Skip to main content

Full text of "Memoirs of the California Academy of Sciences"

See other formats

NINA G. Jablonski, Series Editor 


copy 2 



The Pleistocene 


OF THE New World 

Edited by 

Wattis Symposium Series in Anthropology 

Memoirs of the California Academy 
OF Sciences, Number 27 

academy of sciences 

FEB 2 4 2003 


The First Americans 

The Pleistocene Colonization 
OF THE New World 

Volume Editor 
Nina G. Jablonski 

Wattis Symposium Series in Anthropology 

Memoirs of the California Academy of Sciences 

Number 27 

San Francisco, California 
April 8, 2002 


Alan Leviton, Editor 

Katie Martin, Managing Editor 

© 2002 by the California Academy of Sciences 

Golden Gate Park 

San Francisco, CA 94118 

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any 
means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval 
system, without permission in writing from the publisher. 

Library of Congress Control Number: 2002103062 

ISBN: 0-940228-49-1 (cloth) 
ISBN: 0-940228-50-5 (paper) 

Printed in the United States of America by the AUen Press 
Distributed by the University of California Press 

The First Americans 

The Pleistocene Colonization of the New World 


Bruce Bradley 

Primitive Technology Enterprise, Inc. 

P.O. Box 534 

Cortez, CO 81321 

(970) 565-7618 

L.inda Brown 

Department of Anthropology 

University of Montana 

Missoula, MT 59812 

(406) 243-2693 

(406) 243-4918 FAX 

Tom D. Dillehaj 

Department of Anthropology 

University of Kenmcky 

Lexington, KY 40506 

(773) 702-7729 

(773) 702-4503 FAX 


John Douglas 

Department of Anthropology 

University of Montana 

Missoula, MT 59812 

(406) 243-2693 

(406) 243-4918 FAX 

Scott A.. Elias 

Department of Geography 
University of London 
Royal Holloway 
Egham, Surrey TW20 OEX 
(44) 1784-443-647 
(44) 1784-472-836 FAX 
s.elias @rhul. 

Jon M. Erlandson 

Department of Anthropology 

University of Oregon 

Eugene, OR 97403 

(541) 346-5098 

(541) 346-0668 FAX 


Nina G. Jablonski 
Department of Anthropology 
California Academy of Sciences 
Golden Gate Park 
San Francisco, CA 94118 
(415) 750-7161 
(415) 750-7346 FAX 

David]. Melt^er 

Department of Anthropology 

Southern Methodist University 

DaUas, TX 75275 

(214) 768-2826 

(214) 768-2906 FAX 


D. yindreiv Merriwether 

Department of Anthropology 

Department of Ecology & Evolutionary Biology 

The Center for Statistical Genetics 

University of Michigan 

1020 LSA Building 

Ann Arbor, MI 48109 

(734) 647-6777 

(734) 763-5383 FAX 


Johanna Nichols 

Department of Slavic Languages and Literatures 

Mailcode 2979 

University of California at Berkeley 

Berkeley, CA 94720-2979 

(510) 642-1097 

(510) 642-6220 FAX 

Joseph F. Powell 

Department of Anthropology 
University of New Mexico 
Albuquerque, NM 87131 
(505) 277-1538 

v4. C. Koosevelt 

Department of Anthropology, University of Illinois 

Department of Archaeology, Field Museum of Natural History 

Roosevelt and Lakeshore Drive 

Chicago, IL 60653 

(312) 665-7055 

(847) 869-0423 FAX 

Jack Kossen 

Department of Anthropology 

Ithaca College 

Ithaca, NY 14850 

Dennis Stanford 

Department of Anthropology & Archaeology 

National Museum of Natural History 

MRC 112, Smithsonian Institution 

Washington, DC. 20560 

(202) 357-2672 

(202) 357-2208 FAX 

D. Gentry Steele 

Department of Anthropology 

Texas A&M Universit}^ 

CoUege Station, TX 77843 

(979) 845-5297 

(979) 845-4070 FAX 

Christy G. Turner II 

Department of Anthropology 

Arizona State University, Box 87242 

Tempe, AZ 85287 

(480) 965-6213 

(480) 965-7671 FAX 


Preface xiii 

Introduction 1 

Changing Perspectives of the First Americans: Insights Gained 
and Paradigms Lost, by Nina G. Jahlonski 

Setting the Stage: Environmental Conditions in Beringia as People 9 

Entered the New World, by Scott A. Elias 

What Do You Do When No One's Been There Before? Thoughts on 27 

the Exploration and Colonization of New Lands, by David J. Melt^er 

Anatomically Modern Humans, Maritime Voyaging, and the 59 

Pleistocene Colonization of the Americas, by Jon M. Erlandson 

Facing the Past: A view of the North American Human Fossil Record, 93 

by D. Gentry Steele and Joseph F. Powell 

Teeth, Needles, Dogs, and Siberia: Bioarchaeological Evidence for the 1 23 

Colonization of the New World, by Christy G. Turner II 

The Migrations and Adaptations of the First Americans: Clovis and 159 

Pre-Clovis Viewed from South America, by A. C. Koosevelt, 
John Douglas and Unda Brown 

Plant Food and its Implications for the Peopling of the New World: 237 

A view from South America, by Tom D. Dillehay and Jack Rossen 

Ocean Trails and Prairie Paths? Thoughts about Clovis Origins, 255 

by Dennis Stanford and Bruce Bradley 

The First American Languages, hy Johanna Nichols 273 

A Mitochondrial Perspective on the PeopUng of the New World, 295 

by D. Andrew Mernwether 

Index 311 



^^^^^^ San Francisco is privileged to have as one of her citizens Mrs. Phyllis 
^^^^^^^ Wattis, a woman of tremendous warmth, sincerity and generosity. 
'"^^^^^^K^^^m The support that she has given to the city's educational and cultural 
t V ^Pii institutions in the last thirty years has been nothing short of phe- 
nomenal, and being on the receiving end of some of the largesse that she has bestowed 
on the California Academy of Sciences has been wonderful. The passionate and wide- 
ranging interest that Mrs. Wattis has maintained in anthropology over the years has 
inspired her to support several major programming initiatives at the Academy, including 
the Wattis Symposia in Anthropology, which were begun in 1993. These symposia were 
conceived as occasions for the presentation of new information (to the public) on a topic 
of broad interest and importance in anthropology. The scholars participating were to be 
the best in the world. After the great success of the first symposium, it was decided that 
the edited proceedings of that symposium and all subsequent ones should be published, 
so as to bring the spirit and content of each symposium to a much larger audience. I am 
happy to say that we have succeeded in this enterprise, with each symposium and its 
respective volume seeming to be better than the last. The Fourth Wattis Symposium, 
"The First Americans: The Pleistocene Colonization of the New World," was held on 
October 2, 1999 at the California Academy of Sciences, and was attended by a fuU house 
of over 350 people. We knew that the time was ripe for another airing of this long-debat- 
ed topic. The symposium was a great success, as judged by the reactions of the partici- 
pants and the audience, and the positive "chemistry" that was established at the sympo- 
sium has, happily, extended into this volume. I am grateful, first and foremost to Mrs. 
Phyllis Wattis for making this symposium possible through her generous donations to 
the Department of Anthropology of the California Academy of Sciences over the past 
twenty years. I also wish to thank her for hosting the wonderful dinner which preceded 
the symposium. I am also grateful to all the participants, who not only behaved very well 
at the symposium (archaeologists can be a wild bunch!), but also transformed their oral 
presentations into generally very good written chapters. The quality of this volume has 
been assured by a thorough and exhaustive process of anonymous peer review for each 
chapter, undertaken by three outside experts in the specific field relating to each paper. 
I thank those hardy reviewers, many of whom have asked to remain nameless, very 


Finally, I wish to express my gratitude to three people who were critical to the suc- 
cess of the symposium and to the physical production of this book. Nancy Gee attend- 
ed completely to all the logistics of the symposium, and in doing so relieved me of an 
immense responsibility. The success of the symposium itself was made possible by her 
expert efforts in organi2ing transportation and accommodation for the participants, and 
in making all the arrangements for the symposium itself Vivian Young was responsible 
for the design of the book cover and the chapter format, both of which take their inspi- 
ration from her designs for the symposium brochure and program. It was a joy to work 
with such a gifted artist. Rachel Wolf, who has acted as my editorial assistant in the pro- 
duction of this book, was responsible for the immense tasks of corresponding with the 
authors on editorial matters, and formatting and typesetting the book in preparation for 
publication. The beautiful layout of the text and illustrations, and the punctiliously edit- 
ed text owe everything to her unflagging and good-humored efforts. I hope that you find 
this book as rich and stimulating to read as I found it to edit. 

Nina G. Jablonski 
San Francisco 
January 15, 2002 


Chapter One 


Changing Perspectives of the First Americans: 
Insights Gained and Paradigms Lost 

Nina G. Jablonski 


'Far from being a novel one, some aspect of the question of the origin of the 
American race has come to be almost perennial, and it acquires greater interest 
as increase in our knowledge of the subject offers new points of approach. A.t 
what epoch man came to our continent from a former home; how he made his 
way hither; and his history since he came, are questions that possess greater and greater attrac- 
tion as the science of man becomes broader and deeper. If ive accept the theory that man orig- 
inated in the Old World, it is evident that his colonisation of America is a question of mode 
of miration, which resolves itself into a geographical or a geological one. An adequate solu- 
tion to our problem must draw contributions from several sciences — geology, geography, com- 
parative anatomy, and culture history. " (Fewkes 1912:1—2) 

"The program calls first for a presentation of the historical side of the subject, or, strict- 
ly speaking, for a brief history of the opinions that have been held on the question of the 
nature and origin of the American natives since their discovery. This history, it may be said 
at the outset, is largely one of speculation, fettered on one side by ignorance and on the other 
by ancient traditions. " (HrdUcka 1912:5) 

Twenty-first century anthropologists Hke to think of themselves as integrative 
thinkers who are pioneering new, interdisciplinary and multidiscipHnary approaches to 
some of the perennial questions of the discipline, such as the peopling of the Americas. 
Happily, the ninety-year-old passage from Fewkes disabuses us of that arrogant nodon. 
And, the passage of the same vintage from Hrdlicka will also sound fresh, in content, if 
not in style. Anthropologists, especially those trained in the American academic tradition, 
have been integrative scientists throughout most of their history. The only thing that 
really sets the anthropologists of today apart from those of Fewkes's time is that they 


have more to integrate. Information relevant to an understanding of the peopling of the 
New World comes to us today not only from the disciplines named by Fewkes, but from 
many other areas of study. The fundamental questions posed by Fewkes are, however, 
the same as in his day. They have been visited many times since the beginning of the dis- 
cipline, and the literature on the subject is voluminous (see, e.g., Fewkes 1912; Howard, 
1936; Owen, 1984). The contributions to this volume represent a continuation of this 
long tradition. 

As this chapter is written, in the early 21st century, we have tremendous advantages 
over our colleagues of even just 20 years ago. The most important of these is informa- 
tion. We are no longer restricted to examination of information from the "traditional" 
anthropological disciplines of archaeology or osteology. We have the potential — not 
always exercised, mind you — to synthesize information from these fields as well as 
many others, including paleoecology, geochronology, odontology, linguistics and molec- 
ular biology. The chapters in the current volume represent more of an encouraging 
progress report on the peopling of the Americas rather than the last word. 

The Fourth Wattis Symposium on "The First Americans: The Pleistocene 
Colonization of the Americas" was conceived with the notion of bringing together the 
latest evidence that pertains to all the issues surrounding the earliest peopling of the New 
World. Seemingly intractable controversies and entrenched debates continue to charac- 
terize this area of inquiry: HrdUcka's fetters of ignorance and "ancient tradition" have 
evolved in ninety years, but they have not disappeared. Fortunately, the reader wiU find 
in this volume that the current state of knowledge about this important phase of human 
history reflects many new insights gained from new bodies of data. Even if Fewkes's 
questions haven't been answered completely or HrdHcka's lament rendered unfounded, 
the peopling of the Americas is now understood in greater detail than ever before. 

In the context of the five-miUion-year history of the human lineage, the peopling of 
the Americas is, in some respects, of limited interest. After all, human colonization of 
the New World occurred in the latest Pleistocene and early Holocene, long after anatom- 
ically modern humans were established in the Old World. The peopling of the Americas 
did not, therefore, involve the origin or extinction of a new hominid species or the inter- 
action of two hominid species. And, because of the recency of the event, we also see lit- 
tle evidence of biological adaptation to New World environments (Brues 1977; Brace et 
al. 2001): There hasn't been much time for this to occur, and most of the adaptations 
that humans have made to life in the varied environments of the New World have been 
cultural not biological. What is of great interest and significance, though, is that the 
humans entering the Americas in the Late Pleistocene were entering a continental mass 
that previously knew nothing of humans or human cultures. Like the colonization of 


Australia that preceded it, the peopling of the Americas was an ecological event of enor- 
mous magnitude. Suddenly, modern humans with a sophisticated culture moved onto 
unknown landscapes, where the challenges caused by their ignorance of new habitats 
were more than compensated by their own adaptability and the many riches the new con- 
tinent had to offer. Fascinating too are the smaller-scale biological phenomena that 
occurred as small human groups entering the Americas underwent "microevolutionary" 
changes, resulting in subtie genetic differentiation from their parent stocks. This was 
accompanied by a tremendous diversification of languages, a demonstration of the dra- 
matic difference that can obtain between rates of biological and cultural evolution. 
Perhaps the most engaging, but least discussed, of the significant issues surrounding 
human colonization of the Americas is the topic of the human extinctions that no doubt 
occurred. It is easy to imagine that many small founding groups or populations met with 
hard times, as the result of rapid environmental change, a shortage of appropriate mates, 
or even new mushrooms. Paleontologists often try their best to connect fossil popula- 
tions with one another, or with living populations, in the absence of compelling evidence 
or theoretical reasons for doing so. The peopling of the Americas occurred at the end of 
the last Ice Age, when ecological times were tough, even for clever, adaptable, culture- 
bearing hominids. Not all populations survived, nor did they all give rise to modern 
Native Americans. Some of the populations that are preserved in the paleontological and 
archaeological records of the Americas no doubt represent extinct demes. Many more 
populations that made it to the Americas probably left no trace of their existence. 

Because of the intense interest among, especially, American scholars in the peopling 
of the New World over the last 150 years, many controversies have arisen over as inves- 
tigators have sought to answer Fewkes's simple but difficult questions. The intensit)^ of 
investigation into each of these questions has resulted in the whipping up of veritable 
tempests of debate that have swirled through generations of scholars and their students. 
The stormiest of these certainly concerns the exact date of arrival of the first humans 
into the Americas. Prior to the 1960s, it was widely considered that colonization of the 
Americas occurred 30,000 years ago or more (see Owen 1984, for a summary of this evi- 
dence). With the arrival in the 1960s and 70s of better techniques for determining 
absolute ages of events in the geological record, however, much more careful attention 
has been paid to establishing an accurate and reproducible chronology of human settie- 
ment in the Americas. Since the mid-1960s, the Clovis site of New Mexico and the 
Clovis fluted tool complex, dated between 12,000 and 11,000 yr BP has been the "gold 
standard" against which aU other claims for early arrival in the Americas have been 
judged. Since then, literally dozens of allegedly pre-Clovis sites have been paraded 
through the literature (Owen 1984; Marshall 2001) with only a very few surviving with 


their pre-Clovis hopes intact. Of the fetters that shackle the study of human prehistory 
in the New World, this is one of the strongest, because debates over geochronology 
have, in recent years, come to dominate all other areas of discourse in the study of early 
American prehistory. This is not to say that the insistence upon a robust geochronolog- 
ical framework is not important: It most certainly is. But many of the disputes and the 
disputants have more of the character of Greek tragedies than scholarly exchange. 
Controversies about the dating of the first colonization of the New World by humans 
are given attention in this volume, but they are not the dominant theme. Other issues, 
such as the biological origins of the first Americans, the mode and routes of their colo- 
nization, and their ways of life in the New World have been given relatively greater con- 
sideration. One of the topics of greatest interest to naturalists and anthropologists of 
the 19th and early 20th centuries was that of whether the aboriginal Americans had 
evolved in the Americas in situ or were migrants from other continents, even represent- 
ing remnants of the Lost Tribes of Israel (Hrdlicka 1912; Owen 1984). Over the inter- 
vening years, it has become clear from, especially, the study of biological evidence that 
the original Americans came from Asia. For most of the last century, the questions have 
been narrowed down to how and when they got to America, who exactly they were, and 
how they made a living. That is where this book begins. In "Setting the Stage: 
Environmental Conditions in Beringia as People Entered the New World," by Scott 
Elias, we learn about the nature of the environments that existed in the latest Pleistocene 
and earliest Holocene on Beringia, that vast Ice-Age land bridge between far northeast- 
ern Asia and far northwestern North America plied by many migrants to the New World. 
The task of recreating this landscape is not easy, and EUas does so with, arguably, some 
of the organisms best suited to the task — insects. Many of the insects and, specifical- 
ly, the beetles of the temperate and polar regions are remarkably narrowly adapted to 
their environments. EUas uses this high level of habitat specificit)' to great advantage in 
reconstructing the rapidly changing environments of Beringia in the period between 
13,000 and 10,000 yr B.P., when humans were beginning to make their way overland to 
the Americas. 

There then arises the seemingly innocent question of what the first migrants did 
when they arrived on their new continent. This topic is tackled by David Meltzer in his 
chapter, "What Do You Do When No One's Been There Before? Thoughts on the 
Exploration of New Lands." In this thought-provoking chapter, Meltzer develops two 
main questions: How did the earliest arrivals come to grips with what was what in their 
environment? and how did the initially small human populations maintain contact with 
each other for purposes of finding mates? In his chapter, he develops important ideas 
about how "landscape learning" must have occurred in the earliest human migrants, and 


how occasional communication between groups resulted in the exchange of vital com- 
modities — information about food and mates. 

For most of the last century, most scholars have theorized or at least assumed that 
humans migrated to the New World overland, via the Bering land bridge. Although 
migrations by sea were mooted, most theories of colonization by boat, either via coastal 
or mid-ocean routes, were considered improbable at best (see Owen 1984 for a review 
of the early literature). Fortunately, the theory has been dusted off and given new 
respectability by Jon Erlandson. In his chapter, "Anatomically Modern Humans, 
Maritime Voyaging, and the Pleistocene Colonization of the Americas," Erlandson 
points out that coastal migrations have been dismissed out of hand for lack of evidence. 
He makes the convincing argument, however, that rising sea levels since the Late 
Pleistocene have eradicated much of that evidence. Erlandson has spent much of the last 
two decades exploring steep shoreline sites along the North American coast, and has 
found considerable archaeological evidence for latest Pleistocene and early Holocene 
occupation there. He makes the cogent argument that there must be more such evidence, 
even if it is now inundated. 

The nature of the human populations that first peopled the Americas is the subject 
of the next two chapters. D. Gentry Steele and Joseph Powell examine the cranial oste- 
ological evidence for the nature of the founding populations in their chapter, "Facing the 
Past: A View of the North American Human Fossil Record." In their chapter, Steele and 
PoweU undertake multivariate analyses of the measurements of fossU and modern cra- 
nia. Their sample includes most of the fossil crania identified as latest Pleistocene or 
early Holocene Americans and a wide range of modern human populations. Their analy- 
sis reveals that the ancient Americans were a heterogeneous lot with respect to cranial 
shape, and that not all ancient North American populations have obvious similarities to 
the modern counterparts. 

Christy Turner attacks the same problem from a different angle in his chapter, 
"Teeth, Needles, Dogs, and Siberia: Bioarchaeological Evidence for the Colonization of 
the New World." Turner, who has long utilized dental morphological characteristics to 
infer patterns of human interrelationship and migration, draws upon a combination of 
dental and archaeological evidence in his review. He contends that most, if not aU, of the 
founding populations of the Americas came from Siberia, and that they appear to have 
arrived in three Late Pleistocene migration events. Archaeological evidence pertaining 
directiy to the timing of the arrival and adaptations of the earliest Americans is the sub- 
ject of the next two chapters. In "The Migrations and Adaptations of the First 
Americans: Clovis and Pre-Clovis Viewed from South America," Anna Roosevelt, John 
Douglas, and Linda Brown examine the growing evidence for habitation of South 


America by humans at or before Clovis times. They argue that not only are these sites 
competitive with Clovis in terms of their geochronology, but they preserve peoples that 
followed a different lifestyle. Their thorough review questions the theory of "Clovis as 
Ancestor" and offers in return "Clovis in Context." Marshalling copious geochronolog- 
ical, archaeological, and paleobotanical evidence, they demonstrate that most early South 
American populations were not big game hunters, but foragers and hunters of small 

A similar theme is developed by Tom DUlehay and Jack Rossen in, "Plant Food and 
its Implications for the Peopling of the New World: A View from South America." They 
argue that the absence of preserved plant remains from many archaeological sites has 
prevented scientists from appreciating the importance of plant foods in the diets of 
ancient peoples and has caused them to dwell inordinately on stone tools and the 
butchered remains of large animals. Using the Chilean site of Monte Verde with its well- 
preserved plant foods as their primary example, Dillehay and Rossen demonstrate that 
the earliest Americans inhabited "food affluent" environments such as wetlands and 
forests that contained a great diversity of plant biomass, and that most of their 
economies were based on foraging, not big game hunting. 

All contributors to this volume recognize the great antiquity of the occupation of the 
Clovis people in the New World. Not all, however, agree that their origins were Asian. 
In "Ocean Trails and Prairie Paths? Thoughts About Clovis Origins," Dennis Stanford 
and Bruce Bradley develop the controversial theory that the origins of the Clovis culture 
He in the western part of the Old World, specifically in the Iberian Peninsula. Citing 
strong similarities in the shape and mode of manufacture of fluted stone points, 
Stanford and Bradley argue that the closest affinities between North American Clovis 
culture and another stone tool tradition are with the Solutrean tradition of Iberia. 

Our volume on the earliest Americans concludes with two chapters that examine and 
apply "non-traditional" data to the problems of the nature of the earliest population(s) 
and the timing of their arrival(s). In "The First American Language," Johanna Nichols 
explores the origins of the tremendous linguistic diversity of modern indigenous peoples 
in the Americas by tracing the frequencies of structural elements in various languages 
over vast landscapes, an approach that is best described as combining linguistic typolo- 
gy with geography. Her reconstruction questions the relatively short time scale that 
archaeologists have defined and proposes that the Americas were colonized in the Late 
Pleistocene by Asian peoples, many of whom were originally coastal. 

The molecular evidence for the colonization of the Americas, a topic that Fewkes 
would surely have loved, is explored by Andrew Merriwether in "A Mitochondrial 
Perspective on the Peopling of the New World." This examination develops 


Merriwether's idea that the Americas were colonized in a single wave, followed by a rapid 
differentiation of peoples in situ. This book does not clearly answer any of Fewkes's sim- 
ple questions that opened this chapter. What it does allow readers to do is to learn how 
to judge the evidence relevant to those answers for themselves. Today, the study of the 
human colonization and occupation of the New World brings one into contact with 
many kinds of data, many theories and many scenarios: Evidence exists for and against 
the presence of humans in the New World before 12,000 years ago. Evidence exists for 
and against the h)^othesis of multiple migrations of humans into the New World. 
Evidence exists for and against the hypothesis that humans used coastal as well as hin- 
terland routes to populate the Americans. Evidence exists for and against the source 
population for New World migrants coming from Siberia. The lists of evidence and refu- 
tations go on and on and, not surprisingly, people new to this topic find it all quite bewil- 
dering. It would be presumpmous to state that the surfeit of new and old data present- 
ed in this volume leads to a consensus as to the nature, mode and tempo of the human 
colonization of the New World, but we are closer to achieving this end than ever before. 
Armed with the knowledge derived from chapters of this book, it should be possible to 
face this daunting collection of evidence with hope and equanimity. We are closer to the 
answers to Fewkes's questions than we suspect. 

Uterature Cited 

Brace, C. L., A. R. Nelson, N. Seguchi, H. Oe, L. Sering, Q. -F Pan, Y. -Y. Li, & T. 
Dashtseveg, 2001. Old World sources of the first New World human inhabitants: A 
comparative craniofacial view. I^roc. Natl. Acad. Sci. 98:10017—10022. 

Brues, A. M. 1977. Veopk and Races. MacMillan, New York. 

Fewkes, J. W 1912. Introductory remarks. The problems of the unity or pluralit}' and the 
probable place of origin of the American Aborigines. Am. Anthropol. 14:1^. 

Howard, E. B. 1936. An outline of the problem of man's antiquit)^ in North America. 
Am. Anthropol. 38:394-413. 

Hrdlika, A. 1912. Historical notes. The problems of the unity or pluralit)^ and the prob- 
able place of origin of the American Aborigines. Am. Anthropol 14:5—8. 

Marshall, E. 2001. Pre-Clovis sites fight for acceptance. Science 291:1730—732. 

Owen, R. C. 1984. The Americas: The case against an Ice-Age human population. 
Pages 517-563 in F H. Smith & F Spencer, eds.. The Origins of Modern Humans: A 
World Survey of the Fossil Evidence. Alan R. Liss, Inc., New York. 

Chapter Two 

Setting the Stage: 

Environmental Conditions in Beringiaas 

People Entered the New World 

Scott A. Elias 

^^^^^^k Archaeology is far more than the study of ancient artifacts. 
^ ^y ^^^^^^P Archaeological research aims at unraveling the history of ancient 
^^*P^^^^^ cultures and the forces that brought change to ancient ways of life. 
We know little of the earliest cultures in the New World, because 
these people trod lightly and moved rapidly across the landscape, leaving behind only 
fragmentary evidence of their culture. Apparently they rarely made any permanent 
camps or built any structures beyond simple stone hearths and tent rings. Our knowl- 
edge of them is based mainly on their stone tool kit. From this we deduce that the first 
human inhabitants of the New World were hunter-gatherers who moved about the land- 
scape, mainly in pursuit of big game animals. Independent from the archaeological 
record, we know that the Late Glacial period {ca. 13,000-10,000 yr B.P.) was a time of 
environmental upheaval, especially in Beringia. The aim of this paper is to describe 
Beringian environments during the Paleoindian Period, the transition interval between 
the Late Pleistocene and the Holocene. In so doing, I hope to set the stage for the 
Paleoindian migration into Alaska and the Yukon Territory. In other words, my aim is to 
provide an environmental context in which to examine Paleoindian lifeways at their point 
of entry into the New World. This is a challenging task. As Schweger and colleagues 
(1982) so eloquentiy put it, "To recreate a landscape, green with life or windswept and 
barren, and then repopulate it with animals and men is a formidable task." 

l^ate Pleistocene Physical ^Environments of Beringia 

Alaska, the Yukon Territory of Canada, and northeastern Siberia share a unique Ice 
Age history. Nearly all other high latitude regions of the Northern Hemisphere were 
covered by ice sheets during Pleistocene glaciations, but the lowlands of the aforemen- 
tioned regions remained unglaciated during much of the Pleistocene. Global sea level 


Figure 1. Map of northwestern North America and northeastern Asia, showing the boundaries of 
Beringia during the last glaciation. 

was approximately 120 meters lower than today (during the last glaciation). The conti- 
nental shelves between Alaska and Siberia are mostly less than 100 meters below mod- 
ern sea level, so, during the last glaciation those shelves were exposed as dry land, form- 
ing a land bridge linking the unglaciated lands of Siberia, Alaska, and the Yukon. The 
huge, largely unglaciated region is known as Beringia (Figure 1). 

Full glacial environments 

It is generally believed that Beringian lowlands remained ice-free because the Bering 
Land Bridge blocked Pacific moisture from entering many interior regions. PaleocUmate 
studies have shown that Beringia had sufficiently cold climate to develop and sustain 
glacial ice sheets, but this region lacked sufficient moisture to grow glaciers. For instance, 
much of lowland Alaska bears evidence of ice wedges and other relict permafrost fea- 
tures that formed in the Late Pleistocene (Pewe 1975). These features show that Alaska 
experienced temperamres that were sufficiently cold to foster the growth of glaciers dur- 
ing the Late Pleistocene. On the other hand, large-scale Late Pleistocene loess deposits 



and sand dunes attest to the relative aridity of climate. Pleistocene environmental condi- 
tions in Beringia were unUke any in the modern world, because the environmental fac- 
tors that contributed to the formation of Beringia (lowered sea level and continental ice 
sheets) are no longer in existence. Based on climate models, Bartiein and colleagues 
(1991) listed the important factors that controlled Beringian climate in the Late 
Pleistocene, as follows: 1) cooling effects from the continental ice sheets to the east; 2) 
a split in the atmospheric jet stream, in which one part flowed north of the ice sheets, 
and the other flowed south of them; 3) changes in insolation at high latitudes in the 
Northern Hemisphere through the various Milankovitch cycles; 4) reduced CO2 con- 
centrations in the atmosphere during glacial intervals and increased CO2 concentrations 
during interglacial intervals; 5) sea ice, sea surface temperature and salinity changes; and 
6) cooling effects of late-lying glacial ice on adjacent land areas. These factors interacted 
in many ways, only a few of which are well understood by paleocHmatologists. Some of 
these factors had unique impacts on Beringia. For instance, the spHt flow of the jet 
stream probably brought a stronger flow of southerly air into Alaska and the Yukon 
Territory, making summers cooler than otherwise, but perhaps making winters warmer. 
The combination of sea ice, sea surface temperamres, and salinity of ocean waters off 
the southern coast of Beringia may have brought warmer autumns and winters than exist 
today. Increased CO2 concentrations during interglacial periods served to amplify the 
warming brought on by increased insolation. Conversely, decreased CO2 concentrations 
during glacial periods would have decreased the degree of greenhouse gas warming of 
the atmosphere. Taken together, the various climatic factors interacted to yield intervals 
of interglacial and interstadial warming at 130,000-120,000 yr B.R, 110,000-100,000 yr 
B.R, 87,000-75,000 yr B.R, 60,000-30,000 yr B.R, and the Holocene. Low points in die 
insolation curve correspond with the times of mountain glacial advances in Eastern 
Beringia, and with the build up of continental ice sheets to the east. These intervals 
include 120,000-110,000 yr B.R, 75,000-60,000 yr B.R, and 25,000-18,000 yr B.R 

Estimating paleotemperatures 

Raleotemperature estimates for the Last Glacial Maximum (LGM) have been made 
from fossil beetie assemblages in Alaska and the Yukon Territory (Elias 2000a) and 
northeastern Siberia (Alfimov & Berman 2000). The Eastern Beringian estimates are 
based on the Mutual Climatic Range (MCR) method. This method is based on the deter- 
mination of a set of climatic parameters compatible with all of the species in a fossil 
assemblage. The first step in this procedure is the construction of a climate envelope for 
each species. The climate envelope data developed for tliis study focused on the mean 



temperature of the warmest month of the year (TMAX) and the coldest month of the 
year (TMIN) at the locations where the species are known to occur today. The climate 
envelope of a species represents the set of climatic conditions for which that species is 
adapted. Some species of beetles (stenothermic species) have very precise climatic toler- 
ances. In other words, they are adapted to live in a very narrow range of climatic condi- 
tions. Some of these species are found in only one small region of the world today, 
whereas others are found in many separate regions, such as mountaintops or isolated 
patches of desert, where the same climatic conditions exist in small geographic areas. 
Such species are the most useful for paleoclimatic reconstruction, because their presence 
in a fossil assemblage narrows the overlap between the species' climate envelopes to a 
small set of conditions. Formnately, a number of stenothermic beetle species were found 
in Late Pleistocene fossil assemblages from Alaska and the Yukon Territory. Other bee- 
tles (eurythermic species) are adapted to a broad range of climatic conditions. These 
species are generally found throughout many regions of a continent. Their presence in a 
fossil assemblage rarely provides any useful constraints on the paleoclimatic reconstruc- 

Not many modern beetie collecting localities are in close proximit}? to meteorologi- 
cal stations, so in order to achieve the best estimates of modern mean temperatures for 
these collecting sites, my colleague, Katherine Anderson (INSTAAR, University of 
Colorado) used a computer program that estimates climatic conditions at sites in North 
America, based on the nearest meteorological station to the collecting site. This com- 
puterized estimation is based on a 25 km-grid of the continent, developed by Bartiein 
and colleagues (1994). The climate parameters of the modern beetle collection sites were 
matched with the geographically nearest grid location to each collecting site. Then the 
climatic parameters of each of these localities for a species (usually several hundred col- 
lecting localities) were plotted on diagrams of mean July temperature (TMAX) vs. the 
difference between the mean July and mean January temperature (TRANGE). 

The Eastern Beringian fossil assemblages yielded TMAX and TMIN estimates for 
the LGM and Late Glacial intervals (Elias 2000b). The fossil beetie assemblages used to 
prepare the MCR estimates comprise 101 identified species of predators and scavengers. 
To avoid potential problems with plant migration lag, no plant-feeding species were used 
in the MCR reconstructions. Only predators and scavengers were used to estimate past 
temperatures. These beeties are thought to respond more readily to climatic change, by 
shifting their ranges away from regions of unsuitable cUmate. 

To standardize the results over a broad geographic region, temperature estimates are 
presented primarily as departures from modern TMAX and TMIN at the study sites. The 
paleotemperature estimates were calibrated to a modern data set, using Hnear regression 



equations to fit predicted to observed TMAX and TMIN values for modern localities 
with meteorological records (see Elias et al. 1999). In other words, our previous study 
used modern beetle assemblage data to predict modern temperatures at locations in 
Alaska and the Yukon Territory for which meteorological data are available, as a test of 
the paleoclimate method. The predictions of TJMAX and TMIN that were based on the 
modern beetie assemblage MCR reconstructions did not match the modern TMAX and 
TMIN data exactiy. For the warmest locations, the MCR estimates were too low. For the 
coldest locations, the MCR estimates were too high. We, therefore, developed the linear 
regression equations to adjust our beetie-generated estimates of TMAX and TMIN to 
the actual observed values. 

One difficulty in using MCR estimates from Beringia is that the coastal geography has 
changed with each glacial/interglacial cycle. Glacial and interstadial sites that lie near the 
modern coast were subject to continental, rather than maritime climates. This appears to 
have affected TMIN estimates more than TMAX estimates (Elias et al. 1999). It is, there- 
fore, difficult to interpret the TMIN estimates for coastal and land bridge assemblages, 
because changes in sea level shifted the climatic regime at these sites from continental to 
maritime. The land bridge was inundated by 11,000 yr B.P. (Elias et al. 1996). The MCR 
estimates vary from fairly precise to rather broad, depending on the number of 
stenotherms (species with limited thermal tolerances) in a fossil assemblage. 
Assemblages containing 3—4 stenothermic species often produce very narrow mutual cli- 
matic ranges, whereas assemblages with many eurythermic species (species with broad 
thermal tolerances) often yield only broad temperature estimates. TMAX estimates have 
an average range of 1.5°C, but a few ranges are as large as 5°C. This is an inherent weak- 
ness in the data that cannot be overcome except by additional studies of fossil assem- 
blages from critical regions and time intervals. The calibrated estimates shown in Figure 
2 are the best estimates available from the published data. 

The Alaskan fossil beetie assemblages from this interval (20,000-18,000 yr B.P.) )deld- 
ed a series of TMAX ranging from 5.5 °C colder than modern at 20,000 yr B.P. (Bluefish 
Caves, Yukon) to 0.9 °C warmer than modern at 18,000 yr B.P. (Bering Land Bridge Park, 
Seward Peninsula). The Eastern Beringian data also suggest that TMIN levels were with- 
in a few degrees of modern levels. Alfimov and Berman (2000) suggested that TMAX 
in northeastern Siberia was 12-1 3 °C during the Last Glacial Maximum (LGM). These 
temperatures are essentially the same as modern TMAX in their study region. The 
authors concluded that the most important difference between LGM and recent climates 
in this region was increased continentalit)^ The insect evidence thus points to relatively 
rmld climatic conditions in Beringia during the LGM, compared with far more dramatic 
cooling of summer and winter temperatures in regions south of the continental ice 




, +5' 

^K 0- 

ra E 
9- 0) 
Q o 


Arctic assemblages 
Subarctic assemblages 





'"CyrB.P. X1000 


"CyrB.P. X1000 

Figure 2. Calibrated estimates of Late Pleistocene TMAX for Eastern Beringia. Estimates in upper 
diagram (A) are shown as departures from modern TMAX values at study sites. The points shown in 
the lower diagram (B) are the calibrated results from linear regressions based on modern predicted vs. 
observed values (Elias et a/. 1999). 

sheets. Other lines of evidence, such as periglacial features that developed during the 
LGM, indicate that mean annual temperatures dropped significandy during the LGM 
(Hopkins 1982). Most paleoclimatologists would agree that, on a very broad scale, 
Beringia was relatively dry and cold with cooler summers during the LGM. More 
mesoscale patterns indicate east to west trends in temperamre and moisture gradients 
with colder and drier conditions dominant over eastern Beringia (Anderson & Brubaker 
1994; Lozhkin eta/. 1993; Anderson eta/. 1997; Hamilton 1994). 

Like the paleotemperature reconstructions, there is also considerable debate as to the 
vegetation cover in Beringia during the LGM. Most paleobotanical studies from this 
interval have focused on pollen extracted from cores. The interpretation of ancient mn- 
dra vegetation based on fossil pollen spectra is inherentiy difficult, because of low taxo- 



nomic resolution, poor dispersal of minor poUen types, and the wide ecological toler- 
ances of genera or species that dominated the poUen rain (Anderson et al. 1994). 
Goetcheus and Birks (2000) studied ancient land surface macrofossils preserved beneath 
volcanic ash at sites near the northern coast of the Seward Peninsula. They found that 
the vegetation was a closed, dry, herb-rich tundra with a continuous moss layer, growing 
on calcareous soil that was continuously suppUed with loess. They interpreted the region- 
al soUs during the LGM as relatively fertile, being sustained by nutrient renewal from 
loess deposition and the occurrence of a continuous moss mat. It remains to be demon- 
strated that the vegetation preserved in Goetcheus and Birks' site represents the steppe- 
tundra vegetation envisioned by paleobotanists as having dominated many regions of 
Beringia in the Late Pleistocene. The mystery of the nature and ecology of steppe-tun- 
dra remains unsolved. 

Another approach to the reconstruction of Beringian vegetation is the examination 
of relict patches of steppe, which are mosdy in northeast Asia. Based on this approach, 
Yurtsev (2000) has concluded that Beringia had much greater diversity of herbaceous 
vegetation (grasses, sedges, and forbs) in a mosaic of steppe-tundra landscapes. Yurtsev 
described some of the principal types of steppe-tundra vegetation that may have exist- 
ed in Beringia. Dry watersheds and slopes had cold-adapted steppe and cold and dry- 
adapted herbaceous and prostrate shrub-herbaceous communities. Depressions and val- 
leys supported dry steppe-meadows and brackish-water moist meadows. Valley meadows 
and the bases of slopes were the most productive as pastures for grazing megafaunal 
mammals, because of moisture and nutrients accumulated there. Yurtsev also considered 
that the lowlands of the Bering Land Bridge were covered with shrub tundra, which 
served as a barrier for the dispersal of steppe plants and animals. 

l^ate Glacial environments 

The Late Glacial period (14,000-10,000 yr B. P.) brought a series of large-scale, rapid 
environmental changes throughout Beringia. Climatic fluctuations forced wholesale 
changes in the distribution of Beringian plants and animals, and may have played the 
dominant role in the regional extinction of many megafaunal mammal species. 
According to the fossil beetle data from Eastern Beringia (EUas 2000b), TMAX values 
began rising at least by 12,000 yr B.P., reaching warmer-than-modern levels by 1 1,000 yr 
B.P. In Arctic Alaska, TMAX values rose as much as 7°C warmer than modern at that 
time. This dramatic climate change was probably brought on by a combination of 
increased insolation and the flooding of the Bering Land Bridge, wliich brought rela- 
tively warm, Pacific waters to the northwest coast of Alaska. 



The changing climate also brought about major changes in vegetation across 
Beringia. Pollen evidence (Brubaker et al. 2000; Edwards et al. 2000; Bigelow & Edwards 
2000) shows that the herbaceous tundra vegetation that dominated much of Beringia at 
the end of the last glaciation gave way to shrub mndra in most regions between 
14,000-12,000 yr B.P. This vegetational shift occurred at different times in different 
regions. In northwestern Alaska it began by 14,000 yr B.P. In Western Beringia it took 
place from 13,000-12,500 yr B.P. In the Mackenzie Mountain region, in easternmost 
Beringia, Szeicz and MacDonald (2001) found evidence of rapid climatic amelioration by 
1 1 ,000 yr B.P. Populus (balsam poplar and or aspen) expanded in these mountains from 
1 1 ,000—9000 yr B.P. Populus expanded north of the Brooks Range in Alaska during this 
same interval (Anderson & Brubaker 1994). 

Pollen studies point to a cHmatic oscillation during the Younger Dryas chronozone 
(10,800-10,000 yr B.P.) in Eastern Beringia (Brubaker et al. 1999). Elias (2000b) found 
strong evidence of a decline in TMAX values during this interval, especially in arctic bee- 
tie assemblages. TMAX values cooled by almost 7 °C over a period of about 350 years 
(Figure 2). Bigelow and Edwards (2000) noted a reduction in shrub tundra and an 
increase in herb tundra in central Alaska from 10,500—10,200 yr B.P. Strong evidence for 
a Younger Dryas cooUng has been interpreted also from pollen records from Kodiak 
Island (Peteet & Mann 1994). The history becomes a bit more confused in Western 
Beringia. Lozhkin and colleagues (1999) found no evidence for a Younger Dryas-type cli- 
matic event from most northern and eastern regions of Western Beringia. However, 
Pisaric and colleagues (2000) reported a possible Younger Dryas cooling in the lower 
Lena River basin, based on pollen evidence indicating an increase in herbaceous tundra 
at the expense of shrub tundra. 

Coniferous forest expansion through Eastern Beringia came only in the Holocene. 
The only source populations of conifers available to colonize Alaska were growing in 
distant southern regions during the Late Pleistocene. Paleohydrological modeling by 
Edwards and colleagues (2000) suggests that the Late Glacial period in eastern interior 
Alaska was a time of relative aridit}'^. The combination of extremely cold winters and lit- 
tle effective moisture may have combined to retard the expansion of coniferous forests 
in Eastern Beringia. For whatever reason, spruce forest did not reach the end of its 
migration route in southwestern Alaska until as late as 4500 yr B.P. (Brubaker et al 2000). 



L.ate Pleistocene Megafauna of Beringia 
l^ate Pleistocene fauna 

During the last glaciation, Beringia provided the major refuge for large mammals 
adapted to living in the high latitude regions of the Northern Hemisphere. Based on the 
fossil record, there was an abundance and diversit}' of large mammals that has never 
since been equaled in any high latitude region. The only modern ecosystem that supports 
a similar variet}' of grassland animals is the East African Savannah. There are some inter- 
esting parallels betu^een the Pleistocene steppe-tundra fauna and the modern African 
fauna. In the Beringian ecosystems, the woolly mammoth filled the elephant niche. The 
woolly rhinoceros, probably a rare species in Beringia, fiUed the rhinoceros niche, and 
saiga antelope filled the antelope niche. The modern range of saiga antelope is present- 
ly restricted to Central Asia, where arid grassland habitats persist today. It died out in 
Alaska and the Yukon Territory sometime after 13,000 yr B.P., as changing vegetation 
patterns eliminated its steppe-tundra habitat (Harington & Cinq-Mars 1995). In place of 
African zebras, the steppe-mndra had Pleistocene horses. Instead of water buffalo and 
wildebeest, the steppe-tundra had large-horned bison and two species of musk-ox. The 
dominant grazers on this cold grassland were the woolly mammoths. Pleistocene horses, 
and large-horned bison. Caribou were also present in the Late Pleistocene. Caribou feed 
on grasses and other herbs, shrubs, and lichens. Thus their diet represents a mixture of 
browsing and grazing. 

Mammoths appear to have been more dominant in Arctic regions of Eastern 
Beringia, whereas Pleistocene horses and bison dominated the subarctic regions (Kunz 
et al. 1999). These herbivores either became extinct at the end of the last glaciation, or, 
in the case of saiga antelope, they no longer live in the Beringian region. The only large 
true grazing mammal species stiU living in Alaska is the musk-ox. The other large plant- 
feeding mammals that remain here, such as moose, caribou, dall sheep, and mountain 
goat, are primarily browsers that feed on a mixture of shrubs and herbs. For instance, 
mountain goats feed on alpine grasses, mosses, lichens, and various woody plants and 
herbs. A moose's summer diet consists of a variet}' of plants, including pond weeds, 
sedges, horsetails, grasses, and the leaves of willow, birch, and aspen. In winter its diet 
shifts to Willow, birch, and aspen twigs. 

There is some intriguing evidence that the large grazing mammals of Pleistocene 
Beringia played a strong role in maintaining the grassy nature of their steppe-mndra 
habitat. Zimov and colleagues (1995) used a computer model to simulate the interactions 
of megafaunal mammals and vegetation in Beringia. They hj.'pothesized that trampling 



and grazing by mammalian grazers in a tundra landscape would have caused a vegetation 
shift from mosses to grasses. Grasses reduce soil moisture more effectively than mosses, 
because of their higher rates of evapotranspiration. The results of their model suggest 
that if herds of grazers were reintroduced into northeastern Siberia, Alaska, and the 
Yukon Territory, grass-dominated steppe vegetation would Likewise return. Zimov and 
colleagues believe that the megafaunal grazers had a sufficiently large effect on vegeta- 
tion and soil moisture in Beringia that their extinction may have played a major part in 
the shift from steppe to tundra vegetation at the Pleistocene-Holocene boundary. 

Wherever there are large numbers of herbivores, there will also be predators, and the 
Beringian grazers were preyed upon by a variety of predators. Again, comparisons with 
the African Savannah fauna come readily to mind. The Pleistocene lion looked very 
much like the modern lion, but was somewhat larger. The saber-toothed cat was also a 
powerful predator, armed with dagger-like canine teeth for sUcing into the neck of prey 
animals. Perhaps the most impressive Beringian Pleistocene predator was the giant short- 
faced bear. This animal stood about two meters tall at the shoulder and was three meters 
from nose to tail. Its long legs were adapted for running down prey. Grizzly bears can 
put on bursts of speed of up to 65 km (40 miles) per hour. Giant short-faced bears 
undoubtedly ran faster than that, but more importantly, they were able to sustain their 
speed over greater distances. We have no evidence that these bears sought out humans 
as prey, but some scientists have speculated that the New World was not a safe place for 
people to live until the short-faced bear died out. (See Turner, this volume, for more on 
this subject.) Along with the extinct predators, there were all of the modern Arctic and 
subarctic species, such as the polar and grizzly bear, wolverine, wolf, and lynx. 

Nature and timing of megafaunal extinctions and extirpations 

Recent efforts to pin down the timing of the major megafaunal extinction event at 
the end of the Pleistocene have placed this event at about 11,400 14 C yr B.P., with a sta- 
tistical error of about plus or minus 150 years (Elias 1999). The dating of this event is 
based on 140 AMS radiocarbon dates on purified protein extracts from fossil bones of 
megafaunal mammals that died near the end of the last glaciation, including camels, 
horses, and Pleistocene bison. Mammoths and mastodons persisted beyond 11,400 yr 
B.P Stafford and colleagues (personal communication, 1999) have dated the extinction 
of North American mammoth and mastodon to 10,900—10,850 yr B.P, so it now appears 
that there may have been two distinct extinction episodes. Each event probably took less 
than 100 years. One intriguing exception to this extinction event concerns the wooUy 
mammoth population of Wrangel Island, a small Arctic island off the northeast coast of 



Chukotka, Siberia. Vartanyan and colleagues (1993) obtained numerous radiocarbon 
dates on mammoth bone collagen, showing that the last mammoths died out on Wrangel 
Island only about 3,700 yr B.P. Wrangel Island is unique in several ways. It was connect- 
ed to mainland Siberia until about 12,000 yr B.P., when the intervening continental shelf 
was inundated. The modern climate and vegetation of the island closely resemble 
steppe-tundra environments of Pleistocene Beringia. Also, humans did not enter 
Wrangel Island until the late Holocene. If humans had been there earlier, they might 
have hunted the mammoths to extinction. 

Harly Human l^ifeways in Eastern Beringia 
The timing of touman entry into Eastern Beringia 

In spite of more than fifty years of dedicated research, we still do not know when or 
how people entered the New World. One view is that people crossed into the New World 
via the Bering Land Bridge and then proceeded south along an ice-free corridor in west- 
ern Canada. Another view is that people traveled by boat from Siberia to Alaska, then 
farther south Qosenhans et al 1995; Dixon 2000; Mandryk et d. 2000). (This theory is 
explored in detail by Erlandson, this volume.) Dillehay's (1986) discovery of human 
occupation at Monte Verde in southern Chile by at least 12,500 yr B.P places humans in 
the New World well before the end of the last glaciation. The archaeological evidence 
clearly shows that the inhabitants of Monte Verde had already acquired very sophisti- 
cated knowledge of regional plants and animals. (See Dillehay, this volume, for further 
details.) Such knowledge was not obtained in just a few years, so the Erst occupation of 
the Monte Verde region likely occurred well before 12,500 yr B.P. All of this necessitates 
human migration out of Asia well before 13,000 yr B.P. At the end of the last glaciation, 
the opening up of an ice-free corridor between the Laurentide and Cordilleran ice sheets 
is now believed to have happened only by about 12,400 yr B.P. (Catto 1996). It would 
appear, then, that people had either to move south from Beringia before the LGM {e.g., 
during the mid- Wisconsin interstadial period between 60,000 and 30,000 yr B.P), or that 
they had to enter South America by a different route. 

One possible scenario is that people traveled by boat along the Pacific coast of the 
Americas. The near-shore waters of the North Pacific undoubtedly provided a rich, con- 
stant source of food, including marine mammals, fish, and sheUfish. Successive genera- 
tions of marine-adapted peoples could have traveled from the coasts of northeastern 
Asia to southern South America, harvesting seafoods along the way. Geologic evidence 
Qosenhans et ai 1995; Mandryk et at 2000) indicates that parts of the British Columbia 



coast were ice-free during the late Wisconsin interval, allowing migrating peoples access 
to the resources of coastal landscapes. 

Whether or not Beringia played a key role in the peopling of the rest of the New 
World, the human colonization of interior Eastern Beringia began before 12,000 yr B.P., 
according to archaeological evidence from the Tanana and Nenana River valleys 
(Hoffecker et al. 1996). Three possibly older cave sites have been found in Eastern 
Beringia: Bluefish Caves, Lime Hills Caves, and Trail Creek Caves. Paleoindian archaeol- 
ogists have not reached a consensus about the validit)^ of these sites, although there is a 
reasonably good case for the human occupation of Bluefish Caves by 15,000 yr B.P. 
(Ackerman 1996). Additional archaeological research may push the date of earliest 
human occupation back even further. The bulk of the Paleoindian archaeological evi- 
dence in Beringia falls within the Late Glacial and early Holocene intervals, so this paper 
addresses that time period. 

The big game hunting tradition in Beringia 

As we have seen, the late Wisconsin glacial interval was a period of rapid environ- 
mental change that saw wholesale changes in regional vegetation patterns and the extinc- 
tion of many megafaunal mammals in Beringia. These large mammals were the chief 
prey of early hunters, so their extinction undoubtedly had an impact on human popula- 
tions, forcing the development of new hunting strategies and other changes in lifestyle. 
The climatic changes of this glacial interval may have forced the extinction of such graz- 
ers as mammoth and horse, but it seems to have favored other taxa such as bison, elk, 
and caribou, although the first two of these died out in Alaska and the Yukon during the 
early Holocene. The stone tool kit left behind by the first humans in Eastern Beringia 
can tell us only so much about their lifestyles. Most Late Pleistocene archaeological sites 
in Eastern Beringia contain few faunal remains. The exception to this rule is the Broken 
Mammoth site, in the Central Tanana Valley (Figure 3). Faunal data from this site demon- 
strate that people utilized a wide variet}^ of animals for food, including small game, 
waterfowl, and fish (Yesner et al. 1992). We know more about the food habits of the 
inhabitants of the Broken Mammoth site than we do for most other Beringian sites 
because the Broken Mammoth site has exceptional preservation of organic materials. 
The Broken Mammoth data serve as a reminder of how Uttie we know about these early 
inhabitants and their lifeways. 

I would Uke to end this paper with a more speculative discussion concerning human 
survival in Beringia. It is difficult to imagine hunter-gatherers living in any part of 
Beringia without wood to use for fires. From perhaps 80,000 to 12,000 yr B.P., most of 



— ! ; T— 

— T i 1 \ ! 

IKftV 142^ 136^« 

Wrangel Island 

Figure 3. Beringian archaeological and paleontological sites discussed in the text. 

Beringia lacked both conifers and tall deciduous trees. Shrub birch, alder, and willow may 
have been the only woody plants in most of Beringia during the last glaciation. Even 
though Beringia had abundant game animals to satisfy the dietary needs of hunter-gath- 
erers, it would still have been an exceedingly cold place for humans to live, especially in 
winter. As far as is known, these people Hved in tents made from animal hides. How did 
they keep their tents warm when the air temperatures dipped to -40°C ? In fact there is 
very little evidence that people lived in Western Beringia during the last glaciation, even 
though archaeological evidence has demonstrated human habitation of other regions of 
northeast Asia at that time (Pitul'ko 2000). The most reliable evidence suggests that 
Western Beringia was not inhabited until about 13,000 yr B.P. It could well be that the 
timing of the first appearance of people in Beringia was controlled mainly by the climatic 
amelioration that took place during the late Wisconsin glacial interval. Even if large trees 
were not yet established in most parts of Beringia until a few cenmries later, the inun- 



dation of the Bering Land Bridge would have brought another precious resource to the 
coasts of Beringia: driftu^ood from the Pacific coast regions further south. 

The only other useful source of heat for Late Pleistocene peoples in treeless regions 
of Beringia would have been the bones of large mammals. Burnt bones and ashes have 
been found in many Paleolithic sites in Eastern Europe; many northern regions were 
treeless during the LGM. However, such bone resources may not have been as readily 
available in Western Beringia, except in large basins such as that of the Kolyma River. 
Such basins tend to accumulate the skeletal remains of large mammals, and permafrost 
conditions preserve the fatt}^ contents of bones, allowing them to burn. But perhaps 
human access to the Kolyma basin was blocked by too many kilometers of treeless, 
boneless landscape in eastern Asia during the last glaciation. The highlands that separat- 
ed more southerly Asian human populations from the Kolyma basin were perhaps too 
great a barrier to cross without a ready source of fuel for heating. 

A-bs tract 

l^ike all archaeological issues, the peopling of the Nea> World must he placed 
in a paleoenvironmental framework. ^Environmental pressures probably played the 
dominant role in shaping the timing and direction of human migration into 
Alaska from Siberia. From the archaeological evidence currently at hand, it 
appears that the first human migration across the Bering Fand Bridge took place 
just as regional climates ivere ivarming at the end of the last glaciation, about 
1 2,000 years ago. The cold, arid climate that characteri:(_ed the height of the last 
glaciation (ca. 28 ,000— 1 4,000 yr B.P.) appears to have kept both eastern and 
western Beringia (unglaciated regions of northeastern Siberia) essentially treeless, 
and there is little or no evidence that people inhabited any part of the Beringia 
during that period. 

Between 12,000 and 1 0,000 yr B.P. , bands of hunter-gatherers became estab- 
lished throughout most of Alaska north of the Alaska Kange. This was a peri- 
od of environmental changes on a colossal scale, many of which would have been 
perceived by the human inhabitants. Over a period of perhaps a feiv decades, the 
Bering Tand Bridge was inundated. This, in turn, brought ii'arm. Pacific ivaters 
into contact with the Arctic Ocean, establishing oceanic circulation patterns that 
had been blocked for about 80,000 years. In much of Eastern Beringia 
(unglaciated regions of Alaska and the Yukon Territory), an extremely conti- 
nental climate gave way to a more maritime climate. Tn Arctic Alaska (by 
1 1 ,000 yr B.P.) average summer temperatures rose to levels as much as 7°C 
warmer than they are today This dramatic warming was folloived by an abrupt 



reversal, synchronous ivith the Younger Drjas oscillation in the North Atlantic 
regions. Many of the Pleistocene megafaunal mammals ivere extinct bj 1 1 ,000 jr 
B.P. Their demise probably forced people to adopt neiv hunting strategies and dif- 
ferent game animals. It remains unclear ivhether human hunting contributed sig- 
nificantly to the extinction oj megafaunal mammals in the New World. 


Funds for my paleoecological research in Alaska have been provided by grants from 
the National Science Foundation, DPP-8314957, DPP-8619310, DPP-8921807, 
OPP-9223654, ATM-9612641, and OPP-9424279. Many of the results summarized 
here were initially presented in a 1997 workshop on Beringian paleoenvironments, spon- 
sored by NSF grant OPP-9617429. I thank my INSTAAR coUeague, John Hoffecker, 
for a helpful review of the initial draft of the paper. 

l^iterature Cited 

Ackerman, R. E. 1996. Bluefish Caves. Pages 511-513 /// F. H. West, ed., American 

Beginnings, The Prehistory and Paleoecology of Beringia. University of Chicago Press, 

Chicago, IL. 
Alfimov, A. v., & D. I. Berman. 2001. Beringian climate during the Late Pleistocene and 

Holocene. ^^«/. Sci. Rev. 20:127-134. 
Anderson, P. M., & L. B. Brubaker. 1994. Vegetation liistory of north central Alaska: A 

map summary of late-Quaternary pollen data. ^//i?/. Sci. Rev. 13:71—92. 
Anderson, P. M., P. J. Bardein, & L. B. Brubaker. 1994. Late Quaternary liistory of tun- 
dra vegetation in northwestern Alaska, ^uat Res. 41:306—315. 
Anderson, P M., A.V Lozhkin, B.V Beleya, O.Y. Glushkova, & L. B. Brubaker. 1997. A 

lacustrine pollen record from near altitudinal forest limit, upper Kolyma Region, 

northeastern Siberia. The Holocene 7:331—335. 
Bartlein, P J., P. M. Anderson, M. E. Edwards, & P F. McDowell. 1991. A framework for 

interpreting paleoclimatic variations in Eastern Beringia. Ouat. Int. 10—12:73—83. 
Bartlein, P. J., B. Lipsitz, & R. S. Thompson. 1994. Modern climate data for paleoenvi- 

ronmental interpretations. American Quaternary Association Thirteenth Biennial Meeting, 

Program and Abstracts 197. 
Bigelow, N. H., & M. E. Edwards. 2001. A 14,000 yr paleoenvitronmental record from 

Windmill Lake, Central Alaska: Late glacial and Holocene vegetation in the Alaska 

t2iX\ge. Quat. Sci. Rev. 20:203—215. 



Brubaker, L. B., P. M. Anderson, & F. S. Hu. 2001. Vegetation ecotone dynamics in 
Souttiwest Alaska during the Late Quaternary, jg//^/. Sci. Rev. 20:175-188. 

Catto, N. R. 1996. Richardson Mountains, Yukon-Northwest Territories: the northern 
portal of the postulated "ice-free corridor." ^//^/. Int. 32: 3—19. 

Dillehay, T. D. 1986. The cultural relationships of Monte Verde: A Late Pleistocene set- 
tlement site in the sub-antarctic forest of south-central Chile. Pages 319—337 /// A. L. 
Bryan, ed., Neii' Evidence for the Pleistocene Peopling of the Americas. Center for the Study 
of Early Man, Orono, ME. 

Dixon, E.J. 2001. Human colonization of the Americas: timing, technology and process. 
Quat. Sci. Rev. 20:277-299. 

Edwards, M. E., C. J. Mock, B. P Finney, V. A. Barber, & P J. Bartlem. 2001. Potential 
analogues for paleoclimatic variations in eastern interior Alaska during the past 
14,000 yr: Atmospheric-circulation controls of regional temperature and moisture 
responses. ^//«/. Sci. Rev. 20:189—202. 

EUas, S. A. 1999. Quaternary paleobiology update: debate continues over the cause of 
Pleistocene megafauna extinction. Oimt Times 29:11. 

. 2000. Late Pleistocene climates of Beringia, based on fossil beetle analysis. 

Qmt Res. 53: 229-235. 

. 2001. Mumal climatic range reconstructions of seasonal temperamres based on 

Late Pleistocene fossil beetle assemblages in Eastern Beringia. Oiiat Sci. Rev. 

Elias, S. A., S. K. Short, C. H. Nelson, & H. H. Birks. 1996. Life and times of the Bering 

land bridge. Nature 382:60-63. 
Elias, S. A., J. T Andrews, & K. H. Anderson. 1999. New insights on the climatic con- 
straints on the beetle fauna of coastal Alaska derived from the mutual climatic range 

method of paleoclimate reconstruction. Arctic, Antarctic, and Alpine Research 31:94—98. 
Goetcheus, V. G., & H. H. Birks. 2001. Full-glacial upland tundra vegetation preserved 

under tephra in the Beringia National Park, Seward Peninsula, Alaska. Quat Sci. Rev. 

Hamilton, T. D. 1994. Late Cenozoic glaciation of Alaska. Pages 813-844 in G. Plafker 

& H. Berg, eds.. The Geology of North America, Vol. G— 1^ The Geology of Alaska, 

Geological Society of America, Boulder, CO. 
Harington, C. R., & J. Cinq-Mars. 1995. Radiocarbon dates on saiga antelope {Saiga tatar- 

ica ) fossils from Yukon and the Northwest Territories. Arctic 48:1—7. 
Hoffecker, J. E, W R. Powers, & N. H. Bigelow 1996. Dry Creek. Pages 343-352 in F H. 

West, ed., American Beginnings. The Prehistory and Palaeoecology of Beringia. University of 

Chicago Press, Chicago, IL. 



Hopkins, D. M. 1982. Aspects of the paleogeography of Beringia during the Late 
Pleistocene. Pages 3—28 in D. M. Hopkins, J. V. Matthews, Jr., C. E. Schweger, & S. B. 
Young, eds., Pakoeco/og)! of Beringia, Academic Press, New York, NY. 

Josenhans, H. W, D. W Fedje, K. W Conway, & J. V. Barrie. 1995. Post glacial sea levels 
on the western Canadian continental shelf: evidence for rapid change, extensive sub- 
aerial exposure, and early human habitation. Marine Geology 125:73—94. 

Kunz, M. L., D. H. Mann, P E. Matheus, & P Groves. 1999. The life and times of 
Paleoindians in Arctic Alaska. Arctic Kesearch of the United States 13:33—39. 

Lozhkin, A. v., P M. Anderson, W. R. Eisner, L. G. Ravako, D. M. Hopkins, L. B. 
Brubaker., P. A. CoHnvaux, & M. C. Miller. 1993. Late Quaternary lacustrine pollen 
records from southwestern Beringia. ^//^/. Kes. 39:314—324. 

Mandryk, C. A. S., H. Josenhans, D. W Fedje, & R. W Mathewes. 2001. Late Quaternary 
paleoenvironments of Northwestern North America: implications for inland versus 
coastal migration tovLtts. Ouat Sci. Rev. 20:301—314. 

Peteet, D. M., & D. H. Mann. 1994. Late-glacial vegetational, tephra, and climatic histo- 
ry of southwestern Kodiak Island, Alaska. Ecoscience 1:255—267. 

Pewe, T. L. 1975. Quaternary Geology of Alaska. U. S. Geological Survey Professional Paper 
835, 145 pp. 

Pisaric, M. E J., G. M. MacDonald, A. A. VeUchko, & L. C. Cwynar. 2001. The Late glacial 
and Post glacial vegetation history of the northwestern limits of Beringia, based on 
poUen, stomate and tree stump evidence. Quat Sci. Rev. 20:235—245. 

Pitul'ko, V. 2001. Terminal Pleistocene — Early Holocene occupation in northeast Asia 
and the Zhokhov assemblage. (2^iat. Sci. Rev. 20:267—275. 

Schweger, C E., J. V. Matthews, Jr., D. M. Hopkins, & S. B. Young. 1982. Paleoecology 
of Beringia: a synthesis. Pages 425^44 in D. M. Hopkins, J. V. Matthews, Jr., C. E. 
Schweger, & S. B. Young, eds., Paleoecologji of Beringia. Academic Press, New York, NY. 

Szeicz, J. M., & G. M. MacDonald. 2001. Montane climate vegetation dynamics in east- 
ernmost Beringia during the late Quaternary. Onat. Sci. Rev. 20:247—257. 

Vartanyan, S. L., V. E. Garutt, & A. V. Sher. 1993. Holocene dwarf mammoths on 
Wrangel Island in the Siberian Arctic. Nature 362:337—340. 

Yesner, D. R., C. E. Holmes, & K. L. Crossen. 1992. Archeology and paleoecology of the 
Broken Mammoth Site, central Tanana Valley, interior Alaska, USA. Current Research 
in the Pleistocene 9:53—57. 

Yurtsev, B. A. 2001. The Pleistocene "Tundra-Steppe" and the productivit)' paradox: The 
landscape approach. Quat. Sci. Rev. 20:165. 

Zimov, S. A., V. I. Chuprynin, A. P. Oreshko, E S. Chapin III, J. E Reynolds, & M. C. 
Chapin. 1995. Steppe-tundra transition: A herbivore-driven biome shift at the end of 
the Pleistocene. Amen Nat. 146:765-794. 


Chapter Three 

What Do You Do When No One's Been There 

Before? Thoughts on the Exploration and 

Colonization of New Lands 

David J. Meltzer 

^^^^^^^ Back in the old days, say about three years ago, we had a prett)' com- 

J^^^^^^^ pelling scenario for the peopling of the Americas.^ We believed the 

^^^■^^^^^^ first Americans came out of northeast Asia during the latest 

Pleistocene, crossing the Bering land bridge into Alaska during a 

period of lowered sea-level. For a time, the massive North American ice sheets blocked 

their way south, but after the Laurentide and Cordilleran glaciers melted back (say 

-12,000 years ago-), a group or groups of these hunter-gatherers sped south through 

the newly opened "ice-free" corridor between them (Wright 1991). Once they reached 

the unglaciated lower 48 states, they radiated out across the length and breadth of North 

America with what appears to be archaeologically-breathtaking speed ie.g.^ Haynes 1964, 


By 11,500 years ago, they'd arrived at a spring- fed pond just south of Clovis, New 
Mexico, where their archaeological traces were first found in the 1930s — associated 
with the bones of now extinct Pleistocene megafauna, particularly mammoth (Howard 
1936; see historical summary in Boldurian & Cotter 1999). Soon thereafter, Clovis and 
Clovis-like materials are found throughout North America in a range of environments, 
from the southeastern mixed deciduous forests, through the Plains grasslands, to the 
coniferous forests of the Pacific coast. No subsequent North American occupation 
would ever again achieve such a wide distribution. But they didn't stop here: As the tra- 
ditional mt^ had it {e.g.^ Lynch 1983), descendants of Clovis continued on to South 
America, but developed different artifacts and tools en route — so they were no longer 
readily recognizable as Clovis (Morrow & Morrow 1999; but see DiUehay et al. 1992) — 
and arrived at Tierra del Fuego within a millennium of leaving Alaska (Whitiey & Dorn 

That they traversed North America in what may have been less than 500 years (but 
see below) is all the more remarkable given the landscape: The Clovis colonizers arrived 



at a time of extraordinary environmental upheaval, one that saw massive mammalian 
extinctions, the dissolution of biotic communities that had been in place for tens of 
thousands of years, and rapid changes in climate {e.g., FAUNMAP Working Group 1996; 
Martin & Klein 1984; Wright 1991). And yet they seemingly coped with such adaptive 
challenges with ease: Their tool kit — including its signamre fluted spear points (Figure 
1) — is surprisingly uniform across the continent (Haynes 1982; Kelly & Todd 1988). Its 
apparent lack of variability not only gave added testimony to the radiocarbon evidence 
of a rapid radiation (which evidently lasted such a short time that few new artifacts were 
invented during the process), it also implied an underlying uniformit}' of adaptation — 
big-game hunting. Chasing mammoths, it was argued, enabled these hunter-gatherers to 
hurdle ecological boundaries {e.g., Kelly & Todd 1988; Mason 1962). Some pushed the 
traditional scenario further, arguing that intense predation on mammoth and mastodon, 
and other Late Pleistocene big game animals, was driving these animals to extinction, and 
their disappearance propelled Clovis hunters ever-southward in search of prey {e.g., 
Martin 1973). 

The idea of highly mobile, wide-ranging, big-game 
hunters, whose arrival was tied to the rhythms of conti- 
nental glaciation, made perfect sense — for a while. But 
there were always nagging doubts about this scenario, 
doubts which in the last decade grew much stronger. It 
was difficult to accept the idea that colonizers utterly new 
to a landscape could spread so rapidly under such com- 
plex climatic and environmental conditions (Whitley & 
Dorn 1993; but see Beaton 1991). Their points and tool 

Figure 1. A Clovis projectile point from the t5fpe site 
(Blackwater Localit}' No. 1). The original specimen is ~11 cm in 
length. Wliile these lanceolate forms vary considerably in technolo- 
gy and morphology, they share a suite of attributes, the most diag- 
nostic of which are the flute scars, the remnants of flakes removed 
from the base that traveled toward the tip. Fluting on Clovis and 
closely related eastern forms generally extends no further than a 
third of the way up the face of the point (in the illustrated specimen, 
the flute scar is ~2.4cm long). Finished points are generally ground 
smooth along the base and lower edges, presumably to facilitate their 
hafting (attachment) to a bone, ivory, or wood handle. Like most 
such specimens, the illustrated one is made of high-qualit}^ chert, 
obtained at a source hundreds of miles distant from the site in which 
it was found. (Drawn from a cast by Frederic Sellet.) 



kits proved not to be quite as uniform as they appeared at first glance (Meltzer 1993). 
The archaeological evidence for big-game hunting, the presumed driving force behind 
that expansion, proved to be localized, not continental in scale (Meltzer 1993b). Clovis 
tastes, it appeared, more often involved less risky and slower-moving prey species — Uke 
turdes (Johnson 1991; Stanford 1991). No surprise there: Successful foragers well under- 
stood that to be successful, they had to minimize risk — the risk of coming home 
empty-handed, or the risk of injury or death. While Clovis groups could and did occa- 
sionally bring down mammoths, it was not to the exclusion of other, smaller, and more 
readily exploited prey species. 

Then there was the problem of Clovis origins. Despite half a century of searching, 
no obvious Clovis progenitor has revealed itself in Alaska, let alone Siberia and north- 
east Asia (Bonnichsen 1991; Morlan 1991; but see Goebel et al. 1991; and King and 
Slobodin 1996). (And as a result, there are now very provocative claims of a far differ- 
ent source for these archaeological cultures; I will remrn to the matter of Clovis origins 

But most troublesome of all for the traditional scenario, there were frequent claims 
of a/>r^-Clovis presence in the Americas. These claims came not just from archaeology, 
but also linguistics and genetics {e.g., Nichols 1990; Torroni et al 1993). The latter derived 
their pre-Clovis ages from patterns in the languages and genes of modern Native peo- 
ples (see Merriwether, Nichols, this volume). Actual physical evidence of a pre-Clovis 
presence, of course, must come from the sites and artifacts in the ground, and for the 
last several decades virtually every field season brought forth a new contender. 

Some of those sites, like Calico Hills in California, ostensibly upwards of 100,000 
years old, were championed by the formidable Louis Leakey — fresh from his triumphs 
in Olduvai Gorge and eager to revolutionize New World archaeology as well (Leakey et 
al. 1968; but see Morell 1995:362—368). Others feamred less well known characters, but 
equally old sites (see listings in MacNeish 1976; Morlan 1988). All such sites were her- 
alded with great fanfare, and initial press releases spoke in glowing terms of genuine arti- 
facts in well-dated site contexts, 20-, 35-, or even 50,000 years old. But then, inevitably, 
the site tumbled down the slippery slope of what I came to call the pre-Clovis credibility 
decay curve when it was learned the supposed artifacts were more Ukely namraUy flaked 
stone, the dating technique was experimental and unreliable, or the site's deposits were 
so hopelessly mixed that alleged ancient artifacts were found associated with more recent 

Most pre-Clovis claims had a shelf-life of about a decade (Martin 1987), and so many 
of these contenders failed that the archaeological community grew highly skeptical of 
any and all pre-Clovis claims. In the face of that skepticism, the first site to break the 



Clovis barrier would have to not just meet but exceed the criteria by which early sites are 
judged (Meltzer 1995:33). Now it appears, thanks to Tom Dillehay's recently published 
(1989, 1997) work at the Chilean site of Monte Verde, we have that evidence. 

Although Monte Verde is only slightly more than a thousand years older than Clovis, 
its great distance from Beringia points to an inidal arrival in the Americas much earlier 
than 12,500 years ago. How much earlier depends on a host of variables, like ivhich route 
the ancestral Monte Verdeans took, whether coastal or interior, when those routes were 
open, how rapidly and by what adaptive means these groups traversed an increasingly 
unfamiliar landscape, and so on. Some (me included, see Meltzer 1997) have guessed they 
might have first entered the New World over 20,000 years ago, but that's just a (some- 
what) educated guess. It's not a real number. 

Monte Verde puts the peopling of the Americas in a whole new light, and raises anew 
a raft of questions about where these first Americans came from, who they're related to, 
how early we might be able to push back their entry, whether they came by land or by 
sea, how rapidly and by what methods and through which habitats they traversed the 
hemisphere, and so on. But what it doesn't do, not yet an^avay, is help us much with 

Thinking about Clovis 

Recall that one of the weaknesses of the Clovis-first scenario was its inability to 
explain the sudden archaeological appearance of Clovis. Of course, were there a large 
pre-Clovis population in North America, then one could argue that the sudden appear- 
ance of Clovis points was merely the diffusion of a new technology across an existing 
population. But that implies an unprecedented rate of diffusion not seen again until the 
invention of the hula-hoop, and of a technique (fluting) that (like the hula hoop) "was 
not all that revolutionary in the first place and that died out relatively soon thereafter" 
(Haynes 1971:10; though it's perhaps useful to add that the fluting technology was 
around in some form for a thousand years — a far longer showing than hula hoops). 
Further, diffusion demands a substantial pre-Clovis presence in North America — of 
which we have traces, but which we certainly would hope to have seen a greater presence 
by now, but haven't (Meltzer 1995:29; Storck 1991). 

Complicating the matter, for whatever it is worth comparing assemblages so widely 
separated in time and space, there are no obvious historical or technological links 
between the pre-Clovis age complexes of South America and the Clovis assemblages of 
North America. It looks as though Monte Verde and Clovis represent distinct archaeo- 
logical traditions and separate migratory pulses. The possibility these are separate migra- 



tions comes as no surprise: Given the absence of barriers to cross-Beringian traffic 
(Meltzer 1995), and tlie long period in which potential source populations were present 
in northeast Asia (Derev'anko 1998; Goebel & Askanov 1995; Hoffecker 1996), we 
would expect there to be many migrations — a flow or dribble of populations — from 
Asia to America. And back. 

More complicating still, the geographic expansion of Clovis seemingly occurred 
quickly across a wide area (archaeologicaUy speaking), based on the tight radiocarbon 
ages for this occupation, and the similarity of Clovis and related artifact forms and tool 
kits. These raise the intriguing possibility that their access to large areas of North 
America was essentially unrestricted — that they encountered few (if any) other people 
along the way, despite an apparently earlier migratory pulse into and through North 
America by ancestral Monte Verdeans.-' Perhaps. Of course, one must be mindful of the 
fact that there is variability in Clovis artifact technology, function, and style, the signifi- 
cance of which we (as archaeologists) do not yet understand. This is largely because we 
have yet to securely link patterns of material culture (distinctive artifact styles, say), with 
populations on the ground (be they social or cultural groups, linguistic units, or mating 
networks — not all of which, of course, are even necessarily synonymous with each 
other). We assume they are related, but they may not be in the straightforward ways we 

Furthermore, recent calibration of the radiocarbon curve for this period suggests 
that radiocarbon years may be compressed relative to calendar years, so Clovis groups 
may not have radiated as quickly in real time — a point anticipated by Haynes (1971; see 
Meltzer 1995; Taylor et al. 1996). How much longer the process may have taken is uncer- 
tain because of oscillations in the 14 C reservoir during the Late Pleistocene (Edwards 
et al. 1993; Hughen et al. 1998). Still, I suspect when the calibration issue is ultimately 
resolved, it will be the case that Clovis groups moved relatively rapidly — I consider tra- 
versing a continent in a thousand years a relatively rapid process. 

All that said, for the sake of the discussion I will make just that assumption: That 
Clovis groups did move far and fast across an unknown continent. I do so because at the 
moment, I am more interested in how or why they moved fast, rather than how fast they 
moved. I want to explore just how viable such a rapid expansion may have been, by 
thinking about what might have been the adaptive challenges they faced, and how they 
may have met them. 

So much of our attention up to now has been on the Clovis/pre-Clovis debate, only 
recently have we begun to explore these and other questions of the colonization process 
{e.g., Beaton 1991; Kelly & Todd 1988; Meltzer 1995). While some progress has been 
made, what follows is by no means the final say on any of these matters. What follows 



is based on my own recent model-musings (Meltzer 1998), in which I have tried to under- 
stand what the process of colonization might look like in terms of prey selection, habi- 
tat use, niche expansion, and patch choice (and the answer is, it looks fairly complicat- 

What I want to do here is focus on two components of that larger model, compo- 
nents that I think are critical to the colonization process: demography and landscape learning. 
As you'll see, these matters are relevant to any colonization of an empty landscape, 
though I wiU explore them in terms of Clovis, for that's the archaeological record I know 
best, and because Clovis is an especially interesting, if not unusual colonization episode, 
given the apparently breathtaking speed with which it occurred. 

There are, of course, other examples of archaeological or ethnographic groups mov- 
ing quickly across landscapes; but scarcely any examples of pedestrian groups moving 
that fast across an empty landscape — let alone an empt)' continent."^ The best illustra- 
tions of those come from the Arctic (not surprisingly so, given this is an environment 
that lends itself to rapid traverses). Consider the cases of the Paleo-Eskimo, and the 

The Paleo-Eskimo groups (badly misnamed: As McGhee notes there is nothing to 
demonstrate they are ancestral Eskimo, or even spoke an Eskimo language), moved 
across the high Arctic from Alaska to Greenland, in just a very few centuries around 
4700—4500 years ago (McGhee 1998). Their origins are uncertain, but they carried tool 
assemblages and left behind sites that have affinities to those of the Siberian Neolithic. 
These were groups who exploited seasonally freezing coasts and especially the adjacent 
interior highlands of the Arctic. They traveled far, and fast, but left little behind: the sites 
are few, widely-spaced (implying vast stretches of ground were left unoccupied and 
unsettled), and sporadically occupied by what must have been small populations. How 
they pulled themselves through this territory is uncertain, given their lack of time to 
accumulate knowledge of its complex environments and resources. McGhee raises the 
tantalizing h^rpothesis that they exploited a keystone species — musk-ox — which had 
evolved the unfortunate defensive strategy of responding to threats of predation by 
bunching in a circle. This may have proven effective against wolves (their major preda- 
tor until humans came onto the scene), but it would have enabled humans to come in for 
dangerously close shots. 

The descendants of these groups ultimately disappear from archaeological sight, 
replaced on the landscape by the Thule (who do appear to be ancestral Eskimo). Like the 
Paleo-Eskimo, the Thule move quickly across the High Arctic, but much later (Dumond 
1984; McGhee 1984). Around 900 A.D. they exploded out of western Alaska, moved 
eastward across the Arctic Coast, and arrived in Greenland within just a few centuries. 



Like the Paleo-Eskimo, they were tracking a keystone species, primarily bowhead whales, 
a species they'd hunted for millennia, mosdy around the Bering and Beaufort Seas. There 
they may have stayed, save for a warming trend around this time that melted summer 
pack ice across the high Arctic, enabling bowhead whales to expand east through now- 
unfrozen seas. The Thule hunters went along for the ride, metaphorically speaking, leav- 
ing behind a tool kit remarkably similar from Alaska to Greenland (McGhee 

Yet, unlike the Paleo-Eskimo groups who had an easily exploited keystone prey in a 
relatively stable landscape, or the Thule, who were radiating outward in a familiar, homo- 
geneous, and expanding niche, Clovis groups had no such advantages. They were pio- 
neering a trackless, increasingly exotic, and highly diverse landscape. Not only did the 
landscape vary across space, it varied through time, as longstanding biotic communities dis- 
solved, and plants and animals responded to the climatic changes that accompanied the 
end of the Ice Age. How those changes took place, and whether rapid enough to trigger 
drastic and unpredictable fluctuations in prey populations on a temporal scale (annual or 
decadal) that would have been detectable to and directly impacted human foragers, we 
cannot yet say.*^ Keep in mind that human life spans are (even today) less than a century 
— during Clovis times they were undoubtedly shorter. Biological or ecological process- 
es that played out over several centuries may not have dramatically impacted human for- 
agers, though certainly groups at the latter end of those Late Pleistocene changes faced 
a different world than those who first came to the land. 

Clovis groups also expanded across a much larger area than these Arctic groups — 
and just about everyone else we know of archaeologically. And unUke the landscape 
crossed by those fast-moving Arctic peoples, where the sparseness of resources made it 
advantageous to move quickly. Late Pleistocene North America was a very different 
place. Once groups were south of the ice sheets and beyond the range of near ice, 
periglacial conditions; they were in complex temperate and sometimes resource-rich 
habitats. The speed and frequency and distance that hunter-gatherers move daily, sea- 
sonally, or annually, is a function of many factors, not least of which is the spatial/tem- 
poral availability and abundance and distribution of critical resources (food, as well as 
water and stone), the density and return rates of preferred prey populations (and how 
quickly those changed under predation), the structure and relative richness and patchi- 
ness of the resources on the landscape, and so on. Unfortunately, we cannot measure 
these variables for the Late Pleistocene of North America; what we can say, however, is 
that this was decidedly not the high Arctic, and need not and may not have been a set- 
ting in which rapid movement was necessary or even advantageous. 

A continent the size of North America, if colonized rapidly, must have stretched 



Clovis populations very thinly across the landscape. There are demographic costs to mov- 
ing that far that fast, for it almost certainly involved moving away from other people, 
whether kin or unrelated groups (MacDonald & Hewlett 1999; Meltzer 1995). There can 
be little doubt population density was lower at this time than at any subsequent point in 
American prehistory. And yet in the face of that, colonizers had to maintain a "critical 
mass" of population and an accessible source of potential mates — in order to avoid 
inbreeding or, worse, extinction. These demographic demands would have been more or 
less severe depending on the local group's size, growth rates, age and sex composition, 
degree of environmental uncertaint)^ (and the inverse: knowledge of the landscape's 
resources), as well as how rapidly it was moving from its geographic homeland and/or 
other groups and the local environmental constraints on group size and population den- 
sities (Borgerhoff Mulder 1992:341-342; MacArthur & Wilson 1967:78, 80, 88; Meltzer 
1999; Whallon 1989:434—435). On a diverse and unfamiltar landscape, the risk of extinc- 
tions is greatest soon after dispersal, when population numbers and growth rates tend to 
below(Belovsky 1999).^ 

At the same time, groups moving into increasingly unfamiliar habitats had to learn the 
landscape, and identify and/or locate vital resources, since over the course of time in res- 
idence in an area the preferred or high-ranked resources in that area would decline due 
to foraging pressure. They had to learn what else would feed them, and what would 
clothe them, cure them, or kill them. They had to learn when and where and what other 
resources were available. Each new habitat colonizers entered would, for a time, have 
been unpredictable for the simple reason that the availability and abundance of plant and 
animal resources varies. 

With fixed, permanent and predictable resources — like outcrops of stone suitable 
for making tools — landscape learning could be rapid. Learning about more ephemeral 
resources in unfamiliar settings — plants and animals — took longer, requiring as it did 
that resources be identified, and their properties or behaviors, habitat, location, short and 
long term patterns in abundance and distribution over time and space be learned, which 
in some cases might have required substantial observation or experimentation. Just how 
much of this behavior went on, no one knows, but bear in mind there are over 30,000 
species and varieties of plants in North America alone. It would have been all but impos- 
sible for each new generation to sample anew their properties. Once observation and 
experimentation was complete, there must have been a considerable selective advantage 
to maintaining knowledge of a plant's properties — as the similarity of medicinal plants 
across different areas attests (Daniel Moerman, personal communication, 1999). But still 
knowledge had to be gained anew, as groups moved progressively further south, and into 
areas with unfamiliar and exotic floras.^ 



Would that suggest groups may have then initially focused on big game, as a class of 
prey that would have enabled easy crossing of ecological boundaries or into unknown 
territories (much as musk-ox possibly enabled groups to penetrate the high Arctic)? 
Perhaps, but keep in mind not all animals, let all alone all elephants (mammoth and 
mastodon, in our North American Pleistocene case) are alike in their behavior. Foragers 
harvesting different and perhaps unfamiliar species in the same size class (large game, in 
this example), might have been able to bring generalized weaponry, hunting skills, and 
tactics to bear on all such prey (Kelly & Todd 1988:234), but how those different species 
may have responded would have varied considerably by species, as Prison has argued 
(1991). Even within a species, different animals may behave very differentiy across the 
landscape and at different times. Unless the environment is extraordinarily rich and ani- 
mal encounter rates were very high, locating a specific prey species required knowledge 
of the spatial and temporal and landscape behavior, which varies by the animals' age, sex, group 
size and composition, the season, weather, time of day, competitive relationships, avail- 
able water, physiological stage (breeding, pregnancy, lactation), species composition and 
heterogeneity of the vegetation, topography, and exposure to predation, among other 
factors {e.g., Van Dyne et al 1980:285-298; also Prison 1991:141; Johnson et al 1992). AU 
of which, of course, presupposes some knowledge of the landscape. 

Obviously, one's knowledge of the environment and abiUt}^ to predict resource avaU- 
abilit}^ and abundance increases with the time spent observing resources, and their cycles 
and patterns on the landscape. Colonizers new to a habitat haven't that knowledge base 
— and cannot ask the locals since there aren't any locals to ask — to successfully antic- 
ipate much beyond a seasonal or annual basis when and where resources would be avail- 
able. Hence, patterns of hunting and gathering in an initial foray into a new landscape 
and habitat may have been more experimental and less systematic, more encounter-based 
and less intensive, than such patterns seen among groups after settling into an area. 

In this regard, we see why the ethnographic record of hunter-gatherers is not an 
especially useful analogue for colonizers of empt}^ continents. Virtually all the historical- 
ly known hunter-gatherers have neighbors: Neighbors who are not only convenient 
sources of mates, but also handy sources of information about landscapes and resources 
(and who, in some cases, are also competitors, which can fundamentally alter the calcu- 
lus of landscape use). Our models of hunter-gatherer foraging assume a forager has rel- 
atively complete information about resource distribution and yields, bv \\'hich decisions 
can be made about how long to stay and when to leave a patch (Kaplan & Hill 1992:186; 
Kelly 1995:96-98; Stephens & Krebs 1986:75). They acquire the information by con- 
stantiy monitoring the landscape and, because one group can only cover so much 



ground, by talking to their neighbors and tapping their knowledge about resources avail- 
able in other areas. Clearly, this was not an option for Clovis groups (or, for that matter, 
any new colonizers). 

It has been argued, and I think correctly, that the greatest effort in information acqui- 
sition should occur in patchy and unpredictable environments (Kaplan & Hill 
1992:186-187; also KeUy 1995:151, Stephens & Krebs 1986:103) — the very ecological 
situation in which Clovis groups initially found themselves. Indeed, there are reasons to 
think that on an unknown landscape, natural selection would favor rapid and extensive explo- 
ration to see what's over the next hill. Under the circumstances, one would expect rapid 

Demography and landscape learning are tightiy linked, for several reasons, not least 
that the decision to stay in a patch or move onto the next is in part based on the suit- 
abilit}' of a new patch relative to the current one, factoring in the costs of moving (Baker 
1978:43). Those foragers who can better calculate those costs (by having more knowl- 
edge) will increase their odds of survival. By gaining information about a landscape one 
reduces mortality, and thus can increase population growth rates (MacArthur & Wilson 

Arguably, then, the colonization process on a new landscape involved a trade-off 
between a series of competing demands, including: maintaining resource returns, or keeping 
food on the table, particularly as preferred or high-ranked resources decUned, and in the 
face of limited knowledge of the landscape; mnimi^ing group sit^e, in order to buffer envi- 
ronmental uncertainty or risk on an unknown landscape; maximi^ng Piobility, in order to 
learn as much as possible, as quickly as possible about the landscape and its resources (in 
order to reduce environmental uncertaint}' in space and time), while; staying as long as pos- 
sible in resource-rich habitats, in order to enhance knowledge of specific changes in resource 
abundance and distribution (^^ou learn more by staying longer); and, finally and perhaps 
most critically, maintaining contact between dispersed groups, in order to sustain information 
flow, social relations and, most especially, demographic viability. 

Under the circumstances, we expect to see among colonizers laige scale exploration, in 
order to map the landscape; periodic aggregations of widely dispersed groups in order to 
exchange mates, resources, and information; and extensive mating netivorks, in which spous- 
es can be drawn from distant groups. And central to making all this work are two 
processes that might actually be detectable archaeologically: high settlement mobility to main- 
tain contacts with distant groups, and monitor resources and environmental conditions 
beyond the social and geographic boundaries of the local group; and open social networks, 
which enabled individuals to move easily between and be readily integrated within dis- 
tant groups (see the insightful paper by Anderson [1995] on this matter; Lourandos 



[1997] outlines how these processes may have been manifest in Australia). Under the cir- 
cumstances, strongly territorial behavior would be decidedly disadvantageous. 

Disproving the Case 

Now this is all well and good in theoretical terms, but how do we detect these social 
processes, archaeologically? Not easily, and with very broad brush strokes — as the data 
are not up to the resolution demanded of the models! But let me explore what we might 
see of landscape learning, mobilit)^, and the opening and closing of social networks 
among Clovis and subsequent Paleoindians. 

It has long been known that Paleoindians generally, and Clovis groups in particular, 
preferred high quality stone to manufacture their artifacts (Goodyear 1979; Haynes 
1982). That stone can often be pinpointed to its geological source. As very little of it 
appears to have been acquired via trade or exchange with other groups (Meltzer 1989), 
the distances between the geological source and the site where it was discarded becomes 
an odometer of Paleoindian mobilit}^ Oftentimes, the distances traveled were substan- 
tial: Among western Clovis groups there are several instances of stone procured upwards 
of 500 km from the site in which it was discarded. Average values of 300 km are not 
uncommon (Haynes 1982:392). Probably these groups were even more mobile than the 
figures indicate, for these values only record the straight-line distance from source to site, 
and take no account of side excursions, return trips, or other archaeologically invisible 
travel (Meltzer 1993a). Actual Clovis mileage may vary. 

As Clovis groups looped around the landscape, they strayed far from their favored 
stone quarries, and as their artifacts were used, broken, and their stone supply dwindled, 
they carefully maintained, resharpened, and recycled their remaining tools. It's the sort of 
thing you expect of wide-ranging and highly mobile colonizers who might be exploring 
a new landscape or cannot anticipate what stone resources will be locally available. But 
then that same reliance on distant or exotic high qualit\' stone is also evident among later 
Paleoindians, like Folsom and Plainview groups on the western Plains, who presumably 
had more knowledge of the land (in the eastern woodlands, later Paleoindian groups 
more quickly take up the use of local stone, a difference between east and west that may 
speak to important differences in land use patterns, demographics, and resource rich- 
ness). So the pattern of relying on exotics obviously says more to stone preferences of 
mobile hunter-gatherers, than it does about the landscape famiHarit)^ of colonizers. 

Let's delve a littie deeper. One of the characteristics of the Clovis archaeological 
record that has become apparent lately is the unusual number of artifact caches. These 
contain an^^where from a dozen to a hundred or more pieces, flaked into shapes (bifaces, 



blades) that are relatively easy to carry, and once needed could readily be flaked into any 
one of a varieu^ of useable tools as the circumstances required. The caches rarely occur 
at Clovis camps or kills, but instead turn up in otherwise isolated spots, giving the appear- 
ance of having been stashed — presumably with the intent of being retrieved later. The 
caches are generally comprised of stone from rock outcrops often several hundred kilo- 
meters distant. 

