Geoarchaeology 9780300157345

Table of contents :
Contents
Preface to the Second Edition
Preface to the First Edition
1. Theory and History
2. Sediments, Soils, and Environmental Interpretations
3. Initial Context and Site Formation
4. Methods of Discovery and Spatial Analyses
5. Estimating Time
6. Paleoenvironmental Reconstructions: Landscapes and the Human Past
7. Raw Materials and Resources
8. Sourcing (Provenance)
9. Construction, Destruction, Archaeological Resource Preservation, and Conservation
10. Epilogue
Notes
Glossary
Bibliography
Index

Citation preview

Geoarchaeology

Geoarchaeology The Earth-Science Approach to Archaeological Interpretation Second Edition

George (Rip) Rapp and Christopher L. Hill

Yale University Press New Haven and London

Published with assistance from the Louis Stern Memorial Fund. Copyright © 2006 by Yale University. All rights reserved. This book may not be reproduced, in whole or in part, including illustrations, in any form (beyond that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press), without written permission from the publishers. Set in Janson and Gill Sans types by Tseng Information Systems, Inc. Printed in the United States of America. Library of Congress Cataloging-in-Publication Data Rapp, George Robert, 1930– Geoarchaeology : the earth-science approach to archaeological interpretation / George (Rip) Rapp and Christopher L. Hill. — 2nd ed. p. cm. Includes bibliographical references and index. isbn-13 : 978-0-300-10966-5 (paperbound : alk. paper) isbn-10 : 0-300-10966-0 (paperbound : alk. paper) 1. Archaeology—Methodology. 2. Archaeological geology. I. Hill, Christopher L., 1959– II. Title. cc77.5.r37 2006 930.10285—dc22 2005029099 A catalogue record for this book is available from the British Library. The paper in this book meets the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources. 10

9

8

7

6

5

4

3

2

1

To Malcolm H. Wiener, a philanthropist and archaeologist with a keen vision of the role of natural science in archaeology

Every archaeological problem starts as a problem in geoarchaeology. —Colin Renfrew

Contents

Preface to the Second Edition xi Preface to the First Edition xiii

Soils and Buried Soils 38 The Soil Profile 39 Soil Types 41

1 Theory and History 1

Entisols 41 Vertisols 41

The Scope of Geoarchaeology 1

Inceptisols 41

Archaeology and the Earth Sciences 4

Mollisols 41

Foundational Phase: Before 1900 5

Alfisols 42

Collaborative Phase: 1900–1950 10

Ultisols 42

Integrative Phase: After 1950 16

Spodosols 42

Changing the Guard 20

Aridosols 42 Histosols 42

2 Sediments, Soils, and Environmental Interpretations 25 Sediments 25 Weathering 26 Transportation 27 Postdepositional Changes 28 Archaeological Implications 28 Classification of Sedimentation Products 29 Clastic Deposits 29 Gravel 32

Paleosols and Buried Soils 43 Inferring Environments from Physical and Chemical Parameters 45 Color 45 Cementation and Induration 47 Texture 47 Structure 53 Composition 54 Boundaries 58 Micromorphology 59

Sand 33 Mud 34 Chemical Deposition 35 Calcareous Precipitates 36 Noncalcareous Precipitates 36 Organic Matter 37

3 Initial Context and Site Formation 60 The Creation of the Archaeological Record 60 Stages of Site Formation 62 Initial Landscapes and Original Occupation 62

vii

Contents

Sedimentary Contexts 63 Desert Depositional Systems 64

Seismic Profiling 117 Magnetic Analysis 118

The Effects of Wind 64

Aerial Photography 118

The Effects of Moisture 65

Satellite and Airborne Remote Sensing 120

Desert System Site Formation 66

Dowsing 121

Alluvial Depositional Systems: Flowing Water 67 Depositional Contexts 68 Site Formation in Alluvial Settings 75 Lakes and Associated Basin Settings 78 Basin Deposits 78

Geochemical Prospecting and Analysis 122 Core Drilling 125 Locating Water Resources 129 Geographic Information Systems 129 The Complexities of Scale 130

Site Formation in Basin Settings 80 Cave and Rock Shelter Depositional Systems 81 Limestone Caves 81

5 Estimating Time 132

Sandstone Caves and Rock Shelters 85

Climate Change and Time 133

Igneous Rock Caves 85

Artifacts and Dating 134

Site Formation in Caves 85

Stratigraphy 135

The Glacial System 86

Rhythmites (Varves) 136

Coastal and Marine Depositional Settings 90

Paleosols in Loess and Alluvium 139

Coastal Processes and Site Formation 91 Coastal Landscape Context 95 Postdepositional Processes 98 Mass Wasting 99 Cryoturbation 99 Bioturbation 100

Tephrochronology 142 Dating Using Animal and Plant Fossils 143 Paleontology 143 Dendrochronology 144 Radiometric Dating Methods 144 Potassium-Argon and Argon-Argon Dating 144 Uranium-Series Dating 146

4 Methods of Discovery and Spatial Analyses 103 Maps 103

Other Dating Methods 153 Fission-Track Dating 153

Landform Sediment Assemblages 109

Paleomagnetic and Archaeomagnetic Dating 153

Settlement Patterns 109

Electron Spin Resonance and Luminescence Dating 156

Remote Sensing 110 Geophysical Prospecting 111

Temperature-Affected Dating 160

Magnetometry and Magnetic Properties of Soils and Sediments 113

Amino Acid Racemization and Epimerization 160

Electrical Resistivity 115

Hydration (Obsidian) 160

Electromagnetic Conductivity 116 Ground Penetrating Radar 116

viii

Radiocarbon Dating 147

Contents

Dating Techniques Based on Chemical Accumulation 162 Chemical Analysis 162 Dating Exposed Surfaces 163 Patination and Desert Varnish 163 Cosmogenic Nuclides 163

Human Habitats and Geoecology 186 Tectonics, Climates, Landscapes, and the Human Past 188 Microclimates 191 Human Interaction with Environment and Its Effects on Climate 192

6 Paleoenvironmental Reconstructions: Landscapes and the Human Past 165

7 Raw Materials and Resources 195

Environmental and Landscape Change 165

Minerals 196

Definitions 195

Inferring Environmental Change 166

Chert and Chalcedony 196

Ecology and Landscape Change 167

Semiprecious Stones 197 Other Archaeologically Important Minerals 199

Terrestrial (Non-marine) Geoecologic Data from Lake Records 168

Metals and Ores 202

Plant (Botanical) Indicators 169

Rocks 207

Microfossils 169 Pollen 169 Phytoliths 172 Diatoms 174 Macrofossils 175 Animal Indicators 176 Invertebrates 176 Ostracods 176 Mollusks 178 Insects 179 Vertebrates 180 Mammal Fossil Remains 180

Shells 210 Clays 211 Building Materials 213 Building Stone 214 Burnt Brick 215 Mortar 215 Other Materials 216 Pigments 217 Abrasives 218 Rock and Mineral Recovery 218 Water 219

Bird Fossil Remains 181 Reptile and Amphibian Fossil Remains 182 Fish Fossil Remains 182 Other Ecologic Accumulations 183 Pack-Rat Middens 183 Peat 183 Geochemical Indicators 183 Environmental Change and Archaeological Explanations 186

8 Sourcing (Provenance) 222 Geologic Deposits 224 Materials Used in Sourcing 225 Obsidian 225 Sourcing Other Igneous and Metamorphic Rocks 226 Chert 227 Marble 228

ix

Contents

Clay 229

Pollution 252

Temper 229

Water 254

Amber 230

Erosion and Subsidence in Archaeology 256

Bitumen 230

Erosion 256

Soft Stone, Other Rocks, and Semiprecious Minerals 230

Land Subsidence 256

Native Copper 231 Complex Copper Minerals 232 Tin 232 Lead, Silver, and Gold 234 Sourcing Methods 235 Trace-Element Analyses 235 Instrumental Neutron Activation Analysis 236 X-Ray Fluorescence Spectrometry 237 Isotope Analysis/Mass Spectrometry 237 DNA 238 Mineral Magnetism 238 Other Analytical Methods 238 Petrographic Analysis 238 Statistics and Data Analysis 242

Geologic ‘‘Catastrophes’’ and the Human Past 257 Earthquakes and Seismic Disturbance 257 Floods and Flood Legends 262 Volcanoes 263 Site Preservation 265 Site Preservation Problems 265 Reservoirs 267 Hilltops or Slopes 268 Seismicity 268 Site Stabilization 268 Earth Burial 269 Archaeological/Cultural Resource Management 270 Resource Management 270 Conservation 271 Materials Preservation 271

9 Construction, Destruction, Archaeological Resource Preservation, and Conservation 244

Corrosion 271

Geotechnology 244

10 Epilogue 273

Construction 245

The Future 273

Dams 245 Canals 246

The Future of Geoarchaeology within Archaeology 273

Fields and Raised Fields 247 Roads 248

Notes 277

Excavation (Mines and Quarries) 248

Glossary 287

Natural Burial and Site Formation 249

Bibliography 291

Rock Properties and Weathering 250

Index 319

Destruction 250 Weathering 251

x

Preface to the Second Edition

T

his edition is a significant revision of the original edition. We have benefited from using the first edition as a textbook in geoarchaeology courses, from informal comments of other instructors and practitioners, and from published reviews. In writing this Preface to the Second Edition we assume readers have read the original Preface. Rapp started working in archaeology and geoarchaeology in Greece in 1967. In the nearly four decades since then, the extent to which archaeology has incorporated the earth sciences has increased profoundly. Archaeological excavation and survey are now multidisciplinary endeavors. Geoarchaeology is one of the core disciplines involved. We have attempted to address most of the topics that anyone doing geoarchaeology might find useful in excavation and survey. A geoarchaeologist may be the only expedition member with broad training in the natural sciences and therefore called on to make rapid judgments and carry out studies (including sampling) that cannot be returned to later. Both of us have had extensive field experience, which has led us to make this volume a broad introduction to the practice of geoarchaeology. Most human activities leave behind a physical record. Early archaeologists concentrated on artifacts, features, and human remains to reconstruct past human activities. As archaeology became more multidisciplinary, earth scientists were among the first, if not the first, natural scientists to investigate the remains of human activities using methods they had developed to determine the evolutionary history of the earth. Humanenvironmental interactions can be divided into

human interactions with the biotic world of plants and animals and human interactions with the essentially abiotic world of physical materials and dynamic landscapes. Many aspects of human– biotic world interactions (such as deforestation) leave a geologic record, and the human interactions with earth materials and landscapes almost always leave remains. All of this constitutes the realm of geoarchaeology. We have followed our original philosophy of not writing an ‘‘earth science for archaeologists’’ book but rather presenting as many as possible of the topics that geoarchaeologists and archaeologists currently use. This means embedding geologic concepts, methodologies, and knowledge base in the archaeological context where they apply, with appropriate references. The embedded earth science does not follow the sequence of topics in introductory physical geology, so it may appear disjointed to some readers. We want the archaeological problems that are addressed to govern the introduction of the relevant earth science. Organizational problems will always remain. Geoarchaeological determinations of the nature and extent of human activity areas are frequently based on geophysics and geochemistry. From an earth-science perspective this topic would be considered under geophysics and under geochemistry. Yet from an archaeological perspective, the essential context is activity areas. As topics are covered in this volume we have sometimes taken the earth-science perspective when topics did not fit well into any of the focused archaeological contexts, but more often we were able to cover the topic using an archaeological perspective.

xi

Preface to the Second Edition

We are gratified that the original edition has received a good response. Although there has been a broad readership, our focus remains on first-year graduate students in archaeology and those in the earth sciences considering a focus on geoarchaeology. One reviewer has suggested that we incorporate some archaeological background for readers who are geologists interested in geoarchaeology. Aside from the archaeology inherent in out myriad of examples, we continue to reject using valuable space to offer archaeology that is so well presented in many readily available textbooks and other volumes. Archaeologists, particularly in the United States and the United Kingdom, have been engaged in a rapidly shifting debate over the past few decades concerning anthropological and archaeological method and theory. Our experience has been that this debate has not had a significant effect on the practice of geoarchaeology. However, we have incorporated in Chapters 1 and 10 brief summaries of our view of the impact of shifts in ruling archaeological paradigms on aspects of early-twenty-first-century archaeology. All authors struggle to decide what to include and what to exclude. In terms of the earthscience that archaeologists and geoarchaeologists use, we have attempted to include those fields or subfields normally considered earth science: physical geology, geomorphology (including physical geography), sedimentology, stratigraphy, mineralogy, petrography, pedology, geochemistry, geophysics, climatology, and aspects of geohydrology, palynology, and paleontology. There were some serious omissions in the original edition, including micromorphology. We have tried to remedy this. Generally excluded are biologic remains except where they are important components of sediments and used as climate or environment surrogates or in other ways enhance our understanding of archaeological sediments. The most difficult

xii

decisions relate to analytical techniques. Radiocarbon dating is discussed only briefly except where geology comes into play, as in the case of reservoir effects. Chemical techniques for provenance determination are included as are analytical methods for minor- or trace-element patterns [such as for phosphates] that reflect human impacts. Archaeological sites and geographic names are spelled as they most often appear in the sources we have used in writing this book. Following the new custom, we are using b.c.e. (Before the Common Era) instead of b.c. The number of references in the Bibliography has nearly doubled from the First Edition. Last, we have revised the organization of topics and chapters to reflect an archaeological approach more closely. The sequence in this Second Edition is: theory and history of geoarchaeology; soils, and environmental interpretations; initial context and site formation; methods of discovery and spatial analyses; estimating time; paleoenvironments, landscapes, and the human past; identifying, analyzing and sourcing raw materials; and, finally, construction, destruction, and conservation of sites and materials. In preparing this Second Edition we had vital editorial help from Charisa Homan and, especially, Janice Wallace, to whom we are deeply indebted for putting all of this together. Russell Rothe crafted ten new computer diagrams, and Elaine Nissen drew seven new figures. We are grateful for the excellent work by the people at Yale University Press, especially the expertise of Jean Black, Laura Davulis, and Joyce Ippolito. The revision was strengthened by feedback from students participating in Hill’s course on geoarchaeology and paleoecology at Boise State University. Finally, Hill would like to thank Cheryl Hill for her help and support.

Preface to the First Edition

T

his book was written primarily for archaeologists in the formative stages of their careers and secondarily for geologists who, in increasing numbers, are assisting in the solution of archaeological problems. Finally, we hope it might also be of value to senior archaeologists, as well as to historians, anthropologists, ethnologists, and Quaternary scientists whose scholarship requires some understanding of the physical context of the remains of past material culture, as represented in the artifactual record. It is generally believed that no academic discipline has been spared the experience, at some stage of its development, of sometimes bitter and wide-ranging disputes over method and theory. Archaeology has witnessed such disputes since the 1970s. Because the methodology inherent in most geoarchaeology is geologic, it is possible that this subdiscipline can avoid these disputes. Geology settled its fundamental differences about method and theory a long time ago. (This is not to say that there are not current differences of emphasis and priority.) Rapp co-edited a book published in 1985 by Yale University Press entitled Archaeological Geology. This volume is called Geoarchaeology. We take these names to have somewhat different meanings. A narrow definition of archaeological geology would be geology performed with at least the partial objective of being useful to archaeology. A good example would be the determination (using core drilling, sedimentology, and geomorphology) of the varying positions of shorelines during the Late Quaternary near important archaeological sites. This is mainstream geology, whether or not there are archaeological sites in

the vicinity. In contrast, geoarchaeology, narrowly defined, is the use of geologic concepts, methods, and knowledge base in the direct solution of archaeological problems. Geoarchaeologists do archaeology. Obviously, there is some overlap in actual practice between the two. To distinguish the remains of human activity from those of natural (geologic) agencies is perhaps the first task of the geoarchaeologist. We start from the premise that most of the interpretation of archaeological materials, sediments, and site settings requires a broad understanding of several related methods and concepts. This is the first book to take an integrated and comprehensive approach to these areas. The discussion of each geologic concept or knowledge base is framed within an archaeological context. The principal aim is to provide concepts and data derived from the earth sciences that are essential to solving archaeological problems and aiding interpretations of the archaeological record. We hope that this book demonstrates the necessity of taking an earth-science approach to the examination of diverse archaeological settings, ranging from single occupations to complex, multicomponent habitations. Our examples illustrate how earth science is used to resolve archaeological problems associated with artifact identification and description, the integrity of artifact sets, chronological context, paleo-landscape habitat, and human-environment interactions. When dealing with an interdisciplinary area, it is tempting to explain every phenomenon drawn into the discussion. We have tried to limit our exposition to the geologic aspects of the topic.

xiii

Preface to the First Edition

For example, radiocarbon dating is of great importance in archaeology. However, only certain aspects of carbon-14 ( 14 C) dating are explicitly geologic, and texts abound that cover this field at every level of detail and depth. Therefore we focus on its geologic components. Nomenclature and basic data (such as chronological periods) from many fields must be imported into an interdisciplinary text. We have therefore added a glossary and appendixes so as not to unduly interrupt readers’ concentration by forcing them to consult specialized dictionaries and encyclopedias. Words in the glossary are set in boldface type at first mention. We intend this to be an introductory textbook for advanced-level undergraduates and graduate students in archaeology and geoarchaeology. It is not a scholarly treatise written for colleagues nor a book-length review containing exhaustive references. Some of the material—for example, the discussion of basic rock types in Chapter 7— is rudimentary; more advanced readers may skip this chapter. The authors faced the choice of offering a text with hundreds of references interrupting the discussion or one containing only the specific references from which significant data, unique ideas, or quotations were drawn. Like most textbook authors we chose the latter; in addition, we replaced the standard author-date reference style ( Jones and Bryant 1992) with endnotes and a bibliography to make the reference material readily available but unobtrusive. Scientists have begun to drop the -al from many scientific adjectives, like topographic, stratigraphic, geomorphic, and climatologic, and geologists now tend to prefer geologic to geological. We have followed this practice. Archaeologists who reviewed the manuscript, however, overwhelmingly requested that we retain the traditional archaeological. So we did. Our perspective is unabashedly geologic. Hence, other equally important aspects of the broad interdisciplinary field of archaeology are often not mentioned, even when they contribute to the archaeological examples we have used. This is not to minimize their importance but rather to focus on the geologic aspects. There is no short-

xiv

age of good books and articles covering these other components of archaeology. Rapp has done field geoarchaeology in Greece, Turkey, Cyprus, Israel, Egypt, Tunisia, Ukraine, and China, as well as in North America, and has visited geoarchaeological sites in Italy, Germany, Great Britain, Russia, India, Thailand, Iran, and elsewhere. Hill has done geoarchaeological fieldwork in Egypt, Turkey, and Israel, as well as in North America. We are more comfortable with examples that fall within our range of experience, but this special knowledge has biased the examples given in the text. For this we ask the reader’s indulgence. Early drafts of many chapters were used in a class in geoarchaeology at Boston University in the fall of 1993; later drafts of all the chapters were used as the text for a class in geoarchaeology at the University of Minnesota, Duluth, in the winter of 1995; and still later drafts were used in a geoarchaeology class at the University of Minnesota, Twin Cities, in the spring of 1996. We acknowledge the valuable comments and suggestions of students in these classes. The Duluth class pressed for the inclusion of cost data for chemical analyses and geophysical techniques. We have resisted this because of the many variables involved and because costs constantly change. We are indebted to many colleagues. Doris Stoessel was instrumental in keeping everything under control and editing the final stages of the manuscript. Russell Rothe created most of the computer illustrations. Elaine Nissen drew some of the figures, while Zichun Jing, Emine Kucuk, and Jennifer Shafer assisted with other illustrations. Jean Thomson Black and Edward Tripp of Yale University Press persevered during the long gestation of this book. Mr. Tripp was the spark behind this volume. Our two anonymous reviewers were tough and insightful, and our Yale University Press editor, Susan Laity, made this a much more readable book. Hill would like to express his gratitude to Rapp for providing the opportunity to help contribute to his vision of an introductory text in geoarchaeology, and for more than twenty years of

Preface to the First Edition

support and friendship. In addition, many other individuals have contributed directly or indirectly to this work. Charles Matsch initiated Hill into the wonders of geology and introduced him to Rapp. He also reviewed versions of many of the chapters of this book. David Meltzer provided a critique of an early version of Chapter 1. During graduate studies at Southern Methodist University, Hill was guided and supported by Fred Wendorf. Claude Albritton and the Institute for the Study of Earth and Man provided a framework

for geoarchaeological studies at SMU. Romuald Schild guided Hill in the application of a geoarchaeological approach to field studies in African prehistory. The Museum of the Rockies at Montana State University provided a wonderful setting in which to complete the text. Hill would also like to acknowledge with thanks the help given him by students in the anthropology department at Tulane University and in the department of earth sciences at Montana State University in developing some of the ideas presented in this book.

xv

Geoarchaeology

CHAPTER 1

Theory and History

Knowledge is a continuum, like the sphere of the earth but with the uninterrupted vastness of a universe.—Roald Fryxell 1977

G

eoarchaeology is critical to understanding the archaeological record and the human past. Indeed, the earth sciences play a pivotal role in interpreting evidence of the human past in terms of both concepts and techniques. The antecedents for using an earth-science approach in evaluating the archaeological record lie in the history of the disciplines, beginning with the concern in the eighteenth and nineteenth centuries for an appreciation of ‘‘prehistoric’’ time and the development of basic principles, moving into the twentieth century with the collaboration between natural scientists and archaeologists, and arriving at the present with the convergence of the two disciplines in a number of areas. Throughout the history of this interaction, ideas and methods originating with the earth sciences have been used to study the processes involved in the formation of the sedimentary archaeological record, to infer the paleoenvironmental settings associated with archaeological sites, to develop chronological frameworks, and to measure and classify the physical characteristics of artifacts and geofacts. Here we offer an overview of this history and explore how the geoarchaeological approach attempts to evaluate and understand the archaeological record and the dynamic forces that produced it.

The Scope of Geoarchaeology The term geoarchaeology (or, less commonly, geoarchaeology) has been used with increasing frequency since the 1970s to designate a variety of types of research that use geoscience techniques in the evaluation of the archaeological record. The labeling of research as an aspect or a subsidiary of geoarchaeology depends in part on whether the term is used to designate a narrower, more focused set of concepts and methods or a broader, more inclusive set. There are many viewpoints concerning what can appropriately be called geoarchaeology. Lars-Konig Konigsson, for example, contrasts geoarchaeology with archaeogeology. In his view, archaeogeology is a complementary science that is useful in describing deposits related to archaeological material. It is seen as having an advisory role in archaeological interpretation. Konigsson describes geoarchaeology, in contrast, as a study in which the geologist tries to determine the ‘‘cultural’’ development of an area; geologists are not in direct collaboration with archaeologists and rely exclusively on geologic materials and methods.1 Reid Ferring’s definition of geoarchaeology emphasizes the changes in archaeological perspective since the advent of Lewis Binford’s ‘‘New Archaeology.’’ As practiced, Ferring sees geoarchaeology as a ‘‘grossly empirical approach to archaeological problems’’ or the ‘‘new empiricism.’’ 2 In perhaps its broadest sense and the way the term is used in this book geoarchaeology refers to the application of any earthscience concept, technique, or knowledge base to the study of artifacts and the processes involved in the creation of the archaeological record. Geo-

1

Theory and History

archaeology thus becomes ‘‘the geoscience tradition within archaeology . . . [that] deals with earth history within the time frame of human history’’ or that ‘‘implies archaeological research using the methods and concepts of the earth sciences.’’ 3 However, the term has also been limited specifically to the study and interpretation of sediments and physical landscapes. As Charles French has commented, the more narrow definition of geoarchaeology that emphasizes the study of soils and sediments serves as a basis for the study of landscape evolution and the human past.4 There is also the question of the relation between geoarchaeology and archaeological geology. Our view is that geoarchaeology is part of archaeology. It is the part that uses geoscience methods, concepts, and knowledge base. The geomodifies the noun archaeology. An example of geoarchaeology would be the study of archaeological sediments in an excavation: the framework and questions posed are strictly archaeological with the goal of understanding the human past. In contrast, archaeological geology is geologic research that has direct relevance to one or more archaeological contexts. Coastal change studies that determine shoreline migration, the filling in of estuaries, or the expansion of deltas in archaeologically important areas fit this definition. Geologists routinely investigate the changes in land-sea boundaries that have transpired over hundreds of millions of years. This research is not archaeological per se. A historically interesting example of what we would consider within the scope of archaeological geology is found in the comments by Helmut de Terra, published in Science in 1934: ‘‘I had been working in Northern India and . . . found an artifact . . . due to my ignorance I was unable, temporarily, to use it for my stratigraphic work . . . . [H]ad I known then, as I do now, that the tool belonged to an early Levalloisian type . . . associated with Middle Pleistocene strata, I would have had a definite lead to guide my work in Pleistocene stratigraphy.’’ 5 In this case information derived from archaeological studies would have been applied to answer a geological issue. Also, when paleogeomorphic maps of dramatic landscape changes in such significant archaeological

2

terrain as Ancient Troy, Ancient Carthage, or the east coast of North America are undertaken because of the sites’ archaeological importance, such studies can be considered archaeological geology. Many kinds of geologic investigations in archaeological contexts lie between these two extremes. The geoarchaeology label has been applied by researchers from a wide range of earth-science disciplines (in addition to archaeology itself ), including stratigraphy, sedimentology, geomorphology, pedology, petrology, petrography, geochemistry, geophysics, paleontology, marine geology, geochronology, and climatology. Almost every subdiscipline in geology has concepts, methods, and a knowledge base that can contribute to the solution of archaeological problems. Geoarchaeology in its widest scope or most encompassing form is the application of these earthscience disciplines and subfields to the study of the archaeological record. The key criterion is that archaeological interpretations are produced using earth-science-based ideas or methods. Because geo- modifies archaeology, geoarchaeology must be defined as a type of archaeology, one in which the fundamental goal of archaeology is understood to be the study of the artifactual record and where any physical object, feature, or landscape either made or altered by humans (including extinct hominids) is considered an artifact. We are aware that this is a rather strict, bare-bones definition of archaeology, designed to accommodate research focused on inferring past anthropological behavior as well as research that seeks to evaluate nonbehavioral features of the archaeological record. In our view a major goal of archaeology is to evaluate and understand past human behavior, but a broader goal is to understand all the processes that contribute to the final artifactual context available for observation and study. Where geoarchaeology fits in relation to other disciplines and subdisciplines has been the subject of a surprising amount of discussion. When viewed from the broad perspective, geoarchaeology would include many aspects of archaeometry, environmental archaeology, archaeological science, Quaternary geology and geography (including geomorphology, physical geography,

Theory and History

geoecology, and biogeography), taphonomy, and bridging (middle-range) theory. From another perspective, geoarchaeology is principally a part of the framework for an ecologic or contextual approach to archaeology (not to be confused with the very different ‘‘contextual archaeology’’ as an aspect of postprocessual archaeology).6 There have even been those who would classify archaeology as a subdivision of the natural sciences or natural history or as a branch of Late Quaternary geology, where geology is understood, following Amadeus W. Graubau, as ‘‘the science of the entire earth.’’ 7 The viewpoint expressed by Frederick H. West supports such an archaeogeologic perspective: ‘‘Archaeology . . . is an earth science.’’ 8 Waters has thoughtfully proposed applying the terms geoarchaeology and archaeometry for different aspects of the geoscience-archaeology connection. Whereas the scope of geoarchaeology would include documenting site stratigraphy, determining site formation processes, and reconstructing landscape-human interaction, archaeometry would include the application of the geosciences to archaeological prospecting, provenance, and dating.9 It could be proposed, then, that geoarchaeology represents a particular focus of archaeology that may be considered, along with climatology, hydrology, lithology, and ecology, as a subdivision of earth science in the broadest sense. Indeed, it has been suggested that archaeological sites are really geologic sites containing remains that are of interest to archaeologists.10 Aspects of geoarchaeology, in turn, can be considered from the viewpoints of ‘‘dynamics,’’ ‘‘structure,’’ and ‘‘chronology’’ (when archaeology is viewed as one of the historical natural sciences). Dynamics deals with the effects of physical and chemical forces and is process-oriented. Structure deals with composition and arrangement of materials. Chronology and history are concerned with time, origins, and development. Archaeology has been considered a natural science of the archaeological record,11 and it can be viewed as a division of the earth sciences, especially in terms of some methodological practices.12 However, the archaeological record can be evaluated and interpreted even

more thoroughly by considering dynamics, structure, and history (both chronology and development through time). In contrast to most other physical and natural sciences, geology (in the restricted sense), paleontology, and archaeology share these concerns. The behavioral aspects of archaeology are the focus of anthropological archaeology—which is a traditional field of interest in the United States and Great Britain. From a geoarchaeological perspective the goals of anthropological archaeology —that is, inferring past human behavior from the artifactual record—can be attained only by examining the contextual association of the artifacts and developing models to infer the behavioral and nonbehavioral factors that have produced the record. Under these circumstances, even the application of such terms as culture for what is in reality a collection of items that may or may not reveal patterns created by past human activity may not be always appropriate. The terms geoarchaeology and archaeological geology have also been used since the 1970s to designate the earth-science aspects of archaeological studies, with the recognition that evaluations of prehistoric human behavior rely on contributions from the earth sciences.13 Throughout this period Karl Butzer advocated an ecologic or broadly contextual approach to archaeology that emphasized the application of what he called geoarchaeology: ‘‘It has been said that archaeology is anthropology or it is nothing. . . . I beg to differ with this view. Archaeology . . . has been equally dependent on geology, biology, and geography . . . during its development . . . [and] is heavily dependent on . . . the natural sciences.’’ 14 This is similar to the point of view expressed earlier, in the 1950s, by Mortimer Wheeler, when he wrote: ‘‘Archaeology is increasingly dependent on a multitude of sciences and is itself increasingly adopting the methodology of a natural science. It draws today upon physics, chemistry, geology, biology, economics, political science, sociology, climatology, botany.’’ 15 Another, more restricted use of geoarchaeology would make it analogous to zooarchaeology. From this viewpoint, geoarchaeology is focused primarily on soils, sediments, and stratigraphy

3

Theory and History

(instead of life-forms). This general view is expressed in the writings of both Bruce Gladfelter and Michael Waters. Gladfelter characterizes geoarchaeology as primarily geomorphology and sedimentary petrography, whereas Waters considers the most fundamental aspects of geoarchaeology to be the field aspect of stratigraphy, site formation processes, and landscape reconstruction.16 David Cremeens and John Hart take a similar position, stating that geoarchaeologists attempt to answer two broad questions: (1) why were particular locations selected by prehistoric people, and (2) what events have transformed the original record of prehistoric human activity? 17 Fekri Hassan sees the study of site formation processes as a way of interpreting the ‘‘cultural’’ significance of archaeological remains by elucidating the role cultural and natural factors played in shaping the archaeological record. He argues that the study of formation processes has proven that archaeologists cannot assume that artifact collections correspond to a settlement, ethnic group, or cultural activity. The study of formation processes should convince archaeologists to be careful about making cultural interpretations on the basis of the archaeological record.18 Broader connections between the earth sciences and archaeology such as dating techniques, provenance of artifacts, site location, and the like have been assigned by some practitioners to the realm of archaeometry. In the broader geoarchaeological approach that we advocate, such aspects of archaeometry as dating, provenance, and site location become part of geoarchaeology when they represent the application of earth-science methods to archaeological problems. The fluidness of the definition of geoarchaeology is a manifestation of the classification process within science. One can consider geoarchaeology as a component of prehistoric archaeology that, in turn, may be considered a part of geoecology or paleogeography, which is an aspect of Quaternary geology, and so on. These research fields and subdisciplines can be considered ‘‘facies’’ within a broader framework of natural history and natural science focused on the evaluation of the complete Late Cenozoic record. The primary connecting

4

point may be that these research areas attempt to observe systematically the dynamic processes that occur on the earth and to apply these observations to the inference of past conditions. In this sense archaeology is an aspect of the historical sciences much as paleontology is. Archaeology is distinct from cultural anthropology, by contrast, in the sense that past processes cannot be directly observed but must be inferred from the available record. Both ethnological anthropology and physical geology allow direct observation of dynamic processes, which in turn can be applied to an interpretation of the structural matrix of the archaeogeologic record. From this matrix the chronological and contextual sequence of events that form prehistory can be inferred.

Archaeology and the Earth Sciences The historical connections between archaeology, especially prehistoric archaeology, and the geosciences extend back at least through the nineteenth and eighteenth centuries.19 Several phases in this interaction, which fit into three overlapping periods, can be identified. The first of these periods was characterized by an integrative approach to prehistory; it can be viewed as a period where the primary concern was human antiquity. The nineteenth century traditionally has been characterized as a time when interest focused on evidence concerning the early human occupation of Europe and North America during the Ice Age. From about the 1840s through the 1920s, geology and archaeology were united in the study of evidence for human antiquity. In addition to the interest in relative chronologies, archaeosedimentary sequences (strata containing artifacts) were studied to determine what processes were involved in their formation and how to evaluate associations among artifacts and remains of extinct fauna. During the second phase, from the end of the nineteenth century to about 1950, interest in paleoenvironmental and paleoclimatologic conditions expanded the sphere of interaction between prehistoric archaeology and the geo-

Theory and History

sciences. Collaboration between geologists and archaeologists continued to be an important facet of Paleo-Indian studies in North America after the late 1920s, and a major characteristic of Old World Paleolithic research. Sedimentary sequences or stratified deposits continued to be studied for their site formation and chronological implications, but until about the 1950s, the collaboration appears to have emphasized the use of geoscience methods to evaluate the paleoclimatic and geochronological contexts of archaeological sites. This second period was characterized by a shift from the integrative approach, in which various scientists would together publish a single report, and toward collaboration among specialists, in which each scientist would publish a separate report. There are three variants of ‘‘archaeogeologic’’ efforts during this period: regional geomorphologic studies aimed at developing paleoclimatic and geochronological frameworks; laboratory studies of ecologic and artifactual materials; and site-specific studies of deposits containing artifact sequences. Around the beginning of the second half of the twentieth century, a third phase of interaction began. Although scientists continued to pursue mutual interests in chronology and climatologic change, they also began to combine new emphases and interests in prehistoric archaeology and the geosciences. These new interests had a common theoretical basis that emphasized the ‘‘context’’ of archaeological sites. Archaeologists and prehistorians also believed that the natural sciences could help answer new questions derived from these theoretical developments. This period of convergence owes its existence to a concern for the processes that have transformed the original context of the artifacts into the archaeological record.20

Foundational Phase: Before 1900 In a review of the development of archaeological geology, John Gifford and Rapp delineated a period in which interdisciplinary research was implicit, rather than overt.21 This period had already started by the 1840s. It was characterized by an interest in human antiquity and, consequently, stratigraphic chronology. In both the Old World

and North America the natural science perspective was reflected in the efforts of individuals, mostly geologists by training, who in their studies of the prehistoric record took for granted that artifacts found in geologic deposits had been used by humans. In his history of archaeology, Glyn Daniel notes: ‘‘Many great nineteenth-century archaeologists were in the first instance geologists and natural scientists,’’ and nineteenth-century archaeology ‘‘was essentially geological in outlook.’’ 22 Edward Harris seems to agree; he writes that during this time archaeological research ‘‘was dominated by theories of geological stratigraphy.’’ 23 The early recognition of the importance of understanding the context of artifacts is reflected in the report that John Frere wrote in 1797 about the discovery of stone hand axes within a stratified sedimentary sequence in England. Frere describes the location of the artifacts and the sequence of sediments, and he discusses the implications of these observations: ‘‘The manner in which [the hand axes] lie would lead to the persuasion that it was a place of their manufacture and not of their accidental deposit. . . . It may be suggested that the different strata were formed by inundations happening at distant periods.24 In Frere’s report the horizontal bedding and the density of artifacts in the deposits are delineated. His observations clearly establish that from the beginnings of Paleolithic research, an awareness of the value of evaluating artifacts within their stratigraphic context was understood. The history of the demonstration of the contemporaneity of humans and extinct animals (and consequently the acceptance of the presence of humans during the Ice Age in Europe) provides many examples of the early use of techniques derived from the earth sciences in the study of the human past. The events leading to this demonstration and acceptance have been chronicled since the mid-1800s. Research initiated by Boucher de Crèvecoeur de Perthes in 1837 at Somme Valley in France and continued by Marcel Jérôme Rigollot illustrates the critical interaction between geologists and prehistoric archaeologists.25 Both included profiles that indicated the stratigraphic context of Stone Age artifacts. Their ob-

5

Theory and History

Figure 1.1 Sir Charles Lyell If geoarchaeology has a ‘‘father,’’ it is Charles Lyell, best known for his multivolume, multi-edition Principles of Geology and for his unswerving dedication to the concept of uniformitarianism. His Geological Evidences of the Antiquity of Man is both the first book on geoarchaeology and a defining text: it sets out the archaeological problem and then applies geologic knowledge and principles. An Englishman who, along with Charles Darwin and Georges Cuvier, effected the great nineteenth-century revolution in our understanding of human and natural history, Lyell carefully documented, in a uniformitarian geologic context, all known remains of prehistoric humans and their artifacts. (Drawing by Elaine Nissen, from a photograph in the University of Minnesota Archives)

servations demonstrated the contemporaneity of stone tools and extinct animals, although these observations did not convince everyone. It was after the excavations of Brixham Cave in England in 1858, directed by William Pengelly and supervised by a group that included the geologist Charles Lyell (fig. 1.1), that geologic opinion began to support the archaeological acceptance of the presence of humans during the Ice Age.26 The Brixham Cave excavations were convincing because the remains of extinct fauna such as Elephas (Mammuthus) and Rhinoceros were found

6

in the same strata as stone artifacts, and the strata containing both artifacts and fauna were covered by an undisturbed stalagmite-limestone floor. In addition, the excavations were conducted in a manner that established the spatial relation within the sealed cave sediments. Last, there was close interaction among Pengelly and British geologists Lyell, Hugh Falconer, and Joseph Prestwich. Falconer traveled to the Somme Valley in France during 1858, where he observed the sites studied by Boucher de Crèvecoeur de Perthes. During 1859, Prestwich and Evans traveled first to Abbeville and later to Saint Acheul, in France. At Saint Acheul they viewed stone artifacts embedded in sediments that also contained the remains of extinct animals. These observations led them to conclude that humans had lived at the same time as the extinct animals. Lyell also visited the discoveries at Abbeville in July 1859. He concluded that the stone tools were ‘‘undoubtedly contemporaneous with the mammoth.’’ 27 Lyell’s observations based on the investigations at Abbeville and Brixham Cave were presented at the 1859 meeting of the British Association for the Advancement of Science, and Prestwich’s observations were published the following year in the Proceedings of the Royal Society of London.28 Such communications signaled that members of the scientific establishment were convinced that sufficient documentation existed to validate the association of stone artifacts with the remains of extinct animals; this relation led to the conclusion that humans had been part of the Ice Age world. Just how thoroughly this approach, which from a present-day perspective would be considered geoarchaeological, was ingrained in the ideas of the mid-nineteenth century as exemplified by the works of Lyell and John Lubbock. Both provided detailed reviews of the evidence for the human antiquity in Europe based on evaluations of the contextual integrity of the artifacts. Lyell’s Geological Evidences of the Antiquity of Man clearly established the role of geology in archaeological inquiry.29 Claudine Cohen provides a good review of Lyell’s Geological Evidences and its contribution to prehistoric archaeology and paleoanthropology.30 The label ‘‘geological archaeologist’’ has been applied to John Lubbock, William Dan-

Theory and History

kins, Augustus Pitt Rivers, and John Evans in contrast to the approaches of Lyell, Prestwich, Pengally, and Falconer.31 In his 1865 monograph Pre-Historic Times, Lubbock divided the Stone Age into the Paleolithic and Neolithic, splitting the artifactual record into an Old Stone Age and a New Stone Age. The Paleolithic was subsequently separated into three more subdivisions, Lower, Middle, and Upper, by Edward Lartet and Gabriel de Mortillet, both of whom were greatly influenced by the concepts and practices of geology. A review by James Geikie a few years later deals specifically with the climate and geologic age of Paleolithic deposits.32 The historical importance of geologic thought in prehistoric archaeology has been noted by Daniel, who writes: ‘‘Antiquarianism was . . . securely bedded . . . [in the] advance of natural science. . . . There could be no real archaeology before geology.’’ Daniel describes archaeology as the child of geology and writes about ‘‘the geological beginnings of archaeology.’’ In his view, geology initially ‘‘defined the limits and set the problems of prehistoric archaeology.’’ For some early prehistorians, ‘‘the Paleolithic . . . was a matter of geology’’; ‘‘prehistoric archaeologists . . . [thought] of themselves as natural scientists.’’ The geologic approach was exemplified by de Mortillet’s attitude that prehistory was ‘‘entirely an extension of geology.’’ Daniel reinforces this argument with the suggestion that ‘‘the de Mortillet system with its geological background . . . became the orthodox system of prehistory.’’ 33 During the second half of the nineteenth century other archaeologists became concerned with time and stratigraphy, as is evident in the work of classical archaeology. In 1860 Giuseppe Fiorelli began to conduct excavations at Pompeii that laid emphasis on stratigraphic control. Alexander Conze showed similar concern for stratigraphy in the 1873 excavations of Samothrace, as did Ernest Curtius in the 1875 excavations of Olympia. At Hissarlik (Troy), starting in 1871, Heinrich Schliemann became the first to excavate a multilayered tell. After 1882 the technique of stratigraphic excavation at Troy was firmly established by the efforts of Wilhelm Dörpfeld. During the late 1800s Augustus Pitt Rivers stressed the criti-

cal importance of stratigraphic observation in excavations conducted in England. Another of the main founders of ‘‘modern’’ archaeological methods of excavation was Flinders Petrie, with his excavations in Egypt and Palestine. In 1866 European prehistoric archaeologists held their first international congress, concerned with the application of a scientific treatment of the natural history of humankind. It has been argued that, perhaps because of the Europeans’ early adaptation and refinement of stratigraphic excavation techniques, geologic principles seem to have been more widely applied in Europe than in the United States. As early as the 1840s, however, Ephraim G. Squier and Edwin H. Davis provide a valuable example of how American archaeologists were implicitly using geologic principles during the mid-nineteenth century.34 The two used stratigraphic methods to determine whether anthropogenic or other natural processes had created mounds found in the Mississippi Valley. Their research demonstrates that attention to the geologic context of artifacts (the method favored by Boucher de Crèvecouer de Perthes and Rigollot in France) was being applied to the understanding of the prehistoric record in North America at essentially the same time it was in the Old World. According to Bruce Trigger, ‘‘all the chronological methods used in Europe were known in America and had been successfully applied by archaeologists.’’ 35 From the 1870s onward, shell-mound studies were reported by Jeffries Wyman, Steven Walker, and Clarence B. Moore from the southeastern United States and by William H. Dall in Alaska. During the 1880s, after the work by Squier and Davis, Cyrus Thomas used stratigraphic principles to study mounds, as did William Henry Holmes and Frederick W. Putnam in their research into the possibility of an ‘‘American Paleolithic.’’ 36 North America’s first great mining boom during the 1840s provided early evidence of other forms of prehistoric use of natural resources. In the Keweenaw Peninsula of northern Michigan it became apparent that the miners were not the first to exploit the rich veins and surface nuggets of native copper. The ‘‘ancient diggings,’’ as they were called in the Lake Superior region, were

7

Theory and History

copper mines from a time well before European contact. In the debris of the ancient pits were stone hammers with marks of hard usage. A large proportion were broken, but some had been new or unused when abandoned. These discoveries were reported by professionals in the early 1850s and brought geologists into the picture as they sought to explain the context of these remains.37 A prominent example of the application of geologic methodology in North America during this early period is Lyell’s observations concerning human remains and their association with deposits containing the remains of extinct animals in Mississippi. Lyell conducted field studies of the area in 1846 and concluded that the association was probably a product of redeposition or mixing. This emphasis on describing the processes that contributed to the final state of the artifactual record was a crucial facet of the ‘‘American Paleolithic’’ issue in the late nineteenth century. The American Paleolithic controversy focused on the effect that natural processes might have on the integrity of the archaeological record, and the validity of arguments regarding the antiquity of artifacts or human remains relied mainly on the stratigraphic context of these finds.38 The search for human antiquity and the question of an American Paleolithic involved as much geologic as archaeological knowledge. To understand what happened to the ‘‘early man’’ controversy during the last two decades of the nineteenth century, we can turn to the geologists of the United States Geological Survey (USGS) and the archaeologists of the Bureau of American Ethnology (BAE). Both agencies were created through the efforts of the geologist, geographer, and ethnographer John Wesley Powell (fig. 1.2) in 1879. The BAE, in Powell’s conception, was to provide a permanent anthropological survey for the United States.39 The critical role of geologists in the founding and early development of the BAE should not be underestimated. After founding director Powell left in 1902, the geologist William Henry Holmes became director, a position he held until 1910. Holmes led the BAE into the important new fields of physical anthropology and cultural resource preservation. Holmes himself was an energetic debunker of the

8

Figure 1.2 John Wesley Powell Powell (1834–1902) was a geologist who founded both the United States Geological Survey and Bureau of American Ethnology. Ethnology in the late nineteenth century, often used interchangeably with the term anthropology, meant the study of alien civilizations. Powell’s sensitivity to the Native Americans and their cultures was a driving force in early archaeological and ethnographic research. Powell’s The Exploration of the Colorado River and Its Canyons, which contains a large number of drawings made on the Powell expeditions, is available from Dover Publications, Inc., New York, 1961. This Dover edition is an unabridged and unaltered republication of the work first published by Flood and Vincent in 1895 under the title Canyons of the Colorado. D. Worster has written an excellent biography of Powell (2001). (Drawing by Elaine Nissen, from a photograph in Powell’s The Exploration of the Colorado River and Its Canyons)

American Paleolithic, and he did significant work in excavating ancient mines and quarries.40 The key to the early man controversy lay in the definition of a relative sequence of glacial-age deposits for North America—one of the goals of USGS geologists in the 1880s. European glaciologists had recognized the existence of glacial and interglacial stratigraphy, and Americans also began to discern regional sequences as large areas of the

Theory and History

Midwest and the Rocky Mountains were mapped by state geological surveys and by the USGS. During the last decade of the nineteenth century, the reliability of the geologic context for potential Paleolithic artifacts was considered to be extremely important. Holmes wrote: ‘‘Advocates of a Paleolithic . . . in America have been forced to give up the idea that there is any other reliable test of the age [of an artifact] . . . than that furnished by geology. . . . The antiquity . . . [is a] question germane to the field of geologic research.’’ 41 Despite differences in the actual evaluations of the geologic contexts of ‘‘paleoliths,’’ there seems to have been some basic agreement with this idea. Powell had earlier written: ‘‘The first appearance of mankind in . . . the continent . . . must rest on geologic facts.’’ 42 Regarding Quaternary sediments, Charles C. Abbott stated: ‘‘He is no archaeologist whose training falls short of the ability to study intelligently the history of these superficial deposits.’’ 43 Henry W. Haynes echoed Abbott on the importance of a full understanding of the context of artifacts: ‘‘Whether any . . . object can be identified as a . . . Paleolithic implement . . . is a question for geologists to answer.’’ 44 So did Rollin D. Salisbury: ‘‘Were the ‘Paleolithic implements’ so introduced? This is a geological, not an archaeological question.’’ 45 By 1892, George Frederick Wright had advocated an American Paleolithic based in part on a review of the geologic context of the purported artifacts.46 Critics of Wright’s ideas equally consistently based their arguments on geologic criteria. The application of geologic methods to buttress the argument that humans existed in North America during the Ice Age is exemplified by Holmes’s 1893 Journal of Geology and American Geologist articles concerning contextual problems associated with potential Paleolithic sites in Ohio and Minnesota. These papers represent ‘‘a shift’’ from arguments based on artifact morphology to ‘‘a dependence on geology.’’ 47 Other leading figures of North American geology in the late nineteenth century found the geologic aspects of the early man controversy compelling and made significant contributions to the resolution of questions about the glacial context of human and artifactual remains. Prominent

Figure 1.3 Newton Horace Winchell In 1872 Newton Horace Winchell became the first State Geologist and director of the Geological and Natural History Survey of Minnesota. One of Winchell’s outstanding contributions was his estimate of the length of time since the last ice sheet retreated from Minnesota, based in part on his evaluation of the rate of recession of St. Anthony Falls, which he gave as 8,000 years. After he completed the final survey reports, he devoted much of his time to the archaeology of the state. From 1906 until his death in 1914, he worked at the Minnesota Historical Society, where he was in charge of the Department of Archaeology. (Drawing by Elaine Nissen, from a photograph in the University of Minnesota archives)

among them is the first state geologist of Minnesota, Newton Horace Winchell (fig. 1.3). While working out the glacial geology of Minnesota, Winchell became interested in the ancient remains of the indigenous Native American tribes. Winchell’s interest culminated in a large volume, The Aborigines of Minnesota, published in 1911. In addition, he summarized many of the questions about the geologic context of early humans in North America in his presidential address to the

9

Theory and History

Geological Society of America on 20 December 1902.

Collaborative Phase: 1900–1950 As researchers began to conduct more planned, systematic excavations of archaeological sites and became more specialized in their training, the focus on interaction led to a period of straightforward collaboration. These years show a refinement of the field methodologies employed during the 1800s, such as at Saint Acheul in France. Geoscience specialists increasingly worked with archaeologists in collaborative efforts not only to evaluate the contextual and chronological situation of artifacts and other fossil remains but also to obtain information concerning paleoenvironmental and paleoclimatic changes. Collaborative efforts were also undertaken in the study of artifactual raw materials and remote sensing. The collaboration or multidisciplinary cooperation that characterizes this period can be divided into three types. Most prominent, geoscientists (commonly but not exclusively Pleistocene geologists and vertebrate paleontologists) coordinated their efforts with archaeologists working in the same geographic area. These earth scientists were primarily interested in regional studies that could be used as the basis for paleoclimatic interpretations and building a temporal framework that could also be applied to dating archaeological sites. The emphasis on the use of geomorphic, stratigraphic, and paleontologic criteria was a critical aspect of the chronological studies—indeed, this second period ends with the introduction of ‘‘absolute’’ dating methods (for example, the development of the radiocarbon technique). The second kind of cooperation after 1900 was between specialists involved in the laboratory analysis of geologic samples and artifacts and those who provided the specimens. In the third, the collaboration was between individuals who might be termed ‘‘geoarchaeologists’’ in the sense that they were archaeologists who evaluated the site-specific archaeosedimentary context of artifact localities. In Old World studies this period of cooperation began roughly at the start of the twentieth century. The years 1900–1950 have been identi-

10

fied as a period of growth in interdisciplinary work on the Pleistocene. The investigations of the Grimaldi Caves and the Grotte de Consérvatoire at Monaco represent pioneering efforts in which natural scientists took direct part in archaeological excavations. These studies were led by the paleontologist Ami Boué. This type of effort can be seen as a geoarchaeological perspective to archaeological interpretation. Excavations were conducted on an archaeological site with a Neolithic-Chalcolithic through Iron Age sequence using a refined procedure that included screening and vertical and horizontal control of finds. Specialists in ceramics, vertebrate paleontology and human skeletal remains, botany, the analysis of metal artifacts, geomorphology, and climatology participated in the excavation. Raphael Pumpelly applied what he called the rules of geologic reasoning to archaeology.48 Chronology and environmental and climatic change, along with their relation to human adaptation as reflected in artifactual remains, were among archaeology’s principal concerns after 1900. The pollen-analysis methods developed by Lennart von Post revealed climate intervals associated with postglacial environmental change. Tree rings were employed to estimate chronologies by the dendrochronologist Andrew E. Douglass after 1914, although DeWitt Clinton had used tree rings to date earthworks in New York State as early as 1811.49 In the Old World, the technique of varve analysis was applied by Gerard J. de Geer in 1905 in Sweden and later by Matti Sauramo in Finland. In North America, the Quaternary geologist Ernst Antevs was the first to use varve analysis (see Chapter 5). During the 1920s to the 1940s geoscientists collaborated in a variety of Old World prehistoric investigations. The first phase of geoarchaeological research was archaeological site sedimentologic studies, as described by geoarchaeologists Julie Stein and William Farrand. These are represented in such site sedimentology studies as those conducted by archaeologist Dorothy A. E. Garrod and paleontologist Dorothy M. A. Bate between 1929 and 1934 during their excavations in the southern Levant. Stein and Farrand categorize these as ‘‘archaeological geology.’’ 50 Evi-

Theory and History

dence of multidisciplinary interaction in the Old World is reflected by Frederick E. Zeuner’s use of prehistoric archaeology for its geochronological and climatic implications, in an approach reminiscent of the earlier works of Lyell, Lubbock, Geicke, and Wright.51 Zeuner’s efforts represent specific attempts to work in an area that falls between the natural sciences and human prehistory. Glyn Daniel and Colin Renfrew describe Zeuner’s Dating the Past as ‘‘one of the first books exclusively dedicated to archaeological science. . . . The approach was largely geological, as it had been since the early days of the last century.’’ 52 In Europe, the perspective for the study of prehistory appears to have expanded during the first half of the twentieth century. Jacques de Morgan expressed the view that prehistory was not a discipline narrowly focused on the study of artifacts; rather, ‘‘geology, zoology, botany, climatology, anthropology and ethnography are the bases of prehistory.’’ 53 De Terra also recognized that geology and archaeology are ‘‘border sciences’’ and that there was a ‘‘field of study which lies between the realms of the geological and archaeological sciences’’ and that ‘‘the concerted research methods of these two historic sciences . . . would prove useful in helping . . . decipher the story of ancient man.’’ 54 Remote-sensing techniques were applied during the early twentieth century in the form of aerial photography.55 In 1906 aerial photographs of Stonehenge revealed buried features not visible from the surface. In the Middle East, aerial photographs were used to record the outline of an irrigation system. During the 1920s a major effort to use aerial photography to discern archaeological features was conducted in England, while in North America, Charles Lindbergh took aerial photographs of the Southwest. Richard Atkinson’s use of electrical conductivity in 1948 pioneered the use of resistivity meter surveys for the location of such buried site features as structures, pits, and ditches.56 The application of natural sciences in the analysis and study of materials recovered from archaeological excavations including soils, pollen, metal, stone, and plants also increased. Examples of such applications include Herbert H. Thomas’s

petrographic analysis of stones from Stonehenge, David M. S. Watson’s examination of the fossil faunal remains (bones) from the site of Skara Brae, and the work by Danish scientists on prehistoric grains. In the early part of the twentieth-century investigations of the Paleolithic were under way in many parts of the world. Johan G. Andersson (fig. 1.4), director of the Geological Survey of Sweden, went to China in 1914 as an adviser to the Chinese government on mining. In his decade in China, Andersson made fundamental contributions to Chinese paleontology and archaeology. He is best known as the discoverer (with J. McGregor Gibb) and initial excavator of Zhoukoudian, the famous ‘‘Peking Man Site.’’ 57 Elaboration of the principles involved in stratigraphic studies continued. Mortimer Wheeler’s use of section drawings during the 1920s, which has been considered a turning point in archaeological methodology, is a primary example.58 This approach was more fully explored in Wheeler’s Archaeology from the Earth (1954). Important continuities between the European natural science perspectives that were developed in the nineteenth century and American archaeology in the beginning of the twentieth century can be seen by the expanded use of stratigraphic excavations in North America. Influenced by the European geologic approach to prehistory, Manuel Gamio and Nels C. Nelson employed stratigraphic principles at American sites.59 Alfred V. Kidder, influenced by Nelson’s fieldwork, applied the techniques of stratigraphic excavation at the famous site of Pecos, New Mexico.60 Gamio initiated the stratigraphic excavations at the archaeological site of Atzcapotzalco in Mexico City, Mexico, in 1911. The excavations consisted of a set of trenches that revealed three groups of pottery and their stratigraphic relationships (fig. 1.5). Gamio was able to demonstrate the relative order of these artifacts. Gamio labeled the distinct types of pottery as ‘‘cultures.’’ The lowest deposits containing artifacts had pottery assigned to the De Montana culture. These sediments were overlain by deposits containing Teotihuacan pottery types. The uppermost deposit contained the youngest pottery, designated as Azteca (Aztec).61

11

Theory and History

Figure 1.4 Johan Gunnar Andersson Andersson (1874–1960) was a Swedish geologist who discovered and excavated the famous Paleolithic cave site at Zhoukoudian in China. He also explored and excavated other early sites and defined the first Neolithic (Yangshao) culture in China. Andersson essentially pioneered Chinese prehistoric studies. (Drawing by Elaine Nissen from a photograph supplied by the Museum of Far Eastern Antiquities, Stockholm)

12

Nelson began field studies in the Galisteo Basin, New Mexico, in 1911. In 1913 Nelson visited Spain and observed the methods of stratigraphic excavation conducted by Hugo Obermaier and Henri Breuil at Castillo Cave. By 1914 he was using stratigraphic methods to conduct the excavations at Pueblo San Cristobal. The excavations were conducted in one-foot-thick intervals with artifacts from each level kept separate. This application of the concept of superposition has been described as ‘‘the first exposition of a refined method for determining exactly the time sequence of archaeological materials.’’ 62 If not actually the first use of the concept, Nelson’s excavations at San Cristobal do demonstrate the changing developments in the application of stratigraphic concepts to archaeological field studies that were occurring in the early 1900s and the influence of methodological approaches being applied in Europe at the same time.63 Kidder’s excavations of midden deposits at Pecos Pueblo, in New Mexico, also applied a stratigraphic methodology to evaluate the relative age of archaeological materials (fig. 1.6). Kidder began his excavations in 1915. In some parts of the site he was able to excavate in a way that artifacts could be collected from distinct deposits in contrast to excavations by arbitrary levels. This technique allowed him to document the relative stratigraphic position of pottery types within the archaeological deposits at Pecos Pueblo. He was able to demonstrate that specific pottery types occurred in particular strata at the site and to document the presence of a sequence and relative ordering to these types. The issues concerned with verification of potential Ice Age artifacts and human remains continued to dominate archaeological-geologic studies in the New World in the early twentieth century. Purported Pleistocene human remains or artifacts, for example, had to be verified by unquestionable stratigraphic evidence. Interactions among American archaeologists and geologists during this period are exemplified by the publication in 1917 in the Journal of Geology of a collection of articles dealing with human remains found in association with remains of extinct fauna at Vero, Florida. As with other potential Pleis-

Figure 1.5 Gamio’s Section Valley of Mexico archaeological stratigraphy by Manuel Gamio. Gamio’s profile depicts the locations of three types of artifacts compared to particular sediments.

The geologic approach to prehistory in Europe influenced the application of stratigraphic principles in the Americas during the early part of the twentieth century (from Gamio 1913).

Theory and History

Figure 1.6 Kidder’s Section Pecos, New Mexico, stratigraphy by Alfred Kidder. Stratigraphic designations are based on types of artifacts from the American Southwest. The various strata are determined by specific varieties of pottery. The spatial relationships of the strata can be used to ascer-

tain the temporal relationships of artifact types. This generalized cross-section illustrates the application of stratigraphic principles in archaeological contexts during the first half of the twentieth century (from Kidder 1924).

tocene sites containing evidence of the presence of humans, the discussions revolved around the sedimentologic context of the materials and the validity of their association. Rollin T. Chamberlin and George G. MacCurdy both concluded that there was evidence that the faunal remains lay in secondary deposits, but Chamberlin also believed that there was evidence to support the association between the extinct fauna and the human remains. The physical anthropologist Aleš Hrdlička, in contrast, believed that the Vero discovery could not be supported because of the depositional context of the remains. The discoveries at Folsom, New Mexico, in 1927 showed that human occupation in North America dated to a time when extinct forms of bison were present. It is now known that the Folsom artifacts and associated bison date to the very end of the Pleistocene, around 10,500 radiocarbon years ago. The primary criterion for acceptance of the Folsom evidence was that the

context of the discovery indicated a primary association between undisputed artifacts and extinct fauna, with no evidence that the associations were caused by secondary depositional processes. The resurgence of a geologic interest among American archaeologists can be traced to the Folsom discovery.64 The Southern High Plains region of the United States has been the focus of abundant geoarchaeological studies of Paleo-Indian sites since the 1920s.65 This region contains the Clovis, Folsom, and Lubbock Lake sites. In Texas and New Mexico, geoarchaeologists have reconstructed many site-specific and some regional environments for the Paleo-Indian occupation. The Southern High Plains is the most environmentally homogeneous region of its size in North America with a climate that is continental and semiarid. The Late Quaternary in situ archaeological record in this region is found in draws, playas, salinas, and dunes.

14

Theory and History

One can recognize the strongest natural science orientation in Paleo-Indian studies; within this realm a geoscience perspective has been pervasive in North American archaeology. This growth in interest is exemplified by the efforts of Edgar B. Howard, Elias H. Sellards, Ernst Antevs, and Kirk Bryan. Howard provided a comprehensive study and review of early human occupation in the American Southwest using geologic and archaeological evidence that included excavations at Burnet Cave and Carlsbad, New Mexico, and a study of Late Pleistocene ephemeral lake sediments between the towns of Clovis and Portales, New Mexico.66 His synthesis employed a variety of data derived from natural sciences to infer the Late Pleistocene environment on the Llano Estacado. In addition, he gave a critique of Paleo-Indian artifact typology and a review of the chronological and climatic theories associated with the end of the Pleistocene. Antevs studied the geologic and archaeological aspects of the Clovis-type site when he was a member of Howard’s expedition in 1934. He was influential in employing geologic techniques to develop paleoclimatic models that could be related to archaeological studies.67 Following the research at Vero in Florida, Sellards continued to conduct studies documenting the presence of humans in America during the Pleistocene (fig. 1.7). Vance Holliday described this as the first regional, interdisciplinary group of studies of Paleo-Indian archaeology and their associated environments.68 From about 1925 to 1950 Bryan dominated research at the interface between geology and archaeology in North America. C. Vance Haynes, Jr., designating the period the ‘‘Antevs-Bryan years,’’ demonstrates that like Bryan, Antevs had a great deal of influence in the geochronological study of PaleoIndian sites.69 Bryan and Antevs initiated what Stein and Farrand call the second phase of archaeological sedimentology, in which geoscientists were largely concerned with geochronology and the paleoenvironment of a site. In the 1920s Bryan investigated evidence for environmental change at Chaco Canyon, New Mexico. Later he became active in assessing the geologic setting of Paleo-Indian sites. Quater-

Figure 1.7 Elias H. Sellards Elias H. Sellards is perhaps best known for his advocacy, when he was State Geologist of Florida, of the association of human remains with those of extinct vertebrates in Pleistocene deposits at Vero and Melbourne, Florida. Although Sellards was probably wrong, the well-publicized dispute provoked careful studies of the vertebrate assemblage, studies that applied more sophisticated excavation techniques than was common in American archaeology. (Drawing by Elaine Nissen, from a photograph in the Texas Bureau of Economic Geology archives)

15

Theory and History

nary stratigraphic studies conducted in western Texas made extensive use of Paleo-Indian occurrences. Although a major facet of Bryan’s work was his focus on the application of paleoclimatic chronologies to date Paleo-Indian sites, these efforts were also used to delineate the paleoenvironmental settings of late Pleistocene archaeological sites. The interpretations of the settings were largely based on evaluations of the processes of sedimentary deposition affected by climatic processes. The reports by Bryan and Antevs on the human remains found within possible glacialage deposits in western Minnesota also show their concern for the processes of deposition in evaluating the archaeological record. Bryan’s influence can be seen in the application of geologic techniques to archaeological studies by his students at both New and Old World sites. Sheldon Judson carried out a geologic study of the San Jon site and its Plainview artifacts in northeastern New Mexico under Bryan’s supervision. The Midland study, led by prehistorian Fred Wendorf with the cooperation of geologist Claude Albritton and archaeologist Alex Krieger, has been called a model of the working collaboration between scientists. John Miller and Wendorf extended Bryan’s alluvial chronology to the Tesuque Valley in New Mexico. Bryan had also initiated the investigations of the rock shelter at La Colombiére in southeastern France, which were later completed by Hallam L. Movius and Judson. In 1947 Bryan’s student Herbert E. Wright, Jr., studied the geology of the Paleolithic occurrences at Ksar Akil, Lebanon. It is reasonable to state that by the beginning of the second half of the twentieth century, Bryan and his students had effected a convergence of Pleistocene studies and archaeological studies through an explicitly paleogeomorphic approach to archaeology.70 The nature of the interaction between archaeology and geology in America during the first half of the twentieth century is encapsulated in the volume Early Man published for the 125th anniversary of the Philadelphia Academy of Natural Sciences.71 It contains essays by Hrdlička on the skeletal evidence for a human presence in

16

America during the Pleistocene and an evaluation of the Vero finds by Sellards. There is an essay on the significance of profiles of weathering in stratigraphic archaeology by geomorphologist Morris M. Leighton, another on the use of pollen for dating archaeological deposits by paleobotanist Paul Sears, an essay on climate and ‘‘early man’’ by Antevs, and a discussion of the geology of Folsom deposits by Bryan. The volume demonstrates that many approaches derived from the natural sciences were being applied in earlytwentieth-century prehistoric studies.

Integrative Phase: After 1950 Throughout the nineteenth century and the first part of the twentieth century, earth-science interactions and methodological concerns with archaeology were largely focused on chronological interests. Without question the contribution of radiocarbon dating initiated by Willard Frank Libby in the late 1940s and extensively used from the 1950s onward was a critical contribution of the natural sciences to archaeological interpretation. Other influential technical innovations during the 1950s and 1960s from the field of geochronology include the application of chemical tests like those that verified the Piltdown fraud and research that used potassium-argon (K-Ar) measurements to establish the chronological context of hominid fossils associated with East African igneous rocks. Refinements in radiocarbon dating and the introduction of new chronometric techniques like Ar-Ar and luminescence dating continued in the 1980s and 1990s and into the twenty-first century. Another arena of interaction was in the identification of raw material. H. H. Thomas had initiated these studies at Stonehenge (see Chapters 7 and 8). Other studies of raw-material characterization in the Old World were Frederick R. Matson’s analysis of the clays from which pottery was made, the study of Bronze Age metals by Edward Sangmeister and H. Otto, and the studies by Colin Renfrew and J. R. Cann, into obsidian sources and trade. New World raw-material characterization studies include the work of David Williams on the petrography of pottery tempers and his investigations with Robert E. Heizer of

Theory and History

the rocks used in Olmec monuments from Mexico. Besides the methodological interaction between the earth sciences and archaeology, a trend toward theoretical convergence developed during the last half of the twentieth century. Archaeology went through a period of major change. In part this reflected a renewed emphasis on the development of a theoretical basis for archaeological studies. The change also came from the realization by archaeologists of how much paleoanthropology depended on an understanding of the geologic context of an archaeological deposit. A critical component of the new emphasis on context was the geoscience perspective. The transformation from a period of collaboration to a period of theoretical convergence in archaeology and the earth sciences is evident in Stein and Farrand’s third phase of archaeological sedimentology, ‘‘geoarchaeology.’’ 72 Several prominent developments suggest the explicit integration of an archaeogeologic perspective into archaeological studies during the second half of the twentieth century. The first is a movement toward, and then direct attempts at, formulating a theoretical framework in archaeology that would enable a geoarchaeological perspective. The second was the merging of methodological interests and the formulation and application of these interests in field and laboratory studies. At least two subphases within the development of a theoretical framework can be recognized. The initial subphase is characterized by statements made in the 1940s and 1950s on the potential role of geoscientists in archaeology; the second by more formal statements made in the 1960s through the early 2000s. In the first phase are the writings of Movius, Robert J. Braidwood, H. E. Wright, Troy L. Pewe, and Ian W. Cornwall. Combined, these works indicate that the benefits of interactive collaboration between geologists and archaeologists were established by the end of the 1950s. This view is illustrated in Old World Paleolithic archaeology by Movius, who stressed the bond between prehistoric archaeology and the natural sciences. One goal of

archaeology to investigate human adaptation to natural environments, Movius emphasized, could be achieved through natural science studies concerned with the sequence and correlation of Pleistocene events. He argued for the importance of environmental reconstruction, noting: ‘‘Prehistoric archaeology cannot be divorced from its background with the natural sciences without denying it the very key to the solution of its fundamental problem: the reconstruction and interpretation, insofar as possible, of human activities of the past.’’ 73 A paper published in American Antiquity in 1954 by Pewe emphasized the geologic approach to dating archaeological sites.74 While mentioning the value of tree-ring and radiocarbon dating, he focused on the value of understanding the geologic and biotic response to climate changes and provided a series of examples demonstrating the value of a geologic perspective to help evaluate archaeological contexts. In a volume edited by Walter W. Taylor, Braidwood envisioned a new field of ‘‘Pleistocene ecology,’’ or ‘‘Quaternary geography,’’ consisting of interrelated disciplines that would include archaeology and studies of the environment. In the same volume H. E. Wright emphasized that the most useful contributions geologists could make to archaeological problems lay in the interpretation of the physical and climatic environment. He noted the importance of this approach to the study and evaluation of the whole archaeological site and suggested that Pleistocene geologists would be most interested in the climatic environment and chronology associated with archaeological sites. And Judson also recognized the value of collaboration between Pleistocene geologists and archaeologists.75 An influential study that demonstrated the potential sediments and soils offered in archaeological analysis and interpretation was Cornwall’s Soils for the Archaeologist. Butzer describes Cornwall’s work as the first systematic attempt at geoarchaeology.76 Cornwall argued for contextual studies in archaeology that used the geosciences. Edward Pyddoke wrote in his Stratification for the Archaeologist: ‘‘It should be the task of the archaeologist to interpret and understand not only man’s

17

Theory and History

activities and artifacts but also the nature of the natural strata on, in, and under which they are discovered.’’ 77 The books by Cornwall, Pyddoke, Wheeler, and Zeuner show that a pervasive geoarchaeological perspective was established in the study of Old World prehistory by the 1960s. The conceptual framework produced by the writings of Movius, Braidwood, and Cornwall provided a foundation from which Butzer’s views could be elaborated. Butzer gave the designation ‘‘Pleistocene geography’’ to his view of the natural science approach. He characterized Pleistocene geography as ‘‘environmental reconstruction as applied to an understanding of the ecologic setting to prehistory.’’ 78 In addition to stratigraphy and chronology, Butzer placed an emphasis on environmental reconstruction. The role of geology in Pleistocene paleoecology and archaeology was discussed by C. Vance Haynes, Jr. (fig. 1.8), when he emphasized the importance of the geologic stratigraphy, interdisciplinary cooperation, and the geologist’s role in paleoecologic interpretations. Haynes’s career exemplifies the role of geoarchaeological investigations.79 Beginning in the 1960s, prehistorians, archaeologists, and anthropologists focused on the theoretical and methodological issues inherent in attempting to use the archaeological record to infer prehistoric human behavior. For example, anthropologist Lewis Binford emphasized the fact that archaeological sites vary in their depositional context and history.80 Archaeologists recognized the importance of evaluating all potential processes that could affect the final character of the archaeological record to infer past hominid behavior. Prehistorian Glynn Isaac focused on investigating sedimentologic processes to interpret Acheulian occurrences at Olorgasailie in Kenya.81 Robert Acher looked at disturbances of the archaeological record and the discontinuities between past behavioral systems, the artifacts produced and deposited because of these activities, and subsequent events that affected these materials.82 In Greece during the 1960s, William A. McDonald included earth scientists in his wideranging survey of Messenia, bringing in first Bryan’s student H. E. Wright, a geologist and

18

Figure 1.8 C. Vance Haynes, Jr. One of the few long-time full-time geoarchaeologists in the United States, C. Vance Haynes, Jr., held dual appointments in geology and anthropology/archaeology at the University of Arizona. He has contributed to the solution of a broad range of important geoarchaeological problems in the southwestern United States and Egypt. (Drawing by Elaine Nissen, from a photograph by Christopher Hill)

paleoecologist, then Rapp to complement the physical geographers. The resulting publication was subtitled Reconstructing a Bronze Age Environment. 83 Along the Nile Valley, Wendorf (another student of Bryan’s) organized a multidisciplinary team of scientists (including geologists Jean de Heinzelin and Albritton and paleontologists Francine Martin, Achilles Gautier, and P. H. Greenwood) to study the Paleolithic of Nubia in southern Egypt and northern Sudan.84 Similar efforts at integrating researchers from other natural sciences into archaeology are exemplified by Richard MacNeish’s work in Mesoamerica and Braidwood and Robert M. Adams’s in the Middle

Theory and History

East. The theoretical development of concepts associated with processual archaeology may be strongly linked with the increased tendency to view archaeology as a natural-historic science. It became clear that the archaeological record could not be directly used to observe human behavior but instead that the record could be used to infer the past processes that had created it. The early stages of processual archaeology were influenced by the ecologic orientation of Leslie A. White and Julian H. Steward and the reappraisal of American archaeology by Taylor. The increased interest in ecologic and spatial patterns and their relation to process was noted by Joseph Caldwell in 1959, but Binford’s ideas focused attention on the development of archaeological theory toward a processual view.85 The later theoretical arguments of both Binford and Michael B. Schiffer advocated a focus on the natural transformational processes involved in creating the archaeological record and the usefulness of a natural science approach in assessing them.86 With the goal of providing insights into the accurate interpretation of archaeological record, Binford argued that ‘‘archaeology must face up to the nature of the data it employs . . . [and] adopt the methods of the natural sciences.’’ He added: ‘‘It is much more likely that archaeological remains will be found within geological deposits. . . . [The] materials within (that is, artifacts) derive from events during the deposits formation, none of which is necessarily representative of a behaviorally interrelated set of conditions.’’ 87 On the assumption that the archaeological record was a product of diverse sets of natural and anthropogenic processes, Schiffer differentiated between the ‘‘archaeological context’’ and the ‘‘systemic context,’’ proposed a model based on transformation theory, and focused on identifying the processes involved in creating the archaeological record. Whether advocated as bridging (‘‘middle-range’’) theory or as part of environmental archaeology, the focus on processes has become recognized as an important facet in the study of prehistory. Following the lead of Binford, L. Mark Raab and Albert C. Goodyear argued that the principles of processes have

become virtually synonymous with middle-range theory in archaeology.88 The strong connections between prehistoric archaeology and geology were stressed by Hassan. He emphasized the idea that reconstruction of site development history using sedimentologic analysis was a contribution ‘‘pertinent to the contemporary theoretical orientation of archaeological research.’’ 89 For Hassan, understanding the human past included evaluating artifacts in both their physical context and behavioral matrix in order to interpret their final location. Quaternary geologists, prehistorians, and archaeologists have attempted to develop a conceptual base for geoarchaeology and its relation to archaeology by building on Butzer’s contextual or ecologic archaeology.90 Influenced by the ideas of David L. Clarke, Fedele also used the interactions between the human ecosystem and the external paleo-land system to develop a geoarchaeological approach.91 These ecologic attempts paralleled approaches to employ system theory in the earth sciences and archaeology. This ecologic perspective also can be seen in the environmental and interdisciplinary views expressed by Roald Fryxell.92 Observing that anthropology is a synthetic discipline that draws upon data and ideas from a diverse number of fields, Fryxell argued that the studies conducted by scientists like Bryan, Antevs, Clarke, and Bordes ‘‘clearly demonstrated the value of adding dimensions other than archaeology to prehistory’’ emphasizing that anthropological training required an understanding of environmental interpretation and technical interdisciplinary skills. At a minimum, Fryxell argued, archaeologists should be able to ‘‘recognize, describe and record physical stratigraphy.’’ During the late 1980s and 1990s opinion had consolidated concerning the importance of a natural science perspective in archaeology.93 Our review of historical areas of research and theoretical concepts in this chapter reflects this consensus. Archaeological data appear as the product of varying proportions of human behavioral activities and natural geologic processes. The necessity of differentiating these phenomena is now recognized as crucial to deriving knowledge from the

19

Theory and History

Figure 1.9 Karl Butzer, above; William ‘‘Bill’’ Farrand, right; John ‘‘Chris’’ Kraft, opposite page, left; George ‘‘Rip’’ Rapp, opposite page, right. (Drawings by Elaine Nissen, from photographs)

prehistoric evidence. Archaeological occurrences can be understood more fully when the paleoenvironmental context and the processes involved in creating the archaeological record are documented and evaluated.

Changing the Guard In addition to C. Vance Haynes, mentioned earlier, the new century has seen the ‘‘retirements’’ of an ‘‘old guard’’ in North America. In addition to Haynes, retiring or just retired are four geoarchaeologists (perhaps in some instances better described as archaeological geoscientists) who played important roles in building geoarchaeology into a discipline. These are Karl Butzer, William Farrand, J. C. Kraft, and Rapp (fig. 1.9). Significantly, all were trained in earth science (geology or geography) departments, as were many other important contributors to the recent development of geoarchaeology, among them Tjeerd ( Jerry) Van Andel, Norm Herz, Henry Schwarz,

20

C. Reid Ferring, Richard L. Hay, William Dickinson, Carl Vondra, Larry Agenbroad, Floyd McCoy, Jack Donahue, Bruce Gladfelter, William Johnson, and Gail Ashley. The leaders of the new generation have had more diverse educations. Some were trained in anthropology departments (for example, Michael Waters and Payson Sheets), others still in geology departments (for example, Paul Goldberg, Rob Sternberg, and Art Bettis), some still in geography departments (for example, Vance Holliday and Rolfe Mandel), and even others in interdisciplinary programs (for example, Julie Stein). The field of academic geoarchaeology is in transition. Although most geoarchaeologists at or near retirement have generally remained in geology or geography departments (for example, Farrand, J. C. Kraft, Karl Butzer, Rapp, Van Andel, Schwarcz, Herz, Donahue, McCoy, Sternberg, Hay, Vondra, Ashley, and Gladfelter), most of the new leaders are in departments of anthropology

Theory and History

(for example, Waters, Stein, Holliday, Kenneth Kvamme, Lawerence Conyers, Jeff Eighmy, Pat Julig, C. Russell Stafford, and Stanley Ambrose) or archaeology (for example, Goldberg). The same is true of younger academic geoarchaeologists, who are most often in departments of anthropology (for example, Gary Huckleberry, Zichun Jing, Christopher L. Hill, Rinita Dalan, Sarah Sherwood, R. Lee Lyman, and Loren Davis) or archaeology (for example, Andrea Freeman). A few continue to find academic homes in earth science departments (for example, James Harrell, Richard Dunn, Glenn Fredlund, Lee Nordt, Jamie Woodward, Margaret Guccione, and Michael Wilson) or work jointly in an earth science and an anthropology department (for example, Ervan Garrison, Garry Running, and Craig Feibel). A brief look at the careers of four members of the ‘‘early’’ group, educated in geoscience depart-

ments (figures 1.9a–d), is instructive of the evolution of the field. Karl Butzer received a B.Sc. in mathematics and a M.Sc. in meteorology and geography from McGill University, Montreal, and a D.Sc. in physical geography and ancient history from the University of Bonn, Germany. He worked at the University of Wisconsin, the University of Chicago, and the Swiss Federal Institute of Technology before becoming a professor in the department of geography at the University of Texas at Austin. During the 1950s and 1970s he conducted field studies in Egypt and Nubia, Spain, East Africa, and South Africa. He continued to work in South Africa and Spain during the 1980s and began to work in Mexico. His books include Environment and Archaeology: An Ecologic Approach to Pleistocene Geography (1964) and Archaeology as Human Ecology: Method and Theory for a Contextual Approach (1982). In 1974 he published his paper ‘‘Geo-Archaeological Interpretation of Two Acheulian Pan Sites,’’ and in 1975 American Antiquity published his views in the paper entitled ‘‘The ‘Ecologic’ Approach to Prehistory: Are We Really Trying?’’ Graduate students who completed their degrees under his supervision include

21

Theory and History

Bruce Gladfelter, Joseph Schuldenrein, Arlene Miller Rosen, and Charles Frederick. Butzer is a pioneer of geoarchaeology with his advocacy of a contextual approach to archaeology that links the study of the human past with cultural ecology, applied geomorphology, and environmental history. William (Bill) Farrand studied glaciers in Greenland as part of his M.S. degree from Ohio State University. His doctoral research at the University of Michigan was on the glacial history of the Lake Superior basin and was influenced by the archaeologists James Griffin and Albert Spaulding, who served on his dissertation committee.94 While at Columbia University prehistorian Ralph Solecki convinced him to conduct research in Turkey and Syria. Later he was part of the team led by Movius that conducted studies of the Abri Pataud in France during 1964 and 1965. After returning to the University of Michigan in 1965, Farrand collaborated with Arthur Jelinek in the excavations at Tabun Cave in the 1960s and 1970s. Also in the 1970s he was part of a project that studied Franchthi Cave in Greece. Geoarchaeologists that were trained at the University of Michigan and influenced by Farrand include Goldberg. He has made important contributions to sedimentology, Quaternary geology, archaeological geology, and the geochronology of prehistoric sites. J. C. (Chris) Kraft received his degrees (B.S., M.S., Ph.D.) in geology from Pennsylvania State University and the University of Minnesota. After a short career working for an oil company he became the long-time chairman of the geology department at the University of Delaware. Initially he undertook major studies of coastal change along the eastern seaboard of the United States. In 1970 he joined Rapp in Greece and began a more than three-decade investigation of how significant coastal change has affected such major archaeological sites in Greece and Turkey as Troy, Ephesus, and Ancient Pylos, as well as determining the now well-hidden landscape of the famous battle at Thermopylae between the Spartans and the Persians. His more than thirty papers on Holocene coastal change have defined this important field of study. Kraft exemplifies the consum-

22

mate specialist who knows every aspect of a field (here ranging from sedimentology and geomorphology through paleontology and coring techniques). George (Rip) Rapp received his B.A. in geology and mineralogy from the University of Minnesota and his Ph.D. in geochemistry from the Pennsylvania State University. After about ten years of work in mineralogy Rapp joined a University of Minnesota classical archaeologist colleague, William A. McDonald, as associate director of the interdisciplinary Minnesota Messenia Expedition in archaeological surveying and excavation in southwestern Greece. Since 1967 he has ‘‘focused’’ on a broad range of geoarchaeological investigations, including coastal change, provenance of artifact copper, phytolith studies, and excavation. In contrast to his colleague J. C. Kraft, Rapp has been a generalist who moved where research opportunities took him. He was principal founder of the Archaeological Geology Division of the Geological Society of America, founded the Guide to Graduate Programs in Archaeological Geology and Geoarchaeology, founded the Archaeometry Lab at the University of Minnesota, and (with McDonald) co-founded the Ancient Studies and Interdisciplinary Archaeological Studies graduate programs at the University of Minnesota. These programs turned out some of the first Ph.D.s in geoarchaeology (for example, J. K. Stein and J. A. Gifford) as the field grew rapidly in the 1970s. Before the mid-1970s there were no graduate programs in the United States dedicated to educating geoarchaeologists, so earth scientists were induced or lured into geoarchaeology by the interesting research problems available. In the early twenty-first century it seems likely that most geoarchaeologists in North America will be educated in anthropology or archaeology departments.95 Signs of a growing maturation with the field of geoarchaeology were evident as the twenty-first century began. Within the academic community there appeared to be increased recognition of the value of a geoarchaeological approach. Academic departments and institutions contained people who used a geoarchaeological approach either

Theory and History

by explicitly applying earth-science techniques to obtain information pertinent to understanding the human past or who employed an interdisciplinary earth-science perspective to understand more fully the links between humans and other components of the ecosystem. Within archaeology there was a growing recognition that focusing on human-environmental relations and combining approaches from the social and natural sciences could be beneficial.96 The methodological convergence of the geosciences and archaeology is characterized by the expansion of the application of technical tools in support of archaeological research. In addition to increased interest in applying sedimentologic and stratigraphic principles to locating and delimiting archaeological sites and features, geochemical and geophysical techniques have become critical to archaeological site discovery and analysis. To conclude this appraisal, it seems appropriate to return to the question of what geoarchaeology is and what use it has for creating a theoretical base for archaeology. Several theoretical perspectives are useful for encouraging attempts to understand the archaeological record. From one perspective, archaeology is a branch of earth science that consists of processes, structure, and chronology. In another view, it is an independent natural science parallel to geology. Archaeology is a distinct discipline in that one of its main goals is the study of past human behavior. However, this past human behavior is not directly observable in the archaeological record. In the field of prehistoric archaeology there are no direct observations of actual human activities; all behavior and processes must be inferred from directly observable evidence. As stated by Clarke, archaeology is ‘‘the discipline with the theory and practice for the recovery of unobservable hominid behavior patterns from indirect traces in bad samples.’’ 97 In some sense then, archaeology is closely connected with historical geology and paleontology. These fields seek to understand the history of life on earth. Archaeology may also be understood on its own terms, as many have advocated since the 1960s. The historic sciences of archaeology, paleon-

tology-paleobiology, and historic geology are connected by the fact that they must reconstruct the dynamics of the past through inference and analogy. They use observations of present-day dynamics and the tangible patterns of the past that have survived to develop an empirically based understanding of ancient times. Perhaps the most practical approach is to rely on a combination that can apply some aspects of induction, falsification, and multiple working hypotheses within a framework established by the uniformitarian principles advocated by Charles Lyell. Geoarchaeology provides a methodological and conceptual basis for interpreting the human past anchored in approaches that increase the knowledge base by experiment and observation. This quantification of the evidence from the past can serve as the foundation for archaeological explanation. It can provide the basis for conjecture and multiple working hypotheses that can be tested by further study of the products of past and present interactions between humans and the earth. If the function of geoarchaeology is to solve archaeological problems using earth-science techniques, concepts, and knowledge base, then we must understand the problems that archaeology seeks to resolve. The chapters that follow give an indication of the diverse methods and knowledge derived from the earth sciences that are used to infer past conditions from archaeological data. Among these are the location of sites and features, the distinction of artifacts or archaeological features from naturally occurring forms, the evaluation of assemblage integrity and the environmental setting, the establishment of the ages of artifacts and the duration and intensity of occupations by identifying materials found in the archaeological record, and the sourcing of raw materials and artifacts. This overview of the historical and theoretical interaction between the earth sciences and archaeology prompts a consideration of what the phrase ‘‘archaeological record’’ signifies.98 The archaeological record can be defined as a nonrandom sample of past human activity biased by behavioral and physical factors. Any study of artifacts or use of archaeological evidence or data needs to acknowledge the weaknesses and

23

Theory and History

strengths of the archaeological record. But what, then, is the nature of the sample that comprises the archaeological record? The archaeological record preserves a minute fraction of the initial dynamic, systemic behavioral associations of past humans. Certain conditions are more conducive to the preservation of more complete records of the initial systemic context. What kinds of artifacts are preserved and what factors affect how the record is biased by preferential preservation and subsequent diagenetic alterations? With the exception of sites like Pompeii, primary ‘‘cultural’’ deposits are rare, but many artifactual deposits may have ‘‘cultural’’ consequences. If we describe deposits containing artifacts or archaeological features that may have patterns associated with a human social or cultural system, we may be applying unwarranted interpretations to the data, since the ultimate goal for archaeology is to infer what kind of human behavioral signal is present within the artifactual record. The relation of geology to archaeology can be considered from the complementary perspectives of physical geology and historical geology. As with physical geology, archaeological problems can be approached with an emphasis on understanding the natural (geologic) processes that have affected the record. Here there is an emphasis on process and structure. As with historical

24

geology, the systematic study of past sequences of events, many archaeological problems are related to the continuum that includes stratigraphy, paleontology, geochronology, and Late Cenozoic paleoenvironmental studies. Geoarchaeology may best be considered a meeting ground where the full range of earth sciences is applied to artifactual evidence to infer past processes and events. In the late 1930s, Frederick J. North described the historical recognition of overlap in the fields covered by archaeology and geology. He cited John Aubrey’s paper of 1695, which suggested that the occupations and even the characters of people might be determined by the soils that existed where they lived. Aubrey went even farther: he attempted to explain the distribution of witches and religious instability on the basis of geology. North concluded: ‘‘What, then is the archaeologist to do? Is an advanced course of geology to be regarded as a necessary part of his equipment? I would suggest that what he needs is sufficient knowledge of geology . . . to enable him to deal with straightforward matters . . . to determine whether masses of stone are likely to owe their shape and position to natural agencies or human activities . . . to appreciate the geologic significance of a site . . . [to acquire a] nodding acquaintance with stratigraphy [and the ability] to recognize the rock types most frequently represented by the relics of early man.’’ 99

CHAPTER 2

Sediments, Soils, and Environmental Interpretations Prehistoric archaeology is so intimately bound with the physical and climatic conditions of the past that it demands the closest co-operation between the geologist and archaeologist. —E. W. Gardner 1934

S

ediments and soils on or near the earth’s surface contain the tangible remains of the human past and the ancient world. The deposits where artifacts are found provide information on the age, landscape, and environmental setting of human occupations and on the processes that formed the archaeological record. Most archaeological data are recovered from sedimentary deposits or associated soils. In terms of process, artifacts can be considered sedimentary particles that contribute to the final character of the archaeological record. Sediments accumulate either mechanically or chemically. Mechanical accumulation includes the deposition of fine sands and mud particles that bury human occupations on floodplains alongside rivers. Artifacts in these sites may also be eroded, transported, and redeposited as sedimentary particles. Lime mud in lake basins and other chemical accumulations can cover and protect former sites of human occupation that were once situated near the shoreline. Other natural processes, including human behavior, may alter sediments after deposition. For example, humans may later add chemical elements like phosphorus to sediments and soils. Here, a review is provided of some principles, processes, and attributes of the sediments and soils that may contain artifacts or other archaeological features reflecting past human behavior.1

From a geoarchaeological perspective, artifacts that form archaeological sediments are a special kind of geologic deposit. They are a biostratigraphic deposit. That is, they consist of sediments that contain the remains or traces of past life, either due to the presence of objects modified by people or the remnants of materials—rocks, plants, or animals—used by humans in the past. Because the same principles apply to sedimentary settings regardless of whether they contain artifacts or other archaeological features, archaeologists need to understand sedimentologic concepts. These concepts form the basis of better evaluations of the environmental context of sites and the conditions that affect the final archaeological record. Sediments and soils also provide a systematic framework in which to describe the deposits associated with artifacts. This chapter is an introduction to the origin, classification, and description of sediments and the soils that can develop on and within them.

Sediments An understanding of the steps involved in the formation of sedimentary deposits provides an insight into events that contribute to accumulations of artifacts.2 The major steps involved in the formation of sedimentary deposits are the weathering of the source rocks, the transportation and deposition of particles, and postdepositional alteration (fig. 2.1), which includes changes caused by processes of soil formation. These three basic processes affect archaeological materials as

25

Sediments and Soils

soil formation, bioturbation, and diagenesis (see Chapter 3). An understanding of the processes that form sediments and soils found at and around archaeological discoveries is crucial in any attempt to infer the landscapes and environment that were inhabited by ancient people and to reconstruct the potential sequence of events that led to the archaeological patterns preserved and available for study.

Weathering

Figure 2.1 The Formation of Sedimentary Deposits The initial context consists of unaltered bedrock (igneous, sedimentary, or metamorphic) or the original artifactual patterning derived from human behavior. During the weathering stage, bedrock and artifactual materials begin to break down into smaller particles; sometimes they are destroyed. After weathering, artifactual and other clastic particles can be moved and deposited in secondary context. More weathering or alteration of the materials can occur after the final deposition event, before discovery and study.

well. Pottery, stone artifacts, and archaeological features can be weathered, sometimes even destroyed. Artifacts can also be moved and redeposited.3 Artifacts found in the exact position where they were last used or affected by human behavior are designated as being in primary context. In contrast, archaeological materials that have been removed from the primary context by biologic or geologic agencies are designated as being in secondary context. Most archaeological materials and features can be influenced by postoccupational and postdepositional processes like

26

Sedimentary deposits are composed of materials derived from the weathering of preexisting rocks or sediments. The initial destruction of rocks can be either physical (also called mechanical disintegration) or chemical (also called decomposition). Rocks and artifacts disintegrate in many ways; frost action, development of joints and cracks, and abrasion by organic, aqueous, cryogenic, or eolian (wind) processes can all contribute to weathering. The mechanical disintegration of rocks within caves and rock shelters contributes to the fill on the floor of the cave, burying the remains of prehistoric occupations. For example, caves initially formed principally by chemical dissolution of the carbonate by groundwater acids can then be filled partially by fragments of rocks that break off the cave ceiling. Mechanical disintegration and chemical decomposition can also destroy artifactual records. For example, sediments containing Acheulian hand axes and other Lower Paleolithic artifacts at Bir Sahara in Egypt’s Western Desert have been modified by weathering processes.4 Wind erosion has removed some of the sandy matrix that contained the artifacts. Erosion by the wind also selectively destroyed the smaller stone artifacts, leaving larger hand axes as a deflational lag around an ancient spring mound or pond deposit probably dating to before four hundred thousand years ago. Chemical processes involved in weathering can be useful in dating artifacts (see Chapter 5). Weathering processes produce the clay minerals that humans once used extensively to make ceramics, one of the most important categories of artifacts (see Chapter 7). The in-place alteration of rocks and sediments to produce soils is another

Sediments and Soils

Figure 2.2 Tephra Tephra (volcanic ash) layers are useful as chronostrata —they are excellent time markers. In western North America, volcanic eruptions from the Cascade Range that erupted during the end of the Pleistocene and dur-

ing the Holocene can be found above and below sedimentary layers containing artifacts. Tephra from one of the eruptions of Glacier Peak dates to about 11,250 B.P. while the Mazama tephra dates to middle Holocene, about 6,800 B.P. (Photo by Christopher Hill)

product of weathering. Soils form on exposed, stabilized surfaces that are also locations for human activity and the accumulation of artifacts. When buried by younger deposits, ancient soils are key locations for isolating potential living surfaces of prehistoric human occupation.

fluence which artifacts are removed from their primary, behavioral context. Higher energy levels lead to increased erosion and the transportation of larger or heavier objects. Objects can be removed from their primary contexts and transported in many ways. Rain, which causes surface wash, or running water in streams and along the shores of lakes or seas may result in water transport. Strong winds can transport and deposit sands and finer materials forming dunes and loess deposits. Volcanic eruptions lead to the transport and deposition of ash, such as the final Pleistocene Glacier Peak ash and Early Middle Holocene Mazama ash in western North America (fig. 2.2). Ash deposits from volcanic eruptions that were transported by wind in East Africa are important marker horizons in the dating and correlation of artifact and fossilbearing Quaternary deposits (see Chapter 5).

Transportation Once geologic or archaeological materials have been weathered, they are subject to erosion and transportation by running water, wind, or ice. Transported particles and artifacts are eventually deposited in one of many types of sedimentary environments.5 Whether a particular artifact or particle will be eroded, transported, or deposited is generally dependent on the size of the object and the energy of the transporting agent. Smaller or lighter objects are more easily transported. The amount of energy within the system will in-

27

Sediments and Soils

Glaciers slowly transport particles of all sizes and have been known to bury or move artifact-bearing deposits. Gravity flows can also initiate translocation and deposition, as well as slumping, such as at the Calico Hills site near Yerma, California, in an alluvial fan.6 This is a controversial locality; some scientists contend that it contains evidence of humans in America as long as two hundred thousand years ago (see Chapter 5). At this site, geologic processes of transportation seem to have created naturally fractured rocks, which resemble stone artifacts in Pleistocene fan deposits consisting of mud and debris flows. Skeptics argue that processes of transportation in mud and debris flows can create naturally fractured rocks that have the appearance of objects flaked by humans but that are not artifacts. This type of earth movement can also bury entire human occupations, as happened at the Ozette site along the Pacific coast in Washington.7 Here, during periods of high rainfall, the ground became saturated with water. This caused a hillside to collapse, and the mudslide buried prehistoric and early historic dwellings. One of these mudslides, which happened around 1750, buried and preserved a village. Earth slides can also block streams, causing lakes to form, along which human occupations might be preserved. Biologic processes, such as the trampling of land surfaces by humans or other large animals, also transport sediments and artifacts. In one site study, it was found that in sandy deposits, artifacts of different sizes were displaced by trampling in different ways.8 Larger artifacts were moved upward, in contrast to smaller artifacts, which were pushed downward in the sediments. Mixing, caused by the root action of small plants and the burrowing of animals, is another way biologic agents can move particles. At an Archaic shell mound situated near the Green River in Kentucky, Julie Stein’s research showed that mixing caused by earthworms nearly obliterated the boundaries between archaeological and stratigraphic features.9 People living in the area from about 5,149 b.p. to 4,340 b.p. created the mound, which consists of a 2 m–high pile of discarded debris. The mixing of this debris by earthworms created a gradational

28

boundary between the sediments underlying the mound and the overlying midden material. It also eradicated the boundaries of pits within the midden and made it difficult to distinguish the plow zone.

Postdepositional Changes After deposition a variety of changes can occur within a sediment or archaeological layer. Such alterations are superimposed on the original makeup of the sediment and its constituents. They reflect postdepositional and postoccupational environments and climatic settings. Occasionally the weathering cycle begins again. In the new cycle, the weathering leads to the development of soils and the production of secondary alterations and accumulations. Although they are disruptive of primary relations, secondary accumulations like carbonate crusts on artifacts can provide minimum ages for the artifacts and can furnish evidence of changing environmental conditions. Alterations like diagenesis and lithification include the processes of cementation and recrystallization of the deposited materials. The introduction of substances like calcium carbonate, iron oxide, and silica causes cementing. In aqueous solution these substances infiltrate and lithify deposits. This lithification can slow later weathering, contributing to the preservation of archaeological sites. Such hardening can also hinder the excavation and recovery of artifacts. Some postdepositional accumulations can be helpful in archaeological interpretations. For example, artifacts lying within sediments can become encrusted with secondary carbonates. The carbonates coating the artifacts can be dated (by methods like U-series dating; see Chapter 5). Since the carbonates were formed after the artifacts were deposited within the sediments, the age of the carbonate provides a minimum constraining age for the artifact.

Archaeological Implications The term archaeosediment is used to distinguish deposits that form the part of the sedimentary record resulting directly from past human activities. Archaeosediments are artifacts or ar-

Sediments and Soils

chaeological features that are in primary context. Archaeosediments include the charcoal and burned areas of a hearth or fill deposits that have not suffered postoccupational disturbances. It is sometimes difficult to ascertain when artifacts are not in primary context. A buried floor, a pit fill, and a hearth are all archaeosediments. However, an assemblage of secondarily deposited pottery or stone tool fragments may not be considered archaeosediments. If the artifacts can be inferred to have been redeposited directly by human activity, as with middens or fill, then they are considered archaeosediments. Deposits consisting of stones from fallen walls or objects that have been eroded, transported, and redeposited by other geologic agents would in particular cases be thought of as geologic sedimentary particles. These objects may become so far removed from the original prehistoric behavioral context that important information on past human activity is lost. Although the artifacts themselves may still contain valuable insights into past human behavior, many contextual associations will have been modified. These modifications usually take the form of biologic or geologic restructuring of the previous behavioral patterning created by human activity. Sedimentary deposits can be composed of artifactual materials that have been so greatly disturbed by postoccupational processes that the original patterns produced by human activity have been lost. From a geoarchaeological perspective this is one problem of applying behavioral interpretations to a set of artifacts. The ability to infer past human behavior from isolated or mixed artifacts in redeposited contexts is limited. Bruce Gladfelter coined the term articlasts to describe disturbed artifacts and to compare them to geologic clasts or clastic particles in general.10 Articlasts are objects once used by humans that are found in sedimentologic situations far removed from their original human context.

Classification of Sedimentation Products Sediments and sedimentary rocks consist of three main groups. The first, clastics, are sediments

composed of fragments of other rocks and deposited by physical processes. Objects like stone tools, pottery, and bones can be deposited as clastic sediments. The second group consists of chemical deposits. Chemical and biogenic depositions result in the formation of carbonate, evaporate, and organic deposits. A material like phosphate, which was introduced as a by-product of animal waste, would be a chemical deposit. The final group, organics, includes decomposed artifacts used by humans that were made from plants and animals. Classification provides a means to describe in a structured format the strata and features encountered by the archaeologist. In this format, information can be shared and used by other archaeologists and earth scientists. These classifications provide a foundation for inferring the past environmental conditions and processes that have contributed to the prehistoric record. An introduction to sedimentary rock classifications follows.

Clastic Deposits Clastic (or detrital) deposits are sediments that contain fragments or particles derived from preexisting rocks. Most artifacts (including lithics, metal, and pottery) that are part of a sedimentary deposit are clastic particles. Mineral composition and the range of particle size are the two basic attributes used in geology to classify clastic deposits. The term texture is used in part to describe the range of particle sizes found in a sediment or soil. Several scales are used to differentiate particle sizes. The Ingram-Wentworth scale, which may be used to compare size divisions for soils and artifacts, is in wide use (fig. 2.3).11 Three principal size divisions of clastic materials make up this scale. Deposits composed of large particles are designated gravels, pebbles, cobbles, and boulders. Smaller particles still visible with the unaided eye are termed sand. Very small particles that are not visible without magnification are grouped as mud. Mud is further separated into silt and clay, again based on size. Many terms have been devised to describe deposits containing a mixture of sizes or varying proportions of mineralogic components. Textural and compositional

29

Figure 2.3 Size Categories of Artifacts and Particles Artifacts fall into two categories, with a boundary at 2 mm. Larger artifacts correspond to the gravel size category of sediments. Micro-artifacts correspond to the sediment and soil fractions of sand or smaller par-

ticles. Sherwood (2001) advocates that the span between 6.25 mm and 0.625 mm be defined as the range for micro-artifacts. This is realistic because archaeologists use 6.25 mm screens in sieving. Anything smaller than 6.25 mm would be a micro-artifact.

Sediments and Soils

Figure 2.4 Textural Terms for Sediments and Soils Sedimentologic textural classifications are based on the relative amount of sand and mud (with silt and clay fractions separated). When materials larger than sand are present, terms can be based on the amount of gravel, sand, and mud (with silt and clay fractions combined). The texture classes of soil are based on dif-

ferent designations and employ the term loam for soils containing various mixtures of sand, silt, and clay. An alternate method of categorization is to classify sediments on the basis of the relative amount of sand and mud silicates as well as the relative amount of the sediment composed of carbonate.

classifications of clastic sediments and rock can be visualized by means of triangular diagrams or other similar graphic systems (fig. 2.4). Other terminologies are employed depending on whether the sediments are indurated (hardened) or not. Soil scientists use the same general approach to the size classification of deposits. Unfortunately, the boundaries soil scientists employ to define the divisions between sand and silt and those between silt and clay do not match those used in sedimentology. The division between sand and silt is designated at 0.0625 mm long in sediments and at

0.05 mm long for soils. The separation between clay and silt is 0.004 mm long for sediments and 0.002 mm long for soils. An additional set of terms refers to the relative proportions of these size categories. If archaeologists come across a stratigraphic description that does not explicitly state the taxonomic system being used, however, helpful clues are available. The term loam is commonly employed in soil classifications but is absent from sedimentologic divisions used by geologists (although some European classifications use it). In the same way, the term mud is used in

31

Sediments and Soils

sedimentologic classifications but is not used as a soil class (see fig. 2.4). Gravel Particles that are at least the size of gravel (coarser than 2 mm, larger than sand) indicate high energy levels in aqueous transport or transport in a highly viscous medium. High-energy aqueous conditions are associated with greater possibilities for mixing and mechanical abrasion. Gravel-sized particles can be deposited as screes and debris flows, as rockfall, in streams, along shorelines of basins, and by glaciers. Most objects used by humans, even if broken into fragments, would be in pebble-, cobble-, or boulder-sized fractions. The gravel size category is important because deposits containing these sizes are a source of geofacts.12 Geofacts are nonartifactual objects shaped exclusively by geologic processes. Genuine artifacts found within coarse-grained deposits may have been transported and redeposited from a primary context. Such articlasts may be found as isolated pieces within a coarse-grained matrix or as sets of fairly uniform–sized artifacts in a size-sorted artifact assemblage. Where geofacts are concerned, a variety of processes can so influence the appearance of sedimentary particles that distinguishing them from stone altered by humans may be difficult. These naturally fractured rocks have played an important role in archaeology, especially in the determination of the criteria used to evaluate the existence of early Stone Age sites in the Old World and late Ice Age sites in North America. The critical question is whether particular sets of objects represent lithics manufactured by humans. In Europe and Africa there are collections of objects made by archaeologists who selectively chose specimens that looked as though humans created them but that were actually formed by mechanical abrasion. The question of artifacts versus geofacts will not go away easily. The artifactual status of more than four thousand stones collected from the Early Paleolithic site of Diring Yurikh south of Yakutsk, Russia, was challenged based on the suggestion that they could be geofacts—produced by natural geologic processes. Geoarchaeologists, however, have studied these lithic materials, made

32

lithologic identifications, and pointed out that the lithics in question were wind abraded and polished.13 The geologic context is a lag gravel created by eolian deflation. Deflation is a low-energy geologic process, incapable of breaking quartzite cobbles. A test case for distinguishing artifacts from geofacts was described by Evan Peacock.14 Possible Paleolithic (Clactonian) artifacts were recovered in England from what seemed to be beach cobble deposits affected first by wave action and later by freeze-thaw processes. The battering of the cobbles by waves and the splitting of them induced by thermal change created flakes that showed percussive marks, which gave them the appearance of artifacts. These flakes were compared with lithic material from two known archaeological sites. The comparison of attributes of the known artifact assemblages with those of the potential artifactual flake sets indicated that some flakes were probably artifacts. The question of whether objects are geofacts or articlasts has concerned archaeologists since the late 1800s, and some objects may always be in dispute. Several important criteria are employed to distinguish between artifacts and geofacts. These include assessments of whether the object could exist where it was found had humans not transported it (that is, are there geologic processes that could have introduced the object into the deposit?). Another criterion is whether some other kind of geologic or biologic process that could form what appears to be the flaking on stone objects. Some fragments of chert may show haphazard ‘‘retouch’’ created either by the trampling of animals or by heating. Such objects could be considered artifacts if they were trampled by domesticated animals or if the heating and freezing were caused by human activities, although humans may not have caused the alterations deliberately. In this sense, a burned layer representing an ancient forest fire could be considered an archaeological feature if it was determined that it was caused by human activity. Take, for example, the difficulty in evaluating evidence that humans existed in the New World before about 12,000 b.p. In South America, at the Pedra Furada site in Brazil, natural processes of abrasion may have created objects

Sediments and Soils

that have been identified as pre-Clovis artifacts. Objects from what has been designated the Pedra Furada phase have been dated from about 48,000 to 14,300 b.p., but whether they are artifacts or stone fractured by nonhuman processes is debatable.15 All the objects consist of quartzite cobbles. The source of the quartzite may be a conglomerate that exits about 100 m above the site. It is possible that the flaked stones were created when cobbles eroded out of the conglomerate and fell to the site, where they were flaked and fractured. Sand Sandy deposits contain particles that range in size from 2.0 to 0.0625 mm, or for soils from 2.0 to 0.05 mm long (see fig. 2.4). Although textural terms like sand refer to a size category and are independent of mineralogic composition, in most instances, sand is largely composed of the mineral quartz. Except for very small objects like beads and seeds, few items used by humans are less than 2 mm, but debris derived from manufacture, use, retouch, and breakage of larger artifacts can be this size. Such artifacts are designated microartifacts. The identification of objects as microartifacts may be the first indication of a previously unrecognized artifact zone in a deposit. The presence in a deposit of micro-artifacts alone is a strong indication either of a very specific type of human activity, such as surface sweeping or lithic artifact retouch, or of size sorting by geologic or biologic processes. Micro-artifacts can be used to identify potential surfaces of human occupation, especially when these artifacts come with supporting evidence like larger artifacts, buried soils, or fossil fauna. In such cases, attention should be given to any processes that might form geofacts in the same size range. Mechanical abrasion of sedimentary particles can produce micro-geofacts, which even have bulbs of percussion similar to lithic artifacts. A variety of human activities can cause the deposition of micro-artifacts. One such activity is the flaking of rocks to produce lithic artifacts.16 This activity can create considerable microdebitage—sometimes more than 99 percent in sizes smaller than 1 mm. Even when the larger lithic flakes have been carried away, these accumula-

tions of lithic debris remain within permeable sediments and provide a trace of past human activity. Micro-artifacts can provide evidence of past human activities that larger artifacts do not show. At the Bronze Age site of Nichoria in Greece, techniques for recovering materials smaller than 2 mm long were used that recovered artifacts that influenced archaeological interpretations.17 Although the archaeologists did not discover actual smelting features, the presence of small fragments indicated evidence for metalworking at the site through identification of slag within the sediments. Bone and teeth of microfauna; botanical remains like seeds, flint, and obsidian; and spatters of copper recovered technologically at the site provided a more comprehensive record of human activity. In addition to providing micro-artifacts, sandsize sediments are important in archaeological interpretations for other reasons. They can be assigned to specific depositional contexts or environments like fluvial (river) settings, shore zones of lakes and oceans, and windblown sands in caves or rock shelters. In fluvial settings, sands are an indication of relatively high-energy floods or channel deposits. A major problem of interpretation for geoarchaeologists can be determining whether artifacts were part of the initial deposit or were introduced later, either by human activities or by mixing processes. If there is evidence of soil development or other indications of a surface of past human activity, artifacts found on or in the top part of a sandy deposit may be in primary context. In contrast, if the sands contain larger objects that can confidently be considered part of the sediment, axial-orientation comparisons of the artifacts with the particles known to be transported may indicate whether the artifacts were deposited as part of the initial sedimentologic event or afterward. Another, related problem involves artifacts that form part of a deposit that exhibits soil development. There are at least three possibilities here. First, the artifacts may be associated with the event that deposited the surrounding sedimentary matrix and so be in secondary context. Or the artifacts may have accumulated before the be-

33

Sediments and Soils

ginning of the soil development. In this instance, either they could be in primary context or pedogenic processes could have moved them. A third possibility is that the artifacts could have been deposited during or after the period of soil development within the sediments. Sometimes crusts or other surface features on the objects, or the spatial patterning and size sorting of artifacts, can help us choose among these possibilities. The presence of a range of artifact sizes in sands of a uniform size does not necessarily help the archaeologist decide whether the artifacts are in primary or secondary context. The transportational and depositional energies indicated by the sandy matrix provide only a minimum estimate of the size of particles that could be moved. If feasible, archaeologists should conduct studies of the technological integrity or refit analyses of the artifact assemblages in addition to the sedimentologic analysis to obtain a more comprehensive site formation evaluation. These types of studies can help evaluate whether spatial patterns are a result of past human activity or geologic restructuring. The classification of sands and sandstones can be applied not only to the strata associated with artifacts but also to archaeological materials themselves. For example, the blocks that make up the pyramids in Egypt can be so classified. For archaeologists this type of classification provides a standardized approach to the description of sanddominated rocks and sediments, and it helps in the evaluation of their origin within a stratigraphic sequence. The same kind of classification system could also be used to describe pottery or such construction materials as mud brick. Mud Sediments and rocks dominated by clastic particles less than 0.0625 mm long are muds and mudstone (see fig. 2.4). They are also called argillaceous deposits. Muds can further be divided into larger particles called silts and smaller particles called clay. These are separated at the 0.004 mm boundary. Further subdivisions of muds can be based on mineralogic content. If detrital muds contain significant amounts of calcium carbonate (CaCO 3 ), they are called marls. If they are indurated they are called marlstones. Marls often

34

originate as biogeochemical deposits in the bottoms of lakes and ponds. Lime mud and detrital silts and clays generally are deposited nearer the central part of a lake basin and may be indicators of climate change. One example of the close connection that can exist between carbonates and silts and clays is provided by a stratigraphic sequence at Bir Sahara in Egypt.18 The sequence shows a change in depositional circumstances associated with a Middle Paleolithic site. Sediments deposited during the period when the basin contained its greatest amount of water are high in carbonate and contain the highest amounts of silt and clay. This represents a retrograding sequence of deeper water sediments overlying shallow water sediments (fig. 2.5). These transgressive conditions are reflected on the margin of the basin by calcareous sandstone. The prograding sequence (for example, beach or dune sands deposited above muds in deeper water) is associated with artifacts. At Bir Sahara the lake deposits are overlain by sands containing Middle Paleolithic artifacts. At this site the hominid occupation seems to have occurred after the transgressive event. Wetter climatic intervals can also cause lake expansions that inundate prehistoric occupation sites situated on former shorelines. In these circumstances the sites can be covered with biogenic marls or clastic muds. Depending on the circumstances, fine-grained sediments can be indicators of either low- or highenergy levels of transportation. Fine-grained clastic deposits are indicators of very low energy depositional conditions in aqueous conditions. These are typical of backwater floodplain deposits. The erosion and transport of silt-sized particles by wind would imply higher energy levels. These finer, well-sorted particles typically exhibit little internal stratification and are termed loess. On occasion windblown loess accumulation is interrupted by an interval of relative landscape stability that provides conditions for weathering and organic and chemical accumulation. Soil zones, called pedogenic horizons, that lie within a top section of silt-dominated loess are an indication of climatic change and are potential surfaces of ancient human occupation. Muds have also been used by humans in many

Sediments and Soils

Figure 2.5 Sedimentary Sequences and Facies Sedimentary sequences can contain evidence of both contemporary (lateral) and temporal (usually vertical) variation of depositional environments. Sediments can be deposited simultaneously in a variety of laterally adjacent settings. Coarser clastics may be associated with shore zones of a lake, while deeper parts of the lake may be associated with the deposition of muds or marls. As the size of the lake fluctuates, the

location of characteristic types of deposition changes. Archaeological sites situated on lake shorelines move from place to place as a result of the changing location of the water’s edge. During lake expansion, sites will eventually be buried by sediments deposited nearer the center of the basin. During periods of lake reduction, sites may be buried by basin margin deposits like beach sands and dunes.

important ways. Muds are a common constituent of such construction materials as bricks. Muds are also the primary constituent of pottery and figurines. Although the earliest known ceramic vessels ( Joman pottery) date to about 12,000 b.p. in Japan, figurines composed of baked silts and clays are known from about 26,000 b.p. in Europe.19 Living surfaces composed of mud, whether human-made or deposited by geologic means, can influence the spatial arrangement of artifacts. Artifacts moving vertically through a sandy deposit can be impeded by a mud layer, creating a secondary artifact zone. Because mud layers are more impenetrable and less susceptible to mixing by trampling, many separate occupations may be superimposed on a surface. In addi-

tion, artifacts on hardened mud surfaces are more easily moved and rearranged than those on sandy surfaces.

Chemical Deposition The second major category of sediments consists of minerals that have been deposited by precipitation from solution. The most common of these minerals is calcite (calcium carbonate, or CaCO 3 ). A wide range of chemical deposits exists that can assist us in evaluating paleoenvironmental settings. Many, such as iron oxide, manganese oxide, sulfate, silica, and phosphate, have been important resources for humans. Humans and human activity are also important contributors of chemicals into sedimentary systems.

35

Sediments and Soils

Calcareous Precipitates Sedimentary deposits can consist of a mixture of clastic materials and chemically precipitated minerals, such as calcium carbonate or silica. Like silica (quartz, or SiO 2 ), carbonate can be detrital, but its initial formation was as a chemical precipitate. Calcite is common in some sandstones as a secondary cementing agent, and it is an indicator of diagenetic conditions. Carbonates are also chemically introduced into a sedimentary deposit as a product of pedogenic processes. Constituents like calcium carbonate that are introduced after the primary detrital depositional event provide clues to the later climatic, hydrologic, biologic, and chemical processes that have affected artifact accumulation (see Chapter 3). Various terms have been used to designate the many forms of carbonate found as sediments. These names are applied to sediments and rocks on the basis of their textural and compositional character. Limestone designates rocks that contain more than 50 percent nondetrital carbonate (carbonate formed within the basin of deposition). Marls and marlstones contain approximately 40–60 percent calcareous matter, with the remaining constituents being detrital mud particles. Marl is a synonym for calcareous silts and clays. Special names are given to carbonate rocks with high fossil contents. A slightly cemented sediment composed chiefly of fossil debris that has been sorted is called coquina. Carbonates of marine origin with at least 90 percent calcite are termed chalk. Although compositional and textural criteria have been used to define some types of carbonates, terms like tufa and travertine sometimes have particular genetic connotations. Porous and spongy calcium carbonate deposits in a terrigenous context are called tufa, whereas travertine is massive and relatively dense. Tufa forms by the precipitation of calcium carbonate from spring water on growing plants. Travertine forms as spring deposits and is precipitated from solution in caves. Carbonate precipitated in caves is sometimes called flowstone. Travertine in caves seals underlying deposits, thus separating and protecting artifact-bearing deposits. This in turn provides opportunities for dating (see Chapter 5).

36

Noncalcareous Precipitates Noncalcareous chemical precipitates and evaporates play a major role in archaeological interpretation. Evaporates are produced by precipitation from evaporating solutions. Evaporates include calcite (CaCO 3 ), gypsum (hydrous calcium sulfate, CaSO 4·2H 2 O), anhydrite (anhydrous calcium sulfate, CaSO 4 ), and halite (sodium chloride, or NaCl). These sediments are important indicators of environmental conditions associated with human habitation. Under evaporative conditions minerals precipitate out of solution in a particular order. The identification of the mineral precipitation sequence in a stratigraphic profile is used as an indicator of evaporative conditions and changes in water chemistry within a basin. In addition to their use as indicators of past environmental conditions, such evaporate minerals as salt (halite) are important resources that have been extensively used by humans in the past (see also Chapter 7). Other minerals or mineral groups used as resources by humans are useful environmental indicators. These include iron oxides, silica, phosphates, and manganese oxides. Depositional environments for manganese oxides are near springs, in natural and artificial wells, in bogs, and in lakes. Iron in sedimentary rocks is primarily derived from the weathering of mafic igneous rocks and can be transported as soluble ferrous ions or colloidal ferric oxides. In freshwater conditions iron is deposited as the mineral limonite, while in saline settings iron oxide may be deposited as the mineral hematite (see Chapter 7). Phosphates have been the subject of much study because they are a biologic (including human) contribution to the sedimentary system. Their spatial variation and concentration have been used as indicators of loci of human activity. Robert Eidt proposed that the relative proportions of different kinds of phosphate could be related to past human land use.20 Very soluble phosphates were associated with lands used for crop production—specifically, mixed vegetable cultivation. These lands contained small amounts of two other types of phosphate: tightly bound iron and aluminum phosphate, and apatite and other calcium phosphates. In forests there are small

Sediments and Soils

amounts of apatite and calcium phosphates but roughly equal proportions of easily soluble iron and aluminum phosphates. In abandoned residential land-use areas the amounts of the three types of phosphates were about the same. From this we can infer that some chemicals may produce spatial patterns indicative of particular human uses of the landscape. Chemical elements may be added to a human occupational area by specific activities. Dincauze’s interpretation of the soil chemistry at the eight-thousand-year-old Neville site in New England provides one example.21 In deposits associated with the Archaic occupation of the site, chemical analyses indicated the presence of high amounts of mercury. These high concentrations of mercury were interpreted as indicators that the area had been used during the Archaic to process fish.

Organic Matter Decayed and decomposed plants and animals form another critical component of the sedimentary system. Organic material is vital to soilforming processes and a valuable indicator of past environmental conditions. Because it contains carbon, organic material is critical for radiocarbon dating (see Chapter 5). Organic matter in sediments can reach levels of nearly 100 percent in some Quaternary peatlike deposits and may be an indication of freshwater-marsh conditions. Sediments containing high amounts of organic matter are termed carbonaceous. After it has been deposited, organic matter can be used by plants and animals, undergo biochemical decay, or be destroyed by oxidation. When organic matter shows no signs of extensive alteration or destruction, it can be an indication of an environment with a high rate of sedimentation, where rates of burial were rapid. It is also an indication of anoxic conditions, which inhibit organic activity. Extremely well preserved human remains, wood objects, cloth, and skins have been recovered from bogs where anoxic conditions exist. In Europe depositional circumstances have led to extraordinary preservation of organic materials associated with human prehistoric occupa-

tions. Perhaps the best-known sites are the Swiss lake dwellings, but there are others, such as Starr Carr in northern England.22 Research at this Mesolithic site showed that it was a lake-margin settlement occupied some ten thousand years ago. At both the Swiss Lake dwellings and Starr Carr, organic material was preserved because of inundation by rising water levels. Wet boglands have also preserved prehistoric human remains in northwestern Europe and Florida.23 For example, the remains of a man hanged about two thousand years ago were found near Tollund, Denmark. The organic preservation was so good that it was possible to determine what he ate for his last meal (porridge made of barley, linseed, herb seeds, and wild seeds). Preservation of these organic materials appears to have been the result of several factors. First, the body lay in water deep enough to protect it from scavenging animals and anoxic enough to keep it from bacterial decay. Second, the water in the bog contained enough tannic acid to preserve the body’s outer layers. Last, the water was cool enough to slow decay and rot. In Florida, waterlogged peats at the Windover site have preserved human remains and organic items dating to seven thousand years ago. The site had been used as a burial ground, one where bodies were deliberately placed in the pond. Clothing, wood artifacts, human hair, and brain tissue were preserved as peats filled the pond. Other remarkable examples of preservation of organic materials come from the Paleolithic record in Africa and Europe. In some instances fragments of wood have been preserved with archaeological assemblages that are hundreds of thousands of years old. In Zambia, Africa, at the site of Kalambo Falls, fragments of wood have been recovered in association with Late Acheulian artifacts that likely are older than 300,000–200,000 years.24 Wooden spears dating to around 200,000–150,000 years ago have been discovered at the sites of Clacton in England and Lehringen in Germany. The oldest known wooden artifacts are a group of wood spears dating to about 400,000 years ago found at a site near Schoningen, Germany.25 Anoxic settings can also preserve pollen and macrofossils, which pro-

37

Sediments and Soils

Figure 2.6 Geologic and Biostratigraphic Units Sedimentary layers can be categorized according to their geologic characteristics or their biologically derived components. Geologic units include strata defined on the basis of their rock or chemical composition and size or lithostratigraphic units and pedostratigraphic units (strata based on secondarily derived char-

acteristics associated with soil formation and in situ weathering). Biostratigraphic units include biozones, defined on the basis of organic remains, or sets of artifacts, designated artifact components. The same stratigraphic sequence can be described in more than one way, depending on whether geologic or biologically derived characteristics are available for observation.

vide another set of indicators of paleoenvironmental conditions (see Chapter 6).

ing profiles developed within preexisting sediments. Soils are the result of postdepositional processes superimposed on sediments. Weathering, alteration, and accumulation create soil horizons in the top part of a deposit (figs. 2.6, 2.7). The five soil-forming factors are: (1) parent rock, (2) organisms (including humans), (3) topography, (4) climate, and (5) time. Although parent rock influences soil development, especially by governing the chemical constituents available, the same soil types can form on diverse parent rocks. For example, in Greece, soils called Alfisols have formed on underlying limestones, limestone conglomerates, marine marls, clays, sands, and silts. The onset of cultivation and agriculture during the Neolithic ushered in a more intensive use of the earth’s surface, and available soil types became

Soils and Buried Soils As Vance Holliday has succinctly written: ‘‘A soil . . . is the result of the complex interaction of a variety of physical, chemical, and biologic processes acting on a rock or sediment over time.’’ 26 Soils are a product of biologic activity and weathering. Because they indicate the presence of stabilized landscape surfaces, soils mark locations of possible human occupation and artifact accumulation.27 The upper part of a sedimentary deposit can contain pedogenic materials, which are weather-

38

Sediments and Soils

Figure 2.7 A Standard Soil Profile Some of the ‘‘horizons’’ that can be seen in a soil profile. The unaltered parent material (R) decomposes, forming a C horizon. With the contribution of organic materials, an A horizon is formed, marking the upper part of the soil. Beneath the organic-rich A horizon is a light-colored E horizon. This is common in leaching environments like conifer forests. In contrast to the E horizon, zones characterized by the accumulation of clays or carbonates are designated B horizons.

vital to human affairs. An understanding of soils is equally vital to archaeology. Unfortunately, in archaeology the word soil is used for two very different things. Incorrectly, soil designates many kinds of surface and near-surface sediments. More precisely, soil is the portion of earth-surface material that supports plant life and is altered by continuous chemical and biotic activity and weathering. The lower limit (depth) of a soil zone can often be associated with the lower limit of biologic activity. It may be useful to think of sediments as being, for the most part, biologically dead. In contrast, many soils develop in sediments that are alive, that is, biologically active. Soils often represent a zone of biotic interaction. A cubic meter of

agricultural soil may contain more than a million living creatures, which will affect chemical inputs and mixing. Most sediments do not have active biologic constituents; those that do are being affected by soil-forming processes. Soils often show direct effects of rock weathering. They reflect the composition of the underlying (parent) rock, as it has been modified by climate and/or biologic activity through time. Soil formation results in the development of a ‘‘soil profile’’ that exhibits a vertical series of horizons (see fig. 2.7). A soil horizon is ‘‘a layer of soil, approximately parallel to the soil surface with characteristics produced by soil forming processes.’’ 28 There is also a series of roughly horizontal layers that are due to primary clastic and chemical depositional phenomena. It is imperative that the primary depositional phenomena that form sediments not be confused with postdepositional alterations, including soil-forming processes. In the field, soils (as with sediments) can be described by such properties as color, texture, structure, boundary characteristics, and horizontal continuity. Archaeologists must concentrate on determining which properties represent postdepositional soil-forming phenomena imposed on the primary deposit.29

The Soil Profile The properties of soils change both vertically and horizontally. In general they are more easily observed locally along the vertical dimension by differences in physical, chemical, and biotic attributes (see fig. 2.7). Thus, the soil profile is the vertical sequence ranging from the original underlying parent material to the overlying surface-interaction zone. The causes of the vertical differentiation of earth’s surface materials into soil horizons are both geologic (parent rock composition and stratification) and pedogenetic (biotic and climatic). Soil scientists have given several master horizons standardized symbols designated by capital letters, starting with bedrock (R) or unaltered parent material (C). B horizons often are developed in R or C horizons and are overlain by O, A, or E horizons. The B horizon is a mineral zone that shows few of the characteristics of R or C, either because of the in situ

39

Sediments and Soils

illuvial accumulation or the development of secondary structures. E horizons are characterized by the loss or accumulation of components, while A horizons tend to be darker, primarily because they contain higher amounts of organic matter along with mineral matter. O horizons occur on the surface; because they contain high amounts of organics, they are very dark in color. Petrocalcic zones (K) are well-cemented calcium carbonatedominated horizons. Based on the in situ accumulation of particular components, B horizons can be characterized in additional detail, as can other horizons. A standardized set of symbols is used as suffixes to denote these compositional differences in soils. These lowercase letters describe the master horizon. An accumulation of highly decomposed organic matter is signified by the symbol a. An illuvial accumulation of iron and/or aluminum is denoted by s. Clays are denoted by t. The accumulation of exchangeable sodium is signified by n. B and C horizons with accumulations of calcium carbonate are designated by k (calcic), those with gypsum, by y. The presence of salts that are more soluble than gypsum is indicated by z. Combinations of these subordinate symbols can be used to describe the master horizon. For example, a natric horizon (significant clays with sodium) is denoted Btn. The record of past environments as recorded in sediments is biased toward regions of active deposition, called aggradation. Soils, in contrast, are a record of the ‘‘quieter’’ times of landscape stability that come between events like floods, which cause rapid deposition or erosion. These intervals of nonaggradation result in a relatively stable surface for human activities like agriculture. Any major disturbance caused within a soil by such human activities as plowing or digging creates a visible change in the soil horizons. Thus, it is possible to reconstruct past events by studying the morphology of soils on a site and comparing it with the undisturbed morphology of an offsite setting. New deposition on the surface of a soil can be recognized in the field by several characteristics. An unusually thick A horizon (the mineral horizon formed below the organic surface hori-

40

zon or the surface containing both mineral and organic matter) may indicate the introduction of sediments. Other indicators of renewed deposition are sediments that cannot be related to pedogenic processes, the presence of carbonates in the surface layer that are missing from the horizons immediately below, and distinct color changes. Plowing, a specific kind of bioturbation, mixes the upper horizons of a soil profile and makes it impossible to distinguish them from one another. Nonhuman turbation processes can have the same effect. Artifacts within the disturbed zones may have originally been associated with distinct chronological intervals, but they lose their original temporal and behavioral context because of disturbance. There is often a distinct, nongradational boundary between an overlying plow zone and undisturbed underlying soil horizons or sediments. An ancient burial that was covered with soil from the immediate vicinity often shows a distinct change in character at the foot of the artificial mound. This criterion should distinguish natural mounds from burial features. Frequent disturbances of the soils in habitation areas can prevent the pedogenic horizons from forming, as can such bioturbation as earthworm activity. More than 75 percent of stratified and preserved prehistoric archaeological sites in the eastern United States are located in alluvial valleys. In these places, an understanding of the sedimentation and the soil-development sequences is critical to the interpretation of site formation processes. Soil formation on alluvial terraces is usually quite different from that on floodplains. In his discussion of alluvial soil formation and geoarchaeology, Reid Ferring points to the concentration of sites in alluvial valleys, the importance of soils in defining and correlating stratigraphic units, and the use of alluvial soils for paleoenvironmental reconstructions.30 In addition, the number of soils in a stratigraphic sequence and their degree of development can be important in assessing the duration and integrity of the archaeological record at an alluvial site. The boundaries of a pedostratigraphic unit can be time transgressive, whereas chronostratigraphic units, by definition, are synchronous. It is also important

Sediments and Soils

to note that the stratigraphic Law of Superposition, inviolate in sedimentary sequences, does not apply to soil horizons. Two words encountered frequently in the literature on soil are xeric and mesic. Xeric connotes a habitat characterized by a low or inadequate supply of moisture characteristic of cool, moist winters and warm, dry summers, such as is found in the eastern Mediterranean. Mesic, by contrast, connotes a habitat receiving a moderate amount of moisture.

Soil Types Soil scientists have developed an extensive nomenclature. A few general soil types within this taxonomic nomenclature are especially important in geoarchaeology. Identification of the current soils and reconstruction of past soils is essential for grasping the nature and extent of early agriculture and inferring past environmental and climatic conditions. Among the important soil types are Entisols, Vertisols, Inceptisols, Mollisols, Alfisols, Ultisols, Spodosols, Aridosols, and Histosols. Entisols Present-day soils that are not fully developed or are weakly expressed are called Entisols. These soils may be undeveloped because they are too young or because erosion has removed material as fast as or faster than pedogenic horizons can form. Conversely, newly deposited material may be added faster than soil can develop. Entisols show little evidence of pedogenesis and characteristically have few diagnostic horizons. They usually consist of mineral or organic surface horizons developed on slightly altered parent material or bedrock. For example, major floods of the Yellow River in China deposit so much silt that the new material overwhelms the existing soil profile. Entisols are also common on steep slopes in mountainous regions or in deserts or sandy regions. They currently make up about 20 percent of the earth’s land area. Vertisols Vertisols are usually dark clay soils or soils with a high clay content. Vertisols shrink and swell,

often in accordance with seasonal variations in moisture. This shrinking and swelling is caused by a high clay content, typically more than 35 percent. The clay consists mainly of expanding clays like montmorillonite. (See Chapter 7 for an introduction to clay minerals.) Large cracks may develop for part of the year in these thick, clayey soils, producing a hummock-and-swale topography with a subsurface expression of a disrupted structure. Many Vertisols have formed on igneous terrains with intermediate to basaltic composition.31 Because of their capacity to absorb water, Vertisols retain surface water. Therefore, their agricultural use has been for pasturage rather than forest growth. Worldwide, most Vertisols presently lie between 45° north and 45° south latitude. Vertisols can play a role in artifact taphonomy; the expansion and contraction of clay beds cause vertical movement and the mixing of artifacts. Inceptisols Embryonic or immature soils similar to Entisols (but with slightly more pedogenic development) are designated Inceptisols. Typically Inceptisols have an A horizon with an underlying, weakly developed B horizon. This B horizon has little remnant structure and tends to be redder in color because of the accumulation or leaching of minerals. Inceptisols may not develop ‘‘normal’’ profiles because of resistant parent materials, extensive deposition of volcanic ash, or some other inhibiting condition. Inceptisols form in low, rolling parts of the landscape and in the foothills of mountain fronts. In sequences of alluvial terraces, Inceptisols form between the Entisols nearest the river and the better-developed soils farther away. Inceptisols are widely distributed throughout the world, and many are in equilibrium with their environment. Agricultural uses of Inceptisols have been diverse: Inceptisol vegetation ranges from forest to tundra. Mollisols A dark-colored, humus-rich, deep epipedon (surface horizon) is characteristic of Mollisols. They are the well-known grassland soils of present-day short-grass steppes and long-grass prairies. Many

41

Sediments and Soils

deep, dark soils with relatively fertile topsoil are formed under grassland vegetation. Exceptions include the poorly drained Mollisols associated with low-lying hardwood forests. Most Mollisols are found in low, rolling, or flat country. They form under a wide range of temperatures, in areas ranging from the equator to the poles. In addition, they form on a wide variety of parent rocks, although they are commonly developed within clay, marl, and basalt. Earthworm activity is often extensive in Mollisols, and the consequent mixing affects archaeological sites by altering artifactual spatial patterns. Early farmers realized that these soils could produce a rich variety of foods after the tough sod had been broken by plowing. Alfisols A subsurface horizon with high clay content (Bt) but no humus-rich surface horizon (no Mollic epipedon) is characteristic of an Alfisol overlain by a thin A horizon. Alfisols are the alkaline forest soils (they may sometimes be slightly acid) that form when sufficient clay accumulates in the subsoil to create an argillic (clay-rich) horizon. Alfisols form only in the absence of conditions that favor the formation of Mollisols or Spodosols. They can be found in many present-day climatic regimes but are most common in humid or subhumid environments on young, stable land surfaces that have been free of major erosion or pedoturbation for thousands of years. Alfisols are normally young enough to retain most of their chemical nutrients. Consequently, they are used for cultivating crops, as well as for pasture and forestland. Ultisols Like Alfisols, Ultisols have an argillic horizon, but unlike Alfisols, they are low in chemical bases. These are among the less-alkaline soils formed under hardwood or pine forests. They develop deep, reddish soil profiles in warm, humid regimes on old terrain. Ultisols have a potential for agricultural production but are subject to rapid depletion of nutrients because of deep weathering. They tend to form on older parts of the landscape like exposed bedrock, high alluvial terraces, or

42

tops of plateaus. The natural vegetation of Ultisols is coniferous or hardwood forest. Spodosols Spodosols accumulate subsurface concentrations of aluminum and organics. Often iron oxide and silica cementation is also extensive. These are the ‘‘white earths’’ that used to be called Podzols, in contrast with the ‘‘black earths’’ that were called Chernozems. (‘‘Chernozems’’ included many soils that now have separate names.) Spodosols are often acidic, ashy gray, sandy soils. They respond quickly to changes in vegetation. Hence, they form more rapidly than most other soil types.32 Spodosols may take only a few hundred years to form on quartz-rich sands. Spodosols are naturally infertile and therefore provide only a limited basis for cultivated crops. Coniferous forest is their most common vegetation. Aridosols As the name implies, these are the soils of the arid regions that occupy about a third of the earth’s current land surface. Aridosols are dry for more than three-quarters of each year. Their organic content is low, primarily because of the restrictive moisture conditions, which inhibit biotic activity. Their soil horizons are usually well oxidized. Caliche—horizons of calcium carbonate—may form within and on them. The surface horizons of Aridosols are light-colored, soft, and often vesicular. The lack of water limits their natural vegetation to plants like cactus, yucca, sagebrush, and muhly grass. Aridosols form mostly in lowlying areas; steep slopes in arid regions tend to be eroded down to bedrock. Aridosols can be irrigated for cultivation but at the risk of salinization. Histosols Histosols are the widely distributed organic soils that form wherever the production of organic matter exceeds its conversion or destruction. This comes about when an almost continuous saturation with water inhibits rapid oxidation and slows the decomposition of organic matter. Histosols support bog, swamp, and marsh conditions. They may be drained for cultivation. Peat Histosols

Sediments and Soils

have been used throughout history as a source of fuel.

