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

See other formats

NINA G. JABLONSKI, SERIES EDITOR
HE EDITED PROCEEDINGS OF A PAUL L. AND PHYLLIS WATTIS FOUNDATION ENDOWMENT SYMPOSIUM
MEMOIRS OF THE CALIFORNIA ACADEMY OF SCIENCES NUMBER 27

Serials
ae:
11 |
.C26
no.27

copy 2

Wattis Symposium Series in Anthropology

MEMOIRS OF THE CALIFORNIA ACADEMY
OF SCIENCES, NUMBER 27

| ‘CALIFORNIA
ACADEMY OF SCIENCES

FEB 2 4 2003
LIBRARY va

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

SCIENTIFIC PUBLICATION COMMITTEE:

Alan Leviton, Edztor
Katie Martin, Managing Editor

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 Allen Press

Distributed by the University of California Press

The First Americans

THE PLEISTOCENE COLONIZATION OF THE NEW WORLD

Contributors

Bruce Bradley

Primitive Technology Enterprise, Inc.
IPO) lsioxx 534

Cortez, CO 81321

(970) 565-7618
ptimtech@yahoo.com

Linda Brown

Department of Anthropology
University of Montana
Missoula, MT 59812

(406) 243-2693

(406) 243-4918 FAX

Tom D. Dillehay

Department of Anthropology
University of Kentucky
Lexington, KY 40506

(773) 702-7729

(773) 702-4503 FAX
dilleha@pop.uky.edu

John Douglas

Department of Anthropology
University of Montana
Missoula, MT 59812

(406) 243-2693

(406) 243-4918 FAX

an_jed@selway.umt.edu

Scott A. Ehas

Department of Geography
University of London
Royal Holloway

Egham, Surrey TW20 0EX
(44) 1784-443-647

(44) 1784-472-836 FAX
s.elias@rhul.ac.uk

Jon M. Erlandson

Department of Anthropology
University of Oregon
Eugene, OR 97403

(541) 346-5098

(541) 346-0668 FAX
jerland@oregon.uoregon.edu

Nina G. Jablonski

Department of Anthropology
California Academy of Sciences
Golden Gate Park

San Francisco, CA 94118

(415) 750-7161

(415) 750-7346 FAX
njablonski@calacademy.org

Daud J. Meltzer

Department of Anthropology
Southern Methodist University
Dallas, TX 75275

(214) 768-2826

(214) 768-2906 FAX
dmeltzer@post.cis.smu.edu

Vil

D. Andrew 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

andym@umich.edu

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

A. C. Roosevelt

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

Vill

Jack Rossen

Department of Anthropology
Ithaca College

Ithaca, NY 14850
jrossen@ithaca.edu

Dennis Stanford

Department of Anthropology & Archaeology
National Museum of Natural History

MRC 112, Smithsonian Institution
Washington, D.C. 20560

(202) 357-2672

(202) 357-2208 FAX

stanford.dennis@nmnh.si.edu

D. Gentry Steele

Department of Anthropology
Texas A&M University
College Station, TX 77843
(979) 845-5297

(979) 845-4070 FAX
degsteele@tamu.edu

Christy G. Turner II

Department of Anthropology
Arizona State University, Box 87242
Tempe, AZ 85287

(480) 965-6213

(480) 965-7671 FAX
christy.turner@asu.edu

io , ae 8 oe T. ee fal“
a . > oe 2 — .. ; | i ‘Wye Ap aditth ieee —
| | ma eta ee en

<p sy He = i | Pekises “« Ye genet

iim * ae ‘ ae ta
oh RY aR” a @ ae on Mie ack. =
| carer a sini b 7g jug we Po cae
oa ; perpen Tete or hi ewusieg th —)
ae nt (fe sles dynes sitiewe Ail ee, aoe
ae eee 1A A erry, nig aay ae \
LOen i ik ae
att
7 pyc Cees, Tee onan
d oS — mute | ly ey,
a LARS SAT mae nsalbing 9

ra es Plas Aa aay

CONTENTS

Preface

Introduction

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

Setting the Stage: Environmental Conditions in Beringia as People
Entered the New World, by Scott_A. Exas

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

the Exploration and Colonization of New Lands, by David J. Me/tzer

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 II

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 Nachols

A Mitochondrial Perspective on the Peopling of the New World,
by D. Andrew Merriwether

Index

DY)

123

59)

BND

Sila

: 7 - a a: a
) _ a a 5
a Vien ;
a : E
“Brushed! an
, ® ~ Nee
a ; 4 ’
: = a .* ae
> eee — i ee
x ® i) “
} \ i —_ _
‘yuardee 2 aye | ee aia */@en') an! mA ices
«2 - xt b oats Ho
: . » Ven ebenetndy. ei wvesitanig, Y Rots sare sm
at sl sai 4h pe he. "
71 (Hee eres mcg 7: di jeendl fant) at ‘fait ” wy i ¢
oa ee eK 19a ave ? ey ie celta poplin” (hat a Pe ?
A — ht dwar AFA r¢ ceil we oe nh
7 . a, 4 Mh, es UL ee ge it’ Yer arisen ve
: >. ea. | Sci iH nat ail upess irek, Hen Sh he iio i Hy
7 — | | treet (F ee, Whe ae ;
ae f ee a
" 2 eva
ia ; Git ON: oar bus 1 & sepicuish wee ik weet /- eae AT
7 ’ p Sail al ai a yet Bie ne ot vena
> i bve et, ees ee bee
7 a ill 9 Cashes ts ness a8 area t
| . a8 4 (ama! bie
STE eS OS hie ak” lit PaavN, Sits HAT Gig ae ay bia oe
‘ o4 : | eh SS ore aa +) Any at 2 etch: & Pathe On
13 ae aaaaten) inn is siete aids | aidan Nail Hebi)
an: | eee ne os
28 ee ie itil Gt cui. wed
al

pb (Faw, a em yet aol nal vii alec

eters tl OY Re <.

PREFACE

San Francisco ts privileged to have as one of her citizens Mrs. Phyllis
Wattis, a woman of tremendous warmth, sincerity and generosity.
The support that she has given to the city’s educational and cultural

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 full 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
much.

Finally, I wish to express my gratitude to three people who wete 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 organizing 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

rt ‘ asl VF,

es:

nt : a = a ie) . : :
7 _ at _ : e i.
a > @

= = * i : ;
Agile) o> any gta age anit ee Pea
_ 5 | Oe < *teste we > payee Mate te sony
ee an Me ia ven
— 7 = : a e ete See Pinstripe bale | : *
wom 4 Sy : : J i _ q i Vit ee ie

_— 7 aii a : te, © iy le
L) ‘-* 7 P
7 : P ~
- =
i ;
_ a
asx !
' " ; KS
on i - '
“ —
5
, 7 ae, R
‘ i =
i
an
: ~ - fd '
8 oa ~ r
~~ yt a - :
. io rr
te : Te
a ; 2
{
: .
1
t
5
?
1c — — J
rs Le =
-_
i
;
=
is \
re
on i ae Pn
E ; e
i
_— - 7
a i
>» ’
—
1
:
: ; a 5
7

CHAPTER ONE

INTRODUCTION

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. At
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 we accept the theory that man orig-
inated in the Old World, it is evident that his colonization of America ts a question of mode
of migration, 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, 1s largely one of speculation, fettered on one side by ignorance and on the other
by ancient traditions.” (Ardlicka 1912:5)

Twenty-first century anthropologists like to think of themselves as integrative
thinkers who are pioneering new, interdisciplinary and multidisciplinary 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 notion.
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

JABLONSKI

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 out 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: Hrdlicka's fetters of ignorance and “ancient tradition” have
evolved in ninety years, but they have not disappeared. Fortunately, the reader will 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 Hrdlicka's lament rendered unfounded,
the peopling of the Americas is now understood in greater detail than ever before.

In the context of the five-million-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

CHANGING PERSPECTIVES OF THE FIRST AMERICANS

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 theit 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 subtle 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 intensity 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 settle-
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 “oold
standard” against which all 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

JABLONSKI

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 7” sé#u 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 Elias 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. Elias uses this high level of habitat specificity 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

CHANGING PERSPECTIVES OF THE FIRST AMERICANS

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
Powell undertake multivariate analyses of the measurements of fossil and modern cra-
nia. Their sample includes most of the fossil crania identified as latest Pleistocene or
eatly 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 all, 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
directly 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

JABLONSKI

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
game.

A similar theme is developed by Tom Dillehay 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
lie 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 arrtval(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

CHANGING PERSPECTIVES OF THE FIRST AMERICANS

Merriwether's idea that the Americas were colonized in a single wave, followed by a rapid
differentiation of peoples é 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 hypothesis 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 presumptuous 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.

Literature Cited

Bence, C. Ik Bo IR, Nelsom, IN, Segucil, Isl Os, IL, Sesting, Q, Ae Warm, Wo NG IL, ce IL
Dashtseveg, 2001. Old World sources of the first New World human inhabitants: A
comparative craniofacial view. Proc. Natl. Acad. Sci. 98:10017—10022.

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

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

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

Hrdlika, A. 1912. Historical notes. The problems of the unity or plurality 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.

; : fa > a lee
. i : > ne ni 7
eo 7 . _

, = "a ta : %

_ ie uy ‘on

. cane = a st primp: ee cansigneel

;

i SA eT Ae ie GG Cos oS pat l;.aih sabi %
iereo mG senate Wilber « of WG Mewpecs cule
ee ee ee ee ee ee ee a subelnbiiaial
: agg 7 ee a iceeeaeee
z een ee ies alee dinp ot pe 1 cone alla sneered
, acitetalans aig lich A sitempae TOG) ©) canmchi aie Maite opegmmmvenl Dey ei?
end ar imntenia Geena th Qee Mie PaaS “ate ee eb
an - ® (pours fe i Hd gel “yy ner es, Ai. tal ingot gs tt “nin 98
PERNA AS EAM N\ AT AL ain ae ETE BI
iri garveliy na 4 AE ehh =e a antes inedible wae
Us ones snes A Fue emp) innominate i won asa
(nese ea i og annie. ug iho ey nansca tcabonn iat
ee cd eee eile le" ey <aieip iy. dih pizcal
Sy Fry rene eo a IRN a1 Ke wa, Selendis
ee ae i sith ovat Aisne enepinl aed yh
| enpaents ay - A ihn Sina srhonagde role
“i tm | en pat eh
Rage.” Se, a
—_ P= ig) i AA y
| eu - ‘a ae j 7 a .
pad ee _ _ ona a _ i ioe Ah

: i Ny

i ne E ee | ar. Phyto és
cto, OY | at eae yest he i
| te as i) oan ies a 14 ab ey Ot iN Mae (nes Sh
mu “Qwik @ Ee ea me (PEl. 6tatt= a ant

. a ow reaiia aby qpemiaats a ee vie

_ . nee . tat TR a : ae

idem sett hereogigy i ND 7 ost * wewi ‘pee © p= wlio :
| a ee eee
a, ee Ce i

. >. Ce ty Pa as
. ) lo qengaid Galea. aie al (Mao y bengpes-S <a
ee | ON neh GOS yi ramne 60019

eas a CPF a oe a | yee ee EH Be

CHAPTER TWO

SETTING THE STAGE:
ENVIRONMENTAL CONDITIONS IN BERINGIA AS

PEOPLE ENTERED THE NEw WORLD

Scott A. Elias

Archaeology is far more than the study of ancient artifacts.
Archaeological research aims at unraveling the history of ancient

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 eloquently put it, “To recreate a landscape, green with life or windswept and
barren, and then repopulate it with animals and men 1s a formidable task.”

Late 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

ELIAS

150 160° 170° 180 170° 160° 150° 140°
L

Glaciated regions

0 400
km

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. Paleoclimate
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 (Péwé 1975). These features show that Alaska
experienced temperatures 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

ENVIRONMENTAL CONDITIONS IN BERINGIA

and sand dunes attest to the relative aridity of climate. Pleistocene environmental condi-
tions in Beringia were unlike 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, Bartlein 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 CO, con-
centrations in the atmosphere during glacial intervals and increased CO 5 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 paleoclimatologists. Some of
these factors had unique impacts on Beringia. For instance, the split 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 temperatures, and salinity of ocean waters off
the southern coast of Beringia may have brought warmer autumns and winters than exist
today. Increased CO, concentrations during interglacial periods served to amplify the
warming brought on by increased insolation. Conversely, decreased CO» 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.P., 110,000—100,000 yr
B.P., 87,000—75,000 yr B.P., 60,000—30,000 yr B.P., and the Holocene. Low points in the
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.P., 75,000—60,000 yr B.P., and 25,000—18,000 yr B.P.

Estimating paleotemperatures

Paleotemperature estimates for the Last Glacial Maximum (LGM) have been made
from fossil beetle 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 this study focused on the mean

ELIAS

temperature of the warmest month of the year (TMAX) and the coldest month of the
yeat (IMIN) 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. Fortunately, 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-
tion.

Not many modern beetle collecting localities are in close proximity 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 Bartlein
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 beetle 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 beetles are thought to respond more readily to climatic change, by
shifting their ranges away from regions of unsuitable climate.

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 linear regression

ENVIRONMENTAL CONDITIONS IN BERINGIA

equations to fit predicted to observed TMAX and T’MIN values for modern localities
with meteorological records (see Elias e¢ a/ 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 TMAX and TMIN that were based on the
modern beetle assemblage MCR reconstructions did not match the modern TMAX and
TMIN data exactly. 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 beetle-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 e¢ a/ 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 e¢ a/ 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 beetle assemblages from this interval (20,000—18,000 yr B.P.) yield-
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—-13°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 continentality. The insect evidence thus points to relatively
mild 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

ELIAS

ees

£x

Arctic assemblages
—— Subarctic assemblages
10 15 20
“C yr B.P. X 1000
EC
28
Sk
GE
o
re}
Qa
=0

10 15 20
“C yr B.P. X 1000

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 ef a/, 1999).

sheets. Other lines of evidence, such as periglacial features that developed during the
LGM, indicate that mean annual temperatures dropped significantly 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 temperature and moisture gradients
with colder and drier conditions dominant over eastern Beringia (Anderson & Brubaker
1994; Lozhkin et a/ 1993; Anderson et a/ 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 tun-
dra vegetation based on fossil pollen spectra is inherently difficult, because of low taxo-

ENVIRONMENTAL CONDITIONS IN BERINGIA

nomic resolution, poor dispersal of minor pollen types, and the wide ecological toler-
ances of genera or species that dominated the pollen rain (Anderson ef a/ 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 supplied with loess. They interpreted the region-
al soils 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 mostly 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.

Late 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 (Elias 2000b), TMAX values
began rising at least by 12,000 yr B.P., reaching warmer-than-modern levels by 11,000 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, which brought rela-
tively warm, Pacific waters to the northwest coast of Alaska.

ELIAS

The changing climate also brought about major changes in vegetation across
Beringia. Pollen evidence (Brubaker ef a/, 2000; Edwards ef a/. 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 tundra 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
11,000 yr B.P. Populus (balsam poplar and or aspen) expanded in these mountains from
11,000-9000 yr B.P. Populus expanded north of the Brooks Range in Alaska during this
same interval (Anderson & Brubaker 1994).

Pollen studies point to a climatic oscillation during the Younger Dryas chronozone
(10,800-10,000 yr B.P.) in Eastern Beringia (Brubaker ef a/ 1999). Elias (2000b) found
strong evidence of a decline in TMAX values during this interval, especially in arctic bee-
tle 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 cooling 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 aridity. The combination of extremely cold winters and lit-
tle effecttve moisture may have combined to retard the expansion of coniferous forests
in Eastern Beringia. Por 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 ef a/ 2000).

ENVIRONMENTAL CONDITIONS IN BERINGIA

Late Pleistocene Megafauna of Beringia

Late 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 diversity of large mammals that has never
since been equaled in any high latitude region. The only modern ecosystem that supports
a similar variety of grassland animals is the East African Savannah. There are some inter-
esting parallels between 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, filled 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-tundra 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 still 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 variety 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-tundra
habitat. Zimov and colleagues (1995) used a computer model to simulate the interactions
of megafaunal mammals and vegetation in Beringia. They hypothesized that trampling

laJLIUAS)

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 slicing 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 woolly
mammoth population of Wrangel Island, a small Arctic island off the northeast coast of

ENVIRONMENTAL CONDITIONS IN BERINGIA

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 1s 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.

Early Human Lifeways in Eastern Beringia

The timing of human 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 (Josenhans ef a/ 1995; Dixon 2000; Mandryk ef a/, 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 first 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 (eg.
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 shellfish. 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
(Josenhans ef a/, 1995; Mandryk ef a/ 2000) indicates that parts of the British Columbia

ELIAS

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 e¢ a/ 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 validity 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
eatly 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 variety of animals for food, including small game,
waterfowl, and fish (Yesner ef a/ 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 little we know about these early
inhabitants and their lifeways.

I would like 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

ENVIRONMENTAL CONDITIONS IN BERINGIA

Wrangel Island

Bluefish
Caves
a Caves

Achae 7a
sites

fine Hills ¢ om)

CE ge)

ee
i:

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 lived 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 (Pitu’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 centuries later, the inun-

ELIAS

dation of the Bering Land Bridge would have brought another precious resource to the
coasts of Beringia: driftwood 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 fatty 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,

Abstract

Like all archaeological issues, the peopling of the New World must be 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 Land Bridge took place
just as regional climates were warming at the end of the last glaciation, about
12,000 years ago. The cold, arid climate that characterized the height of the last
glaciation (ca. 28,000—14,000 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 10,000 yr B.P., bands of hunter-gatherers became estab-
lished throughout most of Alaska north of the Alaska Range. 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 few decades, the
Bering Land Bridge was inundated. This, in turn, brought warm, Pacific waters
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. In Arctic Alaska (by
11,000 yr B.P.) average summer temperatures rose to levels as much as 7°C
warmer than they are today This dramatic warming was followed by an abrupt

ENVIRONMENTAL CONDITIONS IN BERINGIA

reversal, synchronous with the Younger Dryas oscillation in the North Atlantic
regions. Many of the Pleistocene megafaunal mammals were extinct by 11,000 yr
B.P. Their demise probably forced people to adopt new hunting strategies and dif-
ferent game animals. It remains unclear whether human hunting contributed sig-
nificantly to the extinction of megafaunal mammals in the New World.

Acknowledgments

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 colleague, John Hoffecker,
for a helpful review of the initial draft of the paper.

Literature Cited

Ackerman, R. E. 1996. Bluefish Caves. Pages 511-513 mm F. H. West, ed., American
Beginnings, The Prehistory and Paleoecology of Beringia. University of Chicago Press,
Chicago, IL.

Alfimoy, A. V., & D. I. Berman. 2001. Beringian climate during the Late Pleistocene and
Holocene. Ouat. Sc. Rev. 20:127-134.

Anderson, P. M., & L. B. Brubaker. 1994. Vegetation history of north central Alaska: A
map summary of late-Quaternary pollen data. Quat. Sc. Rev. 13:71—92.

Anderson, P. M., P. J. Bartlein, & L. B. Brubaker. 1994. Late Quaternary history of tun-
dra vegetation in northwestern Alaska. Ouat. 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. Te Holocene 7:331—335.

Bartlein, P. J., P. M. Anderson, M. E. Edwards, & P. EF 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 paleoenvironmental record from
Windmill Lake, Central Alaska: Late glacial and Holocene vegetation in the Alaska
range. Ouat. Sa. Rev. 20:203—215.

i)
1S)

ELIAS

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

Catto, N. R. 1996. Richardson Mountains, Yukon-Northwest Territories: the northern
portal of the postulated “ice-free corridor.” Quart. 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 i A. L.
Bryan, ed., New 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.
Ouat. Sci. Rev, 20:277—299.

Edwards, M. E., C. J. Mock, B. P. Finney, V. A. Barber, & P. J. Bartlein. 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. Ouat. Sc. Rev. 20:189—202.

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

. 2000. Late Pleistocene climates of Beringia, based on fossil beetle analysis.
Quat. Res. 53: 229-235.

. 2001. Mutual climatic range reconstructions of seasonal temperatures based on

Late Pleistocene fossil beetle assemblages in Eastern Beringia. Ouat. Sci. Rev.
20:77-91.

Blias, S. A., S. K. Short, C. H. Nelson, & H. H. Birks. 1996. Life and times of the Bering
land bridge. Nature 382:60—63.

Blias, 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. Ouat. Sc. Rev.
20:135—-147.

Hamilton, T. D. 1994. Late Cenozoic glaciation of Alaska. Pages 813-844 zn 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. Cing-Mars. 1995. Radiocarbon dates on saiga antelope (Saga tatar-
ica) fossils from Yukon and the Northwest Territories. Arctic 48:1—7.

Hoffecker, J. F, W. R. Powers, & N. H. Bigelow. 1996. Dry Creek. Pages 343-352 zn F. H.
West, ed., American Beginnings. The Prehistory and Palaeoecology of Beringia. University of
Chicago Press, Chicago, IL.

ENVIRONMENTAL CONDITIONS IN BERINGIA

Hopkins, D. M. 1982. Aspects of the paleogeography of Beringia during the Late
Pleistocene. Pages 3—28 zm D. M. Hopkins, J.V. Matthews, Jr., C. E. Schweger, & S. B.
Young, eds., Paleoecology 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. Marie Geology 125:73—94.

Kunz Vee Ei Manne 2 ES Matheusiica 2) Groves: 999s thesliteandetimess ot
Paleoindians in Arctic Alaska. Arctic Research 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. Colinvaux, & M. C. Miller. 1993. Late Quaternary lacustrine pollen
records from southwestern Beringia. Ouat. Res. 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 routes. Ouat. Sc. Rev. 20:301—314.

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

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

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

Pitul’ko, V. 2001. Terminal Pleistocene — Early Holocene occupation in northeast Asia
and the Zhokhov assemblage. Ouat. Sc. Rev. 20:267—275.

Schweger, C. E., J. V. Matthews, Jr., D. M. Hopkins, & S. B. Young. 1982. Paleoecology
of Beringia: a synthesis. Pages 425-444 7m D. M. Hopkins, J. V. Matthews, Jr., C. E.
Schweger, & S. B. Young, eds., Paleoecology 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. Ouat. Scr. 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 productivity paradox: The
landscape approach. Ouat. Sc. Rev. 20:165.

Zimovy, S: A. Vl. Chuprynin, A. P) Oreshko, FS. Chapin Ill, |, F Reynolds, & M. C:
Chapin. 1995. Steppe-tundra transition: A herbivore-driven biome shift at the end of
the Pleistocene. Amer. Nat. 146:765—794.

wid

‘tan? alt Bp in me 4 alii wy OR ale _
-¢ as ae, Sat ag iain \Es> » 7 A ae ts it

aa at fi fee gad mi Cah teste anal
h ow a es | thea ne 7 5 wl Wy s - a ad ny

Sar; Kphavals to

Ps ee ne > 5 7 —_ i Mm See : : ¥
* Bh eciieis semnll sl: cemialinlad oa a Bar
: a
vag 'te eee aad yea Seige et
= we Ny 46 Gs ' ~ : me 4 ;
2 BATES ik . ee
7 7 , ;
. i t ye you -
tl i » F :
= a i a
’ 7 = ay wea
- eas 7 j aL ‘ ‘ Dy ee

ah, nig 7 tipi SS

is.

od ' — ee
. ’ £ Pee > 2*
in foe. ® =e 14 i A oS re ral

eee air4 ws: 1€a

- _o w*} 5) ios I
vai :
7 7 iv aye es 4
a
ou a rj ‘ - es 7 ay Ry ae

_ ae A ere a he igignte a | es. a
Samay We bere ~ Ss yrrngnaigh elie > 2 —_—
Pintle: Aes. eagels ov ee ae nwt

ay te ip Oe 7 yee a ‘hay se ik lt - meh tyre

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 pretty com-
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 ago2), 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 (¢.g., Haynes 1964,
1992).

By 11,500 years ago, theyd arrived at a sprine-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 fra-
ditional view 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 Dillehay ef a/ 1992) —
and arrived at Tierra del Fuego within a millennium of leaving Alaska (Whitley & Dorn
1995)!

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

MELTZER

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 (¢.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 signature fluted spear points (Figure
1) —1s 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 uniformity of adaptation —
big-game hunting. Chasing mammoths, it was argued, enabled these hunter-gatherers to
hurdle ecological boundaries (¢.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 (eg.,
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 type site
(Blackwater Locality No. 1). The original specimen is ~11 cm in
length. While 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-quality chert,
obtained at a source hundreds of miles distant from the site in which
it was found. (Drawn from a cast by Fredetic Sellet.)

THE EXPLORATION AND COLONIZATION OF NEW LANDS

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 — like
turtles (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 e¢ a/ 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 return to the matter of Clovis origins
below.)

But most troublesome of all for the traditional scenario, there were frequent claims
of a pre-Clovis presence in the Americas. These claims came not just from archaeology,
but also linguistics and genetics (¢.g., Nichols 1990; Torroni ef a/, 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
yeats 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 e¢
al. 1968; but see Morell 1995:362—368). Others featured 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-Clowis credibility
decay curve when it was learned the supposed artifacts were more likely naturally 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
remains.

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

IMUSIC IC ZEW

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 shghtly more than a thousand years older than Clovis,
its great distance from Beringia points to an initial arrival in the Americas much earlier
than 12,500 years ago. How much earlier depends on a host of variables, like whzch 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 anyway, is help us much with
Clovis.

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

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-

THE EXPLORATION AND COLONIZATION OF NEW LANDS

tions comes as no surprise: Given the absence of barriers to cross-Beringian traffic
(Meltzer 1995), and the 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 (archaeologically 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
think.

Furthermore, recent calibration of the radiocarbon curve for this period suggests
that radiocarbon yeats 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 e¢ a/ 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 ef a/ 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

Sil

MELTZER

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 ts, it looks fairly complicat-
ed).

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 | will 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 empty 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
Thule.

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 wete 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 hypothesis 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,
teplaced 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.

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Like the Paleo-Eskimo, they were tracking a keystone species, primarily bowhead whales,
a species they’d hunted for millennia, mostly 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
1984:370=373).°

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 vared 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 unlike 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

MELIZER

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 uncertainty (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 unfamiliar landscape, the risk of extinc-
tions is greatest soon after dispersal, when population numbers and growth rates tend to
be low (Belovsky 1999).7

At the same time, groups moving into increasingly unfamiliar habitats had to /arn 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 a/one. 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 progtessively further south, and into
areas with unfamiliar and exotic floras.®

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Would that suggest groups may have then initially focused on big game, as a class of
ptey 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 differently 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 (¢.g., Van Dyne et a/. 1980:285—298; also Frison 1991:141; Johnson ef a/ 1992). All
of which, of course, presupposes some knowledge of the landscape.

Obviously, one’s knowledge of the environment and ability to predict resource avail-
ability 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 empty continents. Virtually all the historical-
ly known hunter-gatherers have mezghbors: 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, by which 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-

stantly monitoring the landscape and, because one group can only cover so much

IVICA ZZTEIRS

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 Kelly 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
dispersal.

Demography and landscape learning are tightly 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-
ability 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
1967:88).

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 declined, and in the
face of limited knowledge of the landscape; minimizing group size, in order to buffer envi-
ronmental uncertainty or risk on an unknown landscape; maximizing mobility, in order to
learn as much as possible, as quickly as possible about the landscape and its resources (in
order to reduce environmental uncertainty 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 (you 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 /arge scale exploration, 1
order to map the landscape; periodic aggregations of widely dispersed groups in order to
exchange mates, resources, and information; and extensive mating networks, 1 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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

[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
ate not up to the resolution demanded of the models! But let me explore what we might
see of landscape learning, mobility, 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,
pteferred 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 mobility. 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 quality 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 familiarity of colonizers.

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

MELE ZER

blades) that are relatively easy to carry, and once needed could readily be flaked into any
one of a variety 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 ts 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 return 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 the 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 future 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 11,000 years later?
In fact, I tend to suspect the opposite was occurring: That Clovis groups were so new 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.

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Stone is especially suitable for caching, since unlike meat or other foods, it won’t
spoil, won't be attractive to scavengers, and barring tectonic activity, 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
mote widely across otherwise unknown and unpredictable terrain. Over time, as new
soutces of suitable stone are located, and as groups are better able to predict where and
when they will be able to replenish their supplies (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
alliance 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 archaeologically
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
setved 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
sutvival 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 mote broadly similar (stylistically, technologically, and typologically) across a larger
area of North America, than any artifact forms in any later cultural period, Paleoindian
ot otherwise (Figure 2).!9 In fact, around the world the only other artifact forms com-

MELIZER

or Bae Gainey 4
Gr. Basin eels { |

dul

: Ne
\r SS,
\y i

~ | _Eastem fluted

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 s¢/arity across that range may well reflect some measure of the soaa/ 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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Great Basin

stemmed | ' ae I Pr HN
Tg | la

i
~ Cumberland

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 tightly 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 (¢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
1996).

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

MELTZER

+ Windust |

Ficure 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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

a vast and empty landscape that was unknown and unpredictable, in which there was a
selective 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 e¢ a/. 1991; Hoffecker et ad 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 be younger than those further south (Reanier 1995). Still, 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 ef a/ 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 ef a/, 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 ef a/ 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 C/ovis 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 Hill, 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

MELTZER

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 passé 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” while 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 hypothesis” — 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 opportunity to work out the details, as proving their case will take time and will be
no easy matter, as they appreciate.!9

Sull, there are obstacles to this hypothesis, 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-lived: Critics pointed out the similarities between
the two continents’ artifacts were so general as to have little meaning (Holmes 1892):
these were different kmds 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 independently, a
convergence 1n 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 turn of the century, resurfaced briefly in the 1960s (¢.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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

structure of the linking argument has not: It 1s 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 obviously, 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-
ity of convergence. After all, red ochre has been used throughout prehistory wherever it
occuts, 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-quality exotic raw materials 1s
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
vety 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 Clovis
site (the Aubrey site, in Texas [Ferring 1994]) is only 11,500 years old.!+ The purported
pte-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 credulity 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 which were labeled (at least initially), as “Caucasoid,” with the strong
implication they represent a migratory pulse from Europe to America. It is now appat-
ent the most famous (or infamous) of those forms, the Kennewick skeleton, is not

MELTZER

“Caucasoid” at all, but an ancestral Native American with affinities to Asian populations
(Powell & 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 colonization, 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 ro/ogica/ 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 sufficiently 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 phenotypic consequences of the reproductive
isolation of these groups, as the spatial scale of the effectrve 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 vatious
forms are unrelated, or that we have to look anywhere 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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

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 (¢.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
archaeology).

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 density) 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 (¢.g., upwatds of a century), what consequences will this have for the scale of
adaptation? Will it shift from shorter-term, small resource fluctuations 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 mote 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

MELTZER

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?) isomorphic 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 ability 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 peopling 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.

Abstract

While much attention has been focused on when the first Americans arrived,
an equally interesting (and perhaps equally controversial) question pertains to
how they colonized the continent. Leaving aside the compelling evidence from
South America of an early entry into the New World (Dillehay’s Monte Verde
work), tt remains the case that in North America the earliest accepted evidence 1s
of Clovis remains (~11,500 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 colonizers 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 new
resources, cope with novel pathogens and diseases, and locate and replenish vital
supplies of high-quality stone and water? And, how effectively they were able to
maintain their population size and reproductive viability while spread thinly on
the landscape? We have plenty of answers to these questions. Unfortunately, we
cannot figure out which of the answers are right. But we can at the very least
explore some of the critical issues related to colonizing empty continents.

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Acknowledgments

I 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.
Collins, Donald KK. 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 ’m on my
own, blame-wise, for this one.

Endnotes

! 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 literature. 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 literature that interested readers may choose to pursue.
Finally, I am only speaking of ages in general, so calibration would be superfluous.

3 Let me stress that this does mot 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 turning 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

MEE IACZIEIRS

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 empty ocean, say ~3500 yr B.P., they moved with surprising
speed across the distant islands of the Pacific Urwin 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 well 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 climatic 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. While there is evidence that change can be rapid in short-lived, temperature sen-
sitive, lake-dwelling 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.

7 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 still 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 require construction of a far better bridge between it and the data than we
now possess.

5 Moerman (personal communication, 1999) 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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

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

? 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
caches.

10 The circles and ovals on this and subsequent distribution maps are intentionally
highly generalized: What I wish to convey ate 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, anything approximating a detailed distributional map (though Anderson e¢
al. {2000] are making important headway in that direction). Second, as Jack Hofman
(1992) and others (¢,g., LeTourneau 1998; O’Brien & Lyman 2000) have rightly argued,
there is considerable variability

spatial, temporal, stylistic, technological, etc. — in
these point forms. Unfortunately, that variability 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
captured 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. All of which
highlights the value of gaining better control on the many dimensions of variability of
these point forms, for when such data are ultimately available, it will provide important
insights into Paleoindian artifact variability, distribution, and so on.

11 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 settlement strategy. Perhaps. But one must always be warty 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?

12 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 antiquity — that the radiocarbon ages were cont-

MELTZER

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 recently reported. There is little 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 all the
mote 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.

13 Tt 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.

'4 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.

'S T 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.

Literature Cited

Abbott, C. C. 1889. Evidences of the antiquity of man in eastern North America. Proc.
LMP ZAG, ZC. SG, DUP).

Adovasio, J. M., J. Donahue, J., & R. Stuckenrath. 1990. The Meadowcroft Rockshelter
radiocarbon chronology 1975-1990. Am. Antiquity 55:348—-354.

Alley, R. B. 2000. The Younger Dryas cold interval as viewed from central Greenland.
Quat. Sc. Rev. 19:213-226.

Anderson, D. 1990. The Paleoindian colonization of eastern North America: a view
from the southeastern United States. Res. 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: Multscalar Analyses and
Interpretations in the Eastern Woodlands. University of Tennessee Press, Knoxville, TN.

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Anderson, D., & M. Faught. 2000. Paleoindian artifact distribution: Evidence and impli-
cations. Antiquity 74:507—-513.

Anderson, D., & K. Sassaman. 1996. The Paleoindian and Early 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 Ameticas.
Pages 209-230 in T. D. Dillehey & D. J. Meltzer. eds.,The First Americans: Search and
Research. CRC Press, Baton Rouge, LA.