On its face, an artifact cache represents a handy solution to a logistical problem: The 
disparity between the location where a stone could be acquired (which is predictable), 
and where it was to be used (which is not always predictable, and in any case may be far 
away). By caching stone around the landscape, you make re-supply available without a 
long remrn trip to the quarry. We now know of at least a dozen or so Clovis caches, scat- 
tered throughout North America — but mostly in the west (Collins 1999). Given that 
later Paleoindian groups are just as mobile as Clovis groups, and just as reliant on high- 
quality distant stone sources, we would also expect to see caches from later Paleoindian 
periods. Yet, we don't: Later Paleoindian caches are extremely rare.*^ 

There are now enough Clovis caches, and few enough later Paleoindian ones, to sug- 
gest the pattern is real and not a sampling effect. How then, can we explain it? The scarci- 
ty (if not absence) of later caches is attributed by Michael Collins (1999, and personal 
communication) to xh& predictability of Clovis movement: He suggests these groups knew 
in advance where they were headed, knew they would be returning to particular spots 
with exhausted tools in need of replenishment, and hence left stone cached behind for 
future use. Later Paleoindians, he argues, could not predict their fumre movements with 
enough confidence to benefit from caching stone. 

Perhaps. Yet, the stone used in Folsom times (immediately post-Clovis) comes from 
a greater diversity of sources, sometimes from greater distances (Meltzer, unpublished 
data). This more complex pattern of stone acquisition, suggests that however unpre- 
dictable their future movements, they certainly knew where the sources were and were 
highly adept at planning their use. Moreover, if Clovis groups knew they were returning 
and would need and use the stone, then why were the caches still there 1 1,000 years later? 
In fact, I tend to suspect the opposite was occurring: That Clovis groups were so ne7v to 
the landscape they didn't know and could not predict where or when they were next 
going to find vital resources. By leaving caches scattered about the landscape, as they 
moved away from known sources, they created artificial resupply depots, and thus antic- 
ipated and compensated for their lack of knowledge. If they found no stone in the area 
where they were headed, they at least would not have to double back completely to the 
outcrop to refurbish their tool kits when those were used up. 



Stone is especially suitable for caching, since unUke meat or other foods, it won't 
spoil, won't be attractive to scavengers, and barring tectonic activit)'^, won't move before 
you return. A stone cache becomes a fixed and predictable and essentially permanent 
location. As such, it serves a useful role in exploring and learning new lands, for now the 
emerging map of the landscape includes not just the newly located natural resources, but 
also some artificially-placed ones. This helps make it possible to venture further and 
more widely across otherwise unknown and unpredictable terrain. Over time, as new 
sources of suitable stone are located, and as groups are better able to predict where and 
when they wiU be able to replenish their suppUes (presumably in the post-Clovis period), 
then caches become less critical to movement, and ultimately unnecessary. 

Of course, one cannot use the absence of caches in later Paleoindian periods as proof 
that landscape learning is complete by then — at least not without getting dizzy travel- 
ling in a logical circle. But I think there might be independent evidence — albeit cir- 
cumstantial — to indicate that might be the case, evidence which gets us back to the issue 
of changing social networks. 

Colonizers on a landscape not only had to be able to track great distances to find 
mates and exchange information and resources, they also had to be able to get along with 
those near and distantly related groups they encountered. Having large and open social 
networks based on flexible and fluid social and kin relations, fewer languages, the easy 
integration of individuals and groups, and sometimes long-distance exchange and 
aUiance networks — all combine to diminish differences among peoples who need to 
stick together under geographic circumstances that might otherwise keep them widely 
separated for years at a time (Anderson 1995; Meltzer 1998). 

One way these open social systems are manifest in the material and archaeologicaUy 
visible realm, is by the widespread distribution, use, and exchange of recognizable, and 
sometimes highly symbolic, artifact forms. These forms had social and ritual functions, 
and over long spans and large areas served to maintain recognition and alliances. In a 
similar situation but a very different setting, Kirch argues that the exchange of Lapita 
ceramics and other items between the widely separated island populations of Oceania 
served as a social binding that helped insure long-distance alliances and a supply of suit- 
able marriage partners in the face of "unpredictable environmental hazards . . . [and] 
demographically small and unstable groups." It might have meant the difference between 
survival and extinction on these remote islands (Kirch 1988:113—114, 1997:254). 

Which brings us back to Clovis points. Early in the Paleoindian period, these artifacts 
are more broadly similar (st5'Listically, technologicall)'^, and t^^pologically) across a larger 
area of North America, than any artifact forms in any later cultural period, Paleoindian 
or otherwise (Figure 2).^*^ In fact, around the world the only other artifact forms com- 



Figure 2. Generalized distribution of Clovis and related fluted points, -11,000 yr B.P. (only Plains 
and southwest sites are radiocarbon-dated to this age). (See endnote 11 for explanatory comments). 

parable in distribution are Lapita ceramics across the islands of Oceania, and the so- 
called "Venus" figurines of Upper Paleolithic Europe and Eurasia (Whallon 1989). The 
extent of the Clovis distribution is a by-product of the size of the colonizing footprint, 
but the si/m/arify across that range may well reflect some measure of the social network on 
the landscape, the symbolic value of that network, such similar forms provide, as well as 
perhaps some measure (or by product) of a cultural founder's effect. 

Whatever the precise cause, the conditions behind it ultimately changed, and did so 
within a matter of centuries. Although the timing varies by area of the country, some- 
time after 10,900 years ago in the central and western portions of the continent, and after 
10,600 years ago in eastern North America, Clovis and its related and highly similar 
forms disappear (Figure 3). By 10,600 years ago, there are in place across North America 
a variety of regionally-distinctive point forms (Anderson & Faught 2000). 

Unlike Clovis, these later Paleoindian points are more restricted in their geographic 
distributions. At least in some areas, these more-regionally specific forms correspond 



Figure 3. Generalized distribution of select Paleoindian point forms after -10,900 (in western 
North America) and -10,600 yr B.P. (in eastern North America). 

with the development of distinctive adaptive strategies. The best known and most read- 
ily recognizable of these is the Folsom complex, which appears on the Plains and in the 
Rocky Mountains of western North America, with a subsistence strategy tightiy tied to 
exploitation of the now-extinct species, Bison antiquus. In other areas, the adaptations of 
these groups is different and less well known. 

That process of shrinking style zones continues, and by 10,000 years ago most areas 
of North America have very distinctive point forms (Figure 4), and perhaps adaptations 
as well. For it is in this window of time that we see the development of new technolo- 
gies {e.g., the Dalton adze in the Midwest), new and sometime prey- or region-specific 
strategies for hunting and processing, and increasing reliance on more locally available 
stone for artifact manufacture (see, for example, the papers in Anderson & Sassaman 

Let's assume that these stylistic forms and traditions mark cultural groups or lineages 
(however those terms are defined), and that different stylistic forms mark changes in 



Figure 4. Generalized distribution of select Paleoindian point forms after -10,500 B.P. 

those cultural groups or lineages. Considered that way, the shift from broad and overall 
homogeneity in stylistic forms, to narrower and more regional ones, in the span from 
11,500 to 10,500 years ago, can be seen as the settling in of colonizers into specific areas 
and/or habitats; a reduction in the scale or openness of the social systems; and the 
development of more region-specific adaptations (also Anderson 1995:5, 11).^^ I wish I 
could say more about this issue, but I don't think I can at the moment without abusing 
the archaeological record even more than it has been. 

Conclusions and Some Implications 

So why did Clovis groups radiate so far and so fast — assuming they did? I don't 
think it was because they were chasing mammoths — or because they were being chased 
by mammoths — or because the extinctions of these animals left Clovis groups with lit- 
tle to eat, so they had to press on. Rather, I think it is because they found themselves on 



a vast and empty landscape that was unknown and unpredictable, in which there was a 
selecdve advantage to widespread movement. During the critical early centuries of their 
dispersal, they still maintained vital long-distance links to other groups that could pro- 
vide mates, resources, and information. As their knowledge of the landscape and its 
resources became ever-more detailed, their foraging systems became more stable. This, 
in turn, increased their population size, lessened their odds of going extinct, and enabled 
them to extend their overall settlement range, and venture further from one another. In 
the end, Clovis and related groups were, with little doubt, highly successful colonizers. 
Within a thousand years, much of the North American continent was populated with 
their descendants. 

But who were their ancestors} We're not altogether sure. As noted above, there are no 
obvious Clovis progenitors in Beringia (but see Goebel et al. 1991; Hoffecker et al. 1993). 
Certainly, the most diagnostic aspect of Clovis — its fluting technology — looks to be 
uniquely American. While some fluted points have been found in Alaska, they are most- 
ly undated, and may \iz younger \}a.2M those further south (Reanier 1995). StiU, most archae- 
ologists look to Alaska and northeast Asia as the ultimate source area for the Clovis pop- 
ulation. Whether they arrived, as the traditional notion has it, fast on the opening of the 
ice free corridor is uncertain, given the uncertainty of our knowledge of the timing of 
that opening (Jackson & Duk-Rodin 1996; Mandryk et al. 2001). 

Some, in fact, suggest they made their way into the continent — or at least parts of 
the continent — sometime earlier. They point to Meadowcroft Rockshelter in 
Pennsylvania, which has produced a lanceolate point with a pre-Clovis age of 12,800 
years ago (Adovasio et al. 1990), and (of late), to Cactus Hill, Virginia, which has pro- 
duced Clovis points and, below them, decidedly non-Clovis forms in sand dune deposits 
dated to upwards of 16,000 yr B.P. (McAvoy et al 2000). Neither site is fully accepted by 
the still-skeptical archaeological community. ^^ Still, there's something curious about east- 
ern North America: There are more Clovis artifacts from here than any other area of 
North America, implying a relatively longer period of occupation — though whether it 
began earlier or lasted later we cannot say (Anderson 1990; Mason 1962; Meltzer 1988). 
Meadowcroft, Cactus HiU, and the great abundance of Clovis in eastern North America 
have led some to suggest that Clovis was invented here, and from whence it rapidly 
spread west. 

But then there's been an even more radical spin to the notion, for it has not gone 
unnoticed that having Clovis radiate out of eastern North America puts it that much 
closer to Europe. And Europe is where Stanford and Bradley (this volume) ultimately see 
Clovis origins (also Boldurian & Cotter 1999). Their argument, briefly, is that there are 
such strong similarities in technology and artifacts of the Solutrean Period of Upper 



Paleolithic Europe, and those of Clovis, that one must be descended from the other. 
Those similarities include: distinctive biface thinning technologies (particularly the use of 
outre passe thinning), common end scraper forms, the presence of bone rods and 
engraved stones, a reliance on exotic raw material, and the use of red ochre. As they 
envision it, Solutrean maritime hunters carrying these technologies and artifacts left the 
Iberian Peninsula in skin canoes heading north and west. They hunted their way through 
the leads fronting the pack ice that had descended across the Pleistocene North Atlantic, 
and ultimately made landfall in eastern North America, from whence sprang Clovis. 

This is a fairly provocative claim (Sellet 1998). Recognizing its controversial status, 
Stanford asks us to "suspend disbelief" wliile he and Bradley work out the details. While 
I'm not willing to suspend my disbelief, I am — in the spirit of Davis's (1926) notion of 
an "outrageous h^'pothesis" — willing to concede our knowledge of Clovis origins is 
limited, and it is worth examining our evidence and arguments closely about what we 
know and do not know on the matter, and how much latitude there may be within the 
current evidence for such alternative hypotheses. So I think it only fair they be granted 
the opportunit}' to work out the details, as proving their case will take time and will be 
no easy matter, as they appreciated-^ 

Still, there are obstacles to this h)^othesis, the most substantial of which surfaced the 
very first time American archaeologists sought to connect the prehistories of America 
and Europe. Over a hundred years ago, archaeologists filled museum cases with stone 
tools from New Jersey and Ohio and Washington, D.C., which, because they looked like 
and appeared to have been made in a similar way to European Paleolithic handaxes, were 
argued to be the same age and historically related (Abbott 1889; Wright 1892; summa- 
rized in Meltzer 1991). Thus, the American Paleolithic was born, and scholars crafted 
scenarios for moving Stone Age Europeans across the Atlantic (where, in the under- 
standing of the times, they were considered ancestors of the modern "Esquimaux") .-Yet, 
the American Paleolithic proved short-Uved: Critics pointed out the similarities between 
the two continents' artifacts were so general as to have Uttie meaning (Holmes 1892): 
these were different kinds of artifacts, with different functions (the American forms 
being merely rejected unfinished artifacts, while those from Europe were finished tools). 
Any similarities between them, it was argued, could have easily resulted independentiy, a 
convergence in the way in which the artifacts were made, or perhaps nothing more than 
random chance. 

The idea of a direct European ancestry for the first Americans disappeared quietly 
by the mrn of the cenmry, resurfaced briefly in the 1960s {e.g., Greenman 1963), and is 
now once again in play. In its current presentation, the details have obviously changed 
(we know a great deal more about the prehistories of both continents), but arguably the 



structure of the linking argument has not: It is still based on identifying a number of pre- 
sumed similarities in the artifact assemblages and technologies across the ocean. Yet, 
Solutrean and Clovis assemblages are really very different — in form, technologies, and 
materials — as Solutrean expert Lawrence Straus and others have argued (Straus 2000; 
also Sellet 1998). Most obviousty, Solutrean points lack the fluting diagnostic of Clovis, 
and their assemblages include stone artifacts and bone tools never found on these shores 
(Sellet 1998; Straus 2000). Where there are similarities, we cannot eliminate the possibil- 
it)' of convergence. After all, red ochre has been used throughout prehistory wherever it 
occurs, for there are few other natural paints so easy to use, and none so richly symbol- 
ic as this blood-red mineral. For that matter, using high-quaHty exotic raw materials is 
characteristic of many highly mobile groups, not just Solutrean and Clovis (see, for 
example, Soffer 1985). 

Even if the similarities were more pronounced and less superficial, we still face the 
very formidable problem that the Solutrean and Clovis cultures are not only separated 
by thousands of miles of space (most of which is ocean), they are also separated by 
roughly 5000 years in time: Solutrean ended -16,500 years ago, while the earliest OXo^'v:, 
site (the Aubrey site, in Texas [Ferring 1994]) is only 11,500 years old.^^ ^\^^ purported 
pre-Clovis age sites might bring us closer to Solutrean times, but the known or suspect- 
ed pre-Clovis lithic assemblages — like those at Cactus Hill or Meadowcroft, for exam- 
ple — are even less like Solutrean than Clovis assemblages. And while negative evidence 
tends not to last in archaeology, it is appropriate to add that there is no evidence 
Solutrean groups had boats, a maritime adaptation, or were living far enough to the north 
in Europe to put them in a geographic position to skirt the pack ice of the North 
Atlantic (Straus 2000). For that matter, no Solutrean points have ever been found in 
North America, let alone in New England where one might expect initial landfall (and it 
pushes the bounds of credulit)' to suggest that any or all Solutrean points brought to 
America were lost on the now-submerged Pleistocene coast). I leave it to others in this 
volume to comment on the genetic, linguistic, and dental evidence that addresses the ori- 
gins of Native American populations (see, for example, the papers in this volume by 
Merriwether, Nichols, and Turner). 

At the moment, then, the archaeological evidence by itself provides no support for 
the idea that ancestral Clovis groups originate in Upper Paleolithic Europe. Now, all of 
this might represent a minor skirmish safely confined to archaeology, were it not for the 
fact that the last several years have also seen a number of human skeletal forms appear 
in the New World wliich were labeled (at least initially), as "Caucasoid," with the strong 
implication they represent a migratory pulse from Europe to America. It is now appar- 
ent the most famous (or infamous) of those forms, the Kennewick skeleton, is not 



"Caucasoid" at all, but an ancestral Native American with affinities to Asian populations 
(PoweU & Rose 1999; also Chatters 2000). 

Yet, if Kennewick is not a "Caucasoid," and Solutrean seafarers were not in America, 
why do we have individuals that look apparently so unlike contemporary Native 
American peoples — so much so that we take them to represent non-Asian popula- 
tions?^^ Let me tie the question back to the earlier discussion of coloni2ation, for view- 
ing the Clovis radiation as I have suggested it might be viewed archaeologically has some 
interesting implications for our understanding of their biology, as well. 

Recall that colonization in its early stages demanded the maintenance of long dis- 
tance linkages to insure a sufficient breeding pool. Yet, once the descendant groups had 
begun to settle into different areas and develop their own adaptations and historical tra- 
jectories, the very vastness of the North American continent and its topographic and 
geographic barriers (glacial ice, major river systems, mountains), conspired to impede 
interaction. In the millennia following Clovis we no longer see archaeological evidence 
for far-reaching linkages across North America, and I think we can assume there wasn't 
much fo'o/<9g;W interaction or gene flow across this vast range either. Evidently, the local 
populations were relatively isolated, and even though continent-wide population density 
was low, it was still sufficientiy large that isolated groups were no longer in demograph- 
ic danger, and hence did not have to maintain long-distance mating patterns. 

Obviously, there would be genetic and phenot^'pic consequences of the reproductive 
isolation of these groups, as the spatial scale of the effective gene pool shrank. If, as my 
colleague geneticist Mike Hammer has suggested (personal communication), there were 
local founder effects accompanying this post-Clovis divergence, and if these groups were 
reproductively isolated for relatively long periods of time, then it's really no surprise why 
apparently anomalous human skeletons like Kennewick look anomalous, and unlike 
modern Native Americans. 

Of course, following nine thousand years of population growth and evolution 
among North American peoples, and the re-establishment of gene flow among distant 
groups (and thus an expansion of the effective gene pool), it's not surprising that a dis- 
tinctive Native American form will evolve, and that it would be unlike Kennewick or 
some other ostensibly "anomalous" skeletal forms. But that doesn't mean the various 
forms are unrelated, or that we have to look an)^where besides far northeast Asia for an 
ancestor of these early individuals or, for that matter, of their archaeological culture. 

So where do we go from here? We must continue to explore in the theoretical realm, 
with testing against the hard anvil of the empirical record, our ideas about range expan- 
sion across new lands. Can we determine approximate population growth rates of colo- 
nizing populations? What are the factors that will control those rates, and will they vary 



across the colonizing space, and through time? What social mechanisms (besides those 
noted above), may have been in place for maintaining long-distance relations and demo- 
graphic viability among colonizers, and did those mechanisms change depending on the 
structure of the environment {e.g., tropical forest versus temperate grassland)? If large 
social networks were in place, what were the effects on language, genes, and material cul- 
ture, and are those reflected in descendant populations? (This raises the questions of 
whether and how we might ultimately reconcile data from linguistics, genetics, and 

We need to investigate how rapidly hunter-gatherer groups could (did) map onto an 
empty and unknown landscape. How fast does resource knowledge accumulate, and is 
the scale in years, generations, or something beyond? To what degree is this learning 
process driven (or otherwise influenced) by population dynamics (population numbers 
and densit}^) and social mechanisms (including degree of dispersion or nucleation of 
groups)? At what temporal scale (beyond the life span of an individual) can human 
groups effectively monitor and store information about environments, environmental 
change, and adaptive responses? If information storage is on the scale of several gener- 
ations (e.g., upwards of a cenmry), what consequences will this have for the scale of 
adaptation? Will it shift from shorter-term, small resource flucmations to longer-term, 
larger resource cycles (Whallon 1989:443-444)? And how can this be modeled or seen 
archaeologically in a colonizing population? 

We must better understand how rapid climatic and environmental changes were at the 
end of the Pleistocene, and whether these would have been detectable on a human scale. 
If the climatic and ecological changes of the Late Pleistocene were detectable to human 
foragers, what impact, if any, did these have on the tempo and spatial patterning of col- 
onization? What may have speeded or slowed movement through particular environ- 
mental zones? 

A more basic issue: We must determine whether the spatial and temporal resolution 
of the archaeological record and our chronological methods are sufficient to see whether 
the colonization process moved in a geographically patterned way across the continent, 
and whether that process was driven by climatic or environmental change. Is the tempo- 
ral resolution of the archaeological and paleoenvironmental records and our chronolog- 
ical methods sufficient to the task? Can we differentiate a colonization process that was 
gradual, from one of relatively long periods of stasis within habitats, interspersed with 
rapid movement between habitats? Or do these happen in the "wink of a radiocarbon 
eye," and would not be detectable twelve thousand years later? 

Some thought must be devoted to what we are looking at when we are looking at arti- 
fact forms (see endnote 10). Are there changes in artifacts and material culture that can 



help us track the flow of and see the adaptive changes taking place within colonizing 
populations across space and through time? Are these units of material culture always 
(ever?) isomorpliic with social or linguistic units (let alone genetic ones?). Just what do 
they mean in terms of adaptation, social networks, and territoriality^? 

Finally, let's not forget the perennial questions: Where in Asia did these first 
Americans came from, and why have their traces so far eluded us, and what does that tell 
us about our search strategies or — possibly — our abilit)^ to tease out historical affini- 
ties from artifacts? Just how early we might be able to push back their entry into the New 
World? Did they come by land or by sea, or both — when? how? and how often? Why 
is there so little evidence of their pre-Clovis presence in North America, and what does 
that say about the archaeological record, or of our techniques for site discovery? 

These are interesting times in the study of the peopUng of the Americas. Much of 
what we knew, or thought we knew, has been turned on its head in the last few years. In 
the scramble to right ourselves many ideas — some controversial, others outlandish — 
are being tried on for size. Even in this paper. It's a natural course of events in scientif- 
ic change, and no cause for alarm. Yet. There is still much to learn. 

A-bs tract 

While much attention has been focused on when the first Americans arrived, 
an equally interesting (and perhaps equally controversial) question pertains to 
hoip they colonized the continent, l^eaving aside the compelling evidence from 
South America of an early entry into the New World (Dillehay's Monte Verde 
ivork), it remains the case that in North America the earliest accepted evidence is 
of Clovis remains (~ 1 1 ,500 years old). And that evidence points to a radiation 
across the length and breadth of North America in an astonishingly brief time. 
If Clovis groups were the first highly successful colonisers of North America, 
and if no one else was home when they arrived (and there are caveats to accepting 
both these notions), then how is it they moved so far and so fast over what became 
an increasingly exotic and unfamiliar New World? How did they map onto neiv 
resources, cope with novel pathogens and diseases, and locate and replenish vital 
supplies of high-quality stone and ivater? And, how effectively they were able to 
maintain their population si^e and reproductive viability while spread thinly on 
the landscape? We have plenty of ansrvers to these questions. Unfortunately, we 
cannot figure out which of the answers are right. But ive can at the very least 
explore some of the critical issues related to coloni':(ing efnpty continents. 




1 would like to thank my old graduate school friend and Tom Robbins' heroine, Dr. 
Nina Jablonski, for the opportunity to participate in the Wattis Symposium; Nancy Gee 
for so ably handling the logistics of the visit; and, of course, Mrs. Phyllis Wattis for mak- 
ing it all possible. My thinking on the issues raised here have benefited from conversa- 
tions with and comments from Drs. James Adovasio, Lewis R. Binford, Michael B. 
CoUins, Donald K. Grayson, Michael Hammer, Daniel Moerman, and Lawrence G. 
Straus. This paper was reviewed by several individuals, at least two of whom (Drs. David 
Anderson and Stuart Fiedel) identified themselves. I thank all for their comments and 
suggestions, only some of which could be incorporated here. Which means I'm on my 
own, blame-wise, for this one. 


^ In preparing my paper for this volume, I had the (self-imposed) choice of either 
leaving it in the conversational form in which it was prepared for the original oral pre- 
sentation, or extensively re -writing it to make it more staid and scholarly. Unable to over- 
come my own slothfulness, I opted for the former, and did little more than add a few 
sections and footnotes here and there, and insert some appropriate references. This 
paper is not intended to be an exhaustive review of the literatare. My apologies in 
advance to any colleagues who see this as a breach of scholarly decorum. 

2 All ages given in this paper are in radiocarbon years before present ("present" by 
convention being defined as 1950). While techniques are now available for calibrating 
radiocarbon ages on a calendrical scale, I will not do so, for several reasons: First, the cal- 
ibrations for this period of the Late Pleistocene are still very much in flux. Second, keep- 
ing the ages in radiocarbon years before present makes them comparable and compati- 
ble with the vast majority of the literamre that interested readers may choose to pursue. 
Finally, I am only speaking of ages in general, so calibration would be superfluous. 

^ Let me stress that this does not contradict the possibility initial peopling of North 
America took place in pre-Clovis times: Monte Verde could testify to a colonizing pulse 
that moved down the Pacific coast of North America without mrning inland, effective- 
ly leaving no archaeological footprints on this continent (but see Surovell 2000a). Or, 
Clovis might be derived from a pre-Clovis presence that had only worked its way into a 
corner of the continent — say, eastern North America (as suggested by Mason 1962; 
also, Meltzer 1988). Finally, it could represent, as the traditional notion has it, an imme- 
diate rapid postglacial arrival down the newly-opened ice-free corridor. Regardless, the 



tight radiocarbon ages on Clovis, and the vast range it covered, speaks of a rapid and 
unhindered radiation. 

^ We have an ocean-going analogy: Once groups in Near Oceania (New Guinea, 
Bismarck Archipelago, Solomon Islands) developed the technology and knowledge to 
navigate vast stretches of empt}' ocean, say ~3500 yr B.P., they moved with surprising 
speed across the distant islands of the Pacific (Irwin 1992; Kirch 1997). While there is 
much to be learned from comparing such cases, this radiation differed in many funda- 
mental respects from that of hunter-gatherers traversing new continents. As the question 
inevitably arises, I should also add that this Pacific expansion has nothing to do with the 
original peopling of the Americas. These Pacific peoples do not appear remotely close 
to South America until weU after that continent was occupied. I remain profoundly skep- 
tical that there were any earlier contacts across the ocean; for that matter, I do not see 
any significant contact between peoples of the Pacific and those of the America in pre- 
historic times. 

^ The Thule case is of interest as a Clovis analogy for another reason: By the time 
they reached central and eastern Canada, they were coming into territory previously 
inhabited by at least one branch of descendants of those Paleo-Eskimo groups. How the 
populations interacted is not known, but is of considerable interest. 

^ We do know from Greenland ice cores and ocean sediments that cUmatic change 
can be rapid and significant — with atmospheric temperature changes of as much as 20° 
— 25°C in less than 50 years in terminal Pleistocene and Younger Dryas times over the 
Greenland ice sheet, for example (Alley 2000; White 2000). But how that change was 
manifest over continents in terrestrial climates, plant, and animal communities is not 
known. Wliile there is evidence that change can be rapid in short-Uved, temperature sen- 
sitive, lake-dwelUng microorganisms {e.g., Birks & Ammann 2000), one should not gen- 
eralize those patterns to the kinds of plants and small and large game resources exploit- 
ed by humans. 

^ In a paper that appeared after this was written, Surovell (2000b) provided a model 
which showed that by using certain strategies for moving about the landscapes, 
hunter/gatherer groups can travel great distances rapidly, yet stiU maintain high fertility 
and population growth rates, and do so without suffering demographic consequences. 
The model is elegant and downright intriguing, but whether it is applicable to the Clovis 
situation will reqiiire construction of a far better bridge between it and the data than we 
now possess. 

° Moerman (personal communication, 1 999) makes the important observation, how- 
ever, that in folk taxonomies plants are more often identified at the genus, rather than 
the species, level. For highly mobile peoples, it is more adaptive to be able to recognize 



generic plants, than particular species which may have much more restricted distribu- 

9 Caches will appear at other times and places much later in prehistory, but under 
conditions far different from, and presumably for different reasons than, Paleoindian 

^0 The circles and ovals on this and subsequent distribution maps are intentionally 
highly generalized: What I wish to convey are very broad spatial patterns, not specific 
point distributions. Let me explain why. First, the data do not exist to provide, at the con- 
tinental scale, anytlxing approximating a detailed distributional map (though Anderson et 
al. [2000] are making important headway in that direction). Second, as Jack Hofman 
(1992) and others {e.g., LeTourneau 1998; O'Brien & Lyman 2000) have rightiy argued, 
there is considerable variability — spatial, temporal, stylistic, technological, etc. — in 
these point forms. Unfortunately, that variabilit}^ is not fully appreciated or understood 
or incorporated in our analytical classifications (Anderson 1995:20). Drawing boundaries 
(be they classificatory or cartographic) around point forms is an arbitrary exercise in any 
case. Thus, rather than give the false impression, by highly detailed maps, that we have 
capmred a set of well-defined and distinctive forms in time/space, I simply drew circles 
and ovals around modal tendencies, with the explicit caveat that these are, at best, circles 
and ovals around modal tendencies: no more, and no less. Without question, there is 
overlap among the forms, and occurrences outside the boundaries show. AU of which 
highlights the value of gaining better control on the many dimensions of variabilit)^ of 
these point forms, for when such data are ultimately available, it will provide important 
insights into Paleoindian artifact variabilit); distribution, and so on. 

^^ This settling in process might also be manifest in the well-documented and appar- 
ently abrupt increase in the use of cave and rockshelter sites in eastern North America 
in Late Paleoindian period — there being a singular lack of such use in Clovis times 
(Walthall 1998). It was later Paleoindian groups who, having mapped their landscape, 
found those locations, shifted their adaptations toward more local resources, and became 
more territorial, could use such fixed and permanent locales as important components 
of their settiement strategy. Perhaps. But one must always be wary in archaeology of 
placing too much weight on negative data: It often doesn't last. It may only reflect 
methodological shortcomings: Have we dug such sites in the right way and thoroughly 
tested for Clovis age components? Collins (1991) gives compelling reasons to think not; 
and even if true, is it necessarily significant? 

^2 Meadowcroft has cheated archaeology's actuarial tables (Meltzer 1993), and 
remains a viable pre-Clovis candidate long after its initial appearance on the scene. Lately, 
one of the key objections to the site's antiquit)' — that the radiocarbon ages were cont- 



aminated by groundwater seeping through the lower deposits on site — was effectively 
rebutted by micromorphological analyses of the sediments (Goldberg & Arpin 1999). 
Cactus Hill has been more recendy reported. There is litde question but that Cactus Hill 
has yielded distinctive artifacts in strata below a Clovis level. That this assemblage bears 
more than a passing resemblance to the early material from Meadowcroft makes it aU the 
more intriguing. However, there is uncertainty about the absolute age of those pre-Clovis 
materials. Certainly they appear pre-Clovis in a relative sense (below Clovis), but how 
much earlier in absolute time seems an open question, given the range (from Pleistocene 
to modern) of radiocarbon ages from that level. It remains to be seen how "pre" the pre- 
Clovis at Cactus Hill will turn out to be. 

^^ It is with some reluctance that I even raise these issues here, but do so for two rea- 
sons. First, the idea of a Late Pleistocene trans-Atlantic crossing has already gotten con- 
siderable media play, and too often (the media being the media) it has been presented as 
highly likely, without any consideration of the possible obstacles to such a hypothesis. 
Those need to be raised, to put the discussion in a larger context. Second, at the panel 
discussion at the end of our Wattis symposium, I was directly asked to speak to this issue, 
in light of my comments on Clovis. This is the long-delayed answer to that question. 

'^^ As there appears to be some confusion on the matter, let me state here that the 
Aubrey site does indeed have an unequivocal Clovis fluted point. 

^^ I would also ask this: Should we necessarily expect skeletal forms of this antiqui- 
ty to look like modern Native Americans, even if they are Native American ancestors? 
Perhaps not. Over 9000 years separate these early forms and modern Native Americans. 
One can imagine a fair amount of evolutionary change in the relatively plastic morpho- 
logical attributes of the skull over that time. 

l^iterature Cited 

Abbott, C. C. 1889. Evidences of the antiquit}^ of man in eastern North America. Proc. 

Am. Assoc. Adv. Sci. 37:293-315. 
Adovasio, J. M., J. Donahue, ]., & R. Stuckenrath. 1990. The Meadowcroft Rockshelter 

radiocarbon chronology 1975-1990. Am. Antiquity 55:348-354. 
AUey, R. B. 2000. The Younger Dryas cold interval as viewed from central Greenland. 

Quat Sci. Reu 19:213-226. 
Anderson, D. 1990. The Paleoindian colonization of eastern North America: a view 

from the southeastern United States. R^j-. Econ. AnthropoL, Suppl. 5:163—216. 
. 1995. Paleoindian interaction networks in the eastern woodlands. Pages 1—26 

in M. Nassaney & K. Sassaman, eds.. Native American Interaction: Multiscalar Analyses and 

Interpretations in the Eastern Woodlands. Universit)^ of Tennessee Press, KnoxviUe, TN. 



Anderson, D, & M. Faught. 2000. Paleoindian artifact distribution: Evidence and impli- 
cations. ^«/i^^//y 74:507— 513. 
Anderson, D, & K. Sassaman. 1996. The Paleoindian and Uarlj Archaic Southeast. University 

of Alabama Press, Tuscaloosa, AL. 526 pp. 
Baker, R. R. 1978. The Evolutionary Ecology of Animal Migration. Holmes and Meier, New 

York, NY. 1012 pp. 
Beaton, J. 1991. Colonizing continents: some problems from Australia and the Americas. 

Pages 209-230 in T. D. Dillehey & D J. Meltzer. cds.J'he First Americans: Search and 

Research. CRC Press, Baton Rouge, LA. 
Belovsky, G., C. MelUson, C. Larson, & P. A. Van Zandt. 1999. Experimental studies of 

extinction dynamics. S cience 2^6: 1175—1177. 
Birks, H., & B. Ammann. 2000. Two terrestrial records of rapid climatic change during 

the glacial-Holocene transition (14,000-9,000 calendar years B.P.) from Europe. Proc. 

Natl Acad Set. 97:1390-1394. 
Boldurian, A., & J. L. Cotter. 1999. Clovis Revisited. Nem Perspectives on Paleoindian 

Adaptations from Blackivater Draw, New Mexico. University of Pennsylvania Press, 

Philadelphia, PA. 145 pp. 
Bonnichsen, R. 1991. Clovis origins. Pages 309-329 in R. Bonnichsen & K. Turnmire, 

eds., Clovis: Origins and Adaptations. Center for the Study of the First Americans, 

Oregon State University, Corvallis, OR. 
Borgerhoff Mulder, M. 1992. Reproductive decisions. Pages 339-374 in E. Smith & B. 

Winterhalder, eds.. Evolutionary Ecology and Human Behavior Aldine de Gruyter, New 

York, NY 
Chatters, J. 2000. The recovery and first analysis of an early Holocene human skeleton 

from Kennewick, Washington. Am. Antiquity 65:291—316. 
Collins, M. B. 1991. Rockshelters and the early archaeological record in the Americas. 

Pages 1 57-1 82 in T Dillehay & D. Meltzer, eds.,T^^ First Americans: Search and Research. 

CRC Press, Boca Raton, FL. 
. 1999. Clovis Blade Technology: A Comparative Study of the Keven Davis Cache, Texas. 

University of Texas Press, Austin, TX. 234 pp. 
Davis, W M. 1926. The value of outrageous geological hypotheses. Science 63:463-468. 
Derev'anko, A. P., ed. 1998. The Paleolithic of Siberia: Neip Discoveries and Interpretations. 

University of Illinois Press, Urbana, IL. 406 pp. 
Dillehay, T. 1989. Monte Verde: A Tate Pleistocene Settlement in Chile, vol. 1. Palaeoenvironment 

and Site Ccontext. Smithsonian Instimtion Press, Washington, DC. 
. 1997. Monte Verde: A Tate Pleistocene Settlement in Chile, vol. 2. The Archaeological 

Context and Interpretation. Smithsonian Institution Press, Washington, DC. 



Dillehay, T., G. Calderon, G. Politis, & M. Beltrao. 1992. Earliest hunters and gatherers 

of South America. 7. World Prebist. 6:145-204. 
Dumond, D. 1984. Prehistory: Summary. Pages 72—79 in D. Damas, ed., Handbook of 

North A.mericc!n Indians, vol. 5, Arctic. Smithsonian Institution Press, Wasliington, D.C. 
Edwards, R., |. Beck, G. Burr, D. Donahue, J. Chappell, A. Bloom, E. Druffel, & F. 

Taylor. 1993. A large drop in atmospheric 14 C/12 C and reduced melting in the 

Younger Dryas, documented with 230Th ages of corals. Science 260:962—968 
FAUNMAP working group. 1996. Spatial response of mammals to Late Quaternary 

environmental fluctuations. Science 272:1601—1606. 
Ferring, C. R. 1994. The role of geoarchaeology in Paleoindian research. Pages 57—72 in 

R. Bonnichsen & D. G Steele, eds., Method and Theory for Investigating the Peopling of the 

Americas. Center for the Study of the First Americans, Oregon State University, 

Corvallis, OR. 
Frison, G 1991. Prehistoric Hunters of the High Plains. Second edition. Academic Press, New 

York, NY 532 pp. 
Goebel, T, & M. Aksenov. 1995. Accelerator radiocarbon dating of the initial Upper 

Paleolithic in southeast Siberia. Antiquity 69:349—357. 
Goebel, T, R. Powers, & N. Bigelow. 1991. The Nenana Complex of Alaska and Clovis 

origins. Pages 49—79 /// R. Bonnichsen & K. Turnmire, eds., Clovis: Origins and 

Adaptations. Center for the Study of the First Americans, Corvallis, OR. 
Goldberg, P., & T. Arpin. 1999. Micromorphological analysis of sediments from 

Meadowcroft Rockshelter, Pennsylvania: Implications for radiocarbon dating. /. Field 

Archaeology 26:325—342. 
Goodyear, A. 1979. A H)^othesis for the Use of Cryptocrystalline Raw Materials 

Among Paleoindian Groups of North America. Research Manuscript Series 156. 

South Caroline Institute of Archaeology and Anthropology, University of South 

Carolina. 9 pp. 
Greenman, E. 1963. The Upper PaleoUthic and the New World. Curr. Anthropol 4:41—66. 
Haynes, C. V. 1964. Fluted points: Their age and dispersion. Science 145:1408—1413. 

. 1971. Time, environment, and early man. Arctic Anthropol. 7:3—14. 

. 1982. Were Clovis progenitors in Beringia? Pages 383—398 in D. Hopkins, ]. 

Matthews, C. Schweger, & S. Young, eds., Paleoecology of Beringia. Academic Press, New 

York, NY. 

1992. Contributions of radiocarbon dating to the geochronology of the peo- 

pling of the New World. Pages 355-374 in R. Taylor, A. Long, & R. Kra, eds., Radio 
After Four Decades: An Interdisciplinary Perspective. Springer- Verlag, New York, NY. 



Hoffecker, J. 1996. Introduction to the archaeology of Beringia. Pages 149-153 in R 

West, ed., A^mencan Beginnings: The Prehistory and Palaeoecology of Beringia. University of 

Chicago Press, Chicago, IL. 
Hoffecker, J., W R. Powers, & T Goebel. 1993. The colonization of Beringia and the 

peopling of the New World. Science 259:46—53. 

Hofman, J. 1992. Recognition and interpretation of Folsom technological variability on 

the Southern Plains. Pages 193-224 in D Stanford & J. Day, eds., Ice Age Hunters of 

the Rockies. Denver Musuem of Natural History, Denver, CO. 

Holmes, W. H. 1892. Modern quarry refuse and the Paleolithic theory. Science 20:295-297. 

Howard, E. B. 1936. The occurrence of flints and extinct animals in pluvial deposits near 

Clovis, New Mexico, Part 1: Introduction. Proc. Phila. Acad. Nat Sci. 87:299-303. 
Hughen, K., J. Overpeck, S. Lehman, M. Kashagarian, J. Southon, L. Peterson, R. AUey, 

& D Sigman. 1998. Deglacial changes in ocean circulation from an extended radio- 
carbon calibration. Nature 391:65—68. 
Irwin, G. 1992. The Prehistoric Exploration and Colonisation of the Pacific. Cambridge 

Universit}'- Press, Cambridge, MA. 240 pp. 
Jackson, L. E. & A. Duk-Rodin. 1996. Quaternary geology of the ice-free corridor: 

glacial controls on the peopling of the New World. Pages 214—227 /;/ T Akazawa & 

E. Szathmary, eds.. Prehistoric Mongoloid Dispersals. Oxford University Press, Oxford. 
Johnson, A. R., J. Wiens, B. Milne, & T. Crist. 1992. Animal movements and population 

dynamics in heterogeneous landscapes, landscape Ecol 7:63—75. 
Johnson, E. 1991. Late Pleistocene cultural occupation on the Southern High Plains. 

Pages. 215-236 in R. Bonnichsen & K. Turnmire, eds., Clovis: Origins and Adaptations. 

Center for the Study of the First Americans, CorvaUis, OR. 
Kaplan, H., & K. HiU. 1992. The evolutionary ecology of food acquisition. Pages 

167-201 in E. Smith & B. Winterhalder, eds., Evolutionary Ecology and Human Behavior. 

Aldine de Gruyter, New York, NY. 
Kelly, R. L. 1995. The Foraging Spectrum: Diversity in Hunter-gatherer Eifeivajs. Smithsonian 

Institution Press, Washington, DC. 446 pp. 
Kelly, R. L., & L. Todd. 1988. Coming into the country: Early Paleoindian hunting and 

mobility. Am. Antiquity 53:231—244. 
King, M. L., & S. Slobodin. 1996. A fluted point from the Uptar site, northeastern 

Siberia. Science 273:634—636. 
Kirch, P. V. 1988. Long-distance exchange and island colonization: The Lapita case. 

Nonvegian ArchaeoL Rev. 21:103—117. 
. 1997. The Lapita Peoples: Ancestors of the Ocean World. Blackwell Publishers, 

Cambridge, MA. 353 pp. 



Leakey, L. S., R. Simpson, & T. Clements. 1968. Archaeological excavations in the Calico 

Mountains, California: Preliminary report. Science 160:1022—1023. 
LeTournau, P. 1998. The "Folsom problem." Pages 52—73 in A. Ramenofsky & A. 

Steffen, eds.. Unit Issues in Archaeology. University of Utah Press, Salt Lake Cit)'^, UT. 
Lourandos, N. 1997. Continent of Hunter-gatherers: NeiP Perspectives in Australian Prehistorj. 

Cambridge University Press, Cambridge, MA. 390 pp. 
Lynch, T. 1983. The Paleo-indians. Pages 86—137 in]. D. Jennings, ed., Ancient South 

Americans. WH. Freeman, San Francisco, CA. 
MacArthur, R. H., & E. O. Wilson 1967. The Theory of Island Biogeographj. Monographs in 

Population Biology 1. Princeton University Press, Princeton, NJ. 203 pp. 
MacDonald, D., & B. S. Hewlett. 1999. Reproductive interests and forager mobility. Curr. 

Anthropol 40:501-523. 
MacNiesh. R. 1976. Early man in the New World. Am. Sci. 63:316-327. 
Mandryk, C. A., H. Josenhans, D. Fedje, & R. Mathewes. 2001. Late Quaternary pale- 

oenvironments of northwestern North America: Implications for inland versus 

coastal migration routes, ^^/i^/. Sci. Kev. 20:301—314. 
Martin, P. S. 1973. The discovery of America. Science 179:969-974. 

. 1987. Clovisia the beautiful. Nat Hist. 96:10-13. 

Martin, P. S., & R.G. Klein, eds. 19M. Quaternafy Extinctions. University of Arizona Press, 

Tucson, AZ. 892 pp. 
Mason, R. J. 1962. The Paleo-Indian tradition in eastern North America. Cum Anthropol 

MacAvoy, J., J. Baker, J. Feathers, R. Hodges, L. McWeeney, & T Whyte. 2000. Summary 

of research at the Cacms Hill archaeological site, 44SX202, Sussex County, VA. Report 

to the National Geographic Society in compliance with stipulations of Grant # 6345—98. 
McGhee, R. 1984. Thule prehistory of Canada. Pages 369—376 in D. Damas, ed.. 

Handbook of North American Indians, vol. 5, Arctic. Smithsonian Institation Press, 

Washington, D.C. 

. 1998. The First Peopling of Arctic North America. Plenary Address, 31 st Annual 

Chacmool Conference, "On Being First: Cultural Innovation and Environmental Consequences of 
First Peoplings. " University of Calgary, Alberta, Canada. 

Meltzer, D. J. 1988. Late Pleistocene human adaptations in eastern North America. /. 
World Prehist 2:1-52. 

. 1989. Was stone exchanged among eastern North American Paleoindians? 

Pages 11—39 in C. EUis & J. Lothrop, eds., Eastern Paleo-Indian Eithic Resource 
Procurement and Processing. Westview Press, Boulder, CO. 

. 1991. On "paradigms" and "paradigm bias" in controversies over human antiq- 



uity in America. Pages 13^9 in T. D Dillehey & D J. Meltzer, eds., The First 
Atnericans: Search and Kesearch. CRC Press, Baton Rouge, LA. 
. 1993a. Search for the First Americans. Smithsonian Books, Washington, DC. 

176 pp. 
. 1993b. Is there a Clovis adaptation? Pages 293-310 in O. Soffer & N. Praslov, 

eds.. From Kostenki to Clovis: Upper Paleolithic — Paleo-indian adaptations. Plenum Press, 

New York, NY. 

. 1995. Clocking the first Americans. Annu. Ret>. Anthropol 24:21-45. 

. 1997. Monte Verde and the Pleistocene peopling of the Americas. Science 


1998. Why we still don't know when the first people came to North America. 

Presented at the 31st Annual Chacmool Conference, "On Being First: Cultural 
Innovations and Environmental Consequences of First Peop lings.'" Calgary Archaeology 
Association, Calgary, Alberta, Canada. 

1999. Comment on "Reproductive interests and forager mobilit}^," by D 

MacDonald & B. S. Hewlett. Cum Anthropol 40:519-520. 
Morell, V. 1995. Ancestral passions: The Leakey Family and the Quest for Humankind's 

Beginnings. Simon & Schuster, New York, NY. 638 pp. 
Morlan, R. E. 1988. Pre-Clovis people: Early discoveries of America? Ethnology 

Monographs 12:31—43. 
. 1991. Peopling of the New World: A discussion. Pages 303-307 in R. 

Bonnichsen & K. Turnmire, Clovis: Origins and Adaptations. Center for the Smdy of the 

First Americans, Corvallis, OR. 
Morrow, J. E., & T. Morrow. 1999. Geographic variation in fluted projectile points: A 

hemispheric perspective. Am. Antiquity 64:215—231. 
Nichols, J. 1990. Linguistic diversit}' and the first settiement of the New World. Language 

O'Brien, M. J., & R. L. Lyman. 2000. Applying Evolutionary Archaeology: A Systematic 

Approach. Plenum Press, New York, NY. 471 pp. 
Powell, J., & J. Rose. 1999. Report on the osteological assessment of the 

"Kennewick Man" skeleton (CENWW.97.Kennewick). Internet citation: kennewick/powell_rose.htm. 
Reanier, R. 1995. The antiquity of Paleoindian materials in northern Alaska. Arctic 

Anthropol 32:31-50. 
Sellet, F. 1998. The French Connection: Investigating a possible Clovis-Solutrean link. 

Current ^search in the Pleistocene 15:67—68. 
Soffer, O. 1985. The Upper Paleolithic of the Central Russian Plain. Academic Press, Orlando. 
FL. 539 pp. 



Stanford, D. J. 1991. Clovis origins and adaptations: an introductory perspective. Pages 

1—13 in R. Bonnichsen & K. Turnmire, eds., Clovis: Origins and Adaptations. Center for 

the Study of the First Americans, CorvaUis, OR. 
Stephens, D., & J. Krebs. 1986. Foraging Theory. Princeton Universit)' Press, Princeton, NJ. 

247 pp. 
Storck, P. 1991. Imperialists without a state: The cultural dynamics of early Paleoindian 

colonization as seen from the Great Lakes region. Pages 153—162 /;/ R. Bonnichsen 

& K. Turnmire, eds., Clovis: Origins and Adaptations. Center for the Study of the First 

Americans, CorvaUis, OR. 
Straus, L. 2000. Solutrean settiement of North America? A review of reality. Am. 

Antiquity 65:219—226. 
Surovell, T. 2000a. Can a coastal migration explain Monte Verde? Paper presented at the 

65th Annual Meeting, Society for American Archaeology, Philadelphia, PA. 
. 2000b. Early Paleoindian women, children, mobility, and fertility. Am. Antiquity 

Taylor, E., C. V. Haynes, & M. Stuiver. 1996. Clovis and Folsom ages estimates: 

Stratigraphic context and radiocarbon calibration. Antiquity 70:515—525. 
Torroni, A., T Schurr, M. Cabell, M. Brown, J. Neel, M. Larson, D. Smith, C. Vullo, & 

D. Wallace. 1993. Asian affinities and continental radiation of the four founding 

Native American mtDNA. Am. J. Hum. Genet 53:563—590. 
Van Dyne, G. M., N. Brockington, Z. Szocs, J. Duek, & C. Ribic. 1980. Large herbivore 

system. Pages 269—537 in A. Breymeyer & G. M. Van Dyne, eds.. Grasslands, Systems 

Analysis, and Man. Cambridge University Press, Cambridge, MA. 
Walthall, J. 1998. Rockshelters and hunter-gatherer adaptation to the 

Pleistocene/Holocene transition. Am. Antiquity 63:223—238. 
WhaUon, R. 1989. Elements of culmre change in the Later Palaeolithic. Pages 433-454 

in P. Mellars & C. Stringer, eds.. The Human Evolution: Behavioral and Biological 

Perspectives on the Origins of Modern Humans. Priceton Universit}^ Press, Princeton, NJ. 
White, J. 2000. Ice core records of rapid and abrupt climate change: data and theory. 

Keynote Address, Climate change and human responses. Fort Burgwin Research 

Center, New Mexico. 
Whitiey, D S ., & R. Dorn. 1993. New perspectives on the Clovis vs. pre-Clovis contro- 
versy. Am. Antiquity 58:626—647. 
Wright, G. F 1892. Man and the Glacial Period D. Appleton, New York, NY. 385 pp. 
Wright, H. E. 1991. Environmental conditions for Paleoindian immigration. Pages 

1 13— 135 /';/ T. D DUlehey & D J. Meltzer, eds.. The First Americans: S search and Research. 

CRC Press, Baton Rouge, LA. 


Chapter Four 

Anatomically Modern Humans, Maritime 
Voyaging, and the Pleistocene Colonization of 

THE Americas 


Jon M. Erlandson 

During most of human history, water must have been 
a major physical and psychological harrier and the 
inability to cope with yvater is shown in the archaeo- 
logical record by the absence of remains of fish, shell- 
fish, or any object that required going deeply into 

water or using boats. There is no evidence that resources of river 

and sea were utili-:^ed until this late pre-agricultural period . . . 

for early man, ^wdter was a barrier and a danger, not a resource 

(Washburn & Lancaster 1968:294). 

For decades, most anthropologists have believed that maritime adaptations and sea- 
faring developed very late in human history and that the Americas were first colonized 
by terrestrial hunters who walked through the interior regions of Beringia and the 
Americas. Since the 1970s, perceptions about the antiquit)^ of human seafaring have 
changed dramatically, but theories about the development of maritime adaptations and 
the peopling of the New World have not kept pace. Recentiy, there has been a veritable 
sea change in our perceptions of the Pleistocene colonization of the Americas. Just a few 
years ago, there was a general consensus that the New World was colonized sometime in 
the very Late Pleistocene by land-based hunters who trekked across Beringia and down 
the fabled ice-free corridor into the heartland of America, gradually spreading from sea 
to shining sea. An alternative theory, long peripheral to the dominant paradigm, pro- 
posed that maritime peoples followed the coastlines of the North Pacific from Asia to 
the Americas — moving by land and by sea from northeast Asia, along the southern 
shores of Beringia, and down the Pacific Coast of North America. Ironically, while this 
"coastal migration theory" has gained adherents in recent years, it has become almost 
conservative compared to scholarly claims and media accounts that the Americas may 
have first been colonized by seafaring Australian Aborigines, Polynesians, or even 



Europeans who crossed the open Pacific or Atlantic oceans by boat (e.g., Dixon 1993; 
Cann 1994:10; Locke 1999:16; Toyne 1999; Stanford, this volume). 

After roughly a century of scientific investigation into how and when the Americas 
were first settied, many scholars and lay people might reasonably ask how we got from 
"Clovis-first" to the "Coastal Migration Theory" to "Almost Anything Goes" in such a 
short amount of time? The answer to that question is complex, but an important com- 
ponent of it clearly grows out of the emerging (but not unanimous) consensus that the 
so-called "Clovis barrier" has been broken by the 12,500 year old pericoastal site of 
Monte Verde in Chile (see DiUehay 1997; Meltzer ef al. 1997). Also contributing is an 
emerging body of data that suggests that the ice-free corridor developed relatively late in 
time and was probably a daunting and difficult landscape for hunter-gatherers to survive 
in {e.g., MacDonald 1987; Clague et al. 1989; Mandryk 1991; Wright 1991; Elias, this vol- 
ume). Research has also shown that the coastlines of Alaska and British Columbia 
deglaciated earlier than once thought and were more suitable for human occupation than 
previously believed [i.e., Mann 1986; Molnia 1986; Barrie et al. 1993; Heaton & Grady 
1993; Mann & Peteet 1994; Mann & Hamilton 1995; Dixon et al 1997; Josenhans et al 
1997). As these issues are discussed elsewhere, they are not the focus of my paper. My 
colleagues in this volume also summarize the genetic, biological, and linguistic evidence 
for the peopling of the New World. 

Instead, I focus on another prime reason why the idea that Pleistocene coastal migra- 
tions into the Americas may have occurred has been marginalized for decades. This is 
the fact that we have long underestimated the sophistication of our distant ancestors, the 
importance of the sea in human history, and the maritime capabilities of Late 
Pleistocene peoples. When did our ancestors first develop relatively complex technolo- 
gies? When did they adapt to the sea and first use relatively sophisticated watercraft? 
When were the Americas first settled and what do current archaeological data tell us 
about the possibility that maritime peoples may have followed the coastlines of the 
North Pacific into the New World? 

In this paper, I bring a relatively global perspective to bear on such questions by 
examining three related topics: the emergence of anatomically modern humans, the 
antiquit}^ of maritime adaptations and seafaring, and the peopling of the Americas. 
Entire books have been written on each of these complex issues, so my discussion of 
each must be relatively brief and focused on recent developments. In the process, I 
explore the maritime capabilities of Pleistocene peoples and their potential for discover- 
ing and settling the New World by sea, by following a North Pacific Rim route from 
northeast Asia to the Americas. 



A^natomically Modern Humans: A.n Emerging Consensus^ 

One of the most intriguing issues in anthropology today involves the evolution of 
anatomically modern humans {Homo sapiens sapiens). Until the late 1980s, anatomically 
modern humans (AMH) were generally believed to have evolved from discrete archaic 
Homo sapiens populations in several geographic regions of Africa, Asia, and Europe. 
These archaic humans were in turn thought to have evolved from ancestral Homo erectus 
populations who migrated out of Africa roughly a million years ago^ (Pagan 1990; Klein 
1995). Advocates of this theory of "multiregional evolution," relying primarily on mor- 
phological studies of human fossils, have proposed a relatively high antiquity {ca. 1—2 mil- 
lion years) for the diversit}^ found among living humans and evolutionary continuity 
between largely discrete lineages Homo erectus., archaic Homo sapiens, and AMH in multiple 
regions of the world. Most multiregionalists believe, for instance, that many modern 
humans of European descent carry the DNA of Neanderthal ancestors within their 
genetic inheritance (see Smith 1994). 

In the late 1980s, molecular biologists and anthropologists proposed a stunning and 
very different theory. Based on the analysis of mitochondrial DNA collected from the 
placentas of 189 living women, Cann and colleagues (1987) proposed that AMH evolved 
first in Africa ca. 150,000 to 200,000 years ago, then spread rapidly around the world 
replacing all archaic humans. Based on data essentially independent of the human fossil 
and archaeological records, this controversial "Out of Africa" (a.k.a. the Eve or 
Replacement) theory proposed just the opposite of the orthodox multiregional scenario. 
Instead of 1—2 million years to account for modern human diversit)'^, the Out of Africa 
model proposed a calendar of 200,000 years or less. Instead of the essentially indepen- 
dent evolution of AMH in multiple regions around the world, the Out of Africa theory 
proposed a complete and relatively recent replacement by a single evolutionary lineage 
of African origin. And instead of Neanderthals contributing their genes to a communal 
European inheritance. Out of Africa proponents see Neanderthals as an evolutionary 
dead end (see Stringer & McKie 1996). 

Seemingly consistent with some archaeological and paleoanthropological data, and in 
conflict with other evidence, the Out of Africa theory has deeply divided paleoanthro- 
pologists and archaeologists studying such problems. It has been criticized on numerous 
counts, from small sample size, to inappropriate statistical analyses, and invalid chrono- 
logical calibration. Among the most vociferous in this chorus of criticisms have been 
biological anthropologists who claim that the use of modern DNA to reconstruct evo- 
lutionary relationships ignores the fragmentary fossil evidence for transitional stages in 
the physical development of archaic Homo sapiens and AMH. Unfortunately for anthro- 
pologists, this last criticism recalls an earlier debate about the ancestry of a 14—17 mil- 



lion year old fossil "hominid" jaw dubbed Ramapitheciis, which molecular biologists study- 
ing the biochemistry of modern apes had the audacity to proclaim was much too old to 
be a hominid. This heated and now historical debate was setded (much to the conster- 
nation of many paleoanthropologists) only when a nearly complete KamapithecMS skull 
was discovered and found to be a fossil orangutan (see Lewin 1987). 

In the current debate about the origins of anatomically modern humans, paleoan- 
thropologists and archaeologists are themselves deeply divided, with each new discovery 
eliciting claims from one camp or the other that the new data support their theory. In 
the meantime, scientists studying human DNA — mitochondrial DNA for maternal lin- 
eages, y-chromosome DNA for paternal Uneages, nuclear DNA, and even DNA extract- 
ed from archaeological samples of Neanderthal and AMH bones — have continued to 
accumulate a compelling body of data that seems to support the Out of Africa theory. 
With much larger samples from a wider range of cultures, better analytical and statistical 
methods, and samples of Neanderthal DNA suggesting that the Neanderthals may have 
been a separate species-^ {e.g., Krings et al. 1997), an emerging consensus argues that some 
version of the Out of Africa theory is probably correct (Klein 1995; MeUars 1998). 

This consensus is supported not only by molecular data, but by some substantial 
paleoanthropological and archaeological evidence, as well. As predicted by the Out of 
Africa theory, for instance, the earUest securely dated AMH skeletal remains have been 
found in Africa, at Klasies River Mouth Caves in coastal South Africa (Singer & Wymer 
1982; Rightmire & Deacon 1991; Klein 1995). These remains, between about 120,000 
and 65,000 years old, are associated with Middle Stone Age stone tool assemblages that 
include the earUest blade technologies, the earUest microUthic geometric assemblages 
(Howieson's Poort), and some of the earliest formal bone tools (Mellars 1998). From the 
Semliki River area of Zaire, recent research also documented the earliest examples of 
complex barbed bone harpoons and intensive fishing from the Katanda sites, dated to 
roughly 90,000 to 80,000 years ago (Brooks et al. 1995; YeUen et al. 1995; Yellen 1998). 
Also dated to about 90,000 years are the earUest AMH remains found outside of Africa, 
the Qafzeh and Skhul skeletons from coastal Israel, suggesting that early modern popu- 
lations had begun to move out of Africa by this time. Curiously, anatomically modern 
humans do not appear to have moved into most of Europe for another 45,000 to 55,000 
years (Klein 1998), possibly held at bay by Neanderthal populations that occupied the 
area. Current evidence suggests, however, that AMH had spread into southern Asia by 
at least 70,000 to 60,000 years ago. From there, within just a few millennia, they proba- 
bly had made the multiple maritime voyages through island southeast Asia that are 
required to settle Sahul, the larger continent that Linked Australia, New Guinea, and 
Tasmania during glacial periods of lower sea level. 



Wherever anatomically modern humans went, at least after about 60,000 years ago, 
they also seem to have carried with them a penchant for art and symbolism, technolog- 
ical innovation, rapid population growth and geographic expansion, and complex prob- 
lem-solving and communication skills (see Davidson & Noble 1992; Klein 1998). This 
creative revolution is evident virtually worldwide during the Late Pleistocene in the 
invention of a whole series of new technologies and behaviors never seen among our 
archaic ancestors.^ These include the first chipped stone blade and microlithic technolo- 
gies, ornaments, complex art and musical instruments, ground stone tools, formal bone 
and shell technologies, ceramics, basketry, cremations, complex projectiles (harpoons, 
throwing sticks and darts, the bow and arrow), intensive fishing, and the domestication 
of plants and animals. Significantiy, as we shall see, they also include the first widespread 
evidence for the development of sophisticated boats and planned maritime voyaging. 

Thus, a convergence of molecular, archaeological, and paleoanthropological evidence 
points toward an emerging consensus — although still vigorously debated — that 
anatomically modern humans evolved in Africa roughly 1 50,000 years ago, then rapidly 
spread around the world in the last 75,000 to 100,000 years, armed with superior intel- 
lectual and technological capabilities (Klein 1995; MeUars 1998). In the process, anatom- 
ically and intellectually modern humans appear to have replaced the vast majority of 
archaic humans, although some mixing between AMH and archaic populations may have 
occurred, with occasional hybrids being swamped in a sea of rapidly expanding anatom- 
ically modern populations. Significantly, the Out of Africa model reduces the amount of 
available time to account for the demographic and geographic expansion of AMH pop- 
ulations by 90 percent or more. As I will show, maritime adaptations and seafaring played 
a key role in the geographic expansion of Ho/m sapiens sapiens, including some of the 
most dramatic migrations in human history. 

The Antiquity of Maritime Adaptations and Seafaring 

For decades, the idea that our Pleistocene ancestors may have made substantial 
migrations by boat suffered from an anthropological theory that maritime and other 
aquatic adaptations and seafaring developed very late in human history (see Washburn & 
Lancaster 1968; Osborn 1977; Bailey 1978; Waselkov 1987; Yesner 1987; and otiiers). 
This view is clearly illustrated by Yesner (1987:285), who stated that the "historical fact 
that maritime resources were not exploited until relatively late in the preliistoric record 
has attracted a general consensus. ... A real commitment to maritime lifeways did not 
precede late Upper Paleolitliic times." Through the 1970s, there was virmal unanimit}' 
that boats, too, were a relatively recent addition to the human technological repertoire 
(Bass 1972:12; Greenhill 1976; Johnstone 1980:xv). 



A.ntiquitj of Coastal A^daptations: Some Problems and Evidence 

On a planet whose surface is almost 75 percent water, where life itself is so depen- 
dent on water to survive, and where opportunistic hominids have successfully adapted 
for milli ons of years, I find it curious that most scholars believe our ancestors did not 
adapt to aquatic environments until very recently. The general perception that humans 
only systematically adapted to marine environments during the last 10,000 to 15,000 
years has long inhibited the study of maritime adaptations, coastal migrations, and boats 
(Erlandson 2001). It tends to peripheralize the significance of aquatic habitats in human 
evolution, relegating them to an incidental role in a broad-spectrum revolution leading 
to agricultural societies and civilizations. Thus, maritime adaptations appear to play a 
marginal role in an important but comparatively brief process in which human ecology 
departs from its natural course as a population explosion forced humans into increas- 
ingly artificial modes of subsistence and production (see Cohen 1977). As Yesner (1987) 
noted, however, this view of aquatic resources as marginal is out of step with historical 
and archaeological data that demonstrate that maritime hunter-gatherers were generally 
more populous and more culturally complex than their non-agricultural interior neigh- 

It is true that there is relatively littie evidence for the intensive use of marine 
resources prior to the end of the Pleistocene (see Waselkov 1987). On a global scale, 
however, it is not clear whether an apparently dramatic expansion in the use of aquatic 
resources over the last 10,000 years accurately reflects changes in human subsistence 
through time. To understand the history of maritime or other aquatic adaptations, glob- 
ally or in any particular region, we must first determine if the patterns evident in the 
archaeological record result from actual changes in human behavior, biases imposed on 
the record by geological or taphonomic forces, or the recovery methods of archaeolo- 
gists themselves (see Erlandson 2001).^ 

Geologically, there is every reason to beUeve the archaeological record of coastal 
adaptations is seriously underrepresented, primarily due to the dramatic swings in glob- 
al sea levels associated with the glacial and interglacial oscillations of the Pleistocene. 
During the Last Glacial about 20,000 years ago, for instance, world sea levels stood 
between about 100 and 125 meters below present, exposing broad coastal plains around 
the world that have virtually all been inundated as seas rose to their present levels. Similar 
cycles have occurred numerous times in the last 3 million years, causing enormous, but 
highly variable, changes in coastal geography around the world. Each time global sea lev- 
els have risen significantiy, the record of human occupation associated with lower shore- 
lines has either been inundated by rising sea levels, destroyed by associated coastal ero- 



sion, or both. Even today, with sea level approximately six meters below the levels of 
about 130,000 years ago, a number of important coastal sites (I-Gasies River Mouth caves. 
Die Kelders Cave, Gorham's Cave, and others) occupied during the Last Interglacial are 
now being destroyed by marine erosion. In North Africa and the Levant, Lower 
Paleolithic artifacts have been found in a number of interglacial beach deposits, testify- 
ing to the destruction of ancient sites located in coastal settings. 

As sea levels rise and fall, coastlines generally move laterally in response to such 
changes. The maximum lateral movements of coastlines during the last 20,000 years, for 
instance, have varied from as much as 1000 km in some areas {i.e., northern Australia) to 
less than a kilometer in others. However, areas where shorelines have moved less than 
about 10 km are unusual and also tend to be correlated with relatively early evidence for 
coastal occupations (Erlandson 2001). Reconstructing the ancient landscape in the vicin- 
ity of coastal sites is crucial, because a cave or open site located on the coast today may 
have been 5 km, 10 km, 50 km, or more from the coast at various times during the last 
25,000 or 125,000 years. Study of modern coastal hunter-gatherers suggests that they 
rarely travel more than about 5 or 10 km from a home base to collect food resources 
(Bigalke 1973:161; Meehan 1982). When they do hunt or forage further afield, the skele- 
tal remains of shellfish, fish, sea mammals, or sea birds are rarely transported much fur- 
ther than this. In most situations, therefore, sites located more than about 5—10 km from 
an ancient shoreline are unlikely to contain substantial evidence for marine resource use 
(Wing 1977). During periods of rising or falling sea levels, which encompass most of the 
Pleistocene, the intensity of aquatic resource use at any given site would have fluctuated 
depending on (among other things) its proximity to coastal habitats. Since sea level has 
risen dramatically in the last 1 8,000 years, most evidence for early marine resource use 
has almost certainly been inundated or destroyed (van Andel 1989). Furthermore, sites 
with long occupational sequences located along the modern coast tj'pically show evi- 
dence for a postglacial intensification of marine resource use that may be primarily relat- 
ed to changes in local environments rather than a diversification or intensification of 
human subsistence (see Parkington 1981; Shackleton 1988). 

Despite such problems, a number of early coastal sites have been found (Figure 1). 
Almost all of these are located in areas where the continental shelf is relatively narrow 
and lateral movements of the coastline have been limited. Geologically, these are just the 
kind of places we should expect evidence for early coastal adaptations to be preserved 
(Richardson 1998; Erlandson 2001). Where shorelines are steep, sites still preserved 
above sea level may sometimes be found within the foraging radius of ancient coastal 
habitats. Consequently, sites like the Klasies River Mouth caves (Singer & Wymer 1982) 
in South Africa, Devil's Tower and Gorham's Cave in Gibraltar (Garrod et al. 1928; 







Waechter 1964), several shell middens located on the Melanesian islands of New Ireland 
and New Britain (AUen et al 1988; Wickler & Spriggs 1988), and Daisy Cave on 
California's Channel Islands (Erlandson et al. 1996) have all produced evidence for the 
Pleistocene use of marine resources. 

Yesner (1987) suggested that such sites are exceptional and reflect rare instances of 
early coastal adaptations in areas of unusually high marine productivity. This is true of 
some early sites, but it does not explain the evidence for early marine resource use at early 
sites in Italy, Lebanon, Libya, and Algeria (see McBurney 1967; Klein & Scott 1986; 
Stiner 1994) along the shorelines of the Mediterranean Sea, where marine productivity^ is 
generally relatively low (Erlandson 2001). Yesner (1980, 1987) also argued that middle 
Holocene sea level stabilization was crucial to the formation of extensive shallow 
nearshore habitats, increased coastal productivity, and the intensification of maritime 
adaptations, which suggests that steep coastlines should be marginal for human exploita- 
tion. Such changes may have enhanced nearshore productivity and marine resource use 
in some areas, but I suspect that evidence for Pleistocene use of marine resources — 
including widespread evidence for inland trade or transport of ornamental shells by 
AMH groups — is the tip of the proverbial iceberg, representing the only visible rem- 
nants of earlier and more extensive coastal settlement and adaptations. Until we system- 
atically search for underwater Pleistocene sites, however, we will not know which sce- 
nario is correct. Given the nearly endless diversit}^ in the relative productivity and acces- 
sibility of marine and terrestrial habitats in coastal zones around the world (see Perlman 
1980), moreover, it seems likely that the antiquity and intensity of coastal adaptations 
varied widely through both space and time. 

At present, it is apparent that Homo erectus and archaic Homo sapiens used marine 
resources to some extent, but the intensity of such use is unknown and appears to be 
Kmited largely to shellfish.^ Not surprisingly, anatomically modern humans currently 
show the earliest evidence for a more intensive use of shellfish and a wider range of 
marine or aquatic resources. At Isdasies River Mouth caves. Die Kelders, and other Last 
Interglacial localities in South Africa, for instance, AMH appear to have regularly eaten 
a variety of shellfish, marine mammals, and flightless birds (see Singer & Wymer 1982; 
Klein & Cruz-Uribe 2000), although some or all of the larger vertebrates may have been 
scavenged rather than hunted (Binford 1984). At Katanda in Zaire, moreover, comes the 
earliest evidence for complex aquatic hunting gear, the roughly 85,000 year old barbed 
harpoons mentioned previously (YeUen et al. 1995). 

Although the situation is still sketchy, current evidence suggests that the earliest sub- 
sistence strategies that included relatively eclectic and intensive use of marine or other 
aquatic resources may well be associated with AMH. When such aquatic adaptations 



were combined with the exploitation of a range of terrestrial plants and animals, the 
result would have been a more diversified and stable resource base. Such economies may 
have contributed significantly to the reproductive success of Homo sapiens sapiens and our 
dramatic demographic and geographic expansion of the last 150,000 years (Erlandson 

The A^ntiquity of Seafaring 

The idea that maritime adaptations may have relatively high antiquity, at least among 
anatomically modern humans, is supported by recent evidence for a relatively early devel- 
opment of seafaring capabilities. It now appears, in fact, that Homo erectus in Southeast 
Asia may have had some seafaring capabilities, apparendy settling the Indonesian island 
of Flores as much as 700,000 to 800,000 years ago (Sondaar et al. 1994; Morwood et al. 
1998). At present, however, there is little other evidence for the use of watercraft by 
Homo erectus or archaic Homo sapiens (see Cherry 1990), however, and it appears that such 
capabilities were relatively rudimentary. 

Evidence for more systematic and sophisticated Pleistocene voyaging comes from 
several coastal regions around the world, but most of it derives from Australia, 
Melanesia, and eastern Asia, where evidence for voyages in excess of 20 to 200 km has 
now been widely documented (Table 1). The proof that seafaring extended well back 
into the Pleistocene requires a major paradigm shift, since maritime voyaging was once 
thought to be strictly a Holocene phenomenon. Two publications in the 1970s pushed 
back the antiquit)' of human seafaring significandy. One of these was the discovery of 
obsidian from the Mediterranean island of Melos in strata at Franchthi Cave in mainland 
Greece dated between about 9500 and 13,000 14 C yr B.P (Cherry 1990). The second 
was the unequivocal demonstration that humans had reached Australia by at least 20,000 
ago (Lampert 1971), a migration that required multiple sea crossings through island 
southeast Asia, even when sea levels were lowered approximately 125 m during the last 
glacial. By the mid-1970s, the Australian archaeological record had been extended to 
about 34,000 years at Lake Mungo (Bowler & Thorne 1976), and subsequent research 
pushed it back to about 40,000 years (Groube et al. 1986), 50,000 years (Roberts et al. 
1990), and now possibly to 60,000 years or more (Thorne et al 1999). 

At first, the Australian data were puzzling for two reasons. First, in historic times 
Australian Aborigines reportedly had no sophisticated watercraft capable of making sub- 
stantial sea crossings (Flood 1990:36), which raised questions about their ability to trav- 
el through island Southeast Asia by boat. Like much of the rest of the world, Australia 
also had no true coastal shell middens or other evidence for intensive marine resource 



Table 1. Islands Colonized or Explored by Pleistocene Seafarers. 


Description of Evidence 

(years B.P.) 


Flores, SE Asia Possible evidence for limited seafaring by 



SondaarelJ. 1984. 

New Guinea 
and Australia 

Crete, Greece 

Oldest sites in Suhul are the earliest evidence 
for planned maritime voyiging, involving 
multiple sea crossings, some up to 90 km long. 

Ham sapiens sapim remains widi poorly 
documented context; calcareous breccia in 
which bones were cemented dated byPa/Uto 
51,000 + 12,000 BP; colonization of Crete 
apparently required several shon sea crossings. 

Bismarck Archipelago, Shell middens, fishing, and seafaring at 
Melanesia several sites dated from 15-35 KYR, with 

voyages up to 140 km long. 

Sicily, Italy Aurignacian assemblage from Mediterranean 

Island; possibly involving a sea crossing. 

Ryul^yu Islands, Human skeletal remains found in Yamashita-cho 

Japan and other caves on Okinawa and other islands; 

involves voyages of ca. 75-150 km. 

Kozushima Island, Upper Paleolithic peoples on Honshu crossing 
Japan 50 km wide channel to obtain obsidian. 

Melos Island, Greece Travel across ca. 24 km of open water to 
obtain obsidian for mainland trade. 

Admiralt}' Islands 


Channel Islands, 

Settlement of Manus Island required 200 km 

Occupation of Aetokremnos site, Akrotiri 
Peninsula on southwest coast of Cyprus. 

Boat and marine resource use by coastal 
Paleoindian groups, with sea crossings of 
at least 10 km 



Groahe a aL 1986; 

Facchini & Giusberti 

Southeast Alaska Presence on islands indicates a maritime 

& British Columbia lifestyle and seafaring capabilities. 


Allenet^ 1988, 1989; 




Matsu'ura 1996. 


Oda 1990:64. 




Allen &Keishaw 1996. 


Cheny 1990:151. 


Johnson efdi 2000; 


Davis e(<?/ 1989; 
Fedje & Quistensen 



use dating to the Pleistocene. In fact, an overwhelming majority of Australia's coastal 
shell middens were less than about 5,000 to 6,000 years old. It was sometimes argued, 
therefore, that Australia was colonized accidentally by castaways, who, stranded in a 
strange new land, abandoned the coast and adapted to terrestrial habitats. As knowledge 
of the Pleistocene geography of island Southeast Asia and Sahul accumulated, the cast- 
away model of the peopling of Australia was rejected on genetic and other grounds 
(Birdsell 1977; Bowdler 1977). Regardless of the route chosen, the colonization of New 
Guinea and Australia required several separate sea crossings, including voyages at least 
80 km long (Clark 1991). As the antiquity of this migration was pushed progressively 
back in time, it became clear that the initial settiement of Sahul represented the earliest 
evidence in the world for planned maritime voyaging. 

Confirmation of this scenario came in the late 1980s, when archaeologists investi- 
gating the roots of the 3500 year old Lapita culmral complex in western Melanesia dis- 
covered several Pleistocene shell middens in the Bismarck Archipelago and the Solomon 
Islands east of New Guinea (AUen et al. 1988, 1989; Wickler and Spriggs 1988). 
Settlement of these islands, now dated to at least 35,000 yr B.P. (Allen & Kershaw 
1996:185), added several significant maritime crossings to those already required to reach 
Australia and New Guinea. Perhaps more importantly, the sites themselves contained the 
marine shellfish, fish, and other remains expected of a maritime people. The presence of 
these Pleistocene shell middens in Melanesia, in contrast to the Australian situation, is 
due to the fact that coastiines adjacent to the sites plunge steeply into deep water and sea 
level fluctuations have had a relatively limited effect on the local geography and coastal 
archaeological record. 

The Melanesian evidence also suggests that maritime voyaging capabilities improved 
significantly between about 35,000 and 15,000 years ago. Although the initial settlement 
of Sahul, New Britain, and New Ireland required voyages of up to 100 km, for instance, 
the settiement of Buka in the Solomon Islands at least 28,000 years ago required a min- 
imum sea crossing of 140 km and perhaps as long as 175 km (Irwin 1992:20). By about 
15,000 years ago, moreover, roughly the time most scholars think the Americas were first 
settied, Melanesian seafarers had reached Manus Island in the Admiralt)^ group, which 
required an uninterrupted voyage of 200—220 km, — km of which was completely out of 
sight of land (Irwin 1992:21). 

Clearly, the relatively warm waters of island Southeast Asia and Melanesia would have 
posed very different challenges to Pleistocene seafarers than those of the North Pacific 
and Beringia (Erlandson 1993, 1994:269). Further evidence for Pleistocene seafaring 
comes from the islands of Japan. Japan itself was connected to the Asian mainland dur- 
ing periods of very low sea level, so its settlement did not necessarily require boats. Pagan 



(1990:191) has suggested, however, that the introduction of new blade and edge-grind- 
ing technologies about 30,000 years ago took place when Japan was separated from the 
mainland and probably involved boats. This idea may be supported by the discovery of 
human bones found beneath a charcoal-rich stratum in Yamashita-cho Cave on Okinawa 
which has been radiocarbon dated to about 32,000 14 C yr B.P. (Matsu'ura 1996:186). 
Human remains dated between about 15,000 and 26,000 yr B.P. have also been found in 
several other limestone caves on Okinawa and the smaller islands of the R\aik)Ti chain 
(Matsu'ura 1996), an island arc that stretches southward from Japan nearly to Taiwan. 
Human remains from Pinza-abu Cave on Miyako Island, found stratigraphically below a 
calcareous flowstone stratum, were associated with charcoal dated to about 26,000 14 C 
yr B.P. (JVIatsu'ura 1996:187). The bathymetry of the R}aikyu Islands area suggests that 
several sea voyages would have been required to reach Okinawa from Japan, including 
one crossing roughly 75 km long. Reaching Miyako Island, from either Japan or Taiwan, 
would have required even longer voyages of up to 150 km. 

In Japan itself, archaeological evidence suggests that by at least 21,000 years ago, mar- 
itime peoples from Honshu were using boats to procure obsidian from Kozushima 
Island located approximately 50 km offshore (Oda 1990). As Fagan (1990:191) noted, 
the first appearance of anatomically modern humans in the Japanese archipelago may 
have "coincided with the beginnings of seafaring in these temperate waters." Despite the 
evidence for Pleistocene maritime voyaging, the oldest shell middens in the Japanese 
islands date to between 9,000 and 10,000 years ago (Aikens & Higuchi 1982). Earlier 
coastal sites may be submerged offshore. 

The presence of Pleistocene maritime peoples in Japan is also significant because it 
places competent mariners in the cool waters and boreal climates of the North Pacitic at 
a date early enough to have contributed to the initial colonization of the Americas. From 
Japan, moreover, the Kurile Islands arc northeastward like stepping stones to the 
Kamchatka Peninsula. From there, the Commander Islands could have provided a base 
for migration through the Aleutian Islands, although a gap of roughly 250 km would 
have represented a formidable obstacle. A more likely route would have been to foUow 
the shorelines of the Kamchatka Peninsula northeastward where they would have 
merged imperceptibly with the south coast of Beringia. 

Given the Pleistocene seafaring capabilities of Howo sapiens sapiens, the presence of 
Pleistocene seafarers in the Japanese archipelago, and the geography of the North 
Pacific, what was once imponderable (Aikens 1990:12) now seems entirely conceivable 
and increasingly likely Near the end of the Pleistocene, in Upper Paleolithic Japan and 
possibly elsewhere in coastal northeast Asia, maritime peoples were simated at the base 
of a maritime pathway to the New World. Did they make such a journey? 



Out of Asia: Migration Koutes into the NeiP World 

Changing perspectives on the peopling of the Americas are interesting, in part, 
because they contrast dramatically with colonization models and recent developments in 
Australian archaeology. The two cases have some interesting parallels — both regions 
were colonized by AMH relatively late in human history, for instance, and involve the 
movement of humans into pristine continents untouched by earlier hominids — but the 
contrasts are even more interesting. Where Australia has been widely viewed as having 
been colonized by sea, the early settiement of the Americas has traditionally been seen 
as a terrestrial affair. Over the last 40 years, moreover, the colonization of Australia has 
been pushed steadily backward in time, while the widely accepted antiquity of human 
settlement in the Americas has contracted significantiy, despite repeated claims for much 
earlier occupations. When I first began seriously studying archaeology in the 1 970s, many 
introductory textbooks suggested that the Americas were first settled 40,000 or more 
years ago and many of the earliest sites were considered to be located along the Pacific 
Coast of North America. At the time, it was widely believed the ice-free corridor was 
closed between about 40,000 and 70,000 years ago, so some archaeology textbooks sug- 
gested that the Americas must have been colonized more than 70,000 years ago. One by 
one, however, purportedly ancient localities like Old Crow in the Yukon and Del Mar 
Man, Los Angeles Man, Texas Street, and Santa Rosa Island in California, all succumbed 
to careful scientific scrutiny, including the redating of several key artifacts or human 
skeletons using improved analytical techniques, which showed them to be Holocene in 
age {e.g., Taylor et al. 1985). Although there has never been any shortage of new claims 
for sites older than about 12,000 years, legitimate questions have been raised about vir- 
mally aU of them, and the 11,000 to 12,000 year old Nenana and Clovis complexes of 
North America came to be widely accepted as the earliest evidence for human occupa- 
tion in the New World (see Hoffecker etcil. 1993). With the widespread acceptance of the 
12,500 year old pre-Clovis site of Monte Verde in Chile (Meltzer et al. 1997), however, 
the "Clovis-first" position has come under attack and left many scientists in search of 
alternative models for the peopling of the New World. 

One if by hand: The Ice-Free Corridor 

Historically, two main routes have been proposed for the initial colonization of the 
Americas. Both involve Late Pleistocene migrations by AMH from northeast Asia into 
North America, but the two models diverge dramatically in the details. The traditional 
view is that terrestrial hunters walked from northeast Asia into North America across the 



arctic plains of Beringia during a glacial period when sea level was relatively low. The 
land-based scenario involves hearty bands of terrestrial hunters spreading across the 
Beringian interior, then winding their way into the American heartland via a long and 
narrow ice-free corridor running between two massive glacial ice sheets. 

This story of an epic journey through a 1,500 km long ice-free corridor has captured 
the imaginations of both archaeologists and the American public for decades. It also 
seemed to make sense as long as the earliest well-documented sites were the Clovis and 
Folsom kill sites where Pleistocene megafauna had been butchered by hunters east of the 
Rocky Mountains. Coastlines were clearly peripheral to this scenario and gave rise to the 
common perception that land-based Paleoindian hunters gradually radiated outward 
from the continental center, following game-rich rivers and valleys until they eventually 
reached the coast and slowly learned to survive in novel marine environments. 

This neat scenario ran into some trouble as Clovis and Folsom sites, both dated to 
relatively narrow periods of time near the end of the Pleistocene, were found further and 
further from the continental core, except to the north where they should have been rel- 
atively abundant and logically the earliest. By the 1980s, i?luted Clovis-like points had 
been found from the Atlantic to the Pacific coasts of North America and from Central 
America to Alaska, raising questions about where Clovis cultural traditions originated 
and how they spread so far so fast. The discovery and definition of the Nenana complex 
in south-central Alaska (Powers & Hoffecker 1989; Goebel et al 1991; Yesner 1996), a 
logical and slightly earlier precursor to Clovis that lacks fluted points, occurred just in 
time to save the Clovis-first, big game hunting, ice-free corridor model from collapse. 
Located far to the south and apparently predating both Clovis and the opening of the 
ice-free corridor, however, the possibility of a 12,500 year old pericoastal occupation at 
Monte Verde in Chile raised new questions about when and how the Americas were first 
settled and brought new attention to the coastal migration theory (see Erlandson 
1994:268; Dixon 1999). 

Two if by Sea: The Coastal Migration Theory 

Recent discoveries and events have breathed new life into the coastal migration the- 
ory, which suggests just the opposite of the ice-free corridor h^'pothesis — that maritime 
peoples first traveled around the North Pacific Coast then followed river valleys leading 
inland from the sea. Having a coastal route available, however, does not prove that such 
a maritime migration took place. Archaeological evidence for early boat use from islands 
along the western margin of the Pacific may support the idea that such a journey was 
technologically feasible, but archaeological data from the Pacific Coast of North and 



South America are presently ambiguous about the origins of the earliest coastal occu- 
pants. I have long argued that many skeptics would not be persuaded that a coastal 
migration was involved in the initial settlement of the New World until true coastal sites 
older than the Nenana and Clovis complexes were clearly documented. Such sites have 
not yet been found, but evidence for the antiquity of coastal settlement, maritime adap- 
tations, and seafaring along the Pacific Coast of the Americas has increased significantiy. 

Over the years, numerous scholars have proposed that a coastal migration may have 
contributed to the initial peopling of the Americas (e.g., Heusser 1960; Chard 1963; 
LaughHn 1967; Fladmark 1979, 1986; Mithun 1979; Gruhn 1988, 1994; Easton 1991; 
Dixon 1993, 1999; Erlandson 1994; Fedje & Christensen 1999; and others). Witia the dis- 
covery in the 1920s of Clovis and Folsom kill sites in the interior, however, models of 
the peopling of the New World shifted inexorably towards the interior and the ice-free 

By the 1960s, prevailing thought on the paleogeography of the North Pacific Coast 
reinforced this shift, suggesting that a coastal route into the New World would have been 
blocked by massive and continuous accumulations of glacial ice that completely covered 
the Alaska Peninsula and the northern Northwest Coast. Consequentiy, the coastal 
migration theory remained a marginal and all but ignored alternative to the ice-free cor- 
ridor model. Fladmark (1979) and others have revived interest in a coastal migration 
route, citing new evidence that glacial refugia existed along the northern Northwest 
Coast even during the last glacial, that most of the outer coast was deglaciated by at least 
13,000 to 14,000 years ago (Mann & Peteet 1994; Mann & Hamilton 1995), and that the 
ice-free corridor of the interior was far less hospitable to humian migration than previ- 
ously thought. Recent evidence even suggests that an ice-free interior route may only 
have become available as recently as 1 1,000 or so years ago (Mandryk 1991; Dixon 1993). 
Thus a coastal route into the Americas once again seems to be theoretically possible. 

There are other reasons for the marginalization of the coastal migration theory, 
including the old biases against early seafaring and maritime adaptations described above. 
The most important reason, however, is the fact that for decades the oldest securely doc- 
umented New World archaeological sites have been found in the American heartland, 
Nenana, Clovis, and Folsom sites located in interior regions and often associated with 
the remains of bison and other large game animals. The oldest widely accepted interior 
sites were dated between about 11,800 and 11,000 14 C yr B.P., while the oldest Pacific 
Coast shell middens dated to about 9,000 14 C yr B.P. This chronological gap of nearly 
3,000 radiocarbon years seemed consistent with a model that big-game hunting and inte- 
rior occupation came first, followed by migrations to the coast and the development of 
coastal foraging economies. 



With further research, however, fluted Clovis or Folsom-like points have now been 
found throughout much of North America, including a number of specimens discov- 
ered along the Pacific Coast of North America (Erlandson & Moss 1996). Unfortunately, 
the vast majority of specimens found in the far west have been surface finds or isolates 
that cannot be precisely dated and have no associated faunal remains to illuminate the 
nature of Paleoindian economies. While the temporal priority of coastal vs. interior 
occupation has essentially closed, it is not yet possible to determine if Pacific Coast 
Clovis peoples were adapted to the sea (Moss & Erlandson 1995). 

The Pacific Coast A.rchaeological Record 

Nonetheless, unequivocal evidence for the antiquity of coastal or maritime adapta- 
tions continues to be pushed back in time along the Pacific Coast of North and South 
America. Today, despite significant biases imposed on coastal archaeological records by 
sea level rise and marine erosion, the gap between Clovis and the earliest documented 
maritime peoples continues to close and has nearly disappeared in some areas. 

A key area in any test of the coastal migration theory is the southern coast of Alaska, 
comparable to the ice-free corridor in the sense that it links Beringia with temperate 
North America. In the Aleutian and southwest Alaska coastal areas, several sites dating 
between about 7500 and 8700 14 C yr B.P. have been found, suggesting that seafaring 
peoples had colonized the eastern Aleutians and Kodiak Island by at least this time. No 
earlier evidence of coastal occupation has yet been found, but archaeological research in 
these areas has been limited and the detailed geological work required to focus any sys- 
tematic search for early coastal sites has only rarely been done. Although portions of the 
Kodiak Archipelago are relatively protected, most of the shorelines of this area are also 
exposed seasonally to very high waves and marine erosion, conditions under which early 
sites are unlikely to have been preserved. 

A different situation exists in the labyrinthine network of islands, estuaries, and heav- 
ily forested shorelines of southeast Alaska and British Columbia. Here, many stretches 
of coast are relatively protected from marine erosion and boat travel is relatively easy in 
sheltered inside waters. Marine productivit}^ is also very high, with a high ratio of coast- 
line length to land area, and terrestrial resources are limited. Finally, geological work has 
shown that postglacial warming and the retreat of glacial ice fields has resulted in iso- 
static rebound of many coastal landforms, leading to unusual circumstances where 
shorelines dating between about 8000 and 13,000 years ago are sometimes found in areas 
above modern sea level. Consequently, the northern Northwest Coast has been the focus 



of recent interdisciplinary efforts to find early coastal sites (see Dixon et al. 1997; 
Josenhans et al. 1997; Fedje & Christensen 1999; Moss & Erlandson 1999). 

One such effort has been the Tongass Caves or Southeast Alaska Caves (SEA Caves) 
project in the southern Alexander Archipelago (Dixon et al. 1997; Moss & Erlandson 
1999). In this ongoing interdisciplinary project, archaeologists, geologists, and paleontol- 
ogists have worked together to reconstruct the sea level, glacial, paleoecological, and 
archaeological histories of the area. Archaeological and paleontological field studies have 
focused on caves and rockshelters on Prince of Wales and other islands west of 
Ketchikan. So far, the earliest secure evidence for human occupation dates to about 9200 
14 C yr B.P. at PET-408, where portions of a human skeleton and a microblade-bearing 
occupational horizon have been documented (Dixon 1999:117-119; Dixon et al. 1997). 
Elsewhere in southeast Alaska, three open air sites dated between about 8000 and 9000 
14 C yr B.P. are also known: Groundhog Bay 2, Hidden Falls, and Chuck Lake II (see 
Ackerman et al 1979, 1985; Davis 1989; Moss 1998). Except for Groundhog Bay, all 
these early southeast Alaskan sites are located on islands that required boats to settle. 

In British Columbia, equally early sites are known, including the mainland site of 
Namu, dated to as much as 9700 14 C yr B.P. (Carlson 1995) and a number of sites on 
the Queen Charlotte Islands now known to be 9000 or more years old (Fedje & 
Christensen 1999). Recent interdisciplinary work on the Queen Charlotte Islands has 
focused on reconstructing the regional paleogeography to narrow the search for early 
coastal sites (Fedje et al. 1996; Josenhans et al 1997). This work has documented several 
intertidal sites dated to between about 8000 and 9500 14 C yr B.P. and even dredged a 
basalt flake from the sea bottom in waters ca. 55 m deep, a surface estimated to have last 
been above sea level about 10,200 years ago (Fedje & Christensen 1999:647). The sheer 
number of early intertidal and raised beach sites documented on the Queen Charlotte 
Islands also implies some time for the demographic expansion of coastal peoples to 
attain that level of archaeological visibilit}^. Unfortunately, it is not currently possible to 
estimate the length of that process. 

