On the Road of the Winds: An Archaeological History of the Pacific Islands before European Contact, Revised and Expanded Edition

By Patrick Vinton Kirch

The Pacific Ocean covers one-third of the earth’s surface and encompasses many thousands of islands that are home to numerous human societies and cultures. Among these indigenous Oceanic cultures are the intrepid Polynesian double-hulled canoe navigators, the atoll dwellers of Micronesia, the statue carvers of remote Easter Island, and the famed traders of Melanesia. Decades of archaeological excavations—combined with allied research in historical linguistics, biological anthropology, and comparative ethnography—have revealed much new information about the long-term history of these societies and cultures. On the Road of the Winds synthesizes the grand sweep of human history in the Pacific Islands, beginning with the movement of early people out from Asia more than 40,000 years ago and tracing the development of myriad indigenous cultures up to the time of European contact in the sixteenth to eighteenth centuries. This updated edition, enhanced with many new illustrations and an extensive bibliography, synthesizes the latest archaeological, linguistic, and biological discoveries that reveal the vastness of ancient history in the Pacific Islands.

• Language: English
• Publisher: University of California Press
• Release date: Nov 7, 2017
• ISBN: 9780520968899

Patrick Vinton Kirch

Patrick Vinton Kirch is Class of 1954 Professor of Anthropology and Integrative Biology at the University of California, Berkeley. He is the author of many books, including Feathered Gods and Fishhooks and On the Road of the Winds (UC Press).

Book preview

On the ROAD of the WINDS

Restored Moai (statue) on Rapa Nui, with replicated obsidian and coral eyes. Photo by Thérèse Babineau.

On the ROAD of the WINDS

An Archaeological History of the Pacific Islands before European Contact

Patrick Vinton Kirch

REVISED AND EXPANDED EDITION

UC Logo

UNIVERSITY OF CALIFORNIA PRESS

University of California Press, one of the most distinguished university presses in the United States, enriches lives around the world by advancing scholarship in the humanities, social sciences, and natural sciences. Its activities are supported by the UC Press Foundation and by philanthropic contributions from individuals and institutions. For more information, visit www.ucpress.edu.

University of California Press

Oakland, California

© 2017 by Patrick Vinton Kirch

Library of Congress Cataloging-in-Publication Data

Names: Kirch, Patrick Vinton, author.

Title: On the road of the winds : an archaeological history of the Pacific islands before European contact / Patrick Vinton Kirch.

Description: Revised and expanded edition. | Oakland, California : University of California Press, [2017] | Includes bibliographical references and index.

Identifiers: LCCN 2017012519 (print) | LCCN 2017015961 (ebook) | ISBN 9780520968899 (ebook) | ISBN 9780520292819 (pbk. : alk. paper)

Subjects: LCSH: Prehistoric peoples—Oceania. | Oceania—Antiquities.

Classification: LCC GN871 (ebook) | LCC GN871 .K573 2017 (print) | DDC 995—dc23

LC record available at https://lccn.loc.gov/2017012519

Manufactured in the United States of America

27  26  25  24  23  22  21  20  19  18

10  9  8  7  6  5  4  3  2  1

FOR DOUGLAS E. YEN

WHOSE OUTSTANDING ETHNOBOTANICAL RESEARCHES CONTINUE TO INFLUENCE INTERPRETATIONS OF PACIFIC PREHISTORY

CONTENTS

List of Maps

List of Figures

List of Tables

Preface

INTRODUCTION

Defining Oceania • Linguistic, Human Biological, and Cultural Variation in Oceania • About This Book • A Note on Dates and Time

1 • DISCOVERING THE OCEANIC PAST

Enlightenment Voyagers • Outposts of Empire: Missionaries, Colonists, and Academic Beginnings • The Problem of Polynesian Origins • Te Rangi Hiroa and the Micronesian Route to Polynesia • The Discovery of Time Depth and Culture Change • The Search for Polynesian Sequences • Broadening Research Horizons • Moving beyond Polynesia: Archaeology in Melanesia and New Guinea • Francophone Archaeology in the Pacific • Not an Ivory Tower: Public Archaeology in the Pacific • Recent Advances in Pacific Archaeology

2 • THE PACIFIC ISLANDS AS A HUMAN ENVIRONMENT

Origins and Development of the Pacific Islands • Types of Islands • Climatic Factors in the Pacific • Island Life and Biogeography • The Microbiotic World and Human Populations • Island Ecosystems • Humans and Island Socioecosystems

3 • SAHUL AND THE PREHISTORY OF OLD MELANESIA

The Pleistocene Geography of Sahul and Near Oceania • Initial Human Arrival in Sahul and Near Oceania • Pleistocene Voyaging in Near Oceania • Near Oceania during the Pleistocene • Cultural Innovations of the Early Holocene • A Paradox and a Hypothesis

4 • LAPITA AND THE AUSTRONESIAN EXPANSION

The Human Landscape of Near Oceania at 2000–1300 b.c. • The Advent of Lapita • Lapita Origins: The Austronesian Expansion • Lapita across Time and Space • Lapita in Linguistic and Biological Perspective • The Lapita Ceramic Series • Lapita Sites and Settlements • Lapita Subsistence Economies • Exchange between Lapita Communities • Ancestral Oceanic Societies

5 • THE PREHISTORY OF NEW MELANESIA

Trading Societies of Papua and the Massim • The Late Holocene in Highland New Guinea • The Bismarck Archipelago after Lapita • The Solomon Islands • Vanuatu • The Polynesian Outliers in Melanesia • Ethnogenesis in La Grande Terre • Fiji: An Archipelago in Between • Larger Themes in Melanesian Prehistory

