Natural Disasters and Cultural Change (One World Archaeology)

NATURAL DISASTERS AND CULTURAL CHANGE The One World Archaeology (OWA) series stems from conferences organised by the W...

Author: John Grattan | Robin Torrence

26 downloads 1444 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

NATURAL DISASTERS AND CULTURAL CHANGE

The One World Archaeology (OWA) series stems from conferences organised by the World Archaeological Congress (WAC), an international non-profit making organisation, which provides a forum of debate for anyone who is genuinely interested in or has a concern for the past. All editors and contributors to the OWA series waive any fees they might normally receive from a publisher. Instead all royalties from the series are received by the World Archaeological Congress Charitable Company to help the wider work of WAC. The sale of OWA volumes provides the means for less advantaged colleagues to attend WAC conferences thereby enabling them to contribute to the development of the academic debate surrounding the study of the past. The World Archaeological Congress would like to take this opportunity to thank all editors and contributors for helping the development of world archaeology in this way.

ONE WORLD ARCHAEOLOGY Series Editor (Volumes 1–37): Peter J. Ucko Academic Series Editors (Volume 38 onwards): Martin Hall and Julian Thomas Executive Series Editor (VoIume 38 onwards): Peter Stone 1. What is an Animal?, T. Ingold (ed.) 2. The Walking Larder: Patterns of domestication, pastoralism and predation, J. Clutton-Brock 3. Domination and Resistance, D. Miller, M.J. Rowlands and C. Tilley (eds) 4. State and Society: The emergence and development of social hierarchy and political centralization, J. Gledhill, B. Bender and M.T. Larsen (eds) 5. Who Needs the Past? Indigenous values and archaeology, R. Layton (ed.) 6. The Meaning of Things: Material culture and symbolic expression, I. Hodder (ed.) 7. Animals into Art, H. Morphy (ed.) 8. Conflict in the Archaeology of Living Traditions, R. Layton (ed.) 9. Archaeological Heritage Management in the Modern World, H.F. Cleere (ed.) 10. Archaeological Approaches to Cultural Identity, S.J. Shennan (ed.) 11. Centre and Periphery: Comparative studies in archaeology, T.C. Champion (ed.) 12. The Politics of the Past, P. Gathercole and D. Lowenthal (eds) 13. Foraging and Farming: The evolution of plant exploitation, D.R. Harris and G.C. Hillman (eds) 14. What’s New? A closer look at the process of innovation, S.E. van der Leeuw and R. Torrence (eds) 15. Hunters of the Recent Past, L.B. Davis and B.O.K. Reeves (eds) 16. Signifying Animals: Human meaning in the natural world, R.G. Willis (ed.) 17. The Excluded Past: Archaeology in education, P.G. Stone and R. MacKenzie (eds) 18.From the Baltic to the Black Sea: Studies in medieval archaeology, D. Austin and L. Alcock (eds) 19. The Origins of Human Behaviour, R.A. Foley (ed.) 20. The Archaeology of Africa: Food, metals and towns, T. Shaw, P. Sinclair, B. Andah and A. Okpoko (eds) 21. Archaeology and the Information Age: A global perspective, P. Reilly and S. Rahtz (eds) 22. Tropical Archaeobotany: Applications and developments, J.G. Hather (ed.) 23. Sacred Sites, Sacred Places, D.L. Carmichael, J. Hubert, B. Reeves and A. Schanche (eds) 24. Social Construction of the Past: Representation as power, G.C. Bond and A. Gilliam (eds) 25. The Presented Past: Heritage, museums and education, P.G. Stone and B.L. Molyneaux (eds)

26. Time, Process and Structural Transformation in Archaeology, S.E. van der Leeuw and J. McGlade (eds) 27. Archaeology and Language I: Theoretical and methodological orientations, R. Blench and M. Spriggs (eds) 28. Early Human Behaviour in the Global Context, M. Petraglia and R. Korisettar (eds) 29. Archaeology and Language II: Archaeological data and linguistic hypotheses, R. Blench and M. Spriggs (eds) 30. Archaeology and Anthropology of Landscape: Shaping your landscape, P.J. Ucko and R. Layton (eds) 31. The Prehistory of Food: Appetites for Change, C. Gosden and J.G. Hather (eds) 32. Historical Archaeology: Back from the edge, P.P.A. Funari, M. Hall and S. Jones (eds) 33. Cultural Resource Management in Contemporary Society: Perspectives on managing and presenting the past, F.P. MacManamon and A. Hatton (eds) 34. Archaeology and Language III: Artefacts, languages and texts, R. Blench and M. Spriggs (eds) 35. Archaeology and Language IV: Language change and cultural transformation, R. Blench and M. Spriggs (eds) 36. The Constructed Past: Experimental archaeology, education and the public, P.G. Stone and P. Planel (eds) 37. Time and Archaeology, T. Murray (ed.) 38. The Archaeology of Difference: Negotiating crosscultural engagements in Oceania, R. Torrence and A. Clarke (eds) 39. The Archaeology of Drylands: Living at the margin, G. Barker and D. Gilbertson (eds) 40. Madness, Disability & Social Exclusion: The archaeology & anthropology of ‘difference’, J. Hubert (ed.) 41. Destruction and Conservation of Cultural Property, R.L. Layton, P.G. Stone and J. Thomas (eds) 42. Illicit Antiquities: the theft of culture and the extinction of archaeology, N. Brodie and K. Walker Tubb (eds) 43. The Dead and their Possessions: repatriation in principle, policy and practice, C. Fforde, J. Hubert and P. Turnbull (eds) 44. Matériel Culture: the archaeology of 20th century conflict, J. Schofield, W.G. Johnson and C.M. Beck (eds) 45. Natural Disasters and Cultural Change, R. Torrence and J. Grattan (eds)

NATURAL DISASTERS AND CULTURAL CHANGE

Edited by Robin Torrence and John Grattan

London and New York

First published 2002 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2003. © 2002 selection and editorial matter, Robin Torrence and John Grattan; individual chapters, the contributors All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book has been requested ISBN 0-203-16510-1 Master e-book ISBN

ISBN 0-203-25949-1 (Adobe eReader Format) ISBN 0-415-21696–6 (Print Edition)

Contents

List of figures List of tables List of contributors Series editors’ foreword Preface

vii xi xii xiv xv

1

The archaeology of disasters: past and future trends Robin Torrence and John Grattan

2

Basic characteristics of disasters Satoru Shimoyama

19

3

Tsunamis and the coastal communities of Papua New Guinea Hugh Davies

28

4

Bacolor town and Pinatubo volcano, Philippines: coping with recurrent lahar disaster K.S. Crittenden and K.S. Rodolfo

43

Maritime archaeology and behaviour during crisis: the wreck of the VOC ship Batavia (1629) Martin Gibbs

66

5

6

7

‘The end is nigh’? Social and environmental responses to volcanic gas pollution John Grattan, Mark Brayshay and Ruud T.E. Schüttenhelm Recurring tremors: the continuing impact of the AD 79 eruption of Mt Vesuvius Penelope M. Allison

1

87

107

vi

CONTENTS

8

Volcanism and early Maori society in New Zealand D.J. Lowe, R.M. Newnham and J.D. McCraw

126

9

Under the volcano: Ni-Vanuatu and their environment Jean-Christophe Galipaud

162

10

Earthquakes, subsidence, prehistoric site attrition and the archaeological record: a view from the Settlement Point site, Kodiak Archipelago, Alaska Patrick Saltonstall and Gary A. Carver

11

Natural disasters and cultural change in the Shumagin Islands Lucille Lewis Johnson

12

Horsemen of the Apocalypse: the relationship between severe environmental perturbations and culture change on the north coast of Peru Kimberly D. Kornbacher

13

14

172

193

204

Climatic change, flooding and occupational hiatus in the lake-dwelling central European Bronze Age Francesco Menotti

235

Towards an archaeology of crisis: defining the long-term impact of the Bronze Age Santorini eruption Jan Driessen

250

15

Volcanoes and history: a significant relationship? The case of Santorini Sturt W. Manning and David A. Sewell

16

What makes a disaster? A long-term view of volcanic eruptions and human responses in Papua New Guinea Robin Torrence

292

The impact of the Kikai-Akahoya explosive eruptions on human societies Hiroshi Machida and Shinji Sugiyama

313

17

18

Index

Volcanic disasters and archaeological sites in Southern Kyushu, Japan Satoru Shimoyama

264

326

343

Figures

2.1 Field and path buried under volcanic tephra at the Hashimuregawa site 24 2.2 Trace of a small river, which was buried by an avalanche 24 3.1 Villages affected by the Aitape tsunami and the resettlement villages 29 3.2 House posts at the site of Warapu village 33 3.3 Tsunami awareness material prepared and distributed by the Papua New Guinea National Disaster Management Office in 1999 38 4.1 Mt Pinatubo and its tectonic setting 44 4.2 The Bacolor area and the extent of lahar activity from 1991 to 1995 46–7 4.3 A house in stages of raising and reconstruction 55 4.4 The Cabetican Archdiocesan shrine 58 5.1 Dutch shipwrecks on the west coast of Western Australia 69 5.2 The massacre of the Batavia survivors 70 6.1 Banham, Norfolk: June–September mortality expressed as standard deviations from the 26-year mean 96 6.2 Cavendish, Suffolk: June–September mortality expressed as standard deviations from the 26-year mean 97 6.3 Castle Donington, Leicestershire: June–September mortality expressed as standard deviations from the 26-year mean 97 6.4 Cranfield, Bedfordshire: June–September mortality expressed as standard deviations from the 26-year mean 97 7.1 View of Mt Vesuvius from Monte Faito to the south, looking across the Bay of Naples 108 7.2 Map of Campania showing locations of Bay of Naples, Mt Vesuvius, Pompeii and Herculaneum 109 7.3 Piles of gypsum in garden of the Casa del Sancello Iliaca in Pompeii 113 7.4 Tourists in the Via dell’Abondanza in Pompeii 120 8.1 North Island volcanoes that have erupted since c. AD 200 and other features or sites mentioned in the text 129 8.2 Summary of eruptions of North Island volcanic centres, and other 130–1 events, since c. AD 200

viii

FIGURES

8.3 Earth oven (umu) on Mt Taranaki at c. 850 m asl 8.4 Archaeological section at Papamoa on the Bay of Plenty coast showing prehistoric Maori shell middens postdating the c. AD 1300 Kaharoa Tephra (‘Ka’) 8.5 Pteridium (bracken) spore profiles from North Island containing the c. AD 1300 Kaharoa Tephra 8.6 Map of the Tarawera area showing locations of the main craters of the 10 June 1886 fissure eruption across the Tarawera Volcanic Complex, Rotomahana Crater and Waimangu craters 8.7 Buried meeting house (wharenui) (‘Hinemihi’) and smaller houses (whare) at Te Wairoa after the AD 1886 Tarawera eruption 8.8 Mayor Island or Tuhua, a rift-related peralkaline rhyolite caldera volcano located c.30 km off the western Bay of Plenty coast, was the pre-eminent source of obsidian (tuhua) for prehistoric Maori 8.9 Locations of volcanic mountains and other features in North Island that are referred to in early Maori oral history 9.1 The New Hebrides island arc showing active volcanoes and the location of the sites discussed in the text 10.1 View of the Settlement Point site with house one excavation in the foreground 10.2 Settlement Point site 10.3 Beach ridges at the Settlement Point site 10.4 House floor elevation data from the Settlement Point site in relation to mean high water (MHHW) in 1995 and extreme high water (EHHW) at various points in time 10.5 The elevation history of the Settlement Point site for the past 1,000 years includes land-level changes resulting from three large earthquakes 11.1 The Aleutian Islands chain and location of the Shumagin Islands 11.2 Radiocarbon dates from lowest levels and therefore earliest habitation of sites in the Shumagin Islands 11.3 Radiocarbon dates for the basal levels of barabaras (houses) at site XSI-040 represent the earliest dates for house construction 11.4 Radiocarbon dates from uppermost levels and therefore earliest habitation of sites in the Shumagin Islands 11.5 Summary of radiocarbon dates for sites in the Shumagin Islands 11.6 Radiocarbon dates for roofs of barabaras (houses) at site XSI-040 represent the latest dates for house construction 12.1 Map of the north coast of Peru showing location of river valleys and major Moche archaeological sites 12.2 Bet-hedging model 12.3 Mortality profiles showing age at death for different archaeological burial assemblages

134

135 136

142 143

146 148 163 178 179 181

182

183 194 198 199 199 200 200 205 208 220

FIGURES

13.1 Lake Constance: locations of Arbon Bay (Switzerland) and Bodman Bay (Germany) 13.2 Lake-level fluctuations on Lake Zurich in the past 4,500 years 13.3 GIS computer simulation of the inferred EBA Lake Constance level in Arbon Bay (392 m asl) 13.4 GIS computer simulation of the inferred MBA Lake Constance level in Arbon Bay (400 m asl) 13.5 GIS computer simulation of the inferred EBA Lake Constance level in Bodman Bay (392 m asl) 13.6 GIS computer simulation of the inferred MBA Lake Constance level in Bodman Bay (400 m asl) 15.1 Simulation of co-ignimbrite phase Santorini tephra fall for (a) March, (b) June, (c) September and (d) December 15.2 Simulation showing the total distal distribution of Santorini tephra for the plinian phase and the co-ignimbrite phase in June 15.3 Average wind speeds for Santorini area, 25.4º E and 36.5º N, for June ( J) and December (D) at 1 km, 10 km, 20 km and 30 km heights for the period 1992 to 1999 16.1 Location of volcanoes and relevant archaeological sites in West New Britain 17.1 Geographical setting of the caldera volcanoes in South Kyushu 17.2 Schematic diagram showing the sequence of the Kikai-Akahoya eruptions 17.3 Eruptive formations of the Kikai-Akahoya tephra outcropping at the harbour on Takeshima Island 17.4 Isopach map of the Kikai-Akahoya ash (thickness in cm) and the approximate distribution of the Koya pyroclastic flows 17.5 Changes in phytolith assemblages at Onejime in the southernmost part of Kyushu 17.6 Changes in Lucidophylous forest inferred from phytolith analysis before the Kikai-Akahoya eruption and after the Kikai-Akahoya eruption 17.7 Clastic dike formed by the co-volcanic earthquake of the Kikai-Akahoya eruption outcropping at Aira town, Osumi Peninsula 17.8 Geographical extension of the regional phases of the earliest Jomon ceramic culture immediately before the Kikai-Akahoya eruption 17.9 Geographical extension of the regional phases of the early Jomon ceramic culture after the Kikai-Akahoya eruption 18.1 Distribution of tephras and principal archaeological sites during the Jomon period in Southern Kyushu 18.2 The sequence of Ryutaimon type pottery in Southern Kyushu 18.3 The spread of the potteries above and below P14 18.4 Tephras and types of pottery on Satsuma and Osumi Peninsula

ix

236 240 241 242 242 243 276–7 278

279 294 315 316 317 318 319

320

321

322 323 327 329 330 331

x

FIGURES

18.5 The movement of Jomon potteries under and above the eruption of Kikai caldera 18.6 Thickness of Ak-2 (km-11) and archaeological sites around Ibusuki City 18.7 Thickness of Mk (km-12) and archaeological sites around Ibusuki City

333 337 338

Tables

2.1 Basic characteristics of natural disasters 2.2 Case study of a disaster: basic characteristics, types of evidence and fields of study 4.1 Summary of lahar activity on the Pasig–Potrero river and its effect on Bacolor from 1991 to 1995 5.1 Predictions for behaviours and archaeological signatures at various stages of a shipwreck crisis 8.1 Volcanic hazards probably experienced or witnessed by prehistoric Maori 8.2 Potential effects and extent of impact of the main volcanic hazards on prehistoric Maori society 8.3 Beneficial volcanic features and products exploited by prehistoric Maori 8.4 Physical and elemental properties of volcanogenic red ochre (kokowai) from Kokowai Springs, Mt Taranaki 9.1 Radiocarbon dates from the Kurvot site on Toga, Vanuatu 10.1 Radiocarbon dates from the Settlement Point site 12.1 Subsistence data for Moche I–IV (Cerro Blanco) and Moche V (Galindo) occupations in the Moche Valley 12.2 Estimate of the maximum number of mud bricks required to construct Huacas A–D at Galindo 16.1 Summary of major Holocene volcanic events in West New Britain 16.2 Average thickness of airfall tephra in archaeological sites 16.3 Comparison between severity of event and cultural response 16.4 Temporal patterning in material culture in West New Britain

21 22 51 82 133 137 145 147 166 180 218 224 296 296 297 298

Contributors

Penelope M. Allison, Research Fellow, School of Archaeology and Anthropology and the School of National Defence, Australian National University, ACT 0200, Australia [email protected] Mark Brayshay, Department of Geographical Science, University of Plymouth, Plymouth, Devon, PL4 8AA, United Kingdom [email protected] Gary A. Carver, P.O. Box 52, Kodiak, Alaska 99615, USA [email protected] K.S. Crittenden, Department of Sociology, M/C 312, University of Illinois at Chicago, 1007 W. Harrison St, Chicago, Illinois 60607, USA [email protected] Hugh Davies, Department of Geology, University of Papua New Guinea, P.O. Box 414, University, NCD, Papua New Guinea [email protected] Jan Driessen, Département d’Archéologie, Université de Louvain, Place B Pascal 1, 1348-Louvain-la-Neuve, Belgium [email protected] Jean-Christophe Galipaud, Laboratoire ERMES, IRD, Technoparc, 5 Rue du Carbone 45072 ORLEANS Cedex 02, France [email protected] [email protected] Martin Gibbs, Department of Archaeology, James Cook University, Townsville, Queensland 4810, Australia [email protected] John Grattan, Institute of Geography and Earth Sciences, University of Wales, Aberystwyth, Ceredigion, SY23 3DB, United Kingdom [email protected] Lucille Lewis Johnson, Department of Anthropology, Vassar College, Poughkeepsie, New York 12604, USA [email protected]

CONTRIBUTORS

xiii

Kimberley D. Kornbacher, Department of Anthropology, Box 353100, University of Washington, Seattle, Washington 98195-0001, USA [email protected] [email protected] D.J. Lowe, Department of Earth Sciences, University of Waikato, Private Bag 3105, Hamilton, New Zealand [email protected] J.D. McCraw, Department of Earth Sciences, University of Waikato, Private Bag 3105, Hamilton, New Zealand [email protected] Hiroshi Machida, Tokyo Metropolitan University, 15-7, Shiratori-dai, Aobaku, Yokohama, Japan 227-0054 [email protected] Sturt W. Manning, Department of Archaeology, University of Reading, P.O. Box 218, Whiteknights, Reading RG6 6AX, United Kingdom [email protected] Francesco Menotti, Institute of Archaeology, 36 Beaumont Street, Oxford OX1 2PG, United Kingdom [email protected] R.M. Newnham, Department of Geographical Sciences, University of Plymouth, Plymouth, Devon PL4 8AA, United Kingdom [email protected] K.S. Rodolfo, Department of Earth and Environmntal Sciences, M/C 186, University of Illinois at Chicago, 845 Taylor St, Chicago, Illinois 60607, USA [email protected] Patrick Saltonstall, Alutiiq Museum, 215 Mission Road, Suite 101, Kodiak, Alaska 99615, USA [email protected] Ruud T.E. Schüttenhelm, Geological Survey of the Netherlands, P.O. Box 157, 2000 AD Haarlem, The Netherlands [email protected] David A. Sewell, Environmental Systems Science Centre, University of Reading, P.O. Box 218, Whiteknights, Reading RG6 6AX, UK Email – see Manning Satoru Shimoyama,The Archaeological Museum of Ibusuki City, 891-0403, 2290 Junichou, Ibusuki City, Kagoshima, Japan [email protected] Shinji Sugiyama, Paleoenvironment Research, 1417, Akaike, Miyazaki, Japan 880-0912 Email – see Machida Robin Torrence, Division of Anthropology, Australian Museum, 6 College Street, Sydney, NSW 2010, Australia [email protected]

Series editors’ foreword

One World Archaeology is dedicated to exploring new themes, theories and applications in archaeology from around the world. The series of edited volumes began with contributions that were either part of the inaugural meeting of the World Archaeological Congress in Southampton, UK in 1986 or were commissioned specifically immediately after the meeting – frequently from participants who were inspired to make their own contributions. Since then WAC has held three further major international Congresses in Barquisimeto, Venezuela (1990), New Delhi, India (1994), and Cape Town, South Africa (1999). Other, more specialised, ‘Inter-Congresses’ have focused on Archaeological ethics and the treatment of the dead (Vermillion, USA, 1989), Urban origins in Africa (Mombasa, Kenya, 1993), The destruction and restoration of cultural property (Brač, Croatia, 1998), Theory in Latin American Archaeology (Olavaria, Argentina, 2000), and The African Disaspora (Curacao, Dutch West Indies, 2001). In each case these meetings have attracted a wealth of original and often inspiring work from many countries. The result has been a set of richly varied volumes that are at the cutting edge of frequently multi-disciplinary new work. The series provides a breadth of perspective that charts the many and varied directions that contemporary archaeology is taking. As series editors we would like to thank all editors and contributors for their hard work in producing these books. We would also like to express our thanks to Peter Ucko, inspiration behind both the World Archaeological Congress and the One World Archaeology series. Without him none of this would have happened. Martin Hall, Cape Town, South Africa Peter Stone, Newcastle, UK Julian Thomas, Manchester, UK November 2001

Preface

The genesis of this book was a one-day session held at the Fourth World Archaeological Congress in Cape Town, South Africa in January 1999. We are very grateful to the organisers for allowing us a whole day for our workshop and for creating a rarefied but relaxed atmosphere that promoted discussion among people from very varied backgrounds. Most of the contributors, including the editors, had never met before and we all benefited enormously from several days of intense but highly enjoyable interaction. The editors would like to thank all those who presented papers as well as the keen audience who actively contributed to the lively debates, which overflowed into the various meeting places throughout the conference. Although given very little warning, Henry Mutoro and Patrick Mbunwe-Samba graciously offered to present important and timely papers about the effects of El Niño on archaeological sites in Kenya and the catastrophic Lake Nyos gas explosion in Cameroon respectively. Although we are unable to present these papers here and therefore lack important African case studies, the experiences they reported at the conference have certainly coloured how many participants view the social effects of disasters. Special thanks to Dave Gilbertson and Pip Rath who took notes from the discussions and to Lucy Johnson for circulating these. We also acknowledge those who contributed abstracts to WAC-4 but were unable to attend. Fortunately, several nevertheless prepared papers for this book. Finally, we were very fortunate in being able to recruit a number of additional papers to broaden the temporal, spatial, and thematic coverage of the papers presented in Cape Town. In this respect we would especially like to thank Chris Newhall and Stephen Athens for their productive suggestions.

1

The archaeology of disasters: past and future trends ROBIN TORRENCE AND JOHN GRATTAN

WHY STUDY DISASTERS? In a landmark book which examined the role of volcanic eruptions in human evolution, Sheets and Grayson (1979: 6) could legitimately note that very few archaeologists had paid significant attention to the potential cultural effects of the natural hazards (e.g. volcanic tephra, earthquake-damaged walls, etc.) whose occurrences were apparent from many of their excavations. The current situation is radically different. In recent years studies stressing the impacts of past natural disasters on ancient societies have increased dramatically, although the majority of these are still authored or inspired by natural scientists and astronomers rather than archaeologists (e.g. Ambrose, 1998; Driessen and Macdonald, 1997; Harris, 2000; Isaacson and Zeidler, 1999; McGuire et al., 2000; McCoy and Heiken, 2000; Newhall et al., 2000; Nur and Cline, 2000; Peiser et al., 1998; Siebe et al., 1996; Stiros and Jones, 1996). Volcanic eruptions have led the way as the most commonly invoked environmental forcing mechanism, but droughts, floods and earthquakes are now also regularly proposed as triggering cultural change. If we look to the modern world as a model for what we might expect to find in the past, we find that severe climatic events that wreak havoc on human communities, destroy homes and livelihoods, and inflict high levels of mortality are surprisingly frequent and widespread. For instance, Tobin and Montz (1997) provide a graphic catalogue of disasters during the single typical year of 1985. An earthquake in Mexico killed 20,000 people; a tropical cyclone killed 11,000 in Bangladesh, and one in Vietnam killed 670; 300 died from landslides in the Philippines; a volcano erupted in Colombia killing 25,000; a flood in China added 500 to the death toll; a storm in Algeria killed 26; cold waves were responsible for 290 deaths in India and 145 in the United States; a heat wave killed 103 in the United States; and 52 died in Egypt in a fire. (Tobin and Montz, 1997: 1) A detailed study by Glickman et al. (1992) found that between 1945 and 1986,

2

R. TORRENCE AND J. GRATTAN

2.34 million people lost their lives to disasters and that 30 disasters and 56,000 deaths occurred on average per year. Consequently, the study and management of natural hazards has become an important concern for the modern world, which now makes large financial investments in hazard prevention and relief. The United Nations went so far as to declare the 1990s the International Decade for Natural Disaster Reduction (IDNDR), an action that stimulated and fostered huge programmes for research and for disaster awareness programmes. Given the importance ascribed to natural disasters in the modern world, it therefore seems reasonable to assume that they were also frequently experienced by past societies. To what extent have severe environmental events had a significant effect on cultural histories? Based on the marked increase in popular and professional archaeological publications on their role in the past (e.g. Keys, 2000; McGuire, 1999; Schoch and Aquinas, 1999), one might assume that disasters have become fairly widely accepted as important agents of cultural change. We feel it is important to question whether the current popularity of external natural forces in accounting for human evolution and social change in the remote past is simply a product of modern concerns or has identified a genuinely important mechanism for change that has been relatively neglected until recently. The critical issue of correlation (an extreme natural event happened about the same time as the observed cultural change) versus causation (the cultural change was dependent on the environmental event) has rarely been satisfactorily addressed by detailed and systematic research (cf. Sadler and Grattan, 1999; Chapters 6 and 18). Too often archaeologists and earth scientists have simply assumed that the occurrence of extreme natural events means that they were the prime movers in cultural change without demonstrating that the latter was solely or largely dependent on the former. Consequently, the overall aim of this book is to critically examine the role of extreme environmental events in causing cultural change. The authors have deliberately taken a sceptical point of view and have carefully examined the evidence in order to distinguish between coincidence and dependence. We begin with a programmatic chapter by Shimoyama which proposes an analytical framework and a set of basic concepts that should guide archaeological disaster studies. Examples from Japan are used to illustrate his methodology. This statement about ideal methodology is followed by case studies with broad coverage in both spatial (North and South America, Europe, Asia and the Pacific) and temporal terms (several thousand years ago up to the present day). They also involve a wide sample of different mechanisms (climatic change, volcanoes, tsunamis, floods, earthquakes and a shipwreck) to present detailed assessments of the relationship between specific natural processes and cultural responses. The inclusion of historical and modern studies illustrates that the widest possible research framework is required in order satisfactorily to evaluate the role of human disasters. The modern studies make a particular contribution because they highlight areas of behaviour that archaeologists cannot monitor effectively. For example, Gibbs’s account in Chapter 5 of a shipwreck off the west coast of Australia provides

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

3

a gripping story of social disintegration following a catastrophe. The detailed reconstruction of the impact of the toxic gases that affected Europe in 1783 (Chapter 6) reminds us that some catastrophic events may not generate certain kinds of data and are therefore ‘invisible’ in archaeological terms. The recent disasters in Papua New Guinea (Chapter 3), the Philippines (Chapter 4) and Japan (Chapter 18) suggest that attachment to land or place may explain why some people do not abandon their homes even when faced with very dangerous and unpleasant conditions. Case studies like these provide explicit models that can help shape future archaeological work and so they form a very important part of this book. The results presented in the wide-ranging case studies highlight the importance of critical, analytical research to determine how and in what situations natural factors create disastrous conditions for humans and whether these have significant, long-lasting effects. On the scales over which archaeology generally deals, the papers emphasise the flexibility and adaptability of past societies and the importance of the social context in determining the ultimate outcome, a point which has also only recently been accepted in modern disaster research (e.g. Blaikie et al., 1994; Oliver-Smith, 1996; Tobin and Montz, 1997). The many substantive and theoretical issues raised by the papers also demonstrate that archaeological analyses of past disasters have a very important role to play in planning for the future.

THEORETICAL IMPORTANCE OF DISASTERS Apart from the current popularity of the concept that catastrophes were a powerful agent for cultural change, there are a number of compelling reasons why studying natural disasters is important for archaeological theory and practice. Archaeological theory about the pace and character of cultural change has generally assumed that the process is mainly internally generated, unfolds slowly through time, and inevitably leads to greater socio-cultural complexity and socalled levels of progress. Although environmental determinism has also been quite influential, various forms of the Functionalist or Processualist theories, which dominated archaeological and anthropological thought from the 1970s until recently, stressed homeostatis and equilibrium, properties which are in conflict with the notion of rapid change induced by external factors. Processual archaeologists are unlikely to have envisaged one-off events as having had a major effect over the very long time scales that archaeologists generally study. Despite experiencing a major catastrophe, societies are expected to have picked themselves up, dusted themselves off, and continued on their relentless social evolutionary path to complexity. As a result, scholars focused on what they saw as ‘normal patterns of behaviour’ and ‘had little to say about systems whose normal coping mechanisms failed’ (cf. Torry, 1979: 518, 521). In contrast, disasters are an important subject for study because, as noted by Oliver-Smith (1996: 303),

4

R. TORRENCE AND J. GRATTAN

they ‘signal the failure of a society to adapt successfully to certain features of its natural and socially constructed environment in a sustainable fashion’. Since they demonstrate what were the limits of adaptive processes, a focus on how societies respond to disasters would seem to be an important way to understand the general processes of evolution. Alternatives to social evolutionary thinking which focus on non-linear change, chaos, punctuated change and catastrophism (e.g. studies in van der Leeuw and McGlade 1997) provide a significant challenge to archaeological theory, but have received very little attention to date, although their role within modern studies of natural hazards has been promoted by Bryant (1991: 5–6). Chance events or what Gould (1989) has called ‘historical contingency’ are also beginning to be recognised as key factors within the process of cultural evolution (e.g. Terrell, 1988; Zeidler and Isaacson, in press). We argue that studying the cultural consequences of natural hazards and the disasters they may have caused in the past may suggest a very productive methodology for breaking out of established patterns of thought. Careful studies of past disasters also provide a useful format for testing alternative approaches to cultural change and may perhaps even lead to new ways for conceptualising non-linear processes. Finally, archaeological research can make a contribution to helping managers cope with contemporary disaster events. From archaeological research we may establish the principal components of a disaster, reconstruct the physical event itself, assess the physical damage it caused, and identify the response strategies of the exposed culture. More importantly, since archaeology operates over a large enough time scale, it can assess the long-term impacts of a disaster that might be overlooked in a modern study. Studies have already shown that long after the world press has moved on, local catastrophes can have profound long-term effects on the lives of the people involved and these have the potential to permeate and eventually alter the society as a whole (cf. Chapters 3, 5 and 12; Mbunwe-Samba, 1999; Grayson and Sheets, 1979: 628; Oliver-Smith, 1986). Furthermore, disasters can accelerate social processes that were in train beforehand (Blong, 1984: 186; Oliver-Smith, 1996: 313; Chapter 14). It is therefore very important to promote research which specifically evaluates the effects of natural disasters over longer time scales than is usually the case in modern disaster studies. Detailed archaeological case studies can make a significant contribution to this goal. With very few exceptions, disasters were widely ignored until the seeming exponential increase in mortality and damages in the recent past created a new awareness of their potential impacts. The danger, however, is that some scholars have gone too far and are making a simplistic analogy between modern concerns about disasters and potential effects in the past. This has led to the adoption of a dangerously uncritical approach when hypothesising the importance of past extreme environmental events. Although we argue that the role of disasters may have been overlooked, we also stress that their role in causing cultural change must be very carefully evaluated on a case-by-case basis.

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

5

CONCEPTS AND DEFINITIONS Risk management, which entails the study of natural hazards and their social impacts, has become increasingly popular in the last decade due to the boost of the United Nations IDNDR and economic challenges to the insurance industry posed by natural disasters. Until recently there were two separate fields of research. On the one hand, earth scientists studied the physical properties of the volcanoes, earthquakes, floods, tsunamis, etc., aiming to predict their occurrences and likely impacts (e.g. Bryant, 1991; Blong, 1984). On the other, social scientists focused on the short-term consequences of disasters and stressed cultural aspects of communities in determining their vulnerability to natural processes and their methods for coping with stress (e.g. Torry, 1979). Unfortunately, the two fields are still relatively separate and distinct (e.g. compare McGuire, 1999 or Harris, 2000 with Blaikie et al., 1994), although there are signs of major changes and recognition that both aspects need to be better incorporated into disaster research. Archaeological research can gain a great deal from the current debates taking place within the broad field of disaster management. Although Sheets and Grayson (1979: 4–6) reviewed this research in the introduction to their book, it was written before social scientists were heavily involved in disaster research (cf. Torry, 1979) and this is reflected in the emphasis in their text on the natural science approach. Previous archaeological studies of disasters have also mainly been influenced by earth scientists (e.g. McGuire et al., 2000; McCoy and Heiken, 2000). The papers in this book represent a significant change toward a more integrated methodology in which the environmental and social variables are considered to be equally relevant. Although they may be initiated by natural factors, ‘disasters are social phenomena’ (Shimoyama, Chapter 2). As emphasised by Blaikie et al., the ‘natural’ and the ‘human’ are so inextricably bound together in almost all disaster situations, especially when viewed in an enlarged time and space framework, that disasters cannot be understood to be ‘natural’ in any straightforward way. (Blaikie et al., 1994: 6) Most scholars agree that the critical ingredient of a disaster is the victims (cf. Chapter 2). Beyond this crucial point the details vary slightly. For example, Tobin and Montz (1997: 6) use 25 deaths as an arbitrary threshold for a disaster. Others require more extensive damage so that ‘all major public and private facilities no longer provide essential social and economic services without extensive replacement or repair’ (Torry, 1979: 518) or that ‘the essential functions of the society are interrupted or destroyed’ (Oliver-Smith, 1996: 305). In other definitions the key factor is the response. For example, a disaster is defined as a situation where ‘recovery is unlikely without external aid’ (Blaikie et al., 1994) or when there is ‘a total breakdown in day-to-day functioning’ and ‘the damage may be so great and so extensive that survivors have nowhere to turn for help’ (Tobin and Montz, 1997: 31). For our purposes the most simple definition – the existence of damage

6

R. TORRENCE AND J. GRATTAN

to individuals or their property – is all that is essential to the definition of a disaster. In this conception disasters can be placed along a continuum ranging from those with minimal consequences to others with economic and social losses. For archaeology the most critical point is not whether a disaster took place but whether it caused cultural change. Unlike most scholars, who consider natural hazards to comprise mainly environmental events, we make a distinction between forcing mechanisms and hazards. The forcing mechanism is defined as the process that initiates the damages. In our scheme the second component of a disaster, the natural hazard, comprises the ‘potential interaction between humans and extreme natural events’ (Tobin and Montz, 1997: 5). In other words, a physical process is not a hazard unless it could potentially impact on a social group. In assessing whether natural processes led to past disasters, the existence and nature of the hazard need to be assessed independently from the occurrence of severe environmental events that have been recorded in the geological record. The potential initiations or forcing mechanisms for disasters can be natural, social (e.g. warfare), or technological (e.g. oil spills, chemical explosions, etc.). In this book we focus on natural forcing mechanisms, which are important environmental events. It is useful to characterise these in terms of their frequency, intensity, duration, areal extent and speed of onset (cf. Bryant, 1991: 9; Tobin and Montz, 1997: 232). Although the importance of frequency and duration are recognised by natural scientists as being important, the cultural impacts of these are rarely studied in much detail because most social science research operates on very short time scales. Archaeology has an important role to play here. Most of the papers deal with processes, which occur suddenly: e.g. volcanoes, earthquakes, tsunamis, floods. Provoking factors with a slow onset, as for example climatic change, have always played an important role in archaeological explanations of cultural change, although the popularity of climate change as a prime mover appears to be on the rise (e.g. Moseley, 1997; Fagan, 1999; Cullen et al., 2000; Weiss and Bradley, 2001; Giller, 2001). It is considered here by Menotti (Chapter 13), and Kornbacher’s discussion of the effects of El Niño (Chapter 12) could be classified in this way, but the problems faced by the prehistoric communities she studied (floods and landslides) mostly arise fairly rapidly. We feel that the issues raised by studying climatic change as a cause of disasters can be quite different from events, which occur suddenly. Following on from the identification of natural hazards, researchers try to assess the risk that a disaster will occur as a consequence of the hazard. This depends on the likelihood of the forcing mechanism occurring as well as the probability that it will happen at a time and place that will affect a community. More important for archaeology is assessing the impact of the disaster once it has taken place. For many anthropologists and social scientists the key variable in a disaster is not the natural event itself but the vulnerability of the society which experiences it (e.g. Oliver-Smith, 1996: 314; Tobin and Montz, 1997: 32, 331). As defined by Blaikie et al. (1994: 9), vulnerability comprises ‘the characteristics of a person or group in terms of their capacity to anticipate, cope with, resist and

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

7

recover from the impact of a natural hazard’. It is clear that the more vulnerable the group, the greater the disaster and, by implication, the larger the potential for cultural change. Within most archaeological writing on disasters, the emphasis has been placed on the environmental forcing mechanism and to a lesser extent on the hazard. Very little attention has been paid to the vulnerability of the population affected. Notable exceptions are papers by Sheets et al. (1991) and Torrence et al. (2000), which compare the effects of volcanic disasters on societies with simple as opposed to complex social organisations. As discussed further below, one of the major outcomes of this book is a refocus on the social component of disasters, particularly with respect to assessing hazards and vulnerability as key elements in cultural change.

COINCIDENCE VS CAUSATION One of the most serious problems dogging archaeological studies of disasters is the lack of critical assessment of whether the relationships between the natural events and the cultural behaviour identified by the researchers were merely a coincidence or whether the latter was actually caused by the former. Generally there is no doubt that a serious environmental event occurred. The question is whether (1) it was contemporaneous with the cultural change observed and/or (2) the cultural change was a necessary consequence of the forcing mechanism. The opportunities offered to, and the problems faced by, archaeologists engaged in assessing the influence of natural disasters can be illustrated by volcanic activity. Volcanic eruptions may influence distant archaeological sites through the generation of climate change, the emission of toxic gases and the deposition of tephra. Archaeological sites close to the source may be influenced by tephra fall, toxic and super-heated gas and lava flow. It is tempting and convenient to use the temporal coincidence of a volcanic eruption or the physical evidence of volcanic ejecta to account for change in the archaeological record, but how may these coincidences be established as cause and effect? There is a real danger that coincidence is taken to imply a causal relationship. In such cases the proposed forcing mechanism is assumed to have been powerful enough to have brought about the change observed in the archaeological record. This is especially dangerous when the volcanological data suggest the eruption to have been of moderate magnitude. This same difficulty is experienced by natural scientists who wish to explain evidence for climate change in terms of the occurrences of major volcanic eruptions (cf. Sadler and Grattan, 1999). In fact the research in this field should sound a word of caution to archaeologists since climatic change generated by any known volcanic eruption during the Holocene has been shown to have been minimal and has never exceeded the variability that occurs normally in response to naturally occurring climatic forcing. To assess the nature of the association between extreme environmental events and a putative cultural response is not a simple matter and requires a critical approach backed up with careful and systematic analyses, as evidenced by the

8

R. TORRENCE AND J. GRATTAN

papers in this book. They demonstrate that the most productive procedure is to conduct independent analyses of the natural and cultural data, rather than to explain one in terms of the other, which has often been the case. Simply obtaining precise enough dating for the hazard and the cultural change can be extremely difficult, as well illustrated by the case of the Santorini eruption (Chapters 14 and 15). Independent dating of the environmental event is likely to continue to be a difficult problem because radiocarbon determinations are frequently taken from archaeological contexts. This can be dangerous because the site could have already been abandoned prior to the damage created by the forcing mechanisms. Furthermore, the standard deviations of most dates do not always allow a straightforward assessment of contemporaneity. The archaeological record contains abundant evidence that cultural groups experienced extreme natural events in the past. Numerous sites have been buried by volcanic tephra, walls collapsed due to earthquakes, dwellings destroyed by tsunamis and landslides. Despite our cautions that care be taken to establish causation in each particular case, disasters were certainly not an uncommon event in the past. It is, however, important to note that not all of the case studies conclude that the disaster which was detected had a noticeable impact on social process. Resilience and persistence are commonly stressed (e.g. Chapters 4, 6, 8, 9–11, 15 and 18). This result makes this book very different from many archaeological accounts of disaster, which we believe have over-emphasised the natural forces over the cultural responses. Modern disaster managers measure the impacts of disasters in terms of deaths and economic costs. Archaeologists, however, are more interested in the implications of these losses for causing cultural change. Their work raises the extremely knotty question about whether a particular disaster had a significant effect on the social group(s) that sustained it or interacted with it either locally or on a larger regional scale (cf. Chapters 6, 11 and 16). Judging from this set of papers, archaeology lacks an agreed definition of ‘significant’ change and lacks a fully satisfying account of what constitutes causation in relation to a disaster. Does the change need to involve the introduction of new behaviour traits or material culture, a total replacement of a culture, a societal collapse, or simply the abandonment of a farmstead? Although most authors here have looked for fairly drastic change as evidence for significant effects, those studying the more recent periods (e.g. Chapters 3 and 4) are impressed by what archaeologists might consider rather small-scale changes in settlement structure and pattern, and a number of authors have dealt with the issue of mythology and oral history as demonstrating important effects (see below; cf. Blong, 1982). Second, does the disaster need to have direct effects or can it be a catalyst for change (cf. Blong, 1984: 180–4)? Furthermore, does the change need to follow immediately or can one also posit long-term, follow-on effects that might last several hundred years? For example, Manning and Sewell (Chapter 15) are only satisfied if the cultural change can be linked in a ‘direct, immediate or quantifiable way’. Most of the rest of the authors are willing to consider a rather longer period of time and a range of effects. For example, Allison’s (Chapter 7) paper demon-

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

9

strates that some disasters can continue to have an effect many hundreds of years after the event. Longer-term effects have also been proposed by Kornbacher (Chapter 12), who argues that cultural evolutionary processes can result from natural disasters. Crittenden and Rodolfo (Chapter 4) point out that in some cases the worst effects are not experienced until long after the forcing event. In their case lahars (mudflows) from the Pinatubo eruption in the Philippines created serious damage five years later and at the time of writing, nearly 10 years after the initial event, they were still a significant threat to life and property. Driessen (Chapter 14) argues that indirect, follow-on effects of a disaster can accelerate a process that is already in motion (cf. Blong, 1984: 186; Oliver-Smith, 1996: 313). Temporal and spatial scales Whether cultural change has been defined as ‘significant’ depends very much on the temporal and spatial scale of the research. In the short term, many of the disasters studied in this book would have been termed ‘catastrophic’ by modern students of natural hazards because they led to large losses of life and/or property, sometimes over very large regions. In many cases a site or region was abandoned following the event. One controversy that arises and has not been handled satisfactorily is establishing whether abandonment of a site or region is a form of cultural change and if so, what length of time is required to label this ‘significant’ cultural change. For example, Torrence (Chapter 16) detected a period of abandonment up to 1,000 years long in Papua New Guinea following a volcanic eruption, but when the region was recolonised, there was virtually no difference in the material culture assemblage. Clearly the local group had been seriously affected by the disaster since it ceased to exist, but the larger regional population which recolonised it many years later had not experienced serious impacts: the basic stone toolkit had remained unchanged. What appears to have been an immediate and catastrophic cultural disaster in the immediate short term appears as nothing of the sort when considered over a longer time scale as the landscape is recolonised and utilised afresh. What we cannot even begin to guess at, however, is whether the lack of change is actually due to the effects of the volcanic eruption and drastic reduction in population. Perhaps the eruption caused a cessation in changes that were previously in motion: i.e. it retarded rather than promoted change. Perhaps the only way to resolve the issue of significance in situations like these is to work on multiple time scales, rather than to impose an arbitrary overall measure of what constitutes ‘significant’ cultural change. There are clear differences between the way volcanic eruptions and other hazards are perceived by those exposed to them during their daily lives, the descendants of those who experienced them, those who report them at a temporal or physical distance, and the excavators and interpreters of the physical debris of past events. The effects of disasters often live on far beyond the time when they occurred. One of the issues we were interested in investigating was the role of oral history and mythology about past disasters in guiding later responses to similar events. Certainly stories about disasters have been passed down in many communities (cf. Blong, 1982), but do they have a more pragmatic effect, as for

10

R. TORRENCE AND J. GRATTAN

example in regulating social behaviour? None of the authors who addressed this topic found clear evidence supporting this hypothesis. For example, the myths in Vanuatu (Chapter 9) depicted disasters as purely social rather than natural events. Johnson (Chapter 11) argues that Aleuts have a folklore concerned with natural hazards, but this is almost entirely dominated by storms at sea – an immediate and ever-present danger to people who rely upon the sea for so much of their livelihood. Earthquakes and volcanic eruptions are mentioned but only impinge on peoples’ lives indirectly. A far more important and immediate threat was the activities of other Aleut groups with whom they may have been in competition. In contrast, Davies (Chapter 3) highlights one of the problems with modern studies of oral history among communities where the social fabric has broken down and traditional stories are no long passed on. The memories and effects of disasters do not always just fade away. Past disasters can be resurrected and used within modern contexts. For example, Bryne (1997; 1999a) has discussed how the physical remains of disasters continue to play a very active part in political struggles long after they have occurred. Memories can be extremely powerful. He has argued that governments in the Philippines and in Bali have taken active measures to obliterate and hide the physical remains of, in this case social, disasters. He has also demonstrated that memories about ancestors can be awakened and preserved through a focus on places where disasters took place (Bryne, 1999b). Allison (Chapter 7) discusses how the events of AD 79, the destruction of Pompeii and Herculaneum, and their burial under many metres of volcanic debris, has coloured modern impressions of the nature of volcanic disasters. Given the predominance of this event in modern scholarship and popular imagination, it is surprising to find out that the wider region affected by this eruption, the Palma Campania, was not abandoned, the culture did not change, and the buffering offered by the Roman empire was considerable. We should keep in mind, however, that the events of AD 79 were not the first such catastrophe to have an impact on this region and its cultures. The problems of temporal scale are especially important for slow-onset events, such as climatic change, which require a different modelling procedure. For example, in Switzerland, Menotti (Chapter 13) shows that climate change appears to have caused a rise in lake levels leading to significant cultural change that occurred over the space of two generations. While the inundation of previous lake shores and consequent destruction of settlements may have been relatively rapid, the cultural response was not immediate. In the face of such events, communities have time to consider a number of strategies before the final response is adopted. A somewhat comparable situation may have arisen with rapid-onset events. The response need not have been immediate if the damage was not devastating. In some instances groups could take their time over considering whether, when and how to resettle. Focus on variability In our view there has been too much emphasis on trying to decide how much impact is enough to be considered as important. This has skewed the research

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

11

such that only the most devastating events are accepted as valid (cf. Chapter 16). The best way forward is to adopt the procedure used by Shimoyama (Chapter 18). He acknowledges that a wide range of effects from disasters can be expected and should be monitored. In this approach all events are considered as ‘significant’ despite a substantial range in the nature and degree of cultural change. In fact, arguments over the ‘significance’ of a disaster in terms of the scale of its impacts on social life detract from the more important task of assessing the nature and degree of the interrelationships among the forcing event, the hazard, vulnerability of communities, and outcome. For understanding why some extreme natural events have a different and perhaps larger impact than others, it would be useful to conduct detailed and systematic analyses which present the cultural outcomes of the extreme environmental event over a number of temporal and spatial scales.