On the southern Northwest Coast in Wasliington, Oregon, and northern California, 
no shell middens more than about 3500 years old were known prior to about 1985. Even 
today, the vast majorit}^ of coastal archaeological sites within this area date to the late 
Holocene, but a few early and middle Holocene sites have now been found (IVIoss & 
Erlandson 1998). An increasingly compelling body of literature now implicates geologi- 
cal forces as playing a significant role in the dearth of early coastal sites (Minor & Grant 
1995; Erlandson et al 1998, 2000; Moss & Erlandson 1998). This area is located adjacent 
to what is known as the Juan de Fuca plate and the Cascadia Subduction Zone, where 
geological evidence suggests that very large earthquakes associated with tsunamis and 



accelerated marine erosion occur every few centuries. It now appears tliat large portions 
of the coast may drop a meter or more during such earthquakes, raising sea levels virtu- 
ally instantaneously and leading to severe marine erosion (Peterson et al. 2000). Despite 
these processes, two shell middens dating to roughly 8000 14 C yr B.P. have been found 
along the southern Northwest Coast in the last decade, the Indian Sands (35-CU-67) 
site on the southern Oregon Coast (Moss & Erlandson 1998) and the Duncan's 
Rockshelter site (CA-SON-348) on the northern California Coast (Schwaderer 1992). 

The earliest evidence of maritime adaptations and seafaring in North America cur- 
rently comes from California's Channel Islands, which have remained separated from the 
mainland throughout the Pleistocene and could only have been settled by boat. At Daisy 
Cave on San Miguel Island, an ephemeral but stratigraphically discrete shell midden lens 
containing a few chipped stone artifacts has produced radiocarbon dates on marine shell 
(10,600 ± 70 and 10,700 ± 90 14 C yr B.P.) and wood charcoal (10,390 ± 130 14 C yr 
B.P.) that indicate that the cave was briefly occupied by maritime Paleoindians during 
Folsom times (Erlandson et al. 1996). Stratigraphically above this terminal Pleistocene 
stramm is a series of much denser shell midden layers dated between about 9700 and 
8000 14 C yr B.P. that have produced abundant shellfish and fish remains, sea mammal 
and sea bird bones, stone tools, bone bipoints (fish gorges), and shell beads. At Arlington 
Springs on the northwest coast of Santa Rosa Island, Orr (1962, 1968) also found human 
bones in a paleosol buried 37 feet {ca. 11m) below the surface. Early dates on charcoal 
and the bones of Arlington Woman suggested that she died roughly 10,000 years ago 
(Erlandson 1994:186), but recent AMS dating using advanced extraction and decontam- 
ination techniques suggests that her true age may be closer to 10,500 or even 11,000 14 
C yr B.P. (Johnson et al. 2000). The Channel Islands have also produced numerous other 
early Holocene sites that corroborate the presence of early seafaring peoples off the 
California Coast. Along the mainland coast of central and southern California, more than 
30 coastal sites have been dated between about 9500 and 8000 14 C yr B.P. (see 
Erlandson & Moss 1996:285-287). 

In South America, even earlier evidence for coastal adaptations has been found (see 
Richardson 1998). This includes apparent evidence for the use of coastal resources at 
Monte Verde in Chile, where seaweed and asphaltum reportedly derived from coastal 
habitats may date to as much as 12,500 14 C yr B.P. (see Dillehay 1997). At the Querero 
site north of Santiago, Nunez and colleagues (1994) reportedly found marine shellfish, 
sea Hon, and whale remains associated with the bones of terrestrial animals, with a series 
of radiocarbon dates ranging between about 11,600 and 10,900 14 C yr B.P. (Richardson 
1998). Llagostera (1979) has also documented the existence of diversified maritime 
economies on the northern Chile Coast dated to as much as 9700 14 C yr B.P. 



In Peru, where Richardson has demonstrated a positive correlation between the pres- 
ence of early coastal sites and steep offshore bathymetry, evidence for coastal settlement 
and marine resource use now appears to extend back over 1 1 ,000 years. The earliest evi- 
dence comes from the Quebrada Jaguay site along the south coast, where charcoal sam- 
ples from a shell midden containing abundant shellfish, fish, and sea bird remains have 
been dated to ca. 1 1,000 14 C yr B.P. (Sandweiss et al. 1998). Keefer and colleagues (1998) 
reported similar remains from another Peruvian midden dated to ca. 10,700 14 C yr B.P. 
and the basal strata at the Ring site shell midden produced a similar date (Sandweiss et al. 
1989). Richardson (1998) also reported the presence of the remains of shellfish collect- 
ed from mangrove habitats at a series of ephemeral campsites in the Talara region of 
northwest Peru, sites dated between about 9000 and 11,200 14 C yr B.P. In Ecuador, 
coastal shell middens of the Las Vegas complex have also been dated between about 
9500 and 10,800 14 C yr B.P (Stothert 1985). 

On the Notion of ^'Negative" Evidence 

Although there is stiU no clear evidence for the presence of fuUy maritime or seafar- 
ing peoples along the Pacific Coast of North and South America before Clovis times {ca. 
11,250 + 250 14 C yr B.P.), the gap has all but disappeared in places and earUer coastal 
sites will almost certainly be found in other areas. As I have noted, there is also reason 
to believe that early Pacific Coast archaeological records have been heavily affected by 
sea level rise, marine erosion, and other factors (Richardson 1998; Erlandson 2001). The 
effects of such processes vary along different stretches of the Pacific Coast, depending 
on variation in offshore bathymetry, wave energy, geology, and other attributes. Contrary 
to earlier assertions {i.e., Osborn 1977), however, there are no known areas of the Pacific 
Coast where shorelines between about 13,000 and 25,000 years old are currently above 
(or even near) modern sea level (Richardson 1998). Thus, we are still missing a key com- 
ponent of the geograpliic and archaeological records relevant to understanding the ini- 
tial colonization of the Americas. I am not suggesting that Clovis or pre-Clovis age sites 
definitely exist along the continental shelves of the Pacific Coast, but without conduct- 
ing a systematic search for such sites we will never be able to answer crucial questions 
about the peopling of the New World. 

The prospect that even older sites may exist on the continental shelf raises an inter- 
esting problem in the debate about the history of maritime adaptations and the coastal 
migration theory. This is the issue of what to do with what is sometimes referred to as 
"negative" evidence. The objection is relatively simple and straightforward: Many schol- 
ars insist, even if we know that the data we have may be biased, that we must buHd our 



models of the past based on the archaeological evidence we have at hand. Therefore, 
even if rising postglacial seas drowned the world's coastlines — and possible evidence 
for early marine resource use or coastal migrations with them — we cannot assume that 
such sites exist and build our models on negative evidence. Without tangible evidence for 
such adaptations or migrations, we must assume that they did not exist, circumstantial 
evidence simply doesn't cut it. 

I sympathize with this view to some extent: Our reconstructions must be governed 
by some reasonable rules of evidence after all. On the other hand, I also believe that as 
scientists we must be more open about what we know and don't know, and more clear- 
ly recognize that there are many major issues about human history that we are unable to 
resolve with currently available data. I believe questions about the antiquity of maritime 
societies and the possibilit)^ that Pleistocene seafaring peoples colonized the Americas are 
among those that cannot currently be answered with any sense of authorit}'. To answer 
such questions we need more evidence and much of the evidence we need is beneath the 
sea, in the presence or absence of submerged Pleistocene archaeological sites. It is one 
thing to suggest that maritime adaptations may have developed late in human history, 
especially with caveats, but quite another to talk of historical facts when the archaeolog- 
ical record remains ambiguous. 

It surprises me, too, that so many archaeologists have essentially disregarded or dis- 
missed the significance of the geological processes (sea level rise, marine erosion, etc.) 
that structure the archaeological record of the world's coastlines. I don't expect everyone 
to endorse the idea that deeply drowned sites will be found, only to accept the possibil- 
ity that archaeological records on land may be significantly biased and that further 
research is needed to determine the nature and extent of such biases. Since the 1970s, 
virmally no one has seriously questioned that the Pleistocene colonization of Australia, 
New Guinea, and western Melanesia involved maritime peoples in seaworthy watercraft. 
Yet the idea that maritime peoples may have colonized the Americas during the 
Pleistocene has been dismissed as relying on negative evidence that may or may not exist 
along drowned Pacific Coast shorelines. Such pronouncements would be easier to accept 
if comparable skepticism was applied to terrestrial alternatives. What evidence is there, 
for instance, that people walked across the now drowned plains of Beringia on foot? 
Why has the unlikely construct of a long and narrow ice-free corridor teeming with game 
remained the dominant paradigm for the peopling of the Americas for decades, despite 
relatively limited evidence for its existence, its suitabilit)^ as a human habitat, or the pres- 
ence of archaeological sites of Clovis or greater age? 

Given the emerging evidence for the technological sophistication and rapid geo- 
graphic spread of anatomically modern humans, the antiquit}^ of maritime adaptations. 



the presence of Pleistocene mariners on the eastern shores of the Pacific Ocean, and the 
nature of late glacial environments in northwestern North America, it is time for a sea 
change. I am not suggesting that we discard the concept of an ice-free corridor that may 
have played a significant role in the peopling and early history of the Americas. I believe, 
however, that the most likely scenario emerging from data derived from a variety of dis- 
ciplines will show that the colonization of the Americas was a complex process that 
involved multiple waves of migration, out of Asia and into the New World, and quite 
possibly by both land and by sea. 

Summary and Conclusions 

To summarize, just over 10 years ago, conventional views on human evolution sug- 
gested that: 1) anatomically modern humans developed independently in several areas of 
the Old World, descended from ancient Homo erectiis populations that spread out of 
Africa about a miUion years ago; 2) coastal adaptations and maritime voyaging played lit- 
tle or no significant role in human history until about 10,000 to 15,000 years ago; and 3) 
the Americas were first settled about 12,000 years ago by small bands of terrestrial 
hunters who marched relatively rapidly across a Beringian land bridge, down the fabled 
ice-free corridor, into the heart of North America, and on to the southern tip of South 

Today, we still have much to learn and large geographic and temporal gaps to fill. 
Nonetheless, data emerging from a variety of disciplines suggest a very different story. 
They suggest that anatomically modern humans probably evolved in Africa about 
1 50,000 years ago then rapidly spread around the world, largely replacing archaic human 
populations. If this Out of Africa model is essentially correct, we are left with only about 
5 to 10 percent of the time allowed for in our old models to account for the demo- 
graphic and geographic expansion of Homo sapiens sapiens. It also appears, at least after 
about 100,000 years ago, that our anatomically modern ancestors were more sophisticat- 
ed intellectually and technologically than previous models gave them credit for. 

The new data also tell us that aquatic adaptations and seafaring played a more signif- 
icant role in the demographic expansion, the geographic spread, and the phenomenal 
success of our species than previously supposed (Erlandson 2001). This notion is sup- 
ported by the earliest evidence for relatively effective use of a wider range of marine 
resources by AMH in a series of Last Interglacial sites located along the coast of south- 
ern and eastern Africa. It appears to be supported by the use of composite harpoons 
associated with abundant freshwater fish remains in Zaire by about 80,000 to 90,000 
years ago. Finally, it is clearly indicated by the evidence for Pleistocene seafaring from 


Australia, New Guinea, Melanesia, Okinawa, and Japan dating between at least 50,000 
and 15,000 years ago. 

Among the last three major colonizing migrations undertaken by humans (Australia, 
the Americas, and the Pacific Islands), at least two appear to have involved planned mar- 
itime voyaging. Circumstantial evidence from the eastern Pacific suggests that several 
Late Pleistocene Asian populations also had watercraft and the ability to make significant 
sea voyages. Given the geography and paleoecology of the North Pacific, Beringia, and 
Northwest North America, I have argued that it is increasingly likely that boats played a 
significant role in the colonization of the Americas. Although it can be argued that such 
a migration may well have taken place, more tangible evidence will be required to prove 
that such a migration acmally occurred (see Yesner 1996). 

I don't know when and how the Americas were first settled, but I do know that a 
coastal migration theory recently considered by many to be imponderable has moved 
into the realm of the possible or even the probable. Ironically, the coastal migration the- 
ory that once was considered by many scholars to be marginal now appears relatively 
conservative when compared to media and scholarly accounts of Pleistocene Australians 
or Polynesians crossing the open Pacific, or Europeans crossing the open Atiantic. 
Despite such claims, I believe the earliest occupants of the New World were the ances- 
tors of the Native Americans who Uved in the Americas when the Vikings and Columbus 
made their pioneering voyages across the Atlantic. This "Out of Asia" scenario is sup- 
ported by such an overwhelming body of linguistic, biological, genetic, and archaeolog- 
ical evidence, that its basic premise seems highly unlikely to change.*^ Thus, if a coastal 
migration was involved in the initial colonization of the Americas, the Pacific coastlines 
of North and South America are the logical places to search for the evidence. 

I believe a compelling case can now be made that we need to increase the levels of 
funding and effort directed towards identifying archaeological evidence for Pleistocene 
coastal occupations in North and South America. The difficulties involved in finding 
coastal sites older than 11,000 or 12,000 radiocarbon years are serious, but not insur- 
mountable. To increase the chances of success, such searches should be focused on areas 
with relatively steep continental shelves, on caves or rockshelters and springs (in arid 
coastal areas) or other features that may have drawn maritime peoples inland from 
Pleistocene shorelines, or on submerged semi-protected coastlines where marine erosion 
is least likely to have destroyed drowned terrestrial sites. Such research should be care- 
fully planned and conducted within an interdisciplinary framework that includes geolog- 
ical analysis of paleoshoreUnes and landscapes, consideration of the sedimentary and 
depositional regimes that may affect the preservation and visibilit)^ of ancient sites, and 
the identification of areas most likely to have been used by early maritime peoples. 



yAfter the appearance of anatomically modern humans about 1 50,000 years 
ago, evidence for technological innovation, marine resource use, and seafaring all 
increase dramatically. Intentional marititne voyaging appears to have developed 
sometime after about 7 5 ,000 years ago, contributing to some of the most impor- 
tant migrations in human history, including the peopling of Australia, western 
Melanesia, and the Pacific Islands. Evidence for Pleistocene seafaring in 
Southeast A-sia and Australia — along with recent archaeological and paleoeco- 
logical evidence from both North and South America — has helped revive the 
notion that the Americas also may have been settled by maritime peoples. 
Evidence that Pleistocene voyaging contributed to the initial colonisation of the 
New World is still largely circumstantial, but a variety of data suggest that it is 
increasingly likely that such a maritime migration took place. I suspect, in fact, 
that the Pleistocene colonisation of the Americas may have included both land- 
based migrations through the Beringian interior and mariti?ne voyaging around the 
North Pacific Kim. 


First and foremost, I wish to thank Mrs. Phyllis Wattis for her generous support of 
the California Academy of Sciences (CAS) symposium in which an early version of this 
paper was first presented. Participating in the Wattis symposium first provided me with 
the opportunity to crystaUize my thoughts on the possible relationships between some 
global problems and the peopling of the New World. I am also grateful to Nina 
Jablonski, program chair and organizer of the symposium, Nancy Gee and the staff of 
CAS for administrative and logistical support, and my fellow participants in the Wattis 
symposium for their stimulating presentations and discussions. I am also indebted to 
Nina Jablonski, Madonna Moss, James Richardson, and an anonymous reviewer for edi- 
torial comments that significantiy improved this paper, and to Ruth Gruhn, Alan Bryan, 
and J. Mithun for providing copies of papers helpful in revising my paper. Over the years, 
my views on the peopUng of the New World have been influenced by discussions with, 
among others, Mel Aikens, Jim Dixon, Jim Haggarty, Richard Jordan, Rick I-Cnecht, 
Madonna Moss, Roger Powers, Jim Richardson, Dennis Stanford, and David Yesner. 
Nonetheless, the opinions and conclusions expressed in this paper are peculiarly my own. 




^ Recent redating of Homo erectus specimens from Java suggests that these hominids 
may have reached Southeast Asia as early as 1.6 to 1.8 million years ago (Swisher et al. 

2 Krings and colleagues (1997) isolated a small segment of DNA from a sample of 
Neanderthal bone from a site in Germany. They found that the Neanderthal DNA was 
three times as different from modern standards as the genetic diversity evident among 
modern humans, suggesting that Neanderthals represented a species separate from our 

3 Recent discussions of the Chatelperronian complex in southwest Europe, long 
regarded as an Upper Paleolithic cultural tradition, suggest that Neanderthals may have 
made some ornaments, formal bone tools, and blades about 35,000 to 40,000 years ago 
as anatomically modern humans were moving into the region. However, it is still not 
clear that the Neanderthal remains found in two Chatelperronian sites represent the 
makers of these typically Upper Paleolithic objects. The Neanderthal remains could have 
been deposited by AMH occupants of the sites. Middle and Upper Paleolithic materials 
may have been mixed within a single stramm by trampling or other disturbance process- 
es, more advanced materials could have been traded or given to Neanderthals by AMH, 
or they could result from Neanderthal mimicry of the material culture of a more sophis- 
ticated society. 

4 Although space does not permit me to discuss these in detail here, it has long been 
known that screening of excavated archaeological sediments, especially screening over 
mesh sizes of 1/8-inch {ca. 3 mm) or smaller, is usually necessary to recover a repre- 
sentative sample of the remains of many aquatic animals, including small fish and shell- 
fish. The fact that faunal remains were not systematically collected from many early 
coastal or riverine sites raises the possibilit}^ that our understanding of early aquatic 
resource use patterns may be seriously biased (see Erlandson 2001). 

^ Fish bones from Hoxne in England, probably dated to ca. 300,000 years ago, may 
be an exception. In distributions that correlate well with those of artifacts and terrestri- 
al fauna, numerous freshwater fish bones were recovered by sieving the site sediments 
(Stuart et al. 1993:163, 198). It is uncertain, however, if the fish remains were of culmr- 
al or natural origin. 

^ Despite relatively intensive research, there currently is no evidence for occupation 
of any Pacific Islands outside of western Melanesia, Okinawa, and Japan earlier than 
about 5,000 years ago. Although coastal migrants around the North Pacific could have 
shared common ancestors with Austronesian peoples, the idea that Polynesians reached 



the Americas in the Pleistocene makes little sense, because distinctive Polynesian cultures 
did not emerge until about 3500 years ago. Despite some interesting technological paral- 
lels, the idea that Europeans crossed the Atlantic to colonize the Americas during the 
Pleistocene (see Stanford, this volume) contradicts a vast body of genetic, linguistic, bio- 
logical, and archaeological evidence and currentiy seems unlikely. 

l^iterature Cited 

Ackerman, R. E., T. D. Hamilton, & R. Stuckenrath. 1979. Early culture complexes on 
the northern Northwest Coast. Canad. J. Archaeol. 3:195—209. 

Ackerman, R. E., K. C. Reid, j. D. Gallison, & M. E. Roe. 1985. Arcbaeo/qgy of Heceta 
Island: A Survey of 16 Timber Harvest Units in the Tongas s 'National Forest, Southeastern 
Alaska. Center for Northwest Anthropology Project Report 3, Washington State 
University, PuUman, WA. 

Aikens, C. M. 1990. From Asia to America: The first peopling of the New World. 
Prehistoric Mongoloid Dispersals 7:1—34. The University of Tokyo Museum. 

Aikens, C. M., & Higuchi. 1982. Prehistory of Japan. Academic Press, New York. 354 pp. 

Allen J., & P. Kershaw. 1996. The Pleistocene-Holocene transition in Greater Australia. 
Pages 175-199 in L. G. Straus, B. V. Eriksen, J. M. Erlandson, & D. R. Yesner, eds.. 
Humans at the End of the Ice Age: The Archaeology of the Pleistocene-Holocene Transition. 
Plenum, New York, NY. 

Allen, J., C. Gosden, &J. P. White. 1989. Human Pleistocene adaptations in the tropical- 
island Pacific: Recent evidence from New Ireland, a Greater Australian outlier. 
Antiquity 63:548-561. 

AUen, J., C. Gosden, R. Jones, & J. P. White. 1988. Pleistocene dates for the human occu- 
pation of New Ireland, northern Melanesia. Nature 331:707—709. 

Bailey, G N. 1978. Shell middens as indicators of post-glacial economies: A territorial 
perspective. Pages 37—63 in P. MeUars, ed., The Early Postglacial Settlement of Northern 
Europe. University of Pittsburgh Press, Pittsburgh, PA. 

Barrie, V., K. Conway, R. Mathewes, H. Josenhans, & M. Johns. 1993. Submerged late 
Quaternary terrestrial deposits and paleoenvironment of northern Hecate Strait, 
British Columbia continental shelf, Canada, ^/w/. Int. 20:123—129. 

Bass, G E 1972. The earliest seafarers in the Mediterranean and the Near East. Pages 
12—31 in G E Bass, ed., The History of Seafaring Based on Underwater Archaeology. Walker 
& Co., New York, NY. 

Bigalke, E. H. 1973. The exploitation of sheUtish by the coastal tribesmen of the 
Transkei. Annual of the Cape Province Museum 9:159—175. 


Binford, L. R. 1984. Faunal K£mams from Klasies River Movi\h. Academic Press, New York, 

NY 283 pp. 
Bifdsell, J. B. 1977. The recalibration of a paradigm for the first peopling of greater 

Australia. Pages 113-167 in]. Allen, J. Golson, & R. Jones, eds., Sunda and Sahul: 

Prehistoric Studies in Southeast Asia, Melanesia, and Australia. Academic Press, New York, 

Bowler, J. M., & A. G. Thome. 1976. Human remains from Lake Mungo. Pages 127-138 

in R. I. Kirk & A. G Thome, eds.. The Origin of the Australians. Humanities Press, NJ. 
Bowdler, S. 1977. The coastal colonization of Australia. Pages 205-246 in]. Allen, J. 

Golson, & R. Jones, eds., Sunda and Sahul: Prehistoric Studies in South East Asia, 

Melanesia, and Australia. Academic Press, New York, NY. 
Brooks, A. S., D. M. Helgren, J. S. Cramer, A. Franklin, W Hornyak, J. M. Keating, R. G 

Klein, W. J. Rink, H. Schwarcz, J. N. L. Smith, K. Stewart, N. E. Todd, J. Verniers, & 

J. E. Yellen. 1995. Dating and context of three Middle Stone Age sites with bone 

points in the Upper Semliki Valley, Zaire. Science 268:548-552. 
Cann, R. L. 1994. MtDNA and Native Americans: a southern perspective. A^//. J. Hum. 

Genet. 55:7-11. 
Cann, R., M. Stoneking, & A. Wilson. 1987. Mtochondrial DNA and human evolution. 

Nature 325:31-36. 
Carlson, R. L. 1995. Early Namu. Pages 83-102 in R. L. Carlson & L. Dala Bona, eds., 

Early Human Occupation in British Columbia. University of British Columbia Press, 

Vancouver, Canada. 
Chard, C. 1963. The Old World roots: Review and speculations. Anthropol Pap. Univ. 

Alaska \0:\\b-\2\. 
Cherry, J. F. 1990. The first colonization of the Mediterranean islands: A review of 

recent research./. Medit Archaeol 3/2:145—221. 
Chilardi, S., D. W Frayer, P Giola, R. Macchiarelli, & M. Mussi. 1996. Fontana Nuova di 

Ragusa (Sicily, Italy): Southernmost Aurignacian site in Europe. Antiquity 70:553-563. 
Clague, J. J., J. M. Ryder, W H. Mathews, O.L. Hughes, N. W Rutter, & C M. McDonald. 

1989. Quaternary geology of the Canadian cordillera. Pages 17-96 /// R.J. Fulton, ed.. 

Quaternary Geology of Canada and Greenland. Geological SocieU^ of Canada, Ottawa. 
Clark, J. T. 1991. Early settlement of the Indo-Pacific. /. Anthropol Archaeol 10:27-53. 
Cohen, M. N. 1977. The Food Crisis in Prehistory: Overpopulation and the Origin of Agriculture. 

Yale Universit}' Press, New Haven, CT 341 pp. 
Davidson, I., & W Noble. 1992. Why the first colonisation of the Australian region is 

the earliest evidence of modern human behaviour. Archaeol Oceania 27:113—119. 



Davis, S. D., ed. 1989. The Hidden Falls Site, Baranof Island, A^laska. Aurora Monographs 

V, Alaska Anthropological Association. 383 pp. 
Dillehay, T. D. 1997. Monte Verde, A Tate ^Pleistocene Settlement in Chile, vol. 2, The 

Archaeological Context and Interpretation. Smithsonian Institution Press, Washington D.C. 

1071 pp. 
Dixon, E. J. 1993. Quest for the Origins of the Tirst Americans. Universit}^ of New Mexico 

Press, Albuquerque, NM. 156 pp. 
. 1999. Tiones, Boats, <&" Bison: Archeology and the First Colonp:^tion of Western North 

America. University of New Mexico Press, Albuquerque, NM. 322 pp. 
Dixon, E. J., T. H. Heaton, T. E. Fifield, T. D Hamilton, D. E. Putnam, & E Grady. 1997. 

Late Quaternary regional geoarchaeology of southeast Alaska karst: A progress 

report. Geoarchaeology 12:689—712. 
Easton, N. A. 1991. Mai de Mer above Terra Icognita, or, "What ails the coastal migra- 
tion xkizor^r Arctic Anthropol 29:28—41. 
Erlandson, J. M. 1993. California's coastal prehistory: A circum-Pacific perspective. 

Proceedings of the Society for California Archaeology 6:23—36. Society for California 

Archaeology, San Diego, CA. 
. 1994. Tarly Hunter-Gatherers of the California Coast. Plenum, New York, NY. 

336 pp. 
. 2001. The archaeology of aquatic adaptations: Paradigms for a new millenni- 

um. /. Archaeol Res. 9:287-350. 
Erlandson, J. M., & M. L. Moss. 1996. The Pleistocene-Holocene transition along the 

Pacific Coast of North America. Pages 277—301 in L. G. Straus, B. V. Eriksen, J. M. 

Erlandson, & D R. Yesner, eds.. Humans at the End of the Ice Age: The Archaeology of the 

Pleistocene-Holocene Transition. Plenum, New York, NY. 
Erlandson, J. M., D |. Kennett, B. L. Ingram, D A. Guthrie, D. P. Morris, M. Tveskoy, 

G. J. West, & P. Walker. 1996. An archaeological and paleontological chronology for 

Daisy Cave (CA— SMI-261), San Miguel Island, California. Radiocarbon 38:355—373. 
Erlandson, J. M., M. A. Tveskov, & R. S. Byram. 1998. The development of maritime 

adaptations on the southern Northwest Coast. Arctic Anthropology 35:6—22. 
Erlandson, J. M., M. A. Tveskov, M. L. Moss, & G. B. Wasson. 2000. Riverine erosion 

and Oregon Coast archaeology: A Pistol River case study. Pages 3—1 8 in R. J. Losey, 

ed., Changing Tandscapes: Proceedings of the 3 Annual Coquille Cultural Preservation 

Conference, 1999. Coquille Indian Tribe, North Bend, OR. 
Facchini, R, & G Giusberti. 1992. Homo sapiens sapiens remains from the island of Crete. 

Pages 189—208 in G Brauer & F. H. Smith, e.d?,.Continuity or Replacement: Controversies in 

Homo sapiens Evolution. A. A. Balkema, Rotterdam. 



Pagan, B. M. 1990. The Journey from Eden: The Peopling of Our World. Thames and Hudson, 

London. 256 pp. 
Fedje, D. W, & T. Christensen. 1999. Modeling paleoshorelines and locating early 

Holocene coastal sites in Haida Gwaii. Am. Antiquity 64:635—652. 
Fedje, D. W., J. B. McSporran, & A. R. Mason. 1996. Early Holocene archaeology and 

paleoecology at the Arrow Creek sites in Gwaii Haanas. Arctic Anthropol 33:116—142. 
Fladmark, K. R. 1979. Routes: Alternate migration corridors for Early Man in North 

America. Km. Antiquity 44:55—69. 

. 1986. Getting one's berings. Nat Hist. 95:8-19. 

Flood, J. 1990. Archaeology of the Dreamtime. Yale University Press, New Haven, CT 

302 pp. 
Garrod, D. A. E., L. H. D. Buxton, G E. Smith, & D. M. A. Bate. 1928. Excavation of 

a Mousterian rock-shelter at Devil's Tower, Gibraltar. /. Royal AnthropoL Inst Great 

Britain and Ireland 58:33— 1 13. 
Goebel, F E., W. R. Powers, & N. H. Bigelow. 1991. The Nenana complex of Alaska and 

Clovis origins. Pages 49-79 in R. Bonnichsen & K. L. Turnmire, eds., Clovis Origins 

and Adaptations. Center for the Study of the First Americans, Oregon State University, 

CorvaUis, OR. 
Greenhill, B. 1976. Archaeology of the Boat. Wesleyan University Press, Middletown, CT. 

320 pp. 
Groube, L., J. Chappell, J. Muke, & D. Price. 1986. A 40,000 year old human occupation 

site at Huon Peninsula, Papua New Guinea. Nature 324:453—435. 
Gruhn, R. 1988. Linguistic evidence in support of the coastal route of earliest entry into 

the New World. Man 23:77-100. 
. 1994. The Pacific Coast route of initial entry: An overview. Pages. 249-256 in 

R. Bonnichsen & D G Steele, eds.. Method and Theory for Investigating the Peopling of the 

Americas. Center for the Smdy of the First Americans, Oregon State Universit}', 

CorvaUis, OR. 
Heaton, T. H., & F. Grady. 1993. Fossil grizzly bears (Ursus arctos) from Prince of Wales 

Island, Alaska, offer new insights into animal dispersal, interspecific competition, and 

age of deglaciation. Cum Res. Pleistocene 10:98—100. 
Heusser, C. J. 1960. Late Pleistocene Environments of North Pacific North America. 

Am. Geogr. Soc. Spe. Pub I 35. 
Hoffecker, J. E., W R. Powers, & F. E. Goebel. 1993. The colonization of Beringia and 

the peopling of the New World. Science 259:46—53. 
Irwin, G. 1992. The Prehistoric Exploration and Colonisation of the Pacific. Cambridge 

University Press, Cambridge, MA. 240 pp. 



Johnson, J. R., T. W. Stafford, H. O. Ajie, & D. P. Morris. 2000. Arlington Springs revis- 
ited. Proceedings of the Fifth Channel Islands Symposium. Santa Barbara Museum of 
Natural History, Santa Barbara, CA. 715 pp. 

Johnstone, P. 1980. The Sea-craft of Prehistory. Harvard University Press, Cambridge, MA. 
260 pp. 

Josenhans, H. W., D. W. Fedje, R. Pienitz, & }. R. Southon. 1997. Early humans and rapid- 
ly changing Holocene sea levels in the Queen Charlotte Islands — Hecate Strait, 
British Columbia, Canada. Science 277:71—74. 

Keefer, D. K., S. D. deFrance, M. E. Moseley, J. B. Richardson, III, D. R. Satterlee, & A. 
Day-Lewis. 1998. Early maritime economy and El Nino events at Quebrada 
Tacahuay, Peru. Science 281:1833—1835. 

Klein, R. G. 1995. Anatomy, behavior, and modern human origins. /. W^orld Prehist 

. 1998. Why anatomically modern people did not disperse from Africa 100,000 

years ago. Pages 509—521 in T. Akazawa, K. Aoki, & O. Bar-Yosef, eds., Neandertals 
and Modern Humans in Western Asia. Plenum, New York, NY. 

Klein, R. G., & K. Cruz-Uribe. 2000. Middle and Later Stone Age large mammal and tor- 
toise remains from Die Kelders Cave 1 , Western Cape Province, South Africa. /. Hum. 
Ei'ol 38:169-195. 

Klein, R. G, & K. Scott. 1986. Re-analysis of faunal assemblages from the Haua Fteah 
and other Late Quaternary sites in Cyrenaican Libya. /. Archaeol Sci. 13:515—542. 

Kjrings, M., A. Stone, R. W Schmitz, H. Krainitzki, M. Stoneking, & S. Paabo. 1997. 
Neandertal DNA sequences and the origin of modern humans. O// 90:19— 30. 

Lampert, R. J. 1971. Burrill Take and Currarong. Terra AustraUsl. Australian National 
University, Canberra. 86 pp. 

Laughlin, W S. 1967. Human migration and permanent occupation in the Bering Sea 
area. Pages 409—450 in D. M. Hopkins, ed., The Bering Tand Bridge. Stanford University 
Press, Stanford, CA. 

Lewin, R. 1987. Bones of Contention: Controversies in the Search for Human Origins. Simon and 
Schuster, New York, NY. 348 pp. 

Llagostera, A. M. 1979. 9,700 years of maritime subsistence on the Pacific: An analysis 
by means of bioindicators in the north of Chile. Am. Antiquity 44:309—324. 

Locke, R. 1999. First American? California bones may be the oldest on the continent. 
Discovering Archaeology (July/ August) . 

MacDonald, G M. 1987. Postglacial vegetation history of the Mackenzie River basin. 
Quat Res. 28:245-262. 


Mandryk, C. S. 1991. Paleoecolog)i as Contextual Archaeology: Human Viability of the L^te 
Quaternary Ice-Free Corridor, Alberta, Canada. Ph.D. dissertation, University of Alberta, 
Edmonton, Canada. 

Mann, D H. 1986. Wisconsin and Holocene glaciation of southeast Alaska. Pages 
237-265 in K. M. Reed, & R. M. Thorson, eds., Glaciation in Alaska: The Geological 
Record, by T. D. Hamilton. Alaska Geological Society, Anchorage, AK 

Mann, D. H., & T. D. Hamilton. 1995. Late Pleistocene and Holocene paleoenviron- 
ments of the North Pacific Co-s&t. Quat. Sci. Rev. \A-AA9-A1\. 

Mann, D H., & D M. Peteet. 1994. Extent and timing of the Last Glacial Maximum in 
southwestern Alaska. ^//^/. Res. 42:136—148. 

Matsu'ura, S. 1996. A chronological review of Pleistocene human remains from the 
Japanese archipelago. Pages 181—197 in K. Omoto, ed.. Interdisciplinary Perspectives on 
the Origins of the Japanese. International Research Center for Japanese Studies. 

McBurney, C. B. M. 1967. The Haua Fteah (Cjrenaica) and the Stone Age of the Southeast 
Mediterranean. Cambridge University Press, Cambridge, MA. 387 pp. 

Meehan, B. 1982. Shell hed to Shell Midden. Australian Institute of Aboriginal Studies, 
Canberra. 189 pp. 

Mellars, P. 1998. Neandertals, modern humans, and the archaeological evidence for lan- 
guage. Pages 89-115 in N. G. Jablonski & L. C. Aiello, eds., The Origin and 
Diversification of Tanguage, Memoirs of the California Academy of Sciences No. 24, 
San Francisco, CA. 

Meltzer, D. J., D K. Grayson, G. Ardila, A. W. Barker, D. F. Dincauze, C. V. Haynes, F 
Mena, L. Nunez, & D. J. Stanford. 1997. On the Pleistocene antiquity of Monte 
Verde, southern Chile. Am. Antiquity 62:659—663. 

Minor, R., & W. C. Grant. 1995. Earthquake-induced subsidence and burial of Late 
Holocene archaeological sites, northern Oregon Coast. Am. Antiquity 61:772—781. 

Mithun, J. 1979. The Bering Strait: A strait jacket h}^othesis. Phoenix: Neiv Directions in the 
Study of Man 3:5—12. 

Molnia, B. F. 1986. Glacial history of the northeastern Gulf of Alaska — a synthesis. 
Pages 219-236 in T D Hamilton, K. M. Reed, & R. M. Thorson, eds., Glaciation in 
Alaska: The Geological Record, Alaska Geological Society, Anchorage, AK. 

Morwood, M. J., P. B. O'SuUivan, F. Aziz, & A. Raza. 1998. Fission-track ages of stone 
tools and fossils on the East Indonesian island of Flores. Nature 392:173-176. 

Moss, M. L. 1998. Northern Northwest Coast Overview. Arctic Anthropol 35:88-111. 

Moss, M. L., & J. M. Erlandson. 1995. Reflections on North American Pacific Coast pre- 
history/. World Prehist. 9:1-45. 

. 1998. Early Holocene adaptation on the southern Northwest Coast./. California 

and Great Basin Anthropol 20:13—25. 



. 1999. The seacaves of southeast Alaska. A.rchaeoL Univ. of Oregon 2:6—9. 

Nunez, L., J. Varla, R. Casamiquela, & C. ViUagran. 1994. Reconstruccion multidiscipli- 
naria de la ocupacion preliistorica de Querero, centro de Chile, l^atin A^merican 
Antiquity S:99-nS. 

Oda, S. 1990. A review of archaeological research in the Izu and Ogasawara islands. Man 
and Culture in Oceania 6:53—79. 

Orr, P. C. 1962. The Arlington Springs site, Santa Rosa Island, California. Am. Antiquity 

. 1968. Prehistory of Santa Rosa Is/and. Santa Barbara Museum of Natural History, 

Santa Barbara, CA. 253 pp. 

Osborn, A. 1977. Strandloopers, mermaids, and other fairy tales: Ecological determi- 
nants of marine resource utilization — the Peruvian case. Pages 157—205 in L. R. 
Binford, ed., For Theory Bui/ding in Archaeology. Academic Press, New York, NY. 

Parkington, J. 1981. The effects of environmental change on the scheduling of visits to 
the Elands Bay Cave, Cape Province, S.A. Pages 341—359 in I. Hodder, G. Isaac, & 
N. Hammond, eds.. Patterns of the Past. Cambridge University Press, Cambridge, MA. 

Perlman, S. 1980. An optimum diet model, coastal variability, and hunter-gatherer 
behavior. Advances in Archaeological Method and Theory 3:257—310. Academic Press, New 
York, NY. 

Peterson, C. D., D. L. Doyle, & E. T. Barnett. 2000. Coastal flooding and beach retreat 
from coseismic subsidence in the central Cascadia Margin, USA. Environmental and 
Ungineering Geoscience 6:255—269. 

Powers, W. R., &; |. F. Hoffecker. 1989. Late Pleistocene settlement in the Nenana Valley, 
central Alaska. Am. Antiquity 54:263—287. 

Richardson, J. B. III. 1998. Looking in the right places: Pre-5000 B.P maritime adapta- 
tions in Peru and the changing environment. Kevista de Arqueologia Americana 

Rightmire, G. P., & H.J. Deacon. 1991. Comparative studies of Late Pleistocene human- 
remains from Klasies River Mouth, South Africa./. Tium. EvoL 20:131-156. 

Roberts, R. G., R.Jones, & M. A. Smith. 1990. Thermoluminescence dating of a 50,000 
year old human occupation site in northern Australia. Nature 345:153—156. 

Sandweiss, D. H., J. B. Richardson III, E. J. Reitz, J. T Hsu, & R. Feldman. 1989. Early 
maritime adaptations in the Andes: Preliminary studies at the Ring site, Peru. Pages 
35—84 in D S. Rice, C. Stanish, & P. R. Scarr, eds.. Ecology, Settlement, and History of the 
Osmore Drainage, Peru. BAR International Series 545. 

Sandweiss, D H., H. Mclnnis, R. L. Burger, A. Cano, B. Ojeda, R. Paretics, M. Del 
Carmen Sandweiss, & M. Glascock. 1998. Quebrada Jaguay: Early South American 
maritime adaptations. Science 281:1830-833. 



Schwaderer, R. 1992. Archaeological test excavation at the Duncans Point Cave, 
CA-SON-348/H. Pages 55-71 /;/ T L. Jones, ed., Essays on the Prehistory of Maritime 
California. Center for Archaeological Research at Davis Publication 10, Universit)^ of 
California, Davis, CA. 

Shackleton, J. C. 1988. Reconstructing past shorelines as an approach to determining fac- 
tors affecting shellfish collecting in the prehistoric past. Pages 11-21 in G. G Bailey 
& J. Parkington, eds.. The Archaeology of Prehistoric Coastlines, Cambridge University- 
Press, Cambridge, MA. 

Singer, R., & J. Wymer. 1982. The Middle Stone Age at Klasies River Mouth in South Africa. 
University of Chicago Press, Chicago, IL. 234 pp. 

Smith, F. 1994. Samples, species, and speculations in the study of modern human origins. 
Pages 227-252 in M. Nitecki & D. Nitecki, eds.. Origins of Anatomically Modern Humans, 
Plenum, New York, NY. 

Sondaar, R Y, G D. van den Bergh, B. Mubroto, F Aziz, J. de Vos, & U. L. Bam. 1994. 
Middle Pleistocene faunal turnover and colonization of Flores (Indonesia) by Homo 
erectus. Comptes Residues de la Academie des Sciences Paris 319:1255—1262. 

Stiner, M. C. 1994. Honor Among Thieves: A Zooarchaeological Studj of Neandertal Ecology. 
Princeton University Press, Princeton, NJ. 447 pp. 

Stothert, K. 1985. The preceramic Las Vegas culmre of coastal Ecuador. Am. Antiquity 

Stringer, C, & R. Mclsje. 1996. African Exodus. Henry Holt, New York, NY 267 pp. 

Suiart, A. J., R. G Wolff, A. M. Lister, R. Singer, & J. M. Egginton. 1993. Fossil verte- 
brates. Pages 163-206 in R. Singer, B., G Gladfelter, & J. J. Wymer, eds.. The Lomr 
Paleolithic Site at Hoxne, England. University' of Chicago Press, Chicago, IL. 

Swisher, C. C, G H. Curtis, T Jacob, A. G Gett)^, A. Suprijo, & Widiasmoro. 1994. Age 
of the earliest known hominids in Java, Indonesia. Science 263:1118—1121. 

Taylor, R. E., L. A. Payen, C. A. Prior, R J. Slota, Jr., R. Gillespie, J. A. J. Gowlett, R. E. 
M. Hedges, A. J. T. JuU, T H. Zabel, D. J. Donahue, & R. Berger. 1985. Major revi- 
sion in the Pleistocene age assignments for North American human skeletons by 
C-14 accelerator mass spectrometry: None older than 11,000 C-14 years B. P. Am. 
Antiquity 50:136-140. 

Thorne, A., R. Grun, G Mortimer, N. A. Spooner,J. J. Simpson, M. McCulloch, L.Taylor, 
& D. Curnoe. 1999. Australia's oldest human remains: Age of the Lake Mungo 3 
skeleton./. Hum. Evol 36:591-612. 

Toyne, S. 1999. Aborigines were the first Americans. The Sunday Times, August 22. Times 

Newspapers, Ltd. 
van Andel, T. H. 1989. Late Quaternary sea level changes and archaeology. Antiquity 



Waechter, J. D. 1964. The Excavation of Gorham's Cave, Gibraltar, 1951-54. i3////. Inst. 

Waselkov, G. A. 1987. Shellfish gathering and shell midden archaeology. Ad. Archaeol. 

Method and Theory 10:93-210. 
Washburn, S. L., & C. S. Lancaster. 1968. The evolution of hunting. Pages 293—303 in B. 

Lee & L DeVore, eds., Man the Hunter. R. Aldine, Chicago, IL. 
Wickler, S., & M. J. T. Spriggs. 1988. Pleistocene human occupation of the Solomon 

Islands, Melanesia. Antiquity 62:703—707. 
Wing, E. S. 1977. Factors influencing exploitation of marine resources. Pages 47—66 in 

E. P. Benson, ed.. The Sea in the Pre-Columbian World, Dumbarton Oaks, Washington 

Wright, H. E. Jr. 1991. Environmental conditions for Paleoindian immigration. Pages 

1 13— 135 in T D. Dillehay & D. J. Meltzer, eds.. The First Americans: Search and Kesearch. 

CRC Press, Boca Raton, PL. 
Yellen, J. E. 1998. Barbed bone points: Tradition and continuity in Saharan and sub- 

S2ih2it2in. Mnc?i. African Archaeol. Kev. 15:173—198. 
Yellen, J. E., A. S. Brooks, E. Cornelissen, M.J. Mehlman, & K. Stewart. 1995. A Middle 

Stone Age worked bone industry from Katanda, Upper Semliki Valley, Zaire. Science 

Yesner, D. R. 1980. Maritime hunter-gatherers: Ecology and prehistory. Curr. Anthropol 

. 1987. Life in the "Garden of Eden": Constraints of marine diets for human 

societies. Pages 285—310 in M. Harris & E. Ross, eds.. Food and Evolution. Temple 

Universit}^ Press, Philadelphia, PA. 
. 1996. Human adaptation at the Pleistocene-Holocene boundary (circa 13,000 

to 8,000 BP) in eastern Beringia. Pages 255-276 in L. G. Straus, B. V. Eriksen, J. M. 
Erlandson, & D. R. Yesner, eds., Humans at the End of the Ice Age: The Archaeology of the 
Pleistocene-Holocene Transition. Plenum, New York, NY. 


Chapter Five 

Facing the Past 

A View of the North American 

Human Fossil Record 

D. Gentry Steele and Joseph F. Powell 

^H^^^ Anthropologists' interest in North America's human fossil record 

^^^^^^^ has risen and fallen since scholars first became interested in who the 

^"^^C^^f^BM ^^^^t Americans were, how the New World continents were colo- 

*» 9 ^ ■ nized, and when. During the first half of the 20th century, for 

example, prehistoric skeletal remains formed the data base for the first models developed 

to explain the peopUng of the Americas {e.g., HrdUcka 1902, 1907, 1918, 1923, & 1937). 

However, during the second half of the century, research focused less often upon the 

ancestry and affinities of the North American Indians, and more upon ecological issues, 

such as how populations adapted to the various environments in which they lived, and 

the lifestyles they followed {e.g., Armelagos et a/. 1982). 

The shift away from issues concerning the origins of American Indians was caused 
by several factors. One major factor was Ales HrdUcka 's conclusion that the colonization 
of the New World was a very recent event. This view was founded upon Hrdlicka's con- 
clusions that the North American human remains being proposed as ancient which he 
had examined were all anatomically modern in appearance compared to the Neanderthal 
remains recovered in Europe. Further, Hrdlicka was convinced that all proposed ancient 
remains were assignable to times following the extinction of the megafauna of the last 
ice age of the Pleistocene (HrdUcka 1902, 1907, 1918, & 1923). In HrdUcka's view, 
Europe had archaic forms of humans who co-existed with the cold-adapted Pleistocene 
megafauna whUe the land of the Americas lay virgin to the human footprint. The reces- 
sion of the Pleistocene ice masses and the spread northward from the equator of 
warmer cUmates saw the demise of the cold-adapted biological populations and their 
replacement by the biological populations whose descendants Uved into historic times. 
This changing face of Europe was the theater in which HrdUcka beUeved the 
Neanderthals evolved into the anatomicaUy modern humans. In the Americas, however, 
it was not until after the end of the Ice Age and the estabUshment of the recent and bio- 
logical communities that the first humans, anatomically modern compared to the 



Neanderthals, crossed from northeastern Asia via the Bering Strait to colonize the New 
World. Linking Hrdlicka's model with the prevalent view of the times that natural selec- 
tion molded populations at an almost imperceptibly gradual rate, there would not be 
enough time since the end of the Pleistocene for anatomical changes to have occurred 
in the modern human populations. 

With the European Neanderthals as his standard of reference for archaic humans, 
Hrdlicka's assessment that all skeletal remains found in the New World were anatomical- 
ly modern and resembled anatomically modern extant Native American populations pro- 
vided the substantive underpinnings to the model. 

Two additional lines of evidence, which supported Hrdlicka's model of the recent 
colonization of the Americas, were the close resemblance of extant Native Americans to 
Northeast Asians compared to other Old World populations, and the inferred high 
degree of similarit}' within and between the populations within the Americas. With the 
above model in place, and the European Neanderthals as the standard of reference for 
early human populations, what could be learned from the continued study of the rela- 
tively recent human remains from the Americas concerning their origins or their disper- 
sion within the Americas? 

The basic tenets of this robust model have survived virtually unscathed even though: 
1) human entrance into the New World occurred at least during the final phase of the 
Pleistocene, earlier than the original model considered possible; 2) many of the North 
American skeletal remains, accepted as ancient and recognized as indistinguishable from 
extant Native Americans, have been found to be much more recent in age than previ- 
ously suspected, and are not representative of the early populations in the New World; 
and 3) that we have gained knowledge about how populations change through time, and 
the rates that changes can occur within an evolving lineage. 

Currentiy, however, there is renewed interest in how, where, and by whom the peo- 
pling of the New World occurred. This renewed interest has been brought about in part 
by challenges to the widely held assumption that the oldest remains were indistinguish- 
able from the most recent populations, and by the discovery of more complete remains 
whose great antiquity (more than 8,500 years old) is firmly established. Examination of 
the recent discoveries such as Kennewick Man (Chatters 2000, Powell & Rose 1999), in 
Washington and the recently dated Spirit Cave, and Wizard's Beach skeletons from 
Nevada (Tuohy & Dansie 1997) have confirmed the initial inferences about the distinc- 
tiveness of the earliest known American remains from more recent North American 
human skeletal remains. Examination of these recent finds have documented their 
resemblances to Old World populations, most commonly those of central and southern 
Asia and other populations of the Pacific Rim. 



We will concentrate upon the recent studies of the oldest remains to indicate the 
nature of the craniofacial shape that is used as evidence to challenge the long held opin- 
ions about how the earliest known colonizers of the Americas looked, and to consider 
the significance of this new perspective. Because of the equivocal nature of results based 
on samples limited in numbers and represented by fragmentary remains, most scholars 
have chosen to compare very recent samples which are represented by more complete 
and less comminuted remains. However, without examining the earliest human remains, 
it is impossible to test the assumption that the historical record has Uttie to add to the 
story of the peopling of the Americas. 

Early Reconsiderations 

When we began to re-evaluate North America's human fossil record (Powell 1993; 
PoweU & Steele 1992; Steele 1989; Steele & PoweU 1992, 1993, 1994), tiie samples we 
considered securely dated, and complete enough for at least limited multivariate com- 
parisons consisted of no more than 10 individuals (Table 1). Of these, only two males 
from Browns Valley and Sauk Valley, Minnesota, (Figure 1) and two females, one from 
Gordon Creek, Colorado, and one from Pelican Rapids, Minnesota were complete 
enough for 13 measurements to be taken. Ironically, two of these proved later to be of 
early Archaic antiquit}^ and the impact of this will be discussed below. 

These early Holocene samples were compared to more recent world populations 
using principal components analysis (PCA) and canonical variate analysis (CVA). Both 
multivariate statistical procedures compare populations on the basis of many traits simul- 
taneously. The samples are dispersed in a multidimensional space based upon the sum of 
the weighted comparative difference of each trait. Within this universe of samples, the 
distance of each sample from the other is a reflection of their dissimilarit)'^. The disper- 
sion of the samples within this universe can be visualized as a conventional three-dimen- 
sional graph where the first two axes identify a plane through the universe that best sep- 
arates the sample. The third axis used for the three dimensional plot is at a angle to the 
other two axes. The relative importance of each trait in establishing the difference 
between the samples is determined by the position of the planes within the universe of 
samples. Of the two multivariate procedures, the CVA requires the researcher to first 
identify related samples. Therefore, the procedure is effective at comparing clusters of 
known populations with one another and comparing the amount of variation within a 
group to the variation between the groups. PCA establishes the distinctiveness of each 
sample by comparing it to all others without defining clusters of similar samples. In prac- 
tice PCA requires the researcher to make fewer assumptions about the data, and this usu- 
ally results in samples which could be grouped together geographically to be more dis- 



persed from one another. This tendency was reflected in our PCA of samples from the 
same geographical region (and therefore, samples presumed to have a greater degree of 
shared genes), while samples from different geographical regions overlapped more 
closely. While this resulted in less discriminating power, we felt the technique more accu- 
rately reflected the biological underpinnings of the statistical measures of similarity. 

Table 1. Paleoindian samples from North and South America. List includes only those specimens 
complete enough for one or more craniofacial measures to be taken and used in multivariate anal^'ses. 
For a complete listing of Paleoindian remains see citations listed. 


Sample Size 

Nature of Remains 


White Water Draw' 




Gordon Creek ' 




Browns Valley ' 



8,700 + 110 

Pelican Rapids ' 



(Early Archaic) 

Sauk Valley' 



(Early Archaic) 

Wilson- Leonard ' 




Horn Sheker ' 




Spirit Cave ^ 



Wizard's Beach ^ 



Kennewick ^ 



Station do Riacho 
Lapa Vermelha ■* 
Cerca Grande ^ 


11,000 -12,( 
9720 + 130 

1. Sample used in early statistical studies of Paleoindian studies by Steele and PoweU (1992, 1994). 
Brown's Valley and Pelican Rapids were considered Paleoindian in age at time of origiaal studies. 

2. Additional North American sample used by Jantz and Owsley (1997) and Steele and Powell 
(1999), and Powell and Neves (1999). 

3. Additional North American sample used in study by Chatters (2000), Powell and Rose (1999), 
and Powell and Neves (1999). 

4. South American Paleoindian skeletons used in Powell and Neves (1999). 



Figures 2 and 3 are the PCA plots based on size corrected data for males and females 
from the two early Holocene males from Minnesota, and the Minnesota and Colorado 
females. A third female skull, the Wilson-Leonard II female (Figure 4), was too crushed 
to be accurately measured, but visually, her features were quite similar to the Minnesota 
and Colorado females. Figure 5 illustrates the process of preparing a facial reconstruc- 
tion of Wilson-Leonard 11. The Minnesota specimens were all recovered during the 
1930s, while the Colorado specimen was recovered in the 1960s. These early Holocene 
samples are compared to recent samples of populations located throughout the World 
(Howells 1969, 1973, 1989). 

For the males (Figure 2), the six recent North American samples are clustered near 
the center of the graph, in proximit)^ to the Northeast Asian samples on one side, and 
European and southern Pacific Rim (including the Australians) samples on the other. 
The Paleoindian female sample (Figure 3) lies along the periphery of the American 
Indian cluster of samples (crosses), within the cluster of southern Pacific Rim and 

Figure 1. Paleoindian female from Pelican Rapids, Minnesota. Initially considered to be of the 
antiquity of other Paleoindian remains (predating 8,500 yr B.P.), the most recent 14 C date indicated the 
age was comparable to North American remains associated with the Early Archaic (post 8,500 yr B.P.). 



Australian samples, aiul a\\a\' from the Northeast Asian samples. The i'Airopean samples 
are widely dispcrsetl in ihe up|ier half of the ^raph: Some appear to be more similar to 
Northeast Asian samples, while others appear more similar to the southern Pacific Rim 
samples. The female sample (Figure 3), analyzed separately, provided a similar picture. 

Bivariate analyses (Figures 6 & 7) of the measurements were used to clearl\ identify 
which Paleoindian cranial features were most disdncdve (Steele & Powell 1992, 1994). By 
using these simpler stadstics requiring only two measurements for analysis, the 
Paleoindians could be compared to a larger number of samples including some samples 
recovered from mid-Holocene deposits. Figure 6 documents that both male and female 
Paleoindians generally have relatively narrower and longer braincases than most recent 
American Indians. Populations with shorter and broader braincases are in the upper left 
portion of the bivariate plot, while populations with relatively longer crania, such as the 
North American Paleoindians, fall toward the lower right quadrant. Notice that samples 
closer, and therefore more similar, to these Paleoindians are from Archaic {e.g., Indian 



PRIN2 -0-38 



Figure 2. Principal Components plot of the first three components (PRIN 1, 2 and 3) illustrating 
relative differences between male Late Pleistocene/Early Holocene and recent samples on basis of 
eight size-corrected measurements of face and braincase. The distance of symbols from one anodier 
reflects their differences. Early samples are Paleoindians (pyramid); Upper Cave, China (cube); 
Minatogawa, Taiwan (flag) and lomon, |apan (cylinder). Samples representing more recent populations 
are North American Indians (cross). Northeast Asians (diamond), peoples of the southern Pacific Rim 
(open circle), and Europeans (open square). (After Steele and Powell 1994.) 





- -1.48 


Figure 3. PCA plot of the first three components (PRIN 1, 2 and 3) illustrating relative differences 
between female samples representing Late Pleistocene/early Holocene and recent populations on the 
basis of eight size-corrected measurements. The distance of symbols from one another reflects their 
differences. Early samples are Paleoindians (pyramid) and Upper Cave, China (cube). Samples of more 
recent populadons are Northeast Asian (diamond), southern Pacific Rim (open circle), and Europe 
(open square). (After Steele and Powell 1994.) 

Knoll and Tennessee Archaic) rather than Late Prehistoric times. When comparing these 
Paleoindians to southern Asians, Europeans, American Indians and northern Asians, the 
Paleoindian samples tend to resemble Australians and southern Pacific Rim populations 
in their long and narrow braincases rather than the northern Asians. Figure 7 documents 
the narrower and shorter faces of Paleoindians compared to more recent Native 
American populations. Again, Paleoindians cluster more closely with Australians and 
samples from the Pacific Rim and southern Asia, than they do with northern Asians (see 
Tables 4 & 5 of Steele & Powell 1994 for a Listing of the sites and mean dimensions of 
measurements used in these figures). 

After publishing these results, we were provided with opportunities to examine two 
virtually complete crania from Nevada that were recentiy found to be approximately 
9,500 years old. The two Nevada specimens, the Spirit Cave mummy (26CH1F) and the 
Wizard's Beach individual (Figure 8) (26WA1605) were recovered in the 1940s and 1950s, 
respectively, but have only recently been radiocarbon dated. The antiquity of the Spirit 



Figure 4. The Wilson-Leonard female cranium from Leander, TX. Although this skuU has been 
securely dated to be 9,500 years old, the skuU was too badly crushed to be measured accuratel}'. Her fea- 
tures, however, closely resembled those of the Gordon Creek and Pelican Rapids females. 

Cave individual is 9410 + 60 14 C yr B.P., based upon seven dates of bone and hair mate- 
rial. The antiquit)' of the Wizard Beach individual is 9515 ± 155 14 C yr B.P., based upon 
one bone date. Both individuals were adult males, and were very well preserved (Jantz & 
Owsley 1997; Steele & PoweU 1999). 

Because the Nevada skulls were much better preserved than the two male Minnesota 
specimens it was possible to compare them to recent crania using 28 measurements of 
the skull and face. In this analysis the two Nevada specimens were treated as represent- 
ing separate populations. Figure 9 illustrates the PCA based on the size-corrected data 
comparing the two Nevada samples to 25 more recent American Indians, Europeans, 
northern Asians, Polynesians, southeastern Asians, and Australians. (Eigenvectors and 
eigenvalues for these PCA are presented in Tables 1 & 2 of Steele & Powell 1999.) The 
most striking feature of this analysis is the distinctiveness of the Nevada individuals 
from more recent samples, as well as their distinctiveness from one another. However, 
they still more closely resemble Potynesians and Australians than thev do American 
Indians and Northeast Asians. 

Figure 10 shows the relationship of Nevada males and the Minnesota males based on 
13 size-corrected measurements which was the maximum number of measurenients 



which could be taken on the Minnesota specimens. Both the Nevada and the Minnesota 
samples again lie on the periphery of the North American Indian samples and close to 
Polynesian and Australasian samples, further from the northern Asians. 

Considering these studies, it appears that the exact relationship of the Paleoindian 
samples to other populations differs depending upon the samples compared, upon how 
many measurements are used in the analysis, and how measurements are made. In spite 
of these variations in the results, a generalized pattern can be seen. First, Paleoindian 
samples fall within a broad cluster of samples composed of samples from the Pacific 
Rim including northern and southern Asians, American Indians, Polynesians, and 
Australians. Second, the Paleoindians consistently appear more similar to southern 
Asians, Australians and populadons of the southern Pacific Rim than they do northern 
Asians. The features wliich the Paleoindians share with these southern Asian and Pacific 
Rim peoples are a long and narrow braincase and a relatively short, narrow face. 


Figure 5. The male cranium from Wizard Beach, Nevada is one of the most complete crania of 
an individual considered of the antiquit}? of Paleoindians (predating 8,500 yr B.P.). 



These resemblances of Paleoindians to more recent populaticms of rlic world, first 
reported in 1992 (Steele & Powell 1992) were affirmed in both the male and female sam- 
ples from two different localities, and samples that were analyzed using different data sets 
of 13 and 28 measurements. When the two Nevada Paleoindian crania were considered 
representative of a single population, and compared to the Minnesota Paleoindian males 
and the more recent comparative samples on the basis of 13 cranial measurements, the 
Nevada early Holocene remains fell within the dispersion of the late Holocene remains. 
However, when the early Nevada remains were compared to the more recent samples on 
the basis of 28 measurements, which provided a more detailed characterization of the 
skulls, the early Holocene Nevada remains were located at the periphery of all the sam- 
ples. This demonstrated that these two specimens could not comfortably be assigned to 
any of the subclusters. 

Kenneivick Man 

The recent discovery of a human skeleton of early Holocene age near Kennewick, 
Wasliington, has focused national and international attention upon America's fossil 
record and the question of the ancestry of the first Americans. The skeleton, that of an 
adult male, later called Kennewick man, was recovered from the Columbia River near 
Kennewick in July 1996. The well-preserved remains, including a skull with the facial 
skeleton complete enough to permit a large body of measurements to be taken, was 
dated at 8410 ± 50 14 C yr B.P., which is approximately 9,500 years old in calendar years 
(Chatters 2000). More recendy, three additional 14 C dates were taken on Kennewick 
Man's bony remains for the Department of the Interior (MacManamon 2000). Two of 
the three dates (8130 ± 40 & 8410 ± 60 14 C yr B.P) are virtually identical to die first 
date obtained and reported by Chatters. The third date (6940 + 30 14 C vr B.P.) confirms 
the antiquit)' of the Kennewick Man even though this date deviates from the other three 
dates by at least 1000 years. Kennewick Man has become one of the most securely dated 
human fossils in the Americas. 

Currendy the focal point in a lawsuit between eight scientists and the Department 
of the Interior, the preliminary analyses of Kennewick Man have yet to be verified. Two 
preliminary smdies have been published: One is by Chatters (2000) based upon his exam- 
ination of the remains before the lawsuit was filed and the other studv is bv Powell and 
Rose (2000) which was authorized by the Department of the Interior after the lawsuit 
was filed. To satisfy'' an issue raised by the lawsuit, Powell and Rose specifically assessed 
whether the Kennewick Man would be identified as an ancestor of a specific extant 
North American Indian group. To address this issue, Kennewick Man was compared to 
extant and late Holocene w^orld populations using 10—52 measurements depending upon 





Upper Cave 

Pecos -k ^ TN Aic+lSi? O 
D Indian— ■K/^'^ ^^Bkimo 

Knoll Tir ^ "*^ 

°, *■ TX Archaic 

160 165 170 175 If 

185 190 195 200 205 210 

155 160 165 170 175 180 185 190 195 


Figure 6. Bivariate plots of male (upper plot) and female (lower plot) samples compared on basis 
of cranial length (GOL) and width (XCB). Diagonal lines mark traditionally recognized boundaries for 
identifying wide crania (upper left), narrow crania (lower right), and crania of moderate width (central 
area) relative to length. Samples representing populations are Paleoindians and more recent samples of 
North American Indians (solid stars), Asians and late Pleistocene/early Holocene samples of 
Minatogawa and Upper Cave (solid circles); Australians and southern Pacific Islands (open stars), 
Africans (open squares), and Europeans (open circles). Individual samples are identified for North 
American Indians, Paleoindians, and the late Pleistocene/early Holocene Upper Cave and Minatogawa 
samples. (After Steele and Powell 1992.) 




Figure 7. Bivariate plots of male (upper plot) and female (lower plot) samples compared on basis 
of facial width (ZYB) and facial height (NPH) Diagonal Unes mark traditionally recognized boundaries 
for identif\ing wide crania (upper left), narrow crania (lower right), and crania of moderate width (cen- 
tral area) relative to cranial length. Samples representing male populations (upper plot) and female pop- 
ulations are Paleoindians (soUd stars) and North American Indians (solid stars), more recent Asians and 
late Pleistocene/early Holocene samples from Upper Cave, China and Minatogawa (solid circles), 
Australians and southern Pacific Islands (open stars), Africans (open squares), and Europeans (open cir- 
cles). Samples individually identified in the figure are North American Indians, Paleoindians, and the 
Late Pleistocene/earlv Holocene Upper Cave and Minatogawa samples. (After Steele and Powell 1992.) 



the completeness of the human remains of the samples compared. The samples were 
compared to Kennewick Man using CVA and PCA based upon primary and size-cor- 
rected craniofacial measurements. Consistently in these studies Kennewick Man was 
structurally midway between North American Indian groups and southern Pacific Rim 
populations. Depending upon the analysis, they were most similar to populations such as 
Maori, Ainu/Jomon, or Polynesian. When Kennewick Man was compared to other geo- 
graphical groups using larger data sets the dissimilarity of Kennewick Man from all 
recent human groups became apparent. However, even in these studies which empha- 
sized Kennewick Mans' dissimilarit}^, the recent populations which most closely resem- 
bled Kennewick Man were consistently southern Asian or Pacific Rim populations. 
Probability studies indicated that Kennewick Man would not be a member of any recent 
North American Indian group, nor would it fall morphologically within any other living 
human worldwide population. Comparative studies using discrete traits of the cranium 
and of the dentition also documented the distinctiveness from recent North American 
Indian populations. Comparing Kennewick Man to Archaic American Indian samples 
and recent human populations, Kennewick Man more closely resembled crania from the 
Archaic Indian Knoll sample, an Ainu/Jomon sample, and pooled Northeast Asian and 
southern Asian populations. Studies of discrete features of the cranium and the denti- 
tion both substantiated the distinctiveness from the extant American Indians. 
Summarizing their results from aU analyses, Powell and Rose concluded (1999): 

Like other Early American skeletons, the Kennewick remains 
exhibit a number of morphological features that are not found in 
modern populations. For all craniometric dimensions, the typicality 
probabilities of membership in modern populations were zero, indi- 
cating that Kennewick is unlike any of the reference samples used. 
Even when the least conservative inter-individual distances are used to 
construct typical probabilities, Kennewick has a low probability of 
membership in any of the late Holocene reference samples.Similar 
results were obtained by Ozolins et al. (1997) for Upper Paleolithic 
samples from Asia, Africa and Europe and Paleoindian groups, and are 
not surprising considering that Kennewick is separated roughly by 
8,000 years from most of the reference samples in Howell's (1989) and 
Hanihara (1996). The most craniometrically similar samples appeared 
to be those from the south Pacific and Polynesia as well as the Ainu of 
japan, a pattern observed in other studies of early American crania 
from North and South America (Steele & Powell 1992, 1994; Jantz & 
Owsley 1997). 



Figure 8. Frontal and lateral views of the Wilson-Leonard II adult female during three stages in 
the process of a preparing facial reconstruction based on the skuU. The top pair shows a reconstruc- 
tion of the crushed skull. The center pair shows the reconstructed skull with markers indicating the pre- 
dicted depth of facial tissue at key points on the skull. The tissue depth at each point is based upon 
mean tissue depths determined for world samples. Asian standards were used in the facial recontruc- 
tion. The bottom pair the reconstructed face prepared by Bett}- Pat Gatliff, a world-renowned forensic 
artist. The facial reconstruction was made for the Houston Museum of Science. Photographs are by 
Betty Pat Gatiiff. 



Early Holocene Kemains from South America 

As the distinctiveness of the face and braincase of early Holocene samples from 
North America was being recognized and reported, Walter Neves and his colleagues 
were documenting the distinctiveness of the earliest Holocene samples recovered in 
South America {e.g., Neves & Pucciarrelli 1989, 1991, 1998; Neves, Meyer, & Pucciarelli 
1996; Neves, Munford & Zanini 1996; Neves, Zunimi, Munford & Pucciarelli 1997; 
Neves, Powell, Prous & Pucciarelli 1998; Neves, Powell & Ozolins 1999; Powell & Neves 
1999). Using bivariate and multivariate statistics, the body of their work clearly docu- 
mented the difference of the early Holocene samples from late Holocene samples from 
South America, and the craniofacial similarities of the early Holocene remains to 
Australian, African, and southern Pacific populations. Recent analysis of the earliest 
known human remains from South America, a young adult female from the site of Lapa 
Vermelha IV thought to be about 12,000 years old, has confirmed the distinctiveness of 






Figure 9. PCA plot of the first three components (PCA 1, 2 ans 3) illustrating the relative differ- 
ence between male samples based on 28 size-corrected measurements of the braincase and face. The 
distance between samples represents their relative difference. Samples representing populations are the 
Spirit Cave and Wizard's Beach, Nevada Paleoindians (spades), more recent American Indians (cubes), 
Europeans (rectangles), Northern Asians (diamonds), Polynesians (cylinders), Southeast Asians (flags) 
and Australians (ovals). (After Steele and Powell 1999.) 



PC A 3 









Figure 10. PCA plot of the first three components (PCA 1, 2 and 3) of male samples based on 14 
size-corrected measurements. Distance between samples represents their relative difference. Samples 
representing populations are Nevada Paleoindians (spades) considered to represent a single populadon, 
the other Paleoindians (pyramids) considered a single population distinct from the Nevada Paleoindians, 
American Indians (cubes,) Europeans (squares), northern Asians (diamonds), Polynesians (cylinders). 
Southeast Asians (flags), and Australasian (ovals). (After Steele and Powell 1999.) 

the early South Americans (Neves et al. 1999). Figure 11 illustrates the structural similar- 
it)' of the face and braincase of the Hominid 1 female from Lapa Vermelha IV compared 
to a worldwide selecdon of populations measured by Howells (1973, 1989). The Lapa 
Vermelha IV Hominid 1 female is markedly similar to the African populations (Doggo, 
Yeit, Zulu), while the Asian and European populations form a loose second cluster. The 
Australian and one Polynesian sample form a third cluster. Previous studies also had doc- 
umented the similarit}' of South American remains to the African and Australian sam- 
ples, particularly the Australian samples. 

Figure 12 (Powell & Neves 1999) illustrates how the structural similarities between 
the major continental populations appear when they are considered at the 90% confi- 
dence limit. The distinctiveness of the Paleoindian samples from the late Holocene 
American Indian samples is still apparent even though the continental populations over- 
lap. The only sample completely separated from the Paleoindians are die American 



Indian samples. On the other hand, both the Paleoindians and North American Indians 
overlap with all other continental clusters of populations. The populations with which 
the present Paleoindians most closely correspond are the Polynesian, then the 
Northeastern and Southeastern Asian samples. 






Figure 11. PCA plot of the first three components (PCA 1, 2 and 3) illustrating differences 
between South American Paleoindians and other samples based on 28 size-corrected measurements. 
Distance between samples is a reflection of their differences. Samples representing populations are 
South American Paleoindian Vermelha IV (crosses), Africans (stars), Asians (diamonds), Australasian, 
American Indians (cubes), Polynesians (cylinders), and Europeans (squares). (After Powell and Neves 

Early and Mid-Holocene Comparisons 

As it became clear that the earliest remains from both New World continents looked 
different from the recent American Indians and Northeast Asians, the next step was to 
compare the Paleoindians with prehistoric North American skeletal remains from the 
middle Holocene to determine when the change occurred. One of the authors (JP) was 
one of the first scholars to specifically address tliis issue. In his dissertation on dental 
variation in prehistoric American Indians, Powell (1995) compared the Paleoindian sam- 



pic from North America with early to mid-Holoccne remains. Powell's conclusions were 
that, whereas Paleoindians were typically distinct from late Holocene samples from 
North America, they much more closely resembled remains from the mid-flolocene 
swath of dme. 

With the similarities between mid-Holocene and Paleoindian remains being docu- 
mented on the basis of dental structures, the next step was to determine if the same sim- 
ilarities could be recognized in the size and shape of the braincase and face. Powell and 
Neves (1999) documented the morphological relationships of the North American 
Paleoindians and South American Paleoindians with the North American Archaic popu- 
lations (Figure 13). The distribution is based on a multidimensional scaling of 
Mahalanobis' distance measurements where, again, the distance between two points is a 
reflection of their structural similarity. 






— ^ "^ 


- ^ 

p, ^ 


NE Asia 

~ m^ 








Africa \ 




1 1 1 

SE Asia 
1 1 

. .1 

1 1 



Figure 12. Bivariate plot of first two components (PC A 1 and 2) of a Principal Component 
Analysis illustrating the extent of variation at 2 90% confidence limit. Distribution of samples is based 
on a Q-mode analysis of 14 variables. Light gray ellipse reflects variation \\ithin North and South 
American Indians, and the dark ellipse represents variation among Australians. Ellipses of Europeans, 
Northeast Asians, Europeans, Southeast Asians and Polynesians are identified. The position of indi- 
vidual North American Paleoindians (solid circles), individual South American Paleoindians (open cir- 
cles), and the centroid positions for Jomon and Ainu samples (crosses) are indi\ndually marked. (After 
Powell and Neves 1999.) 



In Figure 13, three of the North American Paleoindians cluster reasonably well 
together and in a position midway between the South American Paleoindians and the 
North American mid-Holocene to late Holocene human remains. The South American 
samples form a tighter cluster than the North American Paleoindians and the Archaic 
skeletal remains from North America. There are two remains, one from the Archaic 
group and one Paleoindian which are far removed from their cohorts. They clearly indi- 
cate the amount of variation, which probably existed in the early and mid-Holocene 
American populations, that is being underestimated. 

The last point we emphasize here is that the North and South American Paleoindians 
show as great a dispersion as the Archaic and more recent North American Indian sam- 
ples. Previously, the high degree of homogeneity observed in living Native American's 
soft tissue features such as hair color and form, eye color and form, led scholars to 
assume that the total gene pool of Native American Indians was homogeneous. This 

^ LaBrea 

^ WMS 

. Sta. Riacho 



"'' A Lapa Vemieiha 


. Wizards ^ 
JL- Sauk Valley 

W'mdover ^_ Plains 

■ ■ 

SE Archaic 
^ Bird Is. 

HH Indian Knoll 

Figure 13. Distribution of North American Paleoindians (closed stars) and South American 
Paleoindians (open stars) compared with Archaic American Indian samples (solid squares), based on 
multidimensional scaling of Mahalanobis distances between individual Paleoindians and centroids of 
Archaic samples. Paleoindians are plotted individually. The positions of the Archaic samples are the 
centroid position for the sample. (After Powell and Neves 1999.) 



perceived homogeneity was thought to exist because the initial coloni/ers were few in 
number, were from a single region within northeast Asia, and that the earliest colonizers 
and their descendants had not been in the New World long enough for natural selection 
to shape them to local condidons. With our current awareness of the heterogeneit}' of 
American populations, we can no longer assume a recent colonization of the New 
World. Awareness of this heterozygosit}' also means that each Paleoindian and Early 
Archaic find will have an extraordinary potential for providing new information about 
America's earliest colonizers. 


In summary, the research over the past 10 years has documented the following: 1) 
The earliest remains recovered from the Americas have consistentiy been found to have 
a craniofacial shape that is distinguishable from the more recent Prehistoric and extant 
populations; 2) The North American Paleoindians resemble more closely the features of 
living and Late Prehistoric Southeast Asians than they do Northeast Asians; 3) The 
South American Paleoindian remains resemble more closely the North American 
Paleoindian craniofacial form than they do any other American Indian groups (living or 
dead); 4) The South American Paleoindians, while resembling North American 
Paleoindians, differ from them by more closely resembling Australian and African sam- 
ples than do the North American Indian populations; 5) The Paleoindians from both 
American continents are structurally more similar to Archaic (mid-Holocene) popula- 
tions than they are to Late Prehistoric and living American Indians; and 6) There is as 
much variation in early and mid-Holocene populations as there is in Late Prehistoric 

What is the Meaning 


Since the reconsideration of the American fossil record began in tiie 1980s (Neves & 
Pucciarelli 1989; Steele 1989), several explanations have been proposed to explain the 
differences between the Paleoindian remains and the Late Prehistoric and extant 
American Indians. Table 2 summarizes explanations which have focused upon the 
Paleoindian samples to explain their differences from the earliest and most recent pop- 

Table 3 summarizes models proposed to understand the colonization of the New 
World continents. The most mechanical of these explanations is that the characteristics 
of the Paleoindian specimens that have been documented are not an accurate reflection 



Table 2. Alternative explanations focusing upon the Paleoindians to explain why they differ from 
recent American Indians. 



Negating evidence 

Explanation 1: Sampling Error: 
Paleoindians presumed to look like 
late Holocene humans, but sampling 
errors create a false impression of 
Paleoindian population. 

Ver)^ small sample sizes. 

All but one comparison of 
Paleoindians have documented 
their similarity to one another 
and their distinctiveness from 
late Holocene samples. 

Explanation 2: Paleoindians differ 
from extant American Indians 
because they represent a distinct 
population derived from 
Southeast Asia. This Southeast Asian 
derived population later replaced in 
New World by populations derived 
from Northeast Asia. 

Explains distinctiveness of 
recent American populations 
from founding population 
and similarities to recent 
Northeast Asians. 

Assumes that cranial and 
facial features reflect the 
ancestral genotype. Does not 
consider possibility of 
evolutionary forces occurring 
in the Americas which may 
account for the dissimilarity 
of Paleoindians from 
extant American Indians. 

Explanation 3: Paleoindians differ 
from extant Northeast Asians and 
American Indians because they 
represent an earlier Northeast Asian 
population before the evolution of 
the more distinctive facial features of 
more recent populations. 

Explains the distinctiveness 
of Paleoindians and the later 
similarities of Northeast 
Asians and American Indians. 

Does not consider effects of 
evolutionary forces . 

Explanation 4: Distinctive features 
of the braincase and face of living 
American Indians were created by 
the evolutionary forces of genetic 
drift and gene flow, rather than the 
features reflecting an unaltered 
ancestor/ descendant relationships. 

Incorporates evolutionary 
models in explanation. 

Must assume parallel 
evolutionary forces acting 
on Northeast Asian and 
American populations if 
both differ from Paleoindians 
in the same way. 


S'lhULi: AND POWi'JJ. 

of the actLinl Palcoindiaii pojiulaiion as it existed. The reason tor this propt^scd inaccu- 
rate image of the Palcointlians is that the measurements based on the very small samples 
of Paleoindians deviate from the mean measurement of the complete population to such 
an extent that they do not reflect the average or the normal configuration of the parent 
population from which the)^ came. From the inception oi our research (Steele & Powell 
1992, 1994) we acknowledged the potential danger of evaluating such small samples, and 
presented our findings each time as preliminary assessments. However, as each new study 
based on different Paleoindian samples from new localities consistently substantiated our 
initial assessment it became clear that the early Holocene samples did indeed document 
that Paleoindians were different from late Holocene and extant populations. 

Once it was apparent that the differences between the earliest populations in the New 
World differed from the most recent New World populations, the question became, what 
evolutionary forces, or a combination of forces, created the differences? Could we 
assume that the similarities principally were shared feamres with the ancestors? 
Alternatively, could the differences principally be caused by forces such as genetic drift 
or natural selection acting upon the New World populations as they adapted to new envi- 
ronments, new biological communities, and new lifest}des? 

Until recently, the most widely held view was that if the feamres of the face and 
braincase of two skeletal samples, separated in time or space, resembled one another, the 
similarities reflected a genetic link. Such similarities between two species separated in 
time have been the foundation for interpretations by paleontologists unearthing the his- 
tory of all animals. A related inference has been that natural selection has been the pri- 
mary forced creating the differences which have developed through time. These assump- 
tions are reasonable at higher levels of generalizations such as when comparing two 
species or genera. 

Adopting these inferences in smdying the colonization of the Americas, the resem- 
blances of the recent Northeast Asians and American Indians were assumed to reflect 
their genetic ties to one another. When the oldest samples were also thought to resem- 
ble the more recent American Indians and Northeast Asians, as Hrdlicka concluded, the 
evidence was complete. If the descendants looked alike and the ancestor looked like the 
descendants, the only inference considered likely was that they were related, and that the 
most ancient of the samples represented the founding colonizers. 

When the evidence suggests the founding population resembles more closely anodi- 
er population, such as Southeast Asia, or a more generalized ancestral Asian population, 
the reasoning still applies. The assumption is that the founding population is genetically 
linked to the one it resembles. In this model, the Northeast Asians and American Indians 
are assumed to have in common a more recent colonizer to the New World than either 




■> ° 

M O 

2 J 

■^ c 

u > 



Q Ji 

3 D 

.. !2 

is aZ 

gl a a § 
Q « a, aA 

« 3 _ 

.2 c 


o ^ u 
i—i .a 14H t 



b! S i> ^ § 9 
D a, rt-0 j: 

■0 e }j 3^^ 

M O O- 0. aj 

g « & ;a ^ 

II si 1.1^ 

S03(i,<ISl3Z SaZ.1 
























'5 "0 


3 "2 

^ 3 Si -I 

.2 5 c'> 

rt u « 

O C 

fi M' 

PhmS; aT! 3^ 































has ro the earlier colonizers who more closeh' resembled rhe Southeast Asians. The pic- 
ture may become even more complicated if there has been ^ene flow between the first 
founding population and later arrivals and their descendants. However, the guiding 
assumption is the same, the resemblances seen are presumed to reflect genetic ties. 

]>ahr (1995) examined the underlying assumptions of this working principle using 
two New World populations. She compared the craniofacial features of historic South 
American hunters and gatherers of Tierra del Fuego, a population near the Antarctic 
Circle, to Eskimos near the Arctic Circle. Both populations are adapted to cold, dry con- 
ditions, vet in all models each is derived from a separate Pacific Rim population. In both 
populations she found similar superficial craniofacial resemblances which she inferred 
were the results of adaptation through natural selection to similar, but geographically 
separated cold environments. The overall shape and size of the face and braincase of the 
Tierra del Fuegans, however, resembled most closely southern Pacific Asian populations 
rather than South American populations or Eskimos. 

Lahr's research then, supports models that envision basic shape and dimensions of 
the face and braincase as being relatively conservative and tending to reflect ancestral 
descendant relationships while natural selection, acting over a relative short term, molds 
more superficial features. Addressing the issue of the colonizing of the Americas specif- 
ically, Lahr's evidence supports a model that: 1) recognizes the distinctiveness of the 
Paleoindian samples, and infers that the ancestors for Paleoindians were either from 
Southern Asia or southern Pacific Rim, or possibly representative of a proto-Asian pop- 
ulation with more generalized features; and 2) subsequent changes to American Indian 
populations craniofacial features reflect gene flow from more recent Northeast Asian 
ancestors as well as the possible effects of natural selection acting upon local populations 
adapting them to local conditions. 

While Lahr's work indicates how natural selection may work in conjunction with the 
transmission of genes through time, the power of genetic drift and gene flow to shape 
small isolated populations as they spread into a new world was not examined. Powell and 
Neves (1999) in a review of the craniofacial morphology of the First Americans did 
address the effects these evolutionary forces could have on early American populations 
and their descendants. Specifically, they assessed whether the pattern of resemblances 
observed between early, middle, and late Holocene skeletal remains, which we outlined 
above, also could support a model that attributed these differences to evolutionary forces 
such as genetic drift combined with gene flow. 

They proposed that genetic drift acting on small isolated founding populations could 
create as great or a greater amount of variation within the early and middle Holocene as 
seen in late Holocene and extant populations. More recently in time, they suggested it 



would be possible for the isolating mechanisms separating the populations could break 
down and gene flow could tie the populations to one another and create the range and 
pattern of genetic variation we see in American Indians during Late Prehistoric and his- 
toric times. In this model then, the founding population could have had the craniofacial 
shape and dimensions that we and others have found, and then as the founding popula- 
tion expanded into the New World small populations became isolated, drift occurred, 
followed by breakdown of the isolating mechanisms which could have resulted in the 
configuration of the American Indian populations in historic times. 

The foundation for this model is Wright's (1951) measure of within-group genetic 
variation (Fsf) of extant populations modified to permit estimates of heritability of cran- 
iofacial dimensions, shape, and the estimates of the size of prehistoric populations (see 
Powell & Neves 1999 for a discussion of procedures). Using these modifications, Powell 
& Neves (1999) first calculated the Fst, Wright's (1951) measure of within-group varia- 
tion, and used this to estimate between group variation by combining population into 
one sample and estimating that group's Fj/. In cases where the Fj/ of the combined sam- 
ple was in the range of the expected within-group variation for a single population, it was 
estimated that these two samples could represent a single population. If the Fst was 
much greater than the average Fst, it was inferred that the two samples represented sep- 
arate populations, and including them within one sample was an artificial construct. 
When Powell and Neves (1999) computed these values for a selection of world popula- 
tions, and New World Paleoindians, the Paleoindians had a much higher divergence than 
the degree of variation found within later Holocene remains. This supported the 
assumption that the Paleoindians were distinct from American Indians. 

The distinctiveness of the Paleoindians from other populations also was determined 
using Relethford and Blangero's (1990) modified technique of estimating within-group 
variation. In this analysis, the difference between the average variance for all samples was 
compared to the variance within the individual populations. Using the difference as a dis- 
tance measure between samples indicated how likely they were to be members of a sin- 
gle population. Figure 14 (1999) illustrates the distribution of world samples and 
Paleoindians using two different estimates of population size for Paleoindians in the 
computation of within-group variation. In Figure 14A the population size was estimat- 
ed to be equal to all other samples, while in Figure 14B Paleoindian population size was 
estimated to be 0.30 the size of more recent populations. In Figure 14A both Eskimo 
and Paleoindian samples are distinct from all other samples in terms of variance from 
the average indicating they represent separate populations from the others. When the 
amount of within group variance is recalculated using a smaller estimated population size 
for the Paleoindians alone, their variance falls within the range of variation of the other 






-^ A ^ 

O O 

A A 

A A 

A ^ 





Figure 14. Principal coordinates ordination of New World samples using R matrix data from a 
Relethford-Blangero analysis of 14 variables. A shows the distance between any two points reflects their 
differences if all samples were considered equal in size and rate of expansion. B reflects differences 
between samples when North and South American population sizes are estimated as 0.3 of the size of 
other populadons. North American Paleoindians (soUd stars), South American Paleoindians (open 
stars), more recent American Indians (solid triangles). Pacific Northwest Indians (open circles), and 
Eskimos (solid squares). (After Powell and Neves 1999.) 

samples. The Eskimo sample remains distinct since its population size was not adjusted 
in Figure 14B. Assuming that these relative estimates of variance reflect the balance 
between genetic drift and gene flow acting on a population, then Figure 14B indicates 
that the distinctiveness of the Paleoindians could have been achieved through genetic 
drift, rather than assuming that the Paleoindians represent a different founding popula- 
tion. Figure 14A however, indicates that some other cause is creating the extent of vari- 
ance from the average of the populations, plausibly that Paleoindians, like the Eskimo, 
were derived from a different founding event. 


In reviewing the past decade of research on the skeletal remains of the earliest 
known inhabitants of the New World, we feel confident that the early American 
Holocene samples are distinguishable from the late Holocene and extant American 
Indians. It has also been documented that the South American and North American 
Paleoindian samples look more like one another than either does to living American 
Indians, yet they differ to a lesser degree among themselves as well. And research has 
documented that these earliest remains more closely resemble mid-Holocene remains 
than thev do late Holocene remains. Wliile all of these conclusions have been based on 



very small and incomplete samples, the consistent pattern in the way the early Holocene 
remains differ from more recent populations gives us confidence that these differences 
are real and not a result of sampling error. Achieving this level of knowledge about the 
First Americans has been an achievement. 

What has remained more equivocal, however, is an understanding of the cause or 
causes of this variance. The two basic competing explanations for the causes of the vari- 
ation are, that the distinctions reflect different ancestry, or that the distinctions are the 
result of genetic drift and subsequent gene flow within and between populations once 
they arrived in the New World. Both explanations have ardent supporters, and both have 
evidence to support their arguments. It is also true that there are weaknesses in both 

Even we, who have collaborated over a period of 10 years in our research on the 
topic, do not envision the same history and process for the peopling of the Americas. 
We do, however, agree on the broad outline of the events. We agree that the evidence 
does support the view that the Paleoindians were structurally distinct from most recent 
American Indian populations. One of us (DGS) believes that the features of the earliest 
remains generally reflects the features of their Old World ancestors, either a more gen- 
eralized Northeast Asian population prior to the development of the recent Northeast 
Asian feamres; or a more central or southern Asia population, one similar to the Jomon 
of Japan. Both authors accept the evidence of the similarity of the Paleoindians and the 
mid-Holocene samples as evidence that these early Holocene peoples contributed at least 
some of their genes to mid-Holocene populations, and plausibly to some more recent 
North and South American Indians as well. 

When we look from the perspective of today's and Late Prehistoric American Indian 
populations and look back through time, one (DGS) of us sees evidence of more recent 
Northeast Asian colonists adding their genes to the pool that will become the North 
American Indians of today. The other author (JFP) sees that as a possibility^ as well, but 
also sees the possibility that much of the more recent American Indian features are the 
result of genetic drift and gene flow shaping the first founding population to be the 
North American Indians of historic and present times. What is most apparent to us 
both, however, is that while we and other scholars have documented the craniofacial dif- 
ferences between the earliest known Americans and Late Prehistoric and present day 
American Indian populations, the processes that have created these differences have yet 
to be unequivocally revealed. 


S'li:UlJ' AND POWELL tract 

The oldest hiimcin skeletal remains found in North America share broad sim- 
ilarities ivith northern Asians, Southeast Asians, Pacific Islanders, and recent 
Native An/erica ns. /{iiiong these populations North America' s earliest remains 
resemble more closely South East Asians and some Pacific Island populations 
than they do northern Asians and Native Americans. The modest but consistent- 
ly documented differences between the early North American remains and recent 
North American Indians has challenged our traditional vieiv that the Americas 
ivere coloni-:{ed by Northeast Asians in a time too recent for evolutionary forces to 
have altered the features of their Native American descendants. W^hile the dif- 
ferences between the earliest and more recent skeletal remains have become well 
documented, identifying the evolutionary force or forces ivhich created these differ- 
ences is more equivocal. 

l^iterature Cited 

Armelagos, G. J., D. S. Carlson, & D. P. Van Gerven. 1982. The theoretical foundations 
and development of skeletal biology. Pages 305—328 in F. Spencer, ed., A History of 
American Physical Anthropology, 1930—1980. Academic Press, New York, NY. 

Chatters, }. 2000. The recovery and first analysis of an early Holocene human skeleton 
from Kennewick, Washington. Am. Anthropol. 65:291—316. 

Hanihara, T. 1996. Comparisons of craniofacial features of major human groups. Am. J. 
Phys. Anthropol 99:389-412. 

Howells, W. W 1969. The use of multivariate techniques in the study of skeletal popula- 
tions. Am. J. Phys. Anthropol 31:311-314. 

. 1973. Cranial variation in man: A study by multivariate analysis of patterns of 

differences among recent human populations. Paps. Peabody Mus. Arched Hthno. 67, 
Harv Univ. Cambridge, MA. 

. 1989. Skull shapes and the map. Craniometric Analvses in the Dispersion of 

Modern Homo. Paps. Peabody Mus. Archeol Hthno. 79, Harvard Universit}; Cambridge, 

Hrdlicka, A. 1902. The crania of Trenton, New Jersey and their bearing upon the antiq- 

mx^J of man in that region. Bull. Am. Mus. Nat. Hist. 16:23-62. 
. 1907. Skeletal remains suggesting or attributed to early man in North America. 

Am. Bun Am. Ethno. Bull 33:1-113. 
. 1918. Recent discoveries attributed to earlv man in America. Am. Bun Am. 

Ethno. Bull 66:1-67. 



. 1923. The origin and antiquity of the American Indian. Pages 481—493 in 

Annual Rfport of the Board of Regents of the Smithsonian Institution. U.S. Government 
Printing Office, Washington, D.C. 

1937. Early man in America: What have the bones to say? Pages 93—104 in G. 

G. MacCurdy, ed., Early Man as Depicted by leading Authorities at the International 
Symposium at the Academy of Natural Sciences of Philadelphia. J. B. Lippincott, 
Philadelphia. PA. 

Jantz, R. E., & D. W. Owsley. 1997. Pathology, taphonomy, and cranial morphometries 
of the Spirit Cave mummy. Nev. Hist Quart. 40:62-84. 

Lahr, M. M. 1995. Patterns of modern human diversification: Implications for 
Amerindian origins. Yearbook of Physical Anthropology 38:163198. 

MacManamon, E P. 2000. Determination that the Kennewick human skeletal remains are 'Native 
American " for the purpose of the Native American Graves Protection and Repatriation Act 
(NAGPRA). Electronic copy of the original United States Department of Interior, 
National Park Service memorandum to Assistant Secretary, Fish and Wildlife and 
Parks, February 14, 2000. Internet citation: 

Neves, W A., & H. M. Pucciarelli. 1989. Extra-continental biological relationship of early 
South American human remains: A multivariate analysis. Ciencia e Gultura 41:566—575. 