Paleosols and Buried Soils A paleosol is a soil that formed on a landscape at some time in the past (fig. 2.8). Buried soils in a stratigraphic sequence are valuable indicators of an interval of nondeposition. Paleosols can either be buried soils or surface soils that have developed under fluctuating climatic conditions. Paleosols occur frequently in the geologic record and are well represented in archaeological contexts. Archaeologically, they correlate with times of landscape stabilization and surfaces of human occupation. Buried soils are important as position and time markers of these stable landscape environments where ecologic (and sometimes human) forces dominated, rather than erosion or deposition. Rapid burial by windblown loess or overbank floodplain silts can result in the preservation of land surfaces and can create paleosols. For some scientists the definition of buried soils is restricted by the depth of burial. Peter Birkeland defines paleosols as soils that are buried deep enough to be unaffected by current pedogenic processes.33 The U.S. Department of Agriculture classifies a buried soil as one covered by a mantle of new material that is 50 cm or more thick, or between 30 and 50 cm thick if the mantle thickness is at least half that of the diagnostic horizons preserved in the paleosol.34 There is no simple set of criteria by which to recognize a paleosol. Rather, one must identify a developed horizon or weathering zone resulting from past soilforming processes. In addition to clarifying local archaeological stratigraphy, the identification of a paleosol can enable stratigraphic correlation and chronological resolution.35 Paleosols can be recognized by the occurrence of fossil root traces, an accumulation of phytoliths (see Chapter 4), calcareous nodules, bioturbated layers with gradational contacts (relic soil horizons), and an increased organic content in a darker fossil A horizon. Major- and trace-element analyses offer additional clues to the identification of paleosols. Protracted humid weathering can produce enrichment in TiO 2 ,

Figure 2.8 Paleosol Buried soils in upland silts. Buried soils observable as zones high in organics, clays, or carbonates reflect times of slowed sediment deposition and landscape stability. In this instance, dark organic zones are developed within loess deposited in upland regions of the Yellowstone River basin, western North America. Organic materials from the buried soils provide radiocarbon ages from about 11,000 to 9,000 B.P., indicating these were the locations of landscape surfaces that formed at approximately the time of the PleistoceneHolocene boundary (Hill 2006).

Al 2 O 3 , and Fe 2 O 3 , with accompanying loss of CaO and MgO. In addition, the covering vegetation would have absorbed much of the K 2 O and P 2 O 5 . The most important factor in the distribution and redistribution of trace elements is what elements were released from the parent material. Available arsenic shows an affinity for organic materials but is depleted in leached horizons. Both vanadium (V) and zinc (Zn) concentrate in clays. Pedogenic calcium carbonate in paleosols is

43

Sediments and Soils

usually reduced to about half the quantity found in unburied soils, a decrease that is probably related to soil moisture. Most paleosols are wetter than unburied soils (they often occur beneath the water table). Water facilitates the leaching of carbonates and salts from the system. An increase in soil moisture will cause some clayey soils (for example, ancient Vertisols) to swell. Paleosols tend to have yellower hues and lower chromas (lighter colors) than unburied soils. The differences in color result from a lessening of the oxidation processes, gleying, and the changes in the structures of organic compounds. Water-saturated horizons preserve organic carbon. Burial creates compaction, which causes the structure of paleosols to coarsen. The damage to archaeological materials because of this compressive loading depends on the depth of burial, the compressive strengths of the archaeological materials and of the earth matrix, and the orientation of the artifacts and features with respect to the direction of loading. It should be noted that differential strain rather than stress is the primary reason that artifacts break. Hence, where there are differential compressibilities in the matrix or in matrix-artifact combinations, the potential for damage is greatest. In areas where sediments have been newly deposited or where fresh rock has been exposed by erosion, a progressive vegetational succession occurs, first by herbaceous plants, then shrubs, and finally trees. This biotic activity, aided by hydrogeochemical weathering, creates a soil profile by the incorporation of decaying organic matter (A horizon) and also by the migration downward and the redistribution of soluble compounds and fine particles. Features of fossil soils similar to present-day soils include traces of roots, pedogenic horizons, and pedogenic structures. But because of the increased importance of time as a factor, certain attributes of the original soil development, unlike those of current soils, may have been altered or lost. In addition, the vegetational and climatic patterns that form present-day soils in a region cannot be assumed to have existed in the past. Different soil-forming contexts in the past could create paleosols with characteristics differ-

44

ent from those of present-day soils. Overlying deposits may compress paleosols, changing original thicknesses and causing cracks or other deformation structures. A soil that may originally have been loose or friable might become indurated because of the addition of a cementing agent. This can result in the formation of nodules or hardpans. Recrystallization of mineral components, as well as replacement or authigenic mineralization, may occur. Dissolution, dehydration, and oxidation may affect the preservation of paleosols. In some instances, organic matter might not survive in buried A horizons, but the former A horizon can be recognized sometimes by higher amounts of turbation compared with underlying horizons. An eluvial (E, formerly A1 or albic) horizon in a paleosol profile would underlie an A horizon and possibly be lighter in color, more massive, or more indurated, depending on the removal of clays, organics, or sesquioxides. B horizons are composed of illuvial accumulations (see fig. 2.7). Because the classification of present-day soils is based in part on parameters that do not survive in the geologic record, and because certain climatic conditions must be known for proper classification, earth scientists have developed a classification system specifically for paleosols that focuses on features likely to be preserved and attempts to minimize the use of interpretation.36 The presence of dark organic matter (but not coal) designates a carbonaceous paleosol. Two types of paleosols exhibit poor horizonation: Protosols, which are relatively immature soils, and Vertisols, which lack horizonation because of the effects of turbation. Where the horizonation is good but the environmental conditions are oxidizing or reducing, the paleosol may be termed a Gleysol. The accumulation of minimal insoluble minerals can produce paleosols called Calcisols and Gypsisols. Calcisols have a prominent calcic (calcium carbonate) horizon; they include calcretes and caliche. Gypsisols are rich in authigenic sulfate. Paleosols high in clay may be termed Argillsols, while those high in organic matter and iron are termed Spodosols. Extensive in situ alteration of minerals may cause the formation of Oxisols. Some of the subordinate modifiers that can be

Sediments and Soils

applied to paleosols include albic (eluvial horizon present), argillic (t, presence of illuvial clay), calcic (k, pedogenic carbonate), ferric (presence of iron oxides), gleyed (g, presence of reduced iron), gypsic (presence of pedogenic anhydrite or y, gypsum), silicic (q, pedogenic silica), vertic (desiccation cracks, slickensides), and vitric (presence of glass shards or pumice). Despite the alterations that affect soils as they become paleosols, the attributes of present-day soils provide hints about what those attributes found in paleosols indicate about past environmental conditions. It should be kept in mind, however, that consideration of these attributes in making inferences about past conditions relies on the assumption that analogous conditions existed in the past. This is demonstrably not true in many instances. Weakly developed paleosols (with Entisol or Inceptisol characteristics) can result from conditions where biota were in early stages of succession, or possibly from early stages of mixed woodland or grasslands. Mollisol characteristics in paleosols may indicate open grassland conditions. Paleosols that have characteristics similar to Alfisols may be an indicator of past forests and woodlands, while characteristics associated with Oxisols and Ultisols may be an indicator of more humid rain-forest conditions. Spodosol characteristics in paleosols may indicate humid temperate or alpine settings with needleleaf forests. Distinctive postdepositional features like patterned ground, frost heave, or ice wedges in paleosols may be an indication of taiga settings. The past presence of marsh conditions and waterlogged areas (swamps) could be inferred from the presence of carbonaceous paleosols (peat accumulations, for example).

Inferring Environments from Physical and Chemical Parameters Description and analysis of sediments from archaeological sites and the areas surrounding them provide a means of identifying depositional processes and environments, as well as postdepositional alterations. Environmental interpretations can be developed from an evaluation of the con-

texts of the sediments and from the study of their composition. Paleoenvironmental interpretations are based on comparisons of the attributes of sediments with the characteristics of current environments and other sedimentary or stratigraphic contexts. Here, we review some environmental conditions associated with specific sedimentary attributes found in deposits.

Color One of the most obvious attributes of an archaeological deposit is color. The factors that most influence color are the source rocks, the conditions of weathering, the physical and chemical conditions at the site of deposition, and postdepositional changes. Archaeologists have long used the pattern of colors in excavation profiles to differentiate layers and horizontal disruptions. Throughout the history of archaeology, archaeological sediment and related soils have been described by such inexact terms, without the application of a standardized classification system. With the increasingly scientific orientation of archaeology since the 1950s, the Munsell color notation has become the standard. The Munsell color-order system is defined in terms of three coordinates: hue, value, and chroma. Hue is the quality of color described by the words red, green, blue, yellow, etc. It is the color of pigment that must be mixed with black and white (or shades of gray) to produce the color being matched. Value is the quality of lightness or darkness of the color, measured against a series of gray samples that range from white to black. Zero indicates absolute black; ten, absolute white. Chroma defines the degree of color saturation from a gray of the same value. In other words, it is the amount of pigment that must be mixed with a specific value of gray to produce the particular color. Pure gray colors have zero chroma. Munsell color charts were adopted by the U.S. Soil Survey program in 1949, and the International Society of Soil Science recommended their use about ten years later. A chart for colors of wet soils (gleys) has been added since. Although the adoption of Munsell charts and related names has significantly improved field descriptions of soils and sediments, the results should not be considered precise. In addition to the standard problems

45

Sediments and Soils

of operator variability and errors endemic to any observations, soil and sediment color are affected by such factors as the quality of the light, the moisture content, and dimensions of the areas of individual colors. Soil color is best determined in sunlight, with the light coming over the shoulder. Dry soil or sediment is usually about two units higher in value than the same soil or sediment when moist. There may be a difference of one in dry and moist chromas of a soil or sediment. The hue generally remains the same. The color of soils and archaeological sediments usually depends upon the content of the organic matter and ferric oxides. The thicker the organic matter and ferric oxide coating the grains, the darker the soil. Another generalization is that red soils are older than yellow soils, and they indicate the presence of drainage. Iron oxides are good indicators of sediment and soil-forming environments because the iron oxides include several different minerals, which have different colors. As well, the specific mineral formed is a result of the geochemical environment, regardless of whether it is natural or anthropogenic, primary or secondary. It should be emphasized that redness is related not to the amount of iron oxide present but rather to the hematite content. Also, typical iron oxide colors may be missed if they are masked by the darker (blackish) colors of humic materials or manganese oxides. The important sediment and soil iron oxides are goethite, hematite, lepidocrocite, and maghemite. Goethite, FeO(OH), is by far the most common iron oxide in soils. When it occurs as the only iron oxide (which is frequently the case) it can be recognized by its Munsell 10YR to 7.5YR hue, as long as it is finely distributed. Goethite is nearly ubiquitous in soils, irrespective of climatic zone. It is frequently the cement in concretions. Hematite, Fe 2 O 3 , formed at fairly low temperatures, is usually ‘‘blood-red,’’ with Munsell hues ranging from 5YR to 5R. Hematite often coexists with goethite. Higher temperature, lower water availability, a near-neutral pH, a high iron content of the parent rock, and a rapid turnover of biomass all favor hematite over goethite in pedogenic situations. Lepidocrocite, FeO(OH), usually has a hue of 46

7.5YR—which can also be due to a mixture of goethite with a small quantity of hematite. Although lepidocrocite is metastable with respect to goethite, it still occurs in soils, particularly clayey, noncalcareous soils that undergo anaerobic conditions sometime during each year. Goethite is favored in calcareous environments that have a high concentration of carbonate ions in solution. Maghemite, Fe 2 O 3 , has a hue between 2.5YR and 5YR—midway between goethite and hematite. Maghemite is ferrimagnetic and can be extracted from a soil or archaeological sediment with a hand magnet. Oxidation of magnetite (Fe 3 O 4 ) leads to the formation of maghemite. Other iron oxides are transformed into maghemite by fire that is in the presence of organic matter and has a limited supply of oxygen. This can occur at depth through root burning. A color chart for estimating the organic-matter content of Ap horizons using Munsell color notations and laboratory-determined organic content is available.37 The chart consists of five color chips, which correspond to five overlapping organic-matter ranges of 10–20, 15–25, 20–30, 25– 40, and 35–70 gm per kg. John D. Alexander noted that accurate organic-matter estimations were possible more than 95 percent of the time for medium to fine soils, but that more than 50 percent sand content could cause organic matter to be overestimated. Other studies have indicated that the relation between Munsell value and organic matter for Ap horizons is predictable within soil landscapes, if soil textures do not vary widely, but that different landscapes often have different relations. Studies also showed that the relation is not predictable if soil texture varied widely within the landscape (sands versus silts and loams) and that it is similar among landscapes having the same soil textures and parent materials.38 Darker colors often indicate the accumulation of organic matter in a deposit, but they can also mean that manganese oxide or dark-colored rock or mineral fragments are present. Postdepositional alterations, including contributions from human activities, darken sediments. In occupational deposits, dark colors may be attributable to humans depositing organics in refuse or forming charcoal in hearths. Postdepositional conditions are reflected in

Sediments and Soils

color. Grayish, olive-gray, and brown colorations indicate reducing (gleying) conditions and the presence of ferrous iron compounds. An overall mottled appearance in sediments and soils results from the migration in solution of manganese and iron ions, which leads to patchy accumulations of oxides and hydroxides. These are also characteristic of gley soils or conditions. The seepage of colloids, organic matter, or iron compounds may result in colored streaks and tongues in deposits. Mottling also occurs in soil horizons that are incompletely weathered. Anaerobic decay of biologic materials in marshes and stagnant lakes produces deposits that are usually quite dark, but manganese oxides can also form blackish coatings in swamp conditions. Green colors can result from the presence of green minerals (mostly hydrous silicates), including epidote, chlorite, and serpentine. Without detailed laboratory studies, it is sometimes hard to tell whether the color is a product of the original source rock, later additions, or changing geochemical conditions. White or light-gray colors are produced by a variety of conditions. Source minerals may lighten the color of sediments. Sands composed mostly of quartz or calcite are generally light colored, especially when they have not been affected by changing oxidation states of iron oxides associated with hydrologic and geochemical fluctuations. Light colors may indicate leaching by moving water. Reddish and yellowish colors result from oxidation because of good drainage and aeration, in contrast to more mottled grayish, brownish, and yellowish colors that are indicative of reducing conditions. Reddish deposits may also be an indication of intense heat caused by either human activities or fires of other origins. Reds in particular can imply the presence of alternating wet and dry environmental conditions (such as on floodplains and lake margins). Reddish colors have also been related to weathering zones, like those created during intervals of nondeposition and soil development.

Cementation and Induration Natural cementation transforms loose sediments into consolidated rock. The degree of cementation influences the ability to excavate strata and recover artifacts and faunal material. While loose

sediments can make maintaining profiles and retaining information on the location and orientation of artifacts difficult, well-cemented deposits tend to result in the preferential recovery of larger artifacts and bones and more fragmentation of the materials recovered. In caves in southern Africa where australopithecine fossils were recovered, the deposits were so indurated that excavators used dynamite to blast away the matrix. Not surprisingly, this resulted in the loss of materials and made it harder to determine the provenience of a particular find. In these situations the highly cemented nature of the sediments containing artifacts affects the recovery of artifacts, which can influence interpretations. Calcite cement is commonly an indication of initial diagenesis and lithification of sedimentary deposits. Cementation and induration of a sediment may imply that at one time the deposit was saturated with water. Lithified materials may be natural substances like stones used by humans for structures, or they may be materials like mud brick that are consolidated because of human activities. Cementation of deposits overlying artifact-bearing deposits provides protection from weathering and erosion, thus helping to preserve surfaces of occupation intact.

Texture Texture of sediments refers to the size, shape, sorting, and orientation of particles. The principles used to characterize the texture of a deposit are applicable to archaeological components within sedimentary deposits, especially since artifacts are a specific kind of sedimentary particle. The particle-size frequency distribution in a sediment provides an indication of the transport and depositional systems that resulted in the accumulation of the sediments and artifacts (figs. 2.9–2.11). Particle-size distributions provide some information on the general hydrodynamic conditions at the site of deposition. In archaeological contexts they are useful for distinguishing between high-energy and low-energy site formational environments. Thus, they can be used to determine whether there is potential preferential bias in the artifact components. Particle-size distributions of sediments (also known as granulometry) are useful for determin47

Sediments and Soils

Figure 2.9 Histograms and Cumulative Curves That Describe Size Distributions for Sediments and Artifactual Constituents Histograms and frequency curves show the relative amount of particular size categories within an artifac-

tual or other deposit. Cumulative curves can also be used to compare the relative proportion of different size fractions in deposits. They have also been used in comparisons of the relative frequencies of types of artifacts among archaeological assemblages.

ing depositional environments associated with archaeological sites (see also Chapter 3). A good example is the sedimentological context of a Late Acheulian site from South Africa designated as Duinefontein 2.39 The site consists of red, Festained sands overlain by white sands. There are two paleosurfaces within the red sands containing Lower Paleolithic artifacts that date to about 270,000 years ago or earlier. Granulometric analysis was used to infer that the red sands were of eolian origin; the grain-size distributions of the sediments were consistent with windblown sands. The redness of these sands appears to be derived from dissolved iron in underlying bedrock that coated the sand grains during a wetter interval, possibly linked with high water tables. The sedimentologic context of the site appears to

have been a groundwater-supported marsh surrounded by a semiarid landscape. Another example is the use of granulometry to help study the sedimentological context of a human skeleton from near Kennewick, Washington, that appears to date to the Early Holocene.40 Sediments collected from the stratigraphic sequence in the vicinity of where the skeleton was found and from the skeleton were studied. The analyses indicated there was no statistical difference between the grain-size parameters from the two main stratigraphic units at the site. This meant that in this instance grain-size analysis could not be used conclusively to place the skeleton in either of these units, because both deposits contain overbank alluvium and reworked alluvium with similar textures.

48

Sediments and Soils

Figure 2.10 Grain Size versus Energy in Water Transport At different energy levels, different-sized artifacts can be eroded, transported, or deposited. Sand-sized micro-artifacts need lower energies (velocities) to erode, compared with larger artifacts (artifacts in the pebble, cobble, or boulder range). The larger the artifact, the higher the velocity needed to keep it from being deposited. Larger artifacts will be deposited when energy levels are at velocities of 10 cm per second or less, while sand-sized micro-artifacts can still be transported at velocities lower than 1 cm per second.

In some instances the relative proportions of particular size fractions can be used to distinguish sediments at archaeological sites. For example, the textures of deposits at the site of Tel Michal, situated along the eastern Mediterranean coastal plain, were examined in terms of the relative percentages of three sand-size fractions.41 The Hadera dune sands and Dor kurkar samples were generally composed of coarser sands. In contrast Ramat Gan and Nasholim sediments had higher frequencies of smaller-sized sand grains. These types of difference in the textures of the natural sediments in the vicinity of the site could be

Figure 2.11 Velocity and Modes of Transport of Archaeological Particles Artifacts can be moved within a water column or in an air column by similar methods. In water or by aqueous transport, artifacts roll along without leaving the earth’s surface. Similarly, artifacts can be transported by wind by means of surface creep. At progressively higher energies or for smaller artifact particles, saltation or suspension in water or air can be the means of transport.

compared to the materials used for construction purposes at the site, or archaeosediments. The Netanya hamra and Tel Aviv kurkur were used in the construction of leveling platforms during the Middle Bronze Age. Walls composed of hamra or Ramat Gan kurkur were used to build walls on these platform fills. The Tel Aviv kurkur used during the Middle Bronze Age was replaced by the Late Bronze Age with Hadera sediments. Mudbrick from the Persian period strata correlate with samples of Nentanya hamra. Textural properties have also been used to study sedimentary environments associated with strata containing Middle and Late Paleolithic

49

Sediments and Soils

artifact occurrences in the Nile Valley north of Aswan, Egypt, at Wadi Kubbaniya. Here, two main groups of sediments could be distinguished based on grain size. Wadi washes, sand-sheets, and dunes were typically characterized by high proportions of coarse particles, while fine fraction-dominated sediments were associated with fluvial (Nilotic) and lacustrine deposits.42 Size distributions of sediments and artifact accumulations depend on five conditions: (1) the type of source rock and the original sizes of the grains or artifacts; (2) the type of transporting medium; (3) abrasion and solution during transportation; (4) sorting of size fractions before deposition; and (5) the depositional environment. Postdepositional mixing and pedogenic processes alter the initial size distribution of sediments. Examples include the downward migration and accumulation of finer particles and the upward movement of larger particles like artifacts. Various methods of data presentation and different statistical measures have been employed to help determine which combination of these conditions is reflected in a sediment’s particle-size distribution. Measurements of artifact size distribution like skewness (symmetry) and mode may be extremely useful in archaeological interpretations. The size distribution of debitage from the Kalambo Falls Acheulian site in Zambia helped indicate that some of the patterning of the artifactual record may have been produced by stream action. At first glance the presence of a relatively large amount of unmodified flakes and fragments, along with certain reduction-stage categories, seemed to indicate that the artifact set could imply stone flaking at the site. But the size distribution of the debitage revealed a strong unimodal peak in the 4–8 cm fraction. Experiments on stone flaking show that there is a strong negative skew toward smaller debitage. The Kalambo Falls debitage, in contrast, had normal skewness. The median, mean, and mode of the debitage size distribution were all in the 4–8 cm range. There were hardly any particles of less than 2 cm in size, although in experiments large amounts of this size range are created by flaking. The size distribution in-

50

dicated a strong sorting of the larger flake sizes. One likely interpretation is that the size distribution was the result of stream action.43 Typically, particle-size data are presented either as histograms or as cumulative curves, to show the relative abundance of size fractions and to allow statistical measures to be derived (fig. 2.9). Interpretations of these measures of clastic (exogenic) sediments are used to infer the energy of the transporting agent at the point of deposition (fig. 2.10). Principal depositional sites of coarse-fraction clastic sediments include deserts, beaches, river channels, and lake and sea margins (see Chapter 3). The fine-fraction (small-sized particles that compose a sediment) of a sediment is used to interpret both depositional and secondary contexts. These include transportation/deposition agents (if the silts and clays are exogenic), as well as postdepositional processes. The presence of detrital fine sand to clay particles indicates a low-velocity or zerovelocity transport agent like passive suspension. Statistical parameters based on grain-size distributions are used to help infer depositional environments. They have been related to different forms or modes of sediment movement. These include surface creep or rolling, and sliding or traction in continuous contact with bed flow in rivers. In saltation there is intermittent contact and resuspension of particles, while in suspension there is no contact (fig. 2.11). Rolling and traction transport particles by air or water. This usually adds to the coarse fraction of a sediment. These forms of transport move larger artifacts and micro-artifacts. Saltation is a common system of transport for finer sands. Suspension is the common mode of transport for silts and clays. In glaciers, particles of all sizes can be transported within or next to the ice; they are deposited as the ice melts away. Other statistical methods can be used to describe and interpret archaeologically related sediments. These include mean-particle size, modes and bimodality, sorting, and skewness, as well as bivariate comparisons of these attributes. Meanparticle size for particles is related to current velocity or to the overall energy of the environment. Coarser mean sizes in aqueous and eolian

Sediments and Soils

contexts are indicative of high energy, while finer mean sizes are related to lower energy settings. The most common size fraction with a particle distribution is its mode. If there is a single mode, it is an indication that the sediment reflects a single agent of transportation and deposition. Bimodality of size frequencies generally indicates mixing of sediment from two sources. Sorting, the separation of particles by size, is generally a consequence of variations in transport velocity and turbulence. In well-sorted sediments, particles are of a similar size, while poorly sorted deposits consist of a wide range of particle sizes. Sorting is an indication of entrainment and transport by water or air currents. Eolian microdepositional environments have excellent sorting. Dunes typically consist of coarse to fine sand with good sorting. They can exhibit internal sorting, which causes cross bedding, in contrast to loess deposits, which can be so well sorted they appear massive. Because loess deposits are so well sorted, erosional features or soil development are often needed to help the stratigrapher make subdivisions of them. Sandy beaches tend to be well sorted. However, where energy levels fluctuate and sediments are not continuously reworked, sediments on beaches or dunes can be less well sorted. Spring, lake, and marsh sediments commonly exhibit poor to good sorting, primarily because of fluctuating energy levels. Sediments deposited by glaciers, called till, are poorly sorted. They typically contain very large particles in a finer matrix. One measure of sorting that can be useful to archaeologists is reflected in the ratio of different artifact categories that are characterized by particular size ranges. In the reduction stages for the making of stone tools, for instance, two of the byproducts are cores and debitage. Cores are larger particles than most debitage or flaking debris. Extremely well-sorted artifact assemblages could either consist of a preponderance of cores or be composed entirely of smaller debris. In both instances, the well-sorted character of the accumulation may indicate geologic patterning of the artifact set. Artifact sets that are less well sorted or are poorly sorted (where both larger and smaller artifacts are present) may represent arti-

factual patterns that were a direct result of human behavior. The measure of asymmetry of a grade-size distribution, termed skewness, has also been related to selective transport of sediment size fractions. Variations in transport energy seem to affect the extremes of a particle size distribution and influence skewness. Beach sediments can be negatively skewed (contain more coarse particles). Dune and river sands are positively skewed (contain more fine). Besides interpretations based on statistical parameters of a single particle-size distribution, inferences are made by comparing two or more attributes. Relations like the ratios between coarse and fine clastics are used as a means to differentiate sedimentary facies. Relations between kurtosis (peakedness of the distribution curve) and skewness provide an expression of sediment/ energy response which may be applied to lacustrine (lake) and other environments. Diagrams that plot two statistical measures of size distribution have been used to distinguish depositional environments. In deposits consisting of sand-sized particles, for example, eolian dune, lake beach, and river sands have been compared for sorting and skewness. Beach sands are distinguished by negative skewness and good sorting. River sands tend to be positively skewed and less well sorted. Dune sands have positive skewness and are finer grained than beach sands. In addition to size and sorting characteristics, the morphology of sedimentary particles is an important attribute that is useful in examining artifacts as well. Sedimentary particles are described in terms of sphericity and roundness (fig. 2.12). Shape can be an important indicator of conditions of transportation and deposition. Whether the edges of a particle or an artifact are sharp and angular or rounded can be an indication of the amount or intensity of transportation it has undergone. Such surface features as pitting or abrasion marks also indicate the amount and method of transportation. The property of roundness, or the amount of curvature on the edges and corners of an artifact, changes during transportation in an aqueous system. The shape of pottery changes as it is transported by water. In addition to the possibility

51

Sediments and Soils

Figure 2.12 Measurements of Particle Morphology Clastic particles can be described in terms of sphericity and roundness. Particles, including artifacts, become more rounded as they are abraded, unless forces are strong enough to cause fractures. (Numbers are median rho values.)

of being broken, artifacts can be rounded and smoothed by coastal or stream processes. J. Allen used a quantification of the amount of wear as indicated by roundness as a way of analyzing and interpreting Romano-British pottery assemblages. These artifact sets were found in what is now an intertidal coastal deposit and in subfossil wetland deposits. This study indicated that the Wadell projection technique of measuring roundness was suitable for analyzing transposed and transported pottery sherds.44 Another method for studying the properties of abrasion on artifacts caused by stream transport is microscopic measurement of the width of ridges on artifacts. Fresh, unabraided artifacts have narrow, sharp ridges. The ridge width increases or becomes increasingly flattened with abrasion, and this can be measured. The rate of abrasion depends on the shape and hardness of the artifact as well as on the velocity of the current and on the kinds of particles being transported along with the artifact. In the early stages of artifact abrasion during transport, chipping and grinding of the edges is more active. Stress cracks can appear. With further transport the ridge width increases. 52

By using methods of identifying roundedness and abrasion it may be possible to discover the depositional environments associated with artifact accumulation. One could then evaluate whether artifacts are in primary behavioral context. Myra Shackley examined a late Lower Paleolithic assemblage found in a river gravel deposit. Microscopic examination showed that what on a macroscopic level appeared to be unabraided artifacts in fact were artifacts with signs of rounding caused by stream transport. It was concluded that the set of artifacts was a selective accumulation produced by deposition and not a distinctly behaviorally induced typologic assemblage.45 Abrasion in the form of edge damage to stone artifacts that mimics deliberate retouch (pseudoretouch) is created by trampling by animals. For example, human trampling can affect stone tools by causing unintentional abrasion. Such abrasion has been demonstrated to have ramifications for archaeological interpretations, like the classification of classic Paleolithic assemblages.46 Abrasion can occur to artifacts trampled on both coarsegrained and fine-grained substrates if the sediments are relatively impenetrable. Besides the compaction of the substrate on which the artifacts are lying, the density of the artifact concentrations also influences the degree of abrasion. Much of the damage is caused by the impact of one artifact on another artifact. If stone artifacts are in loose sands they will usually be less damaged because they will be more easily dispersed, and thus will be in a less concentrated area. If the underlying sediments are composed of coarse clastics of pebble size or larger it is difficult to determine whether the abrasion is caused by other artifacts or the substrate. Trampling of artifacts can also cause breakage. Breakage would change the character of the assemblage by decreasing artifact size and increasing the total number of specimens within the sample. A study by Sally McBrearty and associates examined these findings in regard to Middle Paleolithic artifact typology and the presence of notched and denticulated forms of stone tools. Archaeological contexts may contain forms that are the result of human activity, but trampling and natural processes can create similar-looking end products that could be geofacts.

Sediments and Soils

Figure 2.13 Primary and Secondary Structures Primary structures are created as a direct result of deposition. Secondary structures are created after the sediment has been deposited. Primary structures provide clues to the conditions present at the time

the particles stopped being transported. Secondary structures can be used to infer the processes or events that affected the sediment sometime after initial deposition.

It appears that for larger artifacts like cores, the properties of sphericity and roundedness play an important role in the amount of potential transport by water. Artifacts that are more spherical, and rounded clasts, such as hammer stones or chopping tools made on fairly spherical cobbles, are more susceptible to transport. They can be more easily moved from primary contexts because it takes lower flow velocities to start them going, and it is easier for them to keep moving. More angular objects do not begin to move as easily and are more easily stopped during transport.47

created during deposition (exogenic), after deposition within sediments (endogenic), by fauna and flora (biogenic), or during pedogenesis or lithification (fig. 2.13). Well-known examples of depositional stratification are layering or bedding. Depositional sedimentary structures can be the result of minor variations in conditions of transport and deposition. These variations result in differences in the size of the particles. In certain sedimentologic situations archaeologists will notice very thin beds, generally less than 1 cm thick. These thin beds reflect slight transportational or depositional variations. Laminae are the smallest, thinnest layers of original clastic deposition. They can have a range of thickness from > 30 to < 1 mm. Lamellae, in contrast, may be postdepositional structures. They can be features

Structure Sedimentary structures are usually small-scale variations in texture or composition. These are

53

Sediments and Soils

produced by translocation or movement of claysized particles within a developing soil profile. Iron oxide deposits can be also be translocated. This movement can produce small-scale banded structures, and it is associated with postdepositional alterations. Postdepositional structures include mud cracks, faults, and convoluted bedding. Mud cracks and other cracks result from sediments drying up. Frost wedges can also produce cracks in the earth as a result of freezing and drying. Stress within strata can lead to the development of faults. Some postdepositional deformational structures occur when the sediments are in a plastic, watersaturated state. Geologists have inferred that these structures reflect regional tectonic events or deformation associated with overlying deposition of sediments. Liquefaction features have been interpreted as indicators of earthquakes. Sediment intrusions, such as sandstone dikes, are caused by the injection of liquefied sand from below. When very dense strata lie over less dense strata, it causes other load structures. The pressure of heavy overlying deposits squeezes the underlying deposits and causes convolutions to form. Lighter sediments expand upward into heavier layers, which creates such deformation features as convolute laminations. Contorted sediments also can be associated with spring deposits. Vertical shafts or horizontal tunnels that appear as structures within sediments or soils may result from biotic activities. These kinds of structures are created by plant root systems or burrowing by large and small animals. The presence of these kinds of biogenic structures is an indication of land stabilization and high levels of moisture. Postdepositional processes also form pedogenetic structures in which soil particles bond together to make various types of aggregate clusters. These soil aggregates are called peds, while aggregates formed by disturbances like digging or plowing are termed clods. Clods are archaeosediments because they result from human activity. The soil aggregates (peds) have been classified into four types. Roughly horizontal peds arranged in a plane are termed platy. Natric B (Btn) horizons have relatively flat vertical surfaces. These vertical surfaces may develop as a result of wetting and dehydration. Blocklike structures that are 54

usually found in argillic B (Bt) horizons have particles aggregated in fairly equidimensional blocks. Spheroidal (granular and crumb) soil structures are associated with A horizons and, as a consequence, with organic matter and bioturbation.

Composition The composition of a deposit refers to the clastic, chemical, and organic constituents of a sediment or soil, as well as to fossil and artifact inclusions. Geologists select physical and chemical properties on the basis of their relation to presentday geologic processes to describe and interpret deposits. The proportion of clastics (sedimentary particles) to authigenic components (minerals formed after clastic deposition) can be used to distinguish different sedimentary depositional environments. Clastic particles derived from an external source area are exogenic. In contrast, endogenic materials are derived from material in solution or near the point of deposition. Figures 2.14, 2.15, and 2.16 illustrate how the proportion of clastics to carbonates can be used to infer changing depositional conditions. These compositional changes are from Pleistocene age basins associated with the Middle Paleolithic occupation of the Sahara in North Africa.48 The types of minerals precipitated from solution can be used to infer environmental settings associated with archaeological accumulations. The principal endogenic phase and dominant carbonate mineral in freshwater and brackish water lacustrine sediments is calcite (CaCO 3 ). The first minerals to form in closed lake basins undergoing evaporative concentration are alkaline earth carbonates. ‘‘Nonpedogenic’’ calcrete (caliche) can form by capillary rise of groundwater and loss by evaporation when the water table lies near the surface. Pedogenic calcrete can be formed by downward migration (illuvation). In early stages, caliche can form powdery or indurated isolated nodules of carbonate. With time, these can become extensive horizons of carbonate. Various stages in this progression have been used as indicators of age and climate. Organic matter in carbonaceous deposits reflects the past presence of plants or other biota. Peats, composed chiefly of organic material, form

Figure 2.14 Compositional and Size-Fraction Data Used to Infer Past Depositional Conditions. Changes in sediment particle size and composition can be used to infer changes in past environmental conditions. Here the proportion of carbonates to clastics composed of silica and the different sizes of the clastic particles indicate the presence of at least three lake cycles. The

peak for each lake cycle, or the time when the lake was at its largest, is indicated by high values of carbonate proportionate to the amount of siliclastics. These same intervals are also associated with higher amounts of smaller particles (silt and clay) in the clastic fraction. Larger, sand-sized particles were deposited when the lake was smaller. (Hill 1993c)

Figure 2.15 The Central Basin and Margin Sediments in Lakes Variation in the composition and size of sedimentary particles can be the result of different energy levels at different places during a particular time interval. The lower sedimentary sequence was near the edge of a lake basin. It is dominated by coarser particles. Higher amounts of carbonates and smaller particles were de-

posited when the lake expanded in size. The upper sedimentary sequence was closer to the center of the basin. Here the quieter, lower energy levels resulted in the deposition of high amounts of carbonate. Notice that there are three peaks of mud (silt and clay) deposition indicating three expansions or maxima associated with three lake cycles. (Hill 1993c)

Sediments and Soils

Figure 2.16 Inference of Lake Cycles from Changes in Relative Amounts of Constituent Materials The beginning, maximum, and ending of a depositional cycle can be inferred from the composition and size distribution of sediments in a stratigraphic sequence. Here the variations can be linked to the initial, early, maximum, and regressive stages of a single lake cycle. The basal sediments not associated with the later lake event consist entirely of clastic particles,

mostly medium or fine sands. During the early stage of the lake, small amounts of carbonate and increasing amounts of fine clastics were deposited in a small body of water. High amounts of carbonate indicate a major expansion of the lake, while two peaks of clay deposition indicate two pulses of increasing lake size during this interval. Coarse particles and a decline in carbonate content mark the end of the lake cycle and the decrease in the size of the lake. (Hill 1993c)

in freshwater environments, although they can also develop in salt marshes or slightly brackish waters. Sapropels, consisting of dark, organically rich sediments, form under more diverse conditions of organic decomposition in lake basins. Plant and organic matter are also introduced into sedimentary contexts as part of the detrital component. This can complicate the use of basin sedimentary sequences for developing paleoecologic and chronological interpretations. Pollen stratigraphies, as an example, may represent regional watershed environmental conditions instead of the local, site-specific setting (see Chapter 6). In other instances, ‘‘dead’’ carbon may be washed in; when this happens, radiocarbon measurements

produce dates that are too old for the archaeological record (see Chapter 5). Fossils and artifacts are also sedimentary components that can be regarded as either unique particles or inclusions in a depositional system. Regardless of whether they are in primary or secondary context, they can provide information about depositional contexts. The artifact or fossil taphonomy and sedimentology of these inclusions are critical aspects of archaeological interpretation (see Chapter 3). Both fossils and artifacts can be used as biostratigraphic markers as long as taphonomic agencies of redeposition are taken into account.

57

Sediments and Soils

Boundaries Sedimentary deposits, as well as archaeological features like walls and floors, are separated from one another by boundaries that can be of great interpretational value. Many terms are used to describe contacts between deposits, among them interfaces and discontinuities. Abrupt contacts in sediments can indicate changes in transportational depositional systems. Gradational contacts, in contrast, reflect more time-transgressive environmental changes. Undulating and broken boundaries indicate postdepositional, postburial deformations like the erosion of the upper surface. Discontinuities between sedimentary deposits can be either conformable or un- or non-conformable. Where there are bedding plains between sedimentary layers and no apparent break in deposition, the contact is conformable. These types of boundaries occur where there is continuous deposition or at least there have been no major erosional events. Conformable boundaries can be abrupt or gradational. Artifacts found along such contacts could either be redeposited sedimentary particles or the result of a past human occupation surface that became buried quickly during an interval of essentially continuous deposition. Nonconformable boundaries are contacts formed by an erosional surface or an interval of nondeposition (like a stabilized or a soil surface). They represent the passage of time between the deposition of the lower layer and the overlying bed. Artifacts found in these settings could have been mixed as a result of being redeposited or, if they seem to be in primary context, could represent many superimposed occupations on a stabilized surface. Unconformities include different forms of discontinuities. They represent an interval of nonsedimentation. Unconformities can be recognized in a stratigraphic sequence by several attributes. The presence of lag deposits may indicate an unconformable surface. The lag deposit may represent a set of artifacts resting on a deflated surface formed by wind erosion. Another indication of an unconformity is the presence of a buried soil or weathered zone. Changes in the dip of the bedding between sedimentary

58

units can also indicate an unconformity. Geologists are usually more explicit than archaeologists in recording and defining unconformities. Although recognizing time gaps is more difficult in archaeological excavation than in most geologic contexts, more attention should be paid to this question. These time gaps may be hints that can help explain time-dependent patterns in the archaeological record. Boundaries may represent time gaps. Where there is a boundary between sedimentary units that cuts across bedding planes, the contact is called a truncation or an erosional disconformity. Disconformities indicate erosional events that might be the result of deflation, running water, or shore processes. They are prevalent in some archaeological settings. A disconformity could result, for example, from the regression of a lake and subsequent subaerial erosion. Disconformities are good indicators that the artifacts associated with these types of contacts may have been eroded and redeposited and are not in primary context. The surface topography of boundaries is of particular help in interpretations of postdepositional processes like erosion, burial, deformation, and mixing. Smooth upper boundaries can indicate erosional truncations. The upper boundaries of depositional units can be buried, eroded, or stabilized. Where the upper surface is stable, soil-formation processes like weathering or the accumulation of organic matter can occur. Artifacts found along the interface between an upper boundary of one deposit and the lower boundary of the deposit above could have resulted from one of two situations. The artifacts could have been deposited on the surface of the lower deposit, perhaps because of a human occupation on that surface, or they could have been part of the lower deposit and eroded from it. Artifacts found in the lower part of the upper unit might have been redeposited from the lower layer. Many surface and near-surface processes can add to, alter, or remove components from sediments (see Chapter 3). Once the observable characteristics of the sediment-soil sequence containing artifacts has been described, ‘‘inferential units’’ can be proposed using Walther’s Law of Correlation of Fa-

Sediments and Soils

cies. Inferential units reflect our interpretations, and these interpretations transform observations about static physical and chemical properties and spatial relationships of deposits into a dynamic story of the past. This transformation of the prehistoric record from the observation of properties to the interpretation of past dynamics connects the study of the human past (archaeology) with the study of past life (paleontology) and the study of the evolution of the earth (geology).

Micromorphology Micromorphology deals with microscopic study of undisturbed soil and sediment.49 The technique involves the use of 30 μ-thick thin sections mounted on glass slides to be analyzed with a polarizing microscope (see Chapter 4). Often micromorphology can provide information not otherwise available, such as that pertaining to pedoturbation, leaching, humification, illuviation, eluviation, soil texture, diagenesis, distribution of micro-artifacts, microfacies analysis, and microstratigraphy. Micromorphology allows study of the three-dimensional context. Normal sediment/soil analyses separate the component parts, destroying the context. An example of the critical role that micromorphology can play is the study of the degradation in the Tell Leilan region of northeast Syria during the ‘‘late 3rd millennium abrupt climate change.’’ 50 At 2200 b.c.e. a marked increase in aridity and wind circulation, subsequent to a volcanic eruption, caused abandonment of Tell Leilan, regional desertification, and collapse of the Akkadian empire. Much of what we know about the early prehistory of the Mediterranean region comes from the study of the sedimentary records in rock shelters and caves. Recently, Jamie Woodward and Paul Goldberg have sought to determine the usefulness of these sedimentary records as archives of environmental change.51 It was determined that the site type (for example, an active karst setting) is important in the sensitivity of sedimentary record as a proxy of climatic change. The

climatic signal may be represented by distinctive suites of micromorphological features. Micromorphology has been used to examine the question of when people first began to inhabit North America.52 Meadowcroft Rock Shelter in Pennsylvania contains a stratigraphic sequence with radiocarbon dates extending from the Late Holocene to the Upper Pleistocene. The oldest radiocarbon ages are about 30,000 b.p. while some of the younger dates indicate the presence of artifacts predating the beginning of the Holocene. Some of the ages from Meadowcroft are controversial. It has been argued that they are possibly contaminated by old carbon, which would explain why artifacts are associated with pre–Clovis era radiocarbon ages. Goldberg and Trina Arpin used thin section samples of the sediments from the site to evaluate site formation processes and the proposal that groundwater had contaminated the site. Their micromorphological analyses were able to confirm that certain depositional and postdepositional processes influenced the stratigraphic record of the cave. Their studies also did not observe any indications of groundwater activity and concluded that there was ‘‘no micromorphological reason to reject the published dates.’’ In part to assess the vitality of micromorphology in defining activity areas (in this case, specifically ‘‘living floors’’), Goldberg and Ian Whitbread studied a Bedouin tent floor in the area of Tell Be’er Sheva, Israel.53 Within the tent complex, they sampled a hearth, a storage area for utensils, the sleeping area, and an abandoned animal pen 5 m from the tent area (which was used at the time as the tent area). The authors present numerous photomicrographs and binary images illustrating the microscopic details. The micromorphological analyses showed clear differences within and between soil thin sections of the activity areas. Another proponent of micromorphology in geoarchaeology is Charles French. Those interested in micromorphology and landscape evolution with a focus on examples from the British Isles should consult his recent book.54

59

CHAPTER 3

Initial Context and Site Formation

Out of more than a hundred flint implements which I obtained at St. Acheul, not a few had their edges more or less fractured or worn, either by use . . . before they were buried . . . or by being rolled in the river’s bed. —Charles Lyell 1863

The Creation of the Archaeological Record

I

t is necessary to study both the human activities and the natural processes that contribute to the formation of the archaeological record to understand it fully. One of the most powerful conceptual approaches in geoarchaeology is the application of artifact taphonomy. This is an interpretational perspective based on the study of formational processes that affect the final spatial pattern and compositional character of the archaeological record. As indicated in Chapter 2, because archaeological materials are recovered from sedimentary deposits, the geomorphic and sedimentologic processes that are associated with these deposits affect the interpretation of the artifacts within them. The initial landscape and resource settings play a critical role in human behavior. Where humans will live and what forms of human behavior will occur are in many circumstances a result of the landscape context. This context also plays a role in affecting how human behavioral patterns embodied in artifacts become the archaeological record. That is to say, the character and spatial distribution of particular activities reflected in the archaeological record are largely dependent upon the original habi-

60

tat setting. This landscape setting also influences the visibility and preservation of the record. As with any set of materials distributed across the earth’s surface, the evidence of human activities recorded in artifact patterning is subject to processes that can transform the original behavioral signal. Weathering, transportation, redeposition, and postdepositional alteration change the patterns imposed on the archaeological record by human behavior. In this chapter, we build on the principles of sediments and soils introduced in Chapter 2 to describe landscape contexts and processes in depositional systems that directly influence the nature of the archaeological record. Interpretations of the archaeological record based on a geoarchaeological site formation approach are founded on methodologies focused on artifact taphonomy. Originally applied to fossil biotic remains in the study of the fossil record, taphonomy involves the study of objects as they move through a trajectory from being part of a living, dynamic context to being a static accumulation or assemblage of materials (fig. 3.1).1 From a strictly paleontologic perspective, taphonomy involves studying the processes which change biotic communities into fossils. The usefulness of a parallel application of the taphonomic approach to the archaeological record should be clear. The principles of artifact taphonomy provide a framework for evaluating the events and processes that affect objects as they travel from the dynamic contexts associated with human behavior through the transforming events after they become part of the geologic context to the point at which they form the patterns that become the archaeological record.

Figure 3.1 The Effects of Human-Environmental Interactions on the Archaeological Record Evidence of human activity in the archaeological record is a result of biosphere and lithosphere interactions. The biosphere includes human behavioral or systemic contexts and the ecological habitats formed by communities of plants and animals. The physical habitat or landscape context is part of the lithosphere and is the result of both strictly geologic and biologic and atmospheric processes. Artifacts are deposited as a result of human behavior. After people leave, these patterns

resulting from human behavior can remain exposed or become buried. If left unburied, the patterns of behavior can remain unmodified, or they can be transformed by surface and pedogenic processes. Artifacts that are buried become part of the sedimentary context. Through erosion, these artifacts can become part of the surface context or be buried again, forcing increased modification of the original systemic or behavioral patterns recorded in artifactual patterning. Artifacts in sedimentary context can also be influenced by postburial alterations.

Initial Context and Site Formation

The development of a theoretical framework using what has been variously called a taphonomic, or transformational, or bridging theory approach to interpreting the archaeological record was discussed briefly in Chapter 1. This framework signaled a convergence between earth-science principles and methodology and the explanatory, process-oriented goals of anthropological archaeology. Behavioral interpretations of the artifactual record should rely on considerations of the transforming events that influence the initial patterns of variability produced by human activities before changing them to the variability observed in the archaeological context. An understanding of the sources of variability and patterning represents a critical contribution of earth sciences to archaeological interpretation. The orientation or bearing of artifactual remains can be used to infer the processes associated with the accumulation of materials at archaeological sites. For example, Larry Todd and George Frison used the direction of mammoth long bones to interpret the Colby site in northern Wyoming that contained the remains of seven mammoths associated with Clovis artifacts. The orientation of the long bones were plotted in the form of a rose diagram. The bones showed a nonrandom pattern: they were oriented roughly parallel to the direction of the drainage channel. A possible explanation of the nonrandom orientation of the mammoth long bones was stream action, but although some of the bones seem to have been rearranged by stream action, some of the larger bone accumulations did not appear to have been created by these types of processes.2

Stages of Site Formation A variety of terms are used to describe the stages in the trajectory of an object or feature as it moves from interaction with the dynamics of human behavior to become part of the archaeological record. In the most straightforward scenario, the trajectory can be separated into two stages. Human activity imposes patterns, and then physical (geologic and biologic) processes influence these patterns or impose additional patterns upon

62

the artifacts. Objects may have participated in both these stages more than once before they are studied as part of the archaeological record. In fact, the archaeological investigation itself is a return to participation in the ‘‘human activity’’ stage (see fig. 3.1). The initial stage of artifact patterning dealing with human activity has traditionally been the focus of anthropologically oriented archaeological interpretations, where the goal has been to infer past human behavior. This stage, where objects are acting as part of a human behavioral system, is called the ‘‘systemic context.’’ 3 The systemic context includes all the processes of human behavior or transformational activity that occur before the site is left unoccupied by people. A variety of behaviors can influence the artifact patterns that exist at the end of this stage, including artifact reuse and reclamation, artifact discard and site abandonment (human deposition), and disturbance (trampling, plowing, or excavation). These factors all result from direct human behavior and can greatly influence the character of the archaeological record. Artifact sets that reflect only processes associated with the systemic context, in which no rearrangement or restructuring by geologic and biologic forces took place, are primary archaeological deposits. In addition to past human activity, a variety of geologic and biologic processes introduce patterns into the archaeological accumulations. The stage where natural physical environmental processes alone affect artifacts is called the archaeological context. After the patterning imposed by human behavior has occurred and humans have abandoned the site, geologic and biologic processes can modify the artifactual record. These are transformations imposed either by surface processes or during or after burial. In contrast to primary archaeological deposits, secondary deposits contain artifacts modified by redepositional or diagenetic processes. Understanding the effect of these postoccupational physicaltransformation processes is critical to deriving inferences and meaning from the archaeological record. There are specific geomorphic controls on artifact size and distribution. A study to quantify

Initial Context and Site Formation

the effects of post-discard processes on Aboriginal stone artifacts in Australia demonstrated the influence of slope gradient, elevation, landform, and contemporary surface processes.4 The results indicate that, even at low gradients, artifact size and slope angle are significantly related. In this instance, it seemed that artifact movement by surface wash is unlikely to affect artifact distribution significantly. The composition and configuration of the archaeological record are greatly influenced by formation processes. Basic properties of individual artifacts like size, shape, spatial orientation, and surface features are all affected by formation processes. Postoccupational factors affect the quantity and diversity of artifactual materials preserved in artifact assemblages. Besides influencing whether relics of past human behavior will be preserved, postoccupational processes affect the visibility of the artifacts. Wherever archaeological remains are preserved, formation processes have directly influenced the spatial distributions (vertical and horizontal distributions and density) of artifacts. Interpretation of the patterns of variability within the archaeological record requires an understanding of both the behavioral and the physical conditions that influence the final character of the prehistoric record.

ous of these are locations related to the availability of a spatially restricted resource or activity context. For instance, procurement activities associated with stone and mineral sources result in the location of quarry sites near sources of raw materials. Similarly, campsites and butchery sites can be expected in settings connected with animal migration routes or near water-related resources. An understanding of the landscape context at the time of human occupation of a particular locality provides important information for determining what types of behavioral activities might have prevailed. In many instances site setting is strongly related to function and use. Whether considered in terms of regional environmental conditions, the physiographic setting immediately surrounding the site, or the local microenvironmental context, the landscape setting directly influences the location of a site and the activities that occur there. The depositional systems discussed below provide contexts for evaluating why humans have used particular settings. They also help us understand what processes might enable primary sites (systemic artifactual patterns) to be left intact, or at least be sealed by burial. Some conditions within these depositional settings are more conducive to the preservation of primary sites, while other circumstances are more closely associated with secondary redeposited assemblages.

Initial Landscapes and Original Occupation Sedimentary Contexts The present distribution and visibility of archaeological sites are based initially on the environmental conditions which existed at the time of human occupation. These conditions made a particular functional activity possible and provided the circumstances which later transformed or preserved the artifactual evidence of human behavior. Many functional activities that take place at a site are related to the landscape context or habitat setting and the resources available at that location. Certain types of activities are restricted to the availability or distribution of resources across a landscape. Specific types of archaeological manifestations are likely to be associated with these landscape patterns. Perhaps the most obvi-

Results of present-day depositional environments are used in comparison with those of past environments to infer past conditions. The attributes like color, texture, composition, micromorphology, and lateral and vertical associations described in Chapter 2 help identify specific sedimentary environments and depositional conditions associated with the presence of artifacts and other archaeological features. The major depositional systems that can be related to past human behavior are settings and processes which affect deserts, lakes (and other water-filled basins), flowing water, caves and rock shelters, glaciers, and sea coasts. In addition, artifacts in postdepositional alterations

63

Initial Context and Site Formation

affect these systems. Humans have been involved with these systems as they use both imbedded and mobile resources, and the remnants of these activities are preserved as part of the artifactual record. The sedimentary contexts discussed here are considered depositional and postdepositional settings in terms of their implications for archaeological site formation. The interpretation of archaeological materials associated with specific depositional settings relies mainly on an evaluation of the destructive and preservational processes prevalent in particular sedimentary contexts and how they have affected artifact accumulations. Where weathering and erosion dominate, the spatial arrangement and composition of the artifact accumulations at a site will be modified. Artifacts will be transported from their original contexts and redeposited. They may be abraded, broken, or destroyed. Destruction by weathering or rearrangement by erosion and redeposition removes some or most of the patterns imposed by the systemic context and introduces other patterns of geologic and biologic structuring. Where low-energy erosional and depositional contexts existed, a stronger possibility for retention of structural patterns directly related to past human activity, the ‘‘behavioral context,’’ also exists. However, even low-energy postburial forces significantly rearrange certain constituents of the archaeological record. As an example, in an overbank flood situation, alluvial sediments may bury larger fragments of pottery and stone, while flood waters may remove smaller artifacts and lighter bone. Every surficial environment is dominated by energy flows and the recycling of material. It is this recycling that leaves a physical and chemical record for the geoarchaeologist to study.

Desert Depositional Systems A diverse set of sedimentary deposits exists in arid or semiarid desert systems because at various times different processes operate, depending on the levels and sources of available moisture. What distinguishes desert systems from other systems is the near-absence of water, which directly affects the habitability of desert regions. Available moisture influences where and when human occupation will occur and restricts most human activity 64

to localized areas that contain water or other important resources. The timespan of human habitation in desert regions can be related to short-term or long-term changes in available moisture. Human occupation and use of arid and semiarid settings can reflect short-term availability of moisture, such as seasonal precipitation. This seasonal variation in available moisture creates habitats that sustain occupation for certain times of the year. Evidence of past human occupation of currently arid settings can also be related to long-term changes in available moisture caused by climate changes (see Chapter 6). Climatic fluctuations transform geographic landscape settings and make them conducive to habitation by plants and animals. Past human activity also has resulted in physical changes of the landscape that affected the availability of moisture and created arid or semiarid settings. The usual lack of water in arid regions alters the focus but not the kinds of sedimentary processes that prevail. The variety of depositional environments and landscape settings integrated into desert systems is reflected in eolian, fluvial, and chemical sedimentary processes. These processes respond to fluctuating moisture availability, causing the erosion, transport, deposition, and burial of sediments and artifacts. Eolian erosion and deposition are the dominant processes in desert systems. This is especially true during the periods of highest aridity. During intervals of increased moisture, fluvial, lacustrine, and pedogenic processes can occur within the desert setting (fig. 3.2). Human occupation of the desert depositional system during extremely arid intervals is usually sporadic, except in areas where surface water or other resources (like a specific stone raw material) are available. Most artifacts in desert settings were initially deposited because of occupation associated with nearby surface water or along routes associated with unique embedded resources. The Effects of Wind Wind can bury artifact arrangements associated with the behavioral occupational context of a site under a dune or other deposit and thus preserve it. But usually eolian processes cause major transformations of the archaeological record. This is

Initial Context and Site Formation

Figure 3.2 Desert System Site Formation In windblown deposits like dunes, artifacts can lie in secondary deflation lags, and smaller artifacts may be absent because they have been abraded away. Artifact accumulations may be associated around the edges of fluctuating desert-lake margins or in related deposits.

Where there is some physiographic relief, occupation surfaces may be buried under alluvial deposits or debris flows, forming paleosols. Younger alluvial fan deposits may incorporate older artifacts as a result of erosion and redeposition.

because of the destruction and abrasion of objects as well as intense erosion. Where stratified sequences of artifact-bearing sediments exist initially, erosion can remove the matrix, a process that permits the collapse of the sequences and the mixing of archaeological components. Smaller artifacts may be destroyed or eroded and redeposited by the wind. Because of the dynamic nature of dunes, the original context of an archaeological site can be severely modified. One example is a Folsom occupation with a buried soil within dunes in southwestern Wyoming, on the American Great Plains.5 The Folsom assemblage appears to have been vertically displaced as a result of postoccupational erosion and deflation. Desert pavement surfaces composed of larger particles (lag gravels) initially produced by deflation can form a protective cover and guard underlying artifact-bearing sediments from more erosion. Dune or sandsheet stabilization after burial may preserve behavioral contexts of occupation surfaces.

Dunes are the most commonly recognized deposit associated with arid settings. A variety of dune shapes can occur, depending on the size of the sediments, the strength and direction of the wind, and the surface on which the material is deposited. Sandsheets and sand seas also occur in arid settings. In addition to these depositional features, erosional features caused by deflation can remove finer materials and leave lag deposits, including artifacts. While finer particles are deflated away, deflation-lag concentrations of coarser particles can form on the surface. The Effects of Moisture During intervals of higher available moisture, streams and lakes can form part of the desert depositional system. Arid and semiarid settings are associated with alluvial fans in places where there is a major difference in the relative elevation on the landscape. These fans consist of cones of sediments that begin in uplands and reach lower elevations. Sometimes they coalesce with 65

Initial Context and Site Formation

other fans to form a bajada or a piedmont alluvial plain (fig. 3.2). Stream-related sedimentary processes dominate alluvial fans. In addition to stream channel deposits, which result in cut-andfill accumulations, sheet flood and debris flow deposits also form alluvial fans. Mass-wasting, the downslope movement of materials induced by gravity, is especially critical for fan development on slopes in arid and semiarid settings. The intermittent depositional events associated with fan development result in the burial of surfaces associated with human occupation. Localized precipitation in arid and semiarid regions may cause the formation of intermittent streams that produce wadis or arroyos. These processes can both erode artifact accumulations and redeposit them or bury and preserve them. Artifact assemblages may be transported and sorted during torrential rains or buried by flash-flood deposition. Desert lakes can seem either intermittent or permanent on the human timescale. Either way, they are focal points for humans and other biologic inhabitants of arid settings. Where there is a high degree of fluctuation in the level of the lake, occupation surfaces can be inundated and preserved or washed away by wave action. Intermittent lakes form playa deposits because of local precipitation, while more permanent lakes may develop where there is regional groundwater recharge. Short-term desert lakes form from surface water runoff or inflowing streams, because of either catastrophic storms or seepage of groundwater. Because current velocity is generally low in these closed basins, deposition occurs mainly from suspension. Human occupations around the margins of these restricted water bodies may be eroded by intermittent stream flow and sheet wash or buried by retransported sediment. When sand dunes are inundated because of groundwater seepage, interdunal ponds can be created. These are locations for the deposition of organics and fine-grained sediments. Desert System Site Formation Although eolian processes are likely to dominate desert systems, processes induced by climatic variation also have a significant influence in desert systems. Archaeological sites in arid settings are

66

commonly preserved around locations associated with the past presence of water. Deposits associated with these settings often contain records of the human occupation of areas that now experience desert conditions. Dune stabilization may accompany increased levels of moisture and, in semiarid conditions, may be associated with increased vegetation cover. Under conditions where there is no vegetation, any local rainfall will have major erosional and redepositional repercussions. Without vegetation cover, ephemeral stream and sheet wash can either erode and redeposit artifacts or bury them. Where seasonal or perennial bodies of water exist in desert systems, fluctuating water levels may erode sites situated along highenergy margins or bury them in quieter, lowenergy depositional regimes. A Middle Paleolithic locality in southern Egypt provides an example of desert system, streamrelated artifactual erosion, transport, and redeposition.6 The artifact assemblage, dated to 100,000–200,000 years ago, was on the edge of a shallow ephemeral stream channel in a setting ideal for stream action to have removed and redeposited artifacts. To determine whether horizontal downstream movement may have occurred, the frequencies and locations of large and small artifacts were plotted. Most of the larger artifacts, consisting of cores and retouched and unretouched flakes, as well as the smaller artifacts, consisting of flaking debris (debitage), were found in a dense upslope concentration. This suggested a degree of horizontal integrity for the artifacts. However, a distribution plot of the ratios of smaller and larger artifacts suggested that more of the smaller artifacts had been eroded and redeposited downslope. In this instance, the smaller artifacts seem to have been transported preferentially downstream during a flood within the channel. Although infrequent and sporadic in desert settings, rainfall can be a major factor in the modification of artifact accumulations. Lack of vegetation enables high-energy runoff to occur, and this in turn affects the erosion, sorting, and repatterning of artifact sets. The nearby Acheulian site at Bir Sahara provides another site formational context associated with desert conditions.7 Here, eolian processes

Initial Context and Site Formation

Figure 3.3 Hand Axe in Limestone Artifacts embedded in freshwater limestone (micrite). The limestones indicate the presence of freshwater ponds or lakes in the Eastern Sahara during the Pleistocene, probably before 300,000 years ago. Acheulian hand axes are found embedded within the limestones.

As the limestone erodes away, the Acheulian artifacts are exposed. The hand axes appear to have been deposited while the carbonate sediments were soft and unindurated; when the artifacts are removed the underlying limestone layer contains an impression of the artifact. (Photo by Christopher Hill)

have deflated sediments which once contained Lower Paleolithic artifacts dating to perhaps 600,000–500,000 years ago. The wind removed the sandy matrix and left the larger artifacts in lag position mantling the remnant of a fossil groundwater pond. Few smaller items were found on the deflated surface containing the Acheulian bifaces. The smaller artifacts seem to have been destroyed by the deflation. Excavations were conducted into the deposits not affected by deflation that remained at the site. In these undeflated sediments there was a much higher percentage of smaller artifacts than of larger Acheulian artifacts (hand axes) compared to the surface collection. In terms of archaeological interpretation, the presence of higher proportions of smaller artifacts in the undeflated deposits suggests that hand axes were being used and resharpened at the site. It also

implies that the site artifacts may still be in primary context. The Acheulian occupation seems to have been overlain by carbonates deposited in a groundwater-fed pond as it expanded in size. The carbonate protected the underlying sediments from erosion, whereas on the margins of the basin, where there was no protective carbonate, wind could easily erode the sandy deposits. Other Acheulian artifacts have been found in the nearby basin of Bir Tarfawi, where they are found actually embedded in remnants of limestone that have been subjected to erosion (fig. 3.3).

Alluvial Depositional Systems: Flowing Water Flowing water is a major force in landscape development and the creation of the habitat context of human prehistoric occupations.8 Sediments de-

67

Initial Context and Site Formation

Figure 3.4 Alluvial Fan Sediments deposited by running water formed this alluvial fan in the northern Rocky Mountains. This geomorphic feature is an example of aggradational or

‘‘filling’’ processes. Contexts such as this can contain buried archaeological landscapes. (Photo by Christopher Hill)

posited by a stream or running water are termed alluvium. Fluvial deposits are a specific type of alluvium that has been transported and deposited in a river system. It is no overstatement to say that every major river system in the world contains important archaeological sites. Not only do these settings frequently attract human habitation, but they also undergo sedimentologic processes that facilitate their initial burial and preservation, as well as their later erosion and exposure.

ing toward the distal segment of the fan. The dominant processes in fan development in humid settings are fluvially related. Distinguishing between alluvial fans in humid climates and braided stream deposits can be difficult, because they are both characterized by multiple channels and bar aggradation. Deposits associated with fluvial conditions vary from gravels transported during periods of high energy to muds deposited while suspended in lowenergy areas. Dried river beds, whose importance in arid settings was discussed earlier, make up a subcategory of alluvial settings. There are four main types of rivers: straight, braided, anastomosing, and meandering. Straight rivers are probably the least common. Braided streams form a network of branching and reuniting channels separated by islands and bars. Based on sedimentary texture and structure, a wide range of microdepositional conditions can be identified with

Depositional Contexts The major depositional contexts of the fluvial system include alluvial fans, stream deposits, and deltas. Fans can be formed in both arid and humid environments (fig. 3.4). In arid settings fans commonly occur along scarps. Fans found in more humid settings may form pediments and generally contain better-sorted deposits. There is a general trend in particle size distributions of fin-

68

Initial Context and Site Formation

Figure 3.5 Meandering Stream Environments Levees are created along the banks of river channels as a result of periodic flooding and depositional events. The river channel moves back and forth within the floodplain region, which is sandwiched between uplands. During large floods the river flows out of the

channel and over the floodplain, resulting in both erosion and burial of artifacts. Old, abandoned sections of the channel form oxbow lakes and backswamp areas that may be associated with functionally specific human prehistoric activity.

braided stream deposition. Gravels with cross bedding are characteristic of bars, channel lag, and channel fill deposits. Artifacts coarser than sand found in these deposits are potential articlasts, having been eroded from their systemic context and transported and redeposited with other nonartifactual clasts. Fine sands and finer fractions (silt and clay) are indicative of very lowenergy regimes like overbank flood plain deposits and standing pools. Artifacts in these settings may have been buried soon after human occupation and may be in primary context. Sedimentologic events in braided streams include flooding and channel aggradation with deposition under decreased velocity, development of bars through lateral accretion, and cut-and-fill episodes. Within an idealized fluvial system, meandering streams are situated between braided streams and deltas. A variety of micro-environments can exist within a meandering stream depositional system (fig. 3.5). Within the channel, coarser lag deposits form in the deeper part of the channel while on the inside (convex side) of the me-

ander loop, point bars can accumulate. Ridge and swale features that resemble lake strandlines can be present on larger point bars. Sediment can be deposited outside the channel in the form of overbank deposits after extreme discharge events. Oxbow lakes form where a meander is abandoned, usually after the river has created a new channel. In these cut-off channels, later deposition contributed directly from the fluvial system will be only low-energy overbank sedimentation. The banks of oxbow lakes are areas with a high potential for containing preserved human prehistoric occupations. The stratigraphic sequence within the lakes may contain such paleoecologic data as fossil pollen, phytoliths, and diatoms (see Chapter 6). Meandering stream conditions are commonly recognized by the textures and structures within a sedimentary sequence. They tend to be texturally upward-fining sequences: larger particles are in the lower section of the sequence and smaller particles are more prevalent in the upper part. Fluvial terraces are benches or shelflike land-

69

Initial Context and Site Formation

Figure 3.6 Terrace Types and Fill Fluvial terrace sequences provide critical information for the interpretation of the archaeological record. Where the highest surfaces are the oldest and the lowest terraces the youngest, they can be used as chronological tools. They can also be used to explain

the spatial distribution patterns of artifacts. Older artifacts will not appear on younger terrace surfaces unless they have been redeposited. Younger deposits may also bury older archaeological material. Stratigraphic excavations of stream valleys make it possible to understand the events that created the terrace sequence.

scape features along a river valley. They are formed by aggradation and degradation (figs. 3.6, 3.7). These settings have been the focus of archaeologists since the discoveries of prehistoric artifacts in France and England during the 1800s (see Chapter 1). Terrace systems can help us understand artifactual assemblages and geomorphicclimatic inferences. However, many traditional techniques of terrace correlation may conflict with detailed lithostratigraphic and chronological research. As can be seen in figures 3.6 and 3.7, a number of different depositional and erosional events can create terrace sequences of similar appearance. One of the longstanding landforms resulting from fluvial environments that is used by archaeologists is the alluvial terrace sequence.9 Alluvial terrace sequences are steplike platforms which mantle the sides of river valleys. In many instances parallel, same-elevation platform sur-

faces that appear to be the same age occur on each side of the valley. When these terrace surfaces are traced upstream they are generally higher topographically, whereas downstream they have lower altitudes. In the classic interpretation, the sets of tilted surfaces represent periods of stream deposition and erosion (incision). The times of stream deposition are ‘‘fill’’ intervals. During these times the valley is filled with sediments by so-called aggradational events. The aggradational fill creates a surface which can form a terrace surface. One way that terraces can form is by erosion of the stream valley fill. Erosion or incision into the fill during a ‘‘down cutting’’ episode results in a new, lower surface. Archaeologists have applied the concepts of stream cut-and-fill episodes to help determine the age of artifacts. Although it is not always the case, usually the highest terrace surfaces are the oldest; the lower and closer

70

Figure 3.7 Alluvial Terrace Development Stages and Artifact Deposition, Mixing There are five stages in the hypothetical development of an alluvial terrace system. These stages have important ramifications on archaeological integrity and visibility. At the outset, stage 1, people drop artifacts on both the upland surface and in areas of the floodplain between the upland and the stream channel. Artifacts in the gravels which form the older terrace system are redeposited. Recurrent deposition of floodplain muds bury human occupation surfaces. During stage 2, fluctuating higher energy levels cause the intermittent

deposition of coarser sediments and increased erosion and redeposition of artifacts. During stage 3, additional artifacts are incorporated into younger stream deposits as well as on the upland surface. There, artifacts are mixed in a way similar to stage 1 occupation. The stream channel cuts into all the previous artifactbearing deposits in stage 4, resulting in the erosion and removal of artifacts and their redeposition downstream as a mixed-age accumulation. In stage 5 people deposit artifacts on the new floodplain surface created by the later migration of the channel composed of eroded and redeposited artifacts.

Initial Context and Site Formation

Figure 3.8 Lower Yellowstone Landscape evolution and archaeological contexts. The ages and spatial relationships of geomorphic features and strata are used to evaluate the temporal and environmental contexts associated with ancient human activities. An understanding of these relationships can also be used to predict the location of archaeological resources. In this example from the Yellowstone River, a major tributary to the Missouri River in west-

ern North America, the chronometric ages are from tephra and radiocarbon samples. The highest gravels are dated by the presence of a 620,000-year-old (Middle Pleistocene) tephra. The lower gravels contain mammoth fossils dated to about 20,000 years ago. Upland loess settings are dated by mammoth remains and buried soils to the Late Pleistocene–Early Holocene (the last glacial interglacial transition) (Hill 2006).

to the present-day stream channel they are, the younger they are (fig. 3.8). Thus, artifact forms and features found exclusively on particular terrace surfaces have value as chronological markers. In addition, the presence or absence of certain artifact types may be explained as the result of the dynamics of terrace formation rather than of past human behavior. For instance, Acheulian artifacts dating to 500,000 years ago might be found on one of the older terrace surfaces but not on a terrace surface created around 30,000 years ago. The absence of Acheulian artifacts on the 30,000-yearold landform surface would not in this instance be the result of a behavioral pattern. Instead, it

would be due to the landform not being in existence at the time of the Acheulian occupation. In this way, so the model goes, younger lower terraces should have artifact forms that are absent on the higher older terraces. In the same way, in the Americas, we would not expect to see PaleoIndian artifacts on terrace surfaces formed later than about 8,000 years ago. Stream valleys can have no terraces or many terraces. There are a variety of ways for more than one terrace to occur. In perhaps the simplest scenario, a single fill or depositional episode can be followed by several down-cutting episodes. Each erosional cut that goes deeper into the fill creates

72

Initial Context and Site Formation

a lower, younger terrace. This progression is illustrated in the top row of figure 3.6, where several incision intervals into a single alluvial fill create first a single terrace (1B) and later a second terrace set (1C). Where there has been more than one episode of aggradation or fill within a fluvial setting, the story becomes somewhat more complicated, especially with regard to archaeological interpretation. There are two complications. First, certain older cut-and-fill episodes may be completely buried by younger fills. When this happens the surface manifestation of the archaeological record is biased toward younger occupations. Older artifactual records will appear to be absent simply because they cannot be observed unless excavations are undertaken. The second complication has to do with the relations between cutand-fill episodes and the number of terraces that can be observed without stratigraphic studies. Because older terraces can be completely buried by younger fill, a valley system can appear to have no terraces at all. On the other hand, sometimes there can be a direct relation between the age of the deposit, the terrace surface, and intervals of valley incision. Some of the ways two depositional events can create different numbers of terraces are shown in figure 3.6. If the earlier cut-and-fill episode is completely buried by a younger fill event no terraces may appear to exist, as in 2A. The same scenario with three alluvial fills is shown in 3A. In both instances the absence of older archaeological materials on the valley surface is explained by their burial by later deposition. In some instances the surfaces of the terraces are the result of two intervals of cut and fill, as shown in 2B. Here older artifacts will be found on the upper higher surface, and younger artifacts will be found associated with the lower terrace. (For three cut-andfill episodes, see 3C.) Sometimes terraces can exist as the result of successive cut-and-fill episodes. Where there are two fills, as shown in 2C, an erosional event after the deposition of an older fill may be completely buried by a single younger fill, which itself has been subjected to several succeeding intervals of incision without intervening fill episodes. For a case where three alluvial fills have

been subjected to a variety of erosional events, but only one terrace can be observed without stratigraphic excavations, see 3B. The artifactual remains of Pleistocene humans that once lived in eastern Asia are found in river sediments. Alluvial and colluvial deposits from the Transbaikal region of Russia contain late Upper Paleolithic deposits associated with two terraces within the Chikoi River Valley.10 The highest terrace in the valley is also the oldest. It contains depositional units composed of channel fill overlain by overbank flood deposits. The overbank sediments contain evidence of Upper Paleolithic humans. These people constructed dwellings on the floodplain. Low-energy floodwaters buried the Upper Paleolithic archaeological features and artifacts in fine sand and silt. Some of the river activity was strong enough to transport charcoal from hearths, but the energy level was not high enough to move larger stones the size of cobbles very far. This cycle of human habitation and flooding and burial occurred around 18,000 to 17,000 years b.p. based on radiocarbon measurements. At about 13,000 years b.p. the Chikoi River cut through these deposits, leaving them as the high terrace. Another fill episode, or aggradation, was begun. The sediments in this new fill episode consisted first of channel deposits and, sometime later, point bar and floodplain deposits. Evidence of people using Upper Paleolithic artifacts is also found in the floodplain sediments. Based on radiocarbon dates, the interval of alluvial deposition had ended by about 10,000 years b.p., and soon after the Chikoi River again downcut through these sediments, forming a second, lower terrace. Holocene colluvial deposits that accumulated on top of these two Late Pleistocene terraces with Upper Paleolithic artifacts contain buried Neolithic and Bronze Age artifacts. Anastomosing rivers commonly exist in deltaic or coastal settings. They resemble braided streams because they also consist of channel networks that separate and reconnect around portions of land; but unlike braided streams, anastomosing rivers have deep, narrow channels; long-lived, vegetated islands; and floodplains with backswamps. They

73

Initial Context and Site Formation

also tend to transport a suspended load of fine clastics. The presence of fine-grained sediments and vegetation retards the lateral stream migration, an action that typically leads to vertical aggradation. Deltas form where sediments are deposited from rivers into still water. Deltas are an alluvial subsetting that is commonly connected with basin-related depositional systems, especially lake and coastal systems. As clastics carried within the stream are moved into a basin, larger particles are deposited first, while smaller particles are carried closer to the center of the basin. Deltas commonly contain a gradation of depositional contexts, consisting of the delta plain, the delta front, and the prodelta. Settings on the delta plain include channels and levees, point bars, crevasse splays, and marshes. Waves and longshore currents dominate in the delta-front zone. Prodelta deposits consist of silt and clay deposited away from the shore. In North America, one of the most well-studied reconstructions of a delta setting is at the mouth of the Mississippi River. Over the past 5,500 years there have been seven discernible lobes of the Mississippi Delta.11 The fan of the Nile Delta is an important alluvial feature that has influenced human activity for thousands of years.12 Three lithofacies sequences exist from the past 35,000 years of sedimentation. The facies reflect changing transport processes associated with the delta. Strata from the lower sequence range in age from older than 35,000 years to about 12,000 years. They consist of alluvial sands that interfinger with overbank and playa muds. The muds were deposited in areas alongside the fluvial channels during high flood events. The top of this lower set of deposits contains an unconformity: the overlying unit consists of coarse sands and marine shells. It seems to be the product of erosion and redeposition of the older Pleistocene alluvium by waves in a highenergy, near-shore setting. These deposits represent a marine transgression dating from about 11,500 to 8,000 b.p. The overlying Holocene deposits include deltaic alluvial plain, delta-front, and prodelta sediments. These deposits help us understand the paleogeographic landscape context associated with the

74

human occupation of this area. From 35,000 to 18,000 b.p., the area seems to have been an alluvial plain with braided channels. As sea level began to rise (about 15,000 to 8,000 years ago), the high-energy shoreline moved inland, reworking the sandy alluvium that had been deposited previously. Archaeological sites that had been part of the braided channel setting would have been subject to erosion and redeposition. The modern Nile Delta began to form about 7,500 years ago. Evidence of human occupation is associated with the Predynastic, dating from about 7,000 to 5,000 years ago. At about 2,000 b.p., humans were actively changing the development of the Nile Delta. Human intervention maintained channel branches, while more intense irrigation and wetland drainage modified the delta. The construction of dams in the twentieth century has eliminated the annual cycle of Nile flooding, which deposited sediment into the delta. Thus, delta deposits can be indicators of former lake or sea levels, because they form at the intersection of the stream with the level of the lake or sea. Many archaeological sites are associated with deltas and with related abandoned shoreline features. Geomorphologists categorize the size of streams and rivers by using the concept of relative rank within a drainage network. This concept is known as stream order. Streams and rivers are ranked from first order (the smallest) to tenth order based on the number of tributaries they have. First-order streams have no tributaries. Second-order streams form by the joining of two first-order streams, and so on. In general, as the stream order and size increase, the area drained also increases. The Mississippi River is a tenthorder stream. Rolfe Mandel has related the location of Archaic Period sites in the Central Great Plains of the United States to stream order.13 He found that alluvial fills in large streams, equal to or greater than fourth order, span the Holocene, whereas nearly all of those from smaller streams, less than or equal to third order, are less than 4,000 years old. The time-space distribution of alluvial deposits in the Central Great Plains explains the paucity of Archaic sites identified in the region. Within large valleys, most sites are

Initial Context and Site Formation

deeply buried. Within small valleys, erosion during Early and Middle Holocene time probably removed Early and Middle Archaic sites. Thus, geomorphic analysis should precede surface archaeological surveys in such terrain. Site Formation in Alluvial Settings Both erosional and depositional processes occur in stream settings. These occurrences that could cause either dramatic modification of archaeological configurations or the preservation of patterns derived from human activity. In the past, humans often have lived on active floodplains and banks next to channels. Although the landscape settings and resources associated with alluvial settings are conducive to the initial presence of artifact accumulations, depositional conditions associated with these settings may result in geologic structuring of the artifact patterns. As alluded to by the quote from Lyell that begins this chapter, there is a good chance that artifacts found in channel stream-related deposits have been redeposited, so that many patterns derived from human behavior have been modified or destroyed. This frequently occurs in depositional settings where either vertical or lateral aggradation is common, as on floodplains or meander belts. We would expect occupation surfaces that could have been occupied by humans to have existed on point bars and stream banks but not in active channel areas, except in ephemeral (seasonal, for example) streams. Artifacts are likely to move downstream in channels and in floodplains near channels when they are removed from their setting as part of bank erosion. Occupations on floodplains next to stream channels can be buried by overbank floodplain deposits. During periods when erosion and deposition do not occur, there are stable surfaces on which occupations can develop. Excellent coherent stratigraphy can be created in contexts where stable surfaces were buried by overbank flood deposits. A set of micromorphological, granulometric, mineralogic, and chemical analyses performed on the sediments from the Kennewick Man skeletal remains and the stream bank adjacent to where the bones were found indicate that the human remains eroded out of Columbia River flood de-

posits. Stratigraphic and sedimentologic evidence support the hypothesis that the body had been interred rather than quickly buried in overbank flooding.14 The importance of understanding the varied sediment accumulation rates in large stratigraphically complex archaeological sites has been pointed out. Issues such as settlement shifts, activity areas, site abandonment and reoccupation, construction of features, and population fluctuations can be understood only if the precise time frames of the various depositional sequences are determined. Sedimentation rates often govern whether or not clear distinctions can be made between artifact assemblages.15 The Indian Creek site in the northern Rocky Mountains consists of twenty-eight stratified prehistoric occupations (zones containing artifacts) within a valley floodplain terrace setting. Excavations exposed 8.5 m of alluvium and colluvium dating back to about 13,000 b.p. 16 The basal layer consists of a coarse gravel bed of Late Pleistocene age. These gravels are overlain by a unit consisting primarily of fluvial sediments with some colluvium, debris flows, and a volcanic ash deposit dating to 11,125 b.p. The unit consists of a thinly bedded sequence of massive and laminated sands, sandy-silty clays, and gravels that seem to indicate mostly braided stream conditions. The presence of hematitic, limonitic, and manganese oxide staining and mottling implies that the unit was intermittently saturated with water. The unit also contains interbedded carbonaceous layers of up to 2 cm thick, some of which contain Paleo-Indian artifacts; but no clearly pedogenic horizons were observed. After about 8,340 b.p., artifact-bearing layers interbedded with alluvial fan and braided stream deposits, as well as organic horizons, were deposited. Middle and Late Holocene strata contain artifactual accumulations dated from 7,200 to 3,000 b.p. They also contain the Mount Mazama volcanic ash dated to around 6,900 b.p. River terrace systems have played an important role in both Old World and New World archaeological studies. For example, the fluvial terraces in northwest Europe have been used to study the Middle and Late Pleistocene archaeological record of the region.17 The first indication of tool-making hominids in the region,

75

Initial Context and Site Formation

such as at Abbeville in France, occur in sediments that were deposited during the Cromerian. The presence of the artifacts in these terrace deposits suggest a maximum age for the youngest Cromerian sediments of about 500,000 years ago. In Great Britain, in the Thames Valley, fluvial deposits containing artifacts made using the Levallois technique are correlated with oxygen isotope stages 8 and 7 (see Chapter 5). Based on the absence of artifacts in primary context within deposits of isotope substage 5e (the last interglacial) it appears that humans did not inhabit Britain during the late Middle Pleistocene and early Late Pleistocene. Evidence for Late Pleistocene migrations associated with Neanderthals and Mousterian artifact assemblages as well as younger, more anatomically modern humans and Upper Paleolithic assemblages can be found within the fluviatile sequences. These deposits are usually found beneath modern floodplains that formed after the incision episode during the last glacial. Also, the presence of Paleolithic artifacts within well-dated terrace sequences in northwest Europe provides an example of archaeological geology: it has been suggested that the Paleolithic artifacts can be used as a means of correlation within the Thames Valley. Another indication of the complexity of alluvial deposition contexts is presented in the Thames River Valley Pleistocene fluvial deposits and associated Paleolithic artifacts.18 The first hand ax ever recorded was found in gravels beneath Gray’s Inn in this region in 1690. Studies of the sedimentologic context of Paleolithic artifacts in the Thames Valley have been used to interpret the landscape during the intervals of prehistoric human occupation and to determine the degree of geologic destructuring that has occurred. Most of the artifacts were incorporated into sediments consisting of gravels that were deposited in a braided river setting under cold climatic conditions. Some sediments containing artifacts seem to have formed during more temperate interglacial climatic conditions. These include artifacts found in finer sediments in the so-called Silt Complex, or ‘‘brickearth,’’ and sometimes in sand. Most of the artifacts are in secondary con-

76

text. They show physical evidence of having been rolled. However, some of the artifacts can be rejoined. Accumulations of unsorted, unabraded artifacts also exist. These seem to imply conditions where transport must have been minimal or of low energy. In terms of artifact taphonomy, it cannot be assumed that the artifacts found in coarser clastics were initially dropped into the sediments from which they have been recovered. Especially under cold climatic conditions, these artifacts could have been incorporated into the sediment by channel migration, overbank floodplain deposition, or eolian or slope-wash deposits. The artifacts could have gone through many cycles of redeposition. It is possible that those found in finer sediments represent actual occupation surfaces. They may represent systemic context patterning that has been buried by younger sediments. However, various processes may have caused geologic restructuring. For example, low-velocity currents may have transported large clasts, including artifacts, on low-angle slopes. Compared to the abrasion that occurs in the collision of gravel-sized clasts in a river channel, the abrasion on these objects was not very extensive. The geologic context of Stone Age sites in northeast Africa illustrates another alluvial situation. In Wadi Kubbaniya, along an old tributary to the Nile River in Upper Egypt, a small, multicomponent Upper Paleolithic site shows the connection between fluvial deposition and the preservation of the archaeological record.19 At this site, stone artifacts, bone fragments, and charcoal were recovered in the upper part of a sand dune that dated to around 19,000 b.p. The dune sediments are interbedded with overbank floodplain silts. More floodplain silts overlie the dune deposits nearby and are themselves overlain by lake deposits dating to about 13,000 b.p. Holocene wash deposits cover the lacustrine sediments. A separate set of younger artifacts seems to have been deflated out of these overlying silts or from deposits that are contemporaneous with the lake sediments. The sedimentary dynamics associated with this stratigraphic series represents a classic example of the encroachment of a dune field onto a floodplain and the simultaneous deposition of

Initial Context and Site Formation

fluvially derived overbank silts. Eolian deposits containing Upper Paleolithic artifacts appear to have been sporadically covered by flood waters that deposited silts. The dunes later created a dam across the floodplain which forms a lake, as well as the deposition of marls and diatomites. One approach to archaeological site formation is the analysis of the spatial arrangement of artifacts to try to determine whether the location of the artifacts is the result of human behavior or whether other processes have transformed the original behavioral record. This has been especially useful in the study of Pleistocene archaeological sites. For example, Michael Petraglia and Richard Potts 20 undertook a study of the spatial distribution of artifacts from sites in Beds I and II at Olduvai Gorge, Tanzania. One of the sites was FLK-22, situated in the middle of Bed I and dating to about 1.8 million years ago. Other sites were from Bed II with dates of about 1.6 to 1.2 million years ago. Mary Leakey described the FLK-22 site as a ‘‘living floor.’’ It was situated in clay deposits indicative of a lake margin environment with no evidence of fluvial sedimentation or river channels. The site was overlain by volcanic ash. In this low depositional energy situation there were high numbers of small artifacts and only a few of the artifacts showed evidence of edge rounding caused by abrasion during transportation. The overall character of the assemblage indicated the positions of the artifacts were relatively undisturbed by flowing water. In contrast, sites from Bed II show more signs of transportation by water flow. At site FCW, for instance, the artifacts were found on clays indicative of low depositional energy environments. This could have led to the interpretation that the site was in primary context. However, the artifacts were found in two concentrations. Such a spatial arrangement could be the product of fluvial activity. The spatial analysis of the assemblage suggested that the artifacts were transported on the clay surface prior to burial. The sedimentological situation of another site in Bed II, termed HWKE-4, consisted of sands, pebbles, and cobbles. This seemed to indicate the presence of a high-energy depositional environment, implying that the artifact assemblage was in secondary context. Com-

pared to the rest of the sites studied from Olduvai, this assemblage contained the fewest number of small artifacts; this was an indication that water transport had occurred. Human-caused changes to the landscape can alter alluvial systems and influence social change. This has been demonstrated in Mesoamerica, where agriculture has affected the landscape context. Arthur Joyce and Raymond Mueller have shown how human habitation of the Rio Verde Valley in Oaxaca, Mexico, during the Late Formative (about 400 to 100 b.c.e.) increased flooding and alluviation, caused a change from meandering to braided stream patterns, and influenced future human populations by increasing the agricultural productivity of the lower valley.21 In the upper part of the valley, land clearance for farming and settlement caused accelerated runoff and erosion. These human-induced changes to the alluvial system of the upper drainage led to several changes in the downstream part of the system. Flooding and alluviation in the lower valley increased, and a meandering stream pattern changed to a braided channel pattern. Three types of sediments provide an indication of the shifts in alluvial conditions. Lateral accretion resulting from high-energy river-channel settings was inferred from the presence of coarse sands. Vertical accretion related to moderate-to-low energy overbank deposition was reflected by the deposition of mud-sized particles. Infilling of oxbow lakes by low-energy deposition resulted in deposits of organic clay. The stratigraphic and geomorphic relations between these sediment types made it possible to distinguish two Late Holocene alluvial patterns for the lower valley. The older one consisted of a meandering channel, while the younger channels were part of a braided stream pattern. The meandering channel was abandoned and became an oxbow lake, gradually filling with sediment. Associated archaeological finds date the abandonment of this earlier channel to about 2,500–2,200 b.p. The younger braided channels are associated with lateral accretion and channel migration. Changes caused by human activity in the upper part of the drainage made the lower drainage more productive for agriculture. During the meandering stage,

77

Initial Context and Site Formation

the floodplain was smaller, and floods that deposited agriculturally useful sediments were less frequent. Only sporadic settlement seems to have occurred before the change in river patterns (during the Early to Middle Formative), although alluviation in the lower valley may have buried or destroyed archaeological sites from this time. The fine clastics deposited on the floodplain as part of the braided stream pattern were well suited for agriculture. In addition, the oxbow lake that formed in the abandoned meander channels was a year-round water resource that attracted fish and birds. Several sites have been found around these infilling ponds. Studies of the alluvial stratigraphy of the Santa Cruz River in Arizona were conducted by Andrea Freeman to help assess the transition from hunting-gathering to agriculture-based societies. In this region of the American Southwest, studies of Middle and Late Holocene stream dynamics indicated that prehistoric people selected specific parts of the Santa Cruz floodplain to grow plants. The Altithermal drought during the middle Holocene appears to have led to low amounts of fluvial deposition. After the Altithermal, there was a period of net aggradation that began around 4,500 and continued until entrenchment of the river was initiated at about 2,000 b.p. 22 The presence of a stable floodplain during the period of the Early Agricultural Period (ca. 3,500–2,000 b.p.) provided landscape settings in the river valley that could be used by prehistoric people to grow plants. Thus, low-energy overbank and slackwater depositional contexts appear to have been more heavily populated compared to localities associated with active, high-energy sediment deposition. The alluvial stratigraphic record also is important in terms of the preservation of archaeological sites. For example, sites are more likely to be preserved in contexts where sediment storage is a dominant process.

Lakes and Associated Basin Settings Lakes are bodies of standing water within basins. Both ancient lake shoreline and basin deposits are areas with high potential for containing archaeological sites. Other archaeologically significant lakelike environments are spring-fed ponds,

78

swamps, marshes, and bogs. These are generally distinguished by water depth and the vegetation associated with it. Lakes are often deep enough to prohibit the growth of vegetation (except for subaqueous plants). Standing water and the presence of trees characterizes swamp settings, while marshes contain grass but no trees. Both are difficult to ascertain in the prehistoric record without the application of fossil indicators. Peat deposits characterize bogs, which may be quite visible in the stratigraphic record. All these settings contain resources humans have used in the past and are associated with archaeological sites. Distinguishing among these types of aqueous environments using only sedimentologic or pedogenic data can be difficult in the absence of fossils or other paleoecologic indicators. Basin Deposits Lake and other water basin deposits can be classified into two major groups, clastic (essentially exogenic) and chemical (endogenic and authigenic). Organic deposits can be important constituents of each. An idealized picture of sediment distribution in lakes corresponds to potential levels of hydraulic energy, in a scale that ranges from coarse particles around the perimeter in the beach zone, grading into sandy marly muds, and finally grading into muds or deposits high in carbonates (fig. 3.9). Organic (carbonaceous) sediments would be expected to occur along the edges of the basins in shallow water marshes, swamps, and bog settings or within ponds. Coarser particles tend to appear in the higher-energy regions around the margin of the basin while fine-grained sediments accumulate nearer the center of the basin. In addition to the changes in particle size of clastic materials, in saline lakes general vertical and lateral patterns may be associated with the deposition of chemical precipitates. Evaporation sequences begin with the deposition of carbonates (calcite, aragonite, dolomite), followed by gypsum, and finally by anhydrite-halite. Modern perennial saline lakes like the Great Salt Lake, the Dead Sea, Lake Magadi, and Lake Chad are generally restricted to climates that are currently dry. Some of these fluctuate greatly in size, based

Initial Context and Site Formation

Figure 3.9 Site Formation in Lake Basin Settings In terms of sediment deposition, higher energies along the edges of the basin result in the deposition of coarser clastics like sands along the beaches. Organics can be deposited in shallow water areas where energy levels are low enough to allow plant growth. Muds and carbonate (limestone and micrite) are deposited nearer the center of the basin. When water levels are low in the basin, human occupation can occur in locations that, with rising waters, can be buried by deeper water sediments. Rising waters can also result

in the erosion and redeposition of artifacts that were once part of lower shoreline margins. In the reverse instance, artifacts associated with lake margins formed during high-water levels in the basin will not be affected by shoreline erosion during intervals of reduced lake size. Artifacts on these higher, abandoned lake margins may be affected by other erosional events, but they may also be dated if their existence can be reliably associated with the presence of humans and the creation of the margin feature.

on seasonal and more long-term climate changes. Calcite is the dominant chemically deposited mineral in freshwater lakes. Evaporation need not cause carbonate precipitates in lakes. Saturation of the water by calcite can be caused by photosynthesis or temperature changes. Lake sediments can also be created by organisms, mostly invertebrates like pelecypods and gastropods (usually depositing the calcium carbonate mineral aragonite), diatoms (siliceous), oncolites (algal), or ostracods. The lower wave energy in lakes and the fluctuation of water level can produce sediments that are more mixed in size. Waves can cause the development of beaches. Waves that rework coarser sediments and move finer sediments offshore affect the shorelines of larger lakes. Freshwater lake and marsh facies are coarser and more heterogeneous near shores, and finer and more homogenous toward the center. Erosion, transport, and deposition of coarse material are mostly confined to the shallow zone of lakes near the shore, which

is also the area where human occupation and artifact deposition are likely to occur. Plant life can be abundant in shallow basins or at the edges of lakes and may be a resource used by humans. Harvesting these plants by humans may provide an explanation for the presence of discarded artifacts in settings that would have contained knee-deep water. Bioturbation is rarer in the deeper parts of lakes: clastic deposition in the center of lake basins usually comes from suspension. Deep-water areas would not tend to contain artifacts unless they had once been the location of basin margins that were later inundated. Artifacts found in deep-water settings may have been deposited on aerially exposed surfaces during a recession of the lake to a lower water level, only to be buried with the next expansion of the lake. Low-lying swamp or marshlike environments are associated with the presence of fossil plant-root structures. These environments contain resources that humans used. Besides holding the possibility of having pri-

79

Initial Context and Site Formation

mary and secondary artifact accumulations interstratified within deposits caused by the expansion and contraction of water bodies, the margins of basins can contain sediments and landforms that may be associated with human occupation. Features that indicate the presence of a former shoreline, such as beaches and deltas, are likely to contain archaeological sites. Because the position of lake-margin features can reflect the changing water levels within a basin, they can help date associated archaeological assemblages and explain archaeological site distribution and visibility. Sedimentologic conditions found in lake and associated environments of deposition (ponds, swamps, springs) are also found in other depositional settings. Depending on the level of groundwater and the amount of seasonal precipitation, playa and more permanent perennial lakes can exist in semiarid and arid desert conditions. Lakes also form in alluvial and glacial settings. Site Formation in Basin Settings Lake-related contexts have a variety of microdepositional settings that both protect and destroy archaeological sites. Low-energy settings associated with the flooding of the margins of water bodies bury and protect sites. Artifacts in higher-energy regimes along shorelines, river mouths, and spring conduits are susceptible to erosion and redeposition. Artifacts in spring contexts are particularly susceptible to redepositional processes, which can cause the mixing of artifacts from different periods, although these mixed-component assemblages may be preferentially preserved because of later diagenic processes. Abrasion of artifacts and size sorting is also common in spring contexts. The later deposition of tufa and other carbonates may serve as a protective seal over primary sites. The geologic conditions of spring formation will be summarized in Chapter 7. Lacustrine shoreline features have a high potential for archaeological occupation, but high energy levels may modify systemic contexts associated with these features. Fluctuating lake margins can either bury sites or erode them. Away from the high energy of spring conduits and shorelines, quieter depositional settings can pre-

80

serve behaviorally related contexts. The infilling of lakes, which creates bogs and swamps, may preserve archaeological localities. Where lake levels are lowered, successive occupations along the margin should become younger toward the center of the basin. The opposite distributional pattern and the potential for burial are associated with transgressive lake events. Spatial analyses of artifacts and bones have been used to infer the formation events at archaeological sites. Sometimes the interpretations show little regard for potential geologic or biotic processes that may have influenced the spatial patterning of the artifact assemblages, while in other instances there has been a deliberate attempt to look at both hominid behavior and geoenvironmental processes that may have contributed to the archaeological record.23 Studies of Middle Paleolithic sites from the Egyptian Sahara in northeast Africa, for example, seem to demonstrate that the original behavioral record had undergone transformation by geologic processes within basin settings prior to burial. At one site the assemblage consisted of over 50,000 stone artifacts embedded in sandy deposits that were interpreted as the beach and shore zone of a shallow lake that existed around 100,000 years ago. The assemblage contained three sets of artifacts based on size. The largest artifacts were cores, tools were intermediate in size, and the smallest set of artifacts consisted of debitage. The spatial distributions of these three sets of artifacts were compared to their relationship with the margins and central part of the ancient lake basin. The spatial patterns showed that the larger, heavier artifacts were in a different location than the smaller, lighter artifacts. The arrangement of artifacts was what might be expected if they were deposited by hominid behavior along the edge of a lake and then were subjected to wave action. The spatial pattern appeared to indicate some size sorting; the heavier artifacts did not seem to have been moved very far from their point of initial deposition, while the smaller artifacts appeared to have been eroded and redeposited closer to the center of the basin. At a nearby site, a Typical Mousterian lithic assemblage was found embedded with fine-grained

Initial Context and Site Formation

sediments of a shallow, perhaps seasonal lake that existed about 125,000 years ago. In this instance there was a concentration of small artifacts (debitage) separated from the location of most of the larger artifacts (flakes and tools). There are at least two potential explanations for the spatial distribution of the different-sized artifacts. The artifacts may represent a nearly intact living surface with the spatial pattern mostly representing the activities of hominids that used Middle Paleolithic tools or, as an alternative, the artifact distribution may represent geologic processes similar to those that produced the two concentrations at the Olduvai site of FWC. In both these cases examinations of the sedimentological context and the spatial distribution led to the conclusion that the Middle Paleolithic archaeological record could reflect a combination of hominid and geologic patterning. Indicators of redistribution of artifacts by the movement of water include: depositional environments, spatial arrangements of artifacts, artifact size distributions, and indications of abrasion on artifacts.

Cave and Rock Shelter Depositional Systems Caves have been a critical source of archaeological information in both the Old and New Worlds. In the Old World, European cave stratigraphies played an early role in the demonstration of human antiquity, as did similar settings in Asia and Africa. Some candidates for the oldest human occupation site in the Americas are contained in caves or rock shelters; some of these also contain artifacts in an excellent state of preservation. Perhaps the best-known cave site in Great Britain is Kent’s Cavern, but there are other caves or cave groups that contain evidence that human habitation started in the Paleolithic.24 Elsewhere in Europe, the best-known caves are in France and Spain—the sites of Lascaux and Altamira, for example. In Australia the best-known sites are in the caves of Devil’s Lair and Kenniff Cave. There are many extraordinary cave sites in Africa, including Sterkfontein and Klassies River Mouth in the south and the Haua Fteah in the north. Important archaeological cave sites in southwest Asia include Mount Carmel, Shanidar, and the Dravid-

ian caves. Another cave with a long archaeological record is Zhoukoudian in China. In northern Asia, sites like the Ust-Kanskaiya Cave in Siberia provide evidence for comparison with the earliest sites in the Americas. Cave settings have been critical to archaeological studies in North America. The extensive study by C. Vance Haynes of the limestone cave of Sandia Cave in New Mexico is a classic, as is the research from Russell Cave in Alabama and Modoc and Graham Caves in Illinois and Missouri. Caves in arid settings, such as Danger Cave in Utah, Leonard Shelter and Gypsum Caves in Nevada, and Ventana Cave in Arizona, have preserved many organic remains. The Meadowcroft Rock Shelter in Pennsylvania contains controversial evidence of human occupation as early as 16,000 b.p. 25 Stratigraphic sequences in caves were an early focus of prehistorians and archaeologists. This was especially true in Europe, where stratified cave and rock-shelter deposits provided objects used by humans in association with extinct animal remains, as well as evidence for the change and development of artifact assemblages and fossils through time.26 Most commonly formed in limestone, caves and rock shelters are significant depositional systems for archaeological studies because they are conducive to burial and potential preservation. They are a critical source of chronostratigraphic and artifactual data (figs. 3.10 and 3.11). Although artifact accumulations may be preserved, taphonomic histories of these types of deposits are complicated. As in other depositional settings, many sedimentary processes are involved in creating the materials that fill caves and rock shelters. Where caves and rock shelters exist, both clastic and chemical deposition occurs. In addition to debris deposited by human or animal occupation, clastic materials include washes, fluvial deposition, rockfall, and eolian deposition. The most common chemical precipitates are various forms of calcite deposition, such as flowstone or dripstone. Limestone Caves In limestone caves, travertine or dripstone in features like stalactites and stalagmites are pro-

81

Initial Context and Site Formation

Figure 3.10 Cross Section of a Typical Limestone Cave This limestone cave is situated near a small stream system with artifact-bearing alluvial sediments and paleosols. On the alluvial and colluvial scarp leading to the cave entrance is a slope composed of rockfall and talus that has buried a site at the location of an earlier cave entrance. Some of the roof fall near the cave entrance also overlies deposits that continue into the cave. The

cave fill consists of several types of sediments. Large rocks are the result of bedrock having fractured and fallen on the surface of the cave. Buried soils and occupation surfaces contain artifacts, while other artifactbearing deposits are in more disturbed contexts. Several flowstone layers composed of precipitated calcium carbonate seal the different layers and provide one means of dating the stratigraphic sequence.

duced when calcite precipitates from permeating water (see fig. 3.10). These deposits have been used extensively for the construction of artifact-related chronologies and paleoclimate. A major step in cave and rock-shelter development is solution by groundwater. Groundwater creates caves primarily by dissolving the carbonate of the limestone bedrock. After solution, removal, and transport, carbonate may be redeposited as travertines and tufas, if evaporation saturates the solution. In southern Africa a series of caves including Sterkfontein, Taung, Swartkrans, and Kromdraai were formed of dolomite (CaMg[CO 3]2). Surface water widened the joints in the bedrock, providing locations for fossil and archaeological debris to accumulate. In some instances travertines formed. The remains of australopithecines have been recovered from these caves. The Haua Fteah site in northern Africa is also in limestone. Here, the stratigraphy consists mostly of roof fall and sediments washed into the cave. Some of the most spectacular archaeological

features ever discovered were found in the deposits and on the walls of caves in Europe. The rock shelters in the Périgord of France are large cavities in the faces of limestone bedrock situated within river valleys (see fig. 3.11). Both chemical solution by water and physical, mechanical weathering have affected these settings. Karstic dissolution by water and frost weathering with resulting rock shatter are major contributors to the formation of limestone caves and rock shelters. Talus slopes may form outside these features. At the site of La Colombière in southern France, sedimentary and artifact-bearing deposits are stratified in limestone bedrock and lie adjacent to the terrace system of the Ain River. Upper Paleolithic (Upper Périgordian) artifacts were recovered in a sandy, angular matrix that was overlain by large rockfall deposits. The rockfall episodes are stratified within clastic sediments and cemented zones. The Mousterian and Aurignacian artifactbearing deposits at El Castillo Cave in Cantabrian Spain are also in a karstic setting (fig. 3.12).

82

Initial Context and Site Formation

Figure 3.11 Evolution of a Rock Shelter Five steps in the hypothetical development of a limestone rock shelter.

El Castillo was one of several cave settings extensively studied by Karl Butzer, who separated the archaeosedimentary column into nineteen units. Stalagmitic flowstone covers the ceiling of the limestone cave and also the present-day surface within the cave. The sedimentary sequence also contains stalagmitic flowstones interstratified with occupational levels and silts and clays. Larger limestone rock fragments also form part of the sequence.27 In Britain, Kent’s Cavern is developed from limestone and contains Middle and Upper Paleolithic artifacts that indicate occupation from about 100,000 to 13,000 b.p. At Kendricks Cave in Wales, a solution cave, faunal materials and artifacts occur in a stratigraphic context. Caves in South Wales contain archaeological sites from both the Stone and Bronze ages. The limestone cave of Drachenloch in Switzerland is famous for its bear remains that appear to be arranged in a ritual manner. There are limestone caves with archaeological remains throughout the world. In Belize, caves

with solution features are associated with Mayan occupations. In Petén, Guatemala, and in the Yucatan Peninsula of southern Mexico, caves associated with limestone bedrock, karst hydrologic settings, and solution features also contain archaeological materials. In highland Guatemala, J. Brady and G. Veni have discovered the existence of pseudo-karst caves that seem to have been ritual activity centers for the Maya.28 These caves, carved out of the bedrock, were formed by human activity; they are not the result of water dissolving away bedrock. One of the best-known caves, La Lagunita, was cut into Quaternary sediments high in volcanic ash. It is situated under the central stairway of a major pyramid and underlies the pyramid’s central plaza. La Lagunita appears to date to the Protoclassic–Early Classic transition, around a.d. 360–400. Other caves, such as those at the Contact Period site of Utatlán (the capital of the Quiche Maya empire), are cut into the sides of mesas. The pattern of the caves underneath the central ceremonial complex has led scientists to believe that they had sacred and ritual signifi83

Initial Context and Site Formation

Figure 3.12 Stratification and Natural Interbedding at El Castillo Cave in Cantabrian Spain The cave is formed in limestone bedrock. Large chunks of the bedrock that once formed the roof of the shelter have fallen and been incorporated into the cave’s stratigraphic sequence. Artifact-bearing deposits denote prehistoric human occupation and use of the

shelter. Prehistoric human use is also reflected in the presence of layers interpreted as hearths. Interlayered between the artifact- and hearth-bearing deposits are sediments that are predominantly composed of muds (clays and silts), as well as numerous layers of stalagmitic flowstone composed of precipitated calcium carbonate. (Based on Bischoff et al. 1992)

cance. Caves appear to have been sites of ritual significance from the Ice Age, with its cave art and bear cults, to the Late Holocene, and the rites of socially complex groups like the Maya. Karst solution features in limestone bedrock have been reported at the Hearth and Mordor caves in Australia, where evidence of human occupation includes artifacts and cave paintings. Fraser Cave in southwest Tasmania, a cave associated with karst and solution features, contains archaeological materials. In China, some caves contain a long record of hominid occupation that provides critical records of long-term development of Paleolithic behavior. Other caves, such as Longgu Cave in northern China, contain Upper Pleistocene rock and cave art. In the Chuandong area (Guizhou Province) of southwestern China, caves contain stratified deposits with a variety of fossil fauna and artifacts. The most important means of sedimentation in limestone caves and rock shelters include freeze-

thaw phenomena, solifluction (see below), the breaking off and collapse of large blocks, stream and sheet wash, eolian deposits, carbonate deposition, and biotic inputs. Karstic limestone caves are widespread and consist for the most part of deposits resulting either from weathered material transported into the cave or physical weathering of the cave walls. One of the most extensively studied karstic cave depositional sequences is in the Peking Man Cave in the Zhoukoudian area southwest of Beijing.29 It is a large, vertical cave formed out of Ordovician limestone and filled with deposits separated into seventeen lithologic units. There seem to have been two main stages of filling. The basal layers are fluvial deposits which filled the cave after karstification processes changed from solution to fluvial deposition. Limestone, breccia, stalactitic, and travertine lenses form the upper part of the sequence. Most of the deposits are Middle Pleistocene in age, and the sequence contains faunal

84

Initial Context and Site Formation

remains (including those of Homo erectus, stone artifacts, and ash lenses). The occupational layers are interbedded with sand and clay deposits that were washed into the cave; with roof fall from periodic collapse of bedrock; and with carbonates formed from crystallization in lime-rich waters. Accumulations in the cave probably span at least a 500,000-year period, ending around 250,000– 200,000 years ago. Sandstone Caves and Rock Shelters One of the most extensive archaeological investigations of sandstone (as distinct from limestone) rock shelters took place at the Meadowcroft Site in southwestern Pennsylvania.30 The well-defined stratigraphy contains artifact-bearing deposits indicating possible human presence in North America by perhaps 16,000 b.p. Jack Donahue and James Adovasio have outlined the development of these types of settings on the basis of the Meadowcroft Rock Shelter and other eastern North American sandstone contexts: the sandstone shelters generally occur along the slopes of relatively young valleys, and the development of the rock shelters is related to less resistant bedrock types interbedded with the sandstone, which leaves sandstone overhangs.31 Mechanisms of sedimentation within sandstone rock shelters incorporate four major processes, which can be recognized by grain-size distributions and sedimentary structures. These are rockfall, attrition, sheetwash, and flooding. In these contexts, large fragments of bedrock released by the development of joints, freeze-thaw conditions, and possibly biotic activity like root action indicate rockfall (including slab failure and rock avalanche). Attrition and granular disintegration also contribute to the stratigraphic sequence in sandstone rock shelters and may be associated with both physical and chemical weathering. Sheetwash, or slopewash, will contribute poorly sorted size fractions, while floodplain deposits dominated by the finer sizes may come from nearby streams. Igneous Rock Caves In addition to the more typical limestone and sandstone caves and rock shelters, there are other

forms of caves used by humans that contain artifacts. In western Iceland, caves formed as lava tubes associated with the Hallmundarhraun lava flow contain faunal materials and artifacts. Caves and shelters in Hawaii formed from such volcanic features as lava tubes have indications that they were used by humans: they contain stratigraphic sequences that include artifacts. Caves and tunnels associated with volcanic features have also been the subject of archaeological study near Teotihuacán, Mexico, and similar contexts have been studied on the Snake River Plain of Idaho, in western North America.32 Site Formation in Caves Site formation processes within caves and rock shelters are eclectic because of the types of bedrock or depositional environments that produce variable depositional, erosional, and occupational events.33 Differential erosion of bedrock can form caves in areas of high energy, such as those along coastal shorelines and along the margins of stream valleys. Caves at the Sidi Abderrahman quarry along the Atlantic coast of Morocco provide an example of archaeological site formation in a coastal setting. The sedimentary sequence in the cave contains Acheulian (Lower Paleolithic) artifacts as well as hominid fossil remains. During a time when the sea level was low, Acheulian artifacts were deposited on the beach. As the sea level continued to drop, dunes encroached along the exposed coastal margin and buried the artifacts on the beach. This sediment became consolidated and formed sandstone. A later rise in sea level eroded the sandstone and formed a cliff wall that contained caves. Beach deposits inside these caves show that about 300,000 years ago the sea level was 27–30 m higher than present sea level. After the beach sediments were deposited in the eroded sandstone cave, hominids lived in the caves.34 Gorham’s Cave, cut into the basal cliffs of the Rock of Gibraltar, is another example of a cave that was directly influenced by coastal formation processes. Artifacts in the cave show that Middle and Upper Paleolithic human populations lived in the cave during the past 100,000 years. A sea-level rise resulting from the melting of large continen-

85

Initial Context and Site Formation

tal glaciers during the last interglacial may have scoured the cave and eroded sediments containing older artifacts. The deposits in the cave display clear connections with the coastal setting.35 Mummy Cave, in the Absaroka Mountains of northwest Wyoming, provides another example of a cave created by high-energy erosion. The cave is an alcove cut into a cliff made of volcanic (tuff-breccia) bedrock on the outside bend of a river. A series of thirty artifact zones covering the past 9,000 years of human prehistory were found in the cave. The cave was formed by lateral cutting of the bedrock through stream erosion along the outside of a meander bend. After the cave was cut, the river shifted its course. More than 12 m of fill, consisting of redeposited volcanic bedrock from outside the cave, accumulated in its interior over the past 10,000 years.36 Many caves containing archaeological materials develop as the result of interactions between bedrock and groundwater, especially in limestone-karst landscapes. The groundwater dissolves the limestone by slow percolation of precipitation through the bedrock or by dissolution along the interface of bedrock with the water table. Sediments within caves are the result of formation processes that produce two categories of deposits: sediments consisting of materials from outside the cave that find their way into the cave and materials that are formed within the cave and deposited there. Areas near the mouths of rock shelters and caves generally have more complicated formation histories than those within the interiors of caves. Sedimentary events connected with the region outside the cave are generally represented in the deposits around the mouth of the cave, while deposits within the cave usually reflect events taking place inside the cave.37 Cave sediments are often subjected to severe physical and chemical alterations that make it difficult to interpret their archaeology. Using micromorphology and studies of the phosphatic minerals that formed in the sediments after deposition it was determined that each depositional unit in Theoptra Cave (Greece) obtained its diagenetic fingerprint fairly soon after burial.38 The prominent ash layers in the older sediments were subjected to severe diagenetic alteration with

86

most of the relatively stable siliceous components of the ash decomposed into amorphous silica. A critical aspect of archaeological stratigraphy is that of lacuna in the depositional record. Geologists, with their grosser scale and fairly sophisticated principles of rock genesis, usually can assess what is missing. This is often not the case in geoarchaeology. Bill Farrand has provided a detailed assessment of six or seven identifiable gaps in a 25,000-year time span in Franchthi Cave, Greece, and presented a framework for approaching such problems.39 Some of the gaps unfortunately coincided with times of significant cultural change. This problem demands caution on the part of archaeologists and geoarchaeologists, especially the former as they attempt an overview of a site or region.

The Glacial System Glaciers have also affected the location of human occupation. Pleistocene glaciers covered large areas of land, where they both eroded and deposited sediments. The growth and melting of continental and mountain glaciers, and the related environmental changes that occurred with these fluctuations, influenced the habitability of certain areas. Indirect glacial-related settings include those formed by both proglacial and periglacial conditions. Each of these settings shows particular sedimentologic regimes and geomorphic contexts that can be related to prehistoric human occupation (fig. 3.13). Here we focus on a short description of glacial settings associated with prehistoric human activities and the associated erosional and depositional processes. Glaciers are moving accumulations of ice which erode and scour the earth’s surface, transporting and depositing sediments and, in some instances, artifacts. The three major types of glaciers are those formed in mountains, those that spread from valleys and flow along the front of mountains, and ice sheets that cover large portions of continents. These moving accumulations of ice result in the formation of erosional and depositional features. Rocks incorporated into bases of glaciers erode the earth’s surface, and the eroded and transported debris is deposited as till. Unsorted, unstratified till can be deposited directly

Initial Context and Site Formation

by glaciers, while glacial meltwater can generate stratified glacial-fluvial deposits. Till deposited along the edges of glaciers creates moraines. Along glacier edges, melting water is associated with outwash-related braided streams, delta deposits, and ice-margin (proglacial) lakes (see fig. 3.14). These sedimentologic contexts were the locations of prehistoric human occupation. Glacial tills may be found to overlie, underlie, or incorporate artifact accumulations. Glacial conditions can destroy, modify, or preserve evidence of human occupation. In the southwestern part of the North American Great Lakes region, for example, there is a stratigraphic sequence consisting of till, proglacial lake sediments, and nonglacial deposits. One till is overlain by laminated clays interbedded with silts and sands of an ice-margin lake. A zone containing the remains of the Two Creeks boreal forest (dated from about 12,000 to 10,500 b.p.) has developed on this lake bed (fig. 3.15). The forest bed is a detrital zone containing organic matter, with plant debris, including logs and stumps, in the parent deposit. Another lake deposit overlies the Two Creeks bed. The lake appears to have flooded the forest before being overrun by another glacial advance that deposited another till. The till has incorporated material from the forest and is itself covered by more lake sands and eolian dune deposits. It is possible that artifacts lie within the deposits that have incorporated remnants of the Two Creeks forest; human occupation of the North American continent had begun by at least 11,000 b.p. In Europe, where there is strong evidence of Middle and Late Pleistocene human occupation, there are instances of both the burial and the incorporation of artifact-bearing deposits into glacial sediments. One early account of a buried forest bed can be found in Charles Lyell’s Antiquity of Man. The Cromerian Forest Bed in England contains the stumps of trees and fossils of mammoths and is overlain by glacial till.40 Glaciers pushed across and deposited till on deposits containing artifacts.41 In Britain, there are artifact-bearing sediments overlain by tills. At Elveden the lowest deposit overlying chalk bedrock consists of a till (‘‘boulder clay’’). The upper

Figure 3.13 Glacial and Postglacial Landscapes In the top drawing, glacial ice is shown advancing. Debris eroded and transported by the moving glacier can contain artifacts. Along the edges of the melting ice sheet, sediments can be deposited, to form outwash plains or delta deposits associated with proglacial lakes. The margins of the glacial lakes are potential locations for human occupations. After glaciation, as shown in the bottom drawing, several geomorphic features may be present that can be used to infer the past presence and character of a glacial event. Large mounds of sediments may be left along the edge of the glacier, forming terminal and recessional moraines. Whale-shaped oblong features within the area where the glacier existed, called drumlins, indicate the direction of ice movement. The sediments that fill kettle lakes formed in the glaciated area can help us interpret the timing of deglaciation and the postglacial environment. This helps us understand when the region was available for human occupation and the ecological context of prehistoric adaptation. Archaeological occurrences may also be associated with the abandoned beaches of former proglacial lakes.

87

Initial Context and Site Formation

Figure 3.14 Proglacial Lake Stratigraphic and chronologic relationships of sediments and geomorphic features. Relative ages are determined by the spatial relationships of geomorphic features and the deposits that form them. The geomorphic and sedimentologic contexts reflect various past physical environments. In this case, from the northern Rocky Mountains and Great Plains of western North

America, fluvial gravels are dated by U-series, sediments associated with proglacial lake setting are dated by luminescence, while post-glacial mammoth remains are associated with radiocarbon-dated pond deposits and tephras. The mammoth remains are associated with landscapes that date to the about the time when Clovis artifacts were used (Hill 2006).

surface of this till is an eroded surface overlain by a marl. Over this marl lies a series of clastic units that contain artifacts, including Acheulian hand axes. These clastics represent an old lake or small channel, and the artifacts seem to come from a hominid occupation along the edge of a water body. The top of the clastic series containing the artifacts appears to represent a lake setting and consists of calcareous sand. The upper section shows signs of cryoturbation, with deformation structures similar to those documented by Lyell in the deposits overlying in the Cromerian Forest Bed. Another till overlies this and contains a superimposed soil in its upper section. Here the artifacts seem not to have been affected by later periglacial conditions or deformed by ice load.

Where glaciers have overridden landscapes that were once occupied by prehistoric humans and the artifacts have become incorporated into glacial sediments, archaeological interpretation becomes more complicated. As with the Two Creeks stratigraphic sequence, older materials have been incorporated into younger glacial tills. Under these conditions, artifacts assigned to an earlier time are recovered in deposits of a younger glacial advance. The British site of High Lodge provides an example of how to apply geologic interpretations of glacial deposits in the interpretation of Old Stone Age artifact accumulations.42 For more than a hundred years, the site of High Lodge has been the subject of much discussion because of the presence of Middle Paleo-

88

Initial Context and Site Formation

Figure 3.15 Idealized Late Quaternary Sequence within Two Creeks Bed, Wisconsin The lowest sedimentary deposit is related to an earlier glacial event. It is covered by a unit largely composed of clays deposited in a lake. The upper part of this lake clay unit contains the roots of plants that grew in the Two Creeks forest. The Two Creeks forest bed over-

lies the lake clay unit and is buried by sands and muds deposited in a younger lake. Glacial deposits above these lake sediments contain fragments of plants that are in a redeposited position. Artifacts found in this deposit would probably be associated with a previous time period, such as when the Two Creeks forest was in existence. (Based on Nilsson 1983)

lithic (Mousterian) artifacts recovered from finegrained deposits overlying till. These artifacts were found under deposits containing what were thought to be older Acheulian hand axes. Here, both the Acheulian and the later Middle Paleolithic artifacts had been subjected to secondary

disturbance, but of different kinds. Apparently the High Lodge Middle Paleolithic artifacts lay in sediments disturbed by the advance of an ice sheet. This ice advance transported large blocks of sediments, still containing artifacts, intact to the top of a younger deposit. Although the arti-

89

Initial Context and Site Formation

facts were fresh, and pieces could be refitted to one another, they were in a disturbed context because they had been transported along with the matrix that enclosed them. The Acheulian hand axe assemblages were also in secondary position. Originally they had been deposited as part of a Lower Paleolithic habitation before the Middle Paleolithic artifacts. After the Middle Paleolithic artifacts had been transported along with their enclosing matrix, the older Acheulian artifacts had been incorporated into glaciofluvial and debris flow deposits that later buried the Middle Paleolithic artifact-bearing sediments. The transportation of large chunks of matrix had removed the Middle Paleolithic artifacts but retained their initial integrity, while the melting of the glaciers had caused the secondary redeposition of the Acheulian materials. The melting of glaciers can create circumstances that are key to archaeological interpretations. Ice-margin or glacial meltwater lakes create landscape contexts that are especially suitable for prehistoric human habitation. Geomorphic features along the margins of the basins and the former drainage into these basins are likely areas to find evidence of human occupation. In the western Great Lakes area, lakes and drainage overlying glacial till and outwash deposits provide evidence of Late Pleistocene and Early Holocene human occupation of the region. In very cold settings postdepositional frost heaving (see below) can result in the formation of rock rings or polygon-shaped features. These form when intervals of extreme cold followed by melting and a freeze-thaw cycle produce contraction cracks at the earth’s surface.

Coastal and Marine Depositional Settings The margins of larger bodies of water, large lakes, seas, and oceans, contain various sedimentary environments that are of archaeological significance. Larger bodies of water have sediments deposited as beaches, backshore dunes, spits and bars, and river-mouth bars and deltas. Changes in the level of the water can either inundate these coastal features or leave them ‘‘stranded,’’ forming abandoned shorelines. In either case, these margins are critical settings of past human occu-

90

pation. Coastal areas that are currently underwater were frequently areas of human habitation during times of lower sea levels. One explanation for the relative absence of sites from certain time periods may be that most of them are now under water. For example, one of the potential migration routes for Pleistocene human groups coming into the New World is along the Pacific coast of western North America. Sites situated along this area dating to the Last Glacial Maximum (20,000 years ago) would have been inundated by the rising sea levels during the last deglaciation. The reconstruction of geologic and ecologic environments has provided an increasingly clear picture of the landscapes and habitats that have sustained human development. In these reconstructions, both the geologist and the archaeologist depend on incomplete stratigraphic records and evidence that is frequently insufficient for an absolute chronology. Few areas of the earth’s land surface have witnessed long and continuous periods of deposition; on the contrary, erosion is the terrestrial norm. In coastal areas three geologic processes combine to drive geomorphic change. Changes in sea level have an immediate impact on the coastal zone. Vertical (up or down) tectonic movements offset or augment eustatic rise or fall. In addition, erosion or deposition may drive the transgressional or regressional migration of the shoreline. Classical writers from the Mediterranean region, including Herodotus, Plato, Strabo, Pausanius, and Livy, noted shoreline changes, but they had no context in which to analyze them. Today we use a number of geologic concepts and methodologies to investigate geomorphic change along active coasts. The initial phase of a paleogeomorphic reconstruction relies on detailed field geomorphology to ascertain the broad scheme of landscape evolution. This may be aided by knowledge of the vertical position of archaeological structures for which we have dates or firsthand reports from ancient texts. However, three decades of work on Mediterranean coasts by Chris Kraft, Rip Rapp, and their colleagues have shown that intensive core drilling with detailed analysis of the sedimentary record is necessary to provide an adequate picture of the sequence

Initial Context and Site Formation

of coastal environments and the associated chronologies. Coastal areas host a high percentage of the world’s population. Where land and water meet there are often large amounts and varieties of nutrients that in turn supply diverse environments like bays, estuaries, beaches, deltas, dunes, and marshes, with adjacent floodplains. Add to this the high energy available from wind-driven waves, and it is easy to see why constant change is the rule. Figure 3.16 shows the geomorphic evolution of the site of Tel Michal on the coast of Israel. This is the same set of sediments from the eastern Mediterranean coastal plain described in Chapter 2. In investigating Holocene coastal change one must view the landscape in terms of the morphology of sedimentary bodies and erosional features, as well as the vertical and lateral sequences of environments created by the processes operating in the coastal environment. These processes include local tectonism, changes in sea level, climatic change, ocean currents, and wave regimes. In addition, one must consider the nature and frequency of catastrophic events; sources, types, and quantities of sediments available; and the nature and intensity of human activity. Environmental processes leave a record of past physiographic change in the local sediments. Microfaunal and microfloral remains in the sediments record such environmental parameters as salinity, water depth, and even pollution. Investigation of them affirms one of the fundamental beliefs of geology, uniformitarianism. Geologists rely on a well-developed concept of sedimentary facies. By facies, geologists mean the characteristics of a rock unit that reflect the conditions of its origin and differentiate it from adjacent units. When coastal sedimentary deposits are considered in terms of their lateral and vertical relations, they provide a full chronologicalgeologic record of the events that resulted in coastal change. For example, the details of the history of a transgression or regression of the sea will be reflected in the vertical sedimentary facies deposited. The lateral and vertical changes that can be present in archaeological sequences and strictly

geologic stratigraphies are described by Walther’s Law of Correlation of Facies, introduced by Johannes Walther in 1894. Where different sedimentary facies reflect different environments, shifts in depositional environments through time will eventually cause the deposition of one type of environment over another (fig. 3.17). Therefore, relations between laterally contiguous facies (representing environments) can be connected to vertically stacked sediments. In other words, for conformable sedimentary sequences, the facies that occurred laterally side by side will appear stacked within a stratigraphic column. Thus, the deposits formed next to one another on a contemporary landscape—starting inland within a shoreline, followed by nearshore and then deepwater settings, and with shifts in environments caused by an expanding water body—would create a vertical sequence where shore-margin sands are covered by deep-water muds. This is important for paleogeographic interpretations because vertical changes preserved in a sedimentologic record provide clues to the types of environments that existed adjacent to these deposits, regardless of whether the lateral deposits are available for study. Walther’s Law can be illustrated in stacked sedimentary sequences when the element of time is introduced. Coastal-change studies draw heavily on Walther’s Law, which provides a means of interpreting environmental changes. According to Walther’s Law, only those sedimentary facies that occur in laterally adjacent environments of deposition can occur in conformable vertical sequence. Walther’s Law gives geologists a powerful tool; it enables them to use three-dimensional stratigraphic sequences in the reconstruction of ancient (sedimentary) landscapes through time and space. Coastal Processes and Site Formation Coastal shoreline processes can erode and redeposit or bury sites of past human activity. Progradational sequences can chronologically order sites located next to shoreline features (the oldest will be away from the present margin; the youngest, on the present margin), while transgressive episodes can result in erosional surfaces and the

91

Initial Context and Site Formation

Figure 3.17 Three-Dimensional Representation of Walther’s Law of Correlation of Facies The internal relations between sedimentary facies and environments are reflected in vertical stratigraphic se-

quences. Here the lateral relations show facies associated with beach-margin, offshore, and deep-basin deposition. The vertical stratigraphy shows these facies after a flooding or transgression.

redeposition of artifacts. Coastal settings provide a wide variety of depositional contexts that can influence human habitation and artifact accumulation. Archaeological accumulations have been found in cave stratigraphic sequences associated with sea margins, as well as in shell midden deposits. In the study of Late Quaternary coastal change around archaeological sites adjacent to the world’s oceans and seas, nothing is more important than the variation in sea level. Coastal

geologists have labored unsuccessfully to construct generic eustatic rise curves. Sea level began rising rapidly at the end of the Pleistocene, when the massive continental glaciers melted. Today, in the northern hemisphere, only the Greenland and Arctic continental ice masses remain, while 20,000 years ago large ice sheets covered most of Canada and Scandinavia. During the last glacial low stand, sea level was more than 100 m below the present mean sea level.

Figure 3.16 Idealized Reconstruction of the Evolution of the Coastal Site of Tel Michal, Israel Initially the location consisted of a coastal ridge formed of kurkar. By 1600 B.C.E., Middle Bronze Age structures had been built on top of the kurkar ridge. Between 1600 and 400 B.C.E., coastal processes caused erosion of the seaside section of the archaeological deposit.

Human habitation continued through the late Persian period, increasing the size of the tell at the same time that coastal erosional processes were destroying it. Present coastal processes have removed a substantial portion of the seaward-facing part of the archaeological deposit. (Based on Gifford et al. 1989)

93

Initial Context and Site Formation

Figure 3.18 Global Sea-Level Change A global sea-level curve depicting eustatic rise, using

radiocarbon-dated samples collected from inferred stable areas. (After Stanley 1995)

Along many continental slope breaks in different parts of the world, there are numerous wave-cut notches at 120 to 125 m below sea level. At the beginning of the Holocene, sea level rose rapidly from those levels until somewhere between 8,000 and 6,500 b.p. (fig. 3.18), at which time the rate of rise decreased substantially.43 These higher sea levels flooded the ‘‘landbridges’’ that had been used by Pleistocene human groups and had connected Siberia to Alaska, England to France, and Japan to the Asian mainland. The shorelines of today are often far from where they were in earlier periods. Norwich, England, was a seaport 700 years ago. Ten thousand years ago, the southern North Sea was a marshy plain. England and continental Europe were joined until about 8,000 b.p. Sea-level rise and fall is a major cause of transgressions and regressions. Above, we presented information on how the enlargement and reduction of water bodies can be inferred from

changes in the characteristics of sedimentary deposits. Transgressions and regressions can also be correlated with changes in both floral and faunal communities. Transgressions seem to take place much more slowly than regressions; therefore, the latter should have more dramatic consequences on plant and animal communities. In regressions, organisms suffer extinction, changes in diversity patterns, and forced migration. In shallow water areas, the principal paleoenvironmental events are due to transgressions (for example, the development of brackish water). Short-term changes in local faunal communities are usually the consequence of environmental perturbations, because any evolutionary changes in these groups would be too slow to detect. Because of their slow pace, we cannot observe marine transgressions directly, but we can easily see some of the consequences, such as a rising water table. The Inupiat site of Pingasagruk in northern

94

Initial Context and Site Formation

Alaska is a prehistoric and historic habitation situated on a coastal sandbar that contains artifact distribution patterns which appear to have resulted in part from storms and consequent erosion and transport.44 Storm waves erode small artifacts from dunes, leaving behind larger objects like fragments of coal and cobbles on the beaches of seas. Artifacts eroded from deposits tend to accumulate in secondary concentrations because eddies and turbulence prevent them from washing away. They can concentrate on the beaches of bays. When breakers submerge areas, more artifacts are likely to be carried away. Artifacts like bone, antler, and ivory become water-borne, tumbled and rolled around before they are redeposited. Artifacts transported from the original location can be temporally mixed as well as spatially displaced. Coastal Landscape Context The region of southern Epirus in western Greece contains major archaeological and historic sites, including Nikopolis (the victory city the Roman emperor Octavian took after he defeated Anthony and Cleopatra) and Actium (the focus of the famous battle). The region has an abundance of major sites ranging chronologically from the Paleolithic to the Medieval Period. Literary and historical references go back at least to the eighth century b.c.e. when Homer and his contemporaries considered the Acheron River Valley to be an entrance to the underworld. Recent studies based on coring provide a detailed description of Holocene coastal change and its human impact in this archaeologically rich region.45 In addition to altering artifact composition and spatial patterns, coastal geologic processes can have a major influence on other aspects of archaeological interpretation. For example, the 1,500–3,000-year gap between the Early Formative Valdivia and the earlier preceramic Vegas occupations of coastal Ecuador may not reflect depopulation of the region, as was once believed, but instead could be associated with changes in the position of coastal shorelines. The abandoned coastline thought to be associated with the Valdivia period contains a variety of features, including uplifted ancient shorelines, cross-bedded

beach deposits, and shells. Pottery associated with beach sands and shells indicate an occupation around 5,500 b.p., overlain by a Late Valdivia site dated to around 3,500 b.p. This indicates that the shoreline environment was previously inland from the current coastline. Other abandoned coastal features near the site of Real Alto in Ecuador indicate that during occupation it was situated along the shoreline. This geoarchaeological information instigated a reevaluation of Real Alto in Andean prehistory. The site had previously been interpreted as an inland location in which agriculture was the dominant feature of the prehistoric economy. The presence of clams at the site was attributed to their use in constructing a pavement. Now it seems possible that the people associated with the Valdivia artifacts had a diverse subsistence economy that included fishing, hunting, and agriculture. Real Alto seems to have been established to take advantage of coastal resources, and the abundance of clam shells surrounding Early Valdivia dwellings may be the result of onsite refuse disposal by people living near coastal mangrove swamps. The site of Real Alto was probably abandoned because of coastal uplift, which made it an inland location. The land between the site and the present-day coast contains later Valdivia sites, and it is possible to relate the distribution of sites to paleo-landforms. Gaps in the archaeological record appear to be caused by restructuring of the shoreline. Some terrestrial areas were not present during Valdivia times and consequently cannot contain sites of this period. Another example of what we can learn from coastal landscape change can be seen in the Late Holocene topography of Greece and its relation to available routes of human movement. Conflicts among historians over the Battle of Thermopylae in 480 b.c.e. between the Greeks (mainly Spartans) and the advancing Persians center around the inconsistencies between ancient sources and the modern topography. There is a discrepancy between the width of the ‘‘pass’’ along the coast at Thermopylae. Kraft, Rapp, and their colleagues used a core drilling program to reconstruct the relevant paleogeography. With the recovery and analysis of materials from seven core

95

Initial Context and Site Formation

Figure 3.19 Geoarchaeological Reconstruction and the Battle of Thermopylae View of the ancient and modern shoreline margins in the vicinity of Thermopylae, Greece. The Thermopylae Pass, or coastal route, was much narrower during ancient times because the level of the Gulf of Malia was higher: there was less land between the sea and the upland regions. A dramatic change of more than

2 km in the Sperchios River drainage is recorded for the interval between 2500 B.C.E. and 480 B.C.E. Around Thermopylae, a large change occurred in the width of the pass between 480 B.C.E. and the twentieth century. The pass increased in size by several kilometers. The smaller size of the Thermopylae coastal route was a crucial factor in the Greeks’ ability to hold the pass. (Based on Kraft et al. 1987)

drilling holes it was possible to delineate considerable variation over time in the coastal geomorphology at Thermopylae. Figure 3.19 illustrates the coastal change of the Gulf of Malia-Sperchios River floodplain at Thermopylae. In 480 b.c.e. the narrowest coastal pass at the place where the small army of Spartans would have had the best chance against the larger Persian army would have been less than 100 m wide. From the point of view of understanding a changing landscape context, it is critical to recognize that the land surface of that time is currently buried under up to 20 m of terrigenous

clastic sediment and hot-spring travertine deposits. These changes underscore the difficulties encountered by investigators who use only observations of the current landscape to reconstruct earlier landscapes in areas of high deposition. Without core drilling to determine the location, elevation, and nature of the earlier landscape features, physiographic reconstructions will remain interesting guesses. The sedimentary sequences and paleogeographic reconstructions at Thermopylae also provided information for interpretations of sites of later local battles: Greeks versus Gauls in 279 b.c.e., Romans versus Antiochus the

96

Initial Context and Site Formation

Figure 3.20 The Harbor Area at Ancient Troy In Homeric times the Aegean Sea coastline near ancient Troy was well inland from its current position. This region contains an extensive depositional record of the evolving Holocene landscape. A decade-long

core drilling program has revealed the Trojan landscape. This research has shown that the descriptions in Homer’s Iliad are substantially correct. (Based on Kraft et al. 2003)

Syrian in 191 b.c.e., and others.46 With additional coring, even more details of the ancient coastal topographies would emerge. However, the basic subsurface geologic data indicate and date the major delta system to the west as well as the sequence of changing environments along the coast (fig. 3.19).