6 • MICRONESIA: IN THE SEA OF LITTLE LANDS

Colonization and Early Settlement in Micronesia • Cultural Sequences in Micronesia • Tuvalu and the Polynesian Outliers in Micronesia • Atoll Adaptations • Later Prehistory in Western Micronesia • Development of Sociopolitical Complexity in the Caroline High Islands

7 • POLYNESIA: ORIGINS AND DISPERSALS

Polynesian Origins • Polynesia as a Phyletic Unit • Ancestral Polynesia • Cultural Sequences in Western Polynesia • The Settlement of Eastern Polynesia • Early Settlement Sites in Eastern Polynesia • Polynesian Voyaging • Summary

8 • POLYNESIAN CHIEFDOMS AND ARCHAIC STATES

Polynesian Chiefdoms: Ethnographic Background and Anthropological Significance • The Traditional Atoll Societies • Sociopolitical Transformation in the Open Societies • The Emergence of Stratified Chiefdoms • From Chiefdom to Archaic State: Tonga and Hawai‘i • Polynesian History: A Concluding Note

9 • BIG STRUCTURES AND LARGE PROCESSES IN OCEANIC PREHISTORY

Voyaging and the Human Conquest of the Pacific • History Written in the Present: Correlations between Language, Biology, and Culture • The Role of Demographic Transitions in Oceanic History • Oceanic Populations on the Eve of European Contact • The Political Economy of Dynamic Landscapes • Intensification and Specialization in Island Economies • Transformations of Status and Power • On Comparison: A Closing Comment

Notes

References

Index

MAPS

1. Oceania, showing the traditional cultural regions of Polynesia, Melanesia, and Micronesia

2. The distribution of the Austronesian and Non-Austronesian languages in Oceania

3. A tectonic map of the Pacific basin

4. The dominant wind and current directions in the Pacific Ocean

5. Approximate configuration of Sunda and Sahul during the Pleistocene

6. Near Oceania, showing the distribution of important Pleistocene and early Holocene archaeological sites

7. The distribution of Lapita sites in Near and Remote Oceania

8. Island Melanesia, excluding Fiji, showing the location of key archaeological sites

9. The Fiji archipelago, showing the location of key archaeological sites

10. The islands of Micronesia

11. The islands of Western Polynesia, showing the location of key archaeological sites

12. The central Eastern Polynesian archipelagoes and islands

13. Rapa Nui (Easter Island), showing the location of key archaeological sites

14. New Zealand, or Aotearoa, showing the location of key archaeological sites

15. The Hawaiian Islands, showing the location of key archaeological sites

FIGURES

TABLES

PREFACE TO THE FIRST EDITION

Born and raised on one Pacific island, I have spent three decades of my adult life exploring, living on, studying, and seeking to understand a plethora of others, especially the history and culture of their inhabitants. While my own ancestry is ultimately European, I sometimes feel that the Pacific is in my blood. This is, naturally, but a metaphor. Yet the many years spent living and working with Pacific islanders, learning to speak more than one of their indigenous languages, and adapting my behavior to fit their cultural canons have—I hope—engendered a certain empathy beyond that typically associated with dispassionate social science. This book is my attempt to distill a lifetime of study—and the insights gained through much tedious sifting of often minute strands of evidence by myself and innumerable colleagues—into a coherent whole, a synthesis for those whose curiosity would take them on an intellectual voyage into the Oceanic past.

Mine is but one of many constructions that could be made from the formidable array of archaeological, historical linguistic, ethnographic, and human biological evidence concerning the Oceanic past, assembled by scholars over more than two centuries. It is necessarily personal in the emphasis I accord particular times, places, and concepts. Despite my acceptance of the idea that history (or prehistory) is constructed rather than reconstructed (as might have been said three decades ago, during the heyday of the New Archaeology), I do not subscribe to the ultrarelativist stance of some late-twentieth-century social scientists. Although each new generation of archaeologists and prehistorians inevitably rewrites the past in their own terms, I prefer to see this, in Paul Veyne’s (1984) words, as a process of lengthening the questionnaire, of constructing multiple, rather than single, historical plots.

With each new generation the empirical base of archaeology advances, constraining what the next generation of scholars may construct of the past. Thus I agree with Ernst Mayr (1997:83) when he claims that science does advance, despite false starts and wrong-minded diversions along the way. Just as Mayr can rightly claim that our understanding of the cell in biology has truly progressed since its first recognition by Robert Hooke in 1667, so our knowledge of Pacific archaeology and long-term history has without question improved throughout the course of the nineteenth and twentieth centuries. The many surveys and excavations of Pacific archaeological sites, enhanced by the patient laboratory analyses of artifacts and other material traces, increasingly improve our ability to come to grips with the human past of Oceania. That does not mean that our current knowledge and understanding constitute truth in any absolute sense; it does mean that our models of what happened in the past are now a better approximation of that historical reality. Was there a real past? Yes, but we shall never know it definitively, as it exists only as material traces in the present. Thus for epistemology.

Peter Bellwood authored the last attempt at a general synthesis of Oceanic prehistory two decades ago: Man’s Conquest of the Pacific (1979). Still useful as a guide to the literature and the interpretations of its time, Peter’s book is long out of date, an encouraging sign of how our knowledge base has increased; it is long out of print as well. Despite the obvious need for a suitable replacement for his opus, I long refrained from undertaking such a work myself. Once committed to the task, however, I found it both stimulating and challenging to grapple with the problems of condensing such a broad field, one that has in recent decades grown by leaps and bounds. To compress within the covers of a 448-page book the panoply of archaeological and anthropological minutiae with which the scholar ordinarily concerns himself is impossible. How to sort and prioritize? What is truly significant? And what can be eliminated? Asking these questions forced me to confront more fundamental issues: underlying assumptions about the very nature of the field, what we think we know, the often-unstated yet subtle organizing categories by which we conduct our research. In the end, the act of writing this book proved far more intellectually engaging than at first I imagined it might be.