. 1991. Morphological affinities of the first Americans: An exploratory analysis 

based on early South American human remains./. Hum. Evol 21:261273. 

. 1998. The Zhoukoudien upper cave skull 101 as seen from the Americas. /. 

Hum. Evol 34:219-222. 
Neves, W A., D. Meyer, & H. M. Pucciarelli. 1996. Early skeletal remains and die peo- 
pling of the Americans. Revista de Antropologia 39:121-139. 
Neves, W A., D. Munford, & M. C. Zanini. 1996. Cranial morphological variation and 

the colonization of the new world: Towards a four-migration model. Am. J. Ploys. 

Anthropol, Suppl. 22:176. 
Neves, W A., J. E Powell, A. Prous, & H. M. Pucciarelli. 1998. Lapa vermelha IV, 

hominid I: Morphological affinities of the earliest known American. Am. J. Phys. 

Anthropol, Suppl. 26:169. 
Neves, W A., J. E Powell, & E. G Ozolins. 1999. Modern human origins as seen from 

the peripheries./. Hum. Evol 34:96—105. 
Neves, W A., M. C. Zunimi, D. Munford, & H. M. Pucciarelli. 1997. Povoamentio da 

America 4a luz da morfologia crania. Revista USP 34:96-1 05m. Sao Paulo. 
Ozolins, E. O., V. H. Stefan, M. L. Rhoads, & J. E Powell. 1997. Craniofacial morpho- 

metric similarity' between modern and Late Pleistocene human populations. Am. J. 

Phys. Anthropol 24:182-183 



Powell, ). F. 1993. DcMital evidence for rhe peopling of the new world: Scjme method- 
ological considerations, th/m. Biol 65:799—815. 

. 1995. Dmtcil variation and biological ajjinity among middle Holocene hitman populations 

in North America. Ph.D. dissertation, Department of Anthropology, Texas A&M 
University, College Station, TX. 

Powell, J. E, & W. A. Neves. 1999. Craniofacial morphology of the first Americans: 
Pattern and processes in the peopling of the new world. Yearbook of Physical 
Anthropology 42: 1 53-1 98. 

Powell, J. P., & }. C. Rose. 1999. Report on the osteological assessment of the 
"Kennewick man" skeleton (CENWW.97.Kennewick). Internet citation: 

Powell, J. P., & D. G. Steele. 1992. A multivariate craniometric analysis of North 
American paleoindian remains. Cur. Kes. Pleist 9:59—62. 

Relethford, J. H., & J. Blangero. 1990. Detection of differential gene flow from patterns 
of quantitative variation. Hum. Biol 62:5-25. 

Steele, D. G. 1989. Recendy recovered paleoindian remains from Texas and the south- 
west (abstract). Am. J. Phjs. Anthropol 78:307. 

Steele, D G, & |. E Powell. 1992. Peopling of the Americas: Paleobiological evidence. 
Hum. Biol 64:303-336. 

. 1993. Paleobiology of the first Americans. Evol Anthropol 2:138-146. 

. 1994. Paleobiological evidence of the peopling of the Americas: A morpho- 

metric view. Pages 141—163 in R. Bonnichsen & D. G Steele, eds.. Method and Theory 
for Investigating the Peopling of the Americas. Peopling of the Americas Publications, 
Center for the Study of the First Americans, Oregon State Universit)", Corvallis, OR. 
1999. Peopling of the Americas: A historical and comparative perspective. 

Pages 91-120 in R. Bonnichsen & R. Gruhn, eds., Who Were the First Americans. 
Peopling of the Americas Publication, Center for the Study of First Americans, 
Oregon State Universit}^, Corvallis, OR. 

Tuohy, D, & A. Dansie. 1997. New information regarding early Holocene in the west- 
ern great basin. Nev. His. Quart. 40:24—53. 

Wright, J. W. 1951. The genetic structure of populations. Ann. Eug 15:323-354. 

Young, D. H., S. Patrick, & D. G Steele. 1987. An analysis of the paleoindian double bur- 
ial from horn shelter no. 2, in central Texas Plains. Anthropology. 32:275—299. 


Chapter Six 

Teeth, Needles, Dogs, and Siberia: 

bloarchaeological evidence for the 

Colonization of the New World 

Christy G. Turner II 

^^^^^ There are many improbable and unscientific views dealing with the 
^^^^^^k origins and peopling of the New World, such as autochthonous cre- 
^ . r-^ B^^^^B ation (numerous Native American religious traditions), lost Israelites 
I^Kp^^ (Mor monism), and unbelievable claims such as space alien colo- 
nization. On the other hand, there are hundreds of thoughtful, 
empirically-based or theoretically-oriented scholarly contributions to understanding the 
colonization of the New World. These fall readily into two classes of temporal view- 
points, Clovis first or pre-Clovis. (Bonnichsen & Turnmire 1991; Dillehay 1999; Dillehay 
& Meltzer 1991; Hall 2000). There are one major and two minor schools of thought on 
origins and routes: Most likely Beringian, much less likely trans-Pacific or trans-Atlantic, 
and impossibly autochthonous. There is a fairly unrancorous set of ideas on the number 
of migrations, basically one to four (Gibbons 1996), with much, much later arrivals start- 
ing around A.D. 1000 with the Vikings whose arrival from Europe via Greenland 
depended on large sailing boats (Boorstin 1983). There is a generally coherent group of 
findings on Beringian paleoenvironmental conditions, resources, hazards, and opportu- 
nities: Arctic steppe with patchy Beringian resources, but megafauna present (Guthrie 
1990); mountain forest-steppe and steppe (Laukin 2000); and coastal maritime (LaughUn 
& Harper 1988). Finally, there is a small but challenging class of ideas concerned with 
learning why the colonization of northern Siberia and Beringia ever occurred in the first 
place, and why it seems from most archaeological evidence to have been so late in the 
Pleistocene history of human dispersal (Pagan 1990; Soffer 1990; Yesner 1998). 

Each of these major and minor domains, be they based on diachronic or synchron- 
ic evaluations, or both, has been championed by workers using primarily a single line of 
evidence. For example, nuclear or mitochondrial DNA (Szathmary 1993); Y chromo- 
somes (Hammer et al. 1999); diachronic studies on crania (Powell & Neves 1999), dental 
morphology (Turner 1992a), and stone tools pillehay & Meltzer 1991). AU try to assess 
Old World origins and affinities: timing, derived from dates of early archaeological sites 



or typological correlations; lant^uagc classification for numljcr of migrations; and so 
forth. However, integrating diverse evidence, in the growing tradition of modern evolu- 
tionary studies, requires that all or most of the domains of evidence be weighed and con- 
sidered together, if not by individuals, then by teams of specialists (Bonnichsen & Steele 

Operadonally, holisitic endeavors udlize at least tv/o scientific principles of evidence 
evaluation — parsimony and concordance. The best single h\-pothesis will be the one 
that integrates the most information (holistic), with the fewest number of assumptions 
(parsimomy), and where several lines of evidence lead to the same or similar inferences 
(concordance). In addition, the quality^ and t}^es of evidence should meet reasonable 
standards of robusticity^, redundancy, independence, and consensus. Realit}', ultimately, is 
that which we agree it is. 

On the other hand, as in all areas of scientific investigation, there can be curious evi- 
dence that fails to fit any existing or imaginative pattern, yet seems intuitively to have 
merit. For example, with respect to the later peopling and culmral evolution in the 
Americas, the possibility of small numbers of trans-Pacific migrants such as rare Chinese 
or Jomon boating events refuses to go away (Meggers 1971). At the present time, in my 
long-standing view, three Late Pleistocene migration events of presumably already bio- 
culturally differentiated stocks can be inferred from a multidisciplinary data base: earlv 
Beringian interior Paleo-Indians (called variously Clovis, "epi-Clovis," Macro-Indian, 
Chinese-Siberian microlithic), later interior Beringian Paleo- Arctic people (Greater 
Northwest Coast, Na-Dene, Diuktai), and south Beringian coastal proto-Aleut-Eskimos 
(lower Amur, Hokkaido). I introduce the term epi-Clovis in recognition that the haUmark 
object in Clovis identification, the fluted spear point, had probably not been fuUv devel- 
oped in the first bands to reach Alaska (such as Component I at Dry Creek, often 
referred to as the Nenana Complex (Hoffecker et al. 1993); however, other correspond- 
ing stone tool types are present in Clovis and the Alaskan epi-Clovis (Hoffecker et al. 
1993:51). Similar tool types are known for the Chinese microlithic tradition that is 
strongly like the widespread Northeast Siberian Diuktai and Alaskan Paleo- Arctic tradi- 
tions, both of which are viewed by Russian archaeologists as derived from north China. 
With these introductory remarks and suggestions of my own inferential preferences, let 
us look at what I consider to be liie eight major questions that, viewed together, provide 
a broad multidisciplinary working hypothesis for the human colonization of the 



Where Was the Kegional yincestral Homeland{s)? 

Aside from the obvious geographic proximity of Alaska and Siberia, which has long 
suggested the proximate homeland of ancestral Native Americans, there are other lines 
of evidence that can be brought to bear on the question. One of these is dental mor- 
phology that has been used for more than a century to characterize and classify mam- 
malian species, assess intergroup genetic relationships, and trace migrations. With respect 
to the peopling of the New World, HrdUcka (1920) was among the first to propose an 
Asian homeland for ancestral Native Americans employing dental morphology, espe- 
cially the shovel-shape polymorphism of the upper incisors. ShoveUng is very common 
in all Native Americans and Northeast Asians but is much less common, to very rare, 
elsewhere in subfossil and modern world populations. Subsequent studies by Pedersen 
(1949), Moorrees (1957), K. Hanihara (1968), Zubov and I<aialdeeva (1979), T Hanihara 
(1991), Turner (1991, 1992a, 1994b), Scott (1994), and many others, have added to the 
dental evidence pointing to an Asian origin for all Native Americans. At a conference 
held in Khabarovsk, U.S.S.R. in 1979 I proposed that East Asia contained two geo- 
graphic variants of the "Mongoloid dental complex" defined by K. Hanihara (1968). 
These were a Southeast Asian pattern of retained and simplified feamres I called 
Sundadonty, and in Northeast Asia the morphological pattern was more specialized and 
complex. This latter pattern I called Sinodonty (Turner 1983, 1987, 1990a). I have found 
only Sinodonty in prehistoric New World teeth. Figures 1—3 show examples of dental 
morphology used in local, regional, continental, and intercontinental comparisons that 
define Sinodont}^ and Sundadonty. Figure 4 shows in synoptic multivariate fashion that 
all Native Americans, ancient and recent, are more Uke one another than like Ainu- 
Jomonese, recent Japanese and Chinese, Polynesians, and Europeans. The close dental 
relationship between the 9000 year-old Archaic period Spirit Cave and Wizards Beach 
crania of Nevada, and recent Paiute Indians of Nevada indicates dental continuit}^ 
through time. Figure 5 shows that the dental relationship in the Americas lies with peo- 
ples of the American Arctic, Northeast Asia, and Neolithic Southeast Asia (a result of 
migration southward of northerly agricultural Sinodonts). There are no close relation- 
ships with modern Southeast Asia (Thailand), Europe, Oceania, or Africa. Like many 
other smdies of teeth and crania, Australians and Africans are closely linked but show 
no relationship to Native Americans. Hence, on the basis of dental morphological com- 
parisons, the homeland of Native Americans continues to be best viewed as having been 
in Northeast Asia. To test this hypothesis requires demonstrating that only Northeast 
Asia was the Pleistocene homeland of the Sinodont dental pattern. To date, Sinodont}^ 
has not been found in any non-Asian skeletal assemblage regardless of geological time 



(Zubov & Khaldccva 1979; Cirinc 1981; Turner & Markowirx 1990; Roller 1992, Irish 
1993; Haeussler 1996; lipschultz 1996; Hawkey 1998; Adler 1999; and many others). 
Turner and colleagues (2000) restudied the dental features retained in plaster casts of the 
north (]hina Zhoukoudian Upper (>ave human crania (the original skulls have been lost). 
We concluded that the Sinodont morphological pattern was present, and depending on 
the age of the Zhoukoudian remains (dates range from 11,000 yr B.P. to 30,000 yr B.P.), 
the ancestors of Nadve Americans were Northeast Asian Sinodonts, who, in turn, had 
evolved from Sundadont or Proto-Sundadont stock that earlier began in Southeast Asia 
or Australasia (Turner 1992c; Hawkey 1998). 

In addition to many synchronic anthroposcopic and anthropometric comparative 
studies of living Native Americans and other worldwide populations that show strong 
affinity between North Americans and Northeast Asians, there have been numerous sim- 
ilar but diachronic osteological investigations into Native American anatomy and its 
bearing on origin questions. Most of the osteological studies have been carried out on 
cranial measurements such as head shape (Howells 1995; Steele & Powell 1992), or on 
nonmetric features such as sumral patterning of the skull (Kozintsev 1988) and other 
traits (Ishida 1993; Ossenberg 1994). Most comparative cranial studies closely link Native 
Americans with Northeast Asians, but a few analyses have been interpreted as suggest- 
ing morphological if not genetic affinity of Archaic New World crania with Australia and 
Africa (Lahr 1995; Neves et al. 1999) and Ainu (Pov^ell & Rose 1999). Such scattered and 
unpatterned results do not fit the larger phenetic and cladistic pattern seen in anthropo- 
scopic, genetic, and dental morphology studies. These anomalous craniological findings 
may be due to error arising from small sample size, unknown environmental effects on 
crania (such as the rapid secular change in head form in many parts of the world), and 
dietary and mastication effects on facial architecture and jaw shapes, etc. It is notewor- 
thy that with the introduction of maize agriculaire and pottery from Mexico, the crania 
of prehistoric Southwestern U.S. Pueblo Indians became less robust (Hrdlicka 1931). In 
my view, this Puebloan cranial change was not determined genetically as it could have 
been if caused by migration from Mexico, rather it more likely was related to the soft- 
ening effects of foodsmffs cooked in pots. This softening reduced the need for the he.2.\T\ 
food chewing stress that contributed to cranial robusticit)% just as muscular activit)' 
affects other areas of the skeleton. Moreover these stews and gruels seem to have con- 
tributed significandy to rampant and painful dental caries, even in children, leading to 
even less chewing or biteforce applied to the jaws and face of children and adults alike. 
There are numerous other examples in the New World of sliifts in food preparation 
techniques, even beginning in Archaic times, that could have affected cranial form. 



Figure 1. Illustrated in these upper teeth are (1) the marked expression of incisor shoveling on the 
lingual surface, (2) double shoveling on the central incisors' labial surface, (3) the similarity in size of 
the central and lateral incisors, (4) the near absence of the hypocone on the second molars, and (5) the 
presence of a parastyle on the second molar, a rare feamre. This individual was a relatively young male 
of the preceramic Basketmaker culture. He died before A.D. 400 (late Archaic period). The skeleton, 
excavated by R. WetheriU from Cave 7, southeast Utah, has a dentition much like all other ancient, 
recent, and unadmixed living North and South American Indians and all known populations of north- 
eastern Siberia and Primoria. These dental characteristics, among others, define the Sinodont division 
of the Mongoloid dental complex. Unlike ancient and modern Europeans, where shoveUng is rare, lat- 
eral incisors are relative small, etc., Sinodonty is characterized by trait intensification and specialization. 
European teeth are characterized by simplification and reduction (CGT neg. 6—26—78:16 Am. Mus. 
Nat. Hist. 99/7371). 

While large sample intergroup comparisons of dental and cranial morphology usual- 
ly show strong correspondences (Hanihara 1991), smaller samples have led to very unex- 
pected findings. For example, the previously mentioned three Zhoukoudian Upper Cave 
crania, when smdied individually (Brown 1998; Kaminga & Wright 1988), instead of as 
a group, have led these researchers to see non- or semi-Mongoloid resemblances. 

In the realm of archaeological research, most of the cultural evidence recovered in 
the Americas has been ia the form of artifacts made of stone, the typologies of which 



Figure 2. The middle tooth in these five upper left teeth is the first premolar with the morpho- 
logical crown shape that D.H. Morris called "Uto-Aztecan premolar." The morphology differs strong- 
ly from the second premolar (second tooth on the left in this view), which is the normal form. 
Subsequent to Morri's study, other workers and I have looked for the trait throughout the Americas. It 
has never been found in Aleut-Eskimos or Indians in the Na-Dene area. It is believed to be due to a 
mutation carried to the New World only by Paleo-Indians, one of the facts which I feel indicates that 
the three New World dental groups had differentiated in Siberia before crossing Beringia. The Uto- 
Aztecan premolar has never been found in Europeans. The pictured individual was a young female who 
lived sometime before A.D. 1700. She was excavated by J.W Fewkes from Awatovi, northeastern 
Arizona (CGT neg. 6-10-78:16 Nat. Mus. Nat. Hist 156324). 

indicate a northern Eurasian homeland, but with no clear-cut regional sourcing identifi- 
able so far (Mochanov 1977; Derev'anko 1990; VasH'ev 1996; Dikov 1997). Blade tools, 
bifaces, burins, adzes, spurred scrapers, chipping methods, and other technological fea- 
tures are more common in early New World sites than are the less formalized tools of 
Southeast Asia, Australmelanesia, and the late Mousterian industries of Europe and 
southern Siberia. Late Pleistocene unifacial blade tools occur in the Amur River basin 
and on Hokkaido that match well with the earliest stone artifact t}-pes found in the 
Aleutian Islands. Throughout much of Late Pleistocene Northeast Siberia, Russian 
archaeologists have recovered microblades and bifacial tools together, an assemblage 



usually referred to as the Diuktai culture (Mochanov 1977; Dikov 1997). This distinctive 
combination matches assemblages called Paleo-Arctic in interior Alaska (Dumond 1987) 
and British Columbia (Carlson & Bona 1996). Hence, stone tool assemblages point to at 
least two cultural homelands for ancestral Native American cultures — Amur-Hokkaido, 
and Primoria-Yakutia-Chukotka. However, so far only one fluted point has been recov- 
ered in northeast Siberia (by archaeologists M.L. King & S.B. Slobodin) in good strati- 
graphic context, but it is not old enough to be considered ancestral to Clovis. In fact, 
there is presently lacking any exactly corresponding Siberian assemblage Uke that well 
established for Clovis, whose hallmark is the fluted point, or the early assemblages of 
Alaska-like Component I of Dry Creek (Hoffecker et al. 1993). For this reason, Stanford 
1999, this symposium) argues for a European origin for Clovis on the grounds that 
Solutrean spear points are more similar to Clovis than anything known for Siberia, north 
China, or Japan. This suggestion has not been well received on both temporal and tech- 

FiCiURE 3. Root form and number is also involved in the definition of human dental patterns, 
although inheritance is less well understood than is crown morphology. Pictured are right lower molars 
with distolingual supernumerary roots (arrows) that are rare in ancient and modern Europeans, but pre- 
sent commonly to frequently in northeast Asians. In the New World, Aleut-Eskimos have the highest 
frequency of this trait [ca. 30—40%); Northern Northwest Coast, intermediate [ca.. 15%); all other 
Indians in North and South America about the same [ca. 5.0%). Again, this hemispheric variation is 
viewed as evidence of three migrations. Pictured is a Shang Dynasty subadult from An-yang, China 
(CGT neg. 9—5—75:36 [color] 122 Academia Sinica, Taipei). 



nological grounds (Straus 2000; G.A. Clark, personal communication, January 4, 2001). 
That there may have been occupation of the New World before Clovis, that is, before 
11,500 yr B.P. condnues to be a hotly debated subject (West 1996; Hall 2000; Roosevelt 
et ai, this volume; Meltzer, this volume). 

Many Native American cultures have oral traditions that envision their ancestors as 
having originated in the New World (Echo-Hawk 2000). There are many well known ori- 
gin legends involving migrations, such as among the Hopi and Navajo, but these are of 
much less historical value than they are for cultural patrimony. Despite the lack of his- 
toric specificit}^, there are oral traditions of heroic shamans and their spirit-helpers, which 
when combined with ethnographic features such as the round flat shaman's drum found 
throughout Siberia and the Americas, points to a north Eurasian homeland for all Native 

In sum, the ultimate regional homeland can reasonably be hypothesized as having 
been in north China, Mongolia, and southern Siberia, a rich ecologically diverse region 
occupied long before humans set foot further north. An easy to remember focal point 
of this region is Lake Baikal, near the headwaters of the Anadar, Ob, Yenisei, Lena, and 
Amur rivers that together drain much of central and eastern Siberia. 




















Figure 4. Dendrogram of worldwide dental reladonships based on the muldvariate Mean Measure 
of Divergence stadstic clustered with Ward's method involving 25 crown and root traits belonging to 
1000s of individuals. As discussed in the text, the dendrogram shows a strong similarit}- benveen den- 
titions of the New World and those of Northeast Asia. There is no close relationship between Native 
Americans and Europeans or Jomon-Ainu (Computer file reference: Small World— 94, 1 9 groups with- 
out central Asians). 





3-CUSP P2 

l-ROOT M2 
CUSP 6 Ml 
l-ROOT Pi 
LaR M3 + 

0^ 20 40 60 80 



Figure 5. Shown are some of the dental traits and their frequencies used to define Sundadont}' and 
Sinodonty. The former is a dental patttern characterized by retention and simpHficadon, the latter by 
intensification and specialization. For example incisor shoveling is more frequent and more pronounced 
in Sinodonts (intensification), less so than in Sundadonts (retention). As individuals age, their dental 
crowns wear out and their roots can acquire excessive cementum, both age-related conditions can con- 
tribute to under-scoring of features such as shoveling, cusp and root numbers, and other traits. 

What Was the Route(s) to the New World? 

Most lines of evidence point to a Bering land bridge route for the initial colonizers 
of the New World. Some of this evidence has been touched on in the homeland con- 
sideration. It is not aU of equal value. However, the pattern and its independent corrob- 
oration is clear and strong. Seemingly, some workers who favor a pre-Clovis settlement 
of the Americas are trying to get around the lack of evidence for a pre-Clovis occupa- 
tion of Northeast Siberia and Alaska by seeking other routes or means of entry. A Pacific 
coast boating scenario was proposed years ago by Fladmark (1978) that would have left 
littie or no pre-Clovis evidence anywhere due to the 100 m rise in sea level at the end of 
the Pleistocene. A similar suggestion, based on a typological similarity between Clovis 
and Solutrean points, for a trans-Atlantic route (Stanford 1 999) would likewise have left 
no evidence if the southern edge of the Atiantic pack ice was followed into North 
America as he suggests. And suggestions for a trans-Pacific route from Australia has the 



same problem of nor l^eing disprovable due to the fl(K)ding oi the continental sheh of 
South America where landings might have taken place. Lacking independent cornjbora- 
tion or conformity with the generally agreed upon Bering land bridge hyp(jthesis, 
claimants need to not only pursue their claim by showing how the prevailing Beringian 
hypothesis is wrong, but also develop the means of falsifying their own alternative pro- 

The Bering land bridge route could have involved not only an interior crossing, but 
also a south coastal cirift or both. There could also have been reladvely early and later 
transits as well, and these could have been more-or-less discrete movements, or a more 
drawn-out and semi-condnuous flow of people (Turner 1985; Scott 1991). The prepon- 
derance of evidence points to a Bering land bridge route for the initial colonization of 
the Americas. On the other hand, at least two thorny problems are evident: First, defin- 
ing archaeologically and biologically what is meant by "migration" has not met with 
much success. While there is a large body of sociological and demographic facts and the- 
ory on migration, it has not been well reviewed by workers interested in the peopling of 
the Americas. Second, while it is reasonable to equate cultural variation with biological 
variation in early small isolated groups, such is not always the case. Because there are 
almost no Pleistocene human remains from archaeological sites in far Northeast Siberia 
(N.N. Dikov provided me with photographs of a few deciduous but racially indiscrimi- 
nate teeth he found at Ushki on the Kamchatka Peninsula), and none in Alaska, we are 
left with mainly stone tool variation and site distribution patterns as hints of biological 
populations. There may well have been very Uttie genetic variation in the waves of 
migrants to Alaska, even though there is substantial geocultural variation, that is, early in 
the settlement of Alaska alone, three traditions can be recognized. At Dry Creek, near 
Fairbanks, the earliest horizon, Component I (ca. 11,000 yr B.P.), has blades and bifaces 
but no microblades; Component II (ca. 10,500 yr B.P.) has microblades in addition to 
blades and bifaces (Powers & Hoffecker 1989; Hoffecker eta/. 1993). In the Aleutians, at 
Anangula (ca. 8,500 yr B.P), there are no bifaces, only unifacial blade tools ranging in size 
from micro- to macroblades (LaughUn 1980). 

When Was the Time of Arrival 

To date, no archaeological sites have been found in Alaska tiiat are older dian 12,000 
yr B.P. (Hoffecker et al. 1993), although there are controversial claims for earlier sites fur- 
ther south such as Meadowcroft in Pennsylvania and Monte Verde in Chile (reyie\\-ed in 
Meltzer [1993] and in nearly every issue of Mammoth Trumpet). Much has been written by 
Russian scholars about the ancient, historic, and modern peoples of Siberia and their 



regionally variable environment and climate (see reviews or translations written in 
English or French in Derev'yanko 1990; Haeussler 1996; Levin & Potapov 1964; Malet 
1998; West 1996). As a generalization, climate is much more stable, less severe, and 
archaeological sites are much earlier and more abundant in southern Siberia than in the 
far north. In the Altai Mountains west of Lake Baikal and north of Mongolia, for exam- 
ple, there are later Pleistocene sites such as Kaminnaya, Ust-Kan, Denisova and 
Okladnikov caves with rich deposits of Mousterian stone artifact types and occupation 
debris that have been dated to at least 30,000 years ago. One thermoluminescence date 
for a deep layer containing Mousterian stone artifacts at Denisova Cave is 151,000 to 
170,000 yr B.P (M. Shunkov, personal communication, July 14, 1999). The oldest 14 C 
date for humans in Kamchatka was obtained at Ushki, being about 14,000 yr B.P. 
(Hoffecker et al. 1993). The 10,260—13,700 year-old Berelekh site is one of the very few 
generally accepted Siberian sites north of the Arctic Circle (Sinitsyn & Praslov 1997; 
Kuzmin 1997, 2000). Since no assemblage of Mousterian stone artifact t\^es have been 
found in Siberia near the Arctic Circle or in any New World site, it would appear that far 
northern Siberia was uninhabitable until sometime around 15,000 years ago. This is the 
time when the world's cUmate began to warm up rapidly, causing continental ice masses 
to melt, sea levels to rise, vegetation zones to shift or go extinct, associated animals to 
go extinct, and other drastic environmental changes throughout Beringia and most other 
regions of the world (Martin & I-Gein 1984). The far northern Siberian and Alaska Arctic 
steppe habitat of dry cold-adapted grasses, sage brush and other shrubby flora ceased as 
did nearly all the megafauna this habitat supported. Thick forest taiga and treeless boggy 
tundra replaced the Arctic steppe, with a resulting reduction in large herbivore and car- 
nivore biomass, and the formation of vast hummocky plains difficult to walk or sled 

One of the many interesting taphonomic findings my Russian co-workers Nicolai 
Ovodov, Olga Pavlova, Nicolai Mart}^novich, and I made in the past three summers 
(1998-2000), was that of all the several thousands of food refuse bone fragments we 
examined from several Late Pleistocene sites in the Altai and Lake Baikal regions, we 
identified very few clear-cut examples of burned bone. Said another way, while we iden- 
tified many excellent examples of stone tool butchering marks and perimortem marrow- 
exposing bone breakage, we found extremely little taphonomic evidence of cooking. 
Moreover, these limestone caves with their marvelous qualities for preservation have 
produced only tiny residues of charcoal, despite paleoenvironmental reconstructions 
based on small and large rodents, such as beaver, that indicate there was forest on the 
ancient landscape that would have been available for firewood. Is it possible that the 
more northern reaches of eastern Siberia were unreachable because these Late 



Plcislocc'iic Moustcrians made ()nl\- limited use oi fire? While it is likeK thai ihe\ made 
fires for cooking, warmth, and repelling insects, it is not well documented thus far in 
their refuse. My point regarding fire is simply this: Those ethnographic and technical fea- 
tures that characterize native populations north of the Arctic Circle at historic contact, 
are not readily apparent in the Mousterian horizons of southern Siberia, whereas thev are 
present in the region's Upper Paleolithic strata in both cave (Kaminnaya, Ust-Kan, 
Denisova, etc.) and open sites (Mal'ta, Kurla I, Gosiderev Log-1, etc.). One reasonable 
inference of these preliminary taphonomic findings is that the far north was uninhabit- 
able by humans until certain developments in technology had come about. 

We also determined from taphonomic analysis of animal bone damage and numer- 
ous well-preseved coprolites that humans were not the sole residents of these Altai and 
Khakatian caves. At various times they were inhabited instead by hyenas and other car- 
nivores. In addition to the cold and patchy northern zone of Siberia, roving packs of 
hyenas might have been another factor in the failure of the Siberian Paleolithic 
Mousterian folk to reach the Arctic Circle. The hyenas may have, not only scavenged the 
remains of Mousterian kill sites, with various negative implications for the duration of 
human encompment at a given kill site, but these large noctural social carnivores may 
also have preyed on the Mousterians or their cliUdren, especially those bands with low 
population densit}' along the northern periphery of their Siberian distribution. Other 
carnivores that Ovodov has identified in many archaeological and paleontological cave 
deposits include lions, tigers, bears, wolves, and various other smaller meat-eaters. 

In sum, my growing knowledge of Siberian prehistory based on smdying published 
archaeological reports, our various site visits in the Ob, Yenisei, and Angara River basins 
and related travel, study of artifact, faunal, and human skeletal collections from these and 
other areas, and some excavation, continues to suggest that there was a very severe envi- 
ronmental barrier to humans reaching western Beringia until almost the end of the 
Pleistocene. To date, I have found no solid evidence for human occupation north of the 
Arctic Circle before 15,000 yr B.P. This is far shorter than suggestions of 30,000 or more 
years of New World occupation proposed by some archaeologists and geneticists. One 
of the leading pioneers of Bering land bridge research is David M. Hopkins. Once an 
emphatic proponent of a pre-Clovis Beringian crossing, Hopkins (1996) now feels that 
the Alaskan and Siberian evidence for pre-Clovis is weak or non-existent. Keeping in 
mind this authoritative shift in opinion, let us turn to the question of migration number. 



Ho2P Many Migrations Were There 


Estimates of the number of migrations range from one (Laughlin & Harper 1988), 
to hundreds (Voegelin 1958). Most workers in archaeology, linguistics, physical anthro- 
pology, and more recentiy, genetics, favor a few migrations rather than many. There is 
not a great deal of difference in one, two, three, or four migrations when one recalls that 
defining the term migration is far from operationally precise, and when one recalls that 
the archaeological, linguistic, physical anthropological, and genetic variation in the New 
World is far less than that of East Asia, despite the New World being a much larger geo- 
graphic area. Much of the archaeological and paleoenvironmental evidence on this topic 
is reviewed by several scholars in West (1996), and paleobiological evidence is discussed 
in Steele and Powell (1992; this volume). Genetic evidence is covered in Szathmary's 
(1993) synthesis and Zegura's (1999) insightful book review. However, I would like to go 
over in some depth the dental evidence because that is what I know best, and it has the 
useful qualit}^ of being able to directly characterize living, recent, and subfossil popula- 

K. Hanihara (1968) defined the "Mongoloid dental complex" (MDC) using a few but 
powerful dental traits, including incisor shoveling, molar cusp numbers, and so forth 
(Figures 1—3). In the course of my dental anthropological studies of peoples of the 
Pacific Basin and adjoining areas, I recognized that the east Asian MDC could be divid- 
ed into southeastern and northeastern groups. These I called Sundadonty for Southeast 
Asia and Sinodonty for Northeast Asia and all the Americas. Dental morphological com- 
parisons between Eurasian, Pacific, African, and New World populations confirm the 
long-held view that the initial colonists to reach the New World came from northeast 
Siberia (Figure 4). All migration waves, whatever the number and migration t}^e, were 
by Sinodont peoples (In Figure 4 these are North-South America, Arctic, NE Asia, and 
Neolithic SE Asia). These people may have not been craniologically "Mongoloid" in the 
typological sense, but they were Sinodonts in the sense of intra- and intergroup odon- 
tological variation. In my view, there is no dental evidence in support of any other route 
or ancestral population. All my analyses of prehistoric and living Native Americans and 
comparative samples (> 25,000 individuals) indicate they possess the Sinodont dental pat- 
tern, which is found only in Northeast Asia and the entire New World. On the other 
hand, Powell and Rose (1999) propose that the controversial Kennewick Man, Initially 
claimed to be a Caucasoid, had a Sundadont dentition. However, these workers have not 
yet indicated which dental traits they used to reach this conclusion, nor have they indi- 
cated how they handled the problem of dental wear. With advanced wear, Sindodonty 
can be mistaken for Sundadonty. This can be appreciated in Figure 5, which shows fre- 



quencies for several rrairs in Sinodonts and Sundadonts. With wear, traits that vary in 
degree of expression, such as marked incisor shoveling and frequent molar cusp num- 
bers in Sinodonts can be misidentified for less marked expressions more common in the 
Sundadont pattern (Burnett et al. 1998). Even traits such as root number can be confused 
in older individuals when cement accretion has been excessive. C^hatters (2000) has also 
called Kennewick a Sundadont, but his assessment can be evaluated since Chatters list- 
ed the six dental traits he used to reach this conclusion. Only three of his traits are part 
of the eight trait suite I originally proposed for differendating between Sinodont}- and 
Sundadonty (Turner 1990a). On this small basis, I personally would not have drawn any 
conclusion about dental affinity, even though the traits Chatters used for affinity' assess- 
ment are not usually affected by wear. 

Claims made for European, Southeast or South Asian, Central Asian, Oceanian, early 
Japan, or African origin for Native Americans can be rejected outright on the basis of all 
presentiy known dental descriptions from these areas that are closely calibrated through 
the use of the Arizona State University Dental Anthropology System (ASU DAS) 
(Turner et al. 1991). A few examples of highly calibrated regional dental anthropology 
studies include: Europe (Zubov & Khaldeeva 1979, Russians; Lipschultz 1996, 
Mediterranean; Adler 1999, Finns); Asia (Liu 1998, China; Haeussler 1985, 1996, 1999, 
Siberia and central Asia; Haeussler & Turner 1992, central Asia and New World); (T. 
Hanihara 1991, East Asia and Pacific); South Asia (Hawkey 1998, India & Sri Lanka); 
Oceania (Harris 1977, Melanesia; Weets 1996, Melanesia); Africa (Irish 1993, Africa; Irish 
& Turner 1990, north Africa; Turner & Markowitz 1990, north and west Africa); New 
World (Morris 1965, Soutiiwest U.S.; Richert 1999, east Canada; G. Scott 1973, 
Euroamericans and Southwest Indians; Larson 1978, Southwest U.S.; Nichol 1990, Pima; 
Crespo 1994, Puerto Rico; Lincoln-Babb 1999, Yaqui, Mexico); Worldwide (Turner 
1990b, 1991, 1992a, 1992b, 1992c, 1992d, 1994a, 1994b, 1995, 1998; Scott & Turner 
1997; Turner & Hawkey 1991, 1998; BaHey-Schmidt 1995; E. Scott 1998). In addition, 
strong concordances have been made beKveen the manv crown morphology studies of 
living groups made by Russian dental anthropologists using the Zubov system through- 
out the former U.S.S.R. and expectations based on ASU DAS. 

Inasmuch as most of my dental anthropological and bioarchaeological studies on the 
peopling of the Americas and the Pacific Rim and Basin are based largely and directiy on 
many samples of dental crown and root morphology of past human populations, aided 
by the aforementioned students and coworkers, I will first discuss mv past and new views 
on the colonization based on teeth. Figures 1—3 illustrate some of the crown and root 
traits that ASU DAS uses in a highly standardized observation schedule. Mv Asian, 
Pacific and New World studies have identified two major dental patterns as previously 



mentioned. Sundadonty is a generalized, retained, and simplified dental pattern that 
occurs throughout Southeast Asia, Polynesia, Micronesia, and in the Jomon-Ainu people 
of Japan (Figure 5). Sinodonty occurs in all pre-Cossack Northeast Asians (Chinese, 
Mongols, Buriats, Siberian Eskimos, the Yayoi-modern Japanese peoples, and others), 
and all Native Americans. There are slight regional differences in both of these two 
broad catagories, and of course, there are minor temporal differences due to local 
microevolutionary and biohistoric events. Something like Sundadont)', which I call 
Proto-Sundadont}/^, is presumed to be the ancestral condition for the Australian and pos- 
sibly the similar Melanesian dental pattern. While neither of the two Asian dental pat- 
terns looks like that of Europeans, Sundadonty is more Uke the European dental pattern 
than is Sinodonty. In fact, Sundadont}' is the most generalized dental pattern in the 
world, showing none of the extremes in trait frequencies as elsewhere (Africa, very high 
frequency of lower molar cusp 7; Europe, almost no incisor shoveling or 3-rooted lower 
first molar; Northeast Asia and the New World, very high frequency of shoveling and 
lower molar cusp 5, etc.). Sinodonty is at least 8,000 years old on the basis of early 
Holocene teeth belonging to a skuU called Shilka found east of Lake Baikal near the cit}^ 
of Chita, teeth of early New World Indians of at least 9,000 years ago, and the three cra- 
nia from Upper Cave, which may be as old as 30,000 yr B.P. Sundadonty can be postu- 
lated as being as old as 20,000 yr B.P. based on the Nlinatogawa crania from Okinawa, 
40,000 years old based on the Niah Cave maxilla found in the Philippines, and perhaps 
50,000 years old based on recent Australian Aborigine teeth which, when averaged for 
the continent as a whole, probably approximate the hypothetical Proto-Sundadont con- 
dition (Turner 1992c, 1992d). Hence, there is no dental evidence that Paleo-Indians or 
any subsequent migrations originated from any dental gene pool other than Northeast 
Asia, which also includes the far eastern region that Russians refer to as Primoria ("near 
the ocean") and is not considered by them as part of Siberia (Nazdratenko 1998). 
However, Primoria might have witnessed an early occupation by Sundadonts as Jomon 
or Jomon-like artifacts have been found (Popov et al. 1997) there as well as in Sakhalin 
(VasiUevsky 1981). 

Sinodont variation in the New World, both past and present, seems to group into 
three divisions, with some widely separated North and South American samples show- 
ing random convergence — useful evidence for thinking that genetic drift is an impor- 
tant factor in dental microevolution. The three New World dental divisions correspond 
rather well with the three linguistic families defined by Greenberg (1987, 1990, 1996). 
These he calls Macroindian (Amerind), Na-Dene, and Aleut-Eskimo, as does Ruhlen 
(1994). Although Greenberg has some detractors, for example, Campbell (1986) and to 
a lesser degree, Nichols (1998), he has parried with compelling reason that these three 



families are genedcally real and that they represent three migrations or waves of migrants 
whose language families had diverged in the Old World prior to their crossing the Bering 
Land Bridge (Greenberg 1990, 1996). I feel there is as much combined evidence in var- 
ious archaeological, osteological, dental, genetic, and natural history analyses to argue in 
favor of the same interpretation, however, some analyses suggest possible admixture 
event(s) responsible for the formation of the Na-Dene ("Greater Northwest Coast") 
dental group (Turner 1985; Scott &c Turner 1997; Scott 1994). This uncertaint)' is a con- 
tinuing area of investigation. Regardless of whether two or three migrations are evenm- 
ally agreed upon, the colonization of the New World was more likely by a few number 
of migrants and migration episodes rather than by many migrants and many migration 
events. I doubt that there was only one migration event as Laughlin first proposed in a 
1985 Arizona State University seminar, simply because there is too much linguistic, 
archaeological, osteological and dental variation in Alaska alone, that in some instances 
matches that found regionally in Siberia. This view of only a few number of migration 
events has support in archaeology (West 1996), DNA and blood group smdies (CavaUi- 
Sforza etal 1994), dentition (Turner 1985 & elsewhere), and at least partly in historic lin- 
guistics. As for parsimony and concordance, the most, and best evidence and analyses 
point to Pleistocene Northeast Asians as having been the ancestral stock of all past and 
present Native Americans. Advocates of an Australian, Southeast Asian, Jomon- Ainu, or 
European origin have never included more than one line of evidence, or have failed to 
consider more parsimonious explanations to explain the nature of their evidence or 

My examinations of the teeth of Late Pleistocene Cro-Magnons of Europe and 
western Russia (1992a, 1998), similar studies by A. Haeussler (Turner 1996, & elsewhere), 
and the vast numbers of studies by Zubov and PChaldeeva (1979, & elsewhere) and their 
Russian co-workers, show without much question the continuit)' of these Cro-Magnons 
with later Mesolithic peoples and modern Europeans, be they English, French, Danish, 
Dutch, Russian, or whatever polity. In no case do these Upper Paleolithic Europeans 
have dental morphology like ancient or modern Northeast Asians or Native Americans. 
Moreover, many reports show that several European dental trait frequencies are more 
like those of Africans than like Sinodonts or Sundadonts (for summary, see Scott & 
Turner 1997). 

Did any Late Pleistocene group other than Sinodonts reach the New World? Cranial 
analyses of some South American crania have suggested that there might have been an 
early migration of "Australoids" (summarized in Lahr 1995; Neves etal 1999), or at least 
a group of early migrants whose crania differed in robustoess from those of more recent 
South American Indians. While these analyses are based on very few crania, and they are 



not unambiguous about the American and non-Asian similarities, they do have a few 
champions who point out that if Australia had been reached from Southeast Asia with 
the use of watercraft 50,000 years ago, the New World might have been similarly 
reached, namely by very early boating people. However, the early stone tool technology 
of Australia and Southeast Asia (horse hoof cores and scrapers, sumatraUths, etc.) have 
not been found with any of the South American "Australoids," as would be expected if 
there had been a Pacific transit 20,000 to 50,000 years ago from Sundaland or Sahulland 
(Green 1992). Moreover, if this "Australoid" trip was by watercraft from the more 
northerly end of the Sundadont population distribution, namely the early Jomonese of 
Japan, then we would expect their distinctive stone tool types to show up instead. As far 
as I know, such is not the case. While I have not examined every dentition from crania 
claimed to be pre-Clovis or non-Clovis in South America, all that I have examined, 
including Archaic collections from Ecuador, Peru, Chile, and Brazil, exhibit only 
Sinodonty (Turner 1992a). 

Based on skull shape, suggestions have been made that early Holocene (9,000 yr B.P., 
Tuohy and Dansie 1997) crania from Nevada — Spirit Cave and Wizards Beach — are 
less like Native Americans and more like some Old World populations, possibly 
Caucasoids or Ainu. While the teeth of these crania are very worn and not much infor- 
mation can be gleaned from them, nevertheless, that which can be safely scored reveal 
the teeth of these skulls to be very similar to those of recent Native Americans and 
unUke Ainu or Europeans (Figure 6). 

The most controversial proposal for the peopling of the New World comes from 
Dennis Stanford (1999 and this symposium), who along with stone tool-making expert, 
Bruce Bradley, envisions a Late Pleistocene "migration" reaching the New World from 
Spain by traveling along the southern edge of the north Atlantic pack ice. Their north 
Atlantic pack ice route is grounded in their belief that a "genetic" relationship exists 
between American Clovis and Iberian Solutrean points, as well as the apparent absence 
of this type of point manufacturing in Asia during Late Pleistocene times. AU known 
European Late Pleistocene and Holocene dental evidence is unsupportive of this idea, 
although, admittedly, I know of no dental studies on whatever Solutrean skeletal remains 
may have been recovered in Spain. Nevertheless, published descriptions and personally- 
studied dental morphology of French Cro-Magnons, and the excellent descriptions of 
Late Pleistocene teeth from western North Africa by Irish (1993) are very much like 
modern Europeans and very dissimilar to Native Americans, ancient and modern. Since 
Spain lies between France and western North Africa, it is fair to assume that the dental 
morphology of the Upper Paleolithic Iberians was like that of the adjacent northern and 
southern populations. 



Tccrh nor (>nl\- proxidc a resource tor identifying chulisric relarionships, rhe\' alsf) pro- 
vide information on environment and tooth-use behaviors. One of the most interesting 
environmental effects on teeth of humans who lived in the Arctic is the great amount of 
dental damage due to chipping, breakage, and ante mortem loss, which is significantly 
greater than in sub- Arctic dentitions. Much of this traumatic damage is thought to have 
been due to near-starvation events where bones were chewed for residual nutrients by 
children and adults alike, as well as the use of teeth as tools, such as vice gripes (Turner 
& Cadien 1969). Arctic dwellers rarely had dental caries, another environmental effect 
due to the near absence of cariogenic carbohydrate foodstuffs in their diets. These peo- 
ple ate almost exclusively products derived from mammals, fish, and birds. 



Figure 6. Dendrogram of representative samples of New World, Asian Sinodonts and 
Sundadonts, Polynesians, and Europeans. The target population consists of the early Holocene Spirit 
Cave and Wizard Beach individuals from Nevada. As this analysis shows the teeth of these early 
Holocene people were most similar to those of recent Indians of Nevada, show no similarit}' with ^•\inu 
or Jomon teeth, and are most dissimilar to the European dental pattern. The analysis shows that all 
American Indians cluster together (the top major branch), indicative of but one common ancestral pop- 
ulation for all North and South American Indians (Computer file reference: Nevada 20 groups— 29 traits 

What is the Genetic Affinity between Native vAmericansand 
Peoples of the Old World Incl/iding the Pacific Basin? 

Most genetic evidence, serological and DNA, points to strong cladistic connections 
between Native Americans and Northeast Asians (Szathmarv 1979, 1985; niany other 
workers). But there are one or two genes that hint svnchronically at some ancient or 
more recent minor relationship with eastern Europe or western Siberia. Neither the Late 



Pleistocene archaeology of southern Siberia nor a new mitochondrial DNA study 
(Voevoda et al. 2000) rule out such a possibility. The monumental study by Cavalli-Sforza 
et al (1994) leaves very little room for doubt that all Native Americans are closely relat- 
ed to Northeast Siberians, making that region the best candidate for the proximate ances- 
tral homeland. Moreover, these workers support the three migration hypothesis based on 
linguistic and dental morphological inferences, as did Powell (1993). 

How Many Migrants Crossed Beringia? 

This is a question that one seldom finds workers unfamiliar with evolutionary 
processes much concerned about. However, population size is a key factor in biological 
and culmral microevolution. The smaller the population, the greater is the opportunit}^ 
for rapid genetic change through time due to random chance, even in the context of 
severe selective pressures as expected under Arctic conditions. It is not unreasonable to 
suppose that the same evolutionary forces that affect biological change operate also on 
cultural variation, even if proof of such has yet to be well established. Hence, in the 
rapid expansion model of P. S. Martin (Martin & Klein 1984) there could have been 
marked changes in the bioculmral characteristics of Siberians as they progressed north 
and eastward to Alaska, with subsequent rapid dispersal throughout the rest of the New 
World. There are numerous examples of small founding populations failing to gain a 
foothold in their newly discovered environments, due presumably to the negative effects 
of small population size. One obvious demographic effect of small population size is on 
the size of the breeding population, usually looked upon as about one quarter of the 
total number of females. In a population of 100, on the average there will be about 50 
females, of which there might be only 12 or so women of breeding age. If only a few of 
these had children who Uved to reproductive age themselves, then the chances of there 
being a representative sample of genes (and to some degree, culmre) transmitted to 
fumre generations is less than 50 percent. If success breeds success, as it often does, then 
a rapid train of culmral and genetic change could be expected. One expectation would 
be a loss of variation. 

One of the fascinating characteristics of Clovis culture is its marked uniformit}^ in 
lithic projectile point st}4e throughout North America. A similar low amount of vari- 
ability characterizes most of the genetic and dental featares of prehistoric and living 
North and South American Indians. If Greenberg (1987) is correct in his New World 
language classification, then the limited amount of linguistic variation at the phylum level 
is also a feamre of New World populations. The parsimonious interpretation of this low 
amount of biocultural variation relative to that of East Asia with its smaller land mass, 



nuich longer period of luiman occLijiation, and c(.|uall\ dixersc cnxifonmcnial zones, is 
thai rhe size of the New World founding population was absolutely small. As such it 
could not possibly have carried a representative sample of the genetic polymorphisms 
and diverse cultural features of Northeast Asia, let alone all of East Asia. If small found- 
ing population size was the case, and I sense that most archaeologists, physical anthro- 
pologists, and geneticists interested in the peopling of the New World would concur, 
then either rapid biocultural changes could be expected in the Beringian transit or there 
was imperfect representation (or both), which resulted in an initial Alaskan population 
that might not be readily traceable to one or another Siberian prehistoric population and 
culture except in a general sense. Given that there was rapid and random change, aggre- 
gations or clusters of neighboring New World groups into culturally and/or biologically 
distinct units would be a sign of mutually common ancestry as much as fission subse- 
quent to arrival in their new territories. Although the amount of archaeological field 
work in Siberia near to and north of the Arctic Circle is much less than in districts fur- 
ther south, nevertheless the limited number of far northern 14 C -dated Late Pleistocene 
sites is one indicator of there likely having been but a very small population in western 
Beringia. If we assume that for every archaeological site known for a given time period 
and geographic district, there are 50 others that have not been discovered or destroyed, 
and assuming that any site represents a band of 20 people, then western Beringia had 
less than 1,000 people from which were drawn the initial and later founders of Alaska. 
This is, of course, only guess work, but estimates of population size at historic contact 
in Siberia and Alaska are supportive, even though environmental changes have occurred 
since the end of the Pleistocene. 

What was the Cultural or Technological Tevel of the Migrants? 

It is easy to identify some of the technology, social behavior, skills, and economic 
strategies necessary for comtemporary high Arctic life by reviewing ethnographic 
accounts of various polar groups. For the western Eurasian Arctic, major groups living 
above the Arctic Circle include the Komi, Nentsey, Saams (Lapps), Entsey, Naganasans, 
Evenks, Dolgans, and Kets. For the eastern Eurasian Arctic there are Yakuts, Evens, 
Yukagirs, Chuckchi, and Koryaks (Levin & Potapov 1964; Malet 1998). For Alaska, 
Canada, and Greenland, there are various bands of Eskimos and Athabaskans (Damas 
1984). AH these groups possess similar skills, values, and equipment needed for li\dng in 
a very cold and chancy environment. For each of the various, readily preservable, culture 
elements made of stone, bone, or earth, such as weaponry, fishing gear, various tools, 
pottery, and housing, the archaeological evidence for trait antiquit)' and evolution is often 









■ 2 



K 3 



spt p. 



1 d 

1 — T 


■ S 



1 ^ 




&i ^— - 




a "^ 

■^ — ' 





Ji^ C3 


•I '^ 




1 f^ 







W ^" 


^ -^ 



ti rt 



feJ*^ "^ 





« Q 












. »\ 































T 1 































, — ,^ 




• S 




• r 












1 ^--^ 


1 " 





1 ^ 





1 ^ 





J s 





1 o 




1 ■ — ' 



1 or. 






































abuiuhiiii on hoth local ami regional scales. Oi more perishable anti behavioral items 
such as clothing and social organization, prehistoric evidence is much less and usually 
indirect and more inferendal. However, I feel it is generally fair to say that high Arctic 
I'pper Paleolithic life was much the same as that in the Eurasian sub-Arctic, but with a 
few absolute needs for survival. Two stand out in my mind, as the title of this article sug- 

Many of the same vertebrate species found m the Late Pleistocene steppe and for- 
est-steppe of Siberia have been found in deposits of similar age in Alaska (Kurt & 
Anderson 1980; Guthrie 1990; n.d. Ovodov, on-going personal communications 
1998—2000). These include mammoth, horse, bison, bovids, lions, and many others, but 
interestingly, not the widespread and seemingly numerous Siberian hyenas (Kurt &c 
Anderson 1980). Human terrestrial hunters and gatherers of southern Siberia would 
have had to make no major technological changes as far as the t}^es of animals hunted 
farther north was concerned. However, intensit;^ and duration of winter cold, mobiHt}', 
patchiness of resources, and possibly other limiting factors such as large predators, had 
to be accomodated in order to come near to and cross the Arctic Circle. The cold would 
have been dealt with by the innovation of windtight fur clotliing assembled \\ith small 
tailorino- needles, one of which Russian associates Ovodov, IMart\'novich, Pavlova, and I 
recovered in a 20,000-year-old stratum during our archaeological testing of a limestone 
grotto called Dvaglaska Cave in Khakatia Autonomous Republic, July, 2000. That stra- 
tum and deeper ones contained a rich representation of extinct Late Pleistocene fauna, 
including hyenas. Here, the environment may have been steppe or forest-steppe. 

Further west, in the Altai Mountains, Ovodov's excavations in a deep limestone 
mountain-top cave called Razboinich'ya turned up a complete dog skull that 14 C -dated 
to about 14,000 yr B.P. This relic of an early domesticated wolf must have been carried 
into Razboinich'ya by hyenas that were its principal Pleistocene occupants. Tliis cave, 
along with those lower down the moutain valley utilized by Mousterians, would have 
been surrounded by forest, as today. Along Siberia's Late Pleistocene east coast, south of 
probably most of the pack ice that would have covered the Sea of Okhotsk further 
north, the hunting of abundant marine mammals (Asiatic fur seal, walrus, sea cow, dol- 
phin, whale, etc.) might have been both a beach and boating endeavor as at the Neolithic 
coastal site of Boisman 11, south of Vladivostok (Popov efa/. 1997). Hence, Siberian nat- 
ural history suggests three major food and shelter resources — marine coast, forest steppe 
and steppe, and mountain valley forest systems, which in turn could have served as three 
regional homelands. I proposed this multiple Siberian and Primorian ecological home- 
land idea in 1979 (Turner 1983), and it has been further elaborated upon by Goebel 



The key Late Pleistocene inventions that are needed to endure and traverse the 
extremely hostile Late Pleistocene winter environment above the Arctic Circle are tai- 
lored windtight fur clothing, and dogs to aid in transporting the bulky equipment need- 
ed for high Arctic survival. The minimal equipment inventory would have had to have 
included for each family in a migratory hunting party: tent poles, ropes, pegs, and bulky 
skin covers; extra clothing and wraps for infants; cooking and other household items; 
hunting and fishing equipment (traps, spears, hooks, lines, clubs, and so forth); various 
filled or empty containers including skin or string bags; heav}^ fur bedding; spare materi- 
als for repairs of weapons, utensils and clothing; emergency food and fuel; and critical- 
ly, carefully protected glowing "slow match," or dust-dry tinder for use with a "fire drill" 
to quickly start cooking and warming fires in hearths or possibly stone oil lamps at each 
winter hunting encampment. Being tireless at nightfall on a midwinter hunting trek 
would have been a serious if not lethal situation. The amount of time spent in any given 
encampment would have depended mainly on the availability of local food and fuel 
resources, either autarchic or under the control of other groups. Similarly, as groups 
attempted to Uve further and further north, the more influential must have been local 
resource control as well as climate and vast distance on population structure, dispersal 
and mating behavior. Mates were likely drawn from the small local populations, which 
would have contributed to the founder effect that I infer was the chief evolutionary 
process that produced the previously mentioned three major New World dental groups. 
Life in the far north was very difficult as seen even today in low population density, short 
life span, high levels of osteoporosis in middle-aged women, and spinal defects (Laughlin 

These two key inventions for colonizing the high Arctic of eastern Eurasia, the small 
tailoring needle for precision sewing, and the domestication of the dog from its north- 
east Asian wolf ancestor, Canis lupus, are according to archaeological discoveries, very late 
developments in Pleistocene prehistory. Tailoring needles are known for several sites in 
Late Pleistocene Eurasia (Soffer 1985). In eastern Siberia there is evidence for wolf 
domestication (Ovodov 1998). No examples of Late Pleistocene dogs have been found 
in Europe (Soffer 1985). While watercraft such as dugouts, birchbark canoes, or hide- 
covered boats, known in one form or another today for most Siberian groups, may well 
have been known and used for summer travel along the northward flowing rivers like the 
Ob, Yenisei, and Lena, and along the eastward flowing Amur, there is only suggestive evi- 
dence of their Late Pleistocene use based on the choice of riverside encampments such 
as Afontova Gora and Kurtak on the Yenisei, Mal'ta on an ancient lake shore, and oth- 
ers. The physical remains of at least one dog dating back to some 14,000 yr B.P (Ovodov 
1998) is about the same age as the dog "burial" discovered by Dikov (1997) at the 



Kamchatkan site of Ushki. Dogs can aid in transporting loads of equipment and sup- 
plies with the use of carrying bags (like saddle bags), the two pole travois, and sleds of 
various size and complexit); Dogs would also have been valuable for locadng, stalking, 
and attacking prey animals; guarding against other large Late Pleistocene carnivores; 
serving as a source of companionship; and in an emergency, food. 

What 'Languages Might the Founders Have Spoken? 

The prehistoric New World populations have been in place long enough to have lost 
all one-to-one language relationships between those spoken in the New" and Old Worlds, 
except possibly for Alaskan Eskimo-Aleut and Siberian Chuckchi (Ruhlen 1975). 
However, language family or phyla relationships have been proposed suggesting con- 
nections between all New World language families and those found in Siberia (Greenberg 
1987; Ruhlen 1994), although much controversy revolves around this proposal 
(Greenberg 1990). Recentiy, Ruhlen (1998) suggested a Linguistic link between Alaskan 
Na-Dene and Ket, whose few remaining settlements along the Yenisei River basin might 
also have been the homeland for Na-Dene. No Linguistic connection has ever been pro- 
posed between Ainu and any New World language, despite the relatively close proximit}' 
of Ainu settlements on Hokkaido and Sakhalin with Alaska. Just as biological and cul- 
tural features might have changed rapidly in the small population context of the Siberian 
Arctic, surely so would have languages. This is exempUtied by the few and seemingly 
genetically unrelated Paleosiberian languages of Northeast Siberia (Ruhlen 1975; Crystal 


In sum, the dental evidence based on highly concordant observers cited previously 
shows that the dental morphology of Native Americans, past and present, is most like 
that of Late Pleistocene and recent Northeast Asians; much less like that of all Southeast 
Asians, Polynesians, and Micronesians; and very dissimilar to that of Australians, 
Melanesians, South Asians (India), Europeans and Africans. While the story of New 
World and Pacific Basin prehistory and human dispersal therein is far from complete, 
there is a useful outUne in hand. Any scenario for the peopling of the New Wbrld that 
does not also consider relevant biocultural and natural liistory events around the world, 
as has been the objective of our dental anthropological surveys and affinit}- assessments, 
fails the evidentiary test of holism for acceptabilit}'. There is no better example of the 
value of worldwide synthesis than the tour deforce by geneticists Cavalli-Sforza and col- 



leagues (1994). These workers find in gene distributions and other biocultural consider- 
ations the same sets of large-scale Old and New World intergroup relationships that are 
revealed by dental morphology. They say: "The genetic patterns in the Americas fully 
confirm the three waves of migration suggested by dental and linguistic evidence ..." 
CavalU-Sforza et al. (1994:340). 

Hence, taken altogether, I find the combined evidence compelling for inferring an 
interior Beringian route for the Clovis or epi-Clovis Paleo-Indians. Ancestral Aleut- 
Eskimo may have boated and trekked along the southern coast of the land bridge at 
about the same time as the epi-Clovis folk were exploiting the resources of, and drifting 
eastward across, the great Beringian plain. Language, ecology, animus, distance, or some 
combination may have kept these two groups from interbreeding while in transit across 
the land bridge. Aleut-Eskimos are viewed as having had a Primorian homeland where 
they developed their wet-cold maritime adaptation in the Amur-Hokkaido sea mammal 
district, whereas interior dry-cold Northeast Siberia was more likely the homeland for 
both ancestral Paleo-Indian and later Paleo-Arctic (Diuktai-Na-Dene) peoples. Thus, I 
continue to envision three more-or-less distinct migration waves made up of small num- 
bers of migrants, a scenario that Powell (1 993) has agreed with, using a different method 
for analyzing my published dental data. Their spoken tongues were ancestral to 
Greenberg's Macro-Indian, Aleut-Eskimo, and Na-Dene phyla. Each of these three pos- 
sessed an advanced level of upper PaleoUthic culmral developments, including dog 
domestication. Their biological and cultural diversification had already evolved before 
they reached Beringia and their different routes, geographic isolation, small population 
sizes, and arrival times caused and maintained their differentiation with the possible 
exception of some manner of post-Pleistocene biocultural mixing in the Gulf of Alaska 
region (see especially G Scott [1994] on this problem). I find nothing in the Siberian, 
Primorian, or Alaskan bioarchaeological record to suggest entry much before Clovis or 
epi-Clovis, nor do I see evidence of Sundadont)^ in the New World, although divergent 
and convergent dental microevolution has occurred within the general Sinodont pattern. 
Claims for Sundadont}^ (Lahr 1995; Powell & Rose 1999), I suggest, are due to wear- 
influenced observation error. Finally, the message of the long occupation of southern 
Siberia and a much shorter record above the Arctic Circle probably signals that the New 
World was unreachable until climate-related technical advances had been made. I suggest 
that two critical advances were the domestication of the wolf and the innovation of the 
fine tailoring needle. Beginning in the early 1960s, when I first began to excavate in the 
Aleutian Islands with W. S. Laughlin, A. P. McCartney, J. A. Turner, G Bair, E. & K. 
Harris, and others, I felt that Late Pleistocene watercraft played some role in the coastal 
peopling of the New World, at least as far as the Alaska Peninsula. However, near-freez- 



ing Arctic water temperatures and infamous Arctic stormy seas are even more hazardous 
to human survival than winter blizzard whiteouts on land. However, periods of starva- 
tion on land last longer than drowning in icy seas, so the chances of crisis survival on 
land are better. 

Finally, can a systems-oriented characterizadon be made of the cultural and biologi- 
cal complex that enabled native Siberians to reach and cross Beringia? At present, one 
can do litde more than recite the biocultural characterisdcs of the far northern peoples 
of Europe, Siberia, Alaska, Canada and Greenland at the time of historic contact. 
Despite the romantic notion of the noble New World savage, I see very few signs of 
environmental conservation, intergroup co-operation, sexual equality, humanitarianism, 
or other idealistic and noble qualities. On the other hand, it seems patentiy evident that 
the human qualities needed for the conquest of the Arctic must have included high intel- 
ligence, toughness, endurance, intragroup co-operation and sharing, creativity', and well- 
tested ways of dealing with the natural environment. Not only were there long periods 
of intense cold stress, and repeated starvation episodes, there were also long periods of 
suffering with mentally-depressing, winter darkness. Teeth not only help inform us about 
who these Arctic pioneers were, and where they came from, but the considerable amount 
of dental chipping, breakage, and loss also exhibits how difficult life was for humans in 
the world's most harsh environment. Even the bone-crushing teeth of Pleistocene 
Siberian hyenas rarely show this kind of tooth damage and loss. tract 

The title of this essay symbolizes mj holistic orientation and current under- 
standing of Siberian and New World anthropology. In this paper I examine 
information from dental morphologj, osteology, archaeology, linguistics, genetics, 
natural history, and oral tradition to assess eight key questions about the origin 
of Native Americans. These questions are: (1 ) Where was the regional ancestral 
homeland(s)?, (2) ivhat route (s) were folloiped?, (3) ivhen was the time(s) of 
arrival?, (4) how many tnigrations ivere there?, (5) what is the genetic affinity 
between Native Americans and peoples of the Old World including the Pacific 
Basin?, (6) how many migrants were there?, (7) what was the cultural or techno- 
logical level of the migrants?, and (8) what languages might they have spoken? It 
is my explicit theoretical view that the best explanation for the peopling of the 
Americas requires integration of all lines of both diachronic and synchronic evi- 
dence, not just one or two as is the most common approach today. As I interpret 
these multiple lines of evidence it would appear that the most powerful single 
hypothesis, among several possible based on one or another line of evidence, envi- 



sions: (1 ) A Siberian homeland, (2) a Bering land bridge route, (3) a relatively 
late time of arrival, (4) a very few number of migrations, (3) a genetic affinity 
primarily with Northeast Siberians, (6) very few migrants, (7) a sophisticated 
Arctic-adapted culture, and (8) languages related to those of eastern Siberia. 
Before humans could live near to, or above, the Arctic Circle (66° 33 '), and 
eventually reach Alaska, they had to have had a biocultural system that supplied 
solutions to several harsh environmental problems. In my view, the next major 
task for students of Native American origins is developing hypotheses about that 


I was honored to be asked by Nina G. Jablonski to participate in this exciting Wattis 
Symposium. She, Mrs. Wattis, Nancy Gee, and other members of the California 
Academy of Sciences are deeply thanked for their support and arrangements. It was a 
pleasure to see colleagues of long standing and to make the acquaintance of other par- 
ticipants in the symposium.. The analysis of the Nevada crania was made possible by 
arrangements and funding from Don D Fowler, University of Nevada, Reno. Computer 
assistance for the Nevada study was provided by Rhea Jacanin. The Siberian taphonomy 
studies were made possible by financial aid from the Exploration and Research 
Committee of the National Geographic Society (grant #6464-99), and the Wenner-Gren 
Foundation for Anthropological Research (grant #6488). Siberian field and laboratory 
studies in 1998 through 2000 were permitted by Academician Anatoli Derevyanko, 
Director, Institute of Archaeology and Ethnology, Siberian Branch, Russian Academy of 
Sciences, Novosibirsk; Nicolai I. Drosdov, President, Krasnoyarsk State Pedagogical 
University; and German I. Medvedev, Director, Laboratory of Archaeology and 
Paleoecology, Irkutsk State University. 

l^iterature Cited 

Adler, A. J. 1999. The Dentition of Contemporary Finns. M.A. thesis, Department of 

Anthropology, Arizona State University, Tempe, AZ. 
Bailey-Schmidt, S. E. 1995. Population Distribution of the Tuberculum Dentale Complex and 

Ano?nalies of the Maxillary Anterior Teeth. M.A. thesis, Department of Anthropology, 

Arizona State University, Tempe, AZ 
Bonnichsen, R., & D G. Steele, eds. 1994. Method and Theory for Investigating the Peopling of 

the Americas. Center for the Study of the First Americans, Oregon State University, 

CorvaUis, OR. 264 pp. 



Bonnichscii, 11., & K. Turnmirc, ctls. 1991. (Joi'is: Orrj^iiis and /Idnp/ti/ioiis. (x-ntcr for the 

Study of the I'irst Americans. Corvallis, OR. 344 pp. 
Boorstin, D. ). 1983. I'hc Discoverers. Random House, New York, NY. 745 pp. 
Brown, P. 1998. The first Mongoloids, another look at Upper Cave 101, IJujiang and 

IVIinatogawa \. Acta Antbropol. Sin. 17:255-275. 
Burnett, S. E., J. D. Irish, & M. R. Pong. 1998. How much is too much? Examining the 

effect of dental wear on studies of dental morphology. Am. j. Phjs. Antbropol., Suppl. 

Campbell, L. 1986. Comment. Onr. Antbropol. 27(5):488. 
Carlson, R. L., & L. D. Bona, eds. 1996. Harlj Human Occupation in Rritisb Columbia. 

Universit)' of British Columbia Press, Vancouver, Canada. 261 pp. 
Cavalli-Sforza, L. L., P. Menozzi, & A. Piazza. 1994. Tbe History and Geography of H/iman 

Genes. Princeton Universit}^ Press, Princeton, NJ. 518 pp. 
Chatters, J. C. 2000. The recovery and first analysis of an early Holocene human skele- 
ton from Kennewick, Washington. Am. Antiquity 65:291—316. 
Crespo. E. F. 1994. Dental Analysis of Human burials Recovered from Pun fa Candelero: A 

Prehistoric Site on the Southeast Coast of Puerto Rico. M.A. thesis, Department of 

Anthropology, Arizona State Universit)'^, Tempe, AZ. 
Crystal, D. 1987. Tbe Cambridge Hncjclopedia of language. Cambridge Universit\' Press, 

Cambridge, MA. 472 pp. 
Damas, D., ed. 1984. Handbook- of Noiih American Indians: Arctic. Smithsonian Institution, 

Washington, D.C. 829 pp. 
Derev'anko, A. P. 1990. Paleolithic of North Asia and tbe Problem of Ancient Migrations. 

Translated by I. Laricheva. Institute of History, Philology, and Philosophv, Academy of 

Sciences of the USSR, Siberian Division, Novosibirsk. 123 pp. 
Dikov, N. N. 1997. Asia at the juncture with America in Antiquity (The Stone Age of tbe Chukchi 

Peninsula). Translated by RL. Bland. National Park Service, Beringia Program, 

Anchorage, AK. 248 pp. 
Dillehay, T D. 1999. The Late Pleistocene cultures of South America. Erol Antbropol 

Dillehay, T D. ,& D. J. Meltzer, eds. 1991. Tbe First Americans: Search and Research. CRC 

Press, Boca Raton, PL. 310 pp. 
Dumond, D. E. 1987. The Eskimos and Aleuts. Revised edition. Thames and Hudson, New 

York, NY. 180 pp. 
Echo-Hawk, R. C. 2000. Ancient history in the New Wbrld: Integrating oral traditions 

and the archaeological record. Am. Antiquity 65:267—290. 



Fagan, B. M. 1990. The Journey from Eden: The Peopling of Our World. Thames and Hudson, 
New York, NY. 256 pp. 

Fladmark, K. R. 1978. The feasibility of the Northwest Coast as a migration route for 
early man. Pages 19-128 in A.L. Bryan, ed., Tiarly Man in America fro?7j a Circpim -Pacific 
Perspective. University of Alberta Department of Anthropology Occasional Paper 1, 
Edmonton, Canada. 

Gibbons, A. 1996. The peopling of the Americas. Science 274:31-33. 

Goebel, T 1999. Pleistocene human colonization and peopling of the Americas: An eco- 
logical approach. Evol AnthropoL 6:208—226. 

Green, R. C. 1992. Changes over time — recent advances in dating human colonisation 
of the Pacific Basin area. Pages 11-42 in D. G. Sutton, ed., Origin of the First Neip 
Zealanders. University of Auckland Press, Auckland, New Zealand. 

Greenberg, |. H. 1987. Language in the Americas Stanford University Press, Stanford, CA. 
438 pp. 

. 1990. The American Indian language controversy. The Kev. Archaeol 11:5—14. 

. 1996. Beringia and New World origins: I. The linguistic evidence. Pages 

525-536 in F. H. West, ed., American Beginnings: The Prehistory and Palaeoecology of 
Beringia. Universit)^ of Chicago Press, Chicago, IL. 

Grine, F. E. 1981. Occlusal morphology of the mandibular permanent molars of the 

South African Negro and the Kalahari San (Bushman). Ann. S. Afr. Mus. 86:157-215. 
Guthrie, R. D 1990. Vro^^en Fauna of the Mammoth Steppe: The Storj of Blue Babe. University 

of Chicago Press, Chicago, IL. 323pp. 
Haeussler, A. M. 1985. Dental Morphology of New World, Eastern Siberia, and Soviet Central 

Asia Populations M.A. thesis. Department of Anthropology, Arizona State University, 

Tempe, AZ. 
. 1996. Dental Anthropology of Russia, Ukraine, Caucasus, Central Asia: The Evaluation 

of Five Hypotheses for Paleo-lndian Origins. Ph.D. dissertation. Department of 

Anthropology, Arizona State University, Tempe, AZ. 

. 1999. Dental anthropology of the Neolithic Russian Far East: I. Eurasian 

Russia. Dental Anthropol 13:5—14. 

Haeussler, A. M., & C. G Turner II. 1992. The dentition of Soviet Central Asians and 
the quest for New World ancestors./ Hum. Ecol {Special Is sui) 2:273-297. 

Hall, D A. 2000. Charting a new era: Clovis and beyond draws over 1,400. Mammoth 
Trumpet 15:1—7. 

Hammer, M. F, S. L. Zegura, A. Bergen, J. C. Long, W Klitz, R. C. Griffiths, A. R. 
Templeton, L. P Osipova, O. L. Posukh, & T M. Karafet. 1999. New World Y chro- 
mosome found haplot)^es and the peopling of the Americas. Am. J. Phys. Anthropol, 
Suppl. 28:144. 



I laiiihara, K. I96S. i\l()iiu;()l()icl tlcntal complex in the permanent dentition. Pages 

298—300 /// Proceedings of the \ 11th Inteniational Congress of Anthropological and I 'J h no logical 

Sciences. Vol. 1 . Science Council of Japan, Tokyo. 395 pp. 
Hanihara, T. 1991. Dental and cranial evidence on the affinities of the Hast Asian and 

Pacific populations. Pages 119—137 /// K. Hanihara, ed., Japanese as a Member of the 

Asian and Pacific Populations: International Syii/posiui// 4. International Research Center 

for Japanese Studies, Kyoto. 
Harris, E. E 1977. Anthropologic and Genetic Aspects of the Dental Morhology of Solon/on 

Islanders, Melanesia. Ph.D. dissertation. Department of Anthropology, Arizona State 

University, Tempe, AZ. 
Hawkey, D. E. 1998. Out of Asia: Dental Hvidence for Microevolution and Affinities of Early 

Populations from India/ Sri hanka. Ph.D. dissertation. Department of Anthropology, 

Arizona State University, Tempe, AZ. 
Hoffecker, J. E, W. R. Powers, & T Goebel. 1993. The colonizadon of Beringia and the 

peopling of the New World. Science 259:46—53. 
Hopkins, D. M. 1996. Introduction: The Concept of Beringia. Pages xvii— xxi in E H. 

West, ed., American Beginnings: The Prehistory and Palaeoecology of Beringia. Universit)' of 

Chicago Press, Chicago, IL. 
Howells, W W 1995. Who's Who in Skulls: Ethnic Identification of Crania from Measurements. 

Papers of the Peabody Museum of Archaeology and Ethnology, Harvard Universit}-^, 

Vol. 82. Cambridge, MA. 108 pp. 
HrdUcka. A. 1920. Shovel-shaped teeth. Am. J. Phjs. Anthropol 3:429^65. 
. 1931. Catalogue of human crania in the United States National Museum col- 
lections: Pueblos, southern Utah Basket-Makers, Navaho. Proc. U.S. Nat. Mus. 

Irish, |. D. 1993. The Dentition of Africa. Ph.D. dissertation. Department of Anthropology, 

Arizona State University, Tempe, AZ. 
Irish, J. D, & C. G. Turner II. 1990. West African dental affinit}' of Late Pleistocene 

Nubians. II. Peopling of the Eurafrican-South Asian triangle. Homo 41:42—53. 
Ishida, H. 1993. Populational affinities of the Peruvian with Siberians and North 

Americans: A nonmetric cranial approach. Anthropol Sci. 101:47—63. 
Kaminga, )., & R. S. V. Wright. 1988 The Upper Cave at Zhoukoudian and the origins of 

the Mongoloids./. Hum. Evol 17:739-767. 
Kozintsev, A. G. 1988. Ethnic Cranioscopy. Racial T ^ariation of Cranial Sutures. Nauka, 

Leningrad (in Russian). 166 pp. 
Kurten, B., & E. Anderson. 1980. Pleistocene Mammals of North America. Columbia 

Universit}- Press, New York, NY. 442 pp. 



Kuzmin, Y. V. 1997. Chronology of Palaeolithic Siberia and the Russian Far East. Rev. 

Archaeol. 18:33-39. 
. 2000. Geoarchaeology of the Lower, Middle and Early Upper Paleolithic of 

Siberia: A review of current evidence. Rev. Archaeol. 21:32—40. 
Lahr, M. M. 1995. Patterns of modern human diversification: Implications for 

Amerindian origins. Yrhk. Phjs. Anthropol. 38:163—198. 
Larson, M. A. 1978. Dental Morphology of the Gran Quivira Indians. M.A. thesis. Department 

of Anthropology, Arizona State University, Tempe, AZ. 
LaughHn, W S. 1980. Aleuts: Survivors of the Bering Ljjnd Bridge. Holt, Rinehart and 

Winston, New York, NY. 151 pp. 
. 1981. Solutions to origins of Bering Sea peoples. Report presented at Working Field 

Conference (Anchorage, Kodiak, Old Harbor, Three Saints Bay, Unalaska, Umnak, 

Pribilovs), September 3—12. 
Laughlin, W S., & A. B. Harper. 1988. Peopling of the continents: Australia and America. 

Pages 14-40 in C. G. Mascie-Taylor & G. W. Lasker, eds., Biological Aspects of Human 

Migration. Cambridge University Press, Cambridge, MA. 
Laukin, S. A. 2000. The settlement of the nothern Asia by Paleolithic Man. Pages 79-80 

in N. I. Drozdov, ed., Paleogeografta Kamennogo Keka. Korreljashiya Prirodnk Sohtii 

Arkeologicheskik Kultyr Paleolita Severnoi A^ia i Sopredelnk Territori. Material 

Meshdunarodnoi Konferensi, Kurtak. Krasnoyarsk. 
Levin, M. G, & L. P. Potapov, eds. 1964. The Peoples of Siberia. Translation edited by S. 

Dunn. University of Chicago Press, Chicago, IL. 948 pp. 
Lincoln-Babb, L. 1999. An Analysis of the Morphological Dental Traits of the Yaqui Indians. 

M.A. thesis. Department of Anthropology, Arizona State University, Tempe, AZ 
Lipschultz, J. G. 1996. Who Were the Natufians? A Dental Assessment of Their Population 

Affinities. M.A. thesis, Department of Anthropology, Arizona State University, 

Tempe, AZ. 
Liu, W. 1998 The characteristics of deciduous teeth of Bronze Age Humans found in 

Changyang, Hubei Province. Acta Anthropol Sin. 17:177—190. 
Malet, C, ed. 1998. Expeditions en Siberie: Anthropologic, archeologie, ecologie, medecine. Boreales 

lA/11 whole volume. Suresnes, France. 336 pp. 
Martin, P. S., & R. G Klein, eds. 1984. Quaternary Extinctions: A Prehistoric Revolution. 

University of Arizona Press, Tucson, AZ. 892 pp. 
Meggers, B.J. 1971. The transpacific origin of Mesoamerican civilization: A preliminary 

review of the evidence and its theoretical implications. Am. Anthropol 11:\—T1. 
Meltzer, D. J. 1993. Pleistocene peopling of the Americas. Evol Anthropol 1:157-169. 
Mochanov, Y A. 1977. The Earliest Stages of Settlement by Man of Northeast Asia. Nauka, 

Novosibirsk (in Russian). 264 pp. 



Moorrees, C. F. A. 1957. The Aleut Denlition: A Correlative Study of Dental Characteristics in 

an Eskimoid People. 1 larvard University Press, (Cambridge, MA. 196 pp. 
Morris, D. H. 1965. 'I'he Anthropological Utility of Dental Morphology. Ph.D. dissertation. 

Department of Anthropology, Universit}' of Arizona, Tucson, AZ. 
Nazdratenko, Y. I., ed. 1998. y\tlas Primorskogo Krai. Vladivostok. 49 pp. (in Russian) 
Neves, W. A., J. E. Powell, & E. G. Ozolins. 1999. Extra-condnental morphological 

affinities of Lapa Vermelha IV, hominid 1: A multivariate analysis with progressive 

numbers of variables. Homo 50:263—282. 
Nichol, C. H. 1990. Dental Genetics and Biological ^relationships of the Pima Indians of Ari:(ona. 

Ph.D. dissertation, Department of Anthropology, Arizona State Urdversit}', Tempe, 

Nichols, ]. 1998. The origin and dispersal of languages: Linguistic evidence. Pages 

127—170 /// N. G. Jablonski & L. C. Aiello, eds.. The Origins and Diversification of 

Tanguage. Memoirs of the California Academy of Sciences No. 24, San Francisco, 

Ossenberg, N. S. 1994. Origins and affinities of the native peoples of northwestern 

North America: The evidence of cranial nonmetric traits. Pages 79—115 in R. 

Bonnichsen & D. G. Steele eds.. Method and Theory for Investigating the Peopling of the 

Americas. Center for the Study of the First Americans, Oregon State University, 

Corvallis, OR. 
Ovodov, N. D. 1998. The A.ncient Dogs of Siberia. ICAZ Symposium on Domestic Dogs, 

Aug. 23-28, Victoria, B.C., Canada. 
Pedersen, P. O. 1949. The East Greenland Eskimo dentition. Meddelelser om Gronland 

Popov, A. N., T A. Chikisheva, & E. G. Shpakova. 1997. Boisman Archaeological Culture in 

Southern Primoria {On the Basis of the Bois??jan II Multilaaaayered Site). Instimte of 

Archaeology and Ethnography, Siberian Branch Russian Academy of Science Press, 

Novosibirsk. 95 pp. (in Russian) 
Powell, J. F. 1993. Dental evidence for the peopling of the New World: Some metholo- 

logical considerations. Hum. Biol 65:799—819. 
Powell, J. E, & W A. Neves. 1999. Craniofacial morphologv of the first Americans: 

Pattern and process in the peopling of the New World. Yearbook of Physical 

Anthropology 42:153-188. 
Powell, |. E, &: J. C. Rose. 1999. Report on the osteological assessment of the 

"Kennewick Man" skeleton (CENWW.97.Kennewick). Internet citation: kennewick/powell_rose.htm. 
Powers, W R., & J. F. Hoffecker. 1989. Late Pleistocene settlement in the Nenana \'alley, 

central Alaska. Am. Antiquity 54:263-287. 



Richer t, S. 1999. The Wood/and Iroqi/ois People of Southern Ontario: A Dental Assment of Their 

Population Affinity. M.A. thesis, Department of Anthropology, Arizona State 

Universit}^, Tempe, AZ. 
RoUer, K. 1992. Near Eastern Dental Variation Past and Present M.A. thesis, Department of 

Anthropology, Arizona State Universit)'^, Tempe, AZ. 
Ruhlen, M. 1976. A Guide to the iMUguages of the World, languages Universals Project Stanford 

University, Stanford, CA. 356 pp. 
. 1 994. The Origin of Tanguage: Tracing the Evolution of the Mother Tongue. John Wiley 

& Sons, New York, NY. 239 pp. 
. 1998. The origin of the Na-Dene. Proc. Natl Acad Sci. U.S. A. 95:13994-13996. 

Scott, E. B. 1998. Maxillary Premolar Accessory Ridges {MxPAK): Worldwide Occurrence and 

Utility in Population Differentiation. M.A. thesis. Department of Anthropology, Arizona 

State University, Tempe, AZ. 
Scott, G. R. 1973. Dental Morphology. A Genetic Study of American White Families and 

Variation in Tiving Southwest Indians. Ph.D. dissertation. Department of Anthropology, 

Arizona State University, Tempe, AZ. 
. 1991. Continuit)^ or replacement at the Uyak site: A physical anthropological 

analysis of population relationships. Pages 1—56 in D. E. Dumond & G R. Scott, eds.. 

The Uyak Site on Kodiak Island: Its Place in Alaskan Prehistory. Universit}' of Oregon 

Anthropological Papers No. 44, Eugene, OR. 
. 1994. Teeth and prehistory on Kodiak Island. Pages 67—74 in T. L. Bray & T 

W. I-GUion, eds., Reckoning with the Dead: The Tarsen Bay Repatriation and the Smithsonian 
Institution. Smithsonian Institution Press, Washington, DC. 

Scott, G. R., & C. G Turner II. 1997. The Anthropology of Modern Human Teeth: Dental 
Morphology audits Variation in Recent Human Populations. University of Cambridge Press, 
Cambridge, MA. 382 pp. 

Sinitsyn, A. A., & N. D Praslov, eds. 1997. Radiocarbon Chronology of the Palaeolithic of 
Eastern Europe and Northern Asia: Problems and Perspectives. Russian Academy of 
Sciences, Insitute of the History of Material Culture, Saint Petersburg. 143 pp. 

Soffer, O. 1985. The Upper Paleolithic of the Central Rjissian Plain. Academic Press. Orlando. 
539 pp. 

. 1990. Before Beringia: Late Pleistocene bio-social transformations and the col- 
onization of northern Eurasia. Pages 230—238 in N. M. Shakhmatova, ed., 
Chronostratigraphy of the Paleolithic in North, Central, East Asia and America. Papers for the 
International Syumposium. Institute of History, Philology and Philosophy, Siberian 
Branch, U.S.S.R. Academy of Science, Novosibirsk. 



Stanford, D. 1999. Who were the first Americans? Pages 6—9 /// C Wood, ed.. Annua/ 

Repor/, "National Miisi'iii?/ of Natural History. Smithsonian histitution, Washington, D.C. 
Steele, D. G., & |. F. Powell. 1992. Peopling of the Americas: Paleobiological evidence. 

Hum. Biol. 64:303-336. 
Straus, L. G. 2000. Solutrean settlement of North America? A review of reality. Am. 

Antiquity 65:2X9-226. 
Szathmary, E. }. E. 1979. Blood groups of Siberians, Eskimos, and Subarctic and 

Northwest Coast Indians: The problem of origins and genetic relationships. Pages 

185—209 in W S. Laughlin & A. B. Harper, eds.. The First Americans: Origins, Affinities 

and Adaptations. Gustav Fischer, New York, NY. 
. 1985. Peopling of North America: Clues from genetic studies. Pages 79-104 in 

R. Isjrk & E. Szathmary, eds.. Out of Asia: Peopling the Americas and the Pacific. Journal 

of Pacific History, Canberra. 
. 1993. Genetics of aboriginal North Americans. EzW. Anthropol. 1:202—220. 

Tuohy, D. R., & A. Dansie. 1997. New information regarding early Holocene manifesta- 
tions in the western Great Basin. Nevada Hist. Soc. Quart. 40:24—53. 

Turner, C. G. II. 1983. Sinodonty and Sundadonty: A dental anthropological view of 
Mongoloid microevolution, origin, and dispersal into the Pacific basin, Siberia, and 
the Americas. Pages 72—76 in R. S. Vasilievsky, ed., Tate Pleistocene and Early Holocene 
Cultural Connections of Asia and America. USSR Academy of Sciences, Siberian Branch, 

. 1985. The dental search for Native American origins. Pages 31—78 /// R. Kirk 

& E. Szathmary, eds.. Out of Asia: Peopling the Americas and the Pacific. Journal of Pacific 
History, Canberra. 

. 1987. Late Pleistocene and Holocene population historv of East Asia based on 

dental variation. Am. J. Pbys. Anthropol. 73:305—321. 
. 1990a. Major feamres of Sundadont)' and Sinodont}', including suggestions 

about East Asian microevolution, population history and Late Pleistocene relation- 
ships with Australian Aboriginals. Am. J. Phys. Anthropol 82:295—317. 
. 1990b. PaleoUthic teeth of the central Siberian Altai Mountains. Pages 239—243 

in N. A. Ahmerova & N. M. Shakhmatova, eds., Chronostratigraphy of the Paleolithic in 
North, Central, East Asia and America. Papers for the International Symposium, 
U.S.S.R. Academy of Sciences, Siberian Branch, Novosibirsk. 

. 1991. The Dentition of Arctic Peoples. Garland Publishing, New York, NY. 281 pp. 

. 1992a. New World origins: New research from the Americas and Soviet Union. 

Pages 7—50 in D. J. Stanford & J. S. Day, eds.. Ice Age Hunters of the Rockies. Denver 
Museum of Natural History and University Press of Colorado, Niwot, CO. 



. 1992b. Sundadonty and Sinodonty in Japan: The dental basis for a dual origin 

hypothesis for the peopUng of the Japanese islands. Pages 95-1 12 in K. Hanihara, ed., 
Japanese as a Member of the Asian and 'Pacific 'Populations. International Research Center 
for Japanese Studies. International Symposium 4. Kyoto. 

. 1992c. The dental bridge between Australia and Asia: Following Macintosh into 

the East Asian hearth of humanity. Archaeology in Oceania {Perspectives in Human Biology 
2) 27:143-152. 

1992d. Microevolution of East Asian and European populations: A dental per- 

spective. Pages 415-438 in T. Akazawa, K. Aoki, & T Kimura, eds., The Evolution and 
Dispersal of Modern Humans in Asia. Honkusen-Sha Pub. Co., Tokyo. 

1 994a. New dental anthropological observations relevant to the human popu- 

lation system of the Greater Beringian Realm. Pages 97-106 in W. W Fitzhugh & V. 
Chaussonnet eds.. Anthropology of the North Pacific Kim. Smithsonian Institution Press, 
Washington, DC. 

. 1994b. Relating Eurasian and Native American populations through dental 

morphology. Pages 131-140 in R. Bonnichsen & D G Steele, eds.. Method and Theory 
for Investigating the Peopling of the Americas. Center for the Study of the First Americans, 
Oregon State University, Corvallis, OR. 

. 1995. Shifting continuity: Modern human origin. Pages 216-243 in S. Brenner 

& K. Hanihara eds.. The Origin and Past of Modern Humans as Viewed from DNA. 
Proceedings of the Workshop on the Origin and Past of Homjo sapiens sapiens as 
Viewed from DNA — Theoretical Approach, Kyoto, December 14-17, 1993, 
International Institute for Advanced Studies. Recent Advances in Human Biology, 
Vol. 1 , C.E. Oxnard, Series Editor. World Scientific, Singapore. 

1998. An update on the dental anthropological interpretation of the peopling 

of Siberia and the Americas. Pages 180-188 in A. P. Derevyanko ed.. Pleistocene 

Paleoecology and Stone Age Cultures of Northern Asia and Adjacent Territories. Proceedings 

of the International Symposium Russian Academy of Sciences, Siberian Branch, 

Institute Archaeology and Ethnography, Novosibirsk. 
Turner, C. G II, & J. D Cadien. 1969. Dental chipping in Aleuts, Eskimos and Indians. 

Am. J. Phjs. AnthropoL 31:303-310. 
Turner, C. G II, & D. E. Hawkey. 1991. World variation in Tome's root. Am. J. Phjs. 

Anthropol, Suppl. 12:175. 
. 1998. Whose Teeth Are These? Carabelli's Trait. Pages 41-50 in]. R. Lukacs, 

ed.. Human Dental Development, Morphology and Pathology. A Tribute to Albert A. Dahlberg. 

University of Oregon Anthropological Papers No. 54. University of Oregon Press, 

Eugene, OR. 



Turner, C. G. II, Y. Manabc, & D. \i. Hawkey. 2000. The Zhcnikoudian Upper Cave den- 
tition. Acta Anthropoiogka Sinica 19:253—268. 

Turner, C. G. II, & M. A. Markowitz. 1990. Dental discontinuit)- between Late 
Pleistocene and recent Nubians. Peopling of Eurafrican-South Asian triangle I. Homo 

Turner, C. G. II, C. H. Nichol, & G. R. Scott. 1991. Scoring procedures for key mor- 
phological traits of the permanent dentition: The Arizona State Universit)' dental 
anthropology system. Pages 13—31 in M. A. Kelley & C. S. Larsen eds., Advances in 
Dental Anthropology. Wiley-Liss, New York, NY. 

Vasil'ev, S. A. 1996. l^te Palaeolithic of the Upper Yenisei {As Represented by Multicomponent 
Sites Near Maina Village) Institute for the History of Material Culture, Russian 
Academy of Sciences, St. Petersburg (in Russian). 223 pp. 

Vasilievsky, R. S. 1981. Report on Primorian prehistory presented at Working Field 
Conference (Anchorage, Kodiak, Old Harbor, Three Saints Bay, Unalaska, Umnak, 
Pribilovs), September 3—12. 

Voegelin, C. R 1958. The dispersal factor in migrations and immigrations of American 
Indians. Pages 454—461 in R. H. Thompson ed., Migrations in NeiP World Culture 
History. Universit)^ of Arizona Social Science Bulletin No. 27. Tucson, AZ. 

Voevoda, M. I., A. G. Romashchenko, V. V. Sitnikova, E. O. Shulgina, & V. R Kobsev 
2000. A comparison of mitochondrial DNA polymorphism in Pazyryks and modern 
Eurasian populations. Archaeology, Ethnolog)i (& Anthropology of Eurasia. 4:88—94. 

Weets, J. D. 1996. The Dental Anthropology of Vanuatu, Eastern Melanesia. M.A. thesis. 
Department of Anthropology, Arizona State Universit}-^, Tempe, AZ. 

