assemblage, showing sequence of use life stages resulting from resharpening and repair

4.9   Distribution of Kirk Corner-Notched hafted bifaces on the Savannah River Site

4.10  Distribution of hafted end scrapers, Edgefield scrapers, and Early Archaic bifacial preforms on the Savannah River Site

4.11  Paleoindian and Early Archaic artifacts from a Carolina bay site in Aiken County, South Carolina

6.1   Survey tracts north of Wallace Dam along the Oconee River drainage

6.2   Distribution of recorded Early and Middle Paleoindian sites along the Oconee River drainage

6.3   Distribution of Dalton and Early Archaic sites along the Oconee River drainage

6.4   Cluster of Paleoindian and Early Archaic sites at Barnett Shoals with piece-plotted points

7.1   Plan view and geological profile of 31CH29 and surrounding area

7.2   Paleoindian/Early Archaic artifacts, 31CH29, block A

7.3   Holocene effective temperature gradients, Haw River project area

7.4   Scatterplots depicting the temporal trends in tool assemblages identified in the original Haw River study

7.5   Scatterplots depicting relationship between sample size and diversity

7.6   Scatterplot of total curated tool diversity by expedient tool diversity, Haw River floors and comparison of curated and expedient tool diversity values for each of the Haw River floors

7.7   Present effective temperature isotherms, southeastern United States

8.1   The upper drainage of the Tennessee River valley, showing the Tellico Reservoir study area

8.2   Artifacts from the Tellico Reservoir area

8.3   Artifacts from the Tellico Reservoir area

8.4   Percentage of blades and bipolar flakes, cores, and pièces esquillées selected for tools in temporally ordered Early Archaic assemblages from Tellico

8.5   Dendrogram of average clustering of 41 Upper Kirk assemblages from the Tellico Reservoir area

8.6   The location of Upper Kirk sites by category in the Tellico study area

8.7   The location of Upper Kirk sites by category in the Great Smoky Mountains study area

9.1   Fluted points from Debert, Vail, and Lamb

9.2   McCary survey fluted points

9.3   Artifacts from the Thunderbird site

9.4   Major Virginia and northeastern fluted point sites

9.5   Quartz and quartzite fluted point distribution maps for Virginia

11.1  Major Paleoindian and Early Archaic sites in the vicinity of Georgia and South Carolina

11.2  South Carolina Paleoindian projectile points

11.3  South Carolina Paleoindian projectile points

11.4  Frequency distribution of Late Paleoindian and Early Archaic components recorded in the South Carolina site files by cultural-historical phase

11.5  Distribution of Early Holocene components by county in South Carolina

12.1  Plan distribution of Palmer and Dalton points in the excavation block at the Taylor site (38LX1)

12.2  Palmer points recovered from the Taylor site (38LX1) excavations

12.3  Artifacts recovered from the Taylor site (38LX1) excavations

12.4  Artifacts recovered from the Taylor site (38LX1) excavations

12.5  Artifacts recovered from the Taylor site (38LX1) excavations

12.6  Paleoindian points recovered from the surface of the Taylor site (38LX1)

13.1  Diagnostic projectile points from the Taylor Hill site

13.2  The Vulcan site assemblage

13.3  Early Paleoindian Clovis points from Georgia

13.4  Middle Paleoindian projectile points from Georgia

13.5  Late Paleoindian projectile points from Georgia

13.6  Early and Middle Paleoindian projectile point distributions in Georgia

14.1  Paleoindian points from Tennessee

14.2  Paleoindian artifacts from Tennessee

14.3  Blade tool assemblage, Carson-Conn-Short site

15.1  Selected Paleoindian and Early Archaic sites

15.2  Paleoindian points from northwestern Alabama

15.3  Unifacial blade tools, vicinity of Pickwick Lake

15.4  Unifacial flake/blade tools, vicinity of Pickwick Lake

16.1  Plan view of Dust Cave showing location of excavation units, and composite cross section of archaeological components, 1989–94

16.2  Major stratigraphic units and positioning of cultural components in the east profile of the entrance trench at Dust Cave

16.3  Plot of radiocarbon determinations by depth for 31 samples from the entrance trench at Dust Cave

16.4  Artifacts associated with the Early Side-Notched component from Dust Cave

16.5  Artifacts from the Late Paleoindian component at Dust Cave

17.1  Left radius of Mammut, UF 60814, used as an abrader

17.2  Left radius of Mammut:: successive cross sections

17.3  Patella of Mammut, BAR 92–508–01, used as an anvil

17.4  Neural spine of thoracic vertebra of Mammuthus, possibly used as a beamer

17.5  Distal left half of tibia of Equus, UF 16822, used as an awl handle

17.6  Left tibia of Equus near midshaft, showing proximal beveled area and enlarged marrow cavity

17.7  Rib of Mammuthus or Mammut, UF 136492, used as a hoe.

17.8  Proximal end of proboscidean rib

18.1  Major archaeological sites and physiographic regions in Mississippi

18.2  Mississippi Paleoindian and Early Archaic projectile points by region and period

18.3  Diagnostic Paleoindian and Early Archaic projectile points from the Yazoo Basin and north-central Mississippi

18.4  Diagnostic Paleoindian and Early Archaic projectile points from northeastern and southern Mississippi

18.5  Notched unifaces from southern Mississippi

18.6  Braided stream surfaces in the Yazoo Basin

18.7  Diagnostic Paleoindian and Early Archaic projectile points from eastern and western braided stream surfaces, Yazoo Basin, Mississippi

18.8  Dalton points from the western braided stream surface, Yazoo Basin, Mississippi

18.9  Percentage of Paleoindian and Early Archaic components by period in the Black Prairie of northeastern Mississippi

18.10 Percentage of Paleoindian and Early Archaic components by period on major stream channels in northeastern, north-central, and southern Mississippi