CRITICAL VARIABLES At this preliminary stage of archaeological research on disasters, it would be worthwhile to move on from the controversy about whether disasters as a general phenomenon are culturally significant or not and focus attention on the very wide range of responses that have been observed. A productive approach would be to examine a range of variables characterising the natural forcing mechanisms and the cultural responses to see if any general patterns can be detected. Historical and modern studies, such as those presented in Chapters 3–7, would play a particularly important role in this exercise. It is hoped that this book will stimulate someone to take on this task as a major piece of research. At this early stage, however, one can already detect some intriguing patterns resulting from the case studies presented in this volume and these could serve as hypotheses for future research. Magnitude It is quite clear that the magnitude of a natural hazard is not the sole or even a straightforward predictor of its cultural impact. It requires a conscious act of discipline on the part of the modern archaeologist or geographer studying the impact of a historical or prehistoric natural hazard to analyse the data carefully. A part of one’s judgement is naturally coloured by the magnitude of the event, as it is uncovered through a research project. The geographical extent of a tephra fall and its thickness, buildings which show signs of damage, tsunami deposits, narrow tree rings and ice-core acidity all tempt us to assume significant environmental forcing and disaster. However, the studies presented in this volume suggest that while the influence of natural hazards is a factor that may be considered in many regions of the world, few severe environmental events have been responsible for major cultural change. Perhaps one of the best examples of the absence of a simple correlation between scale and outcome is presented by Grattan et al. (Chapter 6). In this case the effects of air pollution on a vast continental scale, which was caused by a volcanic eruption, is revealed. All the material which illuminates this event points

12

R. TORRENCE AND J. GRATTAN

to a major disaster: crops were destroyed in the field from Scandinavia to southern Europe; plants were defoliated; people fell sick and even died. Contemporary writers described profound social unease and even panic and it is not exaggerating to suggest that people feared Armageddon and the end of the world. In the historical record the severe air pollution event was shortly followed by the French Revolution and the Napoleonic Wars. Were we considering this material from a distance of 2,000 rather than just over 200 years, we might be comfortable to associate all these events. Reality is rather different. Severe as they were, the events of 1783 did not enter folklore, European agricultural production easily compensated for the destroyed crops and agricultural prices were stable. Even occasional episodes of high mortality do not appear to have unduly troubled the late eighteenth-century communities affected – perhaps because episodes of this kind were seen as the natural order of things. All things considered, although the magnitude of the event was high, the geographical extent of the hazard was continental and the environmental impact was severe, the concentrated sulphuric acid aerosol, which blanketed the continent of Europe in 1783, did not lead to cultural change. European culture and environment were not sufficiently vulnerable to a hazard of this nature and were adequately buffered against its impact. This historical case provides several very important lesson for prehistoric archaeologists, who can rarely detect events on a yearly basis. First, only a very detailed dating programme would show that a major cultural event that followed not long afterwards (warfare) was not caused by the severe environmental event. Second, the magnitude of the event was not the key variable. Third, it emphasises the importance of vulnerability in determining the ultimate outcome of a disaster. Duration and frequency The duration and frequency of the forcing events are likely to be key factors in determining the scale of cultural response. Familiarity with the risks involved should ensure that the environmental and physical risks are continually weighed and socially controlled. People appear to be willing to take quite high risks in the case of rare events in order to reap short-term benefits. Maintaining settlement in locations which are subject to infrequent hazards such as earthquakes and tsunamis is a good example. In Chapters 3, 4 and 9–11 groups are shown to have chosen to ignore natural hazards. In these cases the long-term benefit of locating a settlement in a particular location or the lack of suitable alternatives appear to override any concern relating to rare, if catastrophic, environmental events. In contrast, disasters which occur frequently or over a relatively long period of time can intensify evolutionary development and engender rapid change. For example, Kornbacher (Chapter 12) considers the response of the Moche culture in Peru to a significant range of natural hazards due to the effects of El Niño, including massive flooding and erosion, dune incursion and mass wasting. It is apparent that in response to the influence of a series of environmental catastrophes, these people adapted their subsistence strategies, moved settlements to different locations and developed different, more efficient, building techniques. In essence,

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

13

however, they retained many of the preceding cultural traits. Furthermore, the influence of the natural hazards in this case appears to have been to stimulate rather than devastate the culture. This ‘positive’ response is one that is rarely considered in studies of disasters and should be studied more widely in parts of the world that are subjected to frequent hazards over a relatively long time period. Perception Is it safe to assume that exposed communities view a hazard as a threat? Modern studies have shown that the perception of hazards is critically important to how a community reacts to a forcing mechanism (e.g. Bryant, 1991: 259–60). MbunweSamba (1999) has presented a very graphic study of how the community affected by the catastrophic Lake Nyos gas explosion has failed to come to terms with the event because the survivors have not found an acceptable explanation for the sudden deaths. In this case the absence of a clear perception has markedly delayed recovery from the disaster. On balance, the case studies presented here show that people at risk are more overtly concerned with the social rather than the natural world. For example, Johnson’s (Chapter 11) study of 87 Aleut Eskimo tales and narratives found that only eight mentioned natural hazards, a result that is confirmed by Saltonstall and Carver’s (Chapter 10) review of folklore from the neighbouring Kodiak Archipelago. The perception that earthquakes are not a severe threat in this region is borne out by the archaeological data, which fail to show any correlation between earthquake incidence and cultural change. Lowe et al. (Chapter 8) also found very little oral history among the Maori in New Zealand concerning volcanic hazards. In Chapter 9 Galipaud argues that in Vanuatu natural disasters are perceived as social rather than natural events. Furthermore, since extreme environmental events are seen as caused by humans rather than as natural occurrences, they are not feared. In modern-day Papua New Guinea, people also believe that the disastrous tsunami which they experienced was humanly generated and they were clearly uncomfortable with the explanation provided by the earth scientists (cf. Chapter 3). In other cases the forcing event was ascribed to the actions of supernatural beings or deities (e.g. Chapters 2, 8 and 10). Survivor mentality may also be a critical factor in how people recover from a disaster. Gibbs (Chapter 5) describes the experiences of the survivors of a shipwreck. From these experiences it is evident that in certain survival situations the preceding social order may be subsumed by the activities of a single group and the long-term welfare of the greater community compromised to satisfy the shortterm goals of a single segment of the society. Conflict between sections of the survivor community may then occur, with an unpredictable outcome. Driessen (Chapter 14) and Satoru (Chapter 2) also argue that the community’s perceptions of a disaster can be instrumental in how the population reacts. They also point out that different perceptions of the disaster on the part of the victims and the authorities in charge of disaster relief can lead to conflicts and/or may prolong the suffering (cf. Chapters 3 and 4).

14

R. TORRENCE AND J. GRATTAN

Vulnerability Social scientists emphasise vulnerability as the most important factor in understanding the variability in how societies react to disasters (e.g. Torry, 1979; Blaikie et al., 1994; Oliver-Smith, 1996) and their findings are echoed in most of the studies presented here. Recent works especially focus on how particular social settings create vulnerable communities by reducing their ability to recover (e.g. by creating poverty) or by placing them in hazardous settings, which would not normally be settled (e.g. floodplains, hillsides prone to landslides, etc.). Although attachment to place is identified as a factor in making Bacolor and neighbouring towns highly vulnerable to the mudflows from continuing Pinatubo, Crittenden and Rodolfo (Chapter 4) also note that the people had virtually nowhere else to go. This type of socially induced vulnerability is probably most relevant for highly complex societies in the modern world, but a very broad notion of vulnerability – in terms of the ability of a community to return to its previous state – is clearly most important in explaining the wide range of responses illustrated by the papers in this book. Are simple societies less able to respond to the pressure generated by the occurrence of natural hazards? While this may be the view of some natural scientists considering modern hazards (e.g. Chester, 1993; Blong, 1984: 186, 387), anthropologists have proposed the opposite view (cf. Torry, 1979: 523; Oliver-Smith, 1996), and the archaeological record suggests that ancient cultures were in fact highly resilient. These issues are discussed in detail in Chapter 16, where cases from prehistoric Costa Rica and Papua New Guinea are considered. In terms of the former, 10 volcanic eruptions occurred within a space of 4,000 years, yet archaeologists detected relative cultural stability (Sheets et al., 1991). One might argue, however, that the Costa Rican eruptions were relatively moderate in terms of severity and it is therefore not surprising that little cultural response has been observed. In contrast, during the past 6,000 years prehistoric groups in West New Britain province of Papua New Guinea have been exposed to a series of exceptionally large volcanic eruptions, which probably devastated the vegetation across vast areas of this island. Yet even here severity of the hazard event itself is not the sole factor determining cultural response. Clearly, in addition to the severity and scale of the natural hazard, social variables are critical to the way groups recover from and may change in response to disasters. A number of papers have discussed why the disasters that they studied had very little impact on long-term cultural behaviour. For example, Saltonstall and Carver (Chapter 10) argue that the Alutiiq were not severely affected by the relatively frequent earthquakes in the region. Since they were highly mobile and maintained long-distance contacts, relocation could be undertaken relatively easily. Adaptation of cultures to hazards is also illustrated in Chapter 8. As far as can be established, it appears that Maori culture in New Zealand was not unduly perturbed by volcanic activity. Areas under the most direct threat – i.e. proximal to the volcano – were apparently utilised for transient activities, settlement only

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

15

occurring as a last resort during times of political crisis. Indeed, the designation of zones of known or perceived risk as sacred areas placed out of bounds may be a codification of a practical necessity. In Chapter 18 we can see that in many cases Japanese communities simply adapted to the occurrence of hazards and accommodated their routine to it. Even in Japan, with its many active volcanoes whose eruptions must have caused extreme hardship, as detailed by Machida and Sugiyama (Chapter 17), Shimoyama (Chapter 18) argues that cases where outright abandonment occurs following a volcanic event may have been relatively rare. The problem here is to find out whether the differences in ceramic styles between the archaeological contexts above the volcanic tephra and those buried beneath it were a direct result of the volcanic disaster or were due to normal cultural change.

LOOKING AHEAD The incidence of disasters is said to be increasing in the modern world (e.g. Tobin and Montz, 1997: 2), and consequently the amount of resources invested in the study and mitigation of natural hazards has increased dramatically. It is perhaps not surprising, then, that archaeological speculation about the impact of disasters in the ancient world has also become very popular. Although we argue that the study of past disasters can make a very useful contribution to archaeological method and theory, because it provides an alternative to a previous focus on adaptation and stability, the case studies presented here do not present unmitigated support for the role of disasters in causing cultural change. Many human communities have occupied very risky and hazardous environments and have therefore experienced disasters relatively frequently. The impact of these events on the individuals who survived them must have been overwhelming, judging from our recent case studies. People experienced huge losses in terms of deaths of kin and friends and destruction of property. Despite the enormous damage sustained in the short term, most of the disasters studied by the authors in this volume had very little if any effect on cultural change, when viewed over more than a few generations. Of the exceptions, the Santorini eruption may have only been a catalyst in a process that was already under way (although even this is controversial), settlement change in Switzerland was slow to react, and the nature of cultural replacements in Papua New Guinea and Japan are as yet difficult to interpret. Only Kornbacher (Chapter 12) has provided evidence for a concatenation of disasters caused by El Niño to have led to significant cultural change over the very long term. Her paper raises the important issue that most archaeologists have been expecting disasters to cause cultural collapse or breakdown rather than evolution. Previous work may have focused too narrowly on individual events rather than view the risky environment itself as something which could shape cultural change. Adopting this more positive approach might lead to productive reanalyses of cultural processes among communities living in risk-prone environments.

16

R. TORRENCE AND J. GRATTAN

The key variable identified in these studies has been what is loosely called in modern disaster research the ‘vulnerability’ of the societies in question. Mobile subsistence and settlement patterns and/or extended social networks appear to have been particularly important because they have enabled groups to resist change by moving temporarily to other areas. Alternatively, some groups made huge efforts to resist change or put themselves at risk because they had a strong attachment to place (or nowhere else to go). Why do some groups emigrate and others stay? More detailed research is needed to understand the myriad range of ways that societies have coped with the disasters that they have faced. We also have a very poor understanding of how groups assess risk and how they respond to it. Most of the studies suggest that even when they are recognised, long-term risks are ignored in favour of short-term economic gains (cf. Grayson and Sheets, 1979: 626). If this is the case, do communities take account of the risks by maintaining memory of how to adapt through story-telling or other behaviour patterns that are only rarely called into use? Or are certain behaviours which enable flexible responses, such as mobility and exchange networks, actually long-term outcomes of living with disasters? Much more research is necessary to address these questions using events with suitably long time spans. A more flexible approach to the question of what constitutes a ‘significant’ response to disasters would also provide productive research. To date past disasters have either been totally ignored by archaeologists or used in a very uncritical way to account for cultural change. This ‘all or nothing’ approach glosses over what must have been a wide range of responses to disasters of varying magnitudes and frequencies by groups with different social and economic structures. Some of that range is illustrated in this volume, which demonstrates the value of looking at recent cases as well as those represented in the archaeological record, but many more detailed and critical studies are required before we have enough data to adequately assess the role of disasters in human history. It is hoped that these studies demonstrate the importance of disasters in raising questions about human adaptation and change and will pave the way to further research.

REFERENCES Ambrose, S. (1998) Late Pleistocene human population bottlenecks, volcanic winter, and differentiation of modern humans. Journal of Human Evolution 34: 623–51. Blaikie, P., Cannon, T., Davis, I. and Wisner, B. (1994) At Risk: Natural Hazards, People’s Vulnerability, and Disasters. London: Routledge. Blong, R. (1982) The Time of Darkness: Local Legends and Volcanic Reality in Papua New Guinea. Seattle: University of Washington Press. Blong, R. (1984) Volcanic Hazards: A Sourcebook on the Effects of Eruptions. Sydney: Academic Press. Bryant, E. (1991) Natural Hazards. Cambridge: Cambridge University Press. Byrne, D. (1997) The archaeology of disaster. Public History Review 6: 17–29. Byrne, D. (1999a) Human disasters and heritage lies. Paper presented at the Fourth World Archaeological Congress, Capetown, South Africa.

TRENDS IN THE ARCHAEOLOGY OF DISASTERS

17

Byrne, D. (1999b) In Sad but Loving Memory: Aboriginal Burials and Cemeteries of the Last 2000 Years in NSW. Sydney: NSW National Parks and Wildlife. Chester (1993) Volcanoes and Society. London: Arnold. Cullen, H., de Menocal, P., Hemming, S., Hemming, G., Brown, F., Guilderson, T. and Sirocko, F. (2000) Climate change and the collapse of the Akkadian empire: evidence from the deep sea. Geology 28: 379–82. Driessen, J. and Macdonald, C.F. (1997) The Troubled Island. Minoan Crete Before and After the Santorini Eruption (Aegaeum 17). Liège and Austin: Université de Liège. Fagan, B. (1999) Floods, Famines, and Emperors: El Niño and the Fate of Civilizations. New York: Basic Books. Giller, R. (2001) The Great Maya Drought. Albuquerque: University of New Mexico Press. Glickman, T., Golding, D. and Silverman, E. (1992) Acts of God and Acts of Man: Recent Trends in Natural Disasters and Major Industrial Accidents. Center for Risk Management, Discussion Paper 92-02. Washington, DC: Resources for the Future. Gould, S. (1989) Wonderful life. The Burgess Shale and the Nature of History. New York: W.W. Norton. Grayson, D. and Sheets, P. (1979) Volcanic disasters and the archaeological record. In P. Sheets and D. Grayson (eds) Volcanic Activity and Human Ecology, 587–622. New York: Academic Press. Harris, S. (2000) Archaeology and volcanism. In H. Sigurdsson (ed.) Encyclopedia of Volcanoes, 1301–14. San Diego: Academic Press. Isaacson, J. and Zeidler, J. (1999) Accidental history: volcanic activity and the end of the formative in northwestern Ecuador. In P. Mothes (ed.) Actividad Volcanica y Pueblos Precolombinos en el Ecuador, 41–72. Quito: Ediciones Abya-Yala. Keys, D. (2000) Catastrophe: A Quest for the Origins of the Modern World. New York: Ballantine. Mbunwe-Samba, P. (1999) The Lake Nyos catastrophe. Was it man-made or a natural disaster? What do non-scientists say? Paper presented at the Fourth World Archaeological Congress, Capetown, South Africa. McCoy, R. and Heiken, G. (2000) Volcanic Hazards and Disasters in Human Antiquity. Geological Society of America Special Paper 345. McGuire, B. (1999) Apocalypse. A Natural History of Global Disasters. London: Cassell. McGuire, W., Griffiths, D., Hancock, P. and Stewart, I. (eds) (2000) The Archaeology of Geological Catastrophes. London: Geological Society Special Publication 171. Moseley, M. (1997) Climate, culture, and punctuated change: new data, new challenges. The Review of Archaeology 18: 19–27. Newhall, C. and 17 others (2000) 10,000 years of explosive eruptions of Merapi volcano, Central Java: archaeological and modern implications. Journal of Volcanology and Geothermal Research 100: 9–50. Nur, A. and Cline, E. (2000) Poseidon’s horses: plate tectonics and earthquake storms in the late bronze age Aegean and eastern Mediterranean. Journal of Archaeological Science 27: 43–63. Oliver-Smith, A. (1986) The Martyred City: Death and Rebirth in the Andes. Albuquerque: University of New Mexico Press. Oliver-Smith, A. (1996) Anthropological research on hazards and disasters. Annual Review of Anthropology 25: 303–28. Peiser, B., Palmer, T. and Bailey, M. (eds) (1998) Natural Catastrophes During Bronze Age Civilizations: Archaeological, Geological, Astronomical and Cultural Perspectives. British Archaeological Reports International Series 728. Oxford: Archaeopress. Sadler, J. and Grattan, J. (1999) Volcanoes as agents of past environmental change. Global and Planetary Change 21: 181–96. Schoch, R. and Aquinas, R. (1999) Voices of the Rocks: A Scientific Look at Catastrophes and Ancient Civilizations. New York: McNally.

18

R. TORRENCE AND J. GRATTAN

Sheets, P. and Grayson, D. (1979) Introduction. In P. Sheets and D. Grayson (eds) Volcanic Activity and Human Ecology, 1–8. New York: Academic Press. Sheets, P., Hoopes, J., Melson, W., McKee, B., Sever, T., Mueller, M., Cheanult, M. and Bradley, J. (1991) Prehistory and volcanism in the Arenal area, Costa Rica. Journal of Field Archaeology 18: 445–65. Siebe, C., Abrams, M., Macias, J. and Obenholzner, J. (1996) Repeated volcanic disasters in prehispanic time at Popocatepetl, central Mexico: past key to the future? Geology 24: 399–402. Stiros, S. and Jones, R. (eds) (1996) Archaeoseismology. Fitch Laboratory Occasional Paper 7. Athens: British School of Athens. Terrell, J. (1988) History as a family tree, history as an entangled bank: considering images and interpretations of prehistory in the South Pacific. Antiquity 62: 642–57. Tobin, G. and Montz, B. (1997) Natural Hazards: Explanation and Integration. London: The Guilford Press. Torrence, R., Pavlides, C., Jackson, P. and Webb, J. (2000) Volcanic disasters and cultural discontinuities in the Holocene of West New Britain, Papua New Guinea. In B. McGuire, D. Griffiths and I. Stewart (eds) The Archaeology of Geological Catastrophes, 225–44. London: Geological Society Special Publication 171. Torry, W. (1979) Anthropological studies in hazardous environments: past trends and new horizons. Current Anthropology 20: 517–41. Van der Leeuw, S. and McGlade, J. (eds) (1997) Time, Process, and Structured Transformation in Archaeology. London: Routledge. Weiss, H. and Bradley, R. (2001) What drives societal collapse? Science 291: 609–12. Zeidler, J. and Isaacson, J. (in press) Settlement process and historical contingency in the Western Ecuadorian Formative. In J. Raymond and R. Burger (eds) Dumbarton Oaks Conference on the Archaeology of Formative Ecuador. Washington, DC: Dumbarton Oaks Research Library and Collection.

2

Basic characteristics of disasters SATORU SHIMOYAMA

IMPORTANCE OF DISASTER STUDIES There is no doubt that disasters have affected current and past human behaviour. Historical studies have elevated our awareness of the significance of disasters and provided guides about measures that have been used to protect against future occurrences (Noto, 1993). It is also possible that cultures may share certain characteristics because they have adapted to a series of disasters in the past and have developed similar adaptations to protect themselves from the threat of further occurrences (Shimoyama, 1998). Additional studies of how societies have adapted to disasters are needed to generate general theory concerning the effects of disasters on cultural change. Archaeological discoveries of traces that indicate past disasters are becoming very common. Before these data can contribute to a broader, theoretical understanding of the phenomena in general, a series of basic concepts about what a disaster is and how it occurs is necessary. The establishment of the major components of a disaster is therefore the logical first step in an analysis. Second, it is important that archaeologists go beyond simply reporting that a disaster took place, as is common in many reports. They should also reconstruct the specific conditions present, the nature and extent of damage, the assessments made by the people involved, and the responses of the population, including longer-term adaptations, if any. Obviously, this detailed level of work will require much better excavation techniques and data analysis, but at this stage it is important to set out the general criteria that constitute a disaster as a guide to what archaeologists should be looking for in their data. The purpose of this short chapter, therefore, is to introduce the basic concepts necessary for constructing archaeological research that will assist in building general theory about the role of disasters in cultural change. I will introduce and describe the basic components of a human disaster and provide some examples from Japanese archaeology. More detailed case studies are presented in Chapter 18. One of the important implications of defining the general components of disaster research is that the need for an inter-disciplinary approach is highlighted.

20

S. SHIMOYAMA

Studies focused on the natural origin of a disaster and the kinds of damages sustained have traditionally been carried out in the natural sciences in fields such as volcanology, geology and biology. In contrast, students of the humanities and social sciences, such as archaeology, anthropology, history, folklore, sociology and psychology, have usually reconstructed the processes of how people assess and react to the damage sustained. I argue that the extensive knowledge and technical skills derived from both the natural and humanistic studies of past disasters need to be brought together to understand cultural change following a severe natural event. The final consequence of elucidating the basic characteristics of disasters is that the mechanism of cultural change is specified. This clarifies the relationship between the initial natural event and the cultural outcome. Furthermore, it is also clear that the causes of cultural change following a minor disaster are often easier to study than the massive events that have generally been emphasised in archaeology. It would be beneficial for a general study of natural catastrophes if archaeologists put more effort into understanding a wider range of cases.

BASIC CONCEPTS Disasters are social phenomena. Although they are initially caused by either natural or human activities, the required condition for a disaster is the existence of victims: i.e. there must be direct or indirect damage sustained by humans. When no one suffers a loss, then we are dealing solely with a natural phenomenon. The nature and extent of the damage results from the interactions among the type and severity of the initiating event, local conditions, and the specific cultural context. A summary of the basic components of a disaster is presented in Table 2.1. An example of how a specific case can be analysed using these general concepts is presented in Table 2.2. The example used is a volcanic eruption that took place at Kaimondake, Japan on 25 March, AD 874. Table 2.2 also shows the fields of study that are used in the analysis and the various types of data that are needed. These columns in particular highlight the importance of inter-disciplinary research in the study of past disasters. Initiation Although the initial factor in a disaster may be human action, in line with the theme of this book, I will concentrate on disasters caused in the first instance by extreme natural events. Nevertheless, the basic characteristics of naturally and culturally initiated disasters are fundamentally the same. These are (1) initiation, (2) immediate causes, (3) local conditions, (4) damages, (5) assessment, and (6) actions. The initiation, immediate causes and local conditions have been central to the study of natural disasters (e.g. Katsui, 1979; Endo, 1990), with the initiating factor as foremost. Examples of these include volcanic eruptions, earthquakes, tsunamis, abnormal precipitation leading to flooding, avalanches, etc. The case study presented in Table 2.2 was initiated by a volcanic eruption.

BASIC CHARACTERISTICS OF DISASTERS

21

Table 2.1 Basic characteristics of natural disasters Characteristic

Definition

Initiation

The original process or event that set the scene for the disaster, e.g. extreme natural events such as volcano, flood, earthquake and tsunami

Immediate causes

Specific aspects of the event, which have direct effects on human life and property

Local conditions

Natural and socio-cultural variables that establish the local setting at the time of the event

Damages

Concrete negative effects

Assessment

Process in which victims and observers assess the extent and repercussions of the damages

Actions

Acts carried out following the disaster. These include both short-term actions, such as abandonment and cleaning up, etc., and long-term adaptations, such as moving to a new area, collapse of social system, institution of preventive measures, adoption of ritual practices to avoid further occurrence of disasters, etc.

Immediate causes The immediate causes of a disaster are the specific phenomena, which have direct effects on human life and property. These are obviously specific to the type of event that occurred and the measurement of the severity of these will require a set of variables designed specifically for them. For example, earthquakes can be measured by their strength on the Richter scale, tidal waves have various heights and speeds, and for a volcanic eruption, which is often ranked by the Volcanic Explosivity Index (VEI), it is important to monitor the types of tephra (e.g. pyroclastic flow, lava, bombs, ash, etc.) and their composition, temperature and thicknesss. Studies of the tephra from the Hashimuregawa site, which is about 10 km from the volcano, and other volcanological research has shown that the Kaimondake eruption on 25 March, AD 874 had a VEI of 4. As evidenced by the stratigraphy at the site, we know that there were several eruptive events with 9–15 cm of airfall tephra emplaced during a single day. The first event was the most powerful. The airfall tephra consisted of fine ash and small particles ranging in size from 5 to 10 mm. By the time the airfall material reached the site, it was relatively cool. Local conditions Local conditions can magnify or reduce the effects of the initiating factor. For example, wind direction and amplitude (which may be seasonal) can extend or contain the effects of volcanic airfall tephra and gases. Social conditions, including local knowledge and prior experience, type of subsistence, presence of food stores and/or other support systems, communication networks, nature of the social system, etc. can either ameliorate or amplify the damage done. Furthermore, the time of the day or year and the length of time between the onset of the initiating

By victims

Actions

Actions taken

Actions taken Inspection

Site location, subsistence, social organisation, etc. Previous behaviour

Socio-cultural

By victims By government

Weather conditions (rain, wind velocity, etc.)

Natural

Local conditions

Assessment

Documents

Rate of tephra accumulation

Human facilities such as houses, fields, paths Plants Animals

Tephra Tephra

Extent, pressure, adherence, hardness of tephra Thickness, weight, hardness, range of tephra

Character Degree

Immediate causes

Damages

Mechanism and scale of eruption Documents Chronomentric dates

Eruption of Kaimondake on 25 March, AD 874

Initiation

Site abandoned

Excavated data Documents

Excavated data Leaf stamps Animal bones

Excavated data Documents

Tephra

Evidence

Hashimuregawa site

Characteristic

Table 2.2 Case study of a disaster: basic characteristics, types of evidence and fields of study

Archaeology

Archaeology History

Archaeology Botany Archaeology, Zoology

Volcanology, Archaeology Archaeology History

Volcanology Volcanology, Archaeology History

Volcanology History Archaeology

Discipline

BASIC CHARACTERISTICS OF DISASTERS

23

factor and impact of its effects are just some of the many local conditions that are relevant for causing damage. The Hashimuregawa site, which was impacted by the eruption of Mt Kaimondake, was located on a alluvial fan where agriculture was particularly productive. At the time of the eruption a very strong wind was blowing from the southwest to the northeast. From the analysis of buried dwellings and a small river, it is clear that heavy rain on the day led to destructive mudflows and avalanches. Damages Evidence of damage is often well preserved, especially if it is severe and is followed by abandonment. Consequently, a detailed analysis of the nature and extent of the damages generally form the core of historical and archaeological studies of disasters. Key examples are shipwrecks (Chapter 5), and cities buried under volcanic tephra such as Pompeii in Italy (Chapter 7) and Santorini in Greece (Chapters 14 and 15). The site of Itatsuke, which was covered by sand during a flood in the Yayoi era (c. third century BC), is another good example from Japan (Yamasaki, 1978). Damages to facilities and crops can be understood from the Hashimuregawa site using archaeological, botanical, zoological and historical data. For example, archaeological excavations have shown that the direct impact and weight of the airfall tephra caused widespread destruction to buildings and crops (Fig. 2.1). Moulds of leaves preserved in the tephra demonstrate that the adhesion and hardening of the very fine ash anihilated crops and other vegetation. Mudflows and avalanches also knocked over, buried and infiltrated buildings and completely obliterated a small stream (Fig. 2.2). The damage to crops caused an immediate food shortage which is reported in a historic document entitled ‘Nihon Sandai Jitsuroku’. Assessment How people react and adapt, if at all, to a disaster is obviously determined largely by their assessment of the nature and extent of the damages. People who have sustained the damage, i.e. the victims, make direct assessments of post-disaster recovery and the appropriate strategy to be adopted. In many cases people who are not directly involved in the disaster, but who may have an effect on the response (e.g. relatives, group leaders, or government officials located elsewhere), also assess the situation. Often there are major differences in the actions requested by the victims and those actually taken by government because their assessments are not in accord with one another (cf. conflicts between local people and government officials in Chapters 3 and 4). Differences between those who were involved and those who have the power to make changes can prolong or worsen the disaster. In some respects the victims’ assessment of the damages can be considered as part of their adaptation to the disaster and so the assessment and action stage cannot always be separated into different components. Archaeologists often use the actions taken after a disaster as an indication of what assessments were made.

Figure 2.1 Field and path buried under volcanic tephra at the Hashimuregawa site

Figure 2.2 Trace of a small river, which was buried by an avalanche

BASIC CHARACTERISTICS OF DISASTERS

25

Historical documents and oral history in which people describe the events and the aftermath can provide valuable information about how victims have assessed a disaster, but it is also important to compare these with archaeological information about what action was actually taken as a consequence of this assessment (Nagayama, 1992; Shimoyama, 1997, 1998; cf. Chapters 9, 10 and 11). One also has to be wary of assessments made by people who did not actually suffer the direct effects of the disaster and those writing many years after the event. Studies of documents dating to 30 years after the Kaimondake eruption indicate that government officials had assessed the cause of the disaster to have been the anger of a deity. The assessment of the villagers can only be assumed by their subsequent actions. Actions Actions following a disaster will vary widely, as illustrated by the studies presented in this volume. In the case of the Hashimuregawa site, the historical documents record that the government contributed feudal estates in order to soothe the savage beast which was thought to have caused the eruption. Archaeological excavations show that the victims themselves abandoned the site. In other cases in Japan where a different assessment was made, tephra was cleared away and facilities were reconstructed. Innovative behaviour is also illustrated, with tephra put to a productive use in the construction of new roads. In some cases there will be no perceptible change to the culture following a severe natural event, whereas in other situations there may be marked differences before and after. The scale and nature of changes observed will also depend on whether one is concentrating on a short- or long-term time scale. A group may abandon an area for a short period to avoid some of the effects of the disaster. If damages are inconsequential, then they may return and carry on as if nothing had happened. In this case in archaeological time scales, it will appear that the disaster had no consequences at all. Even if people did experience severe effects, the group might be able to repair damages and continue with seemingly no alterations to their behaviour. In the extreme cases a region may be totally abandoned for a long period and then reoccupied by a totally different cultural group. Some human groups are extremely adaptable in the face of disasters and some may be more tolerant of environmental perturbations than others. It may also happen that the nature of the event may enhance the qualities of the local area, for example by the emplacement of nutrient-rich volcanic tephras or alluvial soils, and so changes after the event are not recognised as a consequence of a disaster. An example is a paddy field made by sand from a flood at the site of Shima-batake in Osaka Prefecture (Eura and Nagahara, 1992). Although there may be no obvious cultural changes immediately after the disaster, over the long term people may make changes to avoid further disasters. An interesting reaction to disasters is the attempt made to avoid further incidents through concrete means, such as reinforcing walls in earthquake zones, developing the means of detecting the onset of natural events, or the disaster organisations in most modern states that monitor situations, make plans for

26

S. SHIMOYAMA

events, and educate people who live in threatened areas. Spiritual means were also taken. As reported in ethnographic and historical sources, many societies undertook ritual acts or ceremonies to prevent disasters (Yamamoto, 1996).

IMPLICATIONS AND CONCLUSIONS For archaeology the most interesting aspect of disaster research is the enormous variation in the responses of groups despite similarities in the initiating factors. Clearly, social groups make differing assessments about how to react to the damages sustained. In some cases people will fight to maintain themselves within an area despite widespread destruction of property. They will return reasonably quickly to their homes, fields and other significant places again and again following a disaster. At other times and places, the response will be totally different. A group faced with what seems a similar disaster may leave or radically change its behaviour in some way. Clearly, the cultural response to a catastrophic natural event is not solely linked to the physical factors. The nature of the group itself and how it assesses the damages are at least as, if not more, important in determining the long-term outcomes that archaeologists study. Although disasters vary widely in scale, I think that it is valuable and important to establish a number of elements that are common to all cases and to propose a standard framework for their analysis. These concepts can be used to describe a particular event or to compare and contrast a number of them. In addition, the variables illustrate the temporal and spatial components of a disaster. The basic characteristics which I have listed include both concrete, natural variables that are traditionally measured by physical scientists as well as human assessments and responses. I have tried to show that an inter-disciplinary approach is necessary for studying disasters and that social sciences and humanities have a very important role to play in this work. It is hoped that this basic framework will be useful in future comparative discussions and will help archaeologists to frame theories to explain the role of disasters in human evolution.

REFERENCES Endo, Kunihiko (1990) Impact of tephra-producing eruption of land surfaces. Quaternary Research (Japan) 30: 399–408. Eura, Hiroshi and Nagahara, Wataru (1992) Kinsei Suidenmen nimiru Saigai Fukkyu. Osaka Bunkazai Kenkyu 8: 35–47. Katsui, Yoshio (1979) Funkasaigai, Iwanamikouza Chikyukagaku 7 – Kazan – Iwanami Shoten, 83–98. Tokyo: Iwanami Shoten. Nagayama, Shuichi (1992) On the evidence of the volcanic activities of Mt Kaimon obtained from the ‘Nihon Sandai Jitsuroku. In The report of the Hashimuregawa Archaeological site 3: 501–10. Ibusuki City: Board of Education. Noto, Takeshi (1993) Koukoiseki ni miru Joushu no Kazansaigai, Kazanbai koukogaku, 54– 82. Tokyo: Kokin Shoin.

BASIC CHARACTERISTICS OF DISASTERS

27

Shimoyama, Satoru (1997) On the range of the disaster archaeology. Hominids 1: 83–103. Tokyo: Congress of Reconstructing Archaeology. Shimoyama, Satoru (1998) Issues on the disaster assessment. Retto no koukogaku, 713–32. Tokyo: Watanabe Makoto Sensei Kanrekikinen Ronshu Kankoukai. Yamamoto, Hiroto (1996) Negai to Inori noShohyousou, Heisei nana nendo Maizoubunkazai Gaiyo, 88–92. Toyama, Japan: Zaidanhouzin Toyama-ken Bunka-shinko Zaidan, Maizou Bunkazai Chousa Jimusho. Yamasaki, Sumio (1978) Fukuoka-ken Itatsuke-iseki no Jomon-jidai Suidenshi. Gekkan Bunkazai 181.

3

Tsunamis and the coastal communities of Papua New Guinea HUGH DAVIES

ORAL HISTORY AND TSUNAMIS Tsunamis represent a rare, rapid-onset phenomenon, which present a hazard in many coastal areas of the world, in particular all areas of the Pacific, the Mediterranean basin, and in rare extreme cases the North Atlantic and the North Sea. The study of recent tsunami events, their impact and the subsequent communal response is vital if archaeologists are to construct robust paradigms by which to interpret and understand the influences that tsunamis may wield upon cultural development. This chapter explores the communal response to tsunami hazard and the degree to which knowledge of the hazard is incorporated in oral history. This chapter, by studying recent events in Papua New Guinea, sheds considerable light on cultural responses and the relevant decision-making process. On the evidence presented here, it appears that, unlike volcanic hazards, knowledge of these events does not play a major part in decisions to build or rebuild in particular areas. If the last 150 years are a reliable guide, a major tsunami has devastated some part of the north coast of the island of New Guinea or the adjacent islands every 15–70 years. This is likely to have been the case throughout the tens of thousands of years of human habitation, given that the geological circumstances, the factors that generate tsunamis, have not changed in that time. If major tsunamis are such regular events, it is logical that they will figure in the culture of the coastal peoples. But just how prominently do tsunamis figure in the traditional cultures in this region? Do communities develop defensive strategies and do they maintain awareness? For example, will a community permanently abandon a site that has been devastated by a tsunami and move to a safer location? Or will they reoccupy the site? Will the warning signs of tsunami become part of the culture and will this knowledge be passed down from generation to generation? Or will the knowledge be dissipated with time? The catastrophic tsunami that struck the Aitape coast in July 1998 (Fig. 3.1) provided an opportunity to explore these questions through interviews with survivors and observation of resettlement trends. In brief, I found that the people affected by the Aitape disaster did have an oral

Reproduced with permission from the University of Papua New Guinea

Note: Tsunami modellers (e.g. Y. Tsuji in Matsumoto et al., 1999) concluded that the submarine Yalingi canyon served to focus the energy of the wave on the 14 km sector of coastline from Mak to Sissano Government Station

Figure 3.1 Villages affected by the Aitape tsunami and the resettlement villages (filled circles)

30

H. DAVIES

history of tsunamis but this knowledge was not widely dispersed amongst the community, although it was preserved amongst the older males: ‘My father told me about these things.’ Despite the fact that this traditional knowledge existed, the villages were not located in areas that were safe from tsunamis. The worst-hit villages were located 50–150 m from the water’s edge on low-lying sand spits backed by swamp or lagoon, in locations from which there was no means of quick escape. They were tsunami death-traps. Nor did the people recognise the warning signs. The giant waves of 17 July 1998 took people completely by surprise. Had they recognised the signs, fewer would have perished. It remains uncertain whether people will return to the devastated area, but the early indications (after 24 months) are that they may do so, in time. Already, despite the fact that permanent facilities such as schools, housing for teachers, and aidposts have been built inland, there is some drift back to the coastal villages, especially those villages that suffered least damage by the waves. From this example, we might conclude that a natural disaster, even one as severe as Aitape, is a fleeting event, a blip on the horizons of time and human memory, that has a negligible effect on long-term settlement patterns. It also seems that lessons learned in one disaster are not preserved and passed on to future generations unless the same disaster recurs with sufficient frequency, e.g. at intervals of perhaps 50 years or less. These conclusions are true for the Aitape coast in modern times, but are not necessarily true for other coasts, nor for earlier times when people were less mobile and traditional societal bonds were stronger. For example, the application of lessons learned at the time of the Ritter Island tsunami of 1878 (Everingham, 1977: 15–20) resulted in reduced loss of life during Madang-Bogia tsunami in 1930 (Ibid.: 40) and may have caused one settlement on northeastern Umboi Island to be permanently relocated inland.