I first participated in Pacific archaeological fieldwork thirty-five years ago, as part of a 1965 Bishop Museum team investigating rockshelter sites in Kona and Ka‘u, Hawai‘i, and I published my first professional article in 1970. When I came into the field as a young student in the mid-1960s, Jack Golson, Roger Green, Doug Yen, and others in Pacific archaeology were inspiring a renaissance in perspective and approach, as well as moving beyond (while not dispensing with) questions of cultural origins and migrations. Their work encompassed the then-new settlement pattern approach, questions of human ecological adaptations, and problems of sociopolitical evolution in island societies. Since then, I have been privileged to participate in many key developments and advances in Pacific prehistory, such as those that came out of the Southeast Solomons Culture History Program of the 1970s and the Lapita Homeland Project of the 1980s. Fieldwork has taken me across the breadth of the Pacific, from the Mussau Islands of Papua New Guinea to Rapa Nui, with stays between these geographic extremes in Palau (Belau), Yap, Majuro, Arno, Kolombangara, Nendö, Vanikoro, Tikopia, Anuta, Futuna, Alofi, ‘Uvea, Niuatoputapu, Vava‘u, Ofu, Olosega, Ta‘u, Mangaia, and all of the main islands of Hawai‘i except Ni‘ihau. My research has encompassed high islands and atolls, islands tropical and sub-tropical, ones large and small, and spanned the classic ethnographic regions of Melanesia, Micronesia, and Polynesia. I trust it not untoward to claim that this diversity of field experiences lends me sufficient geographic and cultural background with which to undertake the synthesis attempted here, incorporating as well the work of a great many other field researchers. The comparative approach has long been a cornerstone of anthropology, and the opportunities to compare the archaeological records of so many diverse islands have been inspirational. Still, this book is as much a synthesis of the work and insights of innumerable others, some of whom I have been privileged to know as colleagues and friends, others only as academic ancestors whom I have met through the legacy of their scholarship.

Organizing this book—structuring it into chapters and subheadings—was perhaps the most agonizing part of writing. Early on, I experimented with less conventional schemes, with various topically organized outlines. In the end, I rejected these approaches in favor of an outline based partly on time and largely on space. Though more conventional in structure, this plan in my view allows for better presentation of the facts of prehistory as we currently conceive them, enabling them to be more readily accessed by students as well as other professionals seeking an introduction to the long-term history of the Pacific. But I am fully cognizant that facts exist—or at least can be interpreted—only within the context of theory, and I have tried to relate archaeological evidence to contemporary theoretical issues and debates wherever possible. My final chapter is an attempt to canvass what I see as some of the grand themes in the prehistory of the Pacific, issues that in the end transcend place and time, and go to the core of anthropology and history at large.

For more than twenty years I have taught Pacific archaeology and prehistory to undergraduates and graduate students at the Universities of Hawai‘i, Washington, and California at Berkeley. Admittedly frustrated by the lack of a suitable introductory text that covered the entire Pacific field, I had not thought of writing such a book myself. Too many other high-priority projects were always to the fore; the undertaking of a general synthesis seemed something that could wait for a more leisured time. William Woodcock, formerly of Princeton University Press, convinced me otherwise, and without his urging I would still be thinking of this as a far-off project. Still, it took four years before I was able to find the appropriate time to concentrate on its writing.

A year-long fellowship (1997–98) at the Center for Advanced Study in the Behavioral Sciences (CASBS), Palo Alto, provided invigorating freedom from normal academic duties. I am grateful for the financial support provided by the National Science Foundation (Grant No. SBR-9601236), which partially supported my fellowship at the Center, and to the University of California at Berkeley for the grant of a year’s sabbatical leave. I especially thank Neil Smelser, director of the CASBS; Robert Scott, associate director; and their wonderfully supportive staff. Librarians Joy Scott and Jean Michel cheerfully tracked down obscure references and obtained rare volumes through interlibrary loans. Virginia MacDonald graciously word-processed my editorial corrections to several drafts. Susan Beach’s delicious lunches helped too! And the stimulating intellectual atmosphere generated by my fellow colleagues of the Center’s Class of 1997–98 made the experience especially memorable.

Several colleagues were generous enough to read and comment on one or more draft chapters. I especially thank Roger Green and Kent Lightfoot for reading the entire manuscript. Jim Allen, Steve Athens, Chris Gosden, Laura Nader, Barry Rolett, Matthew Spriggs, and David Steadman commented on specific chapters. They have saved me from making many errors, and I am most appreciative of their collegial efforts. I also thank the following colleagues for graciously giving me permission to use their photographs or illustrations: Jim Allen, Wal Ambrose, Steve Athens, Janet Davidson, John Flenley, José Garanger, Jack Golson, Roger Green, Geoff Irwin, Pat McCoy, William Morgan, Barry Rolett, Christophe Sand, Yosi Sinoto, Jim Specht, Matt Spriggs, Robert Suggs, Joanne Van Tilburg, Paul Wallin (for the Kon-Tiki Museum), Marshall Weisler, and Peter White.

At the University of California Press, Director Jim Clark and Executive Editor Doris Kretschmer enthusiastically accepted my book manuscript and made its production a high priority. I also thank Nicole Hayward for her superb design and Peter Strupp for his meticulous copyediting.