18.11 Percentage of Paleoindian and Early Archaic components by period in the Loess Hills in southern Mississippi

19.1  Clovis artifacts from the Big Bone Lick, Adams, and Boyd sites, Kentucky

19.2  Clovis Artifacts from the Ezell site, Little River Paleoindian District, Kentucky

20.1  Fluted points from the northeastern Arkansas region

20.2  The distribution of Paleoindian sites across the northeastern Arkansas region

20.3  Buffer map with 5-kilometer zones around potential lithic sources

21.1  Radiocarbon dated eruptions since 14,000 B.P.

23.1  The distribution of Dalton sites across the northeastern Arkansas region

TABLES

6.1   Paleoindian site location on topographic landforms along the Oconee River and its tributaries

6.2   Comparison of Paleoindian sites in shoals and intershoals regions of the Oconee River

7.1   Assemblage diversity values for the Haw River occupation floors

7.2   Curated tool frequencies and diversity values for the Haw River floors

7.3   Expedient tool frequencies and diversity values for the Haw River floors

7.4   Distribution of features and site furniture by living floor in the Haw River block A excavations, 31CH29

7.5   Comparative end scraper proportions for South Carolina sites and Haw River floors

7.6   Curated tool frequencies and diversity values for the South Carolina assemblages

8.1   Percentage of select wood and nut fragments from Early Archaic contexts

8.2   Expected site assemblages for Early Archaic components at Rose Island, Icehouse Bottom, Calloway Island, and Bacon Farm

8.3   Early Archaic site distributions (absolute and relative frequencies) within the sampling strata and by landform in the stratified random sample

8.4   Observed Early Archaic site distribution (absolute and relative frequencies) from nonprobabilistic survey

8.5   Artifact categories defining the morphofunctional categories

8.6   Summary of characteristics of clusters for 41 Upper Kirk assemblages

9.1   Population densities for selected Inuit groups

9.2   Yearly cycle patterns of barren ground caribou hunters

9.3   Yearly cycle patterns of woodland caribou-hunting Indians

16.1  Radiocarbon determinations (uncorrected) from Dust Cave, 1989–94

16.2  Tabulation of projectile points from test unit A, Dust Cave

17.1  Comparative measurements of the worked and an unaltered left Mammut radius

20.1  Soil associations with high occurrence of Paleoindian sites

20.2  Geological deposits with a high occurrence of Paleoindian sites

Preface

The year 1992 marked the 500th anniversary of the initiation of European colonization and settlement of the New World, and to many people, the year 1492 remains the date of the discovery of the Americas. In this volume, we celebrate the true discovery process, a record of exploration and human achievement of epic proportions that is emerging through archaeological research. Our focus is the first 3,500 years or so of human settlement in the lower southeastern United States, spanning the Paleoindian and Early Archaic cultural periods, from approximately 11,500 to 8000 B.P. In the pages that follow, ideas and evidence about this time in the Southeast are summarized, illustrating how our knowledge of these earliest Americans continues to develop. The geographic focus for our research comprises the area roughly south of the Ohio River–West Virginia–Pennsylvania area to just west of the Mississippi River.

This volume is divided into three parts. In Part 1, following introductory remarks about paleoenvironment and chronology, models of Paleoindian and Early Archaic settlement, subsistence, and technological organization are presented from across the region. In Part 2, Paleoindian and Early Archaic research is summarized on a state-by-state basis from across the Southeast by the leading authorities for each area, and primary data are presented on a number of important Paleoindian and Early Archaic sites and artifacts. In Part 3, commentary and directions for future research are offered.

This volume is based on a symposium held at the University of South Carolina in Columbia on 14 September 1991. The meeting was sponsored by the Council of South Carolina Professional Archaeologists and the South Carolina Institute of Archaeology and Anthropology, under the auspices of a Survey and Planning Grant awarded by the South Carolina Department of Archives and History. Christopher Judge was the project principal investigator and handled local arrangements. The symposium was open to the public, and over 70 avocational and professional archaeologists from around the Southeast attended. A limited number of copies of the papers from this meeting were printed in late 1992 in a volume entitled Paleoindian and Early Archaic Period Research in the Lower Southeast: A South Carolina Perspective, edited by David G. Anderson, Kenneth E. Sassaman, and Christopher Judge, and released by the Council of South Carolina Professional Archaeologists. The demand for this volume, which was unavailable within a month of its release, was truly astonishing and prompted us to release this revised and updated version.

The papers herein were produced in final form by their authors in early 1995. While we have tried to make this volume as up to date and comprehensive a regional reader and summary as possible, some gaps occur in our coverage. These typically reflect areas where excellent overviews have been recently presented elsewhere, or else where no researchers are working on these early periods. The Southeast has a tremendously rich and varied Paleoindian and Early Archaic record, furthermore, and new data are coming to light all the time. Accordingly, a primary goal has been to give our colleagues some sense of just how rich the early Southeastern record actually is; we suspect that a regional synthesis a decade or two from now will be far more extensive, and will contain answers to many of the questions that currently puzzle us.