Recent investigations of coastal change relate to significant archaeological sites. Detailed paleographic reconstructions of the region around ancient Troy in eastern Turkey were used to relate the paleolandscape to the features and events described in Homer’s Iliad (fig. 3.20).47 These studies indicate that Homer’s descriptions of the

97

Initial Context and Site Formation

geographic features correlate very well with the geologic evidence. A more precise understanding of the landscape of the ancient Roman harbors of ancient Ephesus in western Turkey was attained by integration of the written record and sediment studies on drill cores.48 Geologic mapping of the sedimentary sequences in the lower Cayster River floodplain allowed relatively precise delineation of the shoreline and the harbor features. Ancient urban systems continually adapted to an ever-changing landscape as the prograding delta moved past the city, infilling harbor installations. In southern Italy, settlement of the sanctuary of Hera Argiva by the sixth century b.c.e. Greeks on the Sele River coastal plain was followed over the past two and a half millennia by movement of the coastline 250 m landward. The paleogeographic reconstructions and stratigraphic succession were based on the analyses of sixteen deep cores, each approximately 25 m long.49 This famous sanctuary and the city of Poseidonia, 5 km south of the river mouth, were founded by the Greeks during a progradational highstand. Poseidonia had a lagoonal harbor, ultimately doomed over time by infilling of the lagoons. Among the earliest, if not the earliest, harbor installation was that built for the Palace of Nestor near the Bay of Navarino in Messenia, Greece.50 A canal was dug from the sea about 500 m to an excavated basin. The Mycenean engineers even developed a system to prevent the silting up of the basin. Most archaeological features, except of course harbor installations, give no evidence about how far from the sea they were constructed. The ages and sequence of proglacial lake levels in the Lake Superior Basin and their relation to Paleo-Indian sites present a good example of coastal change affecting habitation possibilities and context. Geologists, archaeologists, and geoarchaeologists have investigated these relationships.51 Paleo-Indian sites on raised beach terraces and strandlines are well documented. Prehistoric groups favored these shoreline sites because of their biologic resources and supply of fresh water. Along the north shore of Lake Superior there was also an abundant supply of lithic resources. Archaeological interpretations of the sites depend

98

on geologic determinations of the sequence of landscapes and on lake-level chronologies. Habitation possibilities in the area depended not only on water levels in the Lake Superior Basin but on the extent of glacial Lake Agassiz. Lake Agassiz was the largest of the North American proglacial lakes, inundating nearly a million km 2. Fluctuations of the lake greatly affected the size and suitability of the land area between the two lakes. Lake Agassiz beaches and strandlines offered fluctuating travel routes and landscape settings for early human inhabitants.

Postdepositional Processes After artifacts are incorporated into a sedimentary deposit (whether buried as a primary assemblage or secondarily redeposited), physical and chemical processes alter the spatial and compositional character of the artifactual components.52 The major physical (geologic and biologic) processes which can affect the spatial distribution of artifacts include mass wasting, cryoturbation, expansion and contraction of clays, deformation, and bioturbation. Postdepositional chemical conditions can destroy or protect artifacts. Postdepositional alterations of sediments result from several processes. Subaerial weathering processes include biochemical alteration, humification as well as illuviation and eluviation (leaching and accretion). Other processes include turbation, fluctuating groundwater levels (which are reflected in oxidation and reduction mottling), and authigenic carbonate deposition. The swelling and shrinking of clays (usually in Vertisols) cause a form of mixing called argilliturbation. During dry intervals, sediments composed of high amounts of expansible clays will shrink and crack. Artifacts lying on the surface of these sediments can fall into the cracks. During moister intervals the clays may expand and push up sediments. Scientists hypothesize that alternate wetting and drying cycles caused the formation of stone pavements and the size sorting of artifacts through the preferential upward movement of larger objects. One of the postdepositional processes that

Initial Context and Site Formation

alter sediments is the secondary accumulation of carbonate. Pedogenic carbonate enrichment can occur when soil-moisture evaporation is greater than or roughly equal to water infiltration. Chemical processes within a near-surface sediment can form concretions and nodules by accretion. Calcite can occur as isolated nodules or as coalescing layers and can range from a few centimeters to several meters in dimension. The first stage of formation of calcium carbonate–rich horizons of pedogenic origin occurs when carbonate is deposited on the undersides of gravel particles. Accumulation of calcium carbonate may also result from capillary rise from a perched highwater table. The formation of several types of horizons can be associated with high moisture content or nearness to a water table. Poor water drainage, and consequent low oxygen content, leads to reducing conditions that result in the deposition of iron and manganese compounds, which form gray and bluish colors of gleyed horizons. A fluctuating water table level can produce varying oxidizing and reducing conditions that result in mottling. Soluble salts, iron compounds, and manganese compounds may accumulate at the top of either the water table or the capillary fringe. Shell middens present a host of both opportunities and problems for the geoarchaeologist. These middens are found in nearly every coastal area of the world. Such deposits are particularly subject to diagenetic change from sea-level variation and saturation by groundwater. Chemical changes affect the porosity, permeability, and alkalinity. Compaction, translocation, and bioturbation combine with the chemical changes to render stratigraphic and archaeological analysis difficult.53

Mass Wasting Influenced by gravity, sediments can move downslope and mix constituent materials in the deposit. A variety of mass-wasting processes may alter the archaeological record. They can be separated into slow gravity-movement processes like creep, solifluction, and subsidence and rapid gravitymovement processes like mudflows, landslides, and rockfall.

Solifluction is the downslope movement of water-saturated soil and sediments. Under periglacial conditions, solifluction has caused major deformation of archaeological stratigraphic sequences. Solifluction of the sediments overlying an occupation layer caused folding and discontinuities in the artifact layer at the Denbigh site in Alaska, while at the Engigstack site (also in Alaska) this type of movement caused reversals in the stratigraphic sequence. Soil creep can also have a major influence on the spatial distribution of artifacts by causing a movement of deposits in which heavier and denser artifacts tend to be transported farther downslope.54 Artifacts can also be buried by soil creep if they were originally situated at the base of a slope. More rapid massive downslope movements can transport large quantities of sediment very quickly, moving artifacts, producing geofacts, and burying archaeological sites.

Cryoturbation Disturbance of sediments, soils, and artifactdistribution patterns caused by cycles of freezing and thawing is called cryoturbation. Cryoturbated sediments may exhibit several features. Deformation of the earth’s surface during a freeze can cause the upward bending of strata. The creation of patterned ground features, consisting of assorted polygon shapes, during frost heaving can result in the outward movement of objects to the edges, while large objects are thrust to the surface. Cryoturbation is especially critical in periglacial settings but must also be considered as a potential influence in mid-latitude and mountain areas. For example, the present-day maximum frost penetration in midwest North America between Missouri and Minnesota ranges from about 50 to 250 cm.55 Frost heaving is a cryoturbation process connected to freezing and thawing events that can cause upward movement of earth and artifacts; it also reorients artifacts within deposits either by frost pull or frost push.56 When groundwater freezes, it expands and can push artifacts. When the ice melts, water surface tension pulls finer sediments together, while artifacts and larger sedimentary particles remain where the expan-

99

Initial Context and Site Formation

sion of ice pushed them. The presence of size sorting (with larger artifacts closer to the surface) and vertical orientation of the long axes of artifacts within sedimentary sequences may indicate frost heave. Freezing-induced (cryostatic) pressures also cause upward and lateral movement of objects. These in turn can cause contortion and deformation of sediments, such as that found at the Dry Creek Site in Alaska, where the stratigraphic sequence shows evidence of cryostatic pressures in the form of involutions and massdisplacement features.57 Ice and sand wedges are another indication of sediment deformation. Particle size sorting, in which upward sequences become coarser, may result from frost action. This can have critical implications for archaeological components composed of a mixture of artifact sizes. Mechanical abrasion caused by cryoturbation has been proposed as an explanation for geologic ‘‘retouch’’ on flakes in Paleolithic assemblages.58 Another set of cryoturbation features, termed patterned ground, consists of semi-symmetrical shapes formed along the ground surface by frostheaved rocks. These features can take the shape of circles and polygons (among others), and it is important to recognize them because they can be mistaken for structures made or used by humans. Frost action can transform the archaeological record in several ways. It can alter stratigraphic sections, sort both organic and lithic debris, and affect surface distributions of materials. Sediment contortion may obscure stratigraphic boundaries. It may transport larger artifacts upward, and it may transform the patterns of surface artifacts.

Bioturbation Bioturbation can be divided into two major groups: modifications caused by animals (faunalturbation) and disturbances caused by plants (floralturbation). Postoccupational trampling and other alterations by humans and other animals are forms of faunalturbation that change the initial state of the archaeological record. However, it is likely that most postburial faunalturbation in archaeological sites comes from burrowing by small mammals (mainly rodents), insects, and

100

earthworms (fig. 3.21). Different types of animals have different effects on archaeological deposits, depending on their burrowing patterns. Vertical movement of smaller objects may be caused by subsurface foragers like gophers and certain earthworms. Mixing and churning by burrowing animals can cause the disappearance of boundaries between originally distinct layers. Surface foragers like prairie dogs, foxes, ants, termites, and rodents may also cause disturbance by building tunnels and nests. Earthworm activity is a major cause of confused stratigraphy. Earthworms bury stones and seeds, build cairns, produce calcium carbonate nodules, homogenize the layering in sediments and soils, and move soil from depths to the surface.59 Such biologic activity is of major importance to those geoarchaeologists who must deal with the microstratigraphy of excavation trenches. Ants and small vertebrates mix and displace particles, including artifacts, through burrowing and mounding. Larger particles tend to be displaced downward and smaller objects upward.60 The maximum depth of artifact burial will correspond to the base of the major biologic activity. In Mollisols and Alfisols of the midwestern United States, this depth commonly approximates the uppermost part of the B horizon. The effects of pocket-gopher burrowing on the archaeological record illustrate the influence of bioturbation. There are at least four ways that gophers affect artifact accumulations: movement, destruction, impact on sedimentary structures, and organic enrichment. When excavating burrows, gophers push sediment into mounds. This results in vertical and horizontal transportation of sediments and artifacts. Fragile artifacts composed of bone, shell, or plant remains may be broken during burrowing, and materials pushed to the surface may be exposed to weathering. Structures and boundaries within and between strata are disturbed or obliterated by burrowing activity. Pocket gophers collect organic matter from the surface and transport it underground, as well as provide additional organic enrichment by the deposition of feces. In the long term, gopher burrowing influences the patterns of artifact size sorting, strata disruption, and destruc-

Initial Context and Site Formation

Figure 3.21 Effects of Bioturbation This illustration shows what happens to the original

patterning of the soil in three situations, after a lapse of time T1 and after a further period of time T2 .

tion of fragile remains. Occasionally these have produced distinct artifact assemblages that have been interpreted without consideration of the potential effects of bioturbation. Characteristics used to define particular artifact or ‘‘cultural’’ entities may instead be the result of bioturbation,

such as at the Milling Stone archaeological horizon in California. Mixing of archaeologically related soils and sediments by plants can be caused by root growth and decay and by tree fall. Root action can disturb archaeological sites by pushing artifacts or leaving

101

Initial Context and Site Formation

cracks within the sedimentary matrix. If a tree is uprooted, artifacts may be brought to the surface. Discrete artifact components can be mixed, and size sorting of artifacts can result when materials adhering to the root system are redeposited. As with patterned ground there is the possibility that topographic features formed by tree throw may be interpreted as the work of humans. Plants also contribute to the stabilization of sediments and the development of soils. Artifacts within soil horizons may have been deposited as part of the original sedimentary matrix or during a later interval associated with surface stability, vegetation growth, and soil development. Don Johnson contends that the role of bioturbation and related biochemical processes have been undervalued in many areas of geoarchaeological research, particularly those of pedology and geomorphology.61 He advocates a process model (dynamic denudation) that incorporates fully the role of bioturbation in all soil landscapes. Sedimentary contexts provide the initial landscape setting that affects both human habitation and use of a location and what happens to the objects after people use them. In addition to influencing the functional activities that occur at a site, different sedimentary contexts influence the

102

visibility of the site and original site integrity. Many agencies can alter the original primary context by erosion, transportation, and redeposition. As well, secondary processes can also affect the site once the materials are within a depositional matrix. There are several avenues to pursue in trying to determine the kinds of behavioral information available at an archaeological site. Two basic types of information can be used to help interpret the patterning found in the archaeological record. First, one can consider attributes directly associated with artifacts and archaeological features. Second, one can examine the sedimentologic and landscape context of the archaeological materials. The latter is specifically geoarchaeological, and the information can affect how one regards the first type of information. Landscape and sedimentologic data that indicate less disturbance relates to low-energy levels and few opportunities for erosion and redeposition. Greater opportunities for postoccupational site changes occur in higher energy settings. The different sedimentary systems that we have related to archaeological site formation show that the final condition of the archaeological record is a result of a continuum between the primary, systemic context and postoccupational transformational processes.

CHAPTER 4

Methods of Discovery and Spatial Analyses The interpretations of geoarchaeology . . . must be made at the scale of archaeology—about the human past using the human time scale. . . . Scale is an essential factor in understanding . . . the historical sciences, especially when defining . . . geoarchaeology.—Julie Stein 1993

A

geoarchaeological perspective is valuable for discovering and evaluating archaeological sites. Locating evidence of the human past from within stratified sedimentary deposits or on landscapes relies on a fundamental understanding of the methods of geo-spatial documentation and analysis. A general knowledge of maps, landform assemblages and settlement patterns, and remote sensing (including geophysical prospecting, aerial photography, and satellite imaging) should be part of every archaeologist’s training. The collection and analysis of data from coring and geochemical studies also can be used to predict, locate, and evaluate preserved contexts of the human past. The application of these techniques and their integration with geographic information systems as well as an appreciation of the complexities of scale is critical for discovering and interpreting the archaeological record.

Maps Classification, as with many other aspects of geoarchaeology, is central to mapping. Information selected for mapping or display on a map has already been grouped into a small number of categories. Consequently, mapmakers begin with

information which has already been classified. Classification, then, reflects the point of view of those gathering the data and making the maps more than it does the underlying physical reality. Classification schemes are designed to be functional. When they are not, they are altered or abandoned. Maps always reflect a simplification of the variables displayed. Many kinds of maps are available to the archaeologist. Two of these, topographic maps and surficial-geology maps, sometimes called Quaternary maps (first introduced by the Geological Survey of England in 1863), are of major importance in geoarchaeology. Maps of the bedrock geology (the map they give you when you ask for a geologic map), mineral deposits, hydrology, or seismicity are important for some studies. Quaternary maps are quite distinct from soil maps. The former show deposits that are about 0.5 m below the ground and do not take topsoil into consideration. The colors on a Quaternary map represent genetically and lithologically different deposits. Soil scientists make similar maps, on various scales. Field methods are nearly identical to those used in geology and geomorphology, but there is an added emphasis on close examination using coring and trenching (because soils vary more rapidly laterally than do geologic features or materials). A topographic map, as distinct from other kinds of maps, portrays the shape and elevation of the terrain. Most topographic maps represent elevations by contour lines. Contour intervals vary with the relief of the terrain. On quadrangle maps made by the U.S. Geological Survey and similar organizations, small irregularities of the ground

103

Discovery and Spatial Analyses

surface are omitted from the map: the contours are drawn as smooth lines through areas of irregularity. Thus small mounds and other features that are of interest to archaeologists may not be represented on topographic maps, depending on the size of contour interval. Topographic maps contain a great deal of information other than elevation contours. Surface-water features (including swamps and springs); modern or abandoned quarries and mines, buildings, roads, and other cultural features; and areas of vegetation are all recorded on topographic maps. Major geomorphic features like escarpments, terraces, dunes, and waterfalls can easily be discerned, which makes it possible to reconstruct landscapes. When geologists do their field mapping, they often use topographic maps or air photos as base maps on which to record their observational data about geologic materials and features. However, the contours on topographic maps (and even less on elevation maps) do not adequately express the details of roughness and steepness of landscapes. When the scale is appropriate, geoarchaeologists use topographic maps as base maps on which to plot special features or information. That is, they construct a thematic map, such as one that portrays slope stability. In the United States, the largest-scale topographic maps are 1:24,000. Maps compiled before the second half of the twentieth century, particularly those from the nineteenth century, are typically less accurate than current maps. These older maps, however, may contain information very useful to archaeologists interested in landscape change or historic structures. Topographic maps also record the important landscape features that are the first level of geomorphic interpretation. Erosional features like stream valleys, gullies, and washes can easily be identified. In the contours outlining a hill, the shape of the contour lines will show the courses of streams too small to be drawn in with blue lines. Residual features that have resisted erosion, like mesas and terraces, can also be easily identified on topographic maps. Depositional features like alluvial fans and sand dunes have clear topographic expression. In addition, topographic

104

maps locate benchmarks that are critical in establishing vertical control in an excavation or survey. A good example of what can be learned from a topographic map can be seen in figure 4.1. Supplemented by aerial photography, limited drilling, excavation, and 14 C dating, the geomorphic context and paleographic history of a Woodland site called Hannaford was compiled from the Delvin Quadrangle USGS topographic map. The Rainy River is the boundary between the state of Minnesota and Canada. The stratigraphy and geomorphology of the region are the result of geologic events that have affected all of northern Minnesota over the past 15,000 years. The ridge and swale topography, easily seen in figure 4.1, is the geomorphic expression of an eastward movement of the Big Fork River channel before about 2,500 b.p. This was later overlain by sediments deposited in a floodplain regime. The meander scar pattern to the west/southwest of the Hannaford site is clearly older than the ridge and swale topography on which the Hannaford site strata lie. The Hannaford site itself has a pointbar setting. The geoarchaeological interpretation of the site-geomorphic context is shown in figure 4.2. This figure is a generalized cross-section of the A-A' line of figure 4.1. A good general picture of the hydrosphere is also displayed on topographic maps. Because potable water resources, water transportation routes, barriers to transportation (large marshes or peat bogs, rivers), and food resources derived from water bodies are vital, topographic maps are central to understanding sites and settlement patterns. Field archaeologists and geoarchaeologists are strongly advised to become expert in interpreting topographic maps. Manuals at all levels of expertise are readily available.1 Geoarchaeological interpretation of fossilplant records requires assembling information on the broad ecologic conditions prevailing and their variation in different parts of the landscape. This is vital in reconstructing human activity across ancient landscapes. Most multidisciplinary archaeological projects now begin by mapping the present vegetation. In a landscape with natural or semi-natural vegetation, a map of the present vegetation will reflect the natural edaphic con-

Discovery and Spatial Analyses

Figure 4.1 Geomorphology of the Hannaford Site

ditions—the soil and water regimes. When combined with soil maps and geologic, paleoecologic, and paleopedogenic studies from drill cores, the pattern of ecologic change and land use should emerge. In combination with archaeological survey, the human rationale for past land use, settlement patterns, and exploitation of flora and fauna can be determined. Enhanced detail can be added by geochemical surveys tracking phosphate concentrations immediately below the topsoil. Terrain mapping may include many surface and remote survey techniques including coring, test pits, geophysical and geochemical surveys, and laboratory analyses. The terrain, as it exists at present, in many instances will have been modified to some degree since the period of archaeological interest. Terrain features of interest, such as slopes, soil, vegetation, hydrology, and outcrops, can be mapped at a scale appropriate to the study. Terrain patterns will emerge in areas of recurring topographic, soil, and vegetative associations. Terrain evaluation supplements site evaluation and places a site in its environmental context. The slope of the ground, for example,

with its frequent connection to actual or potential human occupation of an area, is a fundamental aspect of terrain that can be measured from a contoured topographic map. Slopes must always be measured at right angles to the contours. Between contours the measurement will show only average slope, but this should suffice. A system of terrain classes has been devised using local relief and the percentage of smooth slopes.2 Specifically, four geomorphic properties are used to define terrain classes: (1) an inclination index—the percentage of an area occupied by slopes with an inclination of less than 8 percent; (2) the maximum difference in elevation within the area—commonly called relief; (3) the general profile character—the percentage of slope of less than 8 percent that falls in the upper and lower halves of the elevation range; and (4) the character of the surface materials. Terrain classes directly affect the distribution of large herbivores, the possibility for various types of agriculture, and the like. Specialized maps are good data sets that are used too sparingly in archaeology. The concept of landscape as a cultural artifact

105

Discovery and Spatial Analyses

Figure 4.2 Generalized Cross Section of the A-A' Line in Figure 4.1

is being increasingly addressed by anthropologists, archaeologists, geographers, and historians, as well as by geoarchaeologists.3 Geographers, in particular, have long recognized that landscapes represent the interactions of humans with the environment. In contrast, until recently, archaeologists have focused on architectural features and settlement plans, rather than on the space between communities. The concept of the continuous archaeological landscape, in contrast to the strong focus on settlements distribution, was developed during the 1970s and well described by John Cherry.4 Tony Wilkinson, in a book with a solid geoarchaeological base, provides a theoretical foundation for studying the archaeological landscapes of the Near East.5 For an anthropological view of landscapes see the volume edited by Peter Ucko and Robert Layton.6 Because bedrock and surficial geology maps are two-dimensional representations of threedimensional structures, they often contain cross

106

sections printed along map borders. Geologic maps, by means of graphic devices indicating, for example, the inclination of nonhorizontal strata, present a great deal of information in concise form. Geologic maps contain the information about natural resources—metal deposits, building stone, some types of clay deposits—as well as the information necessary to gauge the present and past suitability of an area for construction of canals or dams. The ‘‘World’s oldest surviving geological map,’’ the 1150 b.c.e. Turin Papyrus from Egypt,7 has been discussed in terms of the topography and geology of the central Eastern Desert of Egypt. The map appears to accurately depict the special distribution of sedimentary and igneous/metamorphic rocks, a gold-working settlement, goldbearing quartz veins, wadi gravels, and other cultural features. Surficial mapping of the Moche Valley in north coastal Peru helped to show the ways that changes

Discovery and Spatial Analyses

in Andean Holocene landscapes influenced people living in the region.8 Geologic surfaces of different ages were related to archaeological sites and deposits. Because the archaeological remains were cross-cut by braided channels it was possible to date alluvial surfaces numerically. The study of these surfaces could be used to inspect archaeological site patterning. The margins of the valley contained older surfaces while most of the alluvial surfaces in the valley were Late Holocene in age. The paucity of Early Holocene alluvial surfaces and archaeological remains may be related to increased erosion and alluviation produced by changes in Holocene climate that may have resulted in a higher frequency of flood events. Surficial maps provided one important line of evidence in examining the relation between the development of complex irrigation-based societies in the Peruvian desert and the rates of landscape change. A geologic map provides the basis for preparing second-order maps like isopach maps, which show the variation in thickness of a rock unit; structure-contour maps, which show the variation in altitude of the upper surface of a rock unit; and landslide-susceptibility maps, which show landslide potential in relation to bedrock, geomorphic, and climatic factors. With additional data, second-order maps can be constructed for almost any physical (for example, aquifers), chemical (for example, variation in phosphate content), or cultural (for example, quarries) feature that varies systematically with the geology. From geologic mapping and from surface observations, the geologist can predict the subsurface geology. The archaeologist is not as fortunate. The geoarchaeologist should be able to gather the maximum amount of subsurface information from geologic techniques available. For example, if a site is on a ridge composed of a gently dipping sequence of sands, silts, and gravels, the geoarchaeologist can project the depths of a given geologic layer from one part of the site to another. Hydrogeologic conditions, including water-table depth and springs, can also be predicted. The U.S. Geological Survey publishes Hydrologic Investigation Atlas maps, overprinted on the 7.5 minute, 1:24,000 topographic map base.

Soil maps show the distribution of soils on the landscape. This distribution is related to active surface and near-surface processes. The age of a soil can be no greater than that of the land surface upon which it has developed. Geologic maps, by contrast, represent in a given area the history of the earth’s outer crust for a considerable length of time. A lithostratigraphic unit depicted on a geologic map may or may not be related to geologic processes currently active in the area. In the United States the main source of soil-related information is the Soil Conservation Service, which can provide soil maps and soil data forms. The latter contain a soil description, soil properties like permeability and pH, and interpretations of land-use limitations. Land-use limitation maps can be constructed for geoarchaeological reconstruction of paleoenvironmental contexts. Special maps of great significance for geoarchaeological investigations have often been made for very different purposes. A good example is shown in figure 4.3. The U.S. Army Corps of Engineers and other government bodies responsible for waterways investigations compile and often publish data and maps that relate directly to paleogeomorphic change. Figure 4.3 details the wide meandering of the Mississippi River from 1765 to 1961. Using coring techniques, geoarchaeologists can follow meandering river courses back through a considerable amount of time. Every human activity has its own scale: that is, scales differ according to activity area, habitation site, or site-catchment area. Maps of these activities have scales that range from 1:50 to 1:2,500. Geologic phenomena are mapped at vastly different scales—1:25,000 or 1:50,000—and are not made for archaeological purposes. In mapping, the concept of scale is really one of resolution. Small scale means low resolution, and large scale means high resolution. Reading the landscape from topographic maps is an important skill for geoarchaeologists. If a topographic map shows an isolated prominent hill, there is no single probable explanation for this distinctive feature. It may represent an isolated outcrop of rock, more resistant to weathering than surrounding rocks. Commonly these will be outcrops of igneous rocks—for example, a vol-

107

Discovery and Spatial Analyses

Figure 4.3 The Course of the Mississippi River, 1765– 1961

(Reconstructed from data and maps published by the U.S. Army Corps of Engineers)

canic neck like Devil’s Tower, Wyoming, in the western United States. A second possibility is that the hill is not formed from more resistant rock but is merely capped by a more resistant layer. This cap rock can be either igneous or sedimentary. A third explanation is that the hill is merely a residual feature of the landscape, not associated with a resistant lithology. There are many linear hill features that find expression on a topographic map. Asymmetrical hills with one steep face and one gently sloping face may well be caused by the outcrop of a resistant, gently dipping bed. More symmetrical ridges, with both sides steeply sloping (known as hogbacks), are often formed by an outcrop of a resistant bed that dips steeply or even is nearly vertical. Many rock shelters that contain deposits with archaeological materials are formed by hogbacks. Flat-topped hills bounded by steep slopes,

sometimes called mesas, are most often caused by resistant cap rocks that are nearly horizontal. Very flat coastal plains often represent land that has emerged because of regression of the sea. Large, very flat areas inland may be either the floodplain of a large river or the bottom of a former lake. The pattern of rivers and streams exhibited on a topographic map will indicate their nature and origin, as well as aspects of the bedrock and climate. More detailed examinations of river drainage patterns, valley forms, and coastal geomorphologies are available in many excellent geomorphology books. For this overview, suffice it to say that geoarchaeologists responsible for terrain analysis need to be competent geomorphologists, with the ability to ‘‘read the landscape.’’ The larger river valleys in the central United States contain large volumes of Holocene alluvium. C. Russell Stafford and Steven Creas-

108

Discovery and Spatial Analyses

man have documented the association between floodplain Mollisols, Inceptisols, and Entisols and Late Holocene landforms.9 Although paleosols are rare, archaeological materials are abundant in Late Holocene ‘‘Landform Sediment Assemblages.’’ Because many of the valley fills are Late Holocene, Woodland Period sites are frequently buried and invisible in surface surveys.

Landform Sediment Assemblages Landform Sediment Assemblages (LSAs) are informal mappable units that identify specific landforms that tend to be underlain by a sedimentary sequence of characteristic lithofacies. Mapping of LSAs can provide the basis for assigning landscape suitability ranking for the likelihood of finding buried archaeological sites. These mappable units must be recognizable in the field with the aid of topographic maps, aerial photography, and soil survey data. Mappable units should be as archaeologically relevant as possible. For example, a unit might be labeled ‘‘no likelihood of sites’’ because the landscapes were geologically too old, too young, or destroyed by erosion. Such landscapes would be of no interest to archaeologists, except for understanding the context of geomorphic events in a particular area. Landscapes are sets of genetically related landforms. The recognition of different landscapes is based on assemblages of features characteristic of the geologic agent or process that constructed the landform (for example, wind, running water). A distinct landscape would be, for example, a pediment—an erosion surface landform with a thin mantle of alluvium. Landform Sediment Assemblage analysis should work well in large alluvial valleys. Where such valleys have multiple terraces, it is important to construct sections across the valleys and to correlate the terrace levels down the length of the valleys. Alluvial valley landscapes are mosaics of landforms and sedimentary sequences. The spatial and temporal patterns of these LSAs govern many spatial and temporal patterns in the archaeological record. Patterns of erosion and deposition will produce differential preservation and visibility of the original archaeological patterns. Without geoarchaeological studies to assess the dimensional LSA context,

it is not likely that the full archaeological picture will emerge. Matthew Bennett and associates provide a detailed example of the use of LSAs in the Old World.10 The authors describe glaciolacustrine landform assemblages and associated sediments from Hagavatn, Iceland. In the New World, Arthur Bettis and Edwin Hajic have shown that intact Archaic Period deposits in the upper midwestern United States are found only rarely on modern ground surfaces of colluvial slopes, alluvial fans, and flood plains but are frequently found buried and well preserved within these landforms.11 Net aggradation of these LSAs occurred during the Early and Middle Holocene in response to climatic change that impacted water and sediment supply. Archaic deposits are buried in major river valleys except where they occur on late Wisconsin age terraces.

Settlement Patterns Understanding landscape change is crucial to geoarchaeological analysis of settlement patterns. Landscape preservation can vary greatly over small areas, depending on environmental conditions such as weathering, erosion, cultivation, downslope creep, burial, and even geomorphic stability. A site that lies on open ground for long periods of time will suffer destructive effects. The dating of events and deposits—crucial to any analysis of landscape change—can be quite complex and requires a thorough understanding of the regional geomorphology. One can often date the establishment and abandonment of sites archaeologically. When landscape factors affect these processes, geoarchaeologists need to relate rapid (for example, earthquakes or catastrophic/torrential floods) or slow (for example, sediment infilling of harbors) geomorphic changes to the human chronology. An example of this kind of change exists when a site is altered because of erosion. Such landscape features as availability of potable water or communication routes may undergo marked change in relatively short periods of time, which will affect habitation sites. For example, the changes that occurred to the coast of Greece near Thermopylae had a marked effect on the human

109

Discovery and Spatial Analyses

habitation (see Chapter 3).12 Settlement patterns are a mixture of two basic forms: nucleated and dispersed. An available water supply may be the most important of the landscape features that can explain nucleation and/or siting. Where surface water or water from shallow wells is widely available, it is unlikely to be an important factor. In desert areas or areas underlain by wide expanses of chalk, limestone, or loose and permeable sand, lack of available surface water can be key to settlement distribution. Other landscape features that affect settlement patterns that are readily apparent from good topographic maps include areas offering flood protection and defensive advantages. The floodplains of major rivers are all subject to periodic inundation. A major difficulty in using topographic maps to assess the possible importance of the flood factor is that the contour interval may be too large to show relevant altitude differences. During periods of danger, settlements have been concentrated on the defensible tops of steep-sided hills or, in the case of danger from the sea, away from observation by passing ships. Habitation sites are more numerous on coasts and large rivers. Rivers provide water, transport, food, an element of defense, and, at crossing points, an element of control of communication. Coastal sites offer some of the same advantages. Modern topographic maps are likely to misinform concerning the landscapes associated with very ancient sites of human activity. Coastal harbors on estuaries can rapidly become landlocked as sediment fills the estuary. In broad flat floodplains, the major rivers meander rapidly—for example as much as 20 km in two hundred years for the lower Mississippi. Finally, there may be bedrock geology or pedological features that influence settlement patterns and the location of archaeological sites. At different sides of a boundary of bedrock or soil types, different settlement patterns may result from the different exploitable resources. Around the peripheries of the larger poljes in the Yugoslavian karst topography are settlements that combine freedom from flooding on the polje floor with the ability to exploit the deeper soils which accumulate along the break in the slope.

110

Geoarchaeologists need to be aware of the data available on ‘‘prehistoric’’ maps from the Old World. The earliest maps grew out of prehistoric art. Picture maps from Çatal Höyük, a major Neolithic site in Anatolia, and from Maikop, Russia (ca. 3000 b.c.e.) are two striking examples. Both the Babylonians and the Pharaonic Egyptians made maps. A diagrammatic map of the late second millennium b.c.e. from Nippur shows nine settlements with canals and a road between them but no indication of distances. A complete history of ancient maps is provided by John Harley and David Woodward.13 This volume should be consulted by anyone working in Old World prehistory. Finally, the more geologic knowledge, including tectonic history, one has of a region the more one is able to frame a survey/discovery strategy. Although extensive archaeological surveys had been carried out, only a few Mesolithic sites had been discovered in the interior of northern Sweden. Ingela Bergman and others used a model of non-uniform glacio-isostatic uplift and laketilting to reconstruct shoreline displacement of ancient lakes.14 Their work has resulted in the discovery of a significant number of Mesolithic sites and of an early postglacial landscape.

Remote Sensing The term remote sensing was introduced in the 1960s to describe any method of deriving information about an object (or a stratum) from measurements made at a distance—that is, without contact.15 The advent of computer digital processing and data-analysis techniques have greatly increased our remote-sensing capabilities, for they have given us new pattern-recognition methodologies. They help us derive information about materials at or near the earth’s surface from spatial and spectral distributions of energy emanating or reflecting from these materials. An important step is the analysis of the data: through analysis, data become information. Pattern recognition is a two-step process. First, the classes (for example, objects, strata, rock types, vegetation types) are characterized

Discovery and Spatial Analyses

Figure 4.4 Electromagnetic Spectrum (Detail: The Visible Spectrum)

through the analysis of data that are representative of them. In remote sensing, ground truthing is the process of verifying ‘‘on the ground’’ the correspondence of remotely sensed data with classes of materials. Second, all new data are then classified by means of numerical rules that make use of these class characterizations. Remote-sensing systems can be passive or active. In passive systems the sensor receives energy from a target that has been illuminated by an external radiation source—the sun, for example. An active remote sensing system, like radar, generates the radiation. Most remote sensing devices use the electromagnetic spectrum. Electromagnetic energy spans the spectrum of wavelengths from 10 −10 μm (cosmic rays) to 10 10 μm of broadcast wavelengths (fig. 4.4). There are many prospecting techniques available for locating deeply buried archaeological sites: geophysical, geochemical, core drilling, and

aerial satellite remote sensing. Because each method has its special strengths, they must be applied in an ordered and efficient way. The geoarchaeologist should try to confirm results at every step. A map of geophysical anomalies remains only an ordered set of possibilities until drilling or excavation can provide confirmation.

Geophysical Prospecting As excavation becomes increasingly expensive, scientists seek the least destructive methods of gaining information, and archaeology turns more and more to methodologies like geophysical prospecting. Geophysical methods can detail the location, extent, and character of modified terrain.16 Surface geophysical surveys can be combined with magnetic analysis of soil and geochemical prospecting to give a good picture of the extent of the anthropogenic context of a buried site or feature. With the addition of more exten-

111

Discovery and Spatial Analyses

Figure 4.5 Paleotopography of Nichoria Even in areas of moderate relief, the pre-occupation and occupation topography of a site are likely to be quite different from that of the present because of human activities and erosion/deposition. This figure illus-

trates the paleotopography of a large site in southwestern Greece that was heavily modified during and subsequent to its occupation. (Data from Rapp and Aschenbrenner 1978)

sive core drilling and sediment analysis, the preoccupational topography and landscape evolution of a site can be determined (fig. 4.5). Until the 1970s, magnetic susceptibility and related techniques were not viewed as tools capable of investigating cultural landscapes. A notable exception was a study at Maiden Castle, an Iron Age hill fort near Bickerton, England. Magnetic analyses were used to determine whether charred wood within the ramparts had been burned before emplacement or in situ during or after construction.17 Magnetic analyses of soils and archaeological sediments are used to examine variations in soil-forming processes, including past weathering and climatic regimes; correlate stratigraphic

levels, sequences, and paleosols; and determine sediment sources. The approach to any geophysical survey depends on the surficial geology and the presence or absence of interfering systems. The choice of methods should be matched to both archaeological goals and geologic conditions. A geophysical survey can be one of the main techniques of site evaluation, not just a guide to locating archaeological sites of features. A note of caution: the complexity of most urban stratigraphy, combined with interfering materials, structures, and powerlines, presents a deterrent to geophysical surveying in areas of extensive human activity. Geophysical prospecting is both a broad

112

Discovery and Spatial Analyses

and a specialized topic that goes well beyond its geologic aspects. For a fuller introduction to geophysical methods, see the sources in the notes to this chapter.18 Surface geophysical methods depend on the presence of contrasting physical properties of target features and their surroundings, which enable scientists to interpret subsurface deposits. Once the local subsurface geology is understood, any deviation in geophysical response is considered an ‘‘anomaly.’’ Anomalies can be the result of archaeological features, unexpected geologic features, or postoccupation intrusions. To date, geophysical surveys have been used primarily to locate discrete archaeological features before excavation rather than to assist in landscape reconstruction. Exceptions include the work on the mound builders of the Amazon and the reconstruction of landscape modification at the Cahokia Mounds site in Illinois.19 Geophysical methods have become so diverse and pervasive in archaeology that it is impossible to review the field in any depth. Below are brief introductions to geoarchaeological aspects of four geophysical prospecting techniques that have been successful in archaeological contexts. Bruce Bevan and Anna C. Roosevelt have shown that interactive geophysical surveys and excavations furnished more information than either geophysics or excavation alone. Test excavations were conducted at geophysical anomalies.20 Rapp found this type of interaction to be highly fruitful at the Nichoria excavation in Greece in 1969.21 Archaeologists dealing with various smaller geomorphic forms such as mounds, hillocks, swells, and barrows need to determine whether they are natural or anthropogenic. Changes in geophysical properties due to cultural loading can serve as indicators of anthropogenic earthen forms.22 Cultural manipulation of sediments and soils change their density porosity, permeability, and cohesion. These changes affect geophysical responses. Several studies at three prehistoric mound sites in the United States (Cahokia Mounds, Effigy Mounds, and Hopeton Earthwork) illustrate the geophysical techniques that discriminate between cultural and natural origin. Geophysical surveys are now increasingly com-

mon in North America, as they have been in Europe.23 Geophysical surveys can produce detailed maps of subsurface features over large areas. Such maps provide primary information for the study of site content and boundaries, features, spatial relationships, and related aspects of landscape geoarchaeology.24 Ground-based geophysical surveys are becoming an integral part of Cultural Resource Management (CRM) efforts, particularly to guide archaeologists to (or away from) expensive excavations. Improved instrument sensitivity means that smaller, deeper, and less-prominent buried features can be recognized. Under reasonably ideal conditions, geophysical surveys can produce data sufficient for site description and inter-site analysis. Not all landscape contexts are amenable to these surveys. Rapp and Zhichun Jing found ground penetrating radar and magnetometry of little value in their regional survey in the Yellow River Plain, China, but systematic coring provided the data needed (see section on coring in this chapter). New instruments, acquisition and processing techniques, and more sophisticated interpretation seem likely to enhance field archaeology in the coming decades. Multicomponent geophysical studies using electrical resistivity, magnetic profiling, and ground-penetrating radar (GPR) were carried out on the eastern flank of the Pyramid of the Sun at Teotihuacán, Mexico.25 These studies were designed to study the direction of a buried tunnel beneath the western main entrance to the pyramid. Below the pyramid is a natural cave formation that was enlarged and used probably after a.d. 1. The paper details how the subsurface geologic features governed the conditions for and results of the studies. Another example is the resistivity, magnetometry, electromagnetic, and GPR surveys at the Fort Clark State Historic Site, North Dakota, along the Missouri River in the Northern Plains, United States.26 Magnetometry and Magnetic Properties of Soils and Sediments The magnetometer is the survey instrument that has the widest possible application. However, depending on the relative magnetic susceptibility of the subsurface materials, small or poorly magne-

113

Discovery and Spatial Analyses

tized features are not likely to be resolved if they are buried more than a meter or so deep. Magnetic anomalies show a tendency to broaden as they become more deeply buried by soils and alluvium. Magnetometry is based on the properties of the magnetic field of the earth, which can be measured at any geographic location. The earth’s magnetic field strength varies from about 50,000 to 70,000 gamma and is increased locally by magnetic materials. Baked clays, kiln walls, and other materials containing iron become magnetized when heated to a few hundred degrees Celsius. In North America the use of magnetic methods in archaeology goes back more than thirty years, to the survey at Angel Mounds in Indiana.27 Magnetometry surveys are usually employed to locate specific archaeological features on sites that have been identified by other means, such as a scatter of sherds on the surface. Positive magnetic anomalies result from such features as iron objects, fire pits, kilns, forges, storage pits with high humic content, burned features, bricks, and the foundations of historic buildings. Negative magnetic anomalies are usually associated with graves, middens, wells, and various prehistoric structures. Many anomalies may be dipolar. Magnetic contrast is the key to locating buried archaeological features. At the site of Sotira Kaminoudlua in Cyprus, Rapp was successful in locating buried (iron-free) limestone walls in a typical clayey soil matrix that had 5 percent or more iron, but he could not locate buried limestone walls that were resting on bedrock limestone in a high-lime soil. The most commonly used magnetometer in archaeological prospecting uses the precession of spinning protons in a fluid (usually kerosene, alcohol, or water) to measure magnetic field strength. Spinning protons create magnetic dipoles which align in a uniform magnetic field. In the absence of other fields, protons align with the earth’s field. The magnetometer first realigns the protons perpendicular to the earth’s field then allows them to realign parallel to the earth’s field. The protons precess with a frequency proportional to the total local magnetic field, providing a measurement of the field. Magnetic surveying cannot be undertaken in areas where the bedrock is volcanic or where there 114

are extensive powerlines or modern construction because these geologic or industrial materials have strong magnetic properties that will swamp out the weaker contrasts found in archaeological sites. A metal detector may be used to check for interferences produced by modern metallic waste. A distinct advantage of magnetic surveying is that it is not affected by the moisture content of the ground. The magnetic properties of minerals have been used in conjunction with magnetometry to determine the magnetic properties of site sediments, soils, and features and have been used with core drilling and sediment analysis to investigate anthropogenic soil-forming processes and landscape change at archaeological sites. The minerals that contribute most to the magnetic character of typical soils are hematite (alphaFe 2 O 3 ), maghemite (gamma-Fe 2 O 3 ), and magnetite (Fe 3 O 4 ). Many human activities can alter the magnetic character of soils. For example, hematite can be reduced to magnetite during heating in a reducing environment, which can occur in a hearth, resulting in a greater magnetic susceptibility. Offsite magnetic studies can also contribute key information. For example, the magnetic character of lake sediments can reflect local land use, climate, and fire frequency. Magnetic susceptibility (MS) is a measure of the ability of a substance to be magnetized. Two factors determine magnetic susceptibility in a soil sediment context: the presence of iron oxides inherited from parent rocks; and the degree to which these oxides have become enhanced by processes associated with anthropogenic activity, particularly burning. Magnetic-susceptibility studies contribute to a geophysical survey by providing information in support of the interpretation of magnetometer data. They may also be used as a separate prospecting technique. Magneticsusceptibility surveys can detect features of landscape change, anthropogenic disturbances like those resulting from agriculture, and bioturbation. These surveys are not usually adequate to pinpoint individual archaeological features but rather can provide a broader pattern that can be used in combination with other survey techniques. An MS survey should be conducted at a 10-m sampling interval followed by a magne-

Discovery and Spatial Analyses

tometer or resistivity survey of areas of magnetic enhancement. Magnetometers are particularly useful in historic sites, where shallow remains can be easily detected. Magnetometry can be an important technique for assessing the archaeological content of protected sites for Cultural Resource Management (CRM). For prehistoric North America the most common targets of magnetic prospecting have been hearths, fired rock, and pottery. Magnetic prospecting is more successful when the archaeological features are large. Hearths are theoretically easy to detect but may be missed if they are too small, and/or at too great a depth. Electrical Resistivity Electrical resistivity techniques that could be applied to archaeology were developed mainly in England during the 1950s. A detailed handbook of electrical resistivity surveying in archaeology has been prepared by Christopher Carr.28 In resistivity surveying, a series of metal probes is inserted into the ground at measured intervals along a surveyed traverse. A voltage is applied to outer probes, and the inner probes record the resultant current flow in the earth. The depth to which the current will penetrate the earth is one to one-and-a-half times the probe spacing. Resistivity measurements are suitable where there are soil or sediment contrasts involving differing water retention or dissolved ion concentrations—for example, ditches and pits. Historic architectural features like building foundations and house floors usually provide good electrical contrast. Resistivity methods have been successful in locating features that range from Roman walls to a Paleolithic flint mine in Hungary. Electrical resistivity is not an efficient method for regional or non–site specific surveys. Resistivity surveying should be favored where a strong electrical contrast with the enclosing soil sediment matrix is provided, such as building foundations, ditches, and defensive works. A resistivity survey presumes that such features exist in the survey area. Magnetometry and electricalresistivity survey methods complement each other. It is usually best to try magnetometry first, followed by a selected resistivity survey of significant magnetic anomalies.

Electrical resistivity surveying allows the researcher to control the depth of the investigation by varying the distance between the electrodes. The most effective method is to have previous knowledge (for example, by core drilling) of the approximate depth of archaeological features which generate the anomalies as well as the depth of the undisturbed geologic strata. Geologic strata display strong electrical conductivity contrasts among clays, silts, sands, and gravels and between weathered and unweathered rock. If the archaeological sediments are thick enough, one can set the probes close enough together to be free of the underlying geologic effects. This is important because geologic anomalies are most often one or more orders of magnitude stronger than archaeological anomalies and easily overwhelm the signal of the latter. Buried walls, tombs, and related features that restrict the flow of electrons result in resistivity peaks or maxima. Ancient pits, ditches, and the like that were later filled, even with sediment or soil from the surrounding area, will result in resistivity lows or minima because the material of the fill is more loosely packed and consequently retains more moisture. In most geoarchaeological situations, porosity, soil/sediment moisture, and the concentration of ions will govern both conductivity and resistivity in the subsurface. Uneven terrain presents serious difficulties to the geoarchaeologist using prospecting technology. In a depression the current density is constrained to rise. Conversely, over a mound the current can spread out, which causes the current density to fall. Because electrical currents in the ground are controlled largely by moisture content, present-day agricultural practices designed to retain soil moisture may have a strong effect on electrical resistivity measurements. Electrical resistivity profiles that run parallel to or close to sharp boundaries—for example, where flat hilltops fall off into steep valley sides or where parallel excavation trenches exist—will be subject to ‘‘edge effects’’ because the open air of the hillside or trench represents an infinitely resistant volume. The first resistivity survey in archaeology in the New World was carried out in 1938 at Colonial Williamsburg, Virginia, eastern United States, in 115

Discovery and Spatial Analyses

1938.29 It did locate a high-resistivity feature in a churchyard caused by a natural soil contrast. However, this survey was designed to locate a buried stone vault but failed to do so. Electrical resistivity surveying was used in Portugal to successfully detect Paleolithic activity areas containing hearths, fire-cracked rock, and artifacts.30 This study by Paul Thacker and Brooks Ellwood points out the need to understand the local lithostratigraphy and the geologic context of the archaeological remains. Electromagnetic Conductivity Electromagnetic (EM) methods were introduced in archaeology at the end of World War II, when surplus mine detectors became available. The later development of specialized electromagnetic methods was prompted by the desire to replace the electrical-resistivity method, which required good contact between the ground and the electrode system—a major source of problems. The EM conductivity method induces current flow in the ground without actual electrical contact, thus providing a means of traversing an area rapidly to determine changes in the conductivity of the terrain. The chief disadvantage of EM is that if there is only one instrument (with a given frequency range), the vertical resolution (depth dimension) is limited. Different instruments have specific vertical resolution limitations. For instance, the Geonics EM-31 has an effective depth penetration of about 6 m, while Geonics EM-38 has an effective penetration of 1.5 m. In contrast, with the electricalresistivity method the desired penetration can be obtained by varying the configuration and spacing of the electrodes. The Geonics EM-38 can measure both terrain conductivity and magnetic susceptibility. The EM-31 has the ability to reveal stratigraphy and structure (archaeological features, such as pits). Profiles obtained by an EM-38 often have a less clear signal than EM-31 profiles and record generally lower conductivity values. Variations in soil moisture and temperature can affect the conductivity to the point that one must correct for the effects of weather-related survey conditions when comparing results from various EM surveys.

116

Ground Penetrating Radar Magnetic and electromagnetic instruments rely on the presence of magnetic and conductive materials to produce anomalous responses.31 GPR records variations in the dielectric (nonconducting) properties of near-surface sediments and soils, which are usually caused by variation in moisture content. GPR can be used on frozen or snow-covered sites. In GPR, a short pulse of radio energy is generated from an antenna on or near the ground. The downward-moving pulse is partially reflected by any change in the bulk electrical properties of the ground. This change usually correlates with a change in volumetric water content and may indicate a change in bulk density. GPR has successfully located buried structures at many archaeological sites. The instruments are more costly than magnetometers, instruments for EM conductivity, or electrical-resistivity instruments, but the unique, high-resolution capabilities of the method make it an important tool in geoarchaeology. First used in the United States in the early 1970s, GPR is increasingly being used in archaeological prospecting. It works well where sharp dielectric discontinuities exist at buried walls, foundations, and floors. Concentrations of metals and bricks produce strong radar echoes. One advantage of GPR is that it provides fairly direct information on depth. Another advantage is that the results are relatively easy to interpret. One disadvantage is that conductive soils cause a strong attenuation of radar echoes. For example in dry, sandy soils or sediments, 100 MHz GPR can penetrate 15 m, whereas in wet, clayey soils the depth can be as little as 1 m. Increasing the frequency improves the resolution but decreases the depth of vertical penetration. On the positive side, such buried landscape features as river channels, as well as the depth of alluvium, can be determined using GPR. Traditionally, GPR has been successful at locating tombs, mine tunnels, and other voids. The extent to which walls or foundations can be detected depends on contrast with the enclosing matrix. On the negative side, waterlogged soils and sediments present a problem for GPR, owing to signal attenuation. In clay-rich horizons the useful

Discovery and Spatial Analyses

penetration of GPR can be less than half a meter. The use of longer pulses or lower frequencies can overcome this difficulty to some extent, but these tactics result in a tradeoff between depth of penetration and resolution. Similarly, higher-frequency signals are required to resolve small features. Since higher frequencies are more rapidly attenuated, as the depth explored increases so does the size of objects that can be detected. Therefore the choice of frequencies is dictated by the size and depth of the target archaeological feature and the nature of the subsurface matrix. GPR is most effective when applied to specific and localized problems. It is not efficient as a broad exploratory tool. Success using GPR requires favorable conditions. When these are met, GPR is fast and cost effective. It can be used in relation to historic buildings as well as field sites. For example, GPR was used to map the location of earlier structures beneath two churches in Italy.32 GPR surveying combined with enhanced computer processing techniques has been used for three-dimensional mapping of sites and features in the American Southwest.33 GPR has been successful in intrasite investigations, locating shallow graves, mapping historic-period fortifications, defining the outlines of features, and providing an immediate picture of the site stratigraphy. As the price of GPR instrumentation continues to decrease, this technique will become more commonplace in geophysical surveying. Topographic and geophysical surveying combined with sediment coring were used in exploration for the suggested Canal of Xerxes in northern Greece. The canal, reported by ancient writers, would have been across a 2-km isthmus. Resistivity, GPR, and seismic surveys were used to locate this buried structure. GPR detected successive infillings of a canal but not its original sides or bottom. Seismic measurements did provide decisive evidence of the canal with strong confirmation coming from the coring.34

Seismic Profiling Seismology takes its name from the Greek words for the study of earthquakes, and it is from such studies that much of the instrumentation evolved.

Seismic instruments record two types of seismic waves that reveal subsurface structures. One vibrates parallel to the direction of propagation, and the other vibrates perpendicular to it. The seismic disturbance of the ground in prospecting is produced either by controlled explosions or by impacting the earth with a heavy hammer hitting a steel plate. Seismic waves generated by such devices are reflected and/or refracted from subsurface layers with contrasting properties. When these reflected or refracted waves return to the surface, they can be recorded to provide an account of subsurface stratigraphy. Although seismic reflection methods are commonly used in geology for shallow depth prospecting, they have not been successful in archaeological feature location. However, seismic methods can be used to evaluate offshore coastal areas for archaeological site potential and to reconstruct the offshore paleogeomorphology. For example, archaeological sites dating to between 12,000 and 6,000 b.p. are likely to be found on the portion of the Gulf of Mexico’s continental shelf that was subaerially exposed during that period when sea levels were rising and then stabilizing to about their present position (see coastal settings in Chapter 3). Although current methods in highresolution seismic profiling cannot locate archaeological sites themselves, related geomorphic features such as river channels, bays, and lakes are easily detected. Regional studies have dated these geomorphic features to the period when human presence in the Gulf Coast is well documented. Possible locations that might be habitation areas can be tested by analysis of drill cores.35 High-resolution seismic reflection techniques can be used to image buried wooden shipwrecks. For example, wood buried in unconsolidated marine sediments can be readily imaged.36 Seismic surveys of the Mary Rose and the Invincible shipwrecks showed the areal extent of the sites to be greater than was apparent through direct observation. Marine geophysical techniques can be more broadly applied. The study of the location and evolution of ancient harbors is receiving increasing attention. Because Early Holocene coastlines shifted so rapidly and are now offshore

117

Discovery and Spatial Analyses

(due to the rise in sea level), marine geophysical techniques are required for investigations of these transgressions. High-resolution seismicreflection profiling has been used to reconstruct the postglacial transgressive shorelines in southwestern Greece.37 The shorelines of the Late Pleistocene and Holocene and their coastal environments were mapped. The sea-level rise from approximately −115 m to its current level covered a former geomorphic landscape of scarps, beaches, river channels, and lagoons. These are now buried under a few meters of post-transgression deposits. Some features can be roughly dated by reference to known sea-level rise curves; after 6,000 b.p. sparse archaeological data can provide a local sea-level rise curve (see fig. 3.18 for sea levels since 18,000 years b.p.). In a related study, sub-bottom seismic-reflection profiling was used to determine the position and nature of the Holocene embayment at Franchthi Cave in the Argolid, Greece.38 Franchthi Cave is an impressive karstic feature of about 150 m in length whose mouth lies approximately 15 m above the present shoreline. Excavation at the front of the cave revealed a succession of human occupations down to a depth of 10 m and covering a time span of more than 20,000 years— from the Upper Paleolithic through the Mesolithic to the Neolithic. A second site lies along the shore below the mouth of the cave. A number of Neolithic structures have been uncovered there. Offshore seismic investigations revealed that the site was probably much larger originally. This information, along with the related landscape reconstruction, indicates the extent to which marine geophysical techniques can be applied in geoarchaeology. Magnetic Analysis Magnetic properties of soils and sediments based on their ferromagnetic mineral content can reveal evidence of past human activities.39 Magnetic techniques are well developed, of high sensitivity, rapid, economical, and, in archaeological terms, relatively nondestructive. A magnetic susceptibility logger for archaeological application has been described.40 The ferromagnetic minerals involved may be authigenic, biogenic, or anthropo-

118

genic (by firing).41 One example is a medieval site in France where maghemite was the main ferromagnetic mineral in the archaeological sediments. Magnetic susceptibility is generally stronger in archaeological sediment than in the parent material. Burning experiments provide more accurate parameters for the interpretation of magnetic analyses. N. T. Linford and M. G. Canti conducted a series of experimental fires lit over a range of different soil materials (for example, sand versus clay) to examine the magnetic enhancement of the soils.42 The results demonstrate the sensitivity of iron minerals in the soil to surface burning, even at the relatively modest temperatures attained by the deeper samples. Hence, short-term campfires can provide sufficient heat to produce a magnetic anomaly through the production of both a magnetically enhanced ash layer and a thermal alteration of the underlying soil. Clare Peters and associates constructed experimental hearths and conducted controlled and repeated burning using different fuels.43 A range of mineral magnetic measurements was made on the resulting ash samples. The results indicate that the identification of ash from different fuel types can be established to aid in understanding site formation processes.

Aerial Photography Aerial photography is a valuable tool for discovering archaeological sites.44 Photography from tethered balloons predates the American Civil War. Civil engineers and soil scientists have used remote sensing for mapping since the 1930s. In fact, aerial photography for detecting buried features or features not visible from the ground was the earliest remote-sensing technique. Unfortunately, most of the aerial photographs available to archaeologists are made for other purposes and therefore have a limited usefulness in detecting archaeological features. Because a camera records everything it sees, the inventory of features of the earth’s surface found in an aerial photograph is far more complete than that presented on even the largest-scale map. This has its disadvantages as well as advantages. Maps are selective and explicit—therefore clear and easy to

Discovery and Spatial Analyses

understand. However, in many areas aerial photographs have replaced topographic maps as base maps for both geologic and archaeological fieldwork. The use of aerial photography in archaeology has had striking successes; the shadows cast by a low sun can highlight faint trails or roads, ditches, and mounds that are difficult to observe in a ground-level survey. The results have been even more spectacular in cases where all traces are invisible to the ground observer. Most aerial photographs are taken vertically (the camera aims straight down, perpendicular to the earth’s surface, rather than at an oblique angle). For many archaeological features, oblique photographs that are taken when the sun and shadows are at an optimal angle are more revealing. Archaeological aerial photographs may be used either for prospecting or for mapping. The two often require different techniques and different geometric accuracy. For most geologic and archaeological purposes, stereoscopic aerial photographs are preferred because the contrasts in topographic features stand out. Most features of archaeological interest have simple geometric shapes that are well known to archaeologists and geoarchaeologists. However, buried features may have shapes that differ significantly from those surface expressions that result from secondary differences in soil, vegetation, or plowing patterns. It takes training and experience to differentiate among geologic, pedologic, modern agricultural, modern cultural, and archaeological features in an aerial photograph. In fact, although aerial photography can be an important exploration technique, it would not be advisable to publish a site distribution map based solely on it. Repeated flights under varying conditions (for example, differences in season and time of day), combined with field surveys, are necessary to determine the efficiency of aerial photography as an exploration tool for a given area. Usually in aerial photography only black-andwhite or infrared-sensitive film that makes use of the visible light spectrum is used. Because visible light has little penetrating power, surface materials, particularly soils or vegetation, account for most of the variations in the imagery. Even for long-abandoned sites the disturbance of the

Figure 4.6 Effects of Buried Structures on Overlying Crops

surface is likely to have long-term effects that are observable in textural variations (which affect water retention—loosely packed soil retains more water) and color variation. The process of new soil formation can make these variations permanent. Many of them are not reflected in an archaeological section which exhibits differences through depth rather than surface patterns. Soil color is a function of the spectral reflectance of the material components of the soil. This reflectance is primarily a function of the moisture content, iron oxide content, organic matter, major soil minerals, and texture. Organic matter darkens the soil, iron oxide reddens it, and reflectance increases with decreasing particle size. Soil scientists have found that spectral-reflectance curves follow the standard soil classifications. Variations in Munsell soil colors are sometimes enhanced in black-and-white aerial photographs. The greatest contrasts are caused by differences in soil moisture, which is affected by grain size and grain-size distribution. Different types of buried features will result in different vegetation growth (fig. 4.6). In general, a buried structure will either enhance or re-

119

Discovery and Spatial Analyses

tard specific vegetation. For example, the starkness of the contrast in crop growth depends on the plants under cultivation. Plants with deeper roots may be less affected by differences in nearsurface moisture content. Vegetational markings are more common over buried ditches and pits than over buried walls. It should be noted that agricultural treatments like spraying for weed control will make artificial patterns that show up on aerial photographs. One example of the biases in data collection is an aerial reconnaissance in Scotland.45 A program looking for archaeological sites revealed as crop marks has targeted areas of known potential at the expense of ‘‘less rewarding’’ areas. The limitations as well as the biases in this program were analyzed. The problems arising in sample area selection in any archaeological survey are well known. They apply as well to aerial reconnaissance.

Satellite and Airborne Remote Sensing Imaging spectrometry is now practiced from spacecraft as well as aircraft. The wavelengths of the greatest interest in satellite remote sensing are the optical wavelengths from 0.30 to 1.5 μm. At these wavelengths, electromagnetic energy can be reflected and refracted from solid materials. Effective use of remote-sensing data requires a thorough knowledge of the spatial characteristics of the various earth-surface features and the factors that influence these spatial characteristics. LANDSAT data can provide the information for predictive models in archaeology.46 Geologic and ecologic variables form the basis for such models. Aside from spatial features, spectral data can reveal the chemical nature of earth-surface materials. Spectral data from satellites now monitor crop growth throughout the world as each growing season progresses. Spectral-reflectance curves from soil are less complex than those from vegetation. One of the major reflectance characteristics of dry soil is a generally increasing level of reflectance with increasing wavelength, particularly in the visible and near-infrared portions of the spectrum. The moisture content, percentage of organic matter, percentage of iron oxide, and ratio of clay/silt/sand (textural properties) all influence the spectral reflectance of soils. The in-

120

crease of moisture will cause a decreased reflectance. The percentage of organics will have the opposite effect. Thermal infrared images depict the pattern of heat that is emitted or reflected by target materials. The thermal characteristics of earth-surface materials yield information that is not detectable in other regions of the electromagnetic spectrum. Thermal contrasts are indirect indicators of moisture content or heat capacity and direct indicators of heat radiating from volcanic or geothermal phenomena. The strong, solar-induced diurnal temperature flux must be taken into account in projects using thermal infrared imagery. Water has the highest heat capacity (and therefore heat retention) of all earth-surface materials, so it is easily detected in post-sunset imagery. The best resolution available to geoarchaeologists from satellite sensors comes from the 10 m Satellite Positioning and Tracking (SPOT) imagery. This is too coarse for most site-specific archaeological applications, but it is valuable for assessing current and past landscape features. Much finer resolution, down to 2 m or less, can be obtained from sensors mounted on aircraft. Until the early 1980s the main limitation with satellite remote sensing was that no subsurface information could be acquired. Orbital imaging radar can now provide subsurface data in arid regions. The buried river valleys of the southeastern Sahara are systems of aggraded valleys that were first formed in the Middle Tertiary, containing inset drainage channels that have been entirely obscured by windblown sand. These features, first recorded as radar images, are remnants of a moister landscape in the Pleistocene and Early Holocene. Acheulian artifacts can be found in the alluvium that fills these old valleys. The imaging radar carried on the space shuttle Columbia in 1981 penetrated the extremely dry sands of the eastern Sahara, revealing previously unknown buried valleys.47 Sand-filled and alluvium-filled valleys, some nearly as wide as the Nile, were brought to light by radar images. Wadis superimposed on the large valleys provided sites for episodic early human occupation. These ancient drainage networks offer a geologic explanation for the location of current oases (fig. 4.7).

Discovery and Spatial Analyses

Figure 4.7 Radar Rivers Remote sensing using radar. Radar images have been used to view geologic and archaeological features not easily detected using standard surveying techniques. In this case, the pathway of the radar image provided information on the desert region of northern Sudan and

southern Egypt in northeast Africa. The radar images showed the location of ancient stream channels. The locations of the ‘‘radar rivers’’ have been compared to the spatial distribution of archaeological sites. (Based on McCauley et al. 1986)

An airborne radar survey in 1978 and 1979 over the dense rain forest of Guatemala led to the discovery of an elaborate network of Mayan canals from the Classic period, dug apparently between 250 b.c.e. and a.d. 900. The canals extend over a vast region of swampy jungle and are thought to be the basis for extensive lowland Mayan agriculture. The Maya were known to have dug canals in the arid highlands, but this was the first evidence of an extensive canal network in the lowlands. A technique related to orbital imaging radar is Side-Looking Airborne Radar (SLAR) imagery, available from the U.S. Geological Survey. SLAR is applicable to land-use, cartographic, and groundwater studies. The SLAR system has

an active sensor, providing its own microwave energy. It also has cloud-penetration capability, which allows it to collect imagery in situations where the conventional aerial photograph is inadequate, such as in the rain forests of Brazil. SLAR imagery presents an obliquely illuminated view of the terrain that enhances subtle surface features. More than 25 million km 2 of SLAR data have been gathered in the Western Hemisphere.

Dowsing Before we leave the topic of remote sensing we would like to comment on dowsing. Unfortunately, among the general public and to some extent even among archaeologists, there is a belief in

121

Discovery and Spatial Analyses

the abilities of dowsers to locate buried archaeological features. Martijn Van Lusen has reviewed the nature of such beliefs in published material from professional archaeologists in the United Kingdom.48 Archaeologists undertook field tests but ignored statistical biases, lacked training in controlled test design, and apparently held prior beliefs in the validity of dowsing. Van Lusen quotes G. Gaffney and associates, ‘‘This technique has long been practiced by archaeologists.’’ 49 Fortunately, we are not aware of any use of dowsing by geoarchaeologists.

Geochemical Prospecting and Analysis Geochemical prospecting in archaeology is generally analogous to geophysical prospecting. Based on a grid system and samples recovered from coring, a three-dimensional picture of the anthropogenic biogeochemistry can be developed (fig. 4.8). More focused studies can also be undertaken to locate and delineate graves, refuse areas, and agricultural plots. To understand the humaninduced biogeochemical impact, similar analyses must be made of nearby (offsite) profiles for comparison.50 Most chemical nutrients required for life on land are supplied from the soil. Nitrogen, oxygen, and carbon dioxide come from the atmosphere, and water is taken from the hydrosphere. Human activities, even in nonagricultural societies, alter the levels of micro and macro plant nutrients. Macronutrients are the chemical elements nitrogen, phosphorus, potassium, calcium, magnesium, and sulfur. Because all plants use these elements, the removal of vegetation by human activities depletes their concentration in soils and underlying sediments. Substantial amounts of nitrogen, phosphorus, and calcium are added to the soil by food wastes and human and animal wastes. Wood burning raises the amount of magnesium in the soil, and a high pH may be related to fire. Of special interest in archaeology is the biogeochemistry of phosphorus. Separate analytical techniques can distinguish three distinct phosphate fractions: (1) easily extractable, mainly alu-

122

minum and iron phosphate, associated with growing plants (including crops); (2) more tightly bound phosphate, commonly associated with human activity; and (3) natural geologic phosphate.51 Total phosphate concentration of more than 2,000 ppm (parts per million) indicates a burial. Interpretations aside, analytical methods for phosphate determination in both the field and laboratory continue to be improved. Richard Terry and associates report that their methodology for acid—extractable phosphorus concentrations under primitive field conditions (in Guatemala)—deviated by only 7 percent from those made under controlled laboratory conditions.52 At this same site, phosphate and trace metal concentrations were used as indicators of ancient activities.53 It was found that elevated phosphate, barium, and manganese levels indicated areas of organic refuse disposal. Mercury and lead concentrations indicated areas of craft production. These analytical data were compared to and consistent with artifact data. An example of the use of inorganic phosphates in archaeological interpretation is the study of the Caddon (a.d. 1250–1400) Huntsville site in Arkansas.54 The measurement of phosphate levels of a living floor within Mound A at the site was used to delineate activity areas and the boundaries of features. The results of the analysis of the spatial distribution of phosphate intensities were used to argue that the low levels of inorganic phosphates in the sediments indicated their use for nonintensive or possibly ceremonial purposes. The conclusion was that the phosphate values supported the idea that these areas were not the location of intense human habitation and were ceremonial centers that were prepared by clearing.55 Chemical prospecting to locate sites, as distinct from burials, is inefficient but useful in determining horizontal and vertical boundaries of known sites and features within sites.56 Geochemical analyses of sediments provide a broad range of possibilities for recovering information on technological activities. Mark Abbott and Alexander Wolfe have used abundances of lead, antimony, bismuth, silver, and tin in lake sediments near the major silver deposit in Cerro Rico de Potosi to illuminate

Discovery and Spatial Analyses

Figure 4.8 Phosphate Zones Spatial distribution and relative intensities of biochemicals at an archaeological site from the North American Midwest. This example depicts relative values of phosphate from an archaeological site along the Mississippi

River. Phosphate values can reflect human activity that is not observable because of the absence of artifacts or traditional types of archaeological features. (Hill, unpublished data)

Pre-Incan metallurgy in the Bolivian Andes.57 These elements, associated with smelting, are highly enriched in both the terminal stages of the Tiwanaku culture (a.d. 1000–1200) and Inca times through the Colonial Period (a.d. 1400– 1650). Soil phosphate studies are now common on archaeological sites as a reconnaissance tool or to investigate activity areas (fig. 4.8). However, the interpretation of soil phosphate data is not always straightforward. J. Crowther points to and details many of the difficulties, including natural background variation in phosphate concentra-

tion, spatial variation in phosphate retention capacity, vertical variation with soil profiles, and the effects of recent phosphate inputs from fertilizers and grazing animals.58 Phosphate analysis is most useful when integrated with soil magnetic studies and geophysical surveys. The phosphorus signature in anthropogenically altered soils can only be removed by the erosion of the soil. Organic geochemical analyses are used increasingly to indicate ancient activities. Lipid components in the soil at a Minoan site in Crete were analyzed to determine whether the practice of manuring had been employed in antiquity.59

123

Discovery and Spatial Analyses

Analysis of total organic carbon, certain lipids, and sterol components detected the practice of manuring. The manuring had been inferred by sherd scatter. J. Lambert provides a survey of geochemical methods for determining human activities.60 P. Bethell and I. Máté provided a historical review of phosphate analyses in archaeological surveying.61 The authors conclude that there are yet ‘‘major theoretical problems’’ in the archaeological interpretation of the results of phosphate analysis. Eight chemical elements are considered important micronutrients (nutrients that are required in smaller amounts than macronutrients). These are: iron, manganese, zinc, copper, boron, molybdenum, cobalt, and chlorine. One of the ways human activities alter the soil environment is by adding trace amounts of metals and hydrocarbons. These chemicals can depress some nutrients to the point of deficiency or augment others to the point of toxicity. Most of these chemical activities leave clear records in archaeological sediments because trace elements released from anthropogenic sources become part of normal biogeochemical processes. Toxicities are recorded in human paleopathologies. Examples can be seen in the processing and use of lead and arsenic. In historic archaeology the burning of coal and the smelting of iron and nonferrous metals have contributed a heavy load of trace-metal contaminants like mercury and cadmium to the environment. Soils are geochemical sinks for contaminants, and this contamination is often permanent, providing a geoarchaeological record of human activities. Pollution aside, there have always been geographic patterns in deaths that are related to the toxic elements in local bedrock that find their way into soil and water. The trace-metal content of soils and plants varies widely in different geologic provinces. A well-known historic example is the selenium poisoning that has occurred over the past 150 years in parts of Wyoming and South Dakota. Selenium deficiency in soils and the plants growing on them can pose an equal danger to grazing animals like cattle. Unfortunately, only a few of these toxic and/or nutrient elements are retained in bone, teeth, and hair, where they would be recoverable in archaeologi-

124

cal contexts. As recovery methods in archaeology improve, and as biochemical and geochemical analyses become possible on smaller and smaller quantities of material, additional information on ancient geochemical environments and their impact on human society will become available. In addition to nutrient elements, geochemical analyses that are helpful for understanding site stratigraphy, sedimentology, and anthropogenic impacts include studies of organic matter and carbonates as well as pH measurements. Organic matter and carbonate analyses provide indications of activity and refuse areas and site boundaries. Total organic carbon and carbonate can be measured together in a simple loss-onignition method.62 When plotted in conjunction with stratigraphic profiles, these data can help us understand mixed layers, define features and boundaries, display changes through time, and unravel soil horizons and processes. Stable-isotope analyses of soils and sediments also provide evidence of human activities. Physical and biologic processes, as distinct from strictly chemical processes, fractionate isotopes. Where human activities interact with and alter physical or biologic processes, there will be a record for geoarchaeologists. An example would be sulfurisotope fractionation in coal burning and copper smelting. Sulfur has four stable isotopes: 32 S, 33 S, 34 S, and 36 S. These have natural abundances of approximately 95 percent, 0.75 percent, 4.21 percent, and 0.02 percent, respectively. When humans use raw materials that formed deep within the earth and process them by burning or smelting, they alter the existing sulfur-isotope ratios, particularly the 34 S/ 32 S ratio, depending on the source of the raw material. Soils and sediments encapsulate nearly all of the archaeological record, including the new and exciting finds of ancient DNA. Human DNA is now recovered from bones and teeth at sites where the preservation conditions are favorable. Genetic analyses have revealed that plant and animal DNA can be preserved for long periods of time, even in the absence of obvious macrofossils. Samples ranging from 400,000 to 10,000 years old from Siberian permafrost cores have been recovered that contain the DNA sequences

Discovery and Spatial Analyses

from mammoth, bison, horse, and at least nineteen plant taxa.63 Temperate cave sediments from New Zealand yielded DNA sequences of extinct biota, including two species of moa bird from the prehuman period. Currently, the main and very serious problem with ancient DNA is contamination. Geoarchaeologists are likely to have a role in sample collection. Handling remains for DNA analysis with bare hands will contaminate those remains. Handling DNA samples requires a whole new protocol—study up before the situation arises. There is no one sampling strategy in geoarchaeology. Sampling is governed by the questions asked. A good knowledge of the local soils, sediments, and microstratigraphy is essential in devising a sampling strategy. Recently there has been excitement about the recovery of DNA where no macro remains were observable. Sampling for ancient DNA will require special considerations. DNA in sediment is likely to pass through a sieve and be lost. A DNA sampling strategy would include sampling organic carbon as a partial guide, realizing that paleosols are vastly more likely to contain ancient DNA than rapidly deposited flood sediments; volcanic ash would be a poor candidate; highly disturbed sediments are likely to be contaminated; and so forth.

Core Drilling Core drilling has many useful applications in geoarchaeology. In regional environmental reconstruction, especially in areas subject to rapid change like coastal and riverine environments, core drilling is a critical technique for recovering data. Archaeological excavation can be slow and costly. To determine the nature, depth, and extent of habitation, drilling, along with sediment/soil analysis and perhaps geophysical prospecting methods, can be quick, minimally destructive, and cost-effective. It is too little used in archaeological studies, but it is starting to be used more frequently.64 Drilling, coring, and augering each provide different kinds of information. A distinction should be made between methods that recover a con-

tinuous (or nearly continuous) core and methods that merely bring up sediments in the sequence encountered. Although there is some distortion, especially compaction, core-recovery methods give far more comprehensive and undisturbed information than augers, which propel the material to the surface by means of an Archimedes screw. In the latter technique there is general mixing of the sediments, although gross stratigraphic relations are retained. Core recovery uses a hollow cylinder to bring a relatively undisturbed section of the subsurface stratigraphy to the surface. In coring, precise vertical control is sometimes difficult because sediments become compacted within the core recovery tube. In the worst case the top sediments ‘‘freeze’’ in the bottom and ‘‘rodding’’ occurs (that is, the core tube acts like a solid rod that pushes through the sediments without recovering any). In much of the world the water table is within a few meters of the surface. Trenching beneath the water table is impractical at best (requiring massive pumping to lower the water table in the area of the site). But core recovery is not materially affected by the water table, although wellsorted wet sand is difficult to recover without special equipment. Many types of instruments are available for coring, ranging from hand corers (fig. 4.9) to vibracorers, truck- or trailer-mounted Giddings corers and large rotary drill rigs. Each has its place, depending on the nature of the problem and the resources available. As the environmental and geomorphic components of culturalresource management increase, the use of coring techniques to recover subsurface information will become mandatory. Core drilling was employed by Zhichun Jing and his colleagues to reconstruct landscape change in relation to a sequence of archaeological habitation levels and soil development in the Yellow River Plain in China.65 On the basis of stratigraphy and sedimentology, a Holocene landscapeevolution model was constructed (fig. 4.10). The prolonged landscape stability from the very Late Pleistocene or Early Holocene to about 2,000 b.p. provided Neolithic and Bronze Age human inhabitants with a favorable environment. After 2,000 b.p., the hydrologic regime changed, and

125

Discovery and Spatial Analyses

Figure 4.9 Core Drilling Equipment

the floodplain experienced 2–3 m of gradual vertical accretion during the next millennium. In response to a dramatic change in hydrologic regime after the early twelfth century a.d., overbank deposition covered the floodplain by as much as 10 m of younger alluvium. This had a pronounced effect on the preservation, visibility, and discovery of the ancient dynastic sites. Although the phrase landscape stability is used to refer to a period of very slow change that allows good soil formation, landscapes are continuously changing. Geoarchaeologists, even when they are dealing entirely with Holocene sites, must understand the entire Late Quaternary geomorphic history of the region where they work. Many geologic processes throughout the entire Pleistocene have contributed to the relevant geomorphic and sedimentary record. The Thames River, currently flowing through the center of London, has provided a generous sequence of alluvial terraces in response to the succession of Pleistocene climatic events. In

126

southern England, the Thames meandered across a wide area.66 Sometimes natural exposures in pits and river banks take the place of drill cores. Studies of the pebble lithology of gravel pits and of heavy mineral content in river bank sediments helped determine the former courses of the Thames. With the wealth of Paleolithic activity in the Thames drainage, the roving course of the river was a major determinant in settlement patterning. In the late 1930s, archaeologists in the United States began systematic archeological surveys (mapping the locations of the sites) to supplement excavation data. By the 1960s, there was extensive discussion about how to sample large areas for surveying where intensive and complete coverage was impractical (see Chapter 9). Reconnaissance survey gave only a limited view of the regional picture and could not be used in a statistical sense. It required an intensive and carefully controlled archaeological survey to provide an inventory of the nature and distribution of sites.

Discovery and Spatial Analyses

Figure 4.10 Holocene Landscape Evolution Model of the Yellow River Plain, China. (Based on Jing et al. 1997)

Even the intensive survey lacked two features: (1) it was a two-dimensional survey with archaeologists walking over the surface of the land, and (2) it recovered only certain types of strictly archaeological information, with limited attention paid to ecologic and geomorphic components. Broader multidisciplinary surveys have evolved to include ecologic and geologic components. To overcome the limited two-dimensional information obtained by traditional archaeological survey, which cannot locate deeply buried sites, Rapp and colleagues have developed a three-

dimensional survey by systematic coring. Extensive coring has been used to determine Holocene coastal change at archaeological sites in the eastern Mediterranean area.67 These projects provided the sequence of coastal depositional environments and associated chronologies. Important sites were buried 15 m below the present land surface. Some ancient harbors are now more than 10 km from the sea. In the eastern Mediterranean area basically three types of coring units were used: a vibracorer, a truck-mounted rotary drill, and a manual corer (the Dutch auger).

127

Discovery and Spatial Analyses

In the early 1990s, Jing and Rapp used core drilling to determine Neolithic, Bronze Age, and later landscape evolution in the Shangqiu area, China.68 In the Shangqiu area the early Shang layers from the second millennium b.c.e. are buried approximately 10 m below the surface. The record from the drill cores was crucial for evaluating and interpreting, as well as predicting, buried archaeological sites. The Shangqiu region lies on the floodplain of the lower Yellow River. Its archaeological importance lies in its potential location as the center of predynastic Shang cultures. At Shangqiu, they used a Dutch auger but switched to the Luoyang spade (the traditional tool used to detect buried cultural remains in China) to increase the number of core holes per unit time. Coring in this region led to the discovery of a major Zhou city, buried many meters below the surface. More recently, Jing, Tang, and Rapp used core drilling in the discovery of Huanbei Shang City near Anyang. A regional archaeological survey was begun at Anyang in the fall of 1996, undertaken jointly by the Institute of Archaeology of the Chinese Academy of Social Sciences and the Archaeometry Laboratory at the University of Minnesota, Duluth. The survey paid a great deal of attention to the investigation of the Middle Shang settlement pattern because it would provide critical data on the emergence of the last Shang capital at Yinxu. A subsurface archaeological survey based on coring was explicitly designed in this project. It became obvious that the Dutch auger method was too slow to be useful; however, the use of multiple Luoyang spades by welltrained Chinese operators was extremely successful. It was survey coring with the Luoyang spade that led to the discovery of Huanbei Shang City. More than two thousand cores were drilled. In regions of major alluvial sedimentation, many or most early archaeological sites will be covered to a depth that makes the sites invisible to surface archaeological survey. Therefore, some other method of finding deeply buried sites is needed. In some cases, ground-based geophysical exploration methods (electrical resistivity, magnetometer, or GPR) might be useful, but coring has proved to be very successful. The drill cores are logged in the field, including the record128

ing of the variations in Munsell color. Then the cores are sampled for radiocarbon dating where appropriate and for organic carbon (a good indicator of Holocene paleosols), grain size analysis, and X-ray diffraction (for identification of clay minerals). Direct evidence of paleovegetation patterns can be obtained by analyzing paleosols for macrobotanic, pollen, and phytolith remains. Phosphate is the best chemical indicator of human and animal activity. Environmental magnetic techniques (anhysteretic remnant magnetism and low-field magnetic susceptibility) provide additional parameters for characterizing site formation and anthropogenic information. Analyses of drill core sediments to identify paleosols and plant/animal/human activity provide the major framework for predicting and locating deeply buried sites. Using coring data in young sediments that were folded by thrusting on the Blind Puente Hills fault beneath Los Angeles, California, James Dolan and associates have documented at least four large (magnitude 7.2–7.5) earthquakes during the past 11,000 years.69 C. Vance Haynes used coring in geoarchaeological investigations at the Clovistype site in New Mexico.70 Six lines of core holes that were transverse to the outlet channel defined the subsurface configuration and stratigraphy of the ancient spring run. The precise plotting of the microstratigraphy permitted a more detailed interpretation of depositional processes and paleoclimate changes than previously possible. Studies of seventy cores around ancient Pompeii have allowed the reconstruction of Holocene environments prior to the a.d. 79 volcanic eruption that buried Pompeii, Herculaneum, and Stabiae and caused a shoreline progradation of 1 km.71 At the Holocene transgressive maximum, the sea reached an area more than 2 km east of ancient Pompeii. Pompeii was built as the shoreline prograded toward its present position. The coastal plain is today about 1.5 km wider than it was during Roman times. Core drilling is one of the most useful methods in geoarchaeology. Archaeological excavations are often slow and costly. The nature, depth, and extent of a site can be determined by coring with the associated laboratory analyses. For example, coring with the Luoyang spade (or similar

Discovery and Spatial Analyses

equipment) is a rapid, minimally destructive, and cost-effective technique for testing deposits for buried archaeological components and features. We strongly recommend the application of coring for the discovery and evaluation of archaeological contexts.

Locating Water Resources Humans require two to three liters of water each day. Succulents, melons, and some other plants can provide some moisture but regular sources of potable water must be available for habitation. Such sources may include rivers, lakes, springs and seeps, and wells. A major study may need to be undertaken in the geoarchaeological study of a site if the source(s) of potable water is not immediately apparent. Civilizations have flourished when reliable water supplies were available and then collapsed when these supplies failed. In dry regions the penalties for the failure of the water supply could be severe: ‘‘If anyone be too lazy to keep his dam in proper condition, and does not keep it so; if then the dam breaks and all the fields are flooded, then shall he in whose dam the break occurred be sold for money and the money shall replace the corn which he has caused to be ruined’’ (Code of Hammurabi, Section 53 [1760 b.c.e.]).’’ Thick jungle covered much of the region of Maya civilization in the Yucatan and karst topography rendered surface streams totally absent. To ensure a year-round source of water, the Maya constructed deep wells, cisterns, and clay-lined storage depressions. In every region where potable water supplies failed during droughts, societies sought alternate sources. At some point this sometimes included digging wells. Such excavation might well have been located where there had been surface water shortly before. C. Vance Haynes and colleagues have recorded the oldest water well in the New World, at a Clovis site dating to around 11,500 b.p. 72 The feature was a cylindrical pit dug in an attempt to reach shallow ground water. Other wells have been discovered in the desert southwestern United States probably indicating a lowered water table during droughts. In the Old World, Yaacov Nir has excavated wells in Israel

that are as early as the Pre-Pottery Neolithic, approximately 8,000 years b.p. 73 Research on water quality in Olduvai Gorge shows that early hominids gathered around freshwater sources where scavenging opportunities were greater.74 This study indicates where geoarchaeologists may have to turn for evidence of early environments. The research showed that the clay geochemistry correlated well with artifact abundances. The geochemistry indicated leaching of lacustrine claystones beneath freshwater wetlands following lake retreat. The exploitation of land and water resources to sustain population increases involves changes to regional hydrologic regimes. Major land-use changes disrupt the hydrologic cycle, altering the runoff and the evaporation. The introduction of human-made drainage systems can cause wideranging effects in flood volumes and alter even the ground water hydrology. Attempts by geoarchaeologists to account for environmental changes need to take into consideration how human activities have impacted the regional hydrologic regime. Water was a major factor in cultural adaptation to arid Altithermal (7,500–5,000 b.p.) conditions in the North American Great Plains. There is a statistically significant relation between Early Archaic site locations and groundwater sources not apparent before or after the Altithermal.75 Surface water resources from runoff became scarce in the semiarid plains, leading peoples to gravitate to water sources derived from groundwater aquifers. The Great Plains overlie vast sources of such dependable water supplies. These aquifers frequently discharge into perennial streams. In drier conditions, Early Archaic peoples settled in areas of aquifer-replenished streams and springs rather than playa lakes and streams supplied by rainfall runoff. In addition to supplying water for human consumption, these dependable sources attracted game animals harvested by local area inhabitants.

Geographic Information Systems The variety, nature, and volume of geologic data have increased markedly over the past few decades. It has become essential to use databases 129

Discovery and Spatial Analyses

and Geographic Information Systems (GIS) to turn these data into usable information.76 In developing geoarchaeological databases and modeling, it is necessary to separate the recording of descriptive features (empirical data) from processes (concepts). Geoarchaeological databases are four-dimensional—they can involve variations in shape and the relations and distribution of spatial objects through time. A traditional geologic map conveyed an understanding of the geology of an area. In like fashion, a topographic map conveyed geomorphic features. With the advent of GIS and digital databases, a range of new possibilities for data analysis emerges. Many of the data from drill cores that do not find direct expression on the traditional geologic map can be stored and used in regional analyses, taking advantage of GIS and database-analysis software. In terms of the development of geologic databases and GIS, it is important to be aware of the rapidly changing possibilities in the analysis and presentation of geoarchaeological data. Geologic, paleontologic, ecologic, pedologic, hydrologic, climatic, geographic, topographic, and archaeological data can all be integrated within one GIS set.77 GIS provides geoarchaeologists with important tools allowing the processing of vast amounts of information, promoting new possibilities in pattern recognition, landscape analysis, and aiding data visualization not possible before. With the incorporation of more appropriate scales now available through satellite remote sensing, future prospects for GIS seem enormous.78 Software for GIS analysis manipulates data in either a raster or vector format. Rasters are grid cells of a specified dimension, and raster-based software (for example GRASS) is particularly useful in manipulating spatial data. Vector-based software (for example ARC/INFO) is mainly used for linear data, such as drainage networks. Implementing an appropriate GIS should be one of the first undertakings in any major archaeological project. Archaeology is concerned with the spatial arrangement of features and their relation to other parameters (time, climate, soils, land use). The display and manipulation of such data sets are the function of GIS, which evolved as a

130

means of assembling and analyzing diverse spatial data. Most remote-sensing data are fed directly into a GIS. Digital image-processing techniques are an integral component of GIS. Powerful patternrecognition techniques are also available in GIS, but these usually require previous knowledge (ground truth) of the basic elements used in the image analysis. A key application of remotely sensed data is for classification: organizing data into discrete categories in terms of land use or vegetation types, for example.79 For remote regions where few maps are available, the use of satellite imagery combined with GPS and GIS can provide a powerful methodology for archaeological research. National Aeronautics and Space Administration’s (NASA) LANDSAT imagery and GPS data were used to construct a GIS of the Middle Holocene environment to study the roots of agriculture in southern Arabia.80 They were able to document archaeological site locations, organize and analyze field data, perform classifications, and generate maps. NASA’s Web site has images emphasizing how geomorphology and geology influence environmental change. It also has a wealth of geospacial, geographic, topographic, climatic, and land use data that is of interest to geoarchaeologists.

The Complexities of Scale Geologists and archaeologists operate on quite different scales of time and space. Archaeologists work in human timescales of years, decades, or centuries, whereas even those geologists who work exclusively in the Quaternary may deal with more than two million years. Geographically, archaeologists usually focus on a few m 2 in an excavation or a few km 2 in a survey. Most field geologic problems encompass a much larger area. Similarly, in stratigraphy, archaeologists need higher resolution than most geologists require. Geoarchaeologists must bridge these gaps. When studying artifacts, archaeologists are usually measuring objects in centimeters. Geochemists investigating the same artifacts are generally making measurements at the scale of

Discovery and Spatial Analyses

atoms. How chemically homogeneous or inhomogeneous an artifact is can be critical in determining the technology of its manufacture. Geoarchaeologists must work with both scales. Archaeologists, geologists, geographers, ecologists, and related researchers use scales appropriate to the resolution desired for the problem at hand. Some individuals work only, or at least best, at certain scales. A geomorphologist who participated with Rapp on a regional archeological survey was asked to participate in major excavation in the survey area. He declined because working in a small trench was considered an inappropriate scale for him. Most geoarchaeologists, over time, will have to work with a wide variety of scales. Geoarchaeologists whose training is exclusively in the geosciences may come to an archaeological problem with little understanding of the scale of the problem. Julie Stein calls attention to many aspects of picking appropriate scales.81 Differences among disciplines in the scales used for interpretation can cause problems. Geologists use scales appropriate for the whole earth. Archaeological scales vary from those used in regional surveys to those for profiling the walls of a small excavation trench or documenting the location of an artifact or feature. Archaeologists synthesizing some aspect of Late Paleolithic stone tools may think globally, but this is not the usual practice. Scholars studying any aspect of earth history, such as archaeologists and historians, must deal with spatial and temporal scales. Scales are not only important in data collection and analyses but also in publication. Every map should include a scale and usually a north arrow or some indication of global context. For most of its history, archaeology has dealt with artifacts and features on the scale of centimeters and meters. Today, archaeological science has moved archaeology into microscopic and even submicroscopic scales. The scale at which data are collected is not necessarily the scale at which the interpretations are made. In provenance studies based on trace elements, the analytical data are collected at the atomic levels but the interpretations may be at a continental scale. Geologists most often work at temporal

scales of hundreds, thousands, or even millions of years. However, especially for Quaternary geologists, deposits from flash floods, earthquake destruction and faulting, tsunami debris accumulation, volcanic ash falls, and related phenomenon take place on minute to day scales. Many aspects of the effects of scale on archaeological or geoscientific perspectives are presented in a volume edited by Julie Stein and A. R. Linse.82 The scale of a map controls what features it can adequately represent. Few geologic features would need to be shown at a scale of 1:20—a common scale for representing archaeological features. The detail, accuracy, and method of mapping will all change with the scale at which the mapping is carried out. The smaller the scale, the greater the amount of interpretive, as opposed to factual, information on the map. The relation between scale, content, and clarity is fundamental. Each separate geoarchaeological problem requires a careful consideration of scale for maximum clarity. For topographic maps, the landscape to be represented may dictate the scale. Depending on how much detail is being shown, areas with rugged or broken relief will require a larger scale than a featureless plain. The same applies to contour interval. Scale on an aerial photograph is not as explicit as scale on a map. The scale on an aerial photograph will differ somewhat from place to place on the photograph. As historic sciences, archaeology and geology must deal with resolution in timescales. Again, archaeologists typically deal with fairly highresolution timescales. Specialists in the Late Bronze Age pottery of Greece believe that the ceramic chronologies they have created are correct to plus or minus thirty years. For the same period, 14 C dating would be on the order of plus or minus ninety years. Purely geologic dating for this period would be much less precise (see Chapter 5). Soil scientists and archaeologists use similar temporal and spatial scales. In most cases soils develop and stabilize on a scale similar to that of evolutionary and cultural changes in the human sphere. Anthrosols, in particular, are keyed to the human timescale.

131

CHAPTER 5

Estimating Time

A basic question that must be answered for any archaeological site is, how old is it? —Troy Pewe 1954

E

stimating the age of archaeological materials and Quaternary strata is one of geoarchaeology’s primary tasks. Chronology provides the temporal dimension that distinguishes the historic natural-science disciplines of geology, paleontology, and archaeology from disciplines like ethnography and ethology that focus on presentday processes. An understanding of the methods used to date the past is critical to any interpretation of the archaeological record. Before the development of chronometric techniques, temporal control depended either on relative age estimates that were based upon intrinsic features of artifacts or on the correlation of artifacts with evidence for environmental or climatic change. Although these methods are still fundamental, the application of chronometric techniques has considerably affected our understanding of the prehistoric record. Increasingly better-defined age estimates have played a critical role in the development and testing of ideas about past human behavior. ‘‘Relative’’ methods of dating have been employed throughout the history of prehistoric research. Most ‘‘absolute’’ methods of dating were developed during the second half of the twentieth century. Relative dating can be used to determine the temporal order or sequence of events associated with artifacts. In contrast, absolute dating provides age estimates that can be expressed in standard time increments (usually years). Absolute-dating techniques are based on physical or chemical properties or processes that

132

can be measured. Although some techniques described below can be used to date artifactual and fossil materials directly, often one dates the geoarchaeological context—the deposits associated with specific archaeological materials—thus providing an indirect estimate of the age of past human events. In all types of dating that can be applied to archaeological situations, the most critical factors are the existence of a time-dependent quantity that can be documented or measured and that the event or episode being dated has a direct association with artifacts or archaeological features. The geoarchaeologist must determine that the methods employed can translate a temporal signal into a reliable estimate of age and that this age has a firm relation to the archaeological context. Too often dates are discarded because they do not fit a working model of the archaeological context. Although the reliability of an age estimate should be reconsidered if it does not accord with other independent chronological criteria, the researcher is well advised to consider possible explanations for dates that do not ‘‘fit’’ a model, unless there are other reasons to question a date. The choice of dating techniques in a particular situation is limited by the materials available for analysis and the age constraints of the archaeological phenomena. Figure 5.1 shows some materials that can be dated by various techniques and presents the typical age limits. These limits vary according to sample character and technical circumstances. Many, if not most, dating techniques important in archaeology grew out of an earthscience context. This chapter is an introduc-

Estimating Time

Figure 5.1 Archaeological Materials Dating Chart Dating methods, the materials that can be dated, and the general age ranges for these methods. Both organic

and inorganic materials can be dated. Different methods are used to date specific ages of the archaeological record.

tion to the most important earth-science-related dating techniques that geoarchaeologists use.

first part of the Cenozoic, termed the Paleocene or Early Tertiary (Paleocene, Eocene, Oligocene). The fossil record demonstrates the diversification of the primates during the past 65 million years, the Cenozoic. The record of the earliest hominids has been discovered in Late Miocene, Pliocene, and Early Pleistocene deposits. There are two global-scale patterns of climate change associated with the Cenozoic.1 First, since about the early Eocene there has been a general cooling and drying trend. Changes in global climate are linked to major tectonic events, such as the connection of India and Asia during the early Eocene, the accelerated uplift of the Tibetan Plateau in the early Miocene, and the closure of the Panama Seaway resulting in the connecting of North and South America near the end of the Pliocene. Second, the onset of northern hemisphere glaciation and the climatic oscillations of cycles of glacials and interglacials were initiated sometime between about 3 and 2.5 million years ago. This second pattern is reflected in the marine oxygen isotope curve (fig. 5.3). There appears to be a coincidence between the rather dramatic change in cli-

Climate Change and Time Archaeological studies focus on approximately the past three million years of geologic time, since the oldest known artifacts date to about 2.6–2.5 million years ago. In contrast, paleoanthropologists interested in the fossil record of the evolutionary development of the early primates are particularly interested in the time starting around 70 million years ago, while researchers interested in the diversification of the hominoids and the earliest hominids are interested in the time interval from about 24 million to 2 million years ago. By about 2 million years ago, the prehistoric record contains fossil evidence of the genus Homo. The geologic timescale is divided into four large time periods (fig. 5.2). These include the Precambrian, Paleozoic, Mesozoic, and Cenozoic. The early primate record is linked to the end of the Mesozoic, termed the Cretaceous, and the

133

Estimating Time

Figure 5.2 Geologic Time Divisions The geologic time intervals containing evidence of prehistoric life are the Paleozoic, Mesozoic, and Cenozoic. Mammal fossils are found beginning in the Mesozoic while almost the entire primate fossil record is found

within the Cenozoic. Using the time boundaries depicted here, archaeological remains are associated with the last part of the Cenozoic, the Quaternary (Pleistocene and Holocene).

matic patterns, the presence of a variety species of hominids within the genera Australopithicus and Homo, and the earliest evidence of artifacts during the time interval of about 3 to 2 million years ago.2

The major divisions of the archaeological timescale are based on the use of time-diagnostic characteristics of artifacts (figs. 5.4 and 5.5). Initially, during the 1800s and early 1900s, time-diagnostic artifact forms were used to develop relative chronologies. The oldest archaeological occurrences were designated as the Stone Age. These were later divided into the Paleolithic and the Neolithic based on the appearance of traits like pottery and ground stone artifacts. The Paleolithic, now known to extend from about 2.6 million years to about 10,000 years ago in regions of the Old World, has been further subdivided into three major parts also based on the presence or absence of particular artifact forms. Time-diagnostic artifacts are also found in Late Pleistocene– and Holocene-age deposits. For instance, in the Old World, the classic relative sequence of artifact forms led to constructing the succession composed of the Neolithic Age, Bronze Age, and Iron Age (Mesolithic to Modern artifact divisions on fig. 5.4). In the Americas there is also a relative

Artifacts and Dating Since the 1800s it has been known that certain kinds of artifacts can be used to determine the relative age of artifactual assemblages or the strata that contain them. One example of the difference between geoarchaeology and archaeological geology is the use of time-diagnostic artifacts. Where independently dated or time-diagnostic artifacts are found within sediments they can be used by archeological geologists to help determine the age of climatic or landscape-forming events. On the other hand, the application of specific earthscience methods or concepts to estimate the age of an artifact or archaeological feature would be considered geoarchaeology in our terminology.

134

Estimating Time

order of artifact forms. For instance, stone tools are present in North and South America by the end of the Pleistocene and throughout the Holocene while, in contrast, pottery is only present during the later part of the Holocene (fig. 5.5). There are many examples where specific artifact forms have been used as time indicators. For example, in the Old World, Acheulian handaxes occur in deposits known to date from about 1.5 million years ago to about 300,000 years ago. Thus they can be used to infer that deposits containing them are from the Early or Middle Pleistocene. A similar example from the New World is the Clovis point, a fluted pointed biface that has been found with the remains of extinct animals such as mammoth (Mammuthus) and is temporally restricted to the very end of the Pleistocene.

Stratigraphy

Figure 5.3 Paleomagnetism and Oxygen Isotope Records Major intervals of the Quaternary global paleomagnetic record compared to the marine oxygen isotope record. The paleomagnetic and isotope records depicted extend over the past 2.6 million years, the archaeological time range as reflected by the known presence of artifacts. The marine stable isotope record is regarded as an indicator of relative sea levels. Times of low sea levels are associated with glacial stages (odd numbers) while high sea levels are associated with interglacials (even numbers). The Matuyama-Gauss paleomagnetic boundary is dated to about 2.6 million years ago while the Matuyama-Brunhes boundary is at approximately 800,000 years ago. (Based on Lowe 2001)

Stratigraphic principles are fundamental to the study of time in archaeology. These principles are used to evaluate the relationships of sediments, artifacts and archaeological features, and ecofacts within the context of time and space (see fig. 5.6). A chronostratum is a stratigraphic unit that is delimited by lower and upper boundaries that everywhere are the same age; that is, the boundaries are synchronous.3 The chronostratum is restricted to a particular interval of time. In contrast, a lithostratum and a biostratum are not defined by time but instead are defined on the basis of content. In the case of the biostratum, the unit is defined on the basic of fossil content. We would include deposits defined by specific forms of artifacts as particular biozones or biostratigraphic units. A lithostratum is defined on the basis of petrologic or rock/sediment characteristics. Both biostrata and lithostrata can be time transgressive, or diachronic. These are different but equally valuable ways of describing and classifying the prehistoric record. Stratigraphic relations have always been the primary method for inferring the age of artifacts. Some stratigraphic techniques, which use varves or other annual laminae, can provide absolute ages directly, while others rely on independent

135

Estimating Time

Figure 5.4 Geologic and Archaeologic Time Divisions for the Last Three Million Years The last 3 million years of geologic time can be divided into the Early Pleistocene, ending about 800,000 years ago; the Middle Pleistocene, ending around 135,000 years ago; the Late Pleistocene, ending about 10,000 years ago; and the Holocene, continuing to the present. The oldest artifacts, from the Lower Paleolithic, date to about 2.5 million years ago. Lower Paleolithic Acheulian artifacts are found in Early and Middle Pleistocene strata. Middle Paleolithic artifacts are found in Middle and Late Pleistocene geologic units. Upper Paleolithic artifacts are primarily found in Late Pleistocene contexts. Mesolithic and artifacts are found in final Pleistocene and Holocene settings.

dating—for example, dating paleosols in loess, or volcanic ash deposits (tephrochronology). Julie Stein provides an excellent review of stratigraphy and dating in archaeology.4 Two principles are fundamental to stratigraphy: (1) superposition, and (2) original horizonality. Superposition utilizes the observation that in any set of strata, the layers on the bottom were deposited before any layer overlying it; thus, as stated by Sam Boggs,5 the oldest layers are at the

136

bottom and the youngest are at the top. Original horizonality is particularly important when studying sedimentary rocks and sediments. It relies on the observation that sedimentary particles will settle in nearly horizontal layers. The initial application of these principles are traditionally attributed to Nicholas Steno, who conducted his studies in Europe during the late seventeenth century, and the initial application of a uniformitarian perspective as reflected by the studies of James Hutton in England during the eighteenth century. In England during the late 1700s and early 1800s, William Smith demonstrated the practical value of using the idea of superposition as previously observed by Steno and Hutton. The work of Smith was especially important in developing the methodology of stratigraphic correlation. He used the characteristics of strata and the fossil remains contained within sediments to infer connections between separated stratigraphic sequences. As was mentioned in Chapter 1, the publications by Lyell (Principles of Geology, Geological Evidence for the Antiquity of Man) in the mid-1800s demonstrated the potential of applying stratigraphic and uniformitarian principles to understanding earth’s history and the human past.

Rhythmites (Varves) Rhythmic accumulations of sediments forming distinctive laminae (see Chapter 2) are a common occurrence in the geologic record. The geologic term for such sequences is rhythmites. When the laminations form because of annual variations in supply and type of sediment they are called varves. The potential of using annually laminated sediments as a means of dating was first exploited by the Swedish geologist Gerard de Geer in the late nineteenth century. Along with tree rings, varves provide absolute ages when even a single varve date is known. The top layer of sediment at the bottom of a still-active lake belongs to the year just past, a fact that can anchor the chronology. Varves have also provided an absolute calibration of the radiocarbon timescale and can be directly applied to pollen stratigraphies in varved sequences. Varve dating technique is based on variation in sediment deposition during an annual cycle.

Figure 5.5 Relation between Geologic Time Divisions, Paleoclimate Terms, and Archaeological Sequences for North America from about 10,000 B.P. Geologic time and paleoclimatic divisions and the North American archaeological sequence since the end of the Pleistocene.

Estimating Time

Figure 5.6 Chronostrata/Lithostrata/Biostrata The relationships of lithostrata, chronostrata, and biostrata in time and space. Chronostrata designations are unique in that they are defined by time boundaries,

while the nature of the types of rocks are used to designate lithostrata and biotic indicators (fossils and artifacts) are used to define biostrata. (From Holliday 2001)

Commonly, finer particles or chemical precipitates may be deposited during the winter months, while coarser particles are deposited during other periods. The two layers combined represent an annual cycle of deposition. In Scandinavia varve sequences go as far back as 13,000 years. De Geer initially developed the technique in the Baltic region,6 while in North America the work of Ernest Antevs represents an early effort to use varve chronologies.7 In northern lakes that freeze over in the winter, deposition of coarser particles (sand and silt) in the summer is followed in the winter by deposition of finer material (clay and organics). The resulting laminations are easy to distinguish because the light-colored summer sediment alternates with the much darker winter laminae. Unfortunately, bioturbation and other phenomena

often disturb the laminations in all but the deepest lakes. Nevertheless, varve chronologies have long provided a calibration for Late Quaternary events. Approximately 10,000 laminations have been recorded in a five-meter core from Elk Lake, Minnesota, thus extending varve chronology back to the onset of the Holocene in central North America.8 As can be seen in figure 5.7, varve years (which should represent actual calendar years) differ from 14 C years, because of the reservoir effect (see radiocarbon, below). The differences may be as great as 1,950 years. The varve dates at Elk Lake, for instance, are consistently more recent than radiocarbon dates because of older carbon in the carbon reservoir of the lake. The varve chronology was also compared with the pollen stratigraphy and a chronology based on other nearby

138

Estimating Time

Figure 5.7 Varves and Radiocarbon Years Varve years are usually younger than radiocarbon

years because the radiocarbon chronology is measuring older carbon within the system.

pollen records. Rapid and distinctive changes in vegetative patterns occurred several times over the past 11,000 years, including a decrease in spruce around 11,000 b.p., a rise in sagebrush around 8,560 b.p., a rise in birch around 3,390 b.p., and an increase in white pine around 2,700 b.p. Based on the regional pollen records, the uncalibrated ages for the spruce decline and sagebrush rise seem too old, but if the change in spruce in the Elk Lake area can be calibrated to about 10,000 b.p., it would agree with the varve chronology. Varves can also occur in more temperate climates. Sediment load, biomass accumulation, and chemical precipitation all vary seasonally. If bioturbation or current action does not rework the resulting sediments, annual layering will be recorded. The resulting laminations are visible in some lakes. In others they are apparent only through analytical and microscopic study. In limestone regions, light-colored summer layers of calcium carbonate alternate with dark winter layers rich in humus. When one layer is rich in organic material, radiocarbon dating by accelerator mass spectrometry is possible.

Paleosols in Loess and Alluvium Loess deposition interrupted by periods of stabilization and soil formation has led to stratigraphic sequences that can be dated. In North America, Europe, and China the substantial progress made in dating loess deposits and the soils developed within them have considerable value for archaeology.9 These loess-paleosol sequences need to be anchored by independent dating correlations before they can be used as time indicators. Figure 5.8 shows a correlation of the North American and China loess sequences with the marine oxygen isotope curve. In this way, a timescale can be superimposed on the loess sequences. Other criteria have been used to prove such correlations. For example, the China sequence has been dated with magnetic reversal chronology and potassium-argon (K-Ar) dating, while parts of the North American sequences have been dated with thermoluminescence (TL) methods (described below). The loess sequence in Europe has been dated using radiocarbon, TL, and stable-isotope series techniques (see fig. 5.9). William C. Johnson and Brad Logan present an excellent example of the use of loess-soil chrono-

139

Estimating Time

Figure 5.8 Three Analytic Tools That Show Similar Curves and Reversals The similar patterns of climate change recorded in these stratigraphic sequences are interpreted as reflecting major global warming and cooling events during the past 400,000 years and can be related to the human prehistoric record. The oxygen-isotope record from layers of ocean sediment shows changes associated with glacial and interglacial intervals. Conditions were somewhat similar to the present-day intergla-

cial during oxygen isotope stages 5, 7, and 9. Similar patterns in the carbonate content of loess deposits in central North America have been correlated with the ocean isotope record. Intervals associated with higher amounts of carbonate accumulation in the loess coincide with the interglacial intervals. In the windblown loess deposits of China, a similar correlation has been made. There, sediments with high magnetic susceptibility are thought to coincide with interglacial or warmer climatic episodes. (Based on Feng et al. 1994)

stratigraphy.10 The Peoria loess of the Kansas River basin dates from about 23,000 b.p. and is overlain by the Brady soil, dated to about 11,000 b.p. The younger Bignell loess overlies the Brady soil. Both the Peoria loess and the Holocene loess fall within the interval of prehistoric occupation on the American Plains. A discovery of mammoth remains in eastern Montana within the Deer Creek drainage provides another North American instance of loess dating. At this site the remains of a mammoth were overlain by loess silts containing buried soils.11 The silts have been interpreted as loess that began to accu-

mulate around 13,000–12,000 b.p. Using radiocarbon, collagen from the mammoth bones was dated 12,330 and 11,500 b.p. The silt and soil sequence over the mammoth bones may correlate with a wide-ranging Terminal Pleistocene and Early Holocene loess and soil sequence on the Great Plains (the Aggie Brown Member and the Leonard Paleosol). Sedimentary sequences composed primarily of alluvial deposits containing buried soils have been extremely useful in dating archaeological accumulations, especially in areas like the American Midwest. As with paleosols in loess deposits, the

140

Estimating Time

Figure 5.9 Thermoluminescence and Radiocarbon Dating of Microfauna within Loess Deposits Comparison of isotopic variation in mollusk shells with time. In this example, chronological control of the isotopic signal is based on TL and radiocarbon measurements. The Middle Pleistocene older loess sedi-

ments generally have lower isotopic values than the Upper Pleistocene younger loess mollusks. In this instance, the use of TL measurements was crucial to the discovery of the presence of Middle Pleistocene age strata, since radiocarbon techniques cannot date this interval. (Based on Rosseau and Puiseguit 1990)

chronological framework is based either on the relative stratigraphic order of buried soils within a sequence or on independent dating methods (such as 14 C dating of soil humates or bones in the paleosols). Studies by Rolfe Mandel in Kansas, Arthur Bettis in Iowa, and Edwin Hajic in Illinois show how useful soil chronologies can be in interpreting the temporal context of artifactbearing sediments. Based on detailed studies of the characteristics of buried soils and radiocarbon determinations of charcoal, wood, bone, and soil humates, Mandel was able to use paleosols to correlate regional alluvial sequences containing artifacts.12 Late Archaic and Plains Woodland archaeological sites in Kansas are likely to be associated with two buried soils that are valuable stratigraphic markers of valley fill dating

to the Late Holocene. The oldest of the buried soils, called the Hackberry Creek Paleosol, is associated with Late Archaic artifacts. The soil started to form at about 2,800 b.p. and continued to develop until at least 2,000 b.p. The Hackberry Creek Paleosol is buried by alluvium. The Buckner Creek Paleosol developed within this alluvium and contains Plains Woodland artifacts. The Buckner Creek Paleosol began to develop around 1,350 b.p. and continued to form until at least 1,000 b.p. In the Upper Midwest it is possible to separate Holocene alluvial deposits into three groups. Each group is of a different age, so each is potentially related to a different archaeological interval and has a unique set of associated soils. Sediments deposited from about 10,500 to 4,000 b.p.

141

Estimating Time

are designated Early to Middle Holocene (EMH) alluvium. They are characterized by oxidized colors and mottling, and have surface soils with well-developed horizons, especially the subsurface argillic, or Bt, horizons typically formed in Mollisols or Alfisols. Because of organic carbon, Late Holocene (LH) alluvium deposited after 3,500 b.p. is generally darker than EMH alluvium and does not have argillic horizons developed in the surface soils. The youngest deposits are mostly historic-age sediments. They are characteristically lighter in color than LH deposits, exhibit distinct bedding in their lower sections, are oxidized, and show a slight pedogenic alteration. These three alluvial groups have different weathering zone features and different surface soils developed on them, and they can be found throughout the Upper Midwest. Bettis used these distinctive features to help determine the age of artifacts associated with them. Paleo-Indian and Archaic artifacts have been associated with EMH deposits, while LH deposits are likely to contain Late Archaic and Woodland components. At the Koster site in western Illinois there are twenty-three distinct archaeological zones within a Holocene sedimentary and paleosol sequence. The zones are defined by laterally continuous accumulations of artifacts. These artifact accumulations represent a succession from the Early Archaic through the Mississippian periods of human prehistory. Hajic showed that the sequence consists mostly of colluvium and alluvial fan sediments composed of loess that was redeposited as sheetwash and of stream-related deposits. During some intervals, surface processes at Koster became stable enough for soils to form. On the basis of soil morphology and the degree of horizon development, it was possible to discern two groups of soils. Soils that formed over relatively short intervals of time did not have distinct B horizons. Often these soils had distinctive, thick A horizons that had formed during intervals of continuing sedimentation. The time necessary for a cummulic soil horizon to form was roughly several hundred years. The second group of paleosols formed over a longer time on more stable slope and fan surfaces. They have more developed B horizons that may have taken from 500 to

142

1,500 years to form. The first soil with a developed B horizon appears to have started forming around 9,800 b.p. The second major soil formed after about 4,100 b.p. and before 2,500 b.p., since it is older than Early Woodland Black Sand pottery. The third soil with a well-developed B horizon formed sometime between about 2,600 b.p. and 1,150 b.p. Interpretation of the timing of depositional and soil-forming events at Koster related the deposits to diagnostic artifacts (used as time markers) and radiocarbon measurements.

Tephrochronology Volcanic ash deposits have proven effective for relative dating of archaeological sites—and for absolute-age correlations when their ages can be determined through such chronometric techniques as potassium-argon dating (see below). To be useful dating tools, ash deposits need to have been widely dispersed, have a chemical signature that is unique, and be able to fit into a chronologic sequence. The challenge of dating the Bronze Age volcanic eruption on the Aegean island of Thera is an example of the application of tephrochronology. In the Bronze Age, Thera was the site of the extensive Minoan culture of sea traders. In the middle of the second millennium b.p. two-thirds of the island was blown away in a catastrophic eruption. The whole center of the island became a great caldera, open to the sea. A oncethriving seaport on the southern tip of the island was buried under volcanic ash. The Archaeological Society of Greece, under the auspices of the Greek Archaeological Service, is currently excavating the ancient town now called Akrotiri that is situated on one of its remnant islands. This site and its history have tremendous importance in their own right but in addition have received worldwide attention because the cataclysmic eruption and attendant destruction might have been the inspiration for Plato’s legend of Atlantis. Volcanic eruptions eject large quantities of fine particles and sulfur compounds into the upper atmosphere circulation system. These materials return slowly to earth in rain and snow. A significant increase in acidity from this sulfur is found

Estimating Time

in distinct layers in glacier ice cores. A Greenland ice core has a strong acidity peak at 1,644 b.p. plus or minus twenty years. This has been interpreted as recording the eruption of Thera. Artifacts from Akrotiri, linked to the Egyptian calendar, put the Thera eruption at more than a hundred years later. While the controversy remains open, it is our view that the volcanic activity recorded in the Greenland ice core more likely came from nearby Iceland than from the eastern Mediterranean (this may be testable by comparing chemical signatures). In western North America three postglacial volcanic ash deposits are useful marker beds for the Latest Pleistocene and Lower Holocene. Glacier Peak’s G and B and Mount Saint Helens’ J ashes date to Late Pleistocene. The Glacier Peak ashes are from a volcano in north central Washington and date to 11,200 b.p. The Mount Saint Helens J ash comes from southern Washington and consists of ash layers from eruptions that occurred between about 11,500 and 10,800 b.p. The most widespread Holocene ash in western North America, the Mazama ash, came from the volcano at Crater Lake, Oregon, in an eruption dated to approximately 6,845 b.p. The ages of these ashes have been determined primarily by radiocarbon dating, although TL and other techniques were also used.13 Tephrochronology has helped to date many archaeological contexts and stratigraphic sequences in western North America.14 The Clovis cache at the Wenatechee site in Washington was dated by eruptions from the Glacier Peak volcano. At this site, pumice-rich sediments derived from the Glacier Peak tephra were found directly underlying the Clovis artifacts. A nearby stratigraphic sequence contained the Mazama tephra. Glacier Peak Layer G ashfall was identified at the Indian Creek Site in western Montana. The ash was dated by radiocarbon to about 11,125 b.p. and underlies an assemblage of Folsom-related artifacts dated to about 10,980 b.p. The Mazama ash was also found within the Indian Creek sequence. Because these two ash deposits have known ages, they can be used as time markers once they have been identified. Identification can sometimes be made on the basis of thickness or color, but field

identification needs to be checked against mineralogic or geochemical signatures that distinguish the volcanic ashes.

Dating Using Animal and Plant Fossils Paleontology Principles of biostratigraphy have been useful in indicating relative age, although, as with the use of artifacts as ‘‘index’’ fossils, the problems of contemporaneity, time transgression, and habitat differences need to be considered. The most useful indicators of specific age intervals are the remains of organisms with two characteristics: they have a wide geographic distribution, and they appear and then become extinct over a short time. In North America, the presence of the remains of extinct animals associated with artifact-bearing deposits has been decisive for determining the antiquity of humans. Although earlier discoveries of extinct fauna near artifacts had occurred, their contextual associations were questioned. It was the substantiation of the association of extinct bison with undisputed artifacts in a gully near Folsom, New Mexico, that led to the consensus that humans had been in North America for thousands of years. At the time of the discovery, the age of the extinct animals was a matter of extrapolation based on nonabsolute methods of dating. In another example, the first appearance and the evolutionary development of fossil rodents in Europe have been used as paleontologic markers in Quaternary stratigraphic sequences. In Africa the remains of various species of horses, pigs, and elephants have been used as relative time markers for Plio-Pleistocene strata containing hominid remains and Lower Paleolithic artifacts. In addition to their use in evaluating environmental and climatic change, pollen-stratigraphic sequences can also be used as temporal indicators. To be an age indicator, a pollen sequence must be correlated with a previously determined chronology. In turn, this environmental-temporal framework can be correlated with archaeological events. For example, pollen sequences in France from the La Grande Pile and Les Echets peat bogs have been correlated with a chronology

143

Estimating Time

derived from a radio-isotope timescale and can be used as a starting point for evaluating environmental change throughout the Paleolithic of Europe. High pollen abundances correspond to warm climates, associated with interglacials and interstadials. The oldest major peak is attributed to the beginning of isotope stage 5 (the last interglacial), about 130,000 years ago. Changes in arboreal pollen frequency occur throughout the rest of the sequence and can be related to Paleolithic occurrences in the region that were dated independently. Over a wide area, the time-transgressive nature of vegetative change is reflected in pollenstratigraphic sequences. Before radiocarbon dating, pollen zones in northern Europe were dated using varve chronology. In an archaeological context, the presence of a particular pollen zone within a site’s stratigraphic sequence allows it to be correlated to regional chronozones dated by radiocarbon.

Dendrochronology In the early twentieth century, A. E. Douglass developed tree-ring dating.15 It is based on the observation that in many trees a new ring or row of wood cells grows each year. The basic requirement for tree-ring dating is the presence of clearly defined annual rings, while the major problem is that rings may be missing because of situations like extreme climatic conditions. Along with varves, tree-ring dating has proven extremely useful as an independent check on the radiocarbon chronology. The two major tree-ring chronologies come from North America and Europe, although in other areas of the world the beginning date of the regional sequence is continually extended back in time. Tree-ring chronologies have been created that extend throughout the Holocene in North America, Europe, and the Near East. The dendrochronological calibration currently goes back to about 10,000 radiocarbon years b.p. Once a reliable tree-ring series has been obtained, it can be applied to archaeological sites where parts of the series have been preserved. Keep in mind that archaeological application of tree-ring dating generally provides maximum age limits for an

144

archaeological feature; trees older than the archaeological feature that contains the tree remains may have been used or reused.16 Tree-ring dating provided convincing evidence that the Viking period in northern Europe extended back as early as the eighth century.17 Three Viking ship-burial mounds were dated using dendrochronology. The tree-rings were measured from the sapwood structure of the grave chambers of the ships. Dendrochronology is not by nature geologic, but the wood being dated is often found in sedimentary contexts, and tree-ring analysis provides data for paleoclimatology.

Radiometric Dating Methods Potassium-Argon and Argon-Argon Dating Potassium is a major element in many rocks. One of its isotopes, 40 K, decays naturally to 40Ar.18 In such minerals as mica and potassium feldspars, and in volcanic glass, this decay can be used to date igneous and metamorphic rocks. The half-life for 40 K decay is around 1.25 billion years. The youngest published dates for this method are around 35,000 b.p., although it works best when applied to igneous rocks that are more than 100,000 years old. In determining ages of artifacts younger than that, the atmospheric concentration of argon (Ar) disguises the age signal. Source minerals containing potassium (K) are produced in volcanism. This technique can thus be employed when there is a direct relation between volcanic rocks and archaeological phenomena. The amount of trapped 40Ar that has accumulated relative to the amount of remaining 40 K provides an estimate of the age of the volcanic event. Minerals in igneous rocks can be eroded and redeposited. The K-Ar concentrations of the minerals in the sediments measure the age of the original igneous event, not the time of redeposition and formation of the sedimentary strata. When using K-Ar dating, it is assumed that: (1) the argon contained in the mineral was produced solely as a result of the decay of 40 K; (2) no argon has been lost since the mineral was formed; and (3) no argon was added at the time of for-

Estimating Time

mation or during some later event. Similarly, the potassium content of the minerals must be part of a closed system with no changes except for decay of 40 K. Later recrystallization caused by a subsequent metamorphic igneous or melting event may cause loss of argon. Weathering and alteration can cause loss of both potassium and argon and can introduce 40 K. Older dates can result from the incorporation of potassium from older volcanics. Dates based on the decay of potassium to argon have been especially useful in developing a timescale for the Pliocene and Pleistocene. This chronology has been extremely important to paleoanthropology and Paleolithic archaeology. The research by Bernard Brunhes and Montonori Matuyama in the early twentieth century led to the discovery that the earth’s magnetic field had reversed polarity several times during the Quaternary. The timescale for these polarity reversals is based on K-Ar dates of volcanic rocks. Potassium-argon dating has been especially successful for the Pliocene and Pleistocene strata in East Africa, where volcanism has been common. Important archaeological findings that have been dated include the sequence at Olduvai Gorge in Tanzania and the australopithecine remains called Lucy in Ethiopia. Among the earliest applications of K-Ar dating was the work by J. F. Evernden and G. H. Curtis on tuff (volcanic ash) beds at Olduvai Gorge.19 Potassium-argon dates from Olduvai Gorge were at first considered surprising; they were much older than expected. The application of a related technique, 40Ar/ 39Ar, during the early 1990s provided even more precise ages.20 Bed I at Olduvai Gorge contains some of the world’s best-known hominid fossils and is associated with an interval of major biologic and climatic change. Using 40Ar/ 39Ar analyses, single mineral grains from the middle and upper part of Bed I were found to be about 1.8–1.75 million years old. Dates from the lower part of Bed I may be about 100,000 years later than previously proposed. The 40Ar/ 39Ar method was used by J. D. Clark and his associates to date artifacts and fossils of anatomically modern humans from Ethiopia that were stratigraphically related to beds of tuff and redeposited volcanic materials (pumice and

Figure 5.10 Ar-Ar Dating The Ar-Ar dating method has been used to study archaeological occurrences and Pleistocene hominid fossils in contexts that contain igneous strata. In the Afar Rift of Ethiopia fossils and artifacts were dated at between 160,000 and 154,000 years ago using the Ar-Ar technique. Materials that were dated include the minerals anorthoclase and sanadine found in pumice and obsidian. The Ar-Ar dating method has been used to relatively young geoarchaeological events like the explosion of Pompeii in A.D. 79 as well as older contexts, such as discoveries from East Africa dating to millions of years ago. (Based on Clark et al. 2003)

obsidian).21 The stone artifacts show affiliations with the Acheulian and the Middle Stone Age. The artifacts and fossils were found in the Herto Member of the Bouri Formation. Late Acheulian tool types were found in the lower part of the Herto Member, consisting of lacustrine deposits of layers of carbonate and silty clays. Feldspars in a tuff (made of bentonite) from the lower part of the Herto Member were used to obtain a 40Ar/ 39Ar date of about 250,000 years ago (fig. 5.10). These lower deposits were found to grade upward into fluvial and lake margin sandstones capped by another tuff. The upper part

145

Estimating Time

of the Herto contained bifaces made on flakes as well as Levallois flakes. A fluvial sand with clasts composed of pumice and obsidian in the upper part of the Herto Member also contained hominid fossils. An orthoclase feldspar from the pumice dated by 40Ar/ 39Ar yielded an age of about 163,000 years while a date on the sanidine feldspar yielded an age of about 226,000 years ago. An age of 160,00 years ago was obtained from an obsidian clast. The tuff that capped the Herto Member was unable to be directly dated due to contamination by older crystals. The 40Ar/ 39Ar method was used to demonstrate the Late Middle Pleistocene age of fossils of anatomically modern Homo sapiens in deposits containing Middle Stone Age (Levallois) artifacts that were stratigraphically overlying Late Acheulian artifact assemblages. This example shows how a radiometric dating technique can be used to date artifacts found in stratigraphic contexts containing volcanic materials. In this case, the technique provided information on the chronologic relationship between the Lower and Middle Paleolithic (as represented by the change from Acheulian to Middle Stone Age) and it helped to estimate the age of fossils of anatomically modern humans in East Africa.

Uranium-Series Dating The element uranium (U) is radioactive and decays to a sequence of other elements at a known rate.22 This decay sequence, terminating with lead, is known as the uranium series (U series). This chain of decay is the foundation for several ways of estimating age. Two isotopes of uranium, 235 U and 238 U, decay through different element sequences and at different time rates to produce different isotopes of lead (Pb). Methods of uranium-series dating are based on these two decay series. Uranium-series dating is possible only under the following conditions: (1) at the time the uranium-containing mineral is formed, the daughter isotopes were either absent or their concentration can be determined; (2) the activity of the daughter product must reach equilibrium—that is, daughter products have a constant concentration; and (3) the sample has not been chemically disturbed since formation. Usually uranium-series dating relies on the solubility of 146

uranium in water, while the decay products precipitate from hydrous solution.23 Within the decay series of uranium a number of element ratios have been used in dating. The isotope 230 Th (thorium) is within the 238 U sequence, and 231 Pa lies within the 235 U-series. In a closed system, where there was no initial 230 Th and 231 Pa and no U or decay products have entered or left, the ratios of 230 Th/ 238 U or 231 Pa/ 235 U can be used to estimate the age of a material less than 200,000 years old. The ratio of 230 Th/ 238 U was first used to date corals and later shell and carbonate deposits. Uranium-series dating has a range of about 1,000 to 800,000 years. For comparison, radiocarbon is most useful for ages up to about 40,000 years, and potassium-argon is best used for ages greater than about 750,000 years. Uranium-series methods have been useful in dating organic carbonates, corals, and shell (mollusks), and successful attempts have been made to date tooth enamel. Inorganic carbonates such as limestones, speleothems, and travertines from cave, spring, and lake deposits have been dated using uraniumseries. The uranium-series method has been used to date archaeological materials embedded in or encrusted with calcium carbonate. It is especially useful for dating Paleolithic sites. Springdeposited travertine (tufa) deposits are useful chronostratigraphic markers. The 230 Th/ 234 U method has been used to date travertine deposits within the interval between 171,000 and 149,000 b.p. At El Castillo Cave in Spain (discussed in Chapter 3; see fig. 3.12), the uranium-series method was used to evaluate the age of Acheulian and Mousterian artifacts.24 A massive flowstone layer separates two types of artifact assemblages. Beneath the flowstone are artifacts traditionally labeled Acheulian and above the flowstone are artifacts designated Mousterian. The flowstone is mainly composed of a solid travertine with a spongy breccia for the upper part. The base of the unit is a dripstone or stalagmite that was dated to about 89,000 b.p. Dates calculated directly from the daughter-parent ratio of 230 Th/ 234 U gave ages that were too old because of the detrital addition of 230 Th. Since 232 Th is not part of the 238 U/ 235 U decay series, it can be used as an indicator of

Estimating Time

Figure 5.11 Oxygen Isotope Stages for the Middle and Late Pleistocene The last interglacial (isotope stage 5e) marks the end of Middle Pleistocene and the beginning of the Late Pleistocene around 135,000 years ago. The Last Gla-

cial Maximum is within isotope stage 2, between about 32,000 and 13,000 years ago. The beginning of the present interglacial (the Holocene) is marked by the beginning of oxygen isotope stage 1.

the degree of detrital contamination. A date of 89,000 b.p. would correspond to the cool oscillation related to oxygen isotope stage 5b, although at one standard deviation the flowstone could also be associated with either stage 5c or 5a (fig. 5.11). If this is a reasonable estimate for the flowstone, it shows that Acheulian assemblages were in existence before the flowstone’s deposition and that Mousterian artifacts were deposited some time after about 90,000 b.p. For historic archaeology, 210 Pb provides an important dating technique. The isotope occurs

naturally in lake sediments as one of the radioisotopes of the 238 U decay series. It has a half-life of 22.26 years. Because of its short half-life, 210 Pb works only for artifacts dating from the middle of the nineteenth century to the present.

Radiocarbon Dating Carbon-14 is a radioactive isotope of carbon that was initially used by Willard Libby to date organic materials.25 The bombardment of nitrogen atoms by cosmic rays in the upper atmosphere produces the 14 C isotope (fig. 5.12). The 14 C isotopes are 147

Estimating Time

method directly measures the concentration of 14 C relative to the amount of 12 C or 13 C. In the

Figure 5.12 The Carbon-14 Cycle This diagram illustrates the formation of 14 C in the upper atmosphere, the incorporation of the 14 C as part of the CO molecule into living matter, and finally the radioactive decay of 14 C back to 14 N.

incorporated in all geochemical and biochemical systems that are in equilibrium with atmospheric carbon dioxide. The half-life for 14 C is about 5,730 years. Dates determined before 1970 were calculated with a half-life of 5,568 ± 30; the more accurate half-life is 5,730 ± 40. Radiocarbon dating beyond about forty thousand years is possible but it is not as reliable compared to younger samples because of contamination problems and diminishing concentrations of 14 C (fig. 5.13). There are two methods of obtaining a radiocarbon date. The conventional technique relies on measuring beta rays from the radioactive decay. The second method uses an accelerator mass spectrometer (AMS) to measure the number of atoms of 14 C remaining in a sample. The AMS

148

conventional method, 1 gram of present-day carbon would have approximately fifteen disintegrations per minute, but a 22,000-year-old sample would have only about one disintegration per minute. With AMS dating, the speed of counting is faster, much smaller samples can be used (down to less than 1 mg), and under ideal circumstances it may be possible to extend the age range to 60,000 or more years. Several aspects of the carbon system directly influence radiocarbon measurements, including the carbon reservoir, isotopic fractionation, the earth’s magnetic field, sunspot activity, burning of fossil fuels, and nuclear testing. Scientists assume that 14 C is evenly spread within the carbon reservoir. Carbon-14 production varies with cosmic-ray intensity, magnetic-shield intensity, and atomic explosions. If 14 C production is low for a given time period, less of the isotope can be found in samples from that period, and any measurement would indicate a date earlier than the actual age. Several phenomena have influenced the production of 14 C in the past. Because cosmic rays result in the production of 14 C, any change in their intensity alters the natural concentration of 14 C. When the magnetic field is weak, more cosmic rays are present, and the production of 14 C would therefore increase, while the opposite would happen when the magnetic field was stronger, and more cosmic-ray particles were deflected away from earth’s atmosphere. Similarly, during geomagnetic reversals, there is higher production of 14 C. Calibrations based on tree-ring, varve, and uranium-series chronologies have been used to correct 14 C dating (see fig. 5.14). Extreme variations in the production, transport, and deposition of radiocarbon have been documented for the Late Pleistocene. These variations need to be taken into account when evaluating late Middle and Upper Paleolithic archaeologic chronologies.26 During the time interval associated with oxygen isotope 3 (see fig. 5.11), atmospheric radiocarbon appears to have fluctuated, based on plantonic foraminifera, varve, and stalagmite records. Also there appear to be

Estimating Time

Figure 5.13 Age in Radiocarbon Years Chronometric techniques and the half-life curve. In the radiocarbon method the age is calculated based on measuring the amount of 14 C present in a sample and an estimate of the rate of decay. It takes about 5,730

years for half of total 14 C to decay; this is the ‘‘half-life’’ for 14 C. By measuring the amount of 14 C present rela-

patterns of variations in radiocarbon production associated with the Laschamp magnetic excursion around 40,000 calendar years ago as well as several thousand years later at the time of the Mono Lake excursion. These fluctuations need to be taken into consideration when interpreting the hominid fossil and Middle-Upper Paleolithic artifactual records, and the implications of these lines of evidence for the migration of anatomically modern humans into Europe during the interval of 50,000 to 30,000 years ago. For example, in southwest Germany, within the drainage of the Danube River, it might not be easy to calibrate radiocarbon dates to calendar years for Neanderthal fossils or Middle Paleolithic stone tools or evaluate whether there was actual coexistence between Neanderthal and ana-

tomically modern human populations due to the exceptional variations in atmospheric radiocarbon. Early versions of Upper Paleolithic artifact assemblages are in deposits that overlie or postdate the European Middle Paleolithic. For example, the early Aurignacian was present by 40,000 radiocarbon years b.p. and was followed by the Gravettian by about 29,000 b.p. in southwest Germany. Assessment of these ages is complicated by the record of fluctuation in atmospheric concentrations of radiocarbon. The geochemical context of radiocarbon dating includes reservoir effects. For 14 C to provide comparable dates for organic materials, worldwide atmospheric 14 C must be mixed rapidly throughout the world’s carbon-containing reservoirs. This is known not to be the case. One

tive to the original amount, it is possible to determine the age of the sample. Note the small amount of 14 C that is present after 30,000 years.