Belovsky, G., C. Mellison, C. Larson, & P. A. Van Zandt. 1999. Experimental studies of
extinction dynamics. Scence 286: 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. Sc. 97:1390-1394.

Boldurian, A., & J. L. Cotter. 1999. Clovis Revisited. New Perspectives on Paleoindian
Adaptations from Blackwater Draw, New Mexico. University of Pennsylvania Press,
Philadelphia, PA. 145 pp.

Bonnichsen, R. 1991. Clovis origins. Pages 309-329 im 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 i 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—3106.

Collins, M. B. 1991. Rockshelters and the early archaeological record in the Americas.
Pages 157-182 in T. Dillehay & D. Meltzer, eds.,The First Americans: Search and Research.
GRE WresswBocakatomel le:

. 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. Saence 63:463-468.

Derev’anko, A. P., ed. 1998. The Paleolithic of Siberia: New Discoveries and Interpretations.
University of Illinois Press, Urbana, IL. 406 pp.

Dillehay, T. 1989. Monte Verde: A Late Pleistocene Settlement in Chile. vol. 1. Palaeoenvironment
and Site Ccontext. Smithsonian Institution Press, Washington, D.C.

. 1997. Monte Verde: A Late Pleistocene Settlement in Chile. vol. 2. The Archaeological

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

59)

MELIZER

Dillehay, T., G. Calderon, G. Politis, & M. Beltrao. 1992. Earliest hunters and gatherers
of South America. J. World Prehist. 6:145—204.

Dumond, D. 1984. Prehistory: Summary. Pages 72-79 im D. Damas, ed., Handbook of
North American Indians, vol. 5, Arctic. Smithsonian Institution Press, Washington, D.C.

Edwards wRen)) BeckaG, Bure Donahwey || Chappcllyye Blooms Eee Druttc esau
Taylor. 1993. A large drop in atmospheric 14 C/12 C and reduced melting in the
Younger Dryas, documented with 230Th ages of corals. Scence 260:962—968

FAUNMAP working group. 1996. Spatial response of mammals to Late Quaternary
environmental fluctuations. Sczence 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. Aksenovy. 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 mm R. Bonnichsen & K. Turnmire, eds., Clovis: Onrgins 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, J. Fre/d
Archaeology 26:325—342.

Goodyear, A. 1979. A Hypothesis 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 Paleolithic and the New World. Curr. Anthropol. 4:41—66.

Haynes, C. V. 1964. Fluted points: Their age and dispersion. Scence 145:1408-1413.

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

. 1982. Were Clovis progenitors in Beringia? Pages 383-398 zm D. Hopkins, J.

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

Wordly INNZ

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

pling of the New World. Pages 355-374 zm R. Taylor, A. Long, & R. Kra, eds., Radio

After Four Decades: An Interdisciphinary Perspective. Springer-Verlag, New York, NY.

THE EXPLORATION AND COLONIZATION OF NEW LANDS

Hoffecker, J. 1996. Introduction to the archaeology of Beringia. Pages 149-153 i F-
West, ed., American 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. Saence 259:46—53.

Hofman, J. 1992. Recognition and interpretation of Folsom technological variability on
the Southern Plains. Pages 193-224 7m 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. Saence 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. Alley,
& 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
University 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 mm 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 im R. Bonnichsen & K. Turnmite, eds., Clovis: Origins and Adaptations.
Center for the Study of the First Americans, Corvallis, OR.

Kaplan, H., & K. Hill. 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 Lifeways. Smithsonian
Institution Press, Washington, D.C. 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. Sczence 273:634-636.

Kirch, P. V. 1988. Long-distance exchange and island colonization: The Lapita case.
Norwegian Archaeol. Rev. 21:103—117.

. 1997. The Lapita Peoples: Ancestors of the Ocean World. Blackwell Publishers,

Cambridge, MA. 353 pp.

MELCIZER

Leakey, L. S., R. Simpson, & T. Clements. 1968. Archaeological excavations in the Calico
Mountains, California: Preliminary report. Saence 160:1022—1023.

LeTournau, P. 1998. The “Folsom problem.” Pages 52-73 m A. Ramenofsky & A.
Steffen, eds., Unit Issues in Archaeology. University of Utah Press, Salt Lake City, UT.

Lourandos, N. 1997. Continent of Hunter-gatherers: New Perspectives in Austrahan Prehistory.
Cambridge University Press, Cambridge, MA. 390 pp.

Lynch, T. 1983. The Paleo-indians. Pages 86-137 zm J. D. Jennings, ed., Ancient South
Americans. WH. Freeman, San Francisco, CA.

MacArthur, R. H., & E. O. Wilson 1967. The Theory of Island Biogeography. 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. Sc. 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. Ouat. Sc. Rev. 20:301—314.

Martin, P. S. 1973. The discovery of America. Sczence 179:969-974.

. 1987. Clovisia the beautiful. Naz Hest. 96:10-13.

Martin, P. S., & R.G. Nlein, eds. 1984. Ouaternary Extinctions. University of Arizona Press,
Tucson, AZ. 892 pp:

Mason, R. J. 1962. The Paleo-Indian tradition in eastern North America. Curr. Anthropol.
3:227-283.

MacAvoy, J., J. Baker, J. Feathers, R. Hodges, L. McWeeney, & T. Whyte. 2000. Summary
of research at the Cactus Hill archaeological site, 445X202, 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 im D. Damas, ed.,
Handbook of North American Indians, vol. 5, Arctic. Smithsonian Institution Press,
Washington, D.C.

. 1998. The Ferst Peophng of Arctic North America. Plenary Address, 31st Annual
Chacmool Conference, “On Being First: Cultural Innovation and Environmental Consequences of
Forst Peoplings.” University of Calgary, Alberta, Canada.

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

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

Pages 11-39 mm C. Ellis & J. Lothrop, eds., Eastern Paleo-Indian Lithic Resource

Procurement and Processing. Westview Press, Boulder, CO.

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

THE EXPLORATION AND COLONIZATION OF NEW LANDS

uity in America. Pages 13-49 mm T. D. Dillehey & D. J. Meltzer, eds., The Farst
Americans: Search and Research. CRC Press, Baton Rouge, LA.

. 1993a. Search for the First Americans. Sroithsonian Books, Washington, D.C.
176 pp.

. 1993b. Is there a Clovis adaptation? Pages 293-310 zm 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. Rev. Anthropol. 24:21—-A5.

. 1997. Monte Verde and the Pleistocene peopling of the Americas. Science
276:754-755.

. 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 Peoplings.’ Calgary Archaeology

Association, Calgary, Alberta, Canada.

. 1999. Comment on “Reproductive interests and forager mobility,’ by D.
MacDonald & B.S. Hewlett. Curr. 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? Exhnology
Monographs 12:31—43.

. 1991. Peopling of the New World: A discussion. Pages 303-307 m R.
Bonnichsen & K. Turnmire, Clovis: Origins and Adaptations. Center for the Study 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 diversity and the first settlement of the New World. Language
66:475—521.

O’Brien, M. J., & R. L. Lyman. 2000. Apphing Evolutionary Archaeology: A Systematic
Approach. Plenum Press, New York, NY. 471 pp.

Powell, J.. & J. Rose. 1999. Report on the osteological assessment Oi Wave
“Kennewick Man” skeleton (CENWW.97.Kennewick). Internet citation:
http://www.cr.nps.zov/aad/ 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 Research in the Pleistocene 15:67—68.

Soffer, O. 1985. The Upper Paleolithic of the Central Russian Plain. Academic Press, Orlando.
IIL, SID) (ayo

MEEIZER

Stanford, D. J. 1991. Clovis origins and adaptations: an introductory perspective. Pages
1-13 mm R. Bonnichsen & K. Turnmire, eds., Clovis: Origins and Adaptations. Center for
the Study of the First Americans, Corvallis, OR.

Stephens, D., & J. Krebs. 1986. Foragng Theory. Princeton University 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 zm R. Bonnichsen
& K. Turnmire, eds., Clovis: Origins and Adaptations. Center for the Study of the First
Americans, Corvallis, OR.

Straus, L. 2000. Solutrean settlement 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
65:493—508.

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 im A. Breymeyer & G. M. Van Dyne, eds., Grasslands, Systems
Analysis, and Man, Cambridge University Press, Cambridge, MA.

Walthall, jf. 1998. Rockshelters and hunter-gatherer adaptation to the
Pleistocene/Holocene transition. Am. Antiquity 63:223—238.

Whallon, R. 1989. Elements of culture change in the Later Palaeolithic. Pages 433-454
mm P. Mellars & C. Stringer, eds., The Human Revolution: Behavioral and Biological
Perspectives on the Origins of Modern Humans. Priceton University 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.

Whitley, D.S ., & R. Dorn. 1993. New perspectives on the Clovis vs. pre-Clovis contro-
versy. dm. 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
113-135 mT. D., Dillehey & D. J. Meltzer, eds., The First Americans: Ssearch 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 barrier and the
inability to cope with water 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 utilized until this late pre-agricultural period...
for early man, water 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 antiquity 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. Recently, 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
consetvative compared to scholarly claims and media accounts that the Americas may

have first been colonized by seafaring Australian Aborigines, Polynesians, or even

Onl
\S

ERLANDSON

Europeans who crossed the open Pacific or Atlantic oceans by boat (¢.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 settled, 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 Dillehay 1997; Meltzer et a/ 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 a/ 1989; Mandryk 1991; Wright 1991; Elias, this vol-
ume). Research has also shown that the coastlines of Alaska and British Columbia
deelaciated earlier than once thought and were more suitable for human occupation than
previously believed (ze, Mann 1986; Molnia 1986; Barrie e¢ a 1993; Heaton & Grady
1993; Mann & Peteet 1994; Mann & Hamilton 1995; Dixon ef al 1997; Josenhans ef 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
antiquity 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.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Anatomically Modern Humans: An 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! (Fagan 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 diversity 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 diversity, 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-

ERLANDSON

lion year old fossil “hominid” jaw dubbed Ramapithecus, 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 settled (much to the conster-
nation of many paleoanthropologists) only when a nearly complete Ramapithecus 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 lineages, 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? (¢.g., Krings ef a/ 1997), an emerging consensus argues that some
version of the Out of Africa theory is probably correct (Nlein 1995; Mellars 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 earliest 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; Nlein 1995). These remains, between about 120,000
and 65,000 years old, are associated with Middle Stone Age stone tool assemblages that
include the earliest blade technologies, the earliest microlithic 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 ef a/ 1995; Yellen et a/ 1995; Yellen 1998).
Also dated to about 90,000 years are the earliest 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.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

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
atchaic 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. Significantly, 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 150,000 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; Mellars 1998). In the process, anatom-
ically and intellectually modern humans appear to have replaced the vast majority of
atchaic 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 Homo 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 others).
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 prehistoric record
has attracted a general consensus. ... A real commitment to maritime lifeways did not
precede late Upper Paleolithic times.” Through the 1970s, there was virtual unanimity
that boats, too, were a relatively recent addition to the human technological repertoire

(Bass 1972:12; Greenhill 1976; Johnstone 1980:xv).

ERLANDSON

Antiquity of Coastal Adaptations: 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 millions 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-
bors.

It is true that there is relatively little 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
resoutces 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).4

Geologically, there is every reason to believe 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 significantly, the record of human occupation associated with lower shore-
lines has either been inundated by rising sea levels, destroyed by associated coastal ero-

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

sion, ot both. Even today, with sea level approximately six meters below the levels of
about 130,000 years ago, a number of important coastal sites (Klasies 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 (z.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 18,000 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 typically 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 ef a/ 1928;

ERLANDSON

1X9] OY} UT poUONUOUT sv “UY ITIOVg 9y} UO UONLIGKY [vISVOD JO SvoIv PUL SoS IFO sTYAIg “| AUNOLY

|
Sel COOL O 0001

Ag god Zh-G Zh
ania”, suo

UeAIO DYIDe

dg 44 000°0L-LE
Aenfer epeaganiy @ or -
ABNYBIE | Epeiganc
ae Ao00zt-se.
Spuels| uownjog” Mba,
PUBL SA] Mah

BauIN) Anapy

oa 4A goo GL-2e

2 4AQ0S OLE uedep pues] nAyndy «

SpPuels| (AUUE YS
area Asieg ‘ag 4A goo'o2-62
WEdeP PUBS] BLUIUSNZ0s zi

eee

Gee eis
af . g afl ‘a fate A-0 L - ae

i rere 3 pue|s | ULM Bee at
EAE) ERIE H PUe ele
S} ERIEH F Bons

BYSE|y ISB AayINOS =
— = e——4,

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Waechter 1964), several shell middens located on the Melanesian islands of New Ireland
and New Britain (Allen e¢ a/ 1988; Wickler & Spriggs 1988), and Daisy Cave on
California’s Channel Islands (Erlandson ef a/ 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 diversity 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
vatied 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
limited 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 Klasies River Mouth caves, Die Kelders, and other Last
Interglacial localities in South Africa, for instance, AMH appear to have regularly eaten
a vatiety 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 (Yellen e/ a/ 1995).

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

ERLANDSON

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 out
dramatic demographic and geographic expansion of the last 150,000 years (Erlandson

2001).
The Antiquity 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, apparently settling the Indonesian island
of Flores as much as 700,000 to 800,000 years ago (Sondaar et al 1994; Morwood ef al.
1998). At present, however, there is little other evidence for the use of watercraft by
Flomo 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 antiquity of human seafaring significantly. One of these was the discovery of
obsidian from the Mediterranean island of Melos in strata at Pranchthi 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 ef a/ 1986), 50,000 years (Roberts et a/,
1990), and now possibly to 60,000 years or more (Thorne ef a/. 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

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Table 1. Islands Colonized or Explored by Pleistocene Seafarers.

Locality

Flores, SE Asia

New Guinea
and Australia

Crete, Greece

Bismarck Archipelago,
Melanesia

Sicily, Italy

Ryukyu Islands,
Japan

Kozushima Island,
Japan

Melos Island, Greece
Admiralty Islands
Melanesia

Cyprus

Channel Islands,
California

Southeast Alaska
& Bntish Columbia

Descnption of Evidence

Possible evidence for limited seafarmg by

Hom erectus.

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

Honm sajnens sapiens remains with poorly
documented context; calcareous breccia in
which bones were cemented dated by Pa/U to
51,000 + 12,000 BP: colonization of Crete
apparently required several short sea crossings.

Shell middens, fishing, and seafaring at
several sites dated from 15-35 KYR, with
voyages up to 140 km long.

Aurignacian assemblage from Mediterranean
Island; possibly involving a sea crossing.

Human skeletal remains found in Yamashita-cho
and other caves on Okinawa and other islands;
involves voyages of ca. 75-150 km.

Upper Paleolithic peoples on Honshu crossing
50 km wide channel to obtain obsidian.

Travel across ca. 24 km of open water to

obtain obsidian for mainland trade.

Settlement of Manus Island required 200 km

voyage.

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

Boat and marine resource use by coastal
Paleoindian groups, with sea crossings of

at least 10 km.

Presence on islands indicates a maritime
lifestyle and seafanng capabilities.

Age

(years B.P.)

800,000?

60-40,000

50,000?

35,000

30,000

32-15,000

25-20,000

13,000

12,000

10,300

11-10,000

10-9,000

References

Morwood et al. 1998;
Sondaar et al, 1984.

Groube et al, 1986;
Roberts et al. 1990;
Thome et al, 1999.

Facchini & Giusberti
1992

Allen eal 1988, 1989;
Wickler & Spriggs 1988.

Chilardi et al, 1996;
Cherry 1990.

Matsuura 1996.

Oda 1990:64.

Cherry 1990.

Allen & Kershaw 1996.

Cherry 1990:151.

Erlandson et al. 1996;
Johnson et al, 2000;
Orr 1968.

Davis et al, 1989;
Fedje & Chnstensen
1999.

ERLANDSON

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 settlement 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 cultural complex in western Melanesia dis-
coveted several Pleistocene shell middens in the Bismarck Archipelago and the Solomon
Islands east of New Guinea (Allen e¢ a/ 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 coastlines 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 settlement 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
settled, Melanesian seafarers had reached Manus Island in the Admiralty 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. Fagan

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

(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. (Matsuura 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 Ryukyu 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
yt B.P. (Matsw’ura 1996:187). The bathymetry of the Ryukyu 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 mote likely route would have been to follow
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 Homo 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 situated at the base
of a maritime pathway to the New World. Did they make such a journey?

Vik

ERLANDSON

Out of Asia: Migration Routes into the New 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 settlement 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 significantly, despite repeated claims for much
eatlier occupations. When I first began seriously studying archaeology in the 1970s, 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 (¢g., Taylor e¢ a/ 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 vit-
tually all 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 et a/ 1993). With the widespread acceptance of the
12,500 year old pre-Clovis site of Monte Verde in Chile (Meltzer et a/ 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 Land: 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

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

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, fluted 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 ef a/ 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 hypothesis — 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

ERLANDSON

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 significantly.

Over the years, numerous scholars have proposed that a coastal migration may have
contributed to the initial peopling of the Americas (¢g., Heusser 1960; Chard 1963;
Laughlin 1967; Fladmark 1979, 1986; Mithun 1979; Gruhn 1988, 1994; Easton 1991;
Dixon 1993, 1999; Erlandson 1994; Fedje & Christensen 1999; and others). With 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
corridor.

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. Consequently, 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 human migration than previ-
ously thought. Recent evidence even suggests that an ice-free interior route may only
have become available as recently as 11,000 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-
tor occupation came first, followed by migrations to the coast and the development of
coastal foraging economies.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

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 Archaeological 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 1s the southern coast of Alaska,
compatable 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
eatlier 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 ate relatively protected from marine erosion and boat travel is relatively easy in
sheltered inside waters. Marine productivity 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

ERLANDSON

of recent interdisciplinary efforts to find early coastal sites (see Dixon ef a/ 1997;
Josenhans ef 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 ef a/ 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 ef a/ 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 a/ 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 e¢ a/ 1996; Josenhans ef a/, 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 visibility. Unfortunately, it is not currently possible to
estimate the length of that process.

On the southern Northwest Coast in Washington, Oregon, and northern California,
no shell middens more than about 3500 years old were known prior to about 1985. Even
today, the vast majority of coastal archaeological sites within this area date to the late
Holocene, but a few early and middle Holocene sites have now been found (Moss &
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 ef a/, 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

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

accelerated marine erosion occur every few centuries. It now appears that large portions
of the coast may drop a metet or more during such earthquakes, raising sea levels virtu-
ally instantaneously and leading to severe marine erosion (Peterson e/ a/ 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 ef a/ 1996). Stratigraphically above this terminal Pleistocene
stratum 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. 11 m) 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 ef a/. 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 dertved 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 lion, 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.

ERLANDSON

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 11,000 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. 11,000 14 C yr B.P. (Sandweiss e¢ a/. 1998). Keefer and colleagues (1998)
reported similar remains from another Peruvian midden dated to ca. 10,700 14 C yr BP.
and the basal strata at the Ring site shell midden produced a similar date (Sandweiss ef a/.
1989). Richardson (1998) also reported the presence of the remains of shellfish collect-
ed from mangtove habitats at a series of ephemeral campsites in the Talara region of
northwest Peru, sites dated between about 9000 and 11,200 14 C yt BP 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 still no clear evidence for the presence of fully 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 earlier 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 (¢.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 geographic 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 1f we know that the data we have may be biased, that we must build our

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

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 possibility that Pleistocene seafaring peoples colonized the Americas are
among those that cannot currently be answered with any sense of authority. 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,
virtually 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 along 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 suitability 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 antiquity of maritime adaptations,

ERLANDSON

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. | 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 erectus populations that spread out of
Africa about a million 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
America.

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
150,000 years ago then rapidly spread around the world, largely replacing archaic human
populations. If this Out of Africa model 1s 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 mote 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 hatpoons
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

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

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 actually 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
ot Polynesians crossing the open Pacific, or Europeans crossing the open Atlantic.
Despite such claims, I believe the earliest occupants of the New World were the ances-
tors of the Native Americans who lived 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 tockshelters 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 paleoshorelines and landscapes, consideration of the sedimentary and
depositional regimes that may affect the preservation and visibility of ancient sites, and
the identification of areas most likely to have been used by early maritime peoples.

ERLANDSON

Abstract

After the appearance of anatomically modern humans about 150,000 years
ago, evidence for technological innovation, marine resource use, and seafaring all
increase dramatically. Intentional maritime voyaging appears to have developed
sometime after about 75,000 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 Asia 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 colonization of the
New World is still largely circumstantial, but a variety of data suggest that it 1s
increasingly likely that such a maritime migration took place. I suspect, in fact,
that the Pleistocene colonization of the Americas may have included both land-
based migrations through the Beringian interior and maritime voyaging around the

North Pacific Rim.
Acknowledgments

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 crystallize 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 significantly 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 peopling of the New World have been influenced by discussions with,
among others, Mel Aikens, Jim Dixon, Jim Hagegarty, Richard Jordan, Rick Knecht,
Madonna Moss, Roger Powers, Jim Richardson, Dennis Stanford, and David Yesnetr.
Nonetheless, the opinions and conclusions expressed in this paper are peculiarly my own.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Endnotes

1 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 ¢7 a/.
1994).

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
own.

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 stratum 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 possibility 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 e/ a/ 1993:163, 198). It is uncertain, however, if the fish remains were of cultur-
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 eatlier 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

ERLANDSON

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 currently seems unlikely.

Literature Cited

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

Ackerman, R. E., K. C. Reid, J. D. Gallison, & M. E. Roe. 1985. Archaeology of Heceta
Island: A Survey of 16 Timber Harvest Units in the Tongass National Forest, Southeastern
Alaska. Center for Northwest Anthropology Project Report 3, Washington State
University, Pullman, 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 mm L. G. Straus, B. V. Eriksen, J. M. Erlandson, & D. R. Yesner, eds.,
Flumans at the End of the Ice Age: The Archaeology of the Pletstocene-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.

Allen, 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 zm P. Mellars, 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. Ouat. Int. 20:123-129.

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

Bigalke, E. H. 1973. The exploitation of shellfish by the coastal tribesmen of the
Transkei. Annual of the Cape Province Museum 9159-175.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Binford, L. R. 1984. Faunal Remains from Klasies River Mouth. Academic Press, New York,
NY. 283 pp.

Birdsell, J. B. 1977. The recalibration of a paradigm for the first peopling of greater
Australia. Pages 113-167 im J. Allen, J. Golson, & R. Jones, eds., Sunda and Sahut:
Prehistoric Studies in Southeast Asia, Melanesia, and Australa. Academic Press, New York,
NY.

Bowler, J. M., & A. G. Thorne. 1976. Human remains from Lake Mungo. Pages 127-138
in R. I. Kitk & A. G. Thorne, eds., The Origin of the Austrahans. Humanities Press, NJ.

Bowdler, S. 1977. The coastal colonization of Australia. Pages 205-246 zm J. Allen, J.
Golson, & R. Jones, eds., Sanda and Sahul: Prehistoric Studies in South East Asia,
Melanesia, and Australia. Academic Press, New York, NY.

Brooks, A. S., D. M. Helgten, 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. Scence 268:548—552.

Cann, R. L. 1994. MtDNA and Native Americans: a southern perspective. Am. J. Hum.
Genet. 55:7-11.

Cann, R., M. Stoneking, & A. Wilson. 1987. Mitochondrial DNA and human evolution.
Nature 325:31—36.

Carlson, R. L. 1995. Early Namu. Pages 83-102 im 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. Uni.
Alaska 10:115-121.

Cherry, J. F 1990. The first colonization of the Mediterranean islands: A review of
recent research. J. 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—503.

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 zw R. J. Fulton, Gal.
Onuaternary Geology of Canada and Greenland. Geological Society of Canada, Ottawa.

Clark, J. T. 1991. Early settlement of the Indo-Pacific. J. Anthropol. Archaeol. 10:27-53.

Cohen, M. N. 1977. The Food Crisis in Prehistory: Overpopulation and the Origin of Agriculture.
Yale University 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.

ERLANDSON

Davis, S. D., ed. 1989. The Hidden Falls Site, Baranof Island, Alaska. Narora Monographs
V, Alaska Anthropological Association. 383 pp.

Dillehay, T. D. 1997. Monte Verde, A Late Plestocene Settlement in Chik, vol. 2, The
Archaeological Context and Interpretation. Sraithsonian Institution Press, Washington D.C.
1071 pp.

Dixon, E. J. 1993. Quest for the Origins of the First Americans. University of New Mexico
Press, Albuquerque, NM. 156 pp.

. 1999. Bones, Boats, ¢ Bison: Archeology and the First Colonization of Western North
America. University of New Mexico Press, Albuquerque, NM. 322 pp.

Dixon Es) lho rHeaton wie binelds i Ds amilion Das sPutmam sc. EiGradyaslo One
Late Quaternary regional geoarchaeology of southeast Alaska karst: A progress
report. Geoarchaeology 12:689—712.

Easton, N. A. 1991. Mal de Mer above Terra Icognita, or, “What ails the coastal migra-
tion theory.” Arctic Anthropol. 29:28-41.

Erlandson, J. M. 1993. California’s coastal prehistory: A circum -Pacific perspective.
Proceedings of the Socety for California Archaeology 6:23—36. Society for California
Archaeology, San Diego, CA.

— . 1994. Early Hunter-Gatherers of the Cahfornia Coast. Plenum, New York, NY.
336 pp.

. 2001. The archaeology of aquatic adaptations: Paradigms for a new millenni-
um. J. 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 zn 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
Pletstocene-Holocene Transition. Plenum, New York, NY.

Erlandson, J. M., D. J. Kennett, B. L. Ingram, D. A. Guthrie, D. P. Morris, M. Tveskov,
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 3-18 z R. J. Losey,

ed., Changing Landscapes: Proceedings of the 3 rd Annual Coquille Cultural Preservation
Conference, 1999. Coquille Indian Tribe, North Bend, OR.

Facchini, F, & G. Giusberti. 1992. Homo sapiens sapiens remains from the island of Crete.
Pages 189-208 zm G, Brauer & F. H. Smith, eds.Continuity or Replacement: Controversies in
Homo sapiens Evolution. A. A. Balkema, Rotterdam.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Fagan, 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 Gwatt. 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 Gwati Haanas. Arctic Anthropol. 33:116-142.

Fladmark, K. R. 1979. Routes: Alternate migration corridors for Early Man in North
America. Am. Antiquity 44:55—69.

. 1986. Getting one’s berings. Nat. Hast. 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. J. Royal Anthropol. Inst. Great
Britain and Ireland 58:33-113.

Goebel, EF. E., W. R. Powers, & N. H. Bigelow. 1991.’ The Nenana complex of Alaska and
Clovis origins. Pages 49-79 im R. Bonnichsen & K. L. Turnmire, eds., Clovis Origins

and Adaptations. Center for the Study of the First Americans, Oregon State University,
Corvallis, 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 Study of the First Americans, Oregon State University,
Corvallis, 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. Curr. Res. Pleistocene 10:98—100.

Heusser, C. J. 1960. Late Pleistocene Environments of North Pacific North America.
Am. Geogr. Soc. Spe. Publ. 35.

Hoffecker, J. E., W. R. Powers, & F. E. Goebel. 1993. The colonization of Beringia and
the peopling of the New World. Sczence 259:46—53.

Irwin, G. 1992. The Prehistoric Exploration and Colonisation of the Pacific. Cambridge
University Press, Cambridge, MA. 240 pp.

ERLANDSON

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, & J. R. Southon. 1997. Early humans and rapid-
ly changing Holocene sea levels in the Queen Charlotte Islands—Hecate Strait,
British Columbia, Canada. Sczence 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. Scence 281:1833-1835.

Wein, R. G. 1995. Anatomy, behavior, and modern human origins. J. World Prehuist.
9:167—241.

. 1998. Why anatomically modern people did not disperse from Africa 100,000
years ago. Pages 509-521 mm 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. J. Ham.
Evol. 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. J. Archaeol Sci. 13:515—-542.
Krings, M., A. Stone, R. W. Schmitz, H. Krainitzki, M. Stoneking, & S. Paabo. 1997.

Neandertal DNA sequences and the origin of modern humans. Ce// 90:19—30.

Lampert, R. J. 1971. Burrill Lake and Currarong. Terra Australis1. Australian National

University, Canberra. 86 pp.

Laughlin, W. S. 1967. Human migration and permanent occupation in the Bering Sea
area. Pages 409-450 zm D. M. Hopkins, ed., Te Bering Land 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.
QOuat. Res. 28:245—262.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Mandryk, C. S. 1991. Paleoecology as Contextual Archaeology: Human Viability of the Late
Onuaternary Tce-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, AIX

Mann, D. H., & T. D. Hamilton. 1995. Late Pleistocene and Holocene paleoenviron-
ments of the North Pacific Coast. Ouat. Sa. Rev 14:449-471.

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

Matsw’ura, S. 1996. A chronological review of Pleistocene human remains from the
Japanese archipelago. Pages 181-197 i K. Omoto, ed., Interdiscaphinary Perspectives on
the Origins of the Japanese. International Research Center for Japanese Studies.

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

Meehan, B. 1982. She// Bed 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 im N. G. Jablonski & L. C. Aiello, eds., The Ongin and
Diversification of Language, Memoirs of the California Academy of Sciences No. 24,
San Francisco, CA.

Meltzer, D. J., D. K. Grayson, G. Ardila, A. W. Barker, D. FE 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 hypothesis. Phoenix: New 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., G/aciation in
Alaska: The Geological Record, Alaska Geological Society, Anchorage, AK.

Morwood, M. J., P. B. O’Sullivan, 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. J. World Prehist. 9:1—45.
. 1998. Early Holocene adaptation on the southern Northwest Coast. J. Cahjfornia
and Great Basin Anthropol. 20:13—25.

ERLANDSON

. 1999. The seacaves of southeast Alaska. Archaeol Univ. of Oregon 2:6-9.
Nunez, L., J. Varla, R. Casamiquela, & C. Villagran. 1994. Reconstruccion multidiscipli-

natia de la ocupacion prehistorica de Querero, centro de Chile. Latin American
Antiquity 5:99-118.

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
27:417-419

. 1968. Prehistory of Santa Rosa Island. 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 zm L. R.
Binford, ed., For Theory Building in Archaeology. Academic Press, New York, NY.

Parkineton, J. 1981. The effects of environmental change on the scheduling of visits to
the Elands Bay Cave, Cape Province, S.A. Pages 341-359 zm 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
Engineering Geoscience 6:255—269.

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

Richardson, J. B. HI. 1998. Looking in the right places: Pre-5000 B.P. maritime adapta-
tions in Peru and the changing environment. Revista de Arqueologia Americana
15333=56)

Rightmire, G. P., & H. J. Deacon. 1991. Comparative studies of Late Pleistocene human-
remains from Klasies River Mouth, South Africa. J. Hum. 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 Hostory of the
Osmore Drainage, Peru. BAR International Series 545.

Sandweiss, D. H., H. McInnis, R. L. Burger, A. Cano, B. Ojeda, R. Paredes, M. Del
Carmen Sandweiss, & M. Glascock. 1998. Quebrada Jaguay: Early South American
maritime adaptations. Scence 281:1830—833.

MARITIME VOYAGING AND COLONIZATION OF THE AMERICAS

Schwaderer, R. 1992. Archaeological test excavation at the Duncans Point Cave,
CA-SON-348/H. Pages 55-71 i T. L. Jones, ed., Essays on the Prehistory of Maritime
California. Center for Archaeological Research at Davis Publication 10, University 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 wm G. G. Bailey
& J. Parkineton, 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, E 1994. Samples, species, and speculations in the study of modern human ofigins.
Pages 227-252 in M. Nitecki & D. Nitecki, eds., Origins of Anatomically Modern Humans,
Plenum, New York, NY.

Sondaar, P. Y., G. D. van den Bergh, B. Mubroto, F. Aziz, J. de Vos, & U. L. Batu. 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 Study of Neandertal Ecology.
Princeton University Press, Princeton, NJ. 447 pp.

Stothert, K. 1985. The preceramic Las Vegas culture of coastal Ecuador. Am. Antiquity
50:613-637.

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

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

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

Taylor, R. E., L. A. Payen, C. A. Prior, P. J. Slota, Jr., R. Gillespie, J. A. J. Gowlett, R. E.
M. Hedges, A. J. T. Jull, 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.Taylotr,
& D. Curnoe. 1999. Australia’s oldest human remains: Age of the Lake Mungo 3
skeleton. J. Hum. Evol. 36:591—612.

Toyne, S. 1999. Aborigines were the first Americans. The Sanday Times, August 22. Times
Newspapers, Ltd.

van Andel, T. H. 1989. Late Quaternary sea level changes and archaeology. Amiquity

63:733—745.

ERLANDSON

Waechter, J. D. 1964. The Excavation of Gorham’s Cave, Gibraltar, 1951-54. Bull. Inst.
Archaeol. 4189-221.

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 zn B.
Lee & I. 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
IDC,

Wright, H. E. Jr. 1991. Environmental conditions for Paleoindian immigration. Pages
113-135 aT. D. Dillehay & D. J. Meltzer, eds., The Ferst Americans: Search and Research.
CINE Imes, loca, Ikaora, INL,

Yellen, J. E. 1998. Barbed bone points: Tradition and continuity in Saharan and sub-
Saharan Africa. African Archaeol. Rev. 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. Sccence
268:553—556.

Yesner, D. R. 1980. Maritime hunter-gatherers: Ecology and prehistory. Curr. Anthropol.
22735.

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

societies. Pages 285-310 7m M. Harris & E. Ross, eds., Food and Evolution. Temple

University Press, Philadelphia, PA.

. 1996. Human adaptation at the Pleistocene-Holocene boundary (circa 13,000

to 8,000 BP) in eastern Beringia. Pages 255-276 zm L. G. Straus, B. V. Eriksen, J. M.

Erlandson, & D. R. Yesner, eds., Hamans 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

Anthropologists’ interest in North America’s human fossil record
has risen and fallen since scholars first became interested in who the
first Americans were, how the New World continents were colo-
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 peopling of the Americas (¢.g., Hrdlicka 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 ef a/ 1982).