We wish to thank the many colleagues who helped us prepare this manuscript, not the least of whom are Judith A. Knight and the staff of the University of Alabama Press, whose encouragement was crucial to seeing this manuscript through to completion. Albert C. Goodyear and Stephen Williams provided detailed comments on earlier drafts of the manuscript, while Trinket Shaw of Samford University provided extensive help with the copy-editing. Dean Quigley of St. Petersburg deserves special thanks for allowing us to use one of his wonderful reconstructions of Paleoindian life in Florida for the cover/frontispiece. The majority of the text illustrations were drawn or arranged by Julie Barnes Smith who, as usual, has done a superb job. We wish to thank the Society for American Archaeology, the Kentucky Heritage Council, the Eastern States Archaeological Federation, the Society for Georgia Archaeology, the Archeological Society of Virginia, The University of Tennessee Press, the Journal of Middle Atlantic Archaeology, Ernest S. Burch, Jr., Hazel and Paul Delcourt, Arthur S. Spiess, and the Virginia Department of Historic Resources for their permission to reprint, in somewhat modified form, some of the material found herein. The editors wish to thank our employers, the National Park Service and the South Carolina Institute of Archaeology and Anthropology, for continued support of our efforts. In particular, the encouragement of John E. Ehrenhard and Bruce Rippeteau, Mark Brooks, and Richard Brooks is deeply appreciated. The camera-ready for this manuscript was assembled by Virginia Horak of the Southeast Archeological Center, with funding and support provided by the National Park Service and the Savannah River Archaeological Research Program of the South Carolina Institute of Archaeology and Anthropology. Her help, and that of the University of Alabama Press's production editor Rick Cook is deeply appreciated. Finally, we wish to thank all of the contributors for their assistance as this manuscript came together.

Part I

Modeling Paleoindian and Early Archaic Lifeways in the Southeast

1

Environmental and Chronological Considerations

David G. Anderson, Lisa D. O'Steen, and Kenneth E. Sassaman

ENVIRONMENTAL CONDITIONS

The initial human occupation of the Southeast in all probability occurred between 15,000 and 11,000 years before the present (B.P.), during the Late Glacial era. At that time, sea levels were 70 or more meters lower than at the present, and the Atlantic and Gulf shorelines were considerably seaward of their present location. As the continental ice sheets retreated in the north, water was returned to the oceans, and large sections of the continental shelf were inundated. By 9000 B.P. sea level was within a few meters of its present elevation. Widespread extinctions accompanied these environmental changes in North America, specifically the loss of over 30 genera of large mammals, including the Equidae and Camelidae (horses and camels) and all of the members of the order Proboscidea (elephants) (P. Martin 1984:361–63). Contemporary analyses indicate that these extinctions were essentially complete by ca. 10,000 years ago, and possibly as early as 10,500 to 10,800 B.P. (Grayson 1987; Mead and Meltzer 1984:447), shortly after widespread evidence for human settlement appears in the New World archaeological record. The relationship between these human and animal populations is a matter of considerable controversy (Martin and Klein 1984). While human predation of megafauna has been conclusively demonstrated at a number of locations, most notably in the Southwest and on the Great Plains, until recently little evidence for megafaunal exploitation had been recovered from the eastern United States.

This picture is no longer valid. A major accomplishment of Southeastern Paleoindian research in recent years has been the discovery and widespread acceptance of evidence for direct associations between human and now-extinct terminal Pleistocene fauna. Incontrovertible evidence currently exists only from Florida, and includes a speared giant tortoise from Little Salt Springs (Clausen et al. 1979), and the discovery of a Bison antiquus skull in the Wacissa River with a projectile point embedded in its forehead (Webb et al. 1984). In addition, a number of tools carved from green proboscidean ivory and other modified megafaunal bone have been found in Florida waters and have been inventoried by Dunbar and Webb (Dunbar 1991; chapter 17, this volume). Artifacts and human remains have also occasionally been found in cave, sinkhole, salt lick, or rockshelter settings elsewhere in the region, sometimes in deposits with Pleistocene fauna, and increasing care is being directed to understanding the context of these discoveries, to resolve questions of association and contemporaneity (e.g., Elliott and Martin 1991; Tankersley 1985; chapter 19, this volume). On the margins of the Southeast, furthermore, other indisputable associations of humans and mastodon have been found, at Kimmswick in southern Missouri (Graham et al. 1981), Martin's Creek in Ohio (Brush and Smith 1994), and Coates-Hindes in western Tennessee (John Broster, personal communication 1995). There is little doubt that Paleoindian populations in the Southeast hunted megafauna; what remains to be resolved, however, is how important this food source was, and what impact human predation had on the biosphere.

Recent broad-scale paleoenvironmental analyses from the Southeast indicate that major changes in vegetational communities have also occurred over the last 15,000 years. The period from 12,000 to 10,000 B.P., the time of postulated initial human settlement, was one of great change, because the relatively patchy environment was shifting to one of latitudinally and elevationally segregated zones (R. Kelly and Todd 1988:232). In the South Appalachian area north of 33° N, northern hardwoods such as oak, hickory, beech, birch, and elm replaced the Full Glacial spruce-pine boreal forest during this period (H. Delcourt and Delcourt 1985; P. Delcourt and Delcourt 1987) (figure 1.1). Over this same interval, temperatures were becoming warmer in summer and colder in winter, and precipitation was increasing (Watts 1980a–b). The vegetational matrix was thus changing rapidly, trending from a patchy boreal forest-parkland toward a homogeneous, mesic oak-hickory forest. In ecological terms, the vegetation was changing from immature, or coarse grained, to mature, or fine grained (Pianka 1978). The best available evidence suggests that this transition was complete over much of the lower Southeast by shortly after 10,000 B.P., and almost certainly by 9000 B.P. (Anderson et al. 1989; Boyd 1989; M. Davis 1983:172–73; Delcourt and Delcourt 1983:269, 1985:19, 1987; Larsen 1982:208–22; Watts 1971:687, 1980a:195).

South of 33° N, in the South Appalachian area and across much of the Southeast outside of peninsular Florida, evidence suggests that a hardwood canopy was in place considerably earlier, perhaps throughout much of the previous glacial cycle (H. Delcourt and Delcourt 1985; P. Delcourt and Delcourt 1983, 1987; T. Webb 1987). Although traditionally viewed as a time of major paleoenvironmental change, the Late Pleistocene/Early Holocene in this part of the lower Southeast (prior to the Hypsithermal interval) appears to have been characterized by stable regional oak-hickory vegetational communities. Only during the Middle Holocene Hypsithermal interval, from circa 8000–4000 B.P., did southern pine communities begin to emerge in the sandy interriverine uplands; this was also the period when extensive riverine swamps began to emerge (Brooks et al. 1986; M. Davis 1983; H. Delcourt and Delcourt 1985; P. Delcourt and Delcourt 1987; Delcourt et al. 1983; Foss et al. 1985; J. Howard et al. 1980; Knox 1983; Segovia 1985; H. E. Wright 1976).