It gives me distinct pleasure to dedicate this book to Douglas E. Yen, who over three decades has been by turns mentor, co-fieldworker, and colleague, as well as friend and confidant. We have shared never-to-be-forgotten field experiences in Makaha, Hālawa, Kolombangara, Anuta, Tikopia, and Kahikinui, as well as evenings at various watering holes from Nanakuli to Honiara. Doug’s ceaseless adherence to the sound scientific principle of always being alert to the unsuspected alternative hypothesis has more than once been an inspiration. His influence on my generation of Pacific archaeologists has been legion.

PATRICK VINTON KIRCH

Palo Alto

PREFACE TO THE REVISED EDITION

On the Road of the Winds, written nearly twenty years ago, introduced the field of Pacific archaeology and prehistory to a new generation of students and has provided a reference source for scholars and the general public. The book was well received. Norman Hammond, writing in the Times Literary Supplement (February 23, 2001), opined, Comparing Kirch’s excellent book and its scores of new sites, hundreds of radiocarbon dates and bibliography of more than fifty pages, with [previous works] . . . we can see that the fundamentals have not changed much, but that the depth and texture of our understanding of the voyagers on the road of the winds have immeasurably increased. Peter Bellwood, reviewing On the Road of the Winds in the leading journal Antiquity, called it a grand synthesis, adding that Kirch has done Pacific archaeology proud with this book.

A few years ago, however, I began to hear from colleagues who regularly used On the Road of the Winds as a university text that they were spending too much time in their courses explaining where parts of the book had been superseded by new discoveries or revisions of older theories. This did not surprise me, for with the passing of two decades there is no question that our knowledge base and understanding of the ancient Pacific has advanced enormously. In late 2014, with retirement from active teaching and the seemingly endless demands of university service pending, I canvassed nine of these colleagues regarding a possible revised edition. The uniformly enthusiastic response encouraged me to approach my then-editor at the University of California Press, Blake Edgar, with a proposal for a revised edition.

In my proposal, I enumerated a few of the major advances in Pacific historical anthropology that called for a revised edition:

• The discovery of new Lapita sites had extended our knowledge of the geo-graphic range of Lapita (into the southern coast of New Guinea, for example) as well as the nature of the Lapita culture itself, such as the use of ceramic vessels for burials at the Teouma cemetery site in Vanuatu.

• A plethora of new genetic data from both mitochondrial DNA and Y-chromosome analyses of Pacific islander populations had illuminated the biological complexities of early human migrations into the region.

• Extensive fieldwork on the enigmatic terraced landscapes of Palau had elucidated the chronology and function of these massive earthworks.

• The development of high-precision uranium/thorium dating of coral artifacts had pinned down the timing of initial Polynesian settlement as well as revealed the chronologies for temple building in Tahiti and Hawai‘i.

• The previous debate about long versus short chronologies for the expansion of Polynesians into the central and eastern Pacific had been resolved, including the timing of human settlement on Easter Island and in Hawai‘i.

• There was new evidence for contact between Polynesia and the Americas.

• New theoretical work placed the late precontact Hawaiian and Tongan societies among a handful of archaic states that arose in various parts of the world.

• The exciting integration of research in island ecology with archaeology had helped to advance the new discipline of human ecodynamics, in which the Pacific islands as model systems played a key role.

• There was an ongoing debate about whether Easter Island represented a case of societal collapse or a model for indigenous resilience.

These are, of course, only a few examples of how Pacific archaeology had changed and moved forward over the past two decades.

Only after Edgar and UC Press enthusiastically accepted my proposal and issued a contract did I begin to realize how much work I had committed myself to. Even though I have followed the advances in our field over the past two decades—indeed, I have directly participated in some of them—it has been a daunting exercise to compile, review, and synthesize the enormous amount of new research. Literally hundreds of journal articles and book chapters and dozens of new monographs, edited volumes, and books had to be read and absorbed. The task consumed a great deal of my time throughout 2016, prompting my wife, Thérèse, to muse that I lived in my study. Fortunately, the freedom of retirement allowed me to devote uninterrupted periods to the project, much as the freedom of an academic sabbatical had allowed the first edition to be completed in 1997–98. I am quite certain that I will not undertake yet another revision twenty years hence. I bequeath that task to some younger scholar!

I am especially grateful to a number of colleagues who sent me their publication lists and pdf copies of their articles and graciously allowed me to use their illustrations in this revised edition. In particular, I thank Jim Allen, Melinda Allen, Wal Ambrose, Atholl Anderson, Steve Athens, Stuart Bedford, Peter Bellwood, David Burley, Mike Carson, Jeff Clark, Tom Dye, Eric Conte, Geoff Clark, Jeff Clark, Janet Davidson, Scott Fitzpatrick, Travis Freeland, Aymeric Hermann, Geoff Irwin, Dilys Johns, Jennifer Kahn, Thegn Ladefoged, Jolie Liston, Helene Martinsson-Wallin, Lisa Matisoo-Smith, Mark McCoy, Guillaume Molle, Mara Mulrooney, Andrew Pawley, Seth Quintus, Barry Rolett, Christophe Sand, Peter Sheppard, Jim Specht, Matt Spriggs, Glenn Summerhayes, Pam Swadling, Robin Torrence, Jo Anne Van Tilburg, Richard Walter, Marshall Weisler, and Peter White.

At the University of California Press, I thank Blake Edgar and Reed Malcolm for editorial stewardship. Sheila Berg skillfully copyedited my manuscript. Diana Izdebski prepared the maps.