The shift away from issues concerning the origins of American Indians was caused
by several factors. One major factor was Ales Hrdlicka’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 (Hrdlicka 1902, 1907, 1918, & 1923). In Hrdlicka’s view,
Europe had archaic forms of humans who co-existed with the cold-adapted Pleistocene
megafauna while 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 climates saw the demise of the cold-adapted biological populations and their
replacement by the biological populations whose descendants lived into historic times.
This changing face of Europe was the theater in which Hrdlicka believed the
Neanderthals evolved into the anatomically modern humans. In the Americas, however,
it was not until after the end of the Ice Age and the establishment of the recent and bio-

logical communities that the first humans, anatomically modern compared to the

STEELE AND POWELL

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 similarity within and between the populations within the Americas. With the
above model in place, and the European Neanderthals as the standard of reference for
eatly 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 mote 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.

Currently, 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.

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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 little 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;
Powell & Steele 1992; Steele 1989; Steele & Powell 1992, 1993, 1994), the 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
eatly Archaic antiquity 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 dissimilarity. 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-
atates 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-

STEELE AND POWELL

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 analyses.
For a complete listing of Paleoindian remains see citations listed.

Site Sample Size Nature of Remains Age (yr B.P.)

White Water Draw’ 2 skeletons 8,000 - 10,000
Gordon Creek ! 1 skeleton 8,700 + 250
Browns Valley’ 1 skeleton 8,700 + 110
Pelican Rapids ' 1 skeleton (Early Archaic)
Sauk Valley ' 1 skeleton (Early Archaic)
Wilson-Leonard ’ it skeleton 9,000 - 11,000
Hom Shelter’ 2 skeletons 9,000 - 10,000
Spirit Cave * 1 skeleton

Wizard’s Beach’ 1 skeleton

Kennewick ° 1 skeleton 8,410 + 50
Station do Riacho

Lapa Vermelha * 1 skeleton 11,000 - 12,000
Cerca Grande * 8 skeletons 9720 + 130

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

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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 II. 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 proximity 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.).

STEELE AND POWELL

Australian samples, and away from the Northeast Asian samples. The European samples
are widely dispersed in the upper half of the graph: 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 clearly identify
which Paleoindian cranial features were most distinctive (Steele & Powell 1992, 1994). By
using these simpler statistics 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 (¢.g., Indian

PRIN3 o ate
1.76
6) 7 a
eS +e ,
0.52 1] Pa
| Ba
i = °
= 2 (9 at I Me le | eel
Di east thee een melee 8 9.7 ee en ese ~ 0.16
ae PRIN1
-1.96
2.49

PRIN2 -9-38 -3.94

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 another
reflects their differences. Early samples are Paleoindians (pyramid); Upper Cave, China (cube);
Minatogawa, Taiwan (flag) and Jomon, Japan (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.)

THE NORTH AMERICAN HUMAN FOSSIL RECORD

PRIN3 C]

2.49

0.89

-0.71 -

-3.65
-3.19

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 populations are Northeast Asian (diamond), southern Pacific Rim (open circle), and Europe
(open square). (After Steele and Powell 1994.)

KKnoll 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 recently 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

STEELE AND POWELL

FiGuRE 4. The Wilson-Leonard female cranium from Leander, TX. Although this skull has been
securely dated to be 9,500 years old, the skull was too badly crushed to be measured accurately. 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-
tial. The antiquity 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 & Powell 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 Polynesians and Australians than they 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 measurements

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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 populations of the southern Pacific Rim than they do northern
Asians. The features which 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 antiquity of Paleoindians (predating 8,500 yr B.P.).

101

STBELE AND POWELL

These resemblances of Paleoindians to more recent populations of the 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.

Kennewick, Man

The recent discovery of a human skeleton of early Holocene age near Kennewick,
Washington, 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 recently, 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 the first
date obtained and reported by Chatters. The third date (6940 + 30 14 C yr B.P.) confirms
the antiquity 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.

Currently 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 studies have been published: One ts by Chatters (2000) based upon his exam-
ination of the remains before the lawsuit was filed and the other study is by 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 world populations using 10—52 measurements depending upon

102

THE NORTH AMERICAN HUMAN FOSSIL RECORD

160

155 wa
Aleut wicks
150 : “S j
* TN Miss. Minatogawa
145 NW Coast Upper Cave
X Qa x ® O,Pre-Aleut ®
Cc TN Wood. - a K5
B 140 an PoP ee coinsians
Ne hee
Pecos TN Archaic =O
135 Oo Indian os ne sp fettin®
Oo
130 a Hes TX Archaic
2/5
160 165 170 175 180 185 190 195 200 205 210
GOL
150
145 Pe TN Miss.
te TN Wood.
140 2 NW @
Pecos y te Coast
Xx
Gass ~~ * i Upper Cave
B an

130 *& TX Archaic
*
125

120
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
atea) 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.)

103

STEELE AND POWELL

N
p
H
Peru x fc Paleoindians
* Minatogawa
120 125 130 135 140 145 150
ZYB
Pre-Aleut
N
P
H

Minatogawa
®

115 120 125 130 135 140
ZYB

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 lines mark traditionally recognized boundaries
for identifying 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 (solid 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/early Holocene Upper Cave and Minatogawa samples. (After Steele and Powell 1992.)

104

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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’ dissimilarity, 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 crantum and the denti-
tion both substantiated the distincttveness from the extant American Indians.
Summarizing their results from all 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 e¢ a/ (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).

105

STEELE AND POWELL

ee
Sanaa tee

FIGURE 8. Frontal and lateral views of the Wilson-Leonard IT adult female during three stages in
the process of a preparing facial reconstruction based on the skull. 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 Betty Pat Gatliff, a world-renowned forensic
artist. The facial reconstruction was made for the Houston Museum of Science. Photographs are by
Betty Pat Gatliff.

106

THE NORTH AMERICAN HUMAN FOSSIL RECORD

Early Holocene Remains 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 (eg., 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

BEAS

1.731 6) m4 ine

PCAI

PCA2

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.)

107

STEELE AND POWELL

PCA3

1.14 7 O Oo
Al

Ns | if

-1.32 St a Bae? | 2.05
oe ee Ae | | pAtes
; pe ils cal ee
2.55 aol Sam ROAl
3.14 ne
215
0.04 Aes
PCA2

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 population,
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 ef a/ 1999). Figure 11 illustrates the structural similar-
ity of the face and braincase of the Hominid 1 female from Lapa Vermelha IV compared
to a worldwide selection 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 similarity 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 the American

108

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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.

PCA 3
2.18 ey 0860 viet as FB EAST
MOKA Fy Cmori XX TENT
() SJAP ZULU XX
LA
aRIKE# ©} NuaP () TOLA Oynusr
0.52 HAIN
— ISTCR (_) TASM

-1.14

-2.80

PCA 1 -3.58

Ficure 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
1999.)

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 this issue. In his dissertation on dental
variation in prehistoric American Indians, Powell (1995) compared the Paleoindian sam-

109

STEELE AND POWELL

ple from North America with early to mid-Holocene 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-Holocene
swath of time.

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.

PCA II

PCA I

FIGURE 12. Bivariate plot of first two components (PCA 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 within 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 individually marked. (After
Powell and Neves 1999.)

110

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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 ate 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

_a Brea

WMS
c i Sta. Riacho

CG 7B

Windover a Plains A i CG 6A

Spirit Lapa Vermelha
sie cG7A

SE Archaic

x Bird Is. Sauk Valley

Pa 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.)

etl

STEELE AND POWELL

perceived homogeneity was thought to exist because the initial colonizers 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 conditions. With our current awareness of the heterogeneity of
American populations, we can no longer assume a recent colonization of the New
World. Awareness of this heterozygosity also means that each Paleoindian and Early
Archaic find will have an extraordinary potential for providing new information about

America’s earliest colonizers.
Summary

In summary, the research over the past 10 years has documented the following: 1)
The earliest remains recovered from the Americas have consistently 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
populations.

What is the Meaning?

Since the reconsideration of the American fossil record began in the 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-
ulations.

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

THE NORTH AMERICAN HUMAN FOSSIL RECORD

TABLE 2. Alternative explanations focusing upon the Paleoindians to explain why they differ from

recent American Indians.

Explanation

Explanation 1: Sampling Error:

Paleoindians presumed to look like
late Holocene humans, but sampling

errors create a false impression of
Paleoindian population.

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.

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.

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.

Support

Very small sample sizes.

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

Explains the distinctiveness
of Paleoindians and the later
similarities of Northeast

Asians and American Indians.

Incorporates evolutionary
models in explanation.

WS)

Negating evidence

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

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.

Does not consider effects of
evolutionary forces.

Must assume parallel
evolutionary forces acting

on Northeast Asian and
American populations if
both differ from Paleoidians

in the same way.

STEELE AND POWELL

of the actual Paleoindian population as it existed. The reason for this proposed inaccu-
rate image of the Paleoindians 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 they came. From the inception of 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 features 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 lifestyles?

Until recently, the most widely held view was that if the features 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 studying 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 anoth-
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 mote recent colonizer to the New World than either

114

“punoj ueeq ou sey ssad01d
oun sve uoTdezas yemmavu AFaUEp!
pmnom ie) Wierd oanouNsip VW

“HvISW ISeayUON]
queixa ayy] 1OU st uoNepndod
Surpuno} ft suorejndod
uvIpuy UeotouNYy JO AUeTTUNTS
urerdxa Apaanoazyo 1OU Sood

“PpoM Many UT peatue
rare uonepndod uo sunse
soor0f AreuonNyoas afqissod
IapIsuoOd OU SeOP [aPoy/'T

“sueipuy URSHOLUaYy qUEIXS
wort ToyFip Apusisisuod
sureutot uPIpUurose

a ee a a

[SROINE CSS aa
pure souspra gy Sumesapy

‘PHOM
MONI UT SooI0} AreuUOFINTOA
jo Amjiqissod s1aprsucD

-suonepndod Surpuno}

uodn 15¥ 03 uTaYyI 10}

au] YSNOUS YI JEpour ur
sao10F A1vUOTINJOAS Sapnypouy

-suonepndod urisy iseoyuoONy

Aq pedueny fur Apravoy Inq
‘suonepndod uvisy wor;
qounisip ore suonejndod
ueIpUy ULOUOUNYy TRAX

“sonia Surfdures

01 onp suerpurosfeg

jo ssoudATOUNIsICy “TepTUUTS
AIA dIv SUPIPU] WeoLSUy
pue uvisy isvoyvoyy qvixg

[SROUN OF
aouapiay Suri0ddnsg
pure suonduimssy

*SJUSLUTIOMAUD [PIO]

ur suonepndod uo sunoe
wondayas femyeu O1 anp
aounsip st anq ‘suonejndod
SuIpUNOo] YM samy
auros sareys uonendog

“UOIIEZTUOTOS ATE ATOISTY
Areuonnyoaa anbrun jo
asnvsaq 1azfIp nq ‘va1dep
e o1 suonepndod jensaour
ajquiosax suoneyndod
ueoueury WEpUs.seq]

‘uonepndod

yore YIM sonsiaiovaeyo
sazeys inq ‘suonepndod
SuIpunoy ayi JO yora
WIOIF SIOFFIP uonepndod
ueoLeuny ULpUssseq]

“sueISY ISPOYUONT
pure suerpuy uesuoury
quasar JO saimqeay poreys
wro1y uonepndod Surpunoy
Jo soimeay SIayUT “pro0sa1
[SsO} UvoHOUNY YUON Jo
aoueptAs idaaae Jou seod

uonepndog surpuno.y
oi suonepndog
ueoueuny Wweix gy

Jo saouequiasayy

“UONde]es yeameny

“MOT} 9ues ysnoryr
suonendod Sunruna1
yo pouad Aq pomoyjo}
uonezruojos yo aseyd

Ajrea Suunp 1lrp oneueD

*yeuururtu 10 JUONT

*‘yeuuTunu 410 sUOKT

sao10,J AreuonnJOA Fy
PHOM “eN

-odoing Ajqissod pur visy
uro1f suonepndod Burpuno};
afdnynur 10 a{Zurs : [PPO

“uvISy IseayINOS
IO uvIsy isesyuONy
Ajyensn ‘uonejndod

Suripunoy a[Buts =¢ [Epoy|

“eISy iseayON
WOT} Iq O1 PataprIsuod osje

‘suonepndod proxy pjO s10ur

Io au0 Aq pamoyo} ‘uvisy
jsvoyuoNKy IO uviseoioig
se parpiuapr Ayyensn
‘uonepndod pproyy PIO
SurIpuno} af8urs :Z [PPC

“ueIsy IseayUONT
se parynuapr uonendod
SuIpUNO; JISurs =] [PPO

uonendog surpuno.f
PROM PIO

Ils

THE NORTH AMERICAN HUMAN FOSSIL RECORD

EE ae = ee SES OME go Pn ee See ee ee ee ee SS

‘SULIPUT UBOTTOUY YON, JO SOINIVIF [eIeforuRsto IY} povoto JALY YSTYAN SOITOF our ureydxo QO} Sjapouwl IANVUTOTV “C aldv L

STEELE AND POWELL

has to the earlier colonizers who more closely resembled the Southeast Asians. The pic-
ture may become even more complicated if there has been gene 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.

Lahr (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, yet 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.

Laht’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

116

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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
vatiation (Fst) 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 Fs4, Wright’s (1951) measure of within-group vatia-
tion, and used this to estimate between group variation by combining population into
one sample and estimating that group’s Fst. In cases where the Fst 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 F's/, 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

IIL

STEELE AND POWELL

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 populations. North American Paleoindians (solid 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.

Conclusions

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 they do late Holocene remains. While all of these conclusions have been based on

118

THE NORTH AMERICAN HUMAN FOSSIL RECORD

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
ate 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
arguments.

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 features; or a mote 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.

119

STEELE AND POWELL

Abstract

The oldest human skeletal remains found in North America share broad sim-
tlarities with northern Asians, Southeast Asians, Pacific Islanders, and recent
Native Americans. Among 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 view that the Americas
were colonized by Northeast Asians in a time too recent for evolutionary forces to
have altered the features of their Native American descendants. While the dif-
ferences between the earliest and more recent skeletal remains have become well
documented, identifying the evolutionary force or forces which created these differ-
ences 15 more equivocal.

Literature Cited

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

Chatters, J. 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. |.
Phys. Anthropol. 99:389—412.

Howells, W. W. 1969. The use of multivariate techniques in the study of skeletal popula-
tions. Am. ]. 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. Archeol. Ethno. 67,
Harv Univ. Cambridge, MA.

. 1989. Skull shapes and the map. Craniometric Analyses in the Dispersion of
Modern Homo. Paps. Peabody Mus. Archeol. Ethno. 79, Harvard University, Cambridge,
MA.

Hrdlicka, A. 1902. The crania of Trenton, New Jersey and their bearing upon the antiq-
uity of man in that region. Bu/l Am. Mus. Nat. Hist. 16:23-62.

. 1907. Skeletal remains suggesting or attributed to early man in North America.
Am. Bur. Am. Ethno. Bull. 331-113.

. 1918. Recent discoveries attributed to early man in America. Am. Bur. Am.

Ethno. Bull. 66:1—67.

120

THE NORTH AMERICAN HUMAN FOSSIL RECORD

. 1923. The origin and antiquity of the American Indian. Pages 481—493 i
Annual Report 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 zm 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 morphometrics
of the Spirit Cave mummy. New Hest. Quart. 40:62-84.

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

MacManamon, F. 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:
http://www.ct.nps.gov/aad/Kennewick/cl4memo.htm.

Neves, W. A., & H. M. Pucciarelli. 1989. Extra-continental biological relationship of early
South American human remains: A multivariate analysis. Czéncia e Cultura 41:506-575.

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

based on early South American human remains. J. Ham. Evol. 21:261273.

. 1998. The Zhoukoudien upper cave skull 101 as seen from the Americas. J.
Hum. Evol, 34:219-222.

Neves, W. A., D. Meyer, & H. M. Pucciarelli. 1996. Early skeletal remains and the 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. Phys.
Anthropol. Suppl. 22:1706.

Neves, W. A., J. EF 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. F. Powell, & E. G. Ozolins. 1999. Modern human origins as seen from
the peripheries. J. 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—-105m. Sao Paulo.

Ozolins, E. O., V. H. Stefan, M. L. Rhoads, & J. F. Powell. 1997. Craniofacial morpho-
mettic similarity between modern and Late Pleistocene human populations. Am. J.

Phys. Anthropol. 24:182-183

STEELE AND POWEEL

Powell, J. F. 1993. Dental evidence for the peopling of the new world: Some method-
ological considerations. Hum. Biol 65:799-815.
. 1995. Dental variation and biological affinity among middle Holocene human 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 42:153-198.

Powell, J. F, & J. C. Rose. 1999. Report on the osteological assessment of the
“Kennewick man” skeleton (CENWW.97.Kennewick). Internet citation:
http://ww.cr.nps.gov/aad/kennewick/powell_rose.htm.

Powell, J. EF, & DG. Steele. 1992. A multivariate craniometric analysis of North
American paleoindian remains. Cur. Res. Plast. 9:59—62.

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

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

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

. 1993. Paleobiology of the first Americans. Evol Anthropol. 2138-146.

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

metric view. Pages 141-163 zm 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 University, Corvallis, OR.

. 1999. Peopling of the Americas: A historical and comparative perspective.
Pages 91-120 m R. Bonnichsen & R. Gruhn, eds., Who Were the Forst Americans.
Peopling of the Americas Publication, Center for the Study of First Americans,
Oregon State University, Corvallis, OR.

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

Wright, J. W. 1951. The genetic structure of populations. Ayn. 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:
BIOARCHAEOLOGICAL EVIDENCE FOR THE
COLONIZATION OF THE NEw WOorRLD

Christy G. Turner II

There are many improbable and unscientific views dealing with the
otigins and peopling of the New World, such as autochthonous cre-
ation (numerous Native American religious traditions), lost Israelites

(Mormonism), 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 (Laughlin
& 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 (Fagan 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 ¢¢ a/. 1999); diachronic studies on crania (Powell & Neves 1999), dental
morphology (Turner 1992a), and stone tools (Dillehay & Meltzer 1991). All try to assess
Old World origins and affinities: timing, derived from dates of early archaeological sites

W273)

TURNER

or typological correlations; language classification for number 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
1994).

Operationally, holisitic endeavors utilize at least two scientific principles of evidence
evaluation — parsimony and concordance. The best single hypothesis 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 types of evidence should meet reasonable
standards of robusticity, redundancy, independence, and consensus. Reality, ultimately, is
that which we agree it Is.

On the other hand, as in all areas of scientific investigation, there can be curious ev1-
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 cultural 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: early
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, Diukta1), and south Beringian coastal proto-Aleut-Eskimos
(lower Amur, Hokkaido). I introduce the term epi-Clovis in recognition that the hallmark
object in Clovis identification, the fluted spear point, had probably not been fully devel-
oped in the first bands to reach Alaska (such as Component I at Dry Creek, often
referred to as the Nenana Complex (Hoffecker ef a/, 1993); however, other correspond-
ing stone tool types are present in Clovis and the Alaskan epi-Clovis (Hoffecker ef a/,
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 the eight major questions that, viewed together, provide
a broad multidisciplinary working hypothesis for the human colonization of the
Americas.

124

TEETH, NEEDLES, DOGS, AND SIBERIA

Where Was the Regional Ancestral 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, Hrdlicka (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. Shoveling is very common
in all Native Americans and Northeast Asians but is much less common, to very rate,
elsewhere in subfossil and modern world populations. Subsequent studies by Pedersen
(1949), Moorrees (1957), KK. Hanihara (1968), Zubov and Khaldeeva (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 Nattve 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 features 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 Sinodonty and Sundadonty. Figure 4 shows in synoptic multivariate fashion that
all Native Americans, ancient and recent, are more like 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 continuity
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 studies 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, Sinodonty
has not been found in any non-Asian skeletal assemblage regardless of geological time

125

TURNER

(Zubov & Khaldeeva 1979; Grine 1981; Turner & Markowitz 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 China Zhoukoudian Upper Cave 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 Native 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 sutural 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 e¢ a/ 1999) and Ainu (Powell & 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 agriculture 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 foodstuffs cooked in pots. This softening reduced the need for the heavy
food chewing stress that contributed to cranial robusticity, just as muscular activity
affects other areas of the skeleton. Moreover these stews and gruels seem to have con-
tributed significantly 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 shifts in food preparation
techniques, even beginning in Archaic times, that could have affected cranial form.

126

TEETH, NEEDLES, DOGS, AND SIBERIA

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
ptesence of a parastyle on the second molar, a rare feature. 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. Wetherill 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 shoveling ts 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 studied 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 in the form of artifacts made of stone, the typologies of which

127

TURNER

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; Vasil’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 types found in the
Aleutian Islands. Throughout much of Late Pleistocene Northeast Siberia, Russian

archaeologists have recovered microblades and bifacial tools together, an assemblage

128

TEETH, NEEDLES, DOGS, AND SIBERIA

usually referred to as the Dituktai 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 like 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 a/ 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 recetved on both temporal and tech-

FIGURE 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).

29)

TURNER

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. continues to be a hotly debated subject (West 1996; Hall 2000; Roosevelt
et al., 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 specificity, 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
Americans.

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.

NORTH-SOUTH AMERICA
ARCTIC

NE ASIA

NEOLITHIC SE ASIA
CENTRAL EUROPE
CRO-MAGNON

BALTIC

GERMANTIC

SLAVIC

MELANESIA

SRI LANKA

EGYPT

JOMON-AINU
PALEOLITHIC SE ASIA
POLYNESIA
MICRONESIA
AUSTRALIA

WEST AFRICA
PLEISTOCENE NUBIA

FIGURE 4. Dendrogram of worldwide dental relationships based on the multivariate Mean Measure
of Divergence statistic 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 similarity between 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, 19 groups with-
out central Asians).

130

TEETH, NEEDLES, DOGS, AND SIBERIA

Y GROOVE M2
ROCKER JAW
3-CUSP P2 — SINODONTY
MAND. TORUS -- SUNDADONTY
CARABELLI
WINGING
DOUBLE-SHOVEL
I-ROOT M2
3-ROOT MI

CUSP 5 MI

CUSP 6 MI
ENAMEL EXTEN.
HYPOCONE
SHOVEL

I-ROOT Pl

L&R M3 +

CUSP 5 M2

O 20 40 60 80 100%
PATTERN VARIATION IN MONGOLOID DENTITION

FiGuRE 5. Shown are some of the dental traits and their frequencies used to define Sundadonty and
Sinodonty. The former is a dental patttern characterized by retention and simplification, 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 all 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
little 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 1999) would likewise have left
no evidence if the southern edge of the Atlantic pack ice was followed into North
America as he suggests. And suggestions for a trans-Pacific route from Australia has the

IS)Il

TURNER

same problem of not being disprovable due to the flooding of the continental shelf of
South America where landings might have taken place. Lacking independent corrobora-
tion or conformity with the generally agreed upon Bering land bridge hypothesis,
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-
posals.

The Bering land bridge route could have involved not only an interior crossing, but
also a south coastal drift or both. There could also have been relatively early and later
transits as well, and these could have been more-or-less discrete movements, or a more
drawn-out and semi-continuous 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 eatly 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 little 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 ef a/ 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 (Laughlin 1980).

When Was the Time of Arrival?

To date, no archaeological sites have been found in Alaska that are older than 12,000
yt B.P. (Hoffecker et a/ 1993), although there are controversial claims for earlier sites fur-
ther south such as Meadowcroft in Pennsylvania and Monte Verde in Chile (reviewed in
Meltzer [1993] and in nearly every issue of Mammoth Trumpet). Mach has been written by
Russian scholars about the ancient, historic, and modern peoples of Siberia and their

=
LSS)
ii)

TEETH, NEEDLES, DOGS, ANID SIBERIA

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 mote stable, less severe, and
archaeological sites are much earlier and more abundant in southern Siberia than in the
far north. In the Altat 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 e¢ af 1993). The 10,260—13,700 year-old Berelekh site 1s 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 types 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 climate 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 & Klein 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
across.

One of the many interesting taphonomic findings my Russian co-workers Nicolai
Ovodov, Olga Pavlova, Nicolai Martynovich, 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 Altat 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

133

TURNER

Pleistocene Mousterians made only limited use of fire? While it is likely that they 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 they 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 children, especially those bands with low
population density 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 studying 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.

TEETH, NEEDLES, DOGS, AND SIBERIA

How 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 recently, 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 quality of being able to directly characterize living, recent, and subfossil popula-
tions.

Kk. 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 Astan 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 type, 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
ot 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-

TURNER

quencies for several traits 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 e¢ a/ 1998). Even traits such as root number can be confused
in older individuals when cement accretion has been excessive. Chatters (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 differentiating between Sinodonty 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 ate 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
presently known dental descriptions from these areas that are closely calibrated through
the use of the Arizona State University Dental Anthropology System (ASU DAS)
(Turner e¢ a/ 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 (Mortis 1965, Southwest 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
NSO, OIE ISA OA. WOsrre, I Aal IS, Noy OW). Oo Seo Se Ieianveie
1997; Turner & Hawkey 1991, 1998; Bailey-Schmidt 1995; E. Scott 1998). In addition,
strong concordances have been made between the many 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 directly on
many samples of dental crown and root morphology of past human populations, aided
by the aforementioned students and coworkers, I will first discuss my 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. My Asian,
Pacific and New World studies have identified two major dental patterns as previously

TEETH, NEEDLES, DOGS, AND SIBERIA

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 Sundadonty, which I call
Proto-Sundadonty, 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 like the European dental pattern
than is Sinodonty. In fact, Sundadonty 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 skull called Shilka found east of Lake Baikal near the city
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 Minatogawa 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
ot Jomon-like artifacts have been found (Popov ef a/. 1997) there as well as in Sakhalin
(Vasilievsky 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 1s 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 partied with compelling reason that these three

toy

TURNER

families are genetically 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 & Turner 1997; Scott 1994). This uncertainty is a con-
tinuing area of investigation. Regardless of whether two or three migrations are eventu-
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 1s 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 studies (Cavalli-
Sforza ef al. 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
analysis.

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 Khaldeeva (1979, & elsewhere) and their
Russian co-workers, show without much question the continuity 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 &
diumers997)):

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
eatly migration of “Australoids” (summarized in Lahr 1995; Neves ef a/. 1999), or at least
a group of early migrants whose crania differed in robustness from those of more recent
South American Indians. While these analyses are based on very few crania, and they are

138

TEETH, NEEDLES, DOGS, AND SIBERIA

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, sumatraliths, 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 1s 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
less like Native Americans and more like some Old World populations, possibly

are

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
unlike Ainu or Europeans (Figure 6).

The most controversial proposal for the peopling of the New World comes from
Dennis Stanford (1999 and this sympostum), 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. All 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.

159

TURNER

Teeth not only provide a resource for identifying cladistic relationships, they also 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
ereater than in sub-Arctic dentitions. Much of this traumatic damage is thought to have

g
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.

SPIRIT-WIZARD
RECENT NEVADA
ALABAMA SHELL MOUND
ARCHAIC BRAZIL

N CALIF, HUMBOLDT

N CALIF, SACRAMENTO
PREHISTORIC NEVADA
N CALIF, ALAMEDA
GRASSHOPPER, ARIZ
HAWIKUH, NEW MEXICO
AINU, HOKKATDO
EASTER ISLAND
JOMON, HOKKAIDO
KAMAKURA, JAPAN
HONG KONG, RECENT
CENTRAL THAILAND
MOKAPU, HAWAII
MESOLITHIC UKRAINE
DANISH NEOLITHIC
POUNDBURY, ENGLAND

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 similarity with Ainu
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
[= 3=99):

What is the Genetic Affinity Between Native Americansand
Peoples of the Old World Including the Pacific Basin?

Most genetic evidence, serological and DNA, points to strong cladistic connections
between Native Americans and Northeast Asians (Szathmary 1979, 1985; many other
workers). But there are one or two genes that hint synchronically at some ancient or
more recent minor relationship with eastern Europe or western Siberia. Neither the Late

140

TEETH, NEEDLES, DOGS, AND SIBERIA

Pleistocene archaeology of southern Siberia nor a new mitochondrial DNA study
(Voevoda ef a/. 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 Beringiaé

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 cultural microevolution. The smaller the population, the greater is the opportunity
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 biocultural 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 lived to reproductive age themselves, then the chances of there
being a representative sample of genes (and to some degree, culture) transmitted to
future generations is less than 50 percent. If success breeds success, as it often does, then
a tapid train of cultural 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 uniformity in
lithic projectile point style throughout North America. A similar low amount of vati-
ability characterizes most of the genetic and dental features 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 feature 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,

141

TURNER

much longer period of human occupation, and equally diverse environmental zones, is
that the 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 Level 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). All these groups possess similar skills, values, and equipment needed for living 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 antiquity and evolution is often

TEETH, NEEDLES, DOGS, AND SIBERIA

‘aspug puvy] suiog 3svo7y YINOS (g {99] Youg vas
ysIOUFO (V “syed PA] “WSN avo yds (ez SWS “Yoweg prez (1Z “WS “N94 At (OZ “VSA ‘VMsuvuy (1 ‘vIssmy “ETYs_] (BT ‘vIssnYy
‘yyjorog (L] ‘vissmy ‘oars reryniq (QJ ‘erssmy ‘[] UeLUsTOg (G] ‘uvde( ‘emvsoiwury (py ‘oousog “YyeIN (¢] ‘euTyD ‘avez Joddq uempnoynoy7
(ZI serssmy ‘T vprmsy (TT feissmy “vayepy (QT ‘vissmy ‘oavy aoyruprepyo (6 ‘vIssmy ‘vaostuaq (g ‘vissny ‘vAvuuTRYy (1 ‘vIssmy ‘vA yoruTogzey
(9 seissny Suvsp-3scq, (G ‘vissmy ‘vysepseaq (p ‘vissmy “YuIIMyy (¢ {[—SOT aaraprsoy (7 ‘vissmy ‘sIoyy (] ‘soMyeIOT, aS “L ANNO]

¢ «

SS OOO DO0¢ i) DOE V

B/D SyQUE TU

Pdi HEBIO WIE”

, : .
: = ‘i

TOA T URN
addaqs snamy
37 ERSU

SPMD Spout he

= —e ee "=

143

TURNER

abundant on both local and regional scales. Of more perishable and behavioral items
such as clothing and social organization, prehistoric evidence is much less and usually
indirect and more inferential. However, I feel it is generally fair to say that high Arctic
Upper 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-
gests.

Many of the same vertebrate species found in 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 &
Anderson 1980). Human terrestrial hunters and gatherers of southern Siberia would
have had to make no major technological changes as far as the types of animals hunted
farther north was concerned. However, intensity and duration of winter cold, mobility,
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 clothing assembled with small
tailoring needles, one of which Russian associates Ovodoy, Martynovich, 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 Altat 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. This 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 II, south of Vladivostok (Popov ef a/, 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
(1999).

144

TEETH, NEEDLES, DOGS, AND SIBERIA

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; heavy 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 fireless 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
resoutces, either autarchic or under the control of other groups. Similarly, as groups
attempted to live 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
19S).

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 /upus, 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. (OQvodov
1998) is about the same age as the dog “burial” discovered by Dikov (1997) at the

145

TURNER

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 complexity. Dogs would also have been valuable for locating, 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). Recently, 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 proximity
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 exemplified by the few and seemingly
genetically unrelated Paleosiberian languages of Northeast Siberia (Ruhlen 1975; Crystal
1987).

Conclusions

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 outline in hand. Any scenario for the peopling of the New World that
does not also consider relevant biocultural and natural history events around the world,
as has been the objective of our dental anthropological surveys and affinity assessments,
fails the evidentiary test of holism for acceptability. There is no better example of the
value of worldwide synthesis than the ‘our de force by geneticists Cavalli-Sforza and col-

146

TEETH, NEEDLES, DOGS, AND SIBERIA

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... ”
Cavalli-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 (Dtuktai-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 (1993) 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 Paleolithic cultural 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 Sundadonty in the New World, although divergent
and convergent dental microevolution has occurred within the general Sinodont pattern.
Claims for Sundadonty (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-

147

TURNER

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 characterization be made of the cultural and biologi-
cal complex that enabled native Siberians to reach and cross Beringia? At present, one
can do little more than recite the biocultural characteristics 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 patently 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.

Abstract

The title of this essay symbolizes my holistic orientation and current under-
standing of Siberian and New World anthropology. In this paper I examine
information from dental morphology, 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) what route(s) were followed?, (3) when was the time(s) of
arrival?, (4) how many migrations were 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-

148

TEETH, NEEDLES, DOGS, AND SIBERIA

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, (5) 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
SySTCM.

Acknowledgments

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
atrangements 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.

Literature 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
Anomalies 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,
Corvallis, OR. 264 pp.

149

TURNER

Bonnichsen, R., & K. Turnmire, eds. 1991. Clovis: Origins and Adaptations. Center for the
Study of the First Americans. Corvallis, OR. 344 pp.

Boorstin, D. J. 1983. The Discoverers. Random House, New York, NY. 745 pp.

Brown, P. 1998. The first Mongoloids, another look at Upper Cave 101, Liujiang and
Minatogawa 1. Acta Anthropol. Sin. 17:255—275.

Burnett, S. E., J. D. Irish, & M. R. Fong. 1998. How much is too much? Examining the
effect of dental wear on studies of dental morphology. Am. J. Phys. Anthropol., Sapp.
26:115-116.

Campbell, L. 1986. Comment. Curr. Anthropol. 27(5):488.

Carlson, R. L., & L. D. Bona, eds. 1996. Earl) Human Occupation in British Columbia.
University of British Columbia Press, Vancouver, Canada. 261 pp.

Cavalli-Sforza, L. L., P. Menozzi, & A. Piazza. 1994. The History and Geography of Human
Genes. Princeton University 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 65:291-310.

Crespo. E. F. 1994. Dental Analysis of Human Burials Recovered from Punta Candelero: A
Prehistoric Site on the Southeast Coast of Puerto Rico. M.A. thesis, Department of
Anthropology, Arizona State University, Tempe, AZ.