Biotic resource structure has been shown to influence prehistoric group size, technological organization, and mobility patterns. This has been documented on both a global scale (Binford 1980; R. Kelly 1983, 1992; Shott 1986a) and within the lower Southeast (Anderson and Hanson 1988; Cable 1982a). The patchy forest structure north of 33° N and south of the ice sheets–tundra margin shortly after 12,000 B.P. in the Southeast would have been ideally suited for what have been called logistically organized collector adaptations (after Binford 1980). That is, patchy environments are best exploited by groups radiating out from central base camps and staying at short-term camps as long as necessary to collect resources prior to returning to the home base. This adaptation is known as a collector strategy, since task groups go out for extended periods in the collection of resources, which they then bring back to their settlement. While groups practicing collector strategies do move their base camps, they usually do so only when local resources are depressed or exhausted to the point where the costs of moving are less than those of finding food. The archaeological record of collector groups includes base settlements and extended resource procurement camps. These adaptations are commonly characterized by highly formalized tool kits, assemblages that would have been most advantageous during extended resource procurement forays.

Collector or logistically based adaptations are, in fact, assumed by many researchers to characterize initial Paleoindian groups in North America. Paleoindian tool kits over the region are renowned for their well-made artifacts, including superbly executed bifaces as well as hafted end and side scrapers, gravers, spokeshaves, adzes, denticulates, and other tool forms. These artifacts were curated; that is, they were carried about from place to place and reused as necessary until they were worn out. These tools were frequently made of high-quality lithic raw material, which would have facilitated reworking and hence helped conserve raw material (Goodyear 1979). Movement over large areas also characterized these early populations, with group ranges centered on quarries or other particularly desirable environmental features where home bases appear to have been located (W. Gardner 1989). Once resources in the base camp–logistic station procurement zone began to become exhausted, however, relocation of the base settlement may have required a fairly extended move (R. Kelly 1983; Shott 1986a).

In contrast, in the homogeneous hardwood canopy south of 33° N, on the Gulf of Mexico and lower South Atlantic Slope, less evidence for Early Paleoindian settlement might be expected, since the initial founding populations were apparently not technologically and organizationally adapted to such an environment. This is indeed the situation that has been observed archaeologically; much of the lower southeastern Coastal Plain outside of Florida (which has its own unique environmental conditions) appears to have been largely unoccupied until late in the Paleoindian era or even into the subsequent Archaic period (Anderson 1990a, 1995a).

The homogeneous forest cover in the lower Southeast south of 33° N would, however, have been highly conducive to what have been described as residentially mobile foraging adaptations, that is, adaptations where people foraged over the landscape, readily and repeatedly moving their residences as food in their immediate area became exhausted (after Binford 1980). Archaeological assemblages from foraging adaptations are dominated by numerous short-term camps and by what are called expedient assemblages, composed of tools that were casually made, used, and then discarded on an ad hoc or situational basis. Formal, curated tools tend to be rare in such assemblages, as is the use of high-quality lithic material, unless it happens to outcrop locally. While foraging groups may, like collectors, move over large areas, each individual move tended to be fairly limited, typically no greater than necessary to place the residence near undepleted resources (R. Kelly 1983; Shott 1986a).

As the hardwood canopy expanded from its refugia below 33° N in the lower Southeast and as resource structure changed throughout the region, foraging adaptations appear to have been forced upon the resident human populations. This spread of the deciduous canopy, as noted previously, was occurring during the Paleoindian period, from circa 12,000 to 10,000 years ago, and the initial populations thus had to adjust to rapidly changing environmental conditions. The cultural changes that occurred in response are beginning to be recognized in the archaeological record, and it is during the Paleoindian period that foraging adaptations, traditionally assumed to have developed later, in the Early or Middle Archaic periods, are now thought by some to have emerged in the Southeast (Meltzer 1988; Meltzer and Smith 1986; Morse 1975a). The paleoenvironmental record, accordingly, indicates that the Paleoindian period was one of great change, both culturally and environmentally.

CHRONOLOGICAL CONSIDERATIONS

Paleoindian assemblages in the Southeast are commonly, if somewhat arbitrarily, placed into Early, Middle, and Late subperiods, with estimated temporal ranges of from circa 12,500 to 10,900 B.P., 10,900 to 10,500 B.P., and 10,500 to 10,000 B.P. (Anderson 1990a). These subperiods correspond to the occurrence of lanceolate fluted points resembling western Clovis forms; fluted and unfluted forms with broad blades and constricted hafts like the Cumberland, Suwannee, Simpson, Quad, and Beaver Lake types; and resharpened lanceolate corner- and side-notched forms like Dalton, San Patrice, Bolen, and Big Sandy. The subperiods are equated with populations exploring and colonizing the region in the Early Paleoindian, establishing regional population concentrations and cultural variants in the Middle Paleoindian, and making the switch to essentially modern flora and fauna, and the adoption of a way of life that would characterize much of the ensuing Archaic period, during the Late Paleoindian.

All dates in this manuscript, unless otherwise noted, are reported in years B.P.; these dates actually refer to radiocarbon years before the present rather than strict calendar years, reflecting the principle method currently employed to determine the age of the associated assemblages. Calibrations tying the radiocarbon time scale to calendrical dates have not been developed as of yet this far back into the past (e.g., Stuiver and Becker 1986; Stuiver and Reimer 1993), so the relationship between the radiocarbon and actual time scales is unknown. Current evidence suggests the correspondence is not at all close, however, and that an appreciable time range may be represented by the interval radiocarbon dated from circa 11,500 to 10,000 B.P., possibly as much as 2500 or more years (Faught 1996). Precise determination of the amount of time represented is, of course, crucial to the development of accurate models of colonization and settlement.