149

Estimating Time

Figure 5.14 Dendrochronology versus Various Other Dating Methods Tree-ring dating, varves, and uranium-series dating can be used to obtain a more reliable radiocarbon timescale. In both plots, radiocarbon measurements be-

come progressively younger in the interval between 4,000 and 20,000 years ago. For samples younger than about 12,000 years, tree-ring and varve data are used to compare the radiocarbon chronology. The uraniumseries method is used for older materials.

potential influence is called the hard water effect, where old or ‘‘dead’’ carbon containing no 14 C becomes mixed with the carbon in an organic substance, thus making the sample appear older than it is. This is a special problem in areas saturated with groundwater that has been influenced by bedrock limestone. This problem is critical where old carbon has been incorporated into the carbonate in shell. Living samples from a freshwater lake on limestone terrain have been known to give a radiocarbon date of up to 1,600 b.p. The atmospheric 14 C concentrations can be modified by the lack of any 14 C in the carbon contribution emanating from the limestone. In a similar way, the present-day combustion of coal and oil has diluted the concentration of 14 C because of the addition of dead carbon from these sources. On the other hand, detonation of nuclear devices has introduced large amounts of 14 C into the system. To determine accurately the radiocarbon age of a sample, one must know the initial amount of 14 C at time zero. This initial value is quite variable. Volcanic ash layers were used to mark synchronous deposition of charcoal and planktonic foraminifera.27 The difference between the char-

coal dates, reflecting the 14 C content of the atmosphere, and the foraminifera, reflecting surface ocean conditions, from the same core is the reservoir age of the surface ocean. For the Mediterranean reservoir ages were similar (circa 400 years) for most of the past 18,000 years. However, reservoir ages increased by a factor of two at the beginning of the last glaciation. Radiocarbon ages of samples formed in the ocean such as shells are generally several hundred years older than their terrestrial counterparts. This difference is due to the large carbon reservoir in the oceans. A correction is necessary in order to compare marine and terrestrial samples but the correction varies with marine location. An investigation of marine radiocarbon reservoir ages from the western coast of Norway recorded reservoir corrections of from 200–525 years with a weighted average of 380 years.28 The pottery seriation and 14 C chronologies on Tonga in the southwest Pacific were in major disagreement until a lagoon-specific reservoir correction was made. This brought the dating into agreement.29 Comparison between 14 C ages of terrestrial and marine mammals from a Late Jomon shell

150

Estimating Time

midden in Hokkaido, Japan, showed systematic age differences than can be attributed to the 14 C marine reservoir effect (delta R) in the western North Pacific Ocean.30 On the central Queensland coast of Australia, shell/charcoal paired samples were 14 C dated to determine local reservoir effects. The study determined that earlier delta R values needed correction and that nearshore but open-marine environments contrasted strongly in delta R with some estuary environments.31 Major radiocarbon reservoir effects have been formed in paleolakes of the Altiplano of northern Chile.32 The correction could be as high as minus 2,000 years. Carbon in smaller lakes is isotopically fractionated with respect to carbon in the atmosphere, water, and local rock. The mixing of surface runoff with groundwater requires caution in radiocarbon dating of shells from lacustrine deposits. Fluvially transported charcoal can lead to erroneous 14 C dates. The age of the radiocarbon sample could be much older than the deposit if the sample has been eroded out of older deposits. It is important to understand the nature of the carbon used for radiocarbon dating. Measured 14 C ages of soil organic matter are always younger than the true ages of the soil because of continuous input of organic matter into soils. Differences in soil carbon dynamics can result in widely different 14 C dates for soils of the same age. The geochemical fractionation of the stable carbon isotopes 12 C and 13 C in geologic processes is another source of error in radiocarbon dating. In isotope fractionation, 14 C is used at a slower rate than 12 C, creating differences in the ratios of the two isotopes within different organic materials. Figure 5.15 shows some typical differences in isotope fractions. The enrichment or depletion in 14 C because of natural fractionation is twice the δ 13 C values in the same sample. To eliminate this problem, all 14 C values are now normalized to a common δ value. Radiocarbon remains the main chronometer for the period of the past 50,000 years. Much scientific effort ha been expended to improve calibration curves to overcome the atmospheric variability over time of the 14 C/ 12 C ratio. The latest ‘‘official’’ calibration curve INTCAL04 was ratified during the recent (2003) Eighteenth Inter-

national Radiocarbon Conference.33 This curve does not go back further than 26,000 calendar years before a.d. 1950 (that is b.p.) because no consensus was reached. The low concentrations of 14 C remaining in samples older than 22,000 b.p. makes calibration extremely difficult. However, several techniques have been used to extend calibration back to about 50,000 years ago. Datable archaeological materials include wood charcoal, wood, peat, shell, dung, bone, iron, and parchment. Although good chronologies have been developed using shell and soil organics, potential problems need to be considered. Shells of land snails are thought to be poorly suited for radiocarbon dating because 14 C activities in the sources of carbon incorporated into these shells are not always in equilibrium with atmospheric 14 C. Radiocarbon ages from soil-organic fractions may also be poorly suited because the 14 C activities may not be known for the carbon sources in the soil. In addition, the organics in soils may represent an accumulation over a long time, perhaps even several generations of pedogenesis. Potentially the same problem of unknown 14 C activities exists for carbonate incrustation on artifacts and for geologic deposits like caliche and tufa. Reservoir effects can also occur when volcanic (fumarole) gas emissions containing only dead carbon are incorporated into plants growing within approximately 100 m of the fumarole. The major laboratories undertaking radiocarbon dating have developed sophisticated approaches to these problems, so geoarchaeologists should discuss sample type and sample origin with the knowledgeable laboratory personnel responsible for the analyses. Geoarchaeologists must evaluate the geologic and geochemical contexts that relate a radiocarbon sample to a specific geologic or archaeological feature. The integrity of the association of the radiocarbon sample with an event is of primary importance. Potential geochemical contamination from the environmental matrix where the sample was taken is of nearly equal importance. Geoarchaeological problems in sample collection include: samples taken from eroded or reworked deposits, samples taken from deposits mixed by bioturbation or cryoturbation, geochemical contamination from a fossil-carbon source (for ex151

Estimating Time

Figure 5.15 Variation in the Apparent Radiocarbon Age of Different Types of Samples because of Carbon Isotope Fractionation Because of carbon-isotope fractionation, the radiocarbon content in specimens of the same age can vary. This figure demonstrates that, depending on the relative amount of 13 C, which is a measurement of isotopic fractionation, the radiocarbon measurement of artifactual materials can provide an age estimate that is either too old or too young. Because of the rela-

tively low amounts of 13 C in fat and blubber, these types of materials appear younger than they are. On the other hand, shells made of carbonate or mammoth ivory have high 13 C values and appear older. Because the isotopic-fractionation values of different materials can be measured, corrections can be applied to the radiocarbon measurement. This provides a closer estimate of the age of the artifact and allows the radiocarbon ages derived from isotopically different materials to be compared.

ample, limestone or coal), and geochemical contamination from nearby organic-decay products like humus. It is critical to ascertain that no recycled older carbon is in the sample. Three methods of calibration relate radio-

carbon years to real years, or relate radiocarbon chronologies to calendar dates. Radiocarbon chronologies covering the past 10,000 years or so have been calibrated by tree-ring and varve chronologies. Radiocarbon ages of about 30,000 years

152

Estimating Time

have been calibrated by chronologies based on uranium-series dates (fig. 5.14). Tree-ring derived calibration curves are applicable only to samples formed in equilibrium with atmospheric CO 2 . Deep ocean water is not in equilibrium with the atmosphere. Upwelling of deep water occurs near many coastlines, causing disequilibrium in surface waters and in the shells of mollusks and marine microfauna that derive their carbonate from these waters. Upwelling is affected by the shape of the coastline and the bottom topography, local climate, and wind and current patterns. Hence correcting for this disequilibrium is a local problem. For maximum accuracy, each coastal environment (for example, estuary) must be evaluated to determine the magnitude and variability of marineshell carbonates of equivalent age. For a generalized calibration of samples of marine origin a curve has been developed by the University of Washington at Seattle, and published in the journal Radiocarbon (a computer program is also available), derived from the tree-ring information but modified by carbon-reservoir modeling. Uranium-series calibration of the radiocarbon timescale has been used to investigate the differences between calendar years and radiocarbon years for the final part of the Upper Pleistocene, where varve and tree-ring calibrations are not available.34 Before 9,000 b.p. 14 C ages are consistently younger than true calendar ages. Comparisons between U-Th ages and 14 C, as reflected in figure 5.14, indicate this divergence. Figure 5.14 also shows that there is reasonable agreement between the Lake of the Clouds, Minnesota, varve sequence and dendrochronological ages. The extension of the radiocarbon-calibration curve indicates that the deglaciation dated to about 18,000 radiocarbon years b.p. is associated with a calibrated date of 22,000–21,000 b.p.

Other Dating Methods Fission-Track Dating Fission-tracks are produced when alpha particles created by spontaneous fission (of 238 U) leave a trail of damage across minerals (including mica, apatite, and zircon) and glass.35 To examine these,

the researcher polishes and etches a previously unexposed surface of a sample. This enlarges the size of tracks; the researcher then counts the fission tracks, which can be seen through a microscope. The age of the sample is determined by the number of tracks per unit area: the higher the number of tracks, the older the sample. Fissiontrack dates are a measure of the time elapsed since the substance was solid enough to retain tracks. The age range amenable to fission-track dating is from about 20 years to 1.5 million years. The major problem in fission-track dates is that there may be thermal annealing, which causes the tracks to fade. In addition, the concentration of uranium in a sample must be high enough to produce a high-track density, and the uranium distribution must be sufficiently uniform. Materials that can be dated include obsidian, minerals, archaeological glasses, and ceramics.36

Paleomagnetic and Archaeomagnetic Dating Both paleomagnetic and archaeomagnetic dating techniques rely on the phenomenon of the earth’s magnetic poles changing in space and time.37 Paleomagnetism relates to geologic deposits and archaeomagnetism to archaeological materials or features. The basis of the dating techniques is that the earth’s magnetic poles ‘‘wander’’ (show secular variation) and ‘‘flip’’ (reverse direction). Iron dipoles in minerals in crystallizing igneous rocks align with the earth’s magnetic field at the time of their formation or deposition. The earth’s magnetic field also affects magnetic particles during sedimentation, influencing them to become oriented parallel to the field. This creates detrital remnant magnetism, whose direction and intensity mimic the earth’s magnetic field at the time of deposition. By establishing a chronology for the change in the earth’s magnetic field, it is possible to date rocks by matching their magnetic orientation against the master record. There is now a master curve recording secular change of the geomagnetic field for the past 10,000 years (fig. 5.16). Many archaeological features contain magnetic materials. Because the earth’s geomagnetic poles migrate over time, materials that became magnetized at a specific time can be dated by com-

153

Estimating Time

Figure 5.16 Typical Magnetic Inclination and Declination Compared through Roughly Opposite Reversals Two patterns of change can be used to study inclination and declination of the magnetic field. This sequence has

been proposed as the standard curve recording magnetic secular variation during the Holocene in western North America. The roman numerals denote volcanicash deposits within the sequence.

parison with established records of geomagnetic change (fig. 5.17). The fundamental difficulty lies in finding materials that have not been moved after they acquired their primary magnetization and that have not received a secondary magnetization with a different orientation. Archaeomagnetism has been useful in resolving the problem of dating Hohokam canals in southwestern North America.38 The detrital remnant magnetism in the sediments demonstrated that most of the dates clustered around a.d. 900– 1000. At one site, from La Lomita, detrital remnant magnetism results could be compared with diagnostic ceramics. Archaeomagnetic dating indicated that the oldest canal dated to a.d. 910– 1025. A slightly younger canal, dated to a.d.

1000–1100, was associated with late Sacaton phase artifacts and was used around the time of the Sedentary to Classic transition. The youngest canal contained Soho-phase (Classic period) artifacts. Archaeomagnetic results of clay deposits associated with the Soho-phase feature suggested a date of between a.d. 1165 and 1350. In addition to the standard set of techniques for obtaining information on magnetic variation through time, archaeological structures may be of use. In a unique study, the orientation of Turkish structures (which, by Islamic law, must be aligned toward Mecca) built from the late eleventh century onward was used to determine whether they contained information on the earth’s secular magnetic variation. Some of the mosques were no

154

Estimating Time

Figure 5.17 Magnetic Pole Migration Archaeomagnetic dating and changes in the location of magnetic north since A.D. 600. The migration of the magnetic pole can be used to determine the age of archaeological features like burned earth associated with

archaeological features like hearths. The location of the magnetic north determined from the archaeological feature is matched with the closest location on a curve like this to estimate the age of the archaeological materials. (After Eighmy 2000)

longer oriented toward Mecca. Some of the builders’ mistakes seem to have resulted from using a magnetic compass without correcting for magnetic declination. It may thus be possible to use this type of error to study magnetic variation in the past.39 Paleomagnetic analyses have been used to study the ages of Paleolithic artifact occurrences and/or fossil hominid remains in Africa, Europe, and Asia (fig. 5.3).40 In South Africa, the locality of Kromdraai consists of two nearby sites that contained the first specimen and type fossil of Paranthropus (Australopithecus) robustus as well as Oldowan and Acheulian (Lower Paleolithic) artifacts. The fossil remains were discovered in 1938 and are similar in morphology to the ‘‘Zinjanthropus’’ skull termed Australopithecus boisei found within

deposits of Bed I at Olduvai Gorge in East Africa (Tanzania). Breccia similar to the deposit associated with the type specimen of Paranthropus robustus and a flowstone above the breccia are related to a period of reversed paleomagnetic polarity. Talus cone deposits overlying the breccia and flowstone are associated with an interval of normal polarity within the Olduvai Event (the ‘‘Olduvai Geomagnetic Subchron’’) dated at 1.95– 1.77 million years ago. In the Republic of Georgia at the site of Dmanisi, fossils of Homo ergaster/erectus and Oldowan artifacts were discovered in deposits overlying a basalt with an age based on the 40Ar/ 39Ar method of about 2.0 million years. The basalt and the sediments immediately above it had normal geomagnetic polarity and were correlated with the

155

Estimating Time

Olduvai Subchron. The Oldowan artifacts and hominid remains were found in deposits with reversed polarity and correlated with the Matuyama Chron (1.77 to 1.07 million years ago). Combined with the age of the basalt and the late Pliocene– Pleistocene (Villafranchian) affinities of the associated fauna, the paleomagnetic data indicates an age of about 1.7 million years for the artifacts and Homo erectus/ergaster skull. Paleomagnetic methods have also been used to estimate the age of the earliest presence of humans in northeast Asia. Here, in the Nihewan Basin of north China, close to the Sanggan River, flaked stone Paleolithic artifacts were found with Villafranchian-type mammal fossils. The paleomagnetic record consisted of three normal and two reversed magnetozones. The lowermost zone with normal geomagnetic polarity was correlated with the Olduvai Subchron. Deposits overlying this contained the artifact layer and had reversed polarity correlated with the Matuyama Chron. Above this were two normal polarity events linked with the Jaramillo Subchron (1.07–0.99 million years ago) and the Brunhes Chron (starting at 0.78 million years ago). These examples demonstrate some of the ways that paleomagnetic records have been used to assist in determining the ages of Lower Paleolithic artifacts and PlioPleistocene hominids. The Yuanmou Basin in southwest China includes Late Cenozoic sediments containing hominid fossils, notably Homo erectus. 41 A newly proposed depositional sequence consists of, in ascending order, alluvial fan, ephemeral braided river, sandy braided river, ephemeral gravelly braider river, and alluvial fan systems. The age of these sediments is critical to the clarification of the timing of the earliest migration of Homo erectus to Southeast Asia. A reconstructed magnetostratigraphy has revised the age of the Homo erectus deposits to the early Brunhes chron.

Electron Spin Resonance and Luminescence Dating Both electron spin resonance (ESR) and luminescence dating are based on the accumulation of trapped electrons in minerals; the more trapped electrons present, the older the age of the sam-

156

Figure 5.18 Growth Curves for ESR Dating Curves demonstrating changes in ESR with increased time or dose. The accumulated dose or signal intensity increases until it reaches a plateau or point of saturation.

ple.42 The ongoing formation of electron traps in minerals results from the bombardment of the crystal structure by radiation from radioactive elements within and adjacent to the minerals. The increase in paramagnetic defects with time in crystalline solids is the basis of ESR dating, which provides an age estimate by directly measuring the number of trapped electrons in minerals within artifacts or sediments. The ESR signal of a crystal is proportional to the number of paramagnetic radiation-induced defects. Where a constant or known level of past radiation can be assumed or determined, the ESR signal should be proportional to the age of the crystalline sample (fig. 5.18). Two parameters need to be determined to obtain an ESR date: the Total Dose (TD), also called the Accumulated Dose (AD), and the annual radi-

Estimating Time

ation dose rate that a sample has received. The TD is the total radiation dose that a specimen has received since its formation or its last effective heating (which would drive the electrons from the traps). The ESR age is equivalent to the Total Dose divided by the annual dose. An estimate of TD is back extrapolation to zero ESR intensity, which corresponds to the time of the formation of the sample. The TD is generally not difficult to determine using the additive-dose method, but several possible interference factors need to be considered. These include changes that are due to grinding, bleaching, humic acid radicals, pressure effects, and the thermal stability of the trapped electrons. The major application of ESR has been to Quaternary carbonates. Fossil bone, teeth, and shell have also been dated using ESR. The application of ESR to the dating of spring-deposited travertines is more difficult if the content of manganese is high. Variation in the geochemical content of water associated with lacustrine sediments makes it difficult to evaluate the annual dose and, consequently, to determine ESR ages on lacustrine carbonates. Travertine samples from the El Kown archaeological site in Syria have been dated using ESR. The ages were then compared with uraniumseries ages. The oldest of the travertine deposits from this open-air site gave ESR ages of 216,000 years and uranium-series ages of 245,000 years. Some uranium-series and ESR dates did not correlate well, but most fell in the range of 160,000– 80,000 b.p. An unexpectedly young sample of about 18,000 years old provided by ESR was confirmed by a uranium-series age of about 17,000– 15,000 years. It is possible that the disconjunctive dates were the result of postdepositional changes. Dating spring-deposited travertines has been attempted at several sites in Hungary. External dose rates had to be assumed for all the sites, and internal dose rates were calculated using U and Th contents and uranium-series dating. When these were compared with uranium-series dates, the ages were generally close. Travertine spring deposits above and below the ‘‘Paleolithic stratum’’ at the site of Tata, Hungary, gave uraniumseries ages of 98,000 and 101,000 years, and ESR ages of 81,000 and 127,000 years, respectively. The

other dates did not match. Eolian quartz grains and gypsum may be used to obtain ESR signals. Eolian and stream-sedimented quartz grains have been studied from the site of Arago, France, and gypsum has been analyzed from the Mammoth Cave, Kentucky, area. Thermoluminescence (TL) is the light emitted by a crystal as it is heated. This light is proportional to the age of the sample: it increases with older samples. The release of electrons from the traps at defects in the crystal structure produces the light. It is necessary to measure the background radioactivity of a sample to obtain an age estimate with the TL technique. The TL produced by past exposure to radiation can be compared with induced TL in the laboratory. This allows the past radiation dose to be determined (fig. 5.19). The upper age limit for TL dating is about 250,000 years for quartz and 500,000 years for feldspars. It is important to note that exposure to sunlight can free trapped electrons and thus ‘‘reset the clock.’’ 43 Thermoluminescence has been applied to the dating of pottery, continental and marine sediments, burnt rock, igneous rocks, loess and eolian sands, alluvial and lacustrine deposits, calcitic formations, and shell.44 It has been especially useful when dating igneous rocks too young to be dated by the potassium-argon method. Continental volcanics have been dated with TL measurements of feldspars. The reliability of TL dating of stone tools depends on the accuracy with which the internal and external radiation dose rates during the period of burial can be measured. Determining the internal dose rate is straightforward, but measuring the external dose rate can be very difficult depending on the enclosing sediment and the diagenetic changes during prolonged burial. At the important Lower and Middle Paleolithic site of Tabun Cave in Israel, flint artifacts were selected that had been recovered near a sedimentary sequence that could be analyzed through dosimetry. To obtain a measurement of the cave’s external dose rate forty-six dosimeters were placed as close to the locations of the flint artifacts as possible. To measure the paleodose, the outer 2 mm of each artifact was removed, and the remaining fragments were analyzed after treatment

157

Estimating Time

Figure 5.19 Additive Dose Technique in Thermoluminescence Dating Dating an ash fall using thermally stimulated luminescence, or thermoluminescence. The lower graph shows the presence of a plateau for the equivalent

dose. The empty circles that form this plateau are the result of using the additive-dose technique. The upper graph shows a TL build-up curve. The estimated age of this ash is 7,800 years.

to remove carbonate. The paleodose was measured according to the second TL growth curve. The internal dose rate was calculated using instrumental neutron activation analysis (INAA) for concentrations of 238 U, 232 Th, and 40 K. Radiation doses of sediments that surrounded the artifacts were also measured. Based on the TL rates on flints a new chronological timescale was proposed for the site. The timescale was older than it had seemed in earlier proposed chronologies, one of which was based on ESR. Units XIII to II were assumed to date from 330,000 to 210,000 b.p. (correlated with isotope stages 9 and 8). The lower units, XIII—XI, are associated with Acheulian and Yabrudian artifacts, while IX—II contain Mousterian artifacts, indicating a possible transition to the Middle Paleolithic around 300,000 b.p.

Unit I, which also contained Mousterian artifacts, was probably deposited around 171,000 b.p. (correlated with either the end of isotope stage 7 or the early part of stage 6). One of the important consequences of these TL dates is that they place the beginning of the Mousterian back to about 270,000–250,000 b.p., which is similar to the earliest known Middle Paleolithic in the Sahara Desert. Another archaeological consequence is that Neanderthals may have been in the Levant about 170,000 years ago (oxygen isotope stage 6). There is also the possibility that the burnt-flint TL chronology indicates the existence of archaic forms of Homo sapiens more than 250,000 years ago.45 A loess sequence in northern Pakistan dated by TL helped to assign a site to the Paleolithic.

158

Estimating Time

The archaeological remains included conjoinable blade tools, associated debitage, and a structure that was probably a small shelter.46 The site contained no hearths or charcoal, so TL dating of the loess was the only available method. Because the loess overlies the artifact accumulation, it provides a minimum age for the site. The TL dates from the site could be grouped into three sets. The youngest (uppermost) loess was deposited between 27,000 and 24,000 b.p., or just before the Last Glacial Maximum (LGM). Older dates in the range of 47,000–42,000 b.p. and 64,000– 59,000 b.p. were obtained from loess overlying the archaeological zone. The older dates may represent sediments that were redeposited and not entirely zeroed-out before their redeposition. Because of this, they may have ‘‘inherited’’ TL associated with an earlier depositional event. It was concluded that the Paleolithic artifacts dated to around 45,000–42,000 b.p. and that the loess cover dated to the last glacial. Before the use of TL dating, an age assignment for these materials would have been based on stone-tool typology or the tenuous association of sites with river terraces. Thermoluminescence has been extensively used for dating loess deposits containing paleosols.47 Thermoluminescence measurements were used to estimate the age of soils developed within a loess sequence in western Europe. The loess correlated with stadial events and soil formation associated with interglacials and interstadials. The TL dates ranged in age from about 140,000 to 13,000 years. There are two other ways luminescence has been used to derive archaeological chronologies. Whereas TL is stimulated by heat, optical stimulated luminescence (OSL) is stimulated by visible light, and infrared stimulated luminescence (IRSL) by infrared light. An example of IRSL dating of sediments that are directly related to ancient human activity is the work on the Neolithic site of Bruchsal Aue in Germany.48 Exposure to daylight appears to wash out any previous IRSL signal. This resets the luminescence clock, even in redeposited colluvium exposed to light for as little as thirty minutes. The amount of new luminescence energy stored after resetting was used to date colluvial sediments deposited as

a result of erosion. The erosion was thought to have been initiated by human activities like forest clearing and agriculture. The archaeological site of Bruchsal Aue contained traces of both an Early and a Late Neolithic settlement. During the time Bruchsal Aue was occupied, defensive trenches were dug into a late-glacial loess. Later, after the settlement was abandoned, colluvial sediments accumulated in the trenches. Some trenches show periodic episodes of colluvial fill separated by intervals of erosion. Sediment samples were collected and analyzed from the late-glacial loess, the three stratified colluvial layers, and trench and pit fill (dug either into the loess or into one of the colluvial deposits). The IRSL dates on loess ranged from 14,700 to 10,9000 years ago. Pit fill and the earliest colluvial layer dated around 7,000–6,000 years ago. Less luminescence had accumulated in the stratigraphically higher samples, and the IRSL ages of these sediments were consequently younger. The luminescence dates from the site of Bruchsal Aue have been used to argue that colluvial deposition in Germany was related to high population density and intensive land use starting in the Neolithic and continuing through the medieval period. Because the IRSL method dates a sedimentary deposit’s last exposure to light, it is a direct method of dating the accumulation event. It is thus a potentially valuable tool to understanding human-landscape interactions. Furthermore, IRSL is based on feldspars, common minerals in the earth’s crust, not on rare constituents or charcoal or bone fragments. Luminescence dating has been used to resolve the ambiguity often found in dating the construction of ancient earthworks such as mounds.49 M. A. Smith and associates have compared the radiocarbon and luminescence chronologies over 35,000 years from a rock shelter in Australia.50 Both techniques produced self-consistent chronologies, but TL ages were generally older than 14 C ages. The radiocarbon chronology is supported by sedimentary/paleobotanical evidence, stone artifact typology, and other archaeological data.51 Optically stimulated luminescence (OSL) has been applied to sediments deposited in fluvial terraces, lake deposits, and cave sediments.52 This

159

Estimating Time

method is especially promising because it provides an independent method of dating the evolution of Quaternary landscapes potentially linked to the presence of prehistoric humans. In Europe, along the Loire and Arrox Rivers in France, for example, OSL measurements were used to date a series of stepped-terraces produced by intervals of aggradation and incision. The highest three terraces (T8–T6) dated from about 125,000 to 90–40,000 years ago. The next lowest terrace (T5) had ages of about 23,000–18,000 years ago, approximately the same as the Last Glacial Maximum (LGM). Terraces below the LGM deposits were mostly Holocene in age. In North America, OSL and infrared-stimulated luminescence (IRSL) measurements were also used to estimate the age of deposits associated with glacial Lake Great Falls, situated in Montana at the southern end of a potential migration route for Pleistocene humans traveling along the eastern front of the Rocky Mountains.53 Studies undertaken by James Feathers and Christopher Hill were used to help determine the timing and extent of continental glaciation on the Great Plains near the Rocky Mountains.54 Based on the OSL and IRSL measurements it was possible to assign the glacial lake-related deposits to oxygen isotope stage 2 and to place them within a model of landscape evolution connected with potential Late Pleistocene human presence in the region (see fig. 5.11). In southwest Australia, at the cave site called Devil’s Lair, OSL ages were combined with ESR, uranium-series, and 14 C dating to infer that people had migrated to this region by about 50,000 years ago.

Temperature-Affected Dating Amino Acid Racemization and Epimerization Amino acid geochronology relies on the fact that through time L-amino acids (left-handed) convert to a D-amino acid (a right-handed configuration). This conversion is termed racemization. The D/L ratio moves from being wholly L to being an equal mixture of D and L. Early research concentrated on the use of bone (includ-

160

ing human remains) and mollusk shell, but wood, organic sediments, and ostrich eggshell have also been used. Major problems with the application of amino acid geochronology to fossil bone include the potential of geochemical contamination because of the porous nature of the material and complications that may arise from the degradation of collagen in the osteological material. Although relative age estimates have been obtained using mollusks, nonlinear kinetic reactions complicate the conversion of the ratios to absolute ages. The three primary controls in amino acid dating are time, ambient temperature, and moisture. Conversion from one type of amino acid form to another strongly depends on temperature. One method of obtaining current temperature conditions, which provide a way to estimate past temperatures, is to place sensors into the sediments surrounding the object being dated. High moisture promotes the racemization process; therefore, samples subjected to extremely dry conditions may have lower D/L ratios. The ostrich eggshell amino acid isoleucine has been used to estimate the age of deposits that contain artifacts.55 In contrast to mollusks, there is less potential for leaching because the amino acids are within the calcite crystals of the eggshell. Conversion of D/L ratios into absolute ages can be done when the effective ambient temperature and the D/L ratio are available. Calibration can be accomplished by comparison with other dating techniques.

Hydration (Obsidian) The rhyolitic volcanic glass obsidian, extensively used as the raw material for making stone tools, can be useful for relative dating.56 A hydration ‘‘rind’’ can develop on obsidian as a result of its absorbing water. The thicker the hydration rind, the older the sample. A fundamental assumption is that obsidian hydrates at a given rate. Hydration in obsidian follows the diffusion equation x 2= kt, where x = hydration thickness, k = hydration rate, and t = time. Several variables influence the rate of the hydration, including the chemical composition of the obsidian, the available water, and the temperature associated with the environmental

Estimating Time

Figure 5.20 Hydration Dating Dating by accumulation as illustrated by hydration dating. Some dating techniques depend on the rate of addition of a particular property over time. In this case two rates of accumulation are depicted: a linear rate and an exponential rate. In both instances the thickness or depth of hydration can be matched to a particular age. There is a direct relationship between the thickness of the hydration rind and the age of the sample—increasing hydration thickness is linked to increased time. (Based on Pierce and Friedman 2000)

context of the artifact. When the geologic source of the sample is known, a hydration constant for that source plus an estimate of the temperature during hydration will provide a hydration rate for that material at that location. To obtain relative age estimates, the thickness of the hydration rind needs to be compared with a local hydration rate (see fig. 5.20). Where there is reason to believe artifacts have undergone uniform thermal histories, it is possible to construct relative chronologies. Attempts have also been made to construct more absolute timescales using hydration rates. When a piece of obsidian acquires a fresh surface, as when it is fractured during sediment transport or flaked to make a stone tool, it begins to absorb water. If the fresh exposure of the obsidian surface can be related to a specific pre-

historic event (like the flaking), the thickness of the hydration rind provides an indication of the time elapsed since that event. Many factors affect the hydration rate in a specific location, including the site-specific microclimate, the orientation of the site, the burial depth of the artifact, and past changes in the environmental context. These factors affect the temperature and relative-humidity history of the artifact. Heating caused by forest fires, by fires that destroy human occupations, or by ritual activity (like cremation) may alter the hydration rind thickness through dehydration. It has been shown that the hydration rind can be completely eliminated with no change in surface appearance if obsidian is heated to 430°C. Hydration rims can apparently also undergo spalling. When this happens, a fresh hydration rim forms, giving what appears to be a more recent date for the artifact. Hydration measurements can help infer the presence of mixed components in an artifact accumulation if one knows the variation within a component of a single age. D. Clark and A. McFadyen Clark used the example of specimens collected from a single sealed dwelling floor that were all part of the same occupational event.57 Even in this situation the hydration measurements may not be tightly clustered. To accommodate natural variation within a single age set, an average rind thickness can be used. The usual method used to measure the thickness of the hydration rind has been optical microscopy. However, tests conducted among different hydration-measurement laboratories indicate that measurements are subject to operator variables. To reduce this, some laboratories use computer-assisted imaging technology to digitize and measure hydration rims. An effort has been made to use the thicknesses of hydration rims on artifacts to provide absolute-age determinations. The hydration thickness can be calibrated by experimental determination of the hydration rate. Once a temperature-dependent rate constant has been obtained for a given obsidian source, it can be used for all artifacts made from that source. Where the site is already dated by other means, it should be possible to use the thickness of the hydration rind to estimate the temperatures associated with the

161

Estimating Time

site. In this way, hydration measurements are potentially a useful paleoclimatic proxy. The depth of burial of an artifact can have a major influence on the degree of hydration. Artifacts nearer the surface may be subjected to higher variations in temperature, and relative humidity varies with depth. At depth, relative humidity can be 100 percent, but nearer the surface it may be much less. Reported ages in obsidian dating range from greater than 100,000 to about 200 years old. Hydration measurements have also been used to date glacial moraines. Rocks fractured during glacial transport have fresh surfaces that can develop a hydration rind. This allows us to date the moraine and the timing of the glacial advance. In the same way, cobbles and boulders made of obsidian that were fractured during transport as part of glacial outwash can be dated. This helps us determine the time since the glacier melted. Archaeologists can use the age of glacial melting to evaluate the possible timing and duration of human occupations of glaciated regions. One potential use of the obsidian hydration is in dating archaeological site formation. Joe Michels used hydration measurements to test the chronological integrity of midden deposits and to evaluate the stratigraphic implications of artifact mixing and reuse at the Mammoth Junction site in California.58 After measuring the hydration rim on obsidian artifacts from each unit of deposition, a three-dimensional plot of the distribution of hydration-rim values and depositional units was constructed. Based on the premise that artifacts of the same age would have the same rind thickness despite their depth, significant differences in hydration measurements for artifacts in the same depositional units were interpreted as indicating the presence of mixing. To obtain an average age for artifacts within a depositional unit, the median hydration rim value for each unit was calculated. Although the artifacts were extensively mixed, the slope of the trend line showed that they conformed to the principle of superposition. Michels dated artifacts from Kenya as far back as about 120,000 b.p. using experimentally derived hydration rates for a particular obsidian.59 First he determined experimentally the hydration rate for two different obsidians. Then, the source of the obsidians was determined by comparing compo162

sitional characteristics of artifacts with those of quarry sites. Obsidian artifacts from the Prospect Farm site were recovered from four main episodes of occupation. The oldest artifacts were related to the Middle Stone Age, followed by artifacts from the early Later Stone Age, and by the youngest, from the Pastoral Neolithic. Because of spalling, many artifacts from the Middle Stone Age were calculated to have ages that are probably younger than the actual age of the artifacts. The calculated age for unspalled surfaces of Middle Stone Age artifacts was about 120,000 to 50,000 years old. Hydration rates from the early Later Stone Age put the artifacts in the age range of 33,000–22,000 years old, while artifacts from the Later Stone Age have hydration ages of around 10,000 years old. Neolithic artifacts were dated to about 3,000 years ago. In this study, the hydration rates used to calculate the dates from the Prospect Farm site were calibrated from the 14 C chronology. Obsidian hydration measurements were used to help date a Tecep phase burial at the Mayan site of Nohmul in Belize.60 Obsidian artifacts, some freshly struck, were found in a sealed burial. This provided a chance to date the burial and evaluate whether it was the same age as the structure or a later intrusion. First the obsidian source was determined. All of the material matched the El Chayal source near Guatemala City, Guatemala. Based on the source and estimated effective hydration temperatures, hydration dates were calculated. The different ways of estimating effective hydration temperature change the calculated age by about a thousand years. Using the thermalcell estimate of temperature, the dates mostly fell around a.d. 950, while using the meantemperature estimates they tended to fall around a.d. 1050. Timbers of the building containing the burial were radiocarbon-dated to about a.d. 700. Hence the obsidian-hydration dates were interpreted as indicating that the burial was intrusive.

Dating Techniques Based on Chemical Accumulation Chemical Analysis Dating based on chemical content can be either relative or absolute. In relative dating, the prin-

Estimating Time

ciple is straightforward and based on the concept that certain chemical constituents will accumulate in (or on) an object over time. The greater amount of the chemical present, the older the material. The chemical composition of buried bone can change through time, so chemical analysis of bone has been employed to test the validity of human remains and extinct fauna. Human skeletal remains recovered from near Midland, Texas, had a chemical content within the range found in extinct faunal materials, but it differed from the chemical content found in Late Holocene bones. Uranium-series dates have independently confirmed the Terminal Pleistocene age of the Midland human remains, although the radiocarbon and uranium-series dates are not strictly comparable. One of the criteria that led to the conclusion that the human remains from Midland were ‘‘as much a Pleistocene fossil as any of the extinct vertebrates discovered at this site’’ was a comparison of chemical properties.61 Rabbit bones from modern sediments had low levels of fluorine, while those of horse and human remains had higher fluorine values. Some caution needs to be used in interpretations based on relative amounts of elements in bones, because more than one set of postdepositional circumstances can result in bones having the same chemical content. Chemical dating was also used to confirm that the bones from the infamous Piltdown assemblage did not come from a single animal and were not contemporary. The hominid remains were attributed to gravel that contained Pleistocene fauna. Chemical tests demonstrated that jawbone and teeth had essentially modern concentrations of fluorine, while the Piltdown skull bones were Pleistocene in age, as were the fossils of the extinct animals.62

Dating Exposed Surfaces Dating erosional or depositional surfaces, as opposed to the material or deposits underlying those surfaces, is a burgeoning field of study of great importance in geoarchaeology. The investigation of weathering rinds has been particularly useful. The critical rate of weathering must be established by other, independent dating methods.

Weathering rinds can be used to date artifacts (see obsidian hydration dating, above), deposits, and landscape surfaces. All landscape surfaces are subject to weathering processes that result in physical and chemical alteration. Once a geomorphic surface has been stabilized, soil formation and weathering begin immediately. Exposed surfaces of rock outcrops and clasts eroded from the outcrops will exhibit increasing surface alteration with time. Rind-dating is particularly successful in regions where a single rock type is widely distributed.

Patination and Desert Varnish The use of weathering and chemical alteration of artifacts to estimate their age is based on the principle that the thickness of an alteration rind provides an estimate of age. Chemical accumulation on materials is known as patina, and it is produced by weathering. Desert varnish (cation ratio) dating provides another method of directly dating artifact surfaces. The basis of cation ratio dating is that certain elements (cations) leach out from lithic material faster during weathering than others. For instance, potassium (K) and calcium (Ca) are more easily leached than titanium (Ti). The ratio of K and Ca to Ti provides an indication of relative age. The lower the ratio of K and Ca relative to Ti, the older the sample. Absolute ages for these lithics can sometimes be determined from correlations and extrapolations based on other dating techniques. Varnish dating using the cation ratio technique has been cited by proponents of human occupation in the Americas before the Clovis Complex.63 Cation ratio varnish dating was used as supporting evidence for radiocarbon dates to propose dates earlier than 11,500 b.p. for petroglyphs and stone artifacts. In this instance, the dates obtained from radiocarbon and cation ratio dating provide limiting ages for older artifacts.

Cosmogenic Nuclides Six cosmogenic nuclides are available for study of geologically recent chronologies and rates of surficial processes. These nuclides are: 3He, 10 Be, 14 C, 21 Ne, 26Al, and 36 Cl. They span the range of 10 2 to 10 7 years and can be used to date surface rock exposure of nearly all lithologies and at 163

Estimating Time

any latitude and longitude. Currently, only 14 C has been used extensively in archaeology but 10 Be and 36 Cl potentially have important applications. Using cosmogenic nuclides, the ages of major climatic change can be directly determined in alluvial sediments.64 The isotope 10 Be is created in the atmosphere by cosmic-ray activity, but it is precipitated as an aerosol after an atmosphere residence time of two years. This is shorter than the residence time of gaseous 14 CO 2 . Accumulation of 10 Be in sediments is complicated by remobilization and patterns of deposition and climate conditions have a great effect on the precipitation of 10 Be. How-

164

ever, the much longer half-life of 10 Be (1.5 million years) allows it to see much farther back into the past than 14 C. Another landscape dating method is based on the accumulation of cosmogenic chlorine-36 ( 36 Cl) on exposed rock surfaces. Because of its half-life of 100,000 years, the 36 Cl method can be used to date landforms constructed during the past two million years. Intense chemical weathering can complicate the matter by mixing 36 Cl that comes from the rock with 36 Cl from the atmosphere, so appropriate corrections need to be made. Chlorine-36 has been used to date meteorite craters, glacial deposits, and lava flows.65

CHAPTER 6

Paleoenvironmental Reconstructions: Landscapes and the Human Past Archaeologists ought to be grateful to worms, as they protect and preserve for an indefinitely long period every object, not liable to decay, which is dropped on the surface of the land, by burying it.—Charles Darwin 1881

Environmental and Landscape Change

B

road-scale relationships between landscape evolution, climatic fluctuations, and human activities are a critical aspect of archaeological interpretation. Interactions between the physical, geologic environment and biologic organisms, including humans, can be studied from either biogeographic or geoecologic perspectives. A fundamental aspect of geoarchaeology is the interpretation of the prehistoric record in terms of past patterns of interactions and interrelationships between humans, other organisms, and their physical habitats. This area of geoarchaeology is an aspect of paleogeoecology, the study of interactions between prehistoric life and environmental landscapes. Site-specific, regional, and global changes in the physical and biologic environment are reflected in the landscape context and are often directly influenced by climatic factors. Climate change itself is a complicated area of study. Climate change is the result of global, regional, and local geologic changes that influence atmospheric and hydrologic circulation patterns. Inferred connections between changing depositional environments, geomorphic contexts, and climatic processes are valuable tools for evaluating both spatial and temporal patterns of human behavior. In almost all instances, the visi-

bility of an artifactual accumulation will depend on the patterns of landscape development with which it is associated. The sediment-soil system components of the landform environment contain a variety of biologic and chemical components that can be used as paleoecologic indicators (see Chapter 2). These constituents provide a means to infer past environmental conditions and to study their relation to climate change. Human behavior has had an increasing influence on the environmental landscape, as reflected in the sediment-soil record of erosion, deposition, and landscape stability as well as in the paleoecologic indicators found in Quaternary deposits. The geologic and biologic records of paleoclimates are stored in a variety of materials embedded in terriginous and ocean sediments and in ice sheets. Each of these proxy records provides only a limited local picture, but taken together they present a good general representation of regional and, in some cases, global patterns of Quaternary climates. In North America, for instance, paleoecologic data derived from pollen in lake sediments that have accumulated for the past 12,000 years allows us to reconstruct the habitats of early human occupation. In parts of the Old World there are paleoecologic records of the Plio-Pleistocene that provide clues to the environments associated with early hominids (australopithecines and early and archaic Homo), the Late Pleistocene associated with anatomically modern humans, and Holocene human populations. Accurate time control (see Chapter 5) is essential for paleoclimate reconstruction. For the Late Quaternary, especially the past 30,000 years, accelerator mass spectrometry (AMS) dating has helped provide this control. Other methods, such 165

Paleoenvironmental Reconstructions

as potassium-argon (K-Ar), argon-argon (Ar-Ar), luminescence and electron spin resonance (ESR), and uranium series, have been useful for dating older paleoclimate records and relating them to the artifactual evidence of past human activity.

Inferring Environmental Change A variety of physical and biologic information can be used to reconstruct environmental and climate change. For example, our understanding of the postglacial ecology of North America rests on a combination of geologic (geomorphic, sedimentologic), chemical (including isotopic), and paleontologic (biotic) evidence. From these types of data we know that, although the last remnants of the Laurentide Ice Sheet in North America did not disappear until about 7,000 b.p., the Early and Middle Holocene (10,000–4,500 b.p.) was perhaps warmer than the period since 4,500 b.p. (the warmer period is sometimes called the ‘‘climatic optimum’’). Some Early Holocene warmth may have been seasonal rather than year-round. Changes in precipitation also marked the first half of the Holocene. An estimated 20 percent decrease in rain led to an eastern extension of the ‘‘Prairie Peninsula’’ in the Great Plains of North America. Western North America was drier during the climatic optimum. The human response to the changes resulting from this geoecologic, climatic-biotic interaction is reflected in the patterning of the archaeological record. This patterning is partly a result of the geologic processes of erosion, deposition, and soil formation. However, much of the artifactual patterning shows past human response to changing ecologic habitats. Human adaptation to changing plant and animal arrangements instigated by climate change may explain at least some of the variation within the Paleo-Indian–Archaic– Woodland archaeological succession (see fig. 5.5). In protohistoric and historic times many significant climate changes have affected human habitation and agriculture. In Europe, a cool period beginning about 4,500 b.p. and ending about 2,500 b.p. caused an expansion of mountain glaciation. This cooling interval can be seen in the geomorphic evidence of glacial moraine deposits 166

and in such biotic indicators as pollen. Although a misnomer in strict archaeological terms, this period has been called the ‘‘Iron Age cold epoch.’’ The climate became warmer during the beginning of the Roman Empire, and the so-called Dark Ages in Europe (a.d. 500–1000) coincided with a return to colder climates. After the year a.d. 1000, a warmer regime allowed the settlement of Iceland and Greenland. Alpine passes between Germany and Italy became free of ice. Physical and biologic data support the impact of climatic change on human populations during the latest part of the Holocene. One example is the Little Ice Age, in which mountain glaciers advanced. The Little Ice Age began about a.d. 1450 and did not conclude until around a.d. 1890. The suggestion has been made that if not for the possibility of a human-induced greenhouse effect, the Little Ice Age might have been the start of another interval of global glaciation. This period had two main cold stages, which roughly coincided with the seventeenth and nineteenth centuries. Although it is sometimes thought of as a European phenomenon, the Little Ice Age was global. Other parts of the world experienced changes in precipitation and temperature that had an impact on biotic communities. Prior to the Little Ice Age, Europe had a warm period from approximately a.d. 900 to 1200. The Vikings settled in Greenland during this warm period and, briefly, in North America. By the early 1300s, cold and wet weather created famines in Europe. Mountain glaciers advanced, destroying villages and agricultural land. It is not the purpose here to consider behavioral consequences of natural disasters in detail but rather to consider how to identify and describe such disasters within the geologic and archaeological records. The tools to identify significant climate change include the understanding of which sediments and soils encapsulate the chemical and biologic fingerprints of temperature and hydrologic change. Chapters 2 and 3 provided the geologic and pedologic knowledge that forms the foundation for the use of such tools. The first known European migration and occupation of North America shows how the Little Ice Age directly affected human behavior. During the warmer interval before the Little Ice

Paleoenvironmental Reconstructions

Age, seafaring northern Europeans could travel and settle along the coasts of Greenland and Newfoundland. But with the onset of the Little Ice Age, sea ice expanded in the North Atlantic around Iceland, and the Norse colony established in Greenland was abandoned. The archaeological record suggests that northern European populations were also present in North America. The application of the geoarchaeological or paleogeoecologic perspective helps to explain this artifactual record. In the Middle Ages, Europeans could live in the New World habitat only so long as the traditions and technology they carried with them adapted favorably to the new conditions. To survive the climatically induced environmental changes caused by the Little Ice Age, these people needed to change their behavior, either by adapting to the new conditions or by seeking more compatible environments.1 Climate thus has an important effect on geologic and biotic patterns. Geologic events like volcanic eruptions can cause short-term climate cooling. In addition to the effects volcanic activity has on local and regional environmental habitats, volcanic eruptions can result in global climate change. Acidic and sulfur-rich dust and ash from volcanic explosions have been found embedded in the layers of ice in Greenland, Antarctica, and mountain glaciers. Changes in the amounts of windblown volcanic material have been correlated with changes in climate. By measuring the sulfur content of Greenland ice cores, glacial advances in the past 1,400 years have been correlated to the frequency of volcanic eruptions. The work of Reid Bryson indicates that volcanic eruptions are involved in major climatic transitions and intervals like the Younger Dryas event (around 10,500 b.p.), in which the climate deteriorated, slowing the retreat of continental and alpine glaciers.2

Ecology and Landscape Change Landscape evolution involves changes in both the physical and biologic features of the earth’s surface. One must examine the factors associated with the sediment-soil system and the biologic evidence available in the deposits to analyze the

evolution of the prehistoric landscape and place it within a geoecologic context. The mechanisms that influence spatial and temporal variation in habitat contexts are critical to archaeological interpretation. In addition to techniques based on sedimentologic and pedogenic evidence, a variety of paleoecologic methods based on biologically derived contents or chemical signals found in deposits can be employed (see Chapters 2 and 3). The inference of past environmental conditions from the remains of plants and animals found in sediments or soils relies on the same kinds of taphonomic and site formation principles used to interpret the archaeological record. The initial, dynamic, living populations of plants and animals can be altered by a variety of pre- and postdepositional processes.3 Paleontology is the study of fossils preserved in sediments. Any remains or traces of plants or animals preserved in the sedimentary record are fossils. In this broad sense artifacts, as traces of past human behavior, are fossils. When geology was developing in the early 1800s, geologists used the evolution and extinction of fossil groups to develop the geologic timescale (see fig. 5.2), in much the same way that archaeologists have since used ceramic or lithic typology to develop chronological sequences and correlations (see fig. 5.5). At any one period of time, many separate communities of fauna and flora live in different environments or habitats. Current theory postulates that populations of organisms, rather than communities of organisms, had different responses to environmental change during the Quaternary. This ecologic aspect of the study of fossil fauna and flora provides geoarchaeology with a wealth of information on ancient environments. Changing patterns in ecologic materials found within archaeological contexts can reflect environmental fluctuations or evolving human adaptations. These changing patterns can be the result of human behavior on a physical landscape. The opposite can also happen. Changing human behavioral strategies can result from alterations of the environment. The Quaternary prehistoric and geoarchaeological records illustrate interactions that can be better understood by using a geoecologic approach. A variety of paleoenvironmental and paleocli167

Paleoenvironmental Reconstructions

matic indicators are found in sediments and archaeologically related deposits.4 These can be divided into biotic and abiotic indicators. Chapters 2 and 3 introduced many ways to interpret abiotic geomorphic and sedimentologic indicators. Apart from biofunctions of soils, biotic indicators are of two types: plant and animal remains. Common types of fossil plant (botanical) remains that can be recovered as part of geoarchaeological studies include pollen, floral macrofossils, phytoliths, and diatoms. Fungi and mosses have also been used to infer past environmental conditions. The faunal (zoological or animal) fossil groups that are used to assist archaeological interpretations are divided into vertebrates (including mammals, reptiles, amphibians, fish, and birds) and invertebrates (like ostracods, mollusks, and insects). Unique accumulation circumstances can provide a combination of paleoecologic elements, such as coprolites and pack-rat middens.5 In addition, trace-element compositions and isotopic ratios of plant and animal fossils are used as environmental indicators. When integrated with abiotic evidence derived from geologic data, these biotic sources of paleoenvironmental information can illuminate the geoecologic patterns contained within the prehistoric record. The study of the past relationships between organisms and their environment relies on our knowledge of today’s ecosystems. When we interpret the biologic evidence for past ecosystems, we depend on a variety of assumptions, including knowledge of the present-day environmental constraints on comparable biota; how (and whether) these constraints can be applied to past biotic distributions; and whether the fossil assemblages being studied are a true reflection of past living systems.

Terrestrial (Non-marine) Geoecologic Data from Lake Records Lake data offer paleoecologic information that can be particularly helpful in understanding past human behavior and climate change. Because of the special depositional and preservational contexts associated with lakes, lake data provide

168

much of our paleoclimatic and related environmental information for terrestrial (continental) environments. Marine sediments provide information on global climate change, while information on local and regional climate change comes from lake deposits. Organic-rich sediments tend to accumulate on the edges of lakes (or in swamps and bogs), while pollen and other paleoecologic indicators can be found in the sediments deposited in the deep-water portions of lakes. Although the sediments trap trace chemical and particle components from atmospheric deposition, much of this material is derived from the lake catchment area or the landscape surface areas that drain into the lake. In Holocene lake deposits, the trace chemicals incorporated into the basin provide a record of anthropogenic pollution that reflects the changing patterns of human landscape use, including such behavior as farming and the introduction and use of new technologies. Records of lake-level changes are derived from many kinds of geomorphic, sedimentologic, and biostratigraphic data. These changes can be the result of climatic fluctuations, local geologic events (such as those related to tectonics), or biotic intervention. Animals can have a direct influence on lake levels. For instance, beavers create hydrologic changes by building dams. Humans also have directly affected lake levels: either they raise water levels by impounding waters or lower them by diverting water that would normally flow into a basin. One present-day example of humans affecting lake levels is Mono Lake in California, where inflowing tributaries have been diverted for use in agriculture and urban centers. In the past, lake levels in the basin fluctuated as a result of climatic change. Major changes in the climate of western North America are recorded not only by geomorphic indicators of lakelevel changes but also by the presence of tree stumps. These tree stumps, examples of macrobotanical remains (see below), are what is left of trees killed by waters that rose because of a change in climate. In many cases shoreline change and variation in water depth can be reconstructed by either biologic or physical and chemical signals. One important aspect of the various forms of paleoecologic data available from lakes is that

Paleoenvironmental Reconstructions

the different forms of information provide independent checks on the paleoclimatic interpretations. Lake-level interpretations based on remnant shorelines (geomorphic data) or on chemical signals (isotopes or salinity) provide independent sources of climatic information to test climatic inferences from pollen diagrams (a form of biotic data). Together, these geomorphic, chemical, and biotic signals offer a powerful means to help develop a geoecologic perspective of the past. Some of the most successful studies of lakelevel fluctuations and their climatic implications have involved closed lakes (lakes without any outlet) in regions that are currently semiarid or tropical. Many classic studies concern closed basins that held water in the past, like Lakes Bonneville and Lahanton in the Great Basin of western North America. Studies based on lakes with outlets are more complex, and these open basins typically show less dramatic responses to climate change. Hiatuses, mineralogic changes, microfossil or macrofossil distributions, organic material distributions, and changes in gross sedimentary stratigraphy all provide paleoecologic evidence from which lake-level changes can be reconstructed. Because sediment composition reflects both hydrologic and biolimnologic processes, the sedimentary evidence must be bolstered by other kinds of information. For example, wind-driven currents are usually the defining factor in establishing the boundary of sedimentation within a lake, because sedimentation is related to the high or low energy of the current. Therefore, paleowind directions and velocities, and thermal stratification within the water body, will affect the vertical movement of sediments. This is especially true for lakes in temperate to cold climates. Water-level changes in lakes like the Dead Sea in the Levant are good indicators of past climate. Sometimes determining what aspect of climate produced the observable record is difficult, since lake levels can change because of variations in either precipitation or evaporation. A single paleoclimatic factor, increased precipitation, say, could cause a lake level to rise. Cooling temperatures, which lead to lower evaporation rates, may also increase lake levels by decreasing net water

output. In these cases, biotic or chemical paleoecologic signals would have to be used to interpret the actual cause for the fluctuating lake levels that were indicated by the geomorphic and sedimentologic observations (transgressive and regressive stratigraphies or abandoned shorelines, for example). In our time, human activities control the levels of many such lakes. Over the past millennium, the Dead Sea has fluctuated through a range of 50 m. From about 100 b.c.e. to a.d. 40 it showed a dramatic rise and fall of 70 m. Geologic events can also result in changes in lake levels. In these circumstances lake-level changes may not be directly attributable to climatic or biotic conditions. Changes caused by direct geologic events are typically the result of processes that influence basin drainage patterns. Some of these can be associated with tectonic events and related earth movements. For example, faulting can raise outlets to lakes, or earthquakeinduced landslides can block drainages. The tectonically active area of the East African Rift system is an example of a place where faulting, along with climate, has played an important role in producing a record of lake-level fluctuations associated with archaeological remains. In addition, mass movements and earth flows caused by earthquakes are known to block stream drainages and form lakes.

Plant (Botanical) Indicators Microfossils Pollen Pollen and spores from plants can be preserved in both soils and sediments. The study of these fossils is called palynology.6 Except for certain macrofossils, pollen has probably contributed more to archaeological interpretation than any other plant remains. Many issues in the interpretation of fossil pollen assemblages provide insight into the complexities of interpreting other aspects of the prehistoric record, including artifacts and archaeological features. These issues include the application of taphonomic concepts, the connections between the observation of a pollen assemblage and higher levels of interpretation, and

169

Paleoenvironmental Reconstructions

Figure 6.1 Pollen Percentage Biotic change over time. Changes in plant and animal taxa can be used to infer paleoenvironmental changes or as indicators of particular intervals of time. In this instance, changes over the last 10,000 years in the relative amounts of amounts of certain types of pollen from a bog in central North America are depicted. The

relative amounts of pollen reflect changes in a vegetational landscape dominated first by sedge, then birch, spruce and oak, and finally pine. The four pollen zones can be used as relative time-markers as well as indicators of paleoenvironmental conditions. (Based on Huber and Hill 1987)

the possibility of ‘‘disharmonious’’ or nonanalog ecologic patterns. Securely dated stratigraphic pollen diagrams form a basis for paleoclimatic reconstructions and environmental reconstructions of archaeological sites. Pollen diagrams show the relative changes in kinds of pollen found in the sediment (fig. 6.1). However, links between pollen deposition and vegetation are not straightforward. The pollen record is a complex linkage of pollination ecology (especially wind transportation) and site formation processes. Pollen obtained from sedimentary sequences of lakes has been used to infer past vegetational landscape settings and paleoclimatic contexts. Other sedimentary contexts gen-

erally do not contain pollen unless unique physical and chemical circumstances exist that provide a situation for preservation. In addition, certain postdepositional conditions are more suitable for pollen preservation than others. Oxidizing conditions, calcareous settings, and deposits associated with well-drained alkaline deposits generally are not good for pollen preservation. Transportation and abrasion can also affect the final character of the pollen assemblage. Pollen can best be preserved in acidic, waterlogged areas. As well, the method of transport and the dispersal mechanisms influence the composition of a fossil pollen assemblage. Pollen analyses in environmental reconstruc-

170

Paleoenvironmental Reconstructions

tion rely on as set of basic principles. Plants produce pollen, and a relation can be established between the relative amounts of the pollen initially released and the vegetational landscape. Different plants produce different amounts of pollen, so the absolute abundance of each pollen type has to be considered in relation to plant production. Plants that pollinate by means of wind dispersal produce a lot of pollen. Some pine species, for example, produce billions of pollen grains seasonally. Closed flowering and insect-pollinated plants produce much less, perhaps on the order of 100,000 pollen grains in a season. Water can also move pollen. Thus, a specific pollen assemblage may not represent the vegetative setting of a single place but rather a combination of vegetation types within a stream drainage or region of wind transport. Another source of patterning is the size and shape of pollen grains. Particular pollen types may be sorted and settle in different parts of a basin. In this sense, like artifacts, the final pollen assemblage is related to factors associated with clastic deposition-pollen grains are small bioclasts (artifacts are also bioclasts). Some portion of the pollen production will eventually be deposited and preserved. After initial deposition (following dispersal and accumulation), oxidation and biologic activity can destroy pollen. Preservation is enhanced if the pollen grains settle in anoxic conditions or in settings where sediment accumulation is rapid, resulting in quick burial. In addition to lake and bog deposits, pollen may be preserved in soils, tufas, glacial ice, and cave sediments. Pollen grains can be diagnostic to the species level based on a combination of characteristics, including size, surface textures, and types of aperture. It is possible to observe changes through time in the frequencies of pollen types if samples can be obtained from a stratified sequence. These temporal changes can be related to past climatic conditions or to the effect of human alteration of the landscape. Where paleoenvironmental chronologies have been developed for pollen stratigraphies, the changing pattern of pollen types through time has been used as a general chronological indicator (see fig. 6.1). In order to use

pollen assemblages as an indicator of time, it is necessary to document enough pollen sequences from a region so that the pollen sequence from the new core can be compared with a securely dated pollen chronology. Sometimes it is possible to examine temporally equivalent pollen assemblages from different locations in order to document spatial variation in vegetation and climate. Estimates of past vegetational landscapes can be made by comparing past pollen assemblages with similar present-day settings. Estimating past climatic conditions by means of fossil pollen assemblages requires an understanding of the relationship between plant communities and climatic parameters. In order to make these comparisons, one must understand the vegetational and climatic linkages with the presence of particular soil conditions, the amount of precipitation, temperature patterns, and the like. As with any application of present-day observations to the prehistoric record, care must be taken: in some instances no direct analogs can be applied to past ecologic circumstances. We can trace the impact of human activities on the vegetational landscape during the Holocene by applying pollen sequences in Europe and the Near East. Sometimes particular plant types are good indicators of a particular climatic variable, like precipitation or temperature. The onset of the Holocene, around 10,000 years ago, coincided with the beginning of the Neolithic in the Near East. Early Holocene environmental changes caused conditions favorable to the development of agriculture. Ten thousand years ago the Near East had a very different climatic regime from that of the present, which we know in part through interpretations of fossil pollen records. Pollen has been used to infer changes in both precipitation and temperature. From about 20,000 b.p. to about 15,000 b.p., the climate was cool and arid, with much of the Near East landscape in steppes.7 Between 15,000 b.p. and about 5,000 b.p., temperatures rose, reaching their maximum at the end of that period. At the beginning of the Holocene the climate of the Near East was cooler and more humid than today. In Mesopotamia, pollen data indicate that the temperature increased, but rainfall remained relatively low. In

171

Paleoenvironmental Reconstructions

the eastern Mediterranean and Near East the oak pollen frequency curve, considered a proxy for the amount of precipitation, shows that increased rainfall caused an expansion of the forests in Anatolia, the Levant, and the Zagros Mountains. This climatic change may have been the trigger that led to the domestication of plants in this region. Changes associated with Pleistocene to Holocene transition altered the types of plants and animals available for human use. Incipient domestication may have been an adaptive response by prehistoric humans to increase the reliability of food sources as procurement strategies connected with the Pleistocene were no longer useful in the Holocene. The palynology associated with archaeological sites can be used to evaluate the season of occupation, the subsistence, and related aspects of past human behavior. At Shanidar Cave in Iraq, palynologic studies revealed that the remains of a Neanderthal buried in the cave were associated with flowers.8 In addition to airborne pollen, pollen from hyacinth, hollyhock, and groundsel were clustered with a pine-like shrub. The flowers also provide an indicator of the time of year of the burial: probably between late May and early July. Later interpretations of this depositional context have posed the possibility that roof fall in the cave may have crushed the Neanderthal, and that mechanisms other than human ritual activity can explain the presence of the pollen.9 There are many other examples of pollen being used to infer environmental conditions associated with archaeological sites.10 At the Acheulian site of Terra Amata in France, pollen was found in coprolites associated with archaeological remains of what appear to have been seasonal hunting camps. The pollen indicated that this 400,000-year-old (Middle Pleistocene) site may have been occupied during the spring. In North America, pollen has been used to reconstruct changing vegetative and climatic patterns throughout the Holocene. The Late Pleistocene and the onset of the Holocene in North America are associated with Paleo-Indian populations. Pollen data indicate that in central North America a spruce-dominated forest changed to prairie in the west and to pine in the east between 12,000 and 9,000 b.p. This change in vegetative 172

landscape may be reflected in the change in faunal communities and in human adaptations. A series of pollen stratigraphies documents the timetransgressive nature of first the eastward advance and then the westward retreat of prairie during the Middle Holocene. An archaeological site that shows the eastward movement of the prairie and forest boundary is the Itasca bison kill site in northwestern Minnesota, where the presence of an open pine forest around 8,000 b.p. and its replacement by prairie about 7,500 b.p. has been documented.11 Along with other paleoecologic indicators, pollen studies have figured prominently in a synthesis of North American human prehistory with environmental change.12 Changes in the levels of biodiversity are reflected in the archaeological succession from Paleo-Indian to Archaic (see fig. 5.5). Disturbance by human alteration of the landscape can also be observed in pollen data. The presence of high amounts of maize (Zea mays) pollen found in a pollen stratigraphic sequence in Ontario, Canada, provided evidence for agricultural subsistence patterns associated with Iroquois presence in the region.13 Maize and purslane pollen are found in the Crawford Lake pollen diagram starting about five hundred years ago and continuing until about three hundred years ago. There are also examples that demonstrate the direct effect pollen studies can have on the interpretation of artifacts.14 In one study, the age range of artifacts thought to have been used for massive woodworking at archaeological sites in the coastal areas of Washington and British Columbia was studied. The pollen sequences indicated that between 9,000 b.p. and 2,500 b.p., cedar migrated along the Pacific coast northward from central Washington to British Columbia. The pollen indicated a correlation between the presence of artifacts thought to be used for massive woodworking and cedar trees. These studies led to the conclusion that the development of activities associated with massive woodworking by prehistoric human groups in the region was constrained by the availability of cedar. Phytoliths Plants produce microscopic bodies composed of silica or calcium oxalate, which can be used

Paleoenvironmental Reconstructions

Figure 6.2 Grass Phytolith Types (A) Sinuates (long trapezoid) are characteristic of grasses of the Pooid tribe; (B) Rondels (short trapezoid) are widespread in all tribes, particularly in florescences; (C) Saddles are characteristic of the Chloridoid

tribe but also occur in bamboo and some arundinoids; (D) Dumbbells are most often found in Panicoids; bamboo, some arundinoids, and some Pooids also produce dumbbells. Crosses are a particular type of dumbbell.

as paleoenvironmental indicators in accordance with their morphologic assemblages. The microscopic opal or calcium oxalate deposits that form in and between plant cells are called phytoliths (literally, ‘‘plant rocks’’). Phytoliths can encode significant archaeological and paleoenvironmental information. Unfortunately, phytolith studies have as yet received only limited attention from scientists, so considerable research remains to be done in phytolith systematics. Plant families that produce abundant opal silica bodies are grass (which includes the cereals), sedge, elm, bean, squash, and sunflower. Cerealgrain phytoliths can be classified into many types, according to cellular origin and shape. Phytoliths tend to be deposited through decay-in-place mechanisms and are more stable than pollen

in many depositional environments. This makes phytoliths an excellent complement to pollen.15 In addition to being useful in the study of early agriculture and as broad paleovegetational indicators, opal phytoliths can differentiate between C 3 and C 4 photosynthetic pathways in grasses. A method based on the refractive index changes of phytoliths can determine burned plant matter. Burning causes a shift in phytolith refractive indices to higher values.16 Phytoliths are classified according to shape (fig. 6.2). Pooid phytoliths have circular, rectangular, elliptical, crescent, or oblong shapes and occur primarily in C 3 grasses growing in high latitudes or high elevations. A high percentage of Pooid phytoliths tends to indicate cool temperatures. Chloridoid phytoliths are saddle-shaped 173

Paleoenvironmental Reconstructions

and occur primarily in C 4 grasses. Panicoid phytoliths have dumbbells and crosses and also occur primarily in C 4 grasses. Panicoid phytoliths, however, tend to occur in areas with higher moisture than do Chloridoid phytoliths. Therefore, grass phytoliths from soils, paleosols, and archaeological sediments can be used to reconstruct the environment of a site or region. Increased evapotranspiration in hot arid environments promotes increased phytolith formation. Irrigation farming or farming in areas with poor drainage, where additional soluble silica is available from excess soil water, may enhance this process. Phytoliths have been found in a variety of settings. The phytolith record from Natural Trap Cave of northern Wyoming provides an example from Ice Age deposits.17 The depositional sequence at the cave contains an environmental and climatic record that includes the last interglacial and the Pleistocene-Holocene transition. Phytoliths produced from grasses were found in deposits thought to range from later than 110,000 b.p. to around 12,000 b.p. Pooid phytoliths dominate the oldest deposits, which contain no Chloridoid phytoliths, a distribution that may mean that the region was cooler and wetter than now. Greater numbers of Panicoid and Chloridoid phytoliths occur in the upper part of the sequence, but the grass phytolith composition of circa 12,000 b.p. does not have a composition expected from the Holocene. Fauna found in the same sequence suggest the possible presence of a montane conifer parkland, with C 3 grasses changing to C 4 in the Natural Trap area about 11,000 years ago. Phytolith analysis can help solve a specific paleoethnobotanical problem, such as the identification of maize at an archaeological site. Kernels and cobs are common macrofossils at many sites; however, domestication and diffusion of maize are thought to predate the macrofossil evidence in many areas. Maize produces both larger quantities and larger sizes of cross-shaped phytoliths than other grasses (Cross is defined as a body with at least three sides clearly indented in the planar view, and a length of no more than 2 microscope units, or 9.16 micrometers, greater than the width). In addition to noting the percentage of phytoliths that are cross-shaped and 174

the size of the phytoliths, a dumbbell-cross ratio confirmed that most wild grasses produced fewer and smaller crosses than maize. Phytolith threedimensional structure is based on the face opposite the cross-shaped face. Other approaches to identifying corn by phytoliths that were used by Irwin Rovner and John Russ have focused on computer-based recognition systems to sort the three-dimensional assemblages.18 Phytolith analysis was used to help reconstruct the vegetative history and landscape of the nineteenth-century Harpers Ferry site in West Virginia.19 The samples collected from prehistoric levels at the site contained nongrass phytoliths that probably derived from deciduous trees, as well as phytoliths representative of grasses. In the lowest historic levels, dating from about a.d. 1800 to a.d. 1820, there were fewer phytoliths but higher proportions of grass phytoliths (short-cell Panicoid, Festucoid, and plain-rod types). The number of scrub and tree phytoliths decreased, while that of grass phytoliths increased significantly. In samples dating from about a.d. 1820 to a.d. 1832 there was an increase in Festucoid types associated with grass. Phytoliths from flood layers also were associated with grass growth. The presence of Panicoid and Festucoid classes reflected these moist conditions. Phytoliths can provide data about the types of foods eaten by past human populations. At the medieval site of La Olmeda, Spain, which dates from the seventh to the thirteenth centuries, scientists looked at the surfaces of human teeth. It was possible to identify the presence of a phytolith attributable to the common millet. This was a short-cell or dumbbell-cell phytolith. The phytolith was found through scanning electron microscopy; to verify that the structure was a phytolith, its composition was tested using an X-ray microanalysis system. All the phytoliths that could be classified belonged to the Poaceae (Graminae).20 Diatoms Algae, found in aquatic settings, form cell walls composed of silica.21 These biogenic silica organisms are called diatoms and can be preserved in sedimentary sequences. Distinct morphologies provide a means for identifying different species of diatoms. Different algae types have

Paleoenvironmental Reconstructions

distinct habitat tolerances. Their known ecologic niches provide useful information about the hydrologic conditions of bodies of water. Many taphonomic principles that influence a pollen assemblage also affect diatoms. Streams and wind can readily transport diatoms in drainage basins. Freshwater diatoms exposed on dry lakebeds have been eroded and blown by the wind into marine sediments, providing a record of continental environmental conditions in nearby ocean sediments. Diatoms have been recovered from both lake and marine sediments and are especially useful in documenting environmental changes that reflect fluctuations in water levels and transgression-regression sequences. Diatoms can help document sea-level changes because some species are very sensitive to changes in salinity and consequently are an indicator of transgressive and regressive sequences. Particular species are associated with freshwater, brackish, or marine conditions. Because of this, studies of diatoms can help in evaluations of paleoenvironmental settings connected to coastal archaeological sites. Diatoms were used as paleoecologic indicators in archaeological interpretations at the Lubbock Lake, a Paleo-Indian site in west Texas.22 A study was undertaken to determine the sequence of water conditions in the area. Some sediment samples could be correlated with the Folsom artifact-bearing stratum. The results of the diatom study indicated the existence of bog conditions associated with stagnant to slow-flowing water that had considerable variation in salinity levels. One of the more common species of diatom indicated that flowing springs had once existed. The diatoms showed conditions that fluctuated from slightly brackish to much fresher water; these conditions were consistent with climatic changes in precipitation regimes and fluctuating spring flow and water levels. Another example of the geoarchaeological use of diatoms comes from the study of Montezuma Well in northern Arizona. Montezuma Well consists of a collapsed spring-mound composed of travertine; it has a diatom record of more than 11,000 years.23 A radiocarbon-dated core from Montezuma Well was studied to evaluate past environmental conditions at the spring

and their effect on prehistoric human occupation. Variations in diatom taxa and other information from the core seem to show that before 9,000 b.p. and after 5,000 b.p. there was a moister climate in the region. Between these two dates conditions seem to have been drier. The presence of Chaconnes placentula var. Lineata throughout the sequence indicated the continuous presence of submerged aquatic plants. The variation in amounts of Anomoeoneis sphaerophora suggested periods of high aridity during the intervals of 8,700 to 8,400 b.p., 7,800 to 6,900 b.p., and 2,000 to 1,000 b.p. The major changes in the diatom assemblages after about 5,000 b.p. seem to indicate a change in physicochemical conditions. After about 4,000 b.p., the water levels were apparently higher. There is a possible correlation between the diatom species in the sequence and prehistoric human occupation of the area. The higher values of Aulacoseira granulata and A. islandica coincide to some degree with the occupation of the Sinagua people. The Sinagua occupied pueblo, cliff dwellings, and pit-houses near the well between about a.d. 750 and 1400 (about 1250–600 b.p.). These diatoms were more abundant between 1,500 and 1,000 b.p. and are indicators of organic enrichment in the Montezuma Well. Diatoms have also been used to trace landscape change associated with human settlement.24 In addition to their function as environmental indicators, diatoms have helped trace the source of clays used to make pottery and, consequently, are useful in the study of exchange systems (see Chapter 8). Diatoms can accumulate in lakes and swamps, which results in the formation of biogenic siliceous deposits called diatomites. An indicator of lacustrine and paludal environments, diatomites have been used as a raw material by humans.

Macrofossils Macroscopic plant (‘‘macrobotanical’’) remains found in sediments include seeds, nuts, fragments of charcoal, and larger objects like tree trunks. Some depositional environments enable these materials to become part of the sedimentsoil record. Because of the larger size of plant macrofossils, taphonomic trajectories associated 175

Paleoenvironmental Reconstructions

with them are not usually as complicated as those associated with pollen dispersal. However, as with artifacts, depositional and postdepositional circumstances play a part. As an example, seeds of domesticated plants were recovered from charcoal lenses associated with Late Paleolithic artifacts at Wadi Kubbaniya, a tributary to the Nile River in southern Egypt.25 The charcoal dated to about 17,000 b.p., but the domesticated seeds were dated to the Holocene; they had intruded into the earlier deposits. Careful collection of other plant remains from the series of sites at Wadi Kubbaniya led to the recovery of wood (tamarisk) charcoal and a variety of tubers that were dated by radiocarbon as being contemporary with the Late Paleolithic occupations. Larger plant remains like tree trunks are used as indicators of both general prevailing environmental conditions (based on the type of tree and its habitat tolerances) and finer-scale fluctuations (based on the variation in ring-width thicknesses).26 These kinds of materials can be incorporated either as structural features (for example, the wood beams found in Pueblo sites in the American Southwest) or as sedimentary deposits (such as the Two Creeks and Cromerian forests discussed in Chapter 2). Tree remains also have been used as indicators of climate and as chronological tools.27 In the most straightforward cases trees from archaeological sites can provide evidence of climate, because the width of tree rings can be directly related either to rainfall or temperature, or to a combination of both. Variations in isotopic ratios of tree rings also provide signals of climatic parameters. Conifers have been used in chronologies going back to the Early Holocene in the American Southwest (more than 8,000 years). Studies of oak-tree rings in western Europe have pushed back the tree-ring chronologies to around 7,000–6,000 b.p.

Animal Indicators Invertebrates Ostracods Ostracods are very small (about 0.2–0.7 mm in size) bivalve crustaceans. They are bottom-

176

dwelling organisms, most of which crawl, although some can swim. The presence of ostracods in sediments indicates the existence of aquatic habitats in the past, which can be either saline or freshwater. Chemical signatures contained in ostracods have helped in the evaluation of paleoecologic and paleoclimatic settings. Ostracods have also been used to infer hydrologic changes caused by human activity. Their shells can be preserved in most aqueous depositional environments, including oceans, lakes, ponds, and streams. Because not enough time has passed for phylogenetic changes to take place during the Quaternary, freshwater ostracod species cannot be used as chronostratigraphic markers (the same is true for insects). The exception to this would be the occurrence of unique frequency variations in ostracods that can be dated by independent means. Stratigraphic zones based on different ostracod assemblages have been used to document changing environmental conditions in lacustrine settings. As with most paleoecologic methods, the use of freshwater ostracods to infer past environmental settings relies on a comparison of the morphology of fossil ostracods with physical, chemical, and climatic information about present-day ostracods. Often the present-day habitats associated with ostracods are restrictive, so it is possible to make specific inferences about past conditions from species found in sediments. For example, some species are found predominantly in lakes, others in ponds, and others in moving water. Other parameters that can be derived from present-day habitat associations include salinity and the amount of dissolved oxygen in the water body. The presence of calcareous ostracod shells provides an indication of past geohydrologic conditions. To preserve calcareous shells, hydrologic conditions need to be buffered based on pH levels. Below a pH level of about 8.3 the carbonate in the shell will dissolve; ostracods will therefore not be preserved in many bog and marsh deposits, although they may have been present initially. Ostracods are more likely to be preserved in marl deposits. Ostracods generally do not live in highenergy environments associated with the deposi-

Paleoenvironmental Reconstructions

Figure 6.3 Hohokam Canals and Irrigation Reconstruction from Ostracod Chemistry Geoarchaeological evidence from the chemistry of ostracods is used to infer events associated with the canals and irrigation of the Hohokam. Two chemical ratios in ostracods compare the amount of Mg to Ca or the amount of Sr to Ca. Changes in the propor-

tion of the elements were recorded in three types of ostracods. The salinity index shows several fluctuations from low to high salinity from after 600 to after 1800. When imposed on the archaeological chronology, these changes can be used to interpret prehistoric irrigation from the Early Pioneer through the Classic Period. (Based on Palacios-Fest 1994)

tion of coarser clastics. They may be found in fine sands and coarse silt, but they are most common in clays, fine silts, or organically rich areas. The chemical composition of ostracod shells can throw light on certain environmental parameters. The ratio of magnesium and strontium to calcium is related to salinity. Chemical parameters control the strontium-calcium ratio while water chemistry and temperature combine to control the magnesium-calcium ratio. Changes in the strontium-calcium ratio within the ostracod shell are related only to fluctuations of carbonate or sulfate in the water body. Where strontiumcalcium and magnesium-calcium ratios are both stable, it is an indication that water composition and temperature did not change. This may reflect settings of deeper water. M. Palacios-Fest used the levels of magnesiumcalcium and strontium-calcium to study changes in a Hohokam age canal, in Arizona; variations in the water chemistry were related to the effects

of human- and climate-induced environmental change (fig. 6.3).28 The ostracod samples were collected from canals at the site of Las Acequias in Arizona. Between about a.d. 1025 and 1425, increases in salinity, as recorded by ostracods, may have been a product of human-induced environmental alteration. Ostracods also recorded major climatic pulses, including two floods (one at about a.d. 855–910, another at a.d. 1350) and a drought (from about a.d. 1365 to 1425). Based on the ostracod species present at Las Acequias and the environmental conditions associated with them, four assemblages were determined, which indicated waters ranging from chemically dilute to calcium-enriched, dominated by sodium, magnesium, and sulfate ions (either low or moderate salinities). Assemblage 1 was associated with slow-flowing waters, 2 with seepage conditions, 3 with slow-flowing and stagnant waters, and 4 with seepage and flowing waters in a human-disturbed environment. From the Hoho-

177

Paleoenvironmental Reconstructions

kam Early Pioneer Period (before a.d. 700) to the Colonial Period (a.d. 700–910), canal waters became more dilute, while during the Early Sedentary Period (a.d. 910–1025) they became more saline. After a.d. 1350, the trace-element chemistry of the ostracods indicates a dramatic increase in water salinity during the Classic Period (a.d. 1275–1425), followed by more dilute waters in historic times. The Las Acequias Hohokam irrigation system provides an example of the usefulness of ostracods to an understanding of prehistoric human-land relations. Higher stream flows and more dilute waters and flooding that are probably associated with climate are indicated for the Colonial Period. An increase in salinity during the end of Hohokam occupation (Classic Period) indicated by the ostracods would support the contention that higher temperatures and drier conditions associated with the Anasazi Warm Period were responsible for the abandonment of the area by the Hohokam. Increasing salinization, suggested by the trace-element chemistry in ostracods, may also reflect transformations of the landscape because of Hohokam irrigation and agricultural practices. Cladocera, another form of small bivalve Crustacea, can also help us understand environmental conditions associated with human occupations. Cladocera samples collected from Terminal Pleistocene deposits in Lake Zeribar in Iran were used to indicate environmental conditions during an interval when human subsistence patterns changed from hunting to agriculture.29 Mollusks Mollusks are one of the more common types of invertebrate remains associated with Quaternary deposits and archaeological sites.30 The study of mollusks is called malacology. Gastropoda (snails) and bivalves (including clams) are two prominent types of molluskan invertebrates. Gastropods usually consist of a single conical or spiral shell, while bivalves have a hinged shell. Most mollusks are composed of aragonite although some consist of calcite. Different mollusk species live on land and in water. They can be found in a variety of depositional settings, including loess;

178

caves and rock shelters; stream, lake, and spring sediments; and marine coasts (fig. 6.4). Terrestrial mollusks have been recovered from archaeological features like ditches, wells, and occupation debris. Mollusks also form a primary constituent of coastal middens. Several factors need to be considered in determinations of environmental conditions from assemblages of fossil mollusks. These include taphonomic processes of deposition and preservation (including the potential for mixing), the relative abundance of various types of mollusks, sampling, and identification. Malacology has been used to study the climatic transitions associated with prehistoric human occupations. In northern France the mollusks from several sites covering the late glacial and Early Holocene have been related to the Upper Paleolithic (Late Magdalenian) and Mesolithic archaeological sequence. There were five distinct zones. The bottom of the sequence contained redeposited silt, probably consisting of redistributed loess. It included species that exist in cold conditions and in open, short grassland environments. An overlying organic silt contained a higher number and a greater diversity of mollusks, some of which indicated the presence of relatively warm and humid interstadial conditions and more complete vegetative cover. Humid conditions are indicated by changes in mollusk types in the overlying sediments. The fourth zone, which represents the Early Holocene, reflects a major change. Forest species increased at the expense of openground conditions. In the last zone there is a change back to open-land mollusks and a dramatic decrease in forest species. At one site (Holywell Coombe) that had the same characteristics as the fifth zone, the change may reflect human activities associated with forest clearance during the Neolithic and Early Bronze Age. Mollusks have also been used to infer landscape change. Variations in land molluskan fauna helped in reconstructions of prehistoric land-use change at the Pink Hill archaeological site west of London.31 The lower deposits at the site contained shade-sensitive mollusks associated first with woodland habitats and then with initial clearance of the woodlands during the pre–Iron

Paleoenvironmental Reconstructions

Figure 6.4 Microfaunal Diagrams as Indicators of Environmental Change Different kinds of terrestrial gastropods are indicative of specific types of environmental settings. In this example, we see the changes in the landscape through

time. Greater numbers of shade-loving species indicate the presence of a woodland environment, while greater numbers of open-country species indicate a grassland habitat. (Based on Evans 1972)

Age. An open-country phase A was correlated with Iron Age agricultural activity, and a phase B with Romano-British agriculture. Studies of mollusks along the southern New England coast of North America have demonstrated that Late Holocene use of shellfish resources was affected by environmental and climate change.32 Although human preference has been suggested as an explanation for prehistoric shellfish use patterns, climatic factors may have played a critical role. During the Middle Woodland period, dating from about 1,800 to 1,600 b.p., the regional climate cooling produced cooler water temperatures, which resulted in changes in molluskan fauna from warmer to cooler assemblages. This led to a corresponding alteration of human resource use.

Insects Paleoenvironmental and climatic data can also be derived from insect fossils.33 Insects are a remarkably pervasive group of creatures. They make up more than half the faunal and floral species known today. Insect remains can be recovered in such depositional contexts as ponds, bogs, lakes (often near the margins), fluvial deposits, and pack-rat middens. There are two underlying principles in the use of insect remains: first, throughout the Quaternary many insect species seem to have remained unchanged; and second, the habitat associations of insects do not seem to have changed either. Therefore, insects are good indicators of environmental conditions. Because they are so abundant, many insects (flies, bees, beetles, and ants) have been recovered from Quaternary deposits.

179

Paleoenvironmental Reconstructions

Beetles (Coleoptera) have been used extensively in the reconstruction of environmental conditions. Like other Quaternary insects, they combine genetic stability with sensitivity to climate and so live in restricted environmental settings. There are about 350,000 species of beetles. This has been a boon to paleoenvironmental studies, because they are often precisely adapted to narrow environmental niches, but collectively they inhabit nearly all terrestrial and freshwater habitats. In southern Britain, fossil insects from the type site of the last interglacial in Suffolk provided valuable information regarding environmental conditions during the Paleolithic occupation of this region.34 The twenty-one species of beetles found there indicate that during the Paleolithic occupation the climate was warmer than it is now. Some insects from Britain are extinct while others are now distributed in southern Europe, indicating that the climatic zones shifted southward. The insect fauna of Middle-Late Holocene England has been studied in relation to archaeological features associated with forest clearance.35 Underlain by lake deposits of the last glaciation, Thorne Moor is a relict of an extensive bog containing fossil timber that is riddled with insect galleries. A radiocarbon age of about 3,090 b.p. dates this to Bronze Age clearance of the mixed oak forest of the region. Insect fauna were recovered from organic silts and preserved trees. The study found a number of insects known to be associated with oak. That some insects found in the moor do not exist in the area today indicates change in climate or change in vegetative cover caused by forest clearance. The insect fossil record also throws light on the timing and intensity of environmental change associated with the Terminal Pleistocene and Lower Holocene human occupation of western North America.36 Fossil insects collected at the Lamb Spring Paleo-Indian archaeological site in Colorado seem to indicate that there were cooler-thanpresent mean summer temperatures at around 14,000 b.p. Other locations in Colorado and Montana indicate that there was a gradual warming between 13,000 and 10,000 b.p. Rapid warming seems to have occurred after 11,000 b.p. By

180

9,700 b.p. insect fauna from Utah indicate that temperatures were warmer than today. Similar warming intervals are also documented by insect fauna for the late glacial to Holocene transition in Europe and eastern North America. The insect fossil assemblages indicate a rapid change in climate that probably affected all biotic communities, including prehistoric human populations.

Vertebrates Deposits within landforms and archaeological sites also contain a variety of animal remains that may serve as indicators of past environmental conditions. Although the potential value of using vertebrate remains in making general paleoenvironmental inferences is high, attention must first focus on the taphonomic conditions of accumulation. Environmental evaluations of faunal remains are primarily based on comparisons with the geographic ranges and present-day habitats for living species. Morphological attributes and contextual information help to provide indications of past environments of extinct animals. Mammal Fossil Remains When using fossil vertebrate remains to infer past environments and climates, there are several points to consider.37 Generally, many larger mammals have broad habitat tolerances. This is partly because they are warm-blooded (endothermic) but also because they can be highly adaptive. Smaller animals are usually more valuable for environmental reconstruction, because they are more ecologically restricted. In attempting to determine the past habitat tolerances of extinct species, it is helpful to make ecologic inferences that rely on a variety of mammal types, not just a single species. Some habitat associations can be partially determined from anatomical characteristics, but these may not be restrictive enough. For example, the teeth of mammoth and mastodon show marked differences thought to be related to their eating habits and thus the environments in which they usually lived. Mastodons have short-crowned teeth and were probably browsers, while mammoths have high-crowned teeth and were likely to have been grazers. Mastodons are thought to

Paleoenvironmental Reconstructions

be more closely associated with woodland or forest settings while mammoths are associated with steppe and grassland conditions. But the two different taxa perhaps were not limited to these settings. Generalities like this are not always clear indicators of environmental habitats. Presentday elephants, for instance, have teeth similar to mammoths and are grazing animals, but they can live in either savannas or forests and can also browse on trees. Biotic taxa other than humans seem to have adapted and changed their habitat tolerances through time. Some attributes that enable mammals to be used as chronological indicators make them less useful for paleoenvironmental studies (see Chapter 5). The major characteristic of Late Tertiary (Pliocene) and Quaternary mammals is the generally high morphologic diversity observable in the prehistoric record. This diversity is usually considered the result of adaptation to climatic or environmental fluctuations. Throughout this time major morphologic changes occur in many mammal groups, including primates, elephant types, bison types, deer types, rhinoceroses, beavers, and, in particular, smaller rodents. The morphologic changes mean that mammals can be used as chronological indicators, but these same changes complicate mammals’ use in paleoecologic inference. Morphologic changes are usually thought to be the result of adaptations to particular physical and biotic habitats. Where the mammalian forms are similar to those of existing animals, it is reasonable to consider observations of present-day habitat preference and behavior when studying the fossil record. For mammoths, living elephant communities are used as a first approximation for understanding the fossil record. The same kind of logic is employed for carnivore behavior. Thus, the behavior of existing felidae, along with independent paleoecologic information derived from the context of the fossils, provides the basis for reconstructing the habitats and habits of extinct felids, like the saber tooth–scimitar cat forms. Despite the constraints caused by changes in mammalian morphology through time, basic, generalized environmental habitats can be inferred using modern comparisons. Fossil taxa that

are physically similar to present-day elk, reindeer, lemmings, musk ox, and polar bear would indicate the probable presence of an arctic or boreal landscape habitat. In the same way, antelope and horse types would be an indication of steppe-taiga landscapes. However, the possibility of major behavioral changes not directly observable by the morphology of the fossil forms should always be a consideration. Thus, the sometimes subtle differences in the morphologic features of fossils grouped into the hominid genera Australopithecus and early Homo reflect diverse adaptations that are probably related to a variety of habitats. It is thus necessary to use a wide array of geoarchaeological tools when developing a geoecologic model of the past. For hominids, the evaluation of the entire environmental context, using strictly geologic and biotic criteria (including artifactual patterns) forms the basis of a paleoecologic reconstruction. A more reasonable estimate of prehistoric environments will be produced when we can rely on the likely habitats of many fossil mammal forms, if we can assume that the fossil assemblage is from a contemporaneous community and not the result of taphonomic mixing. Generally, smaller mammals are more reliable indicators of local geoecologic conditions than larger mammals. We base this conclusion on the observation of present-day large and small mammals. Larger mammals inhabit more diverse and widespread landscapes. Many have seasonal migration routes or at least extensive areas of potential habitats. Smaller mammals, especially rodents, are typically better indicators of local geoecologic conditions. They are also useful as chronological indicators. Bird Fossil Remains Because the ecologic constraints of birds are often quite narrow, bird remains can prove useful in ecologic reconstructions. Yet despite specific range tolerances, there are problems with using bird remains. These include birds’ potential to adapt to changing habitats and their high mobility, which results in their being found throughout a wide geographic range. This is especially true of birds that migrate seasonally. In archaeological contexts the presence of migratory fowl

181

Paleoenvironmental Reconstructions

can be an indicator of seasonal occupation. Some bird taxa may be linked to specific landscape contexts. We can thus infer the past presence of a particular landscape on the basis of the bird remains. Sometimes the distributions of bird fossils provide clues to past changes in geoecologic patterns. It has been proposed that changing ecologic patterns in North America led to the reduction in the geographic distribution of the California condor.38 During historic times, the California condor was primarily associated with warm-temperate climatic and ecologic contexts. Based on several lines of paleoecologic evidence (including pollen and plant macrofossils), it appears that around 11,000 b.p. the bird’s range was much less restricted. Evidence from the Hiscock site in western New York shows that the California condor was once able to live in an ecologic setting associated with spruce-jack pine woodlands. This boreal, coniferous vegetative landscape may have been associated with colder climatic conditions. Probably condors were able to exist in this environment because of the availability of an important biotic food source, large mammal carrion. The severely restricted distribution of the California condor during the Late Holocene may be linked to the extinction of large mammals during the Pleistocene-Holocene transition that would have reduced the available food for scavenging birds. Reptile and Amphibian Fossil Remains The study of reptiles of the past is called paleoherpetology. In comparison with mammals, little morphologic change is observable in the Late Tertiary and Quaternary fossil record of reptiles. Reptilian fauna have been stable; they have undergone little adaptive radiation or speciation. Reptiles can be useful for interpreting paleoenvironments, although reconstructions need to take into consideration reptiles’ adaptive abilities. Fossil remains of morphologically similar reptiles have been found in contexts different from those of their apparent present-day counterparts. Scientists thus conclude that reptiles may be highly adaptable to ecologic-habitat alterations induced by climate change. Climatic factors account for most of the changes in the distribution of reptilian forms. Reptile distribution seems 182

to reflect changing patterns of temperature and moisture. As has been demonstrated for other aspects of the biostratigraphic record, taphonomic processes are important to the interpretation of amphibian and reptile fossil accumulations. In some instances, the presence of other paleoecologic data has led to the conclusion that the reptilian specimens were not originally part of the same community. The presence of fossil assemblages containing animals that now live in different ecologic settings led scientists to believe that these were mixed assemblages. The work on amphibian and reptile fossils has led to the conclusion that different ecologic patterns existed in the past. Some accumulations, instead of being heterochronic (consisting of bones from different periods), may be indicators of different, nonanalog, geoecologic patterns in the past. Similar interpretations have been made on the basis of other ecologic indicators, like pollen stratigraphies and mammal assemblages. Because reptiles are ectotherms, they have fairly restricted ecologic and geographic ranges, especially compared to large mammals. These ecologic constraints make reptiles potentially useful in geoecologic reconstructions. Trends in the body size of reptiles have been used as an indicator of either climate or ecologic associations. Larger body size has been attributed to higher temperatures (a climate factor), or less stress on populations from predators, while smaller size (dwarfism) has been associated with increased predation. These types of ecologic connections demonstrate that morphologic change in fossil forms cannot always be directly attributable to climatic factors. For instance, the decrease in the size of lizards living on islands has been interpreted to be the result of Holocene human settlement.39 Fish Fossil Remains With due consideration for taphonomic indicators of transport and redeposition, fish remains can be used to infer a variety of aquatic conditions. Freshwater fish are associated with two contexts: well-oxygenated bodies of water, such as streams and larger lakes; and poorly oxygenated water, like ponds and swamps, or low-energy flu-

Paleoenvironmental Reconstructions

vial settings. Because the scales, vertebrae, and ear stones (otoliths) of fish have annual growth rings, it has been possible to use them to study seasonality. For example, R. Casteel used fish scales to infer a relationship between rates of fish growth and water temperature, with higher rates linked to higher temperatures.40 Fish remains can help provide information about prehistoric drainage connections. In southern Egypt, sedimentary deposits containing Middle Paleolithic artifacts also contained the remains of fish. The presence of fish indicated that streams once connected ancient lakes to one another. Wim Van Neer’s studies of the fish remains from these deposits were also used to reconstruct the paleoaquatic environment of the Middle Paleolithic sites.41 The variety of fish found in the deposits was one indication of good hydrologic conditions, but the fish taxa found also suggested that the lakes were large and deep, with sandy bottoms.

Other Ecologic Accumulations Pack-Rat Middens Pack rats (genus Neotoma) are small mammals that received their common name because they pick up a variety of materials and bring them to their dens, where they accumulate as middens. In dry areas, like the Great Basin of western North America, these midden accumulations have been preserved for more than 40,000 years. In 1964 Philip Wells and Clive D. Jorgensen documented the potential uses of pack-rat middens for paleoecologic information.42 Two characteristics make them useful for studying past environments and climates. Pack-rat middens contain a variety of plant and animal remains, including twigs, leaves, seeds, pollen, and bones. These can be dated, using radiocarbon measurements (see Chapter 5). A major difference between these accumulations and continuous stratigraphies that contain pollen or macrofossils is that pack-rat middens generally represent an isolated interval in time. In addition, pollen stratigraphies can represent regional vegetative patterns, while middens consist mostly of materials within several hundred feet of their location. As a result, one must analyze and date

many separate middens in order to interpret regional or chronological variation.

Peat Peat bogs are a major source of information about Quaternary landscapes for northern latitudes. Although the rate of accumulation varies widely, peat accumulates at about 1 m for every 1,000– 2,000 years in a favorable environment. For most peat deposits the sedimentary record is composed of decayed plant remains in a waterlogged environment. Peat deposits extend back as far as 9,000 b.p. and contain a continuous record of plant communities, as well as a record of atmospheric deposition, including pollen. In North America, Florida’s extensive peat deposits have yielded abundant totems, masks, and figurines. A Florida peat bog preserved a skull from 8,000 b.p. that had a relatively undamaged brain. The soft tissue remains recovered from bogs provide an important source for DNA studies. These bogs are some of the most important environments for wetland archaeology. In Europe peat bogs have preserved many tracks, dugout boats, and even fishnets. The best-known discoveries from bogs have been the well-preserved Iron Age bodies like the ‘‘Tollund man’’ of Denmark. Although acid bog waters have dissolved the bones of the ‘‘bog people,’’ the skin and the stomach, along with its contents, usually remain. Other waterlogged environments have preserved oak-coffin burials, Viking ship burials, textiles, and the wooden structures from Swiss lake dwellings. In historic archaeology, waterfronts of harbor towns have been preserved. A major problem with waterlogged wood is that it begins to dry, crack, and disintegrate when removed from its burial environment. It therefore requires special preservation measures, such as keeping the wood submerged or using waxes and chemicals.

Geochemical Indicators The stable isotope compositions of sediments, soil nodules, soil organic matter, vein calcites, and lake carbonates provide evidence of climate conditions.43 Most investigations have concentrated 183

Paleoenvironmental Reconstructions

Figure 6.5 Soil Organic Matter Carbon isotopes and paleoenvironments. The relative amount of the isotope 13 C in soil organic matter can be used to estimate the types of plants that were present on an ancient landscape. The relative amount of 13 C can be regarded as a measure of the percentage of C4 plants. Changes in 13 C content and the relative propor-

tion of C4 plants may reflect changes in climate and the kinds of resources available for use by people in the past. In this instance the last part of the Pleistocene, based on the 13 C content in the soil organic matter, is characterized by lower amounts of C4 plants, while the higher values of C4 plants reflect warmer conditions of the middle Holocene. (After Nordt 2001)

on the stable isotopes of oxygen, carbon, nitrogen, and hydrogen. The oxygen and hydrogen isotopic ratios in groundwater are determined chiefly by the ratios in the precipitation. Most soil organic carbon comes from higher plants. Plants with C 3 metabolism, the most widespread photosynthetic type, have carbon isotope ratios distinct from the C 4 pathway plants. C 4 plants thrive in drier and/or hotter climates (fig. 6.5). Because isotopes of a given chemical element

have different numbers of neutrons, they have different atomic weights (masses). Isotopes of a particular element will have the same chemical properties, but their different masses cause them to be separated or fractionated by certain natural processes. Two of the three isotopes of oxygen ( 16 O and 18 O) have been useful in the reconstruction of past environmental conditions. When water evaporates, the lighter oxygen isotope ( 16 O) is preferentially incorporated into water vapor,

184

Paleoenvironmental Reconstructions

while the heavier isotope ( 18 O) becomes proportionally higher in the remaining water. The fact that 18 O is preferentially left in ocean water during evaporation has been used to infer global climatic fluctuations. This has led to a revolution in our understanding of environmental and climatic change during the time of human physical and behavioral development. When these climate changes are dated, they can sometimes be used to ascertain the age of archaeological sites. Isotopic signals contained in marine sediment, calcite veins, and ice-core sequences appear to provide a continuous record of global climatic change for the interval associated with the archaeological record. The isotopic signals have been related to relative sea-level changes and alternating periods of colder global climates (glacials) and warmer global climates (interglacials). During Ice Ages the 16 O isotope does not immediately recycle back into the ocean but instead becomes part of the large ice sheets. The heavy oxygen isotope ( 18 O) becomes more common in oceans during these colder intervals. This colder isotope ratio is recorded in the shells of oceanliving organisms. When warmer global climatic intervals prevail, the lighter isotope, which had been trapped in the ice, returns to the ocean. Thus, during interglacials, there is proportionately less 18 O in the oceans. The changes in oxygen isotope ratios have been used to connect artifact-bearing deposits with climate chronologies. On the southern coast of the Mediterranean the Paleolithic archaeological sequence at Haua Fteah was correlated and dated with isotopic values. At Klassies River Mouth on the coast of southern Africa, Shackleton used oxygen isotope evidence to date shell-midden deposits associated with Middle Stone Age (MSA) artifacts.44 The variation in oxygen isotopes from the shellmidden deposits was correlated with the deepsea isotope record. The matches indicated that the MSA I midden contained shell that could be isotopically matched with substage 5e (fig. 5.11) of the continuous marine record. This was a period when the oxygen isotope composition of the ocean was as light as at present. Isotopes of the shells from the younger, MSA II horizon seem comparable to either substages 5c or 5a.

Based on these matches, and the estimated age of substages 5e, 5b, and 5c, it is possible to apply the oxygen isotopic composition of the shells to the dating of the MSA I and MSA II horizons using uranium-series techniques (see Chapter 5). MSA I artifacts matched with 5e probably date from about 130,000–120,000 b.p. while MSA II artifacts matched with either the 5a or 5c date from around 105,000–75,000 b.p. 45 Variation in isotopic 18 O measurements has been used to help explain the collapse of the Tiwanaku state in the Andes of South America between a.d. 1000 and 1100.46 Based in part on the oxygen isotope record from the Queleccaya ice cap, temperatures can be inferred to have increased beginning around 1000 and continued to at least 1400 (fig. 6.6). This temperature rise in South America seems to have been part of a global warming interval that lasted until the Little Ice Age. Higher values of 18 O are associated with decrease in ice accumulation. The temperature rise in the Queleccaya region may have been associated with a simultaneous drop in precipitation. These climatic changes appear to have profoundly affected the human populations of South America. The decrease in precipitation may have hurt the agricultural base of the Tiwanaku and ultimately led to the collapse of their political system. Hominids migrated out of Africa by 1.8 million years ago and occupied subtropical areas of southeast Asia. By 1.15 million years ago, hominids migrated from subtropical China into the temperate climate of the Loess Plateau. This may represent the first occupation of a non-tropical environment. Stable isotope rations from the loess paleosols provide dates for reconstructing the hominid paleoenvironments.47 The climate was characterized by cold/cool, dry winters and warm/mild, semi-humid summer and fall periods. Evidence shows that the hominids occupied the region during a warm period at least 1.15 million years ago and during both warm and cold cycles by 650,000 b.p. Strontium isotope analyses combined with petrologic studies on sediment cores collected in the Nile Delta of Egypt indicate that the climate and Nile flow changed considerably about

185

Paleoenvironmental Reconstructions

Figure 6.6 Relation of Ice Core Record to Agricultural Patterns Fluctuations in the oxygen isotope record and accumulation of glacier ice used to interpret the human prehistory of the Andes. High amounts of rainfall before A.D. 880 can be related to an expansion of irriga-

tion agriculture. An interval of increased drought is represented by low amounts of ice accumulation; isotope values above −18 occurred between about A.D. 980 and 1480. The human response to this change appears to be reflected by a collapse in dry farming and irrigation agriculture. (Ortloff and Kolata 1993)

4,200 to 4,000 b.p. in the Nile basin.48 Widespread drought and decreased Nile flow may have contributed to the collapse of the Old Kingdom. Isotope ratios derived from the remains of plants and animals have also been used to interpret prehistoric human subsistence and settlement systems. Stable carbon and nitrogen isotopic values have been used to indicate the presence of maize and the relative proportion of marine and terrestrial food consumed by prehistoric human populations. Changes in the environmental landscape inferred from the character of and constituents within sediment/soil sequences can be used to develop models of climate change. These models can then be employed to evaluate

the habitat context associated with human occupation.

186

Environmental Change and Archaeological Explanations Human Habitats and Geoecology Evidence for alternating glacial and interglacial periods and even finer-scaled (short-term) environmental-climatic intervals can be found in the stratigraphic record using the paleoecologic methods described earlier in this chapter. Since the Late Pliocene changes in global climate are reflected in the fluctuations between glacial and

Paleoenvironmental Reconstructions

interglacial conditions. Major changes in the behavior of humans may be associated with the end of the last glacial interval and the beginning of the present interglacial, the Holocene. The impact of the earth’s environmental and climatic patterns on human physical and behavioral development, and the effect of prehistoric human behavior on changing environmental landscapes, can be pursued by means of a geoarchaeological approach as part of a broader geoecologic perspective. Human-earth interaction is reciprocal.49 Environmental conditions have influenced humans, and humanity has had an increasing impact on landscape habitats. The concept of ‘‘environmental determinism’’ was first promoted to explain the origin of agriculture as being climatically influenced in Gordon Childe’s Most Ancient Near East. 50 At about the same time, Ellsworth Huntington and C. E. P. Brookes argued that major changes in human society were a result of fluctuations in climate. Environmental determinism was out of favor with anthropological archaeologists between 1960 and 1980, but it seems to be regaining some credence. For example, there is strong evidence for a close temporal correlation between climatic change and the origins of agriculture. Gordon Willey and Philip Phillips’ 1955 observation about North America that the shift at the beginning of the Early Archaic was not so much cultural as it was environmental, would not now be considered controversial. Paleoecologic research has supported the general idea postulated by Childe in the 1920s that the agricultural revolution in the Near East was a response by people to climatic events.51 Changes in climate have affected the physical and biotic contexts associated with human occupation throughout the Quaternary.52 For example, large lakes that once existed in currently dry areas disappeared because of major regional and global paleoclimatic events. In the hyperarid eastern Sahara of today, sedimentary deposits indicating past pluvial conditions during the Middle and Early Upper Pleistocene are associated with Acheulian and Middle Paleolithic artifact assemblages. Many remnants are composed of indurated lake marls that show the ear-

lier presence of large, perennial bodies of water. Fish bones within some of the lakebeds indicate the presence of streams connecting the lakes. No younger Upper Paleolithic artifacts have as yet been found. The absence of artifacts might be explained by a hyperarid interval during the late Upper Pleistocene that made the region uninhabitable. Playa deposits associated with the Terminal Pleistocene and Early Holocene contain Neolithic artifacts marking the return of habitats suitable for human occupation. Large lakes also formed in northern Chile around 14,000–9,000 b.p. The earliest human occupations of this region are associated with Terminal Pleistocene and Early Holocene beaches of lakes. High lake stands in the Andes are be related to glacial retreat and widespread aridity that developed from around 8,500 to 5,000 b.p. Human occupation appears to have occurred before and after this arid interval (from 11,000 to 8,000 b.p. and then again from 5,500 to 4,000 b.p. Loess deposits can provide indications of climate change that may have affected prehistoric human populations. In Argentina at Cerro La China, the stratigraphic sequence contains three episodes of loess deposition and several soil horizons associated with archaeological materials. Paleo-Indian materials dated to around 10,600 b.p. were recovered from the lowest loess. A soilforming interval occurred after the deposition of this loess, which lasted to about 5,000 b.p. It was followed by a period of erosion. A second cycle of loess deposition, soil formation, and erosion occurred in the Late Holocene and was followed by a short eolian episode. There are also important Quaternary loess-soil sequences in China, Europe, and central North America that provide clues to geoecologic conditions associated with human occupation of these regions (see figs. 5.8 and 5.9). Stratified Holocene sedimentary and archaeological contexts within the Upper Delaware Valley of Pennsylvania also appear to reflect regional climatic change. This sequence consists of Archaic and Woodland components. Sediments containing evidence of several major changes in depositional processes also reflect paleoenvironmental changes. The Early Archaic artifacts are

187

Paleoenvironmental Reconstructions

found in deposits dated to around 9,000 b.p. A decrease in landscape stability and increase in deposition from flood events occurred during the midHolocene and is associated with Middle Archaic artifacts (around 6,000–5,000 b.p.). Late Archaic and Middle and Late Woodland components are also present in the sequence. During the middle of the tenth millennium b.p., the site of Starr Carr on the sun-exposed north shore of an ancient lake (glacial Lake Pickering) in Yorkshire, England, provided a resource-rich location for a seasonal camp of the prehistoric human inhabitants.53 Almost all of the raw materials (flint, clay, birch bark, glacial boulders) used at the site were no more than an hour’s walk away. Abundant red deer apparently provided most of the animal protein for the humans. Geoarchaeological sediment coring has helped determine the sequence of environments and provided a detailed picture of the geoecologic landscape setting associated with Starr Carr. Where the sediment core showed woody coarse detritus, it was interpreted as representing a former fen carr. Reed peat was interpreted as the deposit of a reed swamp. Fine detrital mud was taken to represent a marginal open-water environment, while calcareous mud deposits were interpreted as indicating open water. The site was located at the mouth of a shallow gully on the southern slope of a low hill of glacial till that extended into glacial Lake Pickering near the entrance to the lake’s narrow outflow channel. The occupation area was situated where the open water most closely approached the shore. It seems that during Early Mesolithic times there was great variation in the vegetation of the lake margin. At about 9,800 b.p., open water lapped the shore, leaving only a narrow fringe of reeds and sedges. By 9,650 b.p., just before human occupation, a reed swamp extended well into the lake. During the period of occupation (beginning about 9,600 b.p.), the reed swamp was replaced by fen, and the area of the site was covered mostly by ferns. The inhabitants put down a wood matte to consolidate ground formerly covered by damp fen. By 9,300 b.p. the human occupation was over and a fen on which willows were growing covered the site again. Other ecofacts were em-

188

ployed to infer past environments and the kinds of resources that had been used at Starr Carr. Pollen analyses indicated that at the time of occupation, pine and birch probably covered the surrounding area. Other pollen types indicated the presence of open areas that may have been forest clearings. Antlers from deer and elk were used to establish the seasons in which the site was inhabited. Interdisciplinary paleoenvironmental investigations of this nature allow a clear picture of the ancient landscape setting to emerge. A NATO conference was held in the mid1990s concerning third-millennium b.c.e. climate change affecting the collapse of eastern Mediterranean and Near Eastern civilizations. Regime collapses included Old Kingdom Egypt, the Early Bronze Age in the Aegean, Troy II and Indus Valley civilizations—all centered around 2,200 b.c.e. plus or minus 100 years.54

Tectonics, Climates, Landscapes, and the Human Past One of the major integrating concepts in the earth sciences is that of tectonics. Many of the geologic and biologic features associated with the archaeological record can be more fully understood within the context of plate-tectonic theory. The basic idea of plate tectonics is that parts of the earth’s crust are being created, moved, and destroyed by the internal processes of the earth. At some places, like East Africa, for example, parts of the earth are separating, spreading, or ‘‘rifting.’’ At other places, like the Himalayas, the collision of pieces of the earth’s crust resulted in the creation of mountain ranges. In yet other areas—along the Pacific coast of South America, for example—oceanic crust has been moved under (or subducted) continental crust, which has led to volcanic activity and mountain building. Throughout the world the tectonic-related effects of mountain building, faulting, earthquakes, and volcanism have influenced the prehistoric record. Earth movements, or faulting, associated with crustal rifting resulted in the creation of lakes used by prehistoric humans, as well as in the exposure and discovery of sediments containing hominid remains and artifacts. Volcanism asso-

Paleoenvironmental Reconstructions

ciated with rifting or subduction led to the interbedding of fossil and artifact-bearing sediments with datable lava flows and ash deposits. The creation of mountain ranges is thought to have dramatically altered the Paleolithic climate system, causing both geomorphic and biologic impacts. There is a consensus that many broad-scale climatic changes inferred from the prehistoric record are a result of tectonic events. The climatic cooling observed in the marine record for the Cenozoic interval (the time since sixty-five million years ago) generally and the fluctuations of climate observed during the Quaternary (approximately the past three million years) in particular can be explained at least in part as the result of crustal movements caused by tectonic processes. Several major climatic cooling events may be the result of the rifting of continental crust or collisions of fragments or plates of the earth’s crust (or lithosphere). One of these tectonic connections resulted in the creation of the Isthmus of Panama. This and other tectonic events has been implicated in changes in climatic patterns which resulted in the interglacial and glacial episodes of the past three million years (fig. 5.4). The climatic fluctuations between glacial and nonglacial episodes affected the physical and biotic habitats associated with human evolution and the formation of the archaeological record. Climate, through direct expression as temperature and precipitation, or indirectly as vegetation, is an important factor in the nature and intensity of geomorphic processes. Major climate changes are particularly significant in the long-term stability of land surfaces and soil formation. Since Holocene climates have deviated significantly from present-day climates, the geomorphic responses will be preserved in sedimentary sequences and in landforms in many places. On the floodplains of the major river systems, flooding plays a key role in the erosion and transportation of sediments and ultimately in the stability of landforms. Floodplain systems develop a quasiequilibrium morphology that a change in climate may destabilize. Data on river runoff and from watersheds in the United States indicate two critical vegetation

boundaries of geomorphic significance. In grasslands in the temperate zone, where the mean annual temperature is about 10° C, vegetation becomes sparse when annual precipitation drops below about 500 mm.55 The second boundary occurs where the mean annual precipitation is about 800–900 mm and mean annual temperature is 10° C. Under these conditions grassland grades into forest. In undisturbed forests, if the mean annual precipitation is more than 900 mm, there is little overland runoff. Grasslands in good condition are nearly as effective as soil cover in reducing overland flow but may lose this effectiveness in reducing runoff and erosion during droughts. Past changes in precipitation are less well understood than past changes in temperature. There is a general tendency for middle latitudes to experience increased precipitation during cool climate episodes, the most recent of which was the Little Ice Age. Regional responses in sediment yields should mirror Holocene climate changes. Such responses vary with the contextual climatic regime. A change to warmer and drier conditions would decrease annual sediment yield in the arid southwest of the United States, while increasing annual sediment yields in the midcontinent grassland. The Holocene flood record can be reconstructed from dated paleochannels. The intensity of lateral channel migration varies with flood magnitudes. This has a major effect on locations of long-term habitation. Relatively modest changes of mean annual temperature and precipitation have been associated with relatively large adjustments in major flooding in the Mississippi flood record. The erosion, transportation, and deposition of fluvial sediments is an episodic process because of dependency on surface runoff and high-stage flows. These characteristics of floods play a key role in the stability and evolution of alluvial landforms and the ability of humans to live in this important environment. The Younger Dryas is a short, cold climate interval just prior to the beginning of the Holocene. High-resolution sediment and vegetation responses to Younger Dryas climate change are recorded in varved lake sediments from Meerfelder Maar, Germany.56 Thus, varve microfacies variations can be a sensitive proxy for environ-

189

Paleoenvironmental Reconstructions

mental changes. The observed varve changes have been quantified by physical and chemical analyses of sediments with a resolution between eight and forty years depending on sedimentation rate. Their investigations revealed that environmental changes at the beginning of the Younger Dryas occurred within twenty to fifty years. Paleolimnological and archaeological records spanning 3,500 years in the Lake Titicaca region of the Andes demonstrate that the emergence of agriculture, circa 1500 b.c.e., and the later collapse of the local Tiwanaku civilization, circa a.d. 1100, coincided with abrupt, profound climate change.57 Evidence for the timing and magnitude of climate changes is based on lake-level variation recorded in 14 C-dated sediment cores. Archaeological evidence established spatial and temporal patterns of agricultural field activity and later abandonment. This study demonstrates that climate change occurred rapidly during the Holocene and had significant hydrologic and ecologic impact on the inhabitants. Using the bulk titanium content of Caribbean sediment as an indicator of the hydrologic cycle for northeastern South America, it was possible to show that the collapse of the Mayan civilization in the Terminal Classic Period occurred during an extended dry period punctuated by more intense multi-year droughts at circa a.d. 810, 860, and 910.58 From circa a.d. 750–950, the Maya experienced a profound demographic decline. Many of the densely populated urban centers were abandoned permanently. The abrupt and extreme drought events certainly contributed to the demise of the Classic Maya civilization. Climate reconstruction continues to develop with more and better measurements of the two key parameters: chronological dating and the various proxies (for example, vegetation or isotope ratios) used to define the climate. However, it should be noted that there are uncertainties in both. Proxy data most often reflect ‘‘average conditions’’ that miss key short-term but critical events. Expressly geologic proxies such as extreme flood events are subject to abrupt changes in frequency. Easily recoverable proxy climate indicators may be difficult to find. For the Southern High Plains of the United States, stratigraphic

190

and sedimentologic data show that during the Clovis occupation (circa 11,000 b.p.), valleys contained perennial streams.59 A few hundred years later, when people were using Folsom artifacts, there was an abrupt change to a much drier environment with episodic drought. The earlier alluvial deposits are overlain by Folsom-age lacustrine and palustrine deposits. What natural events can cause radical cultural change? Except over very small regions, devastating earthquakes historically have not caused cultural change. Similarly, other rapid but geographically limited natural phenomenon like floods, fires, tsunamis, and volcanic cataclysms normally do not prompt dramatic changes in human societies.60 However, the volcanic eruption of the Greek island of Thera around 1600 b.c.e. is an example of geographically limited devastation —the island had to be abandoned. Geoarchaeologists need to seek explanations for cultural change in longer-term natural events. In this volume we do not consider the long history of war and related human-induced cultural change. There is one phenomenon where natural forces and human impact can work together to cause devastation and cultural change—at least depopulation, or abandonment of specific regions by human populations: the onset of catastrophic erosion. Throughout most of human history an adequate amount of fertile soil was necessary for permanent settlements. Deforestation can, and did, cause the erosion of most of the fertile soil in many regions. Today, advanced economies can import whatever is needed. This was not the case for most of the human past. The evidence for rapid erosion can be found both at the locus of the erosion and in the enhanced sedimentation rates downstream. Fires and heavy tilling of the soil also can be responsible for increased erosion. In the eastern Mediterranean region and the Near East, agriculture had spawned significant erosion at least by 6,000 b.p. Plato and other writers from classical antiquity knew of the relation between deforestation and erosion. Earthquakes can be devastating where the active fault zone is very long. An example is the great Anatolian fault zone, running over 600 km across northern Turkey. In a.d. 1688 a magnitude-8

Paleoenvironmental Reconstructions

quake, accompanied by more than 200 associated shocks over a six-week period, caused extensive devastation. Nevertheless, there is no evidence that earthquakes have caused major cultural discontinuities. Environmental change is a relative concept— relative to human values, technologies, demographics, and scale.61 The temporal scale of environmental change governs cultural response. The effects of short-term change depend on the duration, frequency, amplitude, and periodicity of the phenomenon. Unexpected short-term changes are likely to stress human communities and require an immediate response. An example of long-term environmental change is significant warming or cooling of the climate. The geoarchaeological record of such long-term changes often makes it difficult to date both the natural phenomenon and the cultural response. People respond only to perceived change, and gradual change is mostly perceived in hindsight. Climate not only has a direct affect on humans (for example, clothing worn, the availability of plant and animal resources) but is a major factor in geomorphic, pedogenic, and hydrologic processes. One might ask if a long section on climate is necessary in this book. One senior archaeologist puts it this way: An appreciation of the potential of geoarchaeology and the importance of palaeoenvironmental information is lacking at many universities where anthropological graduate degrees are granted. The post-modern paradigms in vogue in today’s graduate programs seem to minimize the role of climatic fluctuations and change on prehistoric societies. How else can you explain a dissertation that examines evidence of shifting political alliances within a Mississippian cultural hegemony but that never notes that the evidence for such shifts was contemporaneous with the start of the Little Ice Age? It is alarming that so little archaeology is published in such journals as Quaternary Research, The Holocene, Quaternary Science Reviews, and Quaternary International, but it is even more alarming how seldom articles published in American Antiquity or the several good regional

archaeological journals cite noteworthy, relevant palaeoenvironmental information from these four interdisciplinary outlets. Current anthropological thinking certainly can stimulate archaeologists to ask new or better questions and then formulate better research designs. But prehistoric societies were closely linked to particular ecologic settings, and these were subject to the effects of climatic vagaries.62 These comments aside, American Antiquity ran a long article putting forth a synopsis on climate change and the beginnings of agriculture during the Holocene. This study proposed that the Late Pleistocene glacial climates were extremely hostile to agriculture—dry, low in atmospheric CO 2 , and extremely variable on quite short time scales. When the climate ameliorated (with the onset of Holocene climates), plant-intensive resource use followed. The authors also considered rapid cultural changes as underpinning the origins of agriculture but state, ‘‘We are not aware of any proposals for major changes in the intrinsic rate of cultural evolution coincident with the Pleistocene-Holocene boundary.’’ 63

Microclimates Microclimates have a far greater influence on life forms than generally realized. Regional interactions between air, water, solar heat, the earth’s surface, rocks and soils, and its vegetation have not received sufficient attention. Although evaluating microclimatic evidence for habitation sites is difficult, some general principles remain, for example: weather is usually much more severe on the windward side of large hills; wind speeds are amplified and accelerated at ridge tops; in the northern temperate zone, solar energy is more abundant on south-sloping hills; and although it is colder higher up in the atmosphere, at ground level cold air and frost settle in low-lying areas. Microclimates are generated within a relatively thin layer of air near the ground. This is where the atmosphere interacts with soil, rocks, water, vegetation, and human-made objects. Here the atmosphere is influenced by surfaces that either

191

Paleoenvironmental Reconstructions

reflect or absorb solar energy during the day and radiate energy back into the air at night. Water retains solar radiation longer than solids, while surface sediments and soils radiate it back into the atmosphere quite quickly. Heat exchanges are most pronounced between land and water bodies. Valleys produce their own wind systems; the greatest variety of microclimates, therefore, is where offshore air meets rugged landforms. In parts of the earth that receive winter snow, lake-effect snow and the ability of vegetation to trap snow in drifts can account for vast differences in microclimates. In agriculture some crops, such as citrus, are sensitive to low temperatures during certain times of the growing cycle—a freeze can destroy not only the crop but the trees. Geoarchaeologists are asked to assist in paleoclimatic reconstruction. To do so, they must assess everything from deforestation and microclimatic niches to broad climatic change in developing an environmental context.

Human Interaction with Environment and Its Effects on Climate In the earth-surface environment, the lithosphere is the most stable component. Despite long-term patterns of tectonic instability associated with plate movement or short-term instability like volcanic eruptions, earthquakes, faulting, and related events, the lithosphere is far less dynamic than the atmosphere and hydrosphere. Powered by the immense energy from the sun, they are not easily modified by human intervention. But the local climate can be altered by human activity. For instance, the change from forest to open field alters the local heat balance and produces greater temperature extremes at the soil surface. The biosphere is the environment most easily modified by human activity. Out of the ‘‘wilderness,’’ Neolithic humans created settled communities and agriculture. This was a step toward increasing stability for human communities, but it led to new instability for the environment. The widespread onset of human use of fire brought enormous alterations to the floral environment. The forests of the subtropical and middle latitudes are now only a shadow of their primeval selves. Humans have affected geoecologic systems

192

through the clearing of forest for agricultural purposes. Deforestation in the Mediterranean basin was a long and complex process. Sedimentary sequences in the Mediterranean region appear to reflect two periods of erosion during the Holocene that are associated with human activity. The first erosional and aggradational interval was initiated during the Late Neolithic/Early Bronze Ages and had its strongest signal between around 3,800 and 3,100 b.p. From about a.d. 200 to 500, a second interval of erosion was initiated because of human abandonment of terraced hillsides. Somewhat contradictory evidence is available concerning the climate during the Bronze Age in central Europe. This was formerly considered a period of cooling after the postglacial thermal maximum. But low groundwater tables during the period strongly indicate a prolonged dry period. (Pollen and insect studies are unreliable as indicators of climate where human activities have significantly altered the natural vegetational landscape, because of activities like forest clearance.) Pedogenic clay mineral assemblages of interstratified smectite-kaolinite can be an indicator of paleoclimatic changes in Holocene from arid to humid. Investigations in the central IndoGangetic plains of north central India showed that brotite weathered to smectite and vermiculite in the soils during the arid conditions but the smectite was unstable during the humid phase and altered to smectite-kaolin.64 The vermiculite, smectite, and smectite-kaolinite were preserved after the humid period ended. Landscapes in Mesoamerica were also modified because of increased agriculture. Increases in human population and expanded agricultural use of the landscape in Guatemala between about 3,500 and 1,000 years ago have been related to changes in pollen, clays, organic matter, carbonates, and phosphates in lacustrine sedimentary sequences. In western Mexico three intervals of erosion have been correlated with increased human agricultural practices. A 1995 study provides an excellent review of landscape-human interactions in the Archaic Period in North America.65 The Archaic (Middle Holocene) was a period of landscape and climate change that accompanied a dramatic evo-

Paleoenvironmental Reconstructions

lution in human social organization, technological innovation, and trade networks. This volume should be consulted for its examples of the reciprocal relationship of change across the human and geomorphic-climatic landscapes. By the later Holocene, along the west coast of North America, prehistoric hunting and plant gathering were facilitated by the deliberate setting of fire. This affected the natural landscape. Burning was used to control the growth of brush and promote the growth of plants and increase deer populations. According to eighteenth- and nineteenthcentury accounts, the Chumash people promoted the growth of specific plant foods by burning. In the Great Basin of the western United States, the Shoshone and other communities burned the natural vegetation to increase the yields of plants, to increase food for game animals, and to drive prey toward ambush. In North America, slightly less than half the area of Canada and the United States was forested when the first European settlers arrived. A third of these forests are now gone. Within the geoecologic system, vegetation and soil are interdependent. If change occurs in the vegetative cover, the character of the supporting soil will be altered. The human alteration of the landscape is recorded in the soils, paleosols, and archaeological and related geologic sediments. The influence people have on the features of the earth is directly connected to the types of behavior they use to adapt to and function in the environment. Usually, human-earth interaction comparisons are based on the effect particular subsistence strategies had on modifying the original landscape. Hunter-gatherers and foragers are often considered to have had a relatively low impact on the natural environment. This is arguably not so in some instances. A very early impact on the ‘‘natural’’ landscape by humans may stem from their control and use of fire. Evidence suggests that humans may have used fire about 1.4 million years ago at the site of Cheswoanja in Kenya. Human use of fire resulted in major transformations of the biotic and physical landscape. Another major impact resulted from hunting. Along with climate change, Late Paleolithic human migration into

the Americas has been implicated as a potential factor in the extinction of biotic communities at the end of the Pleistocene. Human hunting also seems to have had a major impact on the vegetation and landscape of New Zealand. Hunting, the domestication of animals, and the initiation of agriculture had significant consequences for Holocene landscapes. Apart from the instances where direct observations are available, it can be difficult to distinguish between natural modifications to the environment and impacts caused by human behavior. It is often difficult to determine whether a fire was natural or deliberately set by humans. Sometimes it is possible to provide support for one alternative or the other. For example, the charcoal frequencies found in a sediment sequence in Australia associated with the last interglacial before human occupation are much lower than the frequencies found in Holocene sediments, when humans were present. Patterns of Holocene erosion and deposition linked to landscape modification caused by human activities have been found in the sedimentologic record of lakes. The biologic, mineralogic, and chemical components associated with lake sediments can be used to infer human activities. Mesolithic and Neolithic landscape use during the mid-Holocene in Britain led to increases in finegrained clastics and salt concentrations associated with a decrease in forests. Pollen diagrams from northwest Europe have been used to trace patterns of land clearance and cultivation starting in the early Neolithic. In Asia charcoal concentrations from sediments near the central Thailand archaeological site of Khok Phanom Di (around 6,000–3,000 b.p.) may reflect the deliberate burning of mangrove vegetation during the first occupation and later vegetational disturbances of the Middle Holocene landscape.66 Major modifications of the natural environment during the Holocene throughout many parts of the world can be directly related to human behavior. Farming and herding have had an important impact on landscape development because they can disturb the natural hydrologic, pedogenic, and sedimentologic processes that affect the landscape. Human behavior influences

193

Paleoenvironmental Reconstructions

the soil-sediment system in a variety of ways, especially in terms that also affect the processes of erosion and deposition. Devegetation through burning, clearance, and grazing disrupts the natural patterns and increases the effects of precipitation. Soil loosening caused by cultivation or animals also increases surface-water runoff and erosion. Changes in hydrologic and groundwater conditions and soil-water relations (soil moisture) can have a major impact on the environment. Higher water tables can cause increased leaching. Seasonal dehydration can eventually cause the development of waterlogged areas. Extremely dry conditions, with dramatic drops in the groundwater table (as is currently occurring in the Ogallala aquifer under the American Great Plains), may lead to devegetation and the consequent erosion of previously fertile soils. Irrigation can also lead to the concentration of salt in the soil, ultimately making the soil unusable for agriculture. Human construction activities have had direct effects on the soil-sediment system, primarily by modifying the drainage of water. Terraces, dams, and irrigation ditches used to increase moisture retention also cause the accumulation of sediments. Roads and paths lead to increased erosion, and buildings concentrate runoff, which also increases erosion. The construction of burial mounds, the accumulation of tell debris and middens, and the development of spoil heaps associated with mines also modify landscapes. The soil-sediment record demonstrates that fluctuating patterns of erosion, deposition, and stability have occurred, but determining whether human activities or natural processes have caused these patterns is not always as easy. Plowing causes distinct soil horizons to be mixed and leads to increased erosion. Eroded sediments and soils from upland or upslope areas will eventually be deposited in footslope areas, in floodplains, or in deltas. Increased deposition can lead to increases in the thickness of the A horizon or even complete burial of the soil. Intervals of erosion and

194

deposition (aggradation) may be interrupted by periods of stability that can be recognized by the presence of fossil A soil horizons. In addition to environmental modifications to improve hunting or promote plant growth, some changes in animal populations have been attributed to humans who were hunter-gatherers. Again, it can be difficult to distinguish between natural changes and modifications caused by human activity. Perhaps the best-known example of the problem is the controversy over whether climate change or human predation caused the extinction of the large mammals in North America at the end of the Pleistocene. From a geoecologic perspective it could be suggested that Paleolithic humans (including Paleo-Indian populations) and the now-extinct large Pleistocene animals were both affected by changing biotic patterns that prevailed during the PleistoceneHolocene transition. In this sense climatic change can be considered the prime mover in the changes. Different parts of the geoecologic system reacted in independent ways, thus causing a variety of interrelated patterns. Climate change resulted in the extinction of Pleistocene plant communities, as individual plant taxa responded to changes in precipitation, temperature, and soil development. Large herbivores were forced either to change their behavior (for example, migrate with the plants on which they most depended); or to change physically to better adapt to the new ecologic setting; or they became extinct. The animals that depended on the larger herbivores (predators, carnivores) or relied on both plants and animals (omnivores) also had to respond to these changes. In this sense the differences between Paleo-Indian and later postglacial Holocene artifactual patterns can be viewed as either (1) the extinction of human behaviors predicated upon the ecologic patterns of the Pleistocene or (2) the modification (adaptation) of these behaviors to fit the new Holocene ecologic habitats.

CHAPTER 7

Raw Materials and Resources

Geoarchaeologists, especially in Paleoindian studies, traditionally have not dealt with issues of lithic resources in a systematic, regional manner. Although geoscientists have been involved in studies of stone-tool resources, this work usually either is locality-specific ( focused on a single resource or type of resource or a particular type of method) or is done by archaeologists.—Vance Holliday 1997

T

he excavation of an archaeological site usually brings to light only a small portion of the material culture of a social group. Because artifacts made of rocks and minerals as well as the debris produced during their manufacture are stable in the earth-surface environment, they make up a large part of what is recovered. Most inorganic remains are derived from geologic raw materials. From early Paleolithic tools through building materials and ceramics to sophisticated metal alloys, the source materials are rocks and minerals. By the end of Predynastic times the following had been used for stone vessels in Egypt: alabaster, basalt, diorite, granite, gypsum, limestone, marble, schist, serpentine, and steatite. By Neolithic times stone was being used for construction in Cyprus and the Near East. At Tell es-Sultan, near Jericho in Palestine, stone was being used for house walls by about 6000 b.c.e. With the development of agriculture came additional materials for geoarchaeological study such as new tools, vessels for storage, and grinding stones, as well as major biogeochemical alteration of the soil.

Definitions Let us begin with some basic definitions.1 A rock is a specific aggregate of one or more minerals that occurs commonly enough to be given a name (granite, limestone). A mineral is a naturally occurring inorganic element or compound having a specific crystal structure and a characteristic chemical composition (quartz SiO 2 , calcite CaCO 3 ). Obsidian is a volcanic glass that has not yet crystallized but in all other respects is a rock. The three genetic types of rocks recognized by the geologic sciences and a basic nomenclature is provided later in this chapter. The word stone has many meanings. Within a geoarchaeological context the usage should be restricted to building stone and gemstone. Archaeologists use the word lithic (from the Greek lithos, meaning ‘‘stone’’ or ‘‘rock’’) for materials and artifacts made from rocks or minerals. Geologic nomenclature also makes extensive use of this Greek root: lithification (the compaction and cementation of an unconsolidated sediment into a coherent, solid rock), lithology (the description of the characteristics of a rock, such as color, mineralogy, and grain size), and lithosphere (the solid portion of the earth, as contrasted with atmosphere and hydrosphere). Lithic resources have been used since our hominid ancestors first threw stones and made stone tools. The ancient Greeks and Romans had specific names for a large number of common rocks and minerals. The Greek natural scientist Theophrastus published his On Stones in the fourth century b.c.e., and the last five books of Pliny’s Natural History (a.d. first century) are de-

195

Raw Materials and Resources

voted chiefly to the consideration of lithic materials. Pliny discusses approximately 150 separate rock and mineral species and indicates that there are many more that he does not discuss. Many names of common rocks and minerals come down to us from these early times. From Theophrastus come alabaster, agate, amethyst, azurite, crystal (quartz), lapis lazuli, malachite, and obsidian. 2 There is a vast array of archaeological lithic materials. Here, we have concentrated on those rocks and minerals that geoarchaeologists are most likely to encounter. Thousands of pages of articles on lithics give no evidence of the lithology. However, D. Black and Lucy Wilson begin the abstract of their article on chert sources by stating, ‘‘Belyeas Cove on Washademoak Lake, Queens County, is the only primary bedrock source of chert in New Brunswick known to have been exploited by Native people.’’ 3 This is an attractive opening sentence. The authors go on to present a detailed study of the geology (14 pages) followed by two pages on the archaeology. This may indicate a fortunate trend. They conclude, ‘‘Washademoak multicolored chert is a distinctive chert type associated with this source.’’ They do not fall into the ‘‘quick to source on the basis of color’’ category common in North American archaeology for many decades.

Minerals There are more than 3,000 mineral species. The chemical and structural makeup of the various species gives each a distinct set of physical properties, such as color, hardness, cohesiveness, and characteristic fracture. Chemical composition is the key characteristic in determining how certain minerals—like the ore minerals, which provide metals—were used in the past. Long before the development of writing, humans recognized such distinct mineral properties as hardness, cohesiveness, and color. They used hard, cohesive jade for adzes and formed tools and weapons from fine-grained varieties of quartz. Red hematite and black manganese oxide minerals were used in cave paintings and in burials. Brightly colored minerals

196

provided ornaments. Technological development led to the mining and smelting of ore minerals to produce copper, lead, silver, and iron. The hardness of a mineral is defined as its resistance to scratching. Hardness was quantified by a scratch test. The Austrian mineralogist Friedrich Mohs proposed the following scale of relative hardness in 1822: (1) talc, (2) gypsum, (3) calcite, (4) fluorite, (5) apatite, (6) orthoclase, (7) quartz, (8) topaz, (9) corundum, (10) diamond. Each of the minerals lower in the scale can be scratched by those higher in the scale. The scale is not linear; in absolute hardness diamond is three orders of magnitude harder than talc. Also, many minerals have a slightly different hardness in each crystallographic direction. Hardness is an important diagnostic property in the field identification of minerals. In addition to the minerals in the scale, the following materials serve as handy references for hardness (on the Mohs scale): fingernail: 2 to 2.5; copper coin: approximately 3; pocket knife blade: 5 to 5.5; window glass: 5.5; steel file: 6.5. Rocks, per se, as assemblages of minerals do not have Mohs hardnesses, although some monomineralic rocks like quartzite and marble will exhibit the hardness of their constituent mineral. A rock like granite, composed of orthoclase and quartz, would exhibit a Mohs hardness of between 6 and 7. However, the importance of granite for building stone lies not in its hardness but in its cohesiveness and its mechanical strength. Granite was used for hammer stone as much as for its cohesiveness and lack of tendency to fracture as for its hardness.