Crystal, D. 1987. The Cambridge Encyclopedia of Language. Cambridge University Press,
Cambridge, MA. 472 pp.

Damas, D., ed. 1984. Handbook of North American Indians: Arctic. Srnithsonian Institution,
Washington, D.C. 829 pp.

Derevanko, A. P. 1990. Paleohthic of North Asia and the Problem of Ancient Migrations.
Translated by I. Lancheva. Institute of History, Philology, and Philosophy, Academy of
Sciences of the USSR, Siberian Division, Novosibirsk. 123 pp.

Dikov, N. N. 1997. Asva at the Juncture with America in Antiquity (The Stone Age of the Chukchi
Peninsula). Translated by B.L. Bland. National Park Service, Beringia Program,
Anchorage, AK. 248 pp.

Dillehay, T. D. 1999. The Late Pleistocene cultures of South America. Evol Anthropol.
7:206-216.

Dillehay, T. D. ,& D. J. Meltzer, eds. 1991. The First Americans: Search and Research. CRC
Press, Boca Raton, FL. 310 pp.

Dumond, D. E. 1987. The Eskimos and Alents. Revised edition. Thames and Hudson, New
York, NY. 180 pp.

Echo-Hawk, R. C. 2000. Ancient history in the New World: Integrating oral traditions
and the archaeological record. Am. Antiquity 65:267-290.

150

TEETH, NEEDLES, DOGS, AND SIBERIA

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., Early Man in America from a Circum-Pacific
Perspective. University of Alberta Department of Anthropology Occasional Paper 1,
Edmonton, Canada.

Gibbons, A. 1996. The peopling of the Americas. Sczence 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 zn D. G. Sutton, ed., Omgin of the Furst New
Zealanders. University of Auckland Press, Auckland, New Zealand.

Greenberg, J. H. 1987. Language in the Americas. Stanford University Press, Stanford, CA.
438 pp.

. 1990. The American Indian language controversy. The Rev. Archaeol. 11:5-14.

. 1996. Beringia and New World origins: I. The linguistic evidence. Pages
525-536 in FE. H. West, ed., American Beginnings: The Prehistory and Palaeoecology of
Beringia. University of Chicago Press, Chicago, IL.

Grine, EK E. 1981. Occlusal morphology of the mandibular permanent molars of the
South African Negro and the Kalahari San (Bushman). Amn. S. Afr. Mus. 86:157—215.

Guthrie, R. D. 1990. Frozen Fauna of the Mammoth Steppe: The Story 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,
Mempely AZ:

. 1996. Dental Anthropology of Russia, URraine, Caucasus, Central Asia: The Evaluation

of Five Hypotheses for Paleo-Indian 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. J. Ham. Ecol. (Special Issue) 2:273-297.

Hall, D. A. 2000. Charting a new era: Clovis and beyond draws over 1,400. Mammoth
Trumpet 15:1—7.

Bescon, INL I2, S; 1k, Zeauren, a. resaxem, |, C, Ibex, Wh Iii, IN, (Ce Gransciaars, Is IR.
Templeton, L. P. Osipova, O. L. Posukh, & T. M. Karafet. 1999. New World Y chro-
mosome found haplotypes and the peopling of the Americas. Am. J. Phys. Anthropol.
Suppl. 28:144.

Lil

TURNER

Hanihara, IK. 1968. Mongoloid dental complex in the permanent dentition. Pages
298—300 in Proceedings of the VIIth International Congress of Anthropological and E:thnological
Sciences. Vol. 1. Science Council of Japan, Tokyo. 395 pp.

Hanihara, T. 1991. Dental and cranial evidence on the affinities of the East Asian and
Pacific populations. Pages 119-137 m Ik. Hanihara, ed., Japanese as a Member of the
Asian and Pacific Populations: International Symposium 4. International Research Center
for Japanese Studies, Kyoto.

Harris, E. F. 1977. Anthropologic and Genetic Aspects of the Dental Morhology of Solomon
Islanders, Melanesia. Ph.D. dissertation, Department of Anthropology, Arizona State
University, Tempe, AZ.

Hawkey, D. E. 1998. Out of Asia: Dental Evidence for Microevolution and Affinities of Early
Populations from India/ Sn Lanka. Ph.D. dissertation, Department of Anthropology,
Arizona State University, Tempe, AZ.

Hoftecker, J. F., W. R. Powers, & T. Goebel. 1993. The colonization of Beringia and the
peopling of the New World. Saence 259:46—53.

Hopkins, D. M. 1996. Introduction: The Concept of Beringia. Pages xvii—xxi mm F. H.
West, ed., American Beginnings: The Prehistory and Palaeoecology of Beringia. University of
Chicago Press, Chicago, IL.

Howells, W. W. 1995. Whos Who in Skulls: Ethnic Identification of Crania from Measurements.
Papers of the Peabody Museum of Archaeology and Ethnology, Harvard University,
Vol. 82. Cambridge, MA. 108 pp.

Hrdlicka. A. 1920. Shovel-shaped teeth. Am. J. Phys. Anthropol. 3:429-465.

. 1931. Catalogue of human crania in the United States National Museum col-
lections: Pueblos, southern Utah Basket-Makers, Navaho. Proc. U.S. Nat. Mus.
78:1—95.

Irish, J. D. 1993. The Dentition of Africa. Ph.D. dissertation, Department of Anthropology,
Arizona State University, Tempe, AZ.

Irish, J. D., & C. G. Turner I. 1990. West African dental affinity 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. Sc. 101:47—63.

Kaminga, J., & R.S. V. Wright. 1988 The Upper Cave at Zhoukoudian and the origins of
the Mongoloids. J. Hum. Evol, 17:739-7067.

Kozintsev, A. G. 1988. Ethnic Cranioscopy. Racial Variation of Cranial Sutures. Nauka,
Leningrad (in Russian). 166 pp.

Kurten, B., & E. Anderson. 1980. Pleistocene Mammals of North America. Colambia
University Press, New York, NY. 442 pp.

SZ,

TEETH, NEEDLES, DOGS, AND SIBERIA

Kuzmin, Y. V. 1997. Chronology of Palaeolithic Siberia and the Russian Far East. Rez.
Archaeol, 18:33—39.

. 2000. Geoarchaeology of the Lower, Middle and Early Upper Paleolithic of
Siberia: A review of current evidence. Rex. Archaeol, 21:32-40.

Lahr, M. M. 1995. Patterns of modern human diversification: Implications for
Amerindian origins. Yrbk. Phys. 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.

Laughlin, W. S. 1980. Akeuts: Survivors of the Bering Land 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 1440 mm C. G. Mascie-Taylor & G. W. Lasker, eds., Bzologecal 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. Drozdovy, ed., Paleogeografia Kamennogo Reka. Korrelyashiya Prirodnk Sobtu
Arkeologicheskik, Kultyr Paleohta Severnot Axia i Sopredelnk Territori. Material
Meshdunarodnoi Konferensi, Kurtak. Krasnoyarsk.

Levin, M. G., & L. P. Potapoy, 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,

Mempey 7

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: Anthropologie, archeologie, ecologie, medecine. Boreales

74/77 whole volume. Suresnes, France. 336 pp.

Martin, P. S., & R. G. Nlein, eds. 1984. Ouaternary 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. 77:1—27.

Meltzer, D. J. 1993. Pleistocene peopling of the Americas. Evol Anthropol. 1:157—169.
Mochanoy, Y. A. 1977. The Earliest Stages of Settlement by Man of Northeast Asia. Nauka,
Novosibirsk (in Russian). 264 pp.

153)

TURNER

Moorrees, C. F. A. 1957. The Aleut Dentition: A Correlative Study of Dental Characteristics in
an Eiskimoid People. Harvard University Press, Cambridge, MA. 196 pp.

Morris, D. H. 1965. The Anthropological Utihty of Dental Morphology. Ph.D. dissertation,
Department of Anthropology, University of Arizona, Tucson, AZ.

Nazdratenko, Y. I., ed. 1998. Atlas Primorskogo Krat. Vladivostok. 49 pp. (in Russian)

Neves, W. A., J. E. Powell, & E. G. Ozolins. 1999. Extra-continental 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 Arizona.
Ph.D. dissertation, Department of Anthropology, Arizona State University, Tempe,
NZ

Nichols, J. 1998. The origin and dispersal of languages: Linguistic evidence. Pages
127-170 in N. G. Jablonski & L. C. Aiello, eds., The Origins and Diversification of
Language. Memoirs of the California Academy of Sciences No. 24, San Francisco,
GA,

Ossenberg, N. S. 1994. Origins and affinities of the native peoples of northwestern
North America: The evidence of cranial nonmetric traits. Pages 79-115 zm 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.

Ovodoy, N. D. 1998. The Ancient Dogs of Siberia. \CAZ Sympostaum on Domestic Dogs,
Aug, 23-28, Victoria, B.C., Canada.

Pedersen, P. O. 1949. The East Greenland Eskimo dentition. Meddelelser om Gronland
142:1-244.

Popov, A. N., T. A. Chikisheva, & E. G. Shpakova. 1997. Bossman Archaeological Culture in
Southern Primoria (On the Basis of the Boisman Il Muttilaaaayered Site). \nstitate 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. F, & W. A. Neves. 1999. Craniofacial morphology of the first Americans:
Pattern and process in the peopling of the New World. Yearbook of Physical
Anthropology 42:153-188.

Powell, J. BK, & J. C. Rose. 1999. Report on the osteological assessment of the
“Kennewick Man” skeleton (CENWW.97.Kennewick). Internet citation:
http://www.ct.nps.zov/aad/ kennewick/powell_rose.htm.

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

TEETH, NEEDLES, DOGS, AND SIBERIA

Richert, S. 1999. The Woodland Iroquois People of Southern Ontario: A Dental Assment of Their
Population Affinity. M.A. thesis, Department of Anthropology, Arizona State
University, Tempe, AZ.

Roller, K. 1992. Near Eastern Dental Variation Past and Present. M.A. thesis, Department of
Anthropology, Arizona State University, Tempe, AZ.

Ruhlen, M. 1976..A Guide to the Languages of the World. Languages Universals Project. Stanford
University, Stanford, CA. 356 pp.

. 1994. The Origin of Language: 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. Sa. U.S.A. 95:13994—13996.

Scott, E. B. 1998. Masillary Premolar Accessory Ridges (MxPAR): Worldwide Occurrence and
Unihty 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 Fanuhes and
Variation in Living Southwest Indians. Ph.D. dissertation, Department of Anthropology,
Arizona State University, Tempe, AZ.

. 1991. Continuity or replacement at the Uyak site: A physical anthropological

analysis of population relationships. Pages 1—56 zm D. E. Dumond & G. R. Scott, eds.,

The Uyak Site on Kodiak Island: Its Place in Alaskan Prehistory. University of Oregon

Anthropological Papers No. 44, Eugene, OR.

. 1994. Teeth and prehistory on Kodiak Island. Pages 67—74 7m T. L. Bray & T.
W. Nillion, eds., Reckoning with the Dead: The Larsen Bay Repatriation and the Smithsonian
Institution. Smithsonian Institution Press, Washington, D.C.

Scott, G. R., & C. G. Turner Hl. 1997. The Anthropology of Modern Human Teeth: Dental
Morphology and Its 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 Palaeohthic of
Eastern Europe and Northern Asia: Problems and Perspectives. Rassian Academy of
Sciences, Insitute of the History of Material Culture, Saint Petersburg. 143 pp.

Soffer, O. 1985. The Upper Paleolithic of the Central Russian Plain. Academic Press. Orlando.
539) (ys

. 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 Syumpostum. Institute of History, Philology and Philosophy, Siberian
Branch, U.S.S.R. Academy of Science, Novosibirsk.

155

TURNER

Stanford, D. 1999. Who were the first Americans? Pages 6—9 i C. Wood, ed., Annual
Report, National Museum of Natural History. Smithsonian Institution, Washington, D.C.

Steele, D. G., & J. F. Powell. 1992. Peopling of the Americas: Paleobiological evidence.
Hum. Biol. 64:303—330.

Straus, L. G. 2000. Solutrean settlement of North America? A review of reality. Am.
Antiquity 65:219-226.

Szathmary, E. J. 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 im

R. Kirk & E. Szathmary, eds., Out of Asia: Peopling the Americas and the Pacific. Journal

of Pacific History, Canberra.

. 1993. Genetics of aboriginal North Americans. Evol Anthropol. 1:202—220.

Tuohy, D. R., & A. Dansie. 1997. New information regarding early Holocene manifesta-
tions in the western Great Basin. Nevada Hast. Soc. Quart. 40:24—53.

Turner, C. G. H. 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 zm R. S. Vasilievsky, ed., Late Pleistocene and Early Holocene
Cultural Connections of Asta and America. USSR Academy of Sciences, Siberian Branch,
Novosibirsk.

. 1985. The dental search for Native American origins. Pages 31—78 mm R. Kirk

& E. Szathmary, eds., Out of Asta: Peopling the Americas and the Pacific. journal of Pacific

History, Canberra.

. 1987. Late Pleistocene and Holocene population history of East Asia based on
dental variation. Am. J. Phys. Anthropol. 73:305—321.
. 1990a. Major features of Sundadonty and Sinodonty, including suggestions

about East Asian microevolution, population history and Late Pleistocene relation-

ships with Australian Aboriginals. Am. J. Phys. Anthropol. 82:295-317.

. 1990b. Paleolithic teeth of the central Siberian Altai Mountains. Pages 239-243

in N. A. Ahmerova & N. M. Shakhmatova, eds., Chronostratigraphy of the Paleolithic in
Jorth, Central, East Asia and America. Papers for the International Symposium,

US.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 zm 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.

156

TEETH, NEEDLES, DOGS, AND SIBERIA

. 1992b. Sundadonty and Sinodonty in Japan: The dental basis for a dual origin
hypothesis for the peopling of the Japanese islands. Pages 95-112 7 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 iv T. Akazawa, K. Aoki, & T. Kimura, eds., The Evolution and
Dispersal of Modern Humans in Asia. Honkusen-Sha Pub. Co., Tokyo.

. 1994a. New dental anthropological observations relevant to the human popu-
lation system of the Greater Beringian Realm. Pages 97-106 7 W. W. Fitzhugh & V.
Chaussonnet eds., Avthropology of the North Pacific Rim. Smithsonian Institution Press,
Washington, D.C.

. 1994b. Relating Eurasian and Native American populations through dental
morphology. Pages 131-140 i 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 m 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 Homo 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 i A. P. Derevyanko ed., Plestocene
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. Phys. Anthropol. 31:303-310.
Turner, C. G. I, & D. E. Hawkey. 1991. World variation in Tome’s root. Am. J. Phys.
Anthropol. Suppl. 12:175.
. 1998. Whose Teeth Are These? Carabelli’s Trait. Pages 41—50 7 J. 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.

157)

TURNER

Turner, C. G. II, Y. Manabe, & D. E. Hawkey. 2000. The Zhoukoudian Upper Cave den-
tition. Acta Anthropologica Sinica 19:253—268.

Turner, C. G. II, & M. A. Markowitz. 1990. Dental discontinuity between Late
Pleistocene and recent Nubians. Peopling of Eurafrican-South Asian triangle I. Homo
45:32-41.

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 University dental
anthropology system. Pages 13—31 mm M. A. Kelley & C. S. Larsen eds., Advances in
Dental Anthropology. Wiley-Liss, New York, NY.

Vasil’ev, S. A. 1996. Late Palaeolithic of the Upper Yenisec (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. F. 1958. The dispersal factor in migrations and immigrations of American
Indians. Pages 454-461 mm R. H. Thompson ed., Migrations in New World Culture
History. University of Arizona Social Science Bulletin No. 27. Tucson, AZ.

Voevoda, M. I., A. G. Romashchenko, V. V. Sitnikova, E. O. Shulgina, & V. F. Kobsev.
2000. A comparison of mitochondrial DNA polymorphism in Pazyryks and modern
Eurasian populations. Archaeology, Ethnology ¢ 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 University, Tempe, AZ.

West, FE H., ed. 1996. American Beginnings: The Prehistory and Palaeoecology of Beringa.
University 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. Khaldeeva. 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
were identified as the Clovis people, specialized spear-hunters
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 (¢@2.,
Bednarik 1989; Bryan 1991; Dillehay 2000; Dixon 1999; Krieger 1964; Meltzer 1993;
Willey & Phillips 1958) have long claimed that the human colonization was begun as
eatly 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

159)

ROOSEVELT, DOUGLAS

Mesa (lanceolate point) 11,660£80(?); 10,240480
Putu (fluted point?) 11,470£500(?); 8,810%60

Bluefish Caves
Old Crow ca. 14,000(?)

27,000; 1,350%150

Broken Mammoth (Nenana) 11,770£220
Swan Point (Nenana) 11,660260

Walker Road
(Nenana)
11,300£120

MILL IRON
23,720 £220(?);
ANZICK 11,570 +170(?);
10,680 t50(?)@ @ 10,760 130(?
@— COLBY
11,200 +220(?);
UNION PACIFIC © 10,864 £141
DAISY CAVE MAMMOTH pent
10,650 +80 11,280 +350 11,200 +500;
ARLINGTON 10,750 +40
SPRINGS
CLOVIS oo
960 £80 | o> DOMEBO
aE a 11,300 #240(?) @ 11,210 +380;
11,040 +250;
LEHNER © MURRAY SPRINGS @ 10,980 +60
10,950 +40 10,880 £50 i
AUBREY

11,570 +70(?)

AND BROWN

VAIL
MeapowcrorT — 10.5004300)
ROCKSHELTER 0/580 250

19.100 +810(?)
eo

@ SHAWNEE MINNISINK
10,560 +250

CACTUS HILL
15,070 70(?);
10,920 +250

LITTLE SALT SPRING
12,030 +200(?)

LOS TAPIALES

10,710 +

170(2);

@ 9,860 +185(?)

MADDEN LAKE
TURRIALBA @ undated @ TAIMA TAIMA
undated e 33,000+ (2);
13,880 #120(?)
EL ABRA———
12,400 £160(?); TIBITO #40002)
11,210290 v vi
LAS VEGAS ©
10,840 +410 @ ON
‘ SITIO DO MEIO
TALARA 11,200 #115(2) © CAVERNA DA® 18,600 +600(?)
PEDRAPINTADA —_penra FURADA CHAO DO CABOCLO
GUITARRERO CAVE = 11,075#106 55 460 44 000(2) 11,000 +250
WEE!) i : i BREJO DA MADRE
DO DEUS
PIKIMACHAY @ PERNA @ 11,060 £90
20,200 +1,050(?) BERIEG) eye 10,5302110
QUEBRADA SACU Bs T0705, easter i ° Fee
11,105+ ~ a! 2):
= 10,740 £90 6,450 £150
QUEBRADA TACAHUAY aes Sea ntie
10,710 BO ge 12,070 e780):
Teeaetuy: T 11,440 22002);
10,240 +80 OLD CROW ALO
e ca. 27,000(?); LAPA VERMELHA =
5 1,350 150 11,680 +500(?) LAPA DO DRAGAO
wen cathy e ee 11,000 +300
1, 2s 12,770 £220(?
BLUEFISH © QUEREO ’ ? SANTANA DO RIACHO
BSiO Sed CAVES 11,100 #150 ALICE BOER {11,960 #250(?)
ca. 14,000(?) 14,200 #1,150(?)
BROKEN MAMMOTH US
WALKER ROAD® |11.7704220 g @)
CERRO LA CHINA®@ RS-I-68 11,555 +230(?)
11,300 £120 | SWAN POINT 11,150 +135
11,660 +60 RS-I-69 10,985 +100
eo
RS-I-66 10,810 #275
MONTE VERDE
33,370 £530(?):
ca. 12,500(?)
LOS TOLDOS

CUEVA DEL MEDIO

12,290 $180(2);

10,550 £120 @
FELL’S CAVE®.
@TRES ARROYOS
11,000 #170 "43550 #250(2):
10,600 +90

FIGURE 1. Map of possible Paleoindian sites throughout the Americas. Question marks indicate

questionable dates. Drawing by Joyce Pendola of Conde
History.)

160

Nast. (Roosevelt 2000a, courtesy of Natural

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

than ca. 12,000 years fail to withstand careful scrutiny. Furthermore, all well-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 ef a/ 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 lived by broad-
spectrum foraging of diverse modern biota, not by specialized hunting of now-extinct
megafauna (see Dillehay & 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 11,200
yt 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,2 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,

161

ROOSEVELT, DOUGLAS AND BROWN

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
forms.

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 picture in early New World culture histo-
ry and beyond, and we point to possible directions for future research.

Early Paleoindians in North America

The Clovis migration theory

In the original Clovis migration theory (Fagan 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 shortly before 12,000 years ago. They
went into the cool, upland grasslands of the continental interior, and there they preyed
on large mammals such as mammoth, long-horn bison, horse, and camel. Since
American animals were unaccustomed to human hunting, the highly efficient Clovis
speat-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 11,000 yr 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 geographical scenario held firm for a
long time.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

According to the theory, this lifestyle of terrestrial big-game hunting in open, tem-
perate habitats was typical 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 (¢.g., Anderson ef a/, 1996; Frison 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 (¢.g., Fiedel 1987:82—159,
1992:84-165; Lynch 1983:89).

Early Paleoindian cultures: Universal standards of vahdity

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 cultural 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 complexity 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

163

ROOSEVELT, DOUGLAS AND BROWN

and objects in sites (¢.g., Behrensmeyer & Hill 1980). New radiocarbon dating methods
have allowed taxonomically-identified biota and individual artifacts 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 occupations (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 e¢ a/ 1991; Toth
99D).

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 quantity 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 ef 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 stratigraphic contexts or on samples without documented pri-
mary cultural association. Cultures would be questionable 1f 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, reliability scrutiny needs to be even-handed. There has been a common
tendency for archaeologists to scrutinize the reliability of pre-Clovis sites or Clovis rivals,
but, as we describe next, current claims for the age of Clovis rely 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

focus.

164

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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 ef a/, 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 af 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 (¢.g., Gramley 1993;
McAvoy & McAvoy 1997; Di Peso 1965; Ranere & Cooke 1991).

According to the criteria of reliability, 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 solid-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 e¢ aZ 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 earlier 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 ef a/. 1998; Lewisville site, Stanford

165

ROOSEVELT, DOUGLAS AND BROWN

ee a

YEE: 5
dA Git Ae

TD (alan LY ee
rogey iy if} or

FIGURE 2A. The Lehner Clovis site, Arizona. Projectile points (Haury ef a/, 1959:17, figure 13, used

with permission).

1983), materials abundant in Cretaceous strata incised by the streams by which many
Clovis sites lie (Harbour 1975). Most of the others were run on tree charcoal, which by
its nature 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 reliability 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. Forty-six assays on 35 adequate-size samples of rigorously-cleaned, biolog-

166

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

MODERN ARROYO CHANNEL

SOUTH BANK OF “..
MAMMOTH KILL CREEK A eee D.,

A PROJECTILE POINTS

B OTHER TOOLS G-RED BASE CLAY
C-SANDY RED CLAY WITH CALICHE
f-GRAY CALICHIFIEO CLAY, SILT LENSES
9- COARSE SAND AND GRAVEL

¢ LIMIT OF EXCAVATION h-GRAVEL AND REDEPOSITED RED CLAY

i-FINE SAND
J-SAND GRADING TO SILT AND CLAY
k-BLACK SWAMP SOIL
1-GRAY CLAYEY SILT
M-LOOSE SILT

G PRESENT SURFACE Z'-MAMMOTH KILL CREEK EROSION
z=-MODERN EROSION

RADIOCARBON OATES
U. ARIZONA U. COPENHAGEN U. MICHIGAN

ANTEVS’
GEOLOGICAL ESTIMATES

Z3=MODERN ARROYO
STARTS CUTTING

LEHNER SITE
PROFILE

1880'S

Z?=ALTITHERMAL ARROYO
7000-5000 B.P

6356 +450
eh eacel ae
7133 +350 (A-33)
AV. 6789 + 450

k= PLUVIAL

72054450 (A-34)

Z':MAMMOTH KILL CREEK
HUNTS 13,000 8.P.
OR OLDER

7022+450 (A-32)
6330+ 450 (A-30)
10,900 + 450 (A-400)
12,000 + 450 (A-40b)

11,180 +140 (K-554) | 11,2904500 (M-8il)

FIGURE 2B. The Lehner Clovis site, Arizona. Stratigraphy and inconsistent radiocarbon dates
(Haury ef a/., 1959:25, figure 16, used with permission).

167

ROOSEVELT, DOUGLAS AND BROWN

ically identified, cultural carbon from documented stratigraphy in the five sites with doc-
umented Clovis lithics have produced no statistically-consistent dates earlier than the
period ca. 11,200 to ca. 10,800 yr B.P. (see Appendix 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 assertion is based on the inter-
calibration of only the unreliable Clovis dates (those run on unidentified carbon, carbon
from contaminated contexts, or samples without documented cultural or stratigraphic
association)‘, not the re/able dates, which calibrate at ca. 12,906 yr B.P.

Remarkably, the dates from the famous Clovis type site at Blackwater Draw Locality
1 fail to meet reliability 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; Haynes
et al. 1966; Hester ef a/, 1972; Damon ef af 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 1s present at the site

(Bowman 1990:26; Cotter 1938, 1939; Damon et a/ 1964; Hester et a/ 1972; Holliday

EL LLANO DIG |

COTTER 1936, 1937

rae

WENDORF & DITTERT

MAMMOTH
MAMMOTH

11,630 + 400 (A49i)

11,170 + 360 (A 481)
(1,040+ 500 (A 490)
(0,170 + 250 (A 488)
10,490 200 (A 492)
10,2504 320 (A 379-A 380)
10,490 900 (A 386)
9,890+ 290 (A 489)
8470+ 350 (A 512)

0-169 6,300 + 150 BP.
0-170 6230 +150 BP.
0-157 4,950 + 150 8.P. (Prob. 100 young— Haynes 966 pp 26)

FIGURE 3. Stratigraphy and radiocarbon dates on natural carbon at Clovis, Blackwater Draw, New
Mexico. (Hester ef a/ 1972:174, figure 130, used with permission).

168

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

1997:232—233; Roosevelt 1998a; Stafford et a/, 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 a/ 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 ef a/ 1991) in pri-
mary association with artifacts (Hester et a/ 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 e¢ a/ 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 a/ 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 stratigraphically-dispersed, biologically-unidentified carbon
subject to contamination from the peat, old wood, and calctum 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-yeatr-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

169

ROOSEVELT, DOUGLAS AND BROWN

site could, however, still 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 11,500 yr B.P. for
the Clovis 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 how did they 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 by 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 (Haynes 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

170

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Ficure 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. Courtesy of the munic-
ipality of Osaka.

7th

ROOSEVELT, DOUGLAS AND BROWN

reviewed for approximately 25 directly-dated skeletons of terminal Pleistocene and early
Holocene age (Lahr 1995; Powell 1993; Powell & Neves 1998; Powell & Rose 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 (¢bid.). 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 similarity with the east Asian pattern termed
Sinodonty, but a low percentage similarity with Late Pleistocene and Holocene
Europeans’ teeth (Turner, this volume). Mitochondrial DNA “X-factors” claimed to link
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 actoss 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 ef a/.
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

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

entirely unglaciated, allowing easy passage of migrants with watercraft (Dixon
1999:19-43; Dixon ef a/ 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 teliance on marine food resources (Chatters 1998; Dixon ef a/ 1997; Johnson ef 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 ef a/.
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 & Phillips 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
capacity 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 earliest 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 ef
al. 1968; Simpson 1978), never produced early stone specimens that archaeologists could

LIS

ROOSEVELT, DOUGLAS AND BROWN

agree were human-produced. Furthermore, it lacks other secure hallmarks of human
activity: 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 by humans
when fresh (Bonnichsen 1978; Harington ef a/ 1975). However, follow-up dating showed
that the undoubted cultural materials at the site were not even Pleistocene in age (Taylor
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 a/ 1985). At Pendejo Cave, New Mexico, burned ani-
mal bones and charcoal supposed to represent human activity (MacNeish 1991, 1992;
MacNeish e¢ a/ 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-
tia test outlined above: statistical consistency, biological identification, secure cultural
association, and replicability among sites. Some researchers continue to espouse such
sites, without addressing the specific evidence problems. But there still is absolutely no
good evidence in North America for an entry of humans before a. 12,000. All human
remains are anatomically modern, and all valid early Paleoindian cultures have the full
range of basic lithic manufacturing technology of Upper Paleolithic cultures.

Whatever the ecological or social conditions that encouraged people to venture 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 reliability.

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 years
old, but this interpretation belies its radiocarbon date statistics and lithics, both of which
indicate late, not early Paleoindian, age. The only two pre-Clovis Mesa dates are from

174

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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 catbon from prehuman subsoil, not on cultural carbon associated with artifacts
(Reanier 1996). Similarly, the bone dates from Bluefish caves, northern Alaska (Morlan
& Cing 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 cw. 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; Prison ef a/ 1996). Stratigraphically and cultur-
ally, this is a single occupation bison-kill 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. Mill Iron’s actual age, like 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 ef a/ 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-

ROOSEVELT, DOUGLAS AND BROWN

FIGURE 5. Inconsistent radiocarbon dates from Mesa, Alaska (Kunz & Reanier 1994, used with per-
mission).

Locality Date SE Material Sample ID

B 11,660 80 hearth charcoal; Beta-55286 Split sample this one was run
tools associated w/57430; does not overlap at 2

standard deviations

B 11,190 70 hearth charcoal; Beta-57430 Split sample this one was run

tools associated w/55286; does not overlap at 2

standard deviations
Saddle 10,980 280 hearth charcoal; Beta-50429 Split sample this one was run

tools associated conventionally, 50430; does
not overlap at 2 standard
deviations
B 10,240 80 hearth charcoal; Beta-55283

tools associated

A 10,090 85 hearth charcoal; Beta-50428
tools associated

Saddle 10,070 60 hearth charcoal; Beta-69898 Not in Science
flakes associated

B 10,060 70 hearth charcoal; Beta-52606
tools associated

B 10,050 90 hearth charcoal; Beta-69900 Not in Science
flakes associated

B 10,000 80 hearth charcoal; Beta-55285
tools associated

Saddle 9,990 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

Saddle 9,730 80 soil charcoal; Beta-36805
flakes associated

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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 vicinity.

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 minority of the dates from their levels, in
which the majority 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” lithics and insignificantly distant stratigraphically
(McAvoy & McAvoy 1997:169-170, table 6.3). By the dating standards, such cultures are
not currently credible contenders for Paleoindian ancestors.

A New Paleoindian Ancestor: Nenana, Alaska

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 (Godse a al, IODI3 Islonredses a i, NDS, WV; Sdesiaese 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 ef a/ 1991, figure 3, left; Hoffecker ef a/ 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:
mostly 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-

ROOSEVELT, DOUGLAS AND BROWN

Radiocarbon
dates

Es
gs 0
2040 + 65 S
2815 + 180 ‘S)
4525 + 90 fe
4545 + 90
4540 + 90 Ib
4690 + 110 50
D
7200 + 205 eas
7600 + 140 ni |e
7700 + 80 i
cea
e 100
9310 + 165 w
=
9690 + 960 if
10,270 + 110
10,790 + 230
10,290 + 70 B {150
11,040 + 80
41,060 + 90 iva _E
11,040 + 260 ee ive{
11,280 + 190 Wena Geeenee es VN Leyes
41,420 + 70
11,500 + 80 AAA Sod Zone
Pobre es CI
(ZZ B2 Soil Horizon
11,770 + 210 GF Loess w/clay bands
41,770 + 220

NJ B3 Soil Horizon
Ss Loess
22 Layers

mite

— oe

FiGuRE 6. Nenana culture, Alaska. Top: Stratigraphy and radiocarbon dates at Broken Mammoth
site (West 1996:314-315, figure 6—4, 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-
sion).

178

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

tified gave significantly earlier dates than the other biota, indicating use of fossil tvory.
Nenana bifactal 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 e¢ a/ 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 Latin America
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
ot, 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. 11,200 years, and all have some bifacial artifacts, not exclusively unifactal
artifacts.

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-zame 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-

ROOSEVELT, DOUGLAS AND BROWN

gular forms found in terminal Pleistocene Paleoindian sites in Latin America are mor-
phologically different from the triangular points of Archaic, post-Pleistocene Indians in
North America. The North American Archaic points at issue are typically side-notched
for attachment (Anderson e¢ a/ 1996:10, figure 1.2; Frison 1991:79—109), while the early
South American points (illustrated in figure 10) never show this form. In this important
hafting feature, 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 Northwestern 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
lithic tool representations that establish their forms as fluted, parallel-sided “Clovis”
points, as some assert (Ranere & Cooke 1991; Gruhn & Bryan 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 telling. Northern Mexican sites
classified as Clovis simply amount to finds of indeterminate, unassociated bifaces (¢.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 directly (Taylor 1992,
table 25.5). At Los Taptales, 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,

180

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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 parallel-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 align with virtually any Paleoindian or Early
Archaic points in North America; the waisted shapes have no relation to Clovis forms
and could only align, if at all, with North American forms that are late Paleoindian or
Holocene (¢.g., Frison 1991:62—101, figures 2.28, 2.33, 2.37; Anderson ef a/. 1996:10, fig-
ure 1.2). Only the strong faith in the reality of a hemispheric Clovis horizon and the
tools’ lack of contextual 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 Isthmus in Colombia and Ecuador, after decades of research there ts 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
eatliest 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 concte-
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 ef al. 1972; Hutt e¢ al. 1976) were once widely assumed to represent pre-Clovis,
pte-projectile-point foragers, but some archaeologists now consider the evidence equiv-
ocal (Ardila 1991:276—278; Ardila & Politis 1989:18-22; Roosevelt ef a/ 1996:383). The
unifacial industry was an inadequate artifact sample without plot locations, and the
chronology was based on nothing more than two statistically discordant dates on bio-
logically-unidentified materials not recorded in secure primary association with the lithics
(Hurt e¢ a/, 1976:2, 5, 8, 12, and figure 4). The El Abra biota was a mixture of many mod-
ern and few extinct species (ébid.). Although another Bogota area site, Tibito, has been

181

ROOSEVELT, DOUGLAS AND BROWN

identified as a pre-projectile-point megafaunal 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 der Hammen 1977; Nieuwenhuis 1998),
making the existence of a pre-projectile point complex questionable on grounds of sam-
ple statistics.