No satisfactory evidence exists at the present for pre-Clovis occupation(s) in the East. While Meadowcroft rock shelter in southwestern Pennsylvania has yielded a series of pre-Clovis dates, controversy continues about their interpretation and context (Haynes 1992:367). The Natchez pelvis, originally found with Late Pleistocene megafaunal remains, was redated using accelerator mass spectrometry to 5580 ± 80 B.P. (e.g., Cotter 1991; Maria Smith 1993:63). Lithic reduction analyses conducted over the past two decades have led to the dismissal of homegrown pre-Clovis candidates like the so-called Lively pebble tool complex from Alabama, in actuality cores from initial stage lithic reduction activity as well as expedient tools most commonly found in Late Archaic populations (Futato, this volume); the complex is now seen as a speculative construct rather than an assemblage captured in clear stratigraphic context (Steponaitis 1986:368). Two other early sites from the Southeast provide somewhat stronger evidence for the existence of possible pre-Clovis occupations in the region. Both are from Florida, and date to between 12,000 and 12,500 B.P. At Page-Ladson, five dates bracketing this interval have been obtained from a level containing a mastodon tusk with cut marks (chapter 17, this volume; S. David Webb, personal communication, 1995), and at Little Salt Springs, a wooden spear associated with a giant tortoise was dated to 12,030 ± 200 B.P. (Clausen et al. 1979:611). While these dates raise the possibility of early human occupations in the Southeast well before the currently accepted maximum age for western Clovis at 11,200 B.P. (or possibly 11,600 B.P., if the recent dating of Clovis materials at the Aubrey site in Texas holds up), because the associated artifacts are few in number, and because their existence is difficult to reconcile with theories that see Clovis technology as the signature of the region's founding population, their acceptance has been limited. Many more dates with small sigmas—and plenty of associated artifacts—will be needed from secure contexts before there will be widespread acceptance in the professional archaeological community that human occupation in the region predates ca. 11,200 B.P., or even that fluting technology could be earlier in the Southeast than in the western half of the continent. No other firm evidence supporting pre-Clovis occupations has been found in the Southeast, in spite of the extensive research- and CRM-based survey and excavation activity that has taken place in recent years.

The first unequivocal evidence for human occupation in the southeastern United States dates to about 11,500 B.P., during the terminal Pleistocene era, when assemblages characterized by fluted lanceolate projectile points appear widely over the region. These points and other materials found with them are assumed to represent the remains of human groups who entered the region from the west and spread out beyond passages in the northern ice sheets. These first, Paleoindian occupations have been provisionally grouped into three broad temporal groupings, corresponding to Early, Middle, and Late or transitional Paleoindian subperiods (Anderson et al. 1990a; O'Steen et al. 1986:9) (figure 1.2). The Early Paleoindian is thought to date from circa 11,500 to 10,800 B.P., and occupations of this subperiod are identified by the presence of classic Clovis fluted points similar to those found in the Southwest. The points are relatively large lanceolates with nearly parallel sides, ground haft margins, slightly concave bases, and single or multiple flutes that rarely extend more than a third of the way up the body (Sellards 1952; Wormington 1957). Points that resemble the classic Clovis type but for which minor typological uncertainty exists, as is common for many broken or smaller specimens, are typically assigned to a Clovis category locally and are also attributed to the Early Paleoindian subperiod. Other names sometimes used to describe these forms are Eastern Clovis and Gainey (MacDonald 1983; R. Mason 1962; Shott 1986b; Simons et al. 1984).

Early Paleoindian Clovis points are securely dated from circa 11,200 to 10,900 B.P. in the Southwest and lower Plains (Haynes 1987), while fluted forms with deeply indented bases have been dated somewhat later, to circa 10,600 B.P., at Debert and other sites in the Northeast (Levine 1990). Accordingly, if the Southeastern Clovis-like forms are contemporaneous with their western counterparts, an occurrence between circa 11,200 and 10,900 B.P. is likely (Haynes 1987, 1992, 1993; recent dating of the Aubrey Clovis site in Texas to ca. 11,500 B.P. will, if fully corroborated, extend this range appreciably backward). Unfortunately, there are still no reliable radiocarbon determinations on early fluted point assemblages from the Southeast. At the Johnson site near Nashville, for example, deeply buried hearths with associated fluted preforms yielded widely publicized dates of 11,700 ± 980, 11,980 ± 110, and 12,660 ± 970 B.P. (Broster and Barker 1992; Broster and Norton 1992, Broster et al. 1991), although recent redating indicates the feature producing the latter two dates is actually Early Archaic in age (chapter 14, this volume). The remaining date from the site, the only radiocarbon determination from the region on Clovis materials that is even close to plausible, while admittedly early, has a standard deviation or sigma so large as preclude its use in arguments about the antiquity of fluted points in the region.

The Middle Paleoindian is thought to date from circa 10,800 to 10,500 B.P. and is characterized by smaller fluted points, unfluted lanceolate points, and fluted or unfluted points with broad blades and constricted haft elements. Identifiable common southeastern forms include the Suwannee, Simpson, Clovis Variant, and Cumberland types; a number of Plains Paleoindian forms dating from the Middle Paleoindian and after, occur less commonly in the western part of the region. Unfortunately, radiocarbon dates for diagnostic projectile points are nonexistent from this interval, and we must again rely on the cross-dating of diagnostic bifaces with morphologically similar forms dated in other regions, specifically in the Northeast, on the Plains, and in the Southwest.