West, R H., ed. 1996. American Beginnings: The Prehistory and Palaeoecology of Beringia. 
Universit}^ of Chicago Press, Chicago, IL. 576 pp. 

Yesner, D. R. 1998. Special review. American Beginnings. Am. Antiquity 63:169—174. 

Zegura, S. 1999. Review, The Origins of Native Americans: Evidence from. Anthropology Genetics^ 
by M.H. Crawford, ed. Am. Anthropol 101:199-200. 

Zubov, A. A., & N. I. Ivhaldeeva. 1979. Ethnic Odontology of the USSR Nauka, Moscow. 
256 pp. (in Russian) 


Chapter Seven 

The Migrations and Adaptations 

OF THE First Americans 

Clovis and Pre-Clovis Viewed 

FROM South America 

A.C. Roosevelt, John Douglas and Linda Brown 

^^^^^^ Until recently, in most anthropology textbooks, the first Americans 
^^^^^^B were identified as the Clovis people, specialized spear-hunters 
'**'^^VP^fB whose ancestors crossed the land bridge from Siberia through the 
** " ^ • ice-free corridor into the North American plains about 12,000 years 
ago at the end of the Ice Age (Fagan 1987; Fiedel 1987, 1992; Haynes 1964; Martin 1967; 
Wormington 1957).^ Decimating the large game with their deadly, bifacial fluted projec- 
tile points, they moved south through upland grasslands until they reached the tip of 
South America about 11,000 years ago. In opposition to this view, some researchers {e.g., 
Bednarik 1989; Bryan 1991; Dillehay 2000; Dixon 1999; Krieger 1964; Meltzer 1993; 
WiEey & PhiUips 1958) have long claimed that the human colonization was begun as 
early as 15,000 or even 20,000 years ago by opportunistic foragers who did not yet have 
carefully flaked projectile points. Collecting a wide range of foods with simple stone 
tools flaked on only one side, these people migrated south through diverse habitats long 
before the big-game hunters, according to the pre-Clovis theory. But whenever they were 
supposed to have arrived, the first people were thought to have been Asian Mongoloid 
people from northeast Asia {e.g.. Turner, this volume). A map of possible Paleoindian 
sites discussed in this chapter is presented in Figure 1. 

Today, nearly 70 years after the first excavations at Clovis, New Mexico, new sites and 
new dates from both South and North America have created problems for the earlier 
theories about the migrations and adaptations of the first Americans, as we explain in 
this chapter. Consensus has not kept up with new findings, and most current scholarly 
analyses and media scenarios (Begley & Murr 1999; Booth 1999; Dillehay 2000; Dixon 
1999; Egan 1999; Preston 1997; Wilford 1999) do not fit the accumulated data. Although 
the tide of public and scholarly opinion has definitely turned against Clovis as the earli- 
est culture, nonetheless, the data we have reviewed give no definitive evidence for human 
cultures in the Americas before about 12,000 yr B.P. All cultures represented as older 


Rousi^vi':i:r, douglas and brown 


23.720 1220(?1,- 


10,680 J50(?)« , 10,760 1130i?) 



11,200 12 20(?): 

MEADOWCROFT '°« "^ "°^ ncBf 

UNION PACIFIC • 10.864 1141 


Il,a80±350 ,,,200 1500; 

19,100 ±810{?) 

10,650 180 





10,560 1250 

"• 11,300 1240(?), ,1.2,0 



16,070 «70<?): 


,0,920 1250 



11.660 t80{?); 



11.470 tSOO(?); 





11,300 tl20 

11.660 ±60 





33,370 ±530(7): 

ca. 12,500(?) 


116801500(?) TlAPA 00 DRAGAO 


T AUCEBOER ["•'<* "^'-l 

W.200 11,150(?) 

|rS-I-68 11,555 1230(?) 
RS-I-69 10,985 1100 
RS-I-66 10,810 1275 

11,880 1250(7): 
10,600 190 

Figure 1. Map of possible Paleoindian sites throughout the Americas. Quesuon marks indicate 
questionable dates. Drawing by Joyce Pendola of Conde Nast. (Roosevelt 2000a, courtesy of Namral 



than ca. 12,000 years fail to withstand careful scrutiny. Furthermore, all weU-document- 
ed early cultures have at least some artifacts carefully flaked on both sides; there are no 
exclusively unifacial industries. Current evidence summarized in our chapter suggests 
that the habitat, subsistence, artifacts, and chronology of early Paleoindians are not those 
predicted by either the pre-Clovis or Clovis migration theories (Roosevelt 2000a, 2000d). 
In addition, despite claims for early migrations from Europe and elsewhere, the biology 
of Paleoindian populations throughout the hemisphere consistently points to a geo- 
graphic origin in northeast Asia, not in Europe (Chatters et al. 1999), southeast Asia, or 
Africa (Powell 1993; Steele & Powell 1993; Turner, this volume). 

The dissolution of the evidence for pre-Clovis cultures does not resurrect Clovis as 
ancestor, however. Clovis does not appear to have been the earliest culture in the 
Americas and did not set a hemispheric cultural-ecological pattern. First, the early peo- 
ples occupied a much wider range of habitats than that envisioned in the Clovis theory. 
Cool upland plains, the preferred Clovis habitat, was not the habitat of most Paleoindian 
cultures, which by now have been found in diverse regions: northern, wooded river 
basins, temperate and tropical zone seacoasts, tropical lowlands, and southern pampas, 
as well as the upland plains favored by Clovis people. Second, in contradiction to the 
Clovis theory, remains from many early sites indicate that most people Uved by broad- 
spectrum foraging of diverse modern biota, not by specialized hunting of now-extinct 
megafauna (see DiUehay & Rosen, this volume). Most well-documented, published sites 
contain no megafaunal bone contemporary with the human occupations, but often do 
contain such foods as fruits, nuts, vegetables, fish, shellfish, and small game animals. 
Specialized big-game hunting was not the "bread-and-butter" of most Paleoindians but 
merely one regional group's particular adaptation to cool, dry grasslands poor in the 
diverse foods consumed by most other groups. Third, the many new radiometric dates 
from sites throughout the hemisphere make Clovis just one of several, contemporary 
regional cultures developed soon after the initial migration, not the ancestor of the other 
early Paleoindian cultures. With the spate of new measurements, the age of plains Clovis 
culture has been revised significantly downward to an initial age of no more than 1 1 ,200 
yr B.P. and an ending age of about 10,800 yr B.P, as we document below. The hemi- 
spheric Clovis horizon, based mainly on scattered, poorly-dated finds of rare, presumed 
fluted points,^ has evaporated in the face of regional sequences that have now been 
established in South America. Three different regions of South America now have con- 
sistent radiocarbon series beginning just about the same time as Clovis. These southern 
cultures, thousands of miles away and entirely distinct from Clovis culturally and eco- 
logically, cannot very well be derived from Clovis. Contemporary with the earliest secure 
Clovis dates, in South America there already were maritime foragers on the Pacific coast. 



guanaco hunters in southern pampas, and tropical forest foragers in the eastern tropical 
lowlands. Few of the securely dated and documented early Paleoindian southern cultures 
have fluted, parallel-sided points; most have unfluted, triangular points of different basal 

The body of this chapter critically evaluates the two most popular theories about the 
peopling of the Americas — the Clovis and the pre-Clovis theories — in the light of cur- 
rent data on early North, Central, and South American Paleoindian cultures and in their 
stead offers an alternative — the "Clovis in Context" theory. Our discussion and illus- 
trations document key cultures and sites throughout the hemisphere, but focus particu- 
larly on the Clovis and Monte Alegre cultures in southwestern North America and east- 
ern South America, respectively, because their similarities and differences exemplify the 
significance of the new data. The new picture of Paleoindian cultures based on hemi- 
spheric data not only changes understanding of the initial human colonization and adap- 
tive radiation in the Americas but also has interesting implications for general theories 
about human evolution and human nature. In our conclusions, we summarize what we 
see as the possible general significance of this picmre in early New World culture histo- 
ry and beyond, and we point to possible directions for future research. 

Early Paleoindians in North A.menca 

The Clovis migration theory 

In the original Clovis migration theory (Pagan 1987; Fiedel 1987, 1992, 2000; Haynes 
1964; Jennings 1983; Martin 1967, 1984), the first Americans were Mongoloid people 
who hunted large herd game with distinctive, parallel-sided, fluted spear points named 
after the site near Clovis, New Mexico. Their ancestors were thought to have come from 
the steppes of Siberia into the Americas on foot shortiy before 12,000 years ago. They 
went into the cool, upland grasslands of the continental interior, and there thev preyed 
on large mammals such as mammoth, long-horn bison, horse, and camel. Since 
American animals were unaccustomed to human hunting, the highlv efficient Clovis 
spear-hunting methods were thought to have quickly diminished herds, forcing people to 
move rapidly south through the hemisphere in search of more game. Settling the North 
American plains by 12,000 or 11,500 years ago, according to the theory, they moved 
down the Andean uplands and reached the far southern pampas by about 1 1 ,000 vr B.P. 
(Fiedel 1987, 1992; Haynes 1964; Lynch 1983; Martin 1967; Mosiman & Martin 1975). 
At different periods of research on Paleoindians,-^ as they have been called, ideas about 
the timing of migration have varied; but the basic geograpliical scenario held firm for a 
long time. 



According to the theory, this lifest\de of terrestrial big-game hunting in open, tem- 
perate habitats was t\?pical throughout the hemisphere, and only after the large game had 
gone extinct, and climates warmed and sea levels rose at the beginning of the early 
Holocene, the recent climate period about 10,000 years ago, did people move into low- 
land floodplains and coasts (Fiedel 1987:82-159; Lynch 1983; Martin 1984:368-369). 
Only then, did they turn to more laborious forms of foraging upon small game, fish, 
shellfish, and plant foods and in some areas eventually domesticated local plants and ani- 
mals. In this post-Pleistocene cultural period (called the Archaic) new triangular, 
notched, and sometimes stemmed forms of projectile points are supposed to have 
appeared, appropriate to the new economies (e.g., Anderson et al. 1996; Prison 1991; 
Sassaman 1996). The tropical rainforests, which reseachers considered disease-ridden 
and poor in both game and starchy foods, were not thought to have been colonized until 
the spread of agriculture sometime after 5,000 years ago {e.g., Fiedel 1987:82—159, 
1992:84-165; Lynch 1983:89). 

Harly Paleoindian cultures: Universal standards of validity 

This picture of the migration was elegantly simple and fit some tenets of optimal for- 
aging theory in human evolution (see Binford 1968), but it was not based on a body of 
established, definitive data throughout the Americas. Few sites had been excavated and 
published, and most were in the North American southwest, the Clovis heartland, not in 
the zone that Clovis people supposedly colonized. As our discussion below illustrates, 
early archaeological techniques of excavation and analysis were often crude compared to 
today's: lacking microstratigraphic excavation, controlled piece plotting, and computer- 
ized instruments to record and analyze data quantitatively. The early dating methods were 
experimental, unpredictable, and imprecise. Most of the sites of the supposed Clovis 
descendants in the colonization zone were poor in diagnostic culmral and biological 
remains and lacked suites of reliable, well-documented Pleistocene radiocarbon mea- 
surements. Knowledge of site formation processes was not yet well developed, and the 
implications of stratigraphic complexit)^ were not as widely recognized as today, so dis- 
parate objects that had come together in layers through disturbances were often taken to 
be contemporary. Lacking detailed publications for most sites, conclusions of hemi- 
spheric scale lacked adequate data from the specific regions. 

Since the discovery of Clovis, archaeologists' growing experience and changes in 
research technology have led to an improved body of published data for clarifying early 
New World human history and ecology. Critical evaluation of cultures has been aided by 
advances in taphonomy, the study of the complex processes of deposition of sediments 


R()()si:vi'i:r, douglas and brown 

and objccrs in sites (e.g., Bchrcnsmcvcr & I (ill 1980). New radiocarbon dadng methods 
have allowed taxonomically-identified biota and individual ardfacts from sites to be 
directly and precisely dated (Taylor 1987:90—95). Together, these advances have led to the 
development of universal criteria of reliability' against which to evaluate proposed early 
Paleoindian occupadons (Bowman 1990; Aitken 1990; Taylor 1987, 1992, 1997; Taylor & 
Aitken 1997; Haynes 1992; Mead & Meltzer 1984; Meltzer & Mead 1985; Taylor 1987; 
Roosevelt 1998a, 1998b, n.d.a, n.d.b; Roosevelt & Morrow n.d.; Stafford efal 1991; Toth 

According to the universal criteria, valid early Paleoindian sites have consistent series 
of accurate, statistically precise dates with standard-error bars no greater than about 300 
years. The dates are run on single, taxonomically-identified single objects of carefully 
cleaned, biologically identified cultural carbon of adequate quantit}' derived from record- 
ed stratigraphy in primary association with artifacts, documented in peer- reviewed pub- 
lished accounts. Especially reliable are dates on individual fruits of identified botanical 
species or the purified amino acids extracted from individual bones or teeth of prey ani- 
mals (Stafford 1990; Stafford et a/. 1991). Valid regional cultures, moreover, would be 
represented by several such sites with comparable dates, artifacts, and biological remains. 

In contrast, a site would not be reliably dated if it has few or no dates, statistically 
inconsistent dates, or dates with large standard error bars greater than about 300 years. 
Dates would not be valid if they were run on inadequate sized, incompletely cleaned 
samples, biologically-unidentified carbon samples, or samples in contact with older car- 
bon contaminants. Dates would be unreliable if run on aggregate carbon samples 
amassed from disparate stratigrapliic contexts or on samples without documented pri- 
mary cultural association. Cultures would be questionable if represented by only one site, 
especially if the site is a problematic one that has been in existence for decades without 
resolution of reliability problems according to the standards. Cultures, also, would be 
considered invalid if they were reported only in non-specific, non-peer reviewed publi- 
cations or from unpublished, oral presentations or personal communications. 

To be useful, reliabilit)' scrutiny needs to be even-handed. There has been a common 
tendency for archaeologists to scrutinize the reliabilit}' of pre-Clovis sites or Clovis rivals, 
but, as we describe next, current claims for the age of Clovis relv on dates, sites, and cul- 
tures with obvious reliability problems. When a// such obviously invalid sites, dates, and 
cultures are set aside, a very different picture of the Paleoindian migration comes into 



Clovis: Younger than we thought 

In the new picture, Clovis is no longer the earliest established Paleoindian culture it 
was assumed to be before much was known about the archaeology of regions outside of 
the Clovis sphere in North America. Over the decades, as dating methods have devel- 
oped, Clovis has grown progressively younger. Initially said to be from 12,000 to 11,500 
years old (Haury et al. 1959), its age was first trimmed to between 11,500 to 11,200, 
(Haynes 1964, 1966, 1987), and then to the period from 11,200 to 10,900 yr B.P. (Haynes 
1992). Review of the early sources shows that, in each era, the ages scholars quoted for 
Clovis were considerably earlier than that indicated by the statistics of the available dates. 
Recently, in this vein, several scholars called for a return to a > 11,500 yr B.P. beginning 
date for Clovis (Fiedel 1999a, 1999b; Taylor et al. 1996), and, in fact, much of the 
Paleoindian literature cites the span from 11,000 to 11,500 or even to 12,000 yr B.P. as 
the established age of Clovis (Fiedel 1987:56; Ferring 1994; Frison & Bradley 1999; 
Hannus 1990; Jennings 1983:50; Preston 1997) and, by implication, the assumed age of 
any New World complex that has parallel- sided or fluted points {e.g., Gramley 1993; 
McAvoy & McAvoy 1997; Di Peso 1965; Ranere & Cooke 1991). 

According to the criteria of reliabilit)', however, no Clovis site has valid dates that 
early; all such dates fail the criteria on multiple counts (Roosevelt 1998a). The only ones 
at ca. 12,000 years were soHd-carbon dates run early in the history of radiocarbon dating 
and are no longer accepted by dating specialists (Aitken 1990:77; Bowman 1990:31-32; 
Taylor 1987:76-82). Solid-carbon dates from the same site context were unreliable and 
could range as old as 12,000 and as young as 7,000 yr B.P. (see Figure 2). When Clovis 
sites were redated with the improved technique of gas counting, none produced dates as 
early as 12,000 years. Although recently, dates between 11,600 and 11,300 yr B.P. have 
been claimed for the culture (Fiedel 1999a; Taylor et al. 1996), these dates do not with- 
stand scrutiny by universal standards of radiocarbon dating. Some are rare, statistical out- 
liers in direct conflict with large series of consistently later dates on associated samples 
from the same context (as at Lehner, Murray Springs, and Domebo, discussed below); 
others are from nonhuman carbon in soil horizons earUer than the human occupations 
(as at Clovis, see discussion below; Figure 3; Roosevelt 1998a, n.d.b., and Appendix 1). 
Many plains Paleoindian carbon dates were on carbon vulnerable to contamination by 
too-old carbon. Such dates often were run on samples compiled from small, botanically 
unidentified dark flecks in disparate stratigraphic locations in sites, some of which have 
now been shown experimentally to be contaminated with large amounts of geological 
carbon, such as lignite coal and/or carbonate (Holliday 1997:50-148, 232-233; Mill Iron 
site, Haynes 1992:324; Dent site, Dixon 1999, Haynes et al 1998; Lewisville site, Stanford 


RoosiLvi'irr, douglas and brown 

Figure 2A. The Lehner Clovis site, Arizona. ProjectUe points (Haxiry et a/. 1959:17, figure 13, used 
with permission). 

1983), materials abundant in Cretaceous strata incised by the streams bv wliich many 
Clovis sites lie (Harbour 1975). Most of the others were run on tree charcoal, wliich by 
its namre has an inherent age older than the date of use by humans, since trees grow for 
many years before being cut for firewood (Aitken 1990: 8; Bowman 1990:15; Roosevelt 
1998a). Other early Clovis samples were too small or too poorly cleaned to achieve nec- 
essary accuracy (as at Clovis, Domebo, UP Mammoth site, Colby site, see Appendix 1), 
and many dates have error bars from 350 to 600 yr, a range considered too large for pre- 
cise, definitive dating (Bowman 1990; Taylor 1987:102-104, 123-126). 

The process of chronological attrition by reliabilit\' evaluation means that there are 
only five high plains Clovis sites (Lehner, Murray Springs, Domebo, Dent, and Anzick) 
that have consistent series of precise, accurate dates on cultural carbon from document- 
ed contexts. Fort\"-six assays on 35 adequate-size samples of rigorously-cleaned, biolog- 










6877 ±450) 

7133 1350 (A- 33) 

AV. 6789 ±450 

7205*450 (A-34) 

7022 + 450 (A-32) 
8330 ± 450 (A-30) 
IO,900t 450 (A-400) 
l2,OOOt 450 (A-40b) 

ll.ieO ± 140 (K-554) 

ll,290±500 (M-8II) 



7O0O-5O0O B. P 

HUNTS 13,000 B. P. 

Figure 2B. The Lehner Clovis site, Arizona. Stratigraphy and inconsistent radiocarbon dates 
(Haury et ai, 1959:25, figure 16, used with permission). 



icalK' idcnlitlcd, cultural carbon from documcnrcd srrarigraph\ in the five sites with doc- 
Limcnred (Mox'is lithics have produced no stadsdcally-consistent dates earlier than the 
period ca. 11,200 to ca. 10,800 yr B.P. (see y\ppendix 1). Many writers now cite 11,200 yr 
B.P. as the age of Clovis, but this is only the earliest possible beginning age for Clovis, 
not the average age of Clovis sites, since the reliable Clovis date series averages no ear- 
lier than 10,900 years ago (specifically, 10,848 ± 20 yr B.P.). Moreover, Clovis proponents 
have claimed that the culture has a corrected age of 13,500 calendar years and thus is ear- 
lier than other Paleoindian cultures (Fiedel 1999), but this asserdon is based on the inter- 
calibration of 0/?/]' the unreliable Clovis dates (those run on unidentified carbon, carbon 
from contaminated contexts, or samples without documented cultural or stratigraphic 
association)^, not the reliable dates, wliich calibrate at ca. 12,906 yr B.P. 

Remarkably, the dates from the famous Clovis t)'pe site at Blackwater Draw Localit)' 
1 fail to meet reliabilit)' standards on all counts. Only three Pleistocene dates for that cul- 
ture there have been reported from the site: 11,630 ± 400, 11,170 ± 350, and 11,040 ± 
500 yr B.P. (see Appendix 1) (Boldurian & Cotter 1999; Haynes & Agogino 1966; Ha^-nes 
et al. 1966; Hester et al. \^11; Damon et al. 1964:100-101). Said to be carbonized plants 
in secondary literature and thus widely assumed to be cultural charcoal, the primary 
reports show that they were wet-preserved aquatic plants, a noncultural carbon source 
subject to metabolic contamination by geological carbonate, which is present at the site 
(Bowman 1990:26; Cotter 1938, 1939; Damon et al 1964; Hester et al \912; Holliday 


CavPSlTE y'"^ 


.COTTER 1936. 1937 





\/ Ajcji-'^'^ ~7 ^^— - 





^Nuu ^^::> 

— --^-?~-L^ / 

0^11 ^^JpT/nMM STA. 8 



P /e/ / . 



~^--^^^^\// y( *'""<« 8 SHAT 1954 

-_ / 1 / 


) Y ^^^^^i^% 

^ 6/ ' — — — _ / _— -"" 



yty'^^ VfENOOBf a DITTER1 


V' licT- 

^jk^ "j .MAMMOTH / 


i-rf^ iELlNEK 1956 


'^ / 

::^^fer~--i_^ ^ 




\J / 



1 ^ — , 

— " 


1. Il,630i400 IA49II 

2 It, 1701 350 (A 4811 

J 11,0401 500 IA490) 

4, I0il70±250 IA488) 

5. 10,4901 200 (A 4921 

6 0,2501 320 lA 379 -A 3801 

7 0,4901 900 lA 3861 

8 9,8901 290 (A 4891 

9 8,4701350 IA5I2I 

10 0-169 6,300 1150 6 f 

II 0-170 6^30 1 150 BP 

12 0-157 4,950 1 150 B P IProt wo ytuig-HoyX! l%6 n 261 

Figure 3. Stratigraphy and radiocarbon dates on namral carbon at Clovis, Blackwater Draw, New 
Mexico. (Hester et al. 1972:174, figure 130, used with permission). 



1997:232-233; Roosevelt 1998a; Stafford et al. 1987). In addition, these samples were too 
small to be adequately cleaned and thus could not be counted long enough to achieve 
acceptable error bars. Furthermore, the site report shows that two of them (1 and 3) 
came from noncultural associations stratigraphically several feet below the bone bed 
(Figure 3). The only one (sample 2) that had cultural association was in strata that were 
deformed by a spring that had moved materials vertically in the site (Haynes & Agogino 
1966:816, figure 5B; Hester et al. 1972, figure 130; Warnica 1966). As a result, the epony- 
mous Clovis site still remains undated after more than three decades of research 
(Roosevelt 1998a). Its age could readily be resolved by AMS dating of individual amino 
acids extracted from one of the mammoths (Stafford 1990; Stafford et al. 1991) in pri- 
mary association with artifacts (Hester et al. 1972:71-87). 

The ca. 11,500 age of another site, Domebo, Oklahoma (Leonhardy 1966), claimed 
to suppport an earlier age for Clovis (Fiedel 1999a; Taylor et al. 1996), also is invalid 
under the criteria for reliability. This age was based on a single, imprecise outlier date 
with a much too-large error of 450 years. The valid site age is securely established at ca. 
10,900 yr B.P. by seven consistent dates run on four samples of distinct materials (two 
mammoth bones and two wood fragments) found associated with projectile points in the 
bone bed (Appendix 1). 

Aubrey in Texas (Ferring 1989, 1990, 1994, 1995; Humphrey & Ferring 1994), also 
widely claimed as proof for an initial Clovis age of 11,600 yr B.P. (Fiedel 1999a; Stanford 
1991; Taylor et al 1996), also is not a valid Clovis site. The only two supposed Clovis 
dates, 11,590 ± 90 and 11,540 ± 110 yr B.P, meet few of the criteria of reliability. They 
were run on aggregates of stratigraphicaUy-dispersed, biologically-unidentified carbon 
subject to contamination from the peat, old wood, and calcium carbonate recorded at the 
site (Ferring 1994:31-118). Most unfortunate, the samples were not collected from an 
intact, sealed terminal Pleistocene stratum but from a disconformity, an erosion surface 
on a 15,000 to 20,000-year-old geological alluvial stratum composed of materials eroded 
from coal-bearing Cretaceous bedrock of the region (Ferring 1989, 1994: 25, figures 3.8 
and 3.9). Such a context would not be a secure find-place for dates, according to the cri- 
teria. Moreover, the dates were not on individual specimens of culturally utilized botan- 
ical remains or fauna, and neither date has specific stratigraphic unit proveniences in 
Clovis-age, terminal Pleistocene strata, documented with detailed stratigraphic sections 
of specific excavation units. In addition, these dates do not have published measured sta- 
ble carbon isotope ratios, without which the exact age range of radiocarbon dates is 
unclear (Taylor 1987:123). Also troubling, no specifically Clovis-style lithics have been 
documented in stratigraphy at the site, and no stone artifacts have been illustrated or tab- 
ulated in print or related individually to specific excavation layers, features, or dates. This 



site could, h()\\c\cr, srill be dated reliably by assays on the taxonomically-identified cul- 
tural charcoal, animal bone, or burned lithic artifacts present in the site, once the exca- 
vation details are published. 

There is, then, no empirical support at all for an age of 12,000 to 1 1,50(1 yr B.P. for 
the Clevis culture. By standard dating criteria, the age of the Clovis culture is well-estab- 
lished from the large series of dates from the five documented Clovis sites in the time 
range from ca. 11,200 to 10,800 yr B.P., with an average age of ca. 10,900 yr B.P. These 
dates, which are on a wide range of different cultural materials — human bone, prey-ani- 
mal bone, and charcoal — are consistent regardless of material. This revision of the 
chronological position of Clovis means that many other, distinct New World cultures are 
its full contemporaries, not its descendants. It also means that one North American cul- 
ture previously considered a Clovis contemporary must be its predecessor, as we shall 
dscribe further on. 

Where did the Paleoindians come from and hoiv did thej come? 

The current archaeological and biological evidence does not support the assumptions 
of the Clovis migration theory about Paleoindian migration routes. For example, there 
are no instances of Late Pleistocene, pre-Clovis fluted points at the entry-point to the 
Americas; all are early Holocene (Clark 1991; Yesner 1996, n.d.). Despite intensive 
efforts, research also has produced no Clovis-like, big-game-hunting cultures in the inte- 
rior of northeast Asia. Interior Siberian cultures of the relevant periods have unfluted, 
lanceolate or triangular, stemmed bifaces and subsistence by mixed foraging (Figure 4a) 
(Derev'anko 1998; Dikov 1996; Powers 1996). Late Pleistocene cultures along the north 
Asian coasts have broad-spectrum foraging cultures with stemmed, unfluted bifaces 
(Aikens & Higuchi 1982; Aikens & Akazawa 1996), some of which (Figure 4b) were 
shaped bv techniques characteristic of Clovis and other Paleoindian bifaces. The refined 
Clovis bifacial stone flaking techniques (including long, regular, diagonal, parallel soft- 
hammer thinning) and certain tool forms (large curved blades), sometimes claimed as 
evidence for Paleoindian origins in European Late Pleistocene Solutrean culture 
(Stanford 1991, and this volume) were also common in Late Pleistocene east Asia 
(Aikens & Higuchi 1982; Aikens & Akazawa 1996; Derev'anko 1998; Powers 1996). 
However, tool forms particular to Clovis — such as the parallel-sided fluted point — are 
unique local forms with no demonstrated links to Old World complexes in Europe or 
Asia (Havnes 1964), so they do not constitute evidence for Paleoindian geographic ori- 
gins in the Old World. 

Interpretations of the skeletal biology of Paleoindians in terms of geographic origins 
also have changed over the decades. Metric and nonmetric cranial characters have been 




Figure 4. Late Pleistocene stemmed projectile points from east Asia. A. Lithic assemblage of the 
Ushki Lake culture, Siberia (Derev'anko 1998:327, figure 202, used with permission). B. Triangular 
stemmed projectile points, Kusakayana, Higashi, Osaka Town Museum, Japan. Courtes)? of the munic- 
ipality of Osaka. 


R( )( )si':vi:]:r, douglas and brown 

rc\ic\\cd for approximatclv 25 directly-dated skeletons of terminal Pleistocene and early 
Holocene age (l.ahr 1995; Powell 1993; Powell &c Neves 1998; Powell & Rcjse 1999; Scott 
& Turner 1987; Turner, this volume). This review revealed that Paleoindians differ in cra- 
nial morphology from both modern North American Amerindians and modern 
Mongoloid peoples such as Japanese and Han Chinese. However, despite much media 
interest in the suggestions by some archaeologists that Paleoindians might therefore be 
culturally and biologically descended from caucasoid Europeans (Begley & Murr 1999; 
Booth 1999; Egan 1999; Preston 1997; Wilford 1999), the cranial traits at issue are not 
considered to be under strong genetic control (Turner, this volume) and were vanishingly 
common in Europe in the terminal Pleistocene (ibid.). In contrast, multiple characters of 
human teeth known to be under genetic control are very common in both Paleoindians 
and Northeast Asian populations. Paleoindian incisor shape, molar cusp patterns, and 
root numbers have a high percentage similarit)' with the east Asian pattern termed 
Sinodont\% but a low percentage similarit}^ with Late Pleistocene and Holocene 
Europeans' teeth (Turner, this volume). Mitochondrial DNA "X- factors" claimed to Unk 
Paleoindians to White Europeans also occur widely in Asia (Merriwether, this volume). 
There is, therefore, no statistically significant empirical data that relates Paleoindians 
more closely to Europe genetically than to Asia, but there are several bodies of data that 
link them closely to east Asia. Some anthropologists have suggested (Neves &: Pucciarelli 
1991; Neves etal 1996) that Paleoindians must have come across the sea from the Pacific 
Islands because their crania are more like those of Pacific Islanders than they are like pre- 
sent northeastern Asian populations or Amerindians. However, in the Late Pleistocene 
the cranial traits at issue were common throughout east Asia (Turner, this volume). 
Although some researchers have implied that cranial morphology means that a certain 
Brazilian skeleton must be from an African or southeast Asian population (Neves et al. 
1996), the skeleton in question has not been directly dated nor compared to any others 
by dental morphology, the more appropriate feature for genetic relatedness issues. 

There is some new evidence about the possible routes that early migrants took with- 
in the New World. The Clovis migration theory had depicted humans as entering and 
moving through the hemisphere on dry land because they were supposed to have lacked 
watercraft. Several researchers have argued against these assumptions, pointing out the 
advantages of coastal resources and the lack of archaeological evidence along the sup- 
posed Clovis migration route (Bryan 1978a; Fladmark 1978; Gruhn 1994). Over the 
years, evidence has accumulated suggesting that at least some of the first people migrat- 
ed coastwise (Erlandson, this volume). Some researchers now suggest that the interior 
ice-free corridor was not open at the appropriate times to have been a route for the first 
migrants (Roberts & Julig n.d.), and others suggest that the northwest coast was almost 



entirely unglaciated, allowing easy passage of migrants with watercraft (Dixon 
1999:19—43; Dixon et al. 1997). Another body of evidence for coastal adaptations by 
Paleoindians is the stable carbon isotope data from two early Holocene human skeletons 
from the U.S. northwest, which manifest patterns of stable carbon isotopes, suggesting 
some reliance on marine food resources (Chatters 1998; Dixon et al. 1997; Johnson et al. 
in press; Powell & Rose 1999). In addition, in Daisy Cave on San Miguel Island off the 
southern California coast, an ancient garbage midden with abundant coastal food 
remains of fish, shellfish, and marine birds has produced terminal Pleistocene radiocar- 
bon dates between ca. 10,500 to 9,500 yr B.P. (Erlandsen & Moss 1996; Erlandsen et al 
1996; Erlandson, this volume). These sites are not direct proof for an initial coastal 
route, for none are early Paleoindian in age, but they are our only evidence for the time 
being, since the then coast probably is now under ~ 60 meters of water after post- 
Pleistocene sea level rises, and it has been difficult to locate coastal sites of the right peri- 
od (R. Gruhn, personal communication). However, such finds do show that some north- 
ern Paleoindians did include marine adaptations, which the Clovis migration theory did 
not allow for. More conclusively, as we shall see below, faunal evidence from South 
American sites places coastal and lowland riverine rainforest adaptations securely in the 
earliest regional Paleoindian cultures. 

When did they come, before or after the invention of projectile points'^ 

The pre-projectile point horizon theory is a long-standing alternative to the theory of 
the Clovis migration and has appeared in many different forms (Bednarik 1989; Bryan 
1978b, 1986, 1991; Dillehay 2000; Dixon 1999; Krieger 1964; Willey «Sc PhilHps 1958). It 
assumes that people came into the Americas before anatomically modern humans and 
upper Paleolithic cultures had appeared in Europe; they therefore would have lacked the 
capacit}^ and technology to make fine and bifacial projectile points for specialized hunt- 
ing of big game. Instead, they were supposed to have made crude stone flakes, and edge- 
trimmed cobble tools, and tools of perishable materials. 

After many years of research, however, neither the predicted chronology nor history 
of technology has been borne out. No valid American cultures pre-date ca. 12,000 years 
ago and none has exclusively unifacial tools. All verified early Paleoindian cultures pos- 
sessed some kind of bifacially flaked tool, and most made bifaces, although these are rare 
among the formal artifacts of many industries. One important element of the pre-pro- 
jectile point theory, however, has been borne out by recent research: that the earUest peo- 
ple would be generalized foragers, not specialized big-game hunters. 

The Calico site in California, long claimed to be 100,000-45,000 years old (Leakey et 
al 1968; Simpson 1978), never produced early stone specimens that archaeologists could 



agree \\'ere human-produced. I'urrhermore, it lacks other secure hallmarks of human 
activitv: human skeletal remains, utilized subsistence remains, and spent fuels. These 
absences suggest that this is a purely natural deposit pre-dating human occupation 
(Haynes 1973a). The Old Crow site in the Yukon region of Canada also was long 
thought to be evidence of the pre-projectile point horizon on the basis of a single radio- 
carbon date of 26,000 to 29,000 yr B.P. on bone speculated to have been used b\- humans 
when fresh (Bonnichsen 1978; Harington et al. 1975). However, follow-up dating showed 
that the undoubted cultural materials at the site were not even Pleistocene in age (Ta^4or 
1997:88, table 3.5). Similarly, several human skeletons from the tar pits on the southern 
California coast were claimed to be pre-Clovis, pre-projectile point people, but check- 
dating showed they were of much more recent date, instead (Erlandson & Moss 1996; 
Taylor 1997: 88, Table 3.5; Taylor et al. 1985). At Pendejo Cave, New Mexico, burned ani- 
mal bones and charcoal supposed to represent human activity- (IMacNeish 1991, 1992; 
MacNeish et al. 1993) are not securely associated with undoubted tools and unambigu- 
ous evidence of the human presence (Lynch 1990). 

So far, then, all these North American pre-Clovis candidates fail the universal crite- 
ria test outlined above: statistical consistency, biological identification, secure culti.u-al 
association, and repUcabiHt)^ among sites. Some researchers continue to espouse such 
sites, without addressing the specific evidence problems. But there still is absolutelv no 
good evidence in North America for an entry of humans before ca. 12,000. All human 
remains are anatomically modern, and all vaUd early Paleoindian cultures have the fuU 
range of basic hthic manufacturing technology of Upper Paleolithic cultures. 

Whatever the ecological or social conditions that encouraged people to venmre from 
Asia — whether the rapid warming of the terminal Pleistocene, or pressure from the rise 
of sedentary, pottery-making cultures in northeast Asia, or some other factors — the 
results of many decades of research still constrain their entry to at, or around, 12,000 
years ago, no earlier. 

Projectile point cultures in North ^America before Clovis? 

Numerous projectile-point cultures have been claimed to pre-date Clovis, but all fail 
to meet the criteria of reliabiUt}-. 

Some of these cultures have statistical evidence of old-carbon contamination. The 
site on a knoll at Mesa, Alaska, is claimed by its excavators (Kunz & Reanier 1994; 
Reanier & Kunz 1995) to be a pre-Clovis, early Paleoindian ancestor nearly 12,000 ^^ears 
old, but this interpretation belies its radiocarbon date statistics and Uthics, both of which 
indicate late, not early Paleoindian, age. The only two pre-Clovis Mesa dates are from 



samples that gave pairs of statistically inconsistent dates indicating a problem with old 
carbon (Figure 5). The large series of twelve precise, late Paleoindian dates accord with 
the late Paleoindian style of the associated projectile points (Goebel 1999; Roosevelt 
n.d.b). A now-discredited single pre-Clovis date from Putu, Alaska (Alexander 1987), was 
on carbon from prehuman subsoil, not on cultural carbon associated with artifacts 
(Reanier 1996). Similarly, the bone dates from Bluefish caves, northern Alaska (Morlan 
& Cinq Mars 1982) remain questionable because of old-carbon contamination from cal- 
careous rocks, lack of cultural association, and lack of published specifics on the dates 
(Ackerman 1996:511-512). 

At other supposedly early projectile point sites, Lignite coal, a source of too-old car- 
bon, has been identified. One example, Mill Iron, a bison bone bed with adjacent camp- 
site, has repeatedly been claimed as a ca. 11,600-year-old culture either contemporary 
with, or ancestral to, Clovis (Fiedel 2000:55-56; Frison 1996, 1999; Stanford 1991). The 
site is contaminated with Cretaceous coal that lies a few inches below the cultural stra- 
tum (Frison 1996: 8-9, figures 1.12, 1.13; Frison et al. 1996). Stratigraphically and cultur- 
ally, this is a single occupation bison-kiU site, not an occupation sequence, and yet its 
radiocarbon dates range widely over a span of 11,000 years (Frison 1996:8, table 1.1; 
Haynes 1992: 324). Radiocarbon dates from the same site contexts are statistically incon- 
sistent with each other. Characteristically, they were run on aggregates of dispersed, 
botanically-unidentified carbon. Check-dating of one of the samples revealed it to be 
contaminated with large amounts of lignite coal measured at > 24,000 years old (Frison 
1996:11; Haynes 1992:324, 1998). Although the excavator states, "There does not appear 
to be any reason to suspect contamination of the charcoal samples used for the radio- 
carbon dates" (Frison 1996:8), the demonstrated coal contamination of both an under- 
lying stratum and of a radiocarbon sample from the bone bed obviously give such a rea- 
son for suspicion. MiU Iron's actual age, Hke Mesa's, is probably ca. 10,000 yr B.P., judg- 
ing from the late Paleoindian technology and typology of the points (Bradley & Frison 
1996: 66-67). Its age could readily be resolved by AMS dating of individual, botanically- 
identified specimens of burnt plants and the purified amino acids of several individual 
bison bones or teeth. 

The ca. 16,000-year-old Meadowcroft rockshelter site (Adovasio et al. 1983), also rich 
in coal fragments on the surface and in buried strata, also was dated by aggregates of 
stratigraphically-dispersed, taxonomically-unidentified carbon. That contamination has 
occurred is indicated by the fact that the carbon solutes were more abundant than, and 
dated older than, the carbon solids (Haynes 1991). Its lithics were rare, and its paleoen- 
vironmental materials inconsistent with its radiocarbon age. Meadowcroft dating prob- 
lems could be resolved by AMS dating of several individual pieces of botanically-identi- 



I'lcuRi'; 5. 









Inconsistent radiocarbon dates from Mesa, Alaska (Kunz & Reanier 1994, used with per- 

SE Material 

80 hearth charcoal; 
tools associated 

70 hearth charcoal; 
tools associated 

280 hearth charcoal; 
tools associated 


Sample ID 
Beta-55286 Split sample this one was run 

w/57430; does not overlap at 2 

standard deviations 
Beta-57430 Split sample this one v\/as run 

w/55286; does not overlap at 2 

standard deviations 
Beta-50429 Split sample this one was run 

conventionally, 50430; does 

not overlap at 2 standard 


80 hearth charcoal; Beta-55283 
tools associated 


85 hearth charcoal; Beta-50428 
tools associated 

Saddle 10,070 

60 hearth charcoal; Beta-69898 Not in Science 
flakes associated 


70 hearth charcoal; Beta-52606 
tools associated 


90 hearth charcoal; Beta-69900 Not in Science 
flakes associated 


80 hearth charcoal; Beta-55285 
tools associated 



80 hearth charcoal; Beta-55282 
flakes associated 

Saddle 9,945 75 hearth charcoal; Beta-50430 Split sample this one was run 

tools associated AMS, 50429; does not overlap 

at 2 standard deviations 

B 9,930 80 hearth charcoal; Beta-55284 

tools associated 

B 9,900 80 hearth charcoal; Beta-69899 Not in Science 

tools associated 

B 9,900 70 hearth charcoal; Beta-57429 

flakes associated 



80 soil charcoal; Beta-36805 

flakes associated 



fied charcoal for each level and by assessment of the potential old-carbon effect by dates 
directly on the coal from the surface of the site and from its vicinit)'. 

Other claimed pre-Clovis sites with bifaces in the southeastern U.S., such as Cactus 
Hill (McAvoy & McAvoy 1997), also lack consistent series of Clovis and pre-Clovis dates 
on taxonomically-identified cultural carbon securely associated with well illustrated and 
quantified artifacts from documented stratigraphic contexts.^ The "pre-Clovis" and 
"Clovis" ages from Cactus Hill were a tiny minorit)' of the dates from their levels, in 
which the majorit}^ were early Holocene dates. The "Clovis" points were actually not flut- 
ed Clovis points, but vaguely shaped parallel-sided points that could easily be late 
Paleoindian or Archaic, and the "pre-Clovis" lithic samples were too small for valid sta- 
tistical separation from the "Clovis" Uthics and insignificantly distant stratigraphically 
(McAvoy & McAvoy 1997:169-170, table 6.3). By the dating standards, such culmres are 
not currently credible contenders for Paleoindian ancestors. 

A. Neu^ Paleoindian A.ncestor: Nenana, A-laska 

A new candidate for Paleoindian ancestor in the New World is found in the Nenana 
culture of the Nenana and Tenana River basins in Alaska. Given the redating of Clovis, 
Nenana is now the oldest well-documented American culture with secure radiocarbon 
ages (Goebel et al. 1991; Hoffecker et al. 1993, 1996; Yesner 1996, n.d.; West 
1996:293-374). The culture fulfills all the criteria of reliability. It has consistent series of 
dates run on several different biologically-identified materials. The materials had been 
rigorously cleaned and have small error bars. There is secure association of the dates 
with undoubted cultural materials; all but one site are free of old-carbon traces; and the 
patterning of dates and cultural materials is replicated at several sites. 

The Nenana culture is known from seven archaeological sites of which four have 
produced a consistent series of 22 radiocarbon dates ranging from about 11,800 to 
10,700 yr B.P. (see Figure 6 for the series from Broken Mammoth). The dates are on indi- 
vidual specimens of several different materials including taxonomically-identified biota: 
charcoal, willow charcoal, large mammal bone. Wapiti bone, swan bone, as well as asso- 
ciated soil carbon. Only one site, Dry Creek, produced evidence of old carbon, in this 
case, lignite (Goebel et al 1991, figure 3, left; Hoffecker et al. 1996). Unlike the Clovis 
people, Nenana people were not specialized big-game hunters. Subsistence adaptations 
documented at several sites include broad-spectrum collecting of diverse resources: 
mostiy medium-to-small game, such as Dall sheep, Wapiti, and especially wetland birds, 
fish, molluscs, and plants from a diverse habitat mosaic of forest, floodplain, and plains. 
Specimens of now-extinct megafaunal prey are very rare, and the only mammoth iden- 





2040 ± 65 

2815 ± 180 

9690 ± 960 
10.270 ±110 
10,790 ± 230 

10,290 ± 70 


Figure 6. Nenana culture, Alaska. Top: Stratigraphy and radiocarbon dates at Broken Mammoth 
site (West 1996:314—315, figure 6-A, table 6-4, used with permission). Bottom: Early Nenana artifacts 
from Dry Creek and Walker Road, Alaska (West 1996:349, 361, figures 7-8, 7-14, used with permis- 



tified gave significantly earlier dates than the other biota, indicating use of fossil ivory. 
Nenana bifacial lithics include common small, triangular bifaces, probably used as knives, 
but not parallel-sided, fluted bifaces on the Clovis model. Its other tools tend to overlap 
substantially in form and manufacture the other early Paleoindian complexes in North 
America (Goebel et al. 1991). 

As a broad-spectrum foraging culture of mixed habitat, Nenana presaged the char- 
acteristics of the newly discovered Clovis contemporaries in South America. 

Paleoindians in l^atin yimerica 
Clovis and Pre-Clovis theories fail south of the border 

Neither Clovis-like, big-game hunting, fluted-point cultures nor pre-Clovis forager 
cultures have been demonstrated indisputably in Mesoamerica or South America. As we 
shall see, the points originally identified as Clovis either turn out to be Holocene in age 
or, when analysed in terms of lithic manufacture, turn out to lack the fluting and the 
straight-sided shape characteristic of Clovis. And, surprisingly, there are still no indu- 
bitable Latin American cases of megafauna slain by terminal Pleistocene hunters. 
Similarly, neither pre-projectile point or pre-Clovis projectile point cultures have been 
verified in South or Central America. As in North America, the relevant sites and cul- 
tures either are not Pleistocene in age or do not have the claimed cultural characteristics. 
All Pleistocene cultures whose nature has been adequately established and published fail 
to predate ca. 1 1 ,200 years, and all have some bifacial artifacts, not exclusively unifacial 

Instead, diverse broad-spectrum hunting and collecting, regional Paleoindian cultures 
contemporary with Clovis, have been found in diverse environmental contexts, often 
with lithic complexes characterized by triangular, sometimes stemmed bifaces. Such tri- 
angular bifaces and broad-spectrum adaptations had initially been assumed by Clovis the- 
orists to be post-Pleistocene, an analogy with the transition in parts of North America 
from Paleoindian fluted-point big-game hunters to Archaic triangular-point foraging cul- 
tures, described above. However, also as mentioned above, triangular and/or stemmed 
bifaces and broad-spectrum foraging cultures are common Late Pleistocene forms in 
much of the Old World and therefore are forms that would have been known to the ear- 
liest migrants, rather than merely local post-glacial American innovations, as originally 
assumed by Clovis theorists. Furthermore, triangular points are now established by radio- 
metric dating to be of terminal Pleistocene age both in North and South America; thus 
they pre-date the Archaic forms and cannot be their descendants. In addition, the trian- 


ruoslvjj:i, dcjuglas and brown 

^ular forms foiiml in terminal Pleistocene Paleoiiulian sites in Latin America are mor- 
phologicalh' different from the triangLilar points of Archaic, post-Pleistocene Indians in 
North yVmerica. The North American Archaic points at issue are r)'pically side-notched 
for attachment (Anderson et al. 1996:10, figure 1.2; Prison 1991:79-109), while the early 
South American points (illustrated in figure 10) never show this iornx. In this important 
hafting featxire, there is no formal or temporal overlap between the two areas and there- 
fore no reasonable expectation of a historical relationship. 

Incomplete knowledge of the cultural and ecological characteristics of Paleoindians 
and their possible ancestors in the early days of research thus has tended to obscure the 
character and historical significance of South and Central American cultures in the peo- 
pling of the Americas. 

Central America and Northipestern South America 

In Middle America, most sites once thought to represent Clovis occupations were 
either just surface finds of vaguely Paleoindian or early Archaic projectile points or exca- 
vated sites with indeterminate artifacts with few or widely ranging dates that made it 
impossible to determine the nature of cultural remains, their age, and their relationship 
with biota. The artifacts supposed to be Paleoindian either lacked identified biological 
remains and consistent, precise radiocarbon dates on cultural carbon, or lack published 
Uthic tool representations that establish their forms as fluted, parallel-sided "Clovis" 
points, as some assert (Ranere & Cooke 1991; Gruhn & Brvan 1977). There are no 
Pleistocene sites with fluted points, contemporary megafauna, or human skeletons. Some 
examples of the problems at the claimed early sites are teUing. Northern Mexican sites 
classified as Clovis simply amount to finds of indeterminate, unassociated bifaces {e.g., 
Di Peso 1965). In Central Mexico, Tepexpan "Man," a human skeleton once thought 
Paleoindian, produced an age of less than 2000 years when dated directiv (Tavlor 1992, 
table 25.5). At Los Tapiales, Guatemala, a few indeterminate bifaces without identified 
food remains were associated with three disparate Late Pleistocene and Holocene dates 
on botanically-unidentified carbonaceous materials; although the points were described 
as fluted, illustrations show no evidence of fluting (Gruhn & Bryan 1977). 

In the Isthmian area of lower Central America and Northwestern South America, a 
Clovis occupation has long been hypothesized, but neither Pleistocene fluted points nor 
megafaunal kill sites have been documented or dated (Ranere & Cooke 1991; Ardila 
1991; Ardila & Politis 1989). The "Clovis" points from lower Central America (Dillehay 
2000; Lynch 1983; Morrow & Morrow 1999; Ranere & Cooke 1991; Snarskis 1979), are 
not parallel- sided, fluted points and are not dated to the Pleistocene. Some are wide, 



short, concave- sided cutting tools with waists or stems and often ears or deep, basal con- 
cavities and are either undated finds or finds spatially associated with pottery and 
Holocene dates; others are indeterminate paraUel-sided points without Clovis-style flak- 
ing or fluting (Snarskis 1979, 1984; Ranere & Cooke 1991: 49). The flake scars supposed 
to be "flutes" are merely remnants of broad thinning flakes later invaded by pressure- 
trimming. The parallel sided points could aUgn with virtually any Paleoindian or Early 
Archaic points in North America; the waisted shapes have no relation to Clovis forms 
and could only aUgn, if at all, with North American forms that are late Paleoindian or 
Holocene {e.g., Prison 1991:62-101, figures 2.28, 2.33, 2.37; Anderson eta/. 1996:10, fig- 
ure 1.2). Only the strong faith in the reaUty of a hemispheric Clovis horizon and the 
tools' lack of contexmal and chronological data made them seem like evidence for the 
migration when first reported. The few Central American lithic sites with Pleistocene 
dates so far have undiagnostic lithics; triangular stemmed points have turned up in areas 
with palynological evidence of Paleoindian occupation but have not been dated (Ranere 
& Cooke 1991; Pearson in press). 

Across the Istiimus in Colombia and Ecuador, after decades of research there is still 
no faunal or chronological evidence for Pleistocene fluted-point, big-game hunting in 
upland open environments. Colombian projectile points once claimed to be Clovis flut- 
ed points actually now appear to be the laterally- or basally-thinned stems of large, tri- 
angular points (Ardila 1991; Ardila & Politis 1989:10, 13). Ecuadorian lithic industries 
once claimed to be Clovis fluted points also lack parallel- sided fluted points and even the 
earliest of their 23 associated radiocarbon dates are early Holocene (Bell 2000; Mayer- 
Oakes 1986:133-156), not Late Pleistocene. These industries, termed El Inga, are there- 
fore not even Paleoindian. An upland Ecuadorian skull earlier claimed to be pre-Clovis 
on the basis of > 20,000-year-old thermoluminescence (TL) dates on limestone concre- 
tions and amino acid racemization of collagen, produced ages of only 2000 yr B.P by 
radiocarbon dates on collagen (Davies 1978), considered the most reliable of the three 
methods (Taylor 1997). 

Upland Colombian sites at the Sabana de Bogota El Abra rockshelters (Correal 
Urrego et al. 1972; Hurt et al. 1976) were once widely assumed to represent pre-Clovis, 
pre-projectile-point foragers, but some archaeologists now consider the evidence equiv- 
ocal (Ardila 1991:276-278; Ardila & PoUtis 1989:18-22; Roosevelt et al. 1996:383). The 
unifacial industry was an inadequate artifact sample without plot locations, and the 
chronology was based on notliing more than two statistically discordant dates on bio- 
logically-unidentified materials not recorded in secure primary association with the lithics 
(Hurt etal 1976:2, 5, 8, 12, and figure 4). The El Abra biota was a mixture of many mod- 
ern and few extinct species {ibid}). Although another Bogota area site, Tibito, has been 


R( )( )si ■ VI ':i :r, D( )uci .as and brown 

identified as a pre-projectdle-point megafaunai site, it too, has an inadequate lithic sam- 
ple and is dated by a single date on the fauna; and Bogota sites with larger artifact sam- 
ples turned out to have bifacial and unifacial flaking together (Ardila 1991:275-278; 
Correal Urrego 1981; (Correal Urrego & van dcr Hammen 1977; Nieuwenhuis 1998), 
making the existence of a pre-projectile point complex questionable on grounds of sam- 
ple stadstics. 

Points from northern Vene>^uela termed Clovis fluted points (Oliver & yVlexander 
1990: figure 20 and following), are, like the Isthmian points, undated bifaces disdnct 
from Clovis points in shape and manufacture. Some claim a Paleoindian to Archaic cul- 
tural sequence from bipointed El jobo points to Las Casitas stemmed triangular points 
in Venezuela (Barse 1990; Rouse & Cruxent 1963; Dillehay 2000), but neither of those 
point types have been recovered from stratified, dated contexts there, and none were 
associated with subsistence or environmental remains. This "sequence" is a guess with- 
out any empirical confirmation. 

The Venezuelan site of Taima Taima is represented as a Late Pleistocene megafaunai 
kiU site of die El Jobo culture (Bryan et al. 1978a; Dkon 1999:98-99; Ochsenius & 
Gruhn 1979). However, the site does not meet the criteria of reliabiiit}' on several counts 
(Ardila 1991:274-275; Haynes 1974: 378; Lynch 1974:356; Roosevelt et al. 1996:383). It 
is an artesian spring deposit in contact with geological carbon sources: Miocene lime- 
stone, lignite (Ochsenius & Gruhn 1979:54), and possibly petroleum.*^ Although manv 
secondary sources say that this site component has consistent dates, the stratigraphical- 
ly associated radiocarbon dates are highly inconsistent, ranging from ca. 11,000 to > 
41,000 yr B.P. ago (Bryan & Gruhn 1979:53-58). Given the discontinuous stratigraphy 
and inconsistent dates, the association of mastodon bone with so few stone artifacts, 
only one or two biface mid-sections and a flake, is very insecure. Some bones were 
undoubtedly cut and marked by use for anvils (Rouse & Cruxent 1963, plate 4) but 
whether as fresh or fossil bone is an open question. Often said to have associated El Jobo 
bipoints {e.g., Bryan 1978a; Ochsenius & Gruhn 1979:10; DiUehay 2000:128-133; DLxon 
1999:98—99), none has been published from the site. The point midsection(s) from the 
sites lacks both ends (Ardila 1991:275) and so obviously cannot be identified as an El 
jobo bipoint. The bipoints that are often illustrated in discussions of Taima Taima {e.g.., 
Dillehay 2 000:111, figure 5.1) are not from Taima Taima but are undated surface finds 
from elsewhere (Rouse & Cruxent 1963:27—37). 

The only documented, reliably-dated, early preceramic sites in these regions of 
northern South America give evidence for a very different sort of terminal Pleistocene 
occupation from the hypothesized "Clovis" or pre-Clovis horizons. They show the exis- 
tence of a lowland Paleoindian occupation by mixed foragers using bifacial points that 



sometimes are triangular and stemmed. Four stratified sites in tropical forest have pro- 
duced seven terminal Pleistocene charcoal dates between ca. 10,500 and 10,000 yr B.P. 
(Gnecco 1994; Gnecco & Mora 1997; Lopes de Castano 1995; Lopes de Castano & 
Nieuwenhuis n.d.). AU have abundant debitage from biface manufacture and two have 
triangular, stemmed formal bifaces. The food remains identified and dated at two of the 
sites were abundant carbonized plants representing tropical forest foraging on tree and 
palm fruits (Gnecco & Mora 1997). 

Thus, there is still no empirical support from this region for the Clovis culture or for 
pre-projectile point, pre-Clovis, generalized foragers, and the Colombian and Ecuadorian 
highlands still lack securely dated sites with well-defined Paleoindian cultural occupa- 
tions. It seems likely that there were Paleoindian occupations there, but their nature has 
not yet been established. The only northern sites that have a clear association of consis- 
tent dates, abundant artifacts, and identified environmental remains document the pres- 
ence of biface- and uniface-using, broad-spectrum, tropical forest foragers in the termi- 
nal Pleistocene. Such an occupation does not fit the assumptions of the Clovis theory 
about the history of Paleoindian tool industries and ecological adapations. 

The Central A^ndes 

In the Peruvian Andes, on the supposed highland plains route of the Clovis big- 
gamehunters, no finds of fluted points, no megafauna kiU sites, and no pre-projectile 
point forager sites have been verified. 

Pre-Clovis radiocarbon dates on extinct fauna and wood in caves in Ayacucho were 
once claimed to be evidence of pre-Clovis occupations (MacNeish et al. 1970, 1980, 
1981a, 1981b, 1983). However, the only possible artifacts from the sites are amorphous 
flakes of cave rock that other archaeologists do not accept as human-made (Dixon 1999: 
99-101; Lynch 1990; Roosevelt et al. 1996:383). A supposed pre-Clovis site at Amotepe 
on the north coast of Peru (Richardson 1978) was an artifact scatter lying on the surface 
in an area of petroleum seepage. Its crude Uthics could be any age, having no sealed 
stratigraphic association with the only two radiocarbon dates, which are non-overlapping 
dates on marine-shell, a material considered unreliable for dating (Keefer et al. 1998; 
Taylor 1987:49-52). 

Rather than having Clovis-like or pre-Clovis cultures in terminal Pleistocene times, 
Peru appears to have been populated by broad-spectrum foragers with a tradition of tri- 
angular, sometimes stemmed, bifaces. Six highland sites have stratified cultural deposits 
with undoubted artifacts and diverse food remains whose dates fall between ca. 11,000 
and 7,100 yr B.R (Lynch 1980; Lynch et al 1985; MacNeish et ai 1981a, 1981b; Rick 
1980). The earliest Paleoindian points in these sites are triangular, subtriangular, and 



sometimes \;iguel\' stemmed (I'igure 7a). They are nor flLitecl, |iarallel-sided points like 
those of Clovis, and none are associated with extinct megafauna. At the sites in this 
region, narrow bipointed tools are found at the Pleistocene/ Holocene transition, not in 
the early Paeloindian cultures. The food remains vary from broad-spectrum hunting and 
gathering to specialized hunting of guanaco, a small camelid. 

(Contrary to the assumption that ocean resources would only be exploited in the mod- 
ern climate era, excavations on the coast show the existence of widespread exploitation 
of the resources of the sea or coastal land in early Paleoindian times. Two new sites in 
southern Peru reveal an occupation by broad-spectrum, marine-coastal foragers at the 
same time as Clovis (Sandweiss et al. 1998; Keefer et al. 1998). Among their artifacts are 
bifacially flaked, triangular tools of quartz (Figure 7b), perhaps knives. Although some 
late Paleoindian tools were made of obsidian from adjacent highlands, all the foods 
found in the sites were from the sea coast. They included numerous anchovies, other 
fishes, shellfish, and coastal birds, but no extinct megafauna. The > 30 dates on cultural 
charcoal, shell, and faunal bone from the two stratified Paleoindian coastal sites range 
from 11,200 to ca. 9,000 yr B.P ago (see Figure 8 for the early Paleoindian series). 
Elsewhere, later Paleoindian sites of the Paijan culture of the Peruvian north coast have 


Figure 7. Triangular, bipointed, or stemmed bifaces 
from Peru. A. Stemmed and triangular projectile points 
from Guitarrero Cave, Peru (Lynch 1980:180, 184, fig- 
ures 9.2, 9.4, used with permission). B. Quartz point 
from Quebrada Tacahua\-, south coast, Peru (Keefer et 
A al. 1998:1834, figtu-e 3, used with permission). 




^^C yr B.P. 

Cal. yr B.P. 

Lab no. 


Sector 1 

1-2-D, level 3b 

10,274 ± 125 




1970, layer 4 

10,200 ± 140 


[(5), p. 45] 


1-2-B, level 4c 

11,088 ±220 




1-2-D, level 4c 

11,105 ±260 

Sector II 



II-1-D, level lb M 

10,190 ±220 




II-1-C, element ll-5bii 

9,850 ±170 




II-1-D, element ll-5bi 

10,475 ± 125 




ll-l-D, level 2c 

10,700 ± 300 




ll-l-D, level 2c2 

10,600 ± 135 




ll-l-D, level 2c3 

10,560 ± 125 




ll-l-D, level 2c4 

10,725 ±175 




1992, level 3 

10,770 ± 130 




Sector IV-U nit IV- 1-C 

IV-1-C, level 2c 

10,507 ± 125 




Figure 8. Radiocarbon dates from Quebrada Jaguay, south coast, Peru (Sandweiss ei a/. 1998:1830, 
table 1, used with permission). 

consistent radiocarbon dates between ca. 10,500 and 8,500 yr B.P. on human skeletons, 
charcoal, and food remains from broad-spectrum foraging in coastal forests and rivers 
(Chauchat 1988, 1992). Paijan individuals have generalized Asian dentitions and cranial 
morphologies similar to those of other Paleoindians. The Paijan Uthic complex is char- 
acterized by ubiquitous stemmed, triangular points (Figure 9, but no fluted points at aU). 
As in Nenana sites, the few megafaunal bones in Paijan sites are evidence of use of fos- 
sil bone, not of big-game hunting. They date several thousand years earlier than the asso- 
ciated Paijan cultural remains. 

In their cultural-ecological characteristics, the earliest known Paleoindian sites in the 
Central Andes tend to resemble the northwestern North American early Paleoindian cul- 
mre, rather than Clovis culture. No verifiable trace of pre-Clovis or pre-projectile point 
cultures have been found yet. 

The Southern Cone 

In the Southern Cone, the earliest well-established Paleoindian cultures are contem- 
porary with Clovis but, as elsewhere, they lack evidence of specialized big-game hunting 
and fluted point Uthic industries similar to Clovis. As elsewhere, no secure evidence for 
pre-Clovis or pre-projectile point cultures has yet been found. 


R()()si:vi':]:i; douglas and brown 

The site of Monte Verde near rhe I-'acific coast of south central C^hile has been char- 
acterized as a thirteen-thousand-year old, pre-Clovis site with a possible > 33,000-year- 
old occupation, below (Dillehay 1989, 1997). At first the site was described as a pre-pro- 
jectile-point, broad-spectrum forager culture, but at some point a few narrow, bipointed 
projectile points were found. Although accepted by many archaeologists as valid on the 
basis of a small meeting of experts held there, the main questions about the site still are 
unanswered. The site has been questioned for several reasons (Fiedel 1999b; Lynch 1990; 
Roosevelt et al. 1996:383). It was a shallow, near-surface deposit along a boggy stream, 
with complex and discontinuous stratigraphy difficult to interpret in the absence of spe- 
cific scale unit-level section drawings with plotted date samples and artifacts. Streamside 
bogs tend to be catch-alls of materials of different ages, and other bogs in the region 
contained natural deposits of the mired megafauna, logs, and other plants remains, 
according to the researchers (Dillehay 1997:127—134, 174—182, 699). The namre of the 
association of the only four undoubted artifacts, megafauna, and archaeo-botanical spec- 
imens with the radiocarbon dates is unclear, and no debitage from the manufacmre of 
points was recovered, the reverse of the usual situation in which debitage greatly out- 
numbers points. In addition, bipointed bifaces from elsewhere in the southern cone have 
yielded Holocene dates reaching as late as 5,000 years {e.g., Nunez 1992), not Pleistocene 

cm 5 



FiGUR]-. 9. Stemmed projectile points of the Paijan culmre, north coast, Peru. (Chauchat 1988:55, 
figure 2.6, used with permission). 



There are also problems with the dating itself. Two ca. 33,000-year-old dates, although 
treated by the excavators as cultural dates, are characterized in the dating laboratories' 
notation form as infinite dates, which come from carbon too old to retain any 
carbon-14. Possible sources of old-carbon contamination in the site included bitumin on 
tools, a material available in nearby marine petroleum deposits, and detergent and gaso- 
line in polluted water descending the stream from logger settlements upstream. A fur- 
ther chronological problem is that the pollen evidence for habitat at the time of occu- 
pation is in conflict with the insect evidence for the state of the habitat (Dillehay 
1989:94—96), suggesting that materials of different ages have been combined in the site. 
The nature of the occupation could be clarified with further research to date taxonomi- 
cally-identified, archaeobotanical specimens, the handle of a reported hafted tool, and 
the different materials that gave conflicting environmental evidence. Also helpful would 
be a further survey to try to locate and date other comparable sites and create a local 
sequence against which to evaluate Monte Verde. 

Recent review of its data by a team of North American archaeologists resulted in a 
peer- reviewed statement of moral support for Monte Verde as a 12,500-year-old pre- 
Clovis site (Meltzer et al. 1997). However, that consensus has recently fallen apart, as the 
second volume of the published site report has circulated, and members of the meeting 
and others have spoken against the validity of the pre-Clovis occupation (Dincauze, per- 
sonal communication; Fiedel 1999b; Haynes 1999). 

Another megafaunal site in Central Chile is at Tagua Tagua, claimed to be a 11,800- 
year-old, pre-Clovis site (Montane 1968). However, its few dates are stratigraphically 
inconsistent, have large error bars, and equivocal cultural association. Nearby is Quereo, 
also a megafaunal site, but it has no secure evidence for human occupation contempo- 
rary with the fauna (Nunez et al. 1994). 

In the Southern Cone in Patagonia and the southern Pampas well-established early 
Paleoindian occupations named after Fell Cave are characterized by lithic industries with 
stemmed points, not parallel- sided, fluted Clovis points (Bird et al. 1988; Ardila & Politis 
1989; Politis 1991; Flegenheimer 1980, 1987). Although some North American archae- 
ologists have called the southern points fluted and have cited them as possible descen- 
dants of Clovis fluted points (Morrow & Morrow 1999), those who have made com- 
prehensive study of both the stone tools and debitage note that the points were not 
reduced from bifacial preforms, as were most Clovis points; that they are convex in the 
proximal lower third, rather than concave; and that they lack Clovis fluting (Bird et al. 
1988; Borrero & McEwan 1997:40, figure 23; Flegenheimer 1987; Miotta 1999; PoUtis 
1991). The subsistence remains associated with fishtail points indicate not big-game 
hunting but mixed hunting focused on smaller game, such as guanaco and local flight- 


R()()Si:vj:]:r, uol c^las and brown 

less birds. 'I'lic rare extinet mamma! liones associated with the cuitLiral ehareoal and arti- 
facts in 1 ell sites have u,i\en discordant (.lafes considerahK' earlier than the cultural dates; 
their condition, with marks oi" predators, suggests a mixture ot natural and cultural 
materials (Ardila & Polids 1989; Borrero 1996; Borrero & McFAvan 1997:34-44; Nami 
1987), an indicadon that megafauna were probably used as tossil bone, nf)t as prey. The 
ten cultural dates from seven early fishtail-point sites in the Southern Cone run between 
11,000-10,300 yr B.P. (Bird et al. 1988; Polids 1991), which is comparable to the dme 
span of the Clovis-Folsom Paleoindian sequence in North America. In the vicinit)' of 
Los Toldos on the Patagonian plateau, archaeologists have uncovered a poorly-known 
preceramic culture characterized by rare triangular projectile points as w'eU as fishtail 
points, and a rich rock-painting tradition (Cardich 1978; Schobinger 1999). Existence of 
an earlier culture with unifacial, edge-tximmed tools has been hypothesized but this cul- 
ture has yet to be systematically verified and dated; a single date of 12,600 ± 600 yr B.P. 
(Borrero & McEwan 1997; Cardich 1978: 296; Cardich et al. 1973) lacks secure cultural 

It is remarkable that the valid Paleoindian cultures established in western South 
America give no evidence for the hypothetical passage of big-game hunters descended 
from Clovis. What they do show is the existence of a diverse group of regional cultures 
of broad-spectrum foragers whose Uthic industries include triangular and/or stemmed 
projectile points. In these characteristics, these cultures are more parallel to the Late 
Pleistocene northeastern Asian coastal, river, and lake cultures and the pre-Clovis 
Nenana culture of Alaska than they are to the Clovis high plains culture with which they 
are contemporary. Both chronology and culture make it less likely that the western South 
American cultures are historically related to their contemporary Clovis, which has a dif- 
ferent Uthic tradition and subsistence orientation, than to the broad-spectrum hunting 
and gathering Nenana culture, which predates them. 

Eastern South A.merica 

In tropical and subtropical eastern South America, as in the Andean zone, fluted- 
point, big-game-hunting cultures have not been verified. As in the other areas of the 
hemisphere, eastern South America has provided no vaUd examples of pre-Clovis, pre- 
projectile-point cultures, either. Instead, excavations by several independent researchers 
have established the existence of an earlv Paleoindian occupation bv rock-painting, 
broad-spectrum foragers who made triangular and often stemmed projectile-points, as 
well as other tools. 



Southeastern Brazil and Uruguay 

Numerous pre-Clovis sites have been claimed for eastern Brazil, but none meet the 
standards for reliability. Uranium-Thorium dates > 100,000 yr B.P. at Toca da Esperanca 
were run on megafaunal bones found with crudely flaked rocks in Central, but the radio- 
carbon dates on charcoal with the bones were exclusively middle Holocene (Beltrao 
1993; de Lumley et al. 1987). The Middle Pleistocene dates, thus, relate to geological sub- 
strata containing reworked faunal bones, not to the human materials, which appear to be 
late Prehistoric quarry debris. At another site, Alice Boer, there were pre-Clovis dates 
from subsoil, but the TL dates from burned stone artifacts were late Paleoindian (Beltrao 
et al. 1986). At the painted rockshelter site of Pedra Furada (Split Rock) in Piaui, the 
flaked stones, red-stained rock clusters, and charcoal radiocarbon-dated as early as 30,000 
yr B.P. (Guidon & Delibrias 1986), appear to be products of natural processes (Lynch 
1990; Meltzer et al. 1994; Roosevelt et al 1996:374, 383). A fragment of painted cave wall 
found with the charcoal could have fallen from the wall deep into the soft, indistinct sand 
strata. AU the usual signs of prehistoric human habitadon, such as fragments of burned 
bones and carbonized food plants, were absent from the early strata. 

Such claims of exaggerated antiquit)^ have tended to obscure the existence of as 
many as 12 southeastern and northeastern Brazilian Paleoindian sites with more than 30 
valid terminal Pleistocene dates between ca. 11,500 and 8,500 yr B.P. (Gruhn 1991). The 
sites are characterized by the presence of a wide range of bifacial and unifacial stone tool 
types, abundant debitage, pigment, and numerous undoubted food remains from broad- 
spectrum foraging. The radiocarbon dates from Perna shelter, near Pedra Furada, include 
a date of 10,500 yr B.P. from a hearth in undisturbed strata covering rock art panels 
(Pessis 1999). Sites in Goias (Schmitz 1987), in the Lagoa Santa region of Minas Gerais 
(Prous 1980-81, 1986a, 1986b, 1991, 1995, 1999), in Mato Grosso just south of the 
Amazon (Miller 1987; Vialou & Vialou 1994), and far southern Brazil (Miller 1987), have 
produced human skeletons or carbonized plant directly dated between about 11,000 and 
8,500 yr B.P. At several sites, human skeletons are associated with burned fragments of 
diverse foods, pigment, and a wide range of lithic artifacts, including triangular, stemmed 
points. The species identified among the foods include modern tropical forest fruits, 
deer, varied small game, fish, and molluscs. These biological remains show that Late 
Pleistocene habitats were considerably moister than the current savanna woodlands and 
dry forests, whose vegetation has been depleted by centuries of timbering, ranching, and 
plantation agriculmre since the European conquest. The few remains of megafauna at 
some sites are pieces reworked from limestone geological substrata. No use of megafau- 
na as human prey has been verified. 



'rhcrc arc ajijiroxinialcK' fwc I'-asl Bra/.ilian skeletons wiili tliixci icrminal Pleistocene 
ratliocarhoii dates or sealed, well-dated stratis^raphic contexts. They have small, hiu;h 
faces, narrow noses, long, narrow heads, and Sundadont dentition. As mentioned above, 
the particular female cranium from l>apa Vermelha site, claimed by Neves to be a 
Negroid individual 1 1,500 yr B.R old, has not been directU' dated and nf)t compared den- 
tally with other human remains worldwide. It remains, then, a specimen of unclear sig- 
nificance. So far, then, all documented early east Brazilian human skeletons are general- 
ly similar morphologically to early skeletal collections from other parts of the Americas 
(Neves & PucciarelH 1991; Neves d ,i/. 1996). 

T/je "Loiper Ama-i^on 

Triangular points with stemmed or concave bases had also turned up on the surface 
at sites in the Lower Amazon in Brazil and in the Guianas (Figure 10). The Amazonian 
points had been guess-dated to between 10,000 to 7,000 years, on the assumption that 
they had to be Archaic, based on their shape (Boomert 1980; Evans & Meggers 1961; 
Simoes 1976). However, no triangular, finely-flaked projectile points had been dated to 
the Holocene in the Lower Amazon, and no known Holocene cultures in the area had 
such projectile points or debitage (Roosevelt 1999, 2000b; Roosevelt et al. 1991; 
Roosevelt et al. 1996, 1997). In addition, dates from sites in central and southeastern 
Brazil placed such points as early as the terminal Pleistocene, as mentioned in the previ- 
ous section. 

Some researchers had also assumed that lowland South America w^ould have been a 
savanna with small forest refuges under Late Glacial climates, and they interpreted cur- 
rent savanna patches along the Lower Amazon and Guianas coasts as relicts of glacial 
savannas once connected to the cerrado of central and soouthern Brazil and the llanos 
of Venezuela (Lynch 1983; Sales Barbosa 1992; van der Hammen & Absy 1994). Their 
assumptions, however, were not based on empirical data from dated terminal Pleistocene 
contexts and did not acknowledge the strong association of the "savanna" patches with 
intense recent deforestation in the locations. In fact, all pollen profiles from intact, strat- 
ified sediments radiometricallv dated to Late Pleistocene age in the Amazon have spec- 
tra and stable isotope ratios t}^ical of closed-canopv tropical rainforests, not of savan- 
nas (Absy 1979; van der Hammen & Absy 1994; Athens & Ward 1999; Colinvaux et al. 
1994; Haberle & Maslin 1999; Roosevelt 2000c). These prehistoric spectra have small-to- 
moderate amounts of grass poUen, abundant tropical forest pollen, and stable carbon 
isotopes from -20 to -37 per mil. They showed neither the t}^ical savanna pollen spec- 
tra, which have greater than 99% grass pollen, nor savanna carbon isotope ratios, which 



so- «^ 

^ Site in 
Monte Alegre area 





50O Brasilia ,^^ 

Figure 10. Geological map of Amazon and Middle-Lower Amazon projectile point find surface 
sites (redrawn and adapted from Sioli 1984:53, used with permission). Projectile points from left: Upper 
Rio Negro, 15cm; Ireng River (Evans & Meggers 1961: plate 8, d, used with permission); Monte Alegre, 
ca. 8cm, (collection of Victor Fuchs, photograph by A.C. Roosevelt), Tapajos River, 4.7cm, Tapajos 
River, 6cm, (Museu Paraense EmUio, Goeldi 1986:115, 117, used with permission and Roosevelt et al 
1996:373, figure 1), Santarem, 12cm (Roosevelt 1989:44, figure 8a), and Curua River, ca. 20cm (collec- 
tion of Waldemar Caitano, photograph courtesy of Mauro Barreto). 

range from -9 to -15 per mil (Bonnefille 1995; Tieszen 1991; van der Merwe & Medina 
1991). Thus, the Amazonian areas where the triangular points were found probably 
already had tropical rainforest habitats in the Late Pleistocene. 

Monk Alegre 

Among the surface-find-places of triangular, stemmed projectile points in the Lower 
Amazon was the Monte Alegre, Para at the mouth of the Tapajos River opposite 
Santarem (Figure 11), about halfway between Belem and Manaus. Known for the rock 


R( )( )SI Al J :i , D( )UG1.AS AND BROWN 

Figure 11. Views of Monte Alegre, Para, Brazil general vegetation, a sacure palm, a stream, and a 
lake. Upper left photo courtesy of Nigel Smith, Universit}' of Florida. 

art in its sandstone hills (Pereira 1992), Monte Alegre has numerous large paintings of 
geometric images, animals, and stick-figure humans fingerpainted in red, yellow, and 
brown (Figure 12). The presence of colored handprints of people of all ages, including 
infants, in the rock art panels in the region indicated that family groups were involved. 
In some painted panels, abundant drops of liquid pigment had splashed below the paint- 
ings. Although published in the nineteenth cenmry by Alfred Russell Wallace (1889) and 
others, the painted caves and shelters had not been excavated for evidence of human 



Caverna da Pedra Pintada 

To search for stratified deposits, the research team surveyed and GPS-mapped 21 
caves, rock art sites, rock shelters, and open sites. Auger tests at the sites pinpointed 
Caverna da Pedra Pintada (Cavern of the Painted Rock) as the most suitable habitation 
site (Figure 13), near springs in forest overlooking lakes in the Amazon floodplain. Its 
walls and ceiling are covered with images generally similar to others in the region. Twelve 
auger cores placed inside and outside the cave sampled the stratigraphy, then 11 con- 
tiguous meter excavation units were placed on the north side of the cave. The contigu- 
ous units were placed in intact midden deposits directly below rock paintings to find 
well-preserved biota and check the stratigraphic levels of any spilled paint. Excavation 
procedures were designed to address the epistemological problems that had arisen in the 
interpretation of other Paleoindian sites: hand-excavation by strata and features, record- 
ing of finds in stratigraphic context by photographs, plans, and section drawings, and 
extensive dating of associated biota, sediment, and artifacts. 

The excavations encountered ca. 2 m of prehistoric strata lying upon sterile yellow 
sand and bedrock (Figure 14). The Pleistocene preceramic deposit was a ca. 30-cm- thick 
black sand midden near the base of the stratigraphy. This black sand was overlain by tan 
sterile sand, above which were several dark gray, Holocene, ceramic-age deposits dating 
between ca. 7,500 and 1450 yr B.P. In the preceramic deposit were two main strata, 16 
and 17, each subdivided into layers. Excavating the sediment in 1 meter squares with fine 
tools, we uncovered shallow hearths and lenses and abundant artifacts, food remains, and 
pigment. All the soil excavated was fine-screened or floated with 1.5 mm mesh and fil- 

FiGL'Ki. 12. Rock paintings aiul spilled paint, Serra da Lua, Monte Alegre, Para, Brazil. 




Cave Wall 

Drip Channel 
Drip Line 


Contour Line 


Site Datum 




Auger Hole 


Excavation Unit 

• A 

Paint San^ples 

T ^ 3 4 


Contour interval 10 centimeters 

Elevations in meters above sea level 


Figure 13. Caverna da Pedra Pintada. A. View of cave. B. Paintings on north wall of cave. C Map 
of excavations, auger tests, and paint samples. 




10,250 ±70 9274 
10,350 + 709274 
10,390*70 9274 
10,410 + 70 9274 
10,470 ±70 9274 


10,275 ±245 8345 
1 0,275 ±285 8345 
10,305 ±275 8344 
10,392± 78 8314A 
10,450± 608231E 
10,560± 608231E 


10,875 ± 295 8314B 

10,905 + 295 83148 


11,I45±135 8314B 

Figure 14. Caverna da Pedra Pintada stratigraphy. A. During excavation. B. Drawing with radio- 
carbon dates. (Original drawing, Roosevelt et al. 1996:375, figure 5.) 


R()( )si:vi:i:i, ix )Uglas and brown 

More than 3(),()()() lirliic 'artifacts were rcc(n'crcd from the Palcoindian layers 
(Roosevelt I't cil. 1996, table 1, figure 6). In contrast, the ceramic-age layers held only a 
few hundred. The percentages of lithic raw materials, all from sources outside the cave, 
changed significandy through the five stratigraphically-disdnct, phases of occupation 
(Roosevelt el al. 1996:376—377, table 1). Chalcedon\' predominated in the upper prece- 
ramic deposit, reaching 90% and higher in the highest stratum, 16 A, but quartz crystal 
predominated in the lower part, reaching over 70% in the earliest stramm, 17C. The 24 
formal flaked lithic tools included fragments of triangular projectile points, sometimes 
stemmed, slug-shaped unifacial scrapers; a graver; blade-like flakes; and others (Figure 
15). The bifaces, and flakes from making them, were found from the earliest to the lat- 
est Paleoindian levels. There were abundant pigment chunks and paint drops on the 
stone artifacts, primarily in the earlier preceramic levels, except for one pigment chunk 
found at the surface. The excavated pigment, composed of iron oxides from laterite 
(Figure 16), had chemical composition and iron-titanium ratios comparable to the paint 
on the cave walls, above. Thus, at least some of the cave paintings were probably made 
during the early preceramic occupation. 

The preceramic culmral deposit was rich in wood charcoal, burned fruit and legume 
pits and seeds, and fragments of burnt bone and shell, t)qDes of remains that were com- 
pletely absent from the sterile subsoil below the archaeological deposit. Fift}'-six radio- 
carbon dates were run on the carbonized plant remains by gas counting method or mass- 
spectrometry at three different laboratories. To increase accuracy, the date samples were 
run on localized, plotted, and individual botanicals identified as w^ood charcoal or car- 
bonized pits of taxonomically identified trees, not on aggregate, unidentified carbon 
samples combined from dispersed stratigraphic contexts. Fruits have short life-spans 
compared to trees, so dates on fruit pits permit narrower time-spans to be measured. To 
test for possible carbon contamination and old wood, dates were run on both the solids 
and solutes of many carbonized plants, and the results were statistically identical, regard- 
less of material, consistent with a lack of contamination. To check the contemporaneity- 
of the cultural remains found with the biological remains, 10 burned lithic artifacts were 
dated by thermoluminescence (TL) dates, and three sediment samples were dated by 
optically stimulated luminescence (OSL) (Nlichab et al. 1998). To "blind" the dating 
experiment, the TL lab team was not informed of the radiocarbon OSL results in 
advance of dating the stone tools. 

The preceramic radiocarbon date series runs bet^veen ca. 11,200 and 10,000 yr B.P. 
(Figure 17) (Roosevelt et al. 1996, table 3), placing the preceramic occupation in terminal 
Pleistocene times, exactiy contemporary with the North American Paleoindian sequence 
from Clovis through Plainview (Figure 18 and Appendix 1) (Tajlor et al. 1996). Statistical 



<^ TS^j^ ^ ~" • • - ; — Globular fibro-radial chalcedony 

- Cryplocrystalline chalcedony 

- Hematite plasma in flux form 


of quartz crystals 
Goethite plasma 
Red hematite plasma 

Cryptocrystalline chalcedony 
Globular fibro-radial chalcedony 

Aggregates of quartz crystals 

Partly weathered mica 

Hematite after weathered carbonate 


Microcrystalline matrix 

Unweathered carbonate 

Microcrystalline matrix 






Initial A Initial 

Initial B Initial B/Early Early Eariy/Middle Middle Middle/Lale Late 
{11=753) (n=52) {n=6270) {n=894) (n=e264) (n=l50) {n=13l94) 

Carbonate (Dolomite) 
Microcrystalline matrix 


W..M d 

Figure 15. Lithic artifacts from Caverna da Pedra Pintada. A. Thin sections of chalcedony vari- 
eties. B. Histograms of raw material frequencies. C. Chalcedony debitage during excavadon. D. 
Chalcedony point during excavation. E. Quartz crystal point during excavadon. F. Drawings of lithic 
artifacts. By Ruth Sliva of Desert Research. 





Goethilc plasma 


Dark red hematite plasma 



Red cryplocrystaliinc hematite 


Goethite plasma 



FiCiLRi-: 16. Pigment samples from Caverna da Pedra Pintada. A. Thin sections. B. Specimen 1. 
Microphotograph of paint from concentric cross design on north wall, (Fe/Ti ratio 8.52). C. Specimen 
2. Red pigment rock, Excavation 5, Level 13, object 4, prov. 8344, plan 14, (Fe/Ti rado 10.56). D. 
Specimen 3. Red paint from drops on quartz crystal debitage flake. Excavation 6, level 9, object 37, plan 
11, prov. 8344, (Fe/Ti ratio 8.41). Analysis by W. Barnett, then of the American Museum of Natural 
History and now of the Field Museum of Natural History. 



tests established that the five phases of tliis Paleoindian occupation had weighted-mean 
radiocarbon ages significantly different from each other, in correct stratigraphic order 
(Roosevelt et al. 1997). The earliest phase, Initial A (in Stratum 17C) had eight dates: four 
conventional radiocarbon dates on palm endocarps and four luminescence dates on sed- 
iment and lithic artifacts. The palm seeds had been neady cracked open for their edible 
kernels and burned, a pattern of usage that Amazonian zoologists recognize as distinc- 
tively human, not faunal. The Initial A dates' one-sigma error bars ranged between ±135 
and ±310 years. These four dates were statisticallv the same age and had a weighted 
average of 11,075 ± 106 yr B.P. in radiocarbon years. The intercept of the mean in inter- 
calibrated calendar years is ca. 13,088 yr B.P.. From the same Initial A context, the three 

TL and OSL weighted average; outlier dropped 


Initial A radiocarbon dates 
Midpoint of average (1 3,088 cal) I 

13,143 cai ncnnn 

13,135 cal I |.t; v-'->'^ I 
12,950 cal □ [larZL.Z lJ'^ Z: 
12,920 cal Q " LULZl □ I 



Initial B radiocarbon dates 
Midpoint of average (12,552 cal) I 

12,172 cal I L ^"U 

12.745 cal LaiajH 

12,498 cal 

12,531 cal 

12,333 cal 

12,241 cal 

12,016 cal 

12,052 cal 




12000BC 10000BC 8000BC 

Calendar date 

Figure 17. Graph of initial calibrated radiocarbon, TL, and OSL dates from Caverna da Pedra 



Pedra Pintada Initial A dates 

Clovis dates 


Pedra Pintada Initial A dates 

Radiocarbon Dates 

GX17414 10875±295 
GX 17407 10905±295 
GX17406 11110±310 
GX17413 11145±135 

Luminescence Dates (no calibration) 

12536 BP ±4125 
15330 BP ±900 
13106 BP ±1628 
12491 BP ±1409 

I Clovis radiocarbon dates 

6000 BC 14000 BC 12000 BC 10000 BC 8000 BC 6000 BC Blackwater Draw 1 1 300±240 

Lehner 10950±40 
Calendar date A Murray Springs 10880±50 

Dent 10750±40 
Domebo original 11210±380 
Domebo Stafford 1 1 040±250 
Domebo wood 10930±60 
Lange/Ferguson carbon 11140±140 
Lange/Ferguson bone 10730±530 
Anzick 10680±50 
UP Mammoth 1 1280±350 
Aubrey 1 1 570±70 d 

Figl;re 18. A. Graph of Initial calibrated radiocarbon, TL, and OSL dates from Caverna da Pedra 
Pintada compared with Clovis calibrated dates. B. Dates used to produce calibrated distribution compari- 
son in Figure 18. 

TL dates on burned lithics and one OSL date on associated sediment gave a weighted 
average of 13,180 + 509 calendar years. The two sets of Initial A dates on biota and arti- 
facts thus are statistically the same age as each other and significantiv earlier than the 
weighted average of the radiocarbon dates for the subsequent preceramic occupation 
phase, Initial B of stratum 17B, overlying 17C. For the eleven Initial B radiocarbon dates, 
the weighted average is 10,420 + 23 and the mean calibrated value ca. 12,522 vr B.P. The 
Initial A values are statistically the same as those for the Clovis culture; and the Initial B 
values are comparable to those of the Folsom culture. The preceramic foragers, there- 
fore, arrived at the cave no later than Clovis people arrived in the southwestern United 
States, five thousand miles away to the north, and developed their own multiphase cul- 
mral sequence that ended about 10,000 yr B.P. Some Clovis archaeologists have object- 
ed to beginning the Monte Alegre culture at ca. 11,200 yr B.P., the earliest age of the 
Initial A phase, on the grounds that the dates are few, could be on nattirally occurring 
objects, and have too large errors to be distinct from those of Initial B; thev begin the 
sequence instead at ca. 10,500, the average date of Initial B, leaving 1000 years time for 
Clovis people to have descended the hemisphere to the Amazon after arrival in North 
America (Fiedel 1996; Haynes 1997; Reanier 1997). However, their argument implies that 
Clovis sites date to 11,500 vr B.P, have more dated radiocarbon samples, dates with 



smaller errors, and dates with more secure cultural context than those of Monte Alegre. 
But as we saw in the review of Clovis dates, 11,500 yr B.P. is neither the beginning age 
nor the average for Clovis, which are 11,200 and 10,900 yr B.P., respectively. In addition, 
the few Clovis sites with average dates of 11,000 yr B.P. or earlier have fewer site dates 
and much larger date errors (350 to 600 yr) than Pedra Pintada Initial A, whose largest 
error is only 310. Only two of these early Clovis dates are biologically identified, from 
cultural context, and contamination-free, unUke the four Initial A Pedra Pintada dates. In 
addition, the Clovis date series overlaps substantially with that of Folsom, just as Initial 
A dates overlap Initial B dates. Thus, eliminating the Initial A dates does not make the 
Monte Alegre culture younger than Clovis, for the dates they cite for Clovis would be 
eliminated by the same standards that they apply to Monte Alegre. When compared sta- 
tistically by the same criteria, the dates of the Clovis horizon have a nearly identical age 
distribution as those from Pedra Pintada. 

Finally, the h)^othesis that Monte Alegre had to have been savanna in the Late 
Pleistocene was not upheld by the dated materials from Pedra Pintada. An ancient envi- 
ronment of tropical rainforest, lakes, and rivers for the Monte Alegre culture was con- 
firmed by the taxonomy and carbon isotopes of the dated biological remains. Also, 




J 5 mm 


Figure 19. Paleoindian biological remains from Caverna da Pedra Pintada. A. Fragmentar)' sphe- 
notic bone of large fish, interior and exterior. Prov. 9243. B. Carbonized legume seeds, Hymeiieaea c.f. 
pamfolia. Prov. 9290, 9272. C. OtoUth from large fish. Prov. 8347. 



luiiiK'fous klcntitlcd taxa (I'igurc 19), incluclinu palm nuts, legume seeds, brazil nuts, 
diverse fruit pits, small and large fish bones, shellfish, turtles, tortoises, lizards, medium- 
sized rodents, and birds document broad-spectrum foraging of forest products, fish, and 
small game. Only three fragmentary bones among the several hundred preceramic spec- 
imens recovered could have come from larger fauna, of ca. 65 kg weight. The 56 stable 
carbon isotopes of the carbonized plant remains averaged ca. —29 per mil and ranged 
from —24 to —37 per mil (normalized to wood) (Appendix 1), values t)'pical of closed 
canopy tropical rainforests, not of savannas or even of open forests. Thus, these early 
Amazonian Clovis contemporaries were tropical rainforest foragers, not specialized 
savanna big-game hunters. 


The new archaeological data reviewed in this article indicate that the Clovis migration 
theory, which was framed when few areas but the North American high plains were well- 
known, is no longer viable. It is a strong blow to proponents of Clovis as the ancestral 
culture that no trace of terrestrial big-game-hunting, fluted-point-using, Paleoindian 
people have been verified along the proposed interior, upland plains route through 
Central and western South America. However, the accrued evidence does not replace 
Clovis with a much-earlier ancestor, for all the claimed pre-Clovis cultures other than 
Nenana in Alaska have fatal flaws under the standards of reliability; Rather, we see evi- 
dence of a near instantaneous radiation of people throughout the hemisphere soon after 
12,000 years ago. The initial complexes that have come to light seem to be the first estab- 
lished regional adaptations, not the initial migrants to each region, so we still have not 
found the remains of each regional cultures' founders. Furthermore, the initial com- 
plexes have been found in just a few regions: the high plains and far north of North 
America and the west coast, eastern lowlands, and far south of South America. Large 
areas of the Americas thus remain relatively unknown archaeologicallv for the relevant 
periods, so research is needed to trace in detail the movements of the first peoples. Were 
there several source areas and source populations in the Old World and did people return 
there after establishing themselves in the New World? Did several groups with different 
tool cultures and ecological adaptations come in at the same time and take different 
routes through the hemisphere or did one group expand out and diversify in the process? 
We especiallv need to know more about the culture and Iifesr\'le of the earliest peoples 
and cultures of North America. Other than Nenana in Alaska, the Clovis high plains is 
the only well-known area, and even this area could benefit from more precise dating and 
research on subsistence and artifact cult\ire. And Central America, which must have been 



the major conduit for the groups that populated South America, still has no identified 
initial culmres. To complete the earliest hemispheric culture history, it is a priority to 
make a concerted effort to indentify relevant Quaternary geological featares that could 
be surveyed for initial human sites. 