Points from northern Venezuela termed Clovis fluted points (Oliver & Alexander
1990: figure 20 and following), are, like the Isthmian points, undated bifaces distinct
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 megafaunal
kill site of the El Jobo culture (Bryan e¢ a/ 1978a; Dixon 1999:98—99; Ochsenius &
Gruhn 1979). However, the site does not meet the criteria of reliability on several counts
(Ardila 1991:274-275; Haynes 1974: 378; Lynch 1974:356; Roosevelt e¢ a/ 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 many
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 (¢.g¢., Bryan 1978a; Ochsentus & Gruhn 1979:10; Dillehay 2000:128—133; Dixon
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

182

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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.). All 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
pte-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 Andes

In the Peruvian Andes, on the supposed highland plains route of the Clovis big-
gamehunters, no finds of fluted points, no megafauna kill 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 ef a/ 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 e¢ a/ 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 lithics 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 e¢ a/ 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.P. (Lynch 1980; Lynch ef a 1985; MacNeish ef a/ 1981a, 1981b; Rick
1980). The earliest Paleoindian points in these sites are triangular, subtriangular, and

183

ROOSEVELT, DOUGLAS AND BROWN

sometimes vaguely stemmed (Figure 7a). They are not fluted, parallel-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 ef a/. 1998; Keefer et a/ 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 Tacahuay, south coast, Peru (Keefer e¢

al. 1998:1834, figure 3, used with permission).

184

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Stratum UE Var BLP Cal. yr B.P. Lab no. Component
Sector |
|-2-D, level 3b 10,274 + 125 12,265- 11,663 1943 TP
1970, layer 4 10,200 + 140 12,175-11,124 [(5), p. 45] TP
I-2-B, level 4c 11,088 + 220 13,183- 12,749 2024 TP
|-2-D, level 4c 11,105 + 260 13,240-12,728 1942 TP
Sector II
ll-1-D, level 1b M 10,190 + 220 12,271—11,008 1957 TP
Il-1-C, element II-5bii 9,850 + 170 11,199-10,890 1956 TP
Il-1-D, element II-5bi 10,475 + 125 12,502- 12,144 1936 TP
ll-I-D, level 2c 10,700 + 300 12,885- 12,234 1940 TP
I-I-D, level 2c2 10,600 = 135 12,638-12,312 1939 TP
Il-1-D, level 2c3 10,560 + 125 12,590-12,271 1938 TP
Il-1-D, level 2c4 10,725 + 175 12,794-12,419 1937 TP
1992, level 3 10,770 + 130 12,797-12,515 1702 TP
Sector IV—Unit IV-1-C
IV-1-C, level 2c 10,507 = 125 12,536-12,195 2025 TP

FIGURE 8. Radiocarbon dates from Quebrada Jaguay, south coast, Peru (Sandweiss ef 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 Patjan lithic complex is char-
acterized by ubiquitous stemmed, triangular points (Figure 9, but no fluted points at all).
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 Patjan 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-
tute, 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 lithic industries similar to Clovis. As elsewhere, no secute evidence for
pre-Clovis or pre-projectile point cultures has yet been found.

185

ROOSEVELT, DOUGLAS AND BROWN

The site of Monte Verde near the Pacific coast of south central Chile 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 ¢/ a/, 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 nature 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 manufacture 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 (¢.g., Nunez 1992), not Pleistocene
dates.

nan rail
met ll
\ ANN MTA
HH,

Pte

FIGURE 9. Stemmed projectile points of the Paijan culture, north coast, Peru. (Chauchat 1988:55,
figure 2.6, used with permission).

186

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

There are also problems with the dating itself. Two a. 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
catbon-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 teview 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 e¢ a/, 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 ¢e¢ a/. 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 ef a/ 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 e7 a/.
1988; Borreto & McEwan 1997:40, figure 23; Flegenheimer 1987; Miotta 1999; Politis
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-

187

ROOSEVELT, DOUGLAS AND BROWN

less birds. The rare extinct mammal bones associated with the cultural charcoal and arti-
facts in Fell sites have given discordant dates considerably earlier than the cultural dates;
their condition, with marks of predators, suggests a mixture of natural and cultural
materials (Ardila & Politis 1989; Borrero 1996; Borrero & McEwan 1997:34—44; Nami
1987), an indication that megafauna were probably used as fossil bone, not 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 e¢ a/ 1988; Politis 1991), which is comparable to the time
span of the Clovis-Folsom Paleoindian sequence in North America. In the vicinity of
Los Toldos on the Patagonian plateau, archaeologists have uncovered a poorly-known
preceramic culture characterized by rare triangular projectile points as well as fishtail
points, and a rich rock-painting tradition (Cardich 1978; Schobinger 1999). Existence of
an earlier culture with unifacial, edge-trimmed 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 a/ 1973) lacks secure cultural
association.

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 lithic 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 lithic tradition and subsistence orientation, than to the broad-spectrum hunting
and gathering Nenana culture, which predates them.

Eastern South America

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 valid examples of pre-Clovis, pre-
projectile-point cultures, either. Instead, excavations by several independent researchers
have established the existence of an early Paleoindian occupation by rock-painting,
broad-spectrum foragers who made triangular and often stemmed projectile-points, as
well as other tools.

188

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Southeastern Brazil and Uruguay

Numerous pte-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 e¢ a/ 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 Purada (Split Rock) in Piaui, the
flaked stones, red-stained rock clusters, and charcoal radiocarbon-dated as early as 30,000
yt B.P. (Guidon & Delibrias 1986), appear to be products of natural processes (Lynch
1990; Meltzer et a/ 1994; Roosevelt e¢ a/, 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. All the usual signs of prehistoric human habitation, such as fragments of burned
bones and carbonized food plants, were absent from the early strata.

Such claims of exaggerated antiquity 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 agriculture since the European conquest. The few remains of megafauna at
some sites ate pieces reworked from limestone geological substrata. No use of megafau-
na as human prey has been verified.

189

ROOSEVELT, DOUGLAS AND BROWN

There are approximately five Hast Brazilian skeletons with direct terminal Pleistocene
radiocarbon dates or sealed, well-dated stratigraphic contexts. They have small, high
faces, narrow noses, long, narrow heads, and Sundadont dentition. As mentioned above,
the particular female cranium from Lapa Vermelha site, claimed by Neves to be a
Negroid individual 11,500 yr B.P. old, has not been directly dated and not 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 & Pucciarelli 1991; Neves ef a/ 1996).

The Lower Amazon

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 e¢ a/Z 1991;
Roosevelt e¢ a/ 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 teseatchers had also assumed that lowland South America would 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 radiometrically dated to Late Pleistocene age in the Amazon have spec-
tra and stable isotope ratios typical of closed-canopy tropical rainforests, not of savan-
nas (Absy 1979; van der Hammen & Absy 1994; Athens & Ward 1999; Colinvaux ef a/
1994; Haberle & Maslin 1999; Roosevelt 2000c). These prehistoric spectra have small-to-
moderate amounts of grass pollen, abundant tropical forest pollen, and stable carbon
isotopes from -20 to -37 per mil. They showed neither the typical savanna pollen spec-

tra, which have greater than 99% grass pollen, nor savanna carbon isotope ratios, which

190

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Site near Siteon |
Manaus 60° lreng oe

50°
&= Site in
Monte Alegre area

|
in

80°

ATLANTIC
OCEAN

Tapajos sites
“

Curua River

10°
80°

PACIFIC
OCEAN

2 @©
NEOGENE-QUATERNARY ALLUVIUM 5 _ Brasilia
Lake Titicaca Mh <i 50'
PRE-TERTIARY SEDIMENTS = 60°
0 500 1000
EE
CRYSTALLINE BASEMENT Km

Ficure 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 Emilio, Goeldi 1986:115, 117, used with permission and Roosevelt ef a/.
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.

Monte 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

191

ROOSEVELT, DOUGLAS 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, University 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 century by Alfred Russell Wallace (1889) and

others, the painted caves and shelters had not been excavated for evidence of human
occupation.

192

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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-
ters.

eos ee es athe: iss . f ee Wiremr tM z 2

FIGURE 12. Rock paintings and spilled paint, Serra da Lua, Monte Alegre, Para, Brazil.

193)

ROOSEVELT, DOUGLAS AND BROWN

Legend

Cave Wall

Drip Channel
Drip Line
Contour Line
Site Datum
Boulder

Auger Hole
Excavation Unit
Paint Samples

2 Sa:
Meters

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.

194

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

LLL AL

RAK aerial

~ —~ % tA SLI
ROSY st 16b re, Cate DL SLI ISSN LON LOLS
~ ee KK KON

ee Se RRR LR

% eee me
CORA
ODS?
RRR Bes

; 29 =a] Not eee AY
= : _ <« WATE
We rcs Ne K 10,210 + 60 9290
aaa 2 4 SSSA 0360 + 50 9290 10,230+ 60 9142
EARLY EARLY EARLY
%@ rock = charcoal 10,110+609272 10,260+70 9245 10,000 + 60 9294 10,330 + 70 9204
ap 10,120 709272 10,290 +70 9272 10,000 + 60 9296 10,370 +70 9204
& lithic © root 10,180 + 609246 10,290: 70 9272 10,230 + 60 9296 10,380 + 60 9204
& shell . seed 10,190 509272 10,301 + 70 9272 10,390 + 60 9296 10,480 + 70 9204
10,190 + 60.9272 10,3304 70 2272 10,450 60 9294 10,510 + 60 9204
» leaf & palm 10,210 +70 9272 10,360 +60 9272 10,470 +70 9296 10,570 +70 9204
; 10,210 + 60.9245. 10,370- 60 9246
@ stick & wood 10,220:+ 60.9272 10,420+70 9245
10,250+70 9245 NGA/eanty
10,260+ 609174 10,320+709174
10,280+709174 10,330+409178
10,290+809174 10,340+709174
INTIALB 10,300+.609174 10,4504 609174 INTIAL B
1025070 9274 10,275 + 285 8345
10,350-+70 9274 10,275 + 285 8345
10,390+70 9274 10,305 + 275 8346
10,4104 70 9274 10,3924 78 8314A
10,470: 709274 10,4504 60 8231E
10,5604 60 8231E
INTIALA
10,875 + 295 83148,
10,905 + 295 83148
WMl0#310821W «BB
11,145 + 135 83148

FiGureE 14. Caverna da Pedra Pintada stratigraphy. A. During excavation. B. Drawing with radio-
catbon dates. (Original drawing, Roosevelt e¢ a/ 1996:375, figure 5.)

195

ROOSEVELT, DOUGLAS AND BROWN

More than 30,000 lithic artifacts were recovered from the Paleoindian layers
(Roosevelt e¢ a/ 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 significantly through the five stratigraphically-distinct, phases of occupation
(Roosevelt e¢ a/, 1996:376—377, table 1). Chalcedony predominated in the upper prece-
ramic deposit, reaching 90% and higher in the highest stratum, 16A, but quartz crystal
predominated in the lower part, reaching over 70% in the earliest stratum, 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 cultural deposit was rich in wood charcoal, burned fruit and legume
pits and seeds, and fragments of burnt bone and shell, types of remains that were com-
pletely absent from the sterile subsoil below the archaeological deposit. Fifty-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 wood 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 (IL) dates, and three sediment samples were dated by
optically stimulated luminescence (OSL) (Michab ef a/ 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 between ca. 11,200 and 10,000 yr B.P.
(Figure 17) (Roosevelt ef a/ 1996, table 3), placing the preceramic occupation in terminal
Pleistocene times, exactly contemporary with the North American Paleoindian sequence
from Clovis through Plainview (Figure 18 and Appendix 1) (Taylor ef a/ 1996). Statistical

196

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Globular fibro-radial chalcedony 100%
Cryptocrystalline chalcedony
g : 90%
Hematite plasma in flux form
Aggregates of quartz crystals
Goethite plasma 80%
Red hematite plasma
0, —
0 0.2mm 10%
aH
60% |}
Cryptocrystalline chalcedony
yptocry: 50%
Globular fibro-radial chalcedony
40% |-Y
YY
Aggregates of quartz crystals 30%
0 0,04mm
ae 20%
10%
Partly weathered mica
Hematite after weathered carbonate 0%
()

Muscovite

: : Initial A Initial A/B Initial Initial B/Early Early Early/Middle Middle Middle/Late Late
Microcrystalline matrix

(n=1282) (n=580) (n=758) (n=52) (n=6270) (n=894) (n=6264) (n=150) — (n=13194)
Quartz —

Unweathered carbonate

0 0.2mm

Muscovite

Microcrystalline matrix

0 0.2mm

Carbonate (Dolomite)
Microcrystalline matrix

Quartz

bs »») Fae
GS
ee SO)
ss.

FIGURE 15. Lithic artifacts from Caverna da Pedra Pintada. A. Thin sections of chalcedony vati-
eties. B. Histograms of raw material frequencies. C. Chalcedony debitage during excavation. D.
Chalcedony point during excavation. E. Quartz crystal point during excavation. F. Drawings of lithic
artifacts. By Ruth Sliva of Desert Research.

IDy

ROOSEVELT, DOUGLAS AND BROWN

Hematite plasma

— Quartz

Goethite plasma

0 0.2mm

(CAV ILY,
Dark red hematite plasma

Zircon

Gibbsite

R ——— Goethite plasma

IE SRE Gibbsite
- aN pene Red cryptocrystalline hematite
N —— Gypsum

Goethite plasma
Cavity

Chalcedony

FIGURE 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 ratio 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.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

tests established that the five phases of this Paleoindian occupation had weighted-mean
radiocarbon ages significantly different from each other, in correct stratigraphic order
(Roosevelt e¢ a/. 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 neatly 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 statistically 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

Jo ee <2

"TL and OSL weighted average; outlier dropped

|13,180+509 __ ee
|

we a he SAS 9 SL Ee a Oe ee eee ee eee

| Initial A radiocarbon dates

[Midpoint of average (13,088 cal) |
13,143 cal
113,135 cal
12,950 cal

12,920 cal
|

112,172 cal

12,745 cal
12,498 cal

|12,531 cal

12,333 cal

112,241 cal

12,016 cal Be a)

| 12,052 cal =

[58 SSS ose | meen ees eee eed a ee eee eee (ea | ef ee | | ee ee ed

14000BC 12000BC 10000BC 8000BC
Calendar date

FiGuRE 17. Graph of initial calibrated radiocarbon, TL, and OSL dates from Caverna da Pedra
Pintada.

199)

ROOSEVELT, DOUGLAS AND BROWN

Pedra Pintada Initial A dates

Radiocarbon Dates

Pedra Pintada Initial A dates GX17414 108754295

= GX17407 10905+295
GX17406 11110+310
GX17413 111454135

Luminescence Dates (no calibration)

12536 BP +4125
15330 BP +900

13106 BP +1628
ae ay ai 12491 BP +1409

Clovis dates

Clovis radiocarbon dates

i Ga ese i, Teas ae B Blackwater Draw 11300+240
6000 BC 14000 BC 12000 BC 10000 BC 8000 BC 6000 BC eae FOGEGRAG

Calendar date A Murray Springs 10880+50
Dent 10750440
Domebo original 11210+380
Domebo Stafford 11040+250
Domebo wood 10930+60
Lange/Ferguson carbon 11140+140
Lange/Ferguson bone 107304530
Anzick 10680+50
UP Mammoth 112804350
Aubrey 11570+£70 B

Ficure 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 significantly 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 yr 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-
tural 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 naturally occurring
objects, and have too large errors to be distinct from those of Initial B; they 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 yr B.P., have more dated radiocarbon samples, dates with

200

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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
not 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, unlike 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 hypothesis 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,

FiGurE 19. Paleoindian biological remains from Caverna da Pedra Pintada. A. Fragmentary sphe-
notic bone of large fish, interior and exterior. Proy. 9243. B. Carbonized legume seeds, Hymeneaea c.f.
parvifolia. Prov. 9290, 9272. C. Otolith from large fish. Prov. 8347.

201

ROOSEVELT, DOUGLAS AND BROWN

numerous identified taxa (Figure 19), including 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 typical 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 huntets.
Conclusions

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 archaeologically 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 especially need to know more about the culture and lifestyle 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 culture. And Central America, which must have been

202

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

the major conduit for the groups that populated South America, still has no identified
initial cultures. To complete the earliest hemispheric culture history, it is a priority to
make a concerted effort to indentify relevant Quaternary geological features 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 dit-
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 peopling 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 temperate
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 Paleolithic,
when our genus was emerging (Klein 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 nature will need to be revised.

203

ROOSEVELT, DOUGLAS AND BROWN

Abstract

Early in research on the peopling of the Americas, Clovis big-game hunters of the northern high
plains seemed sure to be the first people who colonized the Americas. Anthropologists beheved that Clovis
hunters had rapidly colonized the hemisphere from Alaska to Patagonia in only a thousand years from
12,000 to 11,000 years ago. Thought to have followed the big herd game through their grazing lands
— the North American high plains and the cool, arid uplands of Central and South America — they
avoided the seacoasts on both sides of the continent and also the game-poor, disease-ridden tropical rain-
Jorests 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 years 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. Re-evaluation of evidence also has pushed Clovis
younger than had been thought, no earher than about 11,100 to 10,500 yr B.P. Alaska still has the
earliest solidly dated and investigated cultures in a range from 11,800 to 11,000 yr 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 horizon. Even supposed Clovis descendants at the tip of
South America turn out to be the same age as Clovis, and their subsistence was small-game, not large-
game, hunting. New sites reveal the existence of coastal fishers and tropical rainforest foragers as old as
Clovis hunters and culturally distinct from them. This new evidence on the colonization reveals the evo-
lution of more diverse Ice-Age ecological and cultural adaptations than previously suspected, raising new

questions for research on human origins.
Endnotes

' Dates in our article are given in uncalibrated radiocarbon years before the present
(1950) unless otherwise noted.

2 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., Frison 1991:70, figure 2.34; Morrow
& Morrow 1999; Borrero & McEwan 1997:40, figure 23). The nature of the lithics in val-
idated Clovis sites also has been confused in the literature by the characterization of very

204

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

finely flaked points from undated caches to the Clovis culture on the assumption that the
finest points have to be Clovis (Frison & Bradley 1999). In many cases, however, the
undated points in question actually show flaking customs found in late Paleoindian com-
plexes (¢.g., Frison & Stanford 1982), not in documented and dated Clovis assemblages
(Hester et a/, 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.
Furthermore, parallel-sided points continue long after the time of Clovis in many North
American regions (Anderson ef a/. 1996:10, figure 1.2; Frison 1991; Morse ef a/, 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.

5 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 yeats 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.

4 For 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 e¢ a/, 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 re/-
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-
pteters from inferring a pre-Clovis phase (¢.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.

205

ROOSEVELT, DOUGLAS AND BROWN

Literature Cited

Absy, M. L. 1979. Pahynological Study of Holocene Sediments in the Amazon Basin. Ph.D. thesis,
University of Amsterdam, Netherlands.

Ackerman, R. E. 1996. Bluefish Caves. Pages 505-510 m EF. H. West, ed., American
Beginnings: The Prehistory and Palaeoecology of Beringia. University of Chicago Press,
Chicago, IL.

Adovasio, J. M., J. Donahue, IK. Cushman, R. C. Carlisle, R. Stuckenrath, J. D. Gunn, &
W. C. Johnson. 1983. Evidence from Meadowcroft Rockshelter. Pages 163-190 7m R.
Shutler, ed., Early Man in the New 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 zm 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. Sczence-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 Fraser University, Burnaby, Canada.

Anderson, D. G., L. D. O’Steen, & K. E. Sassaman. 1996. Environmental and chrono-
logical considerations. Pages 3-57 mm D. G. Anderson & K. E. Sassaman, eds., The
Paleoindian and Early Archaic Southeast. University of Alabama, Tuscaloosa, AL.

Ardila, G. I. 1991. The peopling of northern South America. Pages 261—282 7 R.
Bonnichsen & K. L. Trunmire, eds., C/ovis: 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. Bolin 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. Scence
250:1388—1390.

Beaton, J. M. 1991. Colonizing continents: Some problems from Australia and the
Americas. Pages 209-230 im T. D. Dillehay & D. J. Meltzer, eds., The Ferst Americans:
Search and Research. CRC Press, Boca Raton, FL.

Bednarik, R. G. 1989. Pleistocene settlement of South America. Antiquity 63:101—-111.

Begley, S., & A. Murr. 1999. The first Americans. Newsweek April 16:50—57.

206

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Behrensmeyer, A. K., & A. P. Hill, eds. 1980. Foss¢ds in the Making: Vertebrate Taphonomy and
Paleoecology. University of Chicago Press, Chicago, IL. 338 pp.

Bell, R. E. 2000. Archaeological Excavations at the Site of El 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 pleistoceno medio. Paper presented at
the Piaui Congress, Serra do Capybara, Brazil.

JBelwea®, IMG IML C., CG, ke eevee, |, IDeimon, Je, Zullo, ce G, lowe, iY.
Thermoluminescence dating of burnt cherts from the Alice Boer site, Brazil. Pages
203-213 im A. L. Bryan, ed., New Evidence for the Pleistocene Peopling of the Americas.
Center for the Study of Early Man, Orono, ME.

Binford, L. R. 1968. Post-Pleistocene adaptations. Pages 313-341 mm S. R. Binford & L.
R. Binford, eds., New 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, IA. 246 pp.

Boldurian, A. T., & J. L Cotter. 1999. Clovis Revisited: New Perspectives on Paleoindian
Adaptations from Blackwater Draw, New Mexico. University Museum, University of
Pennsylvania, Philadelphia, PA. 145 pp.

Bonnefille, R. 1995. A reassessment of the Plio-Pleistocene pollen records of East
Africa. Pages 299-310 zm E. S. Vrba, G. H. Denton, T. C. Partidge, L. H. Burkle, eds.,
Paleochimate and Evolution with Emphasis on Human Origins. Yale University Press, New
Flavenu@ir

Bonnichsen, R. 1978. Critical arguments for Pleistocene artifacts from the Old Crow
Basin, Yukon: A preliminary statement. Pages 102-118 7 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 Sipaliwini archaeological complex of Surinam: A summary. Neew
West-Indische Gids 54:94-107.

Booth, W. 1999. Early migrants may have come by land and sea. The Los 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.

207

ROOSEVELT, DOUGLAS AND BROWN

Bowman, S. 1990. Radiocarbon Dating. University of California Press, Berkeley, and British
Museum, London. 62 pp.

Bradley, B. A., & G. C. Frison. 1996. Flaked-stone and worked-bone artifacts from the
Mill Iron site. Pages 43—70 7 G. C. Prison, ed., The Mill Iron Site. University of New
Mexico, Albuquerque, NM.

Bryan, A. L. 1978. The contribution of J. M. Cruxent to the study of the Paleoindian
Problem in the New World. Pages 63-76 im E. Wagner & A. Zucchi, eds., Unidad y
| aniedad: Ensayos Antropologicos en Homenajo a Jose M. Cruxent. Ediciones C.E.A. —TVIC,
Caracas, Venezuela.

. 1986. New Evidence for the Pleistocene Peopling of the Americas. Center for the Study

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 zm R. Bonnichsen &
K. L. Trunmire, eds., Clovis: Origins and Adaptations. Center for the Study of the first
Americans. Oregon State University, Corvallis, OR.

Bryan, A. L., ed. 1978. Early Man in America from a Circum-Pacfic Perspective. (Occasional
Papers of the Department of Anthropology, University of Alberta). Archaeological
Researchers International, Edmonton, Canada. 327 pp.

Bryan, A. L., & R. Gruhn. 1979. The radiocarbon dates of Taima-Taima. Pages 53-58 zm
C. Ochsentus & R. Gruhn, eds., Tasma-Vaima: A Late Pleistocene Paleo-Indian Kill Site in
Northermost South America. CIPICS/ South American Quaternary Documentation
Program.

Bryan, A. L., M. Casamiquela, J. M. Cruxent, R. Gruhn, & C. Ochsenius. 1978. An El
Jobo mastodon kill at Taima-Taima, Venezuela. Sczence 200:1275—1277.

Cardich, A. 1978. Recent excavations at Lauricocha (Central Andes) and Los Toldos
(Patagonia). Pages 296-300 mm A. L. Bryan, ed., Early Man in America from a Circum-
Pacific Perspective. Occasional Papers No. 1, Department of Anthropology, University
of Alberta, Edmonton, Canada.

Cardich, A., L. Cardich, & A. Hajduk. 1973. Secuencia arqueologica y cronologia radio-
carbonico de la cueva 3 de Los Toldos. Re/aciones de la Sociedad Argentina de Antropologia
7:85-123.

Chatters, J. C. 1998. Personal communication to A. C. Roosevelt.

Chatters, J. C., W. A. Neves, & M. Blum. 1999. The Kennewick Man: A first multivariate
analysis. Curr. Res. Pleistocene 16:87—90.

Chauchat, C. 1988. Early hunter-gatherers on the Peruvian Coast. Pages 41-66 mm R.
Keatinge, ed., Peruvian Prehistory. Cambridge University Press, Cambridge, MA.

208

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

. 1992 Prehistoire de la Cote Nord du Perou: Le Payanien de Cupisnique. Centre
National de la Rechrche Scientifique Editions, Bordeaux. 391 pp.

Clark, D. W. 1991. The northwestern (Alaska-Yukon) fluted points. Pages 35-48 am R.
Bonnichsen & K. L. Turnmite, eds. Clovis: Orgins and Adaptations. Center for the
Study of the First Americans, Department of Anthropology, Oregon State
University, Corvallis, OR.

Colinvaux, P. A., P. E. de Oliveira, P. E. Moreno, M. C. Miller, & M. B. Bush. 1996. A
long pollen record from lowland Amazonia: Forest and cooling in glacial times. Sczence
274:85-88.

Correal Urrego, G. 1981. Exidenaas Culturales y Megafauna Pleistocenica en Colombra.
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. Sczence 175:1106—1108.

Correal Urrego, G., & T. van der Hammen. 1977. Investigaciones Arqueologicas en los Abrigos
Rocosos del Tequendama. Bibliotheca 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. Sc. 90:113-117.

Damon, P. E. , & A. Long, 1962. Arizona radiocarbon dates HI. 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,
Canada.

de Lumley, H., de Lumley, M.-A., M. M. C. Beltrao, Y. Yokayama, J. Labeyrie, J. Danon,
G. Delibrias, C. Falgueres, & J. L. Bischoff. 1987. Presence d’outils tailles associes a
une faune quaternaite datee du Pleistocene moyen dans la Toca da Esperanca, Region
de Central, Etat de Bahia. L’Anthropologie 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.

209

ROOSEVELT, DOUGLAS AND BROWN

Dikoy, N. N. 1996. The Ushki Sites: Kamchatka Peninsula. Pages 244—250 in EH. West,
ed., American Beginnings: Prehistory and Palaeoecology of Beringia. University of Chicago,
Chicago, IL.

Dillehay, T. D. 1989. Monte Verde: A Late Pleistocene Settlement in Southern 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. Politis, & M. M. C. Beltrao. 1992. Earliest hunters and
gatherers of South America. J. World Prehist. 6:145—204.

Di Peso, C. C. 1965. The Clovis fluted point from the Timmy Site. Kzva 31:83—87.

Dixon, E. J. 1999. Bones, Boats ¢ Bison. University of New Mexico Press, Albuquerque,

NM. 322 pp.

Dixon, E. J., T. H. Heaton, T. E. Fifield, T. D. Hamilton, D. E. Putnam, & F. Grady. 1997.
Late Quaternary regional geoarchaeology of southeast Alaska Karst, A progress
report. Geoarchaeology 12:689—712.

gan, T. 1999. Expert panel recasts origin of fossil man in northwest. Te New York
Times, October 16:A11.

Erlandson, J. M., & M. Moss. 1996. The Pleistocene-Holocene transition along the
Pacific coast of North America. Pages 277—302 i L. G. Straus, B. V. Erisken, J. M.
Erlandson, & D. R. Yesner, eds., Humans at the End of the Ice-Age: The Archaeology of the
Pkestocene-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.

Fagan, B. 1987. The Great Journey: The Peopling of Ancent America. Thames and Hudson,
London. 288 pp.

Ferring, C. R. 1989. The Aubrey Clovis site: A Paleoindian locality in the upper Trinity
river basin, Texas. Curr. Res. Pleistocene 6:9-11.

. 1990. The 1989 investigations at the Aubrey Clovis site, Texas. Curr. Res.

Pleistocene 7:10—12.

. 1994. Late Ouaternary Geology of the Upper Trinity River Basin, Texas. Ph.D. disser-

tation, Department of Geology, University of Texas, Dallas. University Microfilms,

Ann Arbor, MI.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

. 1995. The late Quaternary geology and archaeology of the Aubrey Clovis site,

Texas. Pages 273-282 zn E. Johnson, ed., Ancent Peoples and Landscapes. 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. Sczence 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 Report November/December:1—12.

. 2000 The peopling of the New World: Present evidence, new theories, and
future directions. J. Archaeol, Res. 8:39-103.

Fladmark, K. 1978. The feasibility of the northwest coast as a migration route for early
man. Pages 19-28 mm 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. Relaciones de la Sociedad Argentina de Anthropologia 14:169—176.

. 1987 Recent research at localities Cerro la China and Cerro El Sombrero,
Argentina. Curr. Res. Pleistocene 4:148-149.

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.

Frison, G., ed. 1996. The Mill Iron Site. University of New Mexico Press, Albuquerque,
NM. 248 pp.

Frison, G., & B. Bradley. 1999. The Fenn Cache: Clovis Weapons and Tools. One Horse Land
and Cattle Company, Santa Fe, NM. 111 pp.

Frison, G., & D. J. Stanford, eds. 1982. The Agate Basin Site: A Record of the Paleoindian
Occupation on the Northwestern High Plains. Academic Press, New York, NY. 403 pp.
Frison, G. C., C. V. Haynes, Jr., & M. L. Larson. 1996. Discussion and conclusions. Pages
205-216 im G. C. Frison, 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.

ROOSEVELT, DOUGLAS AND BROWN

Gnecco, C., & S. Mora. 1997. Late Pleistocene/early Holocene tropical forest occupa-
tions at San Isidro and Pena Rojo, Colombia. Antiquity 71:683—-690.

Goebel, T. 1999. Ice-Age Beringia and human colonization for the Americas: A New View from
the North. Poster session paper, Clovis and Beyond Conference, Santa Fe, NM.

Goebel, T., R. Powers, & N. Bigelow. 1991. The Nenana complex of Alaska and Clovis
origins. Pages 49-80 zm R. Bonnichsen and K. L. Turnmire, eds., C/ovzs: 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. Stratified radiocarbon-dated sites of Clovis age and older in Brazil.
Pages 283-286 im 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 im 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 University, 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 321:769-771.

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 932. Ouat. Res. 51:27—-38.

Hannus, L. A. 1990. The Lange-Ferguson site: A case for mammoth-bone butchering-
tools. Pages 86—99 zm L. D. Agenbroad, J. L. Mead, & L. Nelson, eds., Megafauna and
Man: Discovery of Americas Hearthland. The Mammoth Sites of Hot Springs, South
Dakota, Inc. Scientific Papers, Vol. 1.

Harbour, J. 1975. General stratigraphy. Pages 33-56 zm K Wendorf & J. J. Hester, eds.,
Late Pleistocene Environents of the Southern High Plains. Port 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. Saence 145:
408-1413.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

. 1966. Elephant hunting in North America. Pages 44-52 zm R. S. MacNeish, ed.,
Early Man in America, Readings 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

181:305—310.

. 1974. Paleoenvironmental and cultural diversity in Late Pleistocene South

America: Reply to A. L. Bryan. Quart. Res. 4:378—380.

. 1991. More on Meadowcroft chronology. Rex. Archaeol. 12:8—14.

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

pling of the New World. Pages 355-374 im 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. Scence 275:1948.

. 1998. Personal communication. Letter to A.C. Roosevelt, September 28, 1998.

. 1999. Monte Verde and the pre-Clovis situation in America. Dascovering
Archaeology, Special Report: Monte Verde 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 VII. Radiocarbon 9:1-14.

Haynes, C. V., Jt, M. McFeul, R. H. Brunswig, & K. D. Hopkins. 1998. Kersey-Kuner
Terrace investigations at the Dent and Bernhardt sites, Colorado. Geoarchaeology 13:
201-218.

Hemmings, E. T. 1970. Early Man in the San Pedro Valley, Anzona. Ph.D. dissertation,
Department of Anthropology, University of Arizona. University Microfilms, AZ.
268 pp.

Hester, J. J., E. Lundelius, Jr., & R. Fryxell. 1972. Blackwater 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. FE, W. R. Powers, & T. Goebel. 1993. The colonization of Beringia and the
peopling of the New World. Saence 259:46—53.

Hoffecker, J. F, W. R. Powers, & N. H. Bigelow. 1996. Dry Creek. Pages 343-354 mm F. H.

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

Chicago Press, Chicago, IL.

AAS

ROOSEVELT, DOUGLAS AND BROWN

Holliday, V. T. 1997. Paleoindian Geoarchaeology of the Southern High Plains. University of
Texas Press, Austin, TX. 297 pp.

Humphrey, J. D., & C. R. Ferring. 1994. Stable isotope evidence for latest Pleistocene and
Holocene climatic change in north central Texas. Oat. Res. 41:200—213.

Hurt, W. R., TL van der Hammen, & G. Correal. 1976. The Al Abra Rockshelters, Sabana de
Bogota, Colombia South America. (Occasional Papers and Monographs, No. 2). Indiana
University Museum, Bloomington, IN.

Jennings, J. D. 1983. Origins. Pages 25-69 zm J. D. Jennings, ed., Ancient North Americans.
W. H. Freeman and Company, New York, NY.

, ed. 1983. Ancient North Americans. W.H. Freeman, San Francisco, CA. 642 pp.

Johnson, J. R., T. W. Stafford, Jr, H. O. Ajie, & D. P. Morrison. In press. Arlington
Springs revisited. Proceedings of the Fifth Cahforma 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. Sczence 281:1833—-1835.