Unambiguously sorting waisted fluted and unfluted lanceolate points with broad blades and faint-to-pronounced ears similar to the Florida Suwannee and Simpson types is difficult, since the type descriptions and illustrated specimens for these forms exhibit considerable morphological overlap (Bullen 1958, 1962, 1975a:55–56; Simpson 1948:11–15). Simpsons typically refer to fluted and Suwannees to nonfluted waisted and eared lanceolate points otherwise meeting the type criteria. Smaller fluted forms, some of which appear to be extensively resharpened Clovis points, have been noted in the South Appalachian area, mostly in the Piedmont, and have been provisionally called Clovis Variants (Anderson et al. 1990a; Michie 1977:62–65). Their chronological placement is unknown, but is assumed to fall within the Early or Middle Paleoindian subperiods. Clovis Variants are assumed to occupy a transitional Early/Middle Paleoindian temporal category, and the same placement is accorded fluted points where it is impossible to discern whether the artifacts in question are classic Clovis points or later Middle Paleoindian types such as Suwannees or Simpsons. While the period of manufacture of the Clovis Variants cannot be determined at the present, the transitional morphology of the Clovis Variant/Simpson/Suwannee–like forms suggests an early Middle Paleoindian age, some time around or shortly after 10,800 B.P.

The Cumberland type, common in the Mid-South, is a narrow, deeply fluted, slightly waisted lanceolate with faint ears and a slightly concave base (Lewis 1954). Unfluted lanceolate points are also commonly given a Middle Paleoindian period temporal placement, particularly in the eastern part of the region. In the western part of both the Southeast and the lower Midwest, lanceolate projectile points occur in low incidence that resemble classic Great Plains Paleoindian plano-like forms, such as Folsom, Scottsbluff, Midland, and Angostura; these artifacts indicate some form of movement or interaction occurred between these two regions, although the nature of this behavior is not well understood at present (Anderson 1995a; L. Johnson 1989; Munson 1990; Patterson 1994, 1996; Wykoff and Bartlett 1995). Dating of Plains tradition forms typically follows the dating used in that region. Considerable temporal overlap of all of the Middle Paleoindian forms is probable, and it is also possible that some or all of the forms here assigned to the Middle Paleoindian period may have continued in use after 10,500 B.P., into Late Paleoindian times.

The Late Paleoindian dates from circa 10,500 to 10,000 B.P., and is characterized by Dalton, Hardaway, and presumably related points, including the Quad, San Patrice, and Beaver Lake types (Coe 1964; Goodyear 1974, 1982:390; Justice 1987:35–44; Morse 1971a–b, 1973; Webb et al. 1971) and, toward the end of the period, side-notched Taylor and Bolen points, as well as some notched varieties of San Patrice. Classic Dalton points are characterized by a lanceolate blade outline, at least in the earliest stages of tool life, and a concave base that is occasionally well thinned (sometimes to the point of true fluting) and ground on the lateral and basal margins. Blade edges may be incurvate, straight, or excurvate and are frequently serrated. Cross sections are flattened and biconvex. Beaver Lake points are small, slightly waisted lanceolates with very faint ears, a weakly concave base, and moderate basal thinning (Cambron and Hulse 1975; DeJarnette et al. 1962:47, 84; Justice 1987:35–36). Quad points are small lanceolates with distinct ears, a concave base, and pronounced basal thinning, sometimes to the point of appearing fluted (Cambron and Hulse 1975; Justice 1987:35–36; Soday 1954:9).

Beaver Lake and Quad types are assigned a transitional Middle/Late Paleoindian temporal placement, as are Dalton points exhibiting basal fluting or, more properly, pronounced thinning scars. The morphological similarity with earlier forms, particularly the presence of basal thinning, suggests an earlier occurrence than unfluted Dalton forms. The association of fluted and unfluted Dalton points in presumably contemporaneous assemblages, such as the Sloan site in northeastern Arkansas (Morse 1975a), however, suggests this is not invariably the case. Considerable temporal overlap in the range of occurrence of these forms is probable. The Late Paleoindian temporal placement for Dalton forms in general follows from arguments developed by Goodyear (1982), who examined extant radiocarbon determinations for these point types, as well as their stratigraphic occurrence. Late Paleoindian populations lived in a time of environmental change, when Late Pleistocene flora and fauna were being replaced by modern species (and, indeed, the megafaunal extinction may have been over). The Dalton point and accompanying tool kit retains many characteristics of earlier assemblages; however, the presence of serrations and the evidence for resharpening to exhaustion suggest technological differences in the use of these bifaces, when compared with earlier Paleoindian points (Goodyear 1974, 1982). These changes are increasingly linked to the emergence of foraging, generalist adaptations over the region (Claggett and Cable 1982; Goodyear 1982; Meltzer and Smith 1986; Morse 1975a–b; B. D. Smith 1986).

While this tripartite southeastern Paleoindian sequence is generally accepted, its details remain to be confirmed through stratigraphic excavations and absolute dating. Although there is a general consensus that large classic Clovis lanceolates precede the more waisted or eared fluted or nonfluted forms in the region (W. Gardner 1974:18, 1989; W. Gardner and Verrey 1979; Goodyear et al. 1979:90–96; Lepper and Meltzer 1991; McGahey 1987:7–8; Morse and Morse 1983:60–65; O'Steen et al. 1986:9), the temporal range, ordering, and extent of co-occurrence of the various Paleoindian forms remain to be securely established (e.g., see commentary by Barber and Barfield 1989; James Griffin 1977:5; Meltzer 1988:15). While stratigraphic evidence exists supporting all or parts of this relative sequence from several locations in the Southeast, most notably from sites like Hester in Mississippi, Dust Cave in Alabama, and Silver Springs in Florida (Brookes 1979; Driskell 1992, 1994; Neill 1958), only Late Paleoindian Dalton and Early Side-Notched Bolen/Big Sandy assemblages are reasonably well-dated radiometrically from locations within or near the region, at the Dust Cave, Page-Ladson (Florida), and Rodgers Shelter (Missouri) sites (Driskell 1992, 1994; Dunbar 1991; Dunbar et al. 1988; Goodyear 1982).