Chert and Chalcedony Rocks and minerals composed chiefly of quartz make up a large percentage of lithic artifacts. Quartz is the most stable of all major minerals under sedimentary conditions and in the earth’s surface environment. Few areas of lithic nomenclature are as confusing as that of fine-grained varieties of quartz (SiO 2 ), and chert has been used as a general term for any fine-grained siliceous rock of chemical, biochemical, or biogenic origin. Chert is usually a very hard compact material that fractures conchoidally when struck (quartz has a

Raw Materials and Resources

Mohs hardness of 7, and chert is between 75 and 99 percent quartz.) Barbara Luedtke opts for using chert as the general term for all rocks composed primarily of microcrystalline quartz.4 Our approach differs only slightly. For example, we distinguish chalcedony from chert because it has a different structure (fibrous) that is easy to recognize using the petrographic microscope. Chert occurs as bedded deposits, discontinuous lenses, and nodules that are usually interstratified with chalk, limestone, or dolomite. Chert is microcrystalline quartz composed of interlocking, often roughly equigranular, grains. It can be almost any color and accommodate a wide variety of impurities that affect its workability in lithic manufacture. Grain size also has a significant effect on fracture properties. The chief varieties of chert of interest in archaeology are flint, jasper, and novaculite. The term flint should be used primarily for the grayto-black nodular chert found in chalk and marly limestone. It is very fine-grained and ‘‘tough,’’ and it often has directional properties that are of value in flint knapping. Impurities in flint are usually less than 1 percent and consist chiefly of sponge spicules and calcite. The best flints of Europe occur in chalk beds. Jasper is a common, widespread red chert. It is usually fine-grained and dense, with up to 20 percent iron oxide. Jasper from the Eastern Desert has been used in Egypt since Predynastic times for beads, amulets, and scarabs. Jasper drills are reported by A. Biswas from the first half of the fourth millennium b.c.e. For more than 4,000 years, jasper from inland desert areas was traded over distances of more than 175 km to coastal California.5 Novaculite is white, unlaminated microgranular quartz of uniform grain size. It is not common and so is of special archaeological importance when it is found as an artifact. Chalcedony is also microcrystalline quartz, but when examined using a petrographic microscope it exhibits radiating fibers in bundles. It is more porous than chert, and in addition, chalcedony has a more greasy luster than chert. Chalcedony has no significant impurities and so tends to be whitish, although many chalcedonies turn red with heating. The name comes from Chalcedon,

an ancient maritime city on the Sea of Marmara in Turkey. In the Old World, one important example of chert mining is a discovery in the Nile Valley, at Nazlet Khater. This is an early Upper Paleolithic site that has been radiocarbon dated to around 33,000 b.p. Simple mining techniques were used to obtain the chert.6 In the United States, increased interest in technical studies on chert and related materials has led to the discovery of many prehistoric quarries and promoted geologic studies of the stratigraphic and geographic distribution of cherts in limestone.7 The most famous prehistoric chert quarry in eastern United States is the Dover Quarry in Tennessee. The quarry pits cover about two hectares. Large chert nodules and tabular pieces were made into hoes, adzes, and ceremonial objects. Ceremonial swords reached a length of 69 cm.8 The Late Plains Archaic Schmitt chert or ochre mine in the Horseshoe Hills near the Three Forks of the Missouri River in Montana supplied lithic material from 5,100 b.p. on from the chertbearing layers of the Madison Limestone Formation. Many hammerstones used to break up the bedrock have been found at the mine site(s). The ancient Maya of Mesoamerica exploited a large chert- and chalcedony-bearing region in northern Belize for the production of chipped stone artifacts.9 The chert is found as nodules in limestone. Several lithic workshops are located in the region. The occurrence of black chert and chalcedony artifacts at many archaeological sites in Belize has been of special interest. Analyses of P. Cackler and associates indicate that the black coloration is caused by weathering phenomena and is not indicative of geographic source.

Semiprecious Stones Semiprecious stones are gemstones of lesser value than precious stones like diamond and emerald. They usually have hardness of 7 or less on the Mohs scale. The most common semiprecious stones used by ancient craftsmen are varieties of quartz. We shall list here only the most important varieties of macroscopic quartz crystals for archaeology. Amethyst, which is of bluish, reddish, or pur-

197

Raw Materials and Resources

plish violet color, can occur in large crystals (15 cm long). It was used in Hellenistic and Roman times for engraved seal stones. The name comes from the ancient Greek word for ‘‘not drunken.’’ An amethyst amulet was supposed to protect from intoxication. Amethyst was also one of the twelve gemstones worn by ancient Israelite high priests to represent the twelve tribes of Israel,10 and amethyst mines in Egypt are known from Old Kingdom times. As early as the First Dynasty (2920–2770 b.c.e.) in Egypt, amethyst was used for beads (in bracelets and necklaces), amulets, and scarabs. The Romans exploited amethyst from the Eastern Desert (of Egypt), where it occurs in cavities in a reddish granite.11 Rock crystal is clear quartz in large crystals, used extensively in ancient times for everything from small seals to large vases. Diverse ancient societies (Greek, Roman, Chinese, Japanese) thought of these clear crystals in the same way; each group referred to them as ‘‘permanent ice.’’ Predynastic Egyptians mined rock crystal from a locality north of Aswan. Ancient Egyptians coveted rock crystal for beads, vases, and filling in the corneas of eyes in statues and coffins. By the eighth century, rock crystal deposits in Japan were developed on a commercial scale. Because quartz crystals break with a sharp conchoidal fracture, they were used in North America for projectile points. Some archaeologists have used the term rock crystal for the quartz of these points. Varieties of fine-grained crystalline quartz were also used for seal stones, jewelry, amulets, and related objects. Many of the names date from antiquity. The most important varieties were agate, carnelian, and sard. Agate is a varicolored quartz; the colors often occur in irregular or concentric bands. Sometimes agate contains mossy or dendritic inclusions that give the impression of landscapes or vegetation. Known from ancient times—its name comes from the Achates River in Sicily—agate has been used for projectile points and mortars, as well as for cups, bowls, and bottles. The Sumerians used agate in ceremonial ax heads. Engraved agates were highly esteemed by the ancient Romans. One Roman two-handled wine cup made of agate had a capacity of more than 550 ml and an exterior

198

carved with Bacchanalian themes. The prophet Muhammad wore a signet of Yemen agate. Carnelian is a red (sometimes deep blood red) variety of quartz that may be transparent or translucent. Carnelian is found abundantly as pebbles in the Eastern Desert of Egypt. It was used in Egypt from Predynastic times on for beads, amulets, and inlay in jewelry, furniture, and coffins.12 In the classical world carnelian was used for sealing rings and seal stones. Sard is similar to carnelian but more brown and opaque in appearance. It is the sardion of Theophrastus, named from Sardis, the capital of ancient Lydia in Anatolia. Onyx and sardonyx are banded forms of silica. In onyx and sardonyx the bands are usually straight and comparatively regular; in onyx, milk-white bands alternate with blackish bands, while in sardonyx, white bands alternate with reddish brown or red, as the name implies (onyx alternating with sard). A related mineral, opal (SiO 2 .H 2 O), occurs in many colors and has been popular since antiquity. The name is believed to derive from the Sanskrit word meaning ‘‘precious stone.’’ The use of opal as a gemstone in antiquity dates back to the fifth century b.c.e., when it was mined in Slovakia. Its use was prevalent in ancient India, and it was a favorite gem of the Romans, who mined it in Hungary. Opal and onyx were mentioned in the first book of the Bible. In the Americas, gem opal is found in a belt of scattered deposits that stretches from the northwestern United States through Central America to Brazil. Some Mexican opal was taken back to Europe by explorers as early as a.d. 1520. It has also found favor as a material for projectile points. Opal cannot survive long exposure to the effects of surface weathering. It tends to lose water and become cracked and opaque. Opal found in quarry discard piles is often chalk white and crazed. The nonquartz semiprecious stones most commonly found in archaeological sites are lapis lazuli, jade, and turquoise. Lapis lazuli is a mixture of complex silicates that exhibits a range of colors from deep blue to azure blue and from greenish blue to violet blue. It has a hardness of 5–6, depending on impurities. It has been used for at least 7,000 years; in Predynastic Egypt,

Raw Materials and Resources

it was a favorite material for cylindrical and flat seals, scarabs, beads, pendants, and amulets. Lapis lazuli was also used as a pigment. The most important source was at Baldachin, Afghanistan, a mining area that Marco Polo visited and described in a.d. 1271. There it occurs in large blocks and crystals in a white calcite matrix. Although it was popular in antiquity in China, India, Sumer, Israel, Egypt, Greece, and Rome, the modern name dates only from the Middle Ages and derives from the Latin word for stone (lapis) and the Persian word for blue (lazhward ). The Romans and Israelites called it saphiris. Jade can be one of two minerals: jadeite (a pyroxene with a hardness of 6.5–7) or nephrite (an amphibole with a hardness of 6–6.5). Jadeite is somewhat more vitreous, while nephrite is more oily in appearance. Nephrite is generally more cohesive because of its texture of interlocking fibers. The coloring of jadeite comes from minor chromium; it can be a rich emerald green. Jade can also be bluish, purplish, or even reddish. The Chinese have made extensive use of jade since the beginning of the Neolithic; they acquired their nephrite pebbles from Burma (where it occurs in serpentinite in boulders that can weigh up to several tons) and from Chinese Turkistan.13 Burmese jade occurs in a wide range of colors. The Aztecs and Mayans also used jade extensively,14 although it has been found less commonly in archaeological contexts in other parts of North and South America. The occurrence of jadeite-jade in Guatemala was recently confirmed by nondestructive Raman spectroscopy.15 A source for Mesoamerican jadeite, jade has long been sought because of the abundance of jadeite artifacts in the region. Throughout history many names have been used for jade, and many materials have been called jade, so considerable uncertainty exists about early references to it. The name in English dates only to a.d. 1727 and comes from the Spanish term piedra de yjada, meaning ‘‘stone of the side’’—an allusion to its supposed powers to cure side pains.16 Turquoise (CuAl 6 (PO 4 ) 4 (OH) 8·5H 2 O) is a blue or bluish-green copper aluminum phosphate mineral with a hardness of 5–6. It is a secondary mineral usually found in small veins in weath-

ered volcanic rocks in arid regions. Some turquoise will change color as a result of dehydration shortly after mining. Turquoise was highly prized by, among others, prehistoric societies of the American Southwest, the ancient Egyptians (since the Neolithic), and many other groups throughout the ancient Near East. The famous Persian mines were found near Nishapur in the province of Khorasan (present-day eastern Iran). The name turquoise is French, meaning Turkish, the original stones having come into Europe through Turkey from Persia. In the southwestern United States turquoise deposits were mined in prehistoric times. Much turquoise has been recovered from sites in Chaco Canyon, New Mexico.17 In some ancient quarries, stone mauls, pecks, and chisels have been recovered. There is evidence that the common technique of heating and quenching the bedrock was used to free the turquoise. In southern North America (Mesoamerica) turquoise was in common use. Aztec ruins contain many human skulls that were completely covered with turquoise. The feldspars are a group of silicates that are the most common minerals in the earth’s crust. They are major constituents of most igneous rocks. Feldspars are generally dull, cloudy, and fairly opaque, so they have not been sought after as ornamental ‘‘stones.’’ They have, however, been found in archaeological contexts and as important constituents of igneous rocks. The most common feldspars are orthoclase, plagioclase, and microcline. Semiprecious stone varieties of feldspar include sunstone, moonstone, and amazonstone. In America, people used all three as gems: moonstone by various groups in Mexico, sunstone by the Apaches in Arizona, and amazonstone by the Aztecs, Mayans, as well as other groups of people living in Venezuela, Brazil, Trinidad, Wisconsin, and California.

Other Archaeologically Important Minerals In addition to semiprecious stones of considerable hardness a number of softer minerals were widely used in similar fashion. Alabaster, a fine-grained, cohesive variety of the mineral gypsum, usually has a whitish to pinkish color. It is quite soft: its

199

Raw Materials and Resources

Mohs hardness of 2 is less than that of a fingernail. Alabaster was employed in ancient Egypt from early Dynastic times as a subsidiary building material to line passages and rooms. It was also popular for funerary vessels that contained the viscera of mummies. Some Egyptian material that has been called alabaster is actually fine-grained calcite, which has a hardness of 3. Most New Kingdom (circa 1575 to 1078 b.c.e.) stone vessels were alabaster (both true alabaster and calcite). The name comes from the Greek word alabastros; the Greeks used it to make ointment vases. It was used by the Etruscans for vases, urns, and ornaments. Malachite is a dark-green copper-carbonate hydrate (Cu 2 CO 3 (OH) 2 ) with a hardness of 3.5– 4.5. It was most likely the first mineral ever smelted (see the discussion of ores, below), but it has also been used since the Neolithic for beads and other ornaments. It occurs in the oxidized portions of copper ore deposits, especially in regions where limestone is present. Along with many other green minerals, malachite was much in favor with the ancient Egyptians for ornaments. Malachite was used as an eye-paint as far back as 5000 b.c.e. in Egypt. The name comes from the Greek mallow (resembling the color of the leaf of the mallow), in allusion to its green color. Although it is not strictly a mineral, amber has been used as a precious stone since the Neolithic. Amber is fossilized tree resin from evergreens. It lacks crystalline structure. Amber is brittle and breaks with a conchoidal fracture. It is soft enough to be cut easily by any metallic knife. Easy to shape, amber was used for a wide variety of ornaments and amulets. The Greeks and Romans so revered amber that it was reserved for the nobility. Amber is found in many parts of the world, but the principal supply for Europe and the Near East has been the Baltic region, where it has even been collected from the Baltic Sea because it is buoyant enough to float on salt water. Baltic amber has been found in archaeological contexts as disparate as central Russia, Etruscan Italy, Mycenean Greece, Pharaonic Egypt, and early first millennium Mesopotamia (Iraq). The world’s oldest known amber comes from the Appalachian region of the eastern United

200

States. Other sources in the Americas are Manitoba, Alaska, the Atlantic seaboard, Mexico, Ecuador, and Colombia. A primary source is the Dominican Republic, where it is found in the north, the east, and the center of the country, near the village of Cotuí. It was the first American gemstone recognized by Columbus (a necklace of amber was presented to him by the indigenous peoples of the Dominican Republic as a welcoming gift). Amber earplugs and beads were found in Tomb-7 near Monte Alban, Mexico, dating from around a.d. 1100. The Aztecs controlled the amber trade from their capital, Tenochtitlán. This trade network went through Chiapas, an outpost of the Aztec Empire. The Codex Mendoza mentions amber as part of the tribute materials paid to Montezuma by certain districts in Mexico. Amber was also in common use by the Eskimo.18 In southeast Asia the major source of amber was Burma. Burmese amber has been imported into China since the Han Dynasty (202 b.c.e. to a.d. 220). Often mistakenly called ‘‘black amber,’’ jet is a compact and dense form of fossil coal. Geologists think that jet derives from water-logged pieces of driftwood. High-quality jet has a conchoidal fracture and no foreign matter like pyrite (FeS 2 ) (common in most coal). Because it takes a high polish and is a deep, pure, velvety black, jet has often been used for ornaments. Its Mohs hardness varies between 3 and 4. The jet of Whitby, England, appears to have been used in Britain since pre-Roman times. In North America highquality jet has been found in southern Colorado and Pictou, Nova Scotia. The name mica derives from the Latin word meaning ‘‘to shine.’’ Its name derives from its use as a substitute for glass in Old Russia (Muscovy). Mica is not one mineral but a group of related sheet silicates, the most important of which is muscovite (KAl 2 (AlSi 3 O 10 )OH 2 ). Muscovite is colorless and transparent; with one perfect cleavage, it can occur in sheets called books that are as large as a meter across. It was therefore used for mirrors in ancient Nubia (northern Sudan and southern Egypt). Prehistoric societies throughout the Americas used muscovite. It was mined from many deposits in the Appalachian region

Raw Materials and Resources

and Alabama and traded as far west as the Mississippi River. From a single mound of the ‘‘Mound Builders,’’ more than 250 mica objects were recovered. Many of the large sheets of mica found in Hopewellian ceremonial contexts in eastern North America apparently came from the southern Appalachian region. Mica outcrops in western North Carolina were mined extensively in prehistoric times. L. Ferguson provides an annotated list of prehistoric mica mines in the southern Appalachians in ‘‘Prehistoric Mica Mines in the Southern Appalachians.’’ 19 A hydrous calcium sulfate, gypsum is a soft mineral that has many uses. It is a common mineral, widely distributed in sedimentary rocks, often as thick beds. The colorless, transparent variety called selenite can be found in large-cleavage sheets that have been used as windows since Roman times. Alabaster is another variety of gypsum. Gypsum has also been used widely as a plaster (its name comes from the Greek word for plaster), and it is the raw material for plaster of Paris. The Minoans used large gypsum blocks as building stone. Common table salt, the mineral halite (NaCl) has been the source of salt for human nutritional needs from the beginning of human evolution. It has also served as a preservative, a medium of exchange, and a source of tax revenue. Salt oases lay on the caravan route through the Libyan Desert in the time of Herodotus. The salt mines of ancient India were the center of widespread trade. The salt of its harbor Ostia supplied some of Rome’s needs. Those of Caesar’s soldiers who were ‘‘worth their salt’’ received part of the pay (their salarium) in the form of salt. Throughout the world, salt takes little energy to exploit, and it has always been easily transported. It can be recovered by boiling sea water or merely by allowing the water to evaporate. Large-scale boiling of brine to recover salt dates back to the Iron Age in Europe. The journals of the Lewis and Clark expedition record the recovery of salt by boiling sea water while on the Pacific coast near the Columbia River in 1806.20 Halite also occurs in major beds in many Triassic sedimentary rocks in central Europe and England. Halite deposits form when arms of the sea

are cut off from a supply of water and dry up. Sea water is more than 3 percent NaCl. In coastal estuaries evaporation can increase this percentage to about 8 percent. Pliny identified and described the three different raw materials for salt production: rock-salt beds from evaporite deposits, brine, and sea water. In prehistoric Europe the rock-salt mines of the eastern Alps were a major source: the mines at Salzburg, Hallstatt, and Hallein were worked on a scale far exceeding local needs. Miners dug as far as 350 m into the salt beds of the mountain sides. We probably know more about the Mayan production of salt than from other regions.21 The Maya had two major salt production regions: Belize and Yucatan. Of geoarchaeological interest is the question of how important salt was to the Mayan economics of the northern Yucatan. This issue is connected to the changes of both relative sea level and climate during Mayan times. Salt workshops were submerged by a relative sea-level rise that inundated the Yucatan coasts of Belize and Mexico. The importance to salt production of the contrasting coastal geology of northern versus southern Belize has been documented. Yucatan was the greatest producer of salt. Salt beds extended along the coasts, where the salt was collected at the end of the dry season. Underwater excavations off the coast of Belize have uncovered a site where the Mayans produced salt from sea water more than 1,000 years ago. The salt trade was important in the development of Classic Mayan civilization. Plants have very low sodium content, so inland agricultural societies need mineral salt. People in the tropics have high salt requirements because they sweat so much. They therefore were forced to import their salt. Salt has also been recovered from Guatemalan wells. Salt deposits can form from the hot brines of hot springs. Since around the year a.d. 1000, humans have excavated salt from brines and brine deposits originating in the Rift system in East Africa. Of special note are the deposits at Bunyoro on the eastern shore of Lake Albert. In the Krakow region of Poland, the earliest salt-making sites date to the Middle Neolithic. Ditches, storage tanks, hearths, pits, and ceramics have all been recovered in association with salt making from

201

Raw Materials and Resources

brine springs. The brine was channeled through clay-lined ditches to prevent seepage into the sandy soil, into storage tanks, and into ceramic vessels used for heating and evaporation. The Krakow region also has salt mines that have been worked for more than 1,000 years from a layer of salt nearly 400 m thick lying far beneath the surface. In Romania, many saltwater springs provided salt from the Neolithic through the Middle Ages. In Japan the environment for salt production is poor. Although surrounded by salt water, the islands of Japan lack coastal flats for evaporation fields; in addition, they have too little sun. As a result, at least for the past 1,400 years, the Japanese have relied on a two-step process for concentration and evaporation. Seaweed soaked in seawater was dried, and the salt that precipitated was rinsed off into more seawater, which produced a more concentrated brine that was evaporated by being heated in clay pots. It should be noted that the evaporation of sea water will yield not only halite but also the other salts that are dissolved in the sea. The salts produced by evaporation in natural saltpans occur in the following order: calcium carbonate (calcite), calcium sulfate (gypsum), sodium chloride (halite), and potassium magnesium chloride (carnallite, KMgCl 3·6H 2 O), with halite occurring in the greatest quantity. The composition of natural brines varies widely. Many are unsuitable for salt production. Natron is a naturally occurring mixture of sodium carbonate and sodium bicarbonate. It occurs abundantly at Wadi Natrun (a depression in the Libyan Desert, about halfway between Alexandria and Cairo). Each Nile flood vastly increases the supply of water entering the wadi and its string of small lakes. During the dry season evaporation from the lakes causes deposition at the bottom of the lakes and as an incrustation on the ground adjoining them. In Ancient Egypt, as Pliny notes, natron was also prepared artificially, in much the same manner as table salt, except that Nile water rather than sea water was used. The Egyptians used natron in mummification, for purifying the mouth, for making glass and glaze, for bleaching linen, and in cooking and medicine.

202

Metals and Ores Of the approximately seventy metallic chemical elements, eight (gold, copper, lead, iron, silver, tin, arsenic, and mercury) were recognized and used in their metallic state before the eighteenth century. Only gold and copper were sufficiently available in their native (metallic) state to be of importance to early societies. In the Old World, metallurgy had its beginnings in the Near East more than 7,000 years ago with copper and gold. The first uncontested use of metallic copper dates to the late eighth millennium b.c.e. at an aceramic Neolithic site in southeastern Turkey. Here archaeologists found beads made of native copper. The beginnings of metallurgy in the New World may have begun in the Andes of South America at about 1400 to 1100 b.c.e. 22 Metal artifacts dating to this time have been found in coastal Peru. Copper smelting and bronze production appears along the coast of Peru by a.d. 1000. To the east, in Bolivia, the development of metallurgy has been inferred from lake sediments. Metal (Pb, Sb, Bi, Ag, Sn) concentrations within the sedimentary sequences indicate three metallurgical zones over the past 1,300 years. The oldest is linked to Tiwanaku artifacts dating from a.d. 1000 to 1250 while another can be correlated with the Inca and the early Colonial period dating from a.d. 1400 to 1650. The sediments also reflect tin mining in the late 1800s and early 1900s. In the New World, Andean metallurgists developed sophisticated technologies that later moved north and flourished from Panama to Mexico. The Andes Mountains contain some of the richest gold, copper, tin, and silver mines in the world. Unlike the Old World focus on copper, metallurgy in Andean societies focused on gold from the middle of the second millennium b.c.e. The word ore is derived from an Anglo-Saxon word meaning a lump of metal. It is applied to an aggregate of minerals from which one or more metals can be extracted at a profit. Therefore, what may be an ore under one set of economic conditions may not be an ore under other economic circumstances, for example, when it is possible to import a metal more cheaply than to ex-

Raw Materials and Resources

tract it from a local deposit. In recent years there has been a tendency to drop the requirement that the desired substance be metallic. The worthless material from an ore deposit is called gangue. Copper (Cu) and gold (Au) were the first metals to be used by humans, primarily because they occur in the native, uncombined state in nature. The ancient Egyptians particularly prized gold for its eternal sheen, and they had a nearby source, the extensive deposits in Nubia. There is also a gold-bearing region between the Nile and the Red Sea. In the ancient world gold was exploited for purely decorative purposes. Deposits in the Taurus Mountains of Anatolia were extensively mined in the Early Bronze Age, and deposits in Greece were worked throughout the Bronze Age. In the New World, the use of gold came much later but reached great volume and artistic heights south of the Rio Grande, in Central America, and in northern South America. Indeed, stories of the abundance of gold objects were the driving force behind the Spanish conquest. Gold is widely distributed in small amounts. It usually occurs in high silica (SiO 2 ) rocks and quartz veins. Its exceptionally high density (19.3 g/cm 3; more than six times the density of the average rock) causes it to concentrate in what are called placer deposits. Its melting point of 1,063°C means that it can be melted and cast, and its chemical status as a noble metal means that it is free of unsightly corrosion. Most scholars believe the first gold recovered came from placer deposits. This term refers to gold that has been eroded out of bedrock and redeposited in alluvial sediments. About half of the approximately 900 occurrences of gold in Europe’s Bohemian Massif are placer deposits. These placers have been exploited for at least 3,000 years. Most placer gold is in the form of fine dust, but larger lumps called nuggets are found. The size ranges from submicroscopic to the size of a pea. The most common size for placer gold is that of sand. In the Greek legend of the ‘‘Golden Fleece,’’ ancient placer miners extracted alluvial gold by shoveling the gravels into sluice boxes hollowed out of tree trunks. A lining of sheep skins trapped the gold particles. The coarse gold was shaken out, but the fine gold adhered to the

wet wool. The fleece was then hung out to dry so the gold could be beaten out. Jason and the Argonauts set sail in search of this ‘‘Golden Fleece.’’ In the New World, the use of gold came much later but reached great volumes and great artistic heights south of the Rio Grande River, in Central America, and northern South America. Indeed, stories of the abundance of gold objects were the driving force behind the Spanish conquest and led the early European explorers in a search for ‘‘El Dorado.’’ Placer gold was apparently available throughout much of ancient China but seems to have received little attention until the Late Shang period. It was used to add color to other materials such as bronze. By a.d. 23, in the Han period, the Imperial Chinese treasury contained over 6.3 million ounces of gold,23 so Marco Polo’s account of the wealth of China’s Imperial court in the late thirteenth century may be accurate.24 The most productive fossil placer in the world has been the Witwatersrand of South Africa. These deposits were formed over 2.5 billion years ago within fluvial fans in shallow water. Silver (Ag) does occur in the native state, but in ancient times most of it came as a byproduct from the smelting of lead ores. Silver was used in Egypt before 3000 b.c.e. but was comparatively rare until the Eighteenth Dynasty. In North America, native silver has been found in some Mound Builder (Hopewell) sites. Electrum is an alloy of gold and silver. The earliest electrum used was probably natural, although by Greek and Roman times artificial electrum was also used. Silver has been recovered from lead ores for millennia. Mining of lead/silver ores began at least as early as the third millennium b.c.e. in the Aegean area and the Near East.25 Although most of the silver recovered throughout history has come from lead ores, some exploitable native silver deposits do exist. Among the most noteworthy of these is the one at Kongsberg in southern Norway, which has been worked for several centuries. Shortly after Columbus came to the Americas, the great silver deposits of Central and South America were discovered. These have remained the major source of the world’s silver supply. Most silver has always been produced as a byproduct of lead, copper, and zinc mining. Ex-

203

Raw Materials and Resources

ploitation of the large silver deposits in the New World south of the Rio Grande River during the sixteenth century radically changed the monetary climate in Europe. Silver also occurs as argentite (Ag 2 S), which is named after the Latin word for silver, argentum. It is sometimes called silver glance. In west Mexico during the Late Postclassic period (a.d. 1300–1521) argentite and silver sulfosalts were mined as a source of silver.26 Over a considerable span of time, ancient metalsmiths discovered the rewards of applying pyrotechnology to metalliferous rocks. The smelting of copper (Cu) ores goes back six millennia in the Old World. Our knowledge of the origins of copper alloy metallurgy is primarily indirect, derived from analyses and interpretations of the composition and structure of artifacts, slags, and ores. Because archaeology (particularly Old World archaeology) has focused on the excavation of temples, graves, and habitation sites, relatively few examples of Chalcolithic or Early Bronze Age metallurgical or mining sites are known. Most of these have provided scant evidence of ores or technology. Only since the time of the English chemist and physicist John Dalton (a.d. 1766–1844), who made significant contributions to our understanding of what a chemical element is, has the concept of chemical elements been used in a scientific or technical sense. The development of alloy metallurgy, therefore, must have come about without the artisans having any clear notion of elements and compounds. Ancient metalsmiths must have been aware of the results of smelting a mix of different ‘‘ores,’’ but they must have marveled at the outcome of the results of smelting both metallic-looking and nonmetalliclooking stones. A variety of scenarios may be proposed to account for the initial discovery that copper can be separated from such nonmetallic-looking minerals as malachite and azurite. Bright green malachite (Cu 2 O 3 (OH) 2 ) or bright blue azurite (Cu 3 (CO 3 ) 2 (OH) 2 ) may have been applied as decoration on the surface of pottery by Chalcolithic artisans. If the pottery were then fired in a reducing atmosphere, copper beads would have formed. Malachite and azurite begin to decompose below 400°C. It is likely that the earliest

204

copper smelting was done well below the melting temperature of copper (1,083°C). When ancient metallurgists learned how to smelt the more difficult copper sulfide ores, perhaps about 2000 b.c.e. in the eastern Mediterranean, they turned to the much more abundant copper iron sulfide, chalcopyrite (CuFeS 2 ), for their ore. Chalcopyrite is the most widely occurring ore of copper. It is easily recognized by its brass-yellow color. The development of copper sulfide metallurgy—which is more complex than copper oxide metallurgy—allowed the continued expansion of bronze-making in the ancient world. The sulfide ores in copper deposits lie below the oxidized ores, so ancient metallurgists may have been forced into sulfide smelting when the oxide ores were depleted. In North America prehistoric people did not smelt, melt, cast, or alloy metals, relying instead on the relative abundance of native copper. Copper use had begun in the Lake Superior region by 5,500 years ago. The native copper occurrences in the Lake Superior region are by far the most extensive in the world. Masses and sheets of native copper in the Keweenaw Peninsula of Michigan in the United States occur at the surface throughout a zone 5 km wide and nearly 150 km long. Outcrops are also common on Isle Royale and Michipicoten Island in Lake Superior. In prehistoric times, nuggets in a wide range of sizes must have been found in streams and along the Michigan shores of Lake Superior. Supplies available at the surface did not always exceed demand, however, as shown by the thousands of shallow pits dug to recover near-surface pockets of metallic copper.27 One of the largest secondary native copper deposits is located in the Santa Rita district of New Mexico where millions of pounds of native copper were mined in the early days of ‘‘modern’’ mining. Unfortunately, the native copper deposits of the southwestern United States have not received the attention of archaeologists and geoarchaeologists accorded the Lake Superior region deposits. In Mesoamerica, copper ore sources exploited during the Late Postclassic period (a.d. 1300–1521) occur in west Mexico.28 The minerals mined and smelted included copper carbonates and copper sulfides. Copper alloy metallurgy came late to

Raw Materials and Resources

sub-Saharan Africa, not much earlier than the introduction of iron smelting. Native copper is found in lode (original) deposits; in river, shoreline, and lag deposits of rounded nuggets; and in nuggets from glacial till. Subject only to surface alteration, native copper is nearly indestructible in the surface geologic environment. Most of the sources of native copper are associated with mafic volcanic rocks (for example, those in the Lake Superior region, in the Copper River in Alaska, on the Coppermine River in the Northwest Territories, and at Cap d’Or, Nova Scotia) and are found in the oxidized zone of copper sulfide deposits (for example, in southwestern United States and in the Near East, where metallurgy began). A deep blue complex copper hydroxycarbonate related to malachite, but less abundant, azurite is an easily smelted ore of copper that was used both for its copper content and as a blue pigment, probably as early as the seventh millennium b.c.e. It forms in the upper oxidized zone of copper sulfide deposits, along with the more common malachite. Cinnabar is a blood-red mineral, mercury sulfide (HgS). The Mayans prized it highly as a pigment, probably because its color could symbolize blood and blood sacrifice. Most of the cinnabar probably came from the Mayan highlands. Excavation of a Mayan site in Belize dating to the late ninth or early tenth century uncovered an offering vessel containing more than 100 gm of hematite, 19 gm of cinnabar, and other objects floating in a pool of 132 gm of mercury. Two possible sources for the Belize mercury are the Todos los Santos Formation of Guatemala or the Matapan Formation of western Honduras. This suggests that the mercury was acquired locally and not through trade.29 Native mercury is rare in geologic deposits, and it is not known whether the Mayans mined liquid mercury or smelted it from cinnabar. Cinnabar was used for the coloring on the famous oracle bones of ancient China. But all the ancient artisans had great difficulty differentiating among the various red pigments. Pliny’s account of red mineral pigments is quite garbled. As a tin oxide (SnO 2 ), cassiterite is the only ore of tin. (Tin becomes molten at 232°C and was

occasionally used in its metallic form in the ancient Old World.) Because of its specific gravity of 7 g/cm 3, cassiterite is found in placer deposits, often along with gold. Cassiterite is widely distributed in small amounts, but cassiterite deposits of ore grade are rare. Perhaps the greatest unsolved problem in Old World Bronze Age metallurgy is where the tin came from. The plentiful placer cassiterite from Cornwall was certainly mined for British Isles Bronze Age metallurgy, but we do not know how widely that ore was traded. By Roman times cassiterite from Iberia and Cornwall was available throughout the Mediterranean. The name cassiterite comes from the ancient Greek word for tin. Unlike gold, silver, and copper, lead (Pb) does not occur in the native state but must be smelted from its ores, especially lead sulfide (galena), lead carbonate (cerrusite), and lead sulfate (anglesite). Lead becomes molten at 327°C, so it is easily cast. It may have been smelted as early as copper: it is known from the Old World as long ago as the sixth millennium b.p. Lead mining at Rio Tinto, Spain, goes back to about 900 b.c.e., and at Lavrion, Greece, mining extends at least back to the late prehistoric times. Lead-tin pewter has been used since Roman times. The Romans used lead for many purposes: storage vats, water pipes, pewter. It was also used as a preservative in wines. This toxic element then found its way into their bodies, causing severe medical disorders and leaving its archaeological trace in their bones.30 Litharge (PbO), called red lead, has been widely used since ancient times as a red pigment. It too is quite toxic. Natural litharge forms in the oxidized zone of lead ore deposits. By Roman times red lead was also being manufactured from other lead minerals such as cerrusite (lead carbonate). Galena is lead sulfide (PbS). It is recognizable from its perfect cubic cleavage, high density, and silvery metallic color. It is a common metallic sulfide frequently associated with silver minerals. Since Greek and Roman times a large part of the supply of lead has come as a byproduct of ores mined for their silver content. The Romans gave the name galena to this lead-ore mineral. In North America where there is no evidence of prehistoric smelting to recover metals, the bril-

205

Raw Materials and Resources

liant silvery luster of galena nevertheless attracted the indigenous peoples, and it was used extensively in burial practices and for ornaments. Geologic sources of galena are numerous in the region from the southern Appalachians to the western Great Lakes area, and from Virginia to northeastern Oklahoma. Galena has been reported from more than two hundred prehistoric sites in eastern North America. The earliest known use was during the Early Archaic. Of the 232 sites reported by J. Walthall, 60 percent were mortuary sites in which galena was used in a burial association.31 Galena rarely occurs in sites earlier than the Late Archaic. During the Late Archaic and Early Woodland periods galena is found in the Great Lakes and Mississippi Valley regions. During the Middle Woodland, a large quantity of galena was moved through regional and long-distance exchange systems. More than sixty Mississippian sites from Illinois to the southern Appalachians contained galena. These artifacts often exhibited ground, rather than natural, facets. A red iron oxide, hematite (Fe 2 O 3 ) (hardness 5– 6.5) derives its name from the Greek for ‘‘blood red.’’ But although the streak (the color of the finely powdered mineral, so called because of the diagnostic test used by geologists of drawing a mineral across a piece of unglazed porcelain and noting the streak) of hematite is red, the color of the mineral itself can be black or, in the case of specular hematite, a silvery metallic color. Hematite is widely distributed in rocks of all ages and is the most abundant and important ore of iron. Engraved cylinder seals of hematite were found in the ruins of Babylon, and it was used extensively by the ancient Egyptians. Red ocher is hematite and has been commonly used as a pigment throughout history. The ancient Egyptians had plentiful supplies of red ocher near Aswan and in the oases of the Western Desert. References to what iron ore deposits were exploited are scarce in literature. Iron oxide minerals are nearly ubiquitous at the earth’s surface, so early iron metalsmiths did not have the resource problem faced by bronzesmiths. Iron makes up about 5 percent of the earth’s crust, whereas copper is only 0.005 percent and tin is a

206

scant 0.0005 percent. Iron ores have a considerable range of compositions. Common impurities in iron ores include silica, calcium carbonate, phosphorus, manganese (especially in hematite), sulfur, alumina, titanium, and water. Ancient metalsmiths had to have an ore greater than about 60 percent Fe 2 O 3 because the slag itself would have a 2:1 Fe/Si ratio. The best iron ore for smelting is not necessarily the ore with the highest iron content. Good iron ores are the ones low in lime, magnesia, alumina, and silica. The manganese oxide content does not matter, because it will go into the slag. All primitive slags contain at least 50 percent iron plus manganese oxides. Although iron ore deposits are both widespread throughout the world and often of considerable extent, the exacting requirements for smelting, including a temperature not attainable by methods used in early copper, lead, tin, and zinc extraction, led to relatively late iron metallurgy. Although some iron was smelted by about 1200 b.c.e., the technology was not widespread until 800 b.c.e. Iron is the only common metal that will dissolve carbon. Steel is an alloy of iron with small amounts of carbon. Since charcoal (carbon) was used to smelt iron, ancient metallurgists were unwittingly making low-grade steel for about two millennia. Most hematite (Fe 2 O 3 ) can be ground into a red ocher powder. Red ochre was mined during the Paleo-Indian period at a locality in Wyoming, USA.32 Although ochre has been recovered from many Paleo-Indian sites in the western United States, it has been difficult to identify procurement sites. In North America red ocher, either in the form of lumps or ground into pigment, accompanied the dead in their burials. For example, red ochre was used with a Clovis-age burial at Anzick, Montana. Hematite-stained artifacts were found with skeletal materials dating to around 11,000 b.p. In central and eastern North America (the area roughly bounded by Minnesota, Ontario, North Carolina, and Alabama), hematite was used in prehistoric times to make pendants, axes, celts, and edged tools. Celts are the most common and widespread hematite implements in this region. Hematite was recovered from both glacial till and bedrock quarries. Red ocher has

Raw Materials and Resources

been used for decoration at least since Paleolithic times. At Blombos Cave, on the south coast of Africa, ochre was being used by about 100,00 to 70,000 years ago. The pigments for the Upper Paleolithic cave paintings in France contain red ocher. In the Ukraine, at the Mezhirich site, red ochre was used to decorate the skull of a mammoth. AMS radiocarbon measurements indicate this use of red ochre dates to around 14,500– 13,000 b.p. During the Neolithic ochre was used for painted pottery. Limonite is a field-geology term referring to hydrous iron oxides of uncertain identity. Goethite is hydrous iron oxide and a major iron ore. It was named in honor of the poet Johann Wolfgang von Goethe. When it is yellow in color, goethite is called yellow ocher. Goethite is one of the commonest and most widespread of minerals and forms the gossan or ‘‘iron hat’’ that caps oxidized sulfide deposits. Goethite dehydrates to form hematite when heated to about 275° C and the color changes from yellow to red. Therefore, some red ochre may have been originally a yellow goethite. This heat transformation was known at least by Roman times and probably extends farther back into prehistory. Goethite is common in near-surface sediments and soils as a secondary or concretionary material precipitated from circulating ground water. Our lack of archaeological knowledge concerning iron ore mines of the European Iron Age probably stems from the wide occurrence of small, shallow deposits of limonitic and hematitic iron ores. The exploitation of such ores would have had little lasting effect on the landscape. At the Nichoria excavation in southwestern Greece, goethite was found in the shape of small rods in Bronze Age contexts. This created quite a stir, because it was conjectured that these might have been pre–Iron Age oxidized iron nails. However, Rapp demonstrated that the rods had a radial cross-sectional structure similar to other known goethite nodules rather than the concentric or structureless crosssection that would have been consistent with a rusted metallic iron nail. Magnetite (Fe 3 O 4 ) is a magnetic iron oxide; it is iron black with a metallic luster and a black streak. A natural magnet, magnetite is known as

lodestone. It is common in small amounts in most igneous rocks. Occasionally there are concentrations large enough to be classed as an ore deposit. Although most meteorites are stony in composition, many are composed of an iron-nickel alloy that is nearly rustproof and easily recognizable as a metal. Meteoric iron is malleable and easily worked. It can be distinguished from smelted iron by its high nickel content, 5–26 percent. Metallic meteorites were picked up and used by humans long before the Iron Age in the Old World and were prized in prehistoric America. Ancient Sumerian texts mentioned meteoric iron, calling it ‘‘fire from heaven.’’ Artifacts of meteoric iron are known from the end of the Third millennium b.c.e. in Egypt and the Near East.33 The Aztecs made knives from iron meteorites. The most common and widespread of all sulfide minerals, pyrite (FeS 2 ) frequently occurs as crystals and has long been called ‘‘fools’ gold’’ for its color, although it is not as golden as chalcopyrite. An iron sulfide, it is easily distinguished from gold by its brittleness and hardness (6–6.5 on the Mohs scale) and from chalcopyrite by its paler color and greater hardness. It is ubiquitous in copper deposits and contributes to the formation of gossan, which marks most sulfide deposits. Pyrite has been found in many archaeological contexts. The Mayans used it to make mirrors, while the Aztecs used it for inlays in mosaics and eyes in statues. The Arctic Inuit and others employed it as a firestone because of its ability to give off sparks when struck. It was also used to make amulets.

Rocks Just as the silica minerals have always been preferred for tools and weapons, the common durable rocks were the material of monuments, statuary, burial chambers, and buildings. There are three groups of rocks, based on how they originate and form. The major classifications of rocks are igneous (from the Latin for fire): rocks formed from molten magma; sedimentary: rocks formed from the consolidation of deposited clastic particles or by precipitation from solution; and metamorphic (from the Greek for undergoing a change

207

Raw Materials and Resources

Table 7.1 Igneous Rocks Lighter Colored

Coarse grained Fine grained

High SiO 2 (silica) granite, syenite rhyolite, felsite

Darker Colored Intermediate SiO 2 diorite andesite

Low SiO 2 (silica) gabbro basalt

Table 7.2 Major Sedimentary Rocks Shale

The most common sedimentary rock. Shales are composed of clay and silt and often exhibit a finely laminated structure that resembles bedding.

Sandstone

Rounded or angular particles of sand size cemented together by silica, iron oxide, or calcium carbonate. Sandstones are often quartz and may be well cemented and therefore cohesive or poorly cemented and therefore friable.

Conglomerate

Coarse-grained clastic rocks composed of large rounded pebbles, cobbles or boulders set in a fine-grained matrix of sand or silt cemented by silica, iron oxide, calcium carbonate, or hardened clay.

Limestone

Calcium carbonate in the form of calcite, formed by either detrital, biologic, or chemical processes. Many are highly fossiliferous and represent ancient shell banks or coral reefs. Chalk and travertine are also limestones.

of form): rocks formed by a major alteration of preexisting rocks owing to high temperature and pressure. There are hundreds of names for variations in igneous rocks, but the names in table 7.1, with the addition of obsidian, should suffice for most archaeological work. The classification of igneous rocks is based on two main traits: the relative proportions of silica (SiO 2 ), and the size of the crystals. The difference between granite and syenite lies in the percentage of quartz, with syenite having little or none. Rhyolite is the fine-grained mineralogic equivalent of granite. Felsite is a general term for any light-colored, fine-grained igneous rock, with or without phenocrysts, that is composed chiefly of quartz and feldspar. Basalt is sometimes called trap or trap rock. The wide range of igneous rocks used for hammer-stones, axes, and other pounding instruments is too extensive to be discussed here. There are also hundreds of names for the varieties of the common sedimentary rocks (table

208

7.2). Sedimentary rocks are classified on the basis of particle size and chemical composition. For example, sandstone that has feldspar as a major constituent of the clastic particles is called a greywacke. The pebbles or boulders in conglomerate are often very hard chert or quartz; the ability of prehistoric societies to dress blocks of conglomerate cutting neatly across tough chert boulders (for example, at the Lion Gate at Mycenae in southwestern Greece) is extraordinary. On the other end of the hardness scale, shale was used for beads because it was easy to shape and drill. And there are many names in the geologic literature to account for mineralogic or textural variation within the major metamorphic rock types (table 7.3). It should be noted that not all quartzites are of metamorphic origin. A quartz sandstone completely cemented with silica can fracture through (rather than around) the grains, thus emulating metamorphic quartzites. Individual rock types of special interest in geoarchaeology include:

Raw Materials and Resources

Table 7.3 Metamorphic Rocks Transformation of sedimentary rocks through high temperature (®) and high pressure (®) Limestone ® Marble Sandstone ® Quartzite Shale ® Slate ® Phyllite ® Schist ® Gneiss

Andesite A widespread fine-grained volcanic rock, named after the Andes Mountains, it is hard and cohesive. Vesicular andesite has been used extensively throughout the world for grinding stones for such grains as wheat and corn. Basalt Fine-grained and often extremely tough and cohesive, basalt has found widespread favor for lithic tools, querns, and, in Egypt, statues. The earliest stone vessels made in Egypt were Neolithic basalt vases. Basalt is widely distributed in Egypt and was used as early as the Old Kingdom (2650–2134 b.c.e.) as a material for pavements in the necropolis stretching from Giza to Saqqara. Its source apparently was the Fayum, where one can still see the ancient quarry. During Pharaonic times, basalt was used in statues and sarcophagi. Felsite Felsite is a general term for any finegrained, light-colored igneous rock composed chiefly of quartz and feldspar. Felsite can be very hard and cohesive. Like chert, it breaks with a sharp conchoidal fracture, which makes it valuable as a lithic material. Almost the same as felsite, rhyolite is the fine-grained mineralogic equivalent of granite. Thus it contains quartz as an essential constituent. Granite and Diorite Fairly abundant, often mechanically tough and free of cracks, aesthetically pleasing, and capable of taking a high polish, these igneous rocks have been widely used in the construction of large monuments. As early as Predynastic times in Egypt they were also used for bowls and vases, and later for statues, obelisks, and stelae. The Pharaonic Egyptians carved single large obelisks from Aswan granite and used diorite for large statues. Diorite has been used

in Egypt since Neolithic times for axes, palettes, and mace heads. Not all coarse-grained igneous rocks used in Egyptian statuary that are called diorite by archaeologists are actually diorites— granodiorite, granite, and other rock types have been mistakenly identified as diorite. Granite and diorite vary considerably in composition, texture, color, and durability, however. Scoria In regions where there has been extensive metallurgical activity, scoria has been mistaken by archaeologists for slag. Scoria is the name applied to very dark, highly vesicular, sometimes glassy rock of basaltic composition. Coal clinker has the same appearance and is sometimes also called scoria. In the eastern Mediterranean region a blackish residue from the melting of kiln walls under reducing conditions has also been mistaken for slag. This latter material, however, has very low density. Obsidian The name obsidian goes back as far as Pliny, who described obsidian from Ethiopia. It is a volcanic glass, usually black but sometimes of other colors or even variegated. Highsilica volcanic rocks are typically fine-grained or glassy. The molten material from which these rocks formed is so viscous that crystal growth is impeded and noncrystallized rocks often form during rapid cooling. Obsidians contain less than 1 percent H 2 O by weight because magma at high temperature that is extruded onto the earth’s surface, or intruded at very shallow depth, cannot retain much water in solution. However, obsidian can become hydrated (up to nearly 10 percent) by later absorption of groundwater (see Chapter 8 for a discussion of obsidian provenance studies and Chapter 5 for obsidian hydration dating). As a glass, obsidian breaks with a conchoidal fracture. Because it is easily worked into sharp projectile points and other implements, it has had wide use since prehistoric times. Obsidian was in use as far back as Paleolithic times. It was found in level C of Shanidar Cave in Iraq and dated to approximately 30,000 b.p. Obsidian has been recovered from almost every Neolithic site in the eastern Mediterranean area. It was equally important in New World contexts. Prehistoric miners in central America pursued high-quality obsidian at depth, developing under-

209

Raw Materials and Resources

ground mines. The Aztecs made extensive use of obsidian for projectile points, knives, razors, swords, mirrors, and ornaments. There are numerous high-grade obsidian deposits in western North America, such as Obsidian Cliff in Yellowstone National Park and in the volcanic regions of Central America.34 The Obsidian Cliff source is one of the most famous in North America and was widely traded for 10,000 years. Fifty-nine separate quarries have been identified.35 Pumice and Ash Pumice is a light-colored, vesicular, glassy, pyroclastic rock that is commonly composed of rhyolite. It is often porous enough to float and has been widely used as an abrasive. Pyroclastic rocks are volcanic rocks that form as particle deposition from explosive eruptions. When particles are smaller than 2 mm in size, they are called ash. Pumice from archaeological excavations can now almost routinely be traced to the volcano of origin and often dated to a particular eruption (see Chapter 5, K-Ar and Ar-Ar dating). Quartzite Composed chiefly of quartz sandstone with silica cement, this rock fractures conchoidally, which allows it to be worked into sharp tools, just as chert and obsidian are. Quartzite was used in Pharaonic Egypt for statues and sarcophagi. Nubian sandstone, a quartzite, was used by people who made and used Paleolithic (Acheulian and Mousterian) artifacts in the Sahara Desert.36 ‘‘Hixton Silicified Sandstone,’’ a quartzite from western Wisconsin was widely used in tool manufacture in the western Great Lakes region.37 Marble Marble was the preferred statuary and monumental stone of the classical world. In Egypt it was also used for vases. Marble can be found in thick deposits of wide areal extent that are relatively free of cracks and easy to quarry. It takes a high polish. Its chief drawback is its high susceptibility to disintegration under the action of acid rain (modern pollution aside, rain is acidic because CO 2 dissolves in atmospheric water to create a never-ending supply of carbonic acid in rain). Marble is plentiful in western Anatolia, Italy, Greece, and elsewhere in the Mediterranean area, where it was widely used. Serpentinite A rock composed chiefly of the green minerals of the serpentine group (hard210

ness 2.5–3.5), it occurs widely as the alteration product of mafic igneous rocks. It has been used since antiquity in the Old World for stone bowls, vases, carved figures, and occasionally molds. It was popular throughout North America because it was easily worked. Siliceous Shale/Slate As with quartzite, this low-grade metamorphic rock fractures conchoidally. Because it is composed chiefly of silica, it has a hardness approaching 7 on the Mohs scale. It was widely used in North America, especially in northern Minnesota, for projectile points and other sharp tools. Steatite (Soapstone) A fine-grained, compact rock consisting chiefly of the mineral talc (hardness 1) but usually containing many other constituents, it has been favored since antiquity for ornaments and molds because of its softness and cohesion. In North and Central America, prehistoric peoples used steatite for pots, bowls, and pipes. The Inuit and others used steatite for lamps. Steatite outcrops are common in many locations throughout the Appalachian Mountains. Catlinite A red, siliceous, indurated clay that is often called ‘‘pipestone’’ because of its extensive use in North America since at least the late sixteenth century b.c.e. for making tobacco pipes. Catlinite is named after the well-known George Catlin (1796–1872), painter of Native Americans. Although the best-known catlinite quarry is in southwest Minnesota, there are also deposits in Wisconsin, Ohio, and Arizona. The trade in catlinite was countrywide and extended into Canada.

Shells Shells have been exploited throughout history for food, dye, tools, ornaments, trade goods, talismans, money, and ritual objects. They have proven to be important indicators in both the natural geologic record and in the archaeological record. Shell is composed mainly of calcium carbonate with small amounts of calcium phosphate, magnesium carbonate, and silica. Shell was often used as either currency or trade goods. One of the most widely prized shell currencies was the cowrie, of which the most exten-

Raw Materials and Resources

sively circulated species was the Cypraea moneta or ‘‘money cowrie.’’ Money cowrie is found throughout the Indo-Pacific; most of it originated in the Maldive Islands situated in the Indian Ocean, south-southwest of India. Cowries had symbolic and ritual significance in many human societies, which must have contributed to their acceptance as a form of currency. Cowries were also used for human decoration in prehistoric Europe and North America. Cowries are small gastropods of a large number of species, probably two hundred. The word cowrie is from the Hindi and Urdu languages. Cowrie species vary in size from less than 1 cm to more than 4 cm. Most cowries used as money were about 1.5 cm. Used individually as money and strung as beads in a necklace, cowrie shells were quite durable. The earliest references to the use of cowrie shells come from China where they were in circulation as early as the second millennium b.c.e. during the Shang dynasty.38 A number of indigenous peoples of North America had currency or barter systems that included the exchange of shells. The Nootka of British Columbia collected Antalis pretiosum both as food and for its shells, which were traded with neighboring tribes to the south. The Yurok and Tolowa peoples of northern California used strings of tusk shells as money. The value of the string was based on the length of the string, and the number and size of the shells. The Pomo people of the central Californian coast used strings of beads made from the shell of the Pismo clam (Tivela stultorum) as currency. A higher value was placed on cylindrical beads made from the thickest part of the shell. In northeastern United States, beaded belts made from the quahog (Mercenaria mercenaria) were used for ceremonial exchanges by prehistoric peoples. Purple beads were more highly valued than white since only a small part of the quahog shell is purple. These objects were used as currency under Dutch and British colonialism by historic times. Thus, the Dutch were able to ‘‘purchase’’ the island of Manhattan for a nominal amount of trade goods, which may have included shell beads. Along the Atlantic seacoast in North America shells were made into wampum beads that in the Colonial Period were the Native American’s form

Figure 7.1 Murex Shell

of money. Crinoid stems were drilled and worn as beads. Water-tumbled stones of glacial till containing shell fossils were used for such everyday objects as hearth and boiling stones, hammers, anvils, and mullers. Some projectile points from the eastern United States have shell fossils located centrally in the artifact, suggesting its intentional placement as an ornament. Shell fossils and concretions, some stained with hematite, have been found associated with burials and grave goods. In the eastern Mediterranean region, shells from certain species of the marine gastropod genus Murex yielded the royal purple dye of the ancients, especially favored by the Romans. (A substance secreted by murex turns purple when put into heated seawater.) The use of murex purple dye goes back to 1500 b.c.e. at Ugarit on the Levant coast. Murex shells were found associated with Middle Minoan pottery on the small island of Korephronisi just off Crete, and Cretan use of the purple dye was mentioned by Herodotus. Murex may have been bred in tanks along the ancient south coast of Turkey. Pliny gave fairly good descriptions of the gastropods and their dye. Murex shells are highly ornamented and easily identifiable (fig. 7.1).

Clays The primary raw materials for archaeological ceramics were local clay-rich sediments and soils that were used for the paste and coarse sedimentary particles that were used for the temper. Clay 211

Raw Materials and Resources

deposits are of two general types: (1) primary deposits formed in situ by the weathering of bedrock like granite or shale, and (2) secondary deposits formed by river (fluvial) or lake (lacustrine) deposition. The secondary deposits are often referred to as transported clays. In soils, clay minerals form as part of the natural chemical weathering of the parent bedrock. Clay minerals are low-temperature hydrous minerals that are stable in the earth-surface environment. They form in soils from the breakdown of minerals that were created in high-temperature anhydrous environments. These primary minerals are not stable at the earth’s surface. Plant nutrients become available during this chemical weathering. The supply of potassium (K) in a soil correlates well with the rate of clay formation.39 The word clay has two meanings: it is both a particle size (smaller than two microns) and the name of a group of silicate minerals that have a sheetlike structure. Because of their small size and sheet structure, clays become plastic when mixed with water. This plasticity allows the mixture to be shaped and to retain the new shape. Whether a clayey raw material will make good pottery—one of the primary uses of clay—depends on which clay mineral predominates, the shape and size distribution of the nonclay minerals, the organic content, the exchangeable ions present, and the size distribution in the whole mass. Good pottery clays also contain fine-grained quartz, which provides the refractory backbone during firing. Precise identification of clay minerals requires X-ray diffraction methods. The various clay minerals are not equally stable in all surface environments. Often many other sheet silicates are found mixed with clay minerals in deposits, but these other silicates all promote poor plasticity. The addition of fine-grained organic material and many varieties of plastic and aplastic tempers can improve plasticity, strength, and firing properties of clay mixtures for ceramics. Clay is a very adaptable material. The earliest use of clay appears to be in the Upper Paleolithic when people drew designs in wet clay on cave walls. Next came the making of clay figurines, then clay as a building material. By the Neolithic, in many parts of the Old World, artisans discov-

212

ered that fire-hardened clay made durable vessels. Pottery making was an independent invention in the North America during the Woodland Period (see fig. 5.5). Crude pottery is relatively simple to make. Many clayey soils and sediments can be fired to form simple, thick-walled vessels that can withstand moderate thermal and physical stress. Fired clay objects are known from the Upper Paleolithic of central Europe. The earliest known ceramics from North America come from Florida and date to about 1700 b.c.e. In Chapter 8 we cover ceramics in greater detail. The major clay minerals and their important properties for pottery making are: Kaolinite, Al 2 Si 2 O 5 (OH) 4 Kaolinite is the most refractory of the clays and has excellent firing properties. Not only does it have the best high-temperature stability, it can also be heated rapidly, and it has low shrinkage. Because of its restricted chemical composition, kaolinite maintains its white color. It is the clay of fine china. Its particle size and shape—it forms as small hexagonal plates—give it good plasticity. Kaolinite is the only clay that works well alone in pottery manufacture. It is a common and widespread product of the weathering of feldspars in igneous and metamorphic rocks. It can also be found as a secondary clay. Halloysite, Al 2 Si 2 O 5 (OH) 4 Although similar to kaolinite in composition, halloysite has thin platelets that curl up, leading to poor workability (for example, it can crack during forming and drying). Halloysite must be heated slowly to prevent fracture during firing. It is a common product of the alteration of feldspars and often occurs with kaolinite. Montmorillonite/Smectite, (Al,Mg) 8 (Si 4 O 10 ) 4 (OH) 8·2(H 2 O) (composition variable) The smectites are the expanding (swelling) clays that can absorb large amounts of water between the sheets in their structure. Their ceramic properties include high plasticity and moderate refractoriness but also high shrinkage. These clays are widespread alteration products of the weathering of volcanic ash and of many high magnesium igneous and metamorphic rocks. They also form as soil clays.

Raw Materials and Resources

Illite (OH)K 4 Al 4 (Si,Al) 8 O 20 (composition variable, usually containing iron and magnesium) Illite has good plasticity, variable shrinkage, and low refractoriness. It is good for slips. Illites are the dominant clay minerals in shales and mudstones. They also form commonly in soils from the alteration of micas and other clay minerals and from colloidal silica. Illite is the only clay mineral containing significant potassium. Other sheet silicates are found mixed with these clay minerals in many deposits, but they all reduce the plasticity. The addition of fine-grained organic material and many varieties of plastic and aplastic tempers can improve plasticity, strength, and firing properties. The nature of the clay resources can be critical for pottery making. A case in point can be seen in Polynesia. In Fiji and many islands to the west, soils are formed largely on alluvium derived from andesitic rocks. The rocks weather to form kaolinite or smectite clays that, with the addition of sand temper, can be used to make good pottery. In Tonga the clays are halloysite and suitable temper sands are not available. In Tahiti the soils do not contain suitable material for pottery manufacture. Thus, it seems that as early Polynesians migrated across the Pacific, they abandoned their earlier pottery techniques because of the nature of the available raw materials.40 To reduce shrinkage and/or improve workability, temper is added to the clayey material before it is formed. Coarse sediments for temper are normally available in nearby streams. Fossil shells also make good temper because calcium carbonate has the same thermal expansion as the average pottery clay. The same is true for marble (also composed of calcium carbonate). Chips of broken pottery, called grog, were also used as temper. Glaze is a thin layer of material applied to ceramics before firing that becomes a glassy coating at high temperatures. Low-melting materials are necessary for glazes. Sodium (Na) and lead (Pb) are the most common elements used to lower the melting point of the silicates in ceramic glazes. Lead glazes can cause lead toxicity if used for drinking or eating ware. Soluble salts or saline water can be the source of sodium. High-fired glazes usually contain feldspars, calcite, dolomite,

or wood ash. Low-fired glazes are usually alkaline (containing sodium or potassium).

Building Materials The nature of the building materials employed by a society depends primarily on the kind of material available. It was no accident that the great tells of the Middle East resulted from the disintegration of mud-brick buildings. For hundreds of kilometers around, there were neither hardrock outcrops nor trees available for building. At Çatal Höyük in Anatolia, walls and even built-in furniture were made of clay earth by 6800 b.c.e. In more northern regions, also lacking sufficient timber and where mud-brick walls would not survive, stone construction was a necessity. Except for timber, almost all building materials reflect the exploitation of minerals and lithic resources. During the Paleolithic, humans turned to rocks, first to help support tents and huts and to improve a cave or rock shelter, later to construct dwellings. For example, concentrations of rocks at Olorgesailie, Kenya, and Terra Amata, France, may be the remains of Lower Paleolithic windbreaks or huts. At Bilzingsleben, Germany, there are circular concentrations of rocks that might be the remains of structures dating to about 350,000 years ago. Piles of stones at the site of Cueva Merín, Spain, and fragments of limestone at La Ferrassie may be the remains of structures associated with the Middle Paleolithic (Mousterian). Patterned arrangements of stone blocks are more common in Upper Paleolithic sites. For instance, the Châtelperronian site of Grotte du Renne, France, consists of an arrangement of archaeological features including rocks that seem to indicate the presence of a hut structure dating to about 33,000 years ago. In the Near East, rock walls for houses and towns go back more than 7,000 years. In areas where rock masses were scarce, earth plaster and mud bricks were used. Ancient Egypt, Greece, and Rome were responsible for great advances in the use of quarried and dressed stone for building. By 3000 b.c.e., the Egyptians had learned to shape their comparatively soft Tertiary limestones into rectangular

213

Raw Materials and Resources

blocks. Before 2000 b.c.e., they had learned how shape the more difficult granite and diorite. By the time of the Third Dynasty in Egypt (2649– 2575 b.c.e.) there was a marked increase in the use of stone for buildings. Limestone was the primary building material, but large blocks of granite were used in an unfinished pyramid located between Giza and Abusir and in the step pyramid of Djoser at Saqqara. By the Eighteenth Dynasty (1550–1307 b.c.e.), limestone had largely given way to the more durable sandstone and granite. The great sandstone quarries along the Nile about 60 km north of Aswan bear inscriptions dating from the Eighteenth Dynasty down to Greek and Roman times. The Egyptians removed the surface layers of granite outcrops by burning papyrus on them and then drenching them with cold water, which caused the granite to spall and disintegrate. If no natural crack or joint could be exploited in quarrying the granite, a trench was made around the desired block by pounding it with balls of dolerite, a fine-grained basaltic rock. Wooden wedges were inserted in cracks or trenches, driven tight when dry, then wetted. The Greeks and the Romans developed masonry building in shaped (ashlar) blocks to a fine art. During the Late Bronze Age the Mycenaeans in Greece constructed ‘‘Cyclopean’’ walls by fitting angular blocks neatly together. Sometimes mortar was used in masonry: the Egyptians favored gypsum while the Greeks and Romans used burnt lime (calcium oxide) and clay. Lime plaster was formed by heating limestone (calcium carbonate) at temperatures higher than 900°C to drive off the carbon dioxide. This left calcium oxide. Calcium oxide reacts rapidly with water to form calcium hydroxide. When exposed to the atmosphere, the calcium hydroxide takes up carbon dioxide again, becoming a stable calcium carbonate—hard and cement-like. The Greeks and the Romans also discovered that certain volcanic ash produced a hydraulic cement when mixed with slaked lime and water. The Greeks used the volcanic tuffs from Thera (Santorini), while the Romans discovered around 100 b.c.e. that the volcanic ash from the village of Pozzuoli on the slopes of Mount Vesuvius was suited for this purpose. Not all volcanic

214

ash has this property; the Roman recipes call the cements made with volcanic ash Pozzolanic cements. These cements can set under water; they were used as early as 144 b.c.e. to line water channels. In the early nineteenth century, portland cement was invented in Britain (named after the Isle of Portland in Dorsetshire). The basic materials of portland cement are clay and limestone. Some iron is needed, but this is usually supplied by iron minerals in the clay. The most stringent requirement of the limestone is that it not contain more than about 5 percent magnesium carbonate. Portland cement is hydraulic—water is involved in the basic chemical reactions of manufacture. The first portland cement produced in the United States was in the Lehigh Valley of Pennsylvania in 1875.

Building Stone A large number of distinct hard rock types (including limestone, sandstone, granite, and gneiss) make good building stone. The necessary characteristics are: (1) structural strength (ability to carry a load without failure), (2) durability (resistance to weathering), (3) ease of quarrying and dressing, and (4) availability (the energy cost of transportation). By the time very hard rocks like granite, diorite, quartzite, and conglomerate came into use for building stone, the use of abrasives must have become commonplace. The Aztecs worked granite with copper implements and used quartz sand as an abrasive. Emery, a grayish-black mixture of magnetite and corundum (which is second only to diamond in hardness), and pumice must have been used as abrasives from early on in those areas where they were available. Emery is abundant in the Near East and some of the Aegean Islands. On the low end of the structural-strength scale are clays and muds. Yet these were important building materials (with only minor processing) in antiquity. As a geologic term, mud is a mixture of silt and clay-sized particles (see Chapter 2). (The word in common usage describes essentially the same material.) Many natural muds also contain some sand-sized grains. As noted earlier, clay has more than one meaning, but the thing to keep

Raw Materials and Resources

in mind is that most clay-sized material is composed of clay minerals. The optimal sand-silt-clay ratios vary for different types of structures. For rammed earth walls, there should be about 50–75 percent sand in the mixture to prevent excessive shrinking of the plastic components. The size range of the sand and coarse particles is of little importance. In rammed earth walls only a small amount of clay is desirable: more than 30 percent clay results in rapid erosion. The pre-Spanish inhabitants of the American Southwest used the pisé technique to build layered walls in three-foot (vertical) units of adobe-type mud, rather than making mud bricks. To make satisfactory adobe bricks, a lithology lying outside the range used for rammed earth walls is needed. Adobe bricks must have a higher percentage of clay and contain straw for binder. Good sun-dried adobe bricks can be made from earth containing 9–28 percent clay. (The Uniform Building Code in the United States requires that the clay content of adobe bricks be greater than 25 percent and less than 45 percent.) Too much clay causes the brick to develop cracks as it dries; too little makes the brick too weak, so that it crumbles easily. Individual clay minerals have different bonding properties, and the lime content as well as the organic content will affect the physical properties. Sand-silt-clay ratios cannot tell the whole story. Compositions suitable for adobe or sun-dried mud brick can be attained by mixing a range of materials, but it is likely that in prehistoric times raw materials for mud brick were primarily clay and organic binders such as straw. In present-day mud-brick making in Egypt and Turkey, bricks are carved out of the soil directly, without going through a molding step. Enough plant material may exist in the soil there to take the place of added binder.

Burnt Brick For thousands of years stone, clay, or clay plus timber were the dominant construction materials. The use of timber is recorded in archaeology by the abundance of debris in ‘‘destruction levels’’ that result when towns and villages are burnt down. London was destroyed by the Great Fire of

a.d. 1666 because it was built chiefly of timber and mud. It was rebuilt with brick produced from the abundant high-quality clays available in southeast Britain. The art of baking clay to produce bricks dates from the second millennium b.c.e. in Mesopotamia, India, and Egypt. It was the Romans, however, who developed the technology of making burnt brick of high quality by carefully choosing and fine-grinding raw materials, then subjecting them to controlled firing. Since the twelfth century brick making has become ubiquitous; clayey sediments and weathering products are widely distributed. Although all clay minerals begin to recrystallize above 1,200°C, the mineralogy of the clay is almost as critical in brick making as in pottery manufacture. The compositions of brick raw materials can vary widely throughout the world. In Georgia in the southeast United States, brick is made from pure kaolinite, some feldspar, fire clay, and bentonite (a rock composed chiefly of volcanic ash altered to montmorillonite). In Great Britain red marls, deltaic muds, and illitic sediments of Jurassic age (fig. 5.2) have all provided raw materials for brick. Throughout the Old World the most common raw material used for bricks is recent alluvium from the great floodplains of the Yellow, Ganges, Nile, Mekong, and Indus rivers. These alluvial deposits, however, are often high in silt and low in clay, leading to a weak and porous brick—best suited for single-story houses.

Mortar Ancient mortar, before the advent of cement, usually consisted of clay, gypsum, or lime. Both gypsum and lime mortar required preparation in high-temperature kilns. The ancient Greeks and Romans used lime, whereas Pharaonic Egyptians used gypsum because in a land where wood was scarce, the firing of gypsum plaster was accomplished at lower, fuel-saving temperatures. Clay mortar was used primarily with mud bricks. Clay, lime, and gypsum all served a dual purpose as mortars and as plaster. The ancient Egyptians mixed clay and gypsum and also clay and fine limestone for wall plaster. Since at least the third millennium b.c.e., lime has been used for mortar. The most basic compo-

215

Raw Materials and Resources

nent of most mortars is lime, which can be procured by burning limestone, marble, shell, coral, or bones. Limestone (primarily CaCO 3 ) is the most common source, and both calcitic and dolomitic limestones can be used. Lime is simple and cheap to prepare. Limestone is heated in a kiln to over 900°C; the CO 2 is driven off and CaO remains. The resulting product (calcium oxide) is traditionally referred to as ‘‘quicklime’’ or ‘‘unslaked lime.’’ This slakes with water and when mixed with sand, makes mortar or plaster. Lime is the basic binder in most mortars. However, lime was not always available, so mud was also a commonly used binder. Some societies discovered that the mud mortar was easier to work with and stronger if they mixed it with lime. This required much smaller amounts of lime. Thus, combining mud and lime was an alternate solution if there were no kilns for burning calcium carbonate in large quantities, or if limestone was only available through trade. Certain types of soils also contain naturally occurring calcium carbonate, so analyses of these ancient materials must distinguish unaltered mud mortars from those that have been modified with additives such as lime, straw, ash, blood, or other locally available organic.

Figure 7.2 Concretions

Figure 7.3 Chert Nodules These nodules form in carbonate rocks that weather away, leaving the nodules lying loose on the surface. This nodule was collected in the Valley of Kings near Luxor, Egypt, by Rapp. The scale is in centimeters.

Other Materials When considering the geologic raw materials that may be found in an archaeological context, it also pays to become acquainted with objects that can resemble natural materials in shape or some other aspect but that have a different origin. Chief among these are concretions and dendrites. Concretions are hard, compact segregations of mineral matter found in sedimentary rocks, particularly shales and sandstones, as well as in soils. They are formed by precipitation from aqueous solution, growing outward from a nucleus, and usually are of a composition somewhat different from the host rock. Concretions represent a concentration of a cementing material, such as iron oxide, calcite, silica, or gypsum. Most concretions are spheroidal, ellipsoidal, or discoidal, although many attain odd or fantastic shapes, sometimes mimicking turtle shells, bones, leaves, or

216

other fossil material. These concretions are often septarian. Figure 7.2 shows one concretionary shape. Many concretions have an internal structure similar to that of concentric shells. Chert nodules have similar shapes (fig. 7.3). In soils, concretions cemented by iron oxide are found in better-drained soils but require some moisture to form, whereas calcareous concretions form in dry soils, particularly under alkaline conditions. Concretions range in size from a few millimeters to up to 3 m. Concretions have been found in many ancient refuse pits and as grave goods in North American prehistoric archaeological sites. Dendrites are near-surface deposits of manganese oxide that have crystallized in a branching pattern along fractures in rocks. They are often mistaken for fossils. Figure 7.4 shows a common dendrite.

Raw Materials and Resources

Figure 7.4 Dendrites These near-surface deposits of manganese oxide that have crystallized along fractures in the rocks are frequently mistaken for fossils. (Drawing by Elaine Nissen)

Pigments When discussing any archaeological artifact, color is an important characteristic. A pigment is a finely divided insoluble material that is suspended in a medium and acts as a coloring agent. Generally speaking, pigments are mechanically mixed with the medium and maintain their physical properties in the suspension. Pigments can be organic or inorganic, naturally occurring or artificial. Archaeological evidence suggests that the earliest pigments by humans were ochres, clays, and charcoals from burned wood and bone. The range of colors included reds, browns, yellows, and blacks. Probably the first mineral pigment used was red ochre—in places as far apart as Africa, Europe, Australia, and Japan. For a more complete discussion of ochre and ochre pigments see Rapp.41 L. S. B. Leakey stated that he found lumps of ochre at Olduvai in a context dating from more than 500,000 years ago while fragments are common in Upper Paleolithic context, dating to after 30,000 years ago.42 Pigments must have become important as exchange items almost from the beginning, as exclusive access to mineral deposits could have contributed to a group’s economy. As humans organized into extended social units,

minerals and the knowledge required to utilize them became part of a complex web of trade or exchange. The challenge for the geoarchaeologist is successful identification and characterization of iron oxide pigment, particularly where it has survived only in trace amounts. Evidence of iron oxide pigments on many prehistoric artifacts has been obliterated by weathering, careless handling, or even cleaning.43 It is therefore vital to handle objects sparingly and to plan decisions regarding artifact processing and analysis before excavation and recovery. Iron oxides are the most universally employed pigments. They occur naturally in some form in nearly every region of the earth. Iron oxide pigments range in color from dull yellow to reds, purples, and browns. These pigments were used by many societies to decorate ceramics as early as the Neolithic period. The use of manganese oxide pigments also extends back into prehistory. They are found in early cave paintings, including those at Lascaux Cave in France. Manganese oxides are a significant component in the group of pigments known as umbers. Pyrolusite and its dimorph, ramsdellite (MnO 2 ), are the principal manganese oxides. These black minerals have been used in pottery decoration since prehistory. Eventually their application was extended to glazes and glass. Some of the earliest natural pigments included various clays. A wide range of earth tones is available in clays, depending on chemical composition and impurities. Clays were used in paintings, cosmetics, and pottery ‘‘slip’’ glazes. Clays were given numerous traditional and trade designations such as white and red bole, china clay, and pipe clay (all of which are kaolinite). A pink pigment made from heated kaolinite was popular during the Classical Period of Greece. Nearly all the rocks and minerals that hold their color when ground to a powder (most silicates turn white) have been used as pigments. There are too many of these for detailed consideration, but among the most common are hematite (red ocher), goethite (yellow ocher), malachite, azurite, selenite (gypsum), cinnabar, varieties of coal, calcite, huntite (a rare white calcium-magnesium carbonate), magnesite (a white magnesium car-

217

Raw Materials and Resources

bonate MgCO 3 ), and manganese oxides and graphite. Maya Blue is a bluish-green pigment that has been found associated with Mayan artifacts in Central America. Its use appears to have extended into the Colonial Period. Maya Blue is a mixture of minerals and indigo—a blue dye obtained from various plants. The mineral constituents are palygorskite (formerly called attapulgite) and sepiolite (also called meerschaum). The pigment is notable not only for its association with specific archaeological contexts, but also for its unique physical properties. Maya Blue is known to have survived the extremely harsh environmental conditions of the rain forest. It is also remarkably resistant to alkalis and all but hot concentrated acids.44 This pigment was unknown in other parts of the ancient world. It continued to be used in Central America in mural paintings into the nineteenth century.45

Abrasives Since ancient times, many rocks and minerals have been used as abrasives, including emery, pumice, diatomite, feldspar, quartz minerals, chalk, and various metal oxides. Abrasives must be both hard and cohesive—that is, not subject to easy fracturing. Throughout the world, quartz (hardness of 7) in one form or another undoubtedly has been the most common abrasive. It was available nearly everywhere either as quartz sand or crushed quartz. One of the first major uses of abrasive grit was in the construction of the pyramids of Egypt where the surfaces of stone blocks were rubbed smooth for a remarkably good fit. The concept of ground stone objects requires an abrasive material used to shape an object.

Rock and Mineral Recovery Geologists use the term outcrop for rock bodies that jut up at the surface of the earth. It seems a safe assumption that most rock and mineral products used by prehistoric groups were gathered from outcrops. When a particularly high-quality stone or ore deposit was discovered, prehistoric humans developed mining and quarrying tech-

218

niques to maximize its exploitation. The miners often made use of natural vertical joints in the rock in quarrying. At Rudna Glava in Serbia people of the Vinča culture mined copper 4,500 years ago from 15 m vertical shafts. The mining of copper ores and minerals in the Egyptian Sinai date to the Old Kingdom. Underground flint mines that operated in England more than 5,000 years ago during the Neolithic had shafts of more than 4 m deep. In prospecting for ancient mines the geoarchaeologist should look for evidence of waste dumps. The surrounding (waste) rock in metal mining is called country rock or gangue. This material was routinely discarded at the mining site so the miners would not have to transport it to smelting or other metal-processing sites. A troubling problem in Old World prehistoric metallurgy has been the dearth of copper slag dumps from the Bronze Age. Copper smelting generates large quantities of slag—a waste product. Scientists speculate that later societies resmelted or comminuted (pulverized) some of the earlier slags that retained recoverable copper. Or perhaps archaeological surveys have been insufficiently intensive to locate these mounds. Throughout human prehistory people have prospected stream gravels for lithic raw materials, recognizing that the boulders that could survive the rough-and-tumble ride down water courses would not shatter easily. In most parts of the world these boulders were composed largely of chert and related quartzose rocks and minerals. Most of the debris from weathering is moved away from its point of origin by running water (see Chapter 2). What size, shape, and density particles can be transported depends on the velocity of the water. Larger and denser particles tend to lag behind and accumulate in low spots in the channel, where the water velocity is lower. When there is a sharp change in water velocity— for example, where a river emerges from a mountainous area onto a plain—the larger and denser particles are deposited immediately. These concentrations of heavier particles are called placers. The earliest gold and cassiterite deposits used by humans were undoubtedly placers. The Califor-

Raw Materials and Resources

nia Gold Rush of 1849 was based on placer deposits formed by the active erosion of lode deposits and gravels higher up in the Sierra Nevada Mountains. Placer mining has led to extensive excavations of alluvial valleys. Large blocks of building stone were secured through quarrying. Quarries are not easily eroded or removed from the landscape, except when they are replaced by even larger quarries, so many landscapes contain ancient quarries. The quarrying of large blocks of rock for building stone was not possible until the advent of copper and bronze tools for working the softer rocks like limestone. A block of rock was quarried by isolating it on four sides by means of trenches cut into it, and then detaching it from below by means of wooden wedges wetted with water. In Egypt quarrying probably had its beginnings in the cutting away of limestone to make tombs. A Middle Archaic quarry and manufacturing site in Maine, in the northeastern United States, contains low-grade metamorphic rocks, indicating that the local people used a special technique to reduce these rocks by grinding. They were formed into long rods which were used as mortuary objects. Once this use fell out of vogue, the quarry was abandoned.46 The most necessary lithic material in Precolumbian North America was quartzose material, which was used for projectile points. The continent contains innumerable lithics that are good to excellent for this purpose, and large quarries were operated in what are now Ohio and New York for chert, Pennsylvania and New Jersey for jasper and rhyolite, New York for quartzite, and Minnesota for a slightly metamorphosed high-silica siltstone. Artifacts recovered from nearly 12,000 years of occupation of the Cebolleta Mesa region in New Mexico in the American Southwest illustrate the broad range of mineral resources exploited for human needs.47 Most of the materials came from the immediate area or adjacent regions. Use was made of sandstone for masonry blocks and paint mortars; adobe from alluvial soils for building; basalt for tools, mauls, and bowls; clay from clay sediments for pottery; quartzite, chert, chalcedony, obsidian and pitchstone, pipestone, calcite and travertine, limestone, hematite and limonite

for pigments; jet and turquoise for ornaments; halite for food; and felsite and metamorphic rock (other than quartzite) for implements. Also recovered were feldspar, quartz crystals, malachite, azurite, galena, and pyrite. An extensive treatise on the mining of gems and ornamental stones used by people living in America before European contact has been published.48 While mineral products were essential to both native peoples’ and Europeans’ ways of life, the Indians used coal mainly as an ornament and petroleum as a liniment. Eighty-four different geologic materials were used for gems and ornamental stones. Similarly, researchers have located and plotted 142 mine or quarry sites where California Indians exploited rock and mineral resources.49 Evidence for the use of petroleum by prehistoric North American human populations is well documented from archaeological sites. Asphalt was used as a mastic in the Ohio Valley and elsewhere. Many fossil hydrocarbon deposits (oil shale, tar sands, tar springs) were used as fuels in pre-European North America. Fossils, fossiliferous rocks, quartz crystals, galena cleavages, and sheet mica all have been found as grave goods.

Water For geoarchaeologists, water is perhaps the most important and the least understood compound. It is key not only to keeping organisms alive but to climate, weathering, agriculture, and transportation. Habitation sites have always depended on a supply of potable water. The appearance of natural water reveals only part of its total composition and potability. Suspended and dissolved matter imparts color to water: turbidity from abundant suspended clay particles imparts a yellowish color, abundant phytoplankton will generate a deep green, iron oxide a rusty red. The dark colors of swamps and bogs are due to suspended humic material and tannic acid. All surface and near-surface waters contain bacteria in concentrations of up to hundreds of thousands per cm 3. Even unpolluted water contains large quantities of dissolved

219

Figure 7.5 Geologic Formations That Can Result in the Creation of Springs

Raw Materials and Resources

matter and dissolved gases, particularly O 2 and CO 2 , that have been picked up from the atmosphere. Most natural rainwater—unaffected by acid pollution—is acidic because the carbon dioxide it picks up in the atmosphere forms carbonic acid. This phenomenon has major archaeological consequences including the dissolution of buried bone, the corrosion of monuments (especially marble and limestone) and metal artifacts, the decay of organic matter, and the formation of caves and rock shelters (see Chapter 3). Rainfall that does not run off or evaporate seeps into the ground, where it has an important effect on human economy. It provides the medium for chemical and biochemical soil development, chemical and physical transport in soils, and nutrients for plant growth. Water not removed by plants returns to the surface by capillary rise, where it evaporates, returns to rivers or the sea by lateral flow, or, emerging as springs, is retained as groundwater. The top of the zone where the groundwater saturates the earth is called the water table. Because of modern-day drainage of lowlying areas for land use and the heavy pumping of groundwater for agricultural and other uses, contemporary water tables are much lower than in prehistoric or early historic times. The quality of the groundwater is of critical importance for domestic uses. Usually, groundwater that is not near urban areas is free from bacteria because it is naturally filtered as it flows through near-surface strata and soils. Pure water will dissolve only 20 parts per million (ppm) of calcium carbonate and 28 ppm of magnesium car-

bonate. However, as noted above, water entering the groundwater system as rainfall will be acidic. Such water can dissolve hundreds of ppm of sodium, calcium, magnesium, and iron minerals. One of the main criteria for the acceptability of drinking water is its salinity. Below about 400 ppm sodium chloride imparts no taste to water. By about 500 ppm the water begins to taste brackish and by 4,000 ppm, water is undrinkable. The salinity of groundwater depends on the composition of the local rock formations, the depth from which the water is drawn, and the climate. Water also plays a primary role in chemical weathering and alteration by dissolution phenomena and frost action. The accessibility of groundwater is critical if it is to be exploited economically. The structure, porosity, and permeability of the near-surface rock strata in large part govern the accessibility. Most important is how the pervious and impervious beds alternate. If impervious strata lie at the surface, no water will penetrate to the groundwater system. If an impervious layer lies somewhat beneath the surface, it will be a barrier to vertical flow. When pervious and impervious strata lie at an angle to the surface, springs may develop as groundwater moves down because of the pull of gravity. Because both geologic and climatic conditions remain stable for long periods, most springs have a continuous flow. Figure 7.5 illustrates the geologic conditions that lead to the formation of springs. For a comprehensive discussion of springs, see Chapter 9.

221

CHAPTER 8

Sourcing (Provenance)

Just about the most frequently asked question by archaeology students is: Which instrument is best to analyze my stone objects? The answer, unfortunately, is: ‘‘It depends . . .’’ —M. S. Shackley 1998

P

rovenience is a common archaeological term referring to the precise location at which an artifact was recovered (from a survey or excavation). Without provenience data, artifacts have little archaeological value. By provenance, however, geoarchaeologists mean something quite different. The provenance of an artifact is the location, site, or mine that is the origin of the artifact material. In geoarchaeological terms, this means the geographic-geologic source of the raw material from which the artifact was made, that is, a specific geologic deposit—usually a quarry, mine, geologic formation, outcrop, or other coherent and bounded geologic feature. Geoarchaeological provenance studies do not address the question of where the artifact was manufactured but only the source of the raw material. A large number of chemical, physical, and biologic parameters can be used to determine the source of natural materials. Geologists use trace elements, isotopes, diagnostic minerals or assemblages, microfossils, geophysical properties, and many other distinguishing characteristics to determine the source or origin of geologic materials. A few of the geologic techniques that are most useful to archaeologists are examined in this chapter. The underlying assumption for provenance studies is that there is a demonstrable set of physical, chemical, or mineral characteristics in raw-

222

material source deposits that is retained in the final artifact. This assumption can be justified only through empirical work, which requires large data sets of high analytic accuracy.1 The archaeological significance of information concerning the location of the origin of an object, as opposed to where it was found, is considerable. Provenance studies can provide evidence for the reconstruction of the patterns of exchange systems and trade routes, as well as giving the territory, location, and size of resources that may relate to social stratification and organization of crafts and industries. One of the earliest attempts to source archaeological remains by geologic analysis was the effort to trace the origin of the megaliths at Stonehenge in England (fig. 8.1), an undertaking that was begun in the mid-eighteenth century. Although the early phases of construction at Stonehenge date to about 4000–3000 b.c.e. (earlier Neolithic), the stone circle was probably constructed about 2100 b.c.e. Early observers realized that two types of rock had been used in the construction of Stonehenge. The circle is dominated by large ‘‘sarsens,’’ quartzose rocks of local origin, which were also used in the great circle of Avebury. The other stones, termed ‘‘blue stones,’’ are doleritic igneous rocks. In the nineteenth century the first careful petrographic descriptions of the rocks were made, and by the early twentieth century H. H. Thomas was able to trace the exotic blue stones by petrologic and petrographic analyses to the Preseli Mountains in Wales.2 In the late twentieth century, studies using X-ray fluorescence analysis showed that the dolerites came from three sources in the eastern Preseli Moun-

Sourcing (Provenance)

Figure 8.1 Stonehenge (Drawing by Elaine Nissen)

tains, and the rhyolitic rocks from four sources in the northern Preseli Mountains.3 The altar stone came from southwestern Wales. This variety of sources suggests that the monoliths may have been taken from a glacially mixed deposit, rather than from an in-situ quarry. In other words, the long transport of these large stones to Salisbury Plain may have been accomplished by a glacier, not by humans. Provenance determinations have three major components: (1) locating and sampling for analysis all potential source geologic deposits for the artifact material in question; (2) choosing an analytic method that has the sensitivity and scope to provide diagnostic signatures for each geologic deposit as well as for the artifacts; and (3) choosing a statistical- or data-analysis technique that can evaluate the data and then assign artifacts to source deposits. Only the first part of this process is geologic, so we shall concentrate on it.4 Attempts to assign artifactual materials to a particular geologic deposit need to solve two inherent issues. First, it must be established that the artifact has not undergone any chemical or physical alteration that would invalidate direct comparison of it with the same component material from known deposits. Second, all potential source deposits must be adequately represented in the database for a confident assignment of provenance on the basis of chemical or physical patterning. The first issue has many aspects. Artifacts fall into three groups, depending on the nature of the processing needed to transform the raw material into a useful object. Then, it must be determined whether significant chemical changes occurred to the object (artifact) during use or

after burial. Only chemical characterizations unaffected by processing, manufacturing, use, or postburial diagenesis can be used for provenance determination. Minor and trace-element patterns in lithics and ceramics are not usually altered by use or post-use (burial) conditions. There is a large group of rock and mineral raw materials that require no chemical or physical processing during the manufacture of the object that could alter the chemical characterization found in the raw-material deposit. These include various lithic materials like obsidian, chert, jade, quartzite, serpentinite, and marble, as well as native copper, gold, and amber. For a second group of raw materials, the production of an object is more complex. The most important members of this group are clay and temper for the making of pottery. Clay, water, and frequently temper must be selected, prepared, and mixed in appropriate proportions before being shaped, dried, and fired. The final object is a new material that is not found in nature. A third group of materials requires even more advanced processing and change on the way to becoming a manufactured object. The best example of this group is the complex ores of copper. After mining, the ores must be smelted with fuel and usually a flux. With copper sulfide ores an intermediate stage, matte, is produced. In smelting to produce copper metal, the complex of materials in the furnace splits into metallic and slag phases. Hence, it is impossible to track the trace elements in an original copper mineral once they have been mixed with fuel and flux and divided between metal and slag during the processing. Throughout most of its metallurgical history, smelted copper

223

Sourcing (Provenance)

was then alloyed with tin (usually tin and often lead in ancient China) to make bronze or with zinc to make brass. This alloying of metals, plus the use of scrap metal and remelting, also obscures the trace-element content of the original copper. The chemical reactions in smelting do not alter the lead isotope composition of an ore mineral. However, copper minerals contain small amounts of lead, and so may provide provenance data on smelted ores. Determining the chemical characteristics of geologic deposits and artifacts requires careful location or selection of sites or objects, statistical sampling of deposits or objects, selection of the most appropriate analytical techniques, standardization of analytical procedures, establishment of databases, and evaluation of large sets of data. The quantitative requirements for sampling, chemical analysis, and statistical analysis are much smaller for many archaeological problems if the researcher is merely trying to determine that the raw material in an artifact did not come from a specific deposit rather than figure out the specific deposit where the material originated.

Geologic Deposits To establish and define a potential geologic source through trace-element concentrations it is necessary to do two things: establish the geologic uniqueness and boundaries of the deposit; and collect and analyze ten or more samples that are widely dispersed statistically throughout the deposit. In practice, both these tasks range from difficult to impossible to accomplish. The first condition appears clear: What is the geologic deposit being sampled? Some primary and secondary native copper deposits are quite large (both in terms of kilometers in extent and hundreds of meters in depth). Others are small, perhaps no more than 100 m across. Our work has shown that good, nonoverlapping fingerprints are common for small deposits and for smaller parts of large deposits, such as individual mines (for example, in the Keweenaw Peninsula of northern Michigan).5 The larger the deposit or the larger the area or volume represented, the more diffuse the traceelement fingerprint. 224

In practice, defining the three-dimensional geometry of an ore deposit when only the surface outcrop is available is impossible. This ordinarily does not matter because the prehistoric miners did not penetrate to any great depth (usually they went only a few meters). It is necessary, however, to establish the areal integrity of the deposit. The samples should be collected as widely as possible from the extent of the outcrop to ensure coverage and to establish the complete range of variation in trace-element concentrations. If a well-defined ancient quarry is being investigated, one problem is traded for another. The boundaries of the quarry can be adequately established, but, unlike most ancient mines, in quarries most or all of the material that ought to be sampled has been removed. Researchers also need to determine whether modern, large-scale mining or quarrying techniques have obliterated a deposit exploited in ancient times. The number of samples needed to characterize a deposit will increase with the trace-element variation in the deposit and the number of trace elements needed for the provenance determination. Work on native copper has shown that it takes more than ten samples to characterize a deposit even minimally, assuming that eight or more trace elements are used in the characterization.6 Working with trace-element data on archaeological ceramics, it has been shown mathematically that when fewer than ten elements are used, serious overlaps in source characterizations occur.7 With discriminant analysis it is assumed a priori that each trace element has equal weight in the classification scheme. The greater the number of trace elements used, the more reliable the classification will be, if the number of samples analyzed is greater than the number of elements used in the discrimination. It should be emphasized that although abundant data on two or three elements may allow sourcing in some local situations, for example, with obsidian (which generally needs fewer elements for characterization), reliance on too few elements will defeat the provenancing effort. An excellent example of how to trace the source of stone objects is the work that was done on the origin of the rock used to construct the Colossi of Memnon on the plain near Thebes

Sourcing (Provenance)

in Egypt. These are two statues that protect a temple built during the eighteenth dynasty (New Kingdom) by Amenhotep III during the fourteenth century b.c.e. The statues were made of a very hard, ferruginous quartzite. Based on macroscopic examination, the quartzite blocks could have come from at least six sites. Using instrumental neutron activation analysis (INAA) for trace-element characterization, it was shown that the Gebel el Ahmar quarry, lying 676 km downstream on the Nile, was the likely source, rather than the quarries at Aswan (only 200 km upstream) which would have provided much easier transport, since the rocks would have been traveling downriver.8 Finally, for an understanding of the regional geology and the location of specific deposits, both the technical literature and the appropriate geologic survey personnel need to be consulted. Government geologic surveys—at all levels of government—are extremely helpful. Only a small part of geologic knowledge finds its way into the published literature, however. Most of it resides in unpublished reports, maps, field notes, and the brains of field geologists. Go to where the knowledge is.

Materials Used in Sourcing Obsidian Throughout the Neolithic of the Old World and in the early human prehistory of the New World, the main material traded widely over long distances was obsidian. Some of the most successful trace-element sourcing has been accomplished for this material. Obsidian is not common in most regions of the world, and the number of possible sources is limited. A good knowledge of the bedrock geology of a region will allow a quick determination of the location of any potential obsidian sources. Obsidian is formed in lava flows and as blocks in tuff from explosive volcanic eruptions. Obsidian flows do not have the same appearance as normal basaltic flows. Owing to its higher silica content and the resulting viscosity of the lava, obsidian flows are often dome shaped. The result of rapid cooling, obsidian often occurs on the outside of flows. In regions of silicic volcanic rocks a

significant amount of fieldwork may be necessary to locate all possible sources of workable obsidian, including transported material in river beds. Nearly all obsidian originates in volcanic arcs or chains. Volcanic arcs extend from Alaska to Oregon, over much of Mesoamerica, down the west coast of South America, across the island chains of Southeast Asia, in the Caribbean, in East Africa, and in the Aegean. Similar volcanoes are found in the mountain belt extending from the Alps to the Himalayas, including archaeologically important deposits in the Carpathians and in central and eastern Anatolia. In common with other lithic artifacts, the chemical composition of obsidian objects does not change during manufacture. Large databases of chemical analyses of the major obsidian deposits for most of the world are available, and publications of successful sourcing of obsidian are abundant. Both instrumental neutron activation analysis (INAA) and X-ray fluorescence analysis (XRF) techniques are commonly employed in obsidian provenance studies. Some obsidian deposits can also be sourced by means of 87 Sr/ 86 Sr isotope ratios plotted against rubidium traceelement concentrations, by thermoluminescence (TL) analyses, or using magnetic properties. Concentrations of such elements as manganese, barium, scandium, rubidium, lanthanum, and zirconium vary by as much as three orders of magnitude among obsidian flows, while varying by less than 50 percent within a single flow or pyroclastic deposit. The precise suite of discriminating elements must be determined empirically for each regional situation. In at least some regions, obsidian deposits may be distinguished by means of major-element (such as calcium and magnesium) ratios rather than traceelement concentrations. Atomic absorption (AA) techniques can determine calcium and magnesium. This instrumentation is more readily available to most archaeologists than XRF or INAA. Since the 1960s considerable research has been devoted to locating Anatolian obsidian sources and determining chemical fingerprints for them. Archaeologists have developed reconstructions of early trade and cultural exchange from this database. However, this database may be misleading for two reasons: not all potential source deposits 225

Sourcing (Provenance)

have been sampled, and many deposits were not sampled systematically—with full knowledge and coverage of the geology of the site. Some deposits have obsidian flows covering a span of two to four million years. Flows two million years apart in age are unlikely to have the same trace-element composition. The two major central Anatolian deposits of Açigol and Çiftlik have up to a dozen distinct flows varying in age by as much as two million years. Figure 8.2 illustrates this phenomenon. Using INAA to determine twenty-five major and trace elements, Rapp and his colleagues have defined eight separate signatures, not simply one for Açigol and one for Çiftlik. In the eastern part of the Açigol caldera three separate flow signatures can be defined. In the western part of the caldera there is only one distinct signature—from the youngest of the obsidians in central Anatolia. In the Çiftlik area three separate sources can be distinguished. The eighth source is from the obsidians at Nenezi Dag, about halfway between Açigol and Çiftlik.9 Obsidian sourcing using trace-element characterizations is now routinely performed in many parts of the world. Until the 1990s it was assumed that obsidian found in prehistoric American Southwest contexts were derived from the Government Mountain source in northern Arizona, the Picketpost Mountain source in central Arizona, one of the obsidian deposits in northern New Mexico, or Mesoamerica. This assumption was based on lack of archaeometric studies of the obsidian in this region. Now studies have expanded the number of potential sources and assembled a quantitative trace-element database, which allows much better obsidian artifact sourcing for this important archaeological region.10 The development of new research techniques has moved the sourcing of obsidians beyond dependence on trace-element and related compositional analyses. Back-scattered electron petrography has proven a valuable technique for obsidian provenance studies in the southwestern United States and the Mediterranean.11 It relies on the ability to distinguish the different cooling histories of the originally molten obsidian. Back-scattered imaging, energy-dispersive X-ray

226

Figure 8.2 Açigol and Çiftlik Açigol and Çiftlik volcanic areas in central Turkey, showing caldera rims, interior domes, obsidian outcrops, and (numbered) collection sites for a provenance project by Rapp.

analysis, and image analysis with a scanning electron microscope, in conjunction with one another, provide a petrographic method to investigate geologic and archaeological materials that are difficult or impossible to analyze by optical thin-section microscopy.

Sourcing Other Igneous and Metamorphic Rocks Coarse-grained igneous and high-grade metamorphic rocks have served as building and monumental stone for more than five millennia. Sometimes these rocks were quarried hundreds of kilometers distant from the place they

Sourcing (Provenance)

were found. Many granites have jointing patterns which aid in quarrying. Quality granites widely used for building stone occur throughout most of the British Isles, in northwestern France, in much of Scandinavia, in Italy, and in Russia. Perhaps the most famous granitic rock is the syenite outcropping between Aswan and the first cataract of the Nile River. This rock has been used since the First Dynasty for tombs, temples, and sarcophagi. It was exported to other parts of the Mediterranean for obelisks and statues. At Aswan there is also a gray granite that was used for buildings in Egypt and exported. The red porphyry quarried near the first cataract of the Nile in Egypt has been used for sphinxes and statues in Egypt and was shipped by the Romans to Pompeii and other cities of the Roman Empire. Another popular rock used in antiquity is diorite. Unfortunately, many of the igneous rocks from archaeological contexts have been erroneously identified by nonspecialists. These errors have been compounded and perpetuated by archaeological tradition. Once in the literature, incorrect names are hard to correct. Diorite is a good example. Some true diorites were quarried in Egypt for statues and bowls, but many Egyptian monumental and statuary rocks have been misnamed ‘‘diorite.’’ Major- and trace-element whole-rock fingerprints have been used to provenance Egyptian basalt artifacts.12 Analytical data were applied to seven First Dynasty basalt vessels from Abydos, two Fourth Dynasty basalt paving stones from Giza, and two Fifth Dynasty paving stones from Abu Sir. The bedrock source for all the artifacts was the Haddadin flow in northern Egypt. A database of whole-rock analyses may be more available than for other sourcing studies because geologists and geochemists have made uncounted analyses of rocks for other purposes. The identification and classification of fine-grained (aphanitic) lithics of volcanic origin present special problems for the archaeologist. Petrographic and geochemical analyses can overcome these problems and provide the data for firm identification, classification, and sorting. Petrography and/or major-element geochemistry is required for rock classification, and trace-element geo-

chemistry can be used for discriminating among potential source deposits. One good example is from the Northwest Pacific coastal region of North America. Lithic debitage from the British Camp shell midden on San Juan Island, Washington, had been classified for the past hundred years as basalt, with a local source predicted. Petrographic and geochemical analyses have shown that these lithics are dacite rather than basalt and that the source is in the High Cascades, about 200 km from the site.13

Chert Chert (including flint and other varieties—see Chapter 7) is microcrystalline quartz with few trace chemical impurities in the quartz crystals but abundant impurities as micro-inclusions. Chert from even a single deposit can have considerable visual variability, particularly in color. Trace elements within source deposits vary in complex ways, but with adequate sampling traceelement fingerprints of chert sources can be established. Trace-element variation may be great, but the variation usually involves stratigraphic position and horizontal distance between samples. In some cases chert may be sourced by petrographic properties. Some chert deposits have diagnostic fossil impurities or microstructures that can easily be seen under the petrographic microscope. Chert and chalcedony have specific weight percents of quartz and the silica polymorph morganite. Quartz/morganite ratios show promise of being a new tool for chert sourcing. Petrography and X-ray diffraction with Rietveld refinement were used to determine the weight percent of morganite in archaeological samples from southern New England, United States.14 Knife River Flint is a high-quality lithic material that was used by indigenous cultures of the northern Great Plains of North America from Paleo-Indian to historic periods. It has been reported at archaeological sites up to 1,000 km from its geologic source. In the past, identification has been made on the basis of visual criteria—hence, many misidentifications were likely. Some regional brown cherts called Hudson Bay Lowland cherts are visually similar to Knife River

227

Sourcing (Provenance)

Flint but have been shown by INAA to be chemically distinct from it.15 A broad-based study of cherts and flints in Finland used micro- and macro-fossils contained in the lithics as one of the diagnostic properties to trace the geologic sources of the artifacts to southern Sweden, Denmark, and the former USSR. Diagnostic microfossils included foraminifera, bryozoa, crinoids, echinoids, and ostracods.16 The fossils were identified by petrographic analyses of thin sections. Typical of European flints and cherts, the geologic sources were associated with limestone, usually chalk. An underlying problem with chert provenance studies is that there may be hundreds—or possibly thousands—of potential sources for chert throughout the world. As with other geologic materials, it is first necessary to determine the geology of the chert sources in a region, not forgetting glacial or alluvial sources. The fact that not all chert is workable into objects mitigates this problem. As with other geologic materials, local museums and academic departments will have collections of cherts from the region. However, it is unwise to rely completely on such collections, because the full range in variability of each deposit must be analyzed to obtain an adequate fingerprint.

Marble Attempts to source classical marble in the Mediterranean and Anatolia go back at least to the times of Theophrastus and Pliny. Geologists in the nineteenth century tried (unsuccessfully) to distinguish marble quarries by petrographic methods. Scientific efforts in the second half of the twentieth century using XRF, emission spectroscopy, INAA, and TL were also unsuccessful. Finally, stable ratios of carbon and oxygen isotopes, sometimes augmented by strontium isotope analyses, proved to be a key to sourcing marble. The geologic history of the rock, including its sedimentary origin and subsequent metamorphism, governs the carbon and oxygen isotope composition of marble. For isotope provenance to work, the isotope fingerprint must be uniform over the volume of the quarry, and preferably over

228

the entire geologic district. Uniform isotopic compositions will be attained over a broad area if isotopic equilibrium was attained during original formation and metamorphism, the metamorphic gradient was not too steep, and the marble rock body is relatively pure and thick. Because only the purest white marble was quarried, accessory minerals and other impurities do not generally occur. Since Roman times the principal marble quarries of Italy have been those of Carrara, which are located about 50 km northwest of Pisa in the Apuane Alps. This marble is noted for its purity, grain size, and color, which make it a coveted ornamental and statuary stone. Carrara marble was first quarried under Julius Caesar in the first century b.c.e. Both Julius Caesar and Augustus used Carrara marble to replace older brick buildings. In addition, the Romans exported Carrara marble throughout the ancient world. Greeks and Romans of the classical period preferred pure white marble for statuary and monuments bearing inscriptions. Therefore classical lands have extensive remains of buildings, monuments, and statuary composed of pure white marble. In the Aegean area much of this marble came from Proconnesus on the island of Marmara in Turkey, while some came from Carrara, and some from quarries in Greece. Proconnesus marble was perhaps the most commonly used variety in buildings throughout the classical world. It was also used for large sarcophagi. An extensive database of isotope analyses from all the major marble quarries in the classical lands of the eastern Mediterranean (Italy, Greece, Turkey, and Tunisia) has been assembled.17 These data allow sourcing of Mediterranean area marbles in most cases. In addition to sourcing, these databases can be used to detect forgeries, and for associating broken and separated pieces of statues, epigraphy, and monuments. However, attempts to discriminate the three main quarrying areas of the important Italian Carrara marble district have been uncertain on the basis of isotope signatures. A multi-method approach, based on petrography, stable isotope data, and electron spin resonance spectroscopy, has been used successfully to distinguish the quar-

Sourcing (Provenance)

rying areas.18 Their success rate was greater than 80 percent.

Clay For a discussion of clay minerals used in ceramics see Chapter 7. Although clay sources are needed for tile, brick, adobe, and related materials, the most important archaeological questions have centered on sourcing pottery clays. In fact, probably the greatest amount of effort spent in archaeological provenance studies has been to source pottery clays. Provenancing the clay component of pottery to a local deposit has been successful in only a few cases. This is in part because minor and trace elements are not distributed homogeneously in clay beds (particularly in comparison with obsidian and native copper deposits) and in part because potters selected clays from heterogeneous clay deposits. After clays were selected they were prepared, often by removing undesirable impurities which may have affected the ratio of clay minerals. Sometimes the clays were mixed to achieve a better product. Since we can have no direct knowledge of the pottery-making practices of prehistoric potters, we cannot correct for or evaluate the chemical alterations inherent in the pottery manufacture. Consequently, provenancing of ceramic artifacts is most frequently performed on the basis of a comparison between the trace-element composition of the unprovenanced ceramic object and the composition of a group of ceramic objects of known provenance. Provenancing pottery raw materials takes two forms: (1) minor and trace-element analysis and (2) petrographic analysis. For Anatolian pottery and clays, cesium, thorium, scandium, hafnium, tantalum, and cobalt have proven good discriminating elements. Similarly, in the nearby Aegean area, hafnium, manganese, cerium, and scandium have high discriminating power. Sourcing ceramic raw materials has at least one advantage. It can reasonably be assumed that local sources (not more than a half-day’s journey away, and usually less) were used. Contrast this with copper, obsidian, and especially tin, where the nearest source could be hundreds of kilometers away. Examples of the use of petrography to source pottery raw

materials are given below in the section on petrographic analysis. Microfossils have been used as indicators of provenance in sub-Neolithic pottery from Finland. The glacial clays of Finland formed during different stages in the history of the Baltic Sea, and fossil diatom flora are important in the study of this period. The diatoms in the clay are related to the sedimentary environments of the ancestral Baltic. Siliceous valves of diatoms can withstand the firing temperatures of pottery. Prehistoric peoples used clays deposited during two stages of Baltic history, ignoring clays found nearer the settlements. In one study, changes in diatom composition mirrored changes in pottery decoration style.

Temper Grains much larger than clay particles can become incorporated in pottery either by being a constituent in the original source clay bed or by being added as temper by the potter to modify the properties of the clay. Mineral tempers include crushed rock (of many types), shell material, quartz sand, and volcanic ash. For provenance studies, the critical question is whether the coarse-grained material was present naturally or was added by the potter. One way to approach this problem is to characterize the material carefully: in terms of mineral identification, particle shape, size distribution, and amount. This will give some indication of whether the material is a likely constituent of a clay bed or more typical of a local stream deposit or, in the case of angular crushed rock, unlikely to be derived from a natural geologic source. Provenancing temper is best accomplished with petrographic techniques (see below).19 Temper can be a valuable functional variable.20 This has been understood since the pioneering work of Anna Shepard,21 who used petrography to investigate the differences between two functionally distinct classes of pottery based on the temper used. She showed that culinary versus nonculinary wares contained different tempers. Shell tempers offer properties that make shells an important source of temper material. Shell has a coefficient of thermal expansion very similar to clays

229

Sourcing (Provenance)

and shells fracture into broad plates with large surface areas per unit volume.22 These properties provide a barrier to crack propagation. The source of the volcanic temper sands in prehistoric pottery from the South Pacific island of Tonga has been a major provenance problem. The scarcity of noncalcareous sand on most of the inhabited islands in the Tongan group has prompted the suggestion that either temper was imported from a volcanic island to the west or the pottery was imported from Fiji. Although the mineralogy of the tempers is compatible with that of Tongan volcanic rocks, Tonga lacks deposits of the rounded and well-sorted sands found in the tempers. The presence of beach placer sands derived from the geologic reworking of tephra deposits provides a satisfactory source. Using petrographic techniques, compositional analyses of temper sands in many ancient sherds from throughout the island group indicate that pottery making using local raw materials was once widespread.23 One of the most thorough and inclusive case studies of the provenance of ceramic and metal raw materials is focused on the Bronze Age of the Mediterranean island of Cyprus.24 Sourcing was used in efforts to reconstruct prehistoric production and exchange systems. Various analytical techniques were used to source Cypriot metals, pottery, and clays. Geochemical, archaeological, and statistical data are critical for any provenance study. Unlike the highly successful provenancing of metals, sourcing pottery raw materials has proven exceedingly difficult. Patterns in traceelement data on ceramics may relate to the function of the ware rather than location. This study also emphasizes the problems that arise when analytical data sets are obtained by different techniques or by the same technique from different laboratories.

Amber Establishing the provenance of European amber artifacts by infrared spectroscopy has been successful in many studies. The reliability of this technique rests on the fact that Baltic amber has a highly characteristic infrared spectrum. Assignments for known material are about 97.5 percent

230

correct. Limitations of this method became apparent in source studies of artifact amber which was badly weathered. Research on this problem resulted in the creation of a special method for studying weathered amber that uses gas chromatography to determine the diagnostic amount of succinic acid.

Bitumen Bitumen used as a preservative in ancient Egyptian mummies was long thought to have come only from the Dead Sea area. There are Egyptian sources of bitumen on the shores of Egypt’s Gulf of Suez. Biomarkers of bitumen from the Dead Sea, the Gulf of Suez, and five mummies were analyzed. It was found that four of the mummies contained Dead Sea bitumen, but the fifth and oldest (900 b.c.e.) had bitumen from the Gulf of Suez. This provided the first evidence for indigenous Egyptian source of bitumen.25

Soft Stone, Other Rocks, and Semiprecious Minerals Using trace-element analyses, many other lithic materials have been successfully sourced to their geologic deposit of origin.26 These rock and mineral materials include steatite (soapstone), serpentinite, turquoise, and, from Japan, sanukite (an andesite with orthopyroxene, garnet, and andesine in a glassy groundmass).27 For nonarchaeological purposes, geologists have used a wide variety of techniques to source geologic deposits, formations, and minerals. Examples include: isotopic provenancing of sandstones from the Eocene Type Formation in the Oregon coast range; correlation of North American Ordovician bentonites using apatite chemistry; and trace-element sourcing of detrital quartz in a sedimentary formation to the granite where the quartz originated. Soapstone artifacts dating to the first millennium b.c.e. from the James River drainage in Virginia have been successfully traced to their quarries of origin by rare-earth element concentrations using INAA. Many soapstone bowls from habitation sites in Virginia have rare-earth trace-element patterns that match outcrops and quarries from the region. Artifacts from five

Sourcing (Provenance)

North Carolina habitation sites also had rareearth trace-element patterns that matched the Virginia soapstone sources. Three of sixteen soapstone artifacts from as far away as northwestern Mississippi were also traced to the Virginia deposits. However, questions have been raised about early studies using rare earth elements (REEs) to source steatite from the Middle Atlantic region of eastern North America. Many instrumental neutron activation analyses and multivariate statistics indicate the transition elements, not REEs, make the greatest contribution to characterizing the steatite sources in the region.28 Investigation should be considered to be preliminary until samples from all potential source deposits have been analyzed.29

Native Copper In the United States and Canada, utilitarian copper artifacts appear initially in the archaeological record about 5,500 b.p. Use of native copper flourished in the western Great Lakes area because of the wealth of available copper in the form of nuggets and lode deposits outcropping at the surface. Early uses of copper include tools and ornaments, but by the time of the first European contact, the indigenous peoples used copper primarily for decorative purposes. Since the only technology available to North American indigenous societies was hammering and annealing, the size of a piece of copper governed its usefulness. There was not even a technology available to retrieve usable sizes of copper from an oversized mass. The famous Ontonogan (Michigan) copper boulder is a good example: it weighed about one-and-a-half metric tons. Such masses could not be chiseled, sawed, broken, or otherwise comminuted and so remained unavailable to prehistoric miners. In the eastern Mediterranean and Near East, copper has been in use since early in the ninth millennium b.c.e. Until about 4000 b.c.e., the copper used was probably native copper. In Anatolia, where the earliest extensive use of native copper is known from Neolithic Çayonu, it is assumed that copper was native copper from the large deposits at Ergani Maden, located just 20 km from Cayonu. However, extensive modern open-pit mining has removed all or most of

the evidence, including the native copper portion of the deposits, thus limiting the opportunity for trace-element sourcing studies. Compared to work on North American native copper sourcing, little research has been done on the specific sources of native copper from this area, the region where copper metallurgy developed.30 Native copper occurs primarily in three geologic environments: in mafic lavas and mafic and ultramafic intrusives; in the oxidized zone of copper-sulfide deposits; and deposited in clastic sediments associated with mafic igneous rocks. We refer to deposits of the first type as primary, deposits of the second type as secondary, and deposits of the third type as sedimentary. In North America large deposits of native copper occur in basaltic lavas (and related sedimentary rocks) of the Lake Superior region. Most of the native copper available to prehistoric North Americans at or near the surface was derived from mafic igneous rocks. The opposite is true of Old World native copper deposits. There most of the native copper has come from secondary deposits, where it occurs as a secondary alteration mineral within the oxidized zones of copper-sulfide deposits. Strong oxidation takes place in these deposits because of the abundance of pyrite (FeS 2 ), which dissolves in water to form sulfuric acid and ferric sulfate. The associated chemical reactions are quite different from those in native copper formation in basaltic lavas. This oxidized zone is always closer to the surface than the remainder of the deposit and often outcrops at the surface. In the sulfide copper ores of western North America, native copper is a common but minor constituent of the oxidized zone. Chemical characterization of native copper is performed on the basis of trace-element concentrations. Several analytic methods are available to determine the chemical elements in the low parts per million–parts per billion range. By far the most commonly used is INAA. Because trace elements may vary by orders of magnitude among deposits and by more than 100 percent in a single nugget, absolute values are not as critical as the overall pattern of element concentrations. Rapp and his colleagues have been traceelement fingerprinting North American native

231

Sourcing (Provenance)

copper deposits since the 1970s. The principal finding of our research indicates that the chemical fingerprint of an individual native copper source, coming from a single, well-defined deposit, can be firmly established. This is true in cases where the deposits are geographically close, as well as where they are distant.31 The trace elements that contributed the most to the sourcing of North American native coppers were gold (Au), lanthanum (La), arsenic (As), silver (Ag), iron (Fe), and antimony (Sb). Historic archaeologists are interested in whether copper implements made shortly after the first European contacts with indigenous Americans were manufactured from bits of European referred copper.32 In a separate study accelerator mass spectrometry was used to distinguish native copper and copper alloy artifacts of North American origin from those from Europe.33 This research project also was able to distinguish rapidly by metallography items made of native copper from those of smelted (European origin) metal.

Myth has influenced ideas concerning the sources of Old World copper, and not all myths are ancient. Modern commentaries on the Old Testament contain references to King Solomon’s mines, supposed to be located at the head of the Gulf of Aqabah. Copper from this area was said to have been one of the mainstays of Solomon’s wealth. This myth was added to by the American archaeologist Glueck who, from 1938 to 1940, excavated a tell near the head of the Gulf of Aqabah. Although there is no specific reference in the Old Testament to King Solomon’s mines, Glueck reported that during the earliest phase of occupation, the tell had been the site of a copper refinery for the nearby ‘‘King Solomon’s mines.’’ But later research showed that during the early mining operations at Timna, the local copper deposits were under Egyptian control—long before Solomon was born. And later mining activity took place long after Solomon died. Absolutely no evidence of mining activity during Solomon’s time exists. Myths and archaeology come together only when there is sufficient evidence to support oral tradition or myth.35

Complex Copper Minerals The smelting of copper goes back six millennia to the smelting of malachite (Cu 2 CO 3 (OH) 2 ) and perhaps azurite (Cu 3 (CO 3 ) 2 (OH) 2 ). Malachite, azurite, and other oxide copper minerals occur in the upper, oxidized zone of copper sulfide deposits. These oxide-zone deposits are not as rich or as extensive as the sulfide-zone copper deposits. As the oxide-zone copper was depleted by ancient miners and metalsmiths, increasing pyrotechnical skills were required to smelt sulfide copper; this would have had a profound effect on economies and trade relations in early phases of the Bronze Age. The ability to determine whether a copper or copper alloy artifact was derived from native copper, oxide copper, or sulfide copper thus provides archaeologists with important information on the development of ancient technologies and economies. Research has shown that analyses of artifacts for a suite of diagnostic elements (antimony, arsenic, iron, lead, silver, and sulfur) can, in a majority of cases, trace the origin of the copper to ore type, though not to specific deposit.34

232

Tin Tin has been far more important to metal-using societies than is commonly appreciated. Tin is an essential component of bronze and pewter, from which many artifacts were made. (Bronze is approximately 90 percent copper and 10 percent tin.) One mineral, cassiterite (SnO 2 ), accounts for virtually all of the tin that has ever been exploited. Fortunately, cassiterite is a stable oxide that remains unaltered when weathered out of lodes to form placer deposits; this makes it easier to source. One of the most perplexing provenance problems in geoarchaeology is where the tin came from for the bronze in the Mediterranean and Near East during the Bronze Age. To understand the problem let us first contrast the abundances of iron, copper, and tin. Iron makes up approximately 5 percent of the earth’s crust, and deposits of iron oxide are ubiquitous. The abundance of copper in the earth’s crust is about 50 ppm, and copper deposits are widespread. Tin, on the other hand, makes up only 5 ppm in the earth’s crust,

Sourcing (Provenance)

Figure 8.3 Tin Belts in Europe and Asia

and exploitable concentrations of tin occur only in a few metalliferous zones. In the Old World, where tin bronze originated, the only significant known deposits are in Cornwall, Brittany, Iberia, the Erzgebirge Mountains of the northwest Czech Republic, and Tadjikistan (fig. 8.3). Minor deposits are found in the eastern desert of Egypt, Tuscany, Sardinia, and a few other locations. There is no evidence that humans used the tin areas of the Erzgebirge until about the twelfth century a.d. and no evidence of contact between Anatolia, where the first tin bronzes occur in the late fourth millennium b.c.e., and Cornwall, Brittany, or Iberia. For most of history, Europe’s largest producer of tin has been Cornwall, England, where tin mineralization is linked to the intrusion of granites more than 250 million years ago. During Roman times there was more tin mined in the Iberian

Peninsula, and during the fifteenth century the mines at Erzgebirge outstripped Cornwall. The tin deposits of Iberia are similar in most respects to those of Cornwall. The Iberian tin belt extends more than 480 km, from Galacia in the northwest to Extremadura in the south central region, with the whole of northwest Iberia dotted with abandoned mines, including the heavily worked alluvial (placer) deposits. Iberian deposits were exploited in the Bronze Age, but there is no evidence of trade with the eastern Mediterranean. However, the popular myth persists that the Phoenicians went to Cornwall to trade for tin, which they used to supply the ancient civilizations of the eastern Mediterranean and Near East. In the Near East, ancient texts indicate trade in tin at least back to the middle of the third millennium b.c.e. from an eastern source, perhaps beyond the Indus Valley of Pakistan.

233

Sourcing (Provenance)

Malaysia and Thailand have been the largest producers of tin in recent times, but there is as yet no scientific evidence that Southeast Asian tin was imported into the West before the first millennium b.c.e. There is no textual or archaeological evidence of tin coming from the west to the Aegean and Near East in the Bronze Age. Archaeological evidence for the tin trade, from a shipwreck off the coast of Anatolia, takes us back only to the latter half of the second millennium b.c.e. and that evidence indicates an eastern source.36 There are four cassiterite deposits in the Eastern Desert of Egypt. Although it had previously been believed that Egypt came relatively late to bronze making, one of the tin deposits, at Mueilha, has hieroglyphic evidence on nearby outcrops that ancient activity of some sort goes back at least to the Sixth Dynasty, 2200 b.c.e. On a Fifth Dynasty tomb at Giza a scene shows molten metal being poured from what may be a smelting crucible. This indicates that metal technology existed in Egypt from at least sometime in the mid-third millennium b.c.e. Rapp and colleagues have collected and analyzed cassiterite from Cornwall, Erzgebirge, Egypt, and a few other locations and have demonstrated that cassiterite can be sourced using traceelement signatures. However, as with smelted copper, smelted tin has an altered trace-element chemistry, and so far, archaeologists have failed to locate eastern Mediterranean or Near Eastern Bronze Age tin mining or smelting sites (where some cassiterite should be found). They have also failed to identify or recover any cassiterite in the excavated material.37

Lead, Silver, and Gold Lead objects, all of which must be derived from the smelting of lead minerals, appear in the late seventh millennium b.c.e. in Turkey. Because of lead’s low melting point (327°C) and the ease of smelting galena (PbS), which is metallic in appearance (at below 800°C), lead was easy for early metalsmiths to exploit. Galena deposits are fairly abundant throughout the area in the Near East where metalsmithing originated. In the earliest Chinese bronzes, lead was already serving as an al-

234

loying element, perhaps providing the castability desired by Chinese metalsmiths. Lead ores, especially galena and cerrusite (PbCO 3 ), often contain a substantial amount of silver and were probably the chief source of silver in antiquity. We can source most lead and silver artifacts from regions, such as the Mediterranean area, where the geologic deposits have been analyzed for their lead isotope ratios. Lead has four isotopes: 204 Pb, 206 Pb, 207 Pb, 208 Pb. Each has the same chemical properties. Therefore, in the process of smelting, the isotope ratios derived from the ores are not altered. This means that any isotopic signature in a lead or copper ore will be retained intact in the finished artifact (unlike traceelement signatures, which alter during smelting). Lead isotope ratios of ore minerals vary from deposit to deposit, depending on the geologic age of the deposit. This is because the radioactive decay of 238 U, 235 U, and 232 Th (thorium) is constantly producing 206 Pb, 207 Pb, and 208 Pb, respectively. Over geologic time the constant decay of uranium and thorium leads to changing ratios of 206 Pb, 207 Pb, and 208 Pb to 204 Pb, because the latter has not changed throughout geologic time. However, two geologic conditions can complicate this situation. A serious problem can arise when initial lead isotope ratios in a deposit are altered by subsequent episodes of metamorphism, which may introduce lead from surrounding rocks. A second problem stems from the fact that deeper parts of the earth’s crust and outer mantle have different ratios of lead to uranium and thorium than rocks in the upper crust. This can lead to isotope variations within a particular ore deposit, especially in the extensive lead-zinc deposits in sedimentary rocks. Because of early difficulties in measuring the abundance of the least-abundant isotope, 204 Pb, scientists routinely use the ratios 207 Pb/ 206 Pb and 208 Pb/ 206 Pb, which can be plotted graphically (fig. 8.4). Lead-isotope measurements have been used to determine the source of ores used in lead smelting, providing confirmation that in the early eighteenth century, people in central North America used local ores to produce metallic lead objects. Smelting appears not to alter the isotope ratios of source materials, and the lead ores frequently

Sourcing (Provenance)

silvery metallic mineral galena was used as a ceremonial object. Lead-isotope analyses have been used successfully to source galena in artifacts.39 Since the early 1970s, attempts to evaluate the trace-element sourcing of ancient gold objects have been inconclusive. Some research has indicated that major trace impurities in gold complicate the process of characterizing or identifying its source. Other investigations have been able to correlate these trace impurities with metallogenic provinces rather than specific deposits. Still other research has shown that some trace elements, such as indium, can be diagnostic. Tin and platinum have been shown to occur as minor impurities in secondary (alluvial) gold but not in primary (lode) gold deposits from European sources. However, no universal method has yet been established for provenancing gold. Figure 8.4 Typical Lead Isotope Fields for Copper Deposits in the Mediterranean Area At site B the actual plots for the ten samples that make up the field are shown. Except for site D, which is completely contained within the field for site C, the overlaps typically disappear when the third ratio (207 Pb 206 Pb) is factored in.

display different isotope compositions. The Guebert site, near the Mississippi River in southwest Illinois, contained a crude lead smelting operation associated with a village occupied between a.d. 1719 and 1765. Several other sites from about this time period also contained lead objects, ceramics, musket balls, and lead scrap. Isotope analysis identified the local sources of lead and also lead smelted in Europe. The European objects found were finished products like bale seals and musket balls. Lead-isotope analyses have also been used to determine the ore sources for leadrich metallic artifacts from the Bronze Age, for Roman age lead objects from Africa, and for lead in glazes on ceramics from the New World.38 Lead provenancing shows that ancient miners did not always locate and exploit the nearest ore deposits. For example, the Romans at Carthage imported substantial amounts of lead from mines in Spain and Britain. In North America, where indigenous peoples did not smelt or melt metals, the

Sourcing Methods One of the most serious problems for provenancing rock and minerial artifacts has been that nearly all the analytical techniques were destructive, at least of small samples. Portable gamma ray spectrometry has been used to perform nondestructive, in situ, geochemical characterization of granite-type columns from the site of Leptis Magna in North Africa.40 The technique was developed for geologic applications on extensive rock surfaces. This study provided evidence for the use of Turkish columns at the Libyan site, probably an outgrowth of the extensive Roman trade in building stones.

Trace-Element Analyses The rapid growth of accurate, automated techniques for trace-element chemical analysis has made the modern development of provenance studies possible. Some of the new techniques are nondestructive, and most require only small samples, which makes sampling more acceptable to museum curators and excavators. These newer techniques can be used economically on large numbers of samples under standardized conditions so that statistically valid results can be obtained.

235

Sourcing (Provenance)

The most common analytic techniques for provenance have used trace-element patterns or isotopic compositions to fingerprint geologic deposits. Each trace-element technique has its own set of problems with establishing standards; techniques have differing sensitivities and interferences for various elements; and the results often cannot be compared between laboratories. All techniques have the problem of inhomogeneities in the samples, and all require that care be taken to prevent postsampling contamination at trace levels. Each technique has either advantages for certain types of materials or cost benefits. Optical-emission spectroscopy, for example, which was the first instrumental technique used in provenance studies, did not have the sensitivity or precision necessary for trace element fingerprinting. Modern instrumental techniques are capable of multi-element analysis. These techniques can be used to identify more than half the elements in the periodic table. They provide high precision and accuracy over a range of element concentrations. However, each technique has its limitations, so geoarchaeologists must evaluate the relative merits of the competing systems in terms of the provenance problem at hand. The dominant techniques used by geologists for multi-element analyses are atomic absorption spectrometry (AA), inductively coupled plasma spectrometry (ICP), X-ray fluorescence spectrometry (XRF), and instrumental neutron activation analysis (INAA). Analysts doing archaeological provenance studies favor INAA; laboratories that lack access to a nuclear reactor generally use XRF: geoarchaeologists should have some familiarity with both. In AA and ICP, the sample must be dissolved before it is introduced into the instrument. When dealing with trace concentrations, this requirement adds significantly to the problems of accuracy. This brief summary of analytical methods is not meant to be exhaustive but merely representative of available techniques. Instrumental Neutron Activation Analysis Neutron activation analysis is a physical method of determining trace-element concentrations with high precision and sensitivity. Many chemi-

236

cal elements can be detected at the low partsper-million level and some can be detected well into the parts-per-billion range. In addition, a wide range of elements can be measured simultaneously with no loss in precision. Finally, INAA requires only a small sample (50 mg for metals, 200 mg for silicates); there is no complex sample preparation, and there are no extraction techniques. In this technique, the sample is subjected to irradiation by slow (thermal) neutrons in a nuclear reactor. Various constituent atoms capture these neutrons, producing unstable daughter elements. These unstable isotopes emit gamma rays characteristic of the original element present in the sample, and the gamma-ray intensity is a measure of the concentration of each original element present. The gamma-ray spectrum from the decay is measured in a multichannel gamma-ray spectrometer. Neutron activation methods can achieve high accuracy and precision for some elements but only moderate or poor results for others. As a technique, INAA has a different sensitivity (detection limit) for each chemical element. Sensitivities vary with irradiation time and intensity, delay time, counting conditions, and composition of the sample. Where a reactor is accessible and set up for automated INAA, the per-sample cost is low. A major advantage of INAA is the lack of a matrix interference, a problem with AA and ICP. The principal disadvantage for most investigators is the difficulty of gaining access to a reactor. Other disadvantages include the need for compromises when setting the counting routine and the need to monitor the neutron flux. Neutron activation analysis is one of the most common techniques for determining the geologic sources of raw materials used by prehistoric human groups. It has often been employed by geoarchaeologists interested in trading patterns, population territories, and migration. The basic steps involved are the same as those of all other raw-material sourcing techniques: both the artifact and the potential raw material used to produce it need to be analyzed to determine their compositional characteristics and whether there is a statistical match between the artifact and its potential source.

Sourcing (Provenance)

As an example, INAA was used to characterize the rocks from two prehistoric quarries from the North American Great Plains that both contained varieties of the Chadron Formation chalcedony (or the White River Group Silicates, which may be either from the Chadron or Brule Formations). After the two sources were studied, stone artifacts from a Clovis archaeological site in Kansas (the Eckles site) were analyzed by INAA. Because the sources of the Chadron Formation chalcedony apparently lay some distance from the site, locating a likely source provided a way to evaluate the long-distance movement associated with Clovis stone tool use. The two source areas studied were in Colorado and South Dakota. Element analysis was used to distinguish them. Using multivariate statistics, the artifacts from the Clovis site were found to most closely match the Colorado source. Based on INAA, it was possible to demonstrate that the materials at the Kansas Clovis site probably derived from the Colorado lithic source.41 Instrumental neutron activation analysis was also used to obtain trace-element composition data for chipped stone artifacts from the Mayan site of Colha in Belize. The major human occupations occurred between 1000 b.c.e. and a.d. 1250. Studies were undertaken to investigate chert sources and to trace the exchange and distribution patterns of objects made of chert. First, it was determined that there were compositional differences among cherts from different regions. Using discriminant analysis, archaeological chert samples from Colha and other Mayan centers were characterized according to geologic chert samples. Two chert types found at Colha were local, indicating that the people of Colha procured their chert from outcrops surrounding the site. Colhalike chert was also found at other Mayan centers, discoveries that have implications for our understanding of chert exchange within the region.42 X-Ray Fluorescence Spectrometry This technique achieved prominence in the 1960s and has been used widely ever since. A sample irradiated by an X-ray beam emits a secondary X-ray fluorescence spectrum characteristic of

the elements in the sample. The principal advantage of XRF for provenance studies is that it is a bulk technique and, like INAA, can be applied to an unprepared sample. This answers concerns about sample homogeneity and resistance to dissolution. It is possible to automate XRF systems, and they provide high-precision analyses for many elements. One can also design XRF systems to be both nondestructive and portable— highly desirable when dealing with museum artifacts. The principal disadvantages of XRF are matrix and interference problems, an instrument cost that is perhaps four times as great as that of AA, and lack of sensitivity in the ppb range compared to INAA. This has led to its use in those provenance investigations, such as with obsidian, where discrimination can be achieved even at the high ppm and percent levels for diagnostic elements. Isotope Analysis/Mass Spectrometry As the various isotopes of an element vary from one another only in their mass, the determination of isotope abundances or ratios is made with a mass spectrometer. In a mass spectrometer, the sample is positively ionized, then accelerated through a magnetic field. On leaving the magnetic field, ions of specific masses are collected separately and counted electronically. 87 Sr/ 86 Sr isotope ratios have shown to be a good discriminant for porcellanite stone implements from the Neolithic in Ireland. It is the best sourcing method for the porcellanites.43 Investigators worldwide have studied the renowned Alpine Iceman who lived in the Neolithic-Copper Age of Europe. Researchers have used isotopic signatures from teeth and bones compared to those from soil and water to pinpoint his origins to valleys in the southern Tyrol about 60 km southeast of where he was discovered.44 Data from 18 O indicated he came from south, rather than north, of where the mummy was found. This region of the Alps is geologically complex, allowing investigators to use strontium and lead isotopes in rocks and soils—and the plant food he ate that was grown there—to infer his geographic origins. Accelerator Mass Spectrometry (AMS) is com-

237

Sourcing (Provenance)

monly used in archaeology because it can determine accurately 14 C/ 12 C ratios on as little as one milligram of carbon for purposes of dating. In provenance studies the major use of the technique has been determining marble sources by means of ratios of carbon, oxygen, and strontium-stable isotopes and determining lead, silver, and copper sources with stable lead isotopes.

DNA Additional methods of tracing the source of natural and manufactured material continue to come on line. One of the latest is the use of DNA. The excavations of Roman and Early Byzantine deposits in Turkey yielded catfish remains belonging to species that do not occur near the site. Mitochondrial DNA analysis was performed on modern populations of this species from Turkey, Syria, Israel, and Egypt. The results of the ancient and modern DNA analyses indicate that the ancient catfish came from the lower Nile and were traded to the Sagalassos region of Turkey.45

Mineral Magnetism Magnetic properties of soils and sediments have been used widely in paleoenvironmental studies of land use and climate change. Mineral magnetic properties can also be used to characterize and source archaeological ochres. For example, several magnetic parameters were used to characterize ochres from known quarries in Australia.46 These studies showed that simple magnetic parameters (for example, susceptibility) are effective in discriminating among the quarries. However, more sophisticated parameters were required to assign ‘‘unknown’’ ochres to specific sources. All the required measurements were nondestructive, amenable to small samples, and very sensitive. Magnetic susceptibility provides a rapid, nondestructive method of in situ characterization and sourcing for rocks containing magnetic minerals. Although corrections must be made for object size, surface relief, and curvature, more than 350 Roman granite columns were measured with this method. Results showed clear groupings and similarities with potential sources in Italy, Turkey, and Egypt.

238

Other Analytical Methods Scanning Electron Microscopy (SEM) with Energy Dispersive Spectrometer (EDS) has been used to determine the chemical and mineralogical characteristics of some Etrusco-Corinthian ceramics from Italy.47 This study showed that one specimen was local rather than imported, and that two had temper from distinctly different basalts. The authors advise studying the chemistry of the non-temper matrix of ceramics because it can be diagnostic. Electron Microprobe analyses of amphiboles from Bronze Age ceramic vessels thought to be produced on the island of Aegina, Greece, from sources in the local islands revealed that each island volcanic center had its own unique mineralogical signature.48 The authors had sampled potential source rocks on Aegina and nearby Methana and Poros. Some of the ceramics came from outside the Saronic Gulf.