During the tedious research and writing process, my loyal Rhodesian ridgeback, Nani, kept me company on her bed next to my desk, only periodically reminding me of the need to take a break by resting her head heavily on my arm, forcing a halt in my typing. But it is especially to my wife, Thérèse Babineau, that I owe a tremendous debt for her love and support. Me ke aloha pumehana.

As with the first edition, I dedicate this new edition to Douglas E. Yen, still going strong in the ninth decade of his life. His early mentoring has not ceased to influence my scholarship.

PATRICK VINTON KIRCH

Quinta Pacifica

Introduction

Mine is the migrating bird winging afar over remote oceans,

Ever pointing out the sea road of the Black-heron—the dark cloud in the sky of night.

It is the road of the winds coursed by the Sea Kings to unknown lands!

POLYNESIAN VOYAGING CHANT, in

J.F. STIMSON, Songs and Tales of

the Sea Kings (1957)

IN MARCH 1896, AN ENGLISH gentleman-adventurer by the name of F.W. Christian arrived at a place called Madolenihmw, on the southern coast of Pohnpei Island in Micronesia. Having spent some years in Samoa (where he was a neighbor of Robert Louis Stevenson), Christian had heard from that equally famous teller of South Sea tales, Louis Becke, that there existed on Pohnpei an ancient island Venice shrouded in jungle. Relating his first visit to the Nan Madol ruins at Madolenihmw, Christian wrote, Passing the southern barricade of stones, we turned into the ghostly labyrinth of this city of the waters, and straight-away the merriment of our guides was hushed, and conversation died down to whispers (1899:78). The immensity of the ancient town and its stonework, laced with canals, overwhelmed him: Above us we see a striking example of immensely solid Cyclopean stone-work frowning down upon the waterway, a mighty wall formed of basaltic prisms (79). Uncertain what to make of these vestiges of the long, long ages, Christian evoked comparisons with the semi-Indian ruins of Java, and the Cyclopean structures of Ake, and Chichen-Itza in Yucatan (80). Almost a century later, archaeological excavations at Nan Madol would reveal a story of the rise and fall of an island civilization.

Although one of the largest and most dramatic archaeological sites in the Pacific, Nan Madol is not the most famous; the gargantuan statues of Rapa Nui (Easter Island) would surely claim that distinction. They too testify to the rise of another island civilization, one whose fate has been hotly debated in recent years—whether Rapa Nui stands as a testament to ecocide or to adaptability and resilience. Other Pacific archaeological locales, while known primarily to scholars, are no less significant on the scale of world history. Kuk, a stratified succession of clay layers in the swampy floor of a New Guinea Highlands valley—the antithesis of an impressive stone construction like Nan Madol—has produced clues to some of the earliest horticultural activities anywhere in the world, at about 7000 B.C. And at Matemkupkum, a limestone rockshelter on New Ireland, fishbones and shellfish dating back 35,000 years testify to some of the earliest exploitation of coastal marine resources by modern Homo sapiens.

These examples hint at the diversity and richness of the Oceanic archaeological record, a legacy that only began to be thoroughly explored, studied, and interpreted in the second half of the twentieth century. In the process, questions that scholars have posed and puzzled over for two centuries and more—Where did the Pacific islanders come from? How did they discover and settle the thousands of islands? Why did they build great constructions like Nan Madol, or carve the Rapa Nui statues?—are finally being answered. This book chronicles the efforts of archaeologists to discover and understand the archaeological record of the Pacific islands and offers a synthesis of what we have found.

My title—On the Road of the Winds—is meant to evoke the countless voyages that underwrote the discovery and settlement of the myriad Oceanic islands. Some voyages were short, others of great duration and hardship, most often made toward the east, hence upwind along countless trackways stretching off into the dawn. Ultimately, the origins of the Pacific islanders trace back to the west, to a period tens of thousands of years ago when cyclically rising and falling ice age seas wrought great changes in the coastal configurations of the Southeast Asian and Australian continents. Much later, other voyagers followed, propelled in their outrigger canoes with sails of woven mats across previously untracked seaways, ultimately to reach the shores of South America. Then too my title invokes not just voyages undertaken by Oceanic peoples themselves—whether by raft, dugout, or double-hulled sailing canoe—but also another kind of voyage: the intellectual voyage of exploration and discovery of the Oceanic past. For this is a past encoded not in written texts but in potsherds and stone tools unearthed from island soils, in the relationships among Pacific island languages, in the cultural and biological variation of hundreds of Oceanic societies and populations dispersed over one-third of the earth’s surface, from New Guinea to Rapa Nui, from Hawai‘i to New Zealand. Both kinds of voyages—the real voyages of history and the intellectual voyages of the mind—are the concern of the chapters to follow.

Pacific islanders possess their own indigenous forms of history, accounts of ancestors passed down through chants, songs, and oral traditions.¹ These oral narratives also speak of voyages, many of epic proportions. There is, for example, Rata, whose great double-hulled canoe, The-Cloud-Overshadowing-the-Border, carried him and his mother, North Tahiti—after many harrowing adventures—back home to Great Vava‘u in Upper Havaiki. Or the famous canoe Lomipeau, which transported the massive limestone slabs from ‘Uvea Island to Tongatapu, where they were used to build Paepaeotelea (the tomb of Telea). The atoll dwellers of Kiribati recount stories of the voyages of the tropic bird people led by Koura, in the canoe Te-Buki-ni-Benebene (The-Tip-of-a-Coconut-Leaf). Such indigenous traditions provide one source of knowledge regarding the Oceanic past, offering insights into cultural motivations. It is an insider’s history. Western scholars have long drawn on the historical traditions of Pacific islanders; indeed these offered primary evidence for many late nineteenth- and early twentieth-century syntheses.² But as the human sciences matured in the twentieth century, they developed sophisticated methods for extracting historical information from diverse sources lying outside the boundaries of either traditional oral or written histories, sources that open windows on the deep past of the peoples without history. This book offers an explicitly anthropological history that privileges the archaeological record of human material culture and culturally altered landscapes. It is a history that also draws as appropriate upon the evidence of historical linguistics, comparative ethnography, and biological anthropology. Let us call it, then, an archaeological history that is equally informed by its sister anthropological disciplines.