Five major sites from the Southeast that have been examined in recent years are helping us resolve matters of chronology, technological organization, and subsistence for the Middle Paleoindian through Early Archaic periods, a crucial interval marking both the extinction of Pleistocene fauna and the (likely) forced adoption of modern game species, as well as the onset of Holocene climatic conditions and resource structure/vegetational patterns. These include Hester in Mississippi, an open air site excavated by Brookes (1979; see also chapter 18, this volume); the Haw River floodplain open air sites in North Carolina examined by Claggett and Cable (1982; see also chapter 7, this volume); the Hardaway mountaintop workshop/base camp in North Carolina, whose assemblage data was used by Coe (1964) to help formulate the Archaic cultural sequence for the lower Southeast, and that has been the subject of extensive additional excavations in the 1970s and 1980s (Daniel 1994; see also chapter 5, this volume); the Page-Ladson site in Florida, a now-submerged and partially filled-in sinkhole in the Aucilla River bottom being examined by Dunbar and his colleagues (Dunbar 1991; Dunbar et al. 1988; see also chapter 17, this volume); and Dust Cave in Alabama, a deeply stratified rock shelter currently under excavation by Boyce Driskell (1992, 1994; Goldman-Finn and Driskell 1994; see also chapter 16, this volume) and his colleagues. Preservation of floral and faunal remains at the latter two sites, Page-Ladson and Dust Cave, where excavation and reporting are ongoing, is simply remarkable—bone needles and fishhooks, for example, have been found at the latter site. These same two sites have produced large numbers of logically-ordered radiocarbon dates, furthermore, demonstrating, among other things, that side-notched point forms (locally described as Big Sandy, Bolen, Kessell, or Early Side-Notched) first appeared around 10,200 B.P., somewhat earlier than previously thought.

Absolute dating for the Early and Middle Paleoindian subperiods is minimal outside of the Northeast, and some revision of these intervals may be needed in the years to come; changes are also likely in the assignment of specific assemblages to particular subperiods. The close technological affinities that are increasingly evident between Clovis and northeastern Arkansas/central Mississippi River valley Dalton technology, for example, suggest that Dalton may have evolved directly from Clovis in the central Mississippi River valley, perhaps as early as 10,800 years ago (Morse et al. n.d). Thus, traditional dates assigned to Dalton, from ca. 10,500 to 9900 B.P. (Goodyear 1982), may eventually be pushed back. Because appreciable variability is evident in the initial Clovis occupations in the region, furthermore, which occur in areas as diverse as the karstic river valleys of Florida, along the major drainages of the Mid-South, and along the northern glacial lake and sea margins, some of the diversity currently attributed to the Middle Paleoindian period, specifically the emergence of distinctive subregional assemblages and projectile point styles, may have actually begun earlier than circa 10,900-10,800 B.P. A more gradual or transgressive phenomenon may be indicated; even fluted point assemblages continue as late as circa 10,600 B.P. in some parts of the East (Lepper and Meltzer 1991; Levine 1990). Early, Middle, and Late Paleoindian subperiod divisions are used in this volume because the best current evidence indicates these subdivisions encompass major episodes of technological or organizational change within the Southeast. Readers should be aware, however, that these intervals will likely see some adjustment in the years ahead.

The Archaic, like the Paleoindian, has also been arbitrarily divided into three subperiods in the Southeast (e.g., Bense 1994; I. Brown 1994:48; James Griffin 1967; B. D. Smith 1986; Steponaitis 1986:370). The Early Archaic, marking the initial two millennia of the Holocene, is dated from circa 10,000 to 8,000 B.P. The beginning of the period, at 10,000 B.P., is in accordance with conventional geological dating of the Pleistocene/Holocene boundary (Harland et al. 1982:44). While a distinctive and convenient date, it is becoming increasingly apparent that natural conditions and cultural adaptations were not markedly different in the centuries immediately on either side of this boundary in the lower Southeast. A 2000-year span for the Early Archaic is employed in most sequences, with the end of the period usually equated with the onset of the Hypsithermal warming episode, at about 8000 B.P. (Stoltman 1978:714).

Early Archaic components across much of the Southeast are recognized by the occurrence of successive side- and corner-notched and bifurcate-based bifaces (e.g., Bense 1994; Broyles 1966, 1971; J. Chapman 1985a; Coe 1964). These include: Dalton and Hardaway-Dalton types, classic Late Paleoindian forms, dating circa 10,500 to 9900 B.P. and hence extending only into the earliest part of the Early Archaic period; the Taylor-Big Sandy-Bolen-Kessell side-notched types, dating circa 10,200 to 9500 B.P., another point form that first appears in the Late Paleoindian period; Palmer and Kirk Corner-Notched types, and Hardin Stemmed points, dating circa 9500 to 8800 B.P.; and a series of bifurcate forms, including the MacCorkle, St. Albans, LeCroy, and Kanawha types, dating from circa 8900 to 7800 B.P., and occurring primarily in the eastern half of the region. The end of the Early Archaic saw the replacement of these notched and bifurcate forms by square and contracting stemmed Kirk Stemmed, Stanly Stemmed, and Morrow Mountain Type I and Type II forms in the eastern part of the region (Coe 1964; J. Chapman 1975, 1985a–b; Goodyear et al. 1979; Claggett and Cable 1982; Oliver 1985). These types as well as Sykes, Benton, and Ledbetter forms are common in the Mid-South (e.g., Bense 1994; Coe 1964; Jefferies 1988). To the west of the Mississippi, in what is sometimes called the trans-Mississippi South, a wide range of forms are thought to occur during the earlier Archaic, including the San Patrice (some varieties of which appear coeval with Dalton), Yarbrough, Epps, Ensor, Evans, Kent, Wells, Bulverde and Carrollton types, some of which have affinities to the west, in Texas and on the eastern Plains (Campbell et al. 1990; Jeter et al. 1989; Patterson 1996). The dating of these Archaic forms in the western part of the Southeast is much less precise than it is for many of the point types found further to the east, primarily due to an absence of well-dated deeply stratified sites in the Louisiana and Arkansas area. The relative temporal placement of many Late Paleoindian and Early Archaic hafted biface forms was first determined by Coe at the Hardaway site in Piedmont North Carolina. Subsequent excavations across the region, at St. Albans in West Virginia and at a series of sites in the Little Tennessee River valley have led to some refinement of this sequence, which, as we shall see, has been confirmed at a number of sites across the region (e.g., Anderson et al. 1979, 1982; Broyles 1966, 1971; Claggett and Cable 1982; Michie 1969, 1971; Oliver 1985; Tippitt and Marquardt 1982, 1984; see also chapters 6–8 and 16, this volume).