Petrographic Analysis This brief introduction to optical petrography is designed to present only a general picture of the usefulness of the polarizing microscope in geoarchaeology. Books and articles abound that cover all aspects of these methods in great depth.49 In our view such microscopic methods are woefully underused in archaeological investigations. Most lithic and ceramic materials and products are composed of minerals (obsidian—a volcanic glass—is an exception). Because minerals are crystalline, petrography is based on crystal symmetry and crystal chemistry. Crystal symmetry classifies minerals into seven crystal systems: isometric, hexagonal, tetragonal, trigonal, orthorhombic, monoclinic, and triclinic. For petrographic analysis, it is the optical symmetry and optical parameters of minerals that allow identification and interpretation. Isometric minerals are optically isotropic: they have only one index of refraction. Hexagonal, tetragonal, and trigonal minerals are uniaxial—they have one optic axis and two indices of refraction. Orthorhombic, monoclinic, and triclinic minerals are biaxial. They have two optic axes and three indices of refraction. Uniaxial and biaxial crystals are said to be anisotropic (not isotropic). Mineral grains can

Sourcing (Provenance)

Figure 8.5 Polarizing Microscope

be immersed in oils with known indices of refraction to measure the refractive indices and other optical characteristics that help identify them. Clay minerals are too fine-grained to be studied by the methods of optical mineralogy. The most important technique for petrologic and petrographic analysis is thin-section (optical) petrography. Thin sections are thirty-micronthick slices of rock or ceramic that under polarized light reveal textures and chemical alteration as well as mineral identification. (A note on nomenclature: petrology is the study of the origin of rocks, petrography is the description of rocks. American usage follows the strict definition. The description of rocks or ceramics in thin section is petrography. British usage is to call thin-section analysis petrology.) Thin-section petrography requires a polarizing microscope (fig. 8.5). The polarizing microscope is the most important instrument for determining the optical properties of minerals because with it information can be obtained easily and quickly. The polarizing microscope has two functions: it provides an enlarged image of the object placed on the microscope stage, and it provides plane- and crossed-polarized light and convergent light. The polarizing lens below the rotating microscope stage forces light to vibrate in a frontback (north-south) direction. A converging lens

is also mounted in the substage. Above the microscope stage (and above the object under study) is a second polarizer, called an analyzing lens, which forces light to vibrate left-right (east-west). The substage polarizing lens is fixed in place, but the analyzer can be moved in or out of position. Minerals exhibit a host of distinguishing characteristics in plane- and crossed- (both lenses in position) polarized light and in convergent polarized light. As with other compound microscopes, polarizing microscopes contain a variety of other lenses and devices that can modify the transmission of light for specialized studies. Thin sections of pottery allow identification of mineral constituents and their relative abundance, associations, and states of alteration; grain orientations and related fabric features; the size, shape, and orientation of voids; cracking; and post-use (diagenetic) recrystallization. The mineral grains will exhibit distinct size, shape, sorting, roundness, and sphericity characteristics. All these parameters can provide provenance data. Most thin sections are cut parallel to the vertical axis of the ceramic vessel, but different orientations may be useful for some studies. The aims of petrographic study of pottery are to understand the manufacturing technology and characterization. For example, one can usually determine how the clay paste was mixed from observing the types and distribution of inclusions. Also, whether the vessel was made by hand or a wheel can be inferred from observing the orientation of long inclusions. The chief disadvantage of thin-section petrography in ceramic characterization is that it does not permit study of the fired clay mineral particles. The extreme fine-grained nature of clays requires that clay mineralogy be studied by X-ray analysis because X-rays have much shorter wavelengths than visible light. Another drawback is that for many ceramics the only inclusions are quartz. Determining provenance from this ubiquitous geologic material is difficult. Two of the pioneers (beginning in the early 1930s) in the application of ceramic technology to archaeology, Anna O. Shepard, who worked on prehistoric pottery of the American southwest, and Frederick R. Matson, who worked on

239

Sourcing (Provenance)

ceramics from the Near East, depended on thinsection petrography for their analyses. Shepard’s first success was her demonstration that much of the pottery found at Pecos, New Mexico, was not made locally but was imported from adjacent regions.50 This finding established petrography as an effective tool for the study of cultural interaction. Shepard’s Ceramics for the Archaeologist and Matson’s Ceramics and Man formed the basis for most of the later work in this field (fig. 8.6). The increasingly frequent rise of technical provenance studies bring into question previous attributions based on art-historical judgment. Petrographic analyses of Iranian Safavid (a.d. 1550–1700) ceramics were undertaken to source them to production centers.51 The studies were undertaken because the sites of production for these world-famous ceramics were not based on hard scientific evidence. Five groups were characterized, some attributable to specific production sites. The results of the petrographic analyses in some cases supported prior art-historical judgment, but also showed that many earlier attributions were not tenable. The petrography allowed the development of more accurate stylistic profiles for each of the production centers. Additional examples of the use of petrographic methods in ceramic provenance illustrate the value of this method. The first example concerns Early Bronze Age pottery from the site of Akrotiri, on the Aegean island of Thera. Because whole vessels are rarely found, ceramic analysis depended on ascertaining diagnostic features of sherds (pottery fragments). The sherds at Akrotiri came from the first and second destruction Levels.52 Fourteen ware groups were identified, representing coarse local storage vessels and finer decorated wares thought to have been imported. The mineral composition and manufacturing technology of the fabrics were established in order to determine the existence of multiple workshops and more sources of raw materials for each ware group. The composition of the coarse fraction of the assemblage included a wide range of lithic material including discrete carbonate microfossils, angular quartz, mica, black iron oxide, argillaceous clasts, quart-

240

zite, talc, amphibole, and feldspar—a generous suite with which to work. The pottery fabrics represented a variety of raw materials and sources, including clays derived from volcanic pyroclastic sediments, lavas, micaceous clays derived from schistose rocks, and siliceous sediments. Although the origins of some of the raw materials remain problematic because of the lack of compounds or undiagnostic clays, the petrographic analyses identified local fabrics and imported wares. These pointed to contact with other Aegean islands, such as Naxos, Syros, and Crete. Microfossils in one of the fabrics suggested a possible source in the island of Melos. Chemical analysis further supported some of the petrographic data, particularly the presence or absence of fine-grained calcium oxide to distinguish one of the fabrics. Another example is from the Upper Mississippi Valley of the United States. The issue addressed was the extent to which petrographic analysis of ceramics can shed light on contact among prehistoric communities—specifically, two lateprehistoric villages 80 km apart.53 From a total of 331 vessels at one site, 10 were suspected on stylistic grounds of coming from the other site. The thin-section petrographic analysis consisted of a two-step procedure. The first step involved documenting the mineral inclusions and compiling a list of those that were ‘‘natural’’—or original—as opposed to those that had been introduced as temper. The second step consisted of a point-count analysis (in which the frequency of an object in a sample is determined by counting the number of times the object occurs at specified intervals throughout the sample), using a 1-mm counting interval over the entire area of the thin section. Between 100 and 350 nonvoid points were counted for each slide. The petrography characterized each thin section in terms of: kind of temper, temper grain size, temper amount, and relative proportions of natural materials (excluding temper). The petrographic data strongly suggest that the sixteen Late Woodland vessels from one site were all of local manufacture. They constitute discrete groupings and have a close similarity in composition to the B horizon of a soil developed within the loess that blankets the terrace at the

Figure 8.6 Anna O. Shepard (Drawing by Elaine Nissen) Shepard was a pioneer in ceramic technology, concentrating on material from the U.S. Southwest and Mesoamerica. She entered the ceramic field through archaeology and made her lasting contributions by combining optical petrography with the results of chemical analyses. Shepard joined the Carnegie Institution in 1936

which provided her with the facilities for petrographic and microchemical analyses. In 1956 she published Ceramics for the Archaeologist. For a thorough review of her work see The Ceramic Legacy of Anna O. Shepard (Bishop and Lange 1991). (Drawing by Elaine Nissen from a photo supplied by The Shepard Archives, University of Colorado Museum)

Sourcing (Provenance)

site. The primary temper added to these vessels was hematite, which occurs as natural concretions in the local sandstone bedrock. The petrographic data also established that the fired paste and body materials from the vessels were imported from the other site had been, as had been previously proposed.54 Local sand tempers from the eastern Solomon Islands have diverse compositions reflecting the varied geology of the different islands. Indigenous and exotic ceramic wares recovered from archaeological sites can be distinguished by petrographic temper analysis of thin sections, with the caveat that calcareous temper sands composed of reef detritus are not diagnostic of specific provenance.55 Temper petrography provides the physical evidence for cultural contacts and exchange between human groups within the region. Stone axes are the most abundant artifact surviving from prehistoric Ireland. Petrographic analyses, aided by X-ray fluorescence analyses, have been used to indicate that 15,916 Irish stone axes were made predominantly from one rock type, porcellanite.56 It was possible to differentiate porcellanite sources. Petrographic analyses were used to evaluate the compositional variability of a single ceramic type from 14 sites in Texas and Oklahoma, USA.57 They showed that the sands in these vessels had compositions that were similar within each drainage and dissimilar between drainages. They concluded that local production accounted for the observed patterning.58

Statistics and Data Analysis Many statistical methods have been used in provenance studies. The most common and most powerful are multivariate discriminate analysis and various approaches using cluster analysis, such as K-means cluster analysis. The mathematical methods employed in provenance studies must be able to evaluate the statistical characteristics of the large, multivariate chemical databases that are used and to characterize groups and assign unknowns with precision. Most multivariate statistical procedures assume that the variables (for example, trace concentrations of chemical elements) follow multivariate normal distributions.

242

Therefore, the distribution properties of the data need to be analyzed. In addition, it must be determined whether the data should be standardized. The statistical computer software packages used in discriminant analysis Statistical Package for the Social Sciences (SPSS) and Statistical Analysis Systems (SAS) have such features built in. This is not always the case with cluster programs. The decision to use discrimination procedures or clustering procedures depends on whether there is prior knowledge of the existence of groups. If groups are known to exist, discrimination procedures should be chosen; otherwise, cluster procedures should be selected. In other words, discriminant analysis seeks to discover and use those attributes that discriminate between known groups in order to assign unknowns to one of the groups. Cluster analysis is useful for identifying completely separate groups but less helpful for identifying groups that overlap. In most geoarchaeological provenance studies, the first step is to locate and characterize potential source deposits chemically. This means that the existence of groups (deposits) is known and that their traceelement concentrations will have some overlap. Therefore, discriminant methods are preferable. The SPSS X Discriminant Analysis package computes both linear discriminant and classification functions and has a thorough stepwise procedure that offers five different selection criteria. In using SPSS X it must be determined which elements are discriminating, based on the results of the trace-element analyses from the geologic deposits. A number of factors need to be considered in interpreting discriminant-analysis results. The computer assignment of an artifact to a source deposit as the most probable source does not ensure that the artifact raw material did indeed come from that deposit. It is possible that other sources exist that are not represented in the database. It is also possible for two deposits to have similar trace-element fingerprints; both will thus have strong assignment potential, although only one can be assigned. Probability must also be considered in the assignment of a source. The SPSS X package will assign as a source the deposit that has a centroid of points in multidimensional (equal to the number of elements

Sourcing (Provenance)

used) space which lies closest to the point defined by the trace-element concentrations in the unknown. However, if the location of the unknown in the multidimensional space does not indicate that it belongs to this source, the probability will be reported as zero. It should be borne in mind that data analysis cannot improve on the original data. If potential source deposits are inadequately sampled, or there

are errors in the analyses, no amount of statistical analysis will correct for these faults. Finally, remember that when archaeologists puzzle over the source of a new object, the question is, ‘‘What has moved? people (artisans), ideas, the object itself, or the raw material from which it was made?’’ Geoarchaeological provenance methods address only the source of the raw materials.

243

CHAPTER 9

Construction, Destruction, Archaeological Resource Preservation, and Conservation Since the advent of mankind many human societies have lived in volcanically active zones. The geological, archaeological and historical records provide a rich and diverse source of evidence for both archaeology and volcanology concerning the nature of volcanic processes and the effects of volcanism on the environment and on human society. To achieve a balanced understanding of the effects of volcanism on past cultures, it is important to consider the attractions as well as the hazards of life in an actively volcanic zone.—D. Griffiths 2000

T

his chapter concentrates on what happens when humans and nature interact, as well as on how archaeologists use geologic methods and knowledge to try to mitigate the effects of that interaction and protect what remains of the record of the human past.

Geotechnology From the first selection of appropriate types of rock for chipped stone implements through prehistoric techniques for exploiting various metal ores to make complex alloys, people have developed geotechnologies to accommodate their increasingly complex societies. For example, in the Old World, the early societies of Mesopotamia had developed geotechnologies by the third millennium b.c.e. to cope with problems of building in places where stone was difficult to procure and where the local soils and sediments had poor load-bearing capacity. Even with these constraints the Sumerians managed to build massive

244

ziggurats. They used sun-dried bricks, the materials of which must have initially flowed outward at the base of the large structure. Compaction would slowly have increased the load strength of the material. At some stage the Sumerians learned to place woven reed mats at regular intervals to absorb some of the horizontal thrust. Another Old World example of the development of geotechnical knowledge can be seen in the evolution of pyramid construction in Egypt. The first stone pyramids were built in step fashion of carefully dressed hard rock, to resist the horizontal thrusting. Pharaoh Snefru (reigned 2575– 2551 b.c.e.) had to give up his plan of building a 140-m high true pyramid with a slope angle of 60° because the marly soil on which it stood could not support it. He reduced the slope angle twice and was finally obliged to add an outer casing to make the structure a true pyramid. The ‘‘bent’’ shape of the southern stone pyramid at Dashur owes its success to Snefru’s earlier problem. During the construction of the Bent Pyramid, the slope angle was reduced from 54.5° to 43.5° to increase stability, shortening the pyramid from approximately 130 m to 100 m. By the time of the construction of the three great pyramids at Giza (ca. 2591–2536 b.c.e.), the major geotechnical problems of pyramid-building had been overcome. The greatest difficulty faced by the ancient Mayans of Mesoamerica (fig. 5.5) was the destructive nature of their tropical climate. They, too, built large pyramids. However, unlike Egyptian pyramids, their large structures must have required a great deal of repair and upkeep because of climatic and geotechnical problems. Ex-

Construction, Destruction, Preservation

cept for their recognition that large structures could not be built on swampy ground, the Mayans do not seem to have developed a geotechnical understanding of either rock-foundation stability or management of water runoff. The underlying soils included hydroscopic clays and tropical forest humus with high absorption capabilities, which expanded and contracted significantly during the wet-dry cycle and led to the displacement of masonry facing and the cracking of internal plastered surfaces. The technical response of the Mayans to these problems seems to have been to continually rebuild damaged structures. Much of the Mayan civilization was established on karst terrain. Little surface water exists in karst, so the Mayans had to develop groundwater resources in caves and from springs. Karst soils are typically thin and easily depleted of nutrients. In addition, karst soils are easily lost to the subsurface, a condition that not only affects agriculture but leads to the contamination of the potable water supply. The inability of the Mayans to solve their geotechnical problems may have contributed to their downfall.

Construction Archaeologists tend to focus on habitation, ceremonial, or burial sites. However, archaeology has become increasingly concerned with construction that has modified the landscape. Modern construction has also had a major impact on archaeological sites. Salvage archaeology has thus become one of the largest archaeological enterprises in many countries. Ancient construction produced a weird range of archaeological features, including dams, canals, fields and raised fields, and roads, as well as quarries and mines that all form part of the archaeological landscape.

Dams The geoarchaeology of ancient constructions includes many aspects of water management and hydraulic engineering. Irrigation of fields began more than 6,000 years ago. Major flood control structures were constructed more than 3,000 b.p. At Tiryns in Bronze Age Greece, a dam as high

as a three-story building was built to protect the topographically lower portion of the city against devastating floods, as had happened there earlier.1 The Romans built many dams of massive cutstone masonry set in lime mortar. Ancient dams, their geologic settings, and the construction materials used have received too little attention by archaeological scientists. Dams built to impound water impound sediment as well and thus have a finite life. Earth dams may wash away, often catastrophically, along with the impounded sediments. However, careful geologic study should reveal erosion and deposition impacts on the landscape. Evidence of ancient masonry dams should last many millennia. The various causes of dam failure are only tangentially relevant here, but because catastrophic dam failures have dramatic effects on humans, geoarchaeologists should be aware of the array of geologic evidence available about ancient dam construction. Dam construction has a number of harmful effects on the landscape, including siltation (a heavy silt load can cause sediment or soil liquefaction), reservoir bank erosion from wave action, leaching (owing to elevation of the water table), oxidation and other changes resulting from alternating wet and dry cycles, subaqueous slope failures, and biogeochemical alteration of archaeological contexts. The construction of dams to impound or divert water causes more interference with natural conditions than any other civil construction. From the geoarchaeological perspective, earth and rock dams of any age require materials with suitable strength and deformation characteristics, durable rock or earth, and minimal leakage through the foundations or materials from which the dam is constructed. Dams are probably the least well known of ancient structures but they have a long history. A cut-stone masonry dam, whose ruins are still in existence, was built in Egypt about 3000 b.c.e. to divert the waters of the Nile into a canal. It was still in use in Roman times. In the second millennium b.c.e. the Marduk Dam was built across the Tigris River to control flooding. This dam was maintained for thousands of years, finally falling into ruin in about a.d. 1400. A dam was built near Tiryns in Greece to divert water

245

Construction, Destruction, Preservation

into another river system as a flood-prevention measure.2 Finally, freshwater inundation is overwhelmingly detrimental to archaeological remains. The United States’ Reservoir Salvage Act of 1960 (amended in 1974) requires that any federal agency undertaking dam construction and reservoir impoundment provide written notice to the Secretary of the Interior, who shall cause an archaeological survey to be conducted. If significant archaeological resources are found, a salvage excavation or protective burial must be undertaken. The latter is basically a geoarchaeological problem, which will be discussed later in this chapter.

Canals Humans have been extending or connecting inland waterways, and penetrating narrow land barriers between seas, by canals for more than three millennia. In building canals without locks, the main problem is that the canal bed and banks must be impermeable to attain a consistent water level. The banks must also be stable. If there are locks, they need to be anchored to solid foundations. Local geology is of fundamental importance in all these concerns. A canal connecting the Nile and the Red Sea was begun in the fourteenth century b.c.e. but was not completed until the time of Ptolemy II (reigned 285–246 b.c.e.). It was rebuilt four times between the seventh century b.c.e. and the second century a.d., and rehabilitated again in the middle of the seventh century. In the time of Strabo it boasted locks with movable gates. Surviving remains indicate that it was approximately 97 km long, 46 m wide, and up to 5 m deep. This considerable depth was due to varying flood and low-water conditions of the Nile. At the Cairo end of the canal, silting would have been a constant problem. Many canals were also built in ancient Mesopotamia, while the Romans built canals to link their rivers. In the first century b.c.e. a canal traversed 26 km across the Pontine marshes near Rome, parallel to the Appian Way, and carried passengers when the local road was damaged by floods. In China, large-scale building of navigational canals began in the first millennium b.c.e. The

246

rivers traversed the great North China Plain in parallel fashion on the way from the western mountains to the sea. Thus, north-south transport links were needed. In 219 b.c.e., the oldest contour transport canal was built in Guangxi Province. One of the world’s great inland waterways was built in China during the Sui Dynasty (a.d. 581–617), by rebuilding and extending an earlier canal. The waterway began near Hangzhou, ran north across the Yangtze and Yellow Rivers, and ended near Beijing. Part of the route lay along rivers and part near lateral canals. Irrigation canals for water management in agriculture have been with us for millennia. Irrigation canals probably developed out of floodwater farming. The level of technology or geologic knowledge required for simple irrigation canals is not high. Over time, technology improved, populations grew, and a greater sense of geologic concepts was needed to develop earthworks adequate to provide the water for sustainable agriculture. For example, by the a.d. fourteenth century in the Basin of Mexico, the need for food was so great that even marginal lands were cultivated intensely, and every geologic source of water was exploited. Irrigation canals ranged from poor and inefficient canals dug in local porous earth to those lined with stone or stucco. Although most ancient irrigation canals were cut into soil and surface sediment, a few were chiseled into bedrock. Because ancient irrigation proceeded by gravity flow, determining the precise orientation of irrigation canals provides a three-dimensional marker on the former landscape. Canals in floodplain sediments would have needed regular attention. Erosional deterioration of canal walls and bottoms as well as sedimentation would have altered the profile of canals that were not well maintained. As indicated in Chapter 4, ancient ditches and canals are often easy to trace through geophysical methods. The oldest known features in Mexico that could be a canal irrigation system are Olmec (fig. 5.5), dating to perhaps 1400 b.c.e. Sometime before 1000 b.c.e. a storage dam of uncut stone and masonry blocks was built across the natural drainage. A canal lined with unfinished rock slabs led away from the dam to agricultural fields. The

Construction, Destruction, Preservation

Olmec also used similar conduits constructed of U-shaped basalt troughs as drains. Sometime in the early first millennium b.c.e. riprap (rock placed on embankments to protect against erosion and increase stability) was used in this region to prevent lateral erosion in a meandering stream. Stone slabs are available throughout Mexico. In contrast, they are scarce throughout much of the lower Mesopotamian plain, another area of early irrigation. An interesting canal phenomenon occurred in prehistoric Oaxaca in Mexico.3 Spring-fed irrigation water at a canal system was so high in calcium carbonate that travertine accumulated along the canals, thereby fossilizing them over the centuries. The sequence of technological change in canal building at Oaxaca and the Valley of Mexico consists of the following sequence (dates are approximate): 1400 b.c.e.—relocation of rivers; 800 b.c.e.—advanced relocation of ephemeral streams; 400 b.c.e.—rock diversion dams; a.d. 200—use of valley bottoms, advanced channelization; a.d. 350—use of permanent springs; a.d. 550—masonry storage dams with floodgates; a.d. 750—earthen dams. The agricultural Hohokam people of southern Arizona in the United States (fig. 5.5) were noted for greater duration of settlement than other roughly contemporary groups living in similar arid environments. Irrigation from rivers was frequently associated with long-term persistence of individual sites.4 Along the perennial rivers of the Phoenix area, the largest, most densely inhabited sites and the most extensive canal irrigation date to the later part of prehistoric time. Ancient societies also built canals for flood control, to keep excess water away from fields and to protect habitation sites and roadways. They also channeled streams to increase runoff in times of flooding. Careful investigation of the sediment composition, textures, and stratigraphy can reveal the nature of these ancient constructions. Great Britain is laced with historic-period canals. Perhaps the first canals date to the Roman period, but efforts in the 1600s to make rivers navigable led to the construction of a vast network of canals, which had a major effect on the landscape, on river and terrigenous hydrology, and on

the local ecology and microclimatology. Effects of this construction left records in the sediments of ponds and lakes. The Erie Canal in the United States is an example of a historic-period canal, constructed to link New York and Albany on the Hudson River through the Mohawk River Valley to Buffalo on Lake Erie. Begun in 1817, this canal had to rise 198 m on its way to Buffalo. It had ample water, and a local muck called ‘‘the blue mud of the meadows’’ served as a lining to prevent seepage. A high grade of limestone found near Medina, New York, provided excellent facing for locks and other structures. A special variety of limestone provided the raw material for the underwater cements. Future geoarchaeologists seeking to understand the construction of the canal, which has few comprehensive written records, need to keep in mind not only the geologic aspects of excavation but the whole range of raw materials involved.

Fields and Raised Fields Fields are an important part of the constructed landscape. Ideally, fields have boundaries but unless they are geometric and bounded by identifiable markers they may not be easy to delineate in the geoarchaeological record. Those fields that had irrigation canals leave a clear imprint, as does raised-field agriculture, and ancient terraced fields are easy to recognize. There are many categories of fields: cultivated crop fields, gardens, pastures for grazing, and so forth. If singleuse fields (and particularly irrigated fields) were of long-term duration the soils developed under them carry imprints of their use. Artifact distributions also may offer clues to the land use of constructed field areas. In the Andean highlands of South America, the terrain was often intensively human-modified to permit cultivation of marginal lands. One method was the development of raised fields, large earth planting platforms that prevent water logging and flooding, increase soil fertility, conserve moisture, ensure nutrient production and recycling, and improve crop microclimates.5 The platforms were constructed from the soil/sediment recovered from adjacent canals. There are, however,

247

Construction, Destruction, Preservation

negative environmental factors related to raisedfield agriculture including long-term aspects like the disruption of the natural hydrology and massive alteration of the animal communities. Overexploitation often caused erosion of the topsoil on slopes.

Roads There are three geologic aspects to consider in studying land transportation networks: (1) the topography (relief ); (2) the surficial geology— the need of a roadbed capable of carrying the intended traffic and with reasonable permanent stability; and (3) the availability of geologic materials needed in road building. Roman road construction marked a significant break from the earlier practice of following the easiest path. Roman engineers laid out their roads as straight as possible (possibly because their wheeled vehicles were not good at taking corners). This approach required an understanding of the engineering geology. At the height of the Roman Empire, Roman engineers with geologic insight built more than 75,000 km of high-quality roads. Archaeologists can trace most of this network today. For their main highways the Romans excavated a road-cut 45–60 cm deep and lined both sides with stones placed vertically. Where the soil was not sufficiently hard (indurated) a layer of medium-sized stones was placed at the bottom, followed by layers of gravel intercalated with fine sand or soil, frequently mixed with hydraulic lime for hardening. The surface of the road was finished with flagstone, sometimes transported from distant sources. Flagstones paving the Roman roads of the Po River plain were trachytes.6 These volcanic igneous rocks were very resistant to both abrasion and surface weathering. The trachytes were from the Euganean Hills and have been quarried since the sixth century b.c.e.; they were used widely in central and northern Italy. The oldest road in the archaeological record runs between Van and Elâzig in eastern Anatolia. It can be traced on foot for 100 km. This road predates the Persian Empire and was probably built by the Urartians in the late ninth or early eighth century b.c.e. It had a width of greater than 5 m and included bridges over small streams. It was in South America, however, that we find perhaps 248

the greatest skills in road building under difficult geologic conditions. The Incas constructed more than 6,000 km of mountainous roadway, stretching from Quito in Ecuador to Tucuman in central Chile. The road was more than 7 m wide, and some of it was paved with bitumen. It traversed the pathless sierras, crossed rivers and deep ravines, and scaled precipices, using stairways cut into the rock face. Deep ravines were filled with solid masonry to make bridges. The roadbed was often composed of flagstones. A second Inca road, stretching nearly 3,000 km along the coast, paralleled the main road. Here the geology and topography dictated a different construction. Much of the route was sandy, sometimes requiring an elevated causeway, other times the use of piling. Nevertheless, with simple tools and a good sense for engineering geology, the Incas were able to forge an empire in a land that would inspire only small enclaves.

Excavation (Mines and Quarries) Except for the construction of dams, the major earth-moving activities in ancient times were stone quarrying and mining. In some cases, such as for the exploitation of coal, open-pit mining was essentially a quarrying operation. Conversely, some quarrying was done in underground caverns, so quarrying and mining overlap. In ancient Egypt, methods of quarrying and working stone were developed early in the third millennium b.c.e. By the time of the construction of the first large pyramids (the Step Pyramid of Djoser) more than a million tons of limestone were needed. The grand pyramids at Giza each contain about 700,000 limestone blocks weighing roughly 2.5 tons apiece as well as about 200,000 m 2 of casing blocks of Tura limestone. The Tura limestone came from underground quarries in the Mokattam hills on the east bank of the Nile. As discussed in Chapter 7, there are many references by classical authors to the rock types used by the ancient Greeks, Romans, and Egyptians for their buildings and monuments. Despite these sources, we are just beginning to understand the technical side of ancient quarrying. Unless a later quarry obliterates an ancient quarry, the latter

Construction, Destruction, Preservation

may remain a major topographic feature. However, modern quarrying, coastal erosion, and the steady onslaught of rock weathering can remove even large ancient quarries from the landscape. Preliminary shaping of monumental forms, frequently carried out in situ, is a well-known feature of ancient quarry practices. In their quarries at Aswan, the Pharaonic Egyptians created massive obelisks (up to 75 m in length) that they removed in one piece. Hence, a knowledgeable geoarchaeologist can use the scraps of rock sculpting as a guide to the location of ancient quarries. The broken tools of early mining, chert, and hard igneous rocks are also evidence of ancient quarrying sites.7 The catacombs of Rome, which extend more than 850 km, developed as stone quarries. The catacombs of Paris required the excavation of more than 16 million m 3 of rock. The disposal of so much earth probably presented a greater challenge than the excavation.

Natural Burial and Site Formation Site burial results from two conditions: sediment input from fluvial, eolian, or downslope processes and lack of erosion. Sediment derived from upstream erosion has many sources: cultivated areas, grass and forest land, habitation sites, valleys and gullies, and the stream channel. The amount of sediment deposited on a site is due not only to upstream erosion but also to the carrying capacity of the stream and whether topographic conditions are favorable to deposition. The important factors influencing sediment movement are watershed size, land use, topography, bedrock and surficial geology, soil and vegetation cover, and precipitation patterns. Erosion, transport, and deposition are complex processes; the site geoarchaeologist needs to understand the interplay of these factors during site occupation and burial (see Chapters 2 and 3). Eolian deposition varies tremendously from region to region. Both extremes (deposit and removal of sediment) can be found in deserts. In sand deserts, such as the Sahara, eolian processes keep surface particles in near-constant motion, whereas some rock deserts have no particles fine

enough to be carried by the wind. Although there are no wholly reliable criteria by which to identify ancient eolian deposits, the scarcity or absence of clay or gravel, the predominance of fine-tomedium-grained sand, thick cross-bedding, and the presence of ventifacts indicate eolian deposits. Windblown dune sands are unimodal in size distribution with a mean size that is rarely less than 0.20 mm and rarely greater than 0.45 mm. When rock debris moves downslope under the influence of gravity rather than water, masswasting deposits are created. They are common terrestrial sediments—although obviously, there are few of them on broad, flat plains. Masswasting transport can take the form of falling, sliding, flowing, creeping, or subsidence. Sedimentary features, including structures, grain size, and sorting, may indicate the mode of transport. Seasonal creep produces a distinctive deposit with a stratified sequence that becomes attenuated downslope and passes into trails of debris (fig. 9.1). Geoarchaeologists can use this property to assess mass-wasting-driven geomorphic change for help in locating buried sites and in paleogeomorphic reconstruction. A much rarer geologic process that has buried places that are now famous archaeological sites (Pompeii, Akrotiri) is ash flow or ash fall from a volcanic eruption. Most deposits of volcanic ash are elongate in plan, with the long axis of the deposit extending downwind. The horizontal distance the volcanic ash is transported depends on the maximum height the ash column attains in the atmosphere, the direction and velocity of the wind, the size-distribution of the ash particles, and whether there is rainfall in the path of the ash plume. Finally, in addition to depositional forces, the lack of erosional forces—that is, a stable landscape—at a site can allow burial. Erosion is the norm in landscape development. Erosion by water or wind varies according to time and place. A cultivated field may be in the process of eroding while the next field, under pasture or forest, may not. Each soil and surficial sediment is characterized by a particular topography and land use and consequently has its own erosion regime, which changes with time. Slope stability and effective ground cover (vegetation) are the keys to under249

Construction, Destruction, Preservation

Figure 9.1 Composite Diagram of Typical Effects of Downslope Creep Under the influence of gravity there is constant if exceedingly slow movement of the earth’s surface ma-

terial downslope (creep). Rigid monuments tilt; growing trees overcome the tilt by bending back skyward. This slow creep can be contrasted with such rapid downslope movements as landslides.

standing local erosional processes that take place away from meandering rivers and wave-pounded coasts. Geoarchaeologists should be able to detect at least the general form and intensity of current and past erosion at a site.

tent, and durability. Most sedimentary and metamorphic rocks are layered, with each layer differing in composition, texture, and fabric from the adjoining layers. Marble and some quartzites are exceptions. Compositional and textural inhomogeneities, along with joints and fractures, lead to structural weakness. In some cases shape is also important. To be effective, riprap should have a shape that produces a stable interlocking and must be durable under the conditions to which it will be exposed. In particular, riprap needs to be resistive to alternate wetting and drying. Rocks weather in two main ways: physical disintegration, and chemical decomposition (see Chapter 2). The processes involved in decomposition

Rock Properties and Weathering Destruction Geoarchaeologists should be familiar with the properties of rocks that affect their suitability as building stone. The most important of these are compressive strength, shear strength, tensile strength, porosity, permeability, moisture con-

250

Construction, Destruction, Preservation

Weathering All materials have a stable state for the environment in which they exist. A significant change in environment, however, may force the material to transform into a material with a new stable state. Water, ice, and steam are each a stable state of H 2 O, under different temperature and pressure environments. Rocks formed at high temperatures under anhydrous conditions weather rapidly under moist conditions at the earth’s surface. The chemical breakdown of feldspars provides a good example of the weathering process. Feldspars are a series of sodium, calcium, and potassium aluminosilicates. They make up nearly 60 percent of the earth’s crust. Feldspar weathering proceeds by the removal of potassium, calcium, and sodium and the formation of clay minerals. Because of its high solubility, most of the released sodium finds its way to the oceans. Most of the potassium remains in the soil in new minerals like illite. Some becomes part of growing plants. Calcium does both: some ends up in the oceans, some remains in groundwater systems to precipitate in such terrigenous processes as the formation of carbonate and sulfate minerals. Calcium is the most common cation in fresh water, and the precipitation of calcium carbonate crusts on sherds is common. The durability of igneous and metamorphic rocks under atmospheric weathering depends on their mineral constituents. Minerals that formed under high temperatures and anhydrous conditions are generally not stable under atmospheric conditions. The mineral stability series shown in

this chart gives an indication of how easily igneous rocks weather at the earth’s surface: Calcic plagioclase

¯ Augite

¯ Calcic-alkalic plagioclase

¯ Hornblende

¯ Alkali-calcic plagioclase

¯ Biotite

¯ Alkalic plagioclase ¯

Olivine

¯

are oxidation, reduction, hydration, hydrolysis, carbonation, and solution. Rock weathering leads to gradual and gradational alterations that are marked by deterioration in the mechanical and durability properties of the rock. The deterioration of stone artifacts and monuments proceeds from: chemical attack and dissolution; mechanical disintegration caused by water freezing in pores and cracks; abrasion from winddriven particles; exfoliation from rapid heating and cooling; disintegration because of the activities of organisms; crystal-formation on surfaces; and damage from poor conservation or restoration procedures.

Potash feldspar Muscovite Quartz Minerals at the bottom of the chart are most stable under the atmospheric conditions at the earth’s surface. Mafic igneous rocks (gabbro, basalt) are composed of the mineral constituents near the top of the chart; as a result they break down when exposed to the atmosphere. Felsic igneous rocks (for example, granite) are composed of the more stable minerals near the bottom of the chart, and these rocks break down more slowly. After burial an object enters a new environment with which it will eventually come to equilibrium. When it is excavated, the object is exposed to yet another set of physical, chemical, and biologic contexts that may cause it to change quickly to a new stable state—a process that may destroy the object. How much alteration is brought about by the processes of burial or excavation depends on the structure and composition of the material and the severity of the contrast between the old and new environments. Bone survives fairly well in a neutral or somewhat alkaline earth matrix but disintegrates in acidic soils. The oxidizing potential, acidity (pH), moisture, and soluble-salt content are all environmental parameters affecting the stability of an object. Masonry subjected to the presence of chlorides, sulfates, and nitrates of calcium, magnesium, and sodium suffers serious damage because these salts, by repeated solution, crystallization, and hydration, generate sufficient pressure to cause fragmentation and spalling. Although these salts commonly occur as efflorescences on buildings in all climatic conditions, they are far

251

Construction, Destruction, Preservation

more deleterious in arid regions, where the lack of precipitation results in their accumulation. (In wetter climates they are washed away by rain.) Nighttime condensation on stone surfaces in arid regions followed by its evaporation at sunrise leaves behind a small but significant amount of the harmful salts, which crystallize as the evaporation proceeds. In damp conditions, underfired or low-fired earthenware will gradually rehydrate to clay, which crumbles. This is especially true where the fabric is coarse and porous. Crumbling will be exacerbated in acidic conditions by loss of calcite or other carbonate components. High-fired ceramics are reasonably stable under most burial conditions, although even well-fired ceramics may become softened under alkaline conditions by dissolution of the glassy phase. Deposits of both soluble and insoluble salts readily form on buried ceramics. Porous pottery is prone to staining, particularly by iron oxides. Iron oxide encrustation may form if there is a high, localized pH caused by calcium carbonate within the pottery. Such soluble salts as chlorides, sulfates, and carbonates are ubiquitous in groundwater and are absorbed by any porous object in the ground. After excavation, moisture in an object will begin to evaporate, and dissolved salts will crystallize. The wall paintings of the tomb of Nefertari in the Valley of the Queens at the Theban necropolis in Egypt present a classic example of this geoarchaeological conservation problem. The tomb was hewn to a depth of about 12 m in poorquality, fractured, clayey limestone. Layers of plaster were placed on the tomb walls. The exquisite tomb paintings were then painted over the plaster. Salts, especially gypsum and halite, have crystallized behind the plaster layer, pushing it outward. These dissolved salts were brought to the site by groundwater and seeping surface water during rainfall. A significant part of the conservation program must be control of microdrainages to the tomb.8 In Venice, building stone is subject to the severe saltwater environment from the marine lagoon (exacerbated by modern air pollution). Only one of the materials, Istrian stone, used in building the city has successfully resisted rapid deterioration.

252

Istrian stone is a compact microcrystalline limestone (CaCO 3 ) with few natural planes where dissolution can proceed. Exposure to atmospheric sulfuric acid causes the formation of white gypsum powder on its surface, but its low porosity and impermeability protect it from rapid deterioration. The famous Carrara marble (CaCO 3 ), used throughout the eastern Mediterranean since ancient times, has been especially beleaguered by the corrosive environment. Differential thermal expansion and contraction of the megascopic calcite crystals, which are oriented in multiple directions in the marble, cause microcracks along the edges of the crystals and allow pollutants to penetrate.

Pollution Vehicle emissions and road salt used for de-icing are ever-present preservation problems in some regions. Strontium (Sr) and lead (Pb) isotopes have been used to study the impact of pollution on a 4,500-year-old rock carving in Oslo, Norway.9 Harmful emissions of sulfur and nitrogen compounds are global in scale. De-icing salt can contribute to the rapid deterioration of porous stone. Unfortunately, geoarchaeological remedies are scarce. Many harmful pollutants are of fairly recent origin while others may have been around for a long time. Lead, for example, extends back at least to the inception of lead smelting over 3,000 years ago. The city of York in northern England witnessed lead pollution from Roman through medieval times as shown by geochemical analyses of local alluvial deposits.10 Alluvial deposits from the region of most sites contain a history of natural and human chemical debris. Geochemical analyses can reconstruct the history of mining and manufacturing even where macroscopic evidence of these industries is lacking. All materials expand and contract with temperature change. These changes in dimensions are not in themselves harmful. It is the combination of dissimilar expansions, or contractions where two different materials are combined, that produces mechanical disintegration. For example, lime mortar has a coefficient of linear thermal expansion that is approximately 50 percent greater than that for bricks. In making ceramics, it is nec-

Construction, Destruction, Preservation

essary to ensure that any additives, like temper, have similar coefficients of expansion to that of the clay matrix. The amount of water vapor absorbed by a porous rock depends on the relative humidity and porosity of the rock. Most rock damage occurs during the process of drying out, not during the absorption. Moisture in a porous rock causes dissolution and recrystallization of mineral salts, which results in structural damage and often discoloration. Frost damage in building stone is common where temperature variations around the freezing point cause cycles of freezing and thawing. The influence of frost action on dry stone is limited, but it can be substantial on wet stone: when water freezes, it expands by about 10 percent. When this freezing occurs in a confined or limited space, the pressures on the confining material are enormous. Building stone with high porosity is thus more vulnerable than compact stone. A related process that occurs during drying is the crystallization of soluble salts. During evaporation these salts can crystallize, with consequences similar to those of frost action. The salts can also form crusts on the exterior of the material. The most common salts occurring in walls are CaSO 4·2H 2 O (gypsum) and Na 2 SO 4 in various hydration states. Dry deposits of CaSO 4 are difficult to remove. Sulfates cause damage because they exist in different hydration states. Under varying moisture conditions, one hydration state will convert to another. Transformation to more hydrated states leads to expansion and pressure on the walls of the pores in the stone. The general resistance to deterioration of sandstone building materials is affected by their chemical composition. Those which contain carbonate as the natural cementing agent are susceptible to attack by acid rain. The loss of just a small amount of the carbonate, depriving sand grains of their adhesion, results in those grains being loosened and removed. As with other rock types, moisture is a major disruptive agent as well as a means of transporting salts. Coarse-grained and porous sandstones usually withstand freezing and frost action better than fine-grained ones because water escapes more readily.

The historical buildings in Pharaonic Egypt were built largely of local sandstone. The Horus temple at Edfu was completely constructed of local sandstone, while the Abu Simbel temples were carved directly into the local sandstone. These sandstones mainly consist of quartz grains cemented by ferruginous, siliceous, carbonaceous, and clay cements. Diurnal and seasonal changes in temperature and relative humidity are the principal threats to monuments made of these rocks. The result of these temperature and humidity variations is the growth of halite and gypsum crystals, which cause cracking and structural failure. A dramatic change in environment occurred when large obelisks were taken from Egypt. Sculpted in the middle of the second millennium b.c.e., these monuments stood in Egypt for 3,000 years with little surficial damage. Then, three were exported, one to Paris in about 1840, one to London in about 1870, and one to New York in about 1890. In the more humid and polluted atmospheres of these cities, the three obelisks have suffered major deterioration. Sulfur and nitrogen acids in the atmosphere have done most of the damage. The recent increase of vehicle emissions in Cairo, however, does not augur well for the preservation of the obelisks left in Egypt. An important part of our present-day construction is chiseled in stone, and these stones are slowly weathering away. Although low rainfall has retarded the rate the limestone has weathered in the great pyramids of Egypt (now close to 5,000 years old), the recent dramatic increase in the acidity of the rain could reduce their projected survival time from 100,000 years to substantially less than 10,000 years. Their companion structure, the Sphinx, is even more threatened. In recent years the Sphinx has deteriorated rapidly. Carved from the natural limestone of the Giza Plateau, the Sphinx is composed of varying rock types with different weathering abilities. The lowest stratum of the Sphinx consists of member I hard rock of a reef (fig. 9.2). This rock has not weathered appreciably. Most of the body of the Sphinx is carved from much softer layers of member II, lying above the reef limestone. The head is carved from member III, a more durable rock

253

Construction, Destruction, Preservation

Figure 9.2 Geologic Composition of the Sphinx (After Hawass and Lehner 1994)

than member II. Without intensive conservation, the Sphinx could crumble in less than a hundred years. A combination of weathering and erosion has resulted in this deterioration of the Sphinx. It has been suggested that rainfall runoff has played a major role.11 The Sphinx faces east and much of the runoff would have discharged over the western exposures eroding the exposed limestone.

Water Water is the most aggressive weathering agent there is, and it acts as the vehicle for most chemical weathering processes. The property of water that gives it its critical role in weathering derives from its molecular structure: the two hydrogen atoms lie not on opposite sides of the oxygen to which they are linked but on the same side (fig. 9.3). This atomic arrangement makes the water molecule a dipole and allows water to dissolve many natural materials. A water molecule can electrically wedge its way between surface ions in a mineral and ‘‘float’’ the ions away. Water molecules themselves aggregate into a structure. In effect, a water molecule has four

254

electrically charged ‘‘arms’’ extending from the nucleus. Two extend from the positively charged hydrogen atoms and two from the double negativity (to balance the positive charges on the hydrogen) of the oxygen. When H 2 O molecules are packed together, as in water, each negative arm attracts a positive (hydrogen) arm in a neighboring molecule. The hydrogen atoms then join the molecules in what is called a hydrogen bond (fig. 9.4), which is very strong. No other substance absorbs or releases more heat than water. In order for water to evaporate, these strong bonds must be broken. It thus requires a great deal of energy to boil or evaporate water. Approximately five times as much heat energy is needed to change a given volume of water from liquid to vapor as is needed to raise its temperature from freezing to the boiling point. The world’s climate is made more temperate by water’s ability to soak up and store heat from the sun and then release it slowly. Without large bodies of water, the earth’s surface would be an inhospitable place: the daily variation of temperatures would be far more extreme.

Construction, Destruction, Preservation

Figure 9.4 The Hydrogen Bond In the liquid state, each water molecule establishes hydrogen bonds with its four nearest neighbors. These bonds give water many of its unique qualities.

Figure 9.3 Structure of the Water Molecule The hemisphere with the two hydrogens carries a net positive charge; the opposite hemisphere carries a net negative charge.

Most materials contract when cooled. Over a wide range of temperatures, water is no exception. However, close to its freezing point water behaves very differently. As the temperature falls below 4°C, water expands. A rapid expansion occurs at the freezing point during the transformation to ice, which is about 9 percent greater in volume than the water that froze. The consequences of this abnormality control many geologic phenomena at the earth’s surface. As we have seen, a common change is mechanical breakdown of rocks because of frost action. If ice were denser than water, rather than vice versa, lakes in high latitudes would freeze from the bottom up, and most would remain partially frozen throughout the summer. Climates in temperate regions would therefore be much more severe. Oxidation and reduction reactions in groundwater are geochemically important in determining the preservation of buried materials. In many reactions micro-organisms are involved. The presence or absence of free oxygen in groundwater essentially determines whether oxidizing or reducing conditions will prevail. Dissolved oxygen in atmospheric, surface, and underground

waters acts as a powerful weathering agent. Oxidation processes proceed more rapidly in warm climates than in cold and more rapidly in humid or alternating humid and dry climates than in arid climates. In the past few hundred years humaninduced pollution in atmospheric, surface, and underground water has had serious consequences for buried materials. Even simple cultivation of the soil stimulates microbial activity, which raises the carbon dioxide content and causes additional carbonic acid to form. Water moves freely through the soil and sediments because of gravity flow, capillary action, and, in some situations, osmotic pressure. Groundwater chemistry is the chief agent in dissolution, diagenesis, and such changes as the uptake of uranium in bone. Dissolution and precipitation are controlled chiefly by the hydrogen-ion concentration (pH) and the oxidation-reduction potential (Eh). Bone can last more than 100,000 years or longer in some burial contexts but will survive only 10,000 years when the groundwater is neutral (has a pH of 7) and only 100 years or less if the groundwater-soil matrix has an acidic pH of 5. Archaeological excavations are expected to be dry enough to maintain the stability of the baulks and to protect the important finds and features. Occasionally, water flowing above the water table saturates the sediment and soil. This condition obstructs recording and leads to instability. In most cases the water flow is seasonal and may

255

Construction, Destruction, Preservation

force the excavation to be postponed. If the sediment or soil is permeable, pumping may be fruitless. Many buried sites lie below the water table. In small excavations, in low-to-medium permeability materials, adequate drainage may be achieved with a sump pump. For larger excavations, below the water table in permeable materials, the water table will have to be lowered. This is achieved by drilling to a level below the lowest level of the projected excavation and pumping from each bore hole. The pumping must continue until the excavation is completed. When acid groundwater is able to flow, the process of dissolution of carbonate rock is accelerated. A spring at Silver Springs, Florida, flows at the rate of 15 m 3 per second, carrying 274 ppm of dissolved solid. This represents about 400 tons of dissolved rock per day. The result of this dissolution is the formation of caves and sinkholes. Caves have played a major role in human evolution (see Chapter 3). Caves in northern France along the Seine River were formed in chalk strata and have been used by humans for millennia. One of the greatest of the Norman chiefs made his home there in a group of caves, one of which contained a room more than 100 m long. The Romans used the great caves at Pommeroy Park to quarry building stone. Today these caves provide storage for the maturing of champagne.

Erosion and Subsidence in Archaeology Erosion In the past four billion years whole mountain chains have come and gone. Mountains are formed largely by tectonic forces; today, all but the cores of very ancient mountain systems have been obliterated by the powerful forces of weathering and erosion. During the Quaternary, continental and mountain glaciers have come and gone with dramatic effects on weathering, erosion, and climates. Rivers, meandering across broad floodplains, move down, up, and—primarily—laterally over time. Thus, although they provide the resources for development of habitation sites, they often turn destructive and devour these sites with lateral erosion (see Chapter 3).

256

Land Subsidence Ground subsidence that is not part of the main earth movement (along a fault) may accompany earthquakes. The likelihood of earthquakeinduced subsidence is governed by the geology of the surficial rocks and soils. Subsidence and elevation occur with the loading and unloading of the earth’s upper crust by glaciers, deltaic deposits, and marine transgression. Dramatic subsidence may occur in carbonate rock terrain through the development of sinkholes. These phenomena are not tied directly to human activity, but geoarchaeologists need to understand these kinds of phenomena since they may be encountered in the field assessment of an archaeological problem. In the past 150 years serious land subsidences have been created by the removal of oil and water from the ground and from underground mining. Almost 2 m of ground subsided in Santa Clara Valley south of San Francisco as a result of withdrawal of groundwater. Historic archaeologists should also note that falling groundwater tables have led to the decay of wooden pilings for buildings in Boston, San Francisco, and Milwaukee. In addition, the load-bearing capacity of clayey sediments, soils, and fill is changed appreciably by alteration of their moisture content. The troubles experienced with the foundations of St. Paul’s Cathedral in London appear to be related to the nearby excavation for a deep sewer that began in 1831. The Tower of London rises and falls with the tide in the Thames River. Subsidence of more than 3 m has occurred in the city of Long Beach, California, because of oil wells. Subsidence has long been recognized as the aftermath of mining in soft ground. Most subsidence problems in urban areas stem from poor geologic conditions for bearing the heavy loads that buildings impose. Subsoil conditions in Mexico City consist of 50 m of saturated sandy clays with interbedded sand layers that are dangerous in earthquakes and provide little support for heavy buildings. The Palace of Fine Arts there, completed in 1934, has already sunk more than 3 m. Such subsidence problems plague many famous archaeological sites. Venice has been sinking for centuries, and the Leaning Tower of Pisa was closed to tourists recently because of its in-

Construction, Destruction, Preservation

ago, located in low-lying coastal areas of the Nile Delta in Egypt, were unprotected against floods, earthquakes, tsunamis, and consequent subsidence. These sites were discovered recently offshore at Mediterranean water depths of 5–7 m. Structures had been built directly on poorly consolidated sediment—prone to geohazards. Gradual subsidence was due in part to eustatic sea level rise and sediment compaction but also due to lateral sediment displacement, during floods and earthquakes, of the water-saturated material on which the cities were built. The founders of these cities (Leaning and Eastern Canopus) and subsequent inhabitants built and maintained these sites and their monumental structures on unstable wetland sediment. Modern Venice, as discussed elsewhere, faces comparable hazards.

Geologic ‘‘Catastrophes’’ and the Human Past Earthquakes and Seismic Disturbance

Figure 9.5 The Leaning Tower of Pisa Perhaps the most famous example of foundation problems is the Leaning Tower of Pisa. Construction of the tower began in 1174 but was not completed until 1350. It has continued to tilt since then, with a displacement of 5 m in a total height of 55 m. Foundation strata consist of a bed of clayey sand 4 m thick underlain by 6 m of sand. Attempts over the past two hundred years to stop the progressive leaning have not only been futile; most or all have increased the rate of leaning. A thorough knowledge of both the underlying geology and the principles of soil mechanics must be applied to correct the problem. (Drawing by Elaine Nissen)

creasingly dangerous tilt, the result of differential subsidence (fig. 9.5). Geoarchaeological analysis can illustrate the consequences that result where protection measures related to coastal sites are overlooked.12 Two ancient Greek cities that thrived 2,000 years

Archaeoseismology is constrained by a lack of instrumental data before the twentieth century. The long-term seismic history of an area compiled from historical records (often imprecise) and geologic data (too coarse in both time and space) provides only a rough guide to possible earthquake effects. Intensive research in many parts of the world has resulted in a good understanding of earthquake magnitudes but it is earthquake intensities (the seismic affect on human structures) that are important for geoarchaeology. Unfortunately, research on intensities has lagged far behind that on magnitudes. Historic records chronicle earthquakes of tragic consequences for human societies. Catastrophic seismic disturbances are also recorded in geologic and archaeological strata.13 Archaeologists define chronological horizons by stratigraphic and cultural discontinuities. Stratigraphic discontinuities are sometimes defined by ‘‘destruction layers’’ that may be accompanied by a cultural change evident in ceramic or other artifact typologies. For example, in the earthquake-prone eastern Mediterranean region these layers are often attributed to seismic de-

257

Construction, Destruction, Preservation

struction. In one excavation volume for the site of Knossos, Crete, J. Evans uses subheadings that include phrases like ‘‘seismic catastrophe,’’ ‘‘seismic deposit,’’ and ‘‘fresh earthquake shock.’’ 14 However, although his observations were consistent with seismic destruction, they cannot, without a set of diagnostic criteria, confirm seismic destruction as a single cause. It is necessary to consider multiple working hypotheses. For more than a century, archaeologists working in the eastern Mediterranean have debated the role of seismicities in the well-documented destructions of the Late Bronze Age. Evidence from tectonics, geophysics, archaeology, and written records has been used to propose that an ‘‘earthquake storm’’ or major sequence of earthquakes may have occurred in the Aegean and eastern Mediterranean regions during this period.15 Another example is the prominent and disastrous earthquake sequence during the twentieth century in this region along the North Anatolian Fault in Turkey.16 From the geophysical point of view, it is difficult to attribute structural damage to seismic violence. For example, structures built on slopes underlain by shale, unconsolidated sediments, or fill can topple or come apart because of uncommonly heavy rainfall that saturates new parts of the underlying ground, causing major downslope earth movements. Careful analysis of the geologic and geophysical setting of the archaeological strata is necessary before any interpretation of seismic events can be made. The size of an earthquake is defined in two distinct ways. Because the energy released by an earthquake is the most precise measure of its size, seismologists have adopted a related measure, the Richter magnitude scale. Major earthquakes have magnitudes of 5.5 to 8.9. Each increase of one unit in magnitude corresponds to a tenfold increase in the amplitude of the seismic waves and to roughly a thirtyfold increase in energy. A second scale of earthquake size is based on earthquake intensities, which are defined by the observed destruction. The most commonly used intensity scale is the Modified Mercalli Scale. Table 9.1 presents the characteristic effects of earthquakes of designated intensities in the Modi-

258

fied Mercalli Scale. This scale is particularly useful for interpreting seismic damage to sites in the archaeological record. Seismic effects at the earth’s surface reflect not only the strength of the seismic vibrations but also differences in the character of the local bedrock and overlying unconsolidated sediments and soil. Water-saturated alluvium can shake like jelly, causing destruction at great distances from the quake. Table 9.1 indicates that an intensity rating of X can result in large landslides. However, landslides are common in nonseismic areas as well. The poorest supporting ground is unconsolidated earth, particularly recent fill; this is the kind of ground on which many ancient sites developed. Structures contain many separate parts (walls, roofs, pillars). Often these parts are made of different materials with different vibrational characteristics. Mud brick has very different vibrational characteristics from cut stone. In an earthquake of intensity VII, mud-brick superstructures on top of stone walls or foundations will topple, with only moderate damage to the stone structure. Mud-brick and undressed-stone construction with no mortar or with mud mortar are the materials least resistant to seismic events. Mud brick and adobe dwellings will usually be destroyed at intensity VIII. Wooden structures will flex under the stress of strong earth vibrations. Low, rigid masonry structures can withstand strong vibrations if the structure moves with the ground as a single unit. The most common rock materials at the surface of the earth are unconsolidated sediment and soil. These materials provide the base on which structures are built: ancient societies did not carry their foundations to bedrock. The ability of unconsolidated sediments to bear a load, maintain a slope, or transmit a stress varies widely with mineral composition, grain-size distribution, water content, density, and compaction. The clay minerals in these sediments have varying properties, depending on their chemical composition and structure, that dictate their response to seismic phenomena. For example, montmorillonite clays swell because they absorb water readily, which affects the lubrication of the sediment, and wet sediments, especially saturated sediments, are

Construction, Destruction, Preservation

Table 9.1 Modified Mercalli Earthquake Intensity Scale with Approximate Richter Magnitude

Intensity

Characteristic effects

Approximate Richter Magnitude

I

Detected only by seismographs.

.–.

II

Delicately suspended objects may swing.

.–.

III

Standing automobiles may rock slightly. Hanging objects swing. Vibrations resemble those caused by the passing of a light truck. Duration can be estimated.

.–.

IV

Vibrations resemble those caused by the passing of a heavy truck or by a heavy object striking the building. Walls, windows, and doors creak. Hanging objects swing, and standing automobiles rock noticeably.

.–.

V

Some windows broken; some cracked plaster. Unstable objects overturned. Liquids may be spilled. Doors swing, pictures move, motion of tall objects may be noticed. Pendulum clocks may stop or change rate.

.–.

VI

People walk unsteadily. Objects fall off shelves, and pictures off walls. Windows and glassware broken. Some heavy furniture moves. A few instances of fallen plaster or damaged chimneys. Overall damage slight.

. –.

VII

Difficult to stand. Furniture broken. Poorly built structures damaged. Weak chimneys break at roof line. Waves form on ponds. Sand and gravel banks cave in. Damage slight in well-constructed buildings.

.–.

VIII

Difficult to steer automobiles. Considerable damage in ordinary, substantial buildings, partial collapse, great damage to poorly built structures. Some masonry walls fall. Fall of chimneys, factory stacks, monuments, towers. Heavy furniture overturned. Branches broken from trees. Wells change water level. Cracking in wet ground and on steep slopes.

.–.

IX

Poor masonry destroyed; good masonry damaged seriously. Foundations damaged generally. Buildings shifted off foundations. Reservoirs seriously damaged. Conspicuous cracks in ground. In areas of loose sediment, sand, mud and water ejected. Underground pipes broken.

.–.

X

Most masonry and frame structures destroyed. Foundations and some bridges destroyed. Serious damage to dams, dikes, embankments Large landslides occur. Water splashes over banks of rivers, lakes and canals. Flat areas of mud and sand shift horizontally.

.–.

XI

Few, if any masonry structures remain standing. Bridges destroyed. Broad fissures in ground. Extensive landslides on slopes. Underground pipes completely out of service.

.– .

XII

Damage to humanmade structures nearly total. Waves seen on ground surface. Large rock masses displaced. Lines of sight and level distorted. Objects thrown into air.

. or greater

259

Construction, Destruction, Preservation

more intensely affected by seismic phenomena than dry sediments. Bedrock geology, superficial deposits, and the nature of the soil all affect the intensity of the shaking and the resulting damage. Bruce Bolt illustrates the close relation between rock type and the intensity of the 1906 San Francisco earthquake.17 Harder rock underlay areas of small damage, whereas high damage occurred on filled lands and unconsolidated sediments. Unfortunately, standard geologic maps do not contain sufficient data on the nature and depth of unconsolidated sedimentary deposits. Data are needed on thickness, bedding, bulk density, cohesiveness, porosity, texture, and water content. Major seismic events leave clear impressions in the surficial geologic record. The 1964 Good Friday earthquake in Alaska left secondary structures as indelible marks on the sediments of the Copper River Delta.18 A relatively dense pattern of geologic structures, including sand dikes, sand pipes, slumps, faults, and joints, was formed in the sediments. These structures sometimes terminated in an unconformity. Earthquake-generated seiches planed off the upper 2 m of tidal flats; this led to deposits of clam shells, which resulted from the instantaneous destruction of the clam’s habitat. Unless eroded, such evidence remains in the geologic record. The problem for geoarchaeology is to achieve a sufficiently extensive cross-sectional view of the regional picture to reconstruct the detailed sequence of geologic events responsible for the stratigraphy. Recent tectonic movements can often be decoded from a study of drill core data. Regional subsidence or uplift and small faults showing stratigraphic offsets will show up in a systematic drilling program. Geomorphologic indicators are likely to be present where regional uplift has occurred. In coastal areas, elevated beach rock and lagoon or marine deposits can indicate the vertical displacement. J. Sims has shown that the seismic history of an artificial lake can be correlated with deformational structures in the lake sediments.19 His articles contain good illustrations of seismicinduced sedimentary structures. R. Doig interprets silt layers in organic-rich lake sediments as representing five historic earthquakes dating

260

from a.d. 1638 to 1925 in eastern Canada.20 The silt layers were presumably caused by landslides on tributary streams and resuspension of the sediment. Archaeological excavations and recorded earthquake history along the Dead Sea fault running between Israel and Jordan provide an almost continuous record for more than 2,000 years. T. Niemi and Z. Ben-Avraham have found evidence for earthquakes in Jericho from slumped sediments of the Jordan River Delta in the Dead Sea.21 They used seismic-reflection data to show that a long-term record of ancient earthquakes in Jericho can be found in the sedimentary record. Geoarchaeologists working to reconstruct the seismic history of a site or region must often turn to analogous offsite sedimentary records. The strongest earthquake in the contiguous United States during historic times occurred near New Madrid, Missouri. During the winter of a.d. 1811–1812, there was a series of four quakes with estimated surface-wave magnitudes greater than eight. Modified Mercalli intensities near the epicenter ranged from X to XII, indicating almost complete destruction of structures.22 Because of the low attenuation of seismic waves in the central United States, these earthquakes were felt over an area of 5,000,000 km 2. Roughly 50,000 km 2 were affected by ground failure, including fissures, sandblows, landslides, and subsidence. Liquefaction of subsurface sand deposits ejected sand, water, and other materials through fissures, some of which were kilometers in length and tens of meters wide. Massive bank failures along the Mississippi River sent large tracts of land into the river channel. Although the river quickly eroded and obliterated these soft sediments, the seismic events left a huge imprint on the local geology. Roger Saucier has detailed the geoarchaeological evidence for strong prehistoric earthquakes in the New Madrid seismic zone.23 Archaeological excavations at the Towosahgy State Archaeological Site (23M12) revealed a village dating to a.d. 400–1500 constructed on a Late Holocene natural levee ridge that overlay sandy point-bar deposits on the inner side of a large abandoned channel of the Mississippi River. The site lies on the northeastern periphery of the seismic-

Construction, Destruction, Preservation

induced blows and fissures. Excavation showed that although part of this site was occupied, an earthquake-induced fissure of 18 cm in width had formed and broken through a midden deposit. That the site continued to be occupied after the earthquake can be deduced from trash pits dug into the sandblow. Evidence at the Towosahgy site indicates that a seismic event capable of producing liquefaction occurred less than a hundred years before a.d. 539, based on a calibrated radiocarbon date. The largest earthquake felt in historic Europe was the one that destroyed Lisbon in November 1755. The estimated 8.5-magnitude earthquake occurred along a tectonic plate boundary in the Atlantic Ocean southwest of Lisbon. Intensity maps made based on the available data show the geographic distribution of such effects as changes in wells, springs, and rivers; surface cracks; and liquefaction landslides.24 Ground effects do not always provide a reliable measure of the severity of shaking. These studies support the proposition that seismic intensity, as a measure of earthquake damage, is different for buildings of different frequency response. This earthquake killed tens of thousands in Lisbon, southwest Spain, and Morocco. It provided raw material for Voltaire’s Candide. Holocene archaeological deposits usually provide more material suitable for dating (radiometric or typological) than geologic deposits. Holocene tectonic activity in West Africa has been dated by archaeological methods.25 Movements on a local fault amounted to more than 10 m of vertical displacement in the past 3,000 years. The time of the major offsetting was dated by the discovery of an inscription carved in hard quartzite at a depth of 10 m. (The use of the archaeological record to infer tectonic or seismic processes is an example of archaeological geology.) One of the effects of major earthquakes is the disruption and alteration of the groundwater system of springs. W. Hough records tribal movements in the American southwest as a response to the suppression of old springs and the generation of new ones.26 No greater environmental misfortune can befall a population than the loss of its water supply.

In regions where serious earthquake damage occurs at least once a century, archaeologists take seismic effects into consideration in their interpretations of site history and destruction. In regions lacking historical earthquake data, seismic impacts, including tsunamis, may not have been considered. Geoarchaeologists investigating the geologic framework of a site or region should seek both published tectonic research and field evidence for any significant seismicity. Three broad periods of seismic activity in the thirteenth and fifteenth centuries and in the 1750s to 1850s had significant impacts on human settlements.27 In the fifteenth century coastal settlements were abandoned and people moved to inland areas. The seismic activity and resulting tsunamis resulted in a shift in settlement location from sheltered coastal bays to exposed headlands. In September 1601 a strong earthquake (circa 6.2 on the Richter scale) hit the region of Lake Lucerne in Switzerland. The Alpine region has been a major tectonic region for millions of years. Historical records in this region do not record the long history of seismic activity. To assess the longterm seismic history of the region M. Schnellmann and associates used seismic reflection profiles and sediment cores from Lake Lucerne to reveal millennia of severe seismic disturbance, including severe tsunamis on the lake.28 The sanctuary of Hercules built during the fourth century b.c.e. in the southern Apennines in Italy has sustained damage from historic activity associated with the Matese fault system. P. Galli and F. Galadini,29 using archaeoseismic and paleoseismic analyses, revealed offsets of the sanctuary walls and foundation deposits during seismic events in the third century b.c.e. (unknown from historical records and seismic catalogs) and the catastrophic a.d. 1456 and 1805 earthquakes. A good example of historic study coupled with field observation can be seen in the work of N. Ambraseys and C. Melville.30 They contend that landslides, rockfalls, soil failures, and faulting are often of limited value in assessing seismic intensity. Destruction of a village built on a slope frequently occurs from ground deformation unaided by seismic activity. They note that

261

Construction, Destruction, Preservation

many adobe houses and public buildings collapse every year without the assistance of an earthquake. They also record that the earthquake of 26 February 1894 in Shiraz, Iran, caused no damage but that the heavy rainfall that followed destroyed two thousand houses. Geologic and archaeologic evidence as well as native oral traditions suggest that magnitude-8 earthquakes have occurred at least six times in the Late Holocene in the Pacific coastal areas of the state of Washington and the province of British Columbia. Some of these resulted in the abandonment of villages. In addition, deposits of probable tsunami origin are interbedded with cultural strata at several sites.31 It has been hypothesized that the lack of evidence of prehistoric human occupation along the Pacific coast of southern Washington may be explained partly by the geologic evidence for coastal submergence during prehistoric earthquake episodes.32 Seismic activity around a.d. 1700 caused submergence and tsunami inundation with consequent burial and flooding of low-lying coastal sites. Farther south, along the northern Pacific coast of Oregon, there are rapidly buried coastal marshes.33 Coring has revealed several episodes of marsh burial during the Late Holocene. Data from archaeological sites on estuaries were studied. Estuaries rapidly reflect changes in the microenvironment, especially in mollusk and fish populations likely to be preserved in archaeological contexts. Archaeological data from three estuarine sites suggested land subsidence. At two sites, human occupation continued through the period of these catastrophic coastal earthquakeinduced events. Assessing seismic damage in archaeological contexts is difficult. Field evidence cited in support of ancient seismicity shows that individual features are difficult to distinguish from the features of damage that result from poor construction and adverse geotechnical effects. Tilting and other severe distortions of walls are frequently cited by archaeologists as evidence of archaeoseismic damage. But such damage can also be caused by ground conditions beneath the wall, including ground stresses imposed by the construction of the wall or by earth movements unrelated

262

to earthquakes. Several studies provide criteria for evaluating earthquake effects at archaeological sites.34 Archaeoseismology is a good example of the necessity for integrating many disciplines to unravel the complex evidence for destruction at some sites. The basic archaeological evidence must be supplemented by the application of earth science (seismology and tectonics), civil engineering, architecture, and history. Each of these disciplines has its own language and favored methodologies. Just as the now-maturing discipline of geoarchaeology grew out of the need to apply both archaeology and geology to the investigation of sites and their environmental context, archaeoseismology (a subset of archaeological geology) now appears to be a developing discipline in which practitioners can integrate the data from many diverse fields. Thus seismic data can be used to understand aspects of past human society (geoarchaeology) or strictly geologic processes (archaeological geology). Much conjecture about seismic damage may be correct. But a much firmer evidential base is necessary to consider such conjecture likely. Much of this can come from studies of the effects of modern tremors on structures similar to those that existed in ancient times. For best results, detailed intensity maps must be constructed for the local area. Local geology and geomorphology will affect intensity, so specialized geologic maps should also be made where appropriate.

Floods and Flood Legends Floodplains are where the world’s best agricultural land, and the large populations that go with it, are found. Floodplains often develop in the lower reaches of large river systems as rivers meander back and forth and periodically overflow their banks, depositing fertile sediment. These floodplains are extensive and flat, which allows flooding over a very wide area. Each time a river overflows its banks, the current velocity decreases at the channel margin, and the coarsest fraction of the sediment load is deposited there (see Chapter 3). Over time, this builds up to create a natural levee. Beyond the levee the ground slopes down. When rivers are also depositing sediment in the

Construction, Destruction, Preservation

channel, increasing the elevation of the river, this can be catastrophic. When large floods occur, the river, which has been flowing well above the elevation of the surrounding floodplain, destroys the levee and flows unhindered across the land. When the flood abates, a new channel forms in a different location. The Yellow River flows from the easily eroded loess plateau, carrying vast quantities of silt and clay onto the North China Plain. Its riverbed has been elevated as much as 15 m above the floodplain, rising up to 5 cm a year. This great river has had catastrophic floods at least once every two hundred years during recorded history, resulting in millions of deaths and the destruction of thousands of villages and cities. A new course is established after each major flood. Sometimes the Yellow River flows northeast to the Bohai Sea, at other times southeast to the South China Sea. The distance from the farthest north to the farthest south of its mouths is nearly 500 km. The archaeology of the North China Plain is dominated by the deposition and the destruction from the Yellow River. Of all the myths in the world none seems to have attracted more attention than that of Noah’s flood.35 The development of geology in the early nineteenth century played a critical role in our recognition of the lack of historicity in this legend. By the late nineteenth century, scholars had translated a cuneiform tablet excavated at Nineveh. On this tablet was an account of the flood that was written much earlier than the biblical account. This flood took place in Mesopotamia, not ancient Palestine. Even a quick look at the landscape shows that Mesopotamia is a flat land dominated by two great river systems— a land subject to flooding—whereas mountainous Palestine would require more water than was around to incur a deluge of the biblical proportions. Prehistoric flooding is easily identified in the geologic record and should be clear in most archaeological sediments. Yet no less an archaeological light than Leonard Woolley was led astray. In reporting on his excavations in Mesopotamia at Ur, Woolley stated that evidence of a flood was associated with the ‘‘Flood Story of Genesis.’’ 36 One

of the crew at these excavations was Mallowan, who later published an article, ‘‘Noah’s Flood Reconsidered,’’ that backed away from Woolley’s contention and provided a worthwhile table of Mesopotamian floods discovered in archaeological sequences.37

Volcanoes Volcanoes have played a greater role in human misery than is commonly understood. The destructive capacity of volcanism can be seen in the a.d. 1883 eruption on the Indonesian island of Krakatoa. The eruption occurred after the volcano had been dormant for more than two hundred years. On 27 August 1883, two-thirds of the island (about 20 km 2) blew away, forming a caldera 250 m deep and creating three huge tsunamis that reached heights of more than 30 m. A Dutch warship was washed nearly a kilometer inland, coming to rest 10 m above sea level. Thirty-six thousand people were killed, mostly by drowning; 165 coastal villages were destroyed. Thick rafts of floating pumice, some crossing the Indian Ocean and others reaching Melanesia, were still afloat two years later. The explosion was heard 4,500 km away, and the quantity of ash was so great that for the surrounding 450 km, ‘‘day was turned to night.’’ 38 The ash circumnavigated the earth, lowering global temperatures by as much as 0.5°C in the year after the eruption. Temperatures did not return to normal until a.d. 1888. The great Mount Mazama eruption (more than 6,500 b.p.) led to the formation of Crater Lake (Oregon). The surrounding region was blanketed by more than 40 km 3 of volcanic ash from the eruption. The sandals and other artifacts of people living in that region at the time have been recovered from the ash. Tephra from the Mazama eruption (see fig. 2.2) serve as an excellent time marker in Holocene stratigraphic sequences in Washington, Oregon, Idaho, and Montana and parts of southwest Canada. In the middle of the second millennium b.c.e. the Aegean island of Thera witnessed an even larger volcanic cataclysm. The resulting caldera is 83 km 2 and 350 m deep, about five times the size of the Krakatoa caldera. The ash layer covering the remnants of the island is more than 30 m thick.

263

Construction, Destruction, Preservation

Figure 9.6 Major Volcanic Eruptions of the World (Reported by the Center for Short-Lived Phenomena, 1968–1974)

Pumice floated all over the Aegean and eastern Mediterranean Seas. Ash fell on settlements in Crete and Aegean Turkey. Geoarchaeologists have sought evidence of the destructive force of associated seismic and tsunami events but have so far been unsuccessful. F. McCoy and G. Heiken have summarized the local and regional effects of the Late Bronze Age eruption of Thera.39 Early peoples living in volcanic areas attributed the destructive eruptions of volcanoes to malign deities. The Aztecs and Mayans offered human sacrifices to volcanoes. The Hawaiian volcano goddess Pele is well known in Western lore. The legend of Pele, however, clearly shows that the Hawaiians understood the geologic fact that volcanic activity in the islands is progressively younger from northwest to southeast. The seventeenth century started with a bang. On 19 February 1600, the largest historic volcanic eruption in South America dropped 19 km 3 of ashy debris over 300,000 km 2, impacting the inhabitants of southern and west-central Peru, western Bolivia, and northern Chile. Communities within 20 km to the west of the volcano were

264

devastated. The large amount of fine ash in the global atmosphere was a major factor in making the summer of 1601 the coldest since 1400 in the northern hemisphere. The socieoeconomic consequences of this eruption on the region have been studied by S. de Silva and associates.40 They report on the loss of farmland, crops, livestock, and water resources. By 1750, fruit and cereal farming was thriving again but the wine industry was never fully recovered. Figure 9.6 shows the zones of high volcanic activity. Many of these zones coincide with areas of intense current and past human settlement. Hominids evolved in East Africa in the context of intensive periods of volcanism. Volcanic ash layers provide both good preservation of sites and datable Pliocene and Quaternary stratigraphic sequences. As all human societies possess adaptive mechanisms for coping with environmental fluctuations, Payson Sheets and B. McKee and others have used the natural hazards of volcanic destruction to provide a framework for exploring the dynamic relations between human societies and their changing environments.41

Construction, Destruction, Preservation

The East African Rift System is the most prolific area for Plio-Pleistocene hominid fossils and artifacts in the world because of the sequences of geologic events. The volcanic and tectonic activities were responsible for the formation of the rift basins and associated sedimentation that were ideal environments both for life and for preservation after death. Volcanic sediments and ash were responsible for quick burial, and thereby preservation. Volcanic rocks interbedded with the fossil horizons provide material that can be dated by radioactive elements (see Chapter 5).

Site Preservation Geoarchaeology can play a role wherever and whenever the problem involves both archaeology and earth science. This includes the preservation component of archaeology. The three main aspects of site preservation are the geologic and geotechnical, those concerned with materials science, and the architectural. Geoarchaeologists must deal with the first group, in addition to paying attention to the second. Geologic problems in site preservation include: (1) erosion, including the action of waves, meandering rivers, rising sea level, wind, and ice; (2) freshwater inundation from dam construction; (3) land subsidence because of mining or the withdrawal of water, oil, or gas; (4) landslides, mass wasting, and soil creep; (5) sediment and soil compaction; (6) earthquakes, tsunamis, and movement along faults; (7) diagenesis, bioturbation, and frost action; and (8) volcanic hazards. Geotechnical responses to ameliorate these problems include: (1) draining; (2) waterproofing; (3) chemical stabilization; (4) structural stabilization; (5) channeling; (6) riprapping; and (7) reduction in biologic activity. (There is no geotechnical response to volcanoes.) In terms of geotechnical responses geoarchaeologists should adopt their own Hippocratic oath: ‘‘First, do no harm.’’ There is a long history of attempts at building and monument preservation that did more harm than good—unfortunately, most of these well-meaning preservers had little or no geologic knowledge.

A clear example of the use of geoarchaeology in site preservation is that of Venice, Italy. Venice is sinking slowly, allowing the tides to inundate this architectural treasure with increasing frequency. In the first decade of the twentieth century St. Mark’s Square flooded less than ten times a year. By the 1980s it flooded about forty times a year and now it floods sixty times a year. The city is built in a saltwater marsh lagoon on swamp deposits and river sediments. These foundation deposits are slowly compacting. Eustatic sea level rise and the withdrawal of groundwater from industrial wells in the nearby port town of Maghera greatly exacerbate the slow sinking. Roman-period Venice lies almost 2 m below the current streets. Human concern for the situation is not new.42 In the early 1800s Lord Byron wrote in Childe Harold’s Pilgrimage, ‘‘Venice, lost and won, her thirteen hundred years of freedom done, sinks like a sea-weed, into whence she rose.’’ In large-scale preservation problems such as are occurring at Venice, successful mitigation requires a thorough understanding of both the geologic and the archaeological parameters.

Site Preservation Problems Another task falling to geoarchaeologists relates to the burial or reburial of sites (see also Chapters 2 and 3). In many countries when significant features remain in the ground, sites must be carefully reburied. A related situation arises when long-term preservation of cultural resources can be achieved by in situ preservation—without excavation. In both instances, prior or intended land use usually must be allowed to continue (for example, agriculture or highway construction). Different types of geologic materials are needed depending on what features are to be preserved, the local soil/sediment conditions, and the overall hydrologic regime. There are often local geologic materials that can be used although, increasingly, regulations may require the use of relatively inert quartz sand. Any rock product to be used must not contain damaging amounts of chlorides, carbonates, or soluble iron compounds. The need for understanding the environmental context of preservation efforts is particu-

265

Construction, Destruction, Preservation

larly acute in shipwreck preservations. Geologic, chemical, and biologic degradation processes vary widely in the aquatic environment and shipwreck sites are increasingly being managed as sites. Most shipwrecks have reached degradation equilibrium with their environments and any human-induced change is likely to be deleterious. The natural sedimentation rate, currents, wave action, and water depth are important factors when evaluating a shipwreck preservation plan. High clay content in the incoming sediments may provide a low permeability helping to preserve wood objects. The geochemical environment also needs to be understood. Well-oxygenated water leads to the rapid oxidation of iron objects. Salinity, pH, and temperature will also affect the deterioration of artifacts. River, lake, reservoir, and coastal erosion present an ever-present threat to archaeological sites. Bank erosion, its causes and effects, were the subject of a classic study by Grove Karl Gilbert in the nineteenth century.43 This pioneering geomorphologist explained the forms that lake-shore landforms take and the processes that give rise to them. The landforms studied by Gilbert are being destroyed by human activity, although there have been efforts to preserve them. Geoarchaeologists can help with site preservation in the face of natural erosion by understanding the dynamics of the erosion processes (see Chapters 2 and 3). Although the processes are well known, the rates of erosion vary dramatically with time and with landscape parameters unique to each site. Once the local geology is understood, a variety of techniques can be used to establish baselines for rates of erosion. Historical, sequential aerial photographs provide one of the most accurate records (see Chapter 4). Aerial photographs provide a record that covers a considerable length of the historic period in the United States and some other developed countries. Most of these photographs offer stereo coverage. The U.S. Geological Survey has a systematic and repetitive aerial mapping program. The U.S. Army Corps of Engineers has large-scale metric aerial photographs that were taken with a calibrated camera for engineering mapping of its project areas—projects that often affect potential erosion at archaeological sites. 266

A geologic perspective is essential for any archaeological site-protection plan. It is particularly necessary in river floodplain situations, where the river course can change in response to individual floods, and in coastal situations, where one storm can cause more beach recession than a hundred years of normal erosion. Each archaeological site presents a special set of circumstances that depend on such factors as: whether there was instantaneous abandonment rather than slow decrease of human activity (a more complex situation); the rate and depth of burial; and the age of the site. Natural geomorphic and climatic changes over time will result in some stress in the preservational aspects of archaeological remains. Once a site is abandoned, geologic processes are the primary agents responsible for site destruction or burial (see Chapter 2). These geologic processes include both the physical (compaction and cryoturbation) and chemical (weathering and a host of dissolution phenomena). Until the 1980s, little research was done on the effects of burial and diagenesis. For sites residing in the soil zone, the complex pedologic activity will have an impact on archaeological resources. Although burial may decrease weathering and leaching, sites buried beneath the active soil zone are still affected by one or more of the following: pH and changes in pH; physical movement of materials; compression; wet-dry and freeze-thaw phenomena; wet aerobic and anaerobic reactions; other diagenetic changes; and organisms. An example of special geologic conditions can be seen at Thebes in Egypt. The New Kingdom pharaohs built their tombs in the rock cliffs on the west bank of the Nile. Figure 9.7 illustrates in cross section the geologic stratigraphy and lithology of a typical tomb placement. Three marine sedimentary formations outcrop in the area: the Theban limestone, the Esna shale, and the Dakhla chalk. The Esna shale is composed of montmorillonite clay, silt, and fine sand. Most of the royal tombs were cut into the Theban limestone on a descending slope. Those tombs that descended into the Esna shale (as in fig. 9.7) encountered a geologic ‘‘time bomb.’’ The high montmorillonite content of the Esna shale makes it highly expansive when moistened. Such expansions exert tremendous pressures on rock columns and parti-

Construction, Destruction, Preservation

Figure 9.7 Generalized Cross Section of the Valley of the Kings, Egypt

tions in the tombs. Preservation efforts must concentrate on keeping the tombs dry.

Reservoirs When new reservoirs are filled, engineers frequently must deal with bank conditions where there are flat terraces, and the banks against which the waters lap have steep gradients composed of incompetent silt, sand, or gravels. When these reservoirs are filled, erosional forces are directed against new shorelines that have not previously been affected. Until equilibrium is reached, the reservoir bank may suffer extremely rapid erosion. River-bank terraces throughout the world have been preferred by prehistoric peoples for habitation sites, fishing sites, and places for related activities. There are at least thirty-four processes that influence the nature and rate of bank erosion in reservoirs.44 Ten northern U.S. reservoirs were analyzed in terms of bank-recession analysis covering periods of twenty to thirty years. Bank erosion was extremely variable even within small areas. Reces-

sion measurements along two or three transects may not adequately characterize recession for a given area. For reservoirs from the middle stretch of the Missouri River, the calculated rates of recession were between 0 and 12 m per year. From a geomorphic perspective, not only the erosional feature itself but those landscape features adjacent to it need to be part of the study and interpretation. With stereo aerial-photograph coverage, an exposed erosional bank and its immediate geomorphic context is usually apparent. However, ground-truth field investigation of bank sediment types and vegetative cover is necessary for building an adequate model for siteprotection schemes. The factors that influence reservoir-bank erosion operate in several timescales. Water levels change as often as every few days: floods can raise water levels dramatically. Climatic parameters, such as droughts, operate on a scale of years. Reservoirs are deepest near dams, whereas lakes are deepest near their centers. The tendency of sediments to erode along the extended shorelines of reservoirs and large 267

Construction, Destruction, Preservation

lakes varies. The shores of new reservoirs are in disequilibrium with their new ‘‘lacustrine’’ environment. The time they take to reach equilibrium will vary, but in some reservoirs along the Missouri River equilibrium has not been achieved in more than thirty years.

Hilltops or Slopes From the Bronze Age to medieval times, generally for defensive reasons, Mediterranean area habitation sites and fortifications were often built on flat-topped hills delimited on all sides by steep escarpments. This makes rocky hills highly susceptible to erosion. Stabilization attempts on unstable cliffs or steep slopes were carried out as far back as the Renaissance in Italy. Prominent slopes are especially sensitive to wind and water erosion. The eroded profile may show an irregular shape because of the differential resistance of clay, sand, carbonate, or igneous rock layers. Wind can be erosive on steep slopes or cliffs formed by poorly cemented sediments. Wind velocities increase with height, and sand grains entrained by the wind can be abrasive. Geotechnical preservation techniques that can be used to inhibit erosion include: filling voids with concrete under cantilevered rock, filling open joints and cracks with cement grout, treating with pigmented shotcrete, building protective walls of rock similar to those of the cliff or slope, treating rock formations with durable, water-repellent coatings like silicone compounds, consolidating friable rock, and impregnating porous rock with silicates. Major rockfalls of large blocks are usually contained by bolting or otherwise anchoring the rocks. A special case of instability arises when ancient walls stand at the edge of, or on, steep slopes. The problem is compounded by the fact that such walls are often located on fill—a notoriously poor foundation material. When rainwater accumulates inside a wall that lacks adequate drainage and saturates the foundations, a slide is likely to develop. Ensuring proper drainage will help relieve this stress. A good knowledge of rock properties and local stratigraphy is necessary for any of these techniques. Pueblo Bonito in Chaco Canyon, northwest New Mexico, was built and occupied between 268

about a.d. 900 and 1130. The early inhabitants built their village near the north canyon wall, over and between great blocks of sandstone. They may have been unaware of the significance of these fallen blocks. Later inhabitants, recognizing the instability of the nearby canyon wall, braced the cliff during Pueblo III time (a.d. 1050–1060) with timber posts embedded in stone rubble and protected by masonry. The whole structure was supported by a vast buttress of adobe and rubble. The ancient mitigation efforts were successful— the sandstone cliff that had long menaced Pueblo Bonito did not collapse until January 1941.

Seismicity Seismicity problems in geoarchaeology go beyond assessing seismic damage at archaeological sites. They extend to monument preservation. The maximum seismic hazard in high-risk areas needs to be determined. Paleoseismic maps are a starting point in assessing future seismic load, but the evaluation of seismic hazard owing to earthquakes originating from a fault requires a quantitative description of fault activity. Additional information is needed about the possible ground acceleration at a site if at-risk monuments are to be protected. Seismic risk analysis is an advanced science in countries like Japan and the United States. Hence, the data may already be available for assessing the role of paleoseismic activity and seismic hazards in archaeological site protection. Such data, combined with engineering geology investigations common in many geologic hazard areas, can provide the geoarchaeological framework for site-preservation analysis. Engineering geology studies typically reconstruct the morphologic transformations undergone by a site over many centuries or millennia. Geologic time horizons are typically much longer than those in archaeology.

Site Stabilization Although archaeological sites are susceptible to natural ‘‘aging’’ processes, erosion and related mass movements can often be predicted and mitigated. Site-stabilization efforts have a much longer history than the relatively recent archaeological interests. Thus, mature geotechni-

Construction, Destruction, Preservation

cal sciences concerned with site stabilization are available. Geomorphology, engineering geology, hydrology, soil mechanics, and related sciences all come to bear on site-protection problems. However, archaeological site protection often has a unique set of problems associated with the preservation of buried features. Although some surface sites may be protected by heavy riprap, the application of the same material to a site buried in soft sediments may result in differential compaction that all but destroys the spatial integrity of the features it was designed to preserve. Prestabilization site testing should be considered in every geoarchaeological project of site preservation. The objective of site preservation is to strike a balance between systematic recovery of data and artifacts and the long-term protection of cultural resources that can be preserved in situ.

Earth Burial Although burying archaeological sites under a mantle of earth has often been used for site protection, there are few studies that assess longterm effects. There are some basic geologic considerations involved. Among them are: Fill should be sterile and of a composition that does not inhibit vegetation. Sand fill may be unstable and easily eroded. Fine sand and silt may be subject to deflation. Heavy stone (riprap), though durable and resistant to erosion, may compact the site sediments to an unacceptable degree. The sediments and soils at the site must be investigated for soil mechanical and hydrologic properties. Similarly, a knowledge of artifacts and features to be protected is essential. Fill material must be chemically compatible with site matrix, artifacts, and biofacts. The term ‘‘chemically compatible’’ includes the acid-base character of the site, and pH levels of the fill, and the site earth matrix should be similar. Bone deteriorates rapidly in an acid environment and may be damaged by salt-crystal formation if soluble salts are available in a fill material.

Deep burial may lead to undesirable compaction and other diagenetic effects. A coarse sand or pea gravel layer may be an appropriate marker between fill and artifact-bearing matrix. When rock berms are appropriate for exposed banks of streams, lakes, and reservoirs, determination of the ground surface preparation and expedient slope angle must include the nature of the site matrix, artifacts, and features. The toes of rock berms may be subject to erosion and catastrophic failure if water levels drop below the base of the berm. Where soils are involved, easily altered soil properties like organic matter, pH, nitrogen, and soluble salts will move toward a new equilibrium. Laboratory analysis will help to characterize site soils and matrix. Shallow cores can be used to characterize the site sediments and soils and to evaluate the continuity and extent of site strata. Burial has the potential for limiting erosion, weathering, and biotic damage to archaeological sites, but increased loading, increased moisture, and incompatible chemistries may limit its usefulness. The most successful uses of site burial will be those where the original chemical characteristics of the site matrix are retained and where the deleterious effects of freeze-thaw and wetdry cycles are reduced or eliminated. The approximate relative significance of decay factors for buried sites is: Most Severe

Least Severe

Wet-dry/freeze-thaw Wet aerobic conditions Compression Micro-organisms Freezing Wet anaerobic conditions Low pH conditions Micro-organisms Movement High pH conditions Thawing

269

Construction, Destruction, Preservation

Archaeological/Cultural Resource Management Resource Management Many geoarchaeologists are, or will be, employed in archaeological or cultural resource management (CRM). Increasingly today, an earthscience component is required as a part of CRM contracts in the United States. In Minnesota and Iowa, geomorphology must be an integral part of archaeological surveys or excavations. The geology-based techniques that are usually necessary to CRM investigations are coring, sediment/soil analyses, and geophysical prospecting. Working in CRM requires geoarchaeologists to comply with laws regulating archaeology. Federal protection of archaeological resources on public lands was initiated with the Antiquities Act of 1906. The National Historic Preservation Act of 1966 is the federal law under which archaeology is now conducted in the United States. Amendments to expand or clarify the law have been enacted. This law embodies a strong statement supporting historic preservation. One basic tool of the National Historic Preservation Act is the National Register of Historic Places. The National Register is a list containing the names and locations of nationally important historic and prehistoric archaeological sites. Sites listed on the National Register receive some legal protection. Federal and state funds cannot be used to alter improperly or destroy these properties, and both federal and state governments are obligated to protect the sites under their respective jurisdictions. Current and future laws apply to national parks, forests, and tribal property. However, cultural resource legislation also affects other properties. Any state or federal projects that receive government funding or that require state or federal licenses are subject to regulation. These projects include highway construction or modification, bridges, dams, and federal construction projects. Such projects always have geologic components. The Historic Sites Act of 1935 led to the interagency federal program for recovery of archaeological and paleontologic remains and the River

270

Basin Survey program. This program began to address incidental destruction of sites, mandated cultural resource identification, and called for the mitigation of adverse impacts on these resources. This work was continued with the Reservoir Salvage Act of 1960 and the National Historic Preservation Act of 1966. These acts provided an effective tool for preserving historic and archaeological sites. Private lands are also protected if federal funds are used for modification of the land. For example, any private developer who secures federal loan money for the development of property must comply with all federal laws pertaining to archaeological resources. This includes the issuance of an environmental impact study (EIS) or an environmental assessment (EA). Environmental studies are required for certain types of projects regardless of whether they are private, public, or government. This brief introduction to cultural resources management is intended to indicate to CRM geoarchaeologists in the United States that there are special responsibilities and restrictions related to their work. Most countries now have laws designed to protect cultural resources and to regulate archaeological and geoarchaeological projects. Integrated investigations that include archeology, geophysics, geoarchaeology, paleoecology, and cultural resource management are becoming more frequent worldwide.45 Geoarchaeological analyses of landform elements are often central to the archaeological investigations and the development of cultural resource management and evaluation strategies. We firmly believe that two geoarchaeological techniques, core drilling and geophysical surveying, must be a component of nearly all CRM surveys. The geologic aspects of historic property protection differ somewhat from those of most prehistoric (buried) sites. Although lateral river migration, coastal erosion, and related destructive natural forces are common to both historic and prehistoric sites, natural processes like major floods and acid rain have a far greater impact on historic sites where there are standing structures. An elevation of the groundwater table that might serve to preserve buried sites from alternating wet

Construction, Destruction, Preservation

and dry conditions could well be damaging to a historic structure. In general, preservation of historic sites will have fewer geoarchaeological considerations than are attendant at prehistoric sites. The major geologic aspects of many historic site preservation problems revolve around the weathering of stone structures under adverse atmospheric conditions. On other sites such geotechnical mitigation procedures as berm walls, earth burial, and riprap revetment are indicated.

Conservation Materials Preservation The first commandment of conservation is: ‘‘Any preservation or restoration treatment must be preceded by an exhaustive study of the deterioration process.’’ The remedial action usually includes the following sequential steps: diagnosis; cleaning; consolidation. Cleaning and consolidation employ standardized treatments that are well known to conservators. Diagnosis is where geologic knowledge is necessary, for it cannot proceed without a detailed knowledge of the material. In terms of earth materials, this means the mineralogy and geochemistry of the material and its previous and current environments. For example, limestone and marble are exceedingly and immediately vulnerable to acids in the atmosphere—and the typical outdoor atmosphere is fairly acidic—whereas fresh granite will weather slowly. (But weather it does.) The standard carbonic acid in the atmosphere simply dissolves the calcite (the chief component in limestone and marble), while the sulfuric acid in the atmosphere turns the surface layers of the calcium carbonate to calcium sulfate, which causes the material to crumble. Marly limestone (a common building stone in Italy) also suffers from its heterogeneity. The physical process of dissolution of the carbonate component is further accelerated by the penetration of acidic solutions along the clay layers. In sandstones where the clastic particles are composed of quartz—a highly resistant mineral— the cementing agent is likely to be the vulnerable component. Calcite cement suffers from the same chemical problems as limestone and marble. If the cementing is weak, abrasion caused by particles in

wind can generate rapid mechanical disintegration. Even quarrying methods can affect stone deterioration. Ancient manual-quarrying methods will have done less mechanical damage than modern methods that rely on a pneumatic hammer whose vibrations cause microcracks. All worked surfaces acquire adhering carbonaceous particles more easily that are good catalysts for the hydration of atmospheric sulfur dioxide to sulfuric acid, promoting dissolution. An attempt to quantify the relation of shaping processes to the physical characteristics of building stone has been made of Pietraforte—a sandstone widely used as a building stone in Florence. The clastic particles of this sandstone are derived from metamorphic and sedimentary rocks. The rock is a roughly equal mixture of silicate and carbonate components. Decay of this material is principally caused by absorbed water in pores and discontinuities, combined with sudden temperature changes. The porosity appears to be affected by working processes, particularly when the worked surface is parallel or perpendicular to the stratification in the rock. Mineralogic (including X-ray) and petrographic analyses (see Chapter 7), along with chemical analysis, are necessary to characterize rocks and understand rock weathering. Petrographic analysis allows the geologist to determine porosity and permeability, two important parameters in judging the susceptibility of rocks to chemical weathering. The chemical treatments of stone undergoing conservation or restoration are beyond the scope of a book on geoarchaeology.46 What the geoarchaeologist contributes to materials conservation is the knowledge of the stability or instability of rock and other building materials in various environments. It is largely because of insufficient knowledge of materials in given chemical environments that there has been a century of failures in conservation treatment of stone sculpture, monuments, and buildings. Corrosion Fundamental to the deterioration and weathering of metals is the phenomenon of corrosion. Most metals are not stable under the conditions pre-

271

Construction, Destruction, Preservation

vailing at the earth’s surface. The resulting corrosion has two important, contradictory aspects: it leads to deterioration and disfigurement, but, depending on the metal or alloy, a thin layer of corrosion may form a protective film on the surface, inhibiting further deterioration. The characteristic phenomenon associated with weathering of copper and such high-copper alloys as bronze is the formation of a green patina on the surface of the metal. The patina consists of copper hydroxide salts of sulfate, carbonate, or chloride, as well as salts of lead and tin, where these have been alloying elements. It should be emphasized that more than a dozen minerals have been found as corrosion products on bronze. The rate of corrosion will depend upon environmental factors, especially the acidity that comes from sulfuric and carbonic acids in the atmosphere and in soils. From wet sites, including undersea locations, where oxygen is relatively scarce, the excavated copper or copper alloy is often virtually uncorroded. In reducing environments, bluishblack copper sulfides may form on the surface of objects. Copper in solution is toxic to marine life, so the presence of encrusting marine creatures probably indicates that the metal is free from active corrosion. Ancient iron is not pure iron but an alloy with about 0.1 percent carbon, called wrought iron. A later alloy was developed which we call cast iron, with a carbon content of more than 2 percent. Steel is also an alloy of iron and carbon, but it is a very different material. Iron corrodes easily to porous and inhomogeneous iron oxide rust. Highly corroded iron artifacts are fragile and require special conservation techniques. They should be stored in as dry a place as possible. Iron excavated from damp, aerated sites usually appears to be a nondescript, reddish-brown mass. The shape of the mass frequently bears no resem-

272

blance to that of the original object. The mass is composed of iron oxides, oxyhydroxides, and carbonates. Iron objects from wet, oxygen-free deposits may appear black because of the formation of ferrous iron sulfide by the action of sulfurreducing bacteria. From marine sites, ferrous concretions of iron oxides and oxyhydroxides, calcium carbonate and debris, form around the corroding iron. As these concretions form, they can spread across a site, trapping everything in their way. Where oxygen is relatively scarce, lepidocrocite (gamma FeO(OH), magnetite (Fe 3 O 4 ), and pyrrhotite (FeS) may be present. As a noble metal, gold is not affected by normal geochemical processes at or near the earth’s surface. However, in the natural and artificial alloy of gold and silver (electrum), the silver can be subject to corrosion. In gold and copper alloys, especially the Central and South American alloy tumbaga, the copper can corrode so severely that the artifact is covered with the typical green copperalteration products. Gold alloys are also subject to cracking—a phenomenon not found in pure gold. By Roman times lead was in common use in Europe. Fresh lead has a bright metallic bluishgray color. Corrosion first dulls this color then produces a gray and finally a grayish-white surface. Most excavated lead is covered with this grayish-white alteration product. This is usually cerussite (PbCO 3 ) and/or hydrocerussite (2 PbCO 3 .Pb(OH) 2 ). These corrosion products are found on lead from damp, calcareous soils and on lead from the sea. In both cases they protect the metal from greater corrosion. Because lead is so soft, the corrosion products have a higher Mohs hardness than the underlying metal. The tin-lead alloy pewter will have a similar dull, grayish-white corroded surface when excavated.

C H A P T E R 10

Epilogue

From these examples the reader may gain the impression that already everything has been done to insure greater emphasis in the geologist’s field work on prehistoric problems and to make the archaeologist more geologically minded . . . one can only hope that the digging archaeologist . . . will become conscious of the value of geology . . . to archaeology. —Helmut de Terra 1934

The Future

G

eoarchaeology, and more broadly the application of earth science to humanistic and historical areas of scholarship, seems once again to be in an expansive phase. Ancient myths have always been a central aspect of classical humanities. Earth scientists are regularly interested in geologic aspects of these myths. In Chapter 9 we referred to geologic inputs into the understanding of flood myths. We call your attention to the book Legends of the Earth: Their Geologic Origins. 1 There are many similar volumes in the literature. A recent paper reopened the question of whether the Delphi oracle inhaled narcotic vapors arising from a fissure below the temple of Apollo.2 The authors indicate that the fissure (arising from a fault) cuts through a bituminous limestone at shallow depth. Hydrocarbon gases from this formation could have escaped during and after seismic events. Such gases can induce mild narcotic effects. In another example, classical literary texts are inconsistent concerning the time and place of an earthquake in fifth-century b.c.e. Greece.3 Ar-

chaeological evidence from the region is likewise ambiguous. The authors illustrate the level of uncertainty that frequently accompanies literary and archaeological information and stress the importance of geologic origins. As more and more geoarchaeologists are trained to work within the broader aspects of history and the humanities the initially narrow scope of geoarchaeology (the antiquity of humans in the Old and New Worlds) will continue to broaden.

The Future of Geoarchaeology within Archaeology Archaeology is far more than excavating and using a background in anthropology or classics to tell a plausible story about the human past. To get the fullest possible understanding of ancient lifeways, it is necessary to study in context essentially everything humans interacted with: plants, animals, rocks and minerals, the landscape (including water bodies), and so forth. This is beyond the capacity of individual scholars and researchers, and is why modern projects in archaeology are team efforts. Unfortunately, only a small portion of any ancient human environment exists today for study. Geoarchaeology has the luxury of working with rock products that are exceptionally stable, and landscapes, though always undergoing slow change, leave recognizable records of their past. Cultures have been largely defined by their range of material products, mostly geologic in nature. Archaeological theory is much more contentious than theory in the earth sciences, perhaps because archaeological theory is not based en-

273

Epilogue

tirely on ‘‘hard evidence.’’ Archaeological theories come and go as the field of archaeology matures. One reviewer of the first edition of this book suggested that we did not take into account the changes in archaeological method and theory. It is our view that geoarchaeology is based on natural science, not on the currently dominant views within anthropological archaeology. If geoarchaeologists speculate on the behavioral aspects of ancient societies in ways that cannot be tested, they are not engaging in natural science. Geoarchaeology is a natural science concerned with what, how, and when. Archaeology often also seeks to ask why. We are fairly skeptical about answers to why questions even if we recognize their importance. To use one example from recent history, following the assassination of the American president John F. Kennedy, tens of millions of dollars were spent on the study of photos and films, there were hundreds of witnesses, hundreds of interviews, but the why questions still remain. Hence, we question speculation about the why of practices from hundreds or thousands or even millions of years ago connected with the study of humanity. Ethnoarchaeology may offer clues, but no evidence remains to test hypotheses. In his seminal article on archaeological theory and method, Old World prehistorian C. Hawkes suggests four levels of difficulty in making inferences from human material culture: material techniques are easy to infer; subsistence-economics fairly easy; communal organization harder; and spiritual life the hardest of all.4 Hawkes further suggests that when you move beyond subsistence-economic inferences your archaeological rewards diminish sharply. An adjunct to a discussion of geoarchaeology within archaeology is the question of archaeology within anthropology. Archaeologists in the United States have spent five decades discussing whether archaeology is anthropology and, particularly in the last two decades, whether archaeology should be housed in academic anthropology departments or be a free-standing academic department.5 In many parts of the world, particularly in Britain, where there are archaeometry/archaeological science departments, archaeology is in free-standing departments

274

where geoarchaeologists, archaeometallurgists, and other archaeological scientists are a basic component of the teaching and research program. Our view is simply and firmly that not all of archaeology is anthropology so archaeology should be taught as the broad-based discipline that it surely is today. Why should archaeologists study geoarchaeology? The answer is simply because human social frameworks and their natural environments have co-evolved through time. A thorough understanding of culture and culture change is not possible without an appreciation of the environmental context. Archaeologists are trained to recover data through carefully orchestrated procedures of excavation that destroy the site being examined. For many reasons, including costs, the twentyfirst century is likely to see a shift toward the in situ preservation of sites and the preservation of entire cultural and natural landscapes. Geologists and geoarchaeologists frequently focus on larger areas than sites and site-catchment areas and therefore employ different methodologies than archaeologists. This focus also colors their research and preservation strategies. Recent archaeological survey research has tended more toward the recording of ‘‘find spots’’ and other ‘‘nonsite’’ data. Humans wandered and worked across, and made an impact on, a continuous landscape. It is with the continuous space-time landscape that geoarchaeologists must deal. The concentration of archaeological efforts at ‘‘sites’’ ensures that these areas remain important, but the continuum rather than the isolation should attain greater attention in the future. Intensive multidisciplinary surveys should play an everincreasing role in twenty-first-century archaeology. Survey methods need to be improved with more attention paid to high-resolution satellite images with wide spectral range and to the integrating power of GIS. To better understand the cultural landscapes additional attention needs to be paid to natural landscape transformations. Somewhat the same thing can be said for the approach to chronology. Archaeologists have always focused on the specific strata in archaeological sediments where artifacts and features occur. They considered the time interval be-

Epilogue

tween the deposition of these strata as dead time and the deposits as ‘‘sterile,’’ largely out of bounds for archaeological work. Geoarchaeologists need to see time as a continuum. Life went on whether the deposits recorded it or not. Landscape changes, the climate, gross impacts of human activities, soil development, change in vegetative cover—all are time-continuous features in our earth’s surface environment. Paleoecologists have added much to archaeological thought with their careful reconstruction of environmental change. Often data from paleoecology or geomorphology are at a much coarser scale than desired by site-oriented archaeologists. This can and is being remedied with multidisciplinary, archaeology-centered projects in specific areas. The regional orientation of geomorphology and geoarchaeology is well suited to these future needs. Archaeologists use the word landscape somewhat differently than we do in this book. Archaeological usage considers the geomorphic landscape but only in direct relations to specific cultural landscapes.6

Because archaeological site preservation is largely a post–World War II phenomenon, and because few follow-up studies have been undertaken concerning the effectiveness of site-preservation methodologies, site preservation is a fertile field for future research. Short-term protection of a resource, though often expedient, does little to ensure long-term preservation. Until we evaluate long-term effectiveness, we cannot begin the appropriate cultural-resource management planning. With their broad training in archaeological science, geoarchaeologists can play a major role in the preservation, monitoring, and evaluation of our past. We said it before, we say it again: progress in archaeology and geoarchaeology will be increasingly dependent on technological breakthroughs in, for example, molecular biology, geophysics, and materials analysis. We end with an ageless comment from Confucius: Tell me about the past, and I will know the future.

275

Notes

Chapter 1. Theory and History

22. Daniel 1976. 23. Harris 1989.

1. Konigsson 1989. 2. Ferring 1994. 3. Gladfelter 1981; Butzer 1982. 4. French 2003. 5. De Terra 1934. 6. We refer to the contextual archaeology of Butzer 1982, not to be confused with the postprocessual contextual archaeology described by Willey and Sabloff, History of American Archaeology, 306, or Trigger 1990, 348; Mandel, ed., 2000.

24. Frere 1800. 25. C. Lyell 1863; Breuil 1945. 26. Prestwich 1860. 27. K. Lyell 1881. 28. Prestwich 1860; Gruber 1965. 29. Lyell 1863. 30. Cohen 1998. 31. Van Riper 1993. 32. Lyell 1863; Lubbock 1865; Geikie 1877.

7. Graubau 1960.

33. Daniel 1976.

8. West 1982.

34. Squier and Davis 1848.

9. Waters 1999.

35. Trigger 1990.

10. Thorson 1990.

36. Meltzer 1983.

11. For example, by Binford 1989.

37. Foster and Whitney 1850; Whittlesey 1852.

12. Leach 1992.

38. Meltzer 1983; Gifford and Rapp 1985a.

13. Renfrew 1976; Davidson and Shackley 1976; Gladfelter 1977; Butzer 1982; Schoenwetter 1981; Rapp and Gifford 1982; Rapp 1987a, 1987b; Goldberg et al. 2001.

39. Worster 2001.

14. Butzer 1982; see also Butzer 1964, 1975, 1978, 1985.

40. A brief biography and the republication of some of Holmes’s major works in lithic and ceramic studies can be found in Meltzer and Dunnell 1992.

15. Wheeler 1954.

41. Holmes 1892.

16. Gladfelter 1977; Waters 1992.

42. Powell 1890.

17. Cremeens and Hart 2003.

43. Abbott 1892.

18. Hassan 1995.

44. Haynes 1893.

19. Daniel 1976; Butzer 1964, 1982; Rapp and Gifford 1982; Gifford and Rapp 1985a, 1985b; Rapp 1987b; Grayson 1986a, 1986b, 1990; Meltzer 1983; Stein and Farrand 1985; Stein 1987; Daniel and Renfrew 1988.

45. Salisbury 1893.

20. Schiffer 1976; Stein and Farrand 1985.

48. Pumpelly et al. 1905; Pumpelly 1908.

21. Gifford and Rapp 1985a; succinct reviews of the history of interaction between geology and archaeology are provided by Rapp 1987a, 1987b.

49. Daniel 1976.

46. Wright 1892. 47. Meltzer 1983.

50. Stein and Farrand 1985. 51. Zeuner 1946; Zeuner 1959.

277

Notes to Pages 11–29 52. Daniel and Renfrew 1988.

80. Binford 1964.

53. De Morgan 1924.

81. Isaac 1967.

54. De Terra 1934.

82. Acher 1968.

55. See Capper, Archaeologia 60, plates 69 and 70, for photographs of Stonehenge from a military balloon; Beazeley 1920; Crawford 1923, 1929; Judd 1930.

83. McDonald and Rapp 1972.

56. Atkinson 1963; see also for South America, Shippee 1932, and other studies by the Shippee-Johnson Peruvian expedition. 57. For a detailed history of the Zhoukoudian excavations and of ‘‘Peking Man’’ see Jia and Huang 1990.

84. Wendorf 1969. 85. Caldwell 1959. 86. Binford 1964, 1977, 1981, 1983; Schiffer 1972, 1976, 1983, 1987. 87. Binford 1983, 1981. 88. Raab and Goodyear 1984.

58. See, e.g., Piggott 1965.

89. Hassan 1978, 1979.

59. Gifford and Rapp 1985a; Meltzer 1985.

90. Butzer 1982; Gladfelter 1981; Waters 1992.

60. Kidder 1915; Meltzer 1985; O’Brien and Lyman 1999; Broman and Givens 1996.

91. Fedele 1976.

61. R. E. Adams 1960.

93. Stein and Farrand 1985; Schick 1986; Dincauze 1987; Potts 1988; Kolb, Lasca, and Goldstein 1990; Johnson and Logan 1990; Thorson 1990; Waters 1991, 1992.

62. Spier 1931. 63. Nelson 1916; Spier 1931. 64. Gifford and Rapp 1985a. For a fuller history of this period, see Albanese 2000; Artz 2000; Broman and Givens 1996; Ferring 2000; Mandel 2000; May 2000; and Holliday 2001.

92. Fryxell 1977.

94. Stein 1986. 95. A thoughtful overview of the historical context and status of geoarchaeology can be found in Huckleberry 2000.

65. Holliday 1997, 2000a, 2000b; Mandel, ed., 2000; Bettis 2000.

96. Van der Leeuw and Redman 2002.

66. Howard 1935.

98. Dunnell 1971; Leroi-Gourhan 1968.

67. Antevs 1935, 1937, 1948, 1955a.

99. North 1938.

97. Clarke 1979.

68. Holliday 1986. 69. Haynes 1990. 70. Gifford and Rapp 1995a; The wide range of interaction between the earth sciences and archaeology on the North American Great Plains has recently been summarized in Albanese 2000; Artz 2000; Broman and Givens 1996; Ferring 2000; Mandel 2000; May 2000; and Holliday 2001; Wendorf et al. 1955; and Judson 1957. 71. MacCurdy, ed., 1937. 72. Stein and Farrand 1985. 73. Movius 1949; Movius 1957; Braidwood 1957; Wright 1957; Cornwall 1958; Clark 1960.

Chapter 2. Sediments, Soils, and Environmental Interpretations 1. Reinek 1975; Soil Survey Staff 1975; Stein 2001. 2. Boggs 1995. 3. Bettis 1992: 119. As Bettis has written: ‘‘Archaeological remains, being deposits, are subject to the same natural processes of burial, weathering, and erosion that affect the preservation and distribution of noncultural deposits’’ (119). 4. Hill 2001a.

74. Pewe 1954.

5. Davidson 1980.

75. Taylor 1957.

6. Shlemon and Budinger 1990.

76. Butzer 1982.

7. Gleeson and Grosso 1976.

77. Pyddoke 1961.

8. Stockton 1973.

78. Butzer 1964.

9. Stein 1983.

79. Haynes 1964; see Holliday 2000c for an overview.

10. Gladfelter 1985.

278

Notes to Pages 29–80 11. Sherwood 2001.

19. Vandiver et al. 1989.

49. A book, published in 1989 (Courty et al.), and a chapter in a recent book (Courty 2001) offer a broad introduction to micromorphology. Two papers discussing opposite viewpoints, published in 1998 (Carter and Davidson; Macphail) offer geoarchaeologists both the spectrum of technical aspects and a critical analysis of micromorphology in studies of ancient agriculture; Davidson and Simpson 2001; Goldberg and Whitbread 1993; Macphail and Cruise 2001. Soil scientists perfected micromorphology and find it useful in many applications.

20. Eidt 1985.

50. Weiss et al. 1993.

21. Dincauze 1976.

51. Woodward and Goldberg 2001.

22. Clark 1954.

52. Goldberg and Arpin 1999.

23. See, e.g., chapter 7, in Coles and Coles 1989.

53. Goldberg and Whitbread 1993.

24. Clark 2001.

54. French 2003.

12. For an example, see Haynes 1973. 13. Waters et al. 1999. 14. Peacock 1991. 15. Meltzer et al. 1994. 16. Fladmark 1982. 17. Rapp 1975. 18. Hill 1993c.

25. Thiem 1997. 26. Holliday, ed., 1992. 27. Collins 1995; Mandel and Bettis 2001; Pavich and Chadwick 2003. 28. Rapp 1975; Pavich 2003.

Chapter 3. Initial Context and Site Formation 1. See, e.g., Behrensmeyer and Hill 1980; Hoch 1983; Lyman 1994; Nicholson 2001.

29. See French 2003; Cremeens 2003; Reuter 2000; Nettleton et al. 2000; Nettleton et al. 1998; Dahms and Holliday 1998; Cremeens and Hart 1995 for excellent reviews of soil horizons and profiles.

2. Todd and Frison 1986.

30. Ferring 1992.

5. Mayer 2002.

31. Kapp 1969.

6. Wendorf et al. 1993.

32. Eidt 1985.

7. Hill 2001a.

33. Birkeland 1984.

8. Frederick 2001.

34. U.S. Department of Agriculture 1987.

9. Bridgland 2000; Ferring 2001.

35. Kraus and Brown 1986.

10. Buvit et al. 2003.

36. Mack and James 1992; Mack et al. 1993.

11. Kolb and Van Lopik 1966.

37. Alexander 1969.

12. Coleman 1968; Stanley and Warne 1993.

38. Schulze et al. 1993.

13. Mandel 1995.

39. Cruz-Uribe et al. 2003.

14. Huckleberry et al. 2003.

40. Huckleberry et al. 2003.

15. Stein et al. 2003; Ferring 1986.

41. Gifford et al. 1989.

16. Davis and Greiser 1992.

42. Wendorf 1969; Hill 1989.

17. Bridgland 2000.

43. Schick 1992.

18. Gibbard 1994.

44. Allen 1989.

19. Wendorf and Schild 1989.

45. Shackley 1974.

20. Petraglia and Potts 1994.

46. McBrearty et al. 1998.

21. Joyce and Mueller 1992.

47. Schick 1986.

22. Freeman 2000.

48. Hill 1993c.

23. Farizy 1994; Hill 1993a,b, 2001b; Koetje 1994.

3. Schiffer 1972. 4. Fanning and Holdaway 2001.

279

Notes to Pages 81–115 24. Kempe 1988.

57. Thorson and Hamilton 1977.

25. Goldberg and Arpin 1999. 26. Lyell 1863.

58. Corte 1962, 1963; Inglis 1965; Jackson and Uhlmann 1966; Schweger 1985; Bowers et al. 1983.

27. Butzer 1981.

59. Canti 2003.

28. Brady and Veni 1992.

60. Balek 2002.

29. See Liu 1988.

61. Johnson 2002.

30. Straus 1990 31. Donahue and Adavasio 1990. 32. Dort and Miller 1977.

Chapter 4. Methods of Discovery and Spatial Analyses

33. Geoarchaeologists who are new to the excavation or analysis of the sediments and stratigraphy of caves and rock shelters should consult Farrand’s 2001 review of the principles and practice of such investigations.

1. Miller and Westerback 1989; Upton 1970.

34. Phillipson 1994. 35. Barton and Clark 1993. See also Macphail and Goldberg 2000.

2. Hammond 1964. 3. Wagstaff 1987. 4. Cherry 1983. 5. Wilkinson 2003.

36. Moss 1978.

6. Ucko and Layton 1999.

37. Sutcliff et al. 1976.

7. Harrell and Brown 1992a,b.

38. Karkanas et al. 1999.

8. Huckleberry and Billman 2003.

39. Farrand 1993.

9. Stafford and Creasman 2002.

40. Lyell 1863.

10. Bennett et al. 2000.

41. Wymer 1985.

11. Bettis and Hajic 1995.

42. Ashton et al. 1992; Roe 1981, 1993.

12. Kraft et al. 1987.

43. D.-J. Stanley 1995 revived the possibility of determining a global sea-level rise curve.

13. Harley and Woodward 1987.

44. Reinhardt 1992.

15. Donoghue 2001; Gupta 2003.

45. Jing and Rapp 2003; Besonen et al. 2003.

16. Kvamme 2001; Nishimura 2001.

46. Kraft et al. 1987.

17. Oldfield et al. 1983.

47. Kraft et al. 2003.

18. Aspinall 1992; Aitken and Milligan 1992; Bevan 1991; Clark 1992; Scollar et al. 1990; Tite and Mullins 1971; Vaughn 1986; Wynn 1986; Schmidt 2001; Conyers and Goodman 1997; Garrison 2003.

48. Kraft et al. 2000. 49. Barra et al. 1999. 50. Davis 1998. 51. Julig et al. 1990; Larsen 1985; Farrand 1960; Farrand and Dexler 1985; Phillips and Hill 2004. 52. Leigh 2001. 53. A detailed treatment of the problems in deciphering a shell midden is available in Stein, ed., 1992. 54. An often-cited summary is Wood and Johnson 1978. 55. Rick 1976; Culling 1963; Young 1960; Moeyerson 1978; Taber 1930; Wasburn 1980; Johnson and Hansen 1972; Johnson et al. 1977; Reid 1984; Schweger 1985; Thorson 1990; Thorson and Hamilton 1977. 56. Johnson 1952.

280

14. Bergman et al. 2003.

19. Roosevelt 1991; Dalan 1993. 20. Bevan and Roosevelt 2003. 21. Rapp and Henrickson 1972. 22. Dalan and Bevan 2002. 23. A practical guide to geophysical prospection is Schmidt 2001. 24. Kvamme 2003. 25. Chávez et al. 2001. 26. Kvamme 2003. 27. Black and Johnston 1962. 28. Carr 1982.

Notes to Pages 116–143 29. Bevan 2000.

65. Jing et al. 1995.

30. Thacker and Ellwood 2002.

66. Bridgland 1994; Gibbard 1994.

31. Conyers and Goodman 1997 have written a handy guide to GPR for archaeologists.

67. Rapp and Kraft 1994.

32. Leucci 2002.

69. Dolan et al. 2003.

33. Conyers and Cameron 1998.

70. Haynes 1995.

34. Jones et al. 2000.

71. Pescatore et al. 2001.

35. Mullins and Halfman 2001.

72. Haynes et al. 1999.

36. Quinn et al. 1997, 1998.

73. Nir 1997.

37. Van Andel and Lianos 1984.

74. Deocampo et al. 2002.

38. Van Andel and Sutton 1987.

75. Sheehan 1994; Chapman 2001.

39. The techniques of soil magnetism investigations for geoarchaeologists have been summarized and explained by Dalan and Banerjee 1998.

76. Kvamme 1999 has provided a detailed summary of GIS techniques in archaeology, including a comprehensive bibliography.

40. Dalan 2001. 41. Marmet et al 1999.

77. For extensive coverage of the archaeological uses of GIS see Lock and Stancic 1995.

42. Linford and Canti 2001.

78. Kvamme 1999.

43. Peters et al. 2001.

79. Hudak et al. 2002.

44. A manual for air photo interpretation in archaeology was published by Wilson 2000.

80. Harrower et al. 2002.

45. Crowley 2002.

82. Stein and Linze 1993.

68. Jing et al. 1995, 1997.

81. Stein 1993.

46. Custer et al. 1986. 47. McCauley et al. 1982. 48. Van Leusen 1998. 49. Gaffney et al. 1991. 50. Heron 2001. 51. Eidt 1985; Walker 1992. 52. Terry et al 2000. 53. Parnell et al. 2002. 54. Kerr 1995. 55. Schuldenrein 1995. 56. Eidt 1984. 57. Abbott and Wolfe 2003. 58. Crowther 1997.

Chapter 5. Estimating Time 1. Zachos et al. 2001. 2. Lowe 2001; Lambeck 2002. 3. Walsh 2001; Easton et al. 2003; Harris 1989. 4. Stein 2000. 5. Boggs 1995. 6. De Geer 1937; de Geer 1940. 7. Antevs 1925, 1954, 1955b. 8. An excellent presentation is provided in Bradbury et al. 1993.

60. Lambert 1997.

9. For North America, see Feng et al. 1994; for Europe, see Rousseau and Puisseguir 1990; for China, see Kukla 1987 and Kukla and An 1989.

61. Bethell and Máté 1989.

10. Johnson and Logan 1990.

62. Dean 1974.

11. Hill 2001 and 2003.

63. Willerslev et al. 2003.

12. Mandel 1992; Bettis 1992; after Hajic 1990; Mandel and Bettis 2001.

59. Bull et al. 2001.

64. A thorough history of archaeological coring in North America and an introduction to techniques and procedures are presented by Stein 1986. See also Stein 1991 and Schuldenrein 1991.

13. For example, Foit 1993. 14. Some examples are: Mehringer and Foit 1990; Davis and Greiser 1992; Davis 1984.

281

Notes to Pages 144–172 15. Douglass 1919; Douglass 1935; Nash 2000; Nash, ed., 2000; Nash and Dean 2000.

47. See, for example, Wintle et al. 1984.

16. Kuniholm 2001; Nash 2000; Nash, ed., 2000.

49. Feathers 1997.

17. Bonde and Christensen 1993.

50. Smith et al. 1997.

18. Hall and York 1984.

51. For a detailed overview of luminescence dating in archaeology, see Roberts 1997.

19. Evernden and Curtis 1965. 20. Walter et al. 1991.

48. Lang and Wagner 1996.

21. Clark et al. 2003.

52. Colls et al. 2001; Hill 2001; Schwarcz and Rink 2001; Turney et al. 2001; Wallinga 2002.

22. Latham 2001.

53. Feathers 1997a; Hill 2001b; Feathers and Hill 2003.

23. Schwarcz 1980; Schwarcz and Gascoyne 1984; Schwarcz 1984; Schwarcz and Morawska 1993; Stearns 1984; Hennig et al. 1983.

54. Hill 2001b; Ambrose 2001.

24. Bischoff et al. 1992. 25. Libby 1955; Terasmae 1984; Hester 1987. 26. Beck, et al. 2001; Conard and Bolus 2003; Laj et al. 2002; Richards and Beck 2001; Voelker et al. 2000.

55. Brooks et al. 1990; Miller 1993. 56. There are still questions regarding the utility of obsidian hydration dating is ongoing. For pro and con discussion of this method, see Anovitz 1999. For further information see Ambrose 2001 and Beck and Jones 2000. 57. Clark and McFadyen Clark 1993.

27. Siani et al. 2001.

58. Michels 1969.

28. Bondevik et al. 1999.

59. Michels 1983.

29. Spennemann and Head 1998.

60. Hammond 1989.

30. Yoneda et al. 2002.

61. Wendorf et al. 1955.

31. Ulm 2002.

62. Dawson 1913; Spencer 1990; Thompson 1991.

32. Geyh et al. 1999.

63. Whitley and Dorn 1993.

33. Bard et al. 2004; Hughen et al. 2004; Fairbanks et al. 2005.

64. Stuart 2001. 65. A review of techniques is presented in Wintle 1996.

34. Bard et al. 1990. The relevance of this is addressed in Tushingham and Peltier 1993; Naeser 1984. 35. Gleadow 1980.

Chapter 6. Paleoenvironmental Reconstructions

36. Morwood et al. 1998; O’Sullivan et al. 2001; Sternberg 2001.

1. Fagan 2000.

37. See Eighmy and Sternberg 1991.

2. Bryson 1989; see also Bryson and Goodman 1980; Gunn 1992.

38. Eighmy and Howard 1991. 39. Barmore 1985. 40. Gabunia et al. 2000; Zhu et al. 2001; Thackeray et al. 2002. 41. Urabe et al. 2001. 42. Ningawa et al. 1988. 43. Feathers 1997; Grun 2001. 44. For comprehensive reviews of TL methodology, see Berger 1988; Aitken 1985; Lamothe et al. 1984; Wintle and Aitken 1977; Balescu et al. 1991; Balescu et al. 1988; Wintle et al. 1984; Ningawa et al. 1988.

3. Gifford 1981. 4. For a manual on multiple microfossil extractions in archaeology see Coil et al. 2003. 5. Betancourt et al. 1990; Moore et al. 1996. 6. Faegri et al. 1989; Reinhardt and Bryant 1992. 7. See Wright 1976; Roberts and Wright 1993. 8. Solecki 1963; Solecki 1975. 9. Rowley-Conwy 1993. 10. Dimbleby 1985. 11. Shay 1971; Bradbury et al. 1993.

45. Mercier et al. 1995.

12. Aikens 1983; Stoltman and Baerreis 1983.

46. Rendell and Dennell 1987.

13. Finlayson et al. 1973; Byrne and McAndrews 1975.

282

Notes to Pages 172–195 14. Hebda and Mathewes 1984; Piperno and Pearsall 1993.

53. Clark 1953.

15. Rapp and Mulholland 1992; Albert 2003.

55. Langbein and Schumm 1958.

16. Elbaum et al. 2003.

56. Brauer et al. 1999.

17. Pearsall et al. 1994.

57. Binford et al. 1997.

18. Rovner and Russ 1992; Russ and Rovner 1989.

58. Haug et al. 2003.

19. Rovner 1994.

59. Holliday 2000.

20. Fox et al. 1994. 22. Hohn and Hellerman 1961.

60. McCoy and Heiken 2000b; see also De Boer and Sanders 2002; Rampino and Ambrose 2000; and WoldeGabriel et al. 2000.

23. Blinn et al. 1994.

61. Dincauze 2000.

24. Bradbury 1975.

62. Wyckoff 2002. Quotation from p. 15.

25. Wendorf et al. 1984, 1989.

63. Richerson et al. 2001. Quotation from p. 403.

26. LaMarche 1974; Fritts 1976; Fritts et al. 1979; Pilcher and Hughes 1982.

64. Srivastava et al. 1998.

21. Butterbee 1998.

27. Schwengruber 1988. 28. Palacios-Fest 1994. 29. Megard 1967. 30. Preece 2001. 31. Evans 1972. 32. Jones and Fisher 1990. 33. Robinson 2001. 34. Elias 1994. 35. Buckland and Kenward 1973. 36. Elias 1990. 37. Yalden 2001. 38. Steadman and Miller 1987. 39. Pregill 1986. 40. Casteel 1974, 1976. 41. Van Neer 1993.

54. Dalfes et al. 1997.

65. Bettis, ed., 1995. The volume contains the following articles: J. P. Albanese and G. C. Frison, ‘‘Cultural and Landscape Change during the Middle Holocene, Rocky Mountain Area, Wyoming and Montana’’; C. R. Ferring, ‘‘Middle Holocene Environments, Geology, and Archaeology in the Southern Plains’’; R. D. Mandel, ‘‘Geomorphic Controls of the Archaic Record in the Central Plains of the United States’’; J. A. Artz, ‘‘Geological Contexts of the Early and Middle Holocene Archaeological Record in North Dakota and Adjoining Areas of the Northern Plains’’; E. A. Bettis III and E. R. Hajic, ‘‘Landscape Development and the Location of Evidence of Archaic Cultures in the Upper Midwest’’; C. Chapdelaine and P. LaSalle, ‘‘Physical Environments and Cultural Systems in the Saint Lawrence Valley, 8,000 to 3,000 b.p.’’; and M. J. Stright, ‘‘Archaic Period Sites on the Continental Shelf of North America.’’ 66. Maloney et al. 1989. For a review of charcoal evidence for Holocene fires in this region, see Maxwell 2004.

42. Wells and Jorgensen 1964. 43. Wang et al. 1997; Nordt 2001. 44. Shackleton 1982. 45. For age estimates of the marine isotope record, see Martinson et al. 1987. 46. Ortloff and Kolata 1993. 47. Wang et al. 1997. 48. Stanley et al. 2003. 49. Bell 1992; Richter 1980; Goudie 2000; Wells 2001. 50. Childe 1928. 51. Willey and Phillips 1955; Wright 1993. 52. Holliday 2001.

Chapter 7. Raw Materials and Resources For a more complete treatment of this topic, see Rapp 2002. Rapp has worked on aspects of lithic materials and resources for more than thirty-five years. He has made notes on lithologies of artifacts from perhaps one hundred sites and collections in Greece, Turkey, Cyprus, Israel, Jordan, Egypt, Tunisia, China, and North America. The readers can assume that when a reference is not given in this chapter the items are from the author’s thirty-five years of experience, including eighteen years teaching mineralogy courses, or the information is generally known.

283

Notes to Pages 195–227 1. For definitions of mineral terms readers are referred to Roberts et al. 1974, 1990; de Fourestier 1999; and Blackburn and Dennen 1997. Jackson 1997 is perhaps the best guide to rock names as well as general geologic terms. Numerous mineralogy textbooks are also commonly available.

34. See Sheets et al. 1990 for Mesoamerican obsidian sources.

2. For a detailed summary of the use of minerals and rocks in ancient Egypt, see Lucas 1962.

38. Chang 1980.

3. Black and Wilson 1999.

40. Claridge 1984.

4. Luedtke 1992.

41. Rapp 2002.

5. Biswas 1996; Cottrell 1985.

42. Wreschner 1985.

6. Paulissen and Vermeersch 1987.

43. Faulkner 1962.

7. Butler and May 1984.

44. Gettens and Stout 1966.

8. Gramly 1992; Butler and May 1984.

45. Tagle et al. 1990.

9. Cackler et al. 1999.

46. Sanger et al. 2001.

10. Wright and Chadbourne 1970

47. Dittert 1968.

11. For a summary of the early use of amethyst, see Pliny’s Natural History.

48. Ball 1941.

35. Davis et al. 1992. 36. Hill 2001a. 37. Porter 1961. 39. Bishop and Lange 1991; Matson 1966.

49. Heizer and Treganza 1944.

12. Lucas 1962. 13. Wen and Jing 1992. 14. Harlow 1991. 15. Gendron et al. 2002. 16. Desautels 1986. 17. Mathieu 2001. 18. Francis 1988; Mustoe 1985.

Chapter 8. Sourcing (Provenance) 1. Rapp et al. 2000. 2. Thomas 1923. 3. Thorpe et al. 1991.

22. Abbott and Wolfe 2003.

4. Those interested in exploring the second and third steps in provenance determination should consult the journal Archaeometry, which since the 1960s has published many excellent articles on the analytical and statistical problems involved.

23. Louis 1994.

5. Rapp et al. 2000.

24. Bunker 1993 has reviewed the use of gold in ancient China.

6. Rapp et al. 2000.

25. Wagner and Gentner 1979.

8. Heizer et al. 1973.

26. Hosler and Macfarlane 1996.

9. A review article by Williams-Thorpe 1995 covering obsidian provenance studies in the Mediterranean and the Near East should be consulted for a summary of recent findings. It includes data from Carpathian northeastern Hungary and eastern Slovakian sources, as well as the well-known Italian, Aegean, Anatolian, and Armenian deposits. An extensive bibliography is provided.

19. Ferguson 1974. 20. Moulton 2003. 21. McKillop 2002.

27. Rapp et al. 2000. 28. Hosler and Macfarlane 1996. 29. Pendergast 1982. 30. Aufderheide et al. 1992. 31. Walthall 1981. 32. Stafford et al. 2003. 33. Bjorkman 1973 reviews meteorite artifacts from the Near East and La Paz 1969 presents a record of meteorite utilization from the Paleolithic to the present.

284

7. Harbottle 1990.

10. Shackley 1995. 11. Burton and Krinsley 1987; see also Kayani and McDonnell 1995. 12. Greenough et al. 2001.

Notes to Pages 227–260 13. Bakewell 1996.

50. Shepard 1965.

14. Pretola 2001.

51. Mason and Golombek 2003.

15. Julig et al. 1991.

52. Vaughn 1990.

16. Kinnunen et al. 1985.

53. Stoltman 1991.

17. Herz 1985.

54. Good reviews of the use of ceramic petrography in provenance and location of production studies are provided by Stoltman 2001 and by Whitbread 2001.

18. Attanasio et al. 2000. 19. Blackman 1981 provides a quantitative picture of the influence of the temper on the chemical composition of a clay body.

55. Dickinson 2001. 56. Mandal 1997.

20. Stoltman 2001.

57. Ferring and Pertulla 1987.

21. Shepard 1939.

58. Zaykov et al. 1999.

22. Steponaitis 1984. 23. Dye and Dickinson 1996. 25. Harrell and Lewan 2002.

Chapter 9. Construction, Destruction, Archeological Resource Preservation, and Conservation

26. A review of instrumental analysis in sourcing lithic artifacts is given by Herz 2001.

1. Zangger 1994.

27. Warashina et al. 1978; Togashi and Matsumoto 1991.

2. Zangger 1991.

28. Truncer et al. 1998.

3. Doolittle 1990.

29. A review of instrumental analysis in sourcing lithic artifacts is given by Herz 2001.

4. Fish and Fish 1992.

24. Knapp and Cherry 1994.

30. Rapp et al. 1990b; see also Rapp et al. 1990a. 31. Rapp et al. 2000. 32. Hancock et al. 1991. 33. Wilson et al. 1997. 34. Rapp 1988; Rapp 1989. 35. Rothenberg 1978.

5. Erickson 1992. 6. Capedri et al. 2003. 7. In his Ancient Mining, Shepherd presents details on ancient mining practices throughout the Old World (1993). 8. Wall Paintings of the Tomb of Nefertari: Scientific Studies for Their Conservation 1987.

36. A thorough history of tin in antiquity is available in Penhallurick 1986.

9. Aberg et al. 1999.

37. Rapp et al. 1999.

11. Reader 2001.

38. Farquhar et al. 1995.

12. Stanley et al. 2004.

39. Farquhar and Fletcher 1984.

42. Tobey 1986.

13. The book edited by Stiros and Jones 1996 offers many brief chapters detailing seismic effects in major sites in the eastern Mediterranean region. Noller 2001 presents a broad look at archaeoseismology and offers case studies. See also Galli 2003.

43. Curran et al. 2001.

14. Evans 1964.

44. Müller et al. 2003.

15. Nur and Cline 2000.

45. Arndt et al. 2003.

16. Stein et al. 1997.

46. Mooney et al. 2003

17. Bolt 1978.

47. Bossière and Frère 2001.

18. Reimnitz and Marshall 1965.

48. Dorais and Shriner 2002.

19. Sims 1973; Sims 1979.

49. Stoltman 2001; Whitbread 2001.

20. Doig 1990.

40. Williams-Thorpe et al. 2000. 41. Hoard et al. 1992; Church 1992; Hoard et al. 1995.

10. Hudson-Edwards et al. 1999.

285

Notes to Pages 260–275 21. Niemi and Ben-Avraham 1994.

39. McCoy and Heiken 2000a.

22. Street and Nuttli 1984.

40. De Silva et al. 2000.

23. Saucier 1977; Saucier 1989; Saucier 1991.

41. Sheets and McKee 1994.

24. Martinez Solarez et al. 1979. 25. Thompson 1970.

42. For a review of the archaeology of Venice, see Ammerman and McClennen 2001.

26. Hough 1906.

43. Gilbert 1884.

27. Goff and McFadgen 2003.

44. Gatto and Doe 1983.

28. Schnellmann et al. 2004.

45. Passmore et al. 2002.

29. Galli and Galadini 2003.

46. Readers are referred to Amoroso and Fassina 1983.

30. Ambraseys and Melville 1982. 31. Hutchinson and McMillan 1997.

Chapter 10. Epilogue

32. Cole et al. 1996. 33. Woodward et al. 1990.

1. Vitaliano 1973.

34. Karcz and Kafri 1978; Rapp 1987; Stiros 1988, 1996.

2. De Boer and Hale 2000.

35. Dundes 1988. Dundes has provided a detailed look at the flood myth that includes annotated reprints of important contributions from scholars representing many disciplines.

3. Buck and Stewart 2000. 4. Hawkes 1954.

37. Mallowan 1964.

5. Wiseman 1980, 2001, 2002; Dunnel 1982; Gillespie et al. 2003. For perspectives on teaching archaeology in the twenty-first century in North America, see Bender and Smith 2000.