At this stage, however, this much seems clear: the earliest American hunter-gatherers 
were not confined to cool, open, upland habitats nor to the life of specialized big-game 
hunting characteristic of Clovis. Their choice of habitat and subsistence mode in the dif- 
ferent areas that they reached was eclectic and flexible. In the relatively plant-poor far 
north and far south of the hemisphere they focused on hunting of diverse game, sup- 
plemented by gathering. In the arid high plains of North America, they specialized in the 
big-game-hunting lifeway named after Clovis. In the equatorial zone, some settled the sea 
coast and lived mainly on fish, shellfish, and birds, and some ventured into the highlands 
for broad-spectrum hunting and gathering. In the eastern equatorial lowlands, in con- 
trast, they lived by forest collecting, fishing, shellfishing, and the hunting of small game. 

The emerging, though still incomplete, picture of the peopUng of the Americas has 
implications for theories about the origins and development of the human species. The 
New World case suggests that, contrary to the assumptions of Optimal Foraging Theory, 
the Upper Paleolithic big-game-hunting lifeway was not a primeval adaptation (Beaton 
1991), but rather a late and specialized local adaptation by people coming into tem_perate 
habitats from the tropical and subtropical zones of Africa. Revisions of our under- 
standing of the history of environments and human ecology in Africa (Bonnefille 1995; 
Roosevelt n.d.c) suggests that our species arose and developed not in adaptation to spe- 
cialized big-game hunting in savannas but rather in generalized foraging along the rivers 
and lakes of humid tropical forests in sub-Saharan Africa. In fact, the African botanical 
evidence suggests that the savanna woodlands of sub-Saharan Africa are a much later 
anthropogenic climax (Lovett 1993), rather than an Ice-Age environmental climax that 
was the hearth of the species. Recent reviews of paleoanthropological data from Africa 
cast doubt on the evidence for specialized big-game hunting during the early PaleoUthic, 
when our genus was emerging (Kdein 1999). Evidently, future research on the Plio- 
Pleistocene history of humans will need to forge new understandings of the relationship 
of ecology and social development in human evolution. Big-game hunting was the core 
activity around which early human social organization was supposed to have been orga- 
nized. If broad-spectrum foraging, not big-game hunting, was the preferred adaptation 
of humans during much of the Pleistocene, then models of early human social interac- 
tion and evolutionary explanations of human namre will need to be revised. 


R()()Si'Vi':i;r, douglas and brown tract 

Viarly in research on the peopling of the Americas, Clovis big-game hnnters of the northern high 
phiiiis seemed sure to be the first people who colonif^ed the Americas. Anthropologists believed that Clovis 
h/inters had rapidly colonized the hemisphere from Alaska to Patagonia in only a thousand years from 
1 2,000 to / 1 ,000 years ago. lljo/ight to have followed the big herd game through their gracing lands 
— the North American high plains and the cool, arid uplands of Central and South America — they 
avoided the sea coasts on both sides of the continent and also the gamepoor, disease-ridden tropical rain- 
forests of eastern Central America and South America. 

Systematic investigations at both old and new Paleoindian sites have changed the picture of the 
migrations and adaptations of the first Americans. Although some researchers had dreamed of more 
ancient, pre-Clovis cultures from 12,000 to 30,000 jears ago, the evidence for them has failed most cri- 
teria of reliability: consistent date series lacking contamination from geological carbon, documented cul- 
tural remains, and comparable finds at several sites. R£-evaluation of evidence also has pushed Clovis 
younger than bad been thought, no earlier than about 1 f 100 to 10,800 yr B.P. Alaska still has the 
earliest solidly dated and investigated cultures in a range from 1 1,800 to 1 1 ,000yr B.P, but these cul- 
tures are broad-spectrum hunter-gatherers not resembling Clovis greatly in tool culture or ecological adap- 
tation. Excavations at new sites and continued work at old ones have changed the picture in South 
America as well, revealing a lack of a Clovis hori^^on. Even supposed Clovis descendants at the tip of 
South America turn out to be the same age as Clovis, and their subsistence ivas small-game, not large- 
game, hunting. New sites reveal the existence of coastal fiishers and tropical rainforest foragers as old as 
Clovis hunters and culturally distinct from them. This new evidence on the coloni^^ation reveals the evo- 
lution of more diverse Ice- Age ecological and cultural adaptations than previously suspected, raising new 
questions for research on human origins. 


^ Dates in our article are given in uncalibrated radiocarbon vears before the present 
(1950) unless otherwise noted. 

- Many researchers attach the word Clovis to any thin, vaguely parallel-sided, undat- 
ed points and then assert that such points must therefore be 11, 500 years old, due to their 
equation with Clovis points {e.g., McAvoy & McAvoy 1997:169, figure 5.67—8; Morrow 
& Morrow 1999; Sassaman 1996:79, figure 4.11). It also is common to use the word 
"fluted" for points even when there are no basal channel flakes, but only broad thinning- 
flake scars invaded by pressure edge-trimming {e.g.. Prison 1991:70, figure 2.34; Morrow 
& Morrow 1999; Borrero & McEwan 1997:40, figure 23). The nature of the Uthics in val- 
idated Clovis sites also has been confused in the literature by the characterization of very 



finely flaked points from undated caches to tlie Clovis culture on the assumption that the 
finest points have to be Clovis (Prison & Bradley 1999). In many cases, however, the 
undated points in question actually show flaking customs found in late Paleoindian com- 
plexes {e.g.. Prison & Stanford 1982), not in documented and dated Clovis assemblages 
(Hester et al. 1982; Leonhardy 1966). Such usages are problematic at multiple levels. A 
point the age of Clovis would be between 11,200 to 10,800, not 11,500 yr B.P. 
Purthermore, parallel-sided points continue long after the time of Clovis in many North 
American regions (Anderson et al. 1996:10, figure 1.2; Prison 1991; Morse et al. 1996), so 
an undated point of this shape could be as recent as 8000 years ago. Points with very fine 
parallel, overshot thinning are found in many Paleoindian and Early Archaic cultures. 
Attributing such points to Clovis has created a Clovis horizon without a secure chrono- 
logical basis. 

^ Most researchers use Paleoindian as a chronological term referring to the peoples 
and cultures who initially colonized the New World in the terminal Pleistocene, before 
ca. 10,000 years ago. Some others use the term only for Clovis big-game, fluted-point, 
spear-hunters and their presumed descendants, using the term Archaic for broad-spec- 
trum foragers with the triangular points. However, since these two terms were first used 
for a presumed chronological sequence, and since the triangular point foragers have now 
turned out to be contemporary with Clovis, the latter usage is confusing. 

^ Por the period of early Paleoindian occupations, there is as yet no tree-ring calibra- 
tion curve, so the correction factor is estimated from the difference between radiocar- 
bon and uranium /thorium dates on the same coral samples (Taylor et al 1996). Thus, the 
corrected ages for the early dates are not necessarily more accurate than the uncorrect- 
ed radiocarbon ages at this point in dating technology. Intercalibration of the earliest reli- 
able Clovis beginning dates on identified cultural carbon from documented stratigraphic 
contexts is about 13,000 calendar years, and the majority of Clovis dates calibrate at 
about 12,900 or less (see Appendix 1). 

^ The presence of pre-Clovis complexes in northeastern sites is commonly inferred 
in the following manner. A "Clovis" level is defined, due to the presence of parallel- sided 
bifaces so undiagnostic that actually could be any Paleoindian or Archaic phase. Any lith- 
ic find occuring below that level is defined as pre-Clovis. The association of primarily or 
entirely late Paleoindian or Archaic dates with these lithic levels does not deter the inter- 
preters from inferring a pre-Clovis phase {e.g., McAvoy & McAvoy 1997:22-190). 

'^ The Taima Taima site area has been exploited for petroleum, a possible source of 
too-old carbon there; I myself, observed wells pumping there in 1994. 


r()()Si:vi:j;i; douglas and brovc'n 

literature Cited 

Absy, M. L. 1979. Palynological Study of Holocene Sediiiioils in /he Awac^on Bm///. Ph.D. thesis, 
University of Amsterdam, Netherlands. 

Ackerman, R. E. 1996. Bluefish Caves. Pages 505-510 />/ F. H. West, ed., American 
beginnings: 'Vhe Prehistory and Pa/aeoeco/ogy of Henngia. University of (Chicago Press, 
Cliicago, IL. 

Adovasio, J. M.,J. Donahue, K. Cushman, R. C. Carlisle, R. Stuckenrath, J. D. Gunn, & 
W C.Johnson. 1983. Evidence from Meadowcroft Rockshelter. Pages 163-190 in R. 
Shutler, ed., Early Alan in the Neiv World. Sage Publications, Beverly Hills, CA. 

Aikens, C. M., & T. Higuchi. 1982. The Prehistory of Japan. Academic Press, New York, 
NY. 354 pp. 

Aikens, C. Melvin, & T. Akazawa. 1996. The Pleistocene-Holocene transition in japan 
and adjacent northeast Asia: Climate and biotic change, broad-spectrum diet, pottery, 
and sedentism. Pages 216—228 in L. G. Strauss, B. V. Eriksen, J. M. Erlandson, & 
D. R. Yesner, eds., Humans at the End of the Ice Age: The Archaeology of the Pleistocene- 
Holocene Transition. Plenum, New York, NY. 

Aitken, M.J. 1990. Science-Based Dating in Archaeology. Longman, London. 274 pp. 

Alexander, H. L. 1987. Putu: A Fluted Point Site in Alaska. (Department of Archaeology 
Publication 17). Simon Eraser University, Burnaby, Canada. 

Anderson, D. G., L. D. O'Steen, & K. E. Sassaman. 1996. Environmental and chrono- 
logical considerations. Pages 3—57 in D. G. Anderson & K. E. Sassaman, eds.. The 
Paleoindian and Early Archaic Southeast University of Alabama, Tuscaloosa, AL. 

Ardila, G. L 1991. The peopling of northern South America. Pages 261-282 /// R. 

Bonnichsen & K. L. Trunmire, eds., Clovis: Origins and Adaptations. Center for the 
Study of the First Americans, Oregon State University, Corvallis, OR. 

Ardila, G. I., & G. Politis. 1989. Nuevos datos para un viejo problema. Boletin Museo del 
Oro No. 23. Banco de la Republica, Bogota, Columbia. 

Athens, S., & J. V. Ward. 1999. The late Quaternary of the Western Amazon: Climate, 
vegetation and humans. Antiquity 73:287—302. 

Barse, W. 1990. Preceramic occupations in the Orinoco river valley. Science 

Beaton, J. M. 1991. Colonizing continents: Some problems from Australia and the 
Americas. Pages 209—230 in T. D. DiUehay &: D. |. Meltzer, eds.. The First Americans: 
Search and Research. CRC Press, Boca Raton, FL. 

Bednarik, R. G. 1989. Pleistocene settlement of South America. Antiquity 63:101-1 1 1. 

Begley, S., & A. Murr. 1999. The first Americans. Neivsireek April 16:50—57. 



Behrensmeyer, A. K., & A. P. Hill, eds. 1980. Fossils in the Making: Vertebrate Tapbonomj and 

Paleoecology. University of Chicago Press, Chicago, IL. 338 pp. 
Bell, R. E. 2000. A.rchaeological Excavations at the Site of EI Inga, Ecuador. University of 

Oklahoma Sam Noble Oklahoma Museum of Natural History, R. E. Bell 

Monographs in Anthropology No. 1 . 93 pp. 
Beltrao, M. M. C. 1993. A regiao arqueologica de Central, Bahia: A Toca da Esperanca 

(Cave of Hope) — Um sitio arqueologico do pleistocene medio. Paper presented at 

the Piaui Congress, Serra do Capybara, Brazil. 
Beltrao, M. M. C, C. R. Enriquez, J. Danon, E. Zuleta, & G. Poupeau. 1986. 

Thermoluminescence dating of burnt cherts from the AUce Boer site, Brazil. Pages 

203—213 in A. L. Bryan, ed., Neii' Evidence for the Pleistocene Peopling of the A.mericas. 

Center for the Study of Early Man, Orono, ME. 
Binford, L. R. 1968. Post-Pleistocene adaptations. Pages 313-341 in S. R. Binford & L. 

R. Binford, eds., ISlew Perspectives in Archaeology. Aldine, Chicago, IL. 
Bird, J., M. Bird, & J. Hyslop. 1988. Travels and Archaeology in South Chile. University of 

Iowa Press, Iowa City, lA. 246 pp. 
Boldurian, A. T., & |. L Cotter. 1999. Clovis ^visited: New Perspectives on Paleoindian 

Adaptations from Blackwater Draw, Neip Mexico. University Museum, University of 

Pennsylvania, Philadelphia, PA. 145 pp. 
BonnefiUe, R. 1995. A reassessment of the PUo-Pleistocene poUen records of East 

Africa. Pages 299-310 in E. S. Vrba, G. H. Denton, T. C. Partidge, L. H. Burkle, eds., 

Pakoclimate and Evolution with Emphasis on Human Origins. Yale University Press, New 

Haven, CT. 
Bonnichsen, R. 1978. Critical arguments for Pleistocene artifacts from the Old Crow 

Basin, Yukon: A preliminary statement. Pages 102-118 in A. L. Bryan, ed.. Early Man 

in America from a Circum -Pacific Perspective. Occasional Paper No. 1, Department of 

Anthropology, University of Alberta, Edmonton, Canada. 
Boomert, A. 1980. The SipaUwini archaeological complex of Surinam: A summary. Niew 

West-Indische Gids 54:94-107. 
Booth, W. 1999. Early migrants may have come by land and sea. The Eos Angeles Times, 

September 6:A13. 
Borrero, L. A. 1996. The Pleistocene-Holocene transition in southern South America. 

Pages 339-354 in L. G. Straus, B. V. Ericksen, J. M. Erlandson, & D. R. Yesner, eds., 

Humans at the End of the Ice Age. Plenum, New York, NY. 
Borrero, L. A., & C. McEwan. 1997. The peopling of Patagonia: The first human occu- 
pation. Pages 33-45 in C. McEwan, L. A. Borrero, & A. Prieto, eds., Patagona: Natural 

History, Prehistory and Ethnography. Princeton University Press, Princeton, NJ. 


R()()Si:vi;i:r, DOLicii.AS and broxx^ 

Bowman, S. 1990. Rtid/ocdrhoji Da/iiio, Univcrsit)' of (California Press, Berkeley, and British 

Museum, London. 62 pp. 
l^radlew B. /\., & G. C. Frison. 1996. Flaked-stone and worked-bone artifacts from the 

Mill Iron site. Pages 43-70 //; G. C. Frison, ed., llje Mill Iron Site. Universit}- of New- 
Mexico, Albuquerque, NM. 
Brj^an, A. F. 1978. The contribution of ). M. Cruxent to the studv of the Palef^indian 

Problem in the New World. Pages 63—76 in E. Wagner & A. Zucchi, eds., Unidad y 

\ ciriedcid: iinsciyos Antropologicos en Homenajo a Jose M. Cruxent. Ediciones C.E.A. — IVIC, 

Caracas, Venezuela. 
. 1986. Neu-i Evidence for the Pleistocene Peopling of the Americas. Center for the Saidv 

of Early Man, University of Orono, ME. 368 pp. 

. 1991. The fluted-point tradition in the Americas — One of several adaptations 

to Late Pleistocene American environments. Pages 15—33 /// R. Bonnichsen & 

K. L. Trunmire, eds., Clovis: Origins and Adaptations. Center for the Smdv of the first 

Americans. Oregon State Universit}'^, Corvallis, OR. 
Bryan, A. L., ed. 1978. Early Man in America from a Circum-Padfic Perspective. (Occasional 

Papers of the Department of Anthropology, Universit^' of Alberta). Archaeological 

Researchers International, Edmonton, Canada. 327 pp. 
Bryan, A. F., & R. Gruhn. 1979. The radiocarbon dates of Taima-Taima. Pages 53—58 /'// 

C. Ochsenius & R. Gruhn, eds., Taima-Taima: A Tate Pleistocene Paleo-Jndian Kill Site in 

Nofihermost South America. CIPICS/ South American Quaternary Documentation 

Bryan, A. F., M. Casamiquela, |. M. Cruxent, R. Gruhn, & C. Ochsenius. 1978. An El 

Jobo mastodon kill at Taima-Taima, Venezuela. Science 200:1275—1277. 
Cardich, A. 1978. Recent excavations at Fauricocha (Central Andes) and Fos Toldos 

(Patagonia). Pages 296—300 /// A. F. Br)'an, ed., Earlj Man in America from a Circum- 

Pacific Perspective. Occasional Papers No. 1, Department of Anthropology, L^niversirs^ 

of Alberta, Edmonton, Canada. 
Cardich, A., F. Cardich, & A. Hajduk. 1973. Secuencia arqueologica y cronologia radio- 

carbonico de la cueva 3 de Fos Toldos. Relaciones de la Sociedad Argentina de Antropologia 

Chatters, J. C. 1998. Personal communication to A. C. Roosevelt. 
Chatters, J. C, W' A. Neves, & AF Blum. 1999. The Kennewick Man: A first multivariate 

analysis. Cum Kas. Pleistocene 16:87—90. 
Chauchat, C. 1988. Early hunter-gatherers on the Peruvian Coast. Pages 41—66 /// R. 

Keatinge, ed., Peruvian Prehistory. Cambridge L^niversit\' Press, Cambridge, ALA. 



. 1992 Prehistoire de la Cote Nord du Perou: Le Paijanien de Cupisnique. Centre 

National de la Rechrche Scientifique Editions, Bordeaux. 391 pp. 

Clark, D. W 1991. The northwestern (Alaska-Yukon) fluted points. Pages 35^8 in R. 

Bonnichsen & K. L. Turnmire, eds. Clovis: Origins and Adaptations. Center for the 

Study of the First Americans, Department of Anthropology, Oregon State 

University, Corvallis, OR. 
CoUnvaux, R A., P E. de OUveira, P E. Moreno, M. C. MiUer, & M. B. Bush. 1996. A 

long pollen record from lowland Amazonia: Forest and cooUng in glacial times. Science 

Correal Urrego, G. 1981. Evidencias Culturales y Megafauna Pleistocenica en Colombia. 

Fundacion de Investigaciones Arqueologicas Nacionales, Banco de la Republica, 

Bogota, Columbia. 148 pp. 
Correal Urrego, G., T. van der Hammen, & W Hurt. 1972. Preceramic sequences in the 

El Abra Rock Shelters, Colombia. Science 175:1106—1108. 
Correal Urrego, G, & T. van der Hammen. 1977. Investigaciones Arqueologicas en los Ahrigos 

Rocosos del Tequendama. BibHotheca del Banco Popular, Premios de Arqueologia, 

Bogota, Columbia. 194 pp. 
Cotter, J. L. 1938. The occurrence of flints and extinct animals in pluvial deposits near 

Clovis, New Mexico. Part IV, Report on the excavations at the grand pit, 1936. Proc. 

Phila. Acad Nat. Sci. 8(16). 

. 1939 The occurrence of flints and extinct animals in pluvial deposits near 

Clovis, New Mexico. Proc. Phila. Acad Nat. Sci. 90:113-117. 

Damon, P. E. , & A. Long. 1962. Arizona radiocarbon dates III. Radiocarbon 4:239-249. 

Damon, P. E., C. V. Haynes, Jr., & A. Long. 1964. Arizona radiocarbon dates V. 
Radiocarbon 91—107. 

Davies, D. M. 1978. Some observations on the Otavalo skeleton from Imbabura 
Province, Ecuador. Pages 273—274 in A. L. Bryan, ed.. Early Man in America from a 
Circum-Pacific Perspective. (Occasional Papers No. 1, Department of Anthropology, 
University of Alberta). Archaeological Researchers International, Edmonton, 

de Lumley, H., de Lumley, M.-A., M. M. C. Beltrao, Y. Yokayama, J. Labeyrie, J. Danon, 
G. DeUbrias, C. Falgueres, & J. L. Bischoff. 1987. Presence d'outils tallies associes a 
une faune quaternaire datee du Pleistocene moyen dans la Toca da Esperanca, Region 
de Central, Etat de Bahia. EAntbropologie 91:917-942. 

Derev'anko, A. P., ed. 1998. The Paleolithic of Siberia: New Discoveries and Interpretations. 
University of Illinois Press, Urbana, IL. 406 pp. 


rc)osi:vi:j:i', dcjuglas and brown 

Dikov, N. N. 1996. 1'hc Ushki Sites: Kamchatka Peninsula. Pages 244-250 in V. H. West, 

ed., American Beguuiings: Prehistor-y and Pcihicoecolog)! of Beringia. Universit}' of Chicago, 

Chicago, IL. 
Dilleha)', T. D. 1989. Monte \erde: A Lxjte Pleistocene Settlement in So/it hern Chile. Vol. 1. 

Smithsonian Institution, Washington, D.C. 

. 1997 Monte Verde. Vol. 2. Smithsonian Institution, Washington, D.C. 1071 pp. 

. 2000. The Settlement of the Americas: A New Prehistory. Basic Books, New York, 

NY. 371 pp. 
Dillehay, T. D, G. I. Ardila, G. PoHtis, &c M. M. C. Beltrao. 1992. Earliest hunters and 

gatherers of South America./. World Prehist. 6:145—204. 
Di Peso, C. C. 1965. The Clovis fluted point from the Timmy Site. Kiva 31:83-87. 
Dixon, E. J. 1999. Bones, Boats <& Bison. Universit)' of New Mexico Press, Albuquerque, 

NM. 322 pp. 
Dixon, E. J., T. H. Heaton, T. E. Fifield, T. D. Hamilton, D. E. Putnam, & E Grady. 1997. 

Late Quaternary regional geoarchaeology of southeast Alaska Karst, A progress 

report. Geoarchaeology 12:689—712. 
Egan, T. 1999. Expert panel recasts origin of fossil man in northwest. The New York 

Times, October 16:A11. 
Erlandson, J. M., «Sc M. Moss. 1996. The Pleistocene-Holocene transition along the 

Pacific coast of North America. Pages 277—302 in L. G. Stxaus, B. V. Erisken, J. M. 

Erlandson, & D. R. Yesner, eds.. Humans at the End of the Ice Age: The Archaeology' of the 

Pleistocene-Holocene Transition. Plenum, New York, NY. 
Erlandson, J. M., D. J. Kennett, B. L. Ingram, D. A. Guthrie, D. P. Morris, M. A. Tveskov, 

G.J. West, & P.J. Walker. 1996. An archaeological and paleontological chronology for 

Daisy Cave, San Miguel Island, California. Radiocarbon 38:355—373. 
Evans, C, & B. J. Meggers. 1961. Archaeological Investigations in British Guiana. 

Smithsonian Institution, Bureau of American Ethnology, Bulletin 177. Washington, D.C. 

476 pp. 
Pagan, B. 1987. The Great Journey: The Peopling of Ancient Am/erica. Thames and Hudson, 

London. 288 pp. 
Ferring, C. R. 1989. The Aubrey Clovis site: A Paleoindian locaUt}^ in the upper Trinit}' 

river basin, Texas. Curr Kes. Pleistocene 6:9—1 1 . 
. 1990. The 1989 investigations at the Aubrey Clovis site, Texas. Gun: Res. 

Pleistocene 7:10—12. 
. 1994. Eate Quaternary Geology of the Upper Trinity River Basin, Texas. Ph.D. disser- 

tation. Department of Geology, Universit}^ of Texas, Dallas. University- Microfilms, 
Ann Arbor, MI. 



. 1995. The late Quaternary geology and archaeology of the Aubrey Clovis site, 

Texas. Pages 273—282 in E.Johnson, td., A^ncient Peoples and L^andscapes. Museum 

of Texas Tech University, Lubbock, TX. 
Fiedel, S. J. 1987. Prehistory of the Americas. Cambridge University Press, Cambridge, 

England. 386 pp. 
. 1992. Prehistory of the Americas. Second edition. Cambridge University Press, 

Cambridge, England. 400 pp. 

. 1996 Letter. Science 274:1821-1822. 

. 1999a. Older than we thought: Implications of corrected dates for Paleoindians. 

Am. Antiquity 64:95—115. 
. 1999b. Artifact provenience at Monte Verde: Confusion and contradictions. 

Discovery Archaeology, Special Keport November/December: 1—1 2. 
. 2000 The peopling of the New World: Present evidence, new theories, and 

future directions./. Archaeol Kes. 8:39—103. 
Fladmark, K. 1978. The feasibility of the northwest coast as a migration route for early 

man. Pages 19—28 in A. L. Bryan, ed.. Early Man in America from a Circum-Pacific 

Perspective. (Occasional Papers No 1 . of the Department of Anthropology, University 

of Alberta). Archaeological Researchers International, Edmonton, Alberta, Canada. 
Flegenheimer, N. 1980. Hallazgos de puntas "cola de pescado" en la Provincia de Buenos 

Aires. Kelaciones de la Sociedad Argentina de Anthropologia 14:169—176. 
. 1987 Recent research at localities Cerro la China and Cerro El Sombrero, 

Argentina. Curr. Kes. Pleistocene 4:14S— 149. 
Prison, G. 1991. Prehistoric Hunters of the High Plains. Academic Press, San Diego. 532 pp. 
. 1999. The Goshen cultural complex: A Paleoindian cultural group overlapping 

Clovis. Page 3 in Clovis and Beyond Conference, Santa Fe, New Mexico, October 28—31, 

1999. Abstracts. 
Prison, G, ed. 1996. The Mill Iron Site. University of New Mexico Press, Albuquerque, 

NM. 248 pp. 
Prison, G, & B. Bradley. 1999. The Fenn Cache: Clovis Weapons and Tools. One Horse Land 

and Cattle Company, Santa Fe, NM. Ill pp. 
Prison, G, & D. J. Stanford, eds. 1982. The Agate Basin Site: A Record of the Paleoindian 

Occupation on the Northivestern High Plains. Academic Press, New York, NY. 403 pp. 
Prison, G. C, C. V. Haynes, Jr., & M. L. Larson. 1996. Discussion and conclusions. Pages 

205-216 in G C Prison, ed.. The Mill Iron Site. University of New Mexico Press, 

Albuquerque, NM. 
Gnecco, C. 1994. The Pleistocene-Holocene Boundary in the Northern Andes: An Archaeological 

Perspective. Ph.D. dissertation. Department of Anthropology, Washington University, 

St. Louis, Mo. University Microfilms, Ann Arbor, MI. 



Gnecco, (>., & S. Mora. 1997. Late Pleistocene/early Holocenc tropical forest occupa- 
tions at San Isidro and Pcna Rojo, Colombia. Antiquity 7L683-690. 

Goebel, T. 1999. Beringia and human colom:(ation for the Americas: A New Vieiv from 
the North. Poster session paper, Clovis and Beyond Conference, Santa Fe, NM. 

Goebel, T., R. Powers, & N. Bigelow. 199L The Nenana complex of Alaska and Clovis 
origins. Pages 49—80 in R. Bonnichsen and K. L. Turnmire, eds., Clovis: Origins and 
Adaptations. Center for the Study of the First Americans, Department of 
Anthropology, Oregon State University, Corvallis, OR. 

Gramley, M. 1993. The Richey Clovis Cache. Persimmon Press, Buffalo, New York, NY. 

Gruhn, R. 1991. Stradfied radiocarbon-dated sites of Clovis age and older in Brazil. 
Pages 283—286 in R. Bonnichsen & K. L. Turnmire, eds., Clovis: Origins and 
Adaptations. Center for the Study of the First Americans, Department of 
Anthropology, Oregon State University, Corvallis, OR. 

. 1994 The Pacific coast route of initial entry: An overview. Pages 249—256 in R. 

Bonnichsen & D. G. Steele, eds.. Method and Theory for Investigating the Peopling of the 
Americas. Center for the Study of the First Americans, Department of Anthropology, 
Oregon State Universit}'^, Corvallis, OR. 

Gruhn, R., & A. L. Bryan. 1977. Los Tapiales: A Paleoindian campsite in the Guatemalan 
highlands. Proc. Am. Philos. Soc. 121:235—273. 

Guidon, N., & G. Delibrias. 1986. Carbon-14 dates point to man in the Americas 32,000 
years ago. Nature yiX-Hd^—llX. 

Haberle, S. G., & M. A. Maslin. 1999. Late Quaternary vegetation and climate change in 
the Amazon basin based on a 50,000 year pollen record from the Amazon fan, ODP 
site 9?>2. Quat. Res. 51:27-38. 

Hannus, L. A. 1990. The Lange- Fergus on site: A case for mammoth-bone butchering- 
tools. Pages 86—99 in L. D. Agenbroad, J. L. Mead, & L. Nelson, eds., Megafauna and 
Man: Discovery of America's Hearihland. The Mammoth Sites of Hot Springs, South 
Dakota, Inc. Scientific Papers, Vol. 1 . 

Harbour, J. 1975. General stratigraphy. Pages 33—56 in F. Wendorf & J. J. Hester, eds., 
LMte Pleistocene Environents of the Southern High Plains. Fort Burgwin Research Center, 
Inc., Southern Methodist University, Ranchos de Taos, NM. 

Harington, C. R., R. Bonnichsen, & R. E. Morlan. 1975. Bones say man lived in Yukon 
27,000 years ago. Can. Geogr J. 91:42-48. 

Haury, E. W, E. B. Sayles, & W. W. Wasley. 1959. The Lehner mammoth site, southeast- 
ern Arizona. Am. Antiquity 25:2—30. 

Haynes, C. V., Jr. 1964. Fluted projectile points: Their age and dispersion. Science 145: 



. 1966. Elephant hunting in North America. Pages 44—52 in R. S. MacNeish, ed., 

Harly Man in A.merica, ladings from Scientific American. W H. Freeman, San Francisco, 

. 1987. Clovis origin update. The Kiva 52:83—93. 

. 1973. The Calico site: Artifacts or geofacts? Chipped flints are either the old- 

est evidence of man in the New World or products of geological proceses. Science 
. 1974. Paleoenvironmental and cultural diversity in Late Pleistocene South 

America: Reply to A. L. Bryan. Ouat. Res. 4:378-380. 

. 1991. More on Meadowcroft chronology. Kev. Archaeol. 12:8—14. 

. 1992. Contributions of radiocarbon dating to the geochronology of the peo- 

pling of the New World. Pages 355-374 in R. E. Taylor, A. Long, & R. S. Kra, eds., 
14 C Dating and the Peopling of the New World. Springer- Verlag, New York, NY. 

. 1997. Technical comment. Science 275:1948. 

. 1998. Personal communication. Letter to A.C. Roosevelt, September 28, 1998. 

. 1999. Monte Verde and the pre-Clovis situation in America. 

Archaeology, Special ^port: Monte l^erde Revisited December. 17—19. 
Haynes, C. V, Jr., & G. A. Agogino. 1966. Prehistoric springs and geochronology of the 

Clovis site. Am. Antiquity 31:812—821. 
Haynes, C. V, Jr., P. E. Damon, & D. C. Grey. 1966. Arizona radiocarbon dates VI. 

Radiocarbon 8:1—21. 
Haynes, C. V., Jr., D. C. Grey, P. E. Damon, & R. Bennett. 1967. Arizona radiocarbon 

dates \TI. Radiocarbon 9:1—14. 
Haynes, C. V, Jr., M. McFeul, R. H. Brunswig, & K. D Hopkins. 1998. Kersey-Kuner 

Terrace investigations at the Dent and Bernhardt sites, Colorado. Geoarchaeology 13: 

Hemmings, E. T 1970. Early Man in the San Pedro Valley, Arii^ona. Ph.D. dissertation. 

Department of Anthropology, University of Arizona. University Microfilms, AZ. 

268 pp. 
Hester, J. J., E. Lundelius,Jr., & R. FryxeU. 1972. Blackivater Locality No. 1:A Stratified Early 

Man Site in Eastern New Mexico. (Publication No. 8). Fort Burgwin Research Center, 

Southern Methodist University, Ranchos de Taos, NM. 238 pp. 
Hoffecker, J. F, W R. Powers, & T. Goebel. 1993. The colonization of Beringia and the 

peopling of the New World. Science 259:46—53. 
Hoffecker, J. F, W. R. Powers, & N. H. Bigelow. 1996. Dry Creek. Pages 343-354 /// F H. 

West, ed., American Beginnings: The Prehistory and Paleoecology of Beringia. University of 

Chicago Press, Chicago, IL. 


ii()()si:vi'];i; douglas and brc avn 

I lollithu', \'. T. 1997. Wih'oiiuluni CcoLirclHU'ology of /he So/i/hcni I Ugh P/diiis. L'nivcrsity of 

Texas Press, Austin, TX. 297 pp. 
Humphrey, ). D., &; C R. Ferring. 1994. Stable isotope evidence for latest Pleistocene and 

Holocene climatic change in north central Texas. Oiiat. Res. 41:200—213. 
Hurt, W. R., T van der Hammen, &c G. Correal. 1976. The AlAbra Kockshe/tirs, Sabana de 

Bogota, Coloiubici Soiitl) y\merica. (Occasional Papers and Monographs, No. 2). Indiana 

Universit}' Museum, Bloomington, IN. 
Jennings, J. D. 1983. Origins. Pages 25—69 /// j. D.Jennings, cd.. Ancient North Americans. 

W. H. Freeman and Company, New York, NY. 

, ed. 1983. Ancient North Americans. WH. Freeman, San Francisco, CA. 642 pp. 

Johnson, |. R., T. W. Stafford, Jr., H. O. Ajie, & D. P. Morrison. In press. Arlington 

Springs revisited. Proceedings of the Fifth California Islands Symposium. U.S. Minerals 

Management Service, Camarillo, CA. 
Keefer, D., S. deFrance, M. E. Moseley, J. B. Richardson III, D. R. Satterlee, & A. Day- 
Lewis. 1998. Early maritime economy and El Nino events at Quebrada Tacahuay, 

Peru. i^f/Vz/fg 281:1833-1835. 
Klein, R. 1999. The Hitman Career: Human Cultural and Biological Origins. Universit}- of 

Chicago Press, Claicago, IL. 810 pp. 
Krieger, A. D. 1964 Early man in the New World. Pages 23—84 in ]. D. Jennings & E. 

Norbeck, eds.. Prehistoric Man in the Neiv World. Universit}' of Chicago Press, Chicago, 

Kunz, M. L., & R. E. Reanier. 1994. Paleoindians in Beringia: Evidence from Artie 

Alaska. Science 263:660—662. 
Lahr, M. M. 1995. Patterns of modern human variation: Implicadons for Amerindian 

origins. Yrbk. Phjs. Anthropol 38:163—198. 
Leakey, L. S. B., R. D. Simpson, & T. Clements. 1968. Archaeological excavations in die 

Calico Mountains, California: preliminary report. Science 181: 305—310. 
Leonhardy, F. C, ed. 1966. Domebo: A Paleo-lndian Mammoth Kill in the Prairie-Plains. 

(Contributions of the Museum of the Great Plains, No. 1. Lawton, Oklahoma). 

53 pp. 
Lopes de Castano, C. 1995. Dispersion de puntas de provectil bifaciales en la cuenc 

media del rio Magdalena. Pages 73—82 /'/; Ambito j Ocitpaciones Tempranas de America 

Tropical. Fundacion ERIGAIE, Instituto Colombiano de Antropologia, Bogota, 

Lopes de Castano, C, & C. Niewenhuis n.d. Clovis from the perspective of northern 

South America. In preparation for A.C. Roosevelt & |. Morrow, eds., Chris in Context: 

Neip Tight on the Peopling of the Americas. In preparation for submission to Universit)^ 

of Arizona Press, Tucson, AZ. 



Lovett, J. C. 1993. Climate history and forest distribution in eastern Africa. Pages 23-32 
in J. C. Lovett & S. K. Wasser, eds., Biogeography and Ecology of the Rain Forests 
of Eastern Africa. Cambridge University Press, Cambridge, England. 

Lynch, T. F. 1974. The antiquity of man in South America. Ouat. Kes. 4:356-377. 

. 1980. Guitarrero Cave: Earlj Man in the A.ndes. Academic Press, New York, NY. 

48 pp. 

. 1983. The Paleo-Indians. Pages 87-138 in J. D. Jennings, ed.. Ancient South 

Americans. W. H. Freeman, San Francisco, CA. 
. 1990. Glacial-age man in South America? A critical review. Am. 

Lynch, T. F, R. Gillespie, & J. Gowlett. 1985. Chronology of Guitarrero Cave, Peru. 

Science 229:864-867. 
MacNeish, R. S. 1991. The Fort Bliss Archaeological Project by AFAR Excavation of Pintada 

and Pendejo Caves near Orogrande, Neiv Mexico. 1991 Annual Report. Unpublished 

Report, Andover Foundation for Archaeological Research, Andover, MA. 
. 1992. The 1992 Excavations of Pendejo and Pintada Caves near Orogrande, New 

Mexico: An AFAR and Fort Bliss Archaeological Project Unpublished Report, Andover 

Foundation for Archeological Research, Andover, MA. 
MacNeish, R. S., R. Berger, & R. Protsch. 1970. Megafauna and man from Ayacucho, 

highland Peru. Science 168:975—977. 
MacNeish, R. S., R. K. Vierra, A. Nelken-Turner, & C. J. Phagan, eds. 1980. Prehistory of 

the Ayacucho Basin, Peru. vol. 3. University of Michigan Press, Ann Arbor, MI. 333 pp. 
MacNeish, R. S., A. Garcia Cook, L. G. Lumbreras, R. K. Vierra, & A. Nelken-Turner, 

eds. 1981a. Prehistory of the Ayacucho Basin, Peru. Vol. I. University of Michigan Press, 

Ann Arbor, MI. 
. 1981b. Prehistory of the Ayacucho Basin, Peru. vol. 2. University of Michigan Press, 

Ann Arbor, MI. 266 pp. 
MacNeish, R. S., R. K. Vierra, A. Nelken-Turner, R. Lurie, & A. Garcia Cook. 1983. 

Prehistory of the Ayacucho Basin, Peru. vol. 4. University of Michigan Press, Ann Arbor, 

MI. 280 pp. 
MacNeish, R. S., G Cunnar, G. Jessop, & P. Wilner. 1993. A Summary of the Paleoindian 

Discoveries in Pendejo Cave near Orogrande, New Mexico. The Annual Report of AFAR for 

1993. Andover Foundation for Archaeological Research, Andover, MA. 48 pp. 
Martin, P S. 1967. Prehistoric overkill. Pages 75-120 in P S. Martin & H. E. Wright, eds.. 

Pleistocene Extinctions: The Search for a Cause. Yale Universit}^ Press, New Haven, CT. 
. 1984. Prehistoric extinctions: The global model. Pages 354-403 in P. S. 

Martin & R. G. Klein, ^A^., Quaternary Extnctions: A Prehistoric Revolution. University of 

Arizona Press, Tucson, AZ. 


rc3osi^vuj:i; Douglas and brown 

Mayer-Oakes, W. |. 1986. lU-Inga: A l^ilcoimlian site in the Sierra of northwestern 

RcLiador. I'n/j/s. .Iw. Phi/os. Soc. 76(4). 
McAvoy, J. M., & L. D. McAvoy. 1997. Archia'ological Invtstigations of Site 44SX202, Cactus 

Hill, Susses County, \'^irgima. Virginia Department of Historic Resources, Research 

Report No. 8. Nottoway Rivers Survey Archaeological Research Report No. 2. 

Sandston, VA. 477 pp. 
Mead, ). I., & D. }. Meltzer. 1984. North American Quaternary extinctions and the radio- 
carbon record. Pages 440—450 in P. S. Martin & R. G. I-Clein, eds., Quaternary 

Extinctions: A Prehistoric Kevolution. The Universit}^ of Arizona Press, Tucson, AZ. 
Meltzer, D. J. 1993. Search for the First Americans. St. Remy Press, Montreal and 

Smithsonian Books, Washington, D.C. 176 pp. 
Meltzer, D. J., & 1. 1. Mead. 1985. Dating Late Pleistocene extinctions: Theoretical issues, 

analytical bias, and substantive results. Pages 145—173 in]. I. Mead & D |. Meltzer, 

eds., Environments and Extinctions: Man in Eate Glacial North America. Center for the 

Study of Early Man, Universit}^ of Maine, Orono, ME. 
Meltzer, D. J., J. M. Adovasio, & T D. Dillehay. 1994. On a Pleistocene occupation at 

Pedra Furada, Brazil. Antiquity 68:695—714. 
Meltzer, D. J., D. K. Grayson, G. L Ardila, A. W Barker, D. E Dincauze, C. V. Haynes, E 

Mena, L. Nunez, & D. J. Stanford. 1997. On the Pleistocene antiquit}' of Monte 

Verde, Southern Chile. Am. Antiquity 62:659—663. 
Michab, M., J. K. Leathers, J. -L. Joron, N. Mercier, M. Selo, H Valladas, J. -L. Reyss, & 

A. C. Roosevelt. 1998. Luminescence dates for the Paleoindian site of Pedra Pintada, 

Brazil. ^//^/. Sci. Reu 17:1041-1046. 
Miller, E. T. 1987. Pesquisas archaeologicas paleoindigenas no Brasil occidental. Pages 

37—61 in L. Nunez & B. J. Meggers, eds., Jnvestigaciones Paleoinndias al Sur de la Einea 

Ecuatorial. Estudios Atacamenos, Numero Especial, Universidad del Norte, 

Antofagasta, Chile. 
Miotta, L. 1999. Quandary: The Clovis phenomenon, the First Americans, and the View 

from the Patagonian region. In Clovis and Beyond — Peopling of the Americas Conference, 

Abstracts. Santa Fe, New Mexico, October 28—31. 
Montane, J. C. 1968. Paleoindian remains from Laguna de Tagua Tagua, Central Chile. 

Science \6\:n?>l-n?>^. 
Morlan, R. E., & J. Cinq-Mars. 1982. Ancient Beringians: Human occupation in the Late 

Pleistocene of Alaska and the Yukon Territory. Pages 353—381 /// D. M. Hopkins, }. 

V. Matthhews, Jr., C. E. Schweger, & S. B. Young, eds., Paleoecology of Beringia. 

Academic Press, New York, NY. 



Morrow, J., & T. A. Morrow. 1999. Geographic variation in fluted projectile points: A 

hemispheric perspective. Am. Antiquity 64:215—231. 
Morse, D, D. G. Anderson, & A. Goodyear. 1996. The Pleistocene-Holocene transition 

in the eastern United States. Pages 319—338 in L. G. Strauss, B. V. Eriksen, J. M. 

Erlandson, & D R. Yesner, eds., Humans at the End of the Ice age: The Archaeology of the 

Pleistocene-Holocene Transition. Plenum Press, New York, NY 
Mosimann, J. E., & P. S. Martin. 1975. Simulating overkill by Paleoindians. Am. Sci. 

Nami, H. 1987. Cueva del Medio: A significant Paleoindian site in southern South 

America. Current Research in the Pleistocene 4:157—159. 
Neves, W A., D. Munford, & M. do Carmo Zanini. 1996. Cranial and morphological 

variation and the colonization of the New World: Towards a four-migration model. 

Paper presented at the Annual Meeting of the American Asssociation of Physical 

Anthropology, Durham, North Carolina, April 9—1 1 . 
Neves, W A., & H. M. PuciareUi. 1991. Morphological affinities of the first Americans: 

An exploratory analysis based on early South American remains. /. Hum. Evol 

Nieuwenhuis, C.J. 1998. Unattractive but effective: Unretouched pointed flakes as pro- 
jectile points? A closer look at the Abriense and Tequendamiense artifacts. Pages 

163—133 in M. G. Plew, ed., Explorations in American Archaeology: Essays in Honor of 

Wesley R Hurt. University Press of America, Lanham, MD 
Nunez, L. 1992. Occupacion arcaica en la Puna de Atacama: Secuencia, movilidad y cam- 
bio. Pages 283—308 in B. Meggers, ed., Prehistoria Sudamericana: Nuevas Perspectivas. 

Taraxacum, Washington, D. C. 
Nunez, L., J. Varela, & R. Casamiquela. 1994. Reconstruccion multidisciplinaria da la 

occupacion prehistorica de Quereo, Centra de Chile. Eatin Am. Antiquity 5:99—118. 
Ochsenius, C, & R. Gruhn. 1979. Taima-Taima: A Eate Pleistocene PaleoTndian Kill Site in 

Northernmost South America. CIPICS/South American Quaternary Documentation 

Program, Coro, Venezuela. 1 37 pp. 
Oliver, J. R., & C. S. Alexander. 1990. The Pleistocene peoples of western Venezuela: 

The terrace sequence of Rio Pedregal and new discoveries in Paraguana. Proceedings of 

the First World Summit Conference on the Peopling of the Americas. Center for the Study of 

the First Americans, Orono, ME. Unpublished. 
Pearson, G A. In press. The Paleoindian cultural diversit}^ of Costa Rica: Evidence of a 

melting pot between North American Clovis groups and and South American fish 

tail point makers? Occasional Publications Series, Center of Latin American Studies, 

No. 23. University of Kansas, Lawrence, KS. 


R()()si:vi:i;i; douglas and brown 

Pcrcira, E. 1992. Analyse prclimiar das pinturas rupcstres de Monte Alegre (Py\). holetim 
do Mil sen Paraense Emilio Coeldi, serie Antropologie. 8:5—24. 

Pessis, A. -M. 1999. The chronology and evolution of the prchist(;ric rock painungs in 
the Serra da Capivara National Park, Piaui, Brazil. Pages 41-47 //; M. Strecker & P. 
Bahn, eds., Dating and the Harliest Kh aim Rock Art. Oxbow Books, Oxford, England. 

Politis, G. 1991. Fishtail projectile points in the Southern Cone of South America: An 
overview. Pages 287—301 /'// R. Bonnichsen & K. L. Turnmire, eds. C/ovis: Origins and 
Adaptations. Center for the Study of the First Americans, Oregon State University, 
Corvallis, OR. 

Powell, J. F. 1993. Dental evidence for the peopling of the New World: Some method- 
ological considerations. Hum. Biol. 65:799—819. 

Powell, J. F, & W. A. Neves. 1998. Dental diversit}' of early New World populations: 
Taking a bite out of the tripartite model. Paper presented at the annual meeting of 
the American Association of Physical Anthropology. Am. J. Phjs. AnthropoL, Suppl. 

Powell, J. F, & J. Rose. 1999. Report on the Osteological Assessment of the "Kennewick 
Man" Skeleton (CENWW.97. Kennewick). Discovering Archaeology, Special Kepoii 
October 29:1-20. 

Powers, W R. 1996. Siberia in the late glacial and early postglacial. Pages 229—242 in L. 
G. Strauss, B. V. Eriksen, J. M. Erlandson, & D. R. Yesner, eds., Humans at the End of 
the Ice Age: The Archaeology of the Pleistocene-Holocene Transition. Plenum, New York. 

Preston, J. 1997. The lost man. The New Yorker, March 7:69-81. 

Prous, A. 1980—1981. FouiUes du Grand Abri de Santana de Riacho (Minas Gerais), 
Bresil. /. Soc. Americanistes 67:163—183. 

. 1986a. L'Archeologie au Bresil, 300 Siecles d'occupation humain. 

UAnthropologie 90:257-306. 

. 1986b. Os mais antigos vestigios arqueologicos no Brazil Central. Pages 

173—182 in A. L. Bryan, ed., New Evidence for the Pleistocene Peopling of the Americas. 
Center for the Study of Early Man, University of Maine, Orono, ME. 
. 1991. Fouilles de I'Abris du Boquete, Minas Gerais, Bresil. /. Soc. Americanistes 

. 1995. Archaeological analysis of die oldest settlements in the Americas. Kevista 

Brasileira de Genetica 18:689—699. 
. 1999. Dating rock art in Brazil. Pages 29-34 in M. Strecker & P. Balin, eds.. 

Dating and the Earliest Rock Aii. Oxbow Books, Oxford, England. 
Ranere A., & R. Cooke. 1991. Paleoindian occupation in the Central American tropics. 

Pages 237—253 /'// R. Bonnichsen & K. L. Turnmire, eds., Clovis: Origins and 



Adaptations. Center for the Study of the First Americans, Oregon State University, 
Corvallis, OR. 

Reanier, R. E. 1996. Putu and Bedwell. Pages 505—511 in F. H. West, ed., American 
Beginnings: The Prehistory and Palaeoecology of Beringia. University of Chicago, Chicago, 

. 1997. Technical comment. Science 275:1948-1949. 

Reanier, R. E., & M. Kunz. 1995. The Mesa site: A Paleoindian hunting lookout in Arctic 
Alaska. Arctic Anthropo I. 32:5—30. 

Richardson, J. B. 1978. Early man on the Peruvian north coast: Early maritime exploita- 
tion and the Pleistocene and Holocene environment. Pages 274—289 in A. L. Bryan, 
ed.. Early Man in American in a Circum-Paciftc Perspective. (Occasional Papers, 
Department of Anthropology, University of Alberta ). Archaeological Researches 
International, Edmonton, Canada. 

Rick, J. W 1980. Prehistoric Hunters of the High Andes. Academic Press, New York, NY. 
360 pp. 

Roberts, A., & P. Julig. n.d. Paleoindian littoral adaptations. Paper prepared for A. C. 
Roosevelt & J. Morrow, eds., Clovis in Context: NeiP hight on the Peop/ng of the Americas. 
Volume in preparation for submission to the University of Arizona Press, AZ. 

Roosevelt, A. C. 1998a. Clovis clarification: A follow-up. Mammoth Trumpet 13:14—17. 

. 1998b. Paleoindian and Archaic occupations in the Lower Amazon: A summa- 
ry and comparison. Pages 165—191 in M. G. Plew, ed.. Explorations in American 
Archaeology: Essays in Honor of Wesley R Hurt. University Press of America, Lanham, 

. 1999. Dating the rock art at Monte Alegre, Brazil. Pages 35^1 in M. Strecker 

& P. Bahn, eds.. Dating and the Earliest Known Rock Ari. Oxbow Books. Oxford, 
. 2000a. Who's on first? There's still no end to the controversy over when and 

how humans populated the New World. Natural History 7:76-79. 
. 2000b. New information from old collections: The interface of science and sys- 

tematic collections in A. Hitchcock, ed.. Museum Issues and Trends, Cultural Resource 
Management 23:25-29. U.S. Department of the Interior, National Park Service, 
Culmral Resources, Washington, DC. 
. 2000c. The Lower Amazon: A dynamic human habitat. Pages 455^91 in D 

Lentz, ed.. Imperfect Balance: Landscape Transformations in the Precolumbian Americas. 
Columbia Universit)^ Press, New York, NY. 
. 2000d. Clovis in Context: New Light on the Peopling of the Americas. Hum. 

Biol 12:1-19. 


R()c)Sj'Vi:];r, douglas and brown 

. n.d.a. Gciukr in luimaii cxokition: Sociobiologv revisited and ixxiscd. Chapter 

19 /// S. Nelson & M. yVyalon, eds., //; Search of Gender. Altamira Press. In press. 
. n.d.b. Clovis dating, stratigraphy, and subsistence revisited from a hemispheric 

perspective />; A. C. Roosevelt (!s: |. Morrow, eds., C/ovis in Context: Neiv IJght on the 
Peopling of the Americas. In preparadon for submission to Universit}' of Arizona Press, 
Tucson, AZ. 
. n.d.c. Environment, Subsistence, and Behavior in Human Evolution /// B. 

Ferguson & R. Sussman, eds., ^ Critique of Social Danvinism. In preparation for sub- 
mission to University of California Press. 
Roosevelt, A. C, & J. Morrow, eds. n.d. Clovis in Context: New Ught on the Peopling of the 

Americas. In preparation for submission to University of Arizona Press, Tucson, AZ. 
Roosevelt, A. C, R. A. Housley, M. I. da Silveira, S. Maranca, & R.Johnson. 199E Eighth 

millennium pottery from a prehistoric shell midden in the Brazilian Amazon. Science 

Roosevelt, A. C, M. L. da Costa, C. Lopes Machado, M. NOchab, N. Mercier, H. Vailadas, 

J. Feathers, W. Barnett, M. I. da Silveira, A. Henderson, |. Sliva, B. Chernoff, D. S. 

Reese, J. A. Holman, N. Toth, & K. Schick. 1996. Paleoindian cave dwellers in the 

Americas: The peopUng of the Americas. Science 272:373—384. 
Roosevek, A. C, L. Brown, J. Douglas, M. O'Connell, E. Quinn, & J. Kemp. 1997. 

Dating a Paleoindian site in the Amazon in comparison with Clovis culture: Technical 

comments. Science 275:1950—1952. 
Rouse, I., & J. M. Cruxent. 1963. Venet^uelan Archaeology. Yale Universit}' Press, New 

Haven, CT. 179 pp. 
Sales Barbosa, A. 1992. A tradicao Itaparica: Uma compreensao ecologica e cultural 

dopovoamento inicial do planalto central brasileiro. Pages 145—160 in B. Meggers, ed., 

Prehistoria Sndamencana: Nnems Perspectivas. Taraxacum, Washington, DC. 
Sandweiss, D. H., H. Mclnnis, R. L. Burger, A. Cano, B. Ojeda, R. Paretics, M. C. 

Sandweiss, & M. D. Glasscock. 1998. Quebrada Jaguay: Early South American mar- 
itime adaptations. Science 281:1830-1832. 
Sassaman, K. E. 1996. Early Archaic settiement in the South CaroHna coastal plain. Pages 

58—83 in D G. Anderson & K. E. Sassaman, eds.. The Paleoindian and Earl)' Archaic 

Southeast. Universit}' of Alabama, Tuscaloosa, AL. 
Schmitz, P. 1987. Prehistoric hunter-gatherers of Brazil. /. World Prehist. 1:3—126. 
Schobinger, J. 1999. Argentina's oldest rock art. Pages 53—66 in M. Strecker & P. Bahn, 

eds.. Dating the Earliest Known Rock Art. Oxbow Books, Oxford, England. 
Scott, G. R., & C. G. Turner, II. 1997. The Anthropologj' of Modern Human Teeth: Dental 

Morpholog)' and its I "anation in Recent Human Populations. Cambridge University- Press, 

Cambridge, England. 382 pp. 



Sioli, H., ed. 1984. The Amazon: Umnology and landscape 'Ecology of a Mighty Tropical River 

and its Basin. Dr. W Junk Publishers, Dordrecht. 746 pp. 
Simoes, M. 1976. Nota sobre duas pontas de Projetil da Bacia Tapajos (Para). Bo/. Mus. 

Paraense Emilio Goeldi N.S. 62. Belem. 
Simpson, R. D. 1978. The Calico Mountains archaeological site. Pages 218—220 in A. L. 

Bryan, ed., Early Man in A^merica from a Circum-Tacific Perspective. (Occasional Papers No. 

1 of the Department of Anthropology, University of Alberta). Archaeological 

Researches International, Edmonton, Canada. 
Soffer, O., & N. D. Praslov, eds. 1993. From Kostenki to Clovis. Plenum, New York. 334 pp. 
Snarskis, M. 1979. Turrialba: A Paleoindian quarry and workshop site in Eastern Costa 

Rica. Am. Antiquity 44:125-138. 
. 1984. Central America: The Lower Caribbean. Pages 195—232 in E W. Lange & 

D Z. Stone, eds.. The Archaeology of Tower Central America. Universit)^ of New Mexico 

Press, Albuquerque, NM. 
Stafford, T W., Jr., J. T. JuU, K. Brendel, R. C. Duhamel, & D. Donahue. 1987. Study of 

bone radiocarbon dating accuracy at the University of Arizona NSF Accelerator 

Facility for Radioisotope Analysis. Radiocarbon 29:24—44. 
Stafford, T W, Jr. 1990. Late Pleistocene megafauna extinctions and the Clovis culture: 

Absolute ages based on accelerator 14 C dating of skeletal remains. Pages 118—122 in 

L. D Agenbroad, J. I. Mead, & L. W. Nelson, eds., Megafauna and Man: Discovery of 

America's Heartland.. The Mammoth Site of Hot Springs, South Dakota, Inc, Hot 

Springs, SD, & Northern Arizona University, Flagstaff, AZ. 
Stafford, T W, Jr., P E. Hare, L. Currie, A .J. T Hull, & D. Donahue. 1991. Accelerator 

radiocarbon dating at the molecular level. /. Archaeol. Sci. 18:35—72. 
Stanford, D J. 1983. Pre-Clovis occupation south of the ice sheets. Pages 65-72 /// R. 

Shutler, Jr., ed., Tiarly Man in the New World. Sage Publications, Beverly HiUs, CA 
. 1991. Clovis origins and adaptations: An introductory perspective. Pages 1^ 

in R. Bonnichsen & K. L. Turnmire, eds., Clovis: Origins and Adaptations. Center for the 

Study of the First Americans, Department of Anthropology, Oregon State 

University, CorvaUis, OR. 
Steele, D. G., & J. H. Powell. 1993. Paleobiology of the first Americans. Evol Anthn 2: 

Stuiver, M., R J. Rimer, E. Bard, J. W. Beck, G. S. Burr, K. A. Hughen, B. Kromer, F G 

McCormac, J. van der PUcht, & M. Spurk. 1998. INTCAL 98 Radiocarbon age caH- 

bration, 24,000-0 cal BR Radiocarbon 40:1041-1083. 
Taylor, R. E. 1987. Radiocarbon Dating: An Archaeological Perspective. Academic Press, New 

York, NY. 212 pp. 


R( )( )SI A'l 'l :i', D( )UGLAS AND BROWN 

. 1992. Radiocarbon elating of hone: To collagen aiul l)c\'oncl. Pages 375-402 in 

R. E. Taylor, A. Long, & R. Kra, eds., Kud/occirboii Da //no i/f/cr I'o/ir Decades: An 
Interdisciplinary Perspective. Springer- Verlag, New \'ork, N'S'. 

. 1997. Radiocarbon dating. Pages 65-96 /// R. \l. Ta\lor & M. J. Airkcn, eds., 

Cbronometric Dating in yArchaeo/ogy. Plenum Press, New York, NY. 
Taylor, R. E., & M. |. Aitken, eds. 1997. Chrononietric Dating in Archaeology. (Advances in 

Archaeological and Museum Science, vol. 1). Plenum Press, New York. 395 pp. 
Taylor, R. E., C. V. Haynes, Jr., & M. Stuiver. 1996. Clovis and Folsom age estimates: 

Stratigraphic context and radiocarbon calibration. Antiquity 70:515—525. 
Taylor, R. E., L. A. Payne, C. A. Prior, P J. Slota, Jr., R. Gillespie,]. A.J. Gowlett, R. E. 

M. Hedges, A. J. T. JuU, T. H. Zabel, D. J. Donahue, & R. Berger. 1985. Major revi- 
sions in the Pleistocene age assignments for North American human skeletons by C- 

14 accelerator mass spectrometry: None older than 11,000 years B. P. A??i. Antiquity 

Tieszen, L. 1991. Natural variations in the carbon isotope values of plants: Implications 

for archaeology, ecology, and paleoecology. /. ArchaeoL Jr/. 18:227— 248. 
Toth, N. 1991. The material record. Pages 53-76 in T. D. Dillehay & D J. Meltzer, eds.. 

The First Americans: Search and Kesearch. CRC Press, Boca Raton, PL. 
Turnbull, C. 1962. The Forest People. Simon and Schuster, New York, NY. 295 pp. 
van der Hammen, T., & M. L. Absy. 1994. Amazonia during the last glacial. Palaeogeogr., 

Palaeoclimateol, Palaeoecol 109:247—261. 
van der Merwe, N., & E. Medina. 1991. The canopy effect, carbon isotope ratios, and 

foodwebs in Amazonia./. ArchaeoL Sci. 18:249—259. 
Vialou, A. v., & D. Vialou. 1994. Les premiers peuplements prehistoriques du Mato 

Grosso. P)ulL Sod. Prehistorique Francais 19:57—263. 
Wallace, A. R. 1889. A 'Narrative of Travels on the Aw a -y on and RJo Negro. London, Ward, 

Lock. 363 pp. 
Warnica, J. M. 1966. New discoveries at the Clovis site. Am. Antiquity 31:345—357. 
West, F. H., ed. 1996. American Beginnings: The Prehistory and Paleoecology of Beringia. 

Universit}' of Chicago Press, Chicago, IL. 576 pp. 
Wilford, J. N. 1999. New answers to old questions: Who got here first? The New York 

Times ^OY. 9:D1, D4. 
Willey, G. R. & P. Phillips. 1958. Method and Theory in American Archaeology. L^niversit)' of 

Chicago Press, Chicago, IL. 269 pp. 
Wormington, M. 1957. Ancient Man in Norih America. Fourth edition. Denver Museum of 

Natural History, Denver. 322 pp. 



Yesner, D R. 1996. Human adaptation at the Pleistocene-Holocene border. Pages 
255—276 in L. G. Straus, B. V. Eriksen, J. M. Erlandson, & D R. Yesner, eds., Humans 
at the End of the Ice Age. Plenum, New York, NY. 

. n.d. Technological, chronological, and economic variability in Paleoindian 

occupations from eastern Beringa in A. C. Roosevelt & J. Morrow eds., Clovis in 
Context: 'New Ught on the Peopling of the Americas. In preparation for submission to 
University of Arizona Press. 


R( )( )Si:vi':i:i', ix )Lig].as akd brown 


Dates From High Plains Clovis Sites'^ 

llje Clovis Site, Btachmter Draw i^ocatily t , Ctovis, NM 

Weighted average of tijree dates on tfjree samples: 1 1 ,300 ± 240 jr B.P. (13, 1 80 


The dates from the eponymous Clovis site fail five of the seven reiiabiiit}- criteria. 
Thev have too-large error bars, are on material vulnerable to contamination by too-old 
carbon, are on noncultural carbon, and lack of secure cultural association. In the region- 
al sequence, they average 300 vr earlier than dates from sites with high reliability' scores. 

The context was interpreted as a kill or scavenging site. The material dated was humic 
acid and Hgnin from pooled fragments of naturally carbonized water plants. The amino 
acids of bones of mammoth prey could be dated by current AMS methods but have not 
been so far. Abundant calcareous rocks and their solutes are local geological carbon 
sources for uptake by water plants and for post-depositional diagenesis. Sample sizes 
were too small for stringent cleaning, and the error bars range from 350 to 500 yr. The 
series is in reversed sequence but is internally consistent because of the large error 

11,630 ± 400** (A491) (13,773, 13,722, 13,504 cal) From cranium of slain mammoth 
in bone bed, El Lano dig I, from "grey sand" strata disturbed bv nearbv artesian spring 
chimney. Cultural associations of Clovis artifacts with this mammoth noted but not doc- 
umented with detailed plots, distribution tables, or full illustrations. 

11,170 + 350** (A481) (13,148 cal) From pre-cultural "grey sand" layer ca. 0.5 m 
below bone bed. No cultural associations documented. 

11,040 ± 500** (A490) (13,013 cal) From pre-cultural "grey sand" strata ca. 1.0 m 
below bone bed. No cultural associations documented. 

l^ehiner, AZ 

Weigljted average of 12 dates on 12 samples: 10,950 ± 40jrB.P. (12,980 cal) 

The dates fulfill all but two reliability- criteria and are consistent with each other and 
with reliable dates from other Clovis sites. However, the samples were aggregates of dis- 
persed charcoal, and charcoal is subject to the old-wood problem. 



Lehner is interpreted as a kiU site (Haury et al. 1959). Tlie samples were associated 
with megafaunal bone and artifacts. The sample material is problematic: pooled aggre- 
gates of botanicaUy-unidentified charcoal fragments dispersed in the bone bed. All the 
dates have the old wood charcoal problem. All sample sizes were adequate and all errors 
acceptable. The series is internally consistent and matches high reliability dates from 
other Clovis sites. 

11,470 ±110 (SMU308) (13,444 cal) 

11,170 ± 200 (SMU264) (13,148 cal) 

11,080 + 230 (SMU196) (13,125, 13,083, 13,031 cal) 

11,080 ± 200 (SMU181) (13,125, 13,083, 13,031 cal) 

10,950 ±110 (SMU194) (12,980 cal) 

10,950 ± 90 (SMU290) (12,980 cal) 

10,940 ± 100 (A378) (12,976 cal) 

10,860 ± 280 (SMU164) (12,911 cal) 

10,770 ± 140 (SMU168) (12,880 cal) 

10,710 ± 90 (SMU340) (12,847 cal) 

10,700 ± 150 (SMU297) (12,839 cal) 

10,620 ± 300 (SMU347) (12,807, 12,725, 12,659 cal) 

Murray Springs, AZ 

Weighted average of eight dates on eight samples: 10,880 ± 50jrB.P. (12,923 cal) 

As at Lehner, the large date series fulfills all but two criteria: the preference for indi- 
vidual specimen-samples and freedom from old-wood effects. The series is consistent 
internally and with the reliable date series from other Clovis sites. 

The site was interpreted as a kill site. The dated material was associated with 
megafaunal bone and artifacts. The dated material was pooled aggregates of dispersed, 
botanically unidentified wood charcoal flecks, which present potential old wood prob- 
lems. The samples were associated with artifacts and prey-animal bone in the bone bed. 
Only one of eight dates has too-large error bar. All in the series are the same statistical 
age, consistent with reliable dates from other Clovis sites. 

11,190 ± 180 (SMU18) (13,153 cal) 

11,150 ± 450** (A805) (13,144 cal) 

11,080 ± 180 (Txl413) (13,125, 13,083, 13,031 cal) 

10,930 ± 170 (Txl462) (12,970 cal) 

10,890 ± 180 (SMU27) (12,933 cal) 


ROOSI'A'i:):!, 1)( )UGLAS AND BRO\X/N 

10,840 ± 140 (SM1142 (12,902 cal) 
10,840 ± 70 (SMU41) (12,902 cal) 
10,710 ± 160 (Txl459) (12,847 cal) 

Dent, CO 

Weighted average of six dates on tiro samples: 10J50 ± 40jrB.P. (1 2,873 cal) 

The dates from the Dent, Colorado, site fulfills some of the dating reliabilit}' criteria 
but fails several. Scholars disagree about whether the mammoth represents a human site, 
and the stratigraphic context has given both infinite and recent dates indicating redepo- 
sition and potential old carbon effects. 

The site is a probable kill, but it has been argued that the mammoth may not have 
been associated with the Clovis lithics present. The Clovis-age dates were on two sam- 
ples of mammoth bone. The sample with the oldest date (1622) was too small and gave 
too-large error bars. The other sample was adequate (AA 2941 to 2947) and gave accept- 
able error bars. Geological contamination from coal is documented. The purified bone 
series is consistent internally but falls in the younger range of reliable dates from other 
Clovis sites. 

32,260 + 2100** (SMU 120) lignite from the same matrix as the mammoth bone 

170 + 50 (SMU 121) charcoal from the same matrix as the mammoth bone 

11,200 ± 500** (1622) (13,155 cal) mammoth bone bulk, unpurified organic acids 

10,980 ± 90 (AA2941) (12,992 cal) XAD purified hydrolysate from mammoth bone 

10,660 ± 170 (AA2942) (12,820, 12,709, 12,676 cal) asparric acid from same 

10,800 ±110 (AA2943) (12,889 cal) hydrox^^^roUne from same 

10,600 ± 90 (AA2945) (12,800, 12,731, 12,652 cal) hydrox^-proline from same 

10,710 ± 90 (AA2946) (12,847 cal) glycine from same 

10,670 + 120 (AA2947) (12,824, 12,704, 12,682 cal) alamne from same 

Domeho, OK 

Weighted average of seven dates on foirr samples: 10,944 ± 59jrB.P. (12,978 cal) 

The dates fulfill most criteria, but the samples were small, date all very large, and con- 
taminants were present. However, the series is consistent internally and matches reliable 
date series from other Clovis sites. 

This is a mammoth kill site in a stream. Samples listed here are all that came from the 
bone bed. They were associated with projectile points. The dates were run on bone or 



on noncarbonized wood, the different materials providing a cross-check on dating relia- 
bility. [Some wood samples listed in earlier sources were not from the occupation site.] 
The bone was taxonomically identified but not the wood. The contaminants limestone 
and petroleum were both present in the drainage. Bone A was not purified to individual 
amino acids and so could be affected by contaminants; the other bone sample was puri- 
fied to amino acids considered free from contamination. The bone dates are often 
referred to as evidence for a pre-1 1,200 yr B.P. age of Clovis, but all were on samples 
inadequate in size and thus have very large error bars from 450 to 600 yr. Only the two 
wood samples were adequate size and gave small error bars. (Whether they were indi- 
vidual pieces or if aggregates were dated is not specified.) All the dates are statistically 
consistent and average in the range of reliable dates from other Clovis sites. 

11,220 ± 500** (SI172) (13,159 cal) bulk unpurified organic acids from mammoth 
bone A 

11,200 ± 600** (SI175) (13,155 cal) humic acids from same 

11,480 ± 450** (AA825) (13,448 cal) XAD hydrolysate from different mammoth 

10,860 ± 450** (AA811) (12,911 cal) pro-h>'pro from same 

10,810 ± 420** (AA805) (12,892 cal) XAD hydrolysate from same 

11,010 ± 85 (AA12533) A (13,002 cal) botanically-unidentified wood specimens from 
bone bed (soil stratum C2) 

10,845 ± 90 (AA13028) B (12,904 cal) botanically-unidentified wood specimens from 
bone bed (soil stratum C2) 

hange-Ferguson site, SD 

Weighted average of two dates on two samples: 11 ,100 ± 160 jr B.P. 

The dates have low scores on aU reliabilit)^ criteria but one. Sample materials were 
undesirable: pooled botanically-unidentified dark bits and unpurified mammoth bone 
organic matter. Cleaning procedures were not as rigorous as is possible at present. There 
were only two dates, and one has a too-large error. Samples were not piece plotted, and 
no detailed report published. The only positive criteria are that the dates are statistically 
the same age and fit within the range of highly reliable date series from Clovis sites. 

11,140 ± 140 (AA905) (13,142 cal) pooled aggregates of dispersed, botanically 
unidentified carbon flecks 

10,730 ± 530** (113104) (12,864 cal) mammoth bone bulk, unpurified organic acids 


Rc)()Si-:vi-];r, douglas and brown 

Anf(ick, MT 

Weighted a renioc of three dates on one sample: 10,831 ± 56 yr B.P. 

The five dates fulfill four of seven reliabilitv criteria. The earliest three are internally 
consistent, and the latest two are internally consistent, but the rvvo groups are not con- 
sistent with each other. Clevis archaeologists accept the earliest three, not the later two. 
All the dates have small error-bars, and all were on adequate samples of a human bone. 
The purified amino acids that were dated are not considered subject to carbon contam- 
ination. Nevertheless there was only one bone sample from the relevant burial, and the 
burial was not directly associated with artifacts. The burial site, uncovered during con- 
struction, was not excavated and documented scientifically. Another skeleton at the site 
gave Holocene, not Pleistocene dates. This age difference and the lack of directlv asso- 
ciated diagnostic Clovis cultural materials in a multicomponent site means that the cul- 
ture of the earlier dated skeleton is not established. 

10,940 ± 90 (AA2981) (12,976 cal) glycine from purified human bone 

10,820 ± 100 (AA2979) (12,896 cal) glutamic acid from same 

10,710 ± 100 (AA2980) (12,847 cal) hydrox)'proUne from same 

10,370 ± 130 (AA2982) (12,323 cal) alanine from same 

10,240 ± 120 (AA2978) (11,950, 11,786, 11,783 cal) aspartic acid from same 

Co/by, WY 

Weighted average of tivo dates on two samples: 10,960 ± 120 jr B.P. 

The dates fullfill two of the seven reliability' criteria. The site was a possible kill and 
meat storage site. Dates were on unpurified mammoth bone components and thus sub- 
ject to contamination. They are not the same date statisticallv, a problem since the site 
was not a long-term, sequential occupation. However, the site was scientifically excavat- 
ed and published. Projectile points were present, although their base forms are unusual 
ones for Clovis. Both dates have acceptable error bars and fall in the range of reliable 
dates from Clovis sites. 

11,200 ± 220 (RL392) (13,155 cal) bulk, unpuritled collagen from mammoth bone 
10,864 ± 141 (SMU254) (12,913 cal) apatite from mammoth bone 



UP Mammoth site, WY 

No weighted average because only a single date was available. 

The date only meets one of the seven criteria of reUabilitv^. It is from a possible mam- 
moth-kill site, but the human presence has been disputed by Clovis archaeologists, and 
there were no directly associated human materials. The date was on a bulk, unpurified 
bone sample and thus subject to contaminadon. It has a too-large error bar and is slight- 
ly older than the series of high reliability dates from Clovis sites. 

11,280 ± 350** (1449) (13,174 cal) bulk, unpurified organic acids from mammoth 

Goshen Culture 

Mill Iron, MT 

No iveighted average possible due to statistical inconsistency of associated dates. 

The site fails six of the seven reliability criteria. The site was a kiU site made up of a 
bison bone bed and adjacent campsite. The dates from both contexts were internally 
inconsistent although the material seemed contemporary culturally. The excavator 
reported that there was lignite coal in the deposits, and a dating lab found that one sam- 
ple was made up primarily of the lignite. Although the excavator stated that the carbon 
contamination and statistical inconsistency of the series do not present a problem for the 
chronology, the dates are indeed problematic. They were on aggregations of botanical- 
ly-unidentified dispersed samples. They have one old-wood problem, if they are indeed 
charcoal. The samples have a documented association with a second, much older geo- 
logical carbon contaminant. Lignite, and the lignite was found mixed with charcoal in one 
sample. The associated projectile points show st}^listic and technological features charac- 
teristic of later Paleoindian cultures such as Agate Basin, rather than the early 
Paleoindian Clovis and Folsom cultures. The dates thus do not match the dates expect- 
ed for the culmral material. In the future, the age of the site could be verified by dates 
on purified amino acids from the bison bone in the bed. 

23,720 ± 220 (AA-3668) "charcoal" sample composed predominantly of lignite coal 
from Cretaceous Hell Creek Formation 

11,570 ± 170 (NZA 625) (13,480 cal) bison bone bed 
11,560 ± 920** (NZA 624) (13,477 cal) bison bone bed 



I 1,360 ± 130 (Beta 201 1 1) (13,241 cal) camp, processing area 
1 1,340 ±120 (Beta 13026) (13,258 cal) camp, processing area 
11,320 ± 130 (Beta 16179) (13,279 cal) camp, processing area 
11.010 ± 140 (Beta 16178) (13,002 cal) camp, processing area 
10,990 ± 170 (NZA 623) (12,995 cal) bison bone bed 
10,770 ± 85 (AA 3669) (12,880 cal) bison bone bed 
10,760 ± 130 (Beta 20110) (12,876 cal) camp, processing area 

Culture Undetermined 

A.uhrey, TX 

Weighted average of two dates on two samples: 11 ,570 ± 70jr B.P. (1 3,480 cal) 

The dates fail to meet six of the seven dating criteria. There are only two dates. They 
were run on small samples made up of aggregates of botanicallv-unidentified, dispersed 
carbon bits. The two samples' unit-level stratigraphic sections, specific lithic associations, 
and specific stable carbon isotope measurements have not been published. No fluted 
Clovis points were reported from Trench B, from which the date samples came. The 
samples are from a disturbed geological context consisting of a disconformit}' on an ero- 
sion surface at the interface of Stratum G, a 10,900 to 7,600 yr B.P. carbonate-rich flood- 
plain deposit, with Stratum A, a 25,000 to 15,000 yr B.P. alluvial deposit from stream 
beds cut in Cretaceous bedrock. The samples thus had potential contact with several 
potential older contaminants including carbonate, Cretaceous lignite, peat, and animal 
bone. A Clovis age-geological deposit was not found at the localits', suggesting that the 
cultural component is a later Paleoindian culture deposited into the post-Clovis geolog- 
ical Stramm G. 

11,590 ± 90 (AA-5274) (13,488 cal) 

11,540 ± 110 (AA-5271) (13,470 cal) unidentified carbon aggregates 

*(Damon & Long 1962; Damon ef al. 1964:93-98; Damon et at 1966:100-101; 
Ferring 1989; 1994; Prison 1991:25, 39, 41, 1996; Prison & Todd 1986; Hannus 1990; 
Haury et al. 1959, figure 16; Haynes 1987, 1992, 1997; Haynes et al. 1966, 1967; Haynes 
et al 1998; Humphrey and Perring 1994; Taylor et al. 1996; Leonhardy 1966; Hester et al 
1972:174-176; 225, figure 130; Roosevelt 1998a; Stafford et al 1987; Stafford 1990; 



Stuiver et al. 1998a. Weighted averages calculated by Linda Brown and John Douglas, 
using CALIB.) 

**indicates a too-large error. 

Initial A. and B calibrated radiocarbon and calendar Th and OSl^ 
dates from Pedra Pintada cave, Monte A.legre culture. 

The radiocarbon dates fit aU the reliability criteria, and the TL/OSL dates fit all but 
one. The carbon samples were individual pieces of identified carbonized palm or tree 
firuits, wood charcoal chunks, Hthic artifact fragments, or localized sediment samples. 
(Although Clovis archaeologists have suggested that the initial A fruit seeds could be 
food remains of animals, these seeds are identical in nature to the seeds of the later 
Monte Alegre phases in higher strata, whose dates they accept, and such carbonized 
seeds are judged by Amazon zoologists to be evidence of human, not animal, subsis- 
tence.) Chemical tests for solid and soluble contamination were negative. AU radiocarbon 
errors were within acceptable error ranges, unlike many Clovis dates. All TL/OSL errors 
were larger than acceptable for radiocarbon but t}^ical for these techniques. AU samples 
were from specific, mapped locaUties during excavation. Although Clovis archaeologists 
also suggested that there are too few dates for Initial A, this phase has four radiocarbon 
dates, more than many Clovis sites have. The associated dates on different materials faU 
statistically in the same ranges. The radiocarbon series from each phase is statisticaUy 
consistent internaUy, and the weight averages of the carbon dates of the two phases dif- 
fer significantly. The weight averages of the caUbrated radiocarbon dates and the accept- 
ed calendar TL/OSl dates are statisticaUy consistent. (The earUest TL date, 16,190 yr B.P., 
was rejected for inconsistency with the rest of the series.) The radiocarbon range also is 
consistent statisticaUy with the reUable dates from other eastern BraziUan sites with sim- 
Uar cultural and biological materials.* 


Initial A. 

823 ISW base, object 2 next to hearth 
TL of burned chalcedony bifacial reduction flake (outUer, rejected) (16,190 ± 930 

8346 base, object 1 
TL of burned chalcedony bifacial reduction flake 15,330 ± 900 cal** 


IK )( )si :\'i:i:i, ix )UG1 j\s and brown 

823 IW: hcarrh 
OSL of sediment 13,106 ± 1628 cal** 

823 ISE hearth, object 18 

OSL of heavy fraction of sediment 12,536 ± 4125 cal** 

8231CE general level 
OSL of sediment 12,491 ± 1409 cal** 

**Weighted average of four dates on four saniples: 13,180 ± 509 calendar yr B.P. 


Initial B 

8231 far SW; top, object 1 
TL of burned brecciated quartz flake 11,880 ± 760 cal** 

Initial A. 

8314CW, base, hearth, object 4 
carbonized sacuri endocarp 11,145 ± 135 (GX-17413) 13,143 cal 

8231 W base, hearth, object 17 
carbonized tucuma endocarp 11,110 + 310 (GX-17406) 13,135 cal 

8231 SE base, hearth, object 18 
carbonized tucuma endocarp 10,905 ± 295 (GX-17407) 12,950 cal 

8314C base, object 5 
carbonized tucuma endocarp 10,875 ± 295 (GX-17414) 12,920 cal 

Weighted average of four dates on four samples: 11,075 ± 106 13,123, 13,088, 13,02^ 

Initial B 

carbonized jutai seed 10,683 ± 80 (NZA9898) 12,954-12,681 cal 



8345 object 4 

carbonized tucuma endocarp 10,655 ± 285 (GX— 17420) 12,712 cal 

8231CW top, object 
carbonized sacuri endocarp 10,560 ± 60 (B6952CAMS) 12,745 cal 

8231 CW top, object 
wood charcoal fragment 10,450 + 60 (B76952CAMS) 12,498 cal 

9274 object 13 
humate from GX-19537CAMS 10,470 ± 70 (GX-19537CAMS) 12,387 cal 

9274 object 5 
carbonized sacuri endocarp 10,410 + 60 (GX-19538CAMS) 12,531 cal 

8314SW top, object 6 
carbonized sacuri endocarp 10,392 + 78 (GX-1 7400CAMS) 12,333 cal 

9274 object 5 
humate from GX-17400CAMS 10,390 ± 70 (GX-19538CAMS) 12,332 cal 

9274 object 6 
humate from sacuri endocarp 10,360 ± 70 (GX-19536CAMS) 12,203 cal 

8346 object 1 

carbonized sacuri endocarp 10,305 ± 275 (GX-1 7422) 12,241 cal 

8345 object 13 

carbonized sacuri endocarp 10,275 ± 275 (GX-17421) 12,016 cal 

carbonized achua endocarp 10,261 ± 62 (NZA9897) 12,355-11,808 cal 

9274 object 13 
carbonized sacuri endocarp 10,250 ± 70 (GX-19537CAMS) 12,052 cal 

Weighted average of 12 dates on 11 samples: 10,420 ± 23 yr B.P. 12,582, 12,522, 
12433 cal 


Initial A. 

8231 SW base, object 2 next to hearth 
TL of burned chalcedony bifacial reduction flake (outlier, rejected) (16,190 ± 930 cal) 

8346 base, object 1 

TL of burned chalcedony bifacial reduction flake 15,330 ± 900 cal 

8231W, hearth 
OSL of sediment 13,106 ± 1628 cal 

823 ISE hearth, object 18 


K( )()Si<:vi:]:r, douglas and brown 

OSL of heavy fraction of srclimcnt 12,536 ± 4125 cal 

<S231CE general level 
OSi, of sediment 12,491 ± 1409 cal 

Weighted average of five: 13,180 ± 509 calendar \r B.P. 

Initial B 

8231 far SW, top, object 1 
TL of burned brecciated quartz flake 11,880 ± 760 cal 

Initial y^ 

8314CW, base, hearth, object 4 
carbonized sacuri endocarp 11,145 ± 135 (GX— 17413) 13,143 cal 

8231 W base, heardi, object 17 
carbonized tucuma endocarp 11,110 ± 310 (GX— 17406) 13,135 cal 

8231 SE base, hearth, object 18 
carbonized tucuma endocarp 10,905 ± 295 (GX-17407) 12,950 cal 

8314C base, object 5 
carbonized tucuma endocarp 10,875 ± 295 (GX-17414) 12,920 cal 

Weighted average of four: 11,075 ± 106 13,123, 13,088, 13,028 cal 

Initial B 

carbonized jutai seed 10,683 ± 80 (NZA9898) 12,954-12,681 cal 

8345 object 4 
carbonized tucuma endocarp 10,655 ± 285 (GX-17420) 12,712 cal 

8231 CW top, object 
carbonized sacuri endocarp 10,560 ± 60 (B6952CAMS) 12,745 cal 

8231 CW^ top, object 
wood charcoal fragment 10,450 ± 60 (B76952CAMS) 12,498 cal 



9274 object 13 
humate from GX-19537CAMS 10,470 ± 70 (GX-19537CAMS) 12,387 cal 

9274 object 5 
carbonized sacuri endocarp 10,410 ± 60 (GX-19538CAMS) 12,531 cal 

8314SW top, object 6 
carbonized sacuri endocarp 10,392 + 78 (GX-1 7400C AMS) 12,333 cal 

9274 object 5 
humate from same 10,390 ± 70 (GX-19538CAMS) 12,332 cal 

9274 object 6 
humate from sacuri endocarp 10,360 ± 70 (GX19536CAMS) 12,203 cal 

8346 object 1 
carbonized sacuri endocarp 10,305 ± 275 (GX-1 7422) 12,241 cal 

8345 object 13 
carbonized sacuri endocarp 10,275 ± 275 (GX-17421) 12,016 cal 

carbonized achua endocarp 10,261 ± 62 (NZA9897) 12,355-11,808 cal 

9274 object 13 
carbonized sacuri endocarp 10,250 ± 70 (GX-1 9 5 37C AMS) 12,052 cal 

Weighted average of 11: 10,420 ± 23 yr B.P. 12,582, 12,522, 12433 cal 

*Haynes 1997; Hemmings 1970; Pessis 1999; Prous 1980-81, 1986a and b, 1991, 
1995, 1999; Reanier 1997; Roosevelt 1998a, b, 2000a, 2000d, 1999; Roosevelt et al. 1996; 
Roosevelt et al. 1997. 

**indicates a too-large error 


Chapter Eight 

Plant Food and its Implications for the 
Peopling of the New World: 
A View from South America 

Tom D. Dillehay and Jack Rossen 

^^^^^ In recent years, it has become clear that without the important adap- 
^^^^^^k tations that took place among Late Pleistocene foragers in many 
^ ^ ^^ ^B^^^^B parts of the Americas, the subsequent development of more com- 
^*^K|^^P plex human societies and food production would have been impos- 
sible. Whatever the causes and wherever the centers of plant domes- 
tication, these developments had their roots in decisions made by Late Pleistocene for- 
agers to settle in forested and wetiand environments, for instance, and to restructure their 
subsistence patterns to exploit more plant foods, transport them to new environments, 
and perhaps to consciously manipulate them to the benefit of people. Although early 
foragers probably always procured wild plants, the most important period may be in 
post-glacial times, between 13,000 and 10,000 years ago depending upon geography, 
when different climatic conditions, vegetation, and animal life were established in the 

Due to a series of recent archaeological discoveries in North America and South 
America, many aspects of the initial peopling of the New World are under closer scruti- 
ny New data from early sites in Alaska, Pennsylvania, South Carolina, and Virginia in the 
north and Argentina, Brazil, Chile, and Peru in the south offer challenging fresh insights 
on when people first came to the New World, where they came from, and the kinds of 
technologies and economies they practiced. Until recently, the study of Pleistocene 
archeology focused primarily on tracking the movement of hunters (mainly Clovis peo- 
ple) as food resources were depleted due to climatic change, human population growth, 
and/or overexploitation. To put it simply, most archeologists studied unchanging hunters 
in changing environments {rf., Martin 1973; Haynes 1969; Pagan 1987; Carlisle 1988; 
Meltzer 1993, 1997; Dillehay et al. 1993), and saw technological and economic adjust- 
ments primarily as adaptive reactions to climatic changes. No doubt, these changes cre- 
ated new opportunities and imposed new constraints on, and opportunities for, humans 



^^()sr exciting in recent years has Ijeen a greater recognition of widespread techno 
logical and economic diversity throughout the Americas. People not only hunted, but 
thev also exploited a wide array ot plant toods and, in coastal areas, marine foods. This 
dix'ersitv probably relates to a greater time depth of humans in the New Workl than was 
previously recognized and thus more ancient opportunities to develop different k^cal tra- 
ditions and more cultural complexity'. We know that cultural diversit)' in South America 
was widespread by at least 10,500 yr B.P. (Bryan 1973, 1978, 1986; Ardila & PoHtis 1989; 
Dillehay d al. 1993; Dillehay 1999). 

Archaeologists have identified big-game hunting as a fundamental subsistence actdv- 
it}' of many first Americans, because early sites contain stone tools (especially projectile 
points) and the bone remains of extinct animals. Organic remains other than animal 
bones have not been preserved in sites or have not been found by archeologists due to 
inadequate recovery techniques. For instance, plant remains are easily overlooked if 
water flotation is not conducted, but they appear in surprising quantit}" and diversit)' 
when flotation is included in the excavation and analysis strategy (Dillehay 1989; Rossen 
et al. 1996; Dillehay et al. 1997; Ramirez 1989). As a result, archeologists have underesti- 
mated the role of plant foods in early diets, especially in wetiand and wooded environ- 
ments where a high density and diversity of plant foods are usually found {e.g., 
Richardson 1978; Meltzer & Smith 1986; Dillehay 1986; Meltzer 1993). Examples are the 
Monte Verde site in the cool, temperate rainforest of southern Chile, the Las Vegas site 
in southeastern Ecuador, sites on the north coast of Peru, and several rockshelter sites 
in the tropical forests of eastern Brazil (Figure 1). Other examples may be at Guitarrero 
Cave (Lynch 1981) and Tres Ventanas (Engel 1970) in the Peruvian liighlands and other 
rockshelter sites in the Andes. At these sites, in addition to animal remains, a wide vari- 
ety' of culturally selected plant parts were recovered through a combination of good 
preservation and the application of a specialized water-flotation technique. The preser- 
vation and technique permitted the use of direct evidence of food and non-food items 
to reconstruct the resource base of the paleoenvironment of the site's inhabitants, and 
to infer the diet, residential status and social organization. It was, after all, human 
exploitation of plants in Late Pleistocene times that set the stage for later plant domes- 
tication in many areas of the New World. As we employ more sopliisticated plant recov- 
ery techniques and find new sites in different environments, it will become clearer that 
there are a wide range of generalized stone tool industries (particularly unifacial indus- 
tries) and economies in early sites throughout the New World (Bryan 1973; Dillehay et al. 
1993; Dillehay 1999). We expect to find these sites in "food affluent" environments that 
contained a great bulk and high diversit}' of usable plant biomass, such as wetiands and 
forests. Within wetlands, whether they are swamp, mangroves, marsh, ten or bog, salt- 



Figure 1. Of South America showing countries, major river sj^stems, and some areas discussed in 
the text: 1) Monte Verde site in Chile; 2) Talara area on the north coast of Peru; 3) Las Vegas site in 
southwest Ecuador; 4) savanna and forested plains in highland Colombia and 5) eastern Brazil; and 6) 
Nanchoc area on the north coast of Peru. 


[)ll.l.l•.ll.\^• AM) ROSSI'lN 

water or freshwater, lives a wide array of wildlife unseen in other environments. 
Ethnographers and archaeologists have observed that most semisedentary to sedentary 
hunter-gatherer groups with broad-spectrum economies are found in lush temperate and 
boreal environments (Winterhalder & Smith 1981; Binford 1980). 

Of special interest here is the archaeological record of early hunter-gatherers in the 
wetlands and forested areas of South America. Large campsites of the early to middle 
Archaic periods containing direct or indirect evidence of a broad subsistence base and 
long-term use are often found in these temperate environments, as is the case also in 
mid-northern Europe (Butzer 1971). In much of the eastern and midwestern United 
States, wedands were attractive "pull" zones that reduced early Archaic group mobilit\- 
(Brown & Vierra 1983; Collins & Driskell 1979:1036; Lopinot 1982; Nance 1988; Rossen 
1998a). Case studies in those regions emphasize the broad exploitation of various wet- 
lands microniches such as shorelines, shallow water, and deep water zones. Although lit- 
tie research has been carried out in the temperate zones of South America, ecological 
and archaeological patterns of plant-oriented subsistence are emerging at more Late 
Pleistocene and early Holocene sites in Chile, Peru, and Brazil. 

Although the remains of nuts, berries, stems, fruits, and other edible parts have been 
recovered from several early sites, most noteworthy here are below-ground plants or 
plant parts {e.g., mbers, reeds and sedges; Hatiey & Kappelman 1980; Coursey Sc Booth 
1977; see Ramirez 1989 and Rossen & Ramirez 1997 for southern Chile) and their broad- 
er implications for economic viabilit}'. These parts are special because they usually are 
less affected by drought, fire, seasonal flucmations, and grazing than above-ground 
plants, and they are exploited by fewer competing predators. They also store naturally for 
long periods of time that allow greater residential stability. 