Using an elegant metaphor of history as a ceaseless progression of waves of different amplitude, the French historian Fernand Braudel (1980) called the longest of these the longue durée. The long run of history, of deep time, tracks the underlying rhythms of economic production, the fundamental structures of society, the seemingly imperceptible fitting of culture to nature, and the manipulation of nature to reproduce culture. In writing his famous opus on another ocean, La Méditerranée, Braudel followed the lead of his mentor Marc Bloc by incorporating nondocumentary sources of evidence, arguing that the history of the longue durée is as much inscribed in the very fabric of the land, and in the patterns of culture, as it is in the written word. Many archaeologists, myself among them, have come to view our work as the writing of such long-term histories, the unearthing of the longue durée.

The deep, strong currents of the longue durée are akin to the great transoceanic swells that sweep the Pacific Ocean from continent to continent. To take the measure of their wavelength requires that we move beyond the constraints of a narrow documentary history, or even of a particularistic archaeology. A holistic perspective is called for, one that brings to bear the clues derived as much from the study of synchronic linguistic, cultural, and biological variation as from the direct, materialist, properly diachronic evidence of archaeology.

Modern archaeology is still less than a century old in the Pacific, but the islands have begun to take their rightful place in the annals of world history. Their longue durée is a rich story, one that our narratives are only beginning to describe—fascinating in its own right but also replete with plots and themes whose historical significance resonates beyond local place and specific time.

This is where I would now take you, on a voyage to the islands of history.

DEFINING OCEANIA

Vasco Nuñez de Balboa gazed out, in 1513, across the Pacific Ocean; Ferdinand Magellan crossed it in 1520–21. By the late sixteenth century the Spanish were annually sailing from Acapulco to Manila and back to New Spain via the North Pacific, yet Europeans had little real knowledge of the Pacific or its thousands of islands until nearly two centuries later, when the epic voyages of Captain James Cook (1768–80) disproved the theory of a great Terra Australis, a southern continent. Cook for all intents and purposes created the modern map of the Pacific. Moreover, he and the gentlemen-naturalists who sailed with him (Joseph Banks, Sydney Parkinson, Daniel Carl Solander, Johann Reinhold Forster, and George Forster) initiated serious ethnographic inquiry into the peoples and cultures of the Pacific islands. These European explorers—part of the great intellectual movement we call the Enlightenment—were amazed to find the islands of the Great South Sea well populated by indigenous peoples, many (but not all) of whom spoke related languages. Moreover, these islanders were expert sailors and navigators. Tupaia, a Tahitian priest-navigator interviewed by Cook, named no less than 130 islands to which he knew the sailing directions and relative distances.³ Thus, long before the Spanish and later the French and English, other peoples had explored the vastness of the Pacific, discovered virtually every single one of its habitable islands, and founded successful colonies on most. These island societies—of which Otaheite (Tahiti) was the sine qua non—intrigued and tantalized Enlightenment savants, including Jean-Jacques Rousseau and Denis Diderot, who mined the explorers’ journals for evidence to support their provocative theories of the human social condition.

One of the last of the naval commanders of this period of great exploratory voyages, the Frenchman Dumont d’Urville, in his Notice sur les îles du Grand Océan (1832), classified the peoples of the Pacific islands into three major groups. The first of these were the Polynesians (many islands),⁴ a light-skinned people spread over the islands of the eastern Pacific, including Tahiti, Hawai‘i, Rapa Nui, and New Zealand. In the western Pacific north of the equator, Dumont d’Urville defined another major group, the Micronesians (little islands), many of whom occupied small atolls. His third group, whom he called the Melanesians (dark islands), consisted of the generally darker-skinned peoples inhabiting the large islands of New Guinea, the Solomons, Vanuatu (then the New Hebrides), New Caledonia, and Fiji.

Although based on a superficial understanding of the Pacific islanders, Dumont d’Urville’s tripartite classification stuck. Indeed, these categories—Polynesians, Micronesians, Melanesians—became so deeply entrenched in Western anthropological thought that it is difficult even now to break out of the mold in which they entrap us (Thomas 1989; Tcherkézoff 2008). Such labels provide handy geographic referents, yet they mislead us greatly if we take them to be meaningful segments of cultural history. Only Polynesia has stood the tests of time and increased knowledge as a category with historical significance. Probably this is because the Polynesians were defined by Dumont d’Urville as much by their linguistic similarities as by perceived racial affinities. Hence, the Polynesians do form a meaningful unit for culture-historical analysis (see chap. 7).

The labels Micronesia and, most particularly, Melanesia imply no such culture-historical unity. Whether we are looking at language, human biological variation, or culture, the peoples of Melanesia defy categorization; they are among the most diverse and heterogeneous to be found in any comparably sized geographic space on earth. The historical processes underlying such great variety—which can only be disentangled through the holistic methods of anthropological history—are the subject of this book. Suffice it to say that when the terms Melanesia or Micronesia are used in the following pages, the reference will be exclusively to geographic regions, with no implied ethnolinguistic uniformity.