Klein, R. 1999. The Human Career: Human Cultural and Biological Origins. University of
Chicago Press, Chicago, IL. 810 pp.

Krieger, A. D. 1964 Early man in the New World. Pages 23-84 zm J. D. Jennings & E.
Norbeck, eds., Prehistoric Man in the New World. University of Chicago Press, Chicago,
files

Kunz, M. L., & R. E. Reanier. 1994. Paleoindians in Beringia: Evidence from Artic
Alaska. Sczence 263:660—662.

Lahr, M. M. 1995. Patterns of modern human variation: Implications for Amerindian
origins. Yrbk. Phys. Anthropol. 38:163-198.

Leakey, L. S. B., R. D. Simpson, & T. Clements. 1968. Archaeological excavations in the
Calico Mountains, California: preliminary report. Scence 181: 305-310.

Leonhardy, F C., ed. 1966. Domebo: A Paleo-Indian Mammoth Kill in the Pratrie-Plains.
(Contributions of the Museum of the Great Plains, No. 1. Lawton, Oklahoma).
53 pp.

Lopes de Castano, C. 1995. Dispersion de puntas de proyectil bifaciales en la cuenc
media del rio Magdalena. Pages 73-82 in Ambito_y Ocupaciones Tempranas de America
Tropical. Fundacion ERIGAIE, Instituto Colombiano de Antropologia, Bogota,
Columbia.

Lopes de Castano, C., & C. Niewenhuis n.d. Clovis from the perspective of northern
South America. In preparation for A.C. Roosevelt & J. Morrow, eds., Clovis in Context:
New Light on the Peopling of the Americas. In preparation for submission to University
of Arizona Press, Tucson, AZ.

214

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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. Quat. Res. 4:356—377.

. 1980. Guctarrero Cave: Early Man in the Andes. Academic Press, New York, NY.

48 pp.

. 1983. The Paleo-Indians. Pages 87-138 im J. D. Jennings, ed., Axcient South

Americans. W. H. Freeman, San Francisco, CA.

. 1990. Glacial-age man in South America? A critical review. Am. Antiquity
55:199—228.

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 AEAR Excavation of Pintada
and Pendejo Caves near Orogrande, New 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. Sczence 168:975—-977.

MacNeish, R. S., R. K. Vierra, A. Nelken-Turner, & C. J. Phagan, eds. 1980. Prehestory 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, Pern. 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 Swmmary 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 7 P. S. Martin & H. E. Wright, eds.,
Pkeistocene Extinctions: The Search for a Cause. Yale University Press, New Haven, CT.

. 1984. Prehistoric extinctions: The global model. Pages 354—403 i P. S.

Martin & R. G. Klein, eds., Quaternary Extnetions: A Prehistoric Revolution. University of

Arizona Press, Tucson, AZ.

ANS)

ROOSEVELT, DOUGLAS AND BROWN

Mayer-Oakes, W. J. 1986. El-Inga: A Paleoindian site in the Sierra of northwestern
Ecuador. Trans. Am. Philos. Soc. 76(4).

McAvoy, J. M., & L. D. McAvoy. 1997. Archaeological Investigations of Site 445X202, Cactus
Hill, Susses County, Virginia. Virginia Department of Historic Resources, Research
Report No. 8. Nottoway Rivers Survey Archaeological Research Report No. 2.
Sandston, VA. 477 pp.

Mead, J. I., & D. J. Meltzer. 1984. North American Quaternary extinctions and the radio-
catbon record. Pages 440-450 m P. S. Martin & R. G. Klein, eds., Owaternary
Extinctions: A Prehistoric Revolution. Vhe University 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., & J. I. Mead. 1985. Dating Late Pleistocene extinctions: Theoretical issues,
analytical bias, and substantive results. Pages 145-173 zm J. I. Mead & D. J. Meltzer,
eds., Environments and Extinctions: Man in Late Glacial North America. Center for the
Study of Early Man, University 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. I. 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 62:659-6063.

Michab, M., J. K. Feathers, 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. Ouat. Sci. Rev. 17:1041—-1046.

Miller, E. T. 1987. Pesquisas archaeologicas paleoindigenas no Brasil occidental. Pages
37-61 m L. Nunez & B. J. Meggers, eds., Investigaciones Paleoinndias al Sur de la Linea
Ecuatonial. 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 161:1137-1138.

Morlan, R. E., & J. Cing-Mars. 1982. Ancient Beringians: Human occupation in the Late
Pleistocene of Alaska and the Yukon Territory. Pages 353-381 zm D. M. Hopkins, J.
V. Matthhews, Jr, C. E. Schweger, & S. B. Young, eds., Paseoecology of Beringia.
Academic Press, New York, NY.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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 zm 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. Sz.
63:304—313.

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 9-11.

Neves, W. A., & H. M. Puciarelli. 1991. Morphological affinities of the first Americans:
An exploratory analysis based on early South American remains. J. Hum. Evol.
21:261-273.

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 m 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 zm 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. Latin Am. Antiquity 5:99-118.

Ochsenius, C., & R. Gruhn. 1979. Taima-Taima: A Late Pleistocene Paleo-Indian Kill Site in
Northernmost South America. CIPICS/South American Quaternary Documentation
Program, Coro, Venezuela. 137 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 diversity 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.

2G)

ROOSEVELT, DOUGLAS AND BROWN

Pereira, E. 1992. Analyse prelimiar das pinturas rupestres de Monte Alegre (PA). Boktim
do Museu Paraense Enutio Goeldi, serie Antropologie. 8:5—24.

Pessis, A. -M. 1999. The chronology and evolution of the prehistoric rock paintings in
the Serra da Capivara National Park, Piaui, Brazil. Pages 41-47 im M. Strecker & P.
Bahn, eds., Dating and the Earhest Known Rock Art. Oxbow Books, Oxford, England.

Politis, G. 1991. Fishtail projectile points in the Southern Cone of South America: An
overview. Pages 287-301 7 R. Bonnichsen & K. L. Turnmire, eds. Clovis: 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 diversity 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. Phys. Anthropol. Suppl.
26:169.

Powell, J. FE, & J. Rose. 1999. Report on the Osteological Assessment of the “Kennewick
Man” Skeleton (CENWW.97.Kennewick). Dzscovering Archaeology, Special Report
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, Match 7:69-81.

Prous, A. 1980-1981. Fouilles du Grand Abri de Santana de Riacho (Minas Gerais),
Bresil. J. Soc. Americanistes 67:163-183.

1986a. L’Archeologie au Bresil, 300 Siecles d’occupation humain.

L’Anthropologie 90:257-306.

. 1986b. Os mais antigos vestigios arqueologicos no Brazil Central. Pages

173-182 m 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 PAbris du Boquete, Minas Gerais, Bresil. J. Soc. Americanistes

77:77-109.

. 1995. Archaeological analysis of the oldest settlements in the Americas. Revista

Brasileira de Genetica 18:689-699.

. 1999, Dating rock art in Brazil. Pages 29-34 im M. Strecker & P. Bahn, eds.,
Dating and the Earliest Rock Art. Oxbow Books, Oxford, England.

Ranere A., & R. Cooke. 1991. Paleoindian occupation in the Central American tropics.
Pages 237-253 im R. Bonnichsen & K. L. Turnmire, eds., Clovis: Origins and

218

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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

Reanier, R. E. 1996. Putu and Bedwell. Pages 505-511 zw EK H. West, ed., American
Beginnings: The Prehistory and Palaeoecology of Beringa. University of Chicago, Chicago,
0L5

. 1997. Technical comment. Sczence 275:1948-1949.

Reanier, R. E., & M. Kunz. 1995. The Mesa site: A Paleoindian hunting lookout in Arctic
Alaska. Arctic Anthropol. 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 im A. L. Bryan,
ed., Early Man in American in a Circum-Pacific 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: New Light on the Peoplng 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-

ty and comparison. Pages 165-191 m M. G. Plew, ed., Explorations in American

Archaeology: Essays in Honor of Wesley R. Hurt. University Press of America, Lanham,

MD.

. 1999, Dating the rock art at Monte Alegre, Brazil. Pages 35-41 am M. Strecker
& P. Bahn, eds., Dating and the Earhest Known Rock Art. Oxbow Books. Oxford,
England.

. 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 7 A. Hitchcock, ed., Museum Issues and Trends, Cultural Resource
Management 23:25-29. U.S. Department of the Interior, National Park Service,
Cultural Resources, Washington, D.C.

. 2000c. The Lower Amazon: A dynamic human habitat. Pages 455—491 7 D.
Lentz, ed., Imperfect Balance: Landscape Transformations in the Precolumbian Americas.
Columbia University Press, New York, NY.

. 2000d. Clovis in Context: New Light on the Peopling of the Americas. Hum.
[yal 125119).

ROOSEVELT, DOUGLAS AND BROWN

.n.d.a. Gender in human evolution: Sociobiology revisited and revised. Chapter

19 in S. Nelson & M. Ayalon, eds., In Search of Gender, Altamira Press. In press.

_n.d.b. Clovis dating, stratigraphy, and subsistence revisited from a hemispheric
perspective 7 A. C. Roosevelt & J. Morrow, eds., Clovis in Context: New Light on the
Peopling of the Americas. In preparation for submission to University of Arizona Press,
Tucson, AZ.

. n.d.c. Environment, Subsistence, and Behavior in Human Evolution 7 B.

Ferguson & R. Sussman, eds., A Critique of Social Darwinism. In preparation for sub-
mission to University of California Press.

Roosevelt, A. C., & J. Morrow, eds. n.d. Clovis in Context: New Light 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. 1991. Eighth
millennium pottery from a prehistoric shell midden in the Brazilian Amazon. Science
254:1621-1624.

Roosevelt, A. C., M. L. da Costa, C. Lopes Machado, M. Michab, N. Mercier, H. Valladas,
J. Feathers, W. Barnett, M. I. da Silveira, A. Henderson, J. Sliva, B. Chernoff, D. S.
Reese, J. A. Holman, N. Toth, & K. Schick. 1996. Paleoindian cave dwellers in the
Americas: The peopling of the Americas. Saence 272:373—384.

Roosevelt, 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. Scence 275:1950—1952.

Rouse, I., & J. M. Cruxent. 1963. Venezuelan Archaeology. Yale University Press, New
Haven, CT. 179 pp.

Sales Barbosa, A. 1992. A tradicao Itaparica: Uma compreensao ecologica e cultural
dopovoamento inictal do planalto central brasileiro. Pages 145-160 zw B. Meggers, ed.,
Prehistoria Sudamericana: Nuevas Perspectivas. Taraxacum, Washington, D.C.

Sandweiss, D. H., H. McInnis, R. L. Burger, A. Cano, B. Ojeda, R. Paredes, M. C.
Sandweiss, & M. D. Glasscock. 1998. Quebrada Jaguay: Early South American mar-
itime adaptations. Sczence 281:1830—1832.

Sassaman, K. E. 1996. Early Archaic settlement in the South Carolina coastal plain. Pages
58-83 mm D. G. Anderson & K. E. Sassaman, eds., The Paleoindian and Early Archaic
Southeast. University of Alabama, Tuscaloosa, AL.

Schmitz, P. 1987. Prehistoric hunter-gatherers of Brazil. J. World Prehist. 1:3-120.

Schobinger, J. 1999. Argentina’s oldest rock art. Pages 53-66 zm M. Strecker & P. Bahn,
eds., Dating the Earhest Known Rock Art. Oxbow Books, Oxford, England.

Scott, G. R., & C. G. Turner, I. 1997. The Anthropology of Modern Human Teeth: Dental
Morphology and its Variation in Recent Human Populations. Cambridge University Press,
Cambridge, England. 382 pp.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Sioh, H., ed. 1984. The Amazon: Limnology 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 (Pata). Bol Mus.
Paraense Emilio Goeldi N.S. 62. Belem.

Simpson, R. D. 1978. The Calico Mountains archaeological site. Pages 218-220 im 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
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 im F. W. Lange &
D. Z. Stone, eds., The Archaeology of Lower Central America. University of New Mexico
Press, Albuquerque, NM.

Stafford, T. W., Jr, J. T. Jull, 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 zm
L. D. Agenbroad, J. I. Mead, & L. W. Nelson, eds., Megafauna and Man: Discovery of
Ameriwas 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. J. Archaeol, Sci. 18:35—72.

Stanford, D. J. 1983. Pre-Clovis occupation south of the ice sheets. Pages 65—72 zm R.
Shutler, Jr., ed., Early Man in the New World. Sage Publications, Beverly Hills, CA

. 1991. Clovis origins and adaptations: An introductory perspective. Pages 14
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.

Steele, D. G., & J. H. Powell. 1993. Paleobiology of the first Americans. Evol Anthr. 2:
138-146.

Stuiver, M., P. J. Rimer, E. Bard, J. W. Beck, G. S. Burr, K. A. Hughen, B. Kromer, F G.
McCormac, J. van der Plicht, & M. Spurk. 1998. INTCAL 98 Radiocarbon age cali-
bration, 24,000-0 cal BP. Radiocarbon 40:1041—1083.

Taylor, R. E. 1987. Radiocarbon Dating: An Archaeological Perspective. Academic Press, New
York, NY. 212 pp.

ROOSEVELT, DOUGLAS AND BROWN

. 1992. Radiocarbon dating of bone: To collagen and beyond. Pages 375-402 in

R. E. Taylor, A. Long, & R. Kra, eds., Radiocarbon Dating after Four Decades: An
Interdisciplinary Perspective. Springer-Verlag, New York, NY.
. 1997. Radiocarbon dating. Pages 65—96 m R. E. Taylor & M. J. Aitken, eds.,

Chronometric Dating in Archaeology. Plenum Press, New York, NY.

Taylor, R. E., & M. J. Aitken, eds. 1997. Chronometric 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.

Taylors Ro Er le Aw Payne, CUAL Prior, 2) iy Slotayjiesike Gillespie) yA | "Gowletcesens:
M. Hedges, A. J. T. Jull, 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. Am. Antiquity
50:136—140.

Tieszen, L. 1991. Natural variations in the carbon isotope values of plants: Implications
for archaeology, ecology, and paleoecology. J. Archaeol, Sci.18:227—248.

Toth, N. 1991. The material record. Pages 53-76 im T. D. Dillehay & D. J. Meltzer, eds.,
The First Americans: Search and Research. CRC Press, Boca Raton, FL.

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. Pa/aeogeogr.,
Palaeochmateol., Palaeoecol. 109:247—261.

van der Merwe, N., & E. Medina. 1991. The canopy effect, carbon isotope ratios, and
foodwebs in Amazonia. J. Archaeol. Sci. 18:249—259.

Vialou, A. V., & D. Vialou. 1994. Les premiers peuplements prehistoriques du Mato
Grosso. Bull. Soc. Prehistorique Francais 19:57—203.

Wallace, A. R. 1889..A Narrative of Travels on the Amazon and Rio 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.
University of Chicago Press, Chicago, IL. 576 pp.

Wilford, J. N. 1999. New answers to old questions: Who got here first? The New York
Times Nov. 9:D1, D4.

Willey, G. R. & P. Phillips. 1958. Method and Theory in American Archaeology. University of
Chicago Press, Chicago, IL. 269 pp.

Wormington, M. 1957. Ancient Man in North America. Fourth edition. Denver Museum of
Natural History, Denver. 322 pp.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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.

——— . nd. Technological, chronological, and economic variability in Paleoindian
occupations from eastern Beringa 7 A. C. Roosevelt & J. Morrow eds., Clovis in
Context: New Light on the Peopling of the Americas. In preparation for submission to
University of Arizona Press.

223

ROOSEVELT, DOUGLAS AND BROWN

JN NIDILS. ||
Dates From High Plains Clovis Sites*

The Clovis Site, Blackwater Draw Locality 1, Clovis, NM
Weighted average of three dates on three samples: 11,300 £ 240 yr B.P. (13,180
cal)

The dates from the eponymous Clovis site fail five of the seven reliability criteria.
They have too-large error bars, are on material vulnerable to contamination by too-old
catbon, are on noncultural carbon, and lack of secure cultural association. In the region-
al sequence, they average 300 yr 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 lignin 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
ranges.

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 by nearby 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.

Lehner, AZ
Weighted average of 12 dates on 12 samples: 10,950 £ 40 yr B.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.

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Lehner is interpreted as a kill site (Haury ef a/ 1959). The samples were associated
with megafaunal bone and artifacts. The sample material is problematic: pooled aggre-
gates of botanically-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 £ 50 yr BP. (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)

175150! 450** (A805) (13,144 cal)

11,080 + 180 (1x1413) (13,125, 13,083, 13,031 cal)

10,930 + 170 (Tx1462) (12,970 cal)

10,890 + 180 (SMU27) (12,933 cal)

225

ROOSEVELT, DOUGLAS AND BROWN

10,840 + 140 (SMU42 (12,902 cal)
10,840 + 70 (SMU41) (12,902 cal)
10,710 + 160 (Tx1459) (12,847 cal)

The dates from the Dent, Colorado, site fulfills some of the dating reliability 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) aspartic acid from same

10,800 + 110 (AA2943) (12,889 cal) hydroxyproline from same

10,600 + 90 (AA2945) (12,800, 12,731, 12,652 cal) hydroxyproline from same
10,710 + 90 (AA2946) (12,847 cal) glycine from same

10,670 + 120 (AA2947) (12,824, 12,704, 12,682 cal) alanine from same

Domebo, OK
Weighted average of seven dates on four samples: 10,944 £ 59 yr B.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

226

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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-11,200 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
bone

10,860 + 450** (AA811) (12,911 cal) pro-hypro 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)

The dates have low scores on all reliability 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

aeall

ROOSEVELT, DOUGLAS AND BROWN

The five dates fulfill four of seven reliability criteria. The earliest three are internally
consistent, and the latest two are internally consistent, but the two groups are not con-
sistent with each other. Clovis 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 directly asso-
ciated diagnostic Clovis cultural materials in a multicomponent site means that the cul-
ture of the earlier dated skeleton 1s 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) hydroxyproline 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

Colby, WY
Weighted average of two dates on two samples: 10,960 £ 120 yr 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 statistically, 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, unpurified collagen from mammoth bone
10,864 + 141 (SMU254) (12,913 cal) apatite from mammoth bone

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

UP Mammoth site, WY

No weighted average because only a single date was available.

The date only meets one of the seven criteria of reliability. 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 contamination. 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
tusk

Goshen Culture

Mill Iron, MT

No weighted average possible due to statistical inconsistency of associated dates.

The site fails six of the seven reliability criteria. The site was a kill 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 stylistic 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 cultural 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

N
N
No)

ROOSEVELT, DOUGLAS AND BROWN

11,360 + 130 (Beta 20111) (13,241 cal) camp, processing area
11,340
11,320
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

+ 120 (Beta 13026) (13,258 cal) camp, processing area
+ 130 (Beta 16179) (13,279 cal) camp, processing area

Culture Undetermined

Aubrey, TX
Weighted average of two dates on two samples: 11,570 £ 70 yr B.P. (13,480 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 botanically-unidentified, dispersed
catbon 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 disconformity on an ero-
sion surface at the interface of Stratum G, a 10,900 to 7,600 yr B.P. carbonate-tich 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 locality, suggesting that the
cultural component is a later Paleoindian culture deposited into the post-Clovis geolog-
ical Stratum 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 a/ 1964:93-98; Damon ef a/ 1966:100-101;
Ferring 1989; 1994; Frison 1991:25, 39, 41, 1996; Frison & Todd 1986; Hannus 1990;
Haury et a/ 1959, figure 16; Haynes 1987, 1992, 1997; Haynes et a/, 1966, 1967; Haynes
et al. 1998; Humphrey and Ferring 1994; Taylor et a/. 1996; Leonhardy 1966; Hester ef a/.
1972:174-176; 225, figure 130; Roosevelt 1998a; Stafford et a/ 1987; Stafford 1990;

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

Stuiver ef a/, 1998a. Weighted averages calculated by Linda Brown and John Douglas,
using CALIB.)

**indicates a too-large error.

Initial_A and B cakbrated radiocarbon and calendar TL. and OSL
dates from Pedra Pintada cave, Monte Alegre culture.

The radiocarbon dates fit all the reliability criteria, and the TL/OSL dates fit all but
one. The carbon samples were individual pieces of identified carbonized palm or tree
fruits, wood charcoal chunks, lithic 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 ate judged by Amazon zoologists to be evidence of human, not animal, subsis-
tence.) Chemical tests for solid and soluble contamination were negative. All radiocarbon
errors were within acceptable error ranges, unlike many Clovis dates. All TL/OSL errors
were larger than acceptable for radiocarbon but typical for these techniques. All samples
were from specific, mapped localities 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 fall
statistically in the same ranges. The radiocarbon series from each phase is statistically
consistent internally, and the weight averages of the carbon dates of the two phases dif-
fer significantly. The weight averages of the calibrated radiocarbon dates and the accept-
ed calendar TL/OSI dates are statistically consistent. (The earliest TL date, 16,190 yr B.P.,
was tejected for inconsistency with the rest of the series.) The radiocarbon range also 1s
consistent statistically with the reliable dates from other eastern Brazilian sites with sim-
ilar cultural and biological matertals.*

WL) OSIL
Initial A

8231SW base, object 2 next to hearth
TL of burned chalcedony bifacial reduction flake (outlier, rejected) (16,190 + 930
cally

8346 base, object 1
TL of burned chalcedony bifacial reduction flake 15,330 + 900 cal**

231

ROOSEVELT, DOUGLAS AND BROWN

8231W, hearth
OSL of sediment 13,106 + 1628 cal**
8231SE 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 samples: 13,180 + 509 calendar yr B.P.

OL /OSIL
Initial B

8231 far SW, top, object 1
TL of burned brecciated quartz flake 11,880 + 760 cal**

Radiocarbon
Initial A

8314CW, base, hearth, object 4

carbonized sacuri endocarp 11,145 + 135 (GX—17413) 13,143 cal
8231W base, hearth, object 17

carbonized tucuma endocarp 11,110 + 310 (GX—17406) 13,135 cal
8231SE 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, 13,028
cal

Radiocarbon
Initial B

9290
carbonized jutai seed 10,683 + 80 (NZA9898) 12,954-12,681 cal

ie)
iS)
in)

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

8345 object 4

catbonized tucuma endocarp 10,655 + 285 (GX-17420) 12,712 cal
8231CW top, object

catbonized 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

catbonized sacuri endocarp 10,410 + 60 (GX-19538CAMS) 12,531 cal
8314SW top, object 6

carbonized sacuri endocarp 10,392 + 78 (GX—17400CAMS) 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-17422) 12,241 cal
8345 object 13

catbonized sacuri endocarp10,275 + 275 (GX—-17421) 12,016 cal
8345

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

WL OSL,
Initial A

8231SW 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
8231SE hearth, object 18

ROOSEVELT, DOUGLAS AND BROWN

OSL of heavy fraction of sediment 12,536 + 4125 cal
8231CE general level

OSL of sediment 12,491 + 1409 cal
Weighted average of five: 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

Radiocarbon
Initial A

8314CW, base, hearth, object 4

carbonized sacuri endocarp 11,145 = 135 (GX—17413) 13,143 cal
8231W base, hearth, object 17

carbonized tucuma endocarp 11,110 = 310 (GX—17406) 13,135 cal
8231SE 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

9290

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

i)
SN)
aS

CLOVIS AND PRE-CLOVIS VIEWED FROM SOUTH AMERICA

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—17400CAMS) 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—17422) 12,241 cal
8345 object 13

carbonized sacuri endocarp 10,275 = 275 (GX-17421) 12,016 cal
8345

catbonized 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 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 aZ 1996;
Roosevelt e¢ af 1997.

**indicates a too-large error

235

4 a Ne? ENN IT DRE AS NP EI

Pa. sana
or jew lah, Sa SARA PT 20 Wt EO e ad.

Lee ai & ‘ot Jue ub Te } Pi MM +g

‘a ORD Fey ANG bd) OF eS pian lal iy ire i. :
oO Tie A ee ee wh unis ie

TAT

> ly f é ; 7 t 4 ) }
HY if! ; }
; si ;e i i ie if a mc af nt tama
ray ¢

gilqrieie bt

4 ‘a hy

y
*
—m
i 1
i
i
a 7 ,
i
vA
ms
i : Le
j at 4
T * F
\ ,
i * ie dy : * ¥
i) r nt
il i :
. im) |
ey ei

CHAPTER EIGHT

PLANT FOOD AND ITS IMPLICATIONS FOR THE
PEOPLING OF THE NEw WorRLD:
A VIEW FROM SOUTH AMERICA

Tom D. Dillehay and Jack Rossen

In recent years, it has become clear that without the important adap-
tations that took place among Late Pleistocene foragers in many
parts of the Americas, the subsequent development of more com-

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 wetland 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
Americas.

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 (¢., Martin 1973; Haynes 1969; Fagan 1987; Carlisle 1988;
Meltzer 1993, 1997; Dillehay e7 a/, 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

everywhete.

iw)
Oo
=i

DILLEHAY AND ROSSEN

Most exciting in recent years has been a greater recognition of widespread techno-
logical and economic diversity throughout the Americas. People not only hunted, but
they also exploited a wide array of plant foods and, in coastal areas, marine foods. This
diversity probably relates to a greater time depth of humans in the New World than was
previously recognized and thus more ancient opportunities to develop different local tra-
ditions and more cultural complexity. We know that cultural diversity in South America
was widespread by at least 10,500 yr B.P. (Bryan 1973, 1978, 1986; Ardila & Politis 1989;
Dillehay e¢ a/ 1993; Dillehay 1999).

Archaeologists have identified big-game hunting as a fundamental subsistence activ-
ity 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 quantity and diversity
when flotation is included in the excavation and analysis strategy (Dillehay 1989; Rossen
et al. 1996; Dillehay et a/ 1997; Ramirez 1989). As a result, archeologists have underesti-
mated the role of plant foods in early diets, especially in wetland 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 highlands 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 sophisticated 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 e¢ a/.
1993; Dillehay 1999). We expect to find these sites in “food affluent” environments that
contained a great bulk and high diversity of usable plant biomass, such as wetlands and

forests. Within wetlands, whether they are swamp, mangroves, marsh, fen or bog, salt-

bo
(SS)
CO

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

coLtomBia oY?
aS c
CUADOREN

N
ve
‘|

PARAGUAY)

ARGENTINA i.
i, N
(URUGUAY

ie) 400 800 1200 1600 Km

FiGurE 1. Of South America showing countries, major river systems, 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.

2a)

DILLEHAY AND ROSSEN

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, wetlands were attractive “pull” zones that reduced early Archaic group mobility
(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-
tle 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., tubers, reeds and sedges; Hatley & Kappelman 1980; Coursey & Booth
1977; see Ramirez 1989 and Rossen & Ramirez 1997 for southern Chile) and their broad-
er implications for economic viability. These parts are special because they usually are
less affected by drought, fire, seasonal fluctuations, 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 cultural develop-
ments. These cases are not intended to represent all early time periods and regional envi-
ronments.

Monte Verde 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 Monte Verde is characterized by varied surface features: rainforests of
mixed coniferous and broadleaf trees, rolling hills and low and high mountain ranges,

240

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

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. (Dillehay & Collins 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-like 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-like 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 lithic 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

DILLEHAY AND ROSSEN

cool temperate rainforests in southern Chile. Irrespective of location, all early sites char-
acterized by untfacial industries and by good preservation of organic faunal and floral
remains are associated with a broad-spectrum foraging economy.

Hunting and/or scavenging game is suggested by the bone remains of at least seven
individual 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-
ful.

A well-preserved collection of plant remains (¢.¢., 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 cultures (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 down the river system from the
highlands to the coast.

The macrobotanical remains include more than sixty plant species that sequentially
mature during all seasons of the year in different ecological zones (Dillehay 1984, 1986;
Ramirez 1989; Rossen & Ramirez 1997). Grains, seeds, nuts, tuber 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 yielded information
both on the species of above-ground and below-ground plants that were used by the
Monte Verdeans and on the general nature of the diet.

Among the archaeologically recovered plants, a few species are staple plants that are
available year-round, and have edible greens, seeds, rhizomes or tubers. These species
include Juncus procerus, Scirpus californicus, Carex sp., Gunnera sp., and Solanum magha (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 (C/avaria coralloides), a bamboo (Chusquea sp.), and four seaweeds
(Durvillaea antarctica, Gracilaria sp., Porphyra sp., and Sargassum sp.; Figures 2 & 3) have spe-
cific seasonalities to suggest they were utilized during the non-summer months when
overall plant food availability 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

242

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

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 (Lycopodium sp.), which seem to have been used as a fite 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, So/anum, 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 (cf. Bukasov 1930; Correl 1962; Vavilov 1951; Ugent ef a/Z 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-
ologically.

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” (sensu 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 gm, Juncus procerus seeds ten-
der 7.3 gm per 100 gm, and Sarpus caljornicus 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). Micto-use wear
studies of these and replicative digging sticks, suggest that the site’s inhabitants were
probably using these implements to extract underground food patts.

Based on the reconstruction of past and present resource structures in the study area,
we can infer that the Monte Verdeans chose two complementary ways to procure tubers,

DILLEHAY AND ROSSEN

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 types 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 highlands 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 wetland habitats where
an abundant year-round supply of below-ground tubers and rhizomes were concentrat-
ed. The most productive and accessible wetlands 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 wetlands zones, allowing economic

244

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

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 resoutces.

Ficure 3. A desiccated bulb of a wild potato (So/anum maglia) recovered from the floor of the long
tentlike structure dated at 12,500 yr B.P. at Monte Verde.

Las 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 the 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

245

DILLEHAY AND ROSSEN

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-collecting implements, including small mortar stones.
People who viewed almost all living organisms as potential sources of food and were
willing to organize their activities to collect them, would have been developing an aware-
ness and organizational system potentially receptive to the collection and eventual
domestication 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.

Nanchoc 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 ef a/ 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 culture 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 little 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 ef a/ 1997; Rossen 1998b). Faunal assem-
blages are dominated by mollusks and small game.

Several Sites in Central Brazil

Almost thirty 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 wild fruits.
Perishable wooden and reed artifacts are present in at least three sites (Kipnis 1998:587).

246

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

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-
eret 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,

247

DILLEHAY AND ROSSEN

probably produced the staple foundation of the diet. Yet, there are no identifiable stor-
age facilities at Monte Verde. The Monte Verdeans were probably sedentary nonstorers
who relied on the underground storage and nutritional banking of certain tubers and rhi-
zomes with long use-lives and on the occasional exploitation 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,
fruits, 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, mobility, technology, and social struc-
ture 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 yeat.

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 wetlands 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 sufficiently supported by
available game. Although the settlement pattern data are lacking, it 1s highly unlikely, in
our opinion, that such pressure existed in Late Pleistocene times. It is our guess that, in
the case of Monte Verde, the generalized economy and settlement stability was primari-
ly founded on the high diversity and density and year-round availability 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
quality of minerals and carbohydrates, starch, trace elements, and plant and animal pro-
tein.

248

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

Conclusions

Undoubtedly, new discoveries and directions in the interpretation of initial human
settlement 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 apparently 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 Brazil, northern Peru, and southern Ecuador. Future study
surely will find them in many other forested and wetland 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, subculturally, 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 1f we do not bother to collect what is there.

Abstract

Due to preservation and paradigm bias, plant collecting is generally given Iit-
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
Brazil, and their meaning with respect to the exploitation of wetland environ-

249

DILLEHAY AND ROSSEN

ments and the peopling 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.

Literature Cited

Ardila, G., & G. Politis. 1989. Nuevos datos para un viejo problema: investigacion y dis-
cusion en tornodel poblamiento de America del Sur. Revzsta del Museo del Oro 23:3-45.
Bogota, Colombia.

Binford, L. 1978. Dimensional analysis of behavior and site structure: Learning from an
Hskimo hunting stand. Am. Antiquity 43:330-301.

. 1980. Willow smoke and dog’s tails: Hunter-gatherer settlement systems and

archaeological site formation. Am. Antiquity 45:4—20.

. 1987. Researching ambiguity: Frames of reference and site structure. Pages
418-456 m S. Kent, ed., Method and Theory for Activity Area Research. Colambia
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. zn
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 cultural diversity in Late Pleistocene South
America. Ouat. Res. 3:237—256.

. 1978. An overview of Paleo-American prehistory from a circum-Pacific per-

spective. Pages 306-327 im A. L. Bryan, ed., Early Man in America from a Circum-Pacific

Perspective. Occasional Papers No. 1, Department of Anthropology. University of

Alberta, Edmonton, Canada.

. 1986. Paleoamerican prehistory as seen from South America. Pages 1-14 zn
A. L. Bryan, ed., New Evidence for the Pleistocene Peopling of the Americas. Center for the
Study of Early Man. Univesity of Maine, Orono, ME.

Bukasoy, I. 1930. The cultivated plants of Mexico, Guatemala, and Colombia. Bu//. App/.
Bot. Genet. PL-Breed. (Trudy Prikl. Bot. Suppl.) 47:513-523.

Butzer, K. 1971. Environment and Archaeology. 2nd Edition. Aldine Press, Chicago, IL.

Carlisle, R. 1988. 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.

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

Collins, M. B., & B. N. Driskell. 1979. Summary and conclusions. Pages 1023-1042 in M.
B. Collins, ed., Excavations at Four Archaic Sites in the Lower 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 23-44 in C.L. Leakey,
ed., Food Crops of the Lowland Tropics. Oxford University Press, Oxford, England.

Dillehay, T. 1984. A late Ice Age settlement in Southern Chile. Sc. Amer. 251:106-117.

. 1986. The cultural relationship of Monte Verde: A Late Pleistocene settlement

site in the Sub-Antarctic forest of South-Central Chile. Pages 319-337 im 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 Late Pleistocene Settlement in Chile. vol. 1. Smithsonian

Institution Press, Washington, D.C.

. 1997. Monte Verde: A Late Pleistocene Settlement in Chile. vol. 2. Smithsonian

Institution Press, Washington, D.C. 1071 pp.

. 1999. The Late Pleistocene cultures of South America. Evol Anthro. 7:206-217.