Summarized below are a few case studies discussing the general characteristics of the 
paleoenvironment and resource base of some of the best documented Late Pleistocene 
and early Archaic sites, the archeological evidence for a generalized subsistence econo- 
my, and the implications of these sites for studying later social and culmral develop- 
ments. These cases are not intended to represent all early time periods and regional envi- 

Monte l^erde in South-Central Chile 

The site of Monte Verde is located on a high terrace plain along the banks of a small 
tributary of the Maullin River in southern Chile (Figure 1). Today, the cool, temperate 
landscape of Alonte Verde is characterized by varied surface feamres: rainforests of 
mixed coniferous and broadleaf trees, rolling hills and low and high mountain ranges. 



large fertile river plains and river valleys, expansive wetlands and grasslands, freshwater 
and saltwater marshes, lagoons and bogs, and coastal plains. Cool and wet conditions 
similar to the present-day rainforest environment prevailed in the region some 16,000 to 
12,000 years ago (Dillehay 1984, 1986, 1989). 

Two cultural components have been identified at Monte Verde. The youngest com- 
ponent is well-preserved by an overlying peat layer, and is radiocarbon dated around 
12,500 yr B.P. The oldest component is equivocal, lies in a different area of the site, and 
possibly dates as far back as 33,000 yr B.P (DiUehay & ColUns 1988; Dillehay 1997). Only 
the cultural materials from the younger, better known 12,500 yr B.P. component are con- 
sidered here. 

Several characteristics of this component are important. Wooden foundations of a 
long tent-Uke structure were excavated. The architectural base of an isolated wishbone- 
shaped structure also was found in a specialized, nonresidential sector of the site. The 
tent-Hke and wishbone-shaped structures appear to be components of a planned settle- 
ment that was integrated both spatially and functionally. 

A diversified stone tool industry was made up of three basic technologies: selective- 
ly collected, naturally fractured, and culturally utilized stones with one or more sharp, 
usable edges; unifacially and bifacially flaked tools; and pecked-ground bola and grind- 
ing stones. These three simultaneously utilized technologies indicate a diversified econo- 
my of plant gathering and hunting. Residue material appears on the working edges of 
many Uthic tools. Comparison with microscopic cellular fragments of living plants indi- 
cates that seaweeds and aquatic plants were processed regularly. Several tools were also 
used for artisanal or architectural maintenance (hut construction). Other lithic specimens 
exhibit wood carbon on the edges indicating that charred wood was worked. These tools 
were recovered alongside burned and worked wood in all areas of the site. There are also 
several planing and gouging tools, grinding stones and wooden mortars that suggest a 
plant-based economy. In sum, there is plenty of nonplant evidence indicative of the 
exploitation and preparation of wetland and forest plant species. 

The plant-oriented unifacial tool industry at Monte Verde is not unusual. Most syn- 
thesizers of early South American prehistory have recognized an early unifacial flake or 
edge-trimmed tool tradition which was probably associated with working wood, skin, 
plants, bone, fiber, shell, and feathers (Bryan 1978, 1986; Richardson 1978; Gruhn & 
Bryan 1998; Dillehay 2000). Although a wide variety of bifacial industries occur across 
the continent as well, the unifacial industries are most often associated with late glacial 
and post-glacial savanna-parkland and thorn forests in eastern Brazil, with savanna and 
forested plains in Colombia, mangrove swamps and arid savanna-woodlands in coastal 
Ecuador and possibly the Talara area on the north coast of Peru (Figure 1), and with 



cool iL-mpcratL' rainforests in southern (^hilc. Irrespective of location, all early sites char- 
acterized by unifacial industries and by good preservation of organic faunal and floral 
remains are associated with a broad-spectrum foraging econom\'. 

Hunting and/or scavenging game is suggested by the bone remains of at least seven 
intli\idual mastodons, one paleocamelid and small animals. Meat probably contributed a 
significant proportion of the protein, but the supply may not have been regular and vari- 
ation in amount of meat between individually hunted or scavenged animals would have 
been a factor. Hunting or scavenging for small- and medium-sized animals was probably 
not of nutritional significance but other categories of food sources might have been use- 

A well-preserved collection of plant remains {i.e., seeds, pods, leaves, flowers, pollen, 
fruits, nuts, tubers, and stems) and wooden artifacts suggest much about the past econ- 
omy, technology, and ecology. The wood assemblage confirms the long-held idea that 
plant and other organic materials played important technological roles in Late 
Pleistocene culmres (Kreiger 1964; Bryan 1978, 1986; Richardson 1978). These materi- 
als also provide detailed data for reconstructing the economic catchment zone of the 
site's inhabitants and the general configuration of the local ecology, which revealed that 
the Monte Verdeans were gathering resources up and dow^n the river system from the 
highlands to the coast. 

The macrobotanical remains include more than sixt)^ plant species that sequentially 
mamre during all seasons of the year in different ecological zones (Dillehav 1984, 1986; 
Ramirez 1989; Rossen & Ramirez 1997). Grains, seeds, nuts, mber and rhizome skins, 
tubers, bulbs, roots, and soft leafy material recovered from food pits and hearths inside 
hut structures and from cracks and pits in wooden mortars have }ielded information 
both on the species of above-ground and below-ground plants that were used by the 
Monte Verdeans and on the general namre of the diet. 

Among the archaeologically recovered plants, a few species are staple plants that are 
available vear-round, and have edible greens, seeds, rhizomes or mbers. These species 
mchidc jitnais procerus, Scirpus californiciis, Carex sp., Gunnera sp., and Solaniim maglia (Rossen 
& Ramirez 1997:333—338). Numerous other recovered species represent seasonal foods, 
especially summer fruits, that supplemented the basic plant foods. Certain plants, includ- 
ing a white mushroom [Clavaria coralloides), a bamboo {Cbnsqiiea sp.), and four seaweeds 
{Durvillaea antarctica, Gradlaria sp., Porpbyra sp., and Sargassitm sp.; Figures 2 & 3) have spe- 
cific seasonalities to suggest they were utilized during the non-summer months when 
overall plant food availabilit)- was lower. Ten species of halophytes, native to the coastal 
salt marshes or sand dunes 55 km west of Monte Verde, with high salt content suggest 
a focus on salt as a key resource (Rossen & Ramirez 1997:344). Based on their use today 



by Mapuche shamans, it is likely that ten recovered species had exclusive medicinal uses 
and, along with thirteen edible species with corollary medicinal uses, appear to represent 
a versatile pharmacopoeia related to the treatment of a wide variety of ailments (Rossen 
& Ramirez 1997:339—342). A noteworthy plant remain recovered in great frequency is 
the spores of club moss (Lycopodi/tm sp.), which seem to have been used as a fire starter. 
From many perspectives, plant use at Monte Verde was varied and complex, involving an 
intimate knowledge of local and nonlocal microenvironments, distributions, seasonal 
availabilities, and botanical attributes. 

Within the broad spectrum of plant use displayed archaeologically at Monte Verde, 
perhaps the most interesting of this group are tubers. One, Solanum, is grown by the pre- 
sent-day indigenous Mapuche and Huilliche groups. Moreover, there are at least three- 
hundred other species of wild potatoes available in the area today, most of which have 
several economic uses (c£ Bukasov 1930; Correl 1962; Vavilov 1951; Ugent et al. 1986; 
Ramirez, personal communication, 1983; personal observations). It is likely that some of 
these species were also exploited by the Monte Verdeans but were not recovered archae- 

Tubers and rhizomes would have provided a year-round food supply for the site's 
inhabitants. When the underground parts of these plants are left or stored in their nat- 
ural underground habitat, they are foods with a long "use-life" {sensii Labuza 1962; Soffer 
1989) and banking capacity. (The Mapuche and Huilliche extend the use-life of potatoes 
from 2—6 years by leaving them in the ground or storing them in deep pits in wet terrain 
[Alejandro Herrera, personal communication, 1989; personal observation]. If collected 
and stored culturally, these parts rot in less than five months.) 

These plants could have provided a majority of the Monte Verdean's daily intake of 
protein, carbohydrates, starch, minerals, and other elements, particularly during the win- 
ter months when bog fruits were unavailable. To give an example of the nutritional value 
of protein, Solanum maglia tubers yield about 5.2 gm per 100 ^pa,]iinciis procerus seeds ren- 
der 7.3 gm per 100 gm, and Scirpus califormciis roots give 7.2 gm per 100 gm (Laboratorio 
Fitoquimica, 1983). Figures for the other elements are also moderately high. 

Tubers and rhizomes are relatively easy to procure. The indigenous Huilliche living 
in the area today use digging sticks made of tree branches or cow ribs to collect them. 
Both modified wooden sticks and mastodon ribs, characterized by blunted and/or 
burned edges, have been recovered from Monte Verde (Dillehay 1997). Micro-use wear 
smdies of these and replicative digging sticks, suggest that the site's inhabitants were 
probably using these implements to extract underground food parts. 

Based on the reconstruction of past and present resource structures in the smdy area, 
we can infer that the Monte Verdeans chose two complementary ways to procure mbers. 


DliJ.i:ilA^- AND ROSSJ'N 


Figure 2. A masticated cud recovered from the floor of the wishbone-shaped structure dated to 
ca. 12,500 yr B.P. The cud is comprised of fragments of boldo leaves and of three different t\pes of 
seaweed, all of which are medicinal elements. The cud is presumed to have been chewed by the site's 
inhabitants and tossed on the floor. 

rhizomes, and other foods. One way was to exploit the patchiness of a primary set of 
resource zones in the Maullin River drainage. The archaeological presence of a wide 
range of inorganic and organic resources from multiple environmental zones in the river 
basin suggests a linear catchment zone (see Dillehay 1989; Ramirez 1989; Rossen & 
Dillehay 1997), stretching from the littoral in the west to the Andean grasslands to the 
east. Several materials, including discoidal abraded stones, salt, clay, quartzite and other 
stones, four varieties of seaweed, petrified plant stalks from the coast and plants such as 
the aforementioned club moss from the liighlands attest to the exploitation of resources 
from distant poles of the river basin (Dillehay 1984, 1986, 1989; Dillehay & Rossen 1997; 
Rossen & Dillehay 1997). This pattern suggests that the site inhabitants regulated the dis- 
persion, supply, and preservative properties of selected seasonally available plant species. 
The other possible procurement technique was to target certain wetiand habitats where 
an abundant year-round supply of below-ground mbers and rhizomes were concentrat- 
ed. The most productive and accessible wetiands were on the upland terraces where 
numerous bogs, swamps, and lagoons were located. Furthermore, seasonal fluctuation of 
the water table periodically reduces access to certain wetiands zones, allowing economic 



plant populations to rejuvenate and reducing the potential for resource depletion. If the 
Monte Verdeans targeted such areas for exploitation, it might have been to their advan- 
tage to settle in for a year or so to take advantage of the abundance and the diversity of 
wetland resources. 

Figure 3. A desiccated bulb of a wild potato {Solarium maglid) recovered from the floor of the long 
tentUke structure dated at 12,500 yr B.P. at Monte Verde. 

l^as Vegas in Coastal Ecuador 

Another diversified economy is demonstrated at the Las Vegas site in southwest 
coastal Ecuador (Stothert 1998). Although people may have lived at die site around 
10,000 years ago, a later component dated between 9,700 and 7,000 yr B.P. is associated 
with phytoliths of domesticated squash, gourds, maize and edible roots. The early Las 



Vegas people also exploited deer, small animals, peccary, reptiles, and birds in adjacent 
tropical forests and mangrove swamps, as well as fish and shellfish from the sea. 

The simple stone tool industry demonstrates increasing skill in tool kits, and the 
greater significance of plant-coUecdng implements, including small mortar stones. 
People who viewed almost all living organisms as potendal sources of food and were 
willing to organize their acdvides to collect them, would have been developing an aware- 
ness and organizational system potendally receptive to the collection and eventual 
domesdcadon of wild plant foods. Heavier ground stone tools first appear in sites. These 
may have been used to grind minerals, but their uses became familiar with the charac- 
teristics and potential of ground stone implements that later could be adapted to grind- 
ing plant food. 

Nancboc Culture in Northern Peru 

Early Holocene sites throughout northern Peru are known for their diversified 
economies. In particular, coastal sites throughout Peru exhibit an emphasis on various 
maritime and shoreline resources, and incidental recovery of wild and perhaps cultivat- 
ed plants (Richardson 1978). Inland coastal valley sites from this time have been less 
studied, but a series of sites from the upper Zana Valley dating from 8,500 to 6,000 yr 
B.P. exhibit a plant-oriented economy that may include a precocious suite of semi- 
domesticates or cultigens from various altitudes and environments (Rossen et al. 1996). 
These plants include squash, peanuts, quinoa, and Solanaceous and cactus fruits. It 
appears that plants were introduced to middle and upper valley ecotones from highland 
and tropical forest settings, and that plant experimentation was an economic foundation 
and an impetus of culmre change. These sites are among very few in Peru where sys- 
tematic water flotation has been conducted, and direct A. M.S. dates have been inconsis- 
tent. There can be litde doubt, however, on the basis of varied, well-made, exclusively 
unifacial chipped lithic and ground stone industries, that early Holocene economies in 
the region were heavily plant-oriented (Dillehay etal. 1997; Rossen 1998b). Faunal assem- 
blages are dominated by mollusks and small game. 

Several Sites in Central Brazil 

Almost thirt}' Late Pleistocene and early Holocene sites ranging in radiocarbon dates 
from 12,300 to 8,000 years ago have been reported from central Brazil (Schmitz 1987; 
Prous 1996; Kipnis 1998). These sites display diverse economies based on seeds and 
palm nuts (sometimes filling entire pit features and midden lenses), along with wUd fruits. 
Perishable wooden and reed artifacts are present in at least three sites (Kipnis 1998:587). 



The importance of wild plant collecting continues even much later, when maize and 
manioc cultivation was locally introduced. Faunal remains too are diverse, emphasizing 
small game and birds. Even without the use of specialized plant recovery techniques by 
archaeologists, scholars specifically draw a dichotomy between Late Pleistocene plant 
gathering cultures of central Brazil and the "big-game hunting model" that remains pre- 
eminent in North America (Kipnis 1998:590). 

In sum, we suspect that for a few early societies, subsistence pursuits may occasion- 
ally have been concentrated on large animal species as the primary source of meat; for 
the sites discussed above and surely for most early societies, the variety of foods con- 
sumed was broadened to include small animals, snails, waterfowl, fish, mussels, and 
plants. Certain societies, such as the Las Vegas and perhaps the Nanchoc communities, 
which existed at the end of the Pleistocene, possessed traits preadapted to horticulture. 
Those that relied on hunting and gathering for subsistence began to experiment with the 
plants and animals being used as resources. It is likely that the early inhabitants of many 
areas also collected plant material, but their remains have not been found primarily 
because most sites have not been subjected to detailed plant recovery techniques. 

Foraging, Food Storage, Collecting, and Residential Stability 

Several investigators have noted that the absence of storage facilities in hunter-gath- 
erer sites may suggest that site inhabitants were foragers. A central feature of a foraging 
society is that economic activities relate primarily to the immediate needs of the group, 
and not to the anticipated needs of the future. Thus, long-term storage is not considered 
to be a primary aspect of foraging organization (Binford 1978, 1980). It is, however, col- 
lectors who engage in logistical activities in which a task group collects resources and 
brings them back to a base camp. In this case, storage is vital for stockpiling surplus over 
the next several months (Flannery 1969; Binford 1978, 1987; Redman 1977). Exceptions 
to these generalizations and to storage facilities can be foraging hunter-gatherer groups 
in lower latitudes zones. Low latitude zones often contain a high density of foods that 
permit foragers to gather resources each day on an "encounter" basis (Binford 1980). 

Our model of hunting and gathering at Monte Verde, for example, includes both 
"encounter" foraging and "logistical" collecting, depending on the season, environmen- 
tal zone, and combination of resources present (Rossen & Dillehay 1997). In this area, a 
compact mosaic of economic plant microzones produced an unusually varied econom- 
ic plant inventory, along with prolonged seasonal plant availabilities as individual plant 
species matured differentially in different plant microzones (Rossen & Ramirez 1997). 
The logistical collection of dense, patchy resources like tubers and rhizomes, however. 



probably produced the staple foundation of the diet. Yet, there arc no identifiable stor- 
age facilides at Monte Verde. The Monte Verdeans were probably sedentary nonstorers 
who relied on the underground storage and nutridonal banking of certain tubers and rhi- 
zomes with long use-lives and on the occasional exploitadon of big-game animals to 
remain in one location for a long period of the year. By relying on a primary set of plant 
and animal food resources on a daily encounter basis in nearby bogs and marshes, the 
Monte Verdeans did not necessarily need to actively store food for the lean months. 
Supplemental food resources, such as nuts, soft leafy plants, mushrooms, seaweeds, 
frmts, and small game could have been gathered in these same environments. 

There are other significant implications of naturally stored foods. Natural storage 
would not have necessitated changes in the labor, mobilit)'^, technology, and social struc- 
mre of the resident population, changes that might have led to the construction and 
maintenance of storage features in the site. And it would not have forced the Monte 
Verdeans to have mobilized labor during limited seasonal periods. It also would have 
anchored the Monte Verdeans to a particular location, probably limiting their resource 
mobility, and, in some cases, leading to periodic longer stays at the site, depending on cir- 
cumstances from year to year. 

Those who laid the foundation for the human presence at Monte Verde remain 
unknown. The cause and nature of these developments are difficult to discern, because 
we know nothing about what preceded or succeeded it. We can assume that a transition 
occurred from nomadic foragers to a mixed strategy of plant collecting, foraging, scav- 
enging, and hunting. We cannot assume, however, that in the rich temperate wedands of 
south-central Chile animal meat was the preferred food, that a shortage of animal pro- 
tein was a limiting factor on the size, distribution, and permanence of the site, or that 
any intense exploitation of plant food was a shift to less desirable foods. In following 
Cohen's (1977) now familiar argument, this shift would be expected if the human popu- 
lation had expanded to the point where it could no longer be sufficientiy supported by 
available game. Although the setdement pattern data are lacking, it is highly unlikely, in 
our opinion, that such pressure existed in Late Pleistocene times. It is our guess that, in 
die case of Monte Verde, the generalized economy and setdement stabilit)' was primari- 
ly founded on the high diversit}^ and densit}' and year-round avaUabiHt}' of seeds, nuts, 
fruits, and the nutrient-rich, below-ground, plant foods and their large patchy distribu- 
tion in space. These resources, coupled with animal meat, would have provided a high 
quaUt}' of minerals and carbohydrates, starch, trace elements, and plant and animal pro- 




Undoubtedly, new discoveries and directions in the interpretation of initial human 
settiement in unexplored wetland and wooded environments will clarify our understand- 
ing of the role of early technologies and different plant foods in subsistence economies. 
Future research will undoubtedly lead to new inquiries about the cultural transition 
between Late Pleistocene and early Archaic populations in the Americas. To date, the 
Late Pleistocene prehistory of one region — the cool, temperate rainforest of southern 
Chile — is basically the story of one archaeological site, Monte Verde. The inhabitants 
of Monte Verde had a generalized economy, characterized by a primary set of staple 
plant (seeds, fruits, tubers and rhizomes) and animal (mastodon and paleollama) 
resources. These people generally had a low level of stone tool technological develop- 
ment, but apparentiy a high level of organization to meet their economic and other 
needs. Their exploitation patterns included both encounter foraging and logistical col- 
lecting of plants and animals, yet it appears that logistical collecting of staple plants with 
long availabilities was central to their local permanence. Similar patterns are also being 
recorded at sites in eastern BrazH, northern Peru, and southern Ecuador. Future study 
surely will find them in many other forested and wetiand environments and document 
further the presence of early broad-spectrum economies. 

Finally, to more fully understand the first peopling of the Americas, we must devel- 
op research questions and strategies to study early adaptive processes on a comparative 
local and hemispherical basis that can lead to significant insights into the plasticity of 
Late Pleistocene cultures. With more research we may see that early cultures were far 
from being uniform and static, and were individually, subculmrally, and temporally vari- 
able. We also need to search for new models, construct new scenarios that better reflect 
what we know about ethnographic hunter-gatherers, and develop new methodological 
strategies of excavation and analysis that include water flotation and other techniques 
designed to recover minute plant remains. The archaeological record may be fragmen- 
tary, but it is more so if we do not bother to collect what is there. 


Due to preservation and paradigm bias, plant collecting is generally given lit- 
tle attention in modeling the adaptation of early foragers in the Americas. We 
discuss the social and economic implications of early archeological records where 
plant preservation is good, such as Monte Verde in Chile and several sites in 
Bra:(^il, and their meaning with respect to the exploitation of wetland environ- 



nnnits and the pcoplinQ^ of the New World. We demonstrate that reliance on a 
large number and type of plant resources typified the earliest human economies of 
South America. 

l^iterature Cited 

Ardila, G., & G. Politis. 1989. Nuevos datos para un viejo problema: investigacion y dis- 

cusion en tornodel poblamiento de America del Sur. Avista delMuseo delOro 23:3^5. 

Bogota, Colombia. 
Binford, L. 1 978. Dimensional analysis of behavior and site structure: Learning from an 

Eskimo hunting stand. Am. Antiquity 43:330—361. 
. 1980. Willow smoke and dog's tails: Hunter-gatherer setdement systems and 

archaeological site formation. Am. Antiquity 45:4—20. 
. 1987. Researching ambiguity: Frames of reference and site structure. Pages 

418—456 in S. Kent, ed.. Method and Theory for Activity Area ^search. Columbia 
University Press, New York, NY. 

Brown, J. A., & R. K. Vierra. 1983. What happened during the Middle Archaic?: 
Introduction to an ecological approach to Koster site archaeology. Pages 165—195. in 
J. L. Philips & J. A. Brown, eds.. Archaic Hunters and Gatherers in the American Midwest. 
Academic Press, New York, NY. 

Bryan, A. 1973. Paleoenvironments and culmral diversit}' in Late Pleistocene South 
Kvci^nz2i. Quat. R^j-. 3:237—256. 

. 1978. An overview of Paleo-American prehistory from a circum-Pacific per- 
spective. Pages 306—327 in A. L. Bryan, ed., Harlj Man in America from a Circum-Pacific 
Perspective. Occasional Papers No. 1, Department of Anthropolog}; Universits' of 
Alberta, Edmonton, Canada. 

. 1986. Paleoamerican prehistory as seen from South America. Pages 1—14 /;/ 

A. L. Bryan, ed.. New Evidence for the Pleistocene Peopling of the Americas. Center for the 

Study of Early Man. Univesit}^ of Maine, Orono, ME. 
Bukasov, I. 1930. The cultivated plants of Mexico, Guatemala, and Colombia. Bull Appl 

Bot. Genet Pl.-Breed (Trudy Prikl Bot Suppl.j 47:513-523. 
Butzer, K. 1971. Environment and Archaeology. 2nd Edition. Aldine Press, Chicago, IL. 
Carlisle, R. \9^?). Americas Before Columbus: Ice Age Origins. Ethnology Monographs No. 12. 

University of Pittsburgh, Pittsburgh, PA. 
Cohen, M. 1977. The Food Crises in Prehistory: Overpopulation and the Origins of Agriculture. 

Yale University Press, New Haven, CT. 



Collins, M. B., & B. N Driskell. 1979. Summary and conclusions. Pages 1023-1042 in M. 

B. Collins, ed., Excavations at Your Archaic Sites in the Ijower Ohio Valley, Jefferson County, 

Kentucky. Occasional Papers in Anthropology No. 1, University of Kentucky, 

Lexington, KY. 
Correll, D. S. 1962. The Potato and its Wild Relatives. Texas Research Foundation. Renner, 

TX. 606 pp. 
Coursey, D. G., & R.. H. Booth. 1977. Root and tuber crops. Pages 23^4 in C.L. Leakey, 

ed.. Food Crops of the 'Loivland Tropics. Oxford University Press, Oxford, England. 
Dillehay, T. 1984. A late Ice Age settlement in Southern Chile. Sci. Amer 251:106-117. 
. 1986. The cultural relationship of Monte Verde: A Late Pleistocene setdement 

site in the Sub-Antarctic forest of South-Central Chile. Pages 319-337 in A. L. Bryan, 

ed., 'New Evidence for the Peopling of the Americas. Center for the Study of Early Man, 

University of Maine, Orono, ME. 
. 1989. Monte Verde: A Tate Pleistocene Settlement in Chile, vol. 1. Smithsonian 

Institution Press, Washington, DC. 
. 1997. Monte Verde: A Tate Pleistocene Settlement in Chile, vol. 2. Smithsonian 

Institution Press, Washington, DC. 1071 pp. 

. 1999. The Late Pleistocene cultures of South America. Tvol Anthro. 7:206—217. 

. 2000. The Settlement of the Americas: A NeiP Prehistory. Basic Books, New York, 

NY. 371 pp. 

Dillehay, T, & M. CoUins. 1988. Early cultural evidence from Monte Verde in Chile. 
N^///rf 332:150-152. 

Dillehay, T, G. A. Calderon, G. Politis, & M. da Beltrao. 1993. EarUest hunters and gath- 
erers of South America./. World Prehist. 6:145—204. 

Dillehay, T, J. Rossen, & P. Netherly. 1997. The Nanchoc tradition: The beginnings of 
Andean civilization. Am. Sci. 85:46—55. 

Dillehay, T, & J. Rossen. 1997. Integrity and distribution of the archaeobotanical collec- 
tion. Pages 351—382 in T. Dillehay, ed., Monte Verde: A Tate Pleistocene Settlement in Chile. 
vol. 2. Smithsonian Institution Press, Washington, DC. 

Engel, F. 1970. Explorations of Chilca Canyon, Peru. Curr. Anthropol 11:55—58. 

Fagan, B. 1987. The Great journey: The Peopling of Ancient America. Thames and Hudson, 
London. 288 pp. 

Flannery, K. 1969. Origins and ecological effects of early domesticatiom in Iran and the 
Near East. Pages 73—100 in P. J. Ucko «& G W Dimbleby, eds.. The Domestication and 
Exploitation of Plants and Animals. Aldine Press, Chicago, IL. 

Gruhn, R. & A. Bryan. 1998. A reappraisal of the edge-trimmed tool tradition. Pages 
37—54 in Mark G. Plow, ed., Explorations in American Archaeology: Essays in Honor of 
Wesley K Hurt. University Press of the Americas, Washington, DC. 



Haynes, C. V. 1969. The earliest Americans. Science 166:709-715. 

Hatley, T., & J. Kappelman. 1980. Bears, pigs, and Plio-Pleistocene hominids: A case for 

the exploitadon of below ground food resources. Hum. Ecol. 8:371-387. 
Ingold, T. 1987. The Appropriation of Nature. University of Iowa Press, Iowa C^ity. 287 pp. 
Insdtuto de Geociencias. 1977. Injorme sobre el Clima Kegional. Universidad Austral de 

Chile, Valdivia. 17 pp. 
Kipnis, R. 1998. Early hunter-gatherers in the Americas: Perspective from Central Brazil. 

Antiquity. 72:581-92. 
Kreiger, A. D. 1964. Early man in the New World. Pages 23—91 in]. D. Jennings & E. 

Norbeck, ed., Prehistoric Man in the New World. University of Chicago Press, Chicago, 

Laboratorio Fitoquimica. 1983. Informe sobre el valor nutritivo de las papas silvestres y ]uncus. 

Universidad Austral de Chile, Valdivia. 
LaBuza, T. P. 1982. Shelf-life Dating of Foods. Food and Nutrition. Westport, CT. 500 pp. 
Lopinot, N. 1982. Miscellaneous subterranean plant pasts. Pages lAA—lAd in R. Jefferies 

& B. Butler, eds.. The Carrier Mills Archaeological Project: Human Adaptation in the Saline 

l^alley, Illinois. Center for Archaeological Investigations. Research Paper No. 33. 

Southern Illinois University, Carbondale, IL. 
Lynch, T. F. 1981. Guitarreo Cave: Early Man in the Andes. Academic Press, New York. 

. 1983. The Paleo-Indians. Pages 87—137 in J. D. Jennings, ed., Ancient South 

Americans. WH. Freeman, San Francisco, CA. 

Martin, P. S. 1973. The discovery of America. Science 179:969-874. 

Meltzer, D. J. 1993. Search for the First Americans. St. Remy Press, Montreal, Canada. 

176 pp. 

. 1997. Monte Verde and the Pleistocene Peopling of the Americas. Science 


Meltzer, D. ]., & B. D. Smith. 1986. Paleo-Indian and early Archaic subsistence strategies 
in Eastern North America. Pages 1—30 in S. Nessius, ed.. Foraging, Collecting and 
Harvesting: Archaic Period Subsistence and Settlement in the Eastern Woodlands. Center for 
Archaeological Investigations. Southern Illinois Universit}-, Carbondale, IL. 

Nance, J. 1988. The Archaic period in the lower Tennessee-Cumberland-Ohio region. 
Pages 127—152 in C. D. Hockensmith, D. Pollack, & T. N. Sanders, eds., Paleoindian and 
Archaic Research in Kentucky'. Kentucky Heritage Council, Frankfort, KY. 

Prous, A. 1996. L'Archeologie au Bresil: 300 siecles d' occupation humaine. 
LAnthropologie. 190:257-306. 

Ramirez, C. 1989. Macro-botanical remains. Pages />/T. Dillehay, ed.,AIonte Verde: A Tate 
Pleistocene Settlement in Chile, vol. 1. Palaeoenvironment and Site Settino. Smithsonian 
Institution Press, Washington, D.C. 



Redman, C. 1977. Man, domestication, and culture in Southwestern Asia. Pages 523-541 

in C. Reed, ed.. Origins of Agriculture. Mouton Press, The Hague. 
Richardson, J. 1978. Early Man on the Peruvian North Coast, Early Maritime 

Exploitation and the Pleistocene and Holocene Environment. Pages 274—89 in A. L. 

Bryan, ed., Early Man in America from a Circum-Pacific Perspective. Occasional Papers No. 

1, Department of Anthropology, Universit}^ of Alberta, Edmonton, Canada. 
Rossen, J. 1998a. Unifaces in early Andean Culture History: the Nanchoc Lithic Tradition 

of Northern Peru. Andean Past 5:241-299. 
. 1998b. Archaic plant utilization at the Hedden site, McCracken Count}?, 

Kentucky. Pages 33-44 in D Pollack & K. GremiUion, eds.. Current Archaeological 

'Research in Kentucky, vol. 6. Kentucky Heritage Council, Frankfort, KY. 
Rossen, J., & T D Dillehay. 1997. Modeling ancient plant procurement and use at Monte 

Verde. Pages 329—350 in T Dillehay, ed., Monte J^erde: A iMte Pleistocene Settlement in 

Chile, vol. 2. Smithsonian Institution Press, Washington, DC. 
Rossen, J., T. D Dillehay, & D Ugent. 1996. Ancient cultigens or modern intrusions? 

Evaluating plant remains in an Andean Case Study. /. Archaeo. Science 23:391—407. 
Rossen, J., & C. Ramirez. 1997. Observations on the present-day (1983) economic plants 

in the Monte Verde area and their archaeological implications. Pages in T Dillehay, 

ed., Monte Verde: A L.ate Pleistocene Settlement in Chile, vol. 2. Smithsonian Institution 

Press, Washington, DC. 
Schmitz, P. 1987. Prehistoric hunters and gatherers in Brazil./. World Prehist. 1:1—56. 
Soffer, Olga. 1989. Storage, sedentism and the Eurasian Palaeolithic record. Antiquity 

Stothert, K. 1988. Ea Prehistoria Temprana de la Penisula de Santa Elena: Lm Cultura Las Vegas. 

Miscelanea Anthropologica Ecuadoriana. Series Monografico 10, Museo del Banco 

Central del Ecuador. Quito. 271 pp. 
Testart, A. 1983. Reply. Curr. Anthropol 24(1):119. 

Ugent, K., T. Dillehay, & C. Ramirez. 1986. Potato remains from a Late Pleistocene set- 
tlement in South-central Chile. Econ. Bot 41:17—27. 
Vavilov, N. I. 1951. The origin, variation, immunity and breeding of cultivated plants. 

Chron. Bot 13:1-366. 
Winterhalder, B., & E. Smith, eds. 1981. Hunter-Gatherer Foraging Strategies: Ethnographic and 

Archaeological Analyses. University of Chicago Press, Chicago, IL. 268 pp. 


Chapter Nine 

Ocean Trails and Prairie Paths? 
Thoughts about Clovis Origins 

Dennis Stanford and Bruce Bradley 

^^1^. Archaeologists have long believed that artifacts unearthed with 

^^^^^^^ mammoth bones in a gravel pit near Clovis, New Mexico were made 

,^,_^^fl^^^^H by the First Americans (see Bonnichsen & Turnmire 1991). The 

i^Bp^^ people who made these tools, known as Clovis, were hunters and 

gatherers adapted to the diverse and changing environments of 

North America during the transition between the Pleistocene and the Holocene. 

Archaeological sites containing Clovis artifacts are found from coast to coast and from 

Canada to Central America. These sites are identified on the basis of distinctive bifacial, 

fluted projectile points and blade core technology. 

However, the origin of Clovis culture and technology remains a mystery. 
Traditionally, most archaeologists have reasoned that the ancestors of Clovis people 
came from Asia by crossing a land bridge connecting Siberia and Alaska. This land bridge 
formed when Ice Age sea levels were depressed while water was trapped as glacial ice. It 
is thought that after early populations arrived in Alaska, terrestrial glaciers began to melt 
creating an inland route, or "Ice-free Corridor," that joined the Arctic with the rest of 
the Americas. Clovis people presumably spread southward through the continent by 
means of this pedestrian route. This land bridge/interior-corridor hypothesis seems emi- 
nently logical as Siberia lies in geographic proximit}^ to Alaska, and pedestrian routes until 
recently have been favored strongly over maritime routes in North American coloniza- 
tion models. Further, many modern North American Indians share physical similarities 
with current Northeast Asians. Consequentiy, little doubt existed in the minds of many 
scholars that the bearers of Clovis culture were ancestral to most modern Native 
Americans. For proponents of this hypothesis, Siberia and Alaska are logical places to 
search for the roots of Clovis culture. 



E vidence for A. si an O rigins 

Scholars have sought to find archeological evidence of Clovis origins for decades. 
One of the earliest studies aimed directly at finding evidence of early Clovis peoples in 
the ice-free corridor was that of H. M. Wormington and Richard Forbis (1965). In the 
1950s they surveyed collections and sites in Alberta just below the exit point of the cor- 
ridor. Their efforts failed to locate archaeological material suggestive of early or pre- 
Clovis; indeed, the few Clovis specimens they recorded occurred much further south in 
the Province. Canadian archaeologists still have not found archaeological evidence to 
confirm a human presence in the corridor before 11,000 years ago (Beaudoin et al. 1996). 
A few fluted projectile points have been found in the corridor (see Carlson 1991), but 
these undated artifacts appear to be related to the fluted points found in Alaska, which 
are thought to date around 10,500 yr B.P. (Hamilton & Goebel 1999). 

Evidence of an early Clovis lithic technology in Alaska has also eluded researchers. 
Only the Nenana Complex, found in Central Alaska (Powers & Hoffecker 1990), is 
thought by some investigators to be related to Clovis (Goebel et al. 1991). The breakage 
pattern on the small thin bifaces found at Nenana sites suggest to us that these tools 
were likely used as knives. Although there are some other technological correspondences 
betu^een Clovis and Nenana, the evidence for the comparison is weak. While there are 
some early dates for Nenana, most of the radiocarbon evidence suggests that it is virm- 
ally contemporaneous with Clovis. Thus, it appears to us that there is little archeological 
evidence that Clovis ancestors crossed the land bridge or passed down an ice-free corri- 
dor linking Northeast Asia to America. 

Recent research indicates that the ice-free corridor may not have opened up until 
11,000 yr B.P. (Jackson et al. 1996). After deglaciation, the ice-free corridor may have been 
barren of plant and animal life for a long period of time, and this would have rendered 
the area unsuitable for human habitation (Burns 1996; Mandryk 1990), until after Clovis 
times. If this was the case, an interior land route to the Americas was essentiaUv blocked 
for early explorers between 22,000 and 1 1 ,000 years ago. 

Evidence for Ocean Trails 

Given the relatively late timing of viable human travel in the ice-free corridor and the 
early dates at Monte Verde and other archaeological sites in the Americas, it appears that 
early explorers must have used alternative paths to the New World. Some scholars 
(Fladmark 1979; Dixon 1999) believe that it is likely that people using watercraft estab- 
lished a foothold in the Americas while exploiting the coastal environments of the 



Pacific continental shelf, which apparently was ice-free by 13,000 years ago (Josenhans et 
al. 1997). Recent efforts to reconstruct former environments and ecosystems verify that 
Pleistocene animals flourished along the Ice Age coastUne (Heaton 1996). Further, cul- 
tural artifacts have been found along these coastal environments (Erlandson et al. 1998; 
Dixon 1999), as well as from off shore drowned terrestrial and intertidal deposits (Fedje 
& Josenhans 2000). Thus far, the coastal discoveries are too young and they do not 
appear to be technologically related to Clovis. These discoveries indicate, however, that 
humans were indeed utilizing the now-inundated continental shelf, and future research 
may find evidence of earlier occupations. 

Indirect evidence indicates that watercraft were in use on the Pacific Ocean early in 
the known prehistory of the dispersion of modern humans. For example, nearly 50,000 
years ago, the ancestors of the Australian Aborigines crossed a deep ocean trench, the 
Timor Trough, while traveling from Southeast Asia to the ancestral Australian-New 
Guinea continent of Sahul, a distance of 80 kilometers (Mulvaney & Kamminga 1999). 
Over 30,000 years ago Paleolithic flintknappers from Japan traveled by watercraft to col- 
lect obsidian from Kazushima Island, now located some 170 km off shore (Oda 1990). 
It is our opinion that the idea of such an inherently useful technology as watercraft prob- 
ably spread rapidly among populations around the world. Indeed, obsidian from the 
Island of Melos found in the Late Paleolithic levels of Franchthi Cave in Greece (Perles 
1987), suggest seafaring capabilities by that time (van Andel & Shackleton 1982). 
Considering that simple raft and boat technologies underwent "research and develop- 
ment" for nearly 40,000 years prior to Clovis times, it seems reasonable to presume that 
some groups ably navigated the waters of the North Pacific Rim and quite Ukelv those 
of the North Atiantic by Late Pleistocene times. 

Identifying archaeological evidence supporting the former existence of ancient water 
transport is problematic (Bednarik 1997). Materials used in early watercraft construction 
were perishable and would not readily survive in most environments. Further, such craft 
are usually stored near water and would have suffered degradation through erosion by 
the tides of rapidly rising sea levels associated with deglaciation. This situation may also 
hold true for many coastal habitation sites as well. Thus, finding evidence of the former 
use of Paleolithic watercraft or the occupation of early sites on the continental shelf 
requires strategic concerted effort and no small element of luck. 

Evidence for Clovis Ties to A.sici? 

A growing number of researchers, including the present authors, have examined 
many of the Late Pleistocene archaeological sites and museum collections in Eurasia, 



Asia and the Far F^ast in order to seek ties with (^lovis. But, to date, no clear-cut evidence 
is discernible. Asian Upper Paleolitliic technologies are dominated by microblades made 
from wedge-shaped cores (Derev'anko 1998). When used in weaponr)^, multiple microb- 
lade segments were inset into long, narrow, pointed rods of bone, antler or ivory to form 
long, lethal, cutting edges. In contrast, Clovis stone weapon tips consisted of thin, flat 
bifaces with concave bases that attached directly to the end of a shaft or foreshaft. 
Conceptually, these are entirely different approaches to making effective weapons. 

The incidence of bifaces in Northeast Asia is significantiy lower than that docu- 
mented for Clovis. Moreover, when bifacial technology was used in Northeast Asia, the 
resultant artifacts tended to be thick in relationship to their width and thev were usually 
bipointed. Clovis technology on the other hand included two basic approaches to stone 
tool manufacture: biface production, and blades and bladelets made from nonwedged 
shaped cores. These two reduction techniques are infrequently associated with one 
another in Northeast Asia. In the Trans-Baikal region of central Asia, some assemblages 
include bifaces and large blade cores as part of the flintknappers' repertoire (KiriUov & 
Derevanko 1998). Such sites are rare, are older than Clovis by more than twent}' millen- 
nia, and are separated from Alaska by some 5,000 km, across some of the coldest and 
most inhospitable areas of the world. Further, they appear to be more closely related to 
the Streletskayan technology of the Eurasian Plain than to those of the Far East (Bradley 
etal 1995). 

Since known Asian technologies do not appear to satisfy expectations of a Clovis 
technological prototype, if Clovis peoples came out of Northeast Asia, then their dis- 
tinctive technology evidently developed after they colonized the New World. If this were 
the case there should be identifiable links indicating a transition in the manufacture of 
projectile points from the use of inset microblades to hafted bifaces. Or, our precon- 
ceived notion about where Clovis people originated may be incorrect. 

A.lternative Paths for Clovis Origins 

Although a number of authors (Cotter 1935; Greenman 1963; Boldurian & Cotter 
1999) have pointed out similarities between the American Paleoindian technologies and 
the Paleolithic of Europe, it was Jelinek (1971) who identified the Solutrean culture in 
the Vasco-Cantrabrian region of Northern Spain, as having the most striking resem- 
blance to the Clovis technology. He noted that the similarir\' included bifaciaUv tlaked 
lanceolate, projectile points with concave bases and basal and edge grinding, a basic blade 
industry, smaU end scrapers with graver tips, large side scrapers made on flakes, cylindri- 
cal bevel-based bone points, and a relative scarcit}' of burins. However, he felt that a 



direct historical relationsliip between Clovis and Solutrean was impossible because: 1) 
Solutrean is more than 5,000 years older than Clovis; and 2) of the difficult}^ of crossing 
the late glacial Atlantic Ocean (Jelinek 1971:15-21). In a recent article, Straus (2000a), 
also enumerates these same reservations. Thus, the time and space gaps suggest that the 
similarities of these culmres resulted from independent invention. And, unless these 
issues are resolved, the comparison of Clovis and Solutrean will only provide useful data 
on why such a cultural convergence might occur. However, we suggest that perhaps these 
gaps may even now be resolved and certainly if scholars do not look beyond their cur- 
rent biases, new evidence wiU not be forthcoming. 

Filling the Time Gap 

Meadowcroft rock shelter in Pennsylvania includes a stratified deposit that contains 
human occupation levels dating to perhaps a thousand years before Clovis (Carlisle & 
Adovasio 1982). The artifacts found in the older deposits suggest a technology that 
included bladelets and unfluted bifacial projectile points and basketry (Adovasio et al. 
1978). The site has long been a thorn in the side of scholars who advocate that Clovis 
peoples were the first Americans. Their arguments contend that the radiocarbon samples 
were contaminated by coal residue in groundwater percolating through the cave deposits 
or other phenomena mixing prehuman deposits with cultural charcoal (Haynes 1980). 
Recent microstratigraphic research concludes that there was not a contamination prob- 
lem with the dates (Goldberg et al. 1999). It should be pointed out that none of the puta- 
tive pre-Clovis charcoal dates from the Alaskan sites have undergone similar intense 
scrutiny as the Meadowcroft dates even though the contamination and soil mixing are 
major problems in the Arctic. Could this inequity reflect the preconceived notions of 
scholars wishing to find the Alaskan link to Clovis? Nonetheless, Meadowcroft stood out 
as a single anomaly and the existence of such an early occupation would be inconclusive 
until other sites, dating to the same time period and containing similar artifacts, were 
found and documented. 

Along the Nottoway River in Virginia a site known as Cactus Hill has recendy pro- 
duced a classic sequence of Mid- Atlantic archaeology, including Clovis and occupations 
below Clovis (McAvoy & McAvoy 1997). The artifacts from the lower levels, thin bifa- 
cial points, large blades, bladelets, cores, and other tools, are technologically similar to 
those from the Meadowcroft and could be considered a related complex. The radiocar- 
bon dates from a hearth feature and other samples suggest that the site was occupied 
between 15,000 and 17,000 14 C yr B.P. Although the combined artifact samples from 
both sites are small, we suggest that these two assemblages should be considered part of 



the same icchiiological complex. I'urllier, their chr()n()l()iL!,ical placement suggests to us 
ihat the\ arc prime candidates tor developmental Clovis. 

It is important to note that there are significantly more fluted points and sites tound 
in the eastern United States than in the rest of North America (Mason 1962; Anderson 
& Gillam 1999), as well as the greatest variation in fluted point forms and technology 
(Lepper 1999). Although certainly not conclusive, these circumstances indicate to us that 
the (Clovis Tradidon may have a greater dme depth in the east and that Clovis probably 
developed its characterisdc technological aspects in the east. If this is true, along with the 
pre-Clovis dates from Meadowcroft and Cactus Hill, there is a high degree of probabil- 
it}' that the lithic technology from these sites represents a likely precursor of Clovis tech- 

We are also struck with the number of technological similarities shared among 
Meadowcroft, Cactus Hill, Clovis and Solutrean flaked stone artifacts. For example, the 
Meadowcroft/Cactus Hill bifacial weapon tips are technologically very similar to some 
Solutxean points and though they lack the fluting found on Clovis specimens, they are 
basely thinned. Similar too are the blades and smaller blade cores. Furtiier, some of the 
radiocarbon dates for these sites overlap in time with Solutrean. If liiese dates prove to 
be accurate, we see the possibilit)' of a developmental continuum from Solutrean 
through Cacms HiU and Meadowcroft to Clovis, filling in the 5,000-year time gap. 

Coincidentally with the discovery at Cactnas HiU, the results of DNA research on 
ancient human remains have begun to map genetic relationships among the w^orld's pop- 
ulations. From this research it is clear that modern Native Americans have an over- 
whelming genetic contribution from Asian ancestors. However, mixed into the genetic 
composition of a few Indian groups is a mitochondrial DNA haplogroup known as X 
(Schurr 2000). This haplogroup is absent in eastern Asia but occurs in Europe. The hap- 
logroup X marker has been found in several pre-Columbian populations, suggesting to 
Schurr (2000) the possibilit)' that the haplogroup was present in the New World over 
12,000—34,000 years ago. Could the population that carried this genetic marker from 
Europe to the Americas have been Solutrean people? 

The SolMtrean Evidence 

With these new data and the questions they pose, we felt that the h\-pothesis of a 
Solutrean origin for some of the ftrst American human populations needed to be seri- 
ously examined. Thus, we arranged to look at collections in various museums in Spain, 
France and Portugal. We also studied site reports and syntheses of tiie Solutrean archae- 
ological data. 



Our examination of these materials was most illuminating. Tlie degree of similarity 
we found was greater than anticipated. In collection after collection, we viewed artifacts 
that we would not have been surprised to find at a Clovis site. One of the most remark- 
able technological features of stone working was the identical technique that both groups 
used to manufacture thin bifaces. This is a technique that archaeologists call overshot 
flaking (Prison & Bradley 1999), that is, striking a flake that starts on one edge of the 
biface and travels across the entire surface and removes a small portion of the opposite 
edge. This technique rapidly thins a biface with relatively few flakes. In our examination 
of biface thinning around the world, only these two groups consistentiv and purpose- 
fully used the overshot technique. This is not to say that overshot flakes never occurred 
in other technologies. But in these cases, careful examination of assemblages does show 
them to be only occasionally present, and in most cases it is clear that they were knap- 
ping mistakes. 

There are also technological similarities between Clovis and Solutrean blade manu- 
facture. The methods of precore shaping, setting up the core face, and blade detachment 
techniques are virtually identical between the two groups. In fact, we see Solutrean blade 
technology as closer to Clovis than it is to any other European blade core technologies 
used either before or after Solutrean. 

Other traits related to flaked stone that are shared between Clovis and Solutrean, but 
are not present in other Paleolithic technologies, include the preference for exotic raw 
materials in the manufacture of bifaces (e.g., Balout 1958:619), pressure flaking and occa- 
sional heat-treating of fUnt. Caches of over-sized bifaces (Smith 1966), that can be 
argued were not intended to be used, are another trait shared between Clovis and 
Solutrean. These traits are rare to absent in the Far East Upper Paleolithic traditions. 

The similarities in tools went beyond the stone technology. Many Solutrean bone, 
antler and ivory tools have close Clovis analogues, this is especially true for bone projec- 
tile points or bone rods (Smith 1966). The manufacturing and even decoration in some 
instances were nearly identical. Bone shaft wrenches are found in both assemblages, and, 
although eyed bone needles and bones disks with "tally marks" have not yet been found 
associated with Clovis, they do occur in the related Folsom cultare. Along with incised 
decoration on bone points, engraved stones, etched with geometric and zoomorphic 
designs have been found in both culmres (Smith 1966; Collins et a/. 1991). After our 
study of the Solutrean collections, we were deeply impressed with the amount of con- 
currence we observed. Among the questions that we pondered was how much similari- 
ty is necessary before an interpretation of convergence gives way to an historical rela- 
tionship? It was clear to us that if the Solutrean artifacts we observed were found in 
Northeastern Asia, no one would doubt the connection to Clovis. 



Filling the Space Gap 

If there \\;is an historic connection, the remaining questions to be addressed are 
issues of the crossing of the Atlantic Ocean. Did the Scjlutreans possess the necessary 
skills, have the right equipment, and why would they undertake such a journey? These 
quesdons have seldom been seriously considered because of the preconceived biases of 
scholars working in both areas of the world. Scholars presume that these people were of 
such a rude state of technological knowledge that such endeavors would ha\e been 
impossible for them to accomplish. Further, it is commonly thought that during the Last 
Glacial Maximum (LGM) — the time of the Solutrean people — the environment was 
too cold, harsh and stormy for Paleolithic people to venture out onto the Atlantic Ocean 
(Straus 2000a). If they did, there must have been some extreme reason for such deeds. 

PaleocUmatic reconstructions for Western Europe and the North Atlantic do indeed 
suggest that during the LGM, the weather was on average colder and stormier than at 
present. Even though the net climatic conditions over this period may have been harsh, 
there were intervals when this was not the case and there were climatic regimes alternat- 
ing between dry and cold to temperate and humid (Butzer 1986; Rigaud & Simek 1990; 
Straus 1990). 

It is thought that during Solutrean times, the terrestrial climatic conditions resulted 
in lowered temperatures, shortened growing seasons and a reduction in the extent and 
quaUt}^ of natural grasslands (Straus 1977). Game animals may have died out or become 
rare in the interior regions of Western Europe, and forced both animal and human pop- 
ulations into more favorable areas along the rivers and the coastlines of southwestern 
Europe (Straus 1977). Hunting was still possible in the nearby mountains, but more and 
more people likely turned to alternative food resources such as hunting and fishing along 
the river estuaries and beaches (Straus 2000b). 

Sea ice would have been a common sight in the Bay of Biscay during Solutrean times 
(Webb et al. 1993). The summer seawater temperatures are thought to have dropped to 
6°C (Rigaud & Simek 1990), which is within the tempature range for Arctic cod (0°C — 
1 1°C [A. C. Jensen 1972]). The Arctic cod is a pivotal species in the arctic food web, and 
is an important prey item for sea mammals (Tynan & DeMaster 1997). It therefore seems 
likely that the Southwestern Europe pleniglacial seacoast would have provided an 
extremely rich habitat, attracting fish, birds and sea mammals. Since the Solutrean peo- 
ple resided on the coastal areas of Southwest Europe for more than two thousand years, 
we believe that it would not take them long to learn how to target these resources, espe- 
cially if terrestrial resources were stressed. 



Faunal collections from Solutrean sites in northern Spanish sites contain abundant 
evidence that these people were indeed utilizing coastal and estuary resources (Straus 
1992,112) and there is evidence that through time there was an increasing dependence 
on marine resources (Straus 1985:505). For example, humans were transporting sub- 
stantial amounts of limpets up to 10 km from the pleniglacial shore to La Reira Cave in 
Spain (Straus & Clark 1986). This led Straus (1992) to suggest that if many pleniglacial 
coastal sites were inundated, marine resources may have been even more significant as 
dietary supplements than the present evidence would indicate (Straus 1992:112). In 
reviewing the inland evidence, Cleyet-Merle and Madelaine (1995) suggest a very active 
marine exploitation by Paleolithic peoples and that the use of these resources has been 
underestimated by scholars. 

If, as Straus intimates (1979), Solutrean camps were established on the pleniglacial 
beaches and estuaries of northern Spain, the geographical area in which they could 
expand their hunting territory would increase dramatically along that beach and as far 
northward as the south coast of Ice Age Ireland. The exploitation of this area may have 
originally started as seasonal trips, but through time, they could have established more 
semi-permanent campsites along the coastline. In fact, because of potential marine 
resources, the ancient coast and associated plain may have furnished an independent 
resources base beyond that available in the mountains. Further, these resources may have 
supported a larger Solutrean population than represented by the temporary hunting 
camps known from sites found near the present day coast of Spain. 

Maritime Technology 

In the process of adapting to coastal marine resources, people would have no doubt 
modified old tool kits and created new technologies to more efficiently exploit the sea. 
This may have been especially true for Solutrean people, as they have been characterized 
as developers of specialized technologies (Straus 1990). Among the necessary equipment 
to exploit the marine resources were waterproof clothing, nets, harpoon gear and water- 
craft. Once this equipment was developed, it could have opened up a whole new world 
for the Solutrean people to exploit. 

Adovasio et al. (1996) have found that plant- fiber technology such as cordage, bas- 
ketry, netting, and textiles were being produced in central Europe by 25,000 years ago. 
Thus, it is reasonable to assume that these technologies were known to and used by 
Solutrean peoples, allowing them a wider technology than tiiat represented by their stone 
tools. Waterproof clothing could be made from either textiles or animal gut tissue, but 
would have in any case been necessary for survival in the LGM, so we assume that these 
items were already part of their personal gear. 



Willi the knowledge of cordage manufacrure, onl\- simple modifications ro land- 
based weaponr)' would be rec|uired for harpoon equipment and nets. Sharp durable end 
blades (points) are frequently part of harpoon technology in many parts of the world. 
We believe that it is highly probable that the thin concave base points hjund most com- 
monly in Solutrean sites in north coastal Spanish assemblages ma\' have been well suited 
for harpoon end blades. 

As mentioned above, watercraft had been in use tor at least 30,(K)(J years before 
Solutrean dmes, so we feel that it is not too far-fetched that the news had traveled to the 
inhabitants of Western Europe. And cordage for ropes and fiber texdles for sails are also 
within reason. Even with the lack of physical evidence we see no reason to believe that 
these people did not have some type of watercraft. Just as there is no direct evidence that 
the ancestors of the Australian Aborigines or the Paleolithic peoples from Franchthi 
Cave had watercraft. 

We think that a skin boat outfitted with sails, similar to the Inuit umiak or the Irish 
curraghs are excellent models for Solutrean boats. Skin boats make an excellent vehicle 
for northern waters. They are light and durable, if properly treated, they are sea worthy 
for months. For instance the Brendan, a curragh made of ox hide by Severin (1978) was 
used to cross the Atlantic Ocean. The Brendan was coated with sheep tallow, which 
allowed the boat to be on the ocean for several months, without becoming waterlogged. 
Similar hides and fats were available b}"products of animals that Solutrean peoples were 
hunting in the coastal mountains of Spain. 

The Koute 

Paleooceanographic studies of the North Atiantic suggest that during the LGM, a 
permanent arch of sea ice connected the Southwest coast of Ireland to the Grand Banks, 
forming a north Atlantic ice rim (Figure 1). Winter sea ice would have formed each year, 
extending the ice further south to between 45° and 50° north latitude depending on 
annual weather conditions (see Webb III et a/. 1993). The Gulf Current that now extends 
across the North Atiantic shifted southwards, making a clockwise circulation pattern 
moving toward the coast of Portugal and returned westsvard from North Africa. The 
Gulf Current warmed a weak counter clockwise current that circulated in the North 
Atiantic (Robinson ef a/. 1995). This northern current would have moved northwestward 
from the Irish coast toward the Grand Banks and returned to the Bay of Biscay. 

Even though the LGM North Atiantic has been characterized as cold and stormy, 
these hostile weather conditions were not continuous, being interrupted bv periods of 
short term and perhaps even longer term intervals in which sailing conditions on open 



water would have been possible even for less experienced navigators. Regardless of 
weather conditions, it would not take too many generations of coastal dwellers to under- 
stand the signs of impending storms and the timing of these events. Nor would it have 
taken an observant hunter long to discover that floating ice islands and ice leads provid- 
ed excellent havens during storms. These ice features reduced wave formation and pro- 
vide puUout locations for setting up camps during inclement weather. An overturned, 
well-secured skin boat makes an excellent temporary shelter during such events. Further, 
sea mammals provide fuel oil for heating and cooking, and if necessary to assist in dry- 
ing out skin boats. 

We propose that in the two thousand or so years that Solutrean folks lived next to 
and observed these environmental parameters, they no doubt adapted to and capitalized 
on these resources. Further, as they became more skilled in seamanship, they would have 
venmred further and further out along the ice margin. Perhaps first as simple extended 
hunting trips and finally to spend entire summers hunting along the ice front. Eventually 
someone made it the mere 2500 km between the LGM coastlines of the Grand Banks 
and Ireland. When arriving in the area of the Grand Banks they found a fishery that 
would have been unimaginable to a New England fisherman. This would have been big 
news back in old country. Soon we suspect that more and more expeditions were under- 
taken, spreading the X haplotj'pe into the Americas. Once established in North America, 
they spread southward along the now submerged continental shelf to eventually find new 
terrestrial hunting opportunities and perhaps new people with which to intermingle. 

We feel that the diverse amount of similarities noted between Solutrean and Clovis 
suggests a more complex situation than simple convergence. It appears possible that the 
dating and technology represented by Cactus HiU and Meadowcroft sites provide the Unk 
between Solutrean and Clovis. It may be bold of us, in the face of a dearth of proof to 
coin the term "Atlantic Paleomaritime" for these hypothesized ice-edge-adapted people, 
however, we do so to spur future research on these issues and broaden the thinking of 
new generations of archaeologists. Much research needs to be conducted to either sup- 
port or reject these ideas, but regardless of the outcome of future debates over these 
issues, we feel that it is time to start thinking outside the confines of our present para- 



Figure 1. North Atlantic surface circulation reconstruction for the LGIM. Adapted from Robinson 
etal. 1995; Webb ^/^/. 1993. 


This paper discusses some of the current assumptions widely held about Nen^ 
World origins and outlines the reasoning that led the authors to explore an alter- 
native hypothesis that Clovis technology may have had roots in the Solutrean 
Culture of the European Paleolithic. We suggest that Solutrean people began to 
adapt to marine resources in response to the harsh terrestrial environment of 
Western Europe during the hast Glacial Maximum (LGM). A maritime compo- 
nent of their adaptive orientation, termed here North Atlantic Paleomaritine, 
enabled the movement of people ivestward along the margins of North Atlantic 
sea ice ivhere founding populations were eventually established in North America. 
After preliminary assessment and comparison of Solutrean and Clovis technology, 
as well as examination of paleoecological reconstruction of the L.GM for Europe 



and the North Atlantic, we feel that this hypothesis is most parsimonious with 
currently available data and warrants further study and serious consideration. 

l^iterature Cited 

Adovasio, J. M., J. D. Gunn, J. Donahue, R. Stuckenrath, J. Guilday, & K. Lord. 1978. 

Meadowcroft Rockshelter. Pages 140-180 in A. L. Bryan, ed., Early Man in America 

from a Circum-Pacific Perspective. Occasional Papers No.l Department of Anthropology, 

University of Alberta, Edmonton, Canada. 
Adovasio, J. M., O. Soffer, & B. KHma. 1996. Upper Palaeolithic fiber technology: 

interlaced woven finds from Pavlov I, Czech RepubUc, c. 26,000 years ago. Antiquity 

Adovasio, J. M., D. Hyland, & O. Soffer. 1998. Perishable technology and early human 

populations in the New World. Paper prepared for the 31^^ Annual Chacmool 

Conference: On being first: Culmral innovation and environmental consequences of 

first peopling. Calgary, Alberta, Canada. 
Anderson, D. G., & J. C. GiUam. 1999. Paleoindian colonization of the Americas: 

Implications from an examination of physiography, demography, and artifact distri- 
bution. Am. Antiquity 65:43—66. 
Balout, L. 1958. L'Abri Andre-Ragout au Bois-du-Roc. Fouilles de 1957. Bull Soc. 

Prehistor. Francaise. 55:599—627. 
Beaudoin, A. B., M. Wright, & B. Ronaghan. 1996. Late Quaternary landscape history 

and archaeology in the "Ice-Free Corridor": Some recent results from Alberta. Quat 

Int. 32:113-126. 
Bednarik, R. 1997. The earliest evidence of ocean navigation. Int. J. Naut. Archaeol 

Boldurian, A., & J. Cotter. 1999. Clovis revisited: New Perspectives on Paleoindian Adaptations 

from Blackwater Draw, New Mexico. University Museum Monograph 103. The University 

Museum, University of Pennsylvania, Philadelphia, PA. 
Bonnichsen, R., & K. Turnmire, eds. 1991. Clovis: Origins and Adaptations. Center for the 

Study of the First Americans, Oregon State University, Corvallis, OR. 344 pp. 
Bradley, B., M. V. Anikovitch, & E. Y. Giria. 1995. Early Upper Paleolithic in the Russian 

Plain: Streletskayan flaked stone artifacts and technology. Antiquity 69:989-998. 
Burns, J. A. 1995. Vertebrate paleontology and the alleged Ice-Free Corridor: The meat 

of the matter. ^^^/. Int. 32:107-112. 



Burner, K. VV. 1986. Palcolirhic adaptations and settlement in (^antabrian Spain. Pages 
201—252 in F. Wendorf & ]l. C^lose, eds., y\di'cii!ces in World Archaeology Vol. 5. 
Academic Press, Inc., New York, NY. 

Carlisle, R. C, & J. M. Adovasio. 1982. Meadowcroft: Collected Papers on the 
Archaeology of Meadowcroft Rockshelter and the Cross Creek Drainage. Papers 
Prepared for the Symposium: The Meadowcroft Rockshelter Rolling Thunder Review: Tast 
Act,. 47*- Annual Meeting of the Society for American Archaeology, Minneapolis, 

Carlson, R. L. 1991. Clovis from the perspective of the Ice-Free Corridor. Pages 81-90 
in R. Bonnichsen & K. Turnmire, eds., Clovis Origins and Adaptations. Peopling of the 
Americas Publications, Center for the Stnady of the First Americans, Oregon State 
Universit)^ Press, Corvallis, OR. 

Cleyet-Merle, J., & S. Madelaine. 1995. Inland evidence of human sea coast exploitation 
in Palaeolithic France. Pages 303—308 /;/ A. Fischer, ed., Alan and Sea in the Mesolithic. 
Oxbow Books, Oxford, England. 

Collins, M. B., T. R. Hester, & P. |. Headrick. 1991. Engraved cobbles from the Gault Site, 
Central Texas. Curr. R£s. Pleistocene 9:3-4. 

Cotter, J. L. 1935. Yuma and Folsom Artifacts. M.A. thesis. Department of Anthropology, 
Universit}^ of Denver, CO. 

Derev'anko, A. 1998. The Paleolithic of Siberia: New Discoveries and Interpretations. Universit}' 
of Illinois Press, Urbana, IL. 406 pp. 

Dixon, E.J. 1999. Bones Boats and Bison: Archeology and the First Coloni^tion of Western North 
America. The Universit}' of New Mexico Press, Albuquerque, NM. 322 pp. 

Erlandson, J. M., T. C. Rick, R. L. Vellanoweth, & D |. Kenneth. 1998. Maritime subsis- 
tence at a 9300 year old shell midden on Santa Rosa Island, California. /. Yield 
Archaeol 26:255—265. 

Fedje, D W, & H. Josenhans. 2000. Drowned forests and archaeology on the continen- 
tal shelf of British Columbia, Canada. Geolog): 28:99—102. 

Fladmark, K. R. 1979. Routes: Alternative migration corridors for early man in North 
America. Am. Antiquity 44:55—69. 

Frison, G., & B. Bradley. 1999. The Venn Cache, Clovis Weapons and Tools. One Horse Land 
& Catde Co., Santa Fe, NM. Ill pp. 

Goebel, F. E., W. R. Powers, & N. Bigelow. 1991. The Nenana Complex of Alaska and 
Clovis origins. Pages 49—79 in R. Bonnichsen & K. Turnmire, eds., Clovis Origins and 
Adaptations. Center for the Smdv of the First Americans, Oregon State University-, 
CorvaUis, OR. 



Goldberg, P., T Arpin, & J. Donahue. 1999. Micromorphological analyses of sediments 
from Meadowcroft Rockshelter, Pennsylvania: Implications for Radiocarbon Dating. 
/. Field ArchaeoL 26:325-342. 

Greenman, E- F- 1963. The Upper Paleolithic and theNew World. Curr. AnthropolA-A\-9\. 

Hamilton, T. D, & T. Goebel. 1999. Late Pleistocene peopling of Alaska. Pages 156-199 
in R. Bonnichsen & K. Turnmire, eds., Ice Age Peoples of North America: Environments, 
Origins, and Adaptations of the First Americans. Center for the Study of the First 
Americans, Oregon State University Press, Corvallis, OR. 

Haynes, C. V. Jr. 1980. Paleoindian charcoal from Meadowcroft Rockshelter: Is contam- 
ination a problem? Ajn. Antiquity 45:582—587. 

Heaton, T. H. 1996. The Late Wisconsin vertebrate fauna of On Your Knees Cave, 
Northern Prince of Wales Island, Alaska. /. T "ert. Faleontol 16:40-41. 

Jackson, L. E., Jr., & A. Duk-Rodkin. 1996. Quaternary geology of the ice-free corridor: 
Glacial controls on the peopling of the New World. Pages 214-227 in T. Akazawa & 
E. J. E. Szathmary, eds., Prehistoric Mongoloid Dispersals. Oxford University Press, New 
York, NY. 

Jensen, A. C. 1972. The Cod Thomas P Crowell Co., New York, NY. 182 pp. 

JeUnek, A. J. 1971. Early man in the New World: A technological perspective. Arctic 
Anthropol 8:15-21. 

Josenhans, H. W, D W Fedeje, R. Pienitz, & J. Southon. 1996. Early humans and rapid- 
ly changing Holocene sea levels in the Queen Charlotte Islands-Hecate Strait, British 
Columbia, Canada. Science 211:11— 1 A. 

Kirillov, I., & A. Derev'anko. 1998. The Paleolithic of the Trans-Baikal Area. Pages 
137—274 in A. P. Derev'anko, ed.. The Paleolithic of Siberia: New Discoveries and 
Interpretations. University of Illinois Press, Urbana, IL. 

Lepper, B. T 1999. Pleistocene peoples of midcontinental North America. Pages 
362—394 /;/ R. Bonnichsen & K. L. Turnmire, eds.. Ice Age Peoples of North America: 
Environment, Origins, and Adaptations of the First Americans. Oregan State Universit)^ 
Press, Corvallis, OR. 

Mandryk, C. A. 1 990. Could humans survive the ice-free corridor?: Late-Glacial vegeta- 
tion and cHmate in West Central Alberta. Pages 67-79 in L. D Agenbroad, J. Mead, 
& L. Nelson, eds., Megafaima and Man: Discovery of America's Heartland. The Mammoth 
Site of Hot Springs, South Dakota, Inc. Scientific Papers, Vol. 1 . Hot Springs, SD. 

Mason, R.J. 1962. The Paleo-Indian tradition in eastern North America. Curr Anthropol. 

McAvoy, J. M., & L. D McAvoy. 1997. Archaeological investigations of site 44SX202, 
Cacms Hill, Sussex County, Virginia. Virginia Department of Historic Resources, Research 
Report Series No.8. Sandston, VA. 



Mulvaney, )., & ). Kamminga. 1999. Prehistory of Australia. Smithsonian Institution Press, 

Washington, DC. 480 pp. 
Oda, S. 1990. A review of archaeological research in the Izu and Ogsawara Islands. Man 

and Culture in Oceania. 6:53—79. 
Perles, C. 1987. Les Industries lithiques taillees de Franchthi (Argolide, Grece). Tome 

1, Presentadon General et Industries Paleolithiques. Fascicle 3 /;/ T. W. Jacobsen, ed.. 

Excavations at Franchthi Cave, Greece. Indiana University Press. Bloomington and 

Indianapolis, IN. 
Powers, W. R., & J. F. Hoffecker. 1989. The Pleistocene settlement in the Nenana Valley, 

Central Alaska. A.m. Antiquity. 54:263—287. 
Rigaud, J., & J. Simek. 1990. The Last Pleniglacial in the south of France (24,000-14,000 

years ago). Pages 69-89 in L. Soffer & C. Gamble, eds.. The World at 18,000 BP: High 

Latitudes. Unwin Hyman, London. 
Robinson, S., M. MasUn, & I. McCave. 1995. Magnedc susceptibilit}^ variations in Upper 

Pleistocene deep-sea sediments of the NE Atiantic: Implications for ice rafting and 

paleocirculation at the last glacial maximum. Paleoceanography 10:221—250. 
Severin, T. 1978. The Brendan Voyage: Sailing to An7erica in a Teather Boat to Prove the Tegend 

of the Irish Sailor Saints. The Modern Library, New York, NY. 292 pp. 
Schurr, T. 1996. Mitochondrial DNA and the peopling of the New World. Am. Sci. 

Smith, P. E. 1966. Le Solutreen en France. Imprimeries Delmas, Bordeaux. 449 pp. 
Straus, L. G. 1977. Of deerslayers and mountain men: Paleolitliic faunal exploitation in 

Cantabrian Spain. Pages 41—76 in L. R. Binford, ed.. For Theory Building in Archaeology. 

Academic Press, New York, NY. 
. 1979. Cantabria and Vascongadas, 21,000-17,000 B.P.: Toward a Solutrean set- 

tiement pattern. Munib. 3L1 95-202. 

. 1985. Stone age prehistory of northern Spain. Science 230:501—507. 

. 1990. The Last Glacial Maximum in Cantabrian Spain: The Solutrean. Pages 

88-108 in O. Soffer & C Gamble, eds.. The World at 18,000 BP, vol. I. Unwin Hyman, 
. 1992. Iberia before the Iberians: The Stone Age Prehistory of Cantabnan Spain. 

Universit}^ of New Mexico Press, Albuquerque, NM. 336 pp. 
. 2000a. Solutrean settiement of North America? A review of realit\^ Affi. 

Antiquity 65:219—226. 
. 2000b. A quarter-century of research on the Solutrean of Vasco-Cantabria, 

Iberia and beyond. /. Anthropol Kes. 56:39—58. 



Straus, L. G., & G. A. Clark. 1986. La Riera Cave: Stone age hunter-gatherer adaptations 

in northern Spain. Anthropological Research Papers No. 36. Arizona State University, 

Tempe, AZ. 
Tynan, C. T, «& D. P. DeMaster. 1997. Observations and predictions of Arctic climatic 

change: Potential effects on marine mammals. Arctic 50:308—322. 
van Andel, T. J., & J. C. Shackleton. 1982. Late Paleolithic and Mesolithic coastlines of 

Greece and the Aegean./. Field ArchaeoL. 9:445—454. 
Webb III, T, W. E Ruddiman, E A. Street-Perrott, V. Markgraf, J. E. Kutzbach, P J. 

Bartlein, H. E. Wright Jr., & W L. Prell. 1993. Climatic changes during the past 

18,000 years: Regional syntheses, mechanisms, and causes. Pages 514—542 in H. E. 

Wright Jr., J. E. Kutzbach, T Webb III, W. E Ruddiman, E A. Street-Perrott, & P J. 

Bartiein, eds., Global Climates Since the Fast Glacial Maximum. University of Minnesota 

Press, Minneapolis, MN. 
Wormington, H. M., & R. G. Forbis. 1965. An Introduction to the Archaeology of Alberta, 

Canada. Denver Museum of Natural History, Denver, CO. 248 pp. 


Chapter Ten 

The First American Languages 

Johanna Nichols 

^^^^^^ There are about 6,000 languages on Earth, falling into about 300 
J^^^^^^k separate genealogical families and a wide range of structural t}'pes. 
^^**P^^^^B About 150 of the families, and the full range of structural t}^es, are 
native to the Americas. That is, half of the world's Linguistic genet- 
ic diversit}' comes from this hemisphere, a vast bank of the world's linguistic and culmr- 
al wealth and a large body of evidence useful in tracing the origins of languages both 
American and non-American. Many of these American languages, and even several 
entire families, are now extinct, victims of the European conquest and its aftermath. 
Science and humanity are the poorer for these extinctions. However, the evidence avail- 
able — the languages that have been or are being described, ethnonyms and wordlists 
for some that are otherwise unknown, oral traditions about ethnic identity, in some cases 
Uttle more than indications of sheer difference from any other language — makes it pos- 
sible to put together a bare-bones prehistory of the linguistic population of the Western 
Hemisphere. That prehistory seems to be consistent with what the speakers of the lan- 
guages tell us: They have been in this hemisphere for a very long time, and they are above 
all American and not transparently recent imports from abroad. The kind of external ori- 
gin that can be traced appUes not to particular languages and families but, abstractiy and 
probabilistically, to large geographically-based groups of language families. The lan- 
guages of the west have affinities to languages around the Pacific Rim and those of the 
east have affinities farther south, with the languages of Melanesia and Australia. 

The historical linguistic work done so far on the languages of the Americas has most- 
ly involved establishing and reconstructing genealogical families. Such work is limited to 
the time frame within which the comparative-historical method used for detecting relat- 
edness and reconstructing ancestral languages has any useful resolution — up to about 
the average age of the oldest families, defined below, and almost certainly not more than 
about 10,000 years even in the best of circumstances. Such work has almost nothing to 



say about external connections of the native American languages, apart from showing 
that members of the Eskimo-Aleut family have spread from Alaska into Siberia in recent 
millennia and that the Austronesian colonization of the Pacific islands has not brought 
Austronesian languages to the Americas. Similarly, it has almost nothing to say about the 
initial colonizadon of the Americas, other than to point out that the large number of 
separate language families attested in the Americas seems far too high for colonizadon 
models that regard the Clovis culture or something in its time frame as the initial colo- 
nizer. This chapter takes a very different approach, one based on t)^ological comparison 
and linguistic geography, for which external comparison and deep time depths are nat- 
ural tasks, albeit at the cost of greatiy reduced resolution as regards histories of individ- 
ual language families. This approach does not trace or compare individual families at all, 
but deals with larger populations of language families and samples those populations by 
taking, t)^ically, one representative per language family. For statistical comparison to be 
valid, a population needs to be large enough to yield about ten or more sample languages 
(and ideally distinctiy more than ten). To avoid circularity, a population needs to be 
defined strictiy geographically (and not by its component language families). Populations 
compared here include western North America, central to eastern North America, 
southern Mexico and Central America, western South America, eastern South America, 
northeast Asia, southeast Asia, etc. 

The basic method used here is to compare frequencies of some structural element in 
the various populations. Quite often, frequencies of such features are not signiticantiy 
different from population to population, so it is safe to assume that their levels and their 
variation are the result of universals and/or chance. Sometimes, however, the differences 
of frequencies from population to population significantiy exceed what would be expect- 
ed by chance, and when two populations both exceed chance expectation significantiy 
and in the same direction {i.e., both having either significantly more or significantiy fewer 
tokens of the feature than would be expected by chance), then the resemblance is high- 
ly unlikely to be due to chance or universals. This means that chance and/or universals 
should not be our first choices for explanations of the resemblant frequencies. Rather, it 
is safe to hj'pothesize that they are due to some shared liistorv: undetected deep genetic 
connections between some members of each of the populations, or common geo- 
graphical origins of some of the members or indeed of the entire populations, or sepa- 
rate contacts of the populations or some of their members with a tliird part}*. It is gen- 
erally not clear just what the shared history might have involved, but statistically signifi- 
cant sharings nonetheless mean that shared history in the abstract can safely be posited. 
That is the method used here. Below the term affimty will be used of situations in which 
two populations exhibit significant sharings of more than one structural property- (and 



the properties in question are independent of each other). This term contrasts with 
genealogical relatedness and descent, which are used when speaking of sharings due to inher- 
itance from a common ancestor. 

l^inguistic Diversity in the A^m ericas 

The roughly 150 genealogical lineages of American languages are the kind of lin- 
guistic family I call the stock: the oldest traceable family for which standard linguistic 
comparative method can not only establish the fact of relatedness but also detect regu- 
lar sound correspondences and therefore reconstruct sounds, words, and grammar ele- 
ments of the ancestral protolanguage. The tttva familj is a generic term for genealogical 
groups of all kinds (including stocks, which are a particular kind of family). 

The oldest stocks tend to be about 6,000 years in age: Some examples are Indo- 
European, Uralic, Austronesian, Algic, Siouan-Catawba, Uto-Aztecan (the latter three 
American). Many are much younger, but are nonetheless self-standing stocks in that they 
have no provable outside kin and are therefore highest genealogical nodes: Examples are 
Hmong-Mien, Chumashan, and Kiowa-Tanoan (the latter two American). A few are 
older: Nakh-Daghestanian (or Northeast Caucasian) and Austroasiatic (see Blust 1996) 
are examples. Every language isolate (examples are Basque, Burushaski, or North 
American Zuni and Kutenai) is its own stock. It may be useful to distinguish between a 
stock's internal 2ig&, the time elapsed since its initial breakup, and its external 2ig& or lifespan, 
the age up to which it can be traced without detecting external kin. It can be assumed 
that any stock is likely to have deep kin connected to it at an age beyond the external age, 
but we cannot identify them. There is no uncontroversial purely linguistic means for 
determining the internal age of a family, although it is fairly easy to describe families as 
about Romance-Like or about Oceanic-Uke (or Mayan-like) or about Indo-European-like 
in their diversity, using weU-known families as reference points, and in this way to esti- 
mate their internal ages at around 2,000, 4,000, and 6,000 years respectively. Estimating 
stock lifespans is also not uncontroversial, but I will take 6,000 years, the average age of 
old stocks, to be the average stock lifespan. This figure is an average and an approxima- 
tion, and it will be used only for modeling and simulation, procedures in which estimates 
and approximations are appropriate. 

Some language stocks have many daughter languages: An example is Austronesian, 
with over 1,000 daughter languages. Some have few: Examples are isolates, which have 
one descendant each, as well as small families such as Northwest Caucasian (two to four 
living daughter languages and one recentiy extinct) or Yukian (in the Americas; two 
daughter languages). 



Language stocks are nor evcnh- clisrrihLirctl over the L^arth. In the Americas, most of 
the genealogical diversit}' is found in the west. In North y\merica, tu'o-thirds of the lan- 
guage stocks are found along the western edge of the continent, between the coast and 
the coast range ()acobsen 1989). Other high-diversity clusters are in northern South 
America (especially in the Amazon basin), Mesoamerica (southern Mexico and Central 
America), and around the Caribbean. Relatively few families are found in the Great 
Plains, the Northeast, and the Arctic and sub-Arctic. These patterns of diversity' and 
non-diversit\^ are not unique to the Americas but follow general principles that determine 
language-family diversit)-. Diversit}^ is highest at lower latitudes, along coastlines, and in 
simpler societies; diversit)- is low at high latitudes, in continental interiors, in areas of dry 
or cold climate, and in complex societies (it is particularly low in areas with a long histo- 
ry of empire as in much of Eurasia and northern Africa) (Nichols 1990; 1992:232 ff.; 
1998b; map of world language diversit}', 1990:135.) Since the distribution of diversit)' in 
the Americas is the result of climate, geography, and economy, as it is elsewhere, it can- 
not reveal much about the history of colonization of the Americas. 

The yige of the A^merican linguistic Population 

There are several ways in which the statistical parameters of the linguistic population 
of the New World can be used to estimate the age of that population. (It may be impor- 
tant to emphasize that the procedures used here all yield estimates rather than precise 
ages. Therefore they are stated in approximate figures.) One estimate of age is based on 
average rates of language population growth (Nichols 1990, 1992, 1998a, 1998b, 2000b). 
The approximately 150 language stocks of the Americas are the result of immigration, 
subsequent proliferation, and extinction. A plausible rate of immigration from Siberia to 
North America via Beringia (the only entry point) is one language every two or three mil- 
lennia (Nichols 1990), a figure based on attested rates of language succession in Siberia 
and Alaska over the last 5,000 to 6,000 years, language succession elsewhere at high lat- 
itudes, and rates of archeological cultural succession in southwestern Alaska since the 
Pleistocene. These are rates of successful immigrations, as they are based on succession 
rates for languages that have survived to be attested in their daughter languages and cul- 
tures which have survived to spread in Alaska and show up in the archeological record. 
Entry rates during the Pleistocene must have been slower, as glacial conditions made 
entry (and high-latitude living in general) difficult. During the height of glaciation, 
around 16,000—20,000 yr B.P., entry is likelv to have been impossible. 

While language families varv greatlv in their internal ages and in the number ot their 
daughter languages, their modal ages and sizes can be used for modeling purposes. The 



stock lifespan of about 6,000 years, described above, will be used as an estimate of age. 
For stocks and old famiUes, the average number of surviving initial branches is about 1.5 
(Nichols 1990); about one-third of the world's language stocks are isolates. For working 
purposes it can be assumed that each stock represents a branch that split off from some 
ancestor about 6,000 years ago (its own internal differentiation may have begun much 
later) and each such stock has 1.5 surviving branches, each of which develops into a 
stock (with 1.5 branches of its own) in 6,000 years, and so on.^ 

I performed a number of simulations using various rates of entry and proliferation, 
of which only three kinds of scenarios were able to yield approximately 150 daughter 
stocks: scenarios with 10—12 entries beginning 20,000—30,000 years ago (Nichols 1990), 
ones with one immigration about 50,000 years ago and no more after that (ibid.), and 
ones with a few entrants before the height of glaciation, several afterwards, and a faster 
rate of post-entry proliferation for the first entrant (Nichols 2000b). Varying immigra- 
tion and stock proliferation rates within naturalistic parameters made relatively littie dif- 
ference in the outcome. The only simulations with naturalistic assumptions that gave 
appropriate outcomes began before or during the height of glaciation (and recall that 
actual entry during the very height of glaciation was probably impossible, so that a fig- 
ure in this vicinity requires entry before the maximal glacial advance). Though all the 
component figures are approximate, these models indicate strongly that human settie- 
ment of the Americas began over 20,000 years ago. 

The Monte Verde archeological site near Puerto Montt in southern Chile is dated at 
about 12,500 yr B.P. (DiUehay 1997; Meltzer, this volume). Nichols (1998a, 1999, 2000b) 
estimated rates of language spread from just south of the ice sheets to Monte Verde, 
along both coastal and intermontane interior routes, based on attested and recon- 
structible rates of language spread (using only spreads that did not utilize transport such 
as voyaging canoes or horse-drawn wheeled vehicles) through comparable ecologies. A 
beeHne spread to Monte Verde with constant motion at the fastest plausible pre-trans- 
port rates would require at least 7,000 years, thus requiring an immigration by 19,500 
years ago. In this way language spread rates point to entry during or before the height of 
glaciation, supporting the results of simulations based on proliferation rates. Since entry 
during the height of glaciation was probably impossible, the calculations from language 
spread rates can be interpreted as pointing to colonization beginning before the height 
of glaciation. 

Nichols (1992, 1997, 1998a, 1998b, 2000b, 2001; Nichols & Peterson 1996) has 
described stratification in the linguistic population of the Americas. There is a relatively 
recent overlay along the Pacific coast, which is continuous with a similarly coastal popu- 
lation in Asia and Australasia and for that reason is termed the Pacific Rim population. 



Then there is an older population (perhaps with some internal stratification of its own) 
distributed throughout the hemisphere and overlapping geographically with the Pacific 
Rim population. Here the older population will be called the eastern population, since it 
is the only occupant of the eastern parts of both North and South America (though it 
does not occupy only the east but is also found in the west). Evidence that the Pacific 
Rim population is the younger one comes from its more restricted distribution, its local- 
ization in the vicinit}' of the entry point, its continuit}^ with populations which in Asia or 
Australasia spread after glaciation, and its high structural profile with several distin- 
guishing grammatical markers. Especially good evidence is the clumpy distribution of 
those distinguishing markers in the Pacific Rim population. Grammatical features such 
as verb-initial word order and numeral classifiers, which have no inherent linguistic affin- 
it)^ and are not linked outside of the Americas, show a statistically significant tendency to 
co-occur in the same languages in the American Pacific Rim population. This dumpiness 
suggests that accidental combinations of properties that happened to occur in the ances- 
tral languages still co-occur in their descendants. With enough time, language change will 
dissociate some of these clumps. 

The older population has none of these characteristics: its range is the entire hemi- 
sphere, and it has no continuity^ with any nearby Asian or Australasian population, no 
affinity to any population resulting from a postglacial spread, no distinguishing markers 
not also found in the Pacific Rim languages, and no dumpiness of grammatical features. 

The age of the Pacific Rim population will provide a benchmark for computing the 
time of first colonization, since the age of the entire American linguistic population is 
considerably greater than that of the Pacific Rim population. Nichols (in pressj uses a 
sort of survival analysis to estimate the age of the Pacific Rim population. Average rates 
are estimated for inheritance, loss, innovation, and borrowing of several of the distin- 
guishing Pacific Rim markers, and each marker is assumed to be brought in by a single 
ancestral entrant, which proceeds to proliferate at standard rates. Times required to reach 
the attested frequencies of the markers in Pacific Rim stocks range from over 12,000 yr 
to over 1 8,000 and almost 24,000 years. These figures are very rough because actual rates 
of inheritance, loss, innovation, and borrowing must be estimated from a small sample 
and because this technique is new and unrefined. Still, it can be said that these results 
suggest entry of the Pacific Rim population beginning late in the Last Glacial period or 
during the early waning of the ice sheets. On much firmer grounds, the Pacific Rim pop- 
ulation in eastern Asia began to spread out into Melanesia in the same time frame 
(Nichols 1997, 1998:161 ff.). In both hemispheres the Pacific Rim spread seems to have 
represented the expansion and movement of a coastaUy adapted population as the cli- 
mate improved and sea levels changed. 



Another way of estimating the age of the Pacific Rim population is to consider the 
numbers and ages of language families included in it. The North American Pacific Rim 
languages include the large grouping known as Penutian, for which a branching tree is 
shown in Figure 1. (This is based on my reading of DeLancey & Golla [1997] and other 
contributions in the same volume; no tree is given there, but the evidence presented in 
the individual articles, together with most of the prose statements, suggest approximately 
this tree). The subbranch known as California and Plateau Penutian seems to be a clear 
family, though probative evidence has so far been published only for the Miwok- 
Costanoan-Yokuts subset of it (Callaghan 1997; Callaghan's wording is cautious, but the 
evidence she presents impresses me as probative). Callaghan gives ages of Germanic-Uke 
{i.e., ca. 2500 years) for Miwokan (1988:53, cited by Campbell 1997:129) and 4500 years 
for Miwok-Costanoan (Callaghan [1997] on both Linguistic and archeological evidence) 
and perhaps 6,000 years for Miwok-Costanoan plus Yokuts. Golla (2000) considers these 
ages too high. The California-Plateau group consists of this branch plus three others, the 
interbranch articulation uncertain and the dispersal probably achieved by separate and 
successive migrations (DeLancey & Golla 1997 and other contributions in the same vol- 
ume). Given the stock proliferation rate and stock lifespan, addition of three more 
branches in more than one step increases the age of the putative family considerably. 
This California-Plateau group is then possibly related in some fashion to a set of west- 
ern Oregon families, again with uncertain articulation, and the age of the whole is cor- 
respondingly greater. Finally, the Tsimshianic family of central British Columbia is a 
more speculative connection; if related, it further raises the age of the larger group. The 
diagram in Figure 1 attempts to capture the branching structure of the firm core group- 
ing and the likely strucmre if some of the others are taken to be kin. 

At least some part of this large group forms a genuine linguistic family. Perhaps the 
best interpretation is to regard the California-Plateau core as a stock, with likely deeper 
kin whose actual genealogical stams may never be fully specifiable. The older the group- 
ing posited, the less its certainty. Nichols (1998a:163) estimated the age of Penutian at 
about 12,000 years, based on the average rate of stock proliferation; this of course is its 
age if\t is a family, though calculating an age does not prove it is a family. Golla (2000) 
uses a sophisticated relativization of linguistic measures of age to sociolinguistics and 
considers the age of Penutian closer to that of Indo-European, with differences exag- 
gerated by substratal effects and isolation of communities. Campbell (1997:309 ff.) 
rejects genetic relatedness for Penutian and even for the CaUfornia-Plateau core of it. My 
own current impression is that the California-Plateau subgroup group is a likely family 
and close to (and no less than) Indo-European-Uke age, and some of the western Oregon 



California & Plateau 

Coast; Takflman Kalapuyan Chinookaii 

Miwokan Costanoan Yokuts 






Ncz Perce 



Yakonan (AJsea) 

Figure 1. Tree structure for the Penutian languages. Triangle: Probable branch with uncertain inter- 
nal structure. Dashed line: Connection debated. Ellipse: Branch status and internal structure (if any) 
uncertain. Numbers: Approximate dates, based on linguistic evidence, in yr B.P. * = Supported bv arche- 
ological evidence as well. 

languages are possible more distant kin. The Pacific Rim population is then old enough 
to contain this likely deep family and possible deeper kin as one small subpart. 

Nichols (1990) used structural diversity of the linguistic population as a rough mea- 
sure of age for colonized areas. The languages of Australia and those of New Guinea 
show great t}'pological variet}' for most standard metrics, as is unsurprising: These lands 
have been inhabited for about 50,000 years, so there has been ample time for diversity' to 
develop since settlement began. Furthermore, their colonization prehistories have 
involved many entries at different points and have drawn on the structurally and geneti- 
cally diverse linguistic population of western Melanesia and pre-Austronesian insular 
southeast Asia, so the entrants themselves must have been a diverse population. The lin- 
guistic population of the Americas is nearly as diverse as these,- which is surprising, as 
there was a single entry route (via Beringia) and the colonization process drew only on 
the sparse linguistic population of high-latitude Siberia. That the diversity" is so great 
indicates a long period of habitation, long enough for diversit}^ to develop in situ. 

Structural diversiu^, like genetic diversit)^ is lowered in spread zones — areas such as 
continental interiors, deserts, and high latitudes, in which different language families tend 
not to coexist but rather one language tends to spread widely at the expense of all oth- 
ers (for spread zones, see Nichols 1992:12—24). As Australia is almost entirely a vast 



spread zone, its genetic and structural diversity are lower than those of New Guinea, 
despite the identical colonization dates and colonization histories of the two lands. 
Diversity is high in what are called residual ^ones in Nichols (1992; accretion ^nes'm. Nichols 
1997), areas (mostly coastal and piedmont areas, and regions with mild climates) in which 
large language spreads are rare and diversit)' increases with each entrant and continues to 
build up. By accident of geography and climate, the Americas contain extensive stretch- 
es of such suitable territory for the development and preservation of high diversity. The 
Pacific coastline up to the eastern slopes of the coast ranges, the area around the 
Caribbean and the Gulf of Mexico, most of northern and central South America, and 
most of the southeastern U.S. These regions were able to act as a reservoir of diversity, 
preserving representatives of probably the great majority' of stocks that entered or devel- 
oped after colonization. In view of this large diversity bank, survival rates and diversifi- 
cation rates give a good estimate of the time of habitation for the Americas.-^ 

The Geographical Origin of the A^merican linguistic Population 

The Pacific Rim population of American languages has a number of structural affini- 
ties with the languages of the Asian Pacific coast and northern Melanesia, and these 
make it possible to identify a greater Pacific Rim population extending across the 
Melanesian, Asian, and American Pacific coasts. Several grammatical markers are of rel- 
atively high frequency in the greater Pacific Rim population and rare or virmaUy nonex- 
istent elsewhere. Numeral classifiers and personal pronoun systems with first person n 
and second person m are virtually nonexistent outside of the Pacific Rim population, 
verb-initial word order is rare elsewhere, and true cases and suffixal person agreement 
are common in Eurasia but rare in the Americas except in the Pacific Rim. 