Dumont d’Urville’s three groups, taken together, are generally understood to make up Oceania and usually exclude the islands of Southeast Asia (the Indonesian and Philippine archipelagoes in particular). In some usages (e.g., Oliver 1989), Oceania includes Australia, although that is not the sense here. Island Southeast Asia is usually also regarded as a distinct region from Oceania, even though there are close culture-historical relationships between the indigenous peoples of island Southeast Asia and the Pacific. For one thing, the great Austronesian language family spans both regions. However, island Southeast Asia has had a complex historical overlay of cultural influence from the Indian subcontinent, which is not shared with the islands east of the Moluccas in Eastern Indonesia; this has partly influenced the separateness of geographically focused scholarly traditions. For the purposes of this book, I largely confine my scope to Oceania as traditionally defined (excluding Australia and island Southeast Asia), although at times it will be necessary to look beyond its borders in order to understand fully aspects of Oceanic history and culture.

Two other geographic terms require discussion, for these are relatively new concepts, although they are increasingly used by anthropologists and historians who work in Oceania. These are Near Oceania and Remote Oceania, originally proposed by Roger Green (1991b) in reaction against the historical sterility of the Melanesia concept. As seen in map 1, Near Oceania includes the large island of New Guinea, along with the Bismarck Archipelago, and the Solomon Islands as far eastward as Makira and Santa Ana. This is not only the region of greatest biogeographic diversity within Oceania but also that which saw human arrival beginning in the late Pleistocene (ca. 40,000 years ago; see chap. 3). Within Near Oceania, we find peoples who speak both Austronesian and Non-Austronesian (Papuan) languages. Remote Oceania includes all the Pacific islands to the north, east, and southeast of Near Oceania, yet its inhabitants speak exclusively Austronesian languages. The Remote Oceanic islands were not discovered or settled by humans until after about 1200 B.C., and in some cases as recently as A.D. 1000 (Green 1995). Thus the distinction between Near Oceania and Remote Oceania is not merely a geographic division, but one that encapsulates two major epochs in the history of the Pacific islanders.

MAP 1 Oceania, showing the traditional cultural regions of Polynesia, Melanesia, and Micronesia. The heavy dashed line indicates the boundary between Near Oceania (to the west) and Remote Oceania (to the east).

LINGUISTIC, HUMAN BIOLOGICAL, AND CULTURAL VARIATION IN OCEANIA

While primarily a work of history, this book also seeks an anthropologically grounded explanation for—and understanding of—the synchronic cultural, linguistic, and human biological variation exhibited throughout the modern Oceanic world. Thus a brief introduction to the dimensions of such variation is essential.⁵ I begin with language. The indigenous peoples of Oceania speak roughly 1,200 extant or historically recorded languages. Of these, about 450 belong to the well-defined and geographically widespread Austronesian family (map 2). The Austronesian languages (which total approximately 1,200) are found as far west as Madagascar, and include most of the languages of island Southeast Asia (including the aboriginal languages of Taiwan), the majority of languages spoken in Melanesia outside New Guinea, and all the languages spoken within Micronesia and Polynesia. Moreover, with few exceptions (specifically Chamorro and Palauan in western Micronesia) the Austronesian languages spoken in Oceania all belong to one particular subgroup known as Oceanic (Pawley and Ross 1995). This linguistic distribution pattern proves to be of great culture-historical significance, as discussed further in chapter 4.

MAP 2 The distribution of the Austronesian and Non-Austronesian (or Papuan) languages in Oceania. The Non-Austronesian languages are situated in the shaded areas, while the heavy lines delineate several major subgroups of Austronesian languages. SHWNG, South Halmahera–West New Guinea; CMP, Central Malayo-Polynesian.

On the large island of New Guinea, and in a few scattered locales elsewhere in Near Oceania (such as on New Britain and Bougainville Islands), the indigenous languages are Non-Austronesian, or, as they are sometimes called, Papuan (Foley 1986, 2000). There are an estimated 750 Papuan languages, but these emphatically do not form a single, coherent (i.e., genetically related) language family, as with Austronesian. Rather, the linguistic diversity encompassed within the Papuan group is enormous, and several family-level groups (phyla) are included under this rubric. Many of the Papuan languages are nonetheless historically related (such as those of the Trans–New Guinea Phylum), even though not all of them may have descended from a common ancestral or proto-language. In contrast to the Austronesian languages, the Papuan languages display significantly greater variation and diversity, with profound historical implications. In particular, there had to have been a substantially greater time depth for the differentiation of the Papuan languages, correlating with the much deeper time span of human occupation of New Guinea and Near Oceania (see chap. 3).

Perhaps to a greater extent than in any other major region of the world, in the Pacific archaeologists and historical linguists enjoy a fruitful collaboration. Our data and methods are different and our conclusions are derived independently, but both groups of scholars are concerned with cultural history (Blust 1995, 1996, 2013; Pawley and Ross 1993, 1995). Significantly, historical linguistic work on Pacific languages utilizes the comparative method, a theoretically and empirically well-grounded set of techniques for establishing the genetic or historical relationships among a set of related languages, as well as for reconstructing the vocabularies (and associated semantic meanings) of various ancestral or proto-languages (Hoeningswald 1960, 1973; Trask 1996:202–40). Trask (1996:208) calls the comparative method the single most important tool in the historical linguist’s toolkit, and it must not be confused with other methods, such as lexicostatistics, that depend on crude statistical comparisons and may not yield accurate language family histories or relationships (e.g., Dyen 1965). As Pawley and Ross (1995:40–43) explain, the comparative method builds on rigorous comparison of extensive sets of words or morphemes in groups of languages hypothesized to be historically related (cognate sets), thereby determining patterns of regular sound correspondences. Only when such patterns have been carefully confirmed does the historical linguist turn to the task of generating a subgrouping model or family tree of relationships among the languages under consideration. With such a model in hand, in which branches of related languages are robustly marked by sets of shared innovations, one can begin to reconstruct ancient vocabulary sets and their cultural domains. In this book, I periodically refer to the subgrouping models of Pacific linguists, as well as to reconstructed proto-vocabulary, to assess independently how these stack up against the evidence of archaeology.