38. Quoted in Simkin and Fiske 1983.

6. Ucko and Layton 1999.

36. Woolley 1929.

286

Glossary

aggradation The process of building up a surface by deposition of sediment alkaline earth Sediments and/or soils containing a high proportion of alkaline cations like sodium and potassium amphibole An important rock-forming mineral group of ferromagnesian silicates found in granitic rocks anoxic Lacking oxygen anthrosol A soil whose main characteristics are the result of human activity aragonite A carbonate mineral with the same composition as calcite arroyo Steep-sided, flat-bottomed gully in an arid region through which a stream flows only after very heavy rainfall (usually seasonally) authigenic Formed or generated in place. Rock constituents and minerals that have not been transported or that were formed on the spot where they are found. bajada The surface of a system of coalesced alluvial fans calcite Calcium carbonate (CaCO 3 ) calcrete A conglomerate consisting of surficial sand and gravel cemented into a hard mass by calcium carbonate precipitated from solution and redeposited through the agency of infiltrating waters caldera A large, basin-shaped depression resulting from the explosion or collapse of the center of a volcano capillary rise The tendency of a fluid to rise within narrow passages because of surface tension

clastic Fragmental rock material that has been mechanically transported and deposited in a sedimentary environment colloidal Pertaining to a suspension of finely divided particles in a continuous medium crevasse splay A deep fissure or crack in a valley glacier that runs parallel to the direction of flow in the center of the glacier but curves toward the margin downstream crinoid An echinoderm with a disk-shaped or globular body, characterized by the presence of a stem (more common in fossils than in living forms) cross bedding Creation through deposition of inclined beds in a sedimentary rock. Formed by currents of wind or water in the direction in which the bed slopes downward. debris flow A downhill movement of a mass of rock and soil deflation Removal by strong winds diagenesis The physical and chemical change undergone by a sediment during lithification and compaction diatom A microscopic one-celled plant that grows in both marine and fresh waters. It secretes walls of silica in a variety of geometric forms, depending on the species. These silica forms accumulate in sediments in great numbers. diatomite A siliceous chertlike sediment formed from the hard parts of diatoms edaphic Relating to any soil characteristic that affects plant growth, like acidity or alkalinity eluvial Deposited by the action of wind

287

Glossary

eluviation Internal movement of soil particles when rainfall exceeds evaporation

illuvial Pertains to a soil horizon that has been subjected to illuviation

endogenic Derived from within; deposits originating from within the rocks that contain them

illuviation The deposition in an underlying soil layer of colloids, soluble salts, and mineral particles that are transported from an overlying soil layer

epipedon A soil horizon that forms at the upper end of the soil profile. An epipedon can include the dark A horizon as well as the illuvial B horizon when organic matter extends from the surface through the B horizon.

isotropic Pertaining to a substance in which the magnitude of a physical property does not vary with direction within the substance

eustatic Relating to a worldwide change of sea level

jointing Creation of a fracture in a rock body where no movement has occurred

exogenic, exogenetic Pertaining to materials or processes that originate at or near the surface of the earth

karst A region underlain by limestone where solution has developed sinks, underground caverns, and water passageways

facies A portion of a rock unit with a distinctive group of characteristics (mineral assemblage or fossil assemblage, for example) that differs from other parts of the same rock unit

lag Coarse sediment left behind the surface by wind action

fen carr A pool in a waterlogged or marshy ground containing decaying alkaline vegetation ferrous Pertains to iron with a +2 valence; Fe 2+

lithofacies A lateral or vertical subdivision of a stratigraphic unit, distinguished from other adjacent subdivisions on the basis of lithologic characteristics

flux In metallurgy, material added in smelting to form a fluid slag

loess Fine-grained, loosely consolidated sediments, usually formed from wind-deposited silts

fractionated Pertains to a homogenous chemical material that has been separated into its components

mafic Pertaining to low-silica igneous rocks that contain an abundance of iron-magnesium silicates

gamma A unit of magnetic field strength, symbolized by γ

magnetic susceptibility The magnetic permeability of a medium relative to that of a vacuum; it will be positive for a paramagnetic or ferromagnetic medium and negative for a diamagnetic medium

ferric Pertains to iron with a +3 valence; Fe 3+

geofacts Geologic objects that have the appearance of artifacts glaciofluvial Deposited by glacial meltwater gleyed, gleying Mottled: in soils, caused by partial oxidation and reduction of the soil’s ferric iron compounds gossan An iron oxide product created by weathering that overlies a sulfide deposit hardpan A layer of hard subsoil or clay, produced when precipitation of insoluble materials like silica, iron oxide, or calcium carbonate cements the particles humification The process of development of humus or humic acids, essentially by slow oxidation

288

lithification The process by which unconsolidated sediments are converted to rock

matte In metallurgy, an intermediate-stage product in the smelting of copper-sulfide ores composed of iron and copper sulfides moraine Landform produced by an accumulation of glacial till outwash A sediment deposited by meltwater streams from a glacier oxbow lake Geographic feature created when erosion on the outside of a meander loop causes a river to meet itself, and it abandons the waterfilled meander, which becomes the oxbow lake oxidation A chemical reaction in which one or

Glossary

more electrons are removed from an atom, making its charge more positive palustrine Pertaining to material growing or deposited in a marsh or marsh-like environment pediment A planar, sloping rock surface forming a ramp up to the front of a mountain range in an arid region, which may be covered by local alluvium perched The condition in which the upper surface of a local zone of saturation lies above the regional water table

rare-earth elements Chemical elements with the atomic numbers 57 through 71; they were discovered as oxides (earths) in rare minerals reducing atmosphere (1) In soils, an atmosphere that promotes chemical elements to lose electrons; (2) In ceramics, a firing atmosphere that is oxygen poor refractory Having the ability to withstand high temperatures without melting. In ceramics, refractory materials are usually high in alumina and silica.

periglacial Processes, climates, and topographic features that are either on the immediate margins of former or existing glaciers or are influenced by cold temperatures

regression The retreat or withdrawal of water from land areas. Opposite of transgression.

phenocryst A large crystal surrounded by a finer matrix in a porphyry

salina A place where crystalline salt deposits are formed or found, such as a salt flat or salt pan.

phylogenetic Pertaining to the evolutionary development of a species of plant or animal

sand dike A sedimentary dike consisting of sand that has been squeezed or injected upward into a fissure

playa The flat floor of a closed basin in an arid region. It may be occupied by an intermittent lake. pluvial Caused by, or pertaining to rain. A period of increased rainfall and decreased evaporation. point bars Deposits of sediment on the inner banks of meanders; formed because the stream velocity is lower against the inner bank

retrograding When a coastline is being eroded by wave action

sand pipe A tubular cavity in sedimentary rocks filled with sand and gravel sandblow A patch of coarse sandy sediment or soil denuded of vegetation by wind action seiche A wave that oscillates in lakes, bays, or gulfs

polje An interior valley; an elongated basin with a flat floor and steep enclosing walls found in karst regions

septarian Pertaining to an irregular polygonal system of cracks occurring in some rock concretions

porcellanite A dense siliceous rock having the texture, luster, and conchoidal fracture of unglazed porcelain. The term has been applied to a wide variety of rocks from impure chert to baked clay to silicified tuff.

sheet flood A broad expanse of moving storm water that spreads as a continuous sheet over a large area rather than being concentrated into well-defined channels

proglacial Features produced by glacier ice immediately in front of the glacier or ice sheet

shotcrete A mixture of portland cement, sand, and water applied by pneumatic pressure as a sealing agent

progradational sequence A sequence that shows prograding or regression of sediments, a process that results in an upward-fining sequence of sediments

slag The vitreous mass formed as a part of the smelting process

prograding When a deposit is being built forward or outward by deposition and accumulation

slickensides A smoothly striated rock surface that results from friction during movement along a fault plane

pyroxene An iron and magnesium silicate mineral that occurs primarily in mafic igneous rocks

slake Slaking is the treating of lime (CaO) with water to give hydrated (slaked) lime

289

Glossary

slip A suspension of fine clay in water, used as a coating in the decoration of ceramic ware

terrigenous Pertaining to, or derived from the land (as opposed to marine)

sluice box A device used for concentrating heavy minerals in unconsolidated material; the box is equipped with riffles that trap heavier minerals when the sedimentary material is washed through it

till Unstratified and relatively unsorted glacial deposits

slumping A collapse of unlithified sediments or soils caused by gravity

transgressive, transgression Spread or extension of the sea over land areas. Note that as the seas transgress slowly across coastal land areas the sediments deposited in these shallow seas are time transgressive.

spall A fragment removed from the surface of a rock by weathering, in particular, a relatively thin piece of rock produced by exfoliation

tuff A general term for all consolidated pyroclastic rocks, that is, rocks formed of particles originating from a volcanic explosion

taiga Subarctic evergreen forest of Siberia and similar regions elsewhere in Eurasia and North America

turbation Stirring, mixing, or other modifications of a sediment or soil by unspecified agents

talus A pile or sheet of loose rock fragments accumulated at the base of a steep slope taphonomic, taphonomy Pertaining to a branch of paleoecology concerned with the manner of burial and preservation of plant and animal remains. Artifact taphonomy is the study of archaeological site formation processes tectonism The regional or global deformation of the earth’s crust

ultramafic Pertaining to igneous rocks composed mainly of iron-magnesium silicates uniformitarianism The concept that ancient rocks can be understood in terms of the processes presently operating on the earth; sometimes summed up as ‘‘the present is the key to the past’’ ventifacts Rocks that exhibit the effects of sandblasting or ‘‘snowblasting’’ on their surfaces

tell A mound site formed by successive human occupations over a considerable period of time

vesicular A rock texture characterized by abundant vesicles (voids) caused by gas bubbles that occurred when the rock was fluid

tephra Rock material expelled by a volcanic explosion

wadi A valley, gully, or river bed that remains dry except in the rainy season

290

Bibliography

Abbott, C. 1892. ‘‘Paleolithic Man: The Last Word.’’ Science 20: 344–45. Abbott, M., and A. Wolfe. 2003. ‘‘Intensive Pre-Incan Metallurgy Recorded by Lake Sediments from the Bolivian Andes.’’ Science 301: 1893–95. Aberg, G., H. Stray, and E. Dahlin. 1999. ‘‘Impact of Pollution at a Stone Age Rock Art Site in Oslo, Norway, Studied Using Lead and Strontium Isotopes.’’ Journal of Archaeological Science 26: 1483–88. Acher, R. 1968. ‘‘Time’s Arrow and the Archaeology of Contemporary Community.’’ In Settlement Archaeology, ed. K. Chang, 43–53. Palo Alto, Calif.: National Press Books. Adams, R. 1960. ‘‘Manuel Gamio and Stratigraphic Excavation.’’ American Antiquity 26: 99. Aikens, C. 1983. ‘‘Environmental Archaeology in the Western United States.’’ In Late-Quaternary Environments of the United States, vol. 2: The Holocene, ed. H. E. Wright, Jr., 239–51. London: Longman. Aitken, M. 1985. Thermoluminescence Dating. New York: Academic Press. Aitken, M., and R. Milligan. 1992. ‘‘Ground-Probing Radar in Archaeology: Practicalities and Problems.’’ Field Archaeologist 16: 288–91. Albanese, J. 2000. ‘‘Resume of Geoarchaeological Research on the Northwestern Plains.’’ In Geoarchaeology in the Great Plains, ed. R. Mandel, 199–249. Norman: University of Oklahoma Press. Albert, R. 2003. ‘‘Detection of Burning of Plant Materials in the Archaeological Record by Changes in the Refractive Indices of Siliceous Phytoliths.’’ Journal of Archaeological Science 30: 217–26. Alexander, J. 1969. ‘‘A Color Chart for Organic Matter.’’ Crop Soils 21: 15–17. Allen, J. 1989. ‘‘A Quantitative Technique for Assessing the Roundness of Pottery Shards in Water Currents.’’ Geoarchaeology 4(2): 143–55. Ambraseys, N., and C. Melville. 1982. A History of Per-

sian Earthquakes. Cambridge: Cambridge University Press. Ambrose, W. 2001. ‘‘Obsidian Hydration Dating.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 81–92. New York: John Wiley and Sons. Ammerman, A., and C. McClennen. 2001. Venice before San Marco: Recent Studies on the Origins of the City. Hamilton, N.Y.: Colgate University. Amoroso, G., and V. Fassina. 1983. Stone Decay and Conservation: Atmospheric Pollution, Cleaning, Consolidation and Protection. Materials Science Monographs 11. New York: Elsevier. Anovitz, L. 1999. ‘‘The Failure of Obsidian Hydration Dating: Sources, Implications, and New Directions.’’ Journal of Archaeological Science 26: 735–52. Antevs, E. 1925. Retreat of the Last Ice Sheet in Eastern Canada. Canada Department of Geological Survey Memoir 146. Ottawa: F. A. Acland. . 1935. ‘‘The Occurrence of Flints and Extinct Animals in Pluvial Deposits near Clovis, New Mexico.’’ Part 2: ‘‘Age of Clovis Lake Beds.’’ Proceedings of the Philadelphia Academy of Natural Science 87 (1935): 304–11. . 1937. ‘‘Climate and Early Man in the Southwest.’’ In Early Man, ed. G. MacCurdy, 125–32. Philadelphia: Lippincott. . 1948. ‘‘The Great Basin, with Emphasis on Glacial and Postglacial Times.’’ University of Utah Bulletin 38: 20. . 1954. ‘‘Telerrelation of Varves, Radiocarbon Chronology, and Geology.’’ Journal of Geology 62: 516–21. . 1955a. ‘‘Geologic Climate Dating in the West.’’ American Antiquity 20: 317–35. . 1955b. ‘‘Varve and Radiocarbon Chronologies Appraised by Pollen Data.’’ Journal of Geology 63: 495–99. Arndt, A., W. Van Neer, B. Hellemans, J. Robben,

291

Bibliography F. Volckaert, and M. Waelkens. 2003. ‘‘Roman Trade Relationships at Sagalassos (Turkey) Elucidated by Ancient DNA of Fish Remains.’’ Journal of Archaeological Science 30: 1095–105. Artz, J. 2000. ‘‘Archaeology and the Earth Sciences on the Northern Plains.’’ In Geoarchaeology in the Great Plains, ed. R. Mandel, 250–85. Norman: University of Oklahoma Press. Aschenbrenner, S., and S. Cooke. 1978. ‘‘Screening and Gravity Concentration; Recovery of Small-Scale Remains.’’ In Excavations at Nichoria in Southwest Greece. Vol. 1: Site, Environs, and Techniques, ed. G. Rapp and S. Aschenbrenner, 156–65. Minneapolis: University of Minnesota Press. Ashton, N., J. Cook, S. Lewis, and J. Rose, eds. 1992. High Lodge: Excavations by G. De G. Sieveking, 1962– 68, and J. Cook, 1988. London: British Museum Press. Aspinall, A. 1992. ‘‘New Developments in Geophysical Prospection.’’ In New Developments in Archaeological Science, ed. A. Pollard, 233–44. Proceedings of the British Academy 77. Atkinson, R. 1963. ‘‘Resistivity Surveying and Archaeology.’’ In The Scientist and Archaeology, ed. E. Pyddoke, 1–30. London: Phoenix House. Attanasio, D., G. Armiento, M. Brilli, M. Emanuele, R. Plantiania, and B. Turi. 2000. ‘‘Multi-Method Marble Provenance Determinations: The Carrara Marbles as a Case Study for the Combined Use for Isotopic, Electron Spin Resonance and Petrographic Data.’’ Archaeometry 42: 257–72. Aufderheide, A., G. Rapp, L. Wittmers, J. Wallgren, R. Macchiarelli, G. Fornciari, F. Mallegni, and R. Corruccini. 1992. ‘‘Lead Exposure in Italy: 800 b.c.–700 a.d.’’ International Journal of Anthropology 7: 9–15. Bakewell, E. 1996. ‘‘Petrographic and Geochemical Source-Modeling of Volcanic Lithics from Archaeological Contexts: A Case Study from British Camp, San Juan Island, Washington.’’ Geoarchaeology 11(2): 119–40. Balek, C. 2002. ‘‘Buried Artifacts in Stable Upland Sites and the Role of Bioturbation: A Review.’’ Geoarchaeology 17: 41–51. Balescu, S., Ch. Dupuis, and Y. Quinlif. 1988. ‘‘TL Stratigraphy of Pre-Weichselian Loess from NW Europe Using Feldspar Grains.’’ Quaternary Science Reviews 7: 309–13. Balescu S., S. Pacman, and A. Wintle. 1991. ‘‘Chronological Separation of Interglacial Raised Beaches from Northwestern Europe Using Thermoluminescence.’’ Quaternary Research 35: 91–102. Ball, S. 1941. ‘‘The Mining of Gems and Ornamen-

292

tal Stones by American Indians.’’ Anthropological Papers 13. Bureau of American Ethnology Bulletin 128: 1–77. Bard, E., B. Hamelin, R. G. Fairbanks, and A. Zindler. 1990. ‘‘Calibration of the 14 C Timescale over the Past Thirty Thousand Years Using Mass Spectrometric U/Th Ages from Barbados Corals.’’ Nature 345: 405– 10. Bard, E., F. Rostek, and G. Ménot-Combes. 2004. ‘‘A Better Radioactive Clock.’’ Science 303: 178–79. Barmore, F. 1985. ‘‘Turkish Mosque Orientation and the Secular Variation of the Magnetic Declination.’’ Journal of Near Eastern Studies 44(2): 81–98. Barra, D., G. Calderont, M. Cipriani, J. De La Genière, L. Fiorillo, G. Greco, M. Lippi, M. Secci, T. Pescatore, B. Russo, M. Senatore, G. Sciarelli, and J. Thorez. 1999. ‘‘Depositional History and Palaeogeographic Reconstruction of Sele Coastal Plain during Magna Grecia Settlement of Hera Argiva (Southern Italy).’’ Geologica Romana 35: 151–66. Barton, C., and G. Clark. 1993. ‘‘Cultural and Natural Formation Processes in Late Quaternary Cave and Rockshelters of Western Europe and the Near East.’’ In Formation Processes in Archaeological Context, ed. P. Goldberg, D. Nash, and M. Petraglia, 33– 52. Monographs in World Archaeology 17. Madison, Wisc.: Prehistory Press. Battarbee, R. 1988. ‘‘The Use of Diatom Analysis in Archaeology: A Review.’’ Journal of Archaeological Science 15: 621–44. Beazeley, A. 1920. ‘‘Surveys in Mesopotamia during the War.’’ Geographical Journal 55(2): 109–12. Beck, C., and G. Jones. 2000. ‘‘Obsidian Hydration Dating, Past and Present.’’ In It’s About Time: A History of Archaeological Dating in North America, ed. S. Nash, 124–51. Salt Lake City: University of Utah Press. Beck, J., D. Richards, R. Edwards, B. Silverman, P. Smart, D. Donahue, S. Herra-Osterheld, G. Burr, L. Calsoyas, A. Jull, and D. Biddulph. 2001. ‘‘Extremely Large Variations of Atmospheric 14 C Concentration during the Last Glacial Period.’’ Science 292: 2453–58. Behrensmeyer, A., and A. Hill. 1980. Fossils in the Making: Vertebrate Taphonomy and Paleoecology. Chicago: University of Chicago Press. Bell, M., and M. Walker. 1992. Late Quaternary Environmental Change: Physical and Human Perspectives. New York: John Wiley and Sons. Bender, S., and G. Smith, eds. 2000. Teaching Archaeology in the Twenty-First Century. Washington, D.C.: Society for American Archaeology.

Bibliography Bennett, M., D. Huddart, and T. McCormick. 2000. ‘‘An Integrated Approach to the Study of Glaciolacustrine Landforms and Sediments: A Case Study from Hagavatn, Iceland.’’ Quaternary Science Reviews 19: 633–65. Berger, G. 1988. Dating Quaternary Events by Luminescence. Special Paper 227. Boulder, Colo.: Geological Society of America. Bergman, I., T. Påsse, A. Olofsson, O. Zackrisson, G. Hörnberg, E. Hellberg, and E. Bohlin. 2003. ‘‘Isostatic Land Uplift and Mesolithic Landscapes: Laketilting, a Key to the Discovery of Mesolithic sites in the Interior of Northern Sweden.’’ Journal of Archaeological Science 30: 1451–58. Besonen, M., G. Rapp, and Z. Jing. 2003. ‘‘The Lower Acheron River Valley: Ancient Accounts and the Changing Landscape.’’ Landscape Archaeology in Southern Epirus, Greece 1, ed. J. Wiseman and K. Zachos, 199–268. Athens: American School of Classical Studies at Athens. Betancourt, J., T. Van Devender, and P. Martin. 1990. Packrat Middens: The Last Forty Thousand Years of Biotic Change. Tucson: University of Arizona Press. Bethell, P., and I. Máté. 1989. ‘‘The Use of Phosphate Analysis in Archaeology: A Critique.’’ In Scientific Analysis in Archaeology, ed. J. Henderson, 1–29. Monograph 19. Oxford: Oxford University Press. Bettis, E. 1992. ‘‘Soil Morphologic Properties and Weathering Zone Characteristics as Age Indicators in Holocene Alluvium in the Upper Midwest.’’ In Soils in Archaeology: Landscape Evolution and Human Occupation, ed. V. Holliday, 119–44. Washington, D.C.: Smithsonian Institution Press. . 2000. ‘‘A Brief History of Geoarchaeology in the Eastern Plains and Prairies.’’ In Gearchaeology in the Great Plains, ed. R. Mandel, 137–65. Norman: University of Oklahoma Press. , ed. 1995. Archaeological Geology of the Archaic Period in North America. Special Paper 297. Boulder, Colo.: Geological Society of America. Bettis, E., and E. Hajic. 1995. ‘‘Landscape Development and the Location of Evidence of Archaic Cultures in the Upper Midwest.’’ In Archaeological Geology of the Archaic Period in North America, ed. E. Bettis, 87–113. Special Paper 297. Boulder, Colo.: Geological Society of America. Bevan, B. 1991. ‘‘The Search for Graves.’’ Geophysics 56(9): 1310–19. . 2000. ‘‘An Early Geophysical Survey at Williamsburg, U.S.A.’’ Archaeological Prospection 7: 51–58. Bevan, B., and A. C. Roosevelt. 2003. ‘‘Geophysical Ex-

ploration of Guajará, a Prehistoric Earth Mound in Brazil.’’ Geoarchaeology 18: 287–331. Binford, L. 1964. ‘‘A Consideration of Archaeological Research Design.’’ American Antiquity 29(4): 425–41. . 1977. Theory Building in Archaeology. New York: Academic Press, 1977. . 1981. Bones: Ancient Men and Modern Myths. New York: Academic Press, 1981. . 1983. In Pursuit of the Past: Decoding the Archaeological Record. New York: Thames and Hudson. . 1989. Debating Archaeology. New York: Academic Press. Binford, M., A. Kolata, M. Brenner, J. Janusek, M. Seddon, M. Abbott, and J. Curtis. 1997. ‘‘Climate Variation and the Rise and Fall of an Andean Civilization.’’ Quaternary Research 47: 235–48. Birkeland, P. 1984. Soils and Geomorphology. New York: Oxford University Press. Bischoff, J., J. Garcia, and L. Straus. 1992. ‘‘UraniumSeries Isochron Dating at El Castillo Cave (Cantabria, Spain): The ‘Acheulian’/‘Mousterian’ Question.’’ Journal of Archaeological Science 19: 49–62. Bishop, R., and F. Lange. 1991. The Ceramic Legacy of Anna C. Shepard. Boulder: University Press of Colorado. Biswas, A. 1996. Minerals and Metals in Ancient India. Vol. 1: Archaeological Evidence. New Delhi: DK Printworld. Bjorkman, J. 1973. Meteors and Meteorites in the Ancient Near East. Center for Meteorite Studies 12. Tempe: Arizona State University. Black, D., and L. Wilson. 1999. ‘‘The Washademoak Lake Chert Source, Queens County, New Brunswick, Canada.’’ Archaeology of Eastern North America 27: 81–108. Black, G., and R. Johnston. 1962. ‘‘A Test of Magnetometry as an Aid to Archaeology.’’ American Antiquity 28: 199–205. Blackburn, W., and W. Dennen. 1997. Encyclopedia of Mineral Names. Canadian Mineralogist Special Publication 1. Blackman, M. 1981. ‘‘The Mineralogical and Chemical Analysis of Banesh Period Ceramics from Tal-EMalyan, Iran.’’ In Scientific Studies in Ancient Ceramics, ed. M. Hughes, 7–20. British Museum Occasional Paper 19. London: British Museum. Blinn, D., R. Hevly, and O. Davis. 1994. ‘‘Continuous Holocene Record of Diatom Stratigraphy, Paleohydrology, and Anthropogenic Activity in a SpringMound in the Southwestern United States.’’ Quaternary Research 42: 197–205.

293

Bibliography Boggs, S. 1995. Principles of Sedimentology and Stratigraphy. 2nd ed. Englewood Cliffs, N.J.: Prentice Hall. Bolt, B. 1978. Earthquakes: A Primer. San Francisco: Freeman. Bonde, N., and A. Christensen. 1993. ‘‘Dendrochronological Dating of the Viking Age Ship Burials in Oseberg, Gokstad, and Tune, Norway.’’ Antiquity 67: 575–84. Bondevik, S., H. Birks, S. Gulliksen, and J. Mangerud. 1999. ‘‘Late Weichselian Marine 14 C Reservoir Ages at the Western Coast of Norway.’’ Quaternary Research 52: 104–14. Bossière, G., and D. Frère. 2001. ‘‘Petrological EDS Chemical Study in Thin Section of Some EtruscoCorinthian Ceramics: A Contribution to Their Archaeological Knowledge.’’ In Archaeology and Clays, ed. I. Druc, 39–53. British Archaeological Reports, International Series 942. Oxford: John and Erica Hedges. Bowers, P., R. Bonnichsen, and D. Hoch. 1983. ‘‘Flake Dispersal Experiments: Noncultural Transformation of the Archaeologic Record.’’ American Antiquity 48: 553–72. Bradbury, J. 1975. ‘‘Diatom Stratigraphy and Human Settlement.’’ Geological Society of America Special Paper 171: 1–74. Bradbury, J., J. Platt, and W. Dean, eds. 1993. Elk Lake, Minnesota: Evidence for Rapid Climate Change in the North-Central United States. Special Paper 276. Boulder, Colo.: Geological Society of America. Brady, J., and G. Veni. 1992. ‘‘Man-Made and PseudoKarst Caves: The Implications of Subsurface Features within Maya Centers.’’ Geoarchaeology 7(2): 149–67. Braidwood, R. 1957. ‘‘Means towards an Understanding of Human Behavior before the Present,’’ and ‘‘The Old World: Post Paleolithic.’’ In The Identification of Non-Artifactual Archaeological Materials, ed. W. Taylor, 14–15, 26–27. National Academy of Sciences Publication 565. Washington, D.C.: National Academy of Sciences National Research Council. Brauer, A., C. Endres, C. Günter, T. Litt, M. Stebich, and J. Negendank. 1999. ‘‘High Resolution Sediment and Vegetation Responses to Younger Dryas Climate Change in Varved Lake Sediments from Meerfleder Maar, Germany.’’ Quaternary Science 18: 321–29. Breuil, H. 1945. ‘‘The Discovery of the Antiquity of Man Some Evidence.’’ Journal of the Royal Anthropological Institute of Great Britain and Ireland 75(1/2): 21–31. Bridgland, D. 1994. Quaternary of the Thames. London: Chapman and Hall.

294

. 2000. ‘‘River Terrace Systems in North-west Europe: An Archive of Environmental Change, Uplift and Early Human Occupation.’’ Quaternary Science Reviews 19: 1293–303. Broman, D., and D. Givens. 1996. ‘‘Stratigraphic Excavation: The First ‘‘New Archaeology.’’ American Anthropologist 98(1): 80–95. Brooks, A. S., P. E. Hare, J. E. Kokos, G. H. Miller, R. D. Ernst, and F. Wendorf. 1990. ‘‘Dating Pleistocene Archaeological Sites by Protein Diagenesis in Ostrich Eggshell.’’ Science 247: 60–64. Brothwell, D., and A. Pollard, eds. 2001. Handbook of Archaeological Sciences. New York: John Wiley and Sons. Bryson, R. 1989. ‘‘Late Quaternary Modulation of Milankovitch Climate Forcing.’’ Theoretical and Applied Climatology 39: 115–25. Bryson, R. A., and B. M. Goodman. 1980. ‘‘Volcanic Activity and Climatic Changes.’’ Science 27: 1041–44. Buck, V., and I. Stewart. 2000. ‘‘A Critical Reappraisal of the Classical Texts and Archaeological Evidence for Earthquakes in the Atalanti Region, Central Mainland Greece.’’ In The Archaeology of Geological Catastrophes, ed. W. McGuire, D. Griffiths, P. Hancock, and I. Stewart, 33–44. Special Publication 171. London: Geological Society, 2000. Buckland, P., and H. Kenward. 1973. ‘‘Thorne Moor: A Paleo-Ecological Study of a Bronze Age Site.’’ Nature 241: 405–6. Bull, I., P. Betancourt, and R. Evershed. 2001. ‘‘An Organic Geochemical Investigation of the Practice of Manuring at a Minoan Site on Pseira Island, Crete.’’ Geoarchaeology 16: 223–42. Bunker, E. 1993. ‘‘Gold in the Ancient Chinese World: A Cultural Puzzle.’’ Artibus Asiaw 53: 27–50. Burton, J., and D. Krinsley. 1987. ‘‘Obsidian Provenance Determination by Back-Scattered Electron Imaging.’’ Nature 326: 585–87. Butler, B., and E. May. 1984. Prehistoric Chert Exploitation: Studies from the Midcontinent. Carbondale, Ill.: Center for Archaeological Investigations. Butzer, K. 1964. Environment and Archaeology: An Introduction to Pleistocene Geography. Chicago: Aldine. . 1975. ‘‘The Ecological Approach to Archaeology: Are We Really Trying?’’ American Antiquity 40: 106–11. . 1978. ‘‘Toward an Integrated Contextual Approach in Archaeology: A Personal View.’’ Journal of Archaeological Science 5: 191–93. . 1980. ‘‘Holocene Alluvial Sequences: Problems of Dating and Correlation.’’ In Timescales in Geomorphology, ed. J. Lewin, D. Davidson, and R. Cullingford, 131–41. New York: John Wiley and Sons.

Bibliography . 1981. ‘‘Cave Sediments, Upper Pleistocene Stratigraphy and Mousterian Facies in Cantabrian Spain.’’ Journal of Archaeological Science 8: 133–83. . 1982. Archaeology as Human Ecology. New York: Cambridge University Press. . 1985. ‘‘Response to Presentation of the Archaeological Geology Division Award.’’ Bulletin of the Geological Society of America 97: 1397–98. Buvit, I., M. Waters, M. Konstantinov, and A. Konstantinov. 2003. ‘‘Geoarchaeological Investigations at Studenoe, and Upper Paleolithic Site in the Transbaikal Region, Russia.’’ Geoarchaeology 18: 649–73. Byrne, R., and J. McAndrews. 1975. ‘‘Precolumbian Purslane (Portulaca oleracea L.) in the New World.’’ Nature 253: 726–27. Cackler, P., M. Glascock, H. Neff, H. Iceland, K. Pyburn, D. Hudler, T. Hester, and B. Chiarulli. 1999. ‘‘Chipped Stone Artifacts, Source Areas, and Provenance Studies of the Northern Belize Chert-Bearing Zone.’’ Journal of Archaeological Science 26: 389–97. Caldwell, J. 1959. ‘‘The New American Archaeology.’’ Science 129: 303–7. Canti, M. 2003. ‘‘Earthworm Activity and Archaeological Stratigraphy: A Review of Products and Processes.’’ Journal of Archaeological Science 30: 135–48. Capedri, S., R. Grandi, and G. Venturelli. 2003. ‘‘Trachytes Used for Paving Roman Roads in the Po Plain: Characterization by Petrographic and Chemical Parameters and Provenance of Flagstones.’’ Journal of Archaeological Science 30: 491–509. Capper, J. 1907. Archaeologia 60: photographic pls. 69, 70. Carr, C. 1982. Handbook on Soil Resistivity Surveying. Evanston, Ill.: Center for American Archaeology Press. Carter, S., and D. Davidson. 1998. ‘‘An Evaluation of the Contribution of Soil Micromorphology to the Study of Ancient Arable Agriculture.’’ Geoarchaeology 13: 535–37. Casteel, R. 1974. ‘‘On the Remains of Fish Scales from Archaeological Sites.’’ American Antiquity 39: 557–59. . 1976. Fish Remains in Archaeology and Paleoenvironmental Studies. London: Academic Press, 1976. Caton-Thompson, G. 1934. ‘‘Geological Introduction.’’ In The Desert Fayum, ed. G. Caton-Thompson and E. W. Gardner, 12–18. Gloucester, U.K.: John Bellows-Royal Anthropological Institute. Chang, K. 1980. Shang Civilization. 3rd ed., rev. New Haven and London: Yale University Press. Chapman, H., and R. Van de Noort. 2001. ‘‘HighResolution Wetland Prospection, Using GPS and GIS: Landscape Studies at Sutton Common (South

Yorkshire and Meare Village East) Somerset.’’ Journal of Archaeological Science 28: 365–75. Chávez, R., M. Cámara, A. Tejero, L. Barba, and L. Manzanilla. 2001. ‘‘Site Characterization by Geophysical Methods in the Archaeological Zone of Teotihuacan, Mexico.’’ Journal of Archaeological Science 28: 1265–76. Cherry, J. 1983. ‘‘Frogs Round the Pond: Perspectives on Current Archaeological Survey Projects in the Mediterranean Area.’’ In Archaeological Survey in the Mediterranean Area, ed. D. Keller and D. Rupp, 375– 416. BAR International Series 155. Childe, V. 1928. The Most Ancient Near East: The Oriental Prelude to European Prehistory. London: Kegan Paul. Church, T. 1995. ‘‘Comment on ‘Neutron Activation Analysis of Stone from the Chadron Formation and a Clovis Site on the Great Plains,’ by Hoard et al. (1992).’’ Journal of Archaeological Science 22: 1–5. Claridge, G. 1984. ‘‘Pottery and the Pacific: The Clay Factor.’’ New Zealand Journal of Archaeology 6: 37–46. Clark, A. 1992. ‘‘Archaeogeophysical Prospecting on Alluvium.’’ In Alluvial Archaeology in Britain, ed. S. Needham and M. Macklin, 43–49. Oxbow Monographs 27. Oxford: Oxbow. Clark, D., and A. McFadyen Clark. 1993. Batza Tena, Trail to Obsidian: Archaeology at an Alaskan Obsidian Source. Archaeological Survey of Canada, Mercury Series Paper 147. Hull, Quebec: Canadian Museum of Civilization. Clark, G. 1954. Excavations at Starr Carr. [1952.] Rpt. Cambridge: Cambridge University Press. Clark, J. 1960. ‘‘Human Ecology during the Pleistocene and Later Times in Africa South of the Sahara.’’ Current Anthropology 1: 307–24. . 2001. Kalambo Falls Prehistoric Site: The Earlier Cultures: Middle and Earlier Stone Age. Vol. 3. Cambridge: Cambridge University Press. Clark, J. D., Y. Beyene, G. WoldeGabriel, W. K. Hart, P. R. Renne, G. Gilbert, A. DeFleur, G. Suwa, S. Katoch, K. R. Ludwig, J.-R. Boisserie, B. Asfaw, and T. D. White. 2003. ‘‘Stratigraphic, Chronological and Behavioral Contexts of Pleistocene Homo sapiens from Middle Awash, Ethiopia.’’ Nature 423: 747–52. Clarke, D. 1979. Analytical Archaeology. New York: Academic Press. Cohen, C. 1998. ‘‘Charles Lyell and the Evidences of the Antiquity of Man.’’ In Lyell: The Past Is the Key to the Present, ed. D. Blundell and A. Scott, 83–93. Special Publication 143. London: Geological Society. Coil, J., M. Korstanje, S. Archer, and C. Hastorf. 2003. ‘‘Laboratory Goals and Considerations for Multiple

295

Bibliography Microfossil Extraction in Archaeology.’’ Journal of Archaeological Science 30: 991–1008. Cole, S., B. Atwater, P. McCutcheon, J. Stein, and E. Hemphill-Haley. 1996. ‘‘Earthquake-Induced Burial of Archaeological Sites along the Southern Washington Coast about a.d. 1700.’’ Geoarchaeology 11: 165–97. Coleman, J. ‘‘Deltaic Evolution.’’ 1968. In Encyclopedia of Geomorphology, ed. R. W. Fairbridge, 255–60. New York: Reinhold. Coles, B., and J. Coles. 1989. People of the Wetlands. New York: Thames and Hudson. Collins, M., B. Carter, B. Gladfelter, and R. Southard, eds. 1995. Pedological Perspectives in Archaeological Research. Special Publication 44. Madison, Wisc.: Soil Science of America. Colls, A., S. Stokes, M. Blum, and E. Straffin. 2001. ‘‘Age Limits on the Late Quaternary Evolution of the Upper Loire River.’’ Quaternary Science Reviews 20: 743–50. Conard, N. J., and M. Bolus. 2003. ‘‘Radiocarbon Dating the Appearance of Modern Humans and Timing of Cultural Innovations in Europe: New Results and New Challenges.’’ Journal of Human Evolution 44: 331–71. Conyers, L., and C. Cameron. 1998. ‘‘Ground-Penetrating Radar Techniques and Three-Dimensional Computer Mapping in the American Southwest.’’ Journal of Field Archaeology 25: 417–30. Conyers, L., and D. Goodman. 1997. Ground-Penetrating Radar: An Introduction for Archaeologists. Walnut Creek, Calif.: Altameria Press. Cornwall, I. 1958. Soils for the Archaeologist. London: Phoenix House. Corte, A. 1962. ‘‘Vertical Migration of Particles in Front of a Freezing Plane.’’ Journal of Geophysical Research 67: 1085–90. . 1963. ‘‘Particle Sorting by Repeated Freezing and Thawing.’’ Science 142: 499–501. Cottrell, M. 1985. ‘‘Tomato Springs: The Identification of a Jasper Trade and Production Center in Southern California.’’ American Antiquity 50: 833–49. Courty, M.-A. 2001. ‘‘Microfacies Analysis Assisting Archaeological Stratigraphy.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 205–39. New York: Kluwer Academic/ Plenum. Courty, M.-A., P. Goldberg, and R. Macphail. 1989. Soils and Micromorphology in Archaeology. New York: Cambridge University Press. Crawford, O. 1923 ‘‘Air Survey and Archaeology.’’ Geographical Journal 58: 324–66.

296

. 1929. Air-Photography for Archaeologists. London: Ordnance Survey Professional Papers, n.s. Cremeens, D. 2003. ‘‘Geoarchaeology of Soils on Stable Geomorphic Surfaces: Mature Soil Model for the Glaciated Northeast.’’ In Geoarchaeology of Landscapes in the Glaciated Northeast, ed. D. L. Cremeens and J. P. Hart, 49–60. New York State Museum Bulletin 497. Albany, N.Y. Cremeens, D., and J. Hart. 1995. ‘‘On Chronostratigraphy, Pedostratigraphy, and Archaeological Context.’’ In Pedological Perspectives in Archaeological Research, ed. M. E. Collins, B. J. Carter, B. G. Gladfelter, and R. J. Southard, 15–33. Special Publication 44. Madison, Wisc.: Soil Science of America. , eds. 2003. Geoarchaeology of Landscapes in the Glaciated Northeast. New York State Museum Bulletin 497. Albany, N.Y. Crowley, D. 2002. ‘‘A Case Study in the Analysis of patterns of Aerial Reconnaissance in a Lowland Area of Southwest Scotland.’’ Archaeological Prospection 9: 255–65. Crowther, J. 1997. ‘‘Soil Phosphate Surveys Critical Approaches to Sampling, Analysis, and Interpretation.’’ Archaeological Prospection 4: 93–102. Cruz-Uribe, K., R. Klein, G. Avery, M. Avery, D. Halkett, T. Hart, R. Milo, C. Sampson, and T. Volman. 2003. ‘‘Excavation of Buried Late Acheulian (MidQuaternary) Land Surfaces at Duinefontein 2, Western Cape Province, South Africa.’’ Journal of Archaeological Science 30: 559–75. Culling, W. 1963. ‘‘Soil Creep and the Development of Hillside Slopes.’’ Journal of Geology 71: 127–62. Curran, J., I. Meighan, D. Simpson, G. Rogers, and A. Fallick. 2001. ‘‘ 87 Sr/ 86 Sr: A New Discriminant for Provenancing Neolithic Porcellanite Artifacts from Ireland.’’ Journal of Archaeological Science 28: 713–20. Custer, J., T. Eveleigh, V. Klemas, and I. Wells. 1986. ‘‘Application of Landsat Data and Synoptic Remote Sensing to Predictive Models for Prehistoric Archaeological Sites: An Example from the Delaware Coastal Plain.’’ American Antiquity 51(3): 572–88. Dahms, D., and V. Holliday. 1998. ‘‘Soil Taxonomy and Paleoenvironmental Reconstruction: A Critical Commentary.’’ Quaternary International 51/52: 109– 14. Dalan, R. 1993. ‘‘Landscape Modification at the Cahokia Mounds Site: Geophysical Evidence of Culture Change.’’ Ph.D. diss., University of Minnesota. . 2001. ‘‘A Magnetic Susceptibility Logger for Archaeological Application.’’ Geoarchaeology 16: 263– 73. Dalan, R., and S. Banerjee. 1998. ‘‘Solving Archaeologi-

Bibliography cal Problems Using Techniques of Soil Magnetism.’’ Geoarchaeology 13: 3–36. Dalan, R., and B. Bevan. 2002. ‘‘Geophysical Indicators of Culturally Emplaced Soils and Sediments.’’ Geoarchaeology 17: 779–810. Dalfes, H., G. Kukla, and H.Weiss, eds. 1997. Third Millennium BC Climate and Old World Collapse. Berlin: Springer. Daniel, G. 1976. A Hundred and Fifty Years of Archaeology. 2nd ed. Cambridge: Harvard University Press. Daniel, G., and C. Renfrew. 1988. The Idea of Prehistory. 2nd ed. Edinburgh: Edinburgh University Press. Darwin, Charles. 1989. ‘‘The Formation of Vegetable Mould through the Action of Worms, with Observations on Their Habits.’’ [1881.] In The Works of Charles Darwin, vol. 28. Washington Square: New York University Press. Davidson, D. 1980. ‘‘Erosion in Greece during the First and Second Millennia b.c.’’ In Timescales in Geomorphology, ed. J. Lewin, D. A. Davidson, and R. Cullingford, 148–58. New York: John Wiley and Sons. Davidson, D., and M. Shackley, eds. 1976. Geoarchaeology. Boulder Colo.: Westview Press. Davidson, D., and I. Simpson. 2001. ‘‘Archaeology and Soil Micromorphology.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 167–77. New York: John Wiley and Sons. Davis, J. 1998. From Homer’s Sandy Pylos to the Battle of Navarino: An Archaeological Survey. Austin: University of Texas Press. Davis, L. 1984. ‘‘Late Pleistocene to Mid-Holocene Adaptations at Indian Creek, West-Central Montana Rockies.’’ Current Research in the Pleistocene 1: 9–10. Davis, L., and S. Greiser. 1992. ‘‘Indian Creek Paleoindians: Early Occupation of the Elkhorn Mountains’ Southeast Flank, West-Central Montana.’’ In Ice Age Hunters of the Rockies, ed. D. J. Stanford and J. S. Day, 225–83. Denver: Denver Museum of Natural History and the University Press of Colorado. Davis, L., S. Aaberg, and A. Johnson. 1992. ‘‘Archaeological Fieldwork at Yellowstone’s Obsidian Cliff.’’ Park Service Resource Management Bulletin 12: 26–27. Dawson, C. 1913. ‘‘The Piltdown Skull.’’ Hastings and East Sussex Naturalist 2: 73–82. Dean, W. 1974. ‘‘Determination of Carbonate and Organic Matter in Calcareous Sediments and Sedimentary Rocks by Loss-on-Ignition: Comparison with Other Methods.’’ Journal of Sedimentary Petrology 44: 242–48. De Boer, J., and J. Hale. 2000. ‘‘The Geologic Origins of the Oracle at Delphi, Greece.’’ In The Archaeology of Geological Catastrophes, ed. W. McGuire, D. Griffiths,

P. Hancock, and I. Stewart, 399–412. Special Publication 171. London: Geological Society. De Boer, J., and D. Sanders. 2002. Volcanoes in Human History: The Far-Reaching Effects of Major Eruptions. Princeton, N.J.: Princeton University Press. De Fourestier, J. 1999. Glossary of Mineral Synonyms. Canadian Mineralogist Special Publication 2. Ottawa: Mineralogical Association of Canada. De Geer, G. 1937. ‘‘Early Man and Geochronology.’’ In Early Man, ed. G. MacCurdy, 323–26. Philadelphia: Lippincott. . 1940. Geochronology Suecica Principles. Stockholm: Almqvist and Wiksells. De Morgan, J. 1924. Prehistoric Man: A General Outline of Prehistory. New York: Knopf. de Silva, S., and J. Alzueta. 2000. ‘‘The Socioeconomic Consequences of the a.d. 1600 Eruption of Huaynaputina, Southern Peru.’’ In Volcanic Hazards and Disasters in Human Antiquity, ed. F. W. McCoy and G. Heiken, 15–24. Special Paper 345. Boulder, Colo.: Geological Society of America. De Terra, H. 1934. ‘‘Geology and Archaeology as Border Sciences.’’ Science 80: 447–49. Deocampo, D. M., R. J. Blumenschine, and G. M. Ashley. 2002. ‘‘Wetland Diagenesis and Traces of Early Hominids, Olduvai Gorge, Tanzania.’’ Quaternary Research 57(2): 271–81. Desautels, P. 1986. The Jade Kingdom. New York: Van Nostrand Reinhold. Dickinson, W. 2001. ‘‘Petrography and Geologic Provenance of Sand Tempers in Prehistoric Potsherds from Fiji and Vanuatu, South Pacific.’’ Geoarchaeology 16: 275–322. Dimbleby, G. 1985. The Palynology of Archaeological Sites. London: Academic Press. Dincauze, D. 1976. The Neville Site: Eight Thousand Years at Amoskeag. Peabody Museum Monographs 4. Cambridge: Harvard University Press. . 1987. ‘‘Strategies for Paleoenvironmental Reconstruction in Archaeology.’’ In Advances in Archaeological Method and Theory, ed. M. Schiffer, 11:255–336. New York: Academic Press. . 2000. Environmental Archaeology: Principles and Practice. Cambridge: University Press. Dittert, A. 1968. ‘‘Minerals and Rocks at Archaeological Sites: Some Interpretations from Central Western New Mexico.’’ Arizona Archaeologist 3: 1–16. Doig, R. 1990. ‘‘2,300-Year History of Seismicity from Silting Events in Lake Tadoussac, Charlevoix, Quebec.’’ Geology 18: 820–23. Dolan, J., S. Christofferson, and J. Shaw. 2003. ‘‘Recog-

297

Bibliography nition of Paleoearthquakes on the Puente Hills Blind Thrust Fault, California.’’ Science 300: 115–18. Donahue, J., and J. M. Adavasio. 1990. ‘‘Evolution of Sandstone Rockshelters in Eastern North America: A Geoarchaeological Perspective.’’ In Archaeological Geology of North America, ed. N. P. Lasca and J. Donahue, 231–51. Boulder, Colo.: Geological Society of America. Donoghue, D. 2001. ‘‘Remote Sensing.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 555–63. New York: John Wiley and Sons. Doolittle, W. 1990. Canal Irrigation in Prehistoric Mexico: The Sequence of Technological Change. Austin: University of Texas Press. Dorais, M., and C. Shriner. 2002. ‘‘A Comparative Electron Microprobe Study of ‘Aeginetan’ Wares with Potential Raw Material Sources from Aegina, Methana, and Poros, Greece.’’ Geoarchaeology 17: 555–77. Dort, W., Jr., and S. Miller. 1977. Archaeological Geology of Birch Creek Valley and the Eastern Snake River Plain, Idaho. First Annual Field Trip. Idaho Falls, ID: Robco Printing, Division of Archaeological Geology, Geological Society of America. Douglass, A. 1919. Climate Cycles and Tree Growth. Washington, D.C.: Carnegie Institution of Washington. . 1935. Dating Pueblo Bonito and Other Ruins in the Southwest. Contributed Technical Papers, Pueblo B Ser. 1. Washington, D.C.: National Geographic Society. Dundes, A. 1988. The Flood Myth. Berkeley: University of California Press. Dunnell, R. 1971. Systematics in Prehistory. New York: Free Press. . 1982. ‘‘Science, Social Science, and Common Sense: The Agonizing Dilemma of Modern Archaeology.’’ Journal of Anthropological Research 38: 1–25. Dye, T., and W. Dickinson. 1996. ‘‘Sources of Sand Tempers in Prehistoric Tongan Pottery.’’ Geoarchaeology 11(2): 141–64. Easton, R., L. Edwards, and B. Wardlaw. 2003. ‘‘Discussion and Reply: Notes on Geochronologic and Chronostratigraphic Units.’’ Geological Society of America Bulletin 115: 1016–19. Eidt, R. 1984. Advances in Abandoned Settlement Analysis: Application to Prehistoric Anthrosols in Columbia, South America. Milwaukee: Center for Latin America, University of Wisconsin. . 1985. ‘‘Theoretical and Practical Considerations in the Analysis of Anthrosols.’’ In Archaeological Geology, ed. G. Rapp and J. Gifford, 155–90. New Haven and London: Yale University Press.

298

Eighmy, J. 2000. ‘‘Thirty Years of Archaeomagnetic Dating.’’ In It’s About Time: A History of Archaeological Dating in North America, ed. S. Nash, 105–23. Salt Lake City: University of Utah Press. Eighmy, J., and J. Howard. 1991. ‘‘Direct Dating of Prehistoric Canal Sediments Using Archaeomagnetism.’’ American Antiquity 56(1): 88–102. Eighmy, J., and R. Sternberg. 1991. Archaeomagnetic Dating. Tucson: University of Arizona Press. Elbaum, R., S. Weiner, R. Albert, and M. Elbaum. 2003. ‘‘Detection of Burning of Plant Materials in the Archaeological Record by Changes in the Refraction Index of Siliceous Phytoliths.’’ Journal of Archaeological Science 30: 217–26. Elias, S. 1990. ‘‘The Timing and Intensity of Environmental Changes during the Paleoindian Period in Western North America: Evidence from the Insect Fossil Record.’’ In Megafauna and Man, ed. L. Agenbroad, J. Mead, and L. Nelson, 11–14. Scientific Papers, Vol. 1. Hot Springs: Mammoth Site of Hot Springs, South Dakota. . 1994. Insects and Their Environments. Washington, D.C.: Smithsonian Institution Press. Erickson, C. 1992. ‘‘Prehistoric Landscape Management in the Andean Highlands: Raised Field Agriculture and Its Environmental Impact.’’ Population and Environment 13: 285–300. Evans, A. 1964. The Palace of Minos at Knossos. Vol. 4, pt. 2. New York: Biblio and Tannen. Evans, J. 1972. Land Snails in Archaeology. London: Seminar Press. Evernden, J. F., and G. H. Curtis. 1965. ‘‘The Potassium-Argon Dating of Late Cenozoic Rocks in East Africa and Italy.’’ Current Anthropology 6:343–385. Faegri, K., P. Kaland, and K. Krzywinski. 1989. Textbook of Pollen Analysis. 4th ed. Chichester, U.K.: John Wiley and Sons. Fagan, B. 2000. The Little Ice Age: How Climate Made History. New York: Basic Books. Fairbanks, R., R. Mortlock, T.-C. Chiu, L. Cao, A. Kaplan, T. Guilderson, T. Fairbanks, A. Bloom, P. Grootes, M.J-J. Nadeau. 2005. ‘‘Radiocarbon Calibration Curve Spanning 0 to 50,000 Years BP Based on Paired 230 Th/ 234 U/ 238 U and 14 C Dates on Pristine Corals.’’ Quaternary Science Reviews 24: 1781–1796. Fanning, P., and S. Holdaway. 2001. ‘‘Stone Artifact Scatters in Western NSW, Australia: Geomorphic Controls on Artifact Size and Distribution.’’ Geoarchaeology 16: 667–86. Farizy, C. 1994. ‘‘Spatial Patterning of Middle Paleolithic Sites.’’ Journal of Anthropological Archaeology 13: 153–60.

Bibliography Farrand, W. 1960. ‘‘Former Shorelines in Western and Northern Lake Superior Basin.’’ Ph.D. diss., University of Michigan. Ann Arbor: University Microfilm. . 1993. ‘‘Discontinuity in the Stratigraphic Record: Snapshots from Franchthi Cave.’’ In Formation Processes in Archaeological Context, ed. Goldberg, P., D. Nash, and M. Petraglia, 85–96. Madison, Wisc.: Prehistory Press. . 2001. ‘‘Sediments and Stratigraphy in Rockshelters and Caves: A Personal Perspective on Principles and Pragmatics.’’ Geoarchaeology 16: 537–57. Farrand, W. and C. Dexler. 1985. ‘‘Late Wisconsin Man and Holocene History of the Lake Superior Basin.’’ In Quaternary Evolution of the Great Lakes, ed. P. Karrow and P. Calkin, 17–32. Geological Association of Canada Special Paper 30. Saint John’s, NL: GAC Publications. Farquhar, R., and I. Fletcher. 1984. ‘‘The Provenience of Galena from Archaic/Woodland Sites in Northeastern North America: Lead Isotope Evidence.’’ American Antiquity 49: 774–85. Farquhar, R., J. Walthall, and R. Hancock. 1995. ‘‘Eighteenth-Century Lead Smelting in Central North America: Evidence from Lead Isotope and INAA Measurements.’’ Journal of Archaeological Science 22: 639–48. Faulkner, C. 1962. ‘‘The Significance of Some Red Ochre-like Artifacts from Lake County, Indiana.’’ Wisconsin Archaeologist 41: 1–8. Feathers, J. 1997a. ‘‘Luminescence Dating of Early Mounds in Northeast Louisiana.’’ Quaternary Science 16: 333–40. . 1997b. ‘‘The Application of Luminescence Dating in American Archaeology.’’ Journal of Archaeological Method and Theory 4: 1–66. . 2000. ‘‘Luminescence Dating and Why It Deserves Wider Application.’’ In It’s About Time: A History of Archaeological Dating in North America, ed. S. Nash, 152–66. Salt Lake City: University of Utah Press. Feathers, J., and C. Hill. 2003. ‘‘Luminescence Dating of Glacial Lake Great Falls, Montana, U.S.A.’’ XVI INQUA Congress Programs with Abstracts, 299. Reno, Nev.: Desert Research Institute. Fedele, F. 1976. ‘‘Sediments as Paleo-Land Segments: The Excavation Side of Study.’’ In Geoarchaeology, ed. D. Davidson and M. L. Shackley, 23–48. Boulder, Colo.: Westview Press. Feng, Z., W. Johnson, D. Sprowl, and Y. Lu. 1994. ‘‘Loess Accumulation and Soil Formation in Central Kansas, United States, during the Past Four Hundred

Thousand Years.’’ Earth Surface Processes and Landforms 19: 55–67. Ferguson, L. 1974. ‘‘Prehistoric Mica Mines in the Southern Appalachians.’’ South Carolina Antiquities 6: 1–9. Ferring, C. R. 1986. ‘‘Rates of Fluvial Sedimentation: Implications for Archaeological Variability.’’ Geoarchaeology 1: 259–74. . 1992.‘‘Alluvial Pedology and Geoarchaeological Research.’’ In Soils in Archaeology, ed. V. Holliday, 1– 39. Washington, D.C.: Smithsonian Institution Press. . 1994. Review of Principles of Geoarchaeology: A North American Perspective, by Michael R. Waters. American Anthropologist 96(1): 218–19. . 2000. ‘‘Geoarchaeology in the Southern Osage Plains.’’ In Geoarchaeology in the Great Plains, ed. R. Mandel, 44–78. Norman: University of Oklahoma Press. . 2001. ‘‘Geoarchaeology in Alluvial Landscapes.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 77–106. New York: Kluwer Academic/Plenum Publishers. Ferring, C. R., and T. Pertula. 1987. ‘‘Defining the Provenance of Red Slipped Pottery from Texas and Oklahoma by Petrographic Methods.’’ Journal of Archaeological Science 14: 437–56. Finlayson, W. Byrne, and J. McAndrews. 1973. ‘‘Iroquoian Settlement and Subsistence Patterns near Crawford Lake, Ontario.’’ Canadian Archaeological Association Bulletin 5: 134–36. Fish, S., and P. Fish. 1992. ‘‘Prehistoric Landscapes of the Sonoran Desert Hohokam.’’ Population and Environment 13: 269–83. Fladmark, K. 1982. ‘‘Microdebitage Analysis: Initial Considerations.’’ Journal of Archaeological Science 9: 205–20. Foit, F., P. Mehringer, and J. Sheppard. 1993. ‘‘Age, Distribution, and Stratigraphy of Glacier Peak Tephra in Eastern Washington and Western Montana, United States.’’ Canadian Journal of Earth Sciences 30: 535–52. Foster, J., and J. Whitney. 1850. Report on the Geology and Topography of a Portion of the Lake Superior Land District in the State of Michigan. No. 69, pt. 1: Copper Lands. Washington, D.C.: House Executive Documents. Fox, C., A. Perez-Perez, and J. Juan. 1994. ‘‘Dietary Information through the Examination of Plant Phytoliths on the Enamel Surface of Human Dentition.’’ Journal of Archaeological Science 21: 29–34. Francis, P. 1988. ‘‘Simojovel, Mexico: Village of Amber.’’ Lapidary Journal 42(8): 55–62.

299

Bibliography Frederick, C. 2001. ‘‘Evaluating Causality of Landscape Change: Examples from Alluviation.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 55–76. New York: Kluwer Academic/Plenum Publishers. Freeman, A. K. L. 2000. ‘‘Application of High-Resolution Alluvial Stratigraphy in Assessing the HunterGatherer/Agricultural Transition in Santa Cruz River Valley, Southeastern Arizona.’’ Geoarchaeology 15(6): 559–89. French, C. 2003. Geoarchaeology in Action: Studies in Soil Micromorphology and Landscape Evolution. New York: Routledge. Frere, J. 1800. ‘‘Account of Flint Weapons Discovered at Hoxne in Suffolk.’’ Archaeologia 13: 204–5. Fritts, H. 1976. Tree-Rings and Climate. London: Academic Press. Fritts, H., G. Lofgren, and G. A. Gordon. 1979. ‘‘Variations in Climate since 1602 as Reconstructed from Tree-Rings.’’ Quaternary Research 12: 18–46. Fryxell, R. 1977. The Interdisciplinary Dilemma: A Case for Flexibility in Academic Thought. Augustana College Library Occasional Paper 13. Davenport, Iowa: Augustana College. Gabunia, L, V. Abesalom, D. Lordkipanidze, C. Swisher III, C. R. Ferring, A. Justus, M. Nioradze, M. Tvalchrelidze, S. C. Anton, G. Bosinski, O. Joris, M.-A. de-Limley, G. Majsuradze, and A. Mouskhelishvili. 2000. ‘‘Earliest Pleistocene Hominid Cranial Remains from Manisi, Republic of Georgia: Taxonomy, Geologic Setting, and Age.’’ Science 288: 1019–25. Gaffney, C., J. Gater, and S. Ovenden. 1991. The Use of Geophysical Techniques in Archaeological Evaluations. IFA Technical Paper 9. Birmingham, U.K.: Institute of Field Archaeologists. Galli, P., and F. Galadini. 2003. ‘‘Disruptive Earthquakes Revealed by Faulted Archaeological Relics in Samnium (Molise, Southern Italy).’’ Geophysical Research Letters 30: 1266ff. Gamio, M. 1913. ‘‘Arqueologia de Atzcapotzalco, D.F., Mexico.’’ In Proceedings, Eighteenth International Conference of Americanists, 180–87. London. Gardner, E. W. 1934. ‘‘Geological Introduction.’’ In G. Caton-Thompson and E. W. Gardner, The Desert Fayum, 12–18. Gloucester, U.K.: John Bellows-Royal Anthropological Institute. Garrison, E. G. 2003. Techniques in Archaeological Geology. Heidelberg: Springer-Verlag. Gatto, L., and W. Doe III. 1983. Historical Bank Recession at Selected Sites along Corps of Engineers Reservoirs. Special Report 83-30. Hanover, N.H.: U.S. Army Cold Regions Research and Engineering Laboratory.

300

Geikie, J. 1877. The Great Ice Age and Its Relation to the Antiquity of Man. 2nd ed. London: Daldy, Isbista. Gendron, F., D. Smith, and A. Gendron-Badou. 2002. ‘‘Discovery of Jadeite-Jade in Guatemala. Confirmed by Non-Destructive Raman Microscopy.’’ Journal of Archaeological Science. 29: 837–51. Gettens, R., and G. Stout. 1966. Painting Materials: A Short Encyclopedia. New York: Dover. Geyh, M., M. Grosjean, L. Núñez, and U. Schotterer. 1999. ‘‘Radiocarbon Reservoir Effect and the Timing of the Late-Glacial/Early Holocene Humid Phase in the Atacama Desert (Northern Chile).’’ Quaternary Research 52: 143–53. Gibbard, P. 1994. Pleistocene History of the Lower Thames Valley. Cambridge: Cambridge University Press. Gifford, D. 1981. ‘‘Taphonomy and Paleoecology: A Critical Review of Archaeology’s Sister Disciplines.’’ In Advances in Archaeological Method and Theory, ed. M. Schiffer, 4: 365–438. New York: Academic Press. Gifford, J., and G. Rapp. 1985a. ‘‘The Early Development of Archaeological Geology in North America.’’ In Geologists and Their Ideas: A History of North American Geology, ed. E. Drake and W. Jordon, 409–21. Boulder, Colo.: Geological Society of America,. Gifford, J., and G. Rapp. 1985b. ‘‘History, Philosophy, and Perspectives.’’ In Archaeological Geology, ed. G. Rapp and J. Gifford, 1–23. New Haven and London: Yale University Press. Gifford, J., G. Rapp, and C. Hill. 1989. ‘‘Site Geology.’’ In Excavations at Tel Michal, Israel, ed. Z. Herzog, G. Rapp, and O. Negbi, 209–18. Minneapolis: University Minnesota Press. Gilbert, G. 1884. ‘‘The Topographic Features of Lake Shores.’’ Report of the Director of the U.S. Geological Survey 5: 75–123. Gillespie, S., and D. Nichols, eds. 2003. Archaeology Is Anthropology. Archaeological Papers of the American Anthropological Association 13. Arlington, Va.: American Anthropological Association. Gillings, M. 2001. ‘‘Spatial Information and Archaeology.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 671–83. New York: John Wiley and Sons, 2001. Gladfelter, B. 1977. ‘‘Geoarchaeology: The Geomorphologist and Archaeology.’’ American Antiquity 42 (4): 519–38. . 1981. ‘‘Developments and Directions in Geoarchaeology.’’ In Advances in Archaeological Method and Theory, ed. M. Schiffer, 4: 343–64. New York: Academic Press. . 1985. ‘‘On the Interpretation of Archaeologic Sites in Alluvial Settings.’’ In Archaeological Sediments

Bibliography in Context, ed. J. Stein and W. Farrand, 41–52. Orono: University of Maine, Center for the Study of Early Man. Gleadow, A. 1980. ‘‘Fission Track Age of the KBS Tuff and Associated Hominids in Northern Kenya.’’ Nature 284: 225–30. Gleeson, P., and G. Grosso. 1976. ‘‘The Ozette Site.’’ In Excavation of Water-Saturated Archaeological Sites (Wet Sites) on the Northwest Coast of North America, ed. D. R. Croes, 13–44. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper 50. Ottawa: National Museums of Canada. Goff, J., and B. McFadgen. 2003. ‘‘Large Earthquakes and the Abandonment of Prehistoric Coastal Settlements in Fifteenth-Century New Zealand.’’ Geoarchaeology 18: 609–23. Goldberg, P., and I. Whitbread. 1993. ‘‘Micromorphological Study of a Bedouin Tent Floor.’’ In Formation Processes in Archaeological Context, ed. P. Goldberg, D. Nash, and M. Petraglia, 165–88. Madison, Wisc.: Prehistory Press. Goldberg, P. and T. Arpin. 1999. ‘‘Micromorphological Analysis of Sediments from Meadowcroft Rockshelter, Pennsylvania: Implications for Radiocarbon Dating.’’ Journal of Field Archaeology 26: 325–42. Goldberg, P., V. Holliday, and C. R. Ferring. 2001. Earth Sciences and Archaeology. New York: Kluwer Academic/Plenum Publishers. Goudie, A. 2000. The Human Impact on the Natural Environment. Cambridge, Mass.: MIT Press. Gramly, R. 1992. Prehistoric Lithic Industry at Dover, Tennessee. Buffalo, N.Y.: Persimmon Press. Graubau, A. 1960. Principles of Stratigraphy. [1924.] Rpt. New York: Dover. Grayson, D. 1986a. ‘‘Eoliths, Archaeological Ambiguity and the Generation of ‘Middle-Range’ Research.’’ In American Archaeology, Past and Future, ed. D. J. Meltzer, D. D. Fowler, and J. A. Sabloff. Washington, D.C.: Smithsonian Institution Press. . 1986b. The Establishment of Human Antiquity. New York: Academic Press. . 1990. ‘‘The Provision of Time Depth for Paleoanthropology.’’ In Establishment of a Geologic Framework for Paleoanthropology, ed. L. F. Laporte, 1–13. Special Paper 242. Boulder, Colo.: Geological Society of America. Greenough, J., M. Gorton, and L. Mallory-Greenough. 2001. ‘‘The Major-and Trace-Element Whole-Rock Fingerprints of Egyptian Basalts and the Provenance of Egyptian Artifacts.’’ Geoarchaeology 16: 763–84. Griffiths, D. 2000. ‘‘Uses of Volcanic Products in Antiquity.’’ In The Archaeology of Geological Catastro-

phes, ed. W. McGuire, D. Griffiths, P. Hancock, and I. Stewart, 15–23. Special Publications 171. London: Geological Society. Gruber, J. 1965. ‘‘Brixham Cave and the Antiquity of Man.’’ In Context and Meaning in Cultural Anthropology, ed. M. Spiro, 373–402. New York: Free Press. Grun, R. 2001. ‘‘Trapped Charge Dating (ESR, TL, OSL).’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 47–62. New York: John Wiley and Sons. Gunn, J. 1992. ‘‘Regional Climatic Mechanisms of the Clovis-Phase on the Southern High Plains.’’ In Proboscidean and Paleoindian Interactions, ed. J. W. Fox, C. B. Smith, and K. T. Wilkins, 171–89. Waco, Tex.: Baylor University Press. Gupta, R. 2003. Remote Sensing Geology. 2nd ed. Berlin: Springer-Verlag. Hajic, E. 1990. Koster Site Archaeology I: Stratigraphy and Landscape Evolution. Research Series Vol. 8. Kampsville, Ill.: Center for American Archaeology. Hall, C., and D. York. 1984. ‘‘The Applicability of 40Ar/ 39Ar Dating to Younger Volcanics.’’ In Quaternary Dating Methods, ed. W. Mahaney, 67–74. Amsterdam: Elsevier. Hammond, E. 1964. ‘‘Analysis of Properties in Landform Geography.’’ Annals of the Association of American Geographers 54: 11–19. Hammond, N. 1989. ‘‘Obsidian Hydration Dating of Tecep Phase Occupation at Nohmul, Belize.’’ American Antiquity 54(3): 513–21. Hancock, R., L. Pavlish, R. Farquhar, R. Salloum, W. Fox, and G. Wilson. 1991. ‘‘Distinguishing European Trade Copper and Eastern North American Native Copper.’’ Archaeometry 33: 69–86. Harbottle, G. 1990. ‘‘Neutron Activation Analysis in Archaeological Chemistry.’’ In Chemical Applications of Nuclear Probes, ed. K. Yoshihara, 58–91. Berlin: Springer-Verlag. Harley, J., and D. Woodward, eds. 1987. The History of Cartography. Vol. 1: Cartography in Prehistoric, Ancient, and Medieval Europe and the Mediterranean. Chicago: University of Chicago Press. Harlow, G. 1991. ‘‘The Maya Rediscovered Hard Rock.’’ Natural History 100(8): 4–10. Harrell, J., and V. Brown. 1992a. ‘‘The Oldest Surviving Topographical Map from Ancient Egypt: (Turin Papyri 1879, 1899, and 1969).’’ Journal of the American Research Center in Egypt 29: 81–105. . 1992b. ‘‘The World’s Oldest Surviving Geological Map: The 1150 b.c. Turin Papyrus from Egypt.’’ Journal of Geology 100: 3–18. Harrell, J., and M. Lewan. 2002. ‘‘Sources of Mummy

301

Bibliography Bitumen in Ancient Egypt and Palestine.’’ Archaeometry 44: 285–93. Harris, E. 1989. Principles of Archaeological Stratigraphy. 2nd ed. London: Academic Press. Harrower, M., J. McCorriston, and E. Oches. 2002. ‘‘Mapping the Roots of Agriculture in Southern Arabia: The Application of Satellite Remote Sensing, Global Positioning System and Geographic Information System Technologies.’’ Archaeological Prospection 9: 35–42. Hassan, F. 1978. ‘‘Sediments in Archaeology: Methods and Implications for Paleoenvironmental and Cultural Analysis.’’ Journal of Field Archaeology 5(2): 197– 213. . 1979. ‘‘Geoarchaeology: The Geologist and Archaeology.’’ American Antiquity 44: 267–70. . 1995. Review of ‘‘Formation Processes in Archaeologic Context,’’ ed. P. Goldberg, D. Nash, and M. Petraglia. American Antiquity 60(3): 558–59. Haug, G., D. Günther, L. Peterson, D. Sigman, K. Hughen, and B. Aeschlimann. 2003. ‘‘Climate and the Collapse of Maya Civilization.’’ Science 299: 1731– 35. Hawass, Z., and M. Lehner. 1994. ‘‘The Sphinx: Who Built It, and Why?’’ Archaeology 47(5): 30–42. Hawkes, C. 1954. ‘‘Archaeological Theory and Method: Some Suggestions from the Old World.’’ American Anthropologist 56: 155–68. Haynes, C. V. 1964. ‘‘The Geologist’s Role in Pleistocene Paleoecology and Archaeology.’’ In The Reconstruction of Past Environments, assembled by J. Hester and J. Schoenwetter, 61–66. Taos, N.M.: Fort Burgwin Research Center. . 1973. ‘‘The Calico Site: Artifacts or Geofacts?’’ Science 181: 305–10. . 1990. ‘‘The Antevs-Bryan Years and the Legacy for Paleoindian Geochronology.’’ In Establishment of a Geologic Framework for Paleoanthropology, ed. L. Laporte, 55–66. Special Paper 242. Boulder, Colo.: Geological Society of America. . 1995. ‘‘Geochronology of Paleoenvironmental Change, Clovis Type Site, Blackwater Draw, New Mexico.’’ Geoarchaeology 10: 317–88. Haynes, C. V., D. Stanford, M. Jodry, J. Dickenson, J. Montgomery, P. Shelley, I. Rovner, and G. Agogino. 1999. ‘‘A Clovis Well at the Type Site 11,500 b.c.: The Oldest Prehistoric Well in America.’’ Geoarchaeology 14: 455–70. Haynes, H. 1893. Man and the Glacial Period. New York: Appleton. Hebda, R., and R. Mathewes. 1984. ‘‘Holocene History

302

of Cedar and Native Indians of the North American Pacific Coast.’’ Science 225: 711–13. Heizer, R., and A. Treganza. 1944. ‘‘Mines and Quarries of the Indians of California.’’ California Journal of Mines and Geology 40: 291–93. Heizer, R. F., F. Stross, T. R. Hester, A. Albee, I. Perlman, F. Asaro, and H. Bowman. 1973. ‘‘The Colossi of Memnon Revisited.’’ Science 182: 1219–25. Hennig, G., R. Grun, and K. Brunnacker. 1983. ‘‘Speleothems, Travertines and Paleoclimates.’’ Quaternary Research 20: 1–29. Heron, C. 2001. ‘‘Geochemical Prospecting.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 565–73. New York: John Wiley and Sons. Herz, N. 1985. ‘‘Isotopic Analysis of Marble.’’ In Archaeological Geology, ed. G. Rapp and J. Gifford, 331– 51. New Haven and London: Yale University Press. . 2001. ‘‘Sourcing Lithic Artifacts by Instrumental Analysis.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 449–69. New York: Kluwer Academic/Plenum Publishers. Hester, J. 1987. ‘‘The Significance of Accelerator Dating in Archaeological Method and Theory.’’ Journal of Field Archaeology 14: 445–51. Hill, C. 1989. ‘‘Petrography of Quaternary Sediments in the Nile Valley of Upper Egypt.’’ In The Prehistory of Wadi Kubbaniya., Vol. 2: Stratigraphy, Paleoeconomy, and Environment, ed. A. Close, 101–13. Dallas, Tex.: Southern Methodist University Press. . 1993a. ‘‘E-87-2: A Site in Lake-Margin Deposits of the Green Phase.’’ In F. Wendorf, R. Schild, A. Close, et al., Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East, ed., 443–58. New York: Plenum Press. . 1993b. ‘‘E-87-3: A Small, Dry-Season Occupation at the Onset of the Green Phase.’’ In F. Wendorf, R. Schild, A. Close, et al., Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East, 459–70. New York: Plenum Press. . 1993c. ‘‘Sedimentology of Pleistocene Deposits Associated with Middle Paleolithic Sites in Bir Tarfawi and Bir Sahara East.’’ In F. Wendorf, R. Schild, A. Close, et al., Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East, 66–105. New York: Plenum Press. . 2001a. ‘‘Geologic Context of the Acheulian (Middle Pleistocene) in the Eastern Sahara.’’ Geoarchaeology 16: 65–94. . 2001b. ‘‘Pleistocene Mammals of Montana and Their Geologic Context.’’ In Mesozoic and Cenozoic Paleontology in the Western Plains and Rocky Mountains,

Bibliography ed. C. Hill, 127–44. Museum of the Rockies Occasional Paper 3. Bozeman: Montana State University. . 2001c. ‘‘Pleistocene Stratigraphy, Chronology, and Taphonomy of a Typical Mousterian (Middle Paleolithic) Site in Saharan North Africa.’’ Abstracts with Programs, vol. 33, Geological Society of America, A-294. Boulder, Colo.: Geological Society of America. . 2003. ‘‘Pleistocene Stratigraphy and Chronology of the Lower Yellowstone Basin, North America.’’ In XVI INQUA Congress Programs and Abstracts, 228. Reno, Nev.: Desert Research Institute. . 2006. ‘‘Stratigraphic and Geochronologic Contexts of Mammoth (Mammuthus) and other Pleistocene Fauna, Upper Missouri Basin (Northern Great Plains and Rocky Mountains), U.S.A.’’ Quaternary International. 142–143: 87–106. Hoard, R., S. Holen, M. Glascock, H. Neff, and J. M. Elam. 1992. ‘‘Neutron Activation Analysis of Stone from the Chadron Formation and a Clovis Site on the Great Plains.’’ Journal of Archaeological Science 19: 655–65. Hoard, R., S. Holen, M. Glascock, and H. Neff. 1995. ‘‘Additional Comments on Neutron Activation Analysis of Stone from the Great Plains: Reply to Church.’’ Journal of Archaeological Science 22: 7–10. Hoch, D. 1983. ‘‘Flake Dispersal Experiments: Noncultural Transformation of the Archaeologic Record.’’ American Antiquity 48: 553–72. Hohn, M., and J. Hellerman. 1961. ‘‘The Diatoms.’’ In F. Wendorf, Paleoecology of the Llano Estacado. Publication of the Fort Burgwin Research Center, no. 1, 98–104. Taos: Museum of New Mexico Press. Holliday, V. 1986. Guidebook to the Archaeological Geology of Classic Paleoindian Sites of the Southern High Plains, Texas and New Mexico. College Station: Texas A and M University, Department of Geography. . 1997. Paleoindian Geoarchaeology of the Southern High Plains. Austin: University of Texas Press. . 2000a. ‘‘Folsom Drought and Episodic Drying on the Southern High Plains from 10,900–10,200 14 C yr b.p.’’ Quaternary Science 53: 1–12. . 2000b. ‘‘Historical Perspective on the Geoarchaeology of the Southern High Plains.’’ In Geoarchaeology in the Great Plains, ed. R. Mandel, 10–43. Norman: University of Oklahoma Press. . 2000c. ‘‘Vance Haynes and Paleoindian Geoarchaeology and Geochronology of the Great Plains.’’ Geoarchaeology 15: 511–22. . 2001. ‘‘Quaternary Geosciences in Archaeology.’’ In Earth Sciences and Archaeology, ed. P. Gold-

berg, V. Holliday, and C. R. Ferring, 3–35. New York: Kluwer Academic/Plenum. , ed. 1992. Soils in Archaeology. Washington, D.C.: Smithsonian Institution Press. Holmes, W. 1892. ‘‘Modern Quarry Refuse and the Paleolithic Theory.’’ Science 20: 295–97. Hosler, D., and A. McFarlane. 1996. ‘‘Copper Sources, Metal Production, and Metals Trade in Post-Classic Mesoamerica.’’ Science 273: 1819–24. Hough, W. 1906. ‘‘Earthquakes and Tribal Movements in the Southwest.’’ American Anthropologist 8: 436. Howard, E. 1935. ‘‘Evidence of Early Man in North America.’’ University of Pennsylvania Museum Journal 24(2–3): 61–71. Huber, J. K., and C. L. Hill. 1987. ‘‘A Pollen Sequence Associated with Paleoindian Presence in Northeastern Minnesota.’’ Current Research in the Pleistocene 4: 89–91. Huckleberry, G. 2000. ‘‘Interdisciplinary and Specialized Geoarchaeology: A Post-Cold War Perspective.’’ Geoarchaeology 15: 523–36. Huckleberry, G., and B. Billman. 2003. ‘‘Geoarchaeological Insights Gained from Surficial Geologic Mapping, Middle Moche Valley, Peru.’’ Geoarchaeology 18: 505–21. Huckleberry, G., J. Stein, and P. Goldberg. 2003. ‘‘Determining the Provenience of Kennewick Man Skeletal Remains through Sedimentological Analysis.’’ Journal of Archaeological Science 30: 651–65. Hudak, G. J., E. Hobbs, A. Brooks, C. Sersland, and C. Phillips, eds. 2002. A Predictive Model of Precontact Archaeological Site Location for the State of Minnesota. Mn/DOT Agreement no. 73217. SHPO Reference no. 95-4098. Saint Paul: Minnesota Department of Transportation. Hudson-Edwards, K., M. Macklin, R. Finlayson, and D. Passmore. 1999. ‘‘Mediaeval Lead Pollution in the River Ouse at York, England.’’ Journal of Archaeological Science 26: 809–19. Huford, A., and P. Green. 1982. ‘‘A User’s Guide to Fission Track Dating Calibration.’’ Earth and Planetary Science Letters 59: 343–54. Hughen, K., S. Lehman, J. Southon, J. Overpeck, O. Marchal, C. Herring, and J. Turnbull. 2004. ‘‘ 14 C Activity and Global Carbon Cycle Changes over the Past 50,000 Years.’’ Science 303: 202–7. Hutchinson, I., and A. McMillan. 1997. ‘‘Archaeological Evidence for Village Abandonment Associated with Late Holocene Earthquakes at the Northern Cascadia Subduction Zone.’’ Quaternary Research 48: 79– 87.

303

Bibliography Isaac, G. 1967. ‘‘Towards the Interpretation of Occupation Debris: Some Experiments and Observations.’’ Kroeber Anthropological Society Papers 5(37): 31–57. Inglis, D. 1965. ‘‘Particle Sorting and Stone Migration by Freezing and Thawing.’’ Science 148: 1616–17. Jackson, J. A., ed. 1997. Glossary of Geology, 4th ed. Washington, D.C.: American Geological Institute. Jackson, K., and D. Uhlmann. 1966. ‘‘Particle Sorting and Stone Migration Due to Frost Heave.’’ Science 152: 545–46. Jia, L., and W. Huang. 1990. The Story of Peking Man: From Archaeology to Mystery. Beijing: Foreign Languages Press, and Hong Kong: Oxford University Press. Jing, Z., and G. Rapp. 2003. ‘‘The Coastal Evolution of the Ambracian Embayment and Its Relationship to Archaeological Settings.’’ In Landscape Archaeology in Southern Epirus, Greece 1, ed. J. Wiseman and K. Zachos, 157–98. Athens: American School of Classical Studies at Athens. Jing, Z., G. Rapp, and T. Gao. 1997. ‘‘Holocene Landscape Evolution and Its Impact on the Neolithic and Bronze Age Sites in the Shangqiu Area, Northern China.’’ Geoarchaeology 10(6): 481–513. Johnson, A. 1952. Frost Action in Roads and Airfields. Special Report 1. Washington, D.C.: National Research Council, Highway Research Board. Johnson, D. 2002. ‘‘Darwin Would Be Proud: Bioturbation, Dynamic Denudation, and the Power of Theory in Science.’’ Geoarchaeology 17: 7–39. Johnson, D., and K. Hansen. 1972. ‘‘The Effects of Frost-Heaving on Objects in Soils.’’ Plains Anthropologist 19(64): 81–98. Johnson, D., D. Muhs, and M. Barnhardt. 1977. ‘‘The Effects of Frost-Heaving on Objects in Soils, II: Laboratory Experiments.’’ Plains Anthropologist 22 (76), pt. 1: 133–47. Johnson, W. C., and B. Logan. 1990. ‘‘Geoarchaeology of the Kansas River Basin, Central Great Plains.’’ In Archaeological Geology of North America, ed. N. Lasca and J. Donohue, 267–300. Centennial Special Vol. 4. Boulder, Colo.: Geological Society of America. Jones, J., and J. Fisher. 1990. ‘‘Environmental Factors Affecting Prehistoric Shellfish Utilization, Grape Island, Boston Harbor, Massachusetts.’’ In Archaeological Geology of North America, ed. N. Lasca and J. Donohue, 137–47. Centennial Special Vol. 4. Boulder, Colo.: Geological Society of America. Jones, R., B. Isserlin, V. Karastathis, S. Paramarinopoulos, G. Syrides, J. Uren, I. Balatsas, C. Kapopoulos, Y. Maniatis, and G. Facorellis. 2000. ‘‘Exploration of the Canal of Xerxes, Northern Greece: The Role

304

of Geophysical and Other Techniques.’’ Archaeological Prospection 7: 147–70. Joyce, A., and R. Mueller. 1992. ‘‘The Social Impact of Anthropogenic Landscape Modification in the Rio Verde Drainage Basin, Oaxaca, Mexico.’’ Geoarchaeology 7(6): 503–26. Judd, N. 1930. Arizona’s Prehistoric Canals from the Air. Washington, D.C.: Smithsonian Institution Press. Judson, S. 1957. ‘‘Geology.’’ In The Identification of NonArtifactual Archaeological Materials, ed. W. Taylor, 48– 49. National Academy of Sciences Publication 565. Washington, D.C.: National Academy of Sciences National Research Council. Julig, P., J. McAndrews, and W. Mahaney. 1990. ‘‘Geoarchaeology of the Cummins Site on the Beach of Proglacial Lake Minong, Lake Superior Basin, Canada.’’ In Archaeological Geology of North America, ed. N. Lasca and J. Donohue, 21–51. Centennial Special Vol. 4. Boulder, Colo.: Geological Society of America. Julig, P., L. Pavlish, and R. Hancock. 1991. ‘‘INAA Provenance Studies of Lithic Materials from the Western Great Lakes Region of North America.’’ In Archaeometry ’90, ed. E. Pernicka and G. Wagner. Basel: Birkhäuser Verlag. Kamilli, D., and A. Steinberg. 1985. ‘‘New Approaches to Mineral Analysis of Ancient Ceramics.’’ In Archaeological Geology, ed. G. Rapp and J. Gifford, 313–30. New Haven and London: Yale University Press. Kapp, R. 1969. ‘‘Background.’’ In Kapp, How to Know Pollen and Spores, 1–20. Dubuque, Iowa: William C. Brown. Karcz, I., and U. Kafri. 1978. ‘‘Evaluation of Supposed Archaeoseismic Damage in Israel.’’ Journal of Archaeological Science 5: 237–53. Karkanas, P., N. Kyparissi-Apsotilika, O. Bar-Yosef, and S. Weiner. 1999. ‘‘Mineral Assemblages in Theopetra, Greece: A Framework for Understanding Diagenesis in a Prehistoric Cave.’’ Journal of Archaeological Science 26: 1171–80. Kayani, P., and G. McDonnell. 1995. ‘‘The Potential of Scanning Electron Microscope Techniques for Non-Destructive Obsidian Characterization and Hydration-Rim Research.’’ In Archaeological Sciences 1995: Proceedings of a Conference on the Application of Scientific Techniques to Archaeology, Liverpool, June. Kempe, D. 1988. Living Underground: A History of Cave and Cliff Dwelling. London: Herbert. Kerr, J. 1995. ‘‘Phosphate Imprinting within Mound A at the Huntsville Site.’’ In Pedological Perspectives in Archaeological Research, ed. M. Collins, B. Carter, B. Gladfelter, and R. Southard, 133–49. Soil Science

Bibliography Society of America Special Publication 44. Madison, Wisc.: Soil Science Society of America. Kidder, A. 1915. ‘‘Pottery of the Pajarito Plateau and Some Adjacent Regions in New Mexico.’’ Memoirs of the American Anthropological Association 2(6): 407–62. . 1924. An Introduction to the Study of Southwestern Archaeology, with a Preliminary Account of the Excavations at Pecos. New Haven, Yale University Press. Kinnunen, K., R. Tynni, K. Hokkanen, and J.-P.Taavitsainen. 1985. Flint Raw Materials of Prehistoric Finland: Rock Types, Surface Textures and Microfossils.’’ Espoo: Geological Survey of Finland Bulletin 334. Knapp, A., and J. Cherry. 1994. Provenience Studies and Bronze Age Cyprus: Production, Exchange and PoliticoEconomic Change. Monographs in World Archaeology 21. Madison, Wisc.: Prehistory Press. Koetje, T. 1994. ‘‘Intrasite Spatial Structure in the European Upper Paleolithic: Evidence and Patterning from the SW of France.’’ Journal of Anthropological Archaeology 13: 161–69. Kolb, C., and J. Van Lopik. 1966. ‘‘Depositional Environments of the Mississippi River Deltaic Plain, Southeastern Louisiana.’’ In Deltas in Their Geologic Framework, ed. M. Shirley, 17–61. Houston: Houston Geological Society. Kolb, M. F., N. Lasca, and L. Goldstein. 1990. ‘‘A Soil-Geomorphic Analysis of the Midden Deposits at Aztalan Site, Wisconsin.’’ In Archaeological Geology of North America, ed. N. Lasca and J. Donahue, 199– 218. Centennial Special Vol. 4. Boulder, Colo.: Geological Society of America. Konigsson, L. 1989. ‘‘Pollen Analysis in Archaeogeology and Geoarchaeology.’’ In Geology and Paleoecology for Archaeologists, ed. T. Hackens and U. Miller, 81– 104. Revello, Italy: European University Centre for Cultural Heritage. Kraft, J., G. Rapp, G. Szemler, C. Tziavos, and E. Kase. 1987. ‘‘The Pass at Thermopylae, Greece.’’ Journal of Field Archaeology 14: 181–98. Kraft, J., I. Kayan, and G. Rapp. 2000. ‘‘A Geologic Analysis of Ancient Landscapes and the Harbors of Ephesus and the Artemsion in Anatolia.’’ Jahreshefte des Osterreichischen Archaologischen Institutes in Wien 69: 175–233. Kraft, J., G. Rapp, I. Kayan, and J. Luce. 2003. ‘‘Harbor Areas at Ancient Troy: Sedimentology and Geomorphology Complement Homer’s Iliad.’’ Geology 31: 163–66. Kraus, M., and T. Brown. 1986. Pedofacies Analysis: A New Approach to Reconstructing Ancient Fluvial Sequences. Special Paper 216. Boulder, Colo.: Geological Society of America.

Kukla, J. 1987. ‘‘Loess Stratigraphy in Central China.’’ Quaternary Science Reviews 6: 191–219. Kukla, J., and Z. An. 1989. ‘‘Loess Stratigraphy in Central China.’’ Paleogeography, Paleoclimatology, Paleoecology 72: 203–25. Kuniholm, P. 2001. ‘‘Dendrochronology and Other Applications of Tree-Ring Studies in Archaeology.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 35–46. New York: John Wiley and Sons. Kvamme, K. 1999. ‘‘Recent Directions and Developments in Geographical Information Systems.’’ Journal of Archaeological Research 7: 153–201. . 2001. ‘‘Current Practices in Archaeogeophysics: Magnetics, Resistivity, Conductivity, and GroundPenetrating Radar.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 353– 84. New York: Kluwer Academic/Plenum Publishers. . 2003. ‘‘Geophysical Surveys as Landscape Archaeology.’’ American Antiquity 68: 435–57. La Paz, L. 1969. Topics in Meteoritics—Hunting Meteorites: Their Recovery, Use, and Abuse from Paleolithic to Present. Albuquerque: University of New Mexico. Laj, C., C. Kissel, A. Mazaud, E. Michel, R. Muscheler, and J. Beer. 2002. ‘‘Geomagnetic Field Intensity, North Atlantic Deep Water Circulation and Atmospheric Delta 14 C during the Last Fifty KYR.’’ Earth and Planetary Science Letters 200: 177–90. LaMarche, V. 1974. ‘‘Paleoclimatic Inferences from Long Tree-Ring Records.’’ Science 183 (1974): 1043– 48. Lambeck, K., T. M. Esat, and E.-K. Potter. 2002. ‘‘Links between Climate and Sea Levels for the Past Three Million Years.’’ Nature 419: 199–206. Lambert, J. 1997. Traces of the Past: Unraveling the Secrets of Archaeology through Chemistry. Reading, Mass.: Addison-Wesley. Lamothe, M., A. Driemanis, M. Morency, and A. Raukas. 1984. ‘‘Thermoluminescence Dating of Quaternary Sediments.’’ In Quaternary Dating Methods, ed. W. Mahaney, 153–71. Amsterdam: Elsevier. Lang, A., and G. Wagner. 1996. ‘‘Infrared Stimulated Luminescence Dating of Archaeosediments.’’ Archaeometry 38: 129–41. Langbein, W., and S. Schumm. 1958. ‘‘Yield of Sediment in Relation to Mean Annual Precipitation.’’ Transactions of the American Geophysical Union 39: 1076–84. Larsen, C. 1985. ‘‘Geoarchaeological Interpretation of Great Lakes Coastal Environments.’’ In Archaeological Sediments in Context, ed. J. Stein and W. Farrand, 91–110. Orono: University of Maine, Center for the Study of Early Man.

305

Bibliography Latham, A. 2001. ‘‘Uranium-Series Dating.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 63–72. New York: John Wiley and Sons. Leach, E. 1992. ‘‘On the Definition of Geoarchaeology.’’ Geoarchaeology 7(5): 405–17. Leigh, D. 2001. ‘‘Buried Artifacts in Sandy Soils: Techniques for Evaluating Pedoturbation versus Sedimentation.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 269–93. New York: Kluwer Academic/Plenum Publishers. Leroi-Gourhan, A. 1968. The Art of Prehistoric Man in Western Europe. London: Thames and Hudson. Leucci, G. 2002. ‘‘Ground-Penetrating Radar Survey to Map the Location of Buried Structures under Two Churches.’’ Archaeological Prospection 9: 217–28. Libby, W. 1955. Radiocarbon Dating. Chicago: University of Chicago Press. Liu, Z. 1988. ‘‘Paleoclimatic Changes as Indicated by the Quaternary Karstic Cave Deposits in China.’’ Geoarchaeology 3(2): 103–15. Linford, N., and M. Canti. 2001. ‘‘Geophysical Evidence for Fires in Antiquity: Preliminary Results from an Experimental Study.’’ Archaeological Prospection 8: 211–25. Lock, G., and Z. Stancic. 1995. Archaeology and Geographical Information Systems: A European Perspective. London: Taylor and Francis. Louis, F. 1994. ‘‘Gold and Silver in Ancient China.’’ Arts of Asia 24: 88–96. Lowe, J. 2001. ‘‘Quaternary Geochronological Frameworks.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 9–21. New York: John Wiley and Sons. Lubbock, J. 1865. Pre-Historic Times. London: Williams and Norgate. Lucas, A. 1962. Ancient Egyptian Materials and Industries, rev. J. R. Harris. London: E. Arnold. Rpt. London: Histories and Mysteries of Man, 1989. Luedtke, B. 1992. An Archaeologist’s Guide to Chert and Flint. Vol. 7. Los Angeles: Archaeological Research Tools. Lyell, C. 1863. Geological Evidences of the Antiquity of Man. London: Murray. Lyell, K. 1881. Life, Letters and Journals of Sir Charles Lyell. London: Murray. Lyman, R. 1994. Vertebrate Taphonomy. Cambridge: Cambridge University Press. MacCurdy, G., ed. 1937. Early Man. London: Lippincott. Mack, G., and W. James. 1992. Paleosols for Sedimentologists. Boulder, Colo.: Geological Society of America. Mack, G., W. James, and H. Monger. 1993. ‘‘Classifica-

306

tion of Paleosols.’’ Geological Society of America Bulletin 105: 129–36. Macphail, R. 1998. A Reply to Carter and Davidson’s ‘‘An Evaluation of the Contribution of Soil Micromorphology to the Study of Ancient Agriculture.’’ Geoarchaeology 13(6): 549–64. Macphail, R., and J. Cruise. 2001. ‘‘The Soil Micromorphologist as Team Player: A Multianalytical Approach to the Study of European Microstratigraphy.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 241–67. New York: Kluwer Academic/Plenum. Macphail, R. I., and P. Goldberg. 2000. ‘‘Geoarchaeological Investigations of Sediments from Gorham’s and Vanguard Caves, Gibralter: Microstratigraphical (Soil Micromorphological and Chemical) Signatures. In Neanderthals on the Edge, ed. C. B. Stringer, R. N. E. Barton, and C. Finlayson, 183–200. Oxford: Oxbow. Mallowan, M. 1964. Early Mesopotamia and Iran. London: Thames and Hudson. Maloney, B., C. Higham, R. Bannanurag. 1989. ‘‘Early Rice Cultivation in Southeast Asia: Archaeological and Palynological Evidence from the Bang Pakong Valley, Thailand.’’ Antiquity 63: 363–370. Mandal, S. 1997. ‘‘Striking the Balance: The Roles of Petrography and Geochemistry in Stone Axe Studies in Ireland.’’ Archaeometry 39: 289–308. Mandel, R. 1992. ‘‘Soils and Holocene Landscape Evolution on Central and Southwestern Kansas: Implications for Archaeological Research.’’ In Soils in Archaeology: Landscape Evolution and Human Occupation, ed. V. Holliday, 41–100. Washington, D.C.: Smithsonian Institution Press. . 1995. ‘‘Geomorphic Controls of the Archaic Record in the Central Plains of the United States.’’ In Archaeological Geology of the Archaic Period in North America, ed. E. Bettis, 37–66. Special Paper 297. Boulder, Colo.: Geological Society of America. . 2000. ‘‘The History of Geoarchaeological Research in Kansas and Northern Oklahoma.’’ In Geoarchaeology in the Great Plains, ed. R. Mandel, 79–136. Norman: University of Oklahoma Press. , ed. 2000. Geoarchaeology in the Great Plains. Norman: University of Oklahoma Press. Mandel, R., and E. Bettis. 2001. ‘‘Use and Analysis of Soils by Archaeologists and Geoscientists: A North American Perspective.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 173–204. New York: Kluwer Academic/Plenum Publishers. Marmet, E., M. Bina, N. Fedoroff, and A. Tabbagh.

Bibliography 1999. ‘‘Relationships between Human Activity and the Magnetic Properties of Soils: A Case Study in the Medieval Site of Roissy-en-France.’’ Archaeological Prospection 6: 161–70. Martinez Solarez, J., A. Lopez Arroyo, and J. Mezcua. 1979. ‘‘Isoseismal Map of the 1755 Lisbon Earthquake Obtained from Spanish Data.’’ Tectonophysics 53: 301– 13. Martinson, D. G., N. G. Pisias, J. D. Hays, J. Imbrie, T. C. Moore, and N. J. Shackleton. 1987. ‘‘Age Dating and the Orbital Theory of the Ice Ages: Development of a High Resolution 0- to 300,000-Year Chronostratigraphy.’’ Quaternary Research 27: 1–29. Mason, R., and L. Golombek. 2003. ‘‘The Petrography of Iranian Safavid Ceramics.’’ Journal of Archaeological Science 30: 251–61. Mathieu, F. 2001. ‘‘The Organization of Turquoise Production and Consumption by Prehistoric Chacoans.’’ American Antiquity 66: 103–18. Matson, F. 1966. Ceramics and Man. Viking Fund Publications in Anthropology 41. London: Methuen. Maxwell, A. L. 2004. ‘‘Fire Regimes in North-Eastern Cambodian Monsoonal Forests, with a 9300-year Sediment Charcoal Record.’’ Journal of Biogeography 31: 225–39. May, D. 2000. ‘‘Geoarchaeological Research in Nebraska: A Historical Perspective.’’ In Geoarchaeology in the Great Plains, ed. R. Mandel, 166–98. Norman: University of Oklahoma Press. Mayer, J. 2002. ‘‘Evaluating Natural Site Formation Processes in Eolian Dune Sands: A Case Study from the Krmpotich Folsom Site, Killpecker Dunes, Wyoming.’’ Journal of Archaeological Science 29: 1199– 211. McBrearty, M., M. Bishop, T. Plummer, R. Dewar, and N. Conard. 1998. ‘‘Tools Underfoot: Human Trampling as an Agent of Lithic Artifact Edge Modification.’’ American Antiquity 63: 108–29. McCauley, J., G. Schaber, C. Breed, M. Grolier, C. V. Haynes, B. Issawi, C. Elachi, and R. Blom. 1982. ‘‘Subsurface Valleys and Geoarchaeology of the Eastern Sahara Revealed by Shuttle Radar.’’ Science 218: 1004–20. McCoy, F., and G. Heiken. 2000a. ‘‘The Late-Bronze Age Explosive Eruption of Thera (Santorini), Greece: Regional and Local Effects.’’ In Volcanic Hazards and Disasters in Human Antiquity, ed. F. McCoy, and G. Heiken, 43–70. Special Paper 345. Boulder, Colo.: Geological Society of America. McCoy, F., and G. Heiken, eds. 2000b. Volcanic Hazards and Disasters in Human Antiquity. Special Paper 345. Boulder, Colo.: Geological Society of America.

McDonald, W., and G. Rapp. 1972. The Minnesota Messenia Expedition: Reconstructing a Regional Bronze Age Environment. Minneapolis: University of Minnesota Press. McKillop, H. 2002. White Gold of the Ancient Maya. Gainesville: University Press of Florida. Mehringer, P., and F. Foit. 1990. ‘‘Volcanic Ash Dating of the Clovis Cache at East Wenatchee, Washington.’’ National Geographic Research 6(4): 495–503. Meltzer, D. 1983. ‘‘The Antiquity of Man and the Development of American Archaeology.’’ In Advances in Archaeological Method and Theory, ed. M. Schiffer, 6:1– 51. New York: Academic Press, 1983. . 1985. ‘‘North American Archaeology and Archaeologists, 1879–1934.’’ American Antiquity 50(2): 249–60. Meltzer, D., and R. Dunnell, eds. 1992. The Archaeology of William Henry Holmes. Washington, D.C.: Smithsonian Institution Press. Meltzer, D., J. Adovasio, and T. Dillehay. 1994. ‘‘On a Pleistocene Human Occupation at Pedra Furada, Brazil.’’ Antiquity 262(68): 695–714. Mercier, N., H. Valladas, G. Valladas, and J.-L. Reyss. 1995. ‘‘TL Dates of Burnt Flints from Jelinek’s Excavations at Tabun and Their Implications.’’ Journal of Archaeological Science 22: 495–509. Megard, R. 1967. ‘‘Late Quaternary Cladocera of Lake Zeribar, Western Iran.’’ Ecology 48: 179–89. Michels, J. 1969. ‘‘Testing Stratigraphy and Artifact Reuse through Obsidian Hydration Dating.’’ American Antiquity 34(1): 15–22. . 1983. ‘‘Obsidian Dating and East African Archaeology.’’ Science 219: 361–66. Miller, G. 1993. ‘‘Chronology of Hominid Occupation at Bir Tarfawi and Bir Sahara East, Based on the Epimerization of Isoleucine in Ostrich Eggshells.’’ In F. Wendorf, R. Schild, A. Close, et al., Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East, 241–51. New York: Plenum Press. Miller, V., and M. Westerback. 1989. Interpretation of Topographic Maps. Columbus, Ohio: Merrill. Mizoguchi, K. 2002. An Archaeological History of Japan: 30,000 b.c. to a.d. 700. Philadelphia: University of Pennsylvania Press. Moeyerson, J. 1978. ‘‘The Behavior of Stones and Stone Implements Buried in Consolidating and Creeping Kalahari Sands.’’ Earth Surface Processes 3: 115–28. Mooney, S., C. Geiss, and M. Smith. 2003. ‘‘The Use of Mineral Magnetic Parameters to Characterize Archaeological Ochres.’’ Journal of Archaeological Science 30: 511–23.

307

Bibliography Moore, P., J. Webb, and M. Collinson. 1991. Pollen Analysis. 2nd ed. Oxford: Basil Blackwell, 1991. Morwood, M., P. O’Sullivan, F. Aziz, and A. Raza. 1998. ‘‘Fission Track Age of Stone Tools and Fossils on the East Indonesian Island of Flores.’’ Nature 392: 173–76. Moss, J. 1978. ‘‘The Geology of Mummy Cave.’’ In The Mummy Cave Project in Northwestern Wyoming, ed. H. McCracken et al., 35–40. Cody, Wyo.: Buffalo Bill Historical Center. Moulton, G. E., ed. 2003. The Lewis and Clark Journals: An American Epic of Discovery; The Abridgment of the Definitive Nebraska Edition. Lincoln: University of Nebraska Press. Movius, H. 1949. ‘‘Old-World Paleolithic Archaeology.’’ Bulletin of the Geological Society of America 60: 1443–56. . 1957. ‘‘The Old World Paleolithic.’’ In Identification of Non-Artifactual Archaeological Materials, ed. W. Taylor, 26–27. National Academy of Sciences Publication 565. Washington, D.C.: National Academy of Sciences National Research Council. Mullins, H., and J. Halfman. 2001. ‘‘High-Resolution Seismic Reflection Evidence for Middle Holocene Environmental Change, Owasco Lake, New York.’’ Quaternary Research 55: 322–31. Müller, W., H. Fricke, A. Halliday, M. McCulloch, and J. Wartho. 2003. ‘‘Origin and Emigration of the Alpine Iceman.’’ Science 302: 862–66. Mustoe, G. E. 1985. ‘‘Eocene Amber from the Pacific Coast of North America.’’ Geological Society of America Bulletin 96(12): 1530–36. Naeser, C. 1984. ‘‘Fission-Track Dating.’’ In Quaternary Dating Methods, ed. C. Mahaney, 87–100. Amsterdam: Elsevier. Nash, S. 2000. ‘‘Seven Decades of Archaeological TreeRing Dating.’’ In It’s About Time: A History of Archaeological Dating in North America, ed. S. Nash, 60–82. Salt Lake City: University of Utah Press. , ed. 2000. It’s About Time: A History of Archaeological Dating in North America. Salt Lake City: University of Utah Press, 2000b. Nash, S., and J. Dean. 2000. ‘‘The Surprisingly Deficient History of Archaeochronology.’’ In It’s About Time: A History of Archaeological Dating in North America, ed. S. Nash, 2–11. Salt Lake City: University of Utah Press. Nelson, C. 1916. ‘‘Chronology of the Tano Ruins, New Mexico. American Anthropologist 18: 159–80. Nettleton, W., B. Brasher, E. Benham, and R. Ahrens. 1998. ‘‘A Classification System for Buried Paleosols.’’ Quaternary International 51/52: 175–83. Nettleton, W., C. Olson, and D. Wyosocki. 2000.

308

‘‘Paleosol Classification: Problems and Solutions.’’ Catena 41: 61–92. Nicholson, R. 2001 ‘‘Taphonomic Investigations.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 179–90. New York: John Wiley and Sons. Niemi, T., and Z. Ben-Avraham. 1994. ‘‘Evidence for Jericho Earthquakes from Slumped Sediments of the Jordan River Delta in the Dead Sea.’’ Geology 22: 395– 98. Nilsson, T. 1983. The Pleistocene: Geology and Life in the Quaternary Ice Age. Dordrecht: Reidel Publishing. Nir, Y. 1997. ‘‘Middle and Late Holocene Sea-Levels along the Israel Mediterranean Coast—Evidence from Ancient Water Wells.’’ Journal of Quaternary Science 12: 143–51. Nishimura, Y. 2001. ‘‘Geophysical Prospection in Archaeology.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 543–53. New York: John Wiley and Sons. Noller, J. 2001. ‘‘Archaeoseismology: Shaking Out the History of Humans and Earthquakes.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 143–70. New York: Kluwer Academic/Plenum Publishers. Nordt, L. 2001. ‘‘Stable Carbon and Oxygen Isotopes in Soils: Applications for Archaeological Research.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 419–48. New York: Kluwer Academic/Plenum Publishers. North, F. 1938. ‘‘Geology for Archaeologists.’’ Archaeological Journal 94: 73–115. Nur, A., and E. Cline. 2000. ‘‘Poseidon’s Horses: Plate Tectonics and Earthquake Storms in the Late Bronze Age Aegean and Eastern Mediterranean.’’ Journal of Archaeological Science 27: 43–63. O’Brien, M., and R. Lyman. 1999. Seriation, Stratigraphy, and Index Fossils: The Backbone of Archaeological Dating. New York: Kluwer Academic/Plenum Publishers. Ortloff, C., and A. Kolata. 1993. ‘‘Climate and Collapse: Agro-Ecological Perspectives on the Decline of the Tiwanaku State.’’ Journal of Archaeological Science 20: 195–221. O’Sullivan P., M. Morwood, D. Hobbs, F. Aziz, M. Situmorang, A. Raza, and R. Maas. 2001. ‘‘Archaeological Implications of the Geology and Chronology of the Soa Basin, Flores, Indonesia.’’ Geology 29: 607–10. Palacios-Fest, M. 1994. ‘‘Nonmarine Ostracod Shell Chemistry from Ancient Hohokam Irrigation Canals in Central Arizona: A Paleohydrologic Tool for the Interpretation of Prehistoric Human Occupation in

Bibliography the North American Southwest.’’ Geoarchaeology 9(1): 1–29. Parnell, J., R. Terry, and Z. Nelson. 2002. ‘‘Soil Chemical Analysis Applied as an Interpretive Tool for Ancient Human Activities in Piedras Negras, Guatemala.’’ Journal of Archaeological Science 29: 379–404. Passmore, D., C. Waddington, and S. Houghton. 2002. ‘‘Geoarchaeology of the Milfield Basin, Northern England: Towards an Integrated Archaeological Prospection, Research and Management Framework.’’ Archaeological Prospection 9: 71–91. Paulissen, E., and P. M. Vermeersch. 1987. ‘‘Earth, Man and Climate in the Egyptian Nile Valley during the Pleistocene.’’ In Prehistory of Arid North Africa: Essays in Honor of Fred Wendorf, ed. A. E. Close, 29–67. Dallas, Tex.: Southern Methodist University Press. Pavich, M. J., and O. A. Chadwick. 2003, ‘‘Soils and the Quaternary Climate System.’’ In The Quaternary Period in the United States, ed. A. R. Gillespie, S. C. Porter, and B. F. Atwater, 311–29. Amsterdam: Elsevier. Peacock, E. 1991. ‘‘Distinguishing Between Artifacts and Geofacts: A Test Case from Eastern England.’’ Journal of Field Archaeology 18: 345–61. Pearsall, D., B. Gilbert, and L. Martin. 1994. ‘‘Late Pleistocene Fossils of Natural Trap Cave, Wyoming, and the Climate Model of Extinction.’’ In Quaternary Extinction, ed. P. S. Martin and R. G. Klein, 138–47. Tucson: University of Arizona Press. Pendergast, D. M. 1982. ‘‘Ancient Maya Mercury.’’ Science 217: 533–35. Penhallurick, R. 1986. Tin in Antiquity. London: Institute of Metals. Pescatore, T., M. Senatore, G. Captretto, and G. Lerro. 2001. ‘‘Holocene Coastal Environments near Pompeii before the a.d. 79 Eruption of Mount Vesuvius, Italy.’’ Quaternary Research 55: 77–85. Peters, C., M. Church, and C. Mitchell. 2001. ‘‘Investigation of Fire Ash Residues Using Mineral Magnetism.’’ Archaeological Prospection 8: 227–37. Petraglia, M., and R. Potts. 1994. ‘‘Water Flow and the Formation of Early Pleistocene Artifact Sites in Olduvai Gorge, Tanzania.’’ Journal of Anthropological Archaeology 13: 228–54. Pewe, T. 1954. ‘‘The Geological Approach to Dating Archaeological Sites.’’ American Antiquity 19: 51–61. Phillips, B., and C. Hill. 2004. ‘‘Deglaciation History and Geomorphological Character of the Region between the Agassiz and Superior Basins, Associated with the ‘Interlakes Composite’ of Minnesota and Ontario.’’ In The Late Palaeo-Indian Great lakes: Geological and Archaeological Investigations of Late Pleisto-

cene and Early Holocene Environments, eds. L. J. Jackson and A. Hinshelwood, 275–301. Mercury Series Archaeology Paper 165. Quebec: Canadian Museum of Civilization. Phillipson, D. 1994. African Archaeology. Cambridge: Cambridge University Press, 1994. Pierce, K. L., and I. Friedman. 2000. ‘‘Obsidian Hydration Dating of Quaternary Events.’’ In Quaternary Geochronology: Methods and Applications, ed. J. S. Noller, J. M. Sowers, and W. E. Lettis, 223–40. Washington, D.C.: American Geophysical Union. Piggott, S. 1965. Ancient Europe from the Beginnings of Agriculture to Classical Antiquity: A Survey. Chicago: Aldine. Pilcher, J., and M. Hughes. 1982. ‘‘The Potential of Dendrochronology for the Study of Climate Change.’’ In Climatic Change in Later Prehistory, ed. A. Harding, 75–84. Edinburgh: Edinburgh University Press. Piperno, D., and D. Pearsall. 1993. Current Research in Phytolith Analysis: Applications in Archaeology and Paleoecology. Philadelphia: MASCA, University Museum of Archaeology and Anthropology, University of Pennsylvania. Porter, J. 1961. ‘‘Hixton Silicified Sandstone: A Unique Lithic Material Used by Prehistoric Cultures.’’ Wisconsin Archaeologist 42: 78–85. Potts, R. 1988. Early Hominid Activities at Olduvai. New York: A. de Gruyter. Powell, J. 1890. ‘‘Prehistoric Man in America.’’ Forum 8: 489–503. Preece, R. 2001. ‘‘Non-Marine Mollusca and Archaeology.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 135–45. New York: John Wiley and Sons. Pregill, G. 1986. ‘‘Body Size of Insular Lizards: A Pattern of Holocene Dwarfism.’’ Evolution 40: 997–1008. Prestwich, J. 1860. ‘‘On the Occurrence of Flint Implements, Associated with the Remains of Extinct Mammalia, in Undisturbed Beds of a Late Geological Period.’’ Proceedings of the Royal Society of London 10: 50–59. Pretola, J. 2001. ‘‘A Feasibility Study Using Silica Polymorph Ratios for Sourcing Chert and Chalcedony Lithic Materials.’’ Journal of Archaeological Science 28: 721–39. Pumpelly, R. 1908. Explorations in Turkestan: Expedition of 1904: Prehistoric Civilizations of Anau: Origins, Growth and Influence of Environment. Washington, D.C.: Carnegie Institution of Washington Publication 73. Pumpelly, R., W. Davis, and E. Huntington. 1905. Ex-

309

Bibliography plorations in Turkestan: Expedition of 1903. Washington, D.C.: Carnegie Institution of Washington. Pyddoke, E. 1961. Stratification for the Archaeologist. London: Phoenix House. Quinn, R., J. Adams, J. Dix, and J. Bull. 1998. ‘‘The Invincible (1758) Site—An Integrated Geophysical Assessment.’’ International Journal of Nautical Archaeology 27: 126–38. Quinn, R., J. Bull, and J. Dix. 1997. ‘‘Imaging Wooden Artifacts Using Chirp Sources.’’ Archaeological Prospection 4: 25–35. Raab, L., and A. Goodyear. 1984. ‘‘Middle Range Theory in Archaeology: A Critical Review of Origins and Applications.’’ American Antiquity 49: 255–68. Rampino, M., and S. Ambrose. 2000. ‘‘Volcanic Winter in the Garden of Eden: The Toba Supereruption and the Late Pleistocene Human Population Crash.’’ In Volcanic Hazards and Disasters in Human Antiquity, ed. F. McCoy and G. Heiken, 71–82. Special Paper 345. Boulder, Colo.: Geological Society of America. Rapp, G. 1975. ‘‘The Archaeological Field Staff: The Geologist.’’ Journal of Field Archaeology 2: 232–37. . 1987a. ‘‘Archaeological Geology.’’ In Encyclopedia of Physical Science and Technology, 1:688–98. New York: Academic Press. . 1987b. ‘‘Geoarchaeology.’’ Annual Review of Earth and Planetary Sciences 15: 97–113. . 1988. ‘‘On the Origins of Copper and Bronze Alloying.’’ In The Beginning of the Use of Metals and Alloys, ed. R. Maddin, 21–27. Cambridge, Mass.: MIT Press. . 1989. ‘‘Determining the Origins of Sulfide Smelting.’’ Der Anschnitt 107–10. . 2002. Archaeomineralogy. Heidelberg: SpringerVerlag, 2002. Rapp, G., and S. Aschenbrenner, eds. 1978. Excavations at Nichoria in Southwest Greece. Vol. 1: Site, Environs, and Techniques. Minneapolis: University of Minnesota Press. Rapp, G., and J. Gifford. 1982. ‘‘Archaeological Geology.’’ American Scientist 70: 45–53. Rapp, G., and E. Henrickson. 1972. ‘‘Geophysical Exploration.’’ In The Minnesota Messenia Expedition: Reconstructing a Bronze Age Regional Environment, ed. W. McDonald and G. Rapp, 234–39. Minneapolis: University of Minnesota Press. Rapp, G., and J. Kraft. 1994. ‘‘Holocene Coastal change in Greece and Aegean Turkey.’’ In Beyond the Site: Regional Studies in the Aegean Area, ed. P. N. Kardulias, 69–90. Lanham, Md.: University Press of America. Rapp, G., and S. Mulholland. 1992. Phytolith Systematics: Emerging Issues. New York: Plenum Press.

310

Rapp, G., J. Allert, and G. Peters. 1990a. ‘‘The Origins of Copper in Three Northern Minnesota Sites: Pauly, River Point, and Big Rice.’’ In The Woodland Tradition in the Western Great Lakes, ed. G. Gibbon, 233–38. Publications in Anthropology 4. Minneapolis: University of Minnesota Press. Rapp, G., E. Henrickson, and J. Allert. 1990b. ‘‘Native Copper Sources of Artifact Copper in Pre-Columbian North America.’’ In Archaeological Geology of North America, ed. N. Lasca and J. Donohue, 479–98. Boulder, Colo.: Geological Society of America. Rapp, G., R. Rothe, and Z. Jing. 1999. ‘‘Using Neutron Activation Analyses to Source Ancient Tin (Cassiterite).’’ In Metals in Antiquity, ed. S. Young, A. Pollard, P. Budd, and R. Ixer, 153–62. Oxford: Archaeopress. Rapp, G., J. Allert, V. Vitali, Z. Jing, and E. Hinrickson. 2000. Determining Geologic Sources of Artifact Copper: Source Characterization Using Trace Element Patterns. Lanham, Md.: University Press of America. Reader, C. D. 2001. ‘‘A Geomorphological Study of the Giza Necropolis with Implications for the Development of the Site.’’ Archaeometry 43(1): 149–65. Reid, K. 1984. ‘‘Fire and Ice: New Evidence for the Production and Preservation of Late Archaic FiberTempered Pottery in the Middle-Latitude Lowlands.’’ American Antiquity 49: 55–76. Reimnitz, E., and N. Marshall. 1965. ‘‘Effects of the Alaska Earthquake and Tsunami on Recent Deltaic Sediments.’’ Journal of Geophysical Research 70: 2363– 76. Reineck, H., and I. Singh. 1975. Depositional Sedimentary Environments. Ithaca, N.Y.: Springer. Reinhard, K. J., and U. M. Bryant. 1992. ‘‘Coprolite Analysis: A Biologic Perspective on Archaeology.’’ In Archaeological Method and Theory, ed. M. B. Schiffer, 4:245–88. Tucson: University of Arizona Press. Reinhardt, G. 1993. ‘‘Hydrologic Artifact Dispersals at Pingasagruk, North Coast, Alaska.’’ Geoarchaeology 8(6): 493–513. Rendell, H., and R. Dennell. 1987. ‘‘Thermoluminescence Dating of an Upper Pleistocene Site, Northern Pakistan.’’ Geoarchaeology 2(1): 63–67. Renfrew, C. 1976. ‘‘Archaeology and the Earth Sciences.’’ In Geoarchaeology, ed. D. Davidson and M. L. Shackley, 1–5. Boulder, Colo.: Westview Press. Reuter, G. 2000. ‘‘A Logical System of Paleopedological Terms.’’ Catena 41: 93–109. Richards, D., and J. Beck. 2001. ‘‘Dramatic Shifts in Atmospheric Radiocarbon during the Last Glacial Period.’’ Antiquity 75: 482–85. Richerson, P., R. Boyd, and R. Bettinger. 2001. ‘‘Was Agriculture Impossible during the Pleistocene but

Bibliography Mandatory during the Holocene? A Climate Change Hypothesis.’’ American Antiquity 66: 387–411. Richter, G. 1980. ‘‘On the Soil Erosion Problem in the Temperate Humid Area of Central Europe.’’ Geojournal 4: 279–87. Rick, R. 1976. ‘‘Downslope Movement and Archaeologic Intrasite Spatial Analysis.’’ American Antiquity 41: 133–44. Rink, W. 2001. ‘‘Beyond 14 C Dating: A User’s Guide to Long-Range Dating Methods in Archaeology.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 385–417. New York: Kluwer Academic/Plenum Publishers. Rivka, E., S. Weiner, R. Albert, and M. Elbaum. 2003. ‘‘Detection of Burning of Plant Materials in the Archaeological Record by Changes in the Refractive Indices of Siliceous Phytoliths.’’ Journal of Archaeological Science 30: 217–26. Roberts, N., and H. Wright. 1993. ‘‘Vegetational, LakeLevel, and Climatic History of the Near East and Southwest Asia.’’ In Global Climates since the Last Glacial Maximum, ed. H. Wright, J. Kutzbach, T. Webb, W. Ruddiman, F. Street-Perrot, and P. Bartlein, 194– 220. Minneapolis: University of Minnesota Press. Roberts, R. 1987. ‘‘Luminescence Dating in Archaeology: From Origins to Optical.’’ Radiation Measurements 27: 819–92. Roberts, W., G. Rapp, and J. Weber. 1974. Encyclopedia of Minerals. New York: Van Nostrand Reinhold. Roberts, W., T. Campbell, and G. Rapp. 1990. Encyclopedia of Minerals. 2nd ed. New York: Van Nostrand Reinhold. Robinson, M. 2001. ‘‘Insects as Palaeoenvironmental Indicators.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 121–33. New York: John Wiley and Sons. Roe, D. 1981. The Lower and Middle Paleolithic Periods in Britain. London: R. and K. Paul,. . 1993. ‘‘Landmark Sites of the British Paleolithic.’’ Review of Archaeology 14(2): 1–9. Roosevelt, A. 1991. Mound Builders of the Amazon. New York: Academic Press. Rothenberg, B. 1978. ‘‘Calcolithic Copper Smelting: Excavations at Timna Site 39.’’ In Archaeo-Metallurgy 1, ed. B. Rothenberg, 1–51. London: Institute for Archaeo-Metallurgical Studies. Rousseau, D.-D., and J.-H. Puisseguir. 1990. ‘‘A 350,000-Year Climatic Record from the Loess Sequence of Achenheim, Alsace, France.’’ Boreas 19: 203–16. Rovner, I. 1994. ‘‘Floral History by the Back Door:

A Test of Phytolith Analysis in Residential Yards at Harpers Ferry.’’ Historical Archaeology 28(4): 37–48. Rovner, I., and J. Russ. 1992. ‘‘Darwin and Design in Phytolith Systematics: Morphometric Methods for Mitigating Redundancy.’’ In Phytolith Systematics: Emerging Issues, ed. G. Rapp and S. Mulholland, 253– 76. New York: Plenum Press. Rowley-Conwy, P. 1993. ‘‘Was There a Neanderthal Religion?’’ In The First Humans: Human Origins and History to 10,000 B.C., ed. G. Burenhult, 70. New York: Harper Collins. Russ, J., and I. Rovner. 1989. ‘‘Stereological Identification of Opal Phytolith Populations from Wild and Cultivated Zea.’’ American Antiquity 54: 784–92. Salisbury, R. 1893. ‘‘Man and the Glacial Period.’’ American Geologist 11: 13–20. Sanger, D., A. Kelley, and H. Berry. 2001. ‘‘Geoarchaeology at Gilman Falls: An Archaic Quarry and Manufacturing Site in Central Maine, U.S.A.’’ Geoarchaeology 16: 633–65. Saucier, R. 1977. Effects of the New Madrid Earthquake Series in the Mississippi Alluvial Valley. U.S. Army Engineer Waterways Experiment Station Miscellaneous Paper S-77-5. Vicksburg, Ms.: U.S. Army Engineer Waterways Experiment Station. . 1989. ‘‘Evidence for Episodic Sand-Blow Activity during the 1811–1812 New Madrid (Missouri) Earthquake Series.’’ Geology 17: 103–6. . 1991. ‘‘Geoarchaeological Evidence of Strong Prehistoric Earthquakes in the New Madrid (Missouri) Seismic Zone.’’ Geology 19: 296–98. Schick, K. 1986. Stone Age Sites in the Making: Experiments in the Formation of Archaeological Occurrences. British Archaeological Reports International Series 319. Oxford: British Archaeological Reports. . 1992. ‘‘Geoarchaeological Analysis of an Acheulian Site at Kalambo Falls, Zambia.’’ Geoarchaeology 7(1): 1–26. Schiffer, M. 1972. ‘‘Archaeological Context and Systemic Context.’’ American Antiquity 37: 156–65. . 1976. Behavioral Archaeology. New York: Academic Press. . 1983. ‘‘Towards the Identification of Formation Processes.’’ American Antiquity 48: 675–706. . 1987. Formation Processes of the Archaeological Record. Albuquerque: University of New Mexico Press. Schmidt, A. 2001. Geophysical Data in Archaeology: A Guide to Good Practice. Oxford: Oxbow Books. Schnellmann, M., F. Anselmetti, D. Giardini, J. McKenzie, and S. Ward. 2004. ‘‘Ancient Earthquakes at Lake Lucerne.’’ American Scientist 92: 46–53.

311

Bibliography Schoenwetter, J. 1981. ‘‘Prologue to a Contextual Archaeology.’’ Journal of Archaeological Science 8: 367–79. Schuldenrein, J. 1991. ‘‘Coring and the Identity of Cultural-Resource Environments: A Comment on Stein.’’ American Antiquity 56: 131–37. . 1995. ‘‘Geochemistry, Phosphate Fractionation, and the Detection of Activity Areas at Prehistoric North American Sites.’’ In Pedological Perspectives in Archaeological Research, ed. M. Collins, B. Carter, B. Gladfelter, and R. Southard, 107–32. Soil Science Society of America Special Publication 44. Madison, Wisc.: Soil Science Society of America. . 2001. ‘‘Stratigraphy, Sedimentology, and Site Formation at Konispol Cave, Southwest Albania.’’ Geoarchaeology 16: 559–602. Schulze, D. G., J. L. Nagel, G. E. Van Soyoc, T. L. Henderson, M. F. Marion, and D. E. Scott. 1993. ‘‘Significance of Organic Matter in Determining Soil Colors.’’ In Soil Color, ed. J. Bigham and E. Ciolkosz, 71–90. SSSA Special Publication 31. Madison, Wisc.: Soil Science Society of America. Schwarcz, H. 1980. ‘‘Absolute Age Determination of Archaeological Sites by Uranium Series Dating of Travertines.’’ Archaeometry 22: 3–24. . 1984. ‘‘The Site of Vertesszollos, Hungary.’’ Journal of Archaeological Science 11: 327–36. Schwarcz, H., and M. Gascoyne. 1984. ‘‘UraniumSeries Dating of Quaternary Sediments.’’ In Quaternary Dating Methods, ed. W. Mahaney, 33–41. Amsterdam: Elsevier. Schwarcz, H., and L. Morawska. 1993. ‘‘Uranium Series Dating of Carbonates from Bir Tarfawi and Bir Sahara East.’’ In F. Wendorf, R. Schild, A. Close, et al., Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East, 205–17. New York: Plenum Press. Schwarcz, H., and W. Rink. 2001. ‘‘Dating Methods for Sediments of Caves and Rockshelters.’’ Geoarchaeology 16: 355–71. Schweger, C. 1985. ‘‘Geoarchaeology of Northern Regions: Lessons from Cryoturbation at Onion Portage, Alaska.’’ In Archaeological Sediments in Context, ed. J. Stein and W. Farrand, 127–41. Orono: University of Maine, Center for the Study of Early Man. Schwengruber, F. 1988. Tree Rings: Basics and Applications of Dendrochronology. Dordrecht, Holland: Reidel. Scollar, I., A. Tabbaugh, A. Hesse, and I. Herzog, eds. 1990. Archaeological Prospecting and Remote Sensing. Cambridge: Cambridge University Press. Shackleton, N. 1982. ‘‘Stratigraphy and Chronology of the KRM Deposits: Oxygen Isotope Evidence.’’ In The Middle Stone Age at Klasies River Mouth in South

312

Africa, ed. R. Singer and J. Wymer, 194–99. Chicago: University of Chicago Press. Shackley, M. L. 1974. ‘‘Stream Abrasion of Flint Implements.’’ Nature 248: 501–2. Shackley, M. S. 1995. ‘‘Sources of Archaeological Obsidian in the Greater American Southwest: An Update and Quantitative Analysis.’’ American Antiquity 60(3): 531–51. . 1998. ‘‘Current Issues and Future Directions in Archaeological Volcanic Glass Studies: An Introduction.’’ In Archaeological Obsidian Studies: Method and Theory, ed. Shackley, 1–14. New York: Plenum Press. Shay, C. T. 1971. The Itasca Bison Kill Site: An Ecological Analysis. Saint Paul: Minnesota Historical Society. Sheehan, M. 1994. ‘‘Cultural Responses to the Altithermal: The Role of Aquifer-Related Water Resources.’’ Geoarchaeology 9:113–37. Sheets, P., and B. McKee, eds. 1994. Archaeology, Volcanism, and Remote Sensing in the Arenal Region, Costa Rica. Austin: University of Texas Press. Sheets, P., K. Hirth, F. Lage, F. Strops, F. Asaro, and H. Michel. 1990. ‘‘Obsidian Sources and Elemental Analyses of Artifacts in Southern Mesoamerica and the Northern Intermediate Area.’’ American Antiquity 55: 144–59. Shepard, A. 1939. ‘‘Appendix A—Technology of La Plata Pottery.’’ Archaeological Studies of the La Plata District, ed. E. Morris, 249–87. Carnegie Institution of Washington Publication 519. Washington, D.C.: Carnegie Institution of Washington. . 1965. Ceramics for the Archaeologist. Carnegie Institution of Washington Publication 609. Washington, D.C.: Carnegie Institution of Washington. Shepherd, R. 1993. Ancient Mining. London: Elsevier. Sherwood, S. 2001. ‘‘Microartifacts.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 327–51. New York: Kluwer Academic/Plenum Publishers. Shippee, R. 1932. ‘‘The Great Wall of Peru.’’ Geographical Review 22(1): 1–29. Shlemon, R., and F. Budinger. 1990. ‘‘The Archaeological Geology of the Calico Site, Mojave Desert, California.’’ In Archaeological Geology of North America, ed. N. Lasca and J. Donahue, 301–13. Centennial Special Vol. 4. Boulder, Colo.: Geological Society of America. Siani, G., M. Paterne, E. Michel, R. Sulpizio, A. Sbrana, M. Arnold, and G. Haddad. 2001. ‘‘Mediterranean Sea Surface Radiocarbon Reservoir Age Changes since the Last Glacial Maximum.’’ Science 294: 1917– 20. Simkin, T., and R. Fiske. 1983. Krakatau 1883: The

Bibliography Volcanic Eruption and Its Effects. Washington, D.C.: Smithsonian Institution Press, 1983. Sims, J. 1973. ‘‘Earthquake-Induced Structures in Sediments of Van Normsan Lake, San Fernando, California.’’ Science 182: 161–63. . 1979. ‘‘Records of Prehistoric Earthquakes in Sedimentary Deposits in Lakes.’’ U.S. Geological Survey Earthquake Information Bulletin 11: 228–33. Smith, M., J. Prescott, and M. Head. 1997. ‘‘Comparison of 14 C and Luminescence Chronologies at Puritjarra Rock Shelter, Central Australia.’’ Quaternary Science 16: 299–320. Soil Survey Staff. 1975. Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys. Agricultural Handbook 436. Washington, D.C.: U.S. Department of Agriculture, Soil Conservation Service. Solecki, R. 1963. ‘‘Prehistory in the Shanidar Valley, Northern Iraq.’’ Science 139: 179–93. . 1975. ‘‘Shanidar IV: A Neanderthal Flower Burial in Northern Iraq.’’ Science 190: 880–81. Spencer, F. 1990. Piltdown: A Scientific Forgery. New York: Oxford University Press. Spennemann, D., and J. Head. 1998. ‘‘Tongan Pottery Chronology, 14 C Dates and the Hardwater Effect.’’ Quaternary Geochronology 17: 1047–56. Spier, L. 1931. ‘‘N. C. Nelson’s Stratigraphic Technique in the Reconstruction of Prehistoric Sequences in Southwestern America.’’ In Methods in the Social Sciences, ed. S. Rice, 275–83. Chicago: University of Chicago Press. Squier, E., and E. Davis. 1848. Ancient Monuments of the Mississippi Valley. Smithsonian Contributions to Knowledge 1. Washington, D.C.: Smithsonian Institution Press. Srivastava, P., and B. Parkash. 1998. ‘‘Clay Minerals in Soils as Evidence of Holocene Climatic Change, Central Indo-Gangetic Plains, North-Central India. Quaternary Research 50: 230–39. Stafford, C., and S. Creasman. 2002. ‘‘The Hidden Record: Late Holocene Landscapes and Settlement Archaeology in the Lower Ohio River Valley.’’ Geoarchaeology 17: 117–40. Stafford, M., G. Frison, D. Stanford, and G. Zeimans. 2003. ‘‘Digging for the Color of Life: Paleoindian Red Ochre Mining at the Powars II Site, Platte County, Wyoming, U.S.A.’’ Geoarchaeology 18: 71–90. Stanley, D.-J. 1995. ‘‘A Global Sea-Level Curve for the Late Quaternary: The Impossible Dream?’’ Marine Geology 125: 1–6. Stanley, D.-J., and A. Warne. 1993. ‘‘Nile Delta: Recent

Geologic Evolution and Human Impact.’’ Science 260: 628–34. Stanley, D.-J., M. Krom, R. Cliff, and J. Woodward. 2003. ‘‘Short Contribution: Nile Failure at the End of the Old Kingdom, Egypt: Strontium Isotopic and Petrologic Evidence.’’ Geoarchaeology 18: 223–42. Stanley, D.-J., F. Goddio, T. Jorstad, and G. Schnepp. 2004. ‘‘Submergence of Ancient Greek Cities off Egypt’s Nile Delta: A Cautionary Tale.’’ GSA Today 14: 4–10. Steadman, D., and N. Miller. 1987. ‘‘California Condor Associated with Spruce-Pine Woodland in the Late Pleistocene in New York.’’ Quaternary Research 28(3): 415–26. Stearns, C. 1984. ‘‘Uranium-Series Dating and the History of Sea Level.’’ In Quaternary Dating Methods, ed. W. Mahaney, 53–65. Amsterdam: Elsevier. Stein, J. 1983. ‘‘Earthworm Activity: A Source of Potential Disturbance of Archaeological Sediments.’’ American Antiquity 48(2): 277–89. . 1986. ‘‘Coring Archaeological Sites.’’ American Antiquity 51: 505–27. . 1987. ‘‘Deposits for Archaeologists.’’ In Advances in Archaeological Method and Theory, ed. M. Schiffer, 11: 337–95. New York: Academic Press. . 1991. ‘‘Coring in CRM and Archaeology: A Reminder.’’ American Antiquity 56: 138–42. . 1993. ‘‘Scale in Archaeology, Geosciences, and Geoarchaeology.’’ In Effects of Scale on Archaeological and Geoscientific Perspectives, ed. J. Stein and A. Linse, 1–9. Special Paper 283. Boulder, Colo.: Geological Society of America. . 2000. ‘‘Stratigraphy and Archaeological Dating.’’ In It’s About Time: A History of Archaeological Dating in North America, ed. S. Nash, p. 14–40. Salt Lake City: The University of Utah Press, 2000. . 2001. ‘‘A Review of Site Formation Processes and Their Relevance to Geoarchaeology.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 37–51. New York: Kluwer Academic/Plenum Publishers. , ed. 1992. Deciphering a Shell Midden. San Diego: Academic Press. Stein, J., and W. Farrand. 1985. ‘‘Context and Geoarchaeology: An Introduction.’’ In Archaeological Sediments in Context, ed. J. Stein and W. Farrand, 1–4. Orono: University of Maine, Center for the Study of Early Man. Stein, J., and A. Linse, eds. 1993. Effects of Scale on Archaeological and Geoscientific Perspectives. Special Paper 28. Boulder, Colo.: Geological Society of America. Stein, J., J. Deo, and L. Phillips. 2003. ‘‘Big Sites—Short

313

Bibliography Time: Accumulation Rates in Archaeological Sites.’’ Journal of Archaeological Science 30: 297–316. Stein, R., A. Barka, and J. Dietrich. 1997. ‘‘Progressive Failure on the North Anatolian Fault since 1939 by Earthquake Stress Triggering.’’ Geophysical Journal International 128: 594–604. Steponaitis, V. 1984. ‘‘Technological Studies in Prehistoric Pottery from Alabama: Physical Properties and Vessel Function.’’ In The Many Dimensions of Pottery, ed. S. Van der Leeuw and A. Pritchard, 79–122. Amsterdam: University of Amsterdam. Sternberg, R. 2001. ‘‘Magnetic Properties and Archaeomagnetism.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 73–79. New York: John Wiley and Sons. Stiros, S. 1988. ‘‘Earthquake Effects on Ancient Constructions.’’ In New Aspects of Archaeological Science in Greece, ed. Jones and Catling, 1–6. British School at Athens, Occasional Paper 3. Athens: Fitch Laboratory. . 1996. ‘‘Identification of Earthquakes from Archaeological Data: Methodology, Criteria, and Limitations.’’ In Archaeoseismology, ed. S. Stiros and R. Jones, 129–52. Institute of Geology and Mineral Exploration, and British School at Athens, Occasional Paper 7. Athens: Fitch Laboratory. Stiros, S., and R. Jones, eds. 1996. Archaeoseismology. Institute of Geology and Mineral Exploration, and British School at Athens, Occasional Paper 7. Athens: Fitch Laboratory. Stockton, E. 1973. ‘‘Shaw’s Creek Shelter: Human Displacement of Artifacts and Its Significance.’’ Mankind 9: 112–17. Stoltman, J. 1991. ‘‘Ceramic Petrography as a Technique for Documenting Cultural Interaction: An Example from the Upper Mississippi Valley.’’ American Antiquity 56: 103–20. . 2001. ‘‘The Role of Petrography in the Study of Archaeological Ceramics.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 297–326. New York: Kluwer Academic/Plenum Publishers. Stoltman, J., and D. A. Baerreis. 1983. ‘‘Evolution of Human Ecosystems in the Eastern United States.’’ In Late-Quaternary Environments of the United States. Vol. 2: The Holocene, ed. H. E. Wright, Jr., 252–68. London: Longman. Straus, L. 1990. ‘‘Underground Archaeology: Perspectives on Caves and Rockshelters.’’ Archaeological Method and Theory, ed. M. B. Schiffer, 2: 255–304. Street, R., and O. Nuttli. 1984. ‘‘The Central Mississippi Earthquakes of 1811–1812.’’ In Proceedings, Sym-

314

posium on the New Madrid Earthquakes, ed. P. L. Gori and W. W. Hays, 33–63. United States Geological Service Open-File Report 84-770. Stuart, F. 2001. ‘‘In Situ Cosmogenic Isotopes: Principles and Potential for Archaeology.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 92–100. New York: John Wiley and Sons. Sutcliffe, A. J., et al. 1976. ‘‘Cave Paleontology and Archaeology.’’ In The Science of Speleology, ed. T. Ford and C. Cullingford, 495–549. London: Academic Press. Taber, S. 1930. ‘‘The Mechanics of Frost Heaving.’’ Journal of Geology 38: 303–17. Tagle, A., R. Paschenger, and G. Infante. 1990. ‘‘Maya Blue: Its Presence in Cuban Colonial Wall Paintings.’’ Studies in Conservation 35: 156–59. Taylor, R. E. 2001. ‘‘Radiocarbon Dating.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 24–34. New York: John Wiley and Sons. Taylor, W., ed. 1957. The Identification of Non-Artifactual Archaeological Materials. National Academy of Sciences Publication 565. Washington, D.C.: National Academy of Sciences National Research Council. Terasmae, J. 1984. ‘‘Radiocarbon Dating: Some Problems and Potential Developments.’’ In Quaternary Dating Methods, ed. W. Mahaney, 1–14. Amsterdam: Elsevier. Terry, R., P. Hardin, S. Houston, S. Nelson, M. Jackson, J. Carr, and J. Parnell. 2000. ‘‘Quantitative Phosphorus Measurement: A Field Test Procedure for Archaeological Site Analysis at Piedras Negras, Guatemala.’’ Geoarchaeology 15: 151–66. Thacker, P., and B. Ellwood. 2002. ‘‘Detecting Palaeolithic Activity Areas through Electrical Resistivity Survey: An Assessment from Vale De Óbidos, Portugal.’’ Journal of Archaeological Science 29 (2002): 563– 70. Thackeray, J., J. Kirschvink, and T. Raub. 2002. ‘‘Paleomagnetic Analyses of Calcified Deposits from the Plio-Pleistocene Hominid Site of Kromdraai, South Africa.’’ South African Journal of Science 98: 537–40. Thiem, D. 1997. ‘‘Lower Paleolithic Hunting Spears from Germany.’’ Nature 385: 807–10. Thomas, H. 1923. ‘‘The Source of the Stones of Stonehenge.’’ Antiquaries Journal 3: 239–60. Thompson, K. 1991. ‘‘Piltdown Man: The Great English Mystery Story.’’ American Scientist 79: 194–201. Thompson, T. 1970. ‘‘Holocene Tectonic Activity in West Africa Dated by Archaeological Methods.’’ Geological Society of America Bulletin 81: 3759–64. Thorpe, R., O. Williams-Thorpe, D. Jenkins, and J. Watson. 1991. ‘‘The Geological Sources and Trans-

Bibliography port of the Bluestones of Stonehenge, Wiltshire, UK.’’ Proceedings of the Prehistoric Society 57: 103–57. Thorson, R. 1990a. ‘‘Archaeological Geology.’’ Geotimes (February): 32–33. . 1990b. ‘‘Geologic Contexts of Archaeologic Sites in Beringia.’’ In Archaeologic Geology of North America, ed. N. Lasca and J. Donahue, 399–420. Centennial Special Vol. 4. Boulder, Colo.: Geological Society of America. Thorson, R., and T. Hamilton.1977. ‘‘Geology of the Dry Creek Site, a Stratified Early Man Site in Interior Alaska.’’ Quaternary Research 7: 149–76. Tite, M., and C. Mullins. 1971. ‘‘Enhancement of the Magnetic Susceptibility of Soils on Archaeological Sites.’’ Archaeometry 13: 209–19. Tobey, M. 1986. Trace Element Investigations of Maya Chert from Belize. Vol. 1. Papers of the Colha Project. San Antonio: Center for Archaeological Research, University of Texas at San Antonio. Todd, L., and G. Frison. 1986. ‘‘Taphonomic Study of the Colby Site Mammoth Bones.’’ In Frison and Todd, The Colby Mammoth Site: Taphonomy and Archaeology of a Clovis Kill in Northern Wyoming, 27–99. Albuquerque: University of New Mexico Press. Togashi, T., and M. Matsumoto. 1991. ‘‘Presumption of the Sources of Stone Implements from Sites in Yamaguchi Prefecture.’’ Quaternary Research 30: 251–63 (in Japanese with English Abstract). Trigger, B. 1990. A History of Archaeological Thought. Cambridge: Cambridge University Press, 1990. Truncer, J., M. Glascock, and H. Neff. 1998. ‘‘Steatite Source Characterization in Eastern North America: New Results Using Instrumental Neutron Activation Analysis.’’ Archaeometry 40: 23–44. Turney, C., M. Bird, L. Fifield, R. Roberts, M. Smith, C. Dortch, R. Grun, E. Lawson, L. Ayliffe, G. Miller, J. Dortch, and R. Cresswell. 2001. ‘‘Early Human Occupation at Devil’s Lair, Southwestern Australia Fifty Thousand Years Ago.’’ Quaternary Research 55: 3–13. Tushingham, A., and W. Peltier. 1993. ‘‘Implications of the Radiocarbon Timescale for Ice-Sheet Chronology and Sea-Level Change.’’ Quaternary Research 39: 125–29. Ucko, P., and R. Layton, eds. 1999. The Archaeology and Anthropology of Landscape: Shaping Your Landscape. London: Routledge. Ulm, S. 2002. ‘‘Marine and Estuarine Reservoir Effects in Central Queensland, Australia: Determination of Delta R Values.’’ Geoarchaeology 17: 319–48. Upton, W. 1970. Landforms and Topographic Maps. New York: John Wiley and Sons, 1970.