The Pacific Rim features are purely typological properties except for the n : m pro- 
noun system, for which it is natural to suspect inheritance and hence to posit deep genet- 
ic relatedness for the language families exhibiting it. Nichols and Peterson (1996) con- 
sider the statistical value of such a system and show that it is not sufficient evidence on 
which to base a claim of genetic relatedness. On the other hand, the difference in fre- 
quency of such pronominals between the Pacific Rim population and others (including 
between the American Pacific Rim population and the American eastern population) is 
statistically significant, which means that the high frequency in the Pacific Rim languages 
is highly unlikely to be due to chance or the operation of language universals. (Genetic 
relatedness and language universals exhaust the explanations offered of this distribution 
in the prior literature.) Nichols and Peterson conclude that the n : m pronoun system is 
the result of a single historical event, though the surviving evidence does not indicate 



what kind oF event. Nichols (2001) argues that patterns like the // ; /// one are non-uni- 
versal phonosymbolic canons sometimes found in small closed sets of words or forma- 
tives, and that they are favored in both inheritance and areal transmission. That is, they 
are fairly unlikely to arise but, once present, are quite likely to be inherited and somewhat 
likely to diffuse; the likelihood of transmission of some kind (inheritance, diffusion) is 
greater than the likelihood of the feature arising in the first place. If this is so, the // .• ni 
system has spread in the Pacific Rim population by both diffusion and inheritance from 
some single source. Some, but not all, of the stocks with some representation of n : m 
systems are likely to be deeply genetically related. The language in which the first /; .• m 
system originated was probably spoken in northeastern Asia (to judge from the frequen- 
cy of phonosymbolic pronominal systems, albeit not of the // .• m t^'pe, in northeastern 
Asia). There are a few tokens of n : in systems in the Asian and Melanesian Pacific Rim 
languages, but not enough to significandy exceed chance. Therefore the first n : m system 
must have arisen in a language or language population on its way into the New World, 
and the spread of the system by diffusion and inheritance took place mostly in the New 

The other Pacific Rim markers, however, are well attested in the Asian and 
Melanesian Pacific Rim languages and not only in the Americas. (See Maps 9 and 10 in 
Nichols 1998b, or Maps 12 and 13 in Nichols & Peterson 1996.) The population bearing 
these markers must have formed somewhere in coastal eastern Asia, and its component 
languages evidently spread in two directions: northward and eventually into the 
Americas, and southward and eventually into Melanesia. (That these two trajectories of 
language movement and human migration existed is bevond doubt.) Now, millennia later. 
Pacific Rim features ring the Pacific from Melanesia to eastern Asia to the western 
Americas. It is probably worth emphasizing that this pattern is the result of movement 
along coasts (and, in Melanesia and Australasia only, movement from island to island) 
and not overwater contacts between Asia and the Americas. The human settlement of 
the Polynesian islands (along which any such overwater contacts would have had to 
occur) took place only in the last tw^o millennia, and before that the Polynesian islands 
were uninhabited and almost certainly untouched by humans [e.g., Kirch 2000). 

Markers of the eastern population include head-marking morphology (and in partic- 
ular the otherwise rare radically head-marking t)^e), prefixal person agreement, identical 
singular and plural stems in personal pronouns, and simple consonant systems (some- 
times containing only a single manner of stop articulation). These feamres are well attest- 
ed in Melanesia and especially Australasia. Head marking is faiiiv common in New 
Guinea and northern Australia; prefrxal agreement likewise; identical singular and plural 
stems likewise; and simple consonant systems are relatively frequent in New Guinea and 



nearly universal in Australia. Unlike the Pacific Rim features, these markers are common 
in interior (highland) New Guinea and Australia, i.e., farther from the (northern, coastal) 
entry point than the Pacific Rim features, so that in their geography they appear to be 
earlier entrants than the Pacific Rim markers. They are missing from the coastal east 
Asian staging area that fed the migratory trajectories leading on the one hand to 
Melanesia and on the other to the Americas. In aU these respects they appear to be mark- 
ers of earlier colonizers, now found only at the far ends of the two trajectories. 

The best interpretation of aU the facts would seem to be that both strata of the 
American linguistic population originated somewhere in eastern Asia (probably in the 
same general area but at different times). Colonizers of America at all times appear to 
have mostly been coastaUy adapted people. There is no linguistic evidence for a sub- 
stantial contribution from inner or central Asia to the American linguistic population. 
The main connections are to Southeast Asia and Melanesia. 

Post-colonisation Developments 

After entering North America, the Pacific Rim population evidently spread coastally, 
eventually extending to southern South America and hugging the coast with almost no 
eastward spread. In the last few millennia, a north-to-south trajectory is the most com- 
mon one for language families in North America on or near the coast. The coastal 
Athabaskan languages and most of the CaUfornia-Plateau Penutian groups have all 
spread mostly southward from their points of origin. (For languages and families men- 
tioned here and below, see Campbell 1997: chapters 4—6 and Goddard 1996. Their ranges 
are shown on the map accompanying Goddard 1996. For Athabaskan see also Leer 1991. 
For Penutian see also DeLancey & GoUa 1997.) The Athabaskan family originated in the 
subarctic (southern Alaska to northern British Columbia according to Leer 1991), and its 
Pacific Coast languages extend Uke beads on a string from Washington to northern 
California. The Penutian languages probably originated on the Columbia Plateau and 
spread, in a series of migrations, into Oregon and California. This history is accepted for 
the Wintun, Maidun, Miwok-Costanoan, and Yokutsan branches (Whistier 1977) and is 
a plausible trajectory for the possibly Penutian languages of western Oregon regardless 
of whether they are actual genetic kin. 

If the pre-Pacific Rim entries were also coastal — and it seems increasingly unlikely 
that any interior corridor was open to immigration and inviting to immigration during 
glaciation — then it is likely that once in North America they too spread coastally south- 
ward. This would mean that they spread from inhabited to uninhabited lands and that 
they kept to the coastal ecologies with which they were most familiar. Systematic settie- 
ment of the interior must then have begun after the end of glaciation. The spread of the 



(Hox'is cullurc (sec Mclrzer, this volume) marks one of rlie first expansions info the inte- 
rior, and perhaps the very first. 

Colonization of the interior is likely to have come from the south. One piece of evi- 
dence for northward spread into the interior is the fact that in historical and recon- 
structable times the language families of the mid-latitude interior of North y\merica have 
mostlv spread from south to north. Prominent examples are the Uto-Aztecan family, 
which spread from approximately northern Mexico to southern California and the Great 
Basin (again, see relevant sections of Goddard 1996 and (Campbell 1997; also Madsen 
and Rhode 1994); the Siouan-Catawba family, which spread probably from the vicinit)- 
of the lower Mississippi to form enclaves in various parts of the southeastern U.S. and 
well north along the Missouri, Mississippi, and Ohio valleys (the northernmost members 
are Stoney and Assiniboine, both spoken in southern Canada); and Caddoan, which 
probably also originated in the vicinit)^ of the lower Mississippi and spread far north into 
the Great Plains (its northernmost language, Arikara, is spoken in North and South 
Dakota). (Of these, the Uto-Aztecan spread in the Great Basin was a spread of non- 
farming people; the others were spreads of agriculturalists.) Even the Algonquian fami- 
ly, which spread primarily eastward probably from the vicinit}' of the Columbia Plateau 
(Goddard 1994), once east of the Rocky Mountains underwent considerable northward 
expansion so that in liistorical times it occupied much of central Canada. If Leer is right 
about the west central Canadian origin of the Na-Dene family, then its mostlv interior 
Athabaskan branch has spread far north of tliis origin. The only important exception to 
the mostiy northward drift in the interior is offered by Navajo and the Apachean lan- 
guages spoken around the periphery of the pueblo area in the southwestern U.S.; they 
are the result of southward migrations from southern Canada witliin the last millenni- 

These examples of nortiiward drift in the interior are based on language-family 
spreads, and pertain only to the last several millennia. Uto-Aztecan and Siouan-Catawba, 
two old language families of the interior, are both of about Indo-European-Uke age, i.e., 
about 6,000 years old at most. Movements in this time frame do not directly testify to 
directions of movement shortly after the end of glaciation, and furdiermore the largest 
northward spread, that of the Siouan language family, involved farming people. Still, it is 
more parsimonious to assume that this northward drift continues a long-standing ten- 
dency rather than a recent development. Then the standing northward drift among lan- 
guage families in interior North America may provide indirect evidence that humans 
claimed interior North America after glaciation in northward movements. The source of 
these movements must have been in the vicinits' of the Gulf of Mexico. 



There has also been some west-to-east movement bringing representatives of coastal 
and near-coastal families into the interior. The Athabaskan-Eyak-Tlingit family"^ origi- 
nated near the coast or on the eastern slope of the coast range somewhere from south- 
ern Alaska to central British Columbia, and its Athabaskan branch spread mostly east- 
ward to cover most of interior central and northern Canada and all of interior Alaska. 
The Salishan family originated on or near the coast in northern Washington to southern 
British Columbia, and its interior representatives extend to Idaho and Montana. The 
Algic family probably originated on or near the Columbia Plateau, and its descendants 
now comprise Yurok and Wiyot of coastal northern California and the geographically 
vast but temporally shallow Algonquian family, which extends from the western Great 
Plains (Plains Cree, Blackfoot, Cheyenne, Arapaho) to the eastern seaboard (Montagnais, 
Micmac, Delaware, etc.). Thus the entire non-Pacific part of North America — the 
Great Basin, the intermontane area, the Great Plains, the eastern woodlands and 
seaboard — have received some of their language population from the far west, while 
the converse does not hold (Jacobsen 1989). 

To summarize, the settlement of the Americas involved three main trajectories. 
Immigrants from Siberia spread southward, chiefly along the coast; from time to time, a 
language spread eastward into the interior; and, since the end of glaciation there has been 
northward and eastward movement into interior North America. 

Network analysis of the distribution of structural markers in the Americas supports 
this general picture of multiple trajectories and a secondary hub in the vicinit}' of the 
Gulf of Mexico. Nichols (2000a) divided the American languages into five large areal 
populations (western North America extending to the Rockies, eastern North America 
from the Rockies to the Atlantic, Mesoamerica, i.e., southern Mexico and Central 
America, western South America, and eastern South America) and compared the fre- 
quencies of various structural features, comparing each areal population to all the others 
and to 12 areas of comparable size worldwide. For example, the proportion in each area 
of languages of the radically head-marking type was compared (as high, moderate, or 
low relative to the worldwide mean), and two areas were judged to have an affinit)- with 
regard to that marker if they both had high frequencies. The higher the number of such 
affinities, the stronger the overall affinity of the areas. Languages used for comparison 
were those in a 200-language family-based worldwide sample. In the comparison of the 
American areas to others, no strong evidence of Siberian affinit)^ appeared. The five 
American areas, individually and collectively, find their closest affinities in Melanesia, 
with northern Asia a distant third or fourth. A diagram of the connections among just 
the American groups is shown in Figure 2. The five American areas form a netu^ork with 
a hub centered in the vicinit}' of the eastern Gulf of Mexico and/or the eastern 



Caribbean, from which connections reach out in various directions. Mesoamerica, east- 
ern North America, and eastern South America are a closely linked network, with east- 
ern North America in particular so closely connected that it is virtually an offshoot of 
Mesoamerica. Western North America is more distant, and v^-estern South America still 
more distant and very much off by itself. There is as yet no way to assign a chronology 
to the formation of this network; it could reflect the spread of agriculture from its 
Mesoamerican origin, colonization of the interior after the end of glaciation, or both. 
Regardless of its exact chronology, the network diagram shows that the main factor in 
the distribution of linguistic elements in the Americas was not the process of immigra- 
tion from Siberia but much later developments reflecting the formation of and expan- 
sion from a hub far to the south of the Alaskan entry point. 

Eastern North Americc 




South America 

South America 

Figure 2. Network of interareal linguistic connections in the Americas. Based on 8 non-Pacific Rim 
structural features. One line denotes a shared liigh frequency for one feature. The length of a line 
reflects distance in schematic space. 

The languages of the First A^ffested Americans 

Unless there is inscripdonal evidence of the language, there is generally no way to 
know what language was spoken by a given prehistoric individual or culmre. On the 



other hand, it can be asked, for any language or language family, whether its ancestor 
might have been spoken by some particular ancient society, and with what likelihood. 
This section is a rather speculative attempt to list some of the language families whose 
distant ancestor might have been spoken by each of four ancient individuals or cultures: 
the Kennewick Man, the Monte Verde people, the Channel Islands Woman, and the 
Clovis people. Of course, there is always the possibility that the language of any of these 
has gone extinct leaving no modern descendants. 

The Kennewick Man died a littie less than 10,000 years ago near the Columbia-Snake 
River confluence and, thus, on the Columbia Plateau and close to the coastal area. The 
Columbia River itself has served to connect the two cultural provinces in historical and 
reconstructable times, and must have done so in the Kennewick Man's time as well. In 
this time frame and this area his language could have belonged to either the Pacific Rim 
or the eastern population. His language might well have been the ancestor of Penutian 
(or whatever part of that group may eventually prove to be a family), or the distant 
ancestor of the Algic family, both of which are regarded as having spread from the 
Plateau. His language might have been a pre-Penutian, pre-Algic dispersant from the 
Plateau, in which case its descendant(s) might be found anwhere beyond the frontier of 
either Penutian or Algic — southern California, Washington, British Columbia or, given 
the vast eastward stretch of Algonquian, virtually anywhere in intermontane or eastern 
North America. Or his language might have been caught in the northward drift of the 
interior languages and moved far north; it could, for example, have reached northern 
British Columbia or Alaska and ended up as Proto-Eyak-Athabaskan-THngit (some of 
whose descendants have moved south to end up not far from the Kennewick Man's rest- 
ing place). A pre-Penutian dispersal along the Penutian trajectory, into interior California, 
could have produced any of the Hokan languages of the northern and central California 
hiUs and mountains (these include the Pomoan family, Shasta, Washo, Karok, and oth- 
ers). Or, moving at the rate at which the Pacific Coast Athabaskan languages traveled 
from southern California to northern California, or the rate at which the various 
Penutian languages of California have spread, his language might have moved southward 
to anywhere from Mexico to southern Central America or even South America. In short, 
the Kennewick Man's language might with equal plausibility have been ancestral to near- 
ly any language family of North America and possibly Central and South America as 
well. That is, a few different stocks might descend from his language, and nearly any 
American stock is a candidate for one of those descendants. We wiU almost certainly 
never have any indication as to just which stocks those might be. 

The Channel Islands Woman Lived about 13,000 years ago on Santa Rosa Island off- 
shore from southern California. These islands hosted languages of the Chumashan and 



Uto-Aztecan families in the historical pcrif)d, and Santa Rosa Island was specifically 
Chumashan-speaking. The Uto-Aztecan family originated probably in interior northern 
Mexico some five millennia ago. The Chumashan family is a very shallow one which orig- 
inated on the mainland. Though it is possible that the distant ancestor of Proto- 
Chumashan was spoken where its distant descendant was, it is not likely that anv one lan- 
guage has stayed in place for so long. That is, the Channel Islands Woman is not likely 
to have spoken anything ancestral to modern Chumashan, though the possibilit\' cannot 
be excluded. Similarly, it is unHkely, though it cannot be excluded, that a distant ancestor 
of Uto-Aztecan was spoken here and its descendant ended up many millennia later in 
Mexico as Proto-Uto-Aztecan, whose own later descendant happened by coincidence to 
end up 13,000 years later in a place where the distant ancestor had been spoken. y\ny 
other language family in the Americas, but especially the western Americas, might with 
equal likelihood descend from the early Channel Islands language. At 13,000 years ago 
the island language might have been one of the very first Pacific Rim languages to spread 
south, or it might have been a language of the earlier, eastern populadon that remained 
in the coastal area during and after the Pacific Rim spread. 

The Monte Verde people, as far south as they were at 12,500 years ago, could hardly 
have spoken a Pacific Rim language; at that time the Pacific Rim languages had only just 
begun their spread into coastal North America. Too Little is known about trajectories of 
language spread in South America to support a guess as to where the descendants of the 
Monte Verde people's language might now be. As a language near the far end of the col- 
onizing thrust and not in an obvious spread zone that might have impelled its descen- 
dants far in various directions, it is relatively likely to have gone extinct. 

The Clovis people, beginning about 11,500 years ago, spread out rapidly over a large 
area in the interior southern half of the U.S. (see Meltzer, this volume). Their range is 
much larger than usual for languages of pre-farming, pre-transport societies at its lati- 
tude, though it is comparable to the range of Cree, the northernmost Algonquian lan- 
guage, which extends across eastern and central Canada. Of course, conditions in the 
Clovis range at the Clovis time must have been rather like those in central Canada in 
recent millennia. That is, the Clovis range was almost certainly a spread zone, and the 
Clovis people could have spoken a single language and are Likely to have spoken one or 
more languages belonging to a single family. The Clovis language (s) could have left 
descendants in one or more places in or at the frontier of the range of the Clovis cul- 
ture, and these could have been caught up in the trajectories of northward drift, coast- 
ward drift, and southward movement along the coast. The language of the Kennewick 
Man might well, on chronological and geographical grounds, have been a descendant of 
a Clovis language. By now, virtually any language of North America, Mexico, or Central 



America, other than those with Pacific Rim features, might with equal plausibility 
descend from the language (or languages) of the Clovis culture. 

The languages of these four individuals or cultures, if they survived, have had about 
two 6,000-year stock lifespans in which to reproduce. The language of the Kennewick 
Man and (especially) that of the Clovis people were spoken in important spread zones 
and are Likely to have had more than the average 1.5 descendant stocks. The language of 
the ancient Channel Islands and ancient Monte Verde was spoken in residual zones and 
are likely to have had fewer than the average number of descendant stocks, in view of its 
geography. The Monte Verde language, as mentioned, is the most likely of these four to 
have gone extinct. 

The Clovis language in particular could have had several surviving descendants, as 
has generally occurred with languages spoken by societies entering new ecological or 
economic niches. Each of these branches would then have given rise to its requisite stock 
or two, so that by modern times a number of different stocks — perhaps a dozen — 
might be descendants of the Clovis language. At least half might still be in North 
America. At a time depth like 11,500 years one would expect the descendant stocks to 
retain some suggestive resemblances, though not precise enough or numerous enough to 
firmly indicate genetic relatedness. This hypothesis can therefore be suggested to 
Americanists: Half a dozen language stocks, scattered widely around North America, 
plus a few more stocks from points south, descended from the Clovis language. They 
belong to the eastern linguistic population (though some of them may be physically 
located in the coastal area), and lack most or all Pacific Rim feamres. The northernmost 
stocks of this ancient family have mostly been swallowed up in the spread zone of the 
northern North American interior which is responsible for the Athabaskan and 
Algonquian spreads in the northern U.S. and Canada (though Proto-Na-Dene, the ances- 
tor of Proto-Athabaskan, could on geographical and typological grounds belong to the 
Clovis descendants), but the vast accretion zones of the west and southeast should have 
preserved several others. 

It is also possible that the Kennewick Man, the Channel Islands Woman, and any ran- 
domly chosen individual of either the Monte Verde or Clovis culmre was bilingual or 
multilingual, and /or that their larger communities were bilingual or multilingual. The 
Columbia Plateau is a center of not only language spread but also language contact, and 
of all four cases the Kennewick Man is most likely to have been bilingual and to have 
belonged to a bilingual community, a consideration that doubles the number of his pos- 
sible linguistic descendants. The Channel Islands Woman, as a member of a coastal cul- 
ture, is Ukely to have lived in societies so small that intermarriage was common, and is 
particularly likely to have married into a communit)' speaking a different language from 



luT first language and to have raised children bilingual in her language and that of her 
husband. This consideration doubles the number of her possible linguistic descendants. 
All in all, each of these four ancient languages might have a few surviving descen- 
dants among the 150 stocks and 1,000+ languages of the native Americas, and that of 
the C^lovis people might have several. Conversely, virtually any recent native American 
language has a small probabilit)' of having descended from one or another of these 
ancient languages. Descendants of the Channel Islands language are most likely to be 
Pacific Rim languages, descendants of the Monte Verde and Clovis languages arc almost 
certainly not Pacific Rim languages, and descendants of the Kennewick Man's language 
could equally well belong to the Pacific Rim or eastern populadons. 


From the perspective of the method used here, the main events or situations that 
have shaped the identit}? of the American linguistic population are their origin in the lan- 
guage population of Asia, chiefly coastal Asia; an early date of first colonization of the 
Americas; the appearance of the Pacific Rim population in the Americas as the ice sheets 
waned, and its rapid southward spread along the coast to southern South America; the 
large areas of high linguistic diversity' in the Americas, and the consequent interaction 
and convergence that have resulted from long contacts among different languages; the 
probably long-standing importance of the lands around the Gulf of Mexico as a source 
of linguistic influence and movements; the expansion of the Clovis language or lan- 
guages shortiy after the end of glaciation, and its probable (but so far undetected) dis- 
proportionate representation among the language families of the Americas, particularly 
North America; the development of agriculture in Mesoamerica and (independentiy) 
South America; and the probable spread of languages out from these centers of agri- 
culture. The origin of the linguistic population in Asia plays a small part in the distribu- 
tion of structural features among the American languages. The primary shapers of their 
structural t)'pes, their geographical distribution, and their patterns of genetic relatedness 
have been post-colonization factors: the American geographv, the centers of spread and 
diffusion that formed as a result of that geographv, the interaction among languages in 
those centers, the rise and spread of agriculture, and the recent histories of various lan- 
guage families. It is in this sense that the American languages must be regarded as first 
and foremost indigenous and American. 



yibs tract 

At first European contact a tremendous number and variety of languages 
were spoken in the Americas: over 1 ,000 languages falling into nearly 150 sepa- 
rate families, and exhibiting a great diversity of structural types. This linguistic 
diversity sheds light on the origin and prehistory of the speakers of the languages. 
Where did they come from? Did they come in many immigrations, or has the diver- 
sity developed since entry? How long have they been here? Which of them have 
been here longest? Where and how did they enter, and what route did they travel 
once they entered? What languages ivere spoken by the earliest archeologically 
attested Americans? 

linguistic comparison can give increasingly firm answers to these ques- 
tions. The languages (and their speakers) came from Siberia, and they had spread 
into Siberia from the south, perhaps from coastal southeast Asia. They developed 
here by both i?7imigration and slow divergence. The longest-resident types can be 
identified by their grammatical structures and geographical distribution. Entries 
began at least 20,000 years ago. People entered via the Alaskan coast and spread 
coastally southward more often than inland. We cannot identify the individual 
languages spoken by particular ancient peoples, but ive can make probabilistic 
statements about their possible modern descendants. 

The evidence for earlier and later immigrants is subtle and requires close 
examination, as the languages have been shaped and situated primarily by inter- 
actions with each other and by long residence in their present locations. The native 
American languages are first and foremost indigenous and American, and the 
ancestors of even the most recent entrants ivere in America before the ancestor of 
any present European language is likely to have entered Europe. 


^ Both the immigration rate and the proliferation rate are survival rates, i.e., actual 
immigration or proliferation minus extinctions. Nettle (1999) incorrectly assumes they 
are attempt rates rather than success rates, and that both the average stock age and aver- 
age proliferation rate are precise figures and constants rather than approximations and 

2 In their relatively high frequencies of head-marking t^'pes and inclusive/exclusive 
pronouns, the American languages show founder effects reflecting the limited and 
broadly Asian sources of their colonizers. The founder effects shift the frequencies of 
various types, but do not reduce the overall range of types. 



^ Nettle (1999) proposes a model whereby in a given area or continent genetic diver- 
sit)' increases for a short time and then begins to decrease, so that less diversity- points to 
greater age. This might be correct if all continents were like Australia — conducive to 
spreading and not large — so that a combination of spreading and convergence could 
decimate their language stocks. Australia is unusual, however, and except in the case of 
small islands most colonized areas had not yet reached the saturation point, when diver- 
sit\' starts to decline, bv the time of European contact. Even if Nettle is right that diver- 
sit}' eventuallv starts to level off, it does so only in response to ecological and economic 
limits that are so much more important in determining diversit}' that a model based on 
them alone quite efficiently predicts existing linguistic diversit}'. 

^ Often called Na-Dene, though that label is often used of a putative larger group 
also including Haida. 

l^iterature Cited 

Blust, R. 1996. Beyond the Austronesian homeland: The Austric hj^othesis and its impli- 
cations for archaeology. Pages 117—140 in H. Goodenough, ed., Prehistoric Settlement of 
the 'Pacific. Trans. A.m. Philos. Sac. 86(5). 

Callaghan, C. 1988. Proto Utian stems. Pages 53-57 in William Sliipley, ed., /// Honor of 
Mary Haas: Papers fro?n the Haas Festival Conference on Native A.merican linguistics, Mouton 
de Gruyter, Berlin, Germany. 

. 1997. Evidence for Yok-Utian. Int J. Ani. Ling. 63:1.18-64. 

Campbell, L. 1997. American Indian languages: The Historical Tingiiistics of Native America. 
Oxford Universit}' Press, Oxford, England. 512 pp. 

DeLance}', S., & V. Golla. 1997. The Penutian h\^othesis: retrospect and prospect. ////. /. 
Amer Ling. 63:171-202. 

Dillehay, T. D. 1997. Monte Verde: A Late Pleistocene Settlement in Chile: The Archaeological 
Context. Vol. II. Smithsonian Institution Press, Washington, D.C. 1071 pp. 

Goddard, I. 1996. Handbook of North American Indians, vol. 17. Language. Smithsonian 
Institution, Washington, D.C. 957 pp. 

Golla, V. 2000. Language historv and communicative strategies in aboriginal California 
and Oregon. Pages 43-64 in O. Ahyaoka & M. Oshima, eds., Languages of the North 
Pacific Rim, Vol. 5. Facult\' of Informatics, Osaka Gakuin Universit}; Suita, Japan. 

Jacobsen, W H., ]r. 1989. The Pacific Orientation of Western Noifh American Languages. 
Presented at First Circum-Pacitic Prehistory Conference, Seattie, WA. 

Kirch, P. V. 2000. On the Road of the Winds: An Archaeological History of the Pacific Islands 
before European Contact Universit}' of California Press, Berkeley, CA. 424 pp. 



Leer, J. 199L Evidence for a northern Northwest Coast language area: Promiscuous 
number marking and periphrastic possessive constructions in Haida, Eya, and Aleut. 
ht. J. Am. Ung. 57:2.158-93. 

Madsen, D., & D. Rhode, eds. 1994. Across the West: Human Population Movement and the 
Expansion of the Numa. University of Utah Press, Salt Lake Cit\r, UT 255 pp. 

Nettle, D. 1999. Linguistic diversity of the Americas can be reconciled with a recent col- 
onization. Proc. Natl. Acad. Sci. 96:32—39. 

Nichols, J. 1990. Linguistic diversity and the first settiement of the New World, luinguage 

. 1992. 'Linguistic Diversity in Space and Time. University of Chicago Press, Chicago, 

IL. 358 pp. 

1997. Modeling ancient population structures and movement in linguistics. 

Annu. Rev. AnthropoL 26:359-84. 

. 1998a. The first four discoveries of America. Presented at AAAS annual meeting, 

Philadelphia, PA. 

1998b. The origin and dispersal of languages: Linguistic evidence. Pages 

127-170 in N. G. Jablonski & L. C. Aiello, eds.. The Origin and Diversification of 
Language. Memoirs of the California Academy of Sciences No. 24, San Francisco, 

1999. Languages entering new landscapes. Presented at "Entering New Landscapes", 

University of Florida. 
. 2000a. Linguistic evidence on the peopling of the Neiv World. Presented at AAAS 

annual meeting, Washington, D.C., February 18, 2000. 
. 2000b. Estimating dates of early American colonization events. Colin Renfrew, 

Pages 643—664 in A. McMahon & L. Trask, eds.. Time Depth in Historical Linguistics, vol. 
2. McDonald Institute for Archaeological Research, Cambridge. 
. 2001. Why "me" and "thee"? Pages 253—276 in Laurel Brinton, ed.. Historical 

Linguistics 1999. Benjamins, Philadelphia, PA. 
. in press. Diversit}^ and stability' in language. In B. Joseph & R. Janda, eds. 

Handbook of Historical Linguistics. 

Nichols, J., & D. A. Peterson. 1996. The Amerind personal pronouns. Language 

Whistler, K. W 1977. Wintun prehistory: An interpretation based on linguistic recon- 
struction of plant and animal nomenclature. Berkeley Linguistics Society 3:157—74. 


Chapter Eleven 

A Mitochondrial Perspective on the 
Peopling of the New World 

D. Andrew Merriwether 

^^^^^^^ In this volume, much has been discussed about the Pleistocene peo- 
J^^^^^^k pling of the New World, and the different sorts of evidence that 
^"^i^^^^pB could be brought to bear on questions about the number of migra- 
tions, origins of the founders, routes of entry, and especially the 
timing of entry of human populations into the New World. Among the major theses 
proposed in this volume, and elsewhere, are the following somewhat conflicting scenar- 
ios: 1) There were three waves of migration into the New World, corresponding to the 
Amerind, Na-Dene, and Eskaleut language groupings as described originally by 
Greenberg et al. (1987) and Greenberg (1986). 2) There was a single wave of migration 
into the New World, followed by differentiation of languages and peoples in situ, as 
argued by Merriwether et al. (1995), Kolman et al., (1995), Forster et aL (1996) and others. 

3) There was a separate pre-Columbian European migration into the New World, possi- 
bly by boat along the ice-sheets of the North Atlantic, as argued by Stanford (this vol- 
ume), and possibly supported by Brown and colleague's (1998) X mtDNA haplogroup. 

4) The Clovis hunters were the first humans in the New World (see discussion in Meltzer 
1993; Meltzer & Dillehay 1998). 5) There were pre-Clovis, non-Clovis peoples in South 
America before and concurrent with the first Clovis sites in North America, represent- 
ing an earlier migration. 6) There was a coastal migration, probably making use of boats 
(Erlandson, this volume; Nichols 1995). 7) There was an overland migration through an 
ice-free corridor in western Canada (see discussion in Meltzer 1993). 8) The founders 
came from the Mongolia/Southern Siberia/Manchuria/Lake Baikal region of 
Northeastern Asia (Merriwether et al 1996; Kolman et aL 1995; Neel et al 1994; Schurr 
et al 1 999; Starikovskaya et al 1 998). 



Haplogroups, Haplotypes and Founders 

A brief description of mtDNA nomenclature and definitions of the ubiquitous 
mtDNA haplogroups in the New World are necessary before the above scenarios can be 
evaluated using mtDNA data. First of all, a haplotype is any pattern of linked polymor- 
phisms. Since there is no recombination in mtDNA, and since mtDNA is strictly mater- 
nally inherited (Case et a/. 1981; Merriwether et a/. 1991; Giles et a/. 1980), any polymor- 
phisms found on the mtDNA molecule from one individual may be grouped together 
into a pattern called a haplot}^e. This may be either a stretch of continuous sequence of 
mtDNA, or it may be a collection of restriction fragment length polymorphisms 
(RFLPs) from across the entire 16569 nucleotide sequence spanning the circular mtDNA 
molecule (Anderson et a/. 1981). A haplogroup is a collection of closely related haplo- 
t}"pes that all share one or more mutations in common as compared to other hap- 
logroups. In the New World, Wallace and colleagues (1985) were the first to show that 
Native Americans had a subset of mutations found in Asia, and at drastically different 
frequencies, implying a "dramatic founder effect" between Asia and the New \Xbrld. 
Schurr and colleagues (1990) showed that 97% of the variation found in three popula- 
tions from North, Central and South America fell into four distinct clusters. Wallace and 
Torroni (1992) labeled these clusters A, B, C and D. These were initially defined by 
RFLPs and one 9-base pair deletion. Specifically, haplogroup A was defined as having a 
Hae III restriction site gain at nucleotide (nt) 663. No other Native American groups pos- 
sess this mutation. Similarly, haplogroup C was defined by the loss of a H//?c II restric- 
tion site at nt 13259, and haplogroup D was defined by the loss of an A/u I restriction 
site at nt 5176. Haplogroup B was defined by a deletion of 9-base pair of DNA from 
Region V of the mtDNA genome (a short noncoding stretch between the genes for 
cytochrome oxidase II and the tRNA for lysine). Most humans have tw^o copies of this 
9-base pair sequence tandemly repeated, but those in haplogroup B only have one copy 
and are referred to as 9-base pair deleted. Haplogroups A and B (and Brown's X) lack 
cuts at Ah 1 10397 and D^e 1 10394, while haplogroups C and D (and X6 and X7) cut 
at both sides. A/// 1 10397 is very Asian- specific, and is found only in populations or indi- 
viduals of Asian descent (including Pacific Islanders and Native Americans). The Hinc 
II 13259 site loss is always accompanied by the 10394 and 10397 site grains in the New 
World and Asia. While some populations outside of Asia possess the 13259 loss, they do 
not cut at 10394 or 10397. The 9-base pair deletion appears to have arisen a number of 
times, including at least once (and probably several times) in Africa (SoodyaU ef a/. 1996; 
Chen et a/. 1995; VigHant et a/. 1989, 1991; Redd et al. 1995) and at least twice in Asia 
(Ballinger et al. 1992; Torroni et al. 1993a, 1992; Redd et al 1995). However, in Asian- 



derived populations, 99% of those with the deletion are derived from a single deletion 
event, probably 30,000 to 60,000 years ago. The other deletions tend to be rare. In addi- 
tion to the RFLP markers, there are point mutations in the control region (the noncod- 
ing region of the mtDNA located roughly between nts 15975 and 430) that also can be 
used to define the A, B, C, D and X clusters (Merriwether et al. 2000a; Torroni et al. 
1993a; Ward et al. 1991; Ginther et al. 1993; Kolman et al. 1995, 1996; Lorenz et al 1997; 
among many others). 

Which Way? 

So what can mtDNA (or genetic evidence of any kind) say about these h)^otheses? 
Of the scenarios above, proving routes may be the most difficult using genetic data, 
although a caveat would be that ancient DNA from humans along these routes could 
provide powerful evidence for or against some of these ideas. Nonetheless, we can see 
if the mtDNA data are consistent or inconsistent with any of the h^'potheses. Certainly 
99% of all Native American DNAs, from North, Central and South America are 
descended from t}^pes found in Asia. There are between 5 and 10 haplot)^es that have 
been argued to be shared between the New World and Asia, which belong to 5-7 hap- 
logroups (A, B, C, D, X6/X7, Brown's X). Five of these haplogroups appear to be wide- 
ly accepted (A, B, C, D, and Brown's X) while two of the (X6/X7) are not as universal- 
ly accepted and are closely related to C and D (see Merriwether 1999 and Merriwether et 
al 2000). About 98% of all Native Americans sampled to date are A, B, C, or D (as 
defined by Wallace & Torroni 1992). The remaining 2% appear to be X6/X7 (as defined 
by Easton et al 1996 and Merriwether & Ferrell 1996) or Brown's X (Brown et al. 1998). 
So the vast majorit}^ of all living Native Americans, regardless of tribal origin, language, 
or geographical location in the New World, are a variant of A, B, C, or D (see Figure 1). 
The ancient picture is much, much spottier, but almost everyone that has been fuUy t)^ed 
for founding Uneage markers in ancient populations has been shown to be A, B, C, D, or 
one of the X's. Coastal routes and oversea routes may leave no archaeological trace, espe- 
cially if what was the coast is now underwater (or far inland). Certainly the proposed 
cross-Atlantic ice sheet route cannot be tested by almost any method. 


The idea of using a molecular clock to date the migrations into the New World is 
somewhat problematic when the time scale is so short (in evolutionary time 
10,000—30,000 years is very short indeed). This is because the molecular clock assump- 
tion requires mutations to be random, not under selection (except for purifying selection 

















Figure 1. Frequencies of the founding liaplogroups in the Americas. 

against extremely deleterious mutations), and accumulating in a roughly clock-Like fash- 
ion in all lineages. The problems with these assumptions are: 1) The clock is less accu- 
rate over short time scales. This is because mutations do not really occur at set intervals. 
Rather, the biochemical changes that create mutations are caused bv errors made in repli- 
cation and by environmental insults to the DNA and/or fault}' repair of the damage. 
These different sources of error all plav into the mutation rate. So if we expect to see 
one mutation every 6,000 years, as for some parts of the mtDNA, then on average after 



6000 years in a population, we would expect most people to be one mutation different 
from each other, but many wiU have had none and many may have had two. Either way, 
over longer periods of time, the average number of mutations per year per nt becomes 
a stable number, assuming other factors are held constant. 2) A rapidly growing popula- 
tion in a new niche, without population pressures, may allow new mutations to enter the 
population more easily. This is because the new populations would be small, so any new 
mutation is automatically at higher frequency than if it had occurred in a large popula- 
tion {i.e., a new mutant in a mtDNA population of 10 is now at 10%, while a new mutant 
in a population of 1000 is at 0.1%). If everyone is having lots of offspring, since there 
is endless space for them to live (at least early on in the peopling event), a new mutation 
is likely to increase rapidly. 3) As with fossils, we may always find some sequences that 
are divergent from all others that were missed initially due to inadequate sampling. There 
are currently estimated to be over 2.5 miilUon Native Americans in North America 
(Thornton 2000) and probably 10 million or more in North, Central, and South America 
combined. At the time of contact, Butzer (1992) placed the total population of the New 
World at 43—65 million people. By the 1500s, the hemisphere as a whole was estimated 
to have experienced an 89% reduction in the Native American population (Denevan 
1992). So we have mtDNA sampled a few tenths of a percent of the entire living Native 
American population, and less than a hundredth of a percent of the deceased Native 
population. Worse than this, we have sampled mtDNA from only a tiny fraction of the 
individual languages found in the Americas, although at least a third of the higher order 
groupings of languages have been sampled. So missing data, as always, could change the 
picture dramatically. 4) Once we have identified the mutations, and the mutation rate we 
wish to apply, there is still the problem of identifying which mutations occurred in the 
original parent population prior to migration, and which occurred in the daughter pop- 
ulation after the migration. In the most likely scenario, this means which mutations are 
found in both Asia/Siberia and in the New World. The problem with this is that we can 
never be sure we have typed all the relevant individuals and populations. Indeed, if the 
parent population existed 30,000 years ago, some shared lineages between the New 
World and Asia have gone extinct in Asia. Either way, most published estimates of the 
age of the mitochondrial lineages in the New World have not accounted for mutations 
shared on both sides of the Bering Strait, and those have overestimated the time to a 
common ancestor for New World lineages. 

With all these caveats in place, Torroni and colleagues (1992, 1993a, 1993b, 1994a, 
1994b) have suggested that the average time to a common ancestor for each of the New 
World haplogroups is ~ 25,000 yr B.P., although the authors have argued for multiple 
migrations for this scenario, and that haplogroup B appeared to be younger than A, C, 



and D. However, Bonarro and Salzano (1997) showed rhar rhe divergence times for A, B, 
C and D were not statistically different from one another. The 25,000 yr date is the time 
to a common ancestor in Asia, not necessarily the dme to entry into the New World 
(which could be considerably younger). 

Number of Founding Fineages {and Founders) 

As early as 1985, most of the founding haplogroups for the New World had been 
identified (Wallace et al. 1985), and by 1990, Schurr and colleagues had defined the core 
RFLP mutations that define haplogroups A, B, C and D. In 1992, Wallace and Torroni 
labeled these haplogroups A, B, C and D, and 95—98% of all New World natives pos- 
sessed one of those four tN'pes. At the time, the remaining 2—5% of haplot\'pes were 
thought to be due to European admixmre. In the populations that were surveyed at that 
time, this assumption was mostly true. However, in 1994, Bailiiet and colleagues identi- 
fied some other putative haplotypes that were shared with the New World, and at least 
one other potential new haplogroup. These were confirmed by Merriwether and FerreU 
(1996) and Easton and colleagues (1996). The new subt^-pes were labeled Al, A2, Bl, CI, 
C2, Dl, D2, X6 and X7 (jMerriwether & FerreU 1996). Later still. Brown and colleagues 
(1998) identified another founding lineage, confusingly also called X (hereafter Brown's 
X) that was unrelated to X6 and X7 haplot}^es described by Easton and colleagues 
(1996). With the exception of Brown's X, aU are found in Asia, and even all in one coun- 
try (Mongolia, Merriwether et al. 1996, and Tibet, Torroni et al. 1994b), but most are 
absent in Europe, Africa, and Australia. Merriwether has used this to argue for a single 
wave of migration (Merriwether et al 1994, 1995, 1996, 1999) as have others (Kolman et 
al 1995; Bonatto & Salzano 1997; Forster et al 1996). It must be noted the Hae III 16517 
is a hypervariable site, turning on and off many times in evolutionary time, which 
decreases its utility as a phylogenetic marker. Nonetheless, over short time scales, this 
h)-permutabilit)' can be a useful feamre w-hen placed upon the larger background of 
mtDNA genetic variation. It does mean that the sub-t\'pes, w^hich onlv differ bv this 
marker (I.e., Al vs. A2, X6 vs X7, etc.) may be different bv a mutation which has hap- 
pened within the New World, rather than prior to entry, despite the fact that it is found 
widely on both sides of the Bering Strait. Luckily, there are control region haplot}-pe ver- 
sions of Al and A2, and Dl and D2 which are also shared betw^een the New World and 
Asia (see Merriwether etal 2000 for the specific sequence definitions), which also strong- 
ly supports the notion of more than one variant of at least A and D coining into the 
New World in the initial wave of migration. 



Semantics also plays a role in the discussion, in the defininitions of the terms haplo- 
tYpe, haplogroup, and founding lineage. A haplotype is a unique combination of RFLPs 
or mutations in a sequence. A haplotype may be represented by a single individual, or by 
many individuals that share an identical sequence. A haplogroup is a collection of close- 
ly related haplot}qDes which share at least one (and usually several) mutations in common 
that differentiate it from other such groups. A founding lineage is defined as one that ini- 
tially entered the New World, from which the current Native American variation is now 
derived. Torroni and colleagues (1992) listed three criteria for a founding lineage which 
are useful to think about: 1) The lineage should be shared between the New World and 
the founding (parent) population, presumably in Asia or Siberia. 2) The haplotype should 
be nodal/central to the phylogeny of haplotypes in the New World. 3) The haplotype 
should be widespread in the New World. To this, I would add a fourth criterion that is 
helpful: 4) The haplotype should be found in ancient DNA that predates European con- 

Applying the above haplot\'pes, A, B, C, D, X6, X7 and Brown's X meet some or all 
of the criteria. All are found in ancient DNA that predates European contact (Stone & 
Stoneking 1998, 1999; Merriwether eta/. 1994, 1997, etc.). Variants of A, B, C and D are 
clearly nodal/central to phylogenies, and shared between the New World and 
Asia/Siberia. The X6 and X7 lineages have been demonstrated to be derived from C and 
D haplogroups, and indeed share some C and D haplotypes identically when the D-loop 
sequences are compared from RFLP-defined X6, X7, C, and D individuals. Since X6 and 
X7 are found widely in Asia and Siberia, the question remains whether they migrated into 
the New World, or whether they represent revertant mutations that have lost the defin- 
ing C and D RFLP mutations (both of which would require site gains, since C and D are 
defined by site losses at Hindi 13259 and AM 5176, respectively). Their wide dispersal 
in the New World, as demonstrated by Easton and colleagues (1996) and Merriwether 
and Ferrell (1996) would argue in favor of them entering as founding haplot^'pes. 
However, D-loop sequences from all these X6 and X7 individuals are required to evalu- 
ate the two scenarios for the origins and ages of X6 and X7. Hauswirth typed 40 of his 
8000-9000 year old Florida Windover site samples fotA/u I 10397 and Dde 1 10394 after 
learning of the X6 and X7 haplotypes from this author. Twent}' were X6, 20 were X7, 
none were A, B, C, D, or Brown's X (see Hauswirth, personal communication cited in 
Merriwether et al. 1994). While it is certainly possible that X6 and X7 arose in the 
Americas, derived from C and/or D, the fact that they are found across a wide range of 
languages and geography argues that this must either be a very early event or a very fre- 
quent event. Given the frequency is very low, it is harder to imagine that it is a frequent 
event. The broad RFLP definition of X6 and X7 is very widely distributed throughout 



y\sia (using jusf the tnulitional A, B, (], D, markers plus .1/// / 10397, Dc/e I 10394, and 
Hae III 16517). Sequencing of the New World X6 and X7 individuals is required to 
determine if they are shared with Asia. The few X6 and X7 sequences provided by 
Easton and colleagues (1996) were full of errors (a huge excess of transversions) and are 
of little value. Merriwether has been resequencing all of the samples from the l-Laston et 
al. (1996) paper, and some of the corrected sequences and haplotypes appear in 
Merriwether et al. (2000a). The rest will follow in a future paper. 

Brown's X (Brown et al. 1998; Smith et ai 1999) is found widely in North y\merica, 
but has not yet been confirmed in Central or South America. Smith and colleagues 
(1999) show that Brown's X is widespread across North America, including y\thabaskan, 
Algonquian, Iviowa-Tanoan, Wakashan, Plateau-Penutian, Northern Hokan, and Siouan. 
It appears to be present in ancient DNA (Stone & Stoneking 1999). It has not been pos- 
itively identified in Asia or Siberia, but does share a cluster of mutations with one old 
haplogroup in Europe. All the New World mtDNA lineages labeled Brown's X are at 
least 6 mutations different than those found in Asia, although they also share at least four 
mutations with those same Europeans. The implication is that the European Brown's X's 
and the New World Brown's X's have been separated for a very long time. The fact that 
it has yet to be completely identified in any populations between Europe and the New 
World has led to a number of (in my mind) highly questionable speculations about long 
range migrations across Asia, or worse yet, across the Atlantic Ocean (see Stanford and 
Bradley, this volume). The fly in the ointment may be a pair of partially intermediate hap- 
lotypes found in Korea (Lee et ai 1997, HVRbase ID # 05755; Burkhardt et al 1999) and 
Mongolia (Kolman et al 1996) that are identical to a pair of Nootka sequences from the 
New World (Ward et al 1991, samples NCN3 and NCN4) that are closer to the New 
World variants than to the European variants of X. (See Merriwether et al [2000] for a 
discussion of shared haplotypes between Asian and the New World). They are "partial" 
haplot)q3es because they were not sequenced or RFLP t^'ped for all the defining markers, 
but possess 1—2 of the combination of defining mutations. Clearly further study of 
Asian and Siberian populations is required. It is also possible that in the 1 5—30,000 years 
since the migration(s) into the New World, the Brown's X lineages have gone extinct in 

¥rom Where? 

There are a number of hypotheses regarding where the Native Americans came 
from. The most widely accepted, and least specific, is that they came from Asia. Turner 
(this volume) discusses this issue at length, concentrating on dental evidence. Thorough 



screening of the peoples of Siberia and parts of Northern and Southeast Asia have yield- 
ed some possibilities for where the founders of the New World once lived. James Neel 
and coauthors (1994) suggested the Mongolia, Manchuria, Southern Siberia area as the 
source of the founding population for the New World. Merriwether and colleagues 
(1996) and Kolman and colleagues (1995) also suggested MongoUa as a source of the 
founders. Y-chromosomal studies (Lell et al. 1997; Karafet et al. 1999) have suggested the 
Lake Baikal region of Southern Siberia, as have some mtDNA studies (Schurr et al. 1999; 
Starikovskaya 1998). The Wallace lab has suggested fTorroni et al. 1992, 1993a, 1993b, 
1994a, 1994b) that haplogroup B came over more recendy than A, C, and D (after the 
"Amerind" migration into the Americas) because it has a shallower divergence time and 
has an apparently coastal distribution (west coast of the Americas and East Coast of 
Asia). They argue for a coastal migration originating somewhere in East Asia. This 
hypothesis has some support from Johanna Nichols linguistic analyses (1995; this vol- 
ume) which also show a coastal distribution of some linguistic markers. 


The problems with identifying source populations are many. There are different 
approaches taken to identify the source population, which can lead to different conclu- 
sions. A parsimony approach would argue for the minimum number of migration events 
that are consistent with the observed data. A frequency-based approach might argue for 
each lineage coming from the place where it is currentiy found at the highest frequency. 
Both approaches also rely on geographic proximity to the New World to resolve con- 
flicts between competing potential source populations, with the ideal source population 
being located closest to the Bering Straight. These two approaches, by the way that they 
are formed, very often yield different results. 

To put this into context, Merriwether and colleagues. (1994, 1995, 1996) and 
Merriwether and Ferrell (1996) took a parsimony approach to argue that since virtually 
all of the New World haplogroups are found in present day Mongolian populations (or 
populations originally from Mongolia, such as the Tibetans), they are the current best fit 
for being descended from the same parent population as the Native Americans. The pre- 
sent day Native Americans cannot be derived from the present day Mongolians. If the 
present day Mongolians are derived from the ancient Mongolians, then Mongolia is the 
best current source for New World's founders. 

Michael Hammer's lab (Karafet et al 1999) using Y chromosome data, Starikovskaya 
and colleagues (1998) and Schurr and colleagues (1996, 1999) using mtDNA and Y-data 
have applied the frequency-based approach, and argued for separate origins for a num- 



l)cr of the ^ aiul mtl)N;\ haplogroiips based on ihcir ditfcrinu, hxxjLiencics in \arious 
Asian and Siberian i^opulations. Ji,ven though some populations have all the requisite 
haplogroups (be it Y or mtDNA), thev argue that separate waves of migration came f)ut 
of the regions where the frequency is the highest. 

Ancient DNA 

One important component in evaluating the peopling of the y\mericas from a genet- 
ic perspective is ancient DNA. Ancient DNA provides the opportunit)^ to test hj'pothe- 
ses derived from DNA from living populations. A number of studies have now been 
used to examine ancient human DNA variation in the Americas. O'Rourke and cc^l- 
leagues (1996) provide a good overview of these studies. Most of the ancient DNA stud- 
ies in the New World involve pre-Columbian populations that lived within the last 3000 
years (Stone & Stoneking 1996, 1998, 1999; Kaestle & Smith 2001; Matheny et ^/. 1992; 
Carlyle et al 2000). Stone and Stoneking (1996) showed that an 8,000 year old female 
skeleton belonged to haplogroup B, and several others in the 8,000—10,000 year old age 
also belong to A, B, C, D, X6 or X7 (Merriwether unpublished data; Smith ef a/. 2000; 
Hauswirth et al. 1994). Much has been written about the feasibilit}' of doing ancient 
DNA on the so-called Kennewick Man (Tuross & Kolman 2000; Merriwether in press), 
a 9,000 year old skeleton from Washington State that has been claimed for reburial. Much 
has been made of Kennewick Man's supposed European cranial features. Powell and 
Rose (1999) showed it had Asian/Pacific Islander craniometric affinities, as do many liv- 
ing Native Americans, and not European affinities. Three labs, mine included, tried to 
amplify DNA from a metacarpal and a rib from the Kennewick Man, but none were suc- 
cessful, or at least there were no conclusive results free of contamination from modern 
sources (as of the time of this publication). 

It seems likely that lineages may have entered the New World and gone extinct any- 
where up until European contact, and even more likely after contact. Ancient DNA may 
be the only way we will ever discover these lineages. Ancient DNA may allow us to dif- 
ferentiate between the many h^-pothetical peopling scenarios for the Americas, bv telling 
us what haplogroups were present when and where at different times in the past. 


We are stiU making new discoveries about the origins and affinities of Native 
Americans based on genetics data, and there is still much to learn. As marker densities 
get higher on the Y chromsome, tliis information is likely to play a more and more 



important role in telling the story of the peopling of the New World. For now, some 
issues are still being debated, but a clear picture is emerging. These are that mtDNA hap- 
lotypes A, B, C and D all came over together. There were more than one variant of A, 
C, and D present in the group that founded the New World. X6 and X7 probably arose 
in Asia and came over with A, B, C and D, and X6 and X7 are derived from C and D. It 
is, however, also very possible that X6 and X7 arose from C and D in the New World 
sometime early after their arrivals. 

Brown's X is only seen in North America, across many language groups. The only 
related sequences are a distant set of X-Hneages in Europe, and a possible pair of par- 
tial-X-lineages in Korea and Mongolia. So until more data is in. Brown's X could either 
be part of the same wave of migration from Mongolia into the New World, or it could 
represent a separate migration. Since Mongolian (and Mongolian derived) populations 
already possess all the haplogroups for the New World (including the partial Brown's X, 
X6 and X7), it still has to be a leading candidate for the location of the source popula- 
tion for the Americas (using my parsimony criterion). The caveat would be that if pop- 
ulation(s) closer to the point of entry had as many or more shared haplot}^es, they prob- 
ably would be considered better candidates for the source population region. So there is 
still strong support for a single wave of migration into the Americas, originating from 

A-hs tract 

In this paper, mitochondrial DNA evidence for the initial peopling of the 
NeiP World is contrasted with different theories about the peopling of the 
A.m ericas derived from other types of data. While the data have been interpreted 
in a number of ivajs, they are most consistent with a single wave of migration fol- 
lowed by differentiation. The exception to this may be the X haplogroup of Brown 
and colleagues (1998), which may or may not have been part of this initial migra- 

l^iterature Cited 

Anderson, S., A. T Bankier, B. G. Barrel, M. H. L. DeBulin, A. R. Coulson, J. Drouin, I. 
C. Eperon, D. P NierUch, B. A. Roe, F. Sanger, R H. Schreier, A. J. H. Smith, R. 
Staden, & I. G. Young. 1981. Sequence and organization of the human mitochondr- 
ial genome. Nature 290:457-465. 

Bailliet, G., F Rothhammer, F R. Carnes, C. M. Bravi, & N. O. Bianci. 1994. Founder 
mitochondrial haplotypes in Amerindian populations. Am. J. Hum. Genet 54:27-33. 



Ballinger S. W, T. G. Schurr, A. Torroni, Y. Y. Gan, j. A. I lodge, K. Hassan, K.-1 1. Chen, 
&c D. C. Wallace. 1992. Southeast Asian mitochondrial DNA analysis reveals genetic 
continuity' of ancient Mongoloid migrations. Gene/ics 130:139—152. 

Bonatto, S. 1.., & I'. M. Sal/ano. 1997. Diversity and age of the U)vn major mtDNy\ hap- 
logroups, and their implicaitons hjr the peopling of the New World. Aw. /. / ///w. 
Genet. 61:U\3-U23. 

Brown M. D., S. H. Hosseini, A. Torroni, H. J. Bandelt, J. C. Allen, T. G. Schurr, R. 
Scozzari, F. Cruciani, & D. C. Wallace. 1998. mtDNA haplogroup X: An ancient link 
between Europe/ Western Asia and North America? Am. J. Hum. Genet. 63:1852—61. 

Burkhardt, E, A. von Haeseler., &c S. Meyer. 1999. HvrBase: Compilation of mtDNA 
control region sequences from primates. Nucleic Acids Res. 27:138—142. 

Butzer , K. W. 1992. The Americas before and after 1492: An introduction to current 
geographical research. Ann. Assoc. Am. Geographers 82:345—369. 

Carlyle, S. W, R. L. Parr, M. G. Hayes, & D. H. O'Rourke. 2000. Context of maternal lin- 
eages in the Greater Southwest. Am. J. Pbjs. Anthropol. 113:85—101 

Case, J. T, & D.C. Wallace. 1981. Maternal inheritence of mitochondrial DNA polymor- 
phisms in cultured human fibroblasts. Somatic Cell Genet. 7:103—108. 

Chen, Y. S., A. Torroni, L. Excoffier, A. S. Santachiara-Benerecetti, & D. C. Wallace. 
1995. Analysis of mtDNA variation in African populations reveals the most ancient 
of all human continent-specific haplogroups. -/4/;/. /. Hum. Genet 57:133—149. 

Denevan, W M. 1992. The Pristine Myth: The landscape of the Americas in 1492. Ann. 
Assoc. Am. Geographers 82:345—369. 

Easton, R. D., D. A. Merriwether, D. E. Crews, & R. E. Ferrell. 1996. Mitochondrial 
DNA variation in the Yanomami: Evidence for two additional New World founding 
lineages. v4/;a /. Hum. Genet 59:213—225. 

Forster, P., R. Harding, A. Torroni, & H.- J. Bandelt. 1996. Origin and evolution of 
Native American mtDNA variation: A reappraisal. Am. ]. Hum. Genet 59:935—945. 

Giles, R. E., H. Blanc, H. M. Cann, & D. C WaUace. 1980. Maternal inheritance of 
human mitochondrial DNA. Proc. Natl Acad. Sci. U.S.A. 77:6715-6719. 

Ginther, C, D. Corach, G. A. Penacino, J. A. Rey, F. R. Carnese, M. H. Hutz, A. 
Anderson, J. Just, F M. Salzano, & M.-C. King. 1993. Genetic variation among the 
Mapuche Indians from the Patagonian region of Argentina: Mitochondrial DNA 
sequence variation and allele frequencies of several nuclear genes. Pages 21 1—219 /'// 
S. D. J. Pena, R. Chakraborty, J. T. Epplen, & A. J. Jeffreys, eds., DNA Fingerprinting: 
State of the Science. Birkhauser Verlag Basel, Switzerland. 

Greenberg, J. H. 1987. language in the Americas. Stanford Universit}- Press, Stanford, CA. 
438 pp. 



Greenberg, J. H., C. G. Turner II, & S. L. Zegura. 1986. The settlement of the Americas: 
A comparison of linguistic, dental, and genetic evidence. Curr. Anthropol. 27:477^97. 

Hauswirth, W. W., C. D. Dickel, D. J. Rowold, & M. A. Hauswirth. 1994. Inter- and 
intrapopulation studies of ancient humans. Experientia 50:585—91. 

Kaestle, R A., & D. G. Smith. 2001. Ancient mitochondial DNA evidence for prehistoric 
population movement: The Numic expansion. Am. J. Phys. Anthropol. 115:1—12. 

Karafet, T. M., S. L. Zegura, O. Posukh, L. Osipova, A. Bergen, J. Long, D. Goldman, W. 
Klitz, S. Harihara, P de Knijff, V. Wiebe, R. C. Griffiths, A. R. Templeton, & M. R 
Hammer. 1999. Ancestral Asian source(s) of new world Y-chromosome founder 
haplo types. v4z?/. /. Hum. Genet. 64:817—31. 

Kolman, C. J., E. Bermingham, R. Cooke, R. H. Ward, T D. Arias, & E Guionneau- 
Sinclair. 1995. Reduced mtDNA diversity in the Ngobe Amerinds of Panama. Genetics 

Kolman C. J., N. Sambuughin, & E. Bermingham. 1996. Mitochondrial DNA analysis of 
Mongolian populations and implications for the origin of New World founders. 
Genetics 142:1321-1334. 

Lee, S. D., C. H. Shin, K. B. Kim, Y. S. Lee, & J. B. Lee. 1997. Sequence variation of mito- 
chondrial DNA control region in Koreans. Forensic Sci. Int. 87:99—116. 

Lell, J. T, M. D Brown, T. G. Schurr, R. I. Sukernik, Y. B. Starikovskaya, A. Torroni, L. 
G. Moore, G. M. Troup, & D C. Wallace. 1997. Y chromosome polymorphisms in 
native American and Siberian populations: Identification of Native American Y chro- 
mosome haplotypes. Hum. Genet 100:536—43. 

Lorenz, J. G, & D. G Smith. 1997. Distribution of sequence variation in the mtDNA 
control region of Native North Americans. Hum. Biol. 69:749—76. 

Matheny, R. T., & S. R. Woodward. 1992. A ca. 9,500-10,000 year old human DNA 
sequence from Acha— 2, Nothern Chile. Proceeding of the Conference: Ancient 
DNA: 2^" International Conference, Oct 7—9, Smithsonian Institution, Washington, 

Meltzer, D J. 1993. Search for the First Americans. Smithsonian Books, Washington, DC. 
176 pp. 

Meltzer, T. D, & T. D DiUehay. 1998. The search for the earliest Americans. Archaeology 
pp. 60-61. 

Merriwether, D A. 1999. Freezer anthropology: New uses for old blood. Philos. Trans. R. 
Soc. Fond. B 354:121-129. 

Merriwether, D A. In press. Ancient DNA and Kennewick Man: A review of Tuross and 
Koman's Kennewick Man aDNA Report. Curr Res. Pleistocene 17. 



Merriwether, D. A., A. G. Clark, S. W. Ballingcr, T. G. Schurr, f I. ScxKlyall, T. Jenkins, 
S. T. Sherry, & D. C Wallace. 1991. The structure of human mirochondrial DNA 
variation./ Mo/, livol. 33:543-555. 

Merriwether, D. A., E Rothhammer, & R. E. Ferrell. 1994. Genetic variation in the New 
World: Ancient teeth, bone, and tissue as sources of DNA. Experientia 50:592-601. 

. 1995. Distribution of the Four- Founding Lineage haplot}^es in Native 

Americans suggests a single wave of migration for the New World. Am. ]. Phjs. 
AnthropoL 98:411^30. 

Merriwether, D. A., W Hall, A. Vahlne, & R. E. Ferrell. 1996. mtDNA variation indicates 
Mongolia may have been the source for the founding population for the New World. 
Am.]. Hum. Genet. 59:204-212. 

Merriwether, D. A., & R. E. Ferrell. 1996. The Four Founding Lineage h\^othesis: A crit- 
ical re-evaluation. Mo/. Pljy/ogenet. Evo/ 5:241—246. 

Merriwether, D. A., D M. Reed, & R. E. Ferrell. 1997. mtDNA variation in ancient and 
contemporary IVIayans. Pages 208—220 in S. L. Whittington & D M. Reed, eds.. Bones 
of t/oe Ancestors: ^cent Studies of Ancient Maya Ske/etons. Smithsonian Institution Press, 
Washington, DC. 

Merriwether, D. A., & E A. Kaestie. 1999 Mitochondrial recombination? (continued) 
Science 285:837. 

Merriwether, D A., B. M. Kemp, D E. Crews, & J. V. Neel. 2000. Gene flow and genet- 
ic variation in the Yanomama as revealed by mitochondrial DNA. Pages 89—124 /// C. 
Renfrew, ed., America Vast, America Present: Genes andEanguages in t/je Americas andBejond 
(Papers in tJje Prefjistorj of languages). McDonald Institute for Archaeological Research, 

Neel, J. v., R. J. Biggar, & R. I. Sukernik. 1994. Virologic and genetic studies relate 
Amerind origins to the indigenous people of the MongoUa/Manchuria/ southeastern 
Siberia region. Proc. Natl. Acad Sci. LISA. 91:10737-10741. 

Nichols, J. 1995. The spread of language around the Pacific Rim. EvoL Ant/jro. 3:206—215. 

O'Rourke, D. H., S. W Carlyle, & R. L. Parr. 1996. Ancient DNA: Metiiods, progress, and 
perspectives. v4/;/. /. Hum. Bio/. 8:557-571. 

Powell, J. E, & J. C. Rose. 1999. Report on the osteological assessment of the 
"Kennewick Man" skeleton (CENWW.97.Kennewick). Internet citation: 

Redd A. J., N. Takezaki, S. T Sherry, S. T McGarve)^, A. S. M. Sofro, & M. Stoneking 
1995. Evolutionary historv of the COII/tRNALvs tntergenic 9 base pair deletion in 
human mitochondrial DNAs from the Pacific. Mo/. Bio/. EvoL 12:604—615. 



Schurr, T G., S. W Ballinger, Y. Y. Gan, J. A. Hodge, D. A. Merriwether, D. N. Lawrence, 
W. C Knowler, K. M. Weiss, & D. C Wallace. 1990. Amerindian mitochondrial 
DNAs have rare Asian mutations at high frequencies, suggesting they derived from 
four primary maternal Uneages. Am. J. Hum. Genet. 46:613-623. 

Schurr, T G, R. I. Sukernik, Y B. Starikovskaya, & D. C. WaUace. 1999. Mitochondrial 
DNA variation in Koryaks and Itel'men: Population replacement in the Okhotsk Sea- 
Bering Sea region during the Neolithic. Am. J. Phys. Anfhropol. 108:1-39. 

Smith, D. G, R. S. Malhi, J. Eshleman, J. G. Lorenz, & E A. Kaestie. 1999. Distribution 
of mtDNA haplogroup X among Native North Americans. Am. J. Phys. Anthropol. 

Smith D. G, R. S. Malhi, J. A. Eshelman, & B. A. Schultz. 2000. A study of mtDNA of 
early Holocene North American skeletons. Abstract. Am. J. Phys. Anthropol., Suppl. 
page 284. 

Soodyall, H., L. Vigilant, A. V. Hill, M. Stoneking, & T Jenkins. 1996. mtDNA control- 
region sequence variation suggests multiple independent origins of an "Asian- spe- 
cific" 9-bp deletion in sub-Saharan Africans. Am. J. Hum. Genet. 58:595-608. 

Stone, A. C, & M. Stoneking. 1996. Genetic analyses of an 8000 year-old Native 
American skeleton. Ancient Biomolecuks 1:83—87. 

. 1998. mtDNA Analysis of a prehistoric Oneota population: Implications for 

the peopling of the New World. Am. J. Hum. Genet 62:1153-1170. 

. 1999. Analysis of ancient DNA from a prehistoric Amerindian cemetery. Philos. 

Trans. K Soc. Lond. B 354:153-159. 

Starikovskaya, Y B., R. I. Sukernik, T G Schurr, A. M. Kogelmk, & D. C. WaUace. 1998. 
mtDNA diversity in Chukchi and Siberian Eskimos: Implications for the genetic his- 
tory of ancient Beringia and the peopUng of the New World. Am. J. Hum. Genet 

Sukernik, R. I., T. G. Schurr, E. B. Starikovskaya, & D C. Wallace. 1996. Mitochondrial 
DNA variation in native Siberians with special reference to the evolutionary history 
of American Indians: I. Studies on restriction polymorphism. Ri/ss. J. Genet. 

Thornton, R. 2000. Population history of native North Americans. Pages 9-50 /// M. R. 
Haines & R. H. Steckel, eds., A Population History of North America. Cambridge 
University Press, Cambridge, England. 

Torroni, A., T. G Schurr, C. -C.Yang, E.J. E. Szathmary, R. C. WilHams, M. S. Schanfield, 
G. A. Troup, W C. I-Cnowler, D. N. Lawrence, K. M. Weiss, & D C. WaUace. 1992. 
Native American mitochondrial DNA analysis indicates that the Amerind and the 
Nadene populations were founded by two independent migrations. Genetics 



TotToni, A., '1'. C. Scluirr, M. I''. Cahcll, M. D. Brown,). \'. Nccl, M. I.arscn, D. (i. Smith, 
C. M. Vullo, &; D. C. Wallace. 1993a. Asian affinities and continental radiation of the 
four founding Native American Mitochondrial DNAs. A/^/. J. Hum. Genet. 

Torroni, A., R. 1. Sukernik, T. G. Schurr, Y. B. Starikovskaya, M. F. Cabell, M. H. 
Crawford, A. G. Comuzzie, & D. C. Wallace. 1993b. Mitochondrial DNA variati(;n of 
aboriginal Siberians reveals distinct affinides with Native Americans. Am. J. H/i///. 
Genet. 53:591-608. 

Torroni, A, Y. -.S Chen, (). Semino, A. S. Santachiara-Benerecetd, C. R. Scort, M. T. Lott, 
M. Winter, & D. C. Wallace. 1994a. mtDNA and Y-chromosome polymorphisms in 
four native American Populations from Southern Mexico. Am. J. Hum. Genet. 

Torroni, A, J. Miller, L. G. Moore, S. Zamudio, J. Zhuang, T. Droma, & D. C. Wallace 
1994b. Mitochondrial DNA analysis in Tibet: Implications for the origin of the 
Tibetan population and its adaptation to high altitude. Am. J. Phys. Anthropol. 

Tuross, N., & C. Kolman. 2000. Ancient DNA analysis of human populations. Am. J. 
Phjs. Anthropol. 111:5-23. 

Vigilant, L., R. Pennington., H. Harpending, T D. Kocher, & A. C. Wilson. 1989. Proc. 
Natl. Acad. Sci. U.S. A. 86:9350-9354. 

Vigilant, L., M. Stoneking, H. Harpending, K. Hawkes, & A. C. Wilson. 1991. African 
populations and the evolution of human mitochondrial DNA. Science 

Wallace, D C, Garrison, K., & W Knowler. 1985. Dramatic founder effects in 
Amerindian mitochondrial DNAs. Am. J. Phjs. Anthropol 68:149—155. 

Wallace, DC, & A. Torroni. 1992. American Indian prehistory as written in the mito- 
chondrial DNA: A review. Hum. Biol 64:403^16. 

Ward, R. H., B. L. Frazier, K. Dew-|ager, & S. Paabo. 1991. Extensive mitochondrial 
diversity within a single Amerindian tribe. Proc. Natl Acad. Sci. U.S^. 88:8720—8724. 





agriculture 64, 163, 290 

Ainu 105, 110, 138, 139, 146 

Alaska 9, 10, 11, 12, 16, 17, 18, 22, 27, 29, 32, 60, 73, 74, 75, 76, 132, 142, 148, 174, 

175, 176, 178, 202, 204, 237, 255, 256, 274, 276, 283, 285, 287 
Aleutian Islands 71, 132 
Alexander Archipelago 76 
Altai Mountains 133, 144 
Amazon 190, 191, 193, 200 
American Paleolithic 44 
amino acid racemization 181 
anatomically modern humans (AMH) 3, 5, 61, 62, 63, 67, 72, 80, 82, 93, 94 

adaptations 2, 203 

earliest 62, 63 

geographic expansion 63, 80 

maritime voyaging 63 

multiregional evolution 61 

Out of Africa 61, 62, 63 

stone tools 63 
Andes 183, 185, 238 
archaeological record 48, 62, 75, 78, 79 
Arctic populations 142, 144 

clothing 145 

predators 144 
Argentina 237 
Arlington Woman 77 
artifact cache 38 

Asia 68, 119, 171, 274, 278, 280, 285, 290, 291, 295, 296, 299, 300, 301, 303 
Atlantic Ocean 302 
AustraUa 3, 68, 70, 72, 79, 81, 82, 280, 292 

coastal shell middens 68, 70 

colonization of 70, 72, 82 
AustraUan Aborigines 59, 68, 81, 98, 108, 264 




barh\-mctry 78 
bears 134 
Beaufort Sea 33 
beetles 4, 12 

eurythermic species 12 

stenothermic species 12 
Beringia 4, 9, 10, 11, 14, 15, 16, 20, 30, 59, 70, 73, 75, 79, 81, 123, 133, 134, 142, 276, 


archaeoloo-ical evidence from 20, 21 

climate 11, 13 

ecosystems 17 

human entry into 19, 22, 59, 134 

land bridge 4, 10, 13, 15, 22, 80, 131, 132, 134, 147, 149, 159, 255 

megafaunal mammals 15, 17 

paleobotanical studies 14 

paleoenvironmental conditions 15, 123 

paleontological sites 21 

vegetation 15, 16, 17, 18, 19 
big game hunting 6, 20, 28, 29, 159, 163, 170, 173, 179, 187, 202, 203, 238, 247 
birds 177, 184, 188, 246, 262 
bison 17, 20, 74, 144, 162 
Bluefish Caves 20, 175 
boat(s) 5, 19, 60, 63, 68, 70, 73, 75, 76, 81, 139, 145, 147, 173, 256, 257, 263, 264, 265, 


curragh 264 

dugouts 145 

materials used in 257, 264 

skin canoes 44 

travel 75 

umiak 264 
boating 144 
bowhead whales 33 
Brazil 189, 190, 193, 237, 246, 249 
British Columbia 60, 75, 76, 279, 283, 285, 287 
Broken Mammoth site 20 
butchering 133 




Cactus Hill 43, 45, 52, 177, 259, 260, 265 
CaUco Hills 29, 173 
CaUfornia 76, 77, 283, 284 

coast of 77 
camel 162 

Canada 285, 288, 295 
canonical variate analysis (CVA) 95, 105 
caribou 17, 20 
Caverna da Pedra Pintada 193, 194, 195, 197, 198, 199, 200, 231 

artifacts 197 

dates 196, 199, 200, 201, 231, 232, 233, 234, 235 

pigment from 198 
Central America 180, 202, 274, 276, 285, 296, 302 
Channel Islands 67, 77, 289 
Channel Islands Woman 287, 289 

language of 287 
Chatelperronian complex 83 
Chile 72, 77, 187, 237, 238, 240, 242, 277 
China 129, 130 
clothing 145 
Clovis 3, 6, 27, 32, 33, 36, 37, 45, 48, 52, 72, 73, 74, 78, 79, 141, 147, 159, 162, 163, 

175, 179, 187, 188, 200, 201, 203, 237, 255, 257, 259, 261, 266, 274, 287, 288, 295 

age of 161, 164, 165, 166, 168, 169, 170, 200, 201, 205, 224, 225, 226, 227, 228, 

229, 230 

assemblages 30, 45 

blade core technology 255 

caches 37, 38, 205 

Clovis barrier 60 

demographic costs 34 

European origin for 44, 129, 139, 170, 261, 295 

flaking techniques 170, 181, 261 

foraging systems 43 

geographic expansion 31, 33 

habitat 161 

hunters 28, 237 



ll)cri;in Pciiinstila 44 

landscape learning 34, 36 

language of 287, 288, 289, 290 

mating networks 31, 36 

migration theory 162, 170, 172, 202 

movement 33, 38, 255 

origins 6, 29, 43, 44, 255, 256, 257, 266 

population growth rates 46, 50 

progenitors 43 

projectile point 28, 39, 40, 43, 52, 73, 75, 124, 129, 141, 159, 162, 170, 179, 181, 


radiation 46 

radiocarbon ages 50, 165, 168, 169, 200 

reproductive isolation 46 

risk of extinctions 34 

settiement mobilit)' 36 

sites 73, 159, 166, 168, 169 

social networks 37, 39 

South America 159, 183 

technique 30, 204 

territorial behavior 37 

tool complex 3 

tool-kits28, 29, 31, 170, 182 

variability^ in 31 
Clovis-first 72, 73, 123, 255 
coastal adaptations 67, 77, 80, 238, 263 
coastal geography 64 
coastal landforms 75 

coastal migration theory 59, 73, 74, 75, 81, 172 
coastal migrations 5, 60, 74, 79, 131, 148, 256, 285, 303 
coastal sites 5, 65, 66, 76, 78, 144, 184, 257, 263 
coastiines 65, 67, 70, 73, 79, 81, 265, 276, 281 
colonization of the New World 3, 7, 27, 72, 123, 131 

environmental context 9 

models 72, 73, 79, 82, 94, 124, 136, 146, 249 

molecular evidence 6 

timing; 159 



colonization process 36, 47 

Commander Islands 71 

continental shelf 65, 78, 81, 132, 257, 265 

cooking; 133, 145 

crania 5, 98, 99, 100, 101, 102, 105, 125, 126, 138, 170, 172 

environmental effects on 126 

robusticity 126 

sutural patterning of 126 
craniofacial 119 
craniofacial morphology 5, 116 

of recent populations 102, 104 
Cro-Magnons 138, 139 


Daisy Cave 67, 77 

dall sheep 17 

dating methods 163, 165 

radiocarbon 164, 165 

reUability 164, 165, 168, 169 

dental evidence 135, 302 

dental groups 128, 137, 145 

dental morphology 5, 109, 110, 123, 139, 141, 146, 148, 172 

crown morphology 136 

molar cusp numbers 135, 172 

root number 129, 136 

shovel-shape polymorphism 125, 127, 135, 172 

Sinodont}^ 125, 131, 135, 137, 172 

Sundadont)^ 125, 131, 135, 137, 147, 190 

supernumerary roots 129 

traits and their frequencies 131 

Uto-Aztecan premolar 128 

Zhoukoudian remains 1 26 
dental relationships 1 30 
dental trait frequencies 138 
dental wear 135 
Devil's Tower 65 



digging sticks 243 

diseases 48 

Diuktai culture 129 

dogs 5, 145, 146 

domesticadon 145, 147, 237, 238 

animals 145, 163 

plants 163, 237, 238, 245, 247 
Dvaglaska Cave 144 

Ecuador 245, 249 
epi-Clovis 124, 147 
extinctions 3, 28 
human 3 

fire 134, 145 

fish 20, 65, 77, 78, 80, 83, 161, 163, 184, 189, 202, 247, 262 

fishing 63, 142, 145, 203, 262 

fluting technology 30 

folk taxonomies 50 

Folsom 37, 38, 73, 74, 77, 188, 200, 261 

complex 41 

points 75 
food 20 

animals 20 

habits 20 
food storage 247 
foraging 6, 29, 35, 74, 161, 163, 170, 173, 179, 183, 184, 185, 188, 202, 237, 241, 247, 

248, 249 

coastal 74 
fossil beetle assemblages 11, 12, 13, 15 
founder effects 46 

founding population 114, 118, 119, 141, 142 
Franchthi Cave 68 
fruits 161, 164, 248 



gene flow 46, 116, 118, 119 

genetic data 297 

genetic diversit)^ 280, 292 

genetic drift 114, 116, 118, 119, 137 

genetics 47, 142, 148, 274 

geochronology 4 

giant short- faced bear 18 

glacial advances 1 1 

glacial periods 1 1 

glacial refugia 74 

Gorham's Cave 65 

grazing mammals 17, 18 

Great Basin 284 

Greenland 32, 33, 123, 142, 148 

Greenland ice sheet 50 

grizzly bear 1 8 

Groundhog Bay 76 


handaxes 44 
harpoons 67, 263, 264 
heat 22 

source of 22 
Homo erectits 61, 67, 80, 83 
Homo sapiens 61 

archaic 61, 67 
Homo sapiens sapiens 61, 63, 68, 71, 80 
horse 17, 162 
horticulture 247 
Howieson's Poort 62 
HrdUcka 1,93,114 

model of 94 
human fossil record 93, 95, 102, 120, 132, 174, 180, 189, 190 

North American 93, 97, 1 1 1 



South American 107, 108, 111, 185 
hunter-garhcrcrs 20, 21, 22, 27, 28, 33, 35, 47, 60, 203, 204, 240, 247, 249, 255 

base camp 247 

dietary needs 21 

food storage 247 

maritime 64, 65 

population dynamics 47 
hunting 41, 145, 162, 179, 184, 188, 237, 241, 242, 248, 262, 263, 265 

hunting skills 35 
hyenas 134, 144 

Iberian Peninsula 6, 260 

ice sheets 9, 19, 73, 295 

ice wedges 10 

ice-free corridor 19, 27, 43, 49, 59, 60, 72, 73, 74, 75, 79, 80, 159, 172, 255, 256, 295 

environmental conditions of 74, 256 

tools found in 256 
insects 4 

interglacial periods 1 1 
Ireland 263 


Japan 70, 71, 81, 83, 105, 119, 129, 171, 257 

maritime peoples 71 

Miyako Island 71 

Okinawa 71 
Jomon 105, 110, 124, 138 


Kamchatka Peninsula 71 

Katanda 67 

Kennewick skeleton 46, 94, 102, 105, 135, 287, 289 



affinities 46, 105 

antiquity of 102 

dentition 135 

DNA of 304 

language of 287, 289, 290 

lawsuit 102 
keystone species 33 
Klasies River Mouth 62, 65 
Kodiak Arcliipelago 75 
Kodiak Island 75 
Korea 305 


Lake Baikal 130, 133, 137, 295, 303 

Lake Mungo 68 

land bridge 22, 132, 255, 256 

landscape learning 4, 32, 34, 36, 43, 47 

language famiUes 146, 273, 274, 275, 276, 277, 281, 284, 285, 287 

age of 275, 279 

Algic 285, 287 

Algonquian 285, 287, 289 

Athabaskan 283, 289 

Athabaskan-Eyak-Tlingit 285, 287 

Caddoan 284 

Chumashan 288 

Eskimo-Aleut 274 

Miwokan 279 

Na-Dene 284, 289, 292 

Northwest Caucasian 275 

Oregon 283 

Penutian 279, 280, 283, 287 

Pomoan 287 

Salishan 285 

Siouan 284 

Siouan-Catawba 284 

Tsimshianic 279 



Uto-A/tecan 284, 288 

Yukian 275 
language isolate 275, 277 
language population growth 276 
language spread rates 277 
language stock 275, 276, 277, 278, 281, 289 

daughter languages 275, 276, 277 

proliferation rates 277, 279, 291 
language universals 281 
languages 3, 39, 138, 146, 273, 285, 290 

age of 276, 291 

American 6, 146, 148, 273, 277, 281, 286, 290, 291 

ancestral 147, 291 

Asian 282 

Australasian 282 

classification 124 

coastal spread of 283, 288, 291 

diversity' of 6 

Melanesian 282 

Mesoamerican 276 

Pacific Rim 277, 278, 279, 280, 281, 282, 289 

residual zones 281 

Siberian 146 

South American 276 

structural elements of 274, 278, 280, 281, 282, 285, 291 
Lapita ceramics 39 
Lapita culmral complex 70 
Las Vegas 78, 245, 247 
Last Glacial Maximum (LGM) 11, 13, 14, 22, 262, 263, 265, 266, 278 

mean annual temperatures 14 
Late Glacial 9, 15, 16, 20, 64, 190 
Lime Hills Caves 20 
linguistic differentiation 295, 303 
linguistic diversit)' 273, 275, 276, 281 

in the Americas 275, 281 
linguistic history 273 

of the Western Hemisphere 273, 283 



linguistic interrelationsliips 286 

linguistic variation 141 

linguistics 47, 146, 147, 148 

Hon 18, 134 

loess 10 

Lost Tribes of Israel 4 


mammoths 17, 18, 29, 35, 144, 162, 177 

Maori 105 

marine erosion 77, 78 

marine mammals 19, 65, 67, 77, 144, 147, 262 

maritime adaptation 45, 59, 60, 63, 64, 67, 74, 75, 77, 79, 82, 147, 173, 266 

archaeological record of 64 

South America 77 
maritime adaptations 78 
maritime migration 73, 266 
maritime technology 263 
mastodons 18, 28, 35, 182, 242 

Meadowcroft Rockshelter 43, 45, 51, 132, 175, 259, 260, 265 
meat eating 242, 247, 248 
medicinal plants 34, 243, 244 

megafauna 15, 17, 18, 20, 23, 27, 73, 93, 123, 133, 177, 180, 182, 184, 185, 186, 189 
megafaunal extinctions 18, 19, 20, 23, 28, 42 
Melanesia 67, 68, 70, 79, 81, 82, 83, 278, 281 
Mesoamerica 179, 285, 286, 290 
Mexico 180, 274, 276, 284, 285 
microlithic tradition 124 
Middle Stone Age 62 
migrations 19, 30, 82, 123, 132, 135, 138, 159, 161, 297, 299, 300, 305 

by sea 80, 82 

from Europe 161, 295 

multiple 7, 304 

numbers of 124, 135, 138, 141, 148, 285, 295 

single 295, 305 

theories 161 



timing of 162 
Milankovitch cycles 1 1 
Minnesota 100, 101 
mitochondrial DNy\ 61, 62, 123, 141, 172, 260, 295, 299, 305 

ancient 302, 304 

deletions 297 

divergence times 300 

genedc variation 300 

genome 296 

haplogroup X 260, 265, 295, 300, 302, 305 

haplogroups 260, 296, 298, 300, 301, 304 

haplotypes 296, 300, 301, 305 

lineage markers 297 

maternal inheritance of 296 
mutation rate 299 
mutations of 299 

Native American 297 

New World Uneages 302, 303 
molecular clock 297, 298 
mollusks 177, 246 

MongoUa 130, 133, 295, 300, 303, 305 
Mongoloid dental complex 135 
Monte Alegre 162, 191, 192, 200, 201, 231 
Monte Verde 6, 19, 30, 48, 49, 60, 72, 73, 77, 132, 186, 187, 238, 240, 243, 245, 249, 

256, 277, 287, 288, 289 

ageof 77, 186, 187, 241 

artifacts 186, 241 

habitat 187 

language of 287, 288, 289, 290 

plant use at 243 

storage facilities 248 

stratigraphy 186 

subsistence at 247, 249 
moose 17 
mountain goats 17 
Mousterian stone artifacts 133, 134 
musk-ox 17 



Mutual Climatic Range (MCR) method 1 1 
estimates 13 


Nanchoc Culture 246 

Native American 111, 119, 120, 297, 299 

ancestral 125 

gene pool of 1 1 1 

wi thin-group genetic variation 117 
natural selection 112, 114, 116 
Neanderthals 61, 62, 83, 93, 94 
needles 5, 145, 147 
Nenana Complex 124, 177, 178, 179, 188, 202, 256 

adaptations 177 

age of 177 

tools 179 
Nenana River 20, 72, 73, 74, 177 
Nevada 99, 100, 101 
New Guinea 70, 79, 81, 280, 282 
New Jersey 44 
New Mexico 3 


Oceania 39, 50 
Ohio 44 

Okinawa 71, 83, 137 
shell middens 71 
oral traditions 273 
Oregon 76 
ornaments 63 
overshot flaking 261 

Paleo-Eskimo 32, 33 



Palcoindian 9, 20, 37, 38, 73, 77, 97, 98, 105, 118, 159, 160, 161, 162, 168, 174, 182, 


andquity of 101, 164, 165, 174, 175, 196 

artifacts 37 

caches 38, 39, 51 

coastal sites 184 

cranial features 98, 101, 102, 103, 106, 117, 172 

economies 75, 182 

lifeways 9 

Uthic complexes 179, 180, 184, 185 

maritime 77 

migration 9, 164 

mobilit)' 37 

movements 38 

origins 170 

points 40, 41,42, 180, 181, 183 

radiocarbon 165 

remains 94, 96, 97, 109, 180, 201 

sites 159, 160, 164, 180, 183, 189, 204 

skeletal biology 94, 95, 96, 97, 98, 99, 109, 170, 172 

southern cultures 162, 188 

tools 170, 175, 183, 196 
Paleoindian Period 9, 51 
paleotemperamre estimates 11, 12, 14 
pathogens 48 
permafrost 22 
permafrost features 10 

Peru 78, 183, 184, 186, 237, 238, 241, 246, 249 
Plain view 37 
plant collecting 248 
plant foods 6, 163, 177, 189, 237, 238, 242, 246, 248, 249 

aquatic plants 241 

below-ground plants 240, 242, 243, 248 

nutritional value of 243 

procurement technique 244 

seasonal foods 242 

staple plants 242 



plant remains 168, 238, 249 

plant-fiber-technology 263 

pollen records 16 

Polynesians 59, 100, 105, 108, 282 

population density 46 

population growth rates 36, 46 

population size 4, 142, 149 

pottery 181 

pre-Clovis 3, 4, 29, 30, 31, 43, 48, 49, 52, 72, 78, 123, 131, 134, 139, 159, 161, 164, 174, 

177, 179, 181, 183, 185, 186, 187, 202, 204, 205, 260, 295 

Beringian crossing 134 

dates 29, 164, 174, 175, 177, 183, 187, 189, 259 

lithic assemblages 43, 45, 174, 175 

North America 30 

projectile-point cultures 174, 175, 179 

sites 3, 164, 189,204,205,259 

South America 30, 48, 159 

theories 162 
predators 18, 188 

pre-projectile point theory 173, 174, 179, 181, 183, 185, 188 
prey populations 33 

principal components analysis (PCA) 95, 97, 98, 99, 100, 105, 107, 110 
projectile point 28, 118 


Qafzeh 62 

Queen Charlotte Islands 76 

Querero 77 


radiocarbon 31, 47, 77, 142, 163, 167, 174, 175, 176, 181, 183, 185, 189, 199, 200, 259 

calibration 31, 49 

curve 31 

standards of dating 165 
rainforest 163, 173, 190,202 



adaptations 173 
Ramcipithecus 62 
red ochre 44, 45 

rock painting 188, 192, 193, 194, 196 
rodents 133,202 
routes 72 

saber-toothed cat 18 

Sahulland 70, 139, 257 

saiga antelope 17 

Santa Rosa Island 77 

scavenging 242, 248 

sea level 9, 10, 64, 65, 70, 73, 75, 77, 78, 173, 278 

Pleistocene 64 
seafaring 59, 60, 63, 68, 70, 71, 77, 79, 82 

antiquit)^ 68, 77 

archaic Homo sapiens 68 

Homo erectus 68 
seamanship 265 
Semliki River 62 
Seward Peninsula 1 5 
shamans 130, 243 

drum 130 
shell middens 77, 78 

sheUfish 65, 67, 77, 78, 161, 163, 184, 202 
Siberia 5, 9, 10, 29, 123, 128, 129, 130, 132, 137, 140, 159, 274, 276, 280, 291, 295, 299, 

301, 303 

archaeological sites 129, 133 

hunter-gatherers 9, 144 

migrants coming from 7 

Neolithic 32 

paleoenvironmental reconstructions 133, 142 

prehistoric population 142 

prehistory 134 
skeletal remains 94, 109, 118, 120 



African 108 

Australian 99, 101 

craniofacial shape 46, 95, 98, 110, 116, 126 

Early Archaic 97 

food preparation techniques 126 

Late Prehistoric 112, 117 

mid-Holocene remains 110, 111, 118 

Pacific Rim populations 99 

Paleoindian samples 96, 97, 101, 108, 111, 112 

Polynesian 100, 101, 109 
South America 107, 111, 112, 118 
southern Asians 101 
Skhul 62 

smaU game 6, 20, 161, 163, 177, 187, 189, 202, 203, 242, 246, 248 
social networks 47, 48 
Solomon Islands 70 
Solutrean 45, 131, 139, 170, 259, 260, 261, 262, 265 

relationship to Clovis 259, 260, 265 
Solutrean Period 43, 45 
Solutrean points 45, 258, 260 

fluting 45 
South America 5, 6, 27, 161, 179, 188, 202, 204, 238, 239, 240, 274, 285, 296, 302 

sites 160, 239 
Southern Cone 185, 186, 187, 188 
Spirit Cave 94, 99, 125, 139 

dental morphology 1 25 
stone cache 39 
stone tools 6, 40, 123, 132, 170, 174, 179, 184, 238, 246, 258, 260 

assemblages 129 

bifacial industries 173, 179, 241, 258, 260 

blade tools 128 

ground stone implements 246 

stylistic forms 41 

tool types 124, 127, 128, 139, 261 

unifacial industries 179, 238, 246 
subsistence patterns 237, 249, 250 
Sundaland 139 



'niimaTaima 182,205 

Taiwan 71 

taphonomy 163 

teeth 5, 123, 125, 127, 136, 140, 148, 172 

dental caries 126, 140 
tooth-use behaviors 140, 148 
Thule 32, 33 

hunters 33, 50 
Tibet 300 
Tongass Caves 76 
Trail Creek Caves 20 
trans- Atlantic crossins: 52, 131 


Upper PaleoHthic 43, 45, 63, 83, 105, 258 
Asian 258 


Vikings 81, 123 
Virginia 259 


Washington 76, 287 

waterproof clothing 263 

wetland environments 6, 237, 238, 244, 249 

Wisconsin glacial 20, 21 

Wizard's Beach skeleton 94, 99, 125, 139 

antiquit)' of 100 

dental morphology 125 
wolves 32 
woolly mammoth 17, 20, 27, 28, 42 

extinction 18 



Wrangel Island 18, 19 
wooUy rhinoceros 17 


Y chromosome 123, 303, 304 

Younger Dryas 1 6, 23, 50 

Yukon Territory 9, 11, 12, 17, 18, 72 

Zhoukoudian 126, 127 


The First Americans 

The Pleistocene Colonization 
OF THE New World 


Changjng Perspectives of the First Americans: Insights Gained 
AND Paradigms Lost, by /V/ata G. Jablonski 


3 1853 00083 1862 

Setting the Stage: Environmental Conditions in Beringia as People 
enttred the new world, by scott a. elias 

v/hat Do You Do When No One's Been There Before? Thoughts on 

the Exploration and Colonization of New Lands, by David J. Meltzer 


Anatomically Modern Humans, Maritime Voyaging, and the 

Pleistocene colonization of the Americas, by Jon M. Erlandson 


Facing the Past: A view of the North American Human Fossil Record, 
by d. Gentry Steele and Joseph F. Powell 


Teeth, Needles, Dogs, and Siberia: Bioarchaeological Evidence for the 
Colonization of the New World, by Christy g. Turner U 


The Migrations and Adaptations of the First Americans: Clovis and 
Pre-Clovis Viewed from South America, by a. C. Roosevelt, 
John Douglas and Linda Brown 


Plant Food and its Implications for the Peopling of the New World: 
A view from South America, by Tom D. Dillehay and Jack Rossen 


Ocean Trails and Prairie Paths? Thoughts about Clovis Origins, 
BY Dennis Stanford and Bruce Bradley 


The First American Languages, by Johanna Nichols 

A Mitochondrial Perspective on the Peopling of the New World, 
BY d. Andrew Merriwether 


Nina G. jablonski is the Irvine Chair and Curator of anthropology at the 
California Academy of Sciences. 

distriboted by the 
umiversity of california press 
Printed in the 
Unsted States of America