THE NATURE OF TSUNAMIS A tsunami is a rhythmic movement of the sea that originates in any displacement of the sea floor. The movement of the sea floor, whether up or down, displaces the water column above that point. Waves then radiate out in all directions from the place where the water column was disturbed. These waves differ from normal ocean waves in that the entire water column moves, from sea floor to surface, whereas normal wind-driven waves and ocean swells affect only the top layer of the ocean. In deep water the tsunami travels imperceptibly, with low amplitude, long wave length, and high speed. Only when the tsunami reaches shallow water do the waves slow down and become higher and steeper. When the tsunami reaches the coast, it is likely to be travelling at 10–15 m per second (35–50 km per hour). This is significantly faster than a normal wind-driven wave and is faster than most people can run. If a tsunami originates in the local region, for example in the Bismarck Sea or Solomon Sea in the case of Papua New Guinea (PNG), it is referred to as a near-

TSUNAMIS IN PAPUA NEW GUINEA

31

source tsunami. If it originates at a great distance across the Pacific, for example in Alaskan or Chilean waters, it is termed a far-source tsunami, or tele-tsunami. A near-source tsunami generated in PNG waters will strike the nearest coastline within 10 to 20 minutes. A far-source tsunami generated off Chile will reach PNG waters after 20–22 hours and one generated in Alaska, after 10–12 hours. The great majority of the more than 70 tsunamis that have been recorded in PNG are near-source tsunamis (Everingham, 1977). For far-source tsunamis it is possible to give a warning. The Pacific Tsunami Warning Centre in Honolulu monitors all earthquake activity in the Pacific and automatically transmits a tsunami alert whenever an earthquake of magnitude 7 (Richter scale) or greater occurs. Although the programme is primarily directed towards the safety of the west coast of North America, it is a valuable aid for all the Pacific nations. However, for near-source tsunamis there is not enough time to give a warning, because there is such a short interval of time between tsunami triggering and arrival on the coast. This is true everywhere in the Pacific except on certain parts of the Japanese coast where near-shore detection systems can give up to five minutes’ warning of an impending tsunami. Since no official warning can be issued, the only effective protection against near-source tsunamis is for populations to recognise the warning signs and to know that they must move immediately to a safe place. This is the thrust of current public awareness programmes in PNG and elsewhere. The warning signs are ● ●

●

Any strongly felt earthquake; A roaring sound like a jet plane, sometimes described as the sound of gunfire, or of a large helicopter; and An unusual lowering of sea level, to below the normal low-water mark.

The cause of the roaring noise is not agreed amongst the experts. For whatever reason, the negative or low part of the wave tends to arrive before the positive or high part, and there is a drop in sea level. The roaring sound and the drop in sea level are commonly observed for both near- and far-source tsunamis. All three of the warning signs preceded the 1998 Aitape tsunami, but were not recognised and so not heeded. Only when people saw the first wave looming, some hundreds of metres from the shore, did they turn and run. The same warning signs were noted at the time of severe tsunamis in the past. For example, the extreme tsunami that originated from the collapse of Ritter Island in March 1888 was preceded by withdrawal of the sea and accompanied by loud noises that were likened to rifle fire by one observer (Everingham, 1977: 18).

THE 1998 AITAPE TSUNAMI The Aitape tsunami has been studied more than any previous tsunami anywhere in terms of its effects, its origins and its aftermath (e.g. Davies, 1998; Kawata et al., 1999; Tappin et al., 1999). Information about the tsunami is presented here to give the reader some idea of the scale and scope of damage that a major

32

H. DAVIES

tsunami can wreak and of the way people may respond in the months and years after such an event. The tsunami mainly affected the coast west of Aitape. This strip of country comprises a broad and partly swampy coastal plain, interrupted by a number of isolated hills, and traversed by several boulder-filled, braided rivers (Fig. 3.1). The plain is bounded southwards by the steep-fronted Bewani and Torricelli Ranges at a distance of 10–15 km from the coast, and westward by the Sera Hills. Sissano Lagoon, 25 km west-northwest of Aitape, is a large lagoon, about 10 by 5 km but only a few metres deep, and is separated from the ocean by a narrow sand bar, 100–150 m across. The lagoon was formed by subsidence during and after a major earthquake in 1907, as was reported by Neuhauss (1911: 25–6) and A.B. Lewis (Welsch, 1998: 127–31). The submerged stumps of the houses of the original Warapu village, which was flooded at that time, can still be seen beneath waters of the lagoon. In 1998, approximately 12,000 people lived in villages that stretched along the coast from Malol in the east to Sissano in the west (Fig. 3.1). Most were in four major villages or groups of villages, at Malol (4,000 people), Arop (2,500), Warapu (2,500) and Sissano (2,500). The houses were built in traditional materials. Most were within 100–400 m of the ocean and only 1–2 m above high-water mark. The people subsisted on fishing and sago, with some gardening in the drier hinterland, especially at Malol. Fish and shellfish were recovered from the lagoons and fish from the open ocean. Villagers visited the town of Aitape from time to time to sell produce and buy store goods and fuel, travelling by road from Malol or by sea with powered dinghy. On the evening of Friday 17 July 1998, families in the coastal villages were gathered for the evening meal. Young people were playing touch football and planning social functions to mark the holiday long weekend and the beginning of a week of school vacation. Youths from the Malol villages set out on the long walk along the beach to attend a dance at Arop. At 6.49 pm, 12 minutes after sunset, there was a strong earthquake, sufficient to cause some damage at the old church at Sissano and to cause the Sisters at Malol to worry whether their water tanks might topple. In the Arop villages and at Warapu cracks developed in the ground and muddy water bubbled to the surface, as water-bearing sediments just below the surface lost their strength and liquefied. There was a loud boom, like thunder, and some minutes later a roaring sound, variously described as the noise of a low-flying, large jet plane or as the woopwoop-woop of a heavy helicopter. People who had gone to the water’s edge to look for the source of the noise saw the sea recede to below normal low-water mark and then saw a wave develop. Most turned and ran for safety, but almost all were caught by the wave, which approached rapidly, rose to tree-top height, and then crashed down on their villages (Fig. 3.2). Only a few people at Warapu saw the wave approaching from the east and had time to take to canoes or climb trees and escape. Those caught in the waves were vigorously tumbled and turned in water that was laden with sand and debris. Some recalled being lifted to tree-top height, then

TSUNAMIS IN PAPUA NEW GUINEA

33

Figure 3.2 House posts at the site of Warapu village Note: Most traces of the village, which had around 2,500 people, were removed by the tsunami waves Photograph:

H. Davies

descending to the lagoon floor; others that they ‘saw the sky three times’. Most were stripped of their clothing; all were bruised; and most had lost skin by sand abrasion. Those who were fortunate were carried into the lagoon and were able to find floating logs to cling to. An infant was deposited miraculously on the floating roof of a house. Those less fortunate were carried into the mangroves that fringe the lagoon, where some were impaled or were buried under piles of logs and debris. Some who had survived the initial impact were swept out to sea as the waters receded. Most had ingested water into their lungs. Thirty-five minutes after the initial earthquake, at around 7.25 pm, the destruction had ceased and calm had returned. It was now pitch dark. There was no moon and, in the Sissano area, the stars were blotted out by a low haze. The silence was broken only by the cries of the injured and of searchers looking for loved ones.

THE IMMEDIATE AFTERMATH With the passage of time, it is difficult to conjure up the terror and bewilderment of the victims of the tsunami. None had experienced a tsunami before and few had ever heard of them. The shock, distress and pain of the tsunami experience

34

H. DAVIES

were compounded by uncertainty and fear. What was this strange phenomenon and why had it struck them? What would happen next? Would there be more waves? After the waves had passed and the sea and lagoon were calm again, the survivors began to search for family and friends and to make their way to safety. Although some houses were still standing, notably in the Sissano and Malol villages, most survivors did not want to remain near the waterfront because of the fear of another tsunami. Their fears were compounded by the continued shaking of the earthquake aftershocks. Holding to each other because of the darkness, small parties of survivors stumbled through the mangroves that fringe the lagoon, with no light and no bush knife, until they reached dry land. Others who were more fortunate used boats to start rescuing those who were in the water. The dead were moved to the mangroves at the edge of the lagoon, and the survivors ferried to dry land at Olbrum, the Aroporo boat landing (Fig. 3.1) and other points around the lagoon or places accessible by the river channels that enter the lagoon. People from the Arop villages gathered at the former village site and at the Community School. Where houses were still standing, clothing and food were recovered for the needy. For most, however, this would be the first of several nights with limited food, water, clothing and shelter and, in many cases, open wounds and fractures. Survivors at Malol made their way to the Mission, where the Sisters provided temporary shelter and such emergency treatment for the injured as they could. In a letter written on 23 July 1999, Sister Margaret Conway describes how they heard a roaring sound and then the waves striking the beach and entering the lagoon, where a seiche wave developed and the water was swishing back and forth for some time. They ran to help a mother who cried out because she had lost her infant. As they walked back to their house, they saw people walking towards them. It was at that moment that I realised that we had a terrific disaster on our hands. The first family was our prayer leader Bernard and his family. They were in shock. Their house was on the beach, they were inside, Bernard getting the Sunday liturgy ready and his children sleeping. The house just disappeared from under them. By some miracle he was able to rescue them all. They had cuts and bruises but that was nothing. Then the line to our house grew larger and larger. People with huge cuts which needed stitching, big wounds from where some had tried to climb coconut trees and the skin was literally torn off them. There were broken legs, arms, collarbones and bones sticking out in some cases. (Conway, 1999) The sisters worked through the night. All around our house people gathered in stunned silence and all I could do was give them cups of water and move about amongst them.

TSUNAMIS IN PAPUA NEW GUINEA

35

Lots went into the church and prayed all night. It was an inspiration to just be there with them in their distress. (Conway, 1999) People searched through the night for those who were injured or could not move. At 2 am two young men arrived at the mission, their hair matted and ears filled with sand. They were the sole survivors of the party that had been walking to Arop for the dance. After the wave struck, they had searched frantically for their friends but without success. The bravery of those who worked to rescue others through the night has received little recognition subsequently and is all the more remarkable because the rescuers worked in an atmosphere of trauma and uncertainty. All were victims of the tsunami. Few understood what had happened. All were disconcerted by the aftershocks that continued to rock the area through the night. And none was sure whether more waves might follow. Total casualties have been variously estimated at between 1,700 and 2,200. The greatest loss of life was in the Arop villages, where more than 900 of an original population of 2,500 perished and at Warapu, where more than 500 of 2,500 died. The Sissano villages lost 150 of 2,500, and the Malol villages, which were on the fringe of the worst devastation, 70 of 4,000. In addition, more than 1,000 people were treated for serious injuries. Death was caused by drowning and by injuries received from impact with hard objects incorporated within the wave, such as logs, sawn timber and iron; impact with mangrove and other standing trees; by burial under logs and other debris; and by complications arising from ingestion of sea water, infection of wounds, gangrene and malaria. A large proportion of those who died were the very young and the elderly, who were less able to fend for themselves.

THE LONG-TERM EFFECTS The immediate effect of the tsunami was that people from the disaster area abandoned their coastal settlements and moved inland to temporary accommodation in the nearest villages (Rowoi, Ramo, Pou: Fig. 3.1). These and several other locations became designated care centres. Within three to four months each community, in consultation with the authorities, had selected a new location at a safe distance inland (Fig. 3.1). There was pressure from donors to make these decisions quickly so that their projects such as water supply, schools and housing could go ahead. The situation after 24 months is that the new settlements with their wellestablished schools and aidposts have an air of permanence. However, there is still some degree of instability, as is indicated by a high incidence of petty crime and incidents of anti-social behaviour, some of which are more serious than others. For example, the mission station at Malol, which was the first point of refuge for survivors at the time of the tsunami, is now abandoned because persistent crime

36

H. DAVIES

has rendered it unsafe for the Sisters to remain there. A contributing factor to the social unrest is that the survivors are undergoing a transition from being recipients of aid to becoming dependent on their own resources. One unfortunate effect of the generosity of outsiders is that people tend to lose their resilience and independence and to become jealously concerned about whether they are getting their just deserts and whether funds are being properly used and accounted for. Another factor is that people have experienced a dramatic change in lifestyle, from a purely coastal existence, with all that means in general ambience such as access to food and trade and lower risk from malaria, to an inland existence that is generally more difficult and less attractive on all counts. We can imagine that with the passage of time and as memories of the horror of 17 July 1998 fade, permanent settlements will be re-established on the coast. Working against this trend is the fact that schools and medical services are now permanently relocated at the inland sites. Also, there is strong advice from the authorities that the more vulnerable sectors of the coast should not be permanently reoccupied. Indeed, there is talk of declaring the area a national memorial park, and there are plans to construct memorials that will remind future generations of tsunami danger. The advice of experts is that the Sissano Lagoon sand spit should not be reoccupied on a permanent basis. The reason is that the sea floor offshore from the lagoon is so sculpted that it serves to focus tsunami energy on this sector of coast (Matsumoto et al., 1999; Tappin et al., 1999). Thus, any future major tsunami in this area is likely to yield a similar pattern of devastation to the 1998 event, wreaking greatest havoc on the sand spit. On the other hand, no one has any good idea of how frequently major tsunamis occur on this coast, although this is surely a factor in deciding what is acceptable risk in any area. The frequency could be of the order of once in hundreds of years (which from our preliminary investigations I think is most likely) or even thousands of years. All we can say with certainty is that there had been nothing remotely similar to the 1998 event in the previous 100 years, which is the period for which we have written records.

AWARENESS It remains to draw some general conclusions from the Aitape story and other relevant sources about the effects of tsunamis on PNG society. First, what was the level of tsunami awareness in traditional societies? We can begin to answer this question by looking at the level of awareness in the Aitape coastal communities. In the weeks following the tsunami, I began a public information and awareness programme with the object of telling the affected population what we (scientists) thought had happened and trying to put to rest some of the wild rumours that were rife (Davies, 1998). In return I was able to gather information about the tsunami and about how it was viewed. Most of the survivors with whom I spoke had no prior knowledge of tsunamis.

TSUNAMIS IN PAPUA NEW GUINEA

37

For these people the devastating tsunami was a mystical phenomenon and was seen as either a manifestation of the wrath of God or the work of a foreign power or of people from rival villages who were seeking access to ‘cargo’ (valuable goods) that rightfully belonged to themselves. Only some of the older men who had lived most of their lives in the village, rather than in paid employment elsewhere, recalled that their fathers had told them about tsunamis. One recounted for me a traditional story about a great tsunami that had buried houses beneath the sand (‘The smoking crab holes of Sissano’ in Davies, 1998). As recounted, the story did not include information about tsunami warning signs. The lack of awareness of tsunamis – that they are a natural phenomenon, occur all around the Pacific basin, and have recurred throughout the history of PNG – was a major factor in the trauma that people endured after the disaster. Because people had no understanding of what had happened, they remained in great fear of a recurrence of tsunamis in the weeks and months following the event, a fear that was fed by the propaganda of various unscrupulous religious groups. In this information vacuum, rumour and speculation were rife. Even today, despite the best efforts of myself and others, many survivors still favour the ‘bomb theory’: i.e. that the tsunami was a man-made event. Currently, we do not have a lot of information about tsunami awareness on other parts of the PNG coast. We can speculate that all coastal communities on the at-risk coastlines have an oral history of tsunamis. Probably, as at Aitape, the passing down of the oral history and hence awareness has become less effective and all-embracing in parallel with the weakening of the bonds of the traditional village society and the movement of young men away from the village to take up paid employment elsewhere. Another factor that bears on the level of awareness is the recurrence interval, the length of time between major tsunamis. As noted earlier, major tsunamis probably occur at intervals of 15–70 years in the New Guinea region. However, each tsunami affects only a particular, relatively small sector of coast. The average recurrence interval between major tsunamis on any one portion of coast is likely to be greater than 70 years and may be as much as several hundred years or more. If at any one point on the coast major tsunamis follow each other at intervals of less than 50 years or so, then it is likely that survivors of the earlier tsunami will recognise the warning signs and give the alert. This happened when a major tsunami struck the Madang-Bogia coast in 1930. People who had survived the Ritter Island tsunami of 1888, 42 years previously, recognised the warning signs and gave the alert. Natives report that a similar occurrence happened 40 to 50 years ago and there is no doubt that the knowledge of the previous event minimised casualties, in that when the sea began to recede the majority of the natives ran into the bush. (New Guinea Administration report of 1930 quoted by Everingham, 1977: 40) Similarly, although 50 people died on San Cristobal, Solomon Islands, in the

38

H. DAVIES

tsunami of 1931, the loss of life would have been greater had the people not known what to do. ‘On the first shock it is their usual custom to make their way to higher ground and in this way they avoid the wave’ (Grover, 1955, cited by Everingham, 1977: 41). There is a parallel in the voluntary evacuation of the Rabaul harbour area before the volcanic eruption of September 1994. On the day before the eruption began, many of the older people observed that the warning signs were similar to what they had experienced in the 1937 eruption, 57 years previously, and they gave the alert. We should note, however, that in this case public awareness had been extensively augmented by intensive awareness campaigns waged over the preceding decade by the provincial authorities and the Rabaul Volcanological Observatory. From these examples it appears that cultural memory of the warning signs of any natural disaster exists, but is short and most likely is limited to the span of one lifetime. Clear recollections of the warning signs of a disaster are probably not passed on to succeeding generations. The PNG Government has recognised that there is currently little knowledge about tsunamis in the community and is taking steps to promote awareness, including campaigns of public education (Fig. 3.3) and the inclusion of tsunami information in the school curriculum.

Figure 3.3 Tsunami awareness material prepared and distributed by the Papua New Guinea National Disaster Management Office in 1999 Reproduced with permission from the Director General, National Disaster Management Office, Port Moresby

TSUNAMIS IN PAPUA NEW GUINEA

39

ADAPTATION Another question that arises from the Aitape story is whether coastal societies learn from past tsunami experiences and adapt their lifestyles accordingly by permanently moving away from the known tsunami danger zones. In the Aitape case history the question is, will the disaster victims return to the coast? Aitape is not a perfect test case because the survivors are under some pressure from the authorities not to return to the coast and, in particular, not to return to the sand spit, which is regarded as particularly dangerous. Also, the villagers have been encouraged by the provision of new schools and aidposts to remain inland. Despite these arguments and inducements, however, small groups of people are already returning to the coast. One suspects that in a few more years, or decades, there could be a general relocation back to the original village sites. There is some evidence that the common reaction to disaster on the Aitape coast and elsewhere in PNG is to relocate temporarily, perhaps for several years, and then return to the original site. For example, anthropologist A.B. Lewis noted that after the major earthquake and subsidence that formed Sissano Lagoon in 1907 (Fig. 3.1), the Arop people moved inland for some months but later returned (Welsch, 1998: 128–31). In the following extract from Lewis’s diary it is likely that his references to winter and spring are to the northern hemisphere seasons, so that winter is January–February and spring is April–May. January– February is the season for strong winds and high seas on this coast. In the earthquake of Nov. and Dec. of 1907, a large part of the country here sank several feet, submerging the island of Warapu and the belt of forest surrounding the lagoon. The strip of coast by Arop also sank, so that the land above high water mark was much lessened in extent, and as the numerous dead trees back of Arop show, there was considerable subsidence here. Many people in Warapu were killed and the rest fled to the bush . . . In Arop the people also fled to the hills (the earthquake shocks lasted for several weeks). The following winter (the Arop people had returned in the spring when the shocks ceased), nearly ²/³ of the narrow strip on which Arop stood was washed away by the waves which, during high tide broached over into the lagoon. The Arop people have built another village in the ‘bush’ to which they have moved most of their possessions, and where they expect to live during the winter. Herr Schulz thinks that the sea will during the coming winter destroy what is left of Arop. (Welsch, 1998: 128, 131) There also were examples of temporary relocation at the time of the 1998 Aitape tsunami, when villages on the periphery of the disaster area, but not directly affected by the tsunami, were evacuated voluntarily. Examples are the Sera villages west of Sissano, where people remained in temporary quarters inland for five weeks after the tsunami and at Lemieng, east of Aitape, where people stayed inland for several months.

40

H. DAVIES

The possibility remains that elsewhere in PNG and the Solomon Islands villages may have been permanently relocated after a major tsunami. For example, the people of Mararamu village in northeastern Umboi Island observe a tradition that homes are not to be built on land that is seaward of a set of stone markers that record the advance of the Ritter Island tsunami in 1878 (Eric Kwa Lokai, University of PNG, pers. comm., 2001). However, another source (Ian Lilley, University of Queensland, written comm., 1999) concluded that the Ritter Island tsunami had little or no long-term impact on settlement patterns, at least on the big island of Umboi. In the larger picture of New Guinea and adjacent islands, we do not know enough about patterns of past coastal settlement to be sure whether or not people have adapted by moving permanently to safe locations. Nor do we know the history and frequency of damaging tsunamis on any sector of coast. This is a potentially fruitful field for future archaeological and geological research.

READING THE RECORD OF PAST TSUNAMIS In order to be able to advise people on the relative safety of occupying any sector of coast such as the Sissano Lagoon sand spit, we need to know the recurrence interval of major tsunamis in that sector. For this purpose written historical records are inadequate because they cover too short a time span. The alternative sources are oral histories and the study of tsunami sediments. The tsunami sediments may take various forms. On a sandy coastline such as Aitape the tsunami brings ashore and distributes a layer of sand derived from the immediate offshore or the beach. The areas worst hit by the 1998 tsunami retain about 10 cm of sand, and this extends about 400 m inland. The sand shows graded bedding (coarser at bottom, finer at top) and rests with sharp contact upon residual soils. Where the sand rests on older sands, as in the eastern Malol villages, it is more difficult to differentiate from residual sediment. On a rocky coastline boulders will be strewn inland. For example, in the eastern Malol villages pebbles 1–2 cm in diameter that probably originated from the mouth of the nearby Yalingi River were strewn across the pre-existing sandy beach ridges. Similarly, on a coastline protected by coral reefs, reef debris will be carried ashore. The 1895 Buna tsunami was said to have carried coral debris 6 km inland (Everingham, 1977: 22). In the 1998 tsunami, lumps of coral up to 1 m across were lifted up on to the beach and debris was carried 30 m inland on Tumleo Island near Aitape (Davies, 1998). For the worst-hit areas near Aitape, trees were felled and stacked as far as 500 m inland. These great piles of debris might survive through time, although in the tropics we can imagine they would soon rot away unless preserved by burial. People were caught up in the debris, so human remains might be preserved. For western New Britain at the time of the Ritter tsunami, the same dramatic scouring of forest extended as far as 1 km inland (Everingham, 1977: 15–20).

TSUNAMIS IN PAPUA NEW GUINEA

41

CONCLUSIONS Based on the written record of tsunamis for the last 150 years, it is likely that major tsunamis have recurred in the New Guinea region as frequently as every 15–70 years. At any one specific point on the coast this may translate into a recurrence interval of hundreds of years. The long interval between tsunamis at any one location on the coast militates against the development of tsunami awareness in the coastal communities because if tsunamis recur at intervals greater than an average lifetime, the communal memory will weaken or be lost. The exception that proves the rule is the Madang-Bogia tsunami, which occurred only 42 years after the Ritter Island tsunami and thus was within the living memory of the older people. In contrast, interviews with survivors at Aitape, where there had been no major tsunami for more than 100 years, confirmed that tsunami awareness was only weakly preserved or absent amongst the general population. We might also infer from these data that lessons from disasters are learned only by the community that experiences the disaster, and are not effectively transmitted to other groups, regardless of whether they share strong trade links. Our admittedly limited evidence suggests that, at least on the Aitape coast, tsunamis have no long-term effect on patterns of coastal settlement and that, given time, people are likely to return to the devastated areas. A factor that favours return to the coast is that communal memory is short. Another factor is the obvious advantages of living on the coast where there is a pleasant setting and lifestyle, access to resources, recreation, trade, easy waste disposal and lower risk of malaria. These attractions are likely to outweigh considerations of tsunami risk, once the memory of the most recent disaster has faded. This work has illustrated the complex nature of communal response to rare, rapid-onset geohazards. While it is dangerous to extrapolate from this study to build a comprehensive model of cultural response to such events, the study none the less points to the need to explore this hazard and its impact in much greater detail. Cultural responses to volcanic hazards are seen in this volume to be complex and dependent on a careful balance between multiple factors, of which the eruption itself is but one. On the basis of this study it is clear that further examination of the phenomenon and its role in cultural development would yield valuable results. ACKNOWLEDGEMENTS I thank R.M. Davies for drafting Fig. 3.1 and R. Torrence and P. White for helpful reviews of the manuscript. REFERENCES Conway, M. (1999) Unpublished letter. Davies, H.L. (1998) Tsunami PNG 1998 – extracts from Earth Talk. Waigani: University of Papua New Guinea.

42

H. DAVIES

Everingham, I.B. (1977) Preliminary Catalogue of Tsunamis for the New Guinea/Solomons Region, 1768–1972. Bureau of Mineral Resources Report 180. Canberra: Bureau of Mineral Resources. Grover, J.C. (1955) Geology, Mineral Resources and Prospects of Mining Development in the British Solomon Islands Protectorate. Interim Geological Survey British Solomon Islands Memoir 1. London: HMSO. Kawata, Y., Benson, B.C., Borrero, J.C., Borrero, J.L., Davies, H.L., deLange, W.P., Imamura, F., Letz, H., Nott, J. and Synolakis, C.E. (1999) Tsunami in Papua New Guinea was as intense as first thought. EOS. Transactions of the American Geophysical Union 80: 101–5. Matsumoto, T., Tappin, D.R. and shipboard scientists (1999) KR98-13 Cruise Report. Yokosuka: Japan Marine Science and Technology Centre ( JAMSTEC). Neuhauss, R. (1911) Deutsch-Neu-Guinea. 3 vols. Berlin: Dietrich Reimer. Tappin, D.R., Matsumoto, T., Watts, P., Satake, K., McMurtry, G.M., Matsuyama, M., Lafoy, Y., Tsuji, Y., Kanamatsu, T., Lus, W., Iwabachi, Y., Yeh, H., Matsumoto, Y., Nakamura, M., Moihoi, M., Hill, P., Crook, K., Anton, L. and Walsh, J.P. (1999) Sediment slump likely caused the 1998 Papua New Guinea tsunami. EOS. Transactions of the American Geophysical Union 80: 329, 334, 340. Welsch, R.L. (1998) An American Anthropologist in Melanesia. A.B. Lewis and the Joseph N. Field South Pacific Expedition 1909–1913. 2 vols. Honolulu: University of Hawaii Press.

4

Bacolor town and Pinatubo volcano, Philippines: coping with recurrent lahar disaster K.S. CRITTENDEN AND K.S. RODOLFO

INTRODUCTION This chapter illustrates the extraordinary resilience of people in the face of catastrophe and the persistence of settlement even in the face of repeated environmental threat. The behaviour described below is a salutary lesson to all archaeological researchers that it is dangerous to assume that there is a clear correlation between threat and cultural response. In reality, as this chapter illustrates, the relationship is very complicated and conditioned by socio-cultural influences that may not be detectable in the archaeological record. Mt Pinatubo’s 1991 eruption, the world’s most powerful in 89 years, left 5 to 7 km3 of loose, hot debris on the flanks of this Philippine volcano. The greatest damage was not wreaked during the eruption itself, but during the monsoon– typhoon seasons of the next five years. Annual intense and prolonged rains have mobilised huge amounts of the debris into lahars (flowing slurries) that have buried widespread areas in debris, centimetres to metres thick. Often boiling hot, the Pinatubo lahars have flowed down five major channels of the volcano, some as fast as 35 km/h. As of this writing (2000), much debris remains on the volcano and can still be mobilised into lahars. Additionally, the lahar deposits may be remobilised into new flows. Bacolor, a municipality on the southeastern apron of Pinatubo volcano, has suffered most from lahars descending along the Pasig–Potrero river. These have buried all but one of its barangays (villages) in deposits up to 9 m thick. The majority of the people in the town proper have abandoned their homes; however, almost 2,000 have refused to leave. Exemplifying the intensity of people’s attachment to place, even in the face of ongoing disaster, they have raised their houses on high stilts and struggled to keep their town alive. After describing the physical setting and history of Bacolor, the eruption and the ensuing years of lahar events, we consider how the townspeople have been affected by and responded to the recurring disaster and the patterns of adaptation that have emerged as they have become more familiar with the threat. In general, natural hazards are governed by the ‘magnitude-frequency’

44

K.S. CRITTENDEN AND K.S. RODOLFO

principle: larger events occur less frequently (Alexander, 1991). Therefore, people at risk from a major event have little opportunity to become familiar with the hazard. The effort inhabitants will expend to rehabilitate a damaged community is a function of three factors. First, a more chronic and familiar threat will be better integrated with the local culture, which will have developed routine mitigation strategies (Anderson, 1967). Second, a community will be discouraged from attempting rehabilitation in place if no infrastructure survives upon which to rebuild. Third, even if everything is destroyed, people will rebuild the community if they perceive the threat to be over. We illustrate these principles by contrasting the Bacolor experience in the town proper with that of other barangays, with other Pinatubo towns, and with communities near Mt Mayon, another Philippine volcano.

PINATUBO VOLCANO Mt Pinatubo and its 1991 eruption have been described in encyclopedic detail (Newhall and Punongbayan, 1996). Situated in west-central Luzon island in the Philippines (15°08.2 N, 12°21.1 E), Pinatubo is the northernmost of a chain of large, andesite–dacite stratovolcanoes (Fig. 4.1). The southern members of the chain, Mts Natib and Mariveles, form the Bataan peninsula of World War II fame. The magmas of these volcanoes are associated with eastward subduction of the South China Sea lithosphere along the Manila Trench. In this tropical locale, carbonate sediments are abundant atop the relatively shallow subducting

Figure 4.1

Mt Pinatubo and its tectonic setting

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

45

lithosphere, and so the magmas of these volcanoes are rich in carbon dioxide and exceptionally effervescent. Pinatubo had not erupted in recorded history before 1991. It experienced its greatest explosive episode shortly before 35,000 BP, which was an order of magnitude larger than the 1991 event (Newhall et al., 1996). Seismologic evidence (Mori et al., 1996) delineated a large magma chamber 5 to 11 km beneath Pinatubo. Periodically, basaltic magma injected into the bottom of the chamber triggers dacitic eruptions that over a brief period generate a caldera and much pyroclastic debris. Then, extruded slowly over centuries or millennia, less-silicic andesitic lavas fill the caldera with a lava dome that is obliterated by the next episode. Radiocarbon dates of wood from the volcanic apron document additional episodes roughly 17,000, 9,000, 6,000–5,000, and 3,900–2,300 years BP. The last eruption before AD 1991 occurred some time between 500 and 800 years BP. Mt Pinatubo is nestled in older mountains to the north, west and south. However, the volcano is unconfined to the east, where pyroclastic flows and lahars have built a gently sloping fan that merges into the swamps at the head of Manila Bay. The peak, which stood 1,745 m above sea level before the 1991 eruption, was replaced with a caldera 260 m lower and 2.5 km in diameter with its deepest point at least 680 m below its rim ( Jones and Newhall, 1996). Lahars The Pinatubo lahars, although sometimes called ‘mudflows’, are actually composed chiefly of sand and coarser materials. The more dilute lahars are called hyperconcentrated streamflows because their sediment contents – 60 to 75 per cent by weight – are greater than those of the most turbid normal streams. Apart from displaying prominent standing waves, dilute lahars look deceptively little different from normal, dirty floodwaters (Rodolfo, 2000). The larger, most powerful lahars are debris flows, in which the water is no more than about 25 per cent by weight, and as low as 10 per cent – barely enough to mobilise the debris. Owing to their high densities, thicknesses and velocities, they exert powerful shear stresses on their channels and especially on their bases. By eroding and incorporating additional rock material as they travel, they ‘bulk up’ or swell to a remarkable degree. Emerging from a mountain canyon or overflowing its channel banks and spreading out over gentler slopes, a debris flow is rarely more than several metres thick, and so any low hill or slope can serve as a refuge. As it spreads, its basal area and frictional resistance increase, slowing it down. Debris flows are more viscous at lower speeds; this causes them to slow down even more quickly. As a consequence, spreading lahars are commonly not powerful enough to topple even fragile structures such as thatched huts. They merely flow around and into them, and bury their bases. The Pasig–Potrero River A pre-eruption map (Fig. 4.1) shows the head of the Pasig river to be 6 km from the summit of Mt Pinatubo, 800 m above sea level. It flows eastward down a steep

N

➞

Figure 4.2

The Bacolor area, and the extent of lahar activity from 1991 to 1995

N

➞

48

K.S. CRITTENDEN AND K.S. RODOLFO

gorge for 9 km, to where its valley opens out on to the volcanic apron. There, where the farmers call it the Potrero river, it becomes braided, is only shallowly incised, and turns southeastward to flow another 15 km over the one degree slopes of the apron. At an elevation of 30 m above sea level, it discharges into the maze of estuaries of the Pampanga delta. The delta surface slopes only 0.0001 to 0.0002 m/m along the 21 km distance from Bacolor to the head of Manila Bay and suffers from severe floods during the rainy season, typically from June through October. In the 1960s, dikes were built to contain 13 km of the river and protect the towns and villages of Santa Rita, Bacolor and Guagua from flooding. The swampy region has been greatly modified into rice paddies and fishponds and has been a major ‘rice, sugar and fish basket’ for centuries.

BACOLOR Bacolor municipality, located on the east bank of the Pasig–Potrero river in Pampanga province, straddles the transition from volcanic apron to delta. The names of places and people in Pampanga strongly reflect their riparious environment. Pangpang means ‘river bank’; the people, as well as the area, are called Capampangan, ‘of the river banks’, and the province name is a Spanish corruption of the same word. ‘Bacolor’ is derived from baculud, meaning ‘high ground’ – ‘high’ referring to only 1–2 m of local relief above the surrounding swampy flats or lubao (after which a neighbouring town is named). The town proper, 39 km from the volcano summit, stood about 10 m above mean sea level before the eruption. The layouts of Bacolor and neighbouring municipalities reflect their physiographic setting. To the north, the rivers and creeks that drain the volcanic apron southwestward are paralleled by villages and roads, which to the south are less oriented, following the meandering delta streams. The Philippines is a developing country with few roads and an exploding population; barangays continue to elongate and coalesce along each road. Pre-eruption Bacolor The early history of Bacolor is not well known. Capampangans are believed to have arrived more than two centuries before the Spaniards conquered Pampanga in 1571 (Larkin, 1993). They came from Borneo or Sumatra (Henson, 1955) in large, outriggered, wind- and oar-propelled boats called barangay (Scott, 1994), a name still used to mean ‘village’, abbreviated in formal names as ‘Bgy’. Early Spanish chroniclers speculated that each barangay comprised the descendants of a particular boatload of migrants. It is important to note, in the context of the recent Pinatubo experience, that ‘barangay’ originally referred to a group of people rather than a locality (Scott, 1994). Until 1991, the Capampangans had no experience or oral tradition of volcanoes or lahars. In addition to exacerbating the eruption disaster, this fact may bear on the date of the penultimate Pinatubo eruption. Newhall et al. (1996)

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

49

reported that sparse radiocarbon data indicate that this episode occurred any time from 800 to 500 BP, most likely toward the end of this range. However, this event was larger than the 1991 eruption, and had it happened only 500 years ago, it would be surprising if the first Spaniard chroniclers did not hear about it when they arrived a few generations later. Inhabitants of the Papua New Guinea highlands have a strong oral tradition about the Long Island eruption c.1660 (Blong, 1982), and those of Mt Parker in Mindanao have a sketchy oral history of its 1641 eruption (C. Newhall, written comm., 2000). Prehispanic Filipinos did not live in cities or towns. Beyond the family, the only community was the autonomous barangay, typically of some 30 to 100 families. After the conquest the Spaniards worked hard to consolidate them into larger towns to more easily administer and Christianise the Filipinos. Apparently, only the colourful pomp, spectacle and mystery of the Mass and the fiesta finally induced the reluctant Filipinos to comply (Scott, 1994). In the subsequent centuries, however, Filipinos have developed strong attachments to their towns. Bacolor was the capital of Pampanga under the Spanish and Americans from 1746 to 1904, and in 1762–64 was the seat of the Spanish government driven out of Manila by the British (Henson, 1955). The poblacion or town core was laid out in what has been termed the Philippine plaza complex (Hart, 1955; Larkin, 1993), centred on the plaza or town square dominated by San Guillermo parish church. Across the plaza were the municipio or town hall, a two-storied rectangular, stone edifice, the police station and the courthouse. Near the plaza was the concrete, steel-roofed palengke or municipal market. The nearby business section included two banks, a tobacco factory, specialty stores and shops of local craftsmen. Facing the palengke was the Don Honorio Ventura college of arts and trades, founded by the town élite in 1861. Its several two-storey buildings housed the local public high school and a variety of educational programmes that drew over 5,000 students from the province. Nearby Bacolor Central elementary school served the poblacion. The plaza was surrounded by the splendid ancestral homes of élite families. Larkin (1993: 10–11) has described the Pampanga houses in the early twentieth century. As raised off the ground by a stone foundation; upper-class houses had wooden walls and carved balustrades, with large, sliding window panels made of translucent capiz (scallop) shells mounted in wood. The high, peaked roofs were of nipa palm leaves or cogon grass thatch, and the floors were solid wood. Many such fine homes in Bacolor were so well built that they lasted over 100 years. As late as 1972, the poblacion still featured ‘a large cluster of elegant nineteenth century homes, far more than in any other municipality’ in the province (Larkin, 1993: 97). Lower-class houses, made of cogon, nipa and bamboo, traditionally had one or two sparsely furnished rooms on a bamboo-pole frame, with the floor several feet above the ground. In 1991 only a few of the fine, Spanish-era ancestral homes still remained in the town proper and galvanised iron roofs were much more typical than thatch. Concrete had replaced increasingly scarce wood in construction. However, many of the older homes still sported wooden walls, capiz windows, and hardwood

50

K.S. CRITTENDEN AND K.S. RODOLFO

floors. Decorative, wrought iron railings and ceramic posts were used on balconies, verandas and windows. Some of the smaller homes, new and old, sported roofs of thatch and walls of sawali, a thin material woven out of bamboo strips. Bacolor has a total area of 4,570 ha and, according to the 1990 census (National Statistical Office, 1990), had a total population of 67,259. The town proper, with 16,143 residents, comprised the barangays of Santa Ines, San Vicente and Cabetican, in addition to the poblacion barangay Cabambangan. There were 17 outlying barangays, each with several hundred to several thousand people, a few sari-sari or small variety stores and, commonly, a small chapel.

ERUPTION AND LAHAR HISTORY, 1991–95 Intermittent explosions began in early April 1991 and built up to a powerful climactic episode that lasted 9 hours on 15 June. Owing to the east winds of a typhoon that fortuitously passed north of Pinatubo during the paroxysm, Bacolor received only 3 cm of tephra fall and was spared the collapsing roofs that killed about 300 people in the downwind sectors of the volcano. Lesser explosions occurred until early September. Pyroclastic flows – dense, ground-hugging flows of debris and hot gases – deposited 5–7 × 109 m3 of pumiceous debris on all sides of the volcano. Of this, 1.6 × 109 m3 was deposited on the east flank in a field covering about 34 km2, deeply filling and extending 16 and 5 km down the Sacobia and Pasig river valleys, respectively (Pierson et al., 1996). This accumulation was to provide the debris for the lahars of these and the Abacan river. Pasig–Potrero lahars Most Pinatubo lahars, with typical volumes of 5 × 105 to 5 × 106 m3, are triggered by intense, prolonged rain. However, some of the largest, most devastating flows have been generated when lahars or pyroclastic flows have dammed a tributary to a major lahar channel, accumulating a temporary lake. Eventual breaching of the blockage has resulted in flash floods that have eroded and incorporated large amounts of sediment, bulking up into lahars downstream (Umbal and Rodolfo, 1996). Breakouts of an ephemeral lake at the south bank of the Pasig river in the foothills about 26 km upstream of Bacolor generated major lahars in 1991, 1992 and 1994. We constructed the following lahar chronology from published and unpublished reports of the Philippine Institute of Volcanology and Seismology, our own field notes, and newspaper accounts. We interviewed local officials and present and former inhabitants to gather detailed information about events that affected the town proper. Our results are summarised in Table 4.1 and Fig. 4.2. The typhoon passing during the 15 June 1991 paroxysm generated lahars that flowed down every major drainage channel around Pinatubo. Those along the Pasig–Potrero river were not very voluminous and were largely confined by the flood-control dikes, which suffered only a minor breach about 0.5 km upstream

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

51

Table 4.1 Summary of lahar activity on the Pasig–Potrero river and its effect on Bacolor from 1991 to 1995 (unpublished data from the Philippine Institute of Volcanology and Seismology) Year

Volume Area Affected Bacolor communities deposited, covered, × 107 m3 km2

1991

5

38

1992

4

14

1993 5.5 1994 14

18 53

1995

69

7.5

Town proper, Parulog, Potrero, San Antonio, Santa Barbara Mitla, Parulog, Potrero, Santa Barbara Balas All bgys except Cabalantian, Calibutbut and San Isidro All bgys except Calibutbut

Comments

Average thickness 1.3 m. Largest event a lake breakout lahar on 7 September Mild activity. Largest event a lake breakout on 29 August Only mild activity Lake breakouts on 8 August and 22–23 September 11 events; worst on 1 October; 5 × 107 m3 deposited over 25 km2, up to 9 m thick

from the town proper. Monsoon winds and typhoons generated frequent lahars from late June to late October. On 7 September, a lake-breakout lahar buried much of Bacolor town proper, including the national highway, in debris 1–3 m thick (Arboleda and Martinez, 1996). The outer walls of San Guillermo church were buried by less than a metre, but the lahar did not flow into it. After the rains, the government proposed to evacuate Bacolor residents forcibly, and dike most of the area into a catch basin to protect nearby towns, especially San Fernando, the provincial capital and commercial centre. In fact, Bacolor was never evacuated; instead the first in a series of ineffective, so-called ‘protective’ dikes was built. Also, anticipating additional lahars, the government raised the national highway 3 m to re-establish the vital link between Manila and northwestern Luzon, hoping at the same time to protect Bacolor town proper. In February 1992, a major explosion in the eastern pyroclastic fan greatly modified the terrain, rerouting the lahars from the Abacan river to the Sacobia river, which experienced the greatest lahar activity that year. Along the Pasig– Potrero river, localised monsoonal showers from April through June triggered only minor lahars. After a lake breakout in August, flows buried Bgy Mitla to depths of between 1 and 3 m. No flows were large enough to overtop the newly raised national highway, and so the central town was spared. Before the 1993 rainy season, flimsy new dikes intended to funnel lahars into the lower Pasig–Potrero channel were constructed by bulldozing lahar debris. On 19 August, lahars breached the dikes and hit Bgy Balas. After other flows in September and October, 100 houses in the barangay were buried in lahar deposits 2 m thick. Early in October, a large steam explosion and secondary pyroclastic flow drastically modified the topography of the eastern pyroclastic fan. The watershed of the Sacobia river was captured by the Pasig–Potrero river, which more than

52

K.S. CRITTENDEN AND K.S. RODOLFO

doubled its catchment area, from 24 to 55 km2 (Umbal, 1997). This occurred too late in the rainy season to affect the year’s lahar activity, but had serious consequences for the following years. From July through October 1994, passing typhoons generated lahars 14 times (Arboleda et al., 1995). In July, after breaching the new eastern dike, lahars deposited debris 4 m thick at Bgy Potrero and 1 m thick at Bgys Duat and San Antonio. On 8 August, lake-breakout lahars buried Bgys Maliwalu and Dolores in deposits up to 2 m thick. Similar flows deposited an additional metre in Maliwalu in late August. On 22–23 September, two secondary explosions and a catastrophic lake breakout generated large lahars that killed 23 people and displaced 50,000 people. Along with several outlying barangays, Bacolor town proper was severely hit for the first time since the eruption and over 1,000 houses were buried to depths of 1–4 m. San Guillermo church was invaded and its floor buried under 2 m of debris. The lake-breakout lahars eroded most of their sediment from the depositional field of the lahars brought down since 1991, leaving a wide, deep channel that was to serve as the main conduit for the disastrous lahars of 1995. The government officially designated Bacolor as a high-risk area to be abandoned and denied public funds for construction. Yet, faced with an enormous backlog of refugees from previous lahar seasons, no government agency would accept responsibility for mass permanent resettlement. Schools, churches and public buildings all over the province were swamped with refugees. A debate about whether to resettle victims or to build bigger ‘protective’ dikes was never resolved. Efforts were made to do both, but most of the funds went to constructing a third, U-shaped set of unarmoured dikes, consisting of compacted lahar deposits, to transform about 80 km2 of the Pasig–Potrero watershed into a debris basin. The new dikes proved ineffective and exacerbated the loss of life by imparting an unjustified sense of security in the populace. In 1995, 11 lahars reached all Bacolor barangays but one (Umbal et al., 1996). On 30 July, lahars up to 4 m deep breached a western portion of the dike. Even after the rains stopped on 1 August, lahars continued to spill through the breach into previously unaffected areas. Typhoon-induced lahars also affected western Bacolor from mid-August to early September, killing seven people and burying 900 houses. Tens of thousands of Bacoloreños fled to evacuation centres. Helicopters rescued hundreds trapped on rooftops. The poblacion was badly hit, and San Guillermo church was buried under another 4 m of deposits. Pinatubo’s worst lahar calamity occurred on 1 October. Intense, prolonged typhoon rains triggered lahars that buried a 25 km2 area in 5 × 107 m3 of debris (Umbal et al., 1996). Bacoloreños downstream were relieved at dawn to hear radioed reports from the police watchpoint that the flows were only water. The floodwaters, however, bulked up into debris flows downstream from the watchpoint, swelling even more by eroding and incorporating a 1 km stretch of the eastern dike. Overtopping and cascading down the newly repaired national highway, the lahars took only about six hours to obliterate Bgy Cabalantian,

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

53

which had totally escaped lahars since the eruption. In all, 2,300 houses in Cabalantian and another 3,900 in four nearby villages were buried under deposits up to 9 m deep. No habitable houses remained in Cabalantian. Tens of thousands of new refugees joined the backlog of families awaiting resettlement. Lahars swept hundreds from rooftops. Remobilised lahars are cool, and so many survived. Even non-swimmers who did not panic floated easily in the dense lahar. Some were carried several kilometres, including eight children rescued from rafts of water hyacinths by fishermen 3 km away. As many as 10,000 people spent up to 36 hours on rooftops, exposed to the elements, most without food or water. Most fatalities were from drowning or suffocation. Other people died from electrocution and heart attacks, and six succumbed to snake bite. Cobra breeding for antivenin had been a Cabalantian cottage industry. Many of the cobras escaped and, along with indigenous poisonous species, sought refuge on roofs and trees occupied by people. The total number of Cabalantian fatalities remains controversial. On 3 October, the Bacolor mayor confirmed 70 deaths. He increased that number to ‘more than 80’ on 6 October, while the national government was admitting to only 17 fatalities. The next day the mayor reported over 100 dead and 252 missing, while a regional disaster council spokesperson said that as many as 2,000 had died. Four days after the disaster an estimated 1,000 people were still reported missing and search teams had yet failed to locate 83 listed individuals. By 13 October, 35 on the list had been located with relatives elsewhere. No data are available stating how many of the remainder were dead. Entire families or even clans of poor farming folk living in close proximity in low dwellings may have perished, leaving no one to report them missing.