Recently, work on the historical relationships of Oceanic and other Austronesian languages has benefited not only from traditional comparative linguistic methods but also from the application of cladistic procedures derived from biology (Greenhill and Gray 2005; Gray and Jordan 2000; Hurles et al. 2003). Although these methods rely on statistical comparisons between word lists, the phylogenetic, or family tree, relationships they reveal closely match those derived from classical linguistic methods.

Turning to human biological variation, it is risky to summarize the great diversity of Pacific human populations in a few short paragraphs. Earlier in this century, pioneering physical anthropologists sought to classify the diverse populations of the Pacific into a small set of races, such as Negroid, Negritoid, Australoid, or Polynesian (the latter being regarded as a mixed race).⁶ Modern biological anthropologists have shrived themselves of this kind of racial pigeonholing and endeavor to study populations, using an array of both phenotypic (e.g., anthropometric, dermatoglyphic) and genetic (e.g., mitochondrial DNA, nonrecombining Y-chromosome DNA, blood polymorphisms) characteristics.

The peoples who inhabit Polynesia, while displaying considerable differences in body form, are a relatively homogeneous group when compared to other Oceanic populations. Along with the people of Fiji, the Polynesians generally link together robustly in statistical analyses. Populations distributed within Micronesia are somewhat more varied than those of Polynesia. Most diverse of all, however—almost to the point of defying description—are the populations distributed within the geographic area of Melanesia. Melanesian human biological diversity is immense, once having been described by the Harvard anthropologist W.W. Howells (1970:192) as so protean and varied as to resist satisfactory analysis. As with language, Melanesia (and particularly the part we call Near Oceania) proves to be the most diverse sector of the Oceanic world; this is again an observation with considerable historical significance, since diversity frequently implies great time depth.

Yet the immense biological diversity found within Near Oceania is neither wholly random nor unpatterned. Investigations of genetic variation in this region (e.g., Friedlaender, ed., 1987; Hill and Serjeantson, eds., 1989; Lum and Cann 1998; Merriwether et al. 1999) reveal strong correlations between certain genetic markers and populations as defined on the basis of linguistic criteria (especially the distinction between Austronesian and Non-Austronesian speakers). For example, J.W. Froehlich (1987) used fingerprints (dermatoglyphs), which are highly heritable characters, to look at phylogenetic patterns among Solomon Island populations. He found that despite an accumulation of . . . local effects, and with local relationships sometimes obscured by sampling variance, the fingerprint gene pools still reflect a broadly geographical and presumably historical distinction between [Non-Austronesian-] and [Austronesian]-speaking people (1987:206).

A major breakthrough in our understanding of human biological variation in the Pacific came with advances in population genetics and molecular biology, especially through sequencing of mitochondrial DNA (mtDNA) and nonrecombining Y-chromosome (NRY) DNA (Boyce, Harding, and Martinson 1995; Hill and Serjeantson, eds., 1989; Friedlaender et al. 2008; Friedlaender, ed., 2007; Martinson 1996; Martinson et al. 1993; Melton et al. 1998). One of the first major discoveries was that Polynesians and island Southeast Asians shared a particular 9-base pair deletion in mtDNA, sometimes called the Polynesian motif (Hertzberg et al. 1989; Lum and Cann 1998; Lum et al. 1998; Merriwether et al. 1999). This discovery strengthened the linguistic argument that Polynesian origins could be traced back to Southeast Asia, and more specifically to Taiwan. Subsequent work on NRY variation, however, showed that the dominant Y haplotype in Polynesians is likely to be of Melanesian (Near Oceanic) origin (Kayser et al. 2000; Hurles et al. 2002; Su et al. 2000). This apparent contradiction in the mtDNA and NRY evidence can be resolved through a model whereby Polynesian ancestors originated from East Asia but genetically mixed with Melanesians before colonizing the Pacific (Kayser et al. 2008:1362; see also Vilar et al. 2008). A period of sustained gene flow between ancestral Polynesian and Melanesian populations is also evidenced by the presence of an α-thalassemia deletion, which confers resistance to malaria. This genetic mutation must have been transferred to the ancestral Polynesian populations in Near Oceania before their dispersal to Remote Oceania, where malaria is generally absent (Hill et al. 1985; Martinson 1996).

The recent molecular research has also made significant contributions to understanding genetic diversity within Near Oceania. The various studies of mtDNA and NRY variation in northern Melanesia support an interpretation of great time depth in this region; microsatellite diversity gives estimated divergence ages of between 32,000 and 50,000 years for haplotypes that developed in Near Oceania (Friedlaender, ed., 2007:92). Moreover, Papuan-speaking language groups (typically inland populations) are genetically the most distinctive in island Melanesia (232). These findings correlate well with archaeological and linguistic models of Pleistocene settlement of Near Oceania by ancestors of Papuan-speaking groups, as discussed further in chapter 2.

Culturally, the peoples of Oceania also vary greatly, with similarities