2

Modeling Paleoindian and Early Archaic Settlement in the Southeast: A Historical Perspective

David G. Anderson and Kenneth E. Sassaman

INTRODUCTION

The southeastern Paleoindian and Early Archaic record is rich and varied, and Paleoindian and Early Archaic settlement systems differed markedly over the region. Coupled with this, pronounced changes in both culture and environment occurred over the 3,500 or more years these periods encompassed. Recognition of these facts is essential if we are to interpret properly the record of early human occupation over the region. The papers that follow are directed to the development of explanatory frameworks that can help guide Paleoindian and Early Archaic research in the Southeast in the years ahead. They are specifically intended to help shape the way we view and deal with the record of early human settlement.

Controlling for assemblage, climatic, and physiographic variability over time and space is critical to successful archaeological analysis and modeling activity. Variation in climate and resource structure meant that somewhat different adaptive strategies were required by the human populations in each subregion and locality, and from drainage to drainage. Models that deal with specific areas or subperiods or that focus on specific aspects of technological organization, such as the effects of raw material distribution or selection practices on settlement, must, accordingly, be capable of being placed in a larger regional and diachronic perspective.

Prior to discussion of current models of Paleoindian and Early Archaic settlement in the lower Southeast, a brief review of earlier models is appropriate. The discussion that follows is revised and updated from a paper originally prepared by Anderson and Hanson (1985). Major sites mentioned in the text are illustrated in figure 2.1.

PALEOINDIAN AND EARLY ARCHAIC MODELS PRIOR TO CIRCA 1965: SEQUENCE DEFINITION AND CHRONOLOGICAL CONCERNS

The emergence of reasonably secure assemblage and chronological information on Paleoindian and Early Archaic occupations in the Eastern Woodlands dates to the 1950s and early 1960s, with the initiation of extensive excavations at deeply stratified rock shelter sites such as Graham Cave in Missouri (Logan 1952) and Russell Cave in Alabama (Miller 1956, 1958). Artifacts recovered from these and similar sites, dated with the then newly developed radiocarbon process, were used to develop cultural sequences—focused primarily on the stratigraphic occurrence of projectile point forms—an activity that has continued to the present day. With the exception of Coe's pioneering excavations at the Doershuck and Hardaway open-air sites along and near the Yadkin River of North Carolina (initiated in the late 1940s but not published until 1964), most of this early work was conducted at caves and rock shelters. Major published reports from sites of this kind include the work at Stanfield-Worley bluff shelter (DeJarnette et al. 1962) and Russell Cave (John Griffin 1974; Miller 1956) in Alabama; at Modoc rock shelter in Illinois (Fowler 1959); at the Rodgers shelter in Missouri (McMillan 1971; Wood and McMillan 1976); and at Sheep rock shelter in Pennsylvania (Michaels and Smith 1967).

Many of the views that dominated professional archaeological assessment of Paleoindian and Early Archaic settlement systems during the 1950s and 1960s, and in some instances to the present day, were outlined by James B. Griffin (1952:354–55), in a paper entitled Culture Periods in Eastern United States Prehistory, which was largely prepared prior to the extensive (primarily) rock shelter excavations described above. Paleoindian settlement systems in the East were largely unknown at the time, but populations were assumed to have been small and highly nomadic, with a subsistence emphasis on big-game hunting. This perspective was drawn from excavations at sites in the Southwest and on the Plains, where associations with extinct megafauna had been found.

From the limited assemblage data available, together with knowledge of the general cultural stage of these Early Archaic people (James Griffin 1952:354)—an indication that ethnography shaped much of his perspective—was advanced a picture of small (approximately 20-30 member), exogamous, probably patrilineal and patrilocal egalitarian bands moving within specific hunting territories. Seasonal population movement linked to resource procurement as well as periodic aggregation for ceremonial purposes and information sharing were suggested facets of Early Archaic life:

These movements would be adjusted to the seasonal variation in the available food supplies through the natural yearly cycle such as we have in the Eastern United States. . . . Cultural exchange of ideas would be fostered by marriages outside of the local group, bringing in not only new individuals but also new ideas in regard to the surrounding territory. In addition, these groups almost certainly would have had certain times during the year to which some ceremonial importance was attached connected with the marked periodicity of different types of food supply. Probably a number of bands would congregate at food-rich locations and thus provide some degree of cultural interchange which would have enabled larger and larger aggregates of people to become familiar with larger territories. (James Griffin 1952:354–55)

Some parts of Griffin's account, particularly his characterization of Early Archaic groups as small and probably highly mobile, were practically canonized in the literature of the 1950s and 1960s. Other aspects, unfortunately, such as his emphasis on what has come to be known as seasonal resource scheduling and the importance of periodic aggregation for ceremonial, information, and mating network maintenance, received considerably less attention. Importantly, an increasing familiarity with