. 2000. The Settlement of the Americas: A New Prehistory. Basic Books, New York,
NY. 371 pp.

Dillehay, T., & M. Collins. 1988. Early cultural evidence from Monte Verde in Chile.
Nature 332:150-152.

Dillehay, T., G. A. Calderon, G. Politis, & M. da Beltrao. 1993. Earliest hunters and gath-
erers of South America. J. World Prehist. 6:145—204.

Dillehay, T., J. Rossen, & P. Netherly. 1997. The Nanchoc tradition: The beginnings of
Andean civilization. Am. Sc. 85:46—55.

Dillehay, T., & J. Rossen. 1997. Integrity and distribution of the archaeobotanical collec-
tion. Pages 351-382 zn T. Dillehay, ed., Monte Verde: A Late Pleistocene Settlement in Chile.
vol. 2. Smithsonian Institution Press, Washington, D.C.

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 zn P. J. Ucko & G. W. Dimbleby, eds., The Domestication and
Exploitation of Plants and Animals. A\dine 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 R. Hurt. University Press of the Americas, Washington, D.C.

251

DILLEHAY AND ROSSEN

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 exploitation of below ground food resources. Hum. Ecol. 8:371—387.

Ingold, T. 1987. The Appropriation of Nature. University of Iowa Press, lowa City. 287 pp.

Instituto de Geociencias. 1977. Informe sobre el Chma Regional. 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 zm J. D. Jennings & E.
Norbeck, ed., Prehistoric Man in the New World. University of Chicago Press, Chicago,
0G.

Laboratorio Fitoquimica. 1983. Informe sobre el valor nutritivo de las papas stlvestres y Juncus.
Universidad Austral de Chile, Valdivia.

LaBuza, T. P. 1982. She/Pkfe Dating of Foods. Food and Nutrition. Westport, CT. 500 pp.

Lopinot, N. 1982. Miscellaneous subterranean plant pasts. Pages 744-746 zm R. Jefferies
& B. Butler, eds., The Carrier Mills Archaeological Project: Human Adaptation in the Saline
Valley, I/hnois. Center for Archaeological Investigations. Research Paper No. 33.
Southern Illinois University, Carbondale, IL.

Lynch, T. F. 1981. Guctarreo Cave: Early Man in the Andes. Academic Press, New York.

. 1983. The Paleo-Indians. Pages 87-137 zm J. D. Jennings, ed., Ancient South
Americans. WH. Freeman, San Francisco, CA.

Martin, P. S. 1973. The discovery of America. Sczence 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. Scence
276:754—755.

Meltzer, D. J., & B. D. Smith. 1986. Paleo-Indian and early Archaic subsistence strategies
in Eastern North America. Pages 1-30 im S. Nesstus, ed., Foraging, Collecting and
Harvesting: Archaic Period Subsistence and Settlement in the Eastern Woodlands. Center for
Archaeological Investigations. Southern Illinois University, Carbondale, IL.

Nance, J. 1988. The Archaic period in the lower Tennessee-Cumberland-Ohio region.
Pages 127-152 a C. D. Hockensmith, D. Pollack, & T. N. Sanders, eds., Paleoindian and
Archaic Research in Kentucky. Kentacky Heritage Council, Frankfort, KY.

Prous, A. 1996. L’Archeologie au Bresil: 300 siecles d’ occupation humaine.
L’Anthropologie. 190:257—306.

Ramirez, C. 1989. Macro-botanical remains. Pages zm T. Dillehay, ed., Monte Verde: A Late
Pleistocene Settlement in Chile. vol. 1. Palaeoenvironment and Site Setting. Smithsonian
Institution Press, Washington, D.C.

252.

PLANT FOOD AND THE PEOPLING OF THE NEW WORLD

Redman, C. 1977. Man, domestication, and culture in Southwestern Asia. Pages 523-541
in C. Reed, ed., Ongins 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 i A. L.
Bryan, ed., Early Man in America from a Circum-Pacifie Perspective. Occasional Papers No.
1, Department of Anthropology, University 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 County,

Kentucky. Pages 33-44 mm D. Pollack & K. Gremillion, 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 zm T. Dillehay, ed., Monte Verde: A Late Pleistocene Settlement in
Chile. vol. 2. Smithsonian Institution Press, Washington, D.C.

Rossen, J., T. D. Dillehay, & D. Ugent. 1996. Ancient cultigens or modern intrusions?
Evaluating plant remains in an Andean Case Study. J. Archaeo. Science 23:391407.
Rossen, J., & C. Ramirez. 1997. Observations on the present-day (1983) economic plants
in the Monte Verde area and their archaeological implications. Pages 7m T. Dillehay,
ed., Monte Verde: A Late Pleistocene Settlement in Chile. vol. 2. Smithsonian Institution

Press, Washington, D.C.

Schmitz, P. 1987. Prehistoric hunters and gatherers in Brazil. ]. World Prebist. 1:1-506.

Soffer, Olga. 1989. Storage, sedentism and the Eurasian Palaeolithic record. Antiquity
63:719—32.

Stothert, K. 1988. La Prehistoria Temprana de la Penisula de Santa Elena: La 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.

praye)

- ‘ . a eras igo - ar ‘a ii ve mr ssf Hide Fe prction oaecee
| nyn* —_ ws a“ P | Veh " al manila 7 by i oar ae! quit nee as | i *
Do ain homme eee ee te oi vail
sient scree tniheianGihs titionylle dln iia “
iM hgaeieaekt } vdddrarn. 1 Pn ee ype Spl wind d iota
Socata sf Les tal (uae waar cpurtra hen Ma) iis Patty's 1m phere how
ty | . 0G Leen, gli’ i

Perle: > * : a ‘L nit ‘hed ee Nye pal” ot i pdikgllllan iyemalay anal? th r —

rat

Lyeeg gia ee, wad wh inte dublin Me 4 nie the PRE W. ath bie

Oke ; F pin hia ts fuleputire eur (i bn vial be vo ethane

a “ A grigy Tisce jt yee ii Hie sliftiy me ee ALY ula m! ee
' | | ich }

ae AY + ihe ee, be qh ne at

eh alps oma
ah te Zi ag vit | haa avi ; nh ae oo. oe i
Maniac Me AMUN i maT RL a
» Pe ial Nie piers, APO are i i

” thea ia aie aR Aba eal .
‘awl 7% ny ver _— A beinvng au re Ye ii ts f enbine mn a iM i “we ond?

‘ei Te a er Poe a vane i Ne Z aly oa

. ‘ ner Son ‘ oY ‘ Thy rv 1 ae
j bap nredy e ‘eo 7 F Ci
eee

7 ‘ee fa ‘ wikia 4 ‘ ‘Pi sn ka va _ nif i will ae pian er
‘hing oo a a ee ee

ee Sos rn oe isl f onan ail "ee ¢ yf a4" wy abe’ Ties
‘ao aa » ta any’ Rete Analy la Birjttni hana te al eae

q wr ) oe)? a -
i a Y +
‘ a i 2
3 ore ma Ty .\ VANE lle in gta? Pcie
he mci : os him a” -, re

CHAPTER NINE

OCEAN TRAILS AND PRAIRIE PATHS?
THOUGHTS ABOUT CLOVIS ORIGINS

Dennis Stanford and Bruce Bradley

Archaeologists have long believed that artifacts unearthed with
mammoth bones in a gravel pit near Clovis, New Mexico were made
by the First Americans (see Bonnichsen & Turnmire 1991). The
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 proximity 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. Consequently, 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.

7y5)5)

STANFORD AND BRADLEY

Evidence for Asian Origins

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 e a/, 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 ef a/ 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
between 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 virtu-
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 corti-
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 a/ 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 essentially blocked
for early explorers between 22,000 and 11,000 years ago.

Evidence for Ocean Trat/s

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
eatly 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

OCEAN TRAILS AND PRAIRIE PATHS

Pacific continental shelf, which apparently was ice-free by 13,000 years ago (Josenhans ef
al. 1997). Recent efforts to reconstruct former environments and ecosystems verify that
Pleistocene animals flourished along the Ice Age coastline (Heaton 1996). Further, cul-
tural artifacts have been found along these coastal environments (Erlandson ef a/. 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 likely those
of the North Atlantic 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 Asta?

A growing number of researchers, including the present authors, have examined
many of the Late Pleistocene archaeological sites and museum collections in Eurasia,

JI)

STANFORD AND BRADLEY

Asia and the Far East in order to seek ties with Clovis. But, to date, no clear-cut evidence
is discernible. Asian Upper Paleolithic technologies are dominated by microblades made
from wedge-shaped cores (Derew’anko 1998). When used in weaponry, 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 significantly 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 they 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 (Kirillov &
Derevanko 1998). Such sites are rare, are older than Clovis by more than twenty 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
@ al, NYS),

Since known Asian technologies do not appear to satisfy expectations of a Clovis
technological prototype, 1f 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.

Alternative 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 similarity included bifacially flaked
lanceolate, projectile points with concave bases and basal and edge grinding, a basic blade
industry, small end scrapers with graver tips, large side scrapers made on flakes, cylindri-
cal bevel-based bone points, and a relative scarcity of burins. However, he felt that a

OCEAN TRAILS AND PRAIRIE PATHS

direct historical relationship between Clovis and Solutrean was impossible because: 1)
Solutrean is more than 5,000 years older than Clovis; and 2) of the difficulty 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 cultures 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 will 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 ef a/.
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 e¢ a/ 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 recently 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

259

STANFORD AND BRADLEY

the same technological complex. Further, their chronological placement suggests to us
that they are prime candidates for developmental Clovis.

It is important to note that there are significantly more fluted points and sites found
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 Tradition may have a greater time depth in the east and that Clovis probably
developed its characteristic 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-
ity that the lithic technology from these sites represents a likely precursor of Clovis tech-
nology.

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
Solutrean points and though they lack the fluung found on Clovis specimens, they are
basely thinned. Similar too are the blades and smaller blade cores. Further, some of the
radiocarbon dates for these sites overlap in time with Solutrean. If these dates prove to
be accurate, we see the possibility of a developmental continuum from Solutrean
through Cactus Hill and Meadowcroft to Clovis, filling in the 5,000-year time gap.

Coincidentally with the discovery at Cactus Hill, the results of DNA research on
ancient human remains have begun to map genetic relationships among the world’s pop-
ulations. Prom 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 1s 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 possibility 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 Solutrean Evidence

With these new data and the questions they pose, we felt that the hypothesis of a
Solutrean origin for some of the first 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 the Solutrean archae-
ological data.

260

OCEAN TRAILS AND PRAIRIE PATHS

Our examination of these materials was most illuminating. The 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 (Frison & 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 consistently 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 (¢.g., Balout 1958:619), pressure flaking and occa-
sional heat-treating of flint. 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 proyec-
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 culture. Along with incised
decoration on bone points, engraved stones, etched with geometric and zoomorphic
designs have been found in both cultures (Smith 1966; Collins ef 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 necessaty 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.

261

STANFORD AND BRADLEY

Filling the Space Gap

If there was an historic connection, the remaining questions to be addressed are
issues of the crossing of the Atlantic Ocean. Did the Solutreans possess the necessary
skills, have the right equipment, and why would they undertake such a journey? These
questions 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 tude state of technological knowledge that such endeavors would have 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.

Paleoclimatic 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
quality 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 a/, 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 —
11°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.

OCEAN TRAILS AND PRAIRIE PATHS

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 martine 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 mote 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 ef 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 that 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.

STANFORD AND BRADLEY

With the knowledge of cordage manufacture, only simple modifications to land-
based weaponry would be required 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 found most com-
monly in Solutrean sites in north coastal Spanish assemblages may have been well suited
for harpoon end blades.

As mentioned above, watercraft had been in use for at least 30,000 years before
Solutrean times, 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 textiles 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 byproducts of animals that Solutrean peoples were
hunting in the coastal mountains of Spain.

The Route

Paleooceanographic studies of the North Atlantic 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 Atlantic shifted southwards, making a clockwise circulation pattern
moving toward the coast of Portugal and returned westward from North Africa. The
Gulf Current warmed a weak counter clockwise current that circulated in the North
Atlantic (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 Atlantic has been characterized as cold and stormy,
these hostile weather conditions were not continuous, being interrupted by periods of
short term and perhaps even longer term intervals in which sailing conditions on open

OCEAN TRAILS AND PRAIRIE PATHS

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 pullout 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
ventured 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 haplotype 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 Hill and Meadowcroft sites provide the link
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-

digms.

265

STANFORD AND BRADLEY

exposed contenental shelf

ice sheets
Wii, permanant sea ice
polar front
===> summer winds
——— weak warm current
——> _~warm Gulf current

FIGURE 1. North Atlantic surface circulation reconstruction for the LGM. Adapted from Robinson
et al. 1995; Webb et al 1993.

Abstract

This paper discusses some of the current assumptions widely held about New
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 Last Glacial Maximum (LGM). A maritime compo-
nent of their adaptive orientation, termed here North Atlantic Paleomaritine,
enabled the movement of people westward along the margins of North Atlantic
sea ice where 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 LGM for Europe

266

OCEAN TRAILS AND PRAIRIE PATHS

and the North Atlantic, we feel that this hypothesis is most parsimonious with
currently avatlable data and warrants further study and serious consideration.

Literature Cited

Adovasio, J. M., J. D. Gunn, J. Donahue, R. Stuckenrath, J. Guilday, & K. Lord. 1978.
Meadowcroft Rockshelter. Pages 140-180 zw A. L. Bryan, ed., Early Man in America
Jrom a Circum-Pacific Perspective. Occasional Papers No.1 Department of Anthropology,
University of Alberta, Edmonton, Canada.

Adovasio, J. M., O. Soffer, & B. Klima. 1996. Upper Palaeolithic fiber technology:
interlaced woven finds from Pavlov I, Czech Republic, c. 26,000 years ago. Antiquity
70:526—534.

Adovasio, J. M., D. Hyland, & O. Soffer. 1998. Perishable technology and early human
populations in the New World. Paper prepared for the 315¢ Annual Chacmool
Conference: On being first: Cultural innovation and environmental consequences of
first peopling. Calgary, Alberta, Canada.

Anderson, D. G., & J. C. Gillam. 1999. Paleoindian colonization of the Americas:
Implications from an examination of physiography, demography, and artifact distri-
bution. Am. Antiquity 65:43—60.

Balout, L. 1958. L’Abri Andre-Ragout au Bois-du-Roc. Fouilles de 1957. Bul 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. Ouat.
Int. 32:113-126.

Bednarik, R. 1997. The earliest evidence of ocean navigation. Int. J. Naut. Archaeol.
26:183-191.

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. Girta. 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. Ouat. Int. 32:107-112.

267

STANFORD AND BRADLEY

Butzer, IX. W. 1986. Paleolithic adaptations and settlement in Cantabrian Spain. Pages
201-252 in E Wendorf & E. Close, eds., Advances 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: Last
Ad, 47) Annual Meeting of the Society for American Archaeology, Minneapolis,
MN.

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 Study of the First Americans, Oregon State
University Press, Corvallis, OR.

Cleyet-Merle, J., & S. Madelaine. 1995. Inland evidence of human sea coast exploitation
in Palaeolithic France. Pages 303-308 zm A. Fischer, ed., Man and Sea in the Mesolithic.
Oxbow Books, Oxford, England.

Collins, M. B., T. R. Hester, & P. J. Headrick. 1991. Engraved cobbles from the Gault Site,
Central Texas. Curr. Res. Pleistocene 9:3-4.

Cotter, J. L. 1935. Yama and Folsom Artifacts. M.A. thesis, Department of Anthropology,
University of Denver, CO.

Derev’anko, A. 1998. The Paleolithic of Siberia: New Discoveries and Interpretations. University
of Illinois Press, Urbana, IL. 406 pp.

Dixon, E. J. 1999. Bones Boats and Bison: Archeology and the First Colonization of Western North
America. Vhe University of New Mexico Press, Albuquerque, NM. 322 pp.

Erlandson, J. M., T. C. Rick, R. L. Vellanoweth, & D. J. Kenneth. 1998. Maritime subsis-
tence at a 9300 year old shell midden on Santa Rosa Island, California. /. Fre/d
Archaeol. 26:255—265.

Fedje, D. W., & H. Josenhans. 2000. Drowned forests and archaeology on the continen-
tal shelf of British Columbia, Canada. Geo/ogy 28:99—-102.

Pladmark, K. R. 1979. Routes: Alternative migration corridors for early man in North
America. Am. Antiquity 44:55—-69.

Frison, G., & B. Bradley. 1999. The Fenn Cache, Clovis Weapons and Tools. One Horse Land
& Cattle Co., Santa Fe, NM. 111 pp.

Goebel, F. E., W. R. Powers, & N. Bigelow. 1991. The Nenana Complex of Alaska and
Clovis origins. Pages 49-79 zm R. Bonnichsen & K. Turnmire, eds., Cloves Origins and
Adaptations. Center for the Study of the First Americans, Oregon State University,
Corvallis, OR.

268

OCEAN TRAILS AND PRAIRIE PATHS

Goldberg, P., T. Arpin, & J. Donahue. 1999. Micromorphological analyses of sediments
from Meadowcroft Rockshelter, Pennsylvania: Implications for Radiocarbon Dating,
J. Feld Archaeol. 26:325-342.

Greenman, E. F. 1963. The Upper Paleolithic and the New World. Curr, Anthropol.4:41-91.

Hamilton, T. D., & T. Goebel. 1999. Late Pleistocene peopling of Alaska. Pages 156-199
in R. Bonnichsen & K. Turnmite, eds., Ice Age Peoples of North America: Environments,
Ongins, and Adaptations of the Farst 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? Am. Antiquity 45:582—587.

Heaton, T. H. 1996. The Late Wisconsin vertebrate fauna of On Your Knees Cave,
Northern Prince of Wales Island, Alaska. J. Vert. Paleontol. 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 i'T. Akazawa &
FE. 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.

Jelinek, 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. Sence 277:71—74.

Kirillov, I., & A. Derewanko. 1998. The Paleolithic of the Trans-Baikal Area. Pages
137-274 im A. P. Derewanko, ed., The Pakeohthic of Siberia: New Dauscoveries and
Interpretations. University of Illinois Press, Urbana, IL.

Lepper, B. T. 1999. Pleistocene peoples of midcontinental North America. Pages
362-394 in R. Bonnichsen & K. L. Turnmitre, eds., Ice Age Peoples of North America:
Environment, Ongins, and Adaptations of the First Americans. Oregan State University
Press, Corvallis, OR.

Mandryk, C. A. 1990. Could humans survive the ice-free corridor’: Late-Glacial vegeta-
tion and climate in West Central Alberta. Pages 67—79 im L. D. Agenbroad, J. Mead,
& L. Nelson, eds., Megafanna and Man: Discovery of Americas 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.
5:227-278.

McAvoy, J. M., & L. D. McAvoy. 1997. Archaeological investigations of site 445X202,
Cactus Hill, Sussex County, Virginia. |crginia Department of Historic Resources, Research
Report Series No.8. Sandston, VA.

269

STANFORD AND BRADLEY

Mulvaney, J., & J. Kamminga. 1999. Prehistory of Austraha. Smithsonian Institution Press,
Washington, D.C. 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, Presentation General et Industries Paleolithiques. Fascicle 3 zm 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. Am. 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 im L. Soffer & C. Gamble, eds., The World at 18,000 BP: High
Latitudes. Unwin Hyman, London.

Robinson, S., M. Maslin, & I. McCave. 1995. Magnetic susceptibility variations in Upper
Pleistocene deep-sea sediments of the NE Atlantic: Implications for ice rafting and
paleocirculation at the last glacial maximum. Pakoceanography 10:221—250.

Severin, T. 1978. The Brendan Voyage: Sailing to America in a Leather Boat to Prove the Legend
of the Irish Satlor Saints. The Modern Library, New York, NY. 292 pp.

Schurr, T. 1996. Mitochondrial DNA and the peopling of the New World. Am. Sa.
8:246-253.

Smith, P. E. 1966. Le Solutreen en France. Imprimeries Delmas, Bordeaux. 449 pp.

Straus, L. G. 1977. Of deerslayers and mountain men: Paleolithic faunal exploitation in
Cantabrian Spain. Pages 41-76 zm L. R. Binford, ed., For Theory Buzlding in Archaeology.
Academic Press, New York, NY.

. 1979. Cantabria and Vascongadas, 21,000—17,000 B.P.: Toward a Solutrean set-

tlement pattern. Mumnib. 31:195—202.

. 1985. Stone age prehistory of northern Spain. Sczence 230:501—507.

. 1990. The Last Glacial Maximum in Cantabrian Spain: The Solutrean. Pages

88-108 zn O. Soffer & C. Gamble, eds., The World at 18,000 BP, vol. I. Unwin Hyman,

London.

. 1992. Iberia before the Iberians: The Stone Age Prehistory of Cantabrian Spain.

University of New Mexico Press, Albuquerque, NM. 336 pp.

. 2000a. Solutrean settlement of North America? A review of reality. Am.

Antiquity 65:219-226.

. 2000b. A quarter-century of research on the Solutrean of Vasco-Cantabria,

Iberia and beyond. J. Anthropol. Res. 56:39-58.

OCEAN TRAILS AND PRAIRIE PATHS

Straus, L. G., & G. A. Clark. 1986. La Riera Cave: Stone age hunter-gatherer adaptations
in northern Spain. Axthropological 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. J. Fredd Archaeol.. 9:445—-454.

Webb III, T., W. F Ruddiman, F A. Street-Perrott, V. Markeraf, 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 im H. E.
Wright Jr., J. E. Kutzbach, T. Webb III, W. F Ruddiman, F. A. Street-Perrott, & P. J.
Bartlein, eds., Global Chmates Since the Last Glacial Maximum. University of Minnesota
Press, Minneapolis, MN.

Wormington, H. M., & R. G. Forbis. 1965. Am Introduction to the Archaeology of Alberta,
Canada. Denver Museum of Natural History, Denver, CO. 248 pp.

Uf il alae: 4

“ a
Tf a vor : : - r
( Beaifiew |
@ 7 io wy ee ; arn ial oy Me
- i as is e aie ree ‘ib

‘. i
1 wien is a

vei alia ; nial wisi o aM

ti ney ie op iff ats ‘ie vi sided. bok. ic %) Aa jive t ny

oo}
7.

CHAPTER TEN

THE FIRST AMERICAN LANGUAGES

Johanna Nichols

There are about 6,000 languages on Earth, falling into about 300
separate genealogical families and a wide range of structural types.
About 150 of the families, and the full range of structural types, are

native to the Americas. That is, half of the world’s linguistic genet-
ic diversity comes from this hemisphere, a vast bank of the world’s linguistic and cultur-
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, ate 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
little 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 ort-
gin that can be traced applies not to particular languages and families but, abstractly 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

NICHOLS

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 colonization 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 colonization
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 typological comparison
and linguistic geography, for which external comparison and deep time depths are nat-
ural tasks, albeit at the cost of greatly 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, typically, 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 distinctly more than ten). To avoid circularity, a population needs to be
defined strictly 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 significantly
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 significantly exceed what would be expect-
ed by chance, and when two populations both exceed chance expectation significantly
and in the same direction (z.¢., both having either significantly more or significantly 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 hypothesize that they are due to some shared history: 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 third party. 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 affimity will be used of situations in which
two populations exhibit significant sharings of more than one structural property (and

274

THE FIRST AMERICAN LANGUAGES

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.

Linguistic Diversity in the Americas

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 term family 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 Kutenat) is its own stock. It may be useful to distinguish between a
stock’s zuternal age, the time elapsed since its initial breakup, and its external age or Lifespan,
the age up to which it can be traced without detecting external kin. It can be assumed
that any stock 1s 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-like (or Mayan-like) or about Indo-European-like
in their diversity, using well-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 1s an average and an approxima-
tion, and it will be used only for modeling and simulation, procedures in which estimates
and approximations ate 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 recently extinct) or Yukian (in the Americas; two
daughter languages).

Ie)

NICHOLS

Language stocks are not evenly distributed over the Earth. In the Americas, most of
the genealogical diversity is found in the west. In North America, two-thirds of the lan-
guage stocks are found along the western edge of the continent, between the coast and
the coast range (Jacobsen 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-diversity are not unique to the Americas but follow general principles that determine
language-family diversity. Diversity 1s highest at lower latitudes, along coastlines, and in
simpler societies; diversity 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 diversity, 1990:135.) Since the distribution of diversity 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 Age of the American 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 likely to have been impossible.

While language families vary greatly in their internal ages and in the number of their
daughter languages, their modal ages and sizes can be used for modeling purposes. The

276

THE FIRST AMERICAN LANGUAGES

stock lifespan of about 6,000 years, described above, will be used as an estimate of age.
For stocks and old families, 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
tate of post-entry proliferation for the first entrant (Nichols 2000b). Varying immigra-
tion and stock proliferation rates within naturalistic parameters made relatively little 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-
ute in this vicinity requires entry before the maximal glacial advance). Though all the
component figures are approximate, these models indicate strongly that human settle-
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. (Dillehay 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
beeline 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.

NICHOLS

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 vicinity of the entry point, its continuity 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-
ity 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 clumpiness
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 clumpiness 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 press) 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 18,000 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 entty 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 coastally adapted population as the cli-
mate improved and sea levels changed.

THE FIRST AMERICAN LANGUAGES

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-like
(2.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
mote 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 structure 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 status 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 it 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 California-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-like age, and some of the western Oregon

NICHOLS

Penutian

are California & Plateau
~~ __ Penutian

6000

2500

Tsimshianic Coast: Takelman Kalapuyan Chinookan Miwokan Costanoan Yokuts Maidun Klamath- Sahaptin- Wintun

Coosan, Modoc- Nez Perce
Siuslawan, Molala
Yakonan (Alsea)

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 by 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 typological variety 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 diversity to develop 7 s7tu.

Structural diversity, like genetic diversity, 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

THE FIRST AMERICAN LANGUAGES

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 resedual zones in Nichols (1992; accretion xones in Nichols
1997), areas (mostly coastal and piedmont areas, and regions with mild climates) in which
large language spreads are rare and diversity 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 American 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 virtually nonex-
istent elsewhere. Numeral classifiers and personal pronoun systems with first person 7
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 # : 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 # : # pronoun system is
the result of a single historical event, though the surviving evidence does not indicate

281

NICHOLS

what kind of event. Nichols (2001) argues that patterns like the 7 : ~ one are non-unt-
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) 1s
greater than the likelihood of the feature arising in the first place. If this is so, the ” : m
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 7” : 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 type, in northeastern
Asia). There are a few tokens of # : ~ systems in the Asian and Melanesian Pacific Rim
languages, but not enough to significantly exceed chance. Therefore the first 7 : #7 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
World.

The other Pacific Rim markers, however, ate 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 beyond 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 two 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 type), 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 features are well attest-
ed in Melanesia and especially Australasia. Head marking is fairly common in New
Guinea and northern Australia; prefixal agreement likewise; identical singular and plural
stems likewise; and simple consonant systems are relatively frequent in New Guinea and

ie)
(oe)
in)

THE FIRST AMERICAN LANGUAGES

neatly universal in Australia. Unlike the Pacific Rim features, these markers are common
in interior (highland) New Guinea and Australia, z.¢., farther from the (northern, coastal)
entry point than the Pacific Rim features, so that in their geography they appear to be
eatlier entrants than the Pacific Rim markets. 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 all these respects they appear to be mark-
ets of earlier colonizers, now found only at the far ends of the two trajectories.

The best interpretation of all 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 coastally 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-colonization 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 neat the coast. The coastal
Athabaskan languages and most of the California-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 & Golla 1997.) The Athabaskan family originated in the
subarctic (southern Alaska to northern British Columbia according to Leer 1991), and its
Pacific Coast languages extend like 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 (Whistler 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 settle-
ment of the interior must then have begun after the end of glaciation. The spread of the

283

NICHOLS

Clovis culture (see Meltzer, this volume) marks one of the first expansions into 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 America have
mostly 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 vicinity
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 vicinity 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 vicinity of the Columbia Plateau
(Goddard 1994), once east of the Rocky Mountains underwent considerable northward
expansion so that in historical times it occupied much of central Canada. If Leer is right
about the west central Canadian origin of the Na-Dene family, then its mostly interior
Athabaskan branch has spread far north of this origin. The only important exception to
the mostly 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 within the last millenni-
um.

These examples of northward 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-like age, ze.,
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 furthermore 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 vicinity of the Gulf of Mexico.

284

THE FIRST AMERICAN LANGUAGES

There has also been some west-to-east movement bringing representatives of coastal
and neat-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 vicinity 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, z.¢, 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 affinity 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 affinity 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 network with
a hub centered in the vicinity of the eastern Gulf of Mexico and/or the eastern

NICHOLS

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 western 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 immigta-

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.

SS Eastern North Americ
Western | | | | ——

North Mesoamerica Eastern
America

Western
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 high frequency for one feature. The length of a line

reflects distance in schematic space.

The Languages of the First Attested Americans

Unless there is inscriptional evidence of the language, there is generally no way to

know what language was spoken by a given prehistoric individual or culture. On the

THE FIRST AMERICAN LANGUAGES

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 little 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 ate 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 anywhere 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-Tlingit (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
hills 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 will 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

287

NICHOLS

Uto-Aztecan families in the historical period, 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 any 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 possibility cannot
be excluded. Similarly, it is unlikely, 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. Any
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 population 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

THE FIRST AMERICAN LANGUAGES

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 features. 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 culture 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-
tute, is likely to have lived in societies so small that intermarriage was common, and is
particularly likely to have married into a community speaking a different language from

NICHOLS

her 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 Clovis people might have several. Conversely, virtually any recent native American
language has a small probability 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 are almost
certainly not Pacific Rim languages, and descendants of the Kennewick Man’s language
could equally well belong to the Pacific Rim or eastern populations.

Conclusions

From the perspective of the method used here, the main events or situations that
have shaped the identity 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 shortly 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 (independently)
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 types, their geographical distribution, and their patterns of genetic relatedness
have been post-colonization factors: the American geography, the centers of spread and
diffusion that formed as a result of that geography, 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.

THE FIRST AMERICAN LANGUAGES

Abstract

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 were 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 immigration 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 we 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 were in America before the ancestor of
any present European language is likely to have entered Europe.

Endnotes

1 Both the immigration rate and the proliferation rate are survival rates, 2, 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
averages.

2 In their relatively high frequencies of head-marking types 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.

NICHOLS

” Nettle (1999) proposes a model whereby in a given area or continent genetic diver-
sity increases for a short time and then begins to decrease, so that less diversity points to
conducive to

ereater age. This might be correct if all continents were like Australia
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-
sity starts to decline, by the time of European contact. Even if Nettle is right that diver-
sity eventually starts to level off, it does so only in response to ecological and economic
limits that are so much mote important in determining diversity that a model based on
them alone quite efficiently predicts existing linguistic diversity.

+ Often called Na-Dene, though that label is often used of a putative larger group

also including Haida.
Literature Cited

Blust, R. 1996. Beyond the Austronesian homeland: The Austric hypothesis and its impli-
cations for archaeology. Pages 117-140 zm H. Goodenough, ed., Prehistoric Settlement of
the Pacific. Trans. Am. Philos. Soc. 86(5).

Callaghan, C. 1988. Proto Utian stems. Pages 53-57 zm William Shipley, ed., In Honor of
Mary Haas: Papers from the Haas Festival Conference on Native American linguistics, Mouton
de Gruyter, Berlin, Germany.

— . 1997. Evidence for Yok-Utian. Int. J. Am. Ling. 63:1.18-64.

Campbell, L. 1997. American Indian Languages: The Historical Linguestics of Native America.
Oxford University Press, Oxford, England. 512 pp.

DeLancey, S., & V. Golla. 1997. The Penutian hypothesis: retrospect and prospect. Inv. J.
Amer. Ling. 63:171—202.

Dillehay, T. D. 1997. Monte Verde: A Late Pleistocene Settlement in Chile: The Archaeological
Context. Vol. I. 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 history and communicative strategies in aboriginal California
and Oregon. Pages 43-64 zn O. Miyaoka & M. Oshima, eds., Languages of the North
Pacific Rim, Vol. 5. Faculty of Informatics, Osaka Gakuin University, Suita, Japan.

Jacobsen, W. H., Jr. 1989. The Pacific Orientation of Western North American Languages.
Presented at First Circum-Pacific Prehistory Conference, Seattle, WA.

Kirch, P. V. 2000, On the Road of the Winds: An Archaeological History of the Pacific Islands
before European Contact. University of California Press, Berkeley, CA. 424 pp.

292,

THE FIRST AMERICAN LANGUAGES

Leer, J. 1991. Evidence for a northern Northwest Coast language area: Promiscuous
number marking and periphrastic possessive constructions in Haida, Eya, and Aleut.
Int. J. Am. Ling. 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 City, UT. 255 pp.

Nettle, D. 1999. Linguistic diversity of the Americas can be reconciled with a recent col-
onization. Proc. Natl. Acad. Sc. 96:32-39.

Nichols, J. 1990. Linguistic diversity and the first settlement of the New World. Language
66:3.475—-521.

1992. Linguistic Diversity in Space and Time. University of Chicago Press, Chicago,
UL, S518) yoyo

——— . 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 mm 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 New 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 7 A. McMahon & L. Trask, eds., Tze Depth in Historical Linguistics, vol.

2. McDonald Institute for Archaeological Research, Cambridge.

. 2001. Why “me” and “thee”? Pages 253-276 zm Laurel Brinton, ed., Historical

Lingmstics 1999. Benjamins, Philadelphia, PA.

. in press. Diversity and stability in language. In B. Joseph & R. Janda, eds.,
Handbook of Hastorical Linguistics.

Nichols, J., & D. A. Peterson. 1996. The Amerind personal pronouns. Language
72:2.336—71.

Whistler, K. W. 1977. Wintun prehistory: An interpretation based on linguistic recon-
struction of plant and animal nomenclature. Berkey Lingwistics Society 3:157—74.