Urabe, A., H. Nakaya, T. Muto, S. Katoh, M. Hyodo, and X. Shunrong. 2001. ‘‘Lithostratigraphy and Depositional History of the Late Cenozoic HominidBearing Successions in the Yuanmou Basin, Southwest China.’’ Quaternary Science 20: 1671–81. U.S. Department of Agriculture. 1987. Keys to Soil Taxonomy. Technical Monograph 6. Ithaca, N.Y.: Department of Agronomy, Cornell University; Agency for International Development, U.S. Department of Agriculture, Soil Management Support Sources. Van Andel, T., and N. Lianos. 1984. ‘‘High-Resolution Seismic Reflection Profiles for the Reconstruction of Postglacial Transgressive Shorelines: An Example from Greece.’’ Quaternary Research 22: 31–45. Van Andel, T., and S. Sutton. 1987. Landscape and People of the Franchthi Region. Bloomington: Indiana University Press. Van der Leeuw, S., and C. Redman. 2002. ‘‘Placing Archaeology at the Center of Socio-Natural Studies.’’ American Antiquity 67: 597–605. Van Leusen, M. 1998. ‘‘Dowsing and Archaeology.’’ Archaeological Prospection 5: 123–38. Van Neer, W. 1993. ‘‘Fish Remains from the Last Interglacial at Bir Tarfawi (Eastern Sahara, Egypt).’’ In F. Wendorf, R. Schild, A. E. Close, et al., Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East, 144–54. New York: Plenum Press. Van Riper, A. B. 1993. Men among the Mammoths: Victorian Science and the Discovery of Human Prehistory. Chicago: University of Chicago Press. Vandiver, P., O. Soffer, B. Klima, and J. Svoboda. 1989. ‘‘The Origins of Ceramic Technology at Dolni Vestonice, Czechoslovakia.’’ Science 246: 1002–8. Vaughn, C. 1986. ‘‘Ground-Penetrating Radar Surveys Used in Archaeological Investigations.’’ Geophysics 51(3): 595–604. Vaughn, S. 1990. ‘‘Petrographic Analysis of the Early Cycladic Wares from Akrotiri, Thera.’’ In Thera and the Aegean World III. Vol. 1: Archaeology, ed. D. Hardy, 470–87. London: Thera Foundation. Vermeersch, P., and E. Paulissen. 1989. ‘‘The Oldest Quarries Known: Stone Age Miners in Egypt.’’ Episodes 12(1):35–36. Vitaliano, D. 1973. Legends of the Earth: Their Geologic Origins. Bloomington: Indiana University Press. Voelker, A., P. Grootes, M.-J. Nadeau, and M. Sarnthein. 2000. ‘‘Radiocarbon Levels in the Iceland Sea from 25–53 KYR and Their Link to the Earth’s Magnetic Field Intensity.’’ Radiocarbon 42: 437–52. Wagner, G., and W. Gentner. 1979. ‘‘Evidence in Third

315

Bibliography Millennium Lead–Silver Mining on Siphnos Island.’’ Naturwissenshaften 66: 157–58. Wagstaff, J., ed. 1987. Landscape and Culture: Geographical and Archaeological Perspectives. Oxford: Basil Blackwell. Walker, R. 1992. ‘‘Phosphate Survey: Method and Meaning.’’ In Geoprospection in the Archaeological Landscape, ed. P. Spoerry, 61–73. Oxbow Monographs 18. Oxford: Oxbow. Wall Paintings of the Tomb of Nefertari: Scientific Studies for Their Conservation. 1987. Malibu, Calif.: Getty Conservation Institute. Wallinga, J. 2002. ‘‘Optically Stimulated Luminescence Dating of Fluvial Deposits: A Review.’’ Boreas 31: 303– 22. Walsh, S. 2001. ‘‘Notes on Geochronologic and Chronostratigraphic Units.’’ Geological Society of America Bulletin 113: 704–13. Walter, R. C., P. C. Manega, R. L. Hay, R. E. Drake, and G. H. Curtis. 1991. ‘‘Laser-Fusion 40Ar/ 39Ar Dating of Bed I, Olduvai Gorge, Tanzania.’’ Nature 354: 145–49. Walthall, J. 1981. Galena and Aboriginal Trade in Eastern North America. Illinois State Museum Scientific Papers, Vol. 17. Springfield: Illinois State Museum. Wang, H., S. Ambrose, C. Liu, and L. R. Follmer. 1997. ‘‘Paleosol Stable Isotope Evidence for Early Hominid Occupation of East Asian Temperate Environments.’’ Quaternary Research 48: 228–38. Warashina, T., U. Kamaki, and T. Higashimura. 1978. ‘‘Sourcing of Sanukite Implements by X-Ray Fluorescence Analysis II.’’ Journal of Archaeological Science 5: 283–91. Wasburn, A. 1980. Geocryology: A Survey of Periglacial Processes and Environments. 2nd ed. New York: John Wiley and Sons. Waters, M. 1991. ‘‘The Geoarchaeology of Gullies and Arroyos in Southern Arizona.’’ Journal of Field Archaeology 18: 141–59. . 1992. Principles of Geoarchaeology: A North American Perspective. Tucson: University of Arizona Press. . 1999. ‘‘Book Review.’’ Geoarchaeology 14(4): 365– 73. Waters, M., S. Forman, and J. Pierson. 1999. ‘‘Late Quaternary Geology and Geochronology of Diring Yuriakh, an Early Paleolithic Site in Central Siberia.’’ Quaternary Research 51: 195–211. Weiss, H., M.-A. Courty, W. Wetterstrom, F. Guischard, L. Senior, R. Meadow, and A. Curnow. 1993. ‘‘The Genesis and Collapse of Third Millennium North Mesopotamian Civilization.’’ Science 261: 995– 1004.

316

Wells, L. 2001. ‘‘A Geomorphological Approach to Reconstructing Archaeological Settlement Patterns Based on Surficial Artifact Distribution: Replacing Humans on the Landscape.’’ In Earth Sciences and Archaeology, ed. P. Goldberg, V. Holliday, and C. R. Ferring, 107–41. New York: Kluwer Academic/Plenum Publishers. Wells, P., and C. Jorgensen. 1964. ‘‘Pleistocene Wood Rat Middens and Climate Change in Mojave Desert: A Record of Juniper Woodlands.’’ Science 143: 1171– 74. Wen, G., and Z. Jing. 1992. ‘‘Chinese Neolithic Jade: A Preliminary Geoarchaeological Study.’’ Geoarchaeology 7: 251–75. Wendorf, F., ed. 1969. The Prehistory of Nubia. Dallas: Southern Methodist University Press and Fort Burgwin Research Center. Wendorf, F., and R. Schild, assemblers. 1989. The Prehistory of Wadi Kubbaniya. Vol. 2: Stratigraphy, Paleoeconomy, and Environment, ed. A. Close. Dallas, Tex.: Southern Methodist University Press. Wendorf, F., A. Krieger, C. Albritton, and T. Stewart. 1955. The Midland Discovery. Austin: University of Texas Press. Wendorf, F., R. Schild, A. Close, et al. 1993. Egypt during the Last Interglacial: The Middle Paleolithic of Bir Tarfawi and Bir Sahara East. New York: Plenum Press. Wendorf, F., R. Schild, and A. Close, eds. 1989. The Prehistory of Wadi Kubbaniya. Dallas, Tex.: Southern Methodist University Press. Wendorf, F., R. Schild, A. E. Close, D. J. Donahue, A. Jull, T. Zabel, H. Wieckowska, M. Kobusiewicz, B. Issawi, and N. El Hadidi. 1984. ‘‘New Radiocarbon Dates on the Cereals from Wadi Kubbaniya.’’ Science 225: 645–46. West, F. 1982. ‘‘Archaeological Geology, Wave of the Future or Salute to the Past?’’ Quarterly Review of Archaeology 3(1): 9–11. Weymouth, J., and R. Huggins. 1985. ‘‘Geophysical Surveying of Archaeological Sites.’’ In Archaeological Geology, ed. G. Rapp and J. Gifford, 191–235. New Haven and London: Yale University Press. Wheeler, M. 1954. Archaeology from the Earth. Oxford: Clarendon. Whitbread, I. 2001. ‘‘Ceramic Petrology, Clay Geochemistry, and Ceramic Production: From Technology to the Mind of the Potter.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 449–57. New York: John Wiley and Sons. Whitley, D., and R. Dorn. 1993. ‘‘New Perspectives on

Bibliography the Clovis vs. Pre-Clovis Controversy.’’ American Antiquity 58(4): 626–47. Whittlesey, C. 1852. The Ancient Miners of Lake Superior. Cleveland: Academy of Natural Sciences Annals of Science. Wilkinson, T. 2001. ‘‘Surface Collection Techniques in Field Archaeology: Theory and Practice.’’ In Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 529–41. New York: John Wiley and Sons. . 2003. Archaeological Landscapes of the Near East. Tucson: University of Arizona Press. Willerslev, E., A. Hansen, J. Binladen, T. Brand, T. Gilbert, B. Shapiro, M. Bunce, C. Wiuf, D. Gilichinsky, and A. Cooper. 2003. ‘‘Diverse Plant and Animal Genetic Records from Holocene and Pleistocene Sediments.’’ Science 300: 791–95. Willey, G., and P. Phillips. 1955. Method and Theory in American Archaeology. Chicago: University of Chicago Press. Willey, G., and J. Sabloff. 1993. A History of American Archaeology. New York: W. H. Freeman. Williams-Thorpe, O. 1995. ‘‘Obsidian in the Mediterranean and the Near East: A Provenancing Success Story.’’ Archaeometry 37(2): 217–48. Williams-Thorpe, O., P. Webb, and R. Thorpe. 2000. ‘‘Non-Destructive Portable Gamma Ray Spectrometry in Provenancing Roman Granitoid Columns from Leptis Magna, North Africa.’’ Archaeometry 42: 77– 99. Wilson, D. 2000. Air Photo Interpretation for Archaeologists. 2nd ed. Gloucestershire, U.K.: Stroud. Wilson, G., L. Pavlish, G.-J. Ding, and R. Farquhar. 1997. ‘‘Textual an in Situ Analytical Constraints on the Provenance of Smelted and Native Archaeological Copper in the Great Lakes Region of Eastern North America.’’ Nuclear Instruments and Methods in Physics Research Bulletin 123: 498–503. Wintle, A. 1996. ‘‘Archaeologically Relevant Dating Techniques for the Next Century.’’ Journal of Archaeological Sciences 23: 123–38. Wintle, A., and M. Aitken. 1977. ‘‘Thermoluminescence Dating of Burnt Flint: Application to the Lower Paleolithic Site, Terra Amata.’’ Archaeometry 19(2): 111–30. Wintle, A., N. Shackleton, and J. Lautridou. 1984. ‘‘Thermoluminescence Dating of Periods of Loess Deposition and Soil Formation in Normandy.’’ Nature 310: 491–93. Wiseman, J. 1980. ‘‘Archaeology as Archaeology.’’ Journal of Field Archaeology 7: 149–51. . 2001. ‘‘Declaration of Independence.’’ Archaeology 54: 10–12.

. 2002. ‘‘Point: Archaeology as an Academic Discipline.’’ SAA Archaeological Record 2: 8–10. WoldeGabriel, G., G. Heiken, T. D. White, B. Asfaw, W. K. Hart, and P. R. Renne. 2000. ‘‘Volcanism, Tectonism, Sedimentation, and the Paleoanthropological Record in the Ethiopian Rift System.’’ In Volcanic Hazards and Disasters in Human Antiquity, ed. F. McCoy and G. Heiken, 83–99. Special Paper 345. Boulder, Colo.: Geological Society of America, 2000. Wood, W., and D. Johnson. 1978. ‘‘A Survey of Disturbance Processes in Archaeological Site Formation.’’ In Advances in Archaeological Method and Theory, ed. M. Schiffer, 1: 315–81. New York: Academic Press. Woodward, J., and P. Goldberg. 2001. ‘‘The Sedimentary Records in Mediterranean Rockshelters and Caves: Archives of Environmental Change.’’ Geoarchaeology 16: 327–54. Woodward, J., J. White, and R. Cummings. 1990. ‘‘Paleoseismicity and the Archaeological Record: Areas of Investigation on the Northern Oregon Coast.’’ Oregon Geology 52: 57–65. Woolley, L. 1929. ‘‘Excavations at Ur, 1928–29.’’ Antiquaries Journal 9: 305–39. Worster, D. 2001. A River Running West: The Life of John Wesley Powell. Oxford: Oxford University Press. Wreschner, E. 1985. ‘‘Evidence and Interpretation of Red Ochre in the Early Prehistoric Sequences.’’ In Hominid Evolution: Past, Present, and Future, ed. P. Tobias, 387–94. New York: Alan R. Liss. Wright, G. 1892. Man and the Glacial Period. New York: Appleton. Wright, H. E. 1957. ‘‘Geology.’’ In The Identification of Non-Artifactual Archaeological Materials, ed. W. Taylor, 48–49. National Academy of Sciences Publication 565. Washington, D.C.: National Academy of Sciences National Research Council. . 1976. ‘‘Environmental Setting for Plant Domestication in the Near East.’’ Science 194: 385–89. . 1993. ‘‘Environmental Determinism in Near Eastern Prehistory.’’ Current Anthropology 34(4): 458– 69. Wright, R. V., and R. Chadbourne. 1970. Gems and Minerals of the Bible. New Cannon: Deats Publishing. Wyckoff, D. 2002. ‘‘From Peds to Pedology: A History of Geoarchaeology on the Plains.’’ Review of Archaeology 23: 12–16. Wymer, J. 1985. Paleolithic Sites of East Anglia. Norwich, U.K.: Geo. Wynn, J. 1986. ‘‘Archaeological Prospection: An Introduction to the Special Issue.’’ Geophysics: Geophysics in Archaeology 51(3): 533. Yalden, D. 2001. ‘‘Mammals as Climatic Indicators.’’ In

317

Bibliography Handbook of Archaeological Sciences, ed. D. Brothwell and A. Pollard, 147–54. New York: John Wiley and Sons. Yoneda, M., A. Tanaka, Y. Shibata, M. Morita, K. Uzawa, M. Hirota, and M. Uchida. 2002. ‘‘Radiocarbon Marine Reservoir Effect in Human Remains from the Kitakogane Site, Hokkaido, Japan.’’ Journal of Archaeological Science 29: 529–36. Young, A. 1960. ‘‘Soil Movement by Denudational Processes on Slopes.’’ Nature 188: 120–22. Zachos, J., M. Pagani, L. Sloan, E. Thomas, and K. Billups. 2001. ‘‘Trends, Rhythms, Aberrations in Global Climate 65 Ma to Present. Science 292: 686–69. Zangger, E. 1991. ‘‘Prehistoric Coastal Environments in Greece: The Vanished Landscapes of Dimini Bay and Lake Lerna.’’ Journal of Field Archaeology 18: 1–17. . 1994. ‘‘Landscape Changes around Tiryns during the Bronze Age.’’ American Journal of Archaeology 98: 189–212.

318

Zaykov, V., A. Bushmakin, A. Yuminov, E. Zaykova, G. Zdanovich, A. Tairov, and R. Herrington. 1999. ‘‘Geoarchaeological Research into the Historical Relics of the South Urals: Problems, Results, Prospects.’’ In Geoarchaeology: Exploration, Environments, Resources, ed. A. M. Pollard, 165–76. Geological Society of London Special Publications 165. London: Geological Society. Zeuner, F. 1946. Dating the Past: An Introduction to Geochronology. London: Methuen. . 1959. The Pleistocene Period: Its Climate, Chronology, and Faunal Successions. 2nd ed. London: Hutchinson. Zhu, R., K. Hoffman, R. Potts, C. Deng, Y. Pan, B. Guo, C. Shi, Z. Guo, B. Yuan, Y. Hou, and W. Huang. 2001. ‘‘Earliest Presence of Humans in Northeast Asia.’’ Nature 413: 413–17.

Index

Page numbers in italics refer to illustrations

People (named in text) Abbott, Charles C., 9 Acher, Robert, 18 Adams, Robert M., 18 Adovasio, James, 85 Agenbroad, Larry, 20 Albritton, Claude, 16, 18 Alexander, John D., 46 Ambraseys, N., 261 Ambrose, Stanley, 21 Andersson, Johan Gunnar, 11–12, 12 Antevs, Ernst, 10, 15–16, 19, 138 Arpin, Trina, 59 Ashley, Gail, 20 Atkinson, Richard, 11 Aubrey, John, 24 Bate, Dorothy M. A., 10 Ben-Avraham, Z., 260 Bennett, Matthew, 109 Bergman, Ingela, 110 Bethell, P., 124 Bettis, Arthur, 20, 109, 141–42 Bevan, Bruce, 113 Binford, Lewis, 1, 18–19 Birkeland, Peter, 43 Biswas, A., 197 Black, D., 196 Boggs, Sam, 136 Bolt, Bruce, 260 Bordes, F., 19 Boucher de Crèvecoeur de Perthes, Jacques, 6 Boué, Ami, 10

Brady, J., 83 Braidwood, Robert J., 17–18 Brookes, C. E. P., 187 Breuil, Henri, 12 Brunhes, Bernard, 145, 156 Bryan, Kirk, 15, 16, 18–19 Bryson, Reid, 167 Butzer, Karl, 3, 17–22, 20, 83 Cackler, P., 197 Caesar, Julius, 201, 228 Caldwell, Joseph, 19 Cann, J. R., 16 Canti, M. G., 118 Carr, Christopher, 115 Casteel, R., 183 Catlin, George, 210 Chamberlin, Rollin T., 14 Cherry, John, 106 Childe, V. Gordon, 187 Clark, A. McFayden, 161 Clark, D., 161 Clark, J. D., 1455 Clarke, David L., 19, 23 Columbus, Christopher, 200, 203 Confucius, 275 Conyers, Lawrence, 21 Conze, Alexander, 7 Cornwall, Ian W., 17–18, 205, 233–34 Cremeens, David, 4 Crowther, J., 123 Curtis, G. H., 145 Curtius, Ernst, 7 Cuvier, Georges, 6 Dalan, Rinita, 21 Dall, William H., 7 Dalton, John, 7, 204 Daniel, Glyn, 5, 7, 11

Darwin, Charles, 6, 165 Davis, Edwin H., 7 Davis, Loren, 21 de Geer, Gerard J., 136 de Heinzelin, Jean, 18 de Mortillet, Gabriel, 7 de Silva, S., 264 de Terra, Helmut, 2, 273 Dickinson, William, 20 Dincauze, Dena, 37 Doig, R., 260 Donahue, Jack, 20, 85 Dörpfeld, Wilhelm, 7 Douglass, Andrew E., 10, 144 Dunn, Richard, 21 Eidt, Robert, 36 Eighmy, Jeff, 21, 155 Evans, Arthur, 1, 9 Evans, John, 7, 258 Evernden, J. F., 145 Falconer, Hugh, 6, 7 Farrand, William, 10, 15, 17, 20, 22, 86 Feathers, James, 160 Fedele, F. G., 19 Feibel, Craig, 21 Ferring, Reid, 1, 20, 40 Fiorelli, Giuseppe, 7 Frederick, Charles, 22 Fredlund, Glenn, 21 Freeman, Andrea, 21, 78 French, Charles, 2, 59 Frere, John, 5 Frison, George, 62 Fryxell, Roald, 1, 19 Gaffney, C., 122 Galadini, F., 261

319

Index Galli, P., 261 Gamio, Manuel, 11, 13 Gardner, E. W., 25 Garrison, Ervan, 21 Garrod, Dorothy, 10 Gautier, Achilles, 18 Geikie, James, 7 Gifford, John, 5, 22, 93 Gilbert, Grove Karl, 266 Gladfelter, Bruce, 4, 20, 22, 29 Glueck, Nelson, 232 Goldberg, Paul, 20, 59 Goodyear, Albert C., 19 Greenwood, P. H., 18 Griffin, James, 22 Griffiths, D., 244 Guccione, Margaret, 21 Hajic, Edwin, 109, 141–42 Harley, J. B., 110 Harrell, James, 21 Harris, Edward, 5 Hart, John, 4 Hassan, Fekri, 19 Hawkes, C., 274 Hay, Richard L., 20 Haynes, C. Vance, Jr., 9, 15, 18, 20, 81, 128–29 Heiken, G., 264 Heizer, Robert E., 16 Herodotus, 90, 201, 211 Herz, Norman, 20 Hill, Christopher L., 21, 160 Holliday, Vance, 15, 20–21, 38, 195 Holmes, William Henry, 8, 9 Homer, 95, 97 Hough, W., 261 Howard, Edgar B., 15 Hrdlička, Aleš, 14 Huckleberry, Gary, 21 Huntington, Ellsworth, 187 Hutton, James, 136 Isaac, Glyn, 18 Jelinek, Arthur, 22 Jing, Zichun, 21, 113, 125, 128 Johnson, Don, 102 Johnson, William, 20, 139 Jorgensen, Clive D., 6, 183 Joyce, Arthur, 77

320

Judson, Sheldon, 16, 17 Julig, Pat, 21 Kennedy, John F., 274 Kidder, Alfred V., 11–12, 14 Konigsson, Lars-Konig, 1 Kraft, J. C., 20, 21, 22, 90, 95 Krieger, Alex, 16 Kvamme, Kenneth, 21 Lambert, J., 124 Lartet, Edward, 7 Layton, Robert, 106 Leighton, Morris M., 16 Libby, Frank Willard, 16 Lindbergh, Charles, 11 Linford, N., 118 Linse, A. R., 131 Livy, 90 Lubbock, John, 6–7, 11 Luedtke, Barbara, 197 Lyell, Charles, 6, 6–8, 11, 60, 75, 87–88, 136 Lyman, R. Lee, 21 MacCurdy, George G., 14 Mallowan, Max, 263 Mandel, Rolfe, 20, 74, 141 Martin, Francine, 18 Máté, I., 124 Matson, Frederick R., 16, 239–40 Matuyama, Montonori, 145, 156 McBrearty, Sally, 52 McCoy, Floyd, 20, 264 McDonald, William A., 22 McKee, Brian R., 264 Melville, C. P., 261 Michels, Joe, 162 Miller, John, 16 Mohs, Friedrich, 196 Moore, Clarence B., 7 Movius, Hallam L., 16–17 Mueller, Raymond, 77 Nelson, Nels C., 11, 12 Niemi, T. M., 260 Nir, Yaacov, 129 Nordt, Lee, 21 North, Frederick J., 24 Obermaier, Hugo, 12 Otto, H., 16

Palacios-Fest, M., 177 Pausanius, 90 Peacock, Evan, 32 Pengelly, William, 6 Peters, Clare, 118 Petraglia, Michael, 77 Petrie, William Matthew Flinders, 7 Pewe, Troy L., 17, 132 Phillips, Philip, 187 Plato, 90, 142, 190 Pliny, 195–96, 201–2, 205, 209, 211, 228 Polo, Marco, 199, 203 Powell, John Wesley, 8, 8 Prestwich, Joseph, 6, 7 Ptolemy II, 246 Pumpelly, Raphael, 10 Putnam, Frederick W., 7 Pyddoke, Edward, 17–18 Raab, L. Mark, 19 Rapp, George (Rip), 5, 18, 20, 21, 22, 90, 95, 113–14, 127–28, 131, 207, 217, 226, 231, 234 Renfrew, Colin, 11, 16 Rigollot, Marcel Jérôme, 5, 7 Roosevelt, Anna C., 113 Rosen, Arlene Miller, 22 Rovner, Irwin, 174 Running, Garry, 21 Russ, John, 174 Salisbury, Rollin D., 9, 223 Sangmeister, Edward, 16 Saucier, Roger, 260 Sauramo, Matti, 10 Schiffer, Michael B., 19 Schliemann, Heinrich, 7 Schnellmann, M., 261 Schuldenrein, Joseph, 22 Schwarz, Henry, 20 Sellards, Elias H., 15, 15–16 Shackleton, N. J., 185 Shackley, M. S., 52, 222 Sheets, Payson, 20, 264 Shepard, Anna O., 229, 239–40, 241, 242 Sherwood, Sarah, 21 Sims, J. D., 260 Smith, M. A., 159 Smith, William, 136 Snefru, 244

Index Solecki, Ralph, 22 Spaulding, Albert, 22 Squier, Ephraim G., 7 Stafford, C. Russell, 21, 108 Stein, Julie, 10, 15, 17, 20–22, 28, 103, 131, 136 Steno, N., 136 Sternberg, Rob, 20 Steward, Julian H., 19 Strabo, 90, 246 Tang, J., 128 Taylor, Walter W., 17 Theophrastus, 195–96, 198, 228 Thomas, Cyrus, 7, 11 Thomas, Herbert H., 11, 16, 222 Todd, Larry, 62 Trigger, Bruce, 7 Ucko, Peter, 106 Van Andel, Tjeerd, 20 Van Lusen, Martijn, 122 Van Neer, Wim, 183 Veni, G., 83 von Post, Lennart, 10 Vondra, Carl, 20 Walthall, J., 206 Walther, Johannes, 91 Waters, Michael, 3–4, 20–21 Watson, David M. S., 11 Wendorf, Fred, 16, 18 West, Frederick H., 3 Wheeler, Mortimer, 11, 18 Whitbread, Ian, 59 White, Leslie A., 19 Willey, Gordon, 187 Williams, David, 16 Wilkinson, Tony, 106 Wilson, Lucy, 196 Wilson, Michael, 21 Winchell, Newton Horace, 9, 9–10 Woodward, David, 110 Woodward, Jamie, 21, 59 Woolley, Leonard, 263 Wright, George Frederick, 9 Wright, Herbert E., Jr., 16–18 Zeuner, F. E., 11, 18

Places Abbeville (France), 6, 76 Abri Pataud (France), 22 Absaroka Mountains (Wyoming), 86 Abu Simbel (Egypt), 253 Abu Sir (Egypt), 227 Abydos (Egypt), 227 Achates River (Sicily), 198 Acheron River, 95 Açigol (Turkey), 226 Aegean Islands, 214, 240, 263 Aegina (Greece), 238 Afghanistan, 199 Africa, 32, 37, 47, 76, 80–82, 143, 145, 155, 185, 207, 217, 235 Ain River (France), 82 Akrotiri (Thera, Greece), 142–43, 240, 249 Alabama, 81, 201, 206 Alaska, 7, 94–95, 99–100, 200, 205, 225, 260 Alexandria (Egypt), 202 Altamira (Spain), 81 Altiplano (Chile), 151 Anatolia (Turkey), 110, 172, 198, 203, 210, 213, 225–26, 228, 231, 233, 248 Andes Mountains, 123, 185, 187, 190, 202 Angel Mounds (Indiana), 114 Antarctica, 167 Anyang (China), 128 Anzick (Montana), 206 Appalachian Mountains, 200–201, 210 Arago site (France), 157 Arctic, 93 Argentina, 187 Arizona, 78, 81, 175, 177, 199, 210, 226, 247 Arkansas, 122 Arrox River (France), 160 Asia, 73, 81, 133, 155–56, 185, 193, 200, 225 Aswan (Egypt), 50, 198, 206, 214, 225, 227, 249 Atlantic Ocean, 167, 211, 261 Atzcapotzalco (Mexico City, Mexico), 11 Austin (Texas), 21

Australia, 63, 81, 84, 151, 159, 160, 193, 217, 238 Avebury (England), 222 Babylon (ancient), 206 Baldachin (Afghanistan), 199 Baltic Sea, 138, 200, 229 Basin of Mexico, 246 Bay of Navarino (Greece), 98 Beijing (China), 84, 246 Belize, 83, 162, 197, 201, 205, 237 Belyeas Cove (New Brunswick, Canada), 196 Bent Pyramid (Dashur, Egypt), 244 Bickerton (England), 112 Big Fork River (Minnesota), 104 Bilzingsleben (Germany), 213 Bir Sahara (Egypt), 26, 34, 66 Blombos Cave (southern Africa), 207 Bohai Sea (China), 263 Bohemian Massif (Europe), 203 Bolivia, 202, 264 Bonn (Germany), 21 Boston, 256 Brazil, 32, 121, 198–99 British Columbia (Canada), 172, 211, 262 Brittany, 233 Brixham Cave (England), 6 Bruchsal Aue (Germany), 159 Buffalo (New York), 247 Bunyoro (East Africa), 201 Burma, 199–200 Burnet Cave (New Mexico), 15 Caddon Huntsville site (Arkansas), 122 Cahokia Mounds site (Illinois), 113 Cairo, 202, 246, 253 Chalcedon (Turkey), 197 Calico Hills site (Yerma, California), 28 California, 28, 101, 128, 162, 168, 182, 197, 211, 219, 256 Canada, 93, 104, 172, 193, 231, 260, 263 Canal of Xerxes (Greece), 117 Carlsbad (New Mexico), 15 Carpathian Mountains (Hungary), 8 Carrara (Italy), 228, 252 Carthage (Tunisia), 2, 235

321

Index Çatal Höyük (Turkey), 110, 213 Çayonu (Turkey), 231 Cayster River floodplain (Turkey), 98 Cebolleta Mesa (New Mexico), 219 Central America, 198, 203, 210, 218 Central Great Plains (United States), 74 Cerro Rico de Potosi (Bolivia), 122 Chaco Canyon (New Mexico), 15, 199, 268 Châtelperronian site (France), 213 Cheswoanja (Kenya), 193 Chikoi River (Russia), 73 Chile, 151, 187, 248, 264 China, 6, 11, 41, 81, 84, 113 China Sea, 263 Chuandong area (Guizhou Province, China), 84 Çiftlik (Turkey), 226 Clacton (England), 37 Clovis site (New Mexico), 129, 237 Colby site (Wyoming), 62 Colombia, 200 Colorado, 180, 200, 237 Colossi of Memnon (Egypt), 224 Columbia River, 75 Copper River (Alaska), 260 Coppermine River (Canada), 205 Cornwall (England), 205, 233–34 Crater Lake (Oregon), 143, 263 Crete, 123, 211, 240, 258, 264 Cyprus, 114, 195, 230 Danger Cave (Utah), 81 Dashur (Egypt), 244 Dead Sea (Near East), 78, 169, 230, 260 Deer Creek (Montana), 140 Denbigh site (Alaska), 99 Denmark, 37, 183, 228 Devil’s Lair (Australia), 81, 160 Devil’s Tower (Wyoming), 108 Diring Yurikh site (Russia), 32 Djoser, Step Pyramid of (Egypt), 214, 248 Dmanisi (Republic of Georgia), 155 Dominican Republic, 200 Dorsetshire (Britain), 214 Dover Quarry (Tennessee), 197 Drachenloch (Switzerland), 83 Dravidian caves (India), 81

322

Dry Creek Site (Alaska), 100 Duluth (Minnesota), 128

France, 5–7, 10, 16, 22, 70, 76, 81–82, 94, 118, 143, 157, 160, 172, 178, 207, 213, 217, 227, 256 East Africa, 21, 27, 145–46, 155, 188, Franchthi Cave (Greece), 22, 86, 225, 264 118 Eastern Desert (Egypt), 106, 198, 234 Fraser Cave (Tasmania, Australia), Eckles site (Kansas), 237 84 Ecuador, 95, 200, 248 Edfu (Egypt), 237 Galisteo Basin (New Mexico), 12 Effigy Mounds (Iowa), 113 Ganges River (India), 215 Egypt, 7, 18, 21, 26, 34, 50, 66, 76, Gebel el Ahmar (Egypt), 225 106, 176, 183, 185, 197–99, 200, Georgia (United States), 215 202–3, 207, 209–10, 213–15, 218–19, Georgia, Republic of, 155 225, 227, 230, 233–34, 238, 244–45, Germany, 21, 37, 149, 159, 166, 189, 248, 252–53, 257, 266 213 Egyptian Sahara (North Africa), 80 Gibraltar, 85 El Castillo Cave (Spain), 82–84, 146 Giza (Egypt), 209, 214, 227, 234, El Chayal (Guatemala), 5 244, 253 El Kown (Syria), 157 Glacier Peak (Washington), 27, 143 Elâzig (Anatolia), 248 Gorham’s Cave (Gibraltar), 85 Elk Lake (Minnesota), 138–39 Government Mountain (Arizona), Elveden (England), 87 226 England, 5–7, 11, 32, 37, 70, 87, 94, Graham Cave (Missouri), 81 103, 112, 115, 126, 136, 180, 188, Gray’s Inn site (England), 3 200–201, 218, 222, 233, 252. See Great Basin (North America), 169, also Great Britain 183, 193 Engigstack site (Alaska), 99 Great Britain, 3, 76, 81, 83, 87, 180, Ephesus (Turkey), 22, 98 193, 200, 214–15, 235, 247, 274 Epirus (Greece), 95 Great Lakes (North America), 87, Ergani Maden (Turkey), 231 90, 206, 231 Erie Canal (New York), 247 Great Plains (North America), 65, Erzgebirge Mountains (Czech 74, 113, 129, 140, 166, 194, 227, 237 Republic), 233 Great Salt Lake (Utah), 78 Esna (Egypt), 266–67, 267 Greece, 18, 22, 33, 38, 86, 95, 98, 109, Ethiopia, 145, 209 117–18, 131, 142, 199–200, 203, 205, Euganean Hills (Italy), 248 207–8, 210, 213–14, 217, 228, 238, Europe, 4–7, 11–12, 32, 35, 37, 75–76, 245, 273 81–82, 87, 94, 113, 136, 139, 143–44, Green River (Kentucky), 28 149, 155, 159–60, 166, 171, 176, 180, Greenland, 22, 93, 143, 166, 167 183, 187, 192–93, 197–201, 203–4, Grimaldi Caves (Monaco), 10 211–12, 217, 232–33, 235, 237, 261, Grotte du Renne (France), 213 272 Guangxi Province (China), 246 Guatemala, 83, 121–22, 162, 192, 199, Fayum (Egypt), 209 205 Fiji Islands, 213, 230 Guebert site (Illinois), 235 Finland, 10, 229 Guizhou Province (China), 84 Florence (Italy), 271 Gulf Coast, 117 Florida, 12, 15, 37, 183, 212, 256 Gulf of Aqabah, 232 Folsom site (New Mexico), 14, 143 Gulf of Malia–Spherchios River Forest Bed (England), 87 floodplain, 96 Fort Clark State Historic Site Gulf of Mexico, continental shelf, (North Dakota), 113 117

Index Gulf of Suez (Egypt), 230 Gypsum Caves (Nevada), 81 Hagavatn (Iceland), 109 Hannaford site (Minnesota), 104, 105, 106 Harpers Ferry (West Virginia), 174 Haua Fteah (northern Africa), 81, 185 Hawaii, 85 Hearth Cave (Australia), 3 Hera Argiva (Italy), 3 Herculaneum (Italy), 128 High Lodge (England), 88–89 Hiscock site (New York), 182 Hissarlik (Troy), 1. See also Troy (Turkey) Hokkaido (Japan), 151 Holywell Coombe (England), 178 Hopeton Earthworks (Ohio), 113 Honduras, 205 Horseshoe Hills (Montana), 197 Hudson River, 247 Huanbei Shang City (China), 128 Hungary, 115, 157, 198

Jericho (Palestine), 195, 260 Jordan, 260 Jordan River Delta, 260 Kalambo Falls site (Zambia), 17, 50 Kansas, 140–41, 237 Kendricks Cave (Wales), 83 Kennewick (Washington), 48 Kenniff Cave (Australia), 81 Kent’s Cavern (England), 81 Kentucky, 28, 157 Kenya, 18, 162, 193, 213 Keweenaw Peninsula (Michigan), 7, 224 Khok Phanom Di (Thailand), 193 Klassies River Mouth (Africa), 81 Knossos (Crete), 258 Korephronisi (Crete), 211 Koster site (Illinois), 142 Krakatoa (Indonesia), 263 Krakow (Poland), 201–2 Kromdraai (South Africa), 82, 155 Ksar Akil (Lebanon), 16

La Colombière (France), 82 La Grande Pile (France), 143 Iberian Peninsula, 233 La Lagunita (Guatemala), 83 Iceland, 85, 109, 143, 166–67 La Lomita site (North America), 154 Idaho, 85, 263 La Olmeda (Spain), 174 Illinois, 81, 113, 141–42, 206, 235 Lake Agassiz (North America), 98 India, 2, 10, 133, 192, 198–99, 201, Lake Albert (East Africa), 201 211, 215 Lake Bonneville (Great Basin, North Indian Creek site (Montana), 75, 143 America), 169 Indian Ocean, 211, 263 Lake Chad (central Africa), 78 Indiana, 114 Lake Erie (United States), 247 Indus River, 215 Lake Great Falls (Montana), 160 Indus Valley (Pakistan), 233 Lake Lahanton (Great Basin, North Iowa, 141, 270 America), 169 Iraq (Asia), 172, 200, 209 Lake Lucerne (Switzerland), 261 Ireland (Britain), 237, 242 Lake Magadi (Kenya), 78 Isle of Portland (Dorsetshire), 214 Lake of the Clouds (Minnesota), 153 Isle Royale (Michigan), 220 Lake Superior (United States), 98, Israel, 59, 91, 129, 157, 198–99, 238, 204–5, 231 260 Lake Superior Basin (United States), Isthmus of Panama, 189 22, 98 Italy, 98, 117, 166, 200, 210, 227–28, Lake Titicaca (Andes), 190 238, 248, 261, 265, 268, 271 Lake Zeribar (Iran), 178 Itasca bison kill site (Minnesota), 172 Las Acequias (Arizona), 177–78 Lascaux caves (France), 81, 217 James River (Virginia) drainage, 230 Lavrion (Greece), 205 Japan, 35, 94, 151, 198, 202, 217, 230, Leaning Tower (Pisa, Italy), 256, 257 268 Lehigh Valley (Pennsylvania), 214

Lehringen (Germany), 37 Leonard Shelter (Nevada), 81 Leptis Magna (North Africa), 235 Les Echets (France), 143 Levant, 10, 158, 169, 172, 211 Libyan Desert, 201–2 Lisbon, 261 Llano Estacado (United States), 15 Loess Plateau (China), 185 Loire River (France), 5 London, 6, 126, 178, 215, 253, 256 Long Beach (California), 256 Longgu Cave (China), 84 Lubbock Lake site (Texas), 14, 175 Maiden Castle (England), 119 Maikop (Russia), 110 Maine, 218 Malaysia, 234 Maldive Islands (Indian Ocean), 211 Mammoth Cave (Kentucky), 157 Mammoth Junction site (California), 162 Manitoba (Canada), 200 Marduk Dam (ancient Mesopotamia), 245 Matapan Formation (Honduras), 205 Meadowcroft Site (Pennsylvania), 59, 81, 85 Mecca (Saudi Arabia), 154 Medina (New York), 247 Mediterranean, 41, 49, 59, 90– 91, 127, 143, 150, 172, 185, 188, 190, 192, 204–5, 209–11, 226–28, 230–34, 252, 257–58, 264, 268 Mekong River (Vietnam), 215 Melanesia, 263 Melbourne (Florida), 15 Melos (Aegean), 240 Mesopotamia, 171, 200, 215, 244, 263 Messenia (Greece), 22, 98 Methana (Greece), 238 Mexico, 11, 17, 21, 77, 81, 85, 113, 199–200, 202, 204, 246–47 Mexico City (Mexico), 256 Mezhirich (Ukraine), 207 Michigan, 7, 22, 204, 231 Michipicoten Island (Ontario, Canada), 204 Midland (Texas), 163 Midwest (North America), 9, 140, 141

323

Index Milling Stone Horizon (California), 101 Milwaukee (Wisconsin), 256 Minnesota, 9, 22, 99, 104, 128, 138, 153, 172, 206, 210, 219, 270 Mississippi, 189, 231 Mississippi Delta, 74 Mississippi River, 74, 107, 108, 110, 235, 260 Mississippi Valley, 7, 206 Mississippi Valley mounds, 200–201 Missouri, 81, 99, 260 Missouri River, 113, 197, 267–68 Modoc Cave (Illinois), 81 Mohawk River Valley (New York), 247 Mokattam hills (Egypt), 248 Mono Lake (California), 149, 168 Montana, 140, 143, 160, 197, 206, 263 Monte Alban (Mexico), 200 Montezuma Well (Arizona), 175 Montreal (Quebec, Canada), 21 Mordor caves (Australia), 84 Morocco, 85, 261 Mount Mazama (Oregon), 75 Mount Saint Helens (Washington), 143 Mount Vesuvius (Italy), 214 Mueilha (Egypt), 234 Mummy Cave (Wyoming), 86

3, 215, 225, 227, 245–46, 248, 257, 266 Nile valley, 18, 50, 120, 186, 197 Nineveh (ancient), 263 Nohmul (Belize), 162 North America, 7, 15–16, 28, 199, 207, 209, 219 North Anatolian Fault (Turkey), 190, 258 North Carolina, 201, 206, 231 North China Plain, 246, 261 North Dakota, 113 North Pacific Ocean, 151 North Sea, 94 Norway, 203, 252 Norwich (England), 94 Nova Scotia (Canada), 200 Nubia (ancient), 200, 203 Nubia (southern Egypt and northern Sudan), 18, 21

Oaxaca (Mexico), 77, 247 Obsidian Cliff (Yellowstone National Park), 210 Ogallala aquifer (Great Plains, North America), 194 Ohio, 9, 22, 219 Olduvai Gorge (Tanzania), 77, 81, 129, 145, 155 Oklahoma, 206, 242 Olorgasailie (Kenya), 18 Olympia (Greece), 7 Natural Trap Cave (Wyoming), 174 Ontonogan (Michigan), 231 Naxos (Aegean), 240 Ontario (Canada), 172, 206 Nazlet Khater (Nile Valley), 197 Oregon, 143, 225, 230, 262, 263 Near East, 106, 144, 171–72, 187, 195, Oslo (Norway), 252 199–200, 203, 205, 207, 213–14, Ozette site (Washington), 28 231–34, 240 Nenezi Dag (Turkey), 8 Pacific coast (North America), 28, Nevada, 3 90, 172, 188, 201, 262 Neville site (New Hampshire), 2 Pakistan, 158, 233 New England (United States), 6 Palace of Fine Arts (Mexico City, New Jersey, 7 Mexico), 256 New Madrid (Missouri), 9 Palestine, 7, 195, 263 New Mexico, 11, 12, 14–16, 81, 128, Paris, 201, 249, 253 226, 268 Pecos (New Mexico), 11, 240 New York, 10, 182, 219, 247, 253 Pecos Pueblo (New Mexico), 12 New Zealand, 123, 193 Pedra Furada (Brazil) site, 32 Nichoria (Greece), 33, 112, 207 Peking Man Cave (Zhoukoudian, Nihewan Basin (China), 156 China), 84 Nikopolis (Greece), 95 Pennsylvania, 22, 59, 81, 85, 187, 214, Nile River, 74, 76, 176, 185–86, 202– 219

324

Périgord rock shelters (France), 82 Peru, 106, 202, 264 Petén (Guatemala), 83 Pickering, Glacial Lake (England), 188 Picketpost Mountain (Arizona), 226 Pictou (Nova Scotia, Canada), 200 Pingasagruk (Alaska), 94 Pink Hill (London), 178 Poland, 201 Polynesia, 213 Pommeroy Park (England), 256 Pompeii (Italy), 7, 24, 128, 227, 249 Pontine marshes (Italy), 246 Poros (Greece), 258 Portales (New Mexico), 15 Portugal, 116 Poseidonia (Italy), 98 Pozzuoli (Italy), 214 Preseli Mountains (Wales), 222, 223 Proconnesus (Island of Marmara, Turkey), 228 Prospect Farm site (Kenya), 162 Pueblo Bonito (New Mexico), 268 Pylos (Greece), 22 Queens County (New Brunswick), 196 Queensland (Australia), 151 Real Alto (Ecuador), 95 Red Sea, 203, 246 Rio Grande River (North America), 203, 204 Rio Tinto (Spain), 205 Rio Verde Valley (Oaxaca, Mexico), 77 Rocky Mountains (North America), 9, 160 Rome (Italy), 199, 200, 213, 246, 249 Rudna Glava (Serbia), 218 Russell Cave (Alabama), 81 Russia, 32, 73, 110, 200, 227 Sahara Desert, 54, 80, 120, 158, 187, 210, 249 St. Acheul (France), 6, 10, 60 St. Anthony Falls (Minnesota), 9 St. Paul’s Cathedral (London), 256 Salisbury Plain (England), 223 Salzburg (Austria), 201

Index Samothrace (Greece), 7 San Cristobal (New Mexico), 12 San Francisco (California), 256, 260 San Jon site (New Mexico), 16 San Juan Island (Washington), 227 Sandia Cave (New Mexico), 81 Sanggan River (China), 156 Santa Clara Valley (California), 256 Santa Cruz floodplain (Arizona), 78 Santa Cruz River (Arizona), 78 Santa Rita district (New Mexico), 204 Saqqara (Egypt), 209, 214 Sardinia, 233 Sardis (Anatolia), 198 Scandinavia, 93, 138, 227 Schoningen (Germany), 37 Scotland, 120 Sea of Marmara (Turkey), 197 Seattle (Washington), 153 Seine River (France), 256 Sele River coastal plain (Italy), 98 Shangqiu area (China), 128 Shanidar Cave (Iraq), 81, 172, 209 Shiraz (Iran), 262 Siberia, 81, 94 Sidi Abderrahman quarry (Morocco), 85 Silver Springs (Florida), 256 Sinai (Egypt), 218 Skara Brae (Orkney, Scotland), 11 Slovakia, 198 Snake River Plain (Idaho), 85 Solomon Islands, 242 Somme Valley (France), 5, 6 Sotira Kaminoudhia (Cyprus), 114 South Africa, 21, 48, 155, 203 South America, 32, 133, 135, 185, 188, 190, 199, 202–3, 225, 247–48, 264 South Dakota, 124 Southeast Asia, 156, 225, 234 southern High Plains (United States), 190 Southwest (United States), 11, 14–15, 78, 117, 176, 199, 215, 219, 226 Spain, 12, 21, 81, 82, 146, 174, 205, 213, 235, 261 Stabiae (Italy), 128 Starr Carr (England), 37, 188 Sterkfontein (South Africa), 81, 82 Stonehenge (England), 11, 16, 222, 223

Suffolk (Britain), 180 Swartkrans (South Africa), 82 Sweden, 10, 11, 110, 228 Switzerland, 37, 83, 183, 261 Syria, 22, 59, 157, 238 Syros (Aegean), 240 Tabun Cave (Israel), 22, 157 Tadjikistan, 233 Tahiti (Polynesia), 213 Tanzania (Africa), 77, 145, 155 Tata (Hungary), 157 Taung (Africa), 82 Taurus Mountains (Turkey), 203 Tel Leilan (Syria), 59 Tel Michal (Israel), 49, 91 Tell Be’er Sheva (Israel), 59 Tell es-Sultan (Palestine), 195 Tennessee, 197 Teotihuacán (Mexico), 11, 85, 113 Terra Amata (France), 172, 213 Tesuque Valley (New Mexico), 16 Texas, 14, 16, 163, 175, 242 Thailand, 193, 234 Thames River (England), 126, 256 Thames River Valley (England), 76 Thebes (Egypt), 224, 266 Theoptra Cave (Greece), 86 Thera (Aegean), 142–43, 190, 214, 240, 263–64 Thermopylae (Greece), 22, 95, 96, 109 Thorne Moor (England), 180 Tibetan Plateau, 133 Tigris River, 245 Timna (Israel), 232 Tiryns (Greece), 245 Todos los Santos Formation (Guatemala), 205 Tollund (Denmark), 37 Tonga Islands, 150, 213, 230 Tower of London, 256 Towosahgy State Archaeological Site (Missouri), 260–61 Trinidad (Caribbean), 199 Troy (Turkey), 2, 7, 17, 22, 97, 188 Turkey, 22, 97–98, 191, 197, 199, 202, 211, 215, 228, 234, 238, 258, 264 Tuscany (Italy), 233 Two Creeks boreal forest (Wisconsin), 87, 176 Tyrol (Alps), 237

Ugarit (ancient), 211 Ukraine (Russia), 207 United Kingdom, 122 Upper Delaware Valley (Pennsylvania), 187 Upper Egypt (Africa), 76 Upper Midwest (United States), 141 Upper Mississippi Valley (North America), 240 Ur (ancient), 263 Ust-Kanskaiya Cave (Siberia), 81 Utah, 81, 180 Utatlán (Guatemala), 83 Valley of Mexico, 247 Valley of the Queens (Egypt), 252 Van (Turkey), 248 Venezuela, 199 Venice (Italy), 256, 257, 265 Ventana Cave (Arizona), 81 Vero (Florida), 12, 14, 15, 16 Virginia, 115, 174, 230, 231 Wadi Kubbaniya (Egypt) site, 76, 176 Wadi Natrun (Egypt), 202 Wales, 83, 222, 223 Washademoak Lake (New Brunswick), 196 Washington (State), 48, 143, 153, 172, 262, 263 Wenatechee site (Washington), 143 West Africa, 261 West Virginia, 174 Western Desert (Egypt), 26 Whitby (England), 200 Williamsburg (Virginia), 115 Windover site (Florida), 37 Wisconsin, 21, 199, 210 Wyoming, 86, 108, 124 Yakutsk (Russia), 32 Yangtze River (China), 246 Yellow River (China), 41, 127, 128, 215, 246, 263 Yellow River Plain (China), 113, 125 Yellowstone National Park (Wyoming), 7 Yerma (California), 28 York (England), 252 Yorkshire (England), 188 Yuanmou Basin (China), 156

325

Index Yucatan Peninsula (Mexico), 83, 129, 201

alfisols, 38, 41–42, 45, 100, 142 algae (habitat tolerance), 174–75 alkaline, 42, 170, 213, 216, 252; earth, Zagros Mountains (Asia Minor), 54, 251 172 alkalinity, 99 Zambia (Africa), 37, 50 alloys, 195, 244; copper, 272; tinZhoukoudian (China), 11, 81, 84 lead, 272 alluvial, 73–78, 80, 107, 128, 140–42, 189, 203, 219, 228, 235; chronology, Subjects 16; depositional systems, 67–68; deposits, 74, 157, 190, 215, 233, A soil horizon, 39 252; fans, 28, 65–66, 68, 75, 104, abiotic indicators, 168 109, 142, 156; fill, 73–74; plains, abrasion, 26, 51–52, 65, 76, 170, 66, 74; sediments, 64, 164, 203; 248, 271; artifact, 50, 52, 77, 81; sequences, 141; sites, 40; soil formechanical, 32–33, 100, 251 mation, 40; terraces, 40–42, 70, abrasives, 214, 218 72, 126; valleys, 48, 109, 219 accelerator mass spectrometry alluviation, 77–78, 107 (AMS), 139, 148, 165, 232 alluvium, 48, 68, 74–75, 108–9, 114, accretion, 98–99, 126; lateral, 69, 77; 116, 120, 126, 141–42, 213, 215, 258 vertical, 77 alpha particles, 153 Acheulian, 18, 21, 48, 50, 66–67, 72, aluminum, 40, 42; phosphate, 36–97, 90, 145–47, 172, 187; artifacts, 37, 199 67, 85, 88–89, 120, 136, 146, 155, Amazon River mound builders, 113 158, 210; hand axes, 26, 67, 89–90, amazonstone, 199 135 amber, 200, 223, 230 acid rain, 210, 221, 253, 270 American Geologist, 9 acidic soil, 42, 251 amethyst, 196, 198 acidity, 142–43, 251, 253, 272 amino acid epimerization and racemacids: amino, 160; carbonic, 210, ization, 160 221, 255, 271–72; deoxyribonucleic amphibians, 168 (DNA), 124–25, 183, 238; humic, amphibole, 199, 240 157; nitrogen, 253; succinic, 230; Anasazi Warm Period (climatic sulfuric, 231, 252, 271–72 event), 178 adobe, 215, 219, 229; bricks, 215; Andean prehistory, 95 structures, 258, 262, 268 andesine, 230 agate, 196, 198 andesite, 208–9, 230 Aggie Brown Member, 140 anglesite, 205 aggradation, 40, 69, 70, 73, 78, 109, anhydrite, 36, 45 160, 194; bar, 68; lateral, 75; anhydrous conditions, 251 vertical, 74 animals: body size related to climate, aggradational events, 70; interval, 181; browsers v. grazers, 180; bur192 rows, 100; extinction in North agricultural features, 119, 122, 166, America, 8, 143; waste, 29 246, 262 anomalies, 115; dipolar, 114; geologic agriculture, 38, 40, 77–78, 95, 105, v. archaeological, 115; in geophysi114, 130, 159, 166, 168, 171, 173, cal prospecting, 111, 113; magnetic, 178–79, 190, 192, 247–48, 265; 114–15 Mayan, 121; origins, 41, 187, Anomoeoneis sphaerophora, 175 190–91; use of water, 186, 219, 246 anoxic, 37, 171 alabaster, 195–96, 200 anthropogenic, 7, 19, 46, 113, 118, alcohol, 114 124, 128, 168; biogeochemistry,

326

122; context of buried sites or features, 111; disturbances, 114; impacts, 124; soil-forming processes, 114 anthropological archaeology, 3, 62, 274 anthrosols, 131 antimony, 122, 232 Apaches (Native Americans), 199 apatite, 36–37, 153, 196, 290 aragonite, 78–79, 178 archaeogeology, 1 archaeological geology, 2, 3, 5, 10, 22, 76, 134, 262; record, 1–8, 14, 16, 18–20, 23–25, 40, 57–58, 60, 62– 64, 73, 77–78, 80–81, 95, 99–100, 102–3, 109, 124, 132, 166–67, 185, 188–90, 210, 231, 248, 258, 261; sediments, 2, 45–46, 112, 124, 174, 193, 263, 274; strata, 257; survey, 75, 105, 110, 120, 126–28, 246, 274 Archaeological Society of Greece, 142 archaeology: bridging theory approach, 62; classical, 7; connected with historic geology and paleontology, 23; contextual approach to, 3, 22; distinct from anthropology, 4; earth sciences and, 4, 17; goal of excavations, 2; historic, 124, 147, 183; middle-range theory, 19; multidisciplinary, 104, 275; New Archaeology, 1; processual, 19; relation of geology, 24; salvage, 246; subdivision of natural sciences, 3; taphonomic approach, 60; theoretical basis for studies, 17; wetland, 183 archaeosediments, 29, 49, 54, 83 Archaic Period (North America), 28, 37, 74, 109, 129, 141–42, 172, 187–88, 192, 197, 206, 219 Archimedes screw, 125 argillaceous clasts, 240 argillic soil horizon, 42, 54, 142 argilliturbation, 98 aridosols, 41–42 Army Corps of Engineers, U.S., 107, 266 arroyos, 66 arsenic, 43, 124, 202, 232 articlast, 29, 32, 69

Index artifact(s), 2, 5, 9–10, 28, 51–52, 101, 157, 161–62, 197, 211, 217, 222–24, 234, 236, 272; abrasion, 52; and American Paleolithic, 7; assemblages, 32, 34, 51, 63, 66, 75–76, 80–81, 101, 146, 149, 187; carbonate incrusted, 28, 151; chemical alteration of, 163; in desert settings, 64; deterioration of stone, 251–52; and formation processes, 63, 85; geologic interpretation of deposits as ‘‘index’’ fossils, 143; mixing, 41, 65, 80, 162; orientation, 33, 44, 47, 62–63, 100; patterning, 29, 34, 50–51, 60, 62; primary and secondary accumulations, 28; redeposition, 8, 57, 76, 80, 90; secondary context, 33–34; size distribution, 50, 62, 81; size sorting, 33–34, 80, 98, 100, 102; Stone Age, 5; taphonomy, 41, 57, 60, 76 ash, volcanic, 125, 142–43, 145, 212, 214–15, 229, 249, 264–65; column, 249; fall, 131, 249; flow, 249; layers, 86, 118, 143, 150, 263; lenses, 85; plume, 249 ash, wood, 213 ashlar masonry, 214 Atlantis, legend of, 142 atomic absorption spectrometry (AA), 236 attapulgite, 218 augering, 125 Aulacoseira granulata, 175 Aulacoseira islandica, 175 Aurignacian, 82 Australopithecus, 155, 181 authigenic, 78, 118; carbonate deposition, 98; components, 54; mineralization, 44; sulfate, 44 Aztecs, 199, 200, 207, 210, 214, 264 azurite, 196, 204, 219; pigment, 205; smelting, 205, 232

balloons, and aerial photography, 118 bank(s): canal, 246; erosion, 75, 245, 266–67; overflow, 262; stability, 246 bar(s), 68, 69, 73–74, 90 barium, 122, 225 basalt, 41–42, 155–56, 195, 209, 219, 227, 238, 247, 251; trap rock, 208 basaltic: lava, 231; obsidian flows, 225; rock, 214 bases (chemical), 42 basin(s), 32, 34, 36, 66–67, 74, 78– 80, 90, 140, 168–71, 186, 265, 270; drainage, 169, 175; Great Basin (North America), 169, 183, 193; lake, 25, 54, 57, 63; Lake Superior Basin, 22, 98; Mediterranean, 192 bays, 91, 95, 117, 261 beaches, 50–51, 78, 79–80, 85, 90– 91, 95, 98, 118, 187, 260; cobble, 32; deposition, 85; lake, 51; recession, 266; sands, 34, 51, 95, 230; sediments, 51, 85; terraces, 98 beavers, 181 bedrock, 39, 41–42, 82–87, 107–8, 110, 114, 124, 150, 196–97, 199, 203, 212, 227, 242, 246, 249, 258; canals, 246; geology of, 110, 225, 260; maps of, 103, 106; quarries, 206; toxicity, 124 beetles, 179–80 bentonite, 145, 215 berms, 269 beta rays, 148 bioclasts, 171 biofacts, 269 biologic activity, 38–39, 100, 171, 265 biologic processes, 33, 62, 124; and artifacts, 62; mixing, 39; phosphates, 36; trampling, 52, 62 biosphere: modified by human activity, 61, 192 biostratigraphic: deposits, 25; markers, 57; record, 182; units, 38, 135 B soil horizon, 39–41, 44, 54, 100, biostratigraphy, 143 142, 240 biotic, 17, 39, 60, 80, 166, 168–69, Babylonians, 110 180–82, 193–94; activity, 39, 42, back-scattered electron petrography, 44, 54, 85; attributes, 39; con226 texts associated with humans, backswamps, 73 187; damage, 269; extinction of bajadas, 66 communities, 193; habitats, 181,

189; indicators, 168; inputs, 84; intervention, 168; signals, 169 bioturbation, 26, 40, 54, 79, 98–102, 114, 138–39, 151, 265 bird fossil remains, 181–82 bison (extinct), 14, 143, 172 bitumen, 230–31 bivalve crustacea, 178 ‘‘black amber,’’ 200 bleaching, 157, 202 ‘‘blue mud of the meadows,’’ 247 ‘‘blue stones’’ (Stonehenge), 222 bog people, 183 bogs and boglands, 36, 78, 80, 168, 179, 219; peat, 78, 104, 143, 183; preserved prehistoric human remains, 37, 183 bone, 33, 62, 64, 95, 124, 151, 157, 159–60, 217, 251, 255, 269; antlers, 188; buried, 141, 163, 221, 255; rabbit, 163; remains, 60, 100, 160, 163 boron, 124 boulder clay, 87 boulders, 29, 188, 199, 208, 218; obsidian, 162 Brady soil, 140 breccia, 84, 86, 146, 155 brick, 215, 229; adobe, 215, 258; burnt, 215; concentrations, 116; making, 215; mud, 34, 47, 213, 215, 258; sun-dried, 215 brickearth, 76 brine, 201–2; boiling of, 201; springs, 202 bronze, 202, 219, 224, 232, 234, 272; and cassiterite, 205; Chinese, 224; composition, 232 Bronze Age, 16, 18, 33, 49, 73, 83, 125, 128, 131, 134, 142, 180, 188, 192, 203–5, 207, 214, 218, 230, 232–35, 238, 240, 245, 268 Brule Formation, 237 Bt soil horizon, 42, 54, 142 Btn soil horizon, 40, 54 Buckner Creek paleosol, 141 building: materials, 195, 213–16; stone, 196, 201, 214–15, 219, 227, 251–53, 271 Bureau of American Ethnology, 8 burial(s), 37, 40, 43, 58, 62–66, 68, 73, 77, 80–81, 86–87, 100, 109, 157,

327

Index burial(s) (continued) 162, 171–72, 183, 194, 206–7, 223, 245, 249, 251–52, 255, 262, 265– 66, 269, 271; mound construction, 194; Neanderthal, 172; processes, 251; protective, 246; rate and depth, 161–62, 266; Viking ship, 144, 183 buried: bone, 141, 163, 221, 255; features, 11, 113, 118–19, 269; materials, 255; river valleys, 120; sites, 111, 127–28, 249, 256, 269– 70; soils (see paleosols), 33, 38, 43, 140–41; structures, 116; walls, 115–16, 120 burned features, 114 burnt flint, 158 burnt lime, 214

capillary: action, 255; fringe, 99; rise, 54, 99, 221 carbon, 35, 44, 124–25, 138, 150–51, 184, 206, 238; ‘‘dead,’’ 57, 142, 148, 150–51, 228, 238; isotope fractionation, 151–52; old, 59, 150; photosynthetic pathways, 173; reservoir, 138, 148–49, 150, 153; stable isotopic values, 151, 184, 186 carbon-12 ( 12*C), 147–48 carbon-13 ( 13*C), 147–48, 152 carbon-14 ( 14*C), 147–48, 148, 149, 151 carbon dioxide (CO 2*), 122, 148, 214, 221, 255 carbonaceous, 37, 44, 253, 261; deposits, 54; layers, 75; particles, 271 carbonate(s), 26, 28, 34, 36, 40, 43– C soil horizon, 39–41 44, 46, 54, 67, 78–79, 84–86, 99, cadmium, 124 124, 139, 145–46, 153, 157–58, 176– calcareous, 34, 36, 61, 170, 176, 242; 77, 192, 200, 202, 204–5, 210, 214, concretions, 216; mud deposits, 217, 221, 251–52, 265, 268, 271–72; 188; nodules, 43; ostracod shells, authigenic, 98; dissolution, 271; 176; precipitates, 36–37; sand, 88; enrichment, pedogenic, 99; forms sandstone, 34; shells, 176; soils, of, 36; incrustation on artifacts, 272 151; lake, 183; microfossils, 240; calcic soil horizon, 40, 44–45 organic, 124; precipitates in lakes, calcisols, 44 79; proportions to clastics, and calcite, 35–36, 47, 54, 78–79, 81–82, deposition, 54; Quaternary, 157; 99, 160, 178, 195–97, 199–200, 202, rocks, 36, 246, 256; secondary 208, 213, 216, 252, 271; cement, 47, accumulation, 99; shell, 176 271; pigments, 217, 219; veins, 185 carbonation, 251 calcitic formations, 157 carbonic acid, 210, 255, 272 calcium, 122, 163, 177, 214, 221, 225, carnallite, 202 251 carnelian, 198 calcium carbonate, 28, 34–36, 40, casing blocks, 248 42–43, 79, 99–100, 139, 146, 202, cassiterite, 205, 218, 232, 234 206, 208, 210, 213–14, 216, 221, cast iron, 272 247, 251–52, 271–72; crusts on catacombs, 249 sherds, 251; in paleosols, 43 cation, 163, 251 calcium-magnesium carbonate, 217 catlinite, 210 calcium oxalate, 172–73 cave(s), 85–86, 118, 172, 174; art, 84, calcium oxide, 214, 216, 240 146, 196, 207, 217; deposits, 81, calcium phosphate, 28, 37, 210 84, 86; igneous rock, 85–86; limecalcium sulfate, 36, 201–2, 271 stone, 81, 83–84; pseudo karst, 83; calcrete, 44, 54 sediments, 6, 86, 125, 159, 171 caldera, 142, 226, 263 cement: calcite, 47, 271; portland, caliche. See calcrete 214; Pozzolanic, 214; underwater, canal(s), 98, 106, 110, 117, 121, 154, 214, 247 177–78, 245–47 cementation and induration, 47 cap rock, 108 Cenozoic, 4, 24, 133, 156, 189

328

ceramic, 131, 167, 202, 212, 229–30, 238–40, 257; glaze, 213; temper, 242; thin sections, 239 cerium, 229 cerrusite, 205, 234 cesium, 229 Chadron Formation, 237 chalcedony, 196–97, 219, 227, 237 Chalcolithic, 10, 204 chalcopyrite, 204, 207 chalk, 36, 87, 110, 197, 208, 218, 228, 256, 266 channel(s), 33, 66, 69, 72–75, 77–78, 88, 128, 188, 218, 263; aggradation, 68–9; braided, 68, 74, 77, 107; drainage, 62, 120; fill, 69, 73; margins, 262; migration, 76–77, 189; river, 50, 76–77, 116–18, 260 charcoal, 29, 46, 73, 76, 141, 150–51, 159, 175–76, 193, 206 chemical: alteration of artifacts, 163, 223; analysis, 37, 75, 162–63, 190, 224–25, 235, 240; composition of buried bone, 163; composition of ostracod shell, 177; dating techniques, 163; decomposition, 26, 250; deposition, 29, 35, 39, 81; nutrients, 122, 124; precipitates and precipitation, 36, 78, 81, 138–39; prospecting to locate sites, 122; weathering, 39, 164, 212, 221, 251, 254, 271 chernozems, 42 chert, 32, 196–97, 216, 219, 223, 227–28, 237, 249; Hudson Bay Lowland, 227; nodules, 197, 216; red (see also jasper), 197; sourcing, 196 chisels, 199 chloride, 202, 272 chlorine, 124 chlorine-36 ( 36*Cl), 164 chlorite, 47 chroma, 44–45 chronology, 2–3, 5, 10, 136; absolute, 90 chronometric techniques, 16, 132 chronozones, 144 Chumash (Native Americans), 195 cinnabar, 205, 217 Cladocera, 178 clams, 95, 178

Index clastic(s), 34, 39, 50; and deposition, 53; deposits, 29, 34, 78; materials, 29, 36, 78, 81; particle(s), 29, 34, 54, 207–8, 271; sediments, 54, 82, 96, 231; units, 88 clasts, 29, 53, 76, 146; argillaceous, 240; eroded, 163; non-artifact, 69 clay, 41–42, 45, 50, 77, 118, 138, 210– 15, 219, 229, 249, 253, 263, 266, 268, 271; adobe bricks, 215; beds, 41, 229; boulder, 87; burnt brick, 215; deposits, 85, 106, 154; expansion and contraction, 41; as geologic term, 214; glacial, 229; halloysite, 212–13; illite, 213; kaolinite, 212; matrix, 253; micaceous, 240; minerals, 26, 41, 192, 212–15, 229, 239, 251, 258; montmorillonite, 41; mortar, 215; paste, 239; petrographic study of pottery, 239; pottery, 20, 212, 223; sand-silt-clay ratios, 120, 215; size, 29; smectite, 213; sourcing, 240; structural strength, 214 climate, 7, 10, 14, 38–39, 54, 114, 135, 166–67, 169, 175, 178, 182–83, 185, 187, 191–94, 201, 219, 244, 254, 275; and animal body size, 182; Bronze Age, 192; change, 17, 34, 59, 64, 133, 165–69, 179–80, 182, 185–87, 189–94, 238; cold, 169, 185, 189; in oxygen isotope ratios, 185; and pack-rat middens, 183; and pollen assemblages, 171; temperate, 185, 191, 254–55; tree remains as indicator, 176; tropical, 244; and use of diatom taxa, 175; variation, 169, 171 Clovis, 14–15, 33, 59, 62, 129, 135, 143, 163, 190, 237 cluster analysis, 242 coal, 95, 150, 152; burning, 124; clinker, 209; mining, 248 coastal: areas, 90–91, 172, 257, 260, 262; change studies, 91; environments, 91, 118; erosion, 249, 266, 270; harbors, 110; landscape context, 95–98; marine depositional settings, 90–91; plains, 108; processes and site formation, 91–95; settings, 73, 93, 117; sites, 110, 257, 262; systems, 74; uplift, 95

coasts, 63, 90, 110, 167, 178, 201, 250 coating(s), 28, 46–47, 213, 268 cobalt, 124, 229 Coleoptera (beetles), 180 colloidal, 36; ferric oxides, 36; silica, 213 colluvium, 75, 142, 159 color, 39–41, 44–45, 63, 119, 128, 142–43, 195–200, 203–7, 209, 212, 217, 219, 227, 272 compaction, 44, 52, 99, 125, 195, 244, 257–58, 265–66, 269 computer: imaging technology, 161; digital processing, 110; software, 130 concretions, 46, 99, 211, 216, 242, 272; chert nodules, 216; ferrous, 272; septarian, 216 conductivity: electrical, 11, 115; electromagnetic (EM), 116; subsurface, 115 conglomerate, 33, 208, 214 conservation, 251–52, 254, 271–72 construction (large-scale), 49, 74, 106, 114, 159, 194, 209, 218, 222, 244–48, 253, 265, 270 contamination: detrital, 147; of potable water, 245 context: primary, 26, 29, 32–34, 58, 67, 69, 76–77, 102; secondary, 26, 33–34, 57, 76–77; systemic, 19, 24, 62, 64, 69, 76, 102 continental crust, 188 continental shelf (Gulf of Mexico), 117 contour intervals, 103 contraction cracks, 90 convergent light, 239 copper, 8, 22, 33, 124, 196, 199–200, 202–6, 218–19, 223–24, 231–32, 235, 238, 272; alloy metallurgy, 204; complex copper minerals, 224, 232–34; deposits, 204–5, 214, 231–32; native (metallic), 7, 202, 204–5, 218, 224, 229, 231–32; ore, 200, 204, 218, 234; slag dumps, 21; smelting, 124, 196, 202, 204–5, 218, 232; sourcing, 232; sulfide deposits, 205, 231–32; sulfide ores, 204–5, 223, 231 coprolites, 168, 172 coquina, 36

coral, 208, 216 core drilling, 90, 95–96, 111–12, 114–15, 125–29, 126 corn, 129, 174, 209 corrosion, 203, 221, 271–72 corundum, 196, 214 cosmic-ray intensity, 148 cracking, 239, 245, 253, 259, 272 creep: downslope, 109, 250; seasonal, 249; soil, 99, 265; surface, 50 crinoid, 211 crop(s), 36, 42, 120, 122, 192, 247, 264 cross bedding, 51, 58, 249 cryostatic pressures, 100 cryoturbation, 88, 98–100, 151, 266 crystal(s), 146, 156–57, 155–56, 160, 197–99, 207–9, 219, 227, 238, 251–53, 269 cultivation, 36, 38, 42, 109, 120, 193–94, 247, 255 cultural resource management, 113, 115, 270, 275 cummulic soil horizon, 142 currents: longshore, 74; wind driven, 169 cut and fill, 69–70, 72–73 D-amino acid (right-handed), 160 D/L ratio conversion, 160 dacite, 227 dams, 74, 106, 168, 194, 245–46, 259, 267, 270 data analysis, 110, 130, 223, 242–43 databases: absolute, 10; accelerator mass spectrometry (AMS), 148, 165, 207, 237; argon-argon (Ar-Ar), 16, 144, 166; dendrochronology, 144, 150; digital, 130; electron-spin resonance (ESR) and luminescence, 156–60; of exposed surfaces, 163; fission track, 153–56; geoarchaeological, 130; hydration (obsidian), 160–62, 161; independent, 141; infrared stimulated luminescence (IRSL), 159–60; loess deposits, 139, 141, 159; modeling, 130, 153; paleomagnetic and archaeomagnetic, 155, 155–56; paleosols in loess and alluvium, 140–41; pollen stratigraphy, 138; potassium-argon(K/Ar), 16,

329

Index databases (continued) 139, 142, 144–46, 157, 159, 166; radiocarbon, 10, 16–17, 37, 57, 59, 128, 139, 143–44, 147–51; radiometric, 144–46; rate of weathering, 163–64; relative, 132, 142, 160, 162; rind, 163; shell-midden deposits, 185; and travertine, 146, 157, 175; uranium series (U-series), 146– 48, 157, 160, 166; using animal and plant remains, 143–44; varves, 135–36, 144; volcanic ash, 136, 142 debitage (flaking debris), 50–51, 66, 80–81, 159, 227 debris flow, 66 deer, 181, 188, 193 deflation, 32, 58, 65, 67, 269 degradation, 59 dehydration: seasonal, 194 delta(s), 68–69, 74, 80, 87, 90–91, 97–98, 185, 194, 215, 257, 260 deltaic deposits, 256 Delvin Quadrangle topographic map, 104 dendrites, 216, 217 deoxyribonucleic acid (DNA), 124–25, 183, 238 Department of Agriculture, U.S., 43 deposition, 16, 25, 28, 33, 36, 40, 43, 47, 50–54, 53, 57–58, 63–64, 66, 69, 73, 75–77, 79–80, 90, 96, 99– 100, 138, 147–48, 150, 153, 166, 168, 171, 178, 183, 194, 210, 245, 249, 263; ash, 27, 41; calcite, 81; carbonate, 35, 78, 84, 86, 98, 159, 162; chemical, 35, 81; of chemical precipitates, 78, 81; clastic, 54, 79, 171; in coastal and marine settings, 90; contexts, 33, 50, 57, 68, 74, 76, 78, 91; eolian, 81, 249; flood, 66, 188; fluvial, 76, 78, 81, 189; human, 62, 263; lake-related, 80, 202, 212; loess, 139, 187; overbank, 77, 126; patterns, 78, 109, 164, 193; pollen, 170–71; post-depositional conditions, 45; processes (see also depositional systems), 14, 45, 75, 86, 128, 187; sediment, 28, 40, 78, 136; stream, 69–70, 249; of tufa and other carbonates, 80; and Walther’s Law of Correlation of Facies, 91

330

depositional systems, 47, 58, 60, 63–64, 67, 74, 81 deposits: alkaline, 170; alluvial, 74, 140–41, 252; cassiterite, 205; chert, 227; clay, 77, 85, 106, 154, 229; copper ore, 200; copper sulfide, 231–32; deltaic, 256; eroded or reworked, 151; evaporite, 201; fluvial, 68, 78, 87, 179; galena, 234; glacial, 88, 164; gold, 235; Holocene, 74, 168; lag, 58, 65, 206; lake, 34, 76, 146, 159, 168, 180; lode, 219, 231; loess, 27, 51, 139–40, 159, 187; mass wasting, 99; midden, 12, 93, 227; obsidian, 210, 225–56; placer, 203, 205, 219, 232–33; point bar, 260; prodelta, 74; pyroclastic, 225; Quaternary, 27, 163, 179; salt, 201; sedimentary, 25–26, 26, 36, 47, 58, 60, 64, 91, 94, 103, 176, 187, 260; tephra, 230; travertine, 96, 146, 157; volcanic ash, 136, 142–43 desert(s), 41, 50, 63–66, 65, 80, 107, 110, 129, 197, 233, 249 desert varnish (cation-ratio), 163 destruction, 26, 37, 42, 65, 100, 131, 142, 215, 240, 257–58, 260–66, 270 detrital: contamination, 147; zone, 87 diagenesis, 26, 28, 47, 59, 223, 255, 265–66 diagenetic, 24, 36, 62, 86, 99, 157, 239, 266, 269 diamond, 196, 214 diatomites, 77, 175 diatoms, 69, 79, 168, 174–75, 229 diorite, 195, 208–9, 214, 227 discoloration, 253 discontinuities, 18, 58, 99, 116, 191, 257, 271 discriminant analysis, 224, 237, 242 ditches, 11, 115, 119–20, 178, 194, 201–2, 246 dolerite, 214, 222 dolomite, 78, 82, 197, 213 dosimetry, 157 downslope creep, 109, 250 drainage, 46–47, 74, 77, 90, 99, 108, 126, 129, 140, 149, 169, 174, 185, 194, 221, 230, 242, 246, 256, 268; basin, 169, 175; channels, 62, 120; networks (ancient), 120, 130

dripstone, 81, 146 drought, 78, 177, 186, 190 drumlins, 87 dune(s), 14, 27, 34, 50–51, 64–65, 76–77, 85, 90–91, 95, 104, 249; eolian deposits, 51, 87; interdunal ponds, 66; sand, 49, 51, 66, 76, 104; sediments, 76; stabilization, 66 dung, 151 dwellings, 28, 37, 73, 95, 175, 183, 213, 258 dye, 210–11, 218 ‘‘early man’’ controversy, 8 earthenware, 252 earthquake(s), 54, 109, 117, 128, 131, 169, 188, 190–92, 256–62, 265, 268, 273; Copper River Delta (Alaska), 260; fissures, 261; Jericho (Israel), 195; Knossos (Crete), 258; Lisbon (Portugal), 261; Modified Mercalli Scale, 258; New Madrid (Missouri), 260; Richter magnitude scale, 258; San Francisco, 260; Shiraz (Iran), 262 earthworms, 28, 100 East African Rift system, 169, 265 ecofacts, 135, 188 ectotherms, 182 edaphic, 104 edge effects, 115 efflorescence, 51 electrical conductivity, 11, 115 electrical resistivity, 113, 115, 128 electromagnetic (EM) conductivity, 116 electromagnetic spectrum, 111, 120 electron spin resonance (ESR), 156, 156–58, 166 electron traps in minerals, 156 electrum, 203, 272 element concentrations (rare earth), 230–31 elements, 25, 37, 43, 122–24, 130– 31, 146, 156, 163, 168, 204, 213, 222–27, 229, 231–32, 235 elephants, 143, 181 elk, 181, 188 eluvial horizon, 45 eluviation, 59, 98 emery, 214, 218

Index engineering geology, 248, 268–69 entisols, 41, 109 environmental change and archaeological interpretation, 184; evidence for, 15; human habitats and geoecology, 186–88; inferring environmental change, 166–67 environmental determinism, 187 environments, geologic, 231 Eocene, type formation (Oregon), 230 eolian, 26, 32, 48, 50–51, 64, 66, 76–77, 81, 84, 87, 157, 187, 249 epidote, 47 epigraphy, 228 epipedon, 41–42 erosion, 26–27, 34, 40–41, 44, 47, 58, 64–68, 70, 74–75, 77, 79–80, 86, 90, 102, 109, 123, 159, 165, 187, 189–90, 192–94, 215, 219, 245, 247, 249–50, 254, 256–57, 265–66, 268–69; bank, 75, 245, 266–67; coastal, 249, 266, 270; events, 170; features, 51, 65, 91, 104; intervals, 192; lateral, 247, 256; processes, 266; and redeposition, 64, 74, 80, 102, 107; and reservoirs, 267; and slopes, 268; surfaces, 91 (dating), 163; and transport, 27, 34, 189 escarpments, 104, 268 estuary, 110, 151, 153 eustatic rise curves, 93 evaporation, 54, 78–79, 82, 99, 129, 169, 185, 201–2, 252–53 exfoliation, 251 exogenic sediments, 50 extinction, 94, 167, 182, 193–94

80, 91; solution, 83–84; stranded, 90 feldspar(s), 145–46, 157, 159, 199, 208–9, 213, 215, 218–19, 240, 251; alteration, 212; amazonstone, 199; moonstone, 199; potassium, 144; sunstone, 199; weathering, 212, 251 felsite, 208–9, 219 fen carr, 188 ferric: oxides (colloidal), 36; sulfate, 231 ferrous, 36; concretions, 272; iron compounds, 47; iron sulfide, 272 ferruginous quartzite, 225 fill, 26, 29, 69–70, 72–73, 81, 86–87, 115, 141, 159, 256, 258, 268–69 fingerprint(s): chemical, 225, 232; isotope, 228; non-overlapping, 224; trace-element, 224, 227, 231, 242, 236 fire, 32, 46, 114, 122, 192–93, 207, 215 firestone, 207 fish, 37, 78, 168, 182–83, 187, 262 flagstone, 248 flint, 33, 60, 115, 157, 188, 197, 227–28; burnt, 158; defined, 197; knapping, 197; Knife River Flint, 227; mines, 115 flood legends, 262–63 flooding, 69, 73–75, 77, 80, 85, 110, 178, 189, 245, 247, 262–63 floodplain(s), 25, 34, 40, 43, 47, 73, 75–78, 85, 91, 96, 98, 104, 108– 10, 126, 128, 189, 194, 246, 256, 262–63, 266; deposits, 34, 73, 85; Yellow River (China), 127, 215 floods, 33, 40–41, 78, 109, 131, 177, facies, 4, 35, 74; freshwater lake, 79; 189–90, 245–47, 262–63, 265–67, marsh, 79; microfacies, 59, 189; 270 sedimentary, 51, 91; Walther’s floors, 58–59, 115–16 Law of Correlation of Facies, 91 flora (fossil), 167, 229. See also plant(s) faulting, 131, 169, 188, 192, 261 floralturbation, 100 faults, 54, 243, 260, 265 flow: ash, 249; debris, 28, 32, 66, fauna. See animals 75, 90; lava, 85, 164, 189, 225; features: archaeological, 24–25, 73, obsidian, 225–26 98, 102, 113–16, 118–19, 131–32, 135, flowstone, 36, 81, 83, 146–47, 155 153, 169, 245; buried, 11, 113, 118– fluorine, 163 19, 269; burned, 114; depositional, fluorite, 196 65, 104; erosional, 51, 65, 91, 104; fluvial, 33, 50, 64, 68–70, 73–78, 81, karst solution, 84; lake-margin, 84, 87, 145–46, 159, 179, 189, 203, 80; pedological, 110; shoreline, 74, 212, 249

flux, 120, 223, 236 Folsom artifacts, 14, 16, 65, 143, 173, 175, 190 food: resources, 104; waste, 122 footslope areas, 194 forest, 32, 36, 41–42, 87–88, 121, 159, 189, 192, 218, 245, 249; boundaries, 172; clearing, 188, 192; species, 178; spruce, 172 forges, 114 fortifications, 117, 268 fossils, 57, 60, 81, 87, 143, 145– 46, 149, 155–56, 163, 167–69, 179, 181–82, 216, 219, 228; australopithicene, 47; hominid, 16, 145–46, 156, 265; shell, 211 foundations: historic buildings, 114, 256 fractionated, 151, 184 fracture, 196–98, 200, 208–9, 212, 230 fragmentation, 47, 251 freeze-thaw, 32, 85, 266, 269; cycles, 90, 269 frost: action, 26, 100, 221, 253, 255, 265; damage, 253; effects on building stone, 253; heave, 45, 100 fumarole (gas emissions), 151 fungi, 168 galena, 205–6, 219, 234–35 gamma, 114, 235–36 gamma rays, 236, 272 gangue, 203, 218 garnet, 230 gas chromatography, 230 gas emissions (fumarole), 151 gastropods, 79, 178, 211 geoarchaeology, 1–13, 17–19, 19–25, 27, 40–41, 59–60, 86, 103, 113, 118, 125, 128, 132, 134, 163, 165, 167, 191, 208, 232, 238, 245, 257, 260, 262, 265, 268, 270–71, 273–75 geochemical contamination, 151–52, 160 geochemical environments (ancient), 124 geochemical prospecting and analysis, 111, 122–24 geochemical surveys, 105 geochemistry, 2, 22, 129, 227, 271 geochronology, 2, 15, 16, 22, 24, 160

331

Index geoecology, 3–4, 186 geofacts, 1, 32, 52, 99 geographic information systems (GIS), 130, 274 geologic: anomalies, 115; aspects of land transportation, 248; deposits, 5, 151, 153, 223–24, 230, 234, 236, 242, 261; environments, 231; maps and mapping, 98, 106– 7, 119, 260, 262; networks, 248; time divisions, 134, 136, 137; time horizons, 268; units, 38 Geological Society of America, 10, 22 Geological Survey, U.S., 103, 107, 266 Geological Survey of England, 103 geomagnetic field master curve, 153 geomorphic: change and geologic processes, 90; features, 90, 104, 117, 130; stability, 109 geomorphology, 2, 4, 10, 16, 22, 90, 96, 102–4, 108–9, 262, 269–70, 275 geophysical prospecting, 111–13, 122, 195, 270; electrical resistivity (ER), 115; electromagnetic conductivity, 116; ground-penetrating radar (GPR), 116; magnetometry, 113 geotechnology, ancient, 244 Giddings corers, 125 glacial: advance, 87–88, 162, 167; clay, 229; deposits, 87–88, 164; ice, 171; loess, 159; maximum, 90, 159; melting, 162; moraines, 162, 166; outwash, 162; periglacial conditions, 88, 99; proglacial lakes, 87; system, 86; till, 86–88, 90, 188, 205–6, 211 glaciers, 22, 28, 32, 50–51, 63, 86–88, 90, 93, 166–67, 256 glaciofluvial deposits, 90 glass, 45, 59, 144, 153, 160, 155–56, 200, 202, 209, 217, 238 glaze, 202, 213 gley soils, 47 gleyed horizons, 99 gleying, 44, 47 gleysol, 44 global warming and cooling events, 140 gneiss, 209, 214

332

goethite, 46, 207, 217 gold, 106, 202–3, 205, 207, 218–19, 223, 232, 234–35, 272 gophers, 100 gossan, 207 grain size, 48, 50, 85, 119, 128, 195, 228, 240, 249, 258 graminae, 174 granite, 195–96, 198, 208–9, 212, 214, 227, 230, 235, 238, 251, 271 granular disintegration, 85 graphite, 218 grass, 41–42, 78, 173–74, 249 gravel, 32, 52, 75–76, 99, 126, 163, 248–49, 259, 269 graves, 114, 117, 122, 204 gravity, 28, 66, 99, 205, 221, 246, 249, 255 Greek Archaeological Service, 142 greywacke, 208 grog, 213 groundmass, 230 ground-penetrating radar (GPR), 113, 116 gypsisols, 44 gypsum, 26, 40, 45, 78, 81, 157, 195–56, 199, 201–2, 214–17, 252–53 Hackberry Creek Paleosol, 141 hafnium, 229 halite, 36, 78, 201–2, 219, 252–53 Hallein mine, 201 Hallmundarhraun lava flow (Iceland), 85 halloysite, 212–13 Hallstatt mine, 201 harbors, ancient, 98, 117, 127 hardpans, 44 hearths, 46, 73, 115–16, 118, 159, 201 hematite, 36, 46, 114, 196, 205–7, 211, 217, 219, 242 histograms, 50 histosols, 41–42 Hohokam (North America), 154, 177–78, 247 Holocene, 22, 27, 48, 59, 73–78, 84, 90–91, 94–95, 107–9, 117–18, 120, 125–28, 130, 134–35, 138, 141–44, 160, 163, 166, 168, 171–72, 174, 176, 178–80, 182, 184, 187–94, 260–63; climate change, 189, 191;

coastlines, 117; deposits, 74; loess, 140 hominids, 2, 75, 81, 85, 129, 133–34, 156, 165, 181, 185, 264 Homo erectus, 85, 156 Hopewell (culture), 203 horizons. See under soil horses, 143 Hudson Bay Lowland chert (HBL), 227 humates, 141 humic: acid radicals, 157; material, 46, 219 humification, 59, 98 humus, 41–42, 139, 152, 245 huntite, 217 hydration, 44, 54, 160–63, 194, 199, 209, 251, 253, 271 hydraulic cement, 214 hydrocarbons, 124 hydrogen bond, 255 hydrogeologic conditions, 107 hydrology, 3, 103, 105, 129, 247–48, 269 hydrolysis, 257 hydrosphere, 104, 122, 192, 195 ice, 27, 45, 50, 86–89, 93, 99, 143, 165–67, 171, 185, 198, 251, 155, 265 Ice Age, 4–6, 9, 12, 32, 84, 166–67, 174, 189 ice core, 143, 185–86, 186 illite, 213, 251 illuvial, 40, 44–45 illuviation, 59, 98 inceptisols, 41, 109 indigo, 218 indium, 235 inductively coupled plasma spectrometry (ICP), 236 infrared light, 111, 159 infrared spectroscopy, 230 infrared stimulated luminescence (IRSL), 159–60 Ingram-Wentworth size scale, 29 insects, 100, 168, 176, 179–80 instrumental neutron activation analysis (INAA), 158, 225–26, 228, 230–31, 236–37 interglacial and glacial episodes, 189 International Society of Soil Science, 45

Index interstadials, 144, 159 Inuit, 207, 210 Inupiat, 94 invertebrates, 79, 168, 176, 178 iron, 36, 40, 44–48, 88, 112, 114, 124, 134, 151, 153, 166, 178, 183, 196, 201–2, 204–7, 214, 221, 232, 265– 66, 272; ancient, 272; cast, 272; meteoric, 207; ore mines, 207; in sedimentary rocks, 36; wrought, 272 iron oxide, 28, 35–36, 42, 45, 47, 54, 114, 119–20, 187, 206–8, 216–17, 219, 232, 240, 252, 272 iron phosphate, 36–37, 122 irrigation, 11, 74, 107, 174, 177, 178, 186, 194, 245–47 isoleucine (amino acid), 160 isotope analysis, 124, 235, 237 isotopes: carbon, 147–48, 151, 184; fingerprints, 228; fractionated, 124, 184; lead, 146, 224, 234–35, 237–38, 252; oxygen, 76, 133, 139, 147–48, 158, 160, 184–85, 228; provenance, 228; strontium, 225, 228, 252; sulfur, 124 isotopic: fractionation, 148, 151; ratio variation in tree rings, 176; signals in marine sediments, 185; stable value of nitrogen, 186 ivory, 95 jade, 196, 198–99, 223 jasper, 197, 219 jet, 200, 219 jointing patterns, 227 joints, 26, 82, 85, 218, 250, 260, 268 Jurassic, 215 K-means cluster analysis, 242 kaolinite, 192, 212–13, 215, 217 karst: settings, 59, 83–84; soils, 245; solution features in caves, 84; topography, 110, 129 kilns, 114, 215–16 kurkar, 49 kurtosis, 51 L-amino acids (left-handed), 160 lag: deposits, 58, 65, 205; gravels, 65 lagoons, 98, 118 lake(s), 27–28, 33–34, 47, 50–51, 58,

63, 65–66, 69, 77–80, 87–88, 98, 110, 117, 129, 136, 138–39, 146, 151, 168–69, 171, 175–76, 179, 182–83, 187–88, 193, 202, 247, 255, 259, 261, 266–69; artificial, 260; basins, 25, 54, 57, 74, 78–80; beaches, 187; bottoms, 34; clastic deposition in, 79; closed, 169; cycles, 57; deposits, 34, 76, 78, 146, 159, 168, 180; desert, 66; freshwater, 79, 150; glacial meltwater, 90; ice margin, 87; level changes, 168–69; oxbow, 69, 77–78; proglacial, 98; saline, 78; sediments, 15, 51, 78– 79, 87, 114, 122, 147, 165, 168, 175, 178, 189, 193, 202, 260; shoreline features, 80, 91; wave energy in, 79 lamellae, 53 laminae, 53, 135–36, 138 laminations, 54, 136, 138–39 land transportation networks, 248. See also roads land use, 36–37, 105, 107, 114, 121, 129–30, 159, 178, 221, 238, 247, 249 landscape(s): change, 95, 104, 107, 109, 114, 165, 167, 175; as cultural artifact, 105; features, 96, 104, 109–10, 116, 120, 267; glacial and postglacial, 110; initial, and original occupation, 63; paleolandforms, 95; stability, 34, 40, 125–26, 165, 188 landslides, 99, 169, 258–61, 265 lanthanum, 225, 232 lapis lazuli, 196, 198–99 Laurentide Ice Sheet (North America), 166 lava, 85, 164, 189, 225 lava tubes, 85 leaching, 39, 41, 47, 59, 98, 129, 160, 194, 245 lead, 122, 124, 146, 203, 232, 252, 272; carbonate (see cerrusite), 205, 234, 272; deposits, 205; glaze, 213, 235; isotopes, 146, 224, 234–35, 237–38, 252; litharge, 205; mining, 196, 203, 205; pigment, 205; as a preservative in Roman times, 205; provenancing, 235; salts, 213, 272; smelting, 196, 203, 205–6, 234–

35, 252; sulfate (anglesite), 205; sulfide (galena), 205, 234; use, 205 lenses, 84–85, 176, 197, 239 Leonard Paleosol, 140 levee, 260, 262–63 lime, 25, 34–35, 114, 206, 214–16, 245, 248 limestone, 6, 36, 38, 67, 81–86, 110, 114, 139, 146, 150, 152, 195, 197, 200, 208–9, 213–16, 219, 221, 228, 247–48, 252–54, 266, 271, 273 limonite, 36, 207, 219 liquefaction, 54, 245, 260–61 litharge, 205 lithic, 33, 80, 100, 163, 167, 197, 237; debitage (flaking debris), 227; defined, 195; flakes, 33; material, 32, 196–97, 219, 223, 227, 230, 238, 240; raw materials, 185, 218, 223; resources, 98, 195, 213; tools, 209 lithification, 28, 47, 53, 195 lithofacies sequences, 74 lithology, 3, 108, 126, 195–56, 215, 266 lithosphere, 61, 189, 192, 195 Little Ice Age, 166–67, 189 loam, 31, 46 lode(s), 205, 219, 231–32, 235 lodestone, 207 loess, 34, 43, 142, 157, 159, 178, 240; Bignell, 140; dating, 140, 157, 159; deposits, 27, 51, 139, 159, 187; Holocene, 140; paleosols in, 136, 139–40, 185; Peoria, 140; plateau, 185, 263; sequences, 139–40, 158–59, 187 ‘‘Lucy’’ hominid remains, 145 macrofossil(s), 37, 124, 168–69, 174–75, 182–83 macronutrients (plant), 122, 124 mafic: igneous rocks, 36, 210, 231, 251; intrusives, 231; lava, 231; volcanic rocks, 205 Magdalenian, 178 maghemite, 46, 114, 118 magma, 207, 209 magnesite, 217 magnesium, 122, 177, 202, 210, 212–14, 217, 221, 225, 251 magnetic: analysis of soil, 111, 118; anomalies, 114–15; compass, 155;

333

Index magnetic (continued) declination, 155; field of the earth, 114, 145, 148, 153; poles, 153; properties, 112–14, 118, 225, 238; reversal chronology, 139; shield intensity, 148; susceptibility, 112– 14, 118, 128, 238; variation (time transgressive), 154–55 magnetite, 46, 114, 207, 214, 272 magnetometry, 113, 115 maize (Zea mays), 172, 174, 186 malachite, 196, 200, 204–5, 217, 219, 232 malacology, 178 mammal fossil remains, 180–81 mammoth, 6, 62, 87, 125, 135, 140, 180–81, 207 manganese, 35–36, 47, 99, 122, 124, 157, 206, 225, 229 manganese oxide, 36, 46–47, 75, 196, 206, 216–18 maps and mapping: ancient, 110; base, 104, 119; classification in, 103; geologic, 98, 103, 106–7, 260, 262; intensity, 261–62; isopach, 107; landslide susceptibility, 107; Quaternary maps, 103; scale, 107; soil, 103, 105, 107; structure contour, 107; surficial geology maps, 103, 106; terrain, 105; thematic, 104; topographic, 103–4, 109–10, 119, 131 marble, 195, 196, 209–10, 213, 223, 228, 238, 250, 252, 271; of Aegean area, 228; of Carrara (Italy), 228, 252; corrosion, 221; sourcing, 228 marine: depositional settings, 90– 91; geophysical techniques, 117; microfauna, 153; regression, 94, 108, 175; sediments, 157, 168, 175; sites, 272; transgressions, 94 marker horizons, 27 marl(s), 34, 36, 42, 176 marlstones, 34, 36 marshes, 47, 57, 74, 78, 91, 104, 246, 262 mass wasting, 66, 98–99, 249, 265 mastic, 219 mastodon, 180 matrix, 4, 19, 26, 32–34, 44, 47, 51, 65, 67, 82, 90, 102, 114–17, 151, 199, 208, 236–38, 251, 253, 255, 269

334

matte, 188, 223 Maya, 84, 121, 129, 190, 197, 201 Maya blue, 218 meander: belts, 75; channels, 78; scar pattern, 104 mechanical disintegration, 26, 251, 271 meerschaum, 218 megaliths, 222 mercury, 37, 122, 124, 202, 205 mesas, 81, 104, 108 Mesolithic, 37, 110, 118, 134, 178, 188, 193 metal, 10–11, 16, 33, 106, 114, 124, 195, 202–3, 206–7, 218, 223, 230, 232, 234, 272; artifacts, 29, 202, 221, 235; Bronze Age, 16; concentrations, 116, 122, 202; corrosion, 272; deposits, 106; detector, 114; ores, 202, 244; probes, 115; scrap, 224; trace amounts, 124. See also specific metals metalliferous zones, 233 metallurgy, 204–5 metamorphic rocks, 144, 209 (table), 212, 219, 226 meteorites, 207 mica, 144, 153, 200–201, 219, 240 micrite, 67 micro-artifacts, 33, 50, 59 microclimates and microclimatology, 191–92, 247 microcline, 199 microcrystalline quartz, 197, 227 microdebitage, 33 microdepositional conditions, 68 microfauna, 33; in loess deposits, 141; marine, 153; remains, 91 microfaunal diagrams, 179 microfloral remains, 91 microfossil(s), 169, 222, 228–29, 240 micronutrients (plant), 124 micro-organisms, 255, 269 microscope, 59, 153, 174, 197, 226–27, 238–39, 239 midden deposits, 12, 162; pack rat, 168, 179, 183; shell, 93, 185 Middle Ages, 167, 199 Middle Minoan, 211 migration(s), 2, 44, 47, 50, 54, 63, 74, 76–77, 90, 94, 149, 156, 160, 166, 181, 189, 193, 236, 270

millet, 174 minerals, 26, 35–36, 41, 44, 46–47, 54, 86, 114, 118–19, 128, 144–45, 153, 156, 159, 199–200, 202, 204–7, 210, 212–15, 217–18, 221–22, 224, 229–30, 232, 234, 238–39, 251, 258, 272–73; carbonate and sulfate, 251; defined, 195–56; evaporate, 36; in igneous rocks, 144, 153, 156; isometric, 238 mines and mining, 7, 11, 104, 197–99, 201, 204–5, 207, 218–19, 223–24, 233, 245, 248–49, 252, 256, 265; ancient, 8, 218, 224; coal, 248; flint, 115, 218; King Solomon’s, 232; lead, 203, 205; open pit, 231, 248; ore, 196; placer, 219; salt, 201–2; spoil heaps, 194; tin, 202, 234; tunnels, 116; underground, 209–10, 256; waste dumps, 218 Minnesota Messenia Expedition, 22 Minoan, 123, 142, 201, 211 Mississippian (North America) mixing, 142, 191, 266 Modified Mercalli Scale, 258, 259 Mohs scale of hardness, 196, 197, 207, 210 moisture: content of soil, 44; effects of, 41, 44, 46, 99, 114–16, 120 mollisols, 41–42, 100, 109, 142 mollusks, 146, 153, 160, 168, 178–79 molybdenum, 124 montmorillonite, 41, 212, 215, 258, 266 monuments: deterioration, 221, 251, 253; preservation, 268, 271 moonstone, 199 moraines, 87, 162 mortar: clay, 215; gypsum, 214– 15; lime, 215–16, 245, 252; mud, 215–16, 258 mortuary sites, 206, 219 mosques (Turkish), 154 mosses, 168 mottling, 47, 75, 98–99, 142 mound builders (Amazon River), 113 Mound Builders (Hopewell), 203 Mousterian (Middle Paleolithic), 76, 80, 82, 89, 146–47, 158, 210, 213 muck, 247 mud, 25, 28, 34–35, 47, 68, 74, 78, 188, 213–16, 247, 258–59; cracks,

Index 54; deep water, 91; indicators of climate change, 34; particle size, 29, 34; playa, 74; used by humans, 34–35 mudbrick, 49 mudflows, 99 mudslides, 28 mudstone, 213 mummification (Egyptian), 202 Munsell color, 45–46, 119, 128 murex shells, 211 muscovite, 200, 251 musk ox, 181

ocher, 206–7, 217 Old Kingdom (Egypt), 186, 198, 209 Olmec canal irrigation, 246–47 Olmec monuments (Mexico), 17 oncolites, 79 onyx, 198 opal, 198, 173, 193 optical microscopy, 161 optical-emission spectroscopy, 236 optical stimulated luminescence (OSL), 159–60 orbital imaging radar, 236 ore(s), 196, 200, 202–7, 218, 223–24, 231, 234–33 native copper, 7, 204–5, 223–24, orthoclase, 146, 196, 199 229, 231–32; boulder (Michigan), orthopyroxene, 230 231; chemical characterization of, osmotic pressure, 255 224; geologic environments of, ostracod(s), 79, 168, 176–78, 228 205, 224, 231; North American otoliths, 183 deposits of, 204, 231; sourcing, outcrop(s), 105, 107–8, 163, 201, 204, 231; trace-element fingerprinting, 210, 213–14, 218, 222, 224, 227, 231–32 230–31, 234, 237, 266 natric soil horizon, 40, 54 outwash (glacial), 87, 90, 162 natron, 202 overbank: deposition, 76–78, 126; Neanderthal burial, 172 mud, 74 Neolithic, 7, 10, 38, 73, 110, 118, 125, oxbow lakes, 69, 77 128–29, 134, 159, 162, 171, 178, 187, oxidation, 37, 42, 44, 46–47, 171, 231, 192–93, 195, 199–202, 207, 209, 245, 251, 255; mottling, 47, 98; 212, 217–18, 222, 225, 229, 231, 237 processes in groundwater, 44, 255, nephrite, 199. See also jade 266 New Kingdom (Egyptian), 200, 266 oxide ores, 204 nickel, 207 oxisols, 44, 45 nitrates, 251 oxygen, 46, 76, 99, 122, 133, 139, 147– nitrogen, 122, 147, 184, 186, 252–53, 48, 158, 160, 176, 184–85, 228, 238, 269 254–55, 272 noncalcareous: precipitates, 36; sand, oxyhydroxides, 272 230; soils, 46 novaculite, 197 paleoclimates, 165 Nubian sandstone, 210 paleoecologic indicators, 78, 165, 168, nucleated settlement patterns, 110 172, 175 nuts, 175 paleoecology, 18, 270, 275 paleoenvironmental reconstruction, O soil horizon, 40 165–94 oak pollen, 172 paleogeography, 4, 95 obelisks, 209, 227, 249, 253 paleogeomorphic reconstruction, 90, obsidian, 16, 33, 145–46, 153, 162, 249 208–10, 219, 223, 225–26, 229, paleoherpetology, 182 238; cobbles and boulders, 162; Paleo-Indian (North American), 5, deposits, 210, 225–26; flows, 225; 14–16, 75, 98, 142, 172, 175, 180, hydration dating, 160–62; mir194, 206, 227 rors, 210; provenancing, 224–25; Paleolithic, 5, 7, 11, 16, 18, 32, 34, 37, trace-element database, 226 54, 66, 73, 80–82, 84–85, 90, 95,

100, 115–16, 118, 126, 131, 134, 144, 146, 149, 157–58, 176, 178, 180, 183, 185, 189, 193–95, 197, 207, 209– 10, 212–13, 217; American, 7–9; artifacts, 9, 26, 32, 34, 48–50, 52, 67, 73, 76, 80, 83, 85, 89–90, 143, 149, 155–56, 159, 176, 183, 187; archaeology, 17, 145, 148; deposits, 7, 52, 73 paleomagnetism, 135 paleontology, 2–4, 10–11, 22–24, 59, 132, 143, 167 paleoseismic, 261, 268 paleosols, 43, 43–45, 112, 125, 128, 139–42, 174, 193; in archaeological contexts, 43; as time indicators, 43, 136, 159, 185 palygorskite, 218 palynology, 169, 172 particle morphology, 51–52, 52 paste, 211, 239, 242 patina, 163, 272 patination. See desert varnish (cation-ratio) patterned ground, 99, 100, 102 peat, 37, 45, 54, 118, 151, 183; bogs, 104, 143, 183; deposits, 37, 78, 183; histosols, 42; reed, 188 pebbles, 29, 77, 198–99, 208 pediments, 68 pedogenesis, 41, 53, 151 pedogenic, 34, 36, 38, 40–41, 43– 46, 50, 54, 64, 78, 99, 167, 191–93; alteration, 142; carbonate enrichment, 43–44; horizons, 34, 40–41, 44, 75, 99; zones, 34 peds, 54 pelecypods, 79 perched, 99 periglacial, 86, 88, 99 permeability, 99, 107, 113, 221, 250, 252, 256, 266, 271 petrocalcic zones, 40 petroglyphs, 163 petrography, 2, 4, 16, 226–29, 238– 40, 242 petroleum, 219 petrology, 2, 239 pewter, 205, 232, 272 pH, 46, 107, 122, 124, 176, 251–52, 255, 266, 269 phenocrysts, 208

335

Index Phoenicians, 233 phosphates, 36–37, 122, 123, 192 phosphorus, 25, 122–23, 206 photography, aerial, 11, 103–4, 109, 118–19 photosynthesis, 79 photosynthetic pathways of carbon, 184 phylogenetic change, 176 phytoliths, 43, 69, 168, 172–74, 173 phytoplankton, 219 pigments, 205, 207, 217, 219 pigs, 143 Piltdown Man, 16, 163 pine (white), 139 pipestone, 210 pisé, 215 pitchstone, 219 pits, 8, 11, 90, 105, 114–16, 120, 126, 197, 201, 204, 216, 261 placer deposits, 203, 205, 219, 232–33 plagioclase, 199, 251 Plains Woodland sites, 141 plant(s), 11, 25, 28–29, 36–37, 39, 42, 44, 54, 57, 64, 78–79, 87, 100–102, 104, 120, 122, 124–25, 128–29, 143, 151, 166–67, 169, 171–73, 175–76, 182, 184, 191, 193–94, 201, 218, 221, 237, 251, 273; communities, 94, 171, 183, 194; material, 215; nutrients, 122, 212; remains, 100, 167–69, 175–76, 183, 186 plaster, 201, 213–16, 252, 259; of Paris, 201; on surfaces, 245 plasticity, 212–13 plate tectonic theory, 188 platinum, 235 playa, 14, 66, 74, 80, 129, 187 Pleistocene, 2, 10, 12, 14–15, 16–18, 21, 27–28, 43, 54, 59, 73–77, 84, 86–87, 90, 93–94, 118, 120, 125–26, 133–35, 140, 143, 145–46, 148, 153, 156, 160, 165, 172, 178, 180, 187, 191, 193–94, 265; extinction of North American large mammals, 194; transition to Holocene, 172, 174, 182, 191 Pliocene, 133, 145, 156, 181, 186, 264 Plio-Pleistocene, 143, 265 plowing, 40, 42, 54, 62, 119, 194 pluvial, 187 Poaceae (graminae), 174

336

pocket gophers, 100 podzols, 42 point bar, 69, 73–74, 260 polar bear, 181 polarity reversals, 145 polarizing microscope, 59, 238, 239 polje, 110 pollen, 10–11, 16, 37, 69, 128, 139, 143–44, 165–66, 168–73, 175–76, 182–83, 188, 192–93; stratigraphy, 57, 136, 138–39, 143, 171–72, 182 pollution, 91, 124, 168, 210, 221, 252, 255 ponds, 34, 66, 78, 80, 157, 176, 179, 182, 247, 259 porcelain, 206 porosity, 99, 113, 115, 221, 250, 252– 53, 260, 271 porphyry (red), 227 postdepositional processes, 25–26, 38, 54, 59, 98–100 postglacial landscapes, 87 potassium, 122, 144–45, 163, 202, 212–13, 251 pottery, 11–12, 16, 26, 29, 34–35, 51– 52, 64, 95, 115, 129, 131, 134, 142, 150, 157, 175, 204, 207, 211–13, 215, 217, 219, 223, 229, 230, 239–40, 252; temper, 16, 211, 213, 223, 229– 30; thin sections, 59, 228, 239, 242 precipitates: calcareous, 36; chemical, 78, 81, 138; noncalcareous, 36 precipitation, 35–36, 64, 66, 80, 86, 139, 164, 166, 169, 171–72, 175, 184– 85, 189, 194, 207, 216, 249, 251, 255 Predynastic Period (Egypt), 74, 195, 197–98, 209 proglacial, 86–87, 98 prograding sequence, 34 Protoclassic Period (Maya), 83 protons, 114 protosols, 44 provenance, 3, 4, 22, 131, 209, 222– 43; analytic techniques, 225; defined, 222 provenience, 47, 222 Pueblo III (New Mexico), 268 pumice, 45, 143, 145–46, 210, 214, 218, 263–64

pyramids, 34, 218, 244, 248, 253 pyrite, 200, 207, 219, 231 pyroclastic deposit, 225 pyroxene, 199, 230 pyrrhotite, 272 quarries, 8, 104, 107, 197, 206, 210, 214, 219, 224–25, 228, 230, 237– 38, 245, 248–49; ancient, 199, 219, 249; sites, 63, 162, 219 quarrying, 214, 218–19, 224, 227–28, 248–49, 271 quartz, 33, 36, 47, 106, 157, 195–98, 203, 208–10, 212, 214, 218–19, 227, 229–30, 239–40, 251, 253, 265, 271; microcrystalline, 197, 227 quartzite, 32–33, 196, 208–10, 214, 219, 223, 225, 240, 250, 261 quartzose, 218–19, 222 Quaternary, 2–4, 9–10, 14, 17, 19, 22, 83, 93, 126, 130, 132, 138, 145, 157, 160, 165, 167, 176, 179–80, 183, 187, 189, 191, 256, 264; climates, 165; deposits, 27, 131, 137, 165, 178–79; fossil record, 182; maps, 103 Queleccaya ice cap (South America), 185 querns, 209 racemization, 160 radar, 111, 113, 116, 120, 121 radioactivity, 157 rammed earth, 215 rare earth, 230, 231 raster, 130 ravine, 248 reducing atmosphere, 204 reduction (chemical), 98, 251, 255, 265 reed peat, 188 reef limestone, 253 refractory, 212 regression (marine), 90, 91, 94, 108, 175 reindeer, 181 reservoir effects, 149–51 reservoirs, 138, 148–50, 153, 245, 246, 259, 266–69 resin, 200 revetment (riprap), 271 rhinoceros, 6, 181 rhyolite, 208–10, 219

Index rhythmites, 136. See also varve(s) Richter magnitude, 258, 259, 261 ridge, 92, 93, 107, 108, 191, 260 rift system (East Africa), 169, 201, 265 rifting, 188, 189 riprap, 247, 250, 265, 269, 271 river(s), 8, 25, 33, 41, 43, 50–52, 68– 82, 85, 86, 90, 95, 96, 98, 104, 107–10, 113, 120–29, 140, 149, 156, 159, 160, 176, 185, 189, 197, 198, 201–5, 212, 215, 218, 221, 227, 230, 235, 245–48, 256, 259–68, 270; beds, 60, 68, 225, 263; channels, 50, 69, 77, 116–18; drainage, 78, 96, 108; lateral migration, 108; meanders, 107, 108, 110, 250, 256, 265 roads, 104, 110, 119, 194, 245–48, 252 rock(s), 16, 17, 24–26, 28, 32–34, 36, 38–50, 63, 81–87, 90, 91, 100, 103, 106–8, 110, 114–16, 124, 135– 36, 138, 144–45, 150–53, 157, 159, 162–64, 178, 191, 195–231, 234– 40, 242, 244–50, 256–60, 265–73; classification, 29, 31, 195, 207–9, 227; outcrop, 163; weathering, 26, 36, 39, 249–56, 271 rock crystal, 198 rock salt, 201 rock shelters, 26, 33, 59, 63, 81–86, 83, 108, 178, 221 rockfall, 32, 81, 82, 85, 99, 261, 268 root action, 28, 85, 101 routes, 63, 64, 90, 95, 98, 104, 109, 181, 223 rubidium, 225 runoff, 66, 77, 129, 151, 189, 194, 245, 247, 254 rust (iron oxide), 207, 219, 272

60, 265, 269; beach, 35, 51; dune, 51, 66, 104, 249; particle size, 30, 33, 49, 51, 95, 214–15; quartz, 42, 47, 214, 265, 271 sandblows, 260, 261 sandstone, 34, 36, 54, 85, 145, 208– 10, 214, 216, 219, 230, 242, 253, 268, 271; deterioration, 253, 271; quarries, 214 sanukite, 230 sapropel, 57 sard, 198 sardonyx, 198 sarsens, 222 satellite imagery, 103, 111, 120, 130, 274 scandium, 225, 228, 229 scarp, 68, 82, 104, 118, 268 schist and schistose, 195, 209, 240 scoria, 209 sea level, 74, 85, 90–94, 94, 99, 117, 118, 135, 175, 185, 201, 257, 263, 265 seasonal, 41, 64, 66, 75, 79–81, 139, 166, 171, 172, 181, 182, 188, 194, 249, 253, 255 seasonality, 183 sediment(s) and sedimentary, 1–6, 9, 11, 13, 15–16, 25–59, 30, 31, 35, 55, 56, 57, 61, 63–104, 68, 71, 79, 82, 84, 88, 89, 93, 106, 108–18, 122– 28, 134–44, 147, 153–71, 174–76, 178, 183, 185–96, 201–3, 207–16, 219, 228–31, 234, 238, 240, 244– 50, 255–71; archaeological, 4, 10, 29, 49, 54, 83, 115, 118 sedimentology, 2, 10, 14–15, 17–19, 22, 23, 31–34, 48, 53, 57, 60, 68–69, 75, 77, 81, 124–25, 167–69, 193 seeds, 33, 37, 100, 175–76, 183 seiches, 260 seismic, 103, 117–18, 257–62, 264, salinity, 91, 169, 175–78, 221, 266 268, 273 salt(s), 36, 40, 57, 78, 99, 193, 195, selenite, 201, 217 200–202, 204, 213, 251–53, 265, selenium, 124 269, 272 sepiolite (meerschaum), 218 saltpan, 202 serpentine, 47, 195, 210 sand and sandy, 25–35, 38, 41–42, settlement patterns, 103–5, 109, 110 46–57, 65–67, 69, 73–82, 85, 87–89, shale, 208–10, 212, 213, 216, 219, 258, 91, 95, 100, 104, 107, 110, 115–16, 266 118, 120, 125, 138, 142, 146, 156–57, sheetwash, 85, 142 162, 177, 183, 202–3, 208, 213–19, shell(s), 7, 28, 74, 95, 100, 141, 146, 229, 230, 242, 248–49, 253, 256– 150–53, 157, 160, 176–78, 185, 208,

210–11, 216, 260; middens, 93, 99, 185, 227; temper, 229–30 Shoshone (Native Americans), 193 shotcrete, 268 silica, 28, 35–36, 42, 45, 55, 86, 172– 74, 198, 203, 206–10, 213, 216–17, 219, 225, 227 silicone, 268 silt, 29, 31, 34–38, 41, 43, 46, 50, 55– 56, 69, 73–77, 83, 87, 98, 107, 115, 120, 138, 140, 145, 177–78, 180, 208, 214–15, 219, 245–46, 260, 263, 266–67, 269 siltstone, 219 silver, 122, 196, 202–5, 232, 234–35, 238, 256, 272 sinkholes, 256 site(s), 4–6, 9–25, 28, 32–35, 37, 40, 42, 45, 48–50, 57, 59, 104, 105, 107, 109–30, 132, 140–44, 146, 154–63, 167, 170, 172, 174–85, 188, 191, 193, 105, 197–209, 213, 216, 218–19, 222–27, 226, 230–42, 245– 52, 256–64, 274; abandonment, 266; boundaries, 124; catchment, 274; formation, 3–5, 60–102, 128; habitation, 40, 47, 231, 268; historic, 22, 270–71; preservation, 28, 265–72, 274; stratigraphy, 3 skewness, 50–51 slag, 33, 204, 206, 209, 218, 223 slate, 208, 210 slip, 212, 217 slope, 40, 42, 63, 66, 76, 82, 84–85, 94, 99, 104–5, 108–10, 142, 162, 188, 214, 244–45, 248–49, 258–62, 266, 268, 269 slopewash, 76, 85 slumping, 28 smectite, 192, 212, 213 smelting, 33, 123–24, 196, 202–6, 218, 223–24, 232, 234–35, 252 snails, 151, 178 soapstone (steatite), 210, 230–31 sodium, 36, 40, 177, 201–2, 213, 221, 251 Soho phase (Hohokam Classic), 154 soil, 2, 3, 11, 17, 24, 25–59, 60, 65, 72, 82, 88, 99–102, 105, 109–26, 130, 131, 139–42, 151, 159, 165–70, 171, 175, 183, 184, 186–95, 202, 207, 211–16, 219, 221, 237–40, 244–51,

337

Index soil (continued) 255–61, 265, 269, 270, 272; color, 45–46; definition, 39; formation, 25–28, 33, 34, 37–47, 51, 55, 99, 102, 112, 114, 119, 122–26, 131, 139, 142, 159, 187, 189, 193, 195, 213, 221, 247, 275; horizons, 34, 38–43, 47, 102, 124, 142; maps, 103, 105, 107; micromorphology, 59; profile, 39, 39–44, 54, 123; texture, 31, 33, 46, 48, 54, 59 Soil Conservation Service, 107 Soil Survey (U.S.), 45 solifluction, 84, 99 soluble salts, 99, 213, 252, 253, 269 sorting, 33, 34, 47, 50, 51, 66, 80, 98, 100, 102, 227, 239, 249 sourcing. See provenance spalling, 161, 62, 214, 251 speleothems, 146 spits, 90 spodosols, 41, 42, 44 spoil heaps, 194 springs, 36, 80, 104, 107, 129, 175, 201, 202, 219, 220, 221, 245, 247, 256, 261 spruce, 139, 170, 172, 182 stabilization, 43, 54, 65, 66, 102, 139, 265, 268, 269 stable isotopes, 124, 184, 236, 238 stalactites and stalagmites, 81 steatite, 195, 210, 230, 231 steel, 117, 196, 206, 272 steppe, 41, 171, 181 stone, 11, 24, 29, 32, 47, 63–64, 73, 98, 100, 195, 204, 209–11, 219, 222, 228, 230, 268–69; artifacts and tools, 5, 6, 8, 26, 28–29, 32–33, 50–53, 61, 76, 80, 85, 131, 134–35, 149, 156–63, 195, 197, 200, 208, 237, 242, 244–47; building, 106, 116, 195–96, 201, 210, 213–15, 218–19, 222–27, 235, 244, 248, 250–53, 256, 258, 271; deterioration, 251, 253, 271; semi-precious, 197–200 storage, 114 strandlines, 69, 98 stratigraphy and stratigraphic, 2– 19, 13, 14, 23–25, 28, 31, 41, 45, 51, 57, 70, 73–78, 81–85, 88, 88–91, 98–100, 104, 112, 116–17, 124–25, 128, 130, 135–38, 141–46, 156, 159,

338

162, 168–72, 176, 182, 183, 186, 190, 197, 227, 247, 257, 260, 266, 268; markers, 7, 141, 146; principles, 7, 11, 41, 135–36; profiles, 36, 124; sequences, 34, 38, 40, 43, 48, 57–59, 69, 81–87, 91, 93, 99–100, 139–44, 172, 187, 263–64; units, 38, 40, 48, 107, 135 streams, 27, 28, 32, 50, 52, 62, 65–78, 69, 82, 85–87, 104, 108, 121, 129, 142, 157, 169, 171, 175–78, 182–83, 187, 204, 213, 218, 229, 247–49, 260, 269 strontium, 177, 185, 228, 237–38, 252 subsidence, 99, 249, 256–57, 260, 262, 265 succinic acid, 230 Sui Dynasty, 246 sulfur, 122, 124, 142, 167, 206, 232, 252–53, 271–72 sulfuric acid, 231, 252–53, 271–72 sunstone, 199 superposition, 12, 41, 136, 162 surveys: archaeological, 8, 18, 22, 75, 105, 109–10, 119–20, 124, 126–28, 130–31, 194, 218, 222, 246, 270, 274; geologic, 9, 105, 225, 270; geophysical, 11, 105, 112–17, 121, 123, 270 swamps, 42, 45, 47, 69, 73, 78–80, 95, 104, 114, 121, 168, 175, 182, 188, 219, 245, 265 syenite, 208, 227 systemic context, 19, 24, 61–62, 64, 69, 76, 80, 102

tell, 7, 59, 92, 93, 194–95, 213, 232 temper, 16, 211–13, 223, 229, 230, 238, 240, 242, 253 temperature, 42, 46, 79, 116, 118, 120, 160–62, 166, 169–85, 189, 192, 194, 204, 206, 208–9, 212–15, 229, 251–55, 263, 266, 271 tephra, 27, 72, 88, 143, 230, 263 terraces, 70, 72–73, 75, 98, 104, 109, 159, 160, 194, 267; alluvial, 40–42, 69, 75, 126, 159 terrain, 2, 41–42, 75, 103, 108, 111, 115–16, 121, 150, 245, 247, 256; classification, 105; mapping, 105 terrigenous, 36, 96, 247, 251 test excavations, 113 textiles, 183 texture, 29, 31, 39, 46–49, 53, 59, 63, 68–69, 119, 171, 199, 209, 239, 247, 250, 260 thermal, 32, 118, 120, 157, 158, 161, 169, 212–13, 229, 236, 252; annealing, 153; infrared images, 120; maximum (North America), 92 thermoluminescence, 139, 141, 157, 158, 158–59, 225 Third Dynasty (Egypt), 214 thorium, 146, 229, 234 till, 51, 86–90, 188, 205–6, 211 time scale, 66, 130–36, 139, 144–45, 150, 153, 158, 161, 167, 267 tin, 202, 205–6, 224, 229, 232–35, 233 titanium, 163, 190, 206 Tiwanaku civilization (Peru), 123, 185, 190, 202 Tollund man, 37, 183 taiga, 45, 81 topography, 38, 41, 58, 95, 104, 106, talc, 196, 210, 240 110, 112, 129, 153, 248–49 talus, 82, 155 toxic, 124, 205, 213, 272 tannic acid, 37, 219 trace element(s), 43, 124, 131, 168, taphonomic, 60, 62, 169, 175; mixing, 178, 222–38, 242–43 181; processes of deposition, 57, trading patterns, 236 81, 167, 178, 180, 182 trampling, 28, 32, 35, 52, 62, 100 trajectories, 175 transgressions (marine), 94, 128 taphonomy, 3, 57; artifact, 41, 60, 76; transgressive, 34, 40, 58, 80, 91, 118, defined, 60; fossil, 60 128, 135, 169, 175; time, 144, 172 tar, 219 translocation, 28, 54, 99 Tecep phase (Maya), 162 transport, 27–36, 47, 49, 49–53, 58, tectonic, 54, 90, 110, 133, 168–69, 60, 64, 66, 68–69, 73–79, 82–90, 188–89, 192, 256, 258, 260–62, 265 95, 99–100, 102, 104, 110, 148, 151, teeth, 33, 124, 157, 163, 174, 180–81, 161–62, 170–71, 175, 182, 189, 201, 237 212, 214, 218–25, 246, 248–49, 253

trap rock, 208 trash pits, 261 travertine, 36, 81–82, 84, 96, 146, 157, 175, 208, 219, 247 tree rings, 10, 17, 136, 144, 148, 150, 152–53, 176 tree throw, 102 trenches and trenching, 11, 100, 103, 115, 125, 131, 159, 214, 219 troughs, 247 truncation, 58 tsunamis, 131, 190, 257, 261–65 tufa, 36, 80, 82, 146, 151, 171 tuff, 86, 145–46, 214, 225 tumbaga, 272 turbation, 26, 40, 42, 44, 54, 59, 79, 88, 98–102, 101, 114, 138–39, 151, 265–66 turquoise, 198–99, 219, 230 ultisols, 41–42, 45 ultramafic intrusives, 231 unconformity, 58, 74, 260 uniaxial crystals, 238 uniformitarianism, 6, 91 uranium, 146, 148, 150, 153, 157, 160, 163, 166, 185, 234, 255 Valdavia, 95 valley(s): alluvial, 40, 109, 219; buried, 120; forms, 108; river, 70, 73, 76, 78, 82, 95, 108–9, 120, 247; stream, 70, 72, 85, 104 vanadium, 43 varve(s), 10, 135–39, 139, 144, 148, 150, 152–53, 188, 190 Vegas occupation (Ecuador), 95 vegetation, 41–44, 66, 74, 78, 102, 104–5, 110, 119–20, 122, 128, 130,

170, 171, 173, 188–94, 198, 249, 269; boundaries, 189; cover, 66, 249; and climate, 171 ventifacts, 249 vertebrate(s), 79, 100, 163, 168, 176–78, 180 vertisols, 41, 44, 98 vesicular andesite, 209 vibracorers, 125, 127 Vinca culture (Serbia), 218 visible light spectrum, 111, 119 volcanic, 27, 41, 59, 75, 77, 83, 85–86, 114, 120, 125, 128, 131, 136, 142–46, 150–51, 157, 160, 167, 188, 190, 192, 195, 199, 205, 209–10, 212, 214–15, 225, 226, 227, 229–30, 238, 240, 244, 248, 263–64; ash, 27, 41, 75, 77, 83, 125, 131, 136, 142–43, 145, 150, 212, 215, 229, 249, 263–64, 264; destruction, 264; hazards, 265; temper sands, 230 volcanism, 144–45, 188, 244, 263, 264 wadis, 66, 120 walls, 29, 49, 58, 82, 84, 114–16, 120, 131, 195, 209, 212–15, 246, 252–53, 258–62, 268, 271; buried, 115–16, 120; mud brick, 213 Walther’s Law, 58, 91, 93 water, 26–28, 34, 36–37, 41–54, 57– 59, 63–68, 73–95, 98–99, 104–5, 107, 109–10, 114–16, 119–25, 129, 146, 150–51, 153, 157, 160–61, 168– 69, 171, 174–94, 198, 200–203, 206–14, 216, 218–19, 221–23, 231, 237, 245–73; atomic configuration, 255; frost action, 26, 100, 221, 253;

ground, 26, 48, 54, 59, 66–67, 80, 82, 86, 98–99, 121, 150–51, 184, 192, 194, 209, 221, 245, 251–52, 255–56, 261, 265, 270; table, 44, 48, 54, 86, 94, 99, 107, 125, 129, 192, 194, 221, 245, 255–56, 270; resources, 129 waterlogged, 45, 116, 183, 194, 200 watershed, 57, 189 weathering, 16, 25–28, 34–39, 42–47, 58, 60, 64, 82, 84–85, 98, 100, 107, 109, 112, 142, 145, 163–64, 197– 98, 212, 214–21, 248–56, 266, 269, 271–72 wells, 36, 110, 114, 129, 178, 201, 256, 259, 261, 265 wetland, 52, 74, 129, 183, 257 wind, 26–27, 32–34, 43, 48, 49, 58– 59, 64, 65, 67, 91, 109, 120, 140, 153, 167, 169–70, 171, 175, 191–92, 249, 251, 265, 268, 271. See also eolian wood, 37, 112, 117, 122, 141, 144, 151, 160, 172, 176, 183, 188, 200, 213–19, 256, 258, 266 Woodland period (North America), 104, 109, 141–42, 166, 179, 187–88, 206, 212, 240 X-ray analyses, 239 X-ray fluorescence (XRF), 222, 225, 236–37, 242 Yabrudian artifacts, 158 Younger Dryas event, 167, 189–90 zinc, 43, 124, 203, 206, 224, 234 zooarchaeology, 3

339