LAHAR EXPERIENCES IN BACOLOR By 1996 all the residents of most of the outlying barangays had been displaced and their villages erased from the map. However, the town proper was never totally destroyed or abandoned. Evacuation orders were universal in their coverage, but unenforceable. Families often evacuated only the women and children during bad weather, with the menfolk staying to guard the household. Any time the rain subsided, evacuees returned in a constant stream to reclaim their belongings or repair damage and reoccupy their homes. Families extraordinarily attached to the place never left, even during the worst calamities. The demand for resettlement housing far outstripped the government’s ability to provide it. Only homeowners were eligible for resettlement, and so many refugees could not graduate from the ‘temporary’ staging centres. Many eligible families had to languish for years in staging facilities, subject to overcrowding, poor sanitation, disease, sporadic supply of food and necessities, and lack of livelihood. Once they finally received resettlement houses, many families found the permanent communities to be lacking in amenities, most notably space, privacy and gainful employment. Some ‘resettled’ families sold their houses in the

54

K.S. CRITTENDEN AND K.S. RODOLFO

resettlement communities to gain the wherewithal to re-establish residence in Bacolor; others rented them out and returned home. None the less, as of 1996, no residents remained in 15 barangays. Only Bgy Calibutbut to the far north remained unscathed by lahars. Almost all the residents in one small farming barangay, San Isidro, had returned and resumed farming. The population of the town proper had shrunk to approximately one-tenth its pre-eruption size. An unofficial 1996 census of Bacolor town proper (Lacsamana, 1996) reported 381 families consisting of 1,488 individuals. A similar census in 1997 (Lacsamana and Crittenden, 1997) enumerated 384 families, or 1,755 individuals in the four barangays, struggling to maintain their freehold and rebuild their town. Ironically, those who stayed were the pioneer resettlers of the town, the magnets and models that attracted other families to rejoin them. We now focus on their experience. Burial history As part of the 1997 census, household heads answered questions about the laharburial history of their homes and their families’ reactions since 1991. Neighbours agreed well about how much their homes had been buried and their reports were consistent with the less-complete accounts of government geologists. Based on medians of 353 reports for each year, houses were buried by 2.0 m of lahar deposits in 1991, 1.0 m in 1994, and 3.5 m in 1995. Overall, informants reported that their homes had been buried by 6.5 to 8.0 m over the six-year period with total deposition averaging 6.5 m. House-raising To reclaim and protect their houses, families returned to the traditional practice of raising their houses on stilts. For several years, the practice was limited mostly to small houses of lightweight materials such as wood or sawali. New materials or those recycled from the previous house could be used to build a stilted house. However, many existing houses were raised as well. Family members dug with shovels to a level below the floor of the highest storey, which they sawed from the remainder that would remain buried. We observed one family raising their house with a car jack, elevating each corner in turn a few centimetres at a time until the house was high enough to install on concrete stilts. Over time, the town developed the wherewithal to exhume and raise larger houses. After the 1995 flows, Capampangans living in Chicago in the United States donated six large hydraulic jacks in response to appeals from Bacolor. The municipal government instituted a programme for elevating structurally suitable houses, by designating a local contractor and contributing the use of the jacks, cement and forms for moulding the stilts and concrete beams. Homeowners paid only the labour costs. Once stilted, the houses could be raised again if needed. In 1996 and 1997, 60 houses were lifted, drastically altering the visual landscape (Fig. 4.3 a, b). Additional families borrowed the jacks and forms, and over 250 houses eventually were raised. According to our 1997 census, 15 per cent of families had raised their houses on average each year in the period 1991–96. Almost half

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

55

(a) (b)

(c) Figure 4.3 A house in stages of raising and reconstruction: (a) raised once, 1996; (b) raised again, 1996; (c) lower level enclosed, 1997 Photographs:

the authors

(42 per cent) raised their houses on stilts at least once and some families as many as four times. In 1996, a typical stilted house in the town proper had a tentative, bedraggled look. It stood on concrete stilts at least 4 m tall. Constructed mostly of wood, the bases of its walls cut irregularly and stained where they had been buried. Many houses had running water and a functioning ‘comfort room’ connected by a plastic pipe to a buried septic tank. A ladder or wooden stairway provided entry. Recycled materials were used to repair damage. The lower level, either open or partially enclosed with concrete blocks, recycled building materials and tarpaulins, was left unfinished inside and out. Stone and concrete houses could not be raised, but some families built rooms or new stories on the exposed tops of their houses. The 1997 census indicates that about 8 per cent of the families had built on top of their houses from one to four times. House-raising was not limited to families in the 1997 census. In a 1996 sample survey 82 per cent of the families who had remained had raised their house at least once, and 30 per cent had done so more than once (Lamug et al., 1999). However, 60 per cent of those living in permanent resettlements had also raised their Bacolor houses, including 10 per cent who had done so twice or thrice. In addition, half of the families in temporary refugee centres at that time reported having raised their houses at least once. Some of these ‘resettled’ families had valiantly resisted abandoning their homes, only to be defeated eventually by the

56

K.S. CRITTENDEN AND K.S. RODOLFO

danger and hardship or exhaustion of their resources. Others had moved more willingly, but raised their houses to protect their investments in them or to save them for their eventual return. Only among families who had resettled on their own was it uncommon (7 per cent) to have raised their Bacolor house. The practice of digging up houses and raising them on stilts developed over several years of lahar experience. This adaptation is not generally practised outside the town proper, although it draws from the traditional stilting of houses to cope with the annual monsoon floods. Cabalantian was destroyed late in the aftermath by a single calamity, leaving no infrastructure to rehabilitate, and so residents offered little resistance to being resettled en masse. All other eventually depopulated barangays were upstream communities that were hit by lahars early and often. Not only was the infrastructure destroyed, but also the threat continued. Bacolor town proper was downstream, partially protected by the national highway, and had many taller, more substantial buildings. Total deposition was quite significant, but occurred in several episodes, each much smaller than the Cabalantian event. Each small episode, by raising the surface, helped to reduce the likelihood and thickness of subsequent flows, and it taught lessons about living with lahars. As they became more familiar with the threat, some residents also became more determined to stay. Thus the regularity of the lahars and their familiarity with them, combined with the fact that no single event destroyed everything, encouraged them to develop means to stay and rebuild even though the lahar threat was not over. Grounded in an appreciation of the recurrent threat and employing superior technology, the house-raising efforts after 1995 were well designed for sustainability in the face of future lahars. In 1997–98, to make their stilted houses more comfortable, many families in the town proper enclosed all or part of the lower level with concrete blocks (Fig. 4.3 c). Some raised the floor against flooding with truckloads of lahar deposits tamped down by carabao (water buffalo). Functions such as meals were moved to the lower level, whether enclosed or not. There was no effort to finish the exterior, but some families added a grandiose, curving staircase to the upstairs entrance. Yards were planted with ornamentals and fruit trees, and potted plants were abundant. A few brave souls constructed small, single-storey houses at ground level, expressing the opinion that the lahars were over. Home construction noise was omnipresent and continual in 1999. Families returning from resettlement communities were building new homes, constructing them on top of the buried originals with the lower floor 2–3 m above the ground, often on a site that had first been raised 1 m or more with lahar deposits. Some took care to connect the new house to the foundation of the previous house, an important precaution, because lahar deposits are porous and very susceptible to liquefaction during earthquakes. Families continued to enclose the lower levels of their stilted houses. For the first time homeowners were finishing the exteriors by smoothing with mortar, painting, and adding trim. Buried, unoccupied houses were disappearing from view. Resourceful residents stripped exposed recyclable materials to use in their own makeshift structures. Nature was taking its course, as the often-damp ground claimed

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

57

biodegradable housing materials and furnishings, and tall cogon grass concealed the unburied tops of concrete walls. Rehabilitating churches The Catholic Church is interwoven into almost every aspect of community life in the overwhelmingly Catholic lowland Philippines. Bacoloreños have invested heavily, both sentimentally and financially, in their churches. This is particularly true of the San Guillermo parish church. Built before 1645, this heavy, Spanish Baroque stone building had heavy buttresses and low vaulted interiors. It is 56 m long and 15 m wide, with a central nave, large transept, and gilded retablos and pulpit (Galende, 1996). At the time of the eruption its main doorway and the windows of the church and bell tower were tall arches; the church was 12 m high and the cross atop its tower reached a height of 30.6 m. Ionic columns and higharched mouldings enhanced the feeling of interior height. After the flows began entering the church in 1994, townspeople dug up and stored the altar, pulpit, retablos and statuary. By the end of 1995, cumulative deposition was 6.6 m. Easily visible from the national highway and much photographed by the press, San Guillermo church became the most recognisable visual symbol of lahar devastation in the Philippines. The new entry to the church was through the choir-loft window above the original main doorway. With the floor of lahar deposits exposing only the top few centimetres of the side windows, and the huge chandeliers nearly reaching the floor, the church interior resembled a dark cave. In November 1995 the parish expanded the poblacion’s traditional religious festival to include the entire municipality and its scattered populace, and Bacoloreños came from everywhere to participate in the Mass, procession and festivities. This celebration helped to unify the traumatised town residents and to symbolise the town as a living entity. By the time of the 1996 festival, the altar had been placed on the raised lahardeposit floor and Masses were being celebrated again. Using contributions from scattered Bacoloreños and tourists, the parish began to remodel and restore the church in 1997. A partial concrete floor was laid atop of the lahar deposits; the curved ceiling was removed to restore height to the interior, and to enlighten the room large dormer windows were installed in each side of the roof, which was lined with silvery, reflective insulation. In 1998, an art gallery and museum adjacent to the sanctuary was opened. Restoration of the church continues, as resources become available. Each of the other barangays in the town proper had its own chapel. The smaller ones in San Vicente and Santa Ines were abandoned after their burial in 1995. The larger chapel in Cabetican, an Archdiocesan shrine, had two buildings. Its large, modern asymmetrical concrete shrine, built in 1980, remains buried. The older, more traditional structure with a bell tower and cupola was buried in increments to 5 m by 1995 (Fig. 4.4 a). The townspeople reclaimed it in 1996, putting in a new main entrance above the buried one and restoring and furnishing the interior above the lahar-deposit floor (Fig. 4.4 b). Then, in 1997, they raised the roof and cupola another 5 m (Fig. 4.4 c) to accommodate a remodelled sanctuary with a concrete

58

K.S. CRITTENDEN AND K.S. RODOLFO

(a)

(b)

(c) Figure 4.4 The Cabetican Archdiocesan shrine: (a) after 1992 burial; (b) remodelled for use, 1996; (c) raising the roof by 5 m, 1997 Note:

The banner over the entrance reads ‘No guts, no glory’

Photographs:

(a) C.T. Remotigue; (b) and (c) the authors

floor several metres above the surrounding land and side walls of wrought iron grillwork. Schools After 1995, the national government would provide teachers but refused to restore any school buildings because Bacolor had been declared unsafe. In a self-help effort, the town and the barangay collaborated to rebuild the Cabetican elementary school with the help of donations and an informal barangay tax on every truckload of lahar sand and gravel quarried from its lands. The renovated school, with a new storey added atop one of its buried buildings, has about 240 students in kindergarten and grades 1 through 6. At the college of arts and trades, lahars buried all roads and penetrated the buildings in 1994 and 1995. Both times, students, faculty and staff cleared the debris, and the school reopened one month later. In September 1995 lahars completely buried all the first floors and all single-storey shops and labs to their rooftops and a major typhoon the following month added another metre of deposit. The president, faculty and staff decided not to give up on the college and

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

59

town. They sought emergency funding to resume classes as soon as possible. After holding Saturday classes for two months at Bulaon resettlement, they reopened in Bacolor in December with just over half as many students as before. In 1996, the college’s board directed all capital funds to development of small ‘relocation’ campuses in two nearby towns. With capital expenditures prohibited by its board and the national education agency, the conditions on the Bacolor campus are deplorable. Flooding during the rainy seasons makes the school difficult to access and breeds mosquitoes. None the less, with over 4,000 students in Bacolor – over twice the population of the town proper – and 236 faculty and staff, the school is the major primary employer and a major secondary source of livelihood. Water and electric services The electric cooperative has restored electricity to the remaining residents, and typhoons and lahars interrupt service only briefly. Lahars wiped out the water system and destroyed all the pipes in 1995. The national water administration abandoned Bacolor because it was in a high-risk area, but in early 1996 the local water board restored service for the remaining people. By 1998, they were providing potable water to 250 households. A municipal programme in 1996 helped people drill shallow wells for non-drinking water; some households rely on these for all their water needs. Streets and drainage Streams have not yet cut an integrated drainage network through the irregular surface of the 1994 and 1995 lahar deposits and so every rainstorm incapacitates vehicle traffic. However, the ground is high enough now so that floods recede much sooner than from neighbouring towns, and Bacolor would have no flooding problem if it had an organised drainage system. Recognising how lahar accumulation has made Bacolor less flood-prone, residents said in 1996, ‘Just give us two more metres, and our flood problem will be solved.’ In the absence of infrastructure funding, the old highway that served the local traffic between Bacolor and adjacent towns has been partially closed since it was buried for a second time in 1994. The college, the Cabetican church and many residences are located on this major thoroughfare. Government engineers attempted to restore the highway in town but, instead of raising it, they scraped it down, unintentionally making it a new tributary of the Pasig–Potrero river. The government has promised to raise and rehabilitate the highway and restore the drainage system in the town proper. Were this done, it would greatly enhance the development and resettlement of the town. Many former residents say that they will return as soon as this highway is reopened. Loss of commerce, revenue and livelihood Bacolor was more of an agricultural and genteel dormitory community than a major commercial centre. Its commerce, shut down by the 1994 and 1995 lahars, never returned, except for small home-based enterprises.

60

K.S. CRITTENDEN AND K.S. RODOLFO

The national government returns internal revenue to a municipality as a function of its population and land area. Bacolor’s already-decimated population was seriously undercounted in a national census conducted during the 1995 lahars (National Statistical Office, 1995). The town has lost half of its returned tax monies, and about 70 per cent in local revenues. Nevertheless, the mayor tries to address the needs of scattered former residents in staging and resettlement areas as well as the minority now living in the town. Despite the shortage of commerce and livelihood in Bacolor, the lahar crisis has created employment opportunities for town residents in home construction, dike construction, and quarrying of sandy lahar deposits. Essential to the sustainability of life in Bacolor is its proximity to the neighbouring towns of San Fernando and Guagua, which offer shopping, commercial services and employment opportunities. By contrast, some of the resettlement communities are in remote sites, far from livelihood and commerce. Attachment to place Universally, people’s identities and senses of security are rooted in places, particularly neighbourhoods. These common connections to places underlie the strong sense of affiliation and common identity shared by social groups such as families, neighbours or townspeople (Oliver-Smith, 1982; Orum, 1998). This tendency is especially strong in poor, developing countries, where the wealth of most families is the house, land and animals, and mobility is limited. The Philippine government was unwilling to pay residents for leaving their land and homes and did not provide viable resettlement for all. Undoubtedly, this contributed to people’s reluctance to leave and their extraordinary efforts to save their homes. A few current residents are trapped in Bacolor because in their view they have nowhere to go. Many more have expressed strong emotional attachment to the town and have expended much effort and resources to stay despite the lack of amenities. Others residing elsewhere but similarly attached are watching and waiting for the time when they can return. Many of these have stored their houses on stilts against their eventual return. Local institutions – the Catholic parish, the college and the water board – have invested in preserving the town. Attachment to the point of delusion in the face of danger was evident in 1995. Early in the lahar season, the inhabitants of Cabalantian were fully aware that the protective dikes to the west were failing and barangays were being devastated, yet they clung to the belief that the same dike system would protect them. The disaster literature since Erickson (1976) has recognised that the consequences of post-disaster resettlement may be even more grievous than the disaster itself and that these are ameliorated when neighbours are relocated together, thereby preserving as much of the social fabric as possible (Oliver-Smith, 1991). Thus, the government has made every effort to relocate people in barangay groups. Depopulated barangays have been cloned into resettlement neighbourhoods with the original name in more than one resettlement community, and surviving barangays in the town proper have twins in resettlements.

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

61

To the migratory Pre-Hispanic Filipino, a barangay was less a place, and more a group of people; only under the Spaniards did it become an areal sub-unit of a permanent town. The Pinatubo diaspora has brought these conceptions into conflict. Bacoloreños in resettlements cling to and benefit from both, whereas those who stayed subscribe to the legal, Hispanic model. Many who resettled return every day to spend time with their friends and relatives, but those who stayed complain about their civic disloyalty. Far outnumbered, they resent the resettlers who, ignoring residence requirements, vote in the town and have elected their own to all the municipal offices except mayor. Towns that house resettlement communities are reluctant to provide services to ‘resettled’ residents who insist on voting in Bacolor and who view their ‘permanent’ resettlement as temporary. Citizenry and leaders alike have devoted meagre resources to visual symbols to cultivate an image of Bacolor as a town that is alive, where people still live. Visible from the national highway, the rehabilitated San Guillermo parish church and the stilted houses have tremendous symbolic significance. Buried statues of town heroes, even at the still-buried municipal market, have been dug up and elevated on to new pedestals. There is an informal competition among homeowners to give even their humble, unfinished houses a proper, imposing entrance. And the town often uses holidays and political events as occasions to float banners and signs with brave slogans, such as ‘The Town Rises’, along the highway. On a recycled corrugated-steel sheet used on one raised house is emblazoned: ‘Bless the lord for giving us new land, a new life, and a new paradise’. Attachment to Bacolor is strong both for those who stayed and those who resettled. The fact that neighbours have been relocated together has encouraged their feeling of connection to the town as a social community. Physical reconstruction of the town proper preserves the locale for this attachment. Thus, at the same time that it puts these definitions into conflict, the current situation reinforces the identity of the town as both people and place.

OTHER AFFECTED COMMUNITIES In 1992, two barangays of Mabalacat municipality were badly damaged by lahars from the Sacobia river in an event that was to be exceeded only by the Cabalantian disaster of 1995. Although damage was nearly total, the threat of recurrence was eliminated when steam explosions reorganised the eastern pyroclastic fan in late 1993. With the consent of government scientists, these communities have since been rebuilt atop their original locations. The west side of Pinatubo has experienced much larger lahars than those of the Pasig–Potrero river, but is much less populated. Poonbato, a large resort barangay on the Bucao river, was totally buried in 1991–92 and has not been rebuilt. Bucao lahars remain a serious hazard and so the inhabitants were relocated to two resettlements further downstream. Barangay San Rafael on the Santo Tomas river was also buried without trace in 1993. Its people re-established themselves only 2 km away, on a lowland site still vulnerable to lahars. Clinging to their

62

K.S. CRITTENDEN AND K.S. RODOLFO

pre-eruption mode of life, they refused to build on nearby high ground, instead placing their trust in dangerously flimsy dikes. So far, relatively dry years and weak lahars have spared them. Mayon volcano, 300 km from Pinatubo at the south end of Luzon, erupts much more frequently: at least 48 times since 1616 (Rodolfo, 1989). Its eruptions, however, are less energetic by orders of magnitude. Rain lahars, although much less frequent and voluminous than those at Pinatubo, are familiar events and so inhabitants of the Mayon apron are used to eruptions and lahars and live in harmony with them. Typical of volcanoes in the humid tropics, Mayon soils are fertile and heavily farmed. People expect an eruption lasting only a few days or weeks to affect a limited sector of the volcano every decade or so. Those in harm’s way evacuate, then return as soon as possible to resume farming. After typhoontriggered lahars, farmers immediately plant rice in the fresh deposits, skirting the large boulders left by debris flows. The ground floors of larger houses may be buried and abandoned after the larger events. Smaller houses are built on the ground, usually of hollow blocks made with lahar sand, and people are used to building them up again if partially buried, as at Pinatubo. Their behaviour, now considered routine, may be similar to that of the farmers in Bgy San Isidro in Bacolor, whose land is their primary asset. Without any communication with their peers at Mayon, these San Isidro farmers have arrived at similar pragmatic solutions that could become routine should smaller lahars again encroach.

LOOKING TO THE FUTURE Since 1996 few typhoons have approached the Philippines and monsoonal rains have been markedly reduced. As a result there have been no serious Pinatubo lahars. This is a temporary condition. Much pyroclastic debris awaits mobilisation or remobilisation during the sustained, intense rain of more typical rainy seasons, and lahars will recur whenever conditions are right, possibly for another decade or more. In the meantime the area may become valuable for housing tracts, as the Philippine population inexorably doubles in every generation. Ironically, every time an area experiences lahars, the deposits raise it and make it less prone to subsequent lahars and to floods as well. Conversely, unaffected areas become increasingly more vulnerable as successive flows seek low ground. Areas entirely spared from lahars inevitably become more flood-prone, an increasingly severe problem in other Pampanga towns. Global sea level has risen about 2 mm/y during the twentieth century and could rise at least 100 cm over the next century. However, local sea-level rise in Pampanga could be one to two orders of magnitude greater, owing to deltaic subsidence, volcano–tectonic movements and groundwater withdrawal. Virtually all water for domestic and industrial use is extracted from wells, causing 3 to 4 cm of yearly subsidence and increasing marine encroachment at some Pampanga sites (Siringan and Rodolfo, 1999). Subsidence associated with pumping water can only accelerate as the population grows.

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

63

Because of the lahars Bacolor has experienced, the repeated raising of the national highway, and the dikes intended to protect the adjacent towns, the town proper now stands at least six metres higher than either San Fernando to the east or Guagua to the west. The town has already solved its historical problem with the longer-term flooding that increasingly plagues its neighbours. In anticipation of future flows virtually all the extant buildings in the town proper have been raised well above the current ground level. For these reasons and despite government decisions and decrees, Bacolor is now much less vulnerable to lahar threat than any of the towns around it. People and commerce from the lowerlying neighbouring towns may migrate to the raised Bacolor area. The town may eventually have a bright future.

ACKNOWLEDGEMENTS We thank the Center for Integrative and Development Studies and the National Institute of Geological Sciences of the University of the Philippines, and the College of Liberal Arts and Sciences of the University of Illinois at Chicago for supporting this research.

REFERENCES Alexander, D. (1991) Natural disasters: a framework for research and teaching. Disasters 15: 209–26. Anderson, J.W. (1967) Cultural adaptation to threatened disaster. Human Organization 27: 298–307. Arboleda, R.A., Catane, S.G., Delos Reyes, P.J., Martinez, M.L., Mirabueno, H.T., Regalado, T.M., Tubianosa, B.S., Umbal, J.V., Punongbayan, R.S., Newhall, C.G. and Alonso, R.A. (1995) Chronology of the 1994 lahars at Pinatubo volcano and consequent hazards and risks. Unpublished report of the Philippine Institute of Volcanology and Seismology, Manila. Arboleda, R.A. and Martinez, M.M.L. (1996) 1992 lahars in the Pasig–Potrero river system. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines, 1045–54. Seattle: University of Washington Press. Blong, R.J. (1982) The Time of Darkness: Local Legends and Volcanic Reality in Papua New Guinea. Canberra: Australian National University Press. Erickson, K.T. (1976) Everything in Its Path: Destruction of Community in the Buffalo Creek Flood. New York: Simon and Schuster. Galende, P.G. (1996) Angels in Stone: Augustinian Churches in the Philippines. Manila: San Agustin Museum. Hart, D.V. (1955) The Philippine Plaza Complex: A Focal Point in Culture Change. New Haven: Yale University Southeast Asia Studies. Henson, M.A. (1955) The Province of Pampanga and its Towns. Manila: Villanueva Book Store. Jones, J.W. and Newhall, C.G. (1996) Preeruption and post-eruption digital-terrain models of Mount Pinatubo. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines, 571–82. Seattle: University of Washington Press.

64

K.S. CRITTENDEN AND K.S. RODOLFO

Lacsamana, R. (1996) Certified Updated Masterlist of Current Residents of Barangays Cabambangan [the poblacion], Sn. Vicente, Sta. Ines and Cabetican, Bacolor, Pampanga. Unpublished ms (May). Lacsamana, R. and Crittenden, K.S. (1997) Unofficial Census of Current Residents of the Barangays Cabambangan, Sn. Vicente, Sta. Ines and Cabetican, Bacolor, Pampanga. Unpublished ms ( July). Lamug, C.B., Crittenden, K.S. and Nelson, G.L.M. (1999) Processes Through Which Families in Bacolor (Pampanga) Respond to Natural Disaster, with Emphasis on Relocation. Manila: Center for Integrative and Development Studies, University of the Philippines System. Larkin, J.A. (1993) The Pampangans: Colonial Society in a Philippine Province. Manila: New Day Publishers. Mori, J., Eberhart-Phillips, D. and Marlowe, D.H. (1996) Three-dimensional velocity structure at Mount Pinatubo: resolving magma bodies and earthquake hypocenters. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines, 371–82. Seattle: University of Washington Press. National Statistical Office (1990) Pampanga Census of Population and Housing, 1990. Manila: Philippine Government. National Statistical Office (1995) Central Luzon Census of Population, 1995. Manila: Philippine Government. Newhall, C.G., Daag, A.S., Delfin Jr, F.G., Hoblitt, R.P., McGeehin, J., Pallister, J.S., Regalado, M.M.M., Rubin, M., Tubianosa, B.S., Tamayo Jr, R.A. and Umbal, J.V. (1996) Eruptive history of Mount Pinatubo. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines, 165–95. Seattle: University of Washington Press. Newhall, C.G. and Punongbayan, R.S. (eds) (1996) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines. Seattle: University of Washington Press. Oliver-Smith, A. (1982) Here there is life: the social and cultural dynamics of successful resistance to resettlement in post-disaster Peru. In A. Hansen and A. Oliver-Smith (eds) Involuntary Migration and Resettlement: The Problems and Responses of Dislocated People, 85–103. Boulder, CO: Westview. Oliver-Smith, A. (1991) Successes and failures in post-disaster resettlement. Disasters 15: 12–23. Orum, A. (1998) The urban imagination of sociologists: the centrality of place. The Sociological Quarterly 39: 1–10. Pierson, T.C., Daag, A.S., Delos Reyes, P.J., Regalado, M.T.M., Solidum, R.U. and Tubianosa, B.S. (1996) Flow and deposition of posteruption hot lahars on the east side of Mount Pinatubo, July–October 1991. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines, 921–50. Seattle: University of Washington Press. Rodolfo, K.S. (1989) Origin and early evolution of lahar channel near Mabinit, Mayon volcano, Philippines. Geological Society of America Bulletin 101: 414–26. Rodolfo, K.S. (2000) The hazard from lahars and jokülhlaups. In H. Sigurdsson (ed.) Encyclopedia of Volcanoes, 973–95. San Diego: Academic Press. Scott, W.H. (1994) Barangay: Sixteenth Century Philippine Culture and Society. Manila: Ateneo de Manila University Press. Siringan, F.P. and Rodolfo, K.S. (1999). Increased Philippine flooding: more from local subsidence, less from global sea-level rise. Manila, 12 July presentation to the Office of Civil Defense. Umbal, J.V. (1997) Five years of lahars at Mount Pinatubo: declining but still potentially lethal hazard. Journal of the Geological Society of the Philippines 52: 1–19.

LAHAR DISASTERS IN BACOLOR, THE PHILIPPINES

65

Umbal, J.V. and Rodolfo, K.S. (1996) The 1991 lahars of southwestern Mount Pinatubo and evolution of the lahar-dammed Mapanuepe lake. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruption and Lahars of Mount Pinatubo, Philippines, 951–70. Seattle: University of Washington Press. Umbal, J.V., Delos Reyes, P.J. and Tuñgol, N.M. (1996) Assessment of the volcanic activity and lahar situation at Pinatubo Volcano. Manila, unpublished report of the Philippine Institute of Volcanology and Seismology.

5

Maritime archaeology and behaviour during crisis: the wreck of the VOC ship Batavia (1629) MARTIN GIBBS

INTRODUCTION Many of the chapters in this book and indeed most disaster studies emphasise the impact of a disaster on the environment and draw conclusions from their interaction as to the potential cultural response. Few actually address the direct cultural response to the hazard/disaster itself and it is in the nature of archaeological material that evidence for this is necessarily rare. This chapter throws light on the behaviour of one group of survivors; it is a gripping story of survivor mentality in an isolated community and the breakdown of established social order in the face of catastrophe. Archaeological studies of disasters invariably address the consequences of catastrophes by investigating either natural site formation processes or the longterm impacts on settlement patterns, social systems, resource availability and trade. Discussion of human behaviour immediately before, during and after the crisis itself is primarily used as a device to add drama to the discussion or to emphasise the notion of disruption. Most archaeologists seem to implicitly dismiss human actions during the crisis as being somehow ephemeral, unfathomable or, at worst, of little or no real consequence to the grand sweep of the archaeological story. Furthermore, crisis behaviour is considered to be unstructured and driven by panic. Shipwrecks, probably the class of disaster site most commonly investigated by archaeologists, are particularly susceptible to such lack of consideration. However, the heart of the crisis deserves serious study because it is during this time that many of the crucial decisions and actions that create the post-disaster archaeological record are taken. Psychological studies of modern disasters have demonstrated that individual and group behaviour during crises is patterned and to some extent predictable. This paper will argue that a behavioural model developed from these studies can also be applied to the interpretation of past disaster events, providing a structure for the stages by which cultural and natural factors create or impact upon the archaeological record. It also proposes that many of these same behaviours can potentially be recognised in the archaeological record.

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

67

To illustrate this argument, the events leading to and following the 1629 wreck of the Dutch East India Company (VOC) vessel Batavia will be briefly examined. I show how the analysis of the documentary and archaeological data might be informed through an understanding of human behaviours during crisis. While this paper focuses on the relationship between a behavioural model and archaeological interpretation, a more detailed consideration of the archaeological signatures involved with shipwreck sites and survivor camps is provided elsewhere (Gibbs, in press).

DISASTER STUDIES AND DYNAMIC BEHAVIOURAL MODELS The considerable body of literature on disasters can be broadly divided into studies on the structural nature and dimensions of the event(s), and investigations of behavioural responses. The former type are typological, categorising disasters primarily on the basis of physical properties, but with little capacity for understanding human responses. In contrast, the latter adopt a dynamic approach, developing ‘operational models reflecting the progress of a disaster mirrored in human behaviour’ (Leach, 1994: 6). These are also more appropriate to archaeological investigation, which needs to consider both the physical and the cultural aspects of disasters. Viewing disasters through human responses also means that investigation is not bound by the structure of each particular event in terms of whether it is natural or humanly caused. Most of the literature appears to agree that groups exhibit patterned behaviours with a highly consistent internal structure, regardless of the specifics of the crisis (Dynes and Tierney, 1994). That is, collective human behaviour in a crisis proceeds through a series of identifiable and predictable phases. Although various models for disaster response have been proposed (e.g. Wallace, 1956: 2; Frederick, 1987: 76), this paper will follow the five stages proposed by Leach (1994). 1 Pre-impact stage – the period before the disaster event: a. Threat phase – when the possibility of disaster is identified. b. Warning phase – when disaster is imminent. 2 Impact stage – during the disaster ‘event’ and immediately afterwards. 3 Recoil stage – which commences when the immediate threat to life has receded. 4 Rescue stage – when the person or group is removed from danger. 5 Post-trauma stage – the medium- to long-term responses to disaster. It should be noted that in this context, ‘Impact’ refers to the onset of the disaster and not necessarily a physical collision, although a physical aspect of some form is, of course, central to events. A model such as Leach’s allows us to break the course of a disaster into a series of recognisable stages of behaviour, each of which has an obvious relationship to

68

M. GIBBS

the physical progress of the event. More importantly, this disaster model appears to indicate that the basic individual and group responses during each of these stages will transcend cultural boundaries, especially immediately before, during and after the crisis event. With the possibility that these patterns of reaction might also be applicable over the greater span of human history, we have a potentially powerful interpretative tool that allows us to introduce a comparative structure to disparate disaster situations over both space and time. There is already a rich resource of archaeological, documentary, iconographic and oral historical evidence about disasters of various kinds extending over the last several thousand years (Landow, 1982; Hinton, 1992; Staniforth, 1992). The behavioural model creates a framework for the analysis and integration of these fragmentary sources. At the same time, it may be possible to characterise the range of behavioural and physical responses exhibited before, during, and after particular classes of disaster. Shipwrecks can be seen as one such class of disaster for which archaeological expressions are the wreck site itself and the survivor camp(s) that often result from them. In these instances the marine and terrestrial divide, which might include a marked physical separation of the archaeological materials, is irrelevant as all these sites and materials were created and modified by behaviours associated with the same crisis event. This is paralleled in other cases by the relationship between an original settlement destroyed by disaster and any subsequent sites created by survivors situated away from their homes. As I have argued elsewhere (Gibbs, in press), maritime archaeology, like other forms of disaster archaeology, has fallen into the trap of examining shipwrecks (and survivor camps) as unique events, rather than looking for the commonalities that link them. Where such links have been examined, they have usually been on the basis of common attributes of the original vessels or the physical site formation processes that act upon them (e.g. McCarthy, 1996; Veth and McCarthy, 1999; Ward et al., 1999). In particular, despite the considerable time since the publication of Gould’s (1983) Shipwreck Anthropology, there has been a continued reluctance to examine shipwrecks as behavioural events or to match archaeological signatures to human responses during the wreck event. The case study of the wreck of Batavia demonstrates how the relationships between behaviour, shipwrecks and archaeology might be conceived and examined in such a way as to make them comparable to other sites.

THE BATAVIA CASE STUDY On 19 October 1628 the Dutch East India Company (VOC) flagship Batavia departed from Texel in Holland as part of a convoy of seven ships heading to Batavia (Jakarta). Aboard were around 316 sailors, soldiers and passengers and a cargo of trade items including chests of coinage and various items intended as presents for the Mogul emperor at Agra. On 4 June 1629 Batavia, possibly deliberately separated from the convoy as part of an abortive mutiny attempt,

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

69

wrecked on the reefs surrounding the desolate Houtman Abrolhos Islands, located 60 km off the coast of Western Australia (Fig. 5.1). Approximately 250 persons survived the wreck event. The captain led 48 people in a successful openboat voyage 3,000 km northward to Jakarta and returned immediately in a rescue vessel. However, the story was to become notorious in maritime history because, in the several months before the rescuers returned, 125 of the approximately 198 men, women and children left behind were murdered as part of an insurrection by the junior officers left in charge (Green, 1989: 1). Rediscovered in 1963, the wreck site has been the focus of considerable attention by archaeologists from the Western Australian Maritime Museum (Green, 1975; Henderson, 1986). A lesser amount of work has taken place on the terrestrial sites associated with the survivors (Edwards, 1966; Bevaqua, 1974c; Green and Stanbury, 1988; Gibbs, 1992). Repeated historical retelling of the core story of the wreck of Batavia (e.g. Fig. 5.2) and the extensive archaeological investigation have perhaps bred the illusion that there is little more to be gained from further analysis. However, reexamination of these events in terms of Leach’s model will show that it is possible to provide new interpretations and insights. In particular, it will be demonstrated

Dirk Hartog Island

Shark Bay T

Western Australia

Zuytdorp (1712)

Zeewyck (1727)

Batavia (1629) Geraldton

Abrolhos Islands 0

100 km

Indian Ocean

Vergulde Draeck (1656)

Perth

Figure 5.1 Dutch shipwrecks on the west coast of Western Australia

70

M. GIBBS

Figure 5.2 Source:

The massacre of the Batavia survivors

Jansz (1647: fig. 3)

that ‘extraordinary’ events such as the Batavia shipwreck and aftermath are in fact amenable to structural analysis, in a way that allows them to be seen in comparison to a wider range of wreck incidents. The story of Batavia and the survivors of the wreck is complex and cannot be dealt with in full here. Instead, I will analyse a sample of the behaviours before, during and after the wreck event itself and will explain them in terms of the major stages of Leach’s framework. The primary sources for this analysis are the original accounts of the events; the journal of Francisco Pelsaert, the merchant in charge of Batavia, his reports of the interrogations of the mutineers and three short letters from other survivors. Translations of these original records have been published in Drake-Brockman (1963), Stow (1972) and van Huystee (1988). 1 Pre-impact stage The period preceding a disaster may include two parts, the first being the ‘threat’ phase in which the possibility of danger is identified and alertness increases. Following this there may be a ‘warning’ phase during which the danger is imminent and about to strike (Leach, 1994: 12). To use an analogy, this would be the difference between receiving a warning of icebergs in the immediate area versus actually seeing one off the bow. Different disaster situations have different levels of threat and warning.

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

71

In the case of Batavia, after rounding the Cape of Good Hope, the vessel was headed directly eastward on the ‘Brouwer Route’, following the trade winds with the intention of turning northwards as soon as the ‘southland’ (the west coast of Australia) was sighted. Although navigation in this era suffered from the lack of a method to determine precise longitude, there must have been some notion that they were approaching the coast and therefore the possibility of encountering uncharted reefs and shallows, if not the coastline itself. Surely extra caution was called for. However, at ‘two hours before daybreak’ when the captain queried what appeared to be spray a short distance in front of the ship, he simply accepted the lookout’s response of ‘Skipper, it is the shine of the moon’ (van Huystee, 1998: 1). This situation was echoed almost 100 years later when another VOC vessel Zeewyck was wrecked in the same area. Despite improved sailing instructions and the lessons of the earlier, well-known example of Batavia, the lookout admitted that he had seen the surf for at least half an hour but had not alerted anyone, imagining it was caused by reflection from the sky or moon (de Heer, n.d.: 7). In some respects, the responses aboard Batavia correspond to the main behavioural characteristics of the threat phase. These are inactivity and failure to take counter-measures, usually produced through a state of denial and feelings of personal invulnerability. It is even possible to acknowledge a threat while behaving as if it does not exist, a suspension of judgement referred to as cognitive dissonance. For instance, people may choose to live on the slopes of an increasingly active volcano or fail to take proper precautions against fire, flood, or human error, even when such events are highly probable and likely to result in catastrophe. We have no narrative insight into what happened aboard Batavia immediately before the collision, when the captain and crew realised their error. It may be that there was effectively no warning period before the ship drove on to the reef. If there was any recognition of impending disaster, it is probable that there was a series of desperate actions such as turning the rudder to change course, resetting sails, throwing out anchors and other manoeuvres to prevent or minimise wrecking. Behavioural models suggest that when a threat becomes immediate in the warning phase, there can be a switch to over-activity, but often of a non-effective nature. Leach (1994: 19) suggests that the main psychological reaction at this stage is still one of denial, with people frequently engaging in inappropriate behaviours or failing to respond because they are unable to grasp the situation and formulate a coherent plan of action. Even after days of warning such as a smoking volcano, earth tremors, or the advance of an invading army, some people are capable of completely ignoring a situation of imminent danger. Conversely, some individuals can identify a pre-impact threat or warning and are capable of thinking and responding effectively. This may be in part a result of specific training or prior experience of a disaster situation which will affect factors such as states of preparedness, the making of decisions to evacuate, and so on. In the pre-impact stage we see the first of the decisions and actions which will ultimately generate or affect the archaeological record; for instance, heeding warnings or acknowledging evidence of the potential for disaster and then taking

72

M. GIBBS

appropriate action based on this. On a ship this might mean a change in course, extra diligence at watch, or complete avoidance of a known risk area. There can even be longer-term strategies such as the construction of special vessels, introduction of navigation aids, or efforts to remove physical hazards in order to reduce or eliminate danger. In the long term, in terrestrial situations these preimpact strategies might include decisions to settle or not settle areas known to face threat or otherwise to develop appropriate ways to cope with the risk. Conversely, accepting that people are often willing to deny threat allows us to acknowledge that what we may see in the archaeological record is not a lack of understanding of their vulnerability, but certain behaviours despite it. Living on the slopes of a volcano does not indicate that people are unaware of the risk, but that they are willing to play the odds of an eruption against the benefits of access to fertile soil, while perhaps believing that should the worst happen, they will somehow be able to survive. Similarly, it is necessary to consider the extent to which evidence of pre-impact activity or even site abandonment may be indicative of the length of the threat or warning phases. For a shipwreck this may be visible in the extent to which a vessel had been turned before striking a reef. However, a lengthy warning phase does not necessarily mean that people will choose to respond (Leach, 1994: 23). It is possible that in a single site, various levels of response will be archaeologically visible. One only has to look at Pompeii to see that although many people left once the eruption of Vesuvius appeared imminent, the human remains recovered archaeologically testify to the decision by some not to flee (Lazer, 1997). 2 Impact stage Leach (1994: 25) points out that the actual ‘impact’ of a disaster usually lasts only seconds or minutes, although there are instances where this phase may last much longer. Impact-stage responses will also vary depending upon whether the disaster is catastrophic or is a low-intensity event, as well as whether there has been preimpact awareness or opportunities to mitigate the effects. In the case of a shipwreck, the impact response depends upon the nature and circumstances: e.g. how and where the vessel comes to grief, the rapidity with which it sinks or settles, and the opportunity to implement some strategy for saving the ship or escaping. The Batavia wreck event, consisting of the vessel striking a reef, was not particularly catastrophic of itself. Pelsaert’s written account of action aboard during the impact stage is quite restrained. He records being woken by ‘a rough terrible movement, the bumping of the ship’s rudder, and immediately after that I felt the ship held up in her course against rocks, so that I fell out of my berth’ (van Huystee, 1998: 1). Strong sensory impressions are a common feature of the impact experience (Leach, 1994: 23), with many survivor accounts of wrecks using similar impressions of sound, sight or physical feeling to describe their experiences. Leach suggests that at this point, the senses are usually overwhelmed and individuals may not be able to process information effectively, which results in bewilderment and inactivity. Even trained or experienced people may be temporarily overcome.

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

73

Pelsaert raced to the deck to find out what was happening. He confronted the captain with ‘Skipper, what have you done that through your carelessness you have run this noose around our necks?’ (van Huystee, 1998: 1), whereupon the excuse of mistaking surf for reflected moonlight was provided. It does not appear that the captain had taken any definite action at this stage, suggesting the temporary disorientation described above. However, provoked by Pelsaert’s demands, he quickly ordered depth soundings around the ship, followed by the serious step of throwing the cannons overboard to try to lighten the vessel and float it off the reef. When this failed, there was an abortive attempt to use anchors to pull the ship off and, finally, the desperate decision to cut down the huge main mast in order to stop it from grinding through the bottom of the vessel. None of these manoeuvres proved successful and may have even hastened the demise of the ship. Throwing overboard the cannons, anchors, cargo and eventually cutting away the masts were common strategies for sailing vessels in dire jeopardy. While these actions generally occur following the onset of the disaster, after the actual impact event, they may begin in the terminal stages of the pre-impact ‘warning’ phase. If successful, the vessel might limp away, leaving behind the flotsam and jetsam of its brush with disaster and potentially create a mystery for archaeologists as to why they are finding these objects, but not the wreck itself. Of course, jettisoning might be carried out and fail, as with Batavia, leaving a curious distribution of material which cannot be explained solely through the eventual collapse of the hull. In general the cannon and anchor from Batavia are clustered close to the main wreck site (Green, 1989: 6), highlighting the limited effect that the manoeuvre had in freeing the vessel. The one clear anomaly is one anchor (no. 9), which is located nearly 80 m north of the main body of the site and on top of the reef. The excavators have not ventured an explanation of this situation: e.g. whether it was the result of an action associated with an attempt to prevent the wrecking, contemporary salvage of some form, or a later attempt to remove the relic. In terms of other human action aboard the stricken vessel, Pelsaert briefly alludes to ‘the great wailing that there was in the ship, of Women, Children, Sick and anxious people . . .’ (van Huystee, 1998: 1). In addition, the efforts of those trying to salvage essential supplies ‘were impeded by the godless unruly troops of soldiers, as well as sailors, and their likes whom I could not keep out of the hold on account of the liquor or wine, so that one could not get in there’ (van Huystee, 1998: 1). Despite the continuing threat to their lives, these men had broken into the stores and begun a drunken rampage, rioting, looting and even destroying Pelsaert’s papers. Similar drunken mêlées are reported on other wrecked VOC vessels (de Heer, n.d.). The contrast between the different groups – the crew trying to save the ship, those people stricken with terror, and those acting irrationally – provides some insight into the structure of a crisis-stricken population during the impact stage. Leach (1994: 24) suggests that there are three consistent and quantifiable bands of response. The first group (approximately 20 per cent) is able to remain relatively calm and focused, either through natural ability, or because of training and experience. A few people will even become ‘supercool’, collecting their thoughts

74

M. GIBBS

rapidly and acting decisively and rationally. The second group (approximately 75 per cent) will be stunned and bewildered, with their reasoning and thought processes significantly impaired, resulting in reflexive and mechanical behaviour. They will also potentially exhibit perceptual difficulties, a sense of slowed time, diminished emotional response and a range of physiological reactions such as trembling, nausea, sweating and vomiting. The final group (as many as 10–15 per cent) will exhibit a high degree of inappropriate behaviour that is ineffective or may even increase the danger of the situation. Reactions might include confusion, weeping, screaming, or even paralysing anxiety, rendering them unable to make any physical movement, even to save themselves. Although at this level we are dealing with group responses, Leach (1994: 30–57) also goes into some detail on individual reactions during crisis events. He looks at expressions of panic, paralysing anxiety, denial, depression, apathy, hyperactivity, guilt and psychological breakdown. The most interesting individual reaction is irrational behaviour, sometimes associated with hyperactivity. Because people in this state appear decisive, active or purposeful, the more bewildered and passive participants may be led into danger by following foolish examples or instructions. The ability of a single decisive person, whether sane or not, to direct others into an activity regardless of whether it is rational or not, is important to keep in mind when trying to understand apparently illogical group responses and their material consequences. The most common psychological reaction during the impact stage is denial (Leach, 1994: 25). Just as during the pre-impact stage people believe that a disaster cannot happen to them, during the impact stage they experience a sense of disassociation – they still believe it is not happening to them. Most disaster psychologists appear to agree that experience and/or training are decisive factors in the ability to respond. Many sailors experience wrecks of greater or lesser severity several times during their career, which would potentially reduce inertia and assist response when faced with danger. It may also be that the type of vessel, whether naval, commercial, or passenger, needs to be considered as a means of determining the potential level of appropriate training, as well as the demographics of the population aboard. After it appeared that Batavia was doomed, having burst so that water was flooding the hold, Pelsaert decided to abandon the vessel and move to several small coral islands nearby. He ordered the recovery of supplies of bread and fresh water and organised for the safe removal of the money and jewels, although it seems that it was only after some consideration that he allowed the transportation of ‘most of the people’ (van Huystee, 1998: 2) before returning for the valuables. The three shorter, contemporary accounts agree that little was removed in this initial flight, other than a few barrels of biscuit and water (Stow, 1972: 8; DrakeBrockman, 1963: 264). The ship’s boats were used for this evacuation, which proceeded slowly throughout the day because of the worsening surf, wind and the consequent movement of the ship. Although 180 people had been carried to shore, it was not possible to get the boats close enough to remove the remaining 70 men, who were forced to spend the night on the rapidly deteriorating wreck.

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

75

Still unable to reach them in the morning, Pelsaert suggested that they make up two rafts from the timbers and make their own way to shore. It is important to note the physical consequences of the behaviour at the crisis stage. Jettisoning occurs in the first part as an attempt to save the vessel. However, when this is no longer possible, there is a shift towards collecting survival items, boats, water, bread, or the modification of available resources (the construction of rafts, for example) in order to leave as quickly as possible. It is interesting that in Pelsaert’s account the recovery of the money and valuables figures so prominently. However, in the context of the ship’s mission and the ruthless commercial atmosphere of the VOC’s operation, ensuring the safety of these goods may have been seen as essential for their personal survival by Pelsaert and the other responsible officers. The directors (the Heren XVII) and senior officials of the VOC imposed severe penalties, including forfeiture of possessions, jail terms, or even torture and capital punishment, should they find a person negligent of company law (Drake-Brockman, 1963: 99–102). A final physical consequence of the impact stage might be loss of life, through drowning, injury or other causes, although Pelsaert failed to keep a count of how many perished or the fate of the bodies. Based on other figures in the documentary accounts, as many as 70 lives may have been lost in the first few hours of the wreck and abandonment, although no human skeletal material has been recovered from the wreck site itself (Green, 1989). 3 Recoil stage The recoil stage is usually regarded as starting when the immediate threat to life has receded. Aboard a ship, this may even begin for some individuals once it has been established that the vessel will not sink immediately. In the case of the majority of the survivors of Batavia, the recoil stage probably only began once they had been placed ashore on the several small coral islands near the wreck. However, any initial relief at their apparent salvation would have been replaced with a renewed desperation when they realised that they were on a series of small, desolate islands which held no fresh water. The supplies of water and wine salvaged from the ship were exhausted within one or two days, so that some people died of thirst, while others survived only by drinking their own urine until a rain storm five days after the wreck (Drake-Brockman, 1963: 264). Supplies of bread were similarly limited. This raises the question of how physiological conditions might have affected individual and group behaviours in the weeks and months that followed the wreck. As early as the first morning after the wreck, the officers had begun to organise for survival. There are textual references in all of the contemporary accounts which clearly indicate the construction of rafts, searches for water, subsistence hunting for seals and foraging for marine life on the surrounding reefs, as well as the construction of tents and camps. Salvage of the wreck was also organised, retrieving not only items with survival value, but also whatever valuables remained accessible. In another paper (Gibbs, in press) I have outlined the key themes for investigating the physical and organisational attributes of shipwreck

76

M. GIBBS

survivor camps. I have also argued that the study of these camps is really an investigation of adaptation processes and the ways people utilise their social and physical resources in order to endure (cf. Kirch, 1980). Post-wreck survival strategies may be related to the composition of the survivor group, accessibility of the wreck and the nature of the surrounding environment. Having the available skills and resources to construct a small rescue boat and sail away will naturally require a pattern of salvage and other activities. These lead to a different archaeological signature from that of a group stranded and forced to wait for rescue. However, and perhaps more significantly, strategies are related to the psychological state of the survivors. Particularly important is the development of an effective authority structure which can organise the group towards various goals (Gibbs, in press). There is also likely to be re-establishment or increasing influence from culturally dictated behaviours and structures, dependent again upon the composition of the survivor group. Previous studies of the archaeology of Batavia survivor camps have recorded the distributions of material across several islands in the Wallabi group of the Abrolhos Islands (Bevaqua, 1974c; Kirkham, 1980; Orme and Randall, 1987; Gibbs, 1992, 1994). These are obviously a function of both the mundane and the extraordinary aspects of the events that took place. Leach (1994: 26) suggests that the recoil stage starts with confusion and group fragmentation, followed by a slow return to awareness, reasoning ability, recall and emotional expression. Responses may vary between fear, resentment, anxiety and especially anger, and even be manifested as hyperactivity to find family or friends. For some period, there may be a child-like dependence upon others and the formation of loose but unstable groupings, still characterised by the inability to take decisive action without external direction. People continue to suffer denial or an irrational anger and need to blame others. We can see hints of this confusion in the days following the wreck of Batavia, exacerbated by the physical and mental privations of dehydration and being forced to drink urine to survive. This makes it all the more extraordinary when, after several days of unsuccessfully hunting for water on the surrounding islands, Pelsaert, the captain, all the senior officers and some of the passengers (totalling 48 people) sailed away in the ship’s boat to seek help at Batavia (Jakarta). Although Pelsaert later claimed that he had been forced to take this action against his will, this unexpected abandonment must have been a demoralising event for the 198 survivors remaining on the Abrolhos. The junior officers left in charge of the survivors included several who had been contemplating mutiny and the hijack of Batavia and its treasures. Over time these men, possibly spurred on by certain heretical philosophical beliefs held by the ranking officer (Tylor, 1970), developed a plan to capture the rescue ship and depart with the treasure. This plot required the number of survivors to be dramatically reduced in order to conserve resources and eliminate possible resistance. Initially, using their legitimate authority, they separated the survivors into smaller groups, placing them on different islands, while accusing a number of men of crimes against the company as an excuse to exact the death penalty (Drake-Brockman, 1963: 114). Over time the legitimate authority of the con-

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

77

spirators was subverted, and was ultimately replaced by their own code, which allowed them to steal, rape and systematically murder the unarmed populations on each of the islands. The dominant authority of the mutineers, who were in fact the legitimate officers, combined with the disoriented state of a population in the post-disaster recoil stage, may explain the apparent failure of the passengers and crew to mount an effective resistance. In addition, there was not only the threat of physical violence, but also what might be termed moral violence, where the dramatic warping of social rules by the figures in authority probably left the survivors in a heightened state of confusion, denial and ineffectiveness. The only exception to this general state of unresponsiveness was a group of soldiers, isolated on another island, who were able to organise and defend themselves. This may be because their training or prior experience of crisis situations provided them with some buffer to the events taking place: they had their own military authority model to fall back on, and the expertise to develop a defensive plan. There is undoubtedly ample material for a detailed sociological and psychological study of the mechanics of this insurrection, which, in fact, became an object lesson to later sailors and may have conditioned standing orders for later VOC ships. While the specific events associated with the Batavia survivors are unique, they highlight how post-disaster group dynamics might seriously affect the nature of survival strategies and consequently the archaeological record. The mechanisms and outcomes of organising and controlling people and resources are visible in the physical remains of the survivor camps. For instance, Beacon Island, the site of the main survivor camp and later the mutineers’ camp, shows evidence of at least two major occupation areas. One site is located closer to the wreck and adjacent to the only beach where a boat landing might be made, while the other is at the far end of the island (Bevaqua, 1974a; Gibbs, 1992). Although there has been insufficient archaeological material recovered to undertake any sort of a meaningful comparative analysis (Bevaqua, 1974b; Kirkham, 1980), it seems probable, and archaeologically testable, that this division arises from the separation between different classes within the survivor population. More pointedly, there is the possibility that the site closer to the wreck may be associated with the group controlling the resources taken from it, as well as other foraged materials being landed on the island. Based on the documented history, this was most likely the mutineers, who also had exclusive access to weapons and a range of other items. Similarly, the distribution of sites on several of the other islands probably reflect the division of survivors by the mutineers as a means of reducing possible resistance (Bevaqua, 1974c; Orme and Randall, 1987). 4 Rescue stage Remarkably, Pelsaert completed his 3,000-km boat voyage to Jakarta and departed again almost immediately with a rescue vessel, although by the time he returned to the Abrolhos, a total of three months had passed since the wrecking. Fortunately, before the mutineers could carry out their plan to capture the rescue vessel, the soldiers, who had by now successfully resisted several attacks, were able

78

M. GIBBS

to warn Pelsaert and assist in the capture of the miscreants. Pelsaert quickly learned of the insurrection and the depravities that had taken place, as well as how 125 of the 198 people left behind had been murdered. Of the original 316 who had sailed on Batavia only 116 had survived, with the majority dying as a result of factors not associated with the initial wreck event (Green, 1989: 1). Leach (1994) provides no significant information on behaviour during rescue and appears to mark it only as a turning point in the progression of events, rather than a separate stage in itself. However, for survivors the rescue event could involve a dramatic reversion to familiar authority patterns, the need to justify actions during the crisis, and the reality of impending return to the normal social world. The confessions of the mutineers extracted under torture by Pelsaert, and which form the bulk of the documentary record of the Batavia disaster, contain not only detailed accounts of atrocities, but evidence of the confusion that many felt as to their motives for participation (e.g. Drake-Brockman, 1963: 202). Additionally, after months of mistreatment, the remaining passengers and crew must have experienced similar reactions to those of the survivors of protracted terrorist attacks or hostage situations (cf. Frederick, 1987). Some insight is provided in a guilt-ridden letter written by the official Predikant or minister aboard Batavia, whose lack of resolve in stemming the mutineers, despite his official status as moral guardian, not only allowed much of what happened, but resulted in the murder of his own family (Drake-Brockman, 1963: 77, 263). Although the Predikant’s letter was produced after his removal from the Abrolhos, the rescue stage is significant to researchers of historic disaster events as it is the period when the first oral and written testimonies of survivors are likely to have been recorded. The rescue stage also signals a potentially new episode in the archaeological history of the survivor camp and the wreck, should the latter remain accessible. Whereas survivors in the midst of the crisis would be limited in their ability to salvage a vessel, the rescuers might bring with them extra resources of equipment and labour to begin a far more extensive and systematic retrieval of material. In the case of Batavia, when Pelsaert returned with a rescue ship, he was also under explicit orders from the governor-general of Batavia to recover ‘as much money and goods as can be found’ (Drake-Brockman, 1963: 257). Since he was probably hoping to mitigate his own culpability in the loss of the ship and valuables, Pelsaert did not depart the islands until he was ‘wholly convinced that nothing more is to be found . . . seeing that all has been searched through and dived over’ (Drake-Brockman, 1963: 221). In this instance the extensive salvage undoubtedly has consequences for the archaeological record, although in other cases the removal of cargo and ship’s remains may happen to a greater or lesser extent. Abandonment of a wreck and its cargo may leave it open to opportunistic salvage, either by contemporary people, or by later interests, including archaeologists. 5 Post-trauma stage The post-trauma stage is described as the period when the disaster event has reached its conclusion and the participants are completely removed from the

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

79

original threat. In the case of Batavia, the survivors were transported away from the Abrolhos Islands, with little or no chance of ever revisiting what was then a remote part of the world. However, a number of contemporary vessels did try to relocate the wreck and survivor camp (Drake-Brockman, 1963: 81), with subsequent attempts over the next 400 years to find the scene of the now notorious mutiny, culminating in the 1963 rediscovery (Edwards, 1966). The medium- to long-term reaction to disasters such as shipwrecks can include acute psychological dysfunction, anxiety, depression and psychosomatic disorders from which some people never recover (Leach, 1994: 28). Consequently, there are various implications for the archaeological and historical study of crises. The first is that it is during the post-traumatic period when the majority of the documentary accounts were compiled, whether written by the survivors, by official agencies interviewing participants, or by other secondary sources. One might reflect that the preponderance of disaster accounts published by survivors over the last several centuries might have been less to serve as titillation for a curious public as attempts to cope with what had happened to them. The several accounts from Batavia survivors should be analysed in this light, particularly as they represent different perspectives on the incident. The Predikant’s letter, described above, is written by a person at the heart of the atrocities (DrakeBrockman, 1963: 263). In contrast, another is from a person who managed to escape to the soldiers and resist the mutineers (Stow, 1972: 8). A third is written by a sailor who accompanied Pelsaert to Jakarta, and then returned with the rescue party to be confronted with the unhappy fate of his shipmates (ibid.: 10). From a different perspective, the post-trauma stage is when long-term decisions are made as to the reuse of a site or the materials within it. There are several studies of how survivors, descendants and even non-participants assign significance to disaster sites, including the factors which either encourage or discourage them to revisit or reuse such places (Read, 1996; Byrne, 1997). Understanding how people relate to disaster sites might explain past decisions to completely abandon a place, attempt some form of salvage, or eventually reoccupy sites in the shorter or longer term. These sorts of decisions will ultimately affect the nature of the archaeological record. Revisitation of disaster sites might be a purely ephemeral event, related to commemoration and with little or no physical impact. Conversely, there may be extensive salvage and reuse of materials. A wreck is, of course, likely to have a somewhat more limited range of possible responses associated with it, and these will be mostly related to salvage. However, recognition of the past tragedy might mitigate any activities of this nature, especially if human remains are present, as most recently seen with the Titanic (Ballard, 1987: 210). The complexities of the story of the Batavia survivors might appear to bear more relevance to sociology than archaeology; but this case well illustrates the value of Leach’s (1994) behavioural model for interpreting the material remains left behind. In particular, the removal of material from the wreck, the fragmentation of the group into discrete and archaeologically identifiable sites, and the distribution of material within and between sites, are predictable as crisisinspired responses.

80

M. GIBBS

THE ROLE OF ANALOGY As Leach (1994: 1) has noted, the ‘physical, environmental, physiological and psychological facts of survival, are better understood once the architecture or anatomy of a disaster is appreciated’. It would appear that the dynamic psychological models derived from modern group behaviour during disasters can also be used as a framework for the interpretation of past crisis events. With improved knowledge of the likely range of human responses to crises, the documentary, iconographic, oral historical and archaeological records of early catastrophes can be analysed and interpreted at a more sophisticated level. As shown in the Batavia example, not only are the different stages of response evident within the surviving contemporary documentary accounts, but the archaeological record of the disaster (the wreck and survivor camp) has been created as a function of these stages of response. The question is whether it is possible to develop a broader archaeological model based on analogies with either the material consequences of modern disasters or with historical situations such as Batavia, where there is a close concordance between the documentary and archaeological records. In this sense there is value in reviewing the ‘relations of relevance’ argument presented by Lewis-Williams (1991) for his application of a neuropsychological model to the interpretation of rock art. Lewis-Williams (1991: 152) states that the use of analogy in archaeology is based on uniformitarian principles, with researchers trying to identify behavioural systems that consistently produce particular artefact patterns similar (or identical) to those found in the archaeological record of the site(s) being examined. He suggests that [an] argument based on a strong relation of relevance starts by demonstrating the existence of some causal or otherwise determining mechanism that links two features in the source of the analogy. In other words, if feature A is present in the source of the analogy, the enabling mechanism or relation of relevance ensures that B is also present. Then, if we can show that B is present in the archaeological context and if we are confident that the same mechanism existed in the past, we can conclude that A was also present. (Lewis-Williams, 1991: 152) Using this neuropsychological model, Lewis-Williams argues that the enabling mechanism is physiological rather than cultural and that there are similarities in particular basic art motifs which result from a common origin in the human central nervous system. Since it is unlikely that the human nervous system has changed significantly in the last 40,000 years, he suggests that this explains the repeated appearance of particular images produced around the world throughout that period. However, he acknowledges that the interpretation of these motifs in any time or place is strictly cultural. Similarly, it may be possible to establish that the deeper structures of human behaviour during crisis have been present over some period of time and are

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

81

therefore likely to have influenced past reactions to disaster in the same way that they influence present behaviours. The question of whether and why there is both time-depth and a cross-cultural aspect in these behaviours may have resonance with some of the broader sociobiological arguments in the anthropological literature, although I will not discuss these here. However, if we accept that such enabling mechanisms exist in reference to behaviour during shipwreck events, we can apply these analogies to a wide range of archaeological situations. Furthermore, such a framework, with associated archaeological consequences, will, it is hoped, have a predictive quality (Gould, 1978). There is a difference between using this sort of reasoning for explaining prehistoric art and finding common links for a variety of archaeological sites arising out of behaviours associated with disasters. However, there may be ways to focus on the range of likely responses and thus archaeological consequences associated with particular types of catastrophe. Identifying and considering different ‘classes’ of disaster, such as flood, fire, volcanic eruption or shipwreck, may allow us to investigate the range of behaviours expressed towards those particular physical circumstances by different cultural groups or populations over time and space. Differences might be seen as ambiguities and thus fruitful areas for investigation, rather than dismissed as the product of the peculiarities of individual events. As a class of disaster, ships and shipwrecks have a number of defining characteristics that apply across time and space. There are consistencies in the physical nature of ships, as structures designed to pass over the water and in the constraints of being a closed system. There are similarities in the activities necessary to undertake the voyage, as well as in many of the basic social and political structures, which organised and controlled the people aboard. Ultimately, these similarities also limit the range of options that might be undertaken during a crisis. Table 5.1 proposes a correlation between Leach’s (1994) stages, behaviours and physical responses during a wreck event, and the possible archaeological signatures which might arise from these. Further detail on the different forms of salvage and the behaviour and archaeology of shipwreck survivors is provided elsewhere (Gibbs, in press). The crisis responses and possible archaeological signatures suggested in Table 5.1 are, by necessity, broad-brush outlines, which might be expanded and refined for any given situation. However, Table 5.1 creates a structure for looking at variations within and between a range of shipwreck events, allowing us to identify consistent patterns and, by extension, anomalies requiring further investigation. A better understanding of how sailors responded to disasters, developed through using examples where we have good historical and archaeological evidence, might create opportunities for interpretation of wrecks and survivor camps with poor preservation of documentary and physical remains. For instance, along the Western Australian coast the VOC ships Batavia (1629) and Zeewyck (1727) have substantial contemporary documentary accounts that illuminate many aspects of the events surrounding the wrecking and subsequently the life in the survivor camps. These records can be compared with

82

M. GIBBS

Table 5.1 Predictions for behaviours and archaeological signatures at various stages of a shipwreck crisis Stages (Leach, 1994)

Strategies, options and actions

Pre-impact threat phase

Long-term ● Collection of information on the nature of potential threats, so that they can be evaluated and appropriate response formulated ● Decisions to take or avoid certain routes (sailing instructions) ● Design of vessels and equipage suitable to overcome threat ● Training of crew to increase ability to meet and deal with particular situations

Physical and archaeological signatures

●

●

If strategies to avoid impact are successful, this may result in lack of archaeological evidence Where archaeological remains do exist, these may exhibit evidence of pre-impact strategies to diminish or negate risk

Short-term Changes to course and increased awareness for lookouts ● Preparation or stowage of equipment ●

Pre-impact warning phase

● ●

● ●

Impact

Radical changes to course Attempts to slow or stop ship, including dropping anchor Jettisoning of some items Pre-impact abandonment possible but unlikely

Strategies and actions dependent upon the nature of impact (catastrophic vs low intensity). Efforts to save vessel might include: Decision to remain aboard ● Club-hauling (use of anchors) to pull off from reef, shoal, or shore ● Driving the vessel over obstacle ● Jettisoning heavy items or cutting away masts in order to refloat or save the structure ● Patching leaks or holes until substantial repairs can be made Decision to abandon a vessel ● Lowering of the ship’s boats or lifeboats ● Securing a line to shore (if possible)

●

●

●

●

●

●

Pre-impact behaviours, if effective, may result in lack of archaeological evidence, a debris trail, or jettisoned items, but no wreck The disposition of a wreck and the presence or absence of materials may be indicative of the level of pre-impact awareness, preparedness and response If impact is negated, the vessel may be recovered, resulting in no archaeological signature or jettisoned materials only If unsuccessful, archaeological site may include ship’s structure, cargo and human fatalities Evidence of ‘crisis salvage’ – absence of primarily survivaloriented materials, including boats from the wreck site, or evidence of the same at land sites Salvaged materials and other evidence of occupation at site of initial landing

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

Table 5.1

continued

Stages (Leach, 1994) Impact (cont’d)

Strategies, options and actions

●

● ●

●

Recoil

●

●

●

●

Rescue and post-disaster

83

●

Physical and archaeological signatures

Rapid selection and removal of primarily survival-oriented materials (‘crisis salvage’) Removal of people Initial post-disaster survivor landing site (if possible) Discard of human remains resulting from post-impact mortality Establishment of survivor camp Establishment of authority structure and possible reorganisation of population Organisation of subsistence strategy and rescue strategy Further selection and removal of materials (‘survivor salvage’), assuming that a return to the vessel is possible. Limited by available labour and equipment (see Gibbs, 2000) Complete abandonment of wreck and/or survivor campsite

Post-disaster salvage – varying levels of removal of cargo, fittings and structure depending upon accessibility of sites and benefits versus the cost, effort and time required. ● Opportunistic salvage: often by people who do not necessarily have a direct link to the vessel or rights to remove material. Likely to be of short duration and intensity, resulting in focus on particular types of material ● Systematic salvage, possibly by rescuers with access to increased equipment and labour, introduces opportunities for systematic removal of material, including recovery of all or part of the cargo and ship’s structure over an extended period

●

●

●

●

●

Establishment of survivor camp. Distribution of material within land sites, reflecting survival strategies Evidence of ‘survivor salvage’ in the form of further materials absent on wreck site or located within the land site Evidence of adaptation of materials and foraging behaviour Evidence of human fatalities may indicate unsuccessful strategies Evidence of removal or nonremoval of materials from wreck and land sites

84

M. GIBBS

the well-preserved archaeological sites below and above water. However, this situation contrasts to two other VOC wrecks, located in the same region (Fig. 5.1). The Vergulde Draeck (1656) has historical and archaeological evidence of the wreck itself, although neither contemporary nor modern investigators have been able to determine the fate of the 70 survivors last seen awaiting rescue or the location of their campsite. Similarly, the Zuytdorp (1712) vanished without trace until the modern rediscovery of the wreck and an assemblage of related materials situated on the cliff high above the site. Just as the actions of the crew during the wreck remain open to speculation, so too does the question of whether the clifftop sites are the creation of survivors or the consequences of salvage by Aboriginal people. In these instances, identifying consistent patterns of behaviour and archaeological correlates within the known examples might provide insights as to how VOC captains in that time period responded to threat, as well as illuminate the undocumented instance. Similarly, an understanding of how survivors located and organised their campsites, the strategies employed to cope with the environment and the range of materials salvaged from the wrecks might assist in locating the Vergulde Draeck camp, or in confirming that the Zuytdorp sites are consistent with survivors, rather than indigenous salvage.

CONCLUSION Maritime archaeology has been particularly limited in its attempts to develop comparative frameworks of any sort, partially because of its historical, particularistic approach to wrecks and wreck events. Failure to attempt to discern patterns within and between sites has in many ways limited engagement with wider theoretical concerns and in particular the relationships between the archaeological record and past human behaviours. In consequence, maritime archaeology has often been identified as a second-rate and slightly suspicious sub-discipline in contrast to other fields of archaeological endeavour. The use of Leach’s (1994) behavioural model of crisis response presents us with a new way of investigating the archaeology of disasters, including shipwrecks. His framework allows us to interpret human actions during catastrophes and their aftermath, despite their extraordinary nature. The potential of this approach to introduce a clear structure to behaviour during and after a shipwreck is demonstrated by the re-analysis of the well-known events surrounding the Batavia. The framework also allows us to identify archaeological material, which correlates to the different stages of response during crisis, highlighting the pivotal role that each stage plays in the creation of the archaeological record. This opens many exciting possibilities for comparative research on shipwrecks and the chance to identify patterns within both behaviour and archaeology. It also suggests that we may be able to use these patterns to interpret events at sites with poor documentation, or to assist in locating and identifying the sites themselves.

CRISIS BEHAVIOUR: THE WRECK OF THE BATAVIA

85

REFERENCES Ballard, J. (1987) The Discovery of the Titanic. New York: Warner/Madison Press. Bevaqua, B. (1974a) Archaeological survey of sites relating to the Batavia shipwreck. Fremantle: unpublished report to Department of Maritime Archaeology, Western Australian Maritime Museum. Bevaqua, B. (1974b) Report of a test excavation on Beacon Island. Report No. 1, Fremantle: Department of Maritime Archaeology Western Australian Maritime Museum. Bevaqua, B. (1974c) Archaeological survey of sites relating to the Batavia shipwreck. Early Days 7(6): 50–78. Byrne, D. (1997) The archaeology of disaster. Public History Review 5/6: 17–29. de Heer, C. (n.d.) The Wreck of the East-Indiaman Zeewijk at the Abrolhos Islands in the Year 1727. A translation of the ship’s journal, with a short introduction and notes. M.S. on file. Fremantle: Western Australian Maritime Museum. Drake-Brockman, H. (1963)Voyage to Disaster. Sydney: Angus and Robertson. Dynes, R. and Tierney, K. (1994) Disasters, Collective Behaviour and Social Organisation. Newark: University of Delaware Press. Edwards, H. (1966) Islands of Angry Ghosts. Sydney: Angus and Robertson. Frederick, C. (1987) Psychic trauma in victims of crime and terrorism. In G. VandenBos and B. Bryant (eds) Cataclysms, Crises and Catastrophes: Psychology in Action, 59–108. Washington, DC: American Psychological Association. Gibbs, M. (1992) ‘Batavia’s Graveyard’: A Report on Archaeological Survey and Excavations on Beacon Island, Wallabi Group, Houtman Abrolhos. Report No. 59, Fremantle: Department of Maritime Archaeology, Western Australian Maritime Museum. Gibbs, M. (1994) Report on the Excavation of Skeleton SK5, A Victim of the Batavia Massacre of 1629, Beacon Island, Western Australia. Fremantle: Report for the Department of Maritime Archaeology, Western Australian Maritime Museum. Gibbs, M. (in press) The archaeology of crisis: shipwreck survivor camps in Australasia. Historical Archaeology 37(1). Gould, R. (1978) Beyond analogy in ethnoarchaeology. In R. Gould (ed.) Explorations in Ethnoarchaeology, 249–93. Albuquerque: University of New Mexico Press. Gould, R. (ed.) (1983) Shipwreck Anthropology. Albuquerque: University of New Mexico Press. Green, J. (1975) The VOC ship Batavia wrecked in 1629 on the Houtman Abrolhos, Western Australia. International Journal of Nautical Archaeology 4(1): 43–64. Green, J. (1989) The Loss of the Verenigde Oostindische Compagnie retourschip Batavia, Western Australia, 1629: An Excavation Report and Catalogue of Artefacts. International Series 489. Oxford: British Archaeological Reports. Green, J. and Stanbury, M. (1988) Report and Recommendations on Archaeological Land Sites in the Houtman Abrolhos. Report No. 29, Fremantle: Department of Maritime Archaeology, Western Australian Maritime Museum. Henderson, G. (1986) Maritime Archaeology in Australia. Nedlands: University of Western Australia Press. Hinton, P. (ed.) (1992) Disasters: Image and Context. Sydney: Sydney Association for Studies in Society and Culture. Jansz, J. (1647) Ongeluckige Voyagie van’t Schip Batavia. Amsterdam. Kirch, P. (1980) The archaeological study of adaptation: theoretical and methodological issues. In M. Schiffer (ed.) Advances in Archaeological Method and Theory: Volume 3, 101– 56. London: Academic Press. Kirkham, L. (1980) Beacon Island Excavation. Abrolhos Project 1980, Postgraduate Diploma in Maritime Archaeology. Perth: Western Australian Institute of Technology. Landow, G. (1982) Images of Crisis: Literary Iconography 1750 to the Present. London: Routledge and Kegan Paul.

86

M. GIBBS

Lazer, E. (1997) Pompeii AD 79: a population in flux? In S.E. Bon and R. Jones (eds) Sequence and Space in Pompeii, 102–20. Oxford: Oxbow Books. Leach, J. (1994) Survival Psychology. Sydney: Macmillan. Lewis-Williams, J. (1991) Wrestling with analogy: a methodological dilemma in Upper Paleolithic art research. Proceedings of the Prehistoric Society 57(1): 149–61. McCarthy, M. (1996) SS Xantho, an iron steamship wreck: towards a new perspective in Maritime Archaeology. Unpublished Ph.D. dissertation, James Cook University, Townsville. Orme, Z. and Randall, N. (1987) A survey of the historical limestone structures on West Wallabi Island, Houtman Abrolhos. The Bulletin of the Australian Institute for Maritime Archaeology 11(2): 25–31. Read, P. (1996) Returning to Nothing: The Meaning of Lost Places. Melbourne: Cambridge University Press. Staniforth, M. (1992) Shipwrecks: images and perceptions of nineteenth century maritime disasters. In P. Hinton (ed.) Disasters: Image and Context, 45–63. Sydney: Sydney Association for Studies in Society and Culture. Stow, R. (1972) Two letters of 1629 on the Batavia disaster. Westerly, 1, April: 7–11. Tylor, P. (1970) The Batavia mutineers: evidence of an Anabaptist ‘fifth column’ within 17th century Dutch colonialism. Westerly, 4, December: 33–45. van Huystee, M. (1998) The Batavia Journal of Francisco Pelsaert. Algemeen Rijksarchief [ARA], The Hague, Netherlands Document 1630:1098 QQII, fol. 232–316. Translated and edited by M. van Huystee. Fremantle: Department of Maritime Archaeology, Western Australian Maritime Museum. Veth, P. and McCarthy, M. (1999) Types of explanation in maritime archaeology: the case of the S.S. Xantho. Australian Archaeology, 48: 12–15. Wallace, A.F. (1956) An Explanatory Study of Individual and Community Behaviour in an Extreme Situation: Tornado in Worcester. Disaster Study Number 3, Committee on Disaster Studies, Division of Anthropology and Psychology. Philadelphia: University of Pennsylvania. Ward, I., Larcombe, P. and Veth, P. (1999) A new process-based model for wreck site formation. Journal of Archaeological Science, 26: 561–70.

6

‘The end is nigh’? Social and environmental responses to volcanic gas pollution JOHN GRATTAN, MARK BRAYSHAY AND RUUD T.E. SCHÜTTENHELM

INTRODUCTION This chapter revisits themes that occur several times in this volume: that catastrophic events may be invisible in the archaeological record; major environmental trauma need not have a permanent impact on the cultures affected; and we can only understand the nature of events and cultural response by adopting the widest possible research framework. It is clear from earlier chapters that these concerns operate at the level of specific sites; here we explore these issues on a continental scale. This chapter presents compelling evidence to suggest that toxic gases emitted in a volcanic eruption may be transported over great distances and deposited in sufficient concentration to have a severe impact on environments and perhaps cultures in areas far removed from any apparent volcanic threat. This research has implications not only for the better understanding of the relationship between volcanic eruptions and the archaeological record, but also for the impact volcanic eruptions may have on contemporary human societies and environments. Consideration of the interaction of volcanic eruptions and human society has generally focused on the perilous situation of those living within sight of the volcano’s slopes. True, passing reference may be made to the natural fertility of volcanic soils, but most writers will move swiftly on to the dramatic hazards posed to human society by lava flows, lahars, pyroclastic flows, blast and ash falls. One need only consider the eruption of Vesuvius and the destruction of Pompeii to see this relationship starkly illustrated (see Allison, Chapter 7 in this volume). In the light of such obvious perils few authors pause to debate the potential hazard posed by volcanic gases such as sulphur, ammonia and fluorine; yet in Iceland in AD 1783, while no one was directly killed by the vast lava flows of the Laki Fissure eruption, a quarter of the island’s human population perished following the eruption. These deaths were the consequence of the environmental impact of the volcanic gases emitted, the destruction of crops and grazing, the deaths of nearly 75 per cent of the island’s livestock and the subsequent famine and disease (Jackson, 1982; Steingrímsson, 1998; Thórarinsson, 1979).

88

J. GRATTAN ET AL.

Where archaeologists and palaeoenvironmentalists have proposed the eruption of a distant volcano as the cause of cultural change or environmental stress, they have frequently invoked volcanically generated climate change as the mechanism (Baillie, 1988, 1989a and 1989b; Baillie and Munro, 1988; Ball, 1992; Burgess, 1989). However, volcanically induced climate change has been shown to be on a comparatively minor scale (Grattan and Gilbertson, 1994, 2000b; Grattan and Sadler, 1999a and 1999b) and no eruption of the past 3,000 years has reduced hemispheric temperature by more than 1 ºC, which is within normal fluctuation and hardly of itself likely to bring about long-lasting cultural or environmental change (Grattan et al., 1999). In historical times, where poor weather has been coincident with volcanic eruptions and demonstrated social and environmental stress, pre-existing social, cultural, economic, environmental and climatic trends have been in evidence and it is the combination of these that is significant, not the remote influence of a distant volcanic eruption (Ball, 1992; Dodgshon et al., 2000; Ogilvie, 1986). Where these cannot be identified in the archaeological record, volcanogenic climate change is a theoretical tool which must be used with caution (Grattan and Sadler, 1999a; Grattan et al., 1999). Even the palaeoenvironmental record is fraught with ambiguity: Blackford et al. (1992) presented an initially convincing correlation between the Hekla 4 tephra fall and the pine decline in Scotland, but this was not observed elsewhere in northern Scotland (Charman et al., 1995) or further afield in Ireland (Hall et al., 1994). These considerations lead one to consider whether volcanic eruptions exert an unambiguous influence on distant cultures and environments at all! A detailed analysis of the impact of an Icelandic volcanic eruption upon European environment and society in historical times may answer this question. The social and environmental impacts of the Laki Fissure eruption upon Britain and the mainland of Europe in 1783 were profound and illustrate a mechanism by which distant volcanic eruptions may influence distant environments and cultures by the eruption of mainly acid gases into the troposphere (Grattan, 1998). The clearest evidence to date of this phenomenon is to be found in documentary records made in Europe between June and August 1783, some of which have been reviewed by other scholars (Fiacco et al., 1994; Sigurdsson, 1982; Stothers, 1996; Thórarinsson, 1981; Wood, 1984, 1992). This was a period that coincided with the early eruptive phases of the Laki Fissure eruption in Iceland (Thordarson and Self, 1993). In many private journals, letters, scientific papers, newspaper articles and even in poetry, frequent references were made to the worrying presence of a ‘dry fog’, its impact on human health, the strange and often damaging environmental phenomena associated with it and the fear of Armageddon which this fog instilled in many parts of the community. Taken together, these documents demonstrate that toxic volatile gases emitted during volcanic eruptions may exert an influence upon distant peoples and environments in prehistoric and historic times. A selection of these documents is presented and discussed below.

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

89

A DRY FOG IN EUROPE In the summer of 1783 many writers recorded the ‘dry fog’ across large areas of Europe from Aberdeen to Naples and even across the Mediterranean to Malta and Tripoli. This dry fog was formed from the gaseous emissions of the Laki Fissure eruption in Iceland, which had been transported through the atmosphere by air circulation patterns and concentrated near the earth’s surface (Grattan and Pyatt, 1994; Stothers, 1996; Thordarson and Self, 1993; Thordarson et al., 1996). Sulphur gas emissions from the Laki Fissure eruption were amongst the greatest of the Holocene and it is thought that 60 per cent of the c. 90–190 million tons of sulphur emitted by this event (Clausen and Hammer, 1988, Devine et al., 1984; Fiacco et al., 1994) were discharged into the troposphere (Thordarson and Self, 1993). The meteorological research of Kington (1980, 1988) describes the atmospheric conditions that may have led to the concentration of volcanic gases in the air over Europe. From late June through most of July 1783 a relatively stable highpressure air cell was situated over Europe, and it was during this period that most of the dry fogs and associated phenomena described below were noted. Britain The summer of 1783 was an amazing and portentous one, and full of horrible phenomena; for besides the alarming meteors and thunderstorms that affrighted many counties of this kingdom, the peculiar haze or smokey fog, that prevailed for many weeks in this island and in every part of Europe, and even beyond its limits, was a most extraordinary appearance, unlike anything known within the memory of man. By my journal I find that I had noticed this strange occurrence from June 23 to July 20 inclusive, during which period the wind varied to every quarter without making any alteration in the air. (White, 1977: 265) White’s observation of the dates for these phenomena is important as Kington’s (1988) excellent daily review of weather in Europe for the 1780s establishes the presence of a relatively stable high-pressure air mass over Western Europe at this time. Most of the observations that follow were made between those dates. So long in a country not subject to fogs, we have been cover’d with one of the thickest I remember. We never see the sun but shorn of his beams, the trees are scarce discernible at a mile’s distance, he sets with the face of a hot salamander and rises with the same complexion. (Cowper Letters, 29 June 1783, cited in King and Ryskamp, 1981: 148) Two thirds of July was the same thick blue air as June ended with the month mainly hot, sometimes very hot. (Barker, n.d.: 196)

90

J. GRATTAN ET AL.

The state of the atmosphere for this week past has been more remarkably close and thick than was ever observed at this season. Such a haziness has prevailed, that the hills, at two or three miles distance, have not been discernible, and the appearance of the sun has been like that of a faint ball of fire, without a ray darting from it. (Bristol Journal, 19 July 1783: 1) In his detailed annual accounts of the weather, William Gilpin noted that in 1783, during almost all the summer months, the sky was overspread with a ‘dark dry fog’. Gilpin also reported that the ‘gloomy atmosphere was soon found to be general all over England. Advices from the continent: France, Spain, Italy and other parts, showed it had equally overspread all the countries of Europe and by degrees it was found to be universal over the face of the globe’ (Gilpin, n.d.). France Here the fog was equally widespread, reported from the channel coast to the Mediterranean and from the Loire Valley to the foothills of the Alps. During several of the summer months of the year 1783, when the effect of the Sun’s rays to heat the earth in these northern regions should have been greatest, there existed a constant fog over all Europe . . . this fog was of a permanent nature; it was dry, and the rays of the sun seemed to have little effect towards dissipating it. (Franklin, 1784: 374) Paris, July 4th: For a considerable time past the weather has been very remarkable here; a kind of hot fog obscures the atmosphere, and gives the sun much of a dull red appearance which the wintry fogs sometimes produce . . . those who are come lately from Rome say, that it is as thick and hot in Italy, and that even the top of the Alps is covered with it, and travellers and letters from Spain affirm the same. (Bristol Journal, 19 July 1783: 2) Paris, July 8th: Our atmosphere has, for many days past, been covered with a thick and dry fog, owing to the succession of great heat after long and heavy rains. (Morning Herald and Daily Advertiser, 15 July 1783: 2) A report from Salon de Provence in the south of France, dated 11 July, spoke of 20 days of continuous fog. The sun, although hot, does not dissipate it, . . . the countryside appears whitish grey . . . and the fog sometimes emits a strong odour and is so dry, it does not tarnish a looking glass and instead of liquefying salts, it dries them. (Aberdeen Journal, 18 August 1783: 2) M. Oueilhe, Curé of Larmon, near Toulouse:

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

91

This year there appeared, in the month of June, a thick fog, which hung between the sky and the earth, which was often called a dry fog because it did not moisten the earth, and was often mistaken for a thick smoke. (Rabartin and Rocher, 1993: 9) Germany A report from northern Germany stated that ‘The thick dry mist which has continued so long, seems spread over the whole of Europe . . . during the day it veils the sun and in the evening there is a tainted odour’ (Morning Herald and Daily Advertiser, 5 August 1783: 2). Reports in various German newspapers also describe the fog in Frankfurt, Mannheim and Meiningen. Netherlands The month of June was not noteworthy until the 19th when a remarkable fog contaminated all of Europe . . . This fog lasted between the 19th of June and the 30th of July and was different from other fogs in its constancy, its density and its extreme dryness. The hygrometer indicated an excessive dryness after the 23rd when the fog continued increased in strength for a few days. These conditions persisted for the whole month . . . Neither the storms of the 20th nor the strong winds of the 21st led to its abatement. (Swinden, 1786: 119) On the 24th of June the dry fog was very dense. In the northern Netherlands and adjacent parts of northwest Germany it was accompanied by a very pronounced sulphuric odour, while in more southerly provinces the odour was much weaker or absent. . . . it was remarkable that this fog persisted so long with continuing northerly winds. (Brugmans, 1784: 1) Italy The moon appeared ruddy and . . . the sun could be looked at without being blinded. The fog was hot, dry and dense, and this phenomenon was observed not only by us, but also elsewhere in Italy, Germany and France. (Gennari, 1783, cited in Camuffo and Enzi, 1995: 138) A letter from Naples, July 15, 1783. The fogs continue, and are accompanied with so alarming an increase in obscurity that our bargemen do not dare venture on the waters without compass. (Morning Herald and Daily Advertiser, 19 August 1783: 2) In southern Italy, the persistent dry fog may have been further reinforced by emissions from Stromboli, Vulcano and Vesuvius, which are reported to be active in 1783 (Simkin et al., 1981).

92

J. GRATTAN ET AL.

I am quite persuaded . . . that the dreadful eruptions of fire on the coasts of Italy and Sicily should have occasioned some alteration that has extended faintly hither, and contributed to the heats and mists that have been so extraordinary. (Walpole, Letters, 15 July 1783, cited in Cunningham, 1906: 358) Switzerland Fog began on June 17th and was stronger on some days than on others, but was spread equally around the horizon. The vapour obscured the atmosphere and it was possible to look at the sun all day without taking harm. . . . Rain showers did not disperse the fog, nor did it give way to storms, nevertheless, the fog continued to be excessively dry. (Swinden, 1786: 121) Mediterranean and North Africa The dry fog was also reported as having reached Malta and beyond. In Tripoli, in North Africa, the fog also appears to have persisted throughout July: ‘For this month past both land and sea have been covered with a thick fog . . . the sun appears but rarely, and when it does looks very red’ (The Gentleman’s Magazine, October 1783: 881). The strange fog was reported to have disrupted shipping around the Mediterranean. By the late mails from Africa it appears that the fogs in summer were thicker and more suffocating all along their coasts than with us in England, and that in the archipelago and along the Mediterranean Sea they were so thick as to render communication dangerous. (The Gentleman’s Magazine, September 1783: 803) It is clear from these reports, and from data reported by other scholars (Demarée et al., 1998; Lamb, 1970; Thórarinsson, 1981: Stothers, 1996; Fiacco et al., 1994), that the dry fog was distributed across Europe and beyond. The fog attracted considerable attention because of its unique properties: its persistence in the face of wind from all quarters, the inability of rain storms to disperse it, its tainted or sulphurous odour and its dry rather than moist nature. It is also clear from all the descriptions above that the constituent parts of the fog were concentrated near or at ground level, and are therefore not descriptions of a stratospheric aerosol. However, the dry fog was not solely a meteorological curiosity, for in many places its presence was often associated with damage to vegetation, people and occasionally to livestock.

ASSOCIATED ENVIRONMENTAL PHENOMENA The impact upon the environment in Iceland of the toxic material emitted in the Laki Fissure eruption was severe and is well documented (Gunnlaugsson et al.,

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

93

1984; Ogilvie, 1986; Thórarinsson, 1979; Steingrímsson, 1998). Elsewhere in Europe the dry fog also had a marked environmental impact, typically leaf defoliation and scorching and damage to crops. It is also associated with a wide range of health problems and even death. Sir John Cullum has left a particularly detailed account of the damage witnessed in his garden in Suffolk, reviewed in detail in Grattan and Charman (1994). The aristæ of the barley, which was coming into ear, became brown and weathered at their extremities, as did the leaves of the oats; the rye had the appearance of being mildewed; so that the farmers were alarmed for those crops. . . . The Larch, Weymouth Pine, and hardy Scotch fir, had the tips of their leaves withered; the first was particularly damaged and made a shabby appearance the rest of the summer. The leaves of some ashes very much sheltered in my garden suffered greatly. . . . Cherry-trees, a standard peach tree, filbert and hazel-nut-trees, shed their leaves plentifully, and littered the walks as in autumn. . . . All these vegetables appeared exactly as if a fire had been lighted near them, that had shrivelled and discoloured their leaves. (Cullum, 1786: 604) Elsewhere in Britain Monday night last [23 June], a very sudden and extraordinary alteration in the appearance of the grass and corn growing in this neighbourhood . . . in so much that the grazing land, which only the day before was full of juice and had upon it the most delightful verdure, did, immediately after this uncommon event, look as if it had dried up by the sun, and was to walk on like hay. The beans were turned to a whitish colour, the leaf and blade appearing as if dead. (Cambridge Chronicle and Journal, 26 June 1783: 4) On Wednesday June 25th it was first observed here, and in this neighbourhood, that all the different species of grain, viz, wheat, barley, and oats, were very yellow, and in general to have had all their leaves but their upper ones in particular, withered, within two or three inches at their ends; the forward barley and the oats most so . . . their awns appeared . . . withered also. Many of the oats’ . . . chaff husks were withered in like manner . . . About this time, and for 3 days both before and after, there was an uncommon gloom in the air, with a dead calm. The dews were very profuse. The sun was scarce visible even at mid-day, and then entirely shorn of its beams so as to be viewed by the naked eye without pain. (The Ipswich Journal, 12 July 1783: 4) France A number of accounts have been compiled by Rabartin and Rocher (1993).

94

J. GRATTAN ET AL.

The parish of Champseru has been afflicted by a pestilence which afflicted the throat. One believes that the dry fogs of May, June, July and August (1783), that turned the sun as red as blood, delivered this curse. (Rabartin and Rocher, 1993: 14) M. Picard Curé of Oinville wrote ‘This fog was nearly always dry and damaged the corn which yielded hardly any crop’ (ibid.: 13). Netherlands Two Dutch observers, Brugmans and Swinden, have left very detailed descriptions of severe acid damage to vegetation, insects, people and property: Between the 18th of June and the 21st of July the atmosphere was absolutely covered by the fog, and between the 22nd and 28th the air was very calm with little clouds. The sun, seen through the fog appeared to have a red face and was without strength, it proved possible to look directly at the sun at mid-day without damaging ones eyes, exactly as if viewed through smoked glass: nor was it possible to make out objects viewed at a distance without making a considerable effort. Such were the ordinary effects of the fog; from the morning of the 24th it was accompanied by a very perceptible odour of sulphur, which even penetrated into houses. Those people with weak chests experienced a similar sensation to that experienced when exposed to burning sulphur. On this day the sulphurous fog caused dramatic damage to the trees, which seemed to get worse throughout the day. . . . On the morning of the 28th the leaves of many trees were faded, grass and vegetables appeared likewise. Leaves and fruits fell as if in autumn: afterwards the whole countryside looked desolate. (Swinden, 1786: 120) On many days after the 24th June, in both the towns and countryside there was a strong, persistent fog, which attracted attention both because of the extraordinary phenomenon and because of its bad effects. . . . On the 24th the fog was very dense and accompanied by a very strong smell of sulphur, especially in the morning although it was still noticeable in the afternoon; it was noticeable not only because of the smell but also because of the taste. After the 24th, many people in the open air experienced an uncomfortable pressure, headaches and experienced a difficulty breathing exactly like that encountered when the air is full of burning sulphur, asthmatics suffered to an even greater degree: horses, cattle and sheep were not affected but the fog brought about a great extermination of insects, particularly amongst leaf aphids, but only those on the leaves which were affected. On the morning of the 25th the land offered an aspect of severe desolation, the green colour of the plants had disappeared and everywhere the

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

95

leaves were dry, just as in October or November; . . . in general it was not the fruits which were attacked but the leaves, which I observed to commence on the afternoon of the 24th. This affected a wide variety of plants: some were covered in spots, others changed gradually while some leaves dried up completely. Some leaves did not entirely deteriorate and these continued to grow, but their leaf tips were decayed. Another noticeable change was that in a moment the colour could change from green to brown, black, grey or white. Others kept their natural colour but overnight on the 25th their tips were wizened. Afterwards a great quantity of leaves fell. Many flowers were also attacked, but only the petals – not the fruits. The fall of leaves that summer caused many fruits to fail for lack of nourishment. It is demonstrated that the degeneration and leaf fall lasted over many days, and had not ceased by the 3rd of July, but the beginning of the malady is fixed on the 24th of June. The sap of some plants dried up at this time. Elsewhere the odour of sulphur was very strong, in Groningen where the brass doorknockers were tarnished to a whitish colour. (Brugmans, 1784: 11) Brugmans (1784: 15) studied the effects of the ‘sulphuric smog’ on over 200 species of plants immediately after 24 June, in the gardens and surroundings of his hometown of Groningen. He classified the susceptibility of plant species to the ‘sulphuric smog’ into four groups ranging from severely affected to barely affected plants. Highly susceptible series included trees such as Betula alba, Corylus avellana, Fagus castanaea, Pinus sylvestris, Populus alba, Salix, Tilia europea, Cedrus, Juglans regia and garden flowers such as roses, Calendula, Centaurea and Paeonia. Species that were hardly or not affected included Ilex, Quercus, Juniperus and garden plants such as Canna, Digitalis, Lathyrus, Passiflora and vegetables such as potatoes. Italy Camuffo and Enzi (1995) suggested that many people died in Italy of breathing difficulties during the acid fog which formed in the summer of 1783 and that, as in Iceland, cattle may have perished as a result of ingesting fodder and pasture which was contaminated by volcanic aerosols and microscopic tephra particles. Germany Accounts from northern Germany and the Netherlands reported not only the ‘infectious smell’ of the fog, but also that ‘all the trees on the borders of the [river] Ems have been stripped of their leaves in one night’ (The Ipswich Journal, 9 August 1783: 2). It is clear from the descriptions above that in many areas of Europe the dry fog had a severe if not devastating effect upon both ecology and people. The withering of certain crops, the shrivelling or burning of leaves, leaf fall and the destruction of

96

J.P. GRATTAN ET AL.

insects, coupled with the breathing difficulties described by Brugmans, clearly point to an environmental pollution event of unprecedented magnitude.

COINCIDENT HUMAN MORTALITY IN 1783 Many eighteenth-century accounts link the presence of the volcanic fog with headaches, eye irritation, decreased lung function and asthma. These are discussed in detail in Durand and Grattan (1999). Modern studies of anthropogenic air pollution incidents in and around major cities suggest that death rates may rise as vulnerable groups are affected by severe air pollution (Braun-Fahrlander et al., 1992; Brunekreef et al., 1991; Dockery et al., 1992; Durand and Grattan, 2001; Ostro et al., 1991). Several passages appear to link the presence of the dry fog to a rise in human mortality. The Curé of Landelles in France linked the fog to sickness and deaths: ‘The fogs have been followed by great storms and sicknesses which have driven a third of the men in many parishes to their tombs’ (Rabartin and Rocher, 1993: 10), as did the Curé of Broué: ‘While the sun was obscured there was a sickness which caused innumerable deaths’ (ibid.). Similar events are also reported from northern Italy (Camuffo and Enzi, 1995). There is no compelling reason to suggest that the volcanic air pollution of 1783 should not have caused deaths to occur, as it is clear that far smaller modern air pollution events of anthropogenic origin have an impact on human mortality. A consideration of demographic data compiled from English parishes adds quantitative support to this suggestion. Mortality patterns from widely separated English rural parishes suggest that an unusual crisis did occur during the summer of 1783 – awareness of physical process and the circumstantial evidence suggest acid volcanic gases may have been the key agent. Figs 6.1–6.4 are typical of the mortality patterns observed in many rural parishes in eastern England at this time; they show summer mortality far in excess of the mean for 1770–95. While these data are not conclusive proof of causality, it is clear that in these cases increased death rates and 7

Standard Deviation

6 5 4 3 2 1

1795

1794

1793

1792

1791

1790

1789

1788

1787

1786

1785

1784

1783

1782

1781

1780

1779

1778

1777

1776

1775

1774

1773

1772

1771

1770

0

Year

Figure 6.1 Banham, Norfolk: June–September mortality expressed as standard deviations from the 26-year mean

6

Standard Deviation

5 4 3 2 1

1795

1794

1793

1792

1791

1790

1789

1788

1787

1786

1785

1784

1783

1782

1781

1780

1779

1778

1777

1776

1775

1774

1773

1772

1771

1770

0

Year

Figure 6.2 Cavendish, Suffolk: June–September mortality expressed as standard deviations from the 26-year mean

Standard Deviation

6 5 4 3 2 1

1795

1794

1793

1792

1791

1790

1789

1788

1787

1786

1785

1784

1783

1782

1781

1780

1779

1778

1777

1776

1775

1774

1773

1772

1771

1770

0

Year

Figure 6.3 Castle Donington, Leicestershire: June–September mortality expressed as standard deviations from the 26-year mean 6

Standard Deviation

5 4 3 2 1

1795

1794

1793

1792

1791

1790

1789

1788

1787

1786

1785

1784

1783

1782

1781

1780

1779

1778

1777

1776

1775

1774

1773

1772

1771

1770

0

Year

Figure 6.4 Cranfield, Bedfordshire: June–September mortality expressed as standard deviations from the 26-year mean

98

J. GRATTAN ET AL.

observations of respiratory disorders are coincident in time with eruption of the Laki Fissure and the known impacts of acid aerosols, and are typical of modern demographic responses to air pollution incidents. Further work is under way to test this association and data sets from other European countries are being sought. However, modern case studies show that it is not unreasonable to suggest that the volcanic air pollution is at least a feasible hypothesis by which to account for these phenomena. Irrespective of the extent to which this past air pollution event was associated with increased death rates, it remains clear that the volcanic gases erupted from Laki were transported great distances through the atmosphere and had a range of significant impacts on human health. These events cannot have failed to alarm many people in 1783.

SOCIAL RESPONSES TO THE DRY FOG The dry fog, the pale or reddened sun and the various environmental effects described above had a marked impact upon particular sections of the community. In France the reddened sun and smoky air had ‘alarmed the superstitious part of the people, who had been wrought on by their priests to believe that the end of the world was at hand’ (The Edinburgh Advertiser, 15 July 1783: 43); priests were even forced to don their vestments and exorcise the fog. The French astronomer De La Lande even went to the trouble of writing a paper ‘to quiet the minds of the people’ which was widely reported in British magazines and newspapers (Grattan and Brayshay, 1995). But it was not only the people of France who came to fear the end of the world in the summer of 1783. In England, William Cowper wrote, ‘Some fear to go to bed . . . and assert with great confidence that the day of judgement is at hand’, and dismissed ‘the fallibility of those speculations which lead men of fanciful minds to interpret scripture by the contingencies of the day’ (King and Ryskamp, 1981: 148). In a similar passage Gilbert White also alluded to ‘a superstitious kind of dread with which the minds of men are always impressed by such strange and unusual phenomena’ (White, 1977: 265). The appearance of the sun was not the only worrying phenomenon, which turned people’s thought towards meeting their maker (see also Stothers, 1996: 79). The dry fog was accompanied by extreme weather in the form of high temperatures and a spate of ferocious storms of thunder and lightning, which swept across Europe between June and September 1783. It is clear that during this ‘amazing and portentous summer’ (White, 1977: 269) a strong fear of God and apprehension of Armageddon was felt in the minds of many people, despite the intellectual condescension of the better educated. It is not difficult to imagine how such events might have been perceived and interpreted in less enlightened times than the end of the eighteenth century. The various phenomena described in this paper, the feelings of fear and foreboding they generated, and the general sense of relief felt when worse did not follow are encapsulated in the following poem:

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

99

On the late unusual storms being followed By the earliest harvest within the memory of man. As, when the billows of the boiling deep The winds with unremitting fury sweep The crazy ship (all hope of safety lost) Is wafted sooner to the wish’d for coast; So, when of the late impetuous floods of flame In red confusion burst, and rolling came Tremendous peals of thunder; then with dread Shudder’d and look’d aghast each guilty head: But lo! th’alarming storm is heard no more; Lo! nature smiles more gaily than before; The noxious blights no more destruction bring; The fields in earlier season laugh and sing To the great God then be thy will resigned In judgement awful as in mercy kind. J.S. (The Gentleman’s Magazine, September 1783: 734) The authors have been unable to establish the identity of J.S. but his (or her!) poem neatly encapsulates the terror felt by those not in possession of a clear conscience. The documentary evidence presented has clearly demonstrated that the volcanic environmental pollution event triggered panic and alarm across Europe and possibly beyond. The provenance and constituency of the dry fog will now be examined.

COMPOSITION OF THE DRY FOG We may infer much about the nature and composition of the fog by reference to modern studies. Many of the symptoms presented above – the shedding of leaves, chlorosis, and dramatic changes of colour and the shortness of breath experienced by many people – suggest that a suite of toxic volatile gases was present in the dry fog. These are primarily sulphur, fluorine and chlorine, which is in broad agreement with the reported emissions of the Laki Fissure and is further confirmation of a volcanic source for this material. Much of the damage to vegetation is consistent with damage by acid deposition (Lang et al., 1980; Wilcox, 1959; Wisniewski, 1982); in particular the damage to the leaves of the trees is typical of the damage caused by the absorption of sulphur dioxide (Caput et al., 1978). The shedding of leaves is a classic response to concentrations of fluorine and hydrofluoric acid, and charring is typical of damage caused by a sulphuric acid aerosol probably below pH 2.5. In modern studies the condition of asthma sufferers has been observed to worsen when 24-hour mean concentrations of sulphur dioxide exceeded 250 µg/m3 (Koenig et al., 1979; World Health Organisation, 1979).

100

J. GRATTAN ET AL.

Reduced visibility has also been observed to occur when increased levels of sulphates are present in the lower atmosphere (Lee, 1983). Stothers (1996) has calculated that visibility at the surface was reduced to as low as 2 km and that the optical depth of the atmosphere may have been as high as 4 in some areas of Europe. Volcanic activity is undoubtedly the source for the toxic component of the dry fog in 1783.

DISCUSSION The events of 1783 exhibit all the criteria of an environmental forcing mechanism useful to archaeologists; one can observe strange meteorological phenomena, the blighting of crops and other plants, human illness and death, all of which prompted fear of God and some degree of social unease (Brayshay and Grattan, 1999; Grattan and Brayshay, 1999). All of these were caused by the eruption of a remote volcano of which, in the summer of 1783, no European chronicler was aware. The construction of a catastrophic model based on the events of 1783 is a clear, attractive and convenient option: a distant volcano erupts, noxious gases arrive from the heavens and kill crops and people, the culture decides the gods are against them, packs their bags and leaves, or are so weakened that their unaffected neighbours take the opportunity to pay off some old scores! Such a model is certainly more sustainable than those that attempt to invoke extreme climatic change generated by volcanic eruptions as the cause of sudden change in the archaeological record. As shown above, the largest eruptions of the Holocene did not have a massive or long-lived impact on climate, and any climate change caused is unlikely to have lasted long enough to bring about cultural change identifiable in the archaeological record (Grattan and Sadler, 1999a). Where volcanic eruptions, cultural stress and poor climate do occur together, it is important to establish the context of the events; the climatic and social trends, and the pre-existing social, economic and environmental stresses that may have been in operation. However, a hypothesis that used the events of 1783 to support the suggestion of catastrophic change induced by volcanogenic air pollution in the archaeological record is not immune from these contextual considerations and should be constructed with extreme caution. This can be illustrated when we consider the events of the year dispassionately. First, despite the descriptions of crop destruction in many parts of Europe, there were no agricultural crises; in fact rather the opposite was the case, as high temperatures and ample rainfall induced a bumper crop in many areas. Second, while some fragments of society clearly saw the hand of the supernatural in these events, community leaders did not, and took steps to calm the populace; the role of community leaders in archaeological contexts would be equally important. Third, illness and mortality induced by the poor air quality was patchy, and any alarm generated appears to have been entirely local. It may also be telling, that, despite the continental scale of the event and the impact that the dry fog had at the time, there is no evidence that this event entered folk memory to be remembered as a time of stress or hazard. The only accounts of the event to

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

101

remain in the public domain since the eighteenth century were those written by Gilbert White and Benjamin Franklin. White’s Natural History of Selborne has been in print almost continuously since first published in 1793, and the writings of Franklin, an influential man of science and letters, have always been the subject of public interest. In contrast to the accessibility of White and Franklin, the wide range of sources cited above is the result of ten years of scholarship and intensive search in the libraries of Europe. That an event of this scale was effectively forgotten suggests that we must use such material with caution. It is clearly dangerous to infer from bad news reported, even from a wide number of locations, that conditions were terrible everywhere. This problem is illustrated in a recent paper (Grattan and Gilbertson, 2000a), which examines very detailed and plausible descriptions of a volcanic eruption in central Germany in the eighteenth century. The ‘eruption’ was set at the Gleichberg, a mountain now recognised to have been a volcano – but during the Tertiary. While this eruption, the explosions, smoking and fleeing people were reported around Europe, there are no records of the event amongst the communities who actually lived within sight and sound of the eruption. The reports of the eruption are clearly a complex and plausible hoax, perpetrated for unknown reasons. However, the eighteenth-century eruption of the Gleichberg was falsifiable because of its setting in recent history; had this event occurred in ancient times and been reported by a Roman historian we would undoubtedly have accepted the account as true! That the events of 1783 had no obvious effect on settlement or society also indicates that we must use caution when interpreting and extrapolating catastrophes from fragments of ancient text, or folklore. The survival of a fragment of text from Swinden or Brugmans (cited above) for a few thousand years, coupled with some limited knowledge of the French Revolution and the archaeological identification of destruction layers, perhaps from the Napoleonic Wars, could all too easily be linked together into a grand theory of ecological catastrophe, social breakdown and war – yet we know that these are not linked. How securely then can we link the eruption of Hekla in the second millennium BC with the settlement abandonment in northern Scotland, the construction of defensive structures around Europe, the movement of the sea peoples around the Mediterranean world and the destruction of the city of Ugarit in the Levant? Yet this and similar arguments can be found in the archaeological literature (Grattan and Gilbertson, 2000b). While we reject use of this phenomenon in a catastrophic model, the events of 1783 make it possible to construct a Uniformitarian model with which to examine the role played by volcanic gases in prehistory. It is clear from the material presented in this paper that toxic volatile volcanic gases emitted in volcanic eruptions may be transported great distances and concentrated sufficiently to have a severe impact upon plants and humans. The cultural stress of such an event will depend on several factors, including the magnitude of the environmental forcing event and the sensitivity of the culture to the stress introduced by the volcanic gases (Grattan et al., 1998, 1999). Thus in a robust culture human illness and damage to plants and animals may result in no more than the

102

J. GRATTAN ET AL.

introduction of temporary stress from which recovery is rapid and assured. However, where a culture is located at the periphery, where marginal soils are occupied, where famine is a regular phenomenon, and where there is little margin for error, the stresses introduced by volcanic gases may have a profound impact. Such a model also allows an environmental influence for far smaller volcanic events than those necessary to cause climate change. While it is clear that eruptions on the scale of the Laki Fissure are rare, there appear to have been at least five in Iceland during the Holocene (Hjartanson, 1994; Simkin et al., 1981), including two during historical times (Stothers, 1996; Zielinski et al., 1994, 1995). Relatively minor eruptions have been associated with similar environmental damage. In Italy, Camuffo and Enzi (1995) have documented 19 episodes of crop damage caused by sulphurous dry fogs following the eruption of Italian volcanoes between AD 1374 and AD 1819. In an archaeological context, the cultural, social and environmental influences wielded by such event may have been profound.

CONCLUSIONS The events of 1783 clearly demonstrate that volcanic gases emitted in distant volcanic events may impinge upon distant peoples and environments. The presence of volcanic gases in the lower atmosphere, from even minor volcanic events, may result in environmental damage, the impact of which is surely dependent on the ability of the affected culture to adapt to short-term stress. In an archaeological context, where a society or culture is already vulnerable to environmental stress or a combination of adverse factors such as climate change, soil acidification, plague, blight or famine – to name but a few – then the environmental phenomena described, coupled with the fear of the wrath of the gods, may prompt a response in the unfortunate culture.

REFERENCES Baillie, M.G.L. (1988) Irish oaks record volcanic dust veils drama. Irish Archaeology 2(2): 71–4. Baillie, M.G.L. (1989a) Do Irish bog oaks date the Shang Dynasty? Current Archaeology 117: 310–13. Baillie, M.G.L. (1989b) Hekla 3: how big was it? Endeavour. New Series 13(2): 78–81. Baillie, M.G.L. and Munro, M.A.R. (1988) Irish tree rings, Santorini and volcanic dust veils. Nature 322: 344–6. Ball, T. (1992) Climatic change, droughts and their social impact: Central Canada, 1811– 20, a classic example. In C.R. Harington (ed.) The Year Without a Summer? World Climate in 1816, 185–95. Ottawa: Canadian Museum of Nature. Barker, T. (n.d.) Meteorological Register 1777–1789 made at Lyndon Hall, Rutland, unpublished manuscript, National Meteorological Library and Archive, Bracknell, UK. Blackford, J.J., Edwards, K.J., Dugmore, A.J., Cook, G.T. and Buckland, P. (1992) Hekla-4: Icelandic volcanic ash and the mid-Holocene Scots Pine decline in northern Scotland. The Holocene 2(3): 260–5.

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

103

Braun-Fahrlander, C., Ackermann-Leibrich, U., Schwartz, J., Gnehm, H.P., Rutishauser, M. and Wanner, H.U. (1992) Air pollution and respiratory problems in preschool children. American Review of Respiratory Disorders 145: 45–7. Brayshay, M. and Grattan, J.P. (1999) Environmental and social responses in Europe to the 1783 eruption of the Laki Fissure volcano in Iceland: a consideration of contemporary documentary evidence. Geological Society Special Publication 161: 173–88. Brugmans, S.J. (1784) Natuurkundige verhandeling over een zwavelagtigen nevel den 24 Juni 1783 in de provincie van stad en lande en naburigen landen waargenomen (A physical treatise on a sulphuric smog as observed on 24 July 1783 in the province of Groningen and neighbouring countries). Nijmegen: Isaac van Campen. Brunekreef, B., Kinney, P.L., Ware, J.H., Dockery, D., Speizer, F.E., Spengler, J.D. and Ferris, B.G. (1991) Sensitive subgroups and normal variation in pulmonary function response to air pollution episodes. Environmental Health Perspectives 90: 189–93. Burgess, C. (1989) Volcanoes, catastrophe and the global crisis of the late second millennium BC. Current Archaeology 117: 325–9. Camuffo, D. and Enzi, S. (1995) Impacts of clouds of volcanic aerosols in Italy during the last seven centuries. Natural Hazards 11: 135–61. Caput, C., Belot, Y., Auclair, D. and Decourt, N. (1978) Absorption of sulphur dioxide by pine needles leading to acute injury. Environmental Pollution 16: 3–15. Charman, D.J., Grattan, J.P., West, S. and Kelly, A. (1995) Environmental response to tephra deposition in the Strath of Kildonan, northern Scotland. Journal of Archaeological Science 22(6): 799–809. Clausen, H.B. and Hammer, C.U. (1988) The Laki and Tambora eruptions as revealed in Greenland ice cores from 11 locations. Annals of Glaciology 10: 16–22. Cullum, J. (1786) On a remarkable frost on the 23rd of June, 1783. The Royal Society of London Philosophical Transactions 15: 604. Cunningham, P. (1906) The Letters of Horace Walpole, Vol. VIII. London: Richard Bentley. Demarée, G., Ogilvie, A.E.J. and Zhang, D. (1998) Further documentary evidence of northern hemispheric coverage of the Great Dry Fog of 1783. Climatic Change 39: 727–30. Devine, J.D., Sigurdsson, H., Davis, A.N. and Self, S. (1984) Estimates of sulphur and chlorine yield to the atmosphere from volcanic eruptions and potential climatic effects. Journal of Geophysical Research 89: 6309–25. Dockery, D., Schwartz, J. and Spengler, J.D. (1992) Air pollution and daily mortality: associations with particulates and acid aerosols. Environmental Research 59: 362–73. Dodgshon, R., Grattan, J.P. and Gilbertson, D.D. (2000) Endemic stress, farming communities and the influence of volcanic eruptions in the Scottish Highlands. The Geological Society of London Special Publication 171: 267–80. Durand, M. and Grattan, J.P. (1999) Extensive respiratory health impacts of volcanogenic dry fog in 1783 inferred from European documentary sources. Environmental Geochemistry and Health 21: 371–6. Durand, M. and Grattan, J.P. (2001) Volcanoes air pollution and health. The Lancet 357: 164. Fiacco, R.J., Thordarson, Th., Germani, M.S., Self, S., Palais, J., Whitlow, S. and Groutes, P. (1994) Atmospheric aerosol loading and transport due to the 1783–84 Laki Fissure eruption in Iceland, interpreted from ash particles and acidity in the Gisp2 ice core. Quaternary Research 42: 231–44. Franklin, B. (1784) Meteorological imaginations and conjectures. Memoirs of the Literary and Philosophical Society of Manchester 2: 373–7. Gilpin, W. (n.d.) An Historical Account of the Weather During Twenty Years from 1763–1785. Bodleian MS, Eng. Misc. d. 564. Grattan, J.P. (1998) The distal impact of volcanic gases and aerosols in Europe: a review of the 1783 Laki Fissure eruption and environmental vulnerability in the late twentieth century. Geological Society Engineering Geology Special Publications 15: 97–103.

104

J. GRATTAN ET AL.

Grattan, J.P. and Brayshay, M.B. (1995) An amazing and portentous summer: environmental and social responses in Britain to the 1783 eruption of an Iceland Volcano. The Geographical Journal 161: 125–34. Grattan, J.P. and Brayshay, M.B. (1999) Modelling the impact of the Vesusius/Avellino Eruption upon the Bronze Age settlement of the Palma Campania. In Albore Livadie (ed.) L’Eruzione Vesuviana Delle ‘Pornici di Avellino’ e la Facies di Palma Campania, 125– 32. CUBEC. Grattan, J.P. and Charman, D.J. (1994) Non-climatic factors and the environmental impact of volcanic volatiles: implications of the Laki Fissure eruption of AD 1783. The Holocene 4: 101–6. Grattan, J.P. and Gilbertson, D.D. (1994) Acid-loading from Icelandic tephra falling on acidified ecosystems as a key to understanding archaeological and environmental stress in northern and western Britain. Journal of Archaeological Science 21(6): 851–9. Grattan, J.P. and Gilbertson, D.D. (2000a) ‘A fire spitting volcano in our dear Germany’? Documentary evidence for a low intensity volcanic eruption of the Gleichberg in 1783. The Geological Society of London Special Publication 171: 307–15. Grattan, J.P. and Gilbertson, D.D. (2000b) Prehistoric ‘settlement crisis’, environmental changes in the British Isles and volcanic eruptions in Iceland: an exploration of plausible linkages. Volcanoes in Antiquity. Geological Society of America Special Paper 345: 33–42. Grattan, J.P. and Pyatt, F.B. (1994) Acid damage in Europe caused by the Laki Fissure eruption – an historical review. The Science of the Total Environment 151: 241–7. Grattan, J.P. and Pyatt, F.B. (1999) Volcanic eruptions, dust veils, dry fogs and the European Palaeoenvironmental record. Global and Planetary Change 21: 173–9. Grattan, J.P. and Sadler, J. (1999a) An assessment of the effectiveness of volcanic eruptions as an agent of rapid climatic change. Global and Planetary Change 21: 181–96. Grattan, J.P. and Sadler, J. (1999b) Regional warming of the lower atmosphere in the wake of volcanic eruptions: the role of the Laki Fissure eruption in the hot summer of 1783. Geological Society Special Publication 161: 161–72. Grattan, J.P., Brayshay, M. and Sadler, J.P. (1998) Modelling the impacts of past volcanic gas emissions: evidence of Europe-wide environmental impacts from gases emitted in Italian and Icelandic volcanoes in 1783. Quaternaire 9(1): 25–35. Grattan, J.P., Gilbertson, D.D. and Charman, D.J. (1999) Modelling the impact of Icelandic volcanic eruptions upon the prehistoric societies of northern and western Britain. Geological Society Special Publication 161: 109–24. Gunnlaugsson, G.A., Gudbergsson, G.M., Thorarinsson, S., Raffnson, S. and Einarsson, T. (1984) Skáftareldar 1783–1784. Rekyavik: Mal Og Menning. Hall, V.A., Pilcher, J.R. and McCormac, F.G. (1994) Icelandic volcanic ash and the midHolocene Scots pine (Pinus Sylvestris) decline in the north of Ireland: no correlation. The Holocene 4(1): 79–83. Hjartanson, Á. (1994) Environmental changes in Iceland following the great Þjorsá lava eruption 780014C years BP. Münchener Geographische Abandlungen, B12: 147–56. Jackson, E.L. (1982) The Laki eruption of 1783: impacts on population and settlement in Iceland. Geography 67(1): 42–50. King, J. and Ryskamp, C. (1981) The Letters and Prose Writings of William Cowper. Oxford: Clarendon Press. Kington, J.A. (1980) July 1783: the warmest month in the Central England temperature series. Climate Monitor 9: 69–73. Kington, J.A. (1988) The Weather Patterns for the 1780s Over Europe. Cambridge: Cambridge University Press. Koenig, J., Pierson, W.E. and Frank, R. (1979) Acute effects of inhaled sulphur dioxide plus sodium chloride aerosol on pulmonary function in asthmatic adolescents. Environmental Research 22: 145.

SOCIAL–ENVIRONMENTAL RESPONSES TO VOLCANIC POLLUTION

105

Lamb, H.H. (1970) Volcanic dust in the atmosphere; with a chronology and assessment of its meteorological significance. The Royal Society of London Philosophical Transactions 266(A): 425–533. Lang, D.S., Herzfeld, D. and Krupa, S.V. (1980) Reponses of plants to submicron acid aerosols. In T.Y. Toribara, M.W. Miller and P.E. Morrow (eds) Polluted Rain, 273–90. New York: Plenum. Lee, D.O. (1983) Trends in summer visibility in London and southern England 1962– 1979. Atmospheric Environment 17: 151–9. Ogilvie, A.E.J. (1986) The climate of Iceland 1701–1784. Jökull 36: 57–73. Ostro, B.D., Lipsett, M.J., Wiener, M.B. and Selner, J.C. (1991) Asthmatic response to air-borne acid aerosols. American Journal of Public Health 81: 694–702. Rabartin, R. and Rocher, P. (1993) Les Volcans et la Révolution Française. Paris: L’Association Volcanologique Européene. Sigurdsson, H. (1982) Volcanic pollution and climate: the 1783 Laki eruption. EOS 63: 601–2. Simkin, T., Siebert, L., McClelland, L., Bridge, D., Newhall, C. and Latter, J.H. (1981) Volcanoes of the World. Stroudsburg: Hutchinson Ross. Steingrímsson, J. (1998) Fires of the Earth: The Laki Eruption 1783–1784. Trans. K. Kunz. Iceland: University of Iceland Press. Stothers, R.B. (1996) The Great Dry Fog of 1783. Climatic Change 32: 79–89. Swinden, M.V. (1786) Observations sur quelques particularités météorologiques de l’année 1783. Memoires de l’Académie Royale des Sciences, Turin, Anneés 1784–1785: 113–40. Thórarinsson, S. (1979) On the damage caused by volcanic eruptions with special reference to tephra and gases. In P.D. Sheets and D.K. Grayson (eds) Volcanic Activity and Human Ecology, 125–59. New York: Academic Press. Thórarinsson, S. (1981) Greetings from Iceland: ash falls and volcanic aerosols in Scandinavia. Geografiska Annaler 63A: 109–18. Thordarson, Th. and Self, S. (1993) The Laki and Grímsvötn eruptions in 1783–85. Bulletin Volcanologique 55: 233–63. Thordarson, Th., Self, S., Oskarsson, N. and Hulsebosch, T. (1996) Sulfur, chlorine and fluorine degassing and atmospheric loading by the 1783–1784 AD Laki (Skaftár Fires) eruption in Iceland. Bulletin of Volcanology 58: 205–55. White, G. (1977) The Natural History of Selborne. London: Penguin. Wilcox, R.E. (1959) Some effects of the recent volcanic ash falls with special reference to Alaska. U.S. Geological Survey Bulletin, 1028-N: 409–76. Wisniewski, J. (1982) The potential acidity associated with dews, frosts and fogs. Water, Air, and Soil Pollution 17: 361–77. Wood, C.A. (1984) The amazing and portentous summer of 1783. EOS 65: 410–11. Wood, C.A. (1992) Climatic effects of the 1783 Laki eruption. In C.R. Harington (ed.) The Year Without a Summer? World Climate in 1816, 58–77. Ottawa: Canadian Museum of Nature. World Health Organization (1979) Environmental Health Criteria 8, Sulphur oxides and suspended particulate matter. Geneva: World Health Organization. Zielinski, G.A., Fiacco, R.J., Whitlow, S., Twickler, M.S., Germani, M.S., Endo, K. and Yasui, M. (1994) Record of volcanism since 7000 BC from the GISP2 Greenland ice core and implications for the volcano-climate system. Science 264: 948–52. Zielinski, G.A., Germani, G., Larsen, G., Baillie, M.G.L., Whitlow, S., Twickler, M.S. and Taylor, K. (1995) Evidence of the Eldgjá (Iceland) eruption in the Gisp2 Greenland ice core: relationship to eruption processes and climatic conditions in the tenth century. The Holocene 5: 129–40.

106

J. GRATTAN ET AL.

NEWSPAPERS CONSULTED AT THE BRITISH LIBRARY The Aberdeen Journal, 18 August 1783, 2 The Bristol Journal, 19 July 1783, 1 The Bristol Journal, 19 July 1783, 2 The Cambridge Chronicle and Journal, 26 June 1783, 4 The Gentleman’s Magazine, September 1783, 803 The Gentleman’s Magazine, October 1783, 881 The Ipswich Journal, 12 July 1783, 4 The Ipswich Journal, 9 August 1783, 2 The Morning Herald and Daily Advertiser, 15 July 1783, 2 The Morning Herald and Daily Advertiser, 5 August 1783, 2 The Morning Herald and Daily Advertiser, 19 August 1783, 2

7

Recurring tremors: the continuing impact of the AD 79 eruption of Mt Vesuvius PENELOPE M. ALLISON

‘Many disasters have befallen the world, but few have brought posterity so much joy.’ (Goethe in Knight, 1996: 11)

INTRODUCTION Mt Vesuvius (Fig. 7.1) has had numerous eruptions, but the one that took place during the early Roman Empire is the best known. Its notoriety stems largely from its historical recording, which was the earliest written description of any volcanic eruption. The extensive material record preserved by the deposits from the eruption and first revealed to the modern western intellectual world in the early eighteenth century can also take credit for this fame. This particular eruption was a catastrophe for the many inhabitants of the Bay of Naples (Fig. 7.2) at the time, but its impact on the wider Campanian community or on Roman commerce in the first century AD was less dramatic. Rather, its influence has been felt in other ways, both real and perceived, over the many centuries since the original event. The discovery and excavation of sites like Pompeii and Herculaneum, destroyed during the eruption, have captured the imagination of modern visitors and scholars alike and have coloured their perceptions of the original socio-economic significance of this event. Consequently, this ancient eruption and the modern discovery of its resulting debris have had a fundamental impact on modern scholarship concerning the socio-economics of the wider Roman world. The archaeological material has also profoundly influenced European art, European culture and a European sense of identity since the eighteenth century. In turn, through this influence and its social and cultural associations, these remains play an important role in the cultural and the economic development of the region today, thus ensuring that the area will be an active participant in the global village of the twenty-first century AD. The purpose of this chapter is to contrast the impacts of the Vesuvian event on its contemporary world with its effects on later inhabitants and on the world at

108

P.M. ALLISON

Figure 7.1 View of Mt Vesuvius from Monte Faito to the south, looking across the Bay of Naples Photograph: P. Allison

large. I examine conditions in the periods before, during and after the eruption and continuing up to the present day. I begin with a brief outline of Vesuvius’s volcanic history and some of the reasons why the AD 79 eruption, in particular, has received so much attention in recent history. This is followed by summaries of socioeconomic conditions in Campania leading up to and during the event. These are compared to conditions in the region in the later Roman imperial period to demonstrate that there is little evidence of any reduction in the region’s productivity or in its contribution to Roman commerce as a result of this eruption. In contrast, the final sections demonstrate that the repercussions have been most dramatic for modern-day scholars and cultural tourists in their exploitation of the material. As a consequence, the effects on the productivity of the region today are greater, and seemingly more positive, than any negative impact of the original eruption.

THE ERUPTIONS OF MT VESUVIUS Since its apparent formation from Monte Somma some 17,000 years ago (Sigurdsson et al., 1985: 335), Mt Vesuvius has erupted on a number of occasions. Geological and archaeological evidence has been used to date the beginning of the fifth volcanic cycle of Mt Vesuvius at 3760 ± 70 BP (i.e. c.1880–1740 BC: Sigurdsson et al., 1985: 336; Rossi and Santacroce, 1986: 22–3). Written records

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

109

Figure 7.2 Map of Campania showing locations of Bay of Naples, Mt Vesuvius, Pompeii and Herculaneum Map adapted by J. Lovell and reproduced with permission

indicate that twelve eruptions occurred between AD 79 and AD 1631 (Rossi and Santacroce, 1986: 26–9), with further eruptions in the twentieth century preceded by up to eleven years of ongoing seismic activity (Frederiksen, 1984: 7–9). The last major event occurred in 1944, but earthquakes in 1980 were reminders that this is still an active volcano. Besides the scale of Vesuvius’s eruption in AD 79, the impact of which was reputedly felt in North Africa and Palestine (Dio Cassius, 66, 23), there are other reasons for its notoriety. First, while comments by Roman authors such as Strabo (5, 4, 8), Vitruvius (2, 6, 1–2) and Diodorus Siculus (4, 21, 5), writing in the first century BC and early first century AD, indicate a knowledge of Vesuvius’s volcanic nature, the AD 79 eruption is the earliest for which there is a written record of the actual process of the event. The celebrated Roman scientist, Gaius Plinius Secundus, better known as Pliny the Elder, was killed by this eruption and his nephew, Pliny the

110

P.M. ALLISON

Younger, was asked by his historian friend Tacitus to give an account of the events surrounding his uncle’s death (Pliny the Younger, Ep. 16, 22). This account, and that of Dio Cassius (66, 21–3) written in the late second or early third century AD, also established this eruption as the earliest for which a precise date was recorded. This date has been established as 24–25 August AD 79, although it has also been argued that the eruption may actually have occurred in November of that year (for references see Pappalardo, 1990: 209–10). Not only may the precise date for this Vesuvian eruption be inaccurate, but the scientific reliability of Pliny the Younger’s detailed description of its process has been called into question. The stylistic similarities between his description and an anonymous poem about an AD 40 eruption of Mt Etna have suggested to some scholars that Pliny’s description was more a literary exercise, borrowing from this earlier work, than an account of his actual observations (Wilsdorf, 1979: 40–1; Copony, 1987: esp. 219–20, 227; see also Frederiksen, 1984: 9–11; for discussion: Allison, 1992b: 11; Allison, in press). Given that this particular account was written by a school-boy who was busy with his books at the time, rather than by his scientist uncle, it should undoubtedly be used with considerable caution in any interpretations of the stratigraphical and material remains. However, it is generally taken as being intrinsically linked to the physical evidence (e.g. Sigurdsson et al., 1982, 1985). In any event, this description is still the earliest historical record of a volcanic eruption. It is frequently cited in archaeological, historical and volcanological literature and the concept of an archetypal ‘Plinian eruption’ is based on this account (see Blong, 1984: 3, 5, 6 passim). This eruption is also notorious because of the nature of the burial of its victims and their subsequent discovery and recovery. The area affected by the eruption had been heavily populated. The deposit resulting from it – wind-blown volcanic ash followed by pyroclastic flows forming a deposit of up to 8 m, for some 70 km to the southwest (Cerulli Irelli, 1975: 297) and mudflows forming a solid deposit of up to 20 m to the west (D’Arms, 1970: 153) – meant that much of the material remains of these populated areas, including some of the inhabitants, had been completely and irretrievably covered by the volcanic debris. With the exception of a certain amount of disturbance through looting (see e.g. Bechi, 1834: 2; Cerulli Irelli, 1975: 295; for further references and discussion: Allison, 1992b: 17–19, 37– 9; Allison, in press), which seems to have begun soon after the eruption and to have continued for the next 17 centuries, these remains maintained a remarkably high level of preservation before being revealed to the modern western intellectual world in the early to mid-eighteenth century. The investigation of the debris of human activity resulting from this eruption both contributed to and profited from the development of widespread intellectual interest in the classical world.

CAMPANIA BEFORE AD 79 Before the event in question the Campanian region (Fig. 7.2) had had a longstanding importance in the Mediterranean world in both the pre-Roman and

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

111

Roman periods. Historical and archaeological records indicate that the fertile and volcanic soils made this a much sought-after, and squabbled-over, area for agricultural settlement. Its plains, sheltered bays and relatively accessible overland river routes to the north provided important resources for Etruscan and Greek colonists, not to mention local indigenous populations and invading Samnites. A treaty between these Samnites and the Romans in the mid-fourth century, with subsequent revolts and Roman expansionist interests, led to an increasing Roman domination of the area. The main products exported, in both pre-Roman and Roman times, were wine, olive oil and corn, while sheep and pig farming and fisheries probably also produced exportable commodities. Evidence of this Campanian export trade, notably in the form of amphorae, has been recorded in North Africa, Western Europe and the Greek East, dating from the mid-second century BC onwards (Frederiksen, 1984: 300). Also, remains of substantial rural villas and farm sites in the Campanian plains bear witness to the extensive and well-organised agricultural production of this region from at least the early second century (Arthur, 1991: 63; see also Kockel, 1986: 519–66, esp. fig. 23; Moormann, in press a). The Bay of Naples also provided useful and strategic harbours for shipping these products to other parts of Italy and the Roman world. Another reason for the significance of this region to Rome, and our raised consciousness of this, was that from the early second century BC many politically influential Romans acquired land in the area (D’Arms, 1970: 1–17). Many used these properties to seek ‘asylum from the political jostling of Rome’ (Arthur, 1991: 66), particularly during the last years of the Roman Republic (D’Arms, 1970: 61–70). Not only the aristocracy but also the imperial family owned large parts of Campania. For example, Augustus had purchased the entire island of Capri. Tiberius had at least 12 villas there (D’Arms, 1970: 73), to which he fled when life as a Roman emperor became unbearable. Some Roman aristocrats may have owned villas in inland areas which were part of productive agricultural estates. However, most of their properties consisted of luxury villas along the coast. Campania had once been the ‘land of villages’ (Frederiksen, 1984: 31). By the first century AD, the Bay of Naples consisted of an unbroken line of these villas, such that it gave the appearance of a single town (Strabo, 5, 4, 8). As well as acting as retreats for the privileged, ‘who expended their energies in Rome’ (D’Arms, 1970: 160), these seaside properties may have had commercially productive fishponds and oyster beds, and their strategic military positions may also have been used to advantage. Written sources provide evidence that some of these Roman élites were actively involved in local commerce (D’Arms, 1980; Frederiksen, 1984: 305). Many of the dwellings unearthed in Herculaneum are assumed to have been such luxury abodes of wealthy members of the Roman senatorial classes. However, the remains of substantial town houses with colonnaded Hellenising gardens and elaborately decorated in a style dating to the second and first centuries BC have also been discovered in Pompeii. These indicate the also considerable disposable wealth and lavish tastes of some of its citizens, who did not belong to this particular élite

112

P.M. ALLISON

social group (see Zevi, 1995: 23–4). These Pompeians profited directly or indirectly from the expanding markets and connections, not to mention the resulting influx of foreign goods and a heightened awareness of a need for them, which Roman interests brought to the region (Mouritsen, 1998: 92). However, this does not mean that they owed their affluence and urbanisation wholly to the Roman influence and connections. Rather, as indicated in the public and private buildings in Pompeii, there was a ‘strong hellenistic influence’ (Mouritsen, 1998: 63, 65; see also Arthur, 1986), which could equally have emanated from pre-Roman connections in this region. While some Campanian towns became Roman colonies from the late fourth and third centuries BC, it is perhaps significant that the town of Pompeii did not receive this status until 80 BC. This particular Campanian town may not have been of specific political, social or economic importance to the Romans (Zanker, 1998: 4). Equally, its inhabitants were probably not particularly hospitable to a Roman presence. Such sentiments were undoubtedly widespread in the Italian peninsula leading up to the Social War (see Mouritsen, 1998: esp. 28–9, 77). In summary, Campania had been a productive agricultural and maritime region, as well as a retreat for Roman aristocracy, before AD 79. As such it had played a very active role in Roman commerce and in Roman social life. At the same time, however, many of the inhabitants are likely to have maintained wealth and cultural identities founded on pre-Roman traditions and networks.

THE ANCIENT IMPACT OF THE AD 79 ERUPTION The AD 79 eruption of Mt Vesuvius had a devastating effect on the populations of Pompeii, Herculaneum, Stabiae and the rural area to the west and southwest of the mountain. However, this effect did not begin with the final event. Seismic activity, the more extensive of which was probably an earthquake recorded in AD 62, seems to have been plaguing this area for several years. Archaeological evidence for ongoing damage and repair to, and probably abandonment of, buildings in Pompeii can be shown to be related to more than one upheaval before the AD 79 eruption (see Allison, 1992a, 1992b: 8–9, 86–97; Fröhlich and Jacobelli, 1995, for discussion and references), suggesting that the Pompeians and their near neighbours had been subjected to pre-eruption earthquakes, perhaps for sometime. The experiences and the survival strategies of individuals were undoubtedly considerably varied. Some proprietors seem to have stayed and continued to patch their buildings (Fig. 7.3), while others may have retreated to properties in less threatened areas, leaving slaves or freedmen behind to guard and maintain their interests. Obviously, this latter situation applies only to the wealthier inhabitants who had other properties where they could retreat. Some, who had little property to protect, may also have decided that early abandonment of the area was the best option. Some of the abandoned properties may even have been reoccupied by others. Thus, the evidence suggests that the effect of an impending eruption on the local population may have been as significant as the eruption itself. In any event,

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

Figure 7.3

113

Piles of gypsum in garden of the Casa del Sancello Iliaca in Pompeii

Note: This material is likely to have been used for plastering the walls, suggesting that alteration or restoration work was going on in this house, or that it was being used as a workshop for restoration work Photograph:

J. Agee

the landscape of central Campania, that is much of the southern part of the Bay of Naples, was first disrupted, possibly by earthquakes, and then devastated beyond recognition by the subsequent volcanic eruption. It is estimated that, as a result of this eruption, the coastline moved up to a kilometre further into the Bay and that the courses of once navigable rivers were substantially altered (see Ward-Perkins and Claridge, 1980: 11; Descoeudres et al., 1994: 3–4 and esp. fig. 3). It was probably extremely difficult for any survivors to have located or identified any of their own property. After the eruption, according to both Dio Cassius (66, 24) and Suetonius (De Vita Caesarum, Titus 8), the emperor Titus himself went to Campania, and sent two ex-consuls. Their mission was to supervise its restoration and to grant the survivors the land of those who had perished in the catastrophe and left no heirs. It is often assumed that these textual references pertained to the occupants of the buried and archaeologically excavated cities of Pompeii and Herculaneum (e.g. Cerulli Irelli, 1975: 293) but this is by no means established. While Flavian emperors (AD 69–96) were known for their concern for social welfare, it is likely that imperial and senatorial interest from Rome concerned villas of their own kind as much as, if not more so than, the property of these local townspeople. Apart from the two references of Dio Cassius and Suetonius, there are no further reports about the immediate aftermath of the eruption. Pliny the Younger ended his account with his own escape. There was no report of what the emperor or the ex-consuls found when they got there, or what sort of restoration they supervised.

114

P.M. ALLISON

The archaeological evidence in Pompeii suggests that no post-eruption restoration was initiated. Rather, it points to considerable looting having followed the eruption, although not necessarily immediately (see e.g. Parslow, 1995: 113). Similarly, substantial villa complexes, such as that at the comparatively recently excavated Villa A at Oplontis (De Franciscis, 1975), which reputedly belonged to the Poppaeae family, the family of the emperor Nero’s wife, provide no evidence of any attempts to restore them to their former luxurious existence. Agricultural villa complexes, which have been excavated in the area, also appear to give no indication of attempts to revitalise them after the eruption. Lack of evidence for reconstruction and occupation after AD 79 is perhaps partly a result of lack of attention, particularly during earlier excavations. After all, much richer pickings were to be made from the AD 79 deposit. Many excavators of Pompeii may have used techniques which limited their chances of finding evidence of the ‘renascita di Pompei’. Nevertheless, they had long considered the possibility of such (see Cerulli Irelli, 1975: 291). The lack of evidence could therefore conceivably be attributed to a disparity between the theory and practice of the archaeologists rather than a complete lack of concern for the existence of substantial post-AD 79 occupancy. It is likely that the initially sterile volcanic soil would have been restored to an agriculturally productive state within a few years (Cerulli Irelli, 1975: 293; Widemann, 1990: 223–4). Certainly there is evidence that the area affected by this eruption showed at least some occupation in later Roman periods (Cerulli Irelli, 1975: esp. 295–8). However, no traces of any substantial urban centres have been located to replace these buried cities. Thus both archaeological and historical evidence indicates that Pompeii and Herculaneum, and many other coastal, urban and rural properties, were largely, if not totally, abandoned after the AD 79 eruption. The Sarno River, with its port at Pompeii, seems no longer to have been useful for mercantile activity (see Widemann, 1990: 230). Historical reports indicate that Stabiae recovered at least temporarily (D’Arms, 1970: 154). However, in general, the zones which had been buried by this Vesuvian eruption are largely missing from later historical reports of Campania, and indeed from rigorous archaeological investigation. Scholars consider the area to have been largely unoccupied until the Middle Ages (Widemann, 1990: 230).

POST-AD 79 CAMPANIA Despite these catastrophic effects on the population in the immediate vicinity of Mt Vesuvius, the impact of the AD 79 eruption on Campanian social and commercial activities and on the wider Roman world in the first century AD and later was much less dramatic. Certainly Campania’s heyday as a luxury retreat for the Roman aristocracy seems to have come to an end about this time. However, this demise was more the outcome of changing social and political conditions in Rome than anything related to the AD 79 Vesuvian eruption. First, the

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

115

extravagant tastes of the Julio-Claudian family (i.e. Augustus, Tiberius and their successors) were not shared by the Flavian emperors, although the latter had inherited the properties of their predecessors and later emperors continued to own and visit Campanian properties at least into the fourth century (D’Arms, 1970: 118–64). Likewise, leading Roman families continued to acquire and use land in the region. Unfortunately, this activity is less well documented than that in the late first and early second centuries AD, of which writers like Suetonius kept us informed. Nevertheless, the evidence indicates that the eruption had little longterm effect on the Campanian interests of the Roman élite, beyond the loss of family members, friends and colleagues, such as Pliny the Elder. Archaeological evidence, particularly that of amphorae at Ostia and by association that of the wine trade to Rome, has been used to indicate that Italian markets which had once received Campanian products were importing from Spain and Southern Gaul in the late first century AD in increasing quantities (Widemann, 1990). However, Widemann (1986: 108) has argued not only that the growing population of Rome was a factor in this increase, but also that changes in trading patterns during this period resulted from a general development within the Roman economy of the specialisation of particular provinces in certain export products (Widemann, 1990: 227). This highlights a potential consumer preference for French over Italian wine and emphasises the need to consider the consumers as much as the producers in assessing the significance of trade patterns. In any event, the disappearance of amphora types Dressel 2–4 from many Roman provincial archaeological assemblages, which has been taken as evidence of a decline in general Italian wine exports, did not occur until the end of the second century AD (Tchernia, 1981: esp. 306–10), and is therefore unrelated to this Vesuvian eruption. The town of Capua, near the overland access routes to Rome above the vast northern Campanian plain, has received less attention in the modern literature than Pompeii. However, it was a much more important Roman centre, politically, economically and culturally, with its prestige spanning a long period of Roman history from the third century BC until the fourth century AD (Frederiksen, 1984: 285–93). While it may have suffered a decline in the first and second centuries AD, this was related rather to the emerging importance of the coastal town of Puteoli than to any natural forces. From its foundation as a small maritime colony early in the second century BC, the port of Puteoli, located in one of the few natural harbours on Italy’s west coast with access to the agricultural areas of the hinterland, grew dramatically to become the economic centre of Campania (Frederiksen, 1984: 319–37). More importantly, it became Rome’s principal port, through which passed most commodities from around the Mediterranean, destined for the Italian peninsula and vice versa (see e.g. Johannowsky, 1976; Camodeca, 1996). Archaeological evidence indicates that Puteoli and its environs resembled Rome in size, and that it had been embellished with considerable architectural and engineering works during the first centuries of the Roman Empire. Historical records bear witness to the extent of Puteoli’s trade connections and the degree of its autonomy from Rome. It reigned

116

P.M. ALLISON

supreme as a vital port town in an agriculturally productive region at least until the third century AD, when the port of Ostia, situated at the mouth of the Tiber, rose to prominence as a harbour for Rome (see D’Arms, 1974). The eruption of Mt Vesuvius seems to have had little bearing on the supply of Campanian products and hence the affluence and importance of these Campanian cities and the entrepôt which they provided for these products to reach the wider Roman market. Pliny the Elder’s (Naturalis Historia III, 60) pre-AD 79 descriptions of ‘the lucky land of Campania’ with its ‘vine-covered hills whose liquid produce is famous in every land and ennobles tipsiness’ and of the competition of the ‘divine patrons of wine and corn’ for this region compares well with that of Polybius (Historiae III, 91), which praises the plains in the neighbourhood of Capua, in northern Campania, for their fertility, beauty and accessibility to the sea, the local harbours for their trade to all parts of the known world and the cities of Campania for being the finest and most impressive in Italy. Such descriptions highlight that other parts of Campania had long been prosperous and continued to be so after Mt Vesuvius had erupted and devastated some areas. In general, there is no historical or archaeological evidence for a break in Campania’s agricultural productivity or in the importance of the exportation of its products to the wider Roman world, which can be attributed to the destruction of one of its productive areas. According to Frederiksen (1984: 43), by the second century AD, the Bay of Naples had ‘developed a density of population and an intensity of land use which had on this scale few parallels in the ancient world’. The insignificance of this eruption to the Roman world meant that it and the devastated cities became only a vague memory in folklore.

IMPACT ON THE MODERN WORLD In contrast to its impacts on the ancient world, the AD 79 eruption of Mt Vesuvius has had much more wide-ranging consequences for the modern world, especially for European-based artistic and intellectual life of the eigtheenth, nineteenth and even twentieth centuries. Ever since the Duc d’Elbeufs’s early eighteenth-century discovery that high-quality marble and bronze statuary was to be had by simply drilling wells into the volcanic debris to the south of Naples (Parslow, 1995: 22– 3), the region has become a vital pilgrimage for classicists, artists, historians, politicians, royalty, social aspirants, romantics, or just travellers. While many of the first visitors were taken to Herculaneum, the main destination soon became the more accessible and more extensively excavated Pompeii. Goethe was there; Winckelmann was there; and so were Mark Twain and Freud. As British Ambassador to Naples (1764–1800), Sir William Hamilton took a keen and scholarly interest in the behaviour of Mt Vesuvius, in the excavations at Pompeii and Herculaneum, and in the works of art that were coming to light. He introduced many foreign visitors to Naples to the wonders that Vesuvius was continuing to produce through the excavations of its debris and he had a huge impact on the dissemination of knowledge about them throughout the western world (Jenkins and Sloan, 1996).

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

117

Piranesi, Hogarth, William Gell and countless other artists were inspired to draw and paint these ruins, and sometimes the freshly excavated finds. Publications of these images and numerous guidebooks (e.g. Gell and Grandy, 1821; Breton, 1855) further captured the imagination of their extensive audience and stimulated interest in the remains from the Vesuvian eruption. The wall-paintings in the houses of Herculaneum and Pompeii were arduously copied and used by such style-setters as Robert Adams (e.g. Trevelyan, 1976: fig. 31) to decorate the grand houses and palaces of Europe. Heinrich Schliemann’s whole house in Athens was decorated with imitation Roman paintings taken from the patternbook produced by the German artist William Zahn, which had been copied from wall-paintings in Pompeii and Herculaneum. A whole genre of panel paintings, set in Pompeii-like domestic interiors, was also inspired by such paintings (see e.g. Winkes, 1993). The bronze and silver vessels and bronze furniture fittings which were unearthed at these sites were similarly copied and reproduced. Also, poems and novels have been and continue to be written about Pompeii and Mt Vesuvius (e.g. Sontag, 1992). One of the best-selling novels of all time – Bulwer Lytton’s The Last Days of Pompeii (1834) – from which numerous films have been produced, was based on the devastation of Pompeii. The fame which this area has acquired means that even in the closing years of the twentieth century, many who express, or wish to be seen to express, an interest in European culture feel duty bound to visit the remains of Pompeii. However, the nature of this pilgrimage is multi-faceted. Many tourists are inspired to visit this place, not necessarily because it provides the material substance to develop their knowledge of life during the Roman period, but because they are following a 250-year-old tradition of cultural tourism. It is often the contribution to European culture and sense of European identity, which Pompeii has grown to symbolise, that lures many visitors, rather than the actual role that this town plays in our understanding of the Roman past. Similarly, there is growing academic interest among otherwise Roman scholars in exploring how the Vesuvian detritus and its imagery has contributed to more recent history (e.g. Ridley, 1983; Brilliant, 1993; de Vos, 1993; Parslow, 1995; Jenkins and Sloan, 1996; Wyke, 1997: 147–82; Moormann, 2001). In other words, ‘Pompeiana’ and its neoclassical progeny in the modern world, such as Bulwer Lytton’s novel, have acted to compound an enduring, and perhaps escalating, interest in the ruins that have resulted from Vesuvius’s eruption, rather than in the original ancient city itself.

IMPACT ON ROMAN RESEARCH It is this long and widespread history of both popular and scholarly interest in the Vesuvian region, and not its actual role in the socio-economics of the ancient Roman world, which is responsible for this region’s high profile in modern studies of Roman history and archaeology. The major importance of the eruption of Mt Vesuvius to scholarship is that it has produced a wealth of material culture from a past society, the likes of which have not been seen before or since, and

118

P.M. ALLISON

which also have their own long history of investigation. Thus the investigations of these remains play a major part in the development of an archaeology of historical periods and provide a useful case study in which the relationships between historical sources and archaeological data can be explored at length. The excavations and the study of the towns and villas destroyed by Mt Vesuvius cover some 250 years of changing approaches to archaeology. Unlike other historical archaeologies of more modern periods, Roman archaeology, and particularly the role of the Vesuvian region in it, has its foundation in eighteenthcentury scholarship (see Trigger, 1989: esp. 35–72), when archaeology was largely an antiquarian endeavour. Roman archaeology has frequently been driven by inquiries generated by the methodological and theoretical frameworks of the more developed and dominant disciplines of history and art history. As a consequence, the towns destroyed by Vesuvius have been used as quarries for works of art or to provide the pictures for the ancient written sources. The material remains are used to illustrate and elucidate points and issues in ancient texts (e.g. Clarke, 1991: esp. 2–12; George, 1997). The written sources often ‘set the context’ (Foss, 1997: 197) for the investigation of the material remains from this region (see also Wallace-Hadrill, 1994: 3–8). Seldom is the relationship between the two critically investigated (e.g. Leach, 1997). Because of a pre-eminent concern for the Roman élite in most written sources about Campania, the efforts of archaeological investigation have often been concentrated on the elaboration of the lives of such individuals. For example, much effort has been expended in linking houses and other material remains to known Roman individuals or their families, such as Villa A at Oplontis to the Poppaea family, the Villa at Boscoreale to Agrippa Postumus (D’Arms, 1970: 231–2; von Blanckenhagen and Alexander, 1990: 2–3) or the Casa di Julius Polybius in Pompeii to an imperial freedman (De Franciscis, 1988: 20). These material remains are seen to provide many of the answers, not only for the lives of Romans in Campania but also for all Romans. The towns and villas destroyed by Vesuvius have become the virtual reality of daily life in the whole Roman world. For example, the evidence for wine or oil production and storage equipment at the Villa Pisanella at Boscoreale has been used by Peacock and Williams (1986: 32 and fig. 10) to demonstrate how food was produced in the Roman world. The material remains at Pompeii, Herculaneum and the many other excavated complexes in the region, such as these so-called ‘villae rusticae’, can certainly give us greater insight into the functioning of a Roman world. However, the use of the material remains from this particular region to present a generalised picture of Roman practices risks creating a sense of uniformity throughout the Roman world, which is not necessarily justified or justifiable. Certainly scholars have recently become more aware that the practice of using material culture from any particular Roman period site to constitute ‘an authentic body of Roman material’ (Barrett, 1997: 51), as opposed to non-Roman, is naïve. The long and complex pre-Roman history of the Campanian region and the comparative lack of importance of towns like Pompeii to ancient authors should warn us against using these material remains to solve questions which are

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

119

generated through unrelated texts or which lack regional specificity. However, the habit of Roman archaeologists and historians of turning to this Campanian material to elucidate their findings in ancient texts or in often far-reaching parts of the Roman world or to ‘provide an important reference point’ (Laurence and Wallace-Hadrill, 1997: 6) is often too deep seated to be easily broken or even critiqued. So the influence of the Vesuvian eruption on research into the Roman world continues to be more extensive than its impact on the ancient Roman world itself. However, given the wealth of material remains that it has provided to modern scholars, I am not proposing that it should be otherwise. Rather, I believe that the specific geographical and cultural contexts of these remains and the complexity of their relationship to the written sources should be acknowledged.

CURRENT SOCIO-ECONOMIC IMPACTS Those who have perhaps been even more seriously affected by the AD 79 eruption of Mt Vesuvius are generations of modern-day Pompeians and Italian industry. In the mid-eighteenth century, when the Bourbon kings of Naples turned their attentions from well-digging in Herculaneum to the easier task of removing volcanic ash in the area to the south of Mt Vesuvius, there was no contemporary town or urban centre in the immediate vicinity (Parslow, 1995: 44). Rather, the modern town of Pompeii has grown up as a result of the excavations of the ancient one. It has developed to meet the demands both for accommodation by the workforce for the excavations and its management and for the sustenance and souvenirs for the visitors. Many of the current population of modern Pompeii can count among their family generations of members who are or have been employed in the ‘scavi’ (excavations). In the late nineteenth century this growing town’s lack of religious foundation was a point of concern for the Catholic Church, and this was alleviated by the construction of one of the largest basilicas in the region. This complex, with its accompanying orphanage and hospice, ‘il Sanctuario’, brings more visitors to the modern town than do the excavations. Visitors to the ancient site generally do not stay overnight here but come in package tours from Rome, Naples or Sorrento. Just as the cultural tourists, two-thirds of whom are foreigners, do not generally visit the Sanctuary (Soprintendenza di Pompeii, 1999: 14), many of the religious pilgrims often do not visit the ‘scavi’, except perhaps on one of those Sundays when the entry fee is waived for Italian citizens. Thus the growth and prosperity of modern Pompeii is based mainly on the Sanctuary and its visitors, employment in the excavations, and on those cultural tourists who prefer to design their own itinerary. Nevertheless, all this traffic is a ramification of the Vesuvian eruption. Indeed, the AD 79 eruption of Mt Vesuvius has greatly enhanced the socioeconomic activities and international interactions of this particular part of Campania. For over two centuries visitors to the ancient site and the Sanctuary have been eating in local restaurants, staying in local hotels, and buying souvenirs.

120

P.M. ALLISON

Carts outside the gates of the excavations and the basilica, which belong to members of local families with their own memories, have for generations been selling imitation Roman statuary and Greek vases, jewellery reputedly made from local coral or Vesuvian lava, religious paraphernalia, guidebooks, and postcards. The impact of this influx of both national and international visitors has been enormous. The success of this long-term trade can perhaps be witnessed in the continued sale of 1950s postcards, not as curios but as the mementos of twenty-first century visits. In the last decade the profitability of these enterprises to the local community has also been apparent in the encroachment of businesses from outside the region. Many souvenirs are now also sold by non-locals in chrome and glass jewellers’ shops in the main street of modern Pompeii. Armani and Benetton shops have sprung up, as has the ubiquitous McDonalds. The small town of Pompeii, with a foundation in archaeological labour, is now part of the ‘global village’. Ten years ago even the national language was difficult for many of these Neopolitan speakers, but today the children of Pompeii speak English with relish. The modern economic impact of the Vesuvian eruption is much wider than that on the local town. The lure of Pompeii as an icon of European civilisation is an important element in Italy’s status as a major international tourist destination. Each day each of the several thousand visitors pays approximately $US10 to gain entry to the excavations (Fig. 7.4). The special-purpose Circumvesuviana railway

Figure 7.4 Tourists in the Via dell’Abondanza in Pompeii Note: These are a sample of some of the many thousands of tourists, usually foreign, whose visits constitute a volume of traffic such as the city probably never experienced in ancient times and which make a substantial financial contribution to the local and wider Italian economy Photograph: P. Allison

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

121

system brings many of them effortlessly from Naples or Sorrento right to the gates of Pompeii and the entrance turnstiles. Besides the Vatican and its museums, this site is one of the jewels of one of Italy’s principal commercial enterprises – cultural tourism. Since 1874 the funds raised from these entrance fees have been making a substantial contribution to the coffers of the Ministero di Beni Culturali. The ongoing struggle for the Soprintendenza archeologica di Pompeii to have these funds channelled back into the site’s preservation and upkeep was finally won in 1998 when the Soprintendenza was granted financial autonomy (Soprintendenza archeologica di Pompeii, 1999: 2). In the meantime the eruption of Mt Vesuvius has undoubtedly had an important impact on the maintenance of Italy’s wider cultural property and, therefore, on the economy of modern Italy. Its magnetic attraction to foreign visitors and its inclusion in many travel itineraries still contribute to the flow of foreign currency into the country.

CONCLUSIONS As a natural disaster causing cultural change in the Roman world, the AD 79 eruption of Mt Vesuvius can be considered minor. In contrast, its more important impacts have been on modern society. The extensive and devastated material remains resulting from this disaster have had an overwhelming impact on popular imagination and have made a substantial contribution to the body of material remains available for the research of the Roman world. This seems to have coloured the perspectives of classical scholars and the wider public, if only implicitly, of the significance of this area to the wider Roman world. Equally, the frustrating lack of comparable archaeological evidence and of research efforts for this region after AD 79 seems to have created an impression that Romans generally also suffered through the removal, even temporarily, of this once profitable area. But this is not the case. Those who have been most affected by the AD 79 eruption are the dead Pompeians, modern cultural tourists, and both the local and wider Italian economy. A major factor in the modern impact of this eruption is a sense of European cultural identity, which the substantial and accessible debris of this catastrophe has inspired. The disciplines of history, volcanology and archaeology in general have also been much affected. The remains from this particular Vesuvian eruption contribute an enormous database to current scholarship, but the interpretations of this database are also heavily imbued with past emphases and approaches. Thus critical analyses of these interpretations – examination of the traditions and of the interplay of methods and objectives of the various disciplines concerned – are needed for more informed understandings of the eruption process, of its social and economic impact and of the place of its debris in the wider Roman world. The AD 79 eruption of Mt Vesuvius epitomises not only the impact that a natural disaster can have on its immediate victims, but also the changing impact – even over two millennia – that it can have on a wide range of people located over the entire world, as well as on a region’s economic fortunes.

122

P.M. ALLISON

ACKNOWLEDGEMENTS I am grateful to Robin Torrence and John Grattan for inviting me to take part in the stimulating session at WAC4 in Cape Town. I also thank Robin, Eric Moormann and Estelle Lazer for comments on earlier drafts. I am especially grateful to Eric Moormann for providing me with two unpublished papers. Further thanks are also due to Joyce Agee and to Jaimie Lovell for assistance with the illustrations. Much of the final section includes information from personal observations and discussions with local residents during the last 15 years of researching in ancient Pompeii and living in modern Pompeii, generally for several months at a time. I wish to acknowledge all the modern Pompeians who have accommodated me in their town.

REFERENCES Allison, P.M. (1992a) Artefact assemblages: not the Pompeii Premise. In E. Herring, R. Whitehouse and J. Wilkins (eds) Papers of the Fourth Conference of Italian Archaeology, 49–56. London: Accordia Research Centre. Allison, P.M. (1992b) The distribution of Pompeian house contents and its significance. Ph.D. thesis, University of Sydney. Ann Arbor: University Microfilms no. 9400463. Allison, P.M. (in press) Pompeian Households: An Analysis of the Material Culture. Los Angeles: Cotsen Institute of Archaeology, UCLA. Arthur, P. (1986) Problems of the urbanisation of Pompeii: excavations 1980–1981. The Antiquaries Journal 66: 29–44. Arthur, P. (1991) Romans in Northern Campania. London: British School at Rome. Barrett, J. (1997) Romanization: a critical comment. In D.J. Mattingly (ed.) Dialogues in Roman Imperialism: Power, Discourse and Discrepant Experience in the Roman Empire, 51– 64. Portsmouth: Journal of Roman Archaeology, Supplementary Series no. 23. Bechi, G. (1834) Relazione degli scavi di Pompei. Real Museo Borbonico X: 1–7. Blong, R.J. (1984) Volcanic Hazards: A Source Book on the Effects of Eruptions. Sydney: Academic Press. Breton, E. (1855) Pompeia décrite et dessinée. Paris: Gide et J. Baudry. Brilliant, R. (1993) Herculaneum: archaeological, art historical, and cultural properties. In L. Franchi dell’Orto (ed.) Ercolano 1738–1988: 250 anni di ricerca archeologica, 117– 26. Rome: L’Erma di Bretschneider. Camodeca, G. (1996) L’élite muncipale di Puteoli fra la tarda Republica e Nerone. In M. Cébeillac-Gervasoni (ed.) Les Élites Muncipales de l’Italie Péninsulaire des Gracques à Néron, 91–110. Naples and Rome: Centre Jean Bérard, École française de Rome. Cerulli Irelli, G. (1975) Intorno al problema della rinascita di Pompei. In B. Andreae and H. Kyrieleis (eds) Neue Forschungen in Pompeji, 291–8. Recklinghausen: Aurel Bongers. Clarke, J.R. (1991) The Houses of Roman Italy, 100 BC–AD 250: Ritual, Space and Decoration. Los Angeles: University of California Press. Copony, R. (1987) Fortes Fortuna iuvat. Fiktion und Realität im Vesuvbrief des jüngeren Plinius VI, 16. Grazer Breitrage 14: 215–28. D’Arms, J.H. (1970) Romans on the Bay of Naples: A Social and Cultural Study of the Villas and their Owners from 150 BC to AD 400. Cambridge, Massachusetts: Harvard University Press. D’Arms, J.H. (1974) Puteoli in the second century of the Roman Empire: a social and economic study. Journal of Roman Studies 64: 106–24.

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

123

D’Arms, J.H. (1980) Senator’s involvement in commerce in the Late Republic: some Ciceronian evidence. In J.D. D’Arms and E.C. Kopff (eds) The Seaborne Commerce of Ancient Rome: Studies in Archaeology and History, 77–89. Rome: Memoirs of the American Academy in Rome 36. De Franciscis, A. (1975) La villa romana di Oplontis. In B. Andreae and H. Kyrieleis (eds) Neue Forschungen in Pompeji, 9–38. Recklinghausen: Aurel Bongers. De Franciscis, A. (1988) La Casa di C. Iulius Polybius. Rivista di Studi Pompeiani I: 15–36. Descoeudres, J.-P. et al. (1994) Pompeii Revisited: The Life and Death of a Roman Town. Sydney: Meditarch. de Vos, M. (1993) Camillo Paderni, la tradizione antiquaria romana e i collezione inglese. In L. Franchi dell’Orto (ed.) Ercolano 1738–1988: 250 anni di ricerca archeologica, 99– 116. Rome: L’Erma di Bretschneider. Dio Cassius, Roman History (trans. by E. Cary, Loeb Library 1914–27). London: Heinemann; New York: Macmillan. Diodorus Siculus, Bibliotheca Historia (trans. C.H. Oldfather, Loeb Classical Library 1933– 67). London: William Heinemann and Cambridge, Mass.: Harvard University Press. Foss, P. (1997) Watchful Lares: Roman household organization and the rituals of cooking and dining. In R. Laurence and A. Wallace-Hadrill (eds) Domestic Space in the Roman World: Pompeii and Beyond, 196–218. Portsmouth: Journal of Roman Archaeology, Supplementary Series 22. Frederiksen, M. (1984) Campania. Rome: British School at Rome. Fröhlich, T. and Jacobelli, L. (eds) (1995) Archaölogie und Seismologie: la regione vesuviana dal 62 al 79 d.C., problemi archeologici e sismologici (Deutsches Archäologisches Institut Rom, Soprintendenza Archeologica di Pompei, Osservatorio). Munich: Biering and Brinkmann. Gell, W. and Grandy, J. (1821) Pompeiana: The Topography, Edifices and Ornaments of Pompeii (2nd edition). London: Rodwell and Martin. George, M. (1997) Repopulating the Roman house. In B. Rawson and P. Weaver (eds) The Roman Family in Italy: Status, Sentiment and Space, 299–319. Oxford: Clarendon Press. Jenkins, I. and Sloan, K. (1996) Vases and Volcanoes: Sir William Hamilton and his Collection. London: British Museum Press. Johannowsky, W. (1976) La situazione in Campania. In I.E. Zanker (ed.) Hellenismus in Mittelitalien, 267–99. Göttingen: Vanderhoeck und Ruprecht. Knight, C. (1996) William Hamilton and the ‘art of going through life’. In I. Jenkins and K. Sloan (eds) Vase and Volcanoes: Sir William Hamilton and his Collection, 11–23. London: British Museum Press. Kockel, V. (1986) Archäologische Funde und Forschungen in den Vesuvstädten II. Archäologischer Anzeiger 22: 443–581. Laurence, R. and Wallace-Hadrill, A.F. (eds) (1997) Domestic Space in the Roman World: Pompeii and Beyond. Portsmouth: Journal of Roman Archaeology, Supplementary Series 22. Leach, E. (1997) Oecus on Ibycus: investigating the vocabulary of the Roman house. In S.E. Bon and R. Jones (eds) Sequence and Space in Pompeii, 50–72. Oxbow Monograph 77. Oxford: Oxbow Books. Moormann, E. (2001) Una citta mummificata: qualche aspetto della fortuna di Pompei nella letteratura europea ed Americana. In P.G. Guzzo (ed.) Pompei Scienza e Società. 2500 Anniversario degli Scavi di Pompei. Convegno internazionale, Napoli, 25–7 novembre 1998, 9–18. Milan: Electa. Moormann, E. (in press) Villas in the surroundings of Pompeii and Herculaneum. In J.J. Dobbins and P. Foss (eds) Pompeii and the Ancient Settlements Under Vesuvius. London: Routledge. Mouritsen, H. (1998) Italian Unification: A Study in Ancient and Modern Historiography. London: Institute of Classical Studies, University of London.

124

P.M. ALLISON

Pappalardo, U. (1990) L’eruzione pliniana del Vesuvio nel 79 d.C: Ercolano. In C.A. Livadie and F. Widermann (eds) Volcanology and Archaeology, Pact 25, 197–215. Strasbourg: Council of Europe. Parslow, C.C. (1995) Rediscovering Antiquity: Karl Weber and the Excavations of Herculaeum, Pompeii and Stabiae. Cambridge: Cambridge University Press. Peacock, D.P.S. and Williams, D.F. (1986) Amphorae and the Roman Economy: An Introductory Guide. London and New York: Longman. Pliny the Elder, Naturalis Historia (trans. H. Rackham, Loeb Classical Library, 1938–63). London: Heinemann; Cambridge, Mass.: Harvard University Press. Pliny the Younger, Epistulae (trans. W. Melmoth, Loeb Classical Library, 1931). London: Heinemann; Cambridge, Mass.: Harvard University Press. Polybius, Historiae (trans. W.R. Paton, Loeb Classical Library, 1975). London: Heinemann; Cambridge, Mass.: Harvard University Press. Ridley, R.T. (1983) Dumas Père, Director of Excavations. Pompeii, Herculaneum, Stabiae 1: 259–88. Rossi, M. and Santacroce, R. (1986) L’Attività del Somma-Vesuvio precedente l’eruzione del 1631: dati stratigrafici e vulcanologici. In C.A. Livadie (ed.) Tremblement de Terre, Éruptions Volcaniques et Vies des Hommes dans la Campanie Antique, 15–34. Naples: Centre Jean Bérard. Sigurdsson, H., Cashdollar, S. and Sparks, S.R.J. (1982) The eruption of Vesuvius in AD 79: reconstruction from historical and volcanological evidence. American Journal of Archaeology 86: 39–51. Sigurdsson, H., Carey, S., Cornell, W. and Pescatore, T. (1985) The eruption of Vesuvius in AD 79. National Geographic Research 1(3): 332–87. Sontag, S. (1992) The Volcano Lover: A Romance. New York: Farrar Straus Giroux. Soprintendenza di Pompeii (1999) Pompeii – Vesuvius AD 79: A Six-Monthly Publication of Information and Culture, 2. Pompeii: Soprintendenza archeologica di Pompeii. Strabo, Geographia (trans. H.L. Jones, Loeb Classical Library, 1917). London and Cambridge, Mass.: Harvard University Press. Suetonius, De Vita Caesarum (trans. J.C. Rolfe, Loeb Classical Library, 1913). London and Cambridge, Mass.: Harvard University Press. Tchernia, A. (1981) Quelques remarques sur le commerce du vin et les amphores. In J.H. D’Arms and E.C. Kopff (eds) The Seaborne Commerce of Ancient Rome: Studies in Archaeology and History, 301–12. Rome: Memoirs of the American Academy in Rome 36. Trevelyan, R. (1976) The Shadow of Vesuvius: Pompeii AD 79. London: The Folio Society. Trigger, B.G. (1989) A History of Archaeological Thought. Cambridge: Cambridge University Press. Vitruvius, De Architectura, books I–X (trans. F. Granger, Loeb Classical Library, 1934). London and Cambridge, Mass.: Harvard University Press. von Blanckenhagen, P.H. and Alexander, C. (1990) The Augustan Villa at Boscoreale. Mainz am Rhein: Philip von Zabern. Wallace-Hadrill, A.F. (1994) Houses and Society in Pompeii and Herculaneum. Princeton, NJ: Princeton University Press. Ward-Perkins, J.B. and Claridge, A. (1980) Pompeii AD 79. Sydney: Australian Gallery Directors Council. Widemann, F. (1986) L’effet économique de l’éruption de 79. In C.A. Livadie (ed.) Tremblement de Terre, Éruptions Volcaniques et Vies des Hommes dans la Campanie Antique, 15–34. Naples: Centre Jean Bérard. Widemann, F. (1990) Implications économiques des désastres volcaniques: le court et le long terme dans le cas de Pompéi. In C.A. Livadie and F. Widemann (eds) Volcanology and Archaeology, Volcanologie et Archéologie, Pact 25, 217–31. Strasbourg: Council of Europe.

CONTINUING IMPACT OF THE ERUPTION OF MT VESUVIUS

125

Wilsdorf, H. (1979) Antike Theorien über der Vulkanismus. In K. Munze (ed.) Pompeji 79–1979, Beiträge zum Vesuvbruch und seiner Nachwirkung, 4–53. Stendal: Beiträge der Winckelmann Gesellschaft 11. Winkes, R. (1993) The influence of Herculaneum and Pompeii on American art of the eighteenth and nineteenth centuries. In L. Franchi dell’Orto (ed.) Ercolano 1738–1988: 250 anni di ricerca archeologica, 127–32. Rome: L’Erma di Bretschneider. Wyke, M. (1997) Projecting the Past: Ancient Rome, Cinema and History. London: Routledge. Zanker, P. (1998) Pompeii: Public and Private Life (trans. D.L. Schneider). London and Cambridge, Mass.: Harvard University Press. Zevi, F. (1995) Personaggi della Pompei Sillana. Papers of the British School at Rome 63: 1–24.

8

Volcanism and early Maori society in New Zealand D.J. LOWE, R.M. NEWNHAM AND J.D. M CCRAW

This chapter is dedicated to the memory of Jeanette L. Gillespie, a respected colleague and student in the Department of Earth Sciences at the University of Waikato, who died after a short illness on 4 October 2000. Jeannette, a part-time assistant lecturer in this department, was working towards her Ph.D. on the volcanic histories of Mayor Island and White Island, which feature in this chapter, by documenting the tephra record preserved in marine cores in coastal Bay of Plenty. A meticulous and talented researcher and teacher, and a warm and loyal friend, Jeanette is greatly missed.

INTRODUCTION There is much current interest in the impact of geohazards upon global cultural development. Such studies tend to rest upon broad assumptions as to scale of event and response, but before these can effectively be drawn, it is necessary to have a clear understanding of the interaction of hazards upon cultures through time. This chapter addresses this issue by exploring the interaction of volcanic activity and Maori culture in New Zealand. New Zealand is a mid-latitude, temperate, partly volcanic archipelago lying isolated in the South Pacific Ocean nearly 2,000 km eastward of its nearest neighbour, Australia. It is unique because it was the last substantial landmass to be settled by humans (Sutton, 1994a; Newnham et al., 1999a). A consequence of the exceptionally short prehistory is that the record of interactions between volcanic activity and people is brief. The earliest known European contact with the Polynesian (Maori) inhabitants of New Zealand was by Dutchman Abel Tasman in AD 1642 followed, after a 127-year gap, by Englishman James Cook and Frenchman Jean de Surville, who both arrived in AD 1769. New Zealand’s historical period is therefore restricted to barely the last two hundred years. The self-designated term ‘Maori’, literally ‘usual, ordinary’, came into use only after European settlement in the nineteenth century AD and is applied to the descendants of the Polynesians who first settled New Zealand.

VOLCANISM AND EARLY MAORI SOCIETY

127

In this chapter, the interactions between volcanism and early Maori society in New Zealand – both disastrous and beneficial – are examined. Due to the brevity of New Zealand’s written history, our study relies largely on interpretations of volcanological, palaeoenvironmental, archaeological and oral history data, in turn constrained mainly by radiocarbon and other dating techniques including palaeomagnetism, obsidian hydration and thermoluminescence. For the purposes of the review, we have defined ‘early’ Maori society to date from ‘earliest settlement’ of New Zealand (c. AD 1250–1300) until the catastrophic Tarawera eruption on 10 June, AD 1886, the traumatic effects of which have been well described. After discussing briefly the question of origin and timing of initial Polynesian settlement, we summarise in detail the record of volcanic activity in New Zealand’s North Island since c. AD 200, and record the types of volcanic hazards associated with this activity. We next use one of the main products of volcanism – tephra – to date and correlate the palaeoenvironmental impacts of early Maori people with archaeological records via tephrochronology. Tephrochronology is defined as the use of tephra layers, the unconsolidated, primary, pyroclastic products of volcanic eruptions, as chronostratigraphic marker beds to establish numerical or relative ages (e.g. Froggatt and Lowe, 1990; Lowe and Hunt, 2001; Hunt and Lowe, in press). Following this section, the likely effects and impacts that volcanism and tephra deposition had on Maori society are examined. Turning to the other side of the coin, we describe the benefits and exploitation of volcanic features and products by early Maori. Finally, we discuss aspects of Maori mythology and spirituality associated with volcanism.

POLYNESIAN SETTLEMENT OF NEW ZEALAND Current evidence indicates that the most likely ‘homeland’, or Hawaiiki, of the early Polynesian settlers of New Zealand was central Eastern Polynesia (Sutton, 1994b). The primary contenders within this tropical region include the Society Islands (encompassing Tahiti, Ra’iatea and Borabora, and others) and the islands of the Marquesan Archipelago. Additional possibilities include the Southern Cooks, Mangareva, Pitcairn and the islands of the Austral and Tuamotuan Archipelagos (Evans, 1998). Establishing the timing of settlement has been controversial. However, the most recent and reliable evidence, both from archaeological and natural sites, points consistently to initial settlement between c. AD 1250 and 1300 at the earliest (Anderson, 1991; Higham and Hogg, 1997; Newnham et al., 1998a, 1998b; Ogden et al., 1998; Higham et al., 1999; McGlone and Wilmshurst, 1999; Lowe et al., 2000, in press). ‘Settlement’ is used here to mean establishing a more-or-less permanent abode or place or way of life. An earlier, transient contact at c. AD 50–150, based on Pacific rat-bone (Rattus exulans) dates obtained from natural sites, was proposed by Holdaway (1996, 1999) on the premise that the rats, an introduced predator to New Zealand, accompanied the early Polynesian seafarers as a food source or stowaways. However, neither

128

D.J. LOWE ET AL.

archaeological nor palaeoenvironmental (chiefly pollen, charcoal and phytolith) records currently provide any definitive support for settlement at this time (Newnham et al., 1998a; McGlone and Wilmshurst, 1999; Brook, 2000; Lowe et al., 2000), and the reliability of the early rat-bone dates is strongly disputed (Anderson, 1996, 2000). Short-lived, minor disturbances in the pollen record (e.g. small increases in bracken and other seral taxa) prior to c. AD 1250, although attributed to possible early human activities by Sutton (1987, 1994b), are indistinguishable from natural background events (e.g. lightning-induced fires, impacts from volcanic eruptions, storms or droughts) that occur throughout the Holocene and earlier (e.g. McGlone, 1989; Ogden et al., 1998; Newnham et al., 1998b, 1999a; McGlone and Wilmshurst, 1999). In this chapter we therefore adopt the date of c. AD 1250–1300 as the most likely beginning of Polynesian settlement of New Zealand.

VOLCANIC ACTIVITY AND HAZARDS IN NORTH ISLAND Volcanism in New Zealand originates from its position astride an obliquely converging and active boundary between the Australian and Pacific lithospheric plates (Fig. 8.1, inset). The North Island’s main locus of current volcanic activity is the Taupo Volcanic Zone (TVZ), which is a unique type of ‘rifted arc’ dating back to c. 2 MYA. The highly productive rhyolite caldera volcanoes of the Taupo and Okataina volcanic centres occupy the central TVZ, whereas andesitic stratovolcanoes, including those of Tongariro Volcanic Centre and White Island, occur at its southwest and northeast ends (Fig. 8.1; Wilson et al., 1995). In addition to the TVZ centres, Mt Taranaki volcano (known also as Mt Egmont), the Auckland Volcanic Field, and the Tuhua Volcanic Centre (Mayor Island or Tuhua) are also regarded as active or recently active. Fig. 8.2 summarises the history of known volcanic activity and some related events in North Island since c. AD 200. In this compilation we have divided the time-scale into three main periods: a pre-human (pre-settlement) period from c. AD 200 to 1300; a prehistoric (Polynesian) period from c. AD 1300 to 1800; and a historic (European) period since c. AD 1800. The benchmarks at c. AD 200 and c. AD 1300 are defined by two widespread rhyolitic tephra marker beds, the Taupo and Kaharoa tephras, respectively. Kaharoa Tephra, dated by radiocarbon to between c. AD 1300 and 1390 by Lowe et al. (1998), has most recently been dated more precisely at AD 1314 ± 12 by provisional dendrochronological ‘wigglematch’ dating of a carbonized Phyllocladus spp. tree killed in the eruption (Hogg et al., in press). This tephra represents the critical ‘settlement layer’ datum for prehistory in the North Island, as discussed below. Seven spatially separate volcanic centres or fields have been active in the North Island since c. AD 200 (Fig. 8.2). The andesitic centres (Taranaki, Tongariro, White Island) have all erupted very frequently; the basaltic Auckland field is characterised by a single eruptive episode (Rangitoto Island); and the rhyolitic centres (Taupo, Okataina, Tuhua) have each erupted just once or twice. The

Figure 8.1 North Island volcanoes (bold) that have erupted since c. AD 200 and other features or sites mentioned in the text Note: Those volcanoes active since Polynesian settlement (c. AD 1250–1300) are marked with an asterisk (Tuhua uncertain). TVZ, Taupo Volcanic Zone; NLD, Northland; CMD, Coromandel Peninsula; BOP, Bay of Plenty; HKB, Hawke’s Bay. Inset shows plate tectonic setting of New Zealand Source:

After Lowe et al. (2000)

Maero Debris Flows Unit 1A

Eruptions

c. 200 AD

PRE-HUMAN

c. 1300 AD

Kaupokanui Fm. Unnamed tephra(s)

P. fall P. flow(s)

P. fall P. flow P. flow P. flow/fall

P. flow P. flow P. fall P. flow

Other events

~40 eruptive events 1861–1996 AD

Tf19 1995–96 AD

Crater Lake eruptions

c. 600±150 AD

c. 1450 AD FDS

} c. 1200 AD

c. 1200 –1300 AD

c. 1400 AD KAHAROA TEPHRA

TAUPO TEPHRA

Tf1 after c. 200 AD

Tf7 Tf6 Tf5

Tf8

Onetepu Fm. (lahars)

Other events

~5 – 6 eruptive events incl. lavas 1855 –1896 AD

Red Crater-Te Mari craters eruptions

Ngauruhoe Fm. - multiple tephra fall - members undefined

~60 eruptive events incl. lavas 1839 –1975 AD

Mt Ngauruhoe eruptions

TONGARIRO VOLCANIC CENTRE Mt Ruapehu Mt Tongariro

Maero Debris Flows (mainly Tufa Trig Fm. lahars), - multiple tephra fall c. 1655 AD Egmont - members Tf1–19 c. 1585AD or earlier Andesites c. 1500 AD (lavas) Hangatahua Gravels (PEF)

Late 19th C (?)

Date

P. flow/fall c. 1755–1860 AD

P. flow

Type1

MT TARANAKI – EGMONT VOLCANO

Tahurangi Fm. Tahurangi Ash Burrell Fm. Puniho Lapilli-2 Puniho Lapilli-1 Burrell Lapilli Burrell Ash Umu 2,3 Newall Fm. PREHISTORIC Waiweranui Ash Waiweranui Lapilli (Polynesian) Newall Lapilli Newall Ash FDS2 Unnamed tephra Umu 1 Unnamed tephra Unnamed tephra

c. 1800 AD

HISTORIC (European)

Period

(A)

Probably sporadic minor eruptions and near-continuous fumarolic activity

Sporadic minor steam & tephra eruptions, continuous fumarolic & solfataric activity 1826 – 2000 AD

Eruptions

WHITE ISLAND (Whakaari)

FTP MID

Rangitoto c.1400 AD (lava flows, P. fall)

Eruption

AUCKLAND VOLC. FIELD Rangitoto Island

Eruption Z Taupo Tephra/Unit-Y Subunits Y1 – Y7

Eruptions

Type1

FDS

MID

c. 180 – 230 AD

P. fall, P. flow

~30 yr after Taupo/Y Dome growth; PEF

Date

TAUPO VOLCANIC CENTRE

1904 AD

Kaharoa Tephra 1314±11 AD

Tarawera Tephra 10 June,1886 Rotomahana Mud Tarawera Scoria

Dome growth, PEF P. surge, flow, fall

P. surge, P. fall P. fall

PEF

Type

?

Dome emplacement event (lava flows) possibly after c. 200 AD ?

?

Mayor Island ?Eruption

Date

Mt Tarawera Eruptions

TUHUA VOLCANIC CENTRE

OKATAINA VOLCANIC CENTRE

Sources: Data from Gregg (1960), Neall (1972, 1979), Cole and Nairn (1975), Nairn (1979, 1991), Walker et al. (1984), Houghton and Wilson (1986), Hackett and Houghton (1986), Cole et al. (1986), Houghton et al. (1992), Wilson (1993), Donoghue et al. (1995, 1997), Stevenson et al. (1996), Cronin et al. (1997), Hodgson et al. (1997), White et al. (1997), Lecointre et al. (1998), Gillespie et al. (1999), Manville et al. (1999), Lowe and de Lange (2000), Lowe et al. (2000), Price et al. (2000) and Nairn et al. (2001) Notes: 1 Dominant style of eruption: P. flow, pyroclastic flows; P. fall, pyroclastic falls; PEF, post-eruptive flooding. Pyroclastic flows at Mt Taranaki were block-and-ash flows (V.E. Neall, pers. comm., 2000). See Table 8.1 and Sigurdsson (2000) for technical descriptions of eruption types 2 FDS, first (sustained) deforestation signal (inferred to be human-induced); FTP, human and dog footprints; MID, middens. Umu are earth ovens.

Figure 8.2 Summary of eruptions of North Island volcanic centres, and other events, since c. AD 200: (A) andesitic and basaltic volcanoes; (B) rhyolitic volcanoes. The latest Mayor Island lava dome eruption is undated but assumed to post-date c. AD 200

c. 200 AD

PRE-HUMAN

c. 1300 AD

PREHISTORIC (Polynesian)

c. 1800 AD

HISTORIC (European)

Period

(B)

132

D.J. LOWE ET AL.

latest eruption from the Okataina Volcanic Centre (Tarawera eruption) was a basaltic rather than rhyolitic event, as normally occurs at this centre. The uncertain timing of the latest (dome-building) eruption on Mayor Island/Tuhua precludes it from further discussion in this section because it may not have been active at all during the prehistoric period (Fig. 8.2). The main types or styles of eruption, and associated volcano-related events including lahar emplacement and post-eruptive flooding, for each volcanic centre are given in Fig. 8.2. Based on these data, the types of volcanic hazards likely to have been experienced or witnessed by prehistoric Maori are listed in Table 8.1. Additional information on eruption types and historical impacts and fatalities are also reported there.

TEPHROCHRONOLOGY The use of tephra layers to correlate and date archaeological remains and humaninduced environmental impacts in New Zealand has been reviewed recently by Newnham et al. (1998a) and Lowe et al. (2000). Only the key findings are therefore reported here. Supporting information is provided by related studies on pollen and phytolith stratigraphy (e.g. Sase and Hosono, 1996; Wilmshurst, 1997; Ogden et al., 1998; McGlone and Wilmshurst, 1999; Newnham et al., 1999a; Wilmshurst et al., 1999; Horrocks et al., 2000), on the timing of prehistoric predator damage to landsnails (Brook, 2000), on isotopic analyses of speleothems (Hellstrom et al., 1998), and on careful radiocarbon dating at archaeological sites (e.g. Anderson, 1991, 1995, 1996, 2000; McFadgen et al., 1994; Higham and Hogg, 1997; Higham and Lowe, 1998; Petchey and Higham, 2000), including one of the oldest known at Wairau Bar (Fig. 8.1; Higham et al., 1999). In comparison with many parts of the prehistoric world where chronologies are based largely on radiocarbon dating with its attendant imprecision and error sources, the suite of multi-sourced tephras in New Zealand provides valuable isochronous markers to help chart the course of prehistory. Tephras derived from five North Island volcanic centres or volcanoes, together with exotic sea-rafted tephra deposits (Loisels pumice), are known to be relevant to archaeological studies (Lowe et al., 2000). The locally distributed basaltic and andesitic tephras from Rangitoto Island and from Mt Taranaki overlie or contain cultural remains (including footprints and umu) and directly date human occupation to c. AD 1400 and c. AD 1450, respectively (Figs 8.2, 8.3). Distal andesitic tephras (Tufa Trig Formation members Tf5 and Tf8) from Mt Ruapehu help constrain the timing of onset of human impact signals in Hawke’s Bay to c. AD 1400. The sea-rafted Loisels pumice, although of uncertain stratigraphic reliability in places, overlies cultural remains aged c. AD 1350 on coastlines of eastern North Island (Lowe et al., 2000). It is the widespread Kaharoa Tephra, however, that provides the critical ‘settlement layer’ datum (equivalent to the landnám tephra layer in Iceland). This is because no cultural remains or artefacts are known to occur beneath it (e.g. Fig. 8.4), and because palynological evidence for earliest

VOLCANISM AND EARLY MAORI SOCIETY

Table 8.1

133

Volcanic hazards probably experienced or witnessed by prehistoric Maori

Hazard type

Volcano or centre associated with event(s) 1

Pyroclastic fall2 Pyroclastic flows3 Pyroclastic surges3 Lava flows Lava dome building Lahars4 Post-eruptive flooding4 Debris avalanches5 Volcanogenic earthquakes Lightning, forest fires Hydrothermal eruptions6 Volcanogenic tsunami7 Acidic rain/volcanic gases

Taranaki, Tongariro, White Is., Auckland, Okataina Taranaki, Tongariro, Okataina Okataina Tongariro, Auckland, Okataina Taranaki, Tongariro, Okataina, ?Tuhua Taranaki, Tongariro Taranaki, Tongariro, Okataina Taranaki, Tongariro, ?White Is. Taranaki, Tongariro, Auckland, Okataina Taranaki, Tongariro, Okataina High-temp. geothermal systems in TVZ (e.g. Ketetahi) ? ?

1

2

3

4

5

6

7

See Fig. 8.2 for examples of named eruptions or associated events. Some of the hazard types may have occurred at other volcanoes as well. Pyroclastic fall deposits are produced by the rain-out of clasts (which may be of any size or composition) through the atmosphere from an eruption jet and/or plume during an explosive eruption (Houghton et al., 2000). Pyroclastic flow deposits are emplaced as hot, fast-moving, particulate gaseous flows of high particle concentration. They include (a) block-and-ash flow deposits, comprising dense to moderately vesicular blocks in an ash matrix, and of small volume and localised distribution; and (b) ignimbrites, comprising welded to non-welded, predominantly massive and poorly sorted, pumiceous, ash-rich deposits, and of potentially large volume and widespread distribution (Freundt et al., 2000). Pyroclastic surge deposits are strongly bedded and emplaced by high-velocity, turbulent, gaseous flows (‘hurricanes’ or ‘blasts’) of low particle concentration (e.g. Valentine and Fisher, 2000). Lahar is a general (Indonesian) term used for describing gravitationally driven, rapidly moving mudflows of volcanic origin (e.g. Vallance, 2000). End-member types include (a) debris flows, which are mixtures of debris and water with a high sediment concentration (the deposits are termed diamictons); and (b) hyperconcentrated streamflows, which have a higher water concentration than in debris flows but less than in (normal) muddy streamflow. Hyperconcentrated streamflows possess fluvial characteristics yet carry very high sediment loads. Events described as post-eruptive flooding may be transitional to hyperconcentrated streamflows and include catastrophic ‘breakout’ flood events (e.g. White et al., 1997; Manville et al., 1999). In historic times, a lahar/breakout post-eruptive flood event in the Whangaehu River, Mt Ruapehu, resulted in 151 fatalities when the overnight train was derailed at the Tangiwai bridge in 1953. In 1904, a pyroclastic barrier emplaced at Lake Tarawera during the 1886 eruption was swept away, causing a flood that affected communities for 40 km downstream to the coast (White et al., 1997). Debris avalanche deposits result from sectorial collapse of the flanks of a volcano. They differ from debris flows in that they are not water-saturated: the load is entirely supported by particle–particle interactions (Ui et al., 2000). Debris avalanching/landsliding occurred at Waihi, a Maori village in the Waihi–Tokaanu geothermal field on the caldera fault-margin of southwestern Lake Taupo (Fig. 8.1) in AD 1834, AD 1846 (c.60 fatalities) and in 1910 (1 fatality). A similar event took place on the crater wall of White Is. in 1914 (11 fatalities). Geothermal fields occur throughout TVZ (Hedenquist, 1986), and some generated hydrothermal (steam) explosions during prehistoric times, e.g. at Waiotapu (Fig. 8.1), probably associated with the Kaharoa eruption. After the AD 1886 Tarawera eruption, hydrothermal explosions occurred at Waimangu (Fig. 8.6) in 1903 (4 fatalities), 1915, 1917 (2 fatalities), and in the 1970s–1980s. Numerical modelling has shown that White Is. is unlikely to have generated significant tsunamis at the coast (de Lange and Fraser, 1999). A meteorological tsunami (rissaga) resulting from atmospheric coupling during the powerful Krakatau (Indonesia) eruption of AD 1883 generated waves up to 2 m high around the New Zealand coast. The Taupo eruption of c.AD 200 probably generated a similar or larger rissaga (Lowe and de Lange, 2000). Tsunamis resulting from earthquakes, both local and overseas, have affected coastal settlements in New Zealand in historic times and it is probable that some prehistoric Maori communities were destroyed by such tsunamis and rapidly abandoned (Harada 1993a; Goff and McFadgen, 2000).

134

D.J. LOWE ET AL.

Figure 8.3 Earth oven (umu) on Mt Taranaki at c.850 m asl Note: Undisturbed coarse Burrell Lapilli (top) and fine Burrell Ash directly overlie subrounded andesitic cobbles and pebbles, many of which show heat fractures. The cooking stones lie in an excavation into underlying Waiweranui Lapilli which is dated at c. AD 1500 ± 50 (Alloway et al., 1990; Lowe et al., 2000). Total thickness of Burrell Lapilli and Ash is c.40 cm. Lens cap (left) is c.5 cm in diameter Photograph: B.V. Alloway

human-induced impact (start of the sustained rise in Pteridium) occurs stratigraphically just before its deposition (Fig. 8.5). The rise is inferred to start at c. AD 1280. This date matches the earliest reliable radiocarbon dates derived for both settlement and human impacts from archaeological and natural sites (c. AD 1250– 1300), and implies that the onset of deforestation was essentially contemporaneous with initial settlement (Lowe et al., 2000, in press). The widespread Taupo Tephra provides an isochronous benchmark well before earliest settlement, though it may coincide with the putative earlier transient contact in New Zealand, as noted in the introduction.

IMPACTS OF VOLCANISM ON EARLY MAORI SOCIETY Since c. AD 1250–1300, early Maori have witnessed probably only one catastrophic rhyolitic eruption (Kaharoa), two basaltic eruptions (Rangitoto, Tarawera), and numerous andesitic eruptions from the frequently active volcanoes of Tongariro Volcanic Centre, White Island and, to a lesser extent, Mt Taranaki (Fig. 8.2). Apart from the historic AD 1886 Tarawera eruption, the degree of impact on early

VOLCANISM AND EARLY MAORI SOCIETY

135

Figure 8.4 Archaeological section at Papamoa on the Bay of Plenty coast (Fig. 8.1) showing prehistoric Maori shell middens postdating the c. AD 1300 Kaharoa Tephra (‘Ka’). (1) ‘cultural’ topsoil; (2) shell midden; (3) mixed sand/tephric material (with scattered shells); (4) black charcoal layer; (5) Kaharoa Tephra; (6) thin paleosol; (7) Taupo Tephra; (8) aeolian dune sand (to trench base at c.60 cm depth) Note: Radiocarbon dates from the site show that it was occupied initially from c. AD 1450–1550 with a second occupation phase to c. AD 1650 at the latest (W. Gumbley, pers. comm., 2000) Photograph: W. Gumbley

Maori society from these and related volcanic events is generally not known, but some potential or likely effects, which have been hypothesised on the basis of the different hazard types and tephra thicknesses, are listed in Table 8.2. We provide more specific comments relating to the eruptions and impacts for each volcano in the following sections. Mt Taranaki Mt Taranaki volcano has erupted at least a dozen times since c. AD 1300 (Fig. 8.2). Most eruptions have been pyroclastic in nature, both fall and flow deposits being recorded, but some lavas were erupted near the summit as well (e.g. Egmont Andesites, subunit eg2: Neall, 1979). Many of the eruptions were relatively minor in scale but the Newall and Burrell groups of eruption episodes, dated at c. AD 1500 and c. AD 1600–1650, respectively, were substantial events that destroyed forest or disrupted forest canopies on parts of Mt Taranaki and beyond (Topping, 1972; McGlone et al., 1988; Lees and Neall, 1993). The Newall eruptions were directed mainly to the northwest and the Burrell Lapilli towards the east. If Maori

Figure 8.5 Pteridium (bracken) spore profiles from North Island containing the c. 1300 Kaharoa Tephra

AD

Notes: Dates shown at the top of each profile are estimates of the timing of earliest human-induced deforestation impacts based on major changes in Pteridium spore counts and pollen spectra with respect to the Kaharoa datum. The earliest inferred deforestation signal occurs just before Kaharoa Tephra, at c. AD 1280. The c. AD 200 Taupo Tephra provides a pre-impact datum in all but three profiles where underlying materials have been dated by radiocarbon 1 Stratigraphic position of Kaharoa Tephra is inferred from an adjacent core at the Wharau Rd swamp site Sources: After Newnham et al. (1998a) with additional data from Elliot et al. (1997) and Horrocks et al. (1999, 2000)

137

VOLCANISM AND EARLY MAORI SOCIETY

Table 8.2 Potential effects and extent of impact of the main volcanic hazards1 on prehistoric Maori society Hazard Pyroclastic fall

Threat to life

Generally low, except close to vent2 Pyroclastic flow/surge Extremely high Lava flows Low Lahars/flooding Moderate Gases/acid rain Low Forest fires Low 1 2

3

Threat to property

Areas affected

Variable, depends on Local to regional thickness2 Extremely high Local to regional Extremely high Local High Local to regional3 Low Local to regional Low to moderate Local to regional

After Scott et al. (1995). See also Ansell and Taber (1996). Possible physical effects arising from the accumulation of critical thicknesses of pyroclastic material (after Newnham et al., 1999b): 1 mm Little or no effect on people apart from possible minor, short-lived respiratory problems; rapid recovery. 10 mm Fish and insects killed; little or no visibility; infections of respiratory tract, inflamed eyes; crops possibly damaged or rendered unpalatable; possible animal poisoning, e.g. by fluorine. 100 mm Serious respiratory problems, some human fatalities possible; bird life killed; crops destroyed or severely damaged, trees stripped, branches broken; roof collapse likely for dwellings and other buildings, especially with rain; water supplies temporarily contaminated. 1m Fatalities and injuries from building collapse; bush fires; trees, hunting areas destroyed; many waterways blocked; hunting/fishing gear and canoes and other property destroyed or damaged; animals killed directly or by starvation. 10 m Substantial loss of life through building collapse and burns; widespread building and other property destruction; some waterways permanently altered; long-term loss of land use; resettlement of survivors. These processes may result in prolonged devastation, at times well after the initial eruption impacts, and well away from the volcano (e.g. Vallance, 2000).

had been encamped on or near the volcano, these events would have caused significant impacts both directly from pyroclastic flows or from tephra fallout, since at distances up to c. 15 km from source, the tephra deposits from each eruption range in thickness from c. 5–15 cm (Tonkin, 1970; Neall, 1976). Beyond a radius of about 15 km from source, the tephras are generally <5 cm thick but ash fall of the order of millimetres in thickness would have been blown over much of central North Island, depending on wind strength and direction (Neall and Alloway, 1991). For example, distal ash from one of the Burrell eruptions has been detected in Hawke’s Bay (Eden and Froggatt, 1996). In addition, fires were generated near the volcano by many of the eruptions, including both the Newall Ash and Lapilli events and the Puniho Lapilli events (Neall and Alloway, 1986; McGlone et al., 1988). Soon after the Newall eruptions, and with forest cover removed, much debris (e.g. from landsliding) was reworked down river valleys in a substantial post-eruptive flood event that resulted in extensive alluvial deposition (Hangatahua Gravels) in the Okato district (Figs 8.1, 8.2). The Maori name Okato means ‘place of a great tidal wave’, a description which may relate to this catastrophic flooding (Pearce, 1977; Neall et al., 1986). Several dozen debris flows or lahars, originating largely from lava flow collapses around the Taranaki

138

D.J. LOWE ET AL.

summit (especially the western crater rim) or from deep gorges in the middle part of the cone, have been emplaced since c. AD 1450 (Neall, 1979), and these may also have had delayed (and therefore unexpected) impacts on any Maori communities within stream valleys which acted as channel ways. There are few Maori oral accounts of specific volcanic activity on Taranaki, apart from reference to a fortified village ( pa) called Karaka-Tonga, believed to have been situated at about 600 m altitude on Waiwhakaiho River (near the Kokowai Springs) on the northern slopes of the mountain (Oliver, 1931). This village was evidently destroyed by an early (unknown) eruption. However, three in situ Maori earth ovens or umu, one dated at c. AD 1450 and the others at c. AD 1500 (Fig. 8.2), have been excavated on the volcano, and their occurrence within tephra layers (Fig. 8.3) means that eruptions such as the Newall and Burrell episodes must have been observed. Although most narratives indicate that later Maori regarded the Taranaki volcano with reverence, and that the higher slopes beyond the forest margin, rocky and barren, were sacred (tapu), it has been suggested that the umu, despite being at c. 850 m and c. 1075 m altitude, represent substantial rather than casual camp sites (Alloway et al., 1990). One inference is that the inhabitants of these sites either did not feel unduly threatened by the consequences of a renewal in volcanic activity, or that the activities associated with the sites (e.g. catching birds, collecting ochre, collecting fine-grained andesite for making axes, interment of the dead, or use as a remote retreat in times of political need) were seen to be worth the risk (Topping, 1974; Alloway et al., 1990). It is possible that the designation of the upper slopes as a ‘sacred area’ (wahi tapu), perhaps after initially being declared ‘out-of-bounds’ (rahui), was a deliberate societal response to reduce the impacts of future eruptions. Such responses to natural disasters elsewhere are known for early Maori, and have possible parallels in other societies (e.g. for Japan cf. Harada, 1993b, 1999 and below). For example, the fatal debris avalanching in AD 1846 at the Maori village at Waihi, southern Lake Taupo (see Table 8.1, footnotes), resulted in a tapu being placed on the devastated site to avoid a recurrence of the disaster. Unfortunately a ‘transgressor’ paid with his life in the 1910 landslide (Harada, 1993b). Tongariro Volcanic Centre The Tongariro Volcanic Centre is dominated by two large and very active stratovolcanoes, Mts Ruapehu and Tongariro. Since c. AD 1250–1300, activity on Ruapehu has been centred on Crater Lake (in South Crater) and was probably wholly pyroclastic. On Tongariro eruptive activity, which has been centred on the parasitic cone of Mt Ngauruhoe and on Red Crater and the upper and lower Te Maari craters (Figs 8.1, 8.2), has generated both tephra deposits and lavas (e.g. Nairn and Self, 1978). Thus, prehistoric Maori would potentially have witnessed or been aware of dozens of explosive, tephra-generating eruptions, typically plinian–subplinian or smaller and of low magnitude and volume, and possibly a number of local lava flows: at least one occurred at c. AD 1500 (Hobden et al., 1999). However, the likelihood of significant impact on early Maori (other than minor damage or temporary respiratory-related problems arising from the downwind

VOLCANISM AND EARLY MAORI SOCIETY

139

deposition of several millimetres or so of distal ashfall in Hawke’s Bay for example (Froggatt and Rogers, 1990; Eden and Froggatt, 1996) is slight because apparently the whole area was regarded as tapu (Grace, 1959; Pearce, 1977; cf. below). It is doubtful if any Maori lived in close proximity to the volcanoes which, in any event, would generally be windswept, cold and inhospitable much of the time. In some accounts, the eruptions of Ngauruhoe were regarded as a sign of war by prehistoric Maori (R. Taylor in Gregg, 1960). Lahars were generated during and following larger Crater Lake-derived eruptions on Ruapehu in prehistoric times (Fig. 8.2; Cronin and Neall, 1997; Cronin et al., 1997; Lecointre et al., 1998), and by non-eruptive mechanisms (Donoghue and Neall, 2001), and these and associated post-eruptive flooding may have impacted on Maori communities in low-lying beds of river valleys draining away from the volcanoes (e.g. possibly at Turangi, Tongariro River in Fig. 8.1). Three lahars have been dated at between c. AD 1400 and 1650, and another since c. AD 1650 (Lecointre et al., 1998; Donoghue and Neall, 2001). There are no known Maori cultural remains associated with eruptives or laharic deposits from the Tongariro Volcanic Centre. White Island The island is unlikely to have been occupied for any period of time. Its continual volcanic activity was clearly well known because this has an important place in Maori oral legends (e.g. Luke, 1959; Pearce, 1977; Orbell, 1995). Sporadic eruptions of low volume and magnitude during prehistory probably had little or no impact on nearby coastal settlements in eastern Bay of Plenty and towards East Cape (Fig. 8.1), apart from that resulting from occasional wind-blown ash fall which probably amounted to a few millimetres at most. Furthermore, since it lies in deep water beyond the continental shelf and has a history of only minor eruptions, White Island is unlikely to have generated significant tsunami at the coast (de Lange and Fraser, 1999). There are no known Maori cultural remains associated with White Island eruptives. Rangitoto Island (Auckland Volcanic Field) Rangitoto Island, a small shield volcano comprising basalt lava fields capped by a steep-sided scoria cone, is the youngest and by far the largest volcano in the Auckland Volcanic Field (Fig. 8.1; Kermode, 1992; Nichol, 1992; Allen and Smith, 1994; Cassidy et al., 1999). Its spectacular emergence from the sea in c. AD 1400 (Lowe et al., 2000) was the result of effusive eruptions of lava and explosive fire-fountaining of scoriaceous pyroclasts. Alongside Rangitoto, Motutapu Island was blanketed with basaltic ash-fall deposits from the eruption (Fig. 8.1). At the Sunde site on West Point, Motutapu Island, human (both adult and juvenile) and Maori dog (kuri) footprints, together with other cultural remains, have been preserved beneath and within the ash layers derived from early phases of the eruption (Fig. 8.1; Nichol, 1981, 1982). According to Nichol (1981, 1988), the local Maori were apparently undeterred by the nearby Rangitoto eruption, and sanguinely engaged in horticultural activities by digging and gardening the freshly fallen ash deposits. Nevertheless, a

140

D.J. LOWE ET AL.

tephra-fall deposit in excess of c. 0.6 m thickness on the western third of the island (Nichol, 1988) must have engulfed and damaged or destroyed many Maori dwellings and other items there, as well as having other impacts (Table 8.2) within and beyond this part of the island (Cameron et al., 1997). There would have been only minor effects on Maori communities on nearby islands (e.g. Rakino, Waiheke, Motuihe and Motukorea) or the Auckland isthmus (mainland) that were within range of wind-blown ash. Subsequently, Motutapu Island may have attracted increased numbers of Maori settlers because the new cover of finegrained and easily worked tephra deposits was more suitable for gardening than the pre-existing clayey and enleached soils, which had formerly supported hunting, fishing and stone artefact manufacture (Davidson, 1978; Bulmer, 1996). Mt Tarawera (Okataina Volcanic Centre) The Tarawera Volcanic Complex, forming approximately the southern half of the Okataina Volcanic Centre (Fig. 8.1; Nairn, 1989), has erupted mainly rhyolitic material for most of its history. The c. AD 1300 Kaharoa eruption is the youngest rhyolitic eruption in New Zealand and the largest (volumetrically c. 7.5 km3 as deposited, c. 5 km3 as dense-rock-equivalent) since the powerful c. AD 200 Taupo event (Lowe et al., 1998). The eruption, triggered by an injection of hot basalt (Leonard et al., 2002), was a complex event that included initial vent-clearing blasts, the generation of at least 12 plinian to subplinian pyroclastic fall units and pyroclastic flow deposits, and the slow effusion of three large lava domes. The partial collapse of several domes produced extensive block-and-ash flows around Mt Tarawera (Nairn et al., 2001). The most widely dispersed tephra-fall deposits (‘Kaharoa Tephra’) were distributed in a northwest–southeast oriented lobate pattern, with the 3 cm isopach covering an area of c. 30,000 km2 in eastern and northern North Island including Northland, Coromandel, Bay of Plenty and Hawke’s Bay (Fig. 8.5; Lowe et al., 1998). These cataclysmic phases were followed by ‘breakout’ post-eruptive flooding down the Tarawera River (fed by breaches of water from Lake Tarawera) towards the Bay of Plenty coast and attained an extremely high peak discharge estimated at c. 105 m3/s (Hodgson et al., 1997; White et al., 1997; Manville et al., 1999). Based on similar types of largescale rhyolitic eruptions elsewhere (e.g. Machida, 1990; Rodolfo, 2000), there would have been little or no warning of this flooding. Some landsliding and probably debris avalanching triggered by aftershocks or rainstorms may have occurred as well. Thus, it is likely that the eruption and subsequent events would have had enormous impacts on any early Maori living in eastern North Island, including probable annihilation for those (if any) who had been close to the Mt Tarawera massif, or within or adjacent to the Tarawera River channel and floodplain downstream from the volcano. Around one quarter of the eastern North Island was affected by tephra fallout of 3 cm or more in thickness, and it is likely that little of the North Island escaped some fallout from the eruption (Newnham et al., 1998a). Even a few centimetres of distal, acid-coated ash fall would have been sufficient to damage forest vegetation for some decades (e.g. Giles et al., 1999). Newnham et al. (1998a) and Lowe et al. (2000, in press) have inferred from

VOLCANISM AND EARLY MAORI SOCIETY

141

pollen records that the earliest Maori settlers may have arrived in New Zealand just prior to the Kaharoa eruption. If so, it is possible that only a tiny population saw or experienced the eruption and its consequences. Several oral legends that seemingly refer to an ‘ancient’ eruption of Tarawera (i.e. before AD 1886) support the suggestion that early Maori were witnesses to the event (Lowe et al., 2000). Tarawera eruption The Tarawera eruption of 10 June 1886 was the biggest and most destructive eruption in New Zealand during the historical (European) period. It was a basaltic rather than rhyolitic event, but was nevertheless very explosive: the resulting scoria fall (‘Tarawera Scoria’) has a dispersal similar in extent to that of the Vesuvius AD 79 pumice fall and is one of the few known examples of a basaltic deposit of plinian type from a fissure source (Walker et al., 1984). The eruption cored out a series of craters in a 7 km long fissure through the antecedent rhyolite domes (including those emplaced during the Kaharoa event) of Mt Tarawera, and then generated more craters along an 8 km long southwest extension of the fissure across the Rotomahana basin (which contained two shallow lakes and large silica sinter aprons, the ‘Pink’ and the ‘White’ terraces, asssociated with extensive hydrothermal activity) to Waimangu (Fig. 8.6; Walker et al., 1984). Narratives (summarized by Keam, 1988) indicate that after a series of precursory earthquakes from c. 12.30 am, the eruption began at Ruawahia Dome (Fig. 8.6) at c. 2.00 am on 10 June 1886, and then gradually extended both northeastward and southwestward. At c. 2.10 am the eruption intensified with the ascent of a tephra plume from the vicinity of Ruawahia Dome up to c. 9.5 km (Walker et al., 1984; Keam, 1988). By 2.30 am craters along the whole length of the fissure were erupting, with the Rotomahana extension beginning to erupt possibly at c. 3.20 am. By 3.30 am, craters along the entire 17 km length of the fissure from Wahanga to Waimangu were in eruption. This paroxysmal stage of the eruption was over by 6.00 am, when most activity ceased. The erupted products were exclusively pyroclastic (no lava flows were generated, although basalt dikes were emplaced). The total volume (as deposited) of Tarawera Scoria is c. 2 km3 (Walker et al., 1984). The eruption along the Rotomahana and Waimangu extension was mainly phreatomagmatic (the result of interaction between basalt magma and hydrothermal water) and phreatic. The explosive expansion of superheated water fragmented the country rock containing the hydrothermal system, plus subordinate lake sediment, to produce surge beds and fall deposits (‘Rotomahana Mud’) that rained out over much of the Bay of Plenty and beyond (c. 0.5 km3 as deposited) (Fig. 8.6, inset). Near Rotomahana, the surge beds were emplaced violently by hot and fast-moving turbulent pyroclastic surges or density currents up to c. 6 km from source (Figure 8.6; Nairn, 1979; Keam, 1988). Lightning during the eruption set fire to a house in Te Wairoa and to the forest on the north shore of Lake Tarawera; strong winds flattened many trees at Lake Tikitapu; and suffocating gases and falling mud and ash made breathing difficult at Te Wairoa (Fig. 8.6), where most buildings were buried or collapsed under the weight of c. 1 m of mudfall (Fig. 8.7).

142

D.J. LOWE ET AL. i ve r

l l l l ll

l l ll

l l l l l l l l l l l l l l l l l l

l

l l

l l

l

l l l

l l

ll

l

l l l l l

l

l

l l l

l l

l l l l l l l l lll l l l l

120

l l

ll l l l

l

l

l l l

l l l l l l l l l l

l

l

l

l l l l l l l l

s

l l l l l l l l l l l l l l l l l l l

l l ll l ll l l l

l

l l

er

l l l

l l l l

ll l ll

l

l

l

l

l

l

l l l l l l l l l l l

l l l l l l l l l

l l l l l l l l

l l l l l l l

l l l

l

l l l l l l

ll

l l

l l

l l l l ll l l

l ll

l l l l l

l l l l l l l l l l

t ra

l l

60

l

ba se

ll l l l

l l l l l l l l l l l

ima ctic

l l l l l l l

l

l l l l l l l l l

ll

Approx. bounda ry c l

l l l l l l l l l l l l

l l l l

l l

l l l l l

l l

l

l

l l l l l

l l l

l l

ll l l

l l l

c era

l l l l ll l l l l l l l l l l l l l

l l

l l l l l l l l l l l l l l

l l l l l

aw Tar

l l l l l

l

Southern Dome

Waingongongo (19)

30

ters cra l l

l l l l l l

l l l l

l l

l l l

l

l l l l

l l l l l

l ll l l l l l l

l

ll

l l l

l

l l l

l l l l

l ll l l l

5

PACIFIC OCEAN 15

Aldermen Is Tairua

l l l

Lake Rotomahana (infills Rotomahana Crater)

Mayor I. Tauranga Whakatane

an

aR

d

Hamilton

Lake Rerewhakaaitu

ah

Lake Okaro

tom

L. Tarawera Mt Tarawera

Ro

38

Rotorua

5

Hi gh wa y

Ash fall on Julia Pryce ~300 km N of East Cape

Murupara

Rere

whaka

aitu R oad

L. Taupo

East Cape Tolaga Bay Gisborne

0

3 km

l l l

ngu ima Wa

l l l l

2

l l l ll l l l

l l

l

l l

l l l

l l l l l l l l

l l l

Rd

l l l

l l

1

l l

l

l l l l l l l l l l l

l l

l l

l l l l l l l l l l 60 l l l l l l l l l l l l l l l

l l l l

l

l

a an ah om ot it R m . li ox pr Ap

l l l l l l

l

Lake Ngahewa 0

l l l

l l l l l

l l l l l l l l l l

The Chasm

Lake Rotomakariri Lake Rotomahana (pre-1886) (pre-1886)

(~12)*

l l l l l

fa llo ut

l l l l l l l l l l

l l ll l l l l l

P

ngu

M ud

Tarawera Dome

l

l l l

W

Waim a

St at e

l l l l ll

Te Ariki (~15) *

l l l l l l l l l l l l l l

Western l Dome

Totorariki (7)

l

ll

l l l l l l l l l l l

To Rotorua

Wahanga Dome

Ruawahia Dome l l

l l l l l ll

l l l l l ll

Tutaeheke

l l l l l l

Moura (25)

po de

ll

e rg su

l l l l l l l l l l l l l l l l l

Lake Rotokakahi (Green Lake)

l l l l l l l l l l l l

l

0 12

Tarawera Volcanic Complex

30

l l l l ll

Te Wairoa (ÒBuried VillageÓ) (17) sit

t (cm) fallou coria aS wer a r Ta

l l l l ll

l l l l

l l l l l l l l l l l ll l l l l l l l l l l l l l l l l l l l l

Lake Tarawera

l

Rd

Tokiniho (11)

R

ll

l

e ra aw Tar

l l l

Haroharo Volcanic Complex

l l l l l l l l l l l l l l ll l l l l l l l l l ll l l l l l l l l l

l l

l

l l

l l l l l l l l

Waitangi (1)

era aw Tar

Lake Tikitapu (Blue Lake)

l l

l l ll l l l l l

l l l l l l

Lake Okareka

Maungakakaramea (Rainbow Mountain)

0

100 km

Figure 8.6 Map of the Tarawera area showing locations of the main craters of the 10 June 1886 fissure eruption across the Tarawera Volcanic Complex, Rotomahana Crater (including pre-eruption lakes Rotomahana and Rotomakariri) and Waimangu craters Notes: The approximate boundary of the base surge deposit resulting from the initial cataclysmic Rotomahana explosion is based on data in Smith (1886) and Keam (1988); bold arrows represent current directions inferred from cross-bedding of proximal surge deposits (Nairn, 1979). The southwestern fallout limit for Rotomahana Mud and the isopachs (cm) for Tarawera Scoria, both members of Tarawera Tephra Formation (Froggatt and Lowe, 1990), are from Thomas (1888) and Cole (1970), respectively. The location of villages and associated fatalities (numbers in parentheses) are based on Keam (1988). There was an additional death at an unknown locality. The fatalities were all Maori apart from six Europeans at Te Wairoa and one European and three part-Maori at Waingongongo. W, White Terrace (Te Tarata); P, Pink Terrace (Otukapuarangi) * On the night of the eruption nearly half of Te Ariki’s 27 residents were encamped at Pink Terrace at Lake Rotomahana (Keam