293

ie) oer — deidath taint an ead
ra sian V5 Linceaeneenenanaee > orrt me
A 1. ” ee tit SAAN Gas’

th sgt aml tal A gpa ad, ahaa fled
ana Fy Ay » ahh. via imal oto pi tate a ,
semua phi tle =, vai a ap —_- oy uy rae
| - eer i. i Gam «ll rota, ABE wait
woes a par ats re eighty ee ae

4 une sah i

aie -) may

ae aanitst ity writ): ie pun ut eel os aac Soest ee

gh) ae pie | coy ‘ie ‘me - rh ny hig
raat ee sotto waa * teen aul vat wl ants
ee ore Pisarabi saint neilegil so:seapt palette
Pu! : pea ctanaripee alae Aaa os ake Oe pik. erehegreh Lael x
ee (storied di ae Tien ee gh cai Bey,

- _ wi ih o - ae on

Pe a gl Vga is ns berwe, &
ee intone PS

aie . , .

meant as a are
mn ite Ms ‘bie pti bps y eal (so

rr
4 i
. sh Tne

ROK, ; hae ih 7 =

- 5 i “eg F are ; seal ¢ Pa ~ ¥

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-
pling of the New World, and the different sorts of evidence that
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 ef a/, (1987) and Greenberg (1986). 2) There was a single wave of migration
into the New World, followed by differentiation of languages and peoples 7 s7/u, as
argued by Merrtwether e¢ a/ (1995), Kolman ef a/, (1995), Forster et a/. (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 a/, 1996; Kolman ef a/ 1995; Neel et a/. 1994; Schurr
et al. 1999; Starikovskaya et a/, 1998).

295

MERRIWETHER

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 ef a/ 1981; Merriwether e¢ 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 haplotype. This may be either a stretch of continuous sequence of
mtDNA, or it may be a collection of restriction fragment length polymorphisms
(RELPs) from across the entire 16569 nucleotide sequence spanning the circular mt DNA
molecule (Anderson ef a/ 1981). A haplogroup is a collection of closely related haplo-
types 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 World.
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
Torront (1992) labeled these clusters A, B, C and D. These were initially defined by
RELPs 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 Nattve American groups pos-
sess this mutation. Similarly, haplogroup C was defined by the loss of a Hence II restric-
tion site at nt 13259, and haplogroup D was defined by the loss of an A/ 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 two 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 A/ I 10397 and Dde I 10394, while haplogroups C and D (and X6 and X7) cut
at both sides. A/v I 10397 1s very Asian-specific, and is found only in populations or indi-
viduals of Asian descent (including Pacific Islanders and Native Americans). The Hine
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 (Soodyall et aZ 1996;
Chen ef a/ 1995; Vigilant et a/ 1989, 1991; Redd et a/ 1995) and at least twice in Asia
(Ballinger et al 1992; Torroni et a/ 1993a, 1992; Redd et a/ 1995). However, in Asian-

296

MITOCHONDRIAL PERSPECTIVES

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 e¢ ai 2000a; Torroni ef al.
1993a; Ward e¢ af 1991; Ginther et a/ 1993; Kolman ef a/, 1995, 1996; Lorenz et al 1997;
among many others).

Which Way?

So what can mtDNA (or genetic evidence of any kind) say about these hypotheses?
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 hypotheses. Certainly
99% of all Native American DNAs, from North, Central and South America are
descended from types found in Asia. There are between 5 and 10 haplotypes 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 ef
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 ef a/ 1996 and Merrtwether & Ferrell 1996) or Brown’s X (Brown ef a/ 1998).
So the vast majority 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 fully typed
for founding lineage 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.

When?

The idea of using a molecular clock to date the migrations into the New World ts
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

DOT

MERRIWETHER

A%
B%
C%
D%
X6%
X7%

O
O
O
fi
ww
O
fal

Other%

FIGURE 1. Frequencies of the founding haplogroups 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 by errors made in repli-
cation and by environmental insults to the DNA and/or faulty repair of the damage.
These different sources of error all play 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

298

MITOCHONDRIAL PERSPECTIVES

6000 years in a population, we would expect most people to be one mutation different
from each other, but many will 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 (z.¢., a new mutant ina 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 miillion 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,

BS!)

MERRIWETHER

and D. However, Bonatto and Salzano (1997) showed that the 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 time to entry into the New World

(which could be considerably younger).
Number of Founding Lineages (and Founders)

As early as 1985, most of the founding haplogroups for the New World had been
identified (Wallace e¢ a/. 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 types. At the time, the remaining 2—5% of haplotypes were
thought to be due to European admixture. In the populations that were surveyed at that
time, this assumption was mostly true. However, in 1994, Bailliet 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 Ferrell
(1996) and Easton and colleagues (1996). The new subtypes were labeled Al, A2, B1, C1,
C2, D1, D2, X6 and X7 (Merrtwether & Ferrell 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 haplotypes described by Easton and colleagues
(1996). With the exception of Brown’s X, all are found in Asia, and even all in one coun-
try (Mongolia, Merrtwether ef a/ 1996, and Tibet, Torroni ef a/ 1994b), but most are
absent in Europe, Africa, and Australia. Merriwether has used this to argue for a single
wave of migration (Merriwether et a/ 1994, 1995, 1996, 1999) as have others (Kolman ef
al. 1995; Bonatto & Salzano 1997; Forster et a/, 1996). It must be noted the Hae II 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
hypermutability can be a useful feature when placed upon the larger background of
mtDNA genetic variation. It does mean that the sub-types, which only differ by this
marker (ze, Al vs. A2, X6 vs X7, etc.) may be different by 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 haplotype ver-
sions of Al and A2, and D1 and D2 which are also shared between the New World and
Asia (see Merrtwether ef a/ 2000 for the specific sequence definitions), which also strong-
ly supports the notion of more than one variant of at least A and D coming into the
New World in the initial wave of migration.

300

MITOCHONDRIAL PERSPECTIVES

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 haplotypes 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-
tact.

Applying the above haplotypes, 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 e¢ a/, 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 Hzncll 13259 and A/ul 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 haplotypes.
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 for Adv 1 10397 and Dde I 10394 after
learning of the X6 and X7 haplotypes from this author. Twenty were X6, 20 were X7,
none were A, B, C, D, or Brown’s X (see Hauswirth, personal communication cited in
Merriwether e¢ a/ 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

301

MERRIWETHER

Asia (using just the traditional A, B, C, D, markers plus A/ I 10397, Dde I 10394, and
Hae I11 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 Easton e¢
al. (1996) paper, and some of the corrected sequences and haplotypes appear in
Mertiwether ef a/, (2000a). The rest will follow in a future paper.

Brown’s X (Brown ef a/ 1998; Smith e¢ a/, 1999) is found widely in North America,
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 Athabaskan,
Algonquian, Kiowa-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 1s 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 ef a/ 1997, HVRbase ID # 05755; Burkhardt et a/ 1999) and
Mongolia (Kolman ef a/, 1996) that are identical to a pair of Nootka sequences from the
New World (Ward ef a/ 1991, samples NCN3 and NCN4) that are closer to the New
World variants than to the European variants of X. (See Merriwether e¢ a/ [2000] for a
discussion of shared haplotypes between Asian and the New World). They are “partial”
haplotypes because they were not sequenced or RFLP typed 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 15—30,000 years
since the migration(s) into the New World, the Brown’s X lineages have gone extinct in

Asia.

From W here?

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

302

MITOCHONDRIAL PERSPECTIVES

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. Mertiwether and colleagues
(1996) and Kolman and colleagues (1995) also suggested Mongolia as a source of the
founders. Y-chromosomal studies (Lell ef a/. 1997; Karafet e¢ al, 1999) have suggested the
Lake Baikal region of Southern Siberia, as have some mtDNA studies (Schurr e/ a/, 1999;
Starikovskaya 1998). The Wallace lab has suggested (Torroni ef a/, 1992, 1993a, 1993b,
1994a, 1994b) that haplogroup B came over more recently 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.

Problems

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 1s currently 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, Merrtwether 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 Hammet’s lab (Karafet e¢ a/ 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-

303

MERRIWETHER

ber of the Y and mtDNA haplogroups based on their differing frequencies in various
Asian and Siberian populations. Even though some populations have all the requisite
haplogroups (be it Y or mtDNA), they argue that separate waves of migration came out

of the regions where the frequency is the highest.
Ancient DNA

One important component in evaluating the peopling of the Americas from a genet-
ic perspective is ancient DNA. Ancient DNA provides the opportunity to test hypothe-
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 col-
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 ef a/ 1992;
Carlyle e¢ a/, 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 e¢ a/ 2000;
Hauswirth et a/ 1994). Much has been written about the feasibility 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 hypothetical peopling scenarios for the Americas, by telling
us what haplogroups were present when and where at different times in the past.

Conclusions
We are still 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, this information is likely to play a more and more

304

MITOCHONDRIAL PERSPECTIVES

important role in telling the story of the peopling of the New World. For now, some
issues ate still being debated, but a clear picture is emerging. These are that mt DNA 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-lineages 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 haplotypes, they prob-
ably would be considered better candidates for the source population region. So thete is
still strong support for a single wave of migration into the Americas, originating from
Mongolia.

Abstract

In this paper, mitochondrial DNA evidence for the initial peopling of the
New World is contrasted with different theories about the peopling of the
Americas derived from other types of data. While the data have been interpreted
in a number of ways, 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-
110Nn.

Literature Cited

Anderson, S., A. T. Bankier, B. G. Barrel, M. H. L. DeBulin, A. R. Coulson, J. Drouin, I.
C, 1Bjsesrora, ID) 12) INbeidtieln, 18) Jal, INOS, Jeb Swine, 12 Isk Seas, in, | Jnl, Sioauiin, Ik
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. Ham. Genet. 54:27—33.

305

MERRIWETHER

Ballinger S. W., ’T. G. Schurr, A. Torroni, Y. Y. Gan, J. A. Hodge, KK. Hassan, K.-H. Chen,
& D. C. Wallace. 1992. Southeast Asian mitochondrial DNA analysis reveals genetic
continuity of ancient Mongoloid migrations. Genetics 130:139-152.

Bonatto, S. L., & F. M. Salzano. 1997. Diversity and age of the four major mtDNA hap-
logroups, and their implicaitons for the peopling of the New World. Am. J. Hum.
Genet. 61:1413-1423.

Brown M.D! S: El. Hosseini, A’ Lorront, El |, Bandeleyiy@ Allen) deyGs Sehucemke
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, F., A. von Haeseler., & 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. ]. Phys. 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-Benetecetti, & D. C. Wallace.
1995. Analysis of mtDNA variation in African populations reveals the most ancient
of all human continent-specific haplogroups. Am. J. Hum. Genet. 57:133-149.

Denevan, W. M. 1992. The Pristine Myth: The landscape of the Americas in 1492. Ayn.
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. Am. J. 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. J. Hum. Genet. 59:935—945.

Giles, R. E., H. Blanc, H. M. Cann, & D. C. Wallace. 1980. Maternal inheritance of
human mitochondrial DNA. Proc. Natl. Acad. Sa. U.S.A. 77:6715—6719.

Ginther (Cs); Corach, G2 AS Renacino, /y7an Rey; Ew Res@arnese- sibs Einsteins
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 211—219 zn
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 University Press, Stanford, CA.
438 pp.

306

MITOCHONDRIAL PERSPECTIVES

Greenberg, J. H., C. G. Turner I, & S. L. Zegura. 1986. The settlement of the Americas:
A comparison of linguistic, dental, and genetic evidence. Curr. Anthropol. 27:477-497.

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, F 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 Kniyjff, V. Wiebe, R. C. Griffiths, A. R. Templeton, & M. E
Hammer. 1999. Ancestral Asian source(s) of new world Y-chromosome founder
haplotypes. Am. J. 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 Neobe Amerinds of Panama. Genetics
140:275-83.

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.

ees Sy DCSE Shins Ke Bakmy, YS; ees & J) Bi ees 1997) Sequence variationion miuito-
chondrial DNA control region in Koreans. Forensic Sc. Int. 87:99-116.

Lell, J. T., M. D. Brown, T. G. Schurr, R. I. Sukernik, Y. B. Staritkovskaya, 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. Ham. Bzol. 69:749—70.

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,
IDG,

Meltzer, D. J. 1993. Search for the First Americans. Smithsonian Books, Washington, D.C.
176 pp.

Meltzer, T. D., & T. D. Dillehay. 1998. The search for the earliest Americans. Archaeology
pp. 60-61.

Merriwether, D. A. 1999. Freezer anthropology: New uses for old blood. Philos. Trans. K.
Soc. Lond. B 354121129.

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.

307

MERRIWETHER

Merriwether, D. A., A. G. Clark, S. W. Ballinger, T. G. Schurr, H. Soodyall, T. Jenkins,
S. T. Sherry, & D. C. Wallace. 1991. The structure of human mitochondrial DNA
variation. |. Mo/ Evol, 33:543—555.

Merriwether, D. A., EF 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 haplotypes in Native

Americans suggests a single wave of migration for the New World. Am. J. Phys.
Anthropol. 98:411—430.

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. J. Hum. Genet. 59:204-212.

Merriwether, D. A., & R. E. Ferrell. 1996. The Four Founding Lineage hypothesis: A crit-
ical re-evaluation. Mo/ Phylogenet. Evol. 5:241—246.

Merriwether, D. A., D. M. Reed, & R. E. Ferrell. 1997. mtDNA variation in ancient and
contemporary Mayans. Pages 208—220 7 S. L. Whittington & D. M. Reed, eds., Bones
of the Ancestors: Recent Studies of Ancient Maya Skeletons. Smithsonian Institution Press,
Washington, D.C.

Merrtwether, D. A., & FE A. Kaestle. 1999 Mitochondrial recombination? (continued)
Sczence 285:837.

Merrtwether, 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 zn C.
Renfrew, ed., America Past, America Present: Genes and Languages in the Americas and Beyond
(Papers in the Prehistory of Languages). McDonald Institute for Archaeological Research,
Cambridge.

Neel, J. V., R. J. Biggar, & R. I. Sukernik. 1994. Virologic and genetic studies relate
Amerind origins to the indigenous people of the Mongolia/Manchuria/southeastern
Siberia region. Proc. Natl. Acad. Sci. U.S.A. 91:10737-10741.

Nichols, J. 1995. The spread of language around the Pacific Rim. Evol Anthro. 3:206-215.

O’Rourke, D. H., S. W. Carlyle, & R. L. Parr. 1996. Ancient DNA: Methods, progress, and
perspectives. Am. ]. Hum. Biol. 8:557—571.

Powell, J. FE, & J. C. Rose. 1999. Report on the osteological assessment of the
“Kennewick Man” skeleton (CENWW.97.Kennewick). Internet citation:
http:// www.discoveringarchaeology.com/ken-special-3.shtml

Redd A. J., N. Takezaki, S. T. Sherry, S. T. McGarvey, A. S. M. Softo, & M. Stoneking.
1995. Evolutionary history of the COII/tRNALys tntergenic 9 base pair deletion in
human mitochondrial DNAs from the Pacific. Mo/ Bzol Evol. 12:604—615.

308

MITOCHONDRIAL PERSPECTIVES

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 lineages. Am. J. Hum. Genet. 46:613-623.

Schurr, T. G., R. I. Sukernik, Y. B. Starikovskaya, & D. C. Wallace. 1999. Mitochondrial
DNA variation in Koryaks and Itel’men: Population replacement in the Okhotsk Sea-
Bering Sea region during the Neolithic. Am. J. Phys. Anthropol, 108:1-39.

Smith, D. G., R. S. Malhi, J. Eshleman, J. G. Lorenz, & F. A. Kaestle. 1999. Distribution
of mtDNA haplogroup X among Native North Americans. Am. J. Phys. Anthropol.
110:271—284.

Smith D. G., R. S. Malhi, J. A. Eshelman, & B. A. Schultz. 2000. A study of mtDNA of
eatly 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. ]. Hum. Genet. 58:595—608.

Stone, A. C., & M. Stoneking. 1996. Genetic analyses of an 8000 year-old Native
American skeleton. Ancient Biomolecules 1:83-87.

. 1998. mtDNA Analysis of a prehistoric Oneota population: Implications for

the peopling of the New World. Am. ]. Hum. Genet. 62:1153-1170.

. 1999. Analysis of ancient DNA from a prehistoric Amerindian cemetery. Philos.
Trans. R. Soc. Lond. B 354:153-159.

Starikovskaya, Y. B., R. I. Sukernik, T. G. Schurr, A. M. Kogelnik, & D. C. Wallace. 1998.
mtDNA diversity in Chukchi and Siberian Eskimos: Implications for the genetic his-
tory of ancient Beringia and the peopling of the New World. Am. J. Hum. Genet.
62:1473-1491.

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. Russ. J. Genet.
32:376—382.

Thornton, R. 2000. Population history of nattve North Americans. Pages 9—50 7 M. R.
Haines & R. H. Steckel, eds., A Population History of North America. Cambridge
University Press, Cambridge, England.

Torroni, A., T. G. Schurt, C. -C. Yang, E. J. E. Szathmary, R. C. Williams, M. S. Schanfield,
G. A. Troup, W. C. Knowler, D. N. Lawrence, K. M. Weiss, & D. C. Wallace. 1992.
Native American mitochondrial DNA analysis indicates that the Amerind and the
Nadene populations were founded by two independent migrations. Generics

130:153-162.

309

MERRIWETHER

Torroni, A., 'T. G. Schurr, M. F. Cabell, M. D. Brown, J. V. Neel, M. Larsen, D. G. Smith,
C. M. Vullo, & D. C. Wallace. 1993a. Asian affinities and continental radiation of the
four founding Native American Mitochondrial DNAs. Am. J. Hum. Genet.
53:563—590.

Torroni, A., R. I. Sukernik, T. G. Schurr, Y. B. Starikovskaya, M. EK Cabell, M. H.
Crawford, A. G. Comuzzie, & D. C. Wallace. 1993b. Mitochondrial DNA variation of
aboriginal Siberians reveals distinct affinities with Native Americans. Am. J. Hum.
Genet. 53:591—608.

Torroni, A, Y. -.S Chen, O. Semino, A. S. Santachiara-Benerecetti, C. R. Scott, 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.
54:303-318.

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.
ep ile— 199),

Tuross, N., & C. Kolman. 2000. Ancient DNA analysis of human populations. Am. J.
Phys. Anthropol. 111:5—23.

Vigilant, L., R. Pennington., H. Harpending, T. D. Kocher, & A. C. Wilson. 1989. Proc.
Natl. Acad. Sct. 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. Scence
255 1508= 050k

Wallace, D. C., Garrison, K., & W. Knowler. 1985. Dramatic founder effects in
Amerindian mitochondrial DNAs. Am. J. Phys. Anthropol. 68:149—155.

Wallace, D.C., & A. Torroni. 1992. American Indian prehistory as written in the mito-
chondrial DNA: A review. Hum. Bzol. 64:403-416.

Ward, R. H., B. L. Frazier, kK. Dew-Jager, & S. Paabo. 1991. Extensive mitochondrial
diversity within a single Amerindian tribe. Proc. Natl. Acad. Sc. U.S.A. 88:8720—8724.

INDEX

Index

agriculture 64, 163, 290
Abra, 105, WO, 1B, a9, IAG
Nasa, 9), WO), tl, 12, 16, 17, Us, 22, 2, 29, 32, CO, 13, 14, 15, 16, 132, 12D, WAG. 1174,
V7D> L1G; INS, AVA, AVA, 27, 55, 256, 27144, 216, 283), 235, V37/
Aleutian Islands 71, 132
Alexander Archipelago 76
Altai Mountains 133, 144
mimazonm ION IOT 1935200
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
eatliest 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 ZU, 230), 235, 20; 291, 295, 2916, 2D, SWO, Bil, BO
Atlantic Ocean 302
Mogsalia, 3, 68, 10), 72, 19, sil, v4, A830, 292
coastal shell middens 68, 70
colonization of 70, 72, 82
Australian Aborigines 59, 68, 81, 98, 108, 264

313}

INDEX

bathymetry 78
bears 134
Beaufort Sea 33
beetles 4, 12
eurythermic species 12
stenothermic species 12
laborer 4h). IO), til, W4b WS), he, 40, 30, 52), M0), 12), 15; 12, toil, W222), ais, ilsvh A 276.
280
archaeological evidence from 20, 21
climate 1113
ecosystems 17
human entry into 19, 22, 59, 134
levarel |ovatelexe 45, 110), 3) kS, QZ, (lO), Wail, sy, Sh, AY, MA), ISD), ZS)
megafaunal mammals 15, 17
paleobotanical studies 14
paleoenvironmental conditions 15, 123
paleontological sites 21
vesetation 15s lGs 1775 los 19
Josle eahaars lorbuayataver ©), 20), Zier, Zo), WS), OS), NO, WS), WO), West, ANZ, AVS, 230, 247
birds 177, 184, 188, 246, 262
bison 17, 20, 74, 144, 162
Bluefish Caves 20, 175
lxozid@) 5, 19) GO, GS, Ger, 10), 7D; 75, 16, Sil, 1S, 4s, IAN, WS, ZG, ZI, AOS, 264, 265,
D5
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

314

INDEX

Cactus Hill 43, 45, 52, 177, 259, 260, 265
Calico Ishillg 29), 173
California 76, 77, 283, 284
COASE Or 1/7
camel 162
CanacdaZss283n295
canonical variate analysis (CVA) 95, 105
catibou 17, 20
Caverna da Pedra Pintada 193, 194, 195, 197, 198, 199, 200, 231
artifacts 197
dawes 196, 199, AO, AV, ZI, 22, 23S), 234, 23S
pigment from 198
CentwaleAniericanlS0s 20202745 276s 2855 2964302
Channel Islands 67, 77, 289
Channel Islands Woman 287, 289
language of 287
Chatelperronian complex 83
Cale 72, 17, 1ST, 231, 253, 220, 222, ZV
Clara 129, 130
clothing 145
Clomis 3, 6, 21, 32, 33, 36, 31, BS, BS, 34, V2, WSs 14, VS, 79, 14, 1A, 159, G2, 163,
17d, 17D, UST, isto, AO, 20, 203, ZB, 255; 251, 259, AOI, 266, ZA, 2ST, 288, ZOD
AAS Ot IOI, WO 165, LOG, IG, G69, IO, AO, AVil, 20s, 222 22S, ZO, 227, V8.
2D IIIS)
assemblages 30, 45
blade core technology 255
Cacacs 37, 33, 205
Clovis barrier 60
demographic costs 34
KRuropean 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

INDEX

Iberian Peninsula 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
otigins 6, 29, 43, 44, 255, 256, 257, 266
population growth rates 46, 50
progenitors 43
projectilespoimt 28) 395405 4352, 13a) W2A 2S IA SOS NO2e 707 Os eile
187, 204, 255, 261
radiation 46
radiocarbon ages 50, 165, 168, 169, 200
reproductive isolation 46
risk of extinctions 34
settlement mobility 36
Ses 73, 159, NGO, 1G, 169
social networks 37, 39
South America 159, 183
technique 30, 204
territorial behavior 37
tool complex 3
tool-kits 28, 29, 31, 170, 182
variability in 31
Clovis-fitst 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
consuiines O35, OF, 10), 13, 19) Sil, 20S, 276, 2Zsith
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

INDEX

colonization process 36, 47
Commander Islands 71
COmuaneamll Sache OS, 73, Sil, IS2, 25, AOS
cooking 133, 145
Cura 5, 99, 9D, ICO, NOL, MOA, OS, 15, 126, 138, 170, 1172
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

reliability 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
Sinodontyal Zo. S15 1355 [Bie 172
Sundadonty 125, 131, 135, 137, 147, 190
supernumerary roots 129
traits and their frequencies 131
Uto-Aztecan premolar 128
Zhoukoudian remains 126

dental relationships 130

dental trait frequencies 138

dental wear 135

Devil’s Tower 65

INDEX

digging sticks 243

diseases 48

Diuktai culture 129

dogs 5, 145, 146

domestication 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
Sn AO OS, 77, 1s, SO, 3, Gil, 163, Wes eo, 202, 247, Wor
fishing 63, 142, 145, 203, 262
fluting technology 30
folk taxonomies 50
ln@isora BW, Ste, 1, 14, 17, USS, ZOO, 2G
complex 41
points 75
food 20
animals 20
habits 20
food storage 247
iomenne 6, 29, 35, 145 Gl, 163, 170, 17S, 179, 193, 164. 185, 1a, 20
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

INDEX

gene flow 46, 116, 118, 119
genetic data 297

genetic diversity 280, 292
genetic drift 114, 116, 118, 119, 137
genetics 47, 142, 148, 274
geochronology 4

giant short-faced bear 18

glacial advances 11

glacial periods 11

glacial refugia 74

Gorham’s Cave 65

grazing mammals 17, 18

Great Basin 284

GreenlancdiS2Z Noon 1264243
Greenland ice sheet 50

erizzly bear 18

Groundhog Bay 76

handaxes 44
harpoons 67, 263, 264
heat 22
soutce of 22
Homo erectus 61, 67, 80, 83
Homo sapiens 61
anaclagie Oil, O7/
Homo sapiens sapiens 61, 63, 68, 71, 80
horse 17, 162
horticulture 247
Howieson’s Poort 62
Hrdlicka 1, 93, 114
model of 94
human fossil record 93, 95, 102, 120, 132, 174, 180, 189, 190
North American 93, 97, 111

319

INDEX

South American 107, 108, 111, 185
hunter-gatherers 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 slaaeis 9, 19, 73, 295

ice wedges 10

ieee Corcatelore 19), 277, 43). 4) 59). G0), 72, 12, 144, 13, 15 30, 159, 17/2, 255, 256, 295
environmental conditions of 74, 256
tools found in 256

insects 4

interglacial periods 11

Ireland 263

Jajoein 10), Tl, Sil, 3), IOS, WY), 129), 7, 27
maritime peoples 71
Miyako Island 71
Okinawa 71

Jomon 105, 110, 124, 138

IKkamchatka Peninsula 71
Katanda 67
Kennewick skeleton 46, 94, 102, 105, 135, 287, 289

320

INDEX

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 Archipelago 75
Kodiak Island 75
Korea 305

Wakeppailcalits OMISSIIBb7 9295s 303
Lake Mungo 68
land bridge 22, 132, 255, 256
landscape learning 4, 32, 34, 36, 43, 47
language families 146, 273, 274, 275, 276, 277, 281, 284, 285, 287
AGETONZTO» 219)
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
Renwtdany2 799250) 23835237
Pomoan 287
Salishan 285
Siouan 284
Siouan-Catawba 284
‘Tsimshianic 279

INDEX

Uto-Aztecan 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
AS Ost ZIKO, YOM
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 cultural 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 diversity 273, 275, 276, 281
in the Americas 275, 281
linguistic history 273
of the Western Hemisphere 273, 283

INDEX

linguistic interrelationships 286
linguistic variation 141
linguistics 47, 146, 147, 148
lion 18, 134

loess 10

Lost Tribes of Israel 4

onvroavonondas I, Its, 29. Sis), WAG IG. I a/a/
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
raneeaneaerovn 15. ICT, Ish, AAO, Zen ANG 1S, So WS), Nay, AT, MelO), Mya, sh, MIS, TsO, IY
megafaunal extinctions 18, 19, 20, 23, 28, 42
Migbnesa OF, G8, 10, 19. Sil, 62,03, Zils, Zs
Mesoamerica 179, 285, 286, 290
Mexico 180, 274, 276, 284, 285
microlithic tradition 124
Middle Stone Age 62
raaneneanorns 19) S10), cA, We, SYA, 1135), ists, USO) OIL, AVI, 299), 3100), 305
by sea 80, 82
fromiuuropel lol; 295
multiple 7, 304
MUMDeKSLOle A+ os IS ou IAI IAS 2850295
single 295, 305
theories 161

INDEX

timing of 162
Milankovitch cycles 11
Minnesota 100, 101
mitochondrial DNA 61, 62, 123, 141, 172, 260, 295, 299, 305
ancient 302, 304
deletions 297
divergence times 300
genetic 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 lineages 302, 303
molecular clock 297, 298
mollusks 177, 246
Mongolia 130, 133, 295, 300, 303, 305
Mongoloid dental complex 135
Mioraie INieere OZ, IIL, WZ, A200), A011, 2a!
Monte Verde 6, 195 30, 48, 49° 60) 72, 73, 77, 132, 186, 187, 238; 240, 243, 245, 249°
DINO, ZV, 2S, ASS 2D
BIS Oe 17, Ms, WUS7/, Zab
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

324

INDEX

Mutual Climatic Range (MCR) method 11
estimates 13

Nanchoc Culture 246
Native American 111, 119) 120, 297, 299
ancestral 125
gene pool of 111
within-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
INenanaiRiver! 20% 725 73; 74, 177
Nevada 99, 100, 101
New Guinea 70, 79, 81, 280, 282
New Jersey 44
New Mexico 3

@ceaniars9e5 0)

Ohio 44

Okinawa 71, 83, 137
shell middens 71

oral traditions 273

Oregon 76

ornaments 63

overshot flaking 261
P

Paleo-Eskimo 32, 33

(SS)
No
Ol

INDEX

Paleoindian 9, 20, 37, 38, 73, 77, 97, 98, 105, 118, 159, 160, 161, 162, 168, 174, 182,
196, 201, 204, 205
antiquity 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
lithic complexes 179, 180, 184, 185
maritime 77
migration 9, 164
mobility 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) 995 11095 170; 172
southern cultures 162, 188
HOOKS 170), WS, Wes, 196
Paleoindian Period 9, 51
paleotemperature estimates 11, 12, 14
pathogens 48
permafrost 22
permafrost features 10
Peru 78, 183, 184, 186, 237, 238, 241, 246, 249
Plainview 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

326

INDEX

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
jexxe-Clommls 2, 4, AD, SO, SI; ais, Ahh, 2) 52, 72, WS, 12S, Sil, 4, SYD, SD), UGA, eh, 17
UT, UD, Moi, WB, WIS, Wee, 137, 202, 204, 205, 260, 295
Beringian crossing 134
Glatres 2D), OA, WIE WSS WT, WSS, M7, US, ZED)
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

eachiocamdorom Sil, Aq, Wi, IA, MOS. WOWs IZ WIS, IG, West, ieshSy, itis, We), GY), 20010), ZSw)
calibration 31, 49
curve 31
standards of dating 165

faimtorest loos IN 5s L9ON202

INDEX

adaptations 173
Ramapithecus 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

sepnievatae 59), G0) OS), Oss 10, Vil, 1, 9,
antiquity 68, 77
archaic Homo sapiens 68
Homo erectus 68

seamanship 265

Semliki River 62

Seward Peninsula 15

shamans 130, 243
drum 130

shell middens 77, 78

shellfish 65, 67, 77, 78, 161, 163, 184, 202

Swlorssatal 5), 9), NO, AO), i223), Ws WAS 110), BH, Se AO), ISO, 27ZE 26, ZO, Zi, QOS, 299).
0M, S103}
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

328

INDEX

African 108
Australian 99, 101
craniofacial shape 46, 95, 98, 110, 116, 126
Early Archaic 97
food preparation techniques 126
[ate 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
Sourln Avanearen WOW, Will, 12, Wilts)
southern Asians 101
Skhul 62
small game 6, 20, 161, 163, 177, 187, 189, 202, 203, 242, 246, 248
social networks 47, 48
Solomon Islands 70
Solumeanetowlolen oO IiOE2595 26052 612625265
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
Spisite Cave I45 99F 1255139.
dental morphology 125
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
(eyo) Wi a'4 OX 2, Ba VATE 23 ole a0)
unifacial industries 179, 238, 246
subsistence patterns 237, 249, 250
Sundaland 139

INDEX

Taima Taima 182, 205
Taiwan 71
taphonomy 163
teeth 5, 123, 1125. 127, 136, 140; 1485 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 crossing 52, 131

UO)

Upper Paleolithic 43, AE) 3), BO), KOS, 25x)
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
antiquity of 100
dental morphology 125

wolves 32

woolly mammoth 17, 20, 27, 28, 42

extinction 18

330

INDEX

Wrangel Island 18, 19
woolly rhinoceros 17

Y chromosome 123, 303, 304
Younger Dryas 16, 23, 50

Valor Newatworny O, Wil, WA 7, Wes 7
ZL,

Zhoukoudian 126, 127

33)

i
iT ;
T
i
i
i
i
i
u 1
i i
\ 1
i
im
i
i Af
i
: i
\ if
4 ‘
ee
he i GE F
1
{
vie
te
it eT,
{ -
enya
.
af * ‘ ,
.
op A
~
i 1
1 |
i
i
moo
i
ii
i
i
i
i
i
i
i
i
ell
ui
" ri
* ir
5,
nae hal
A fans aU
i
{
i
, n
i
i ‘i
7 i}
“t =
i
i
i i! A
7 ri if
i
i
i
i
iy 1
i
0 1
il
n
i tr
- 7 tuaeyy
i
fe
: nt: i
i
7 t) 1
all
1 Sait
Dhaiaah

{
i ’
i
i i
Ki
ul
i
nai

i
ean
i
i
Ter
nl
i
{
ree
i
i
ve

PES ca 1%

iM ee ft Mink rats
FS OA ange

() be? pee te

oe 7

THE FIRST AMERICANS

THE PLEISTOCENE COLONIZATION
OF THE NEW WORLD

INTRODUCTION
CHANGING PERSPECTIVES OF THE FIRST AMERICANS: INSIGHTS GAINED
AND PARADIGMS LOST, BY NINA G. JABLONSKI

UNIVERSITY OF CALIFORNIA PRESS
PRINTED IN THE

UNITED STATES, OF AMERICA

23Te* al

273.

SETTING THE STAGE: ENVIRONMENTAL CONDITIONS 1N.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 DAVip J. MELTZER

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 123

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 Bhan AND tts IMPLiaMTONS 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, ~ 255
BY DENNIS STANFORD AND BRUCE BRADLEY 2

THE FIRST AMERICAN LANGUAGES, BY JOHANNA NICHOLS

"A MITOCHONDRIAL PERSPECTIVE ON THE PEOPLING. OF THE NEW WORLD, 295
BY D. ANDREW MERRIWETHER

NINA G. JABLONSKI 1S, THE IRVINE CHAIR AND CURATOR OF ANTHROPOLOGY AT THE

CALIPORNIA ACADEMY. OF SCIENCES.

DistRIBUTED BY THE ISBN O- i 50-5

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

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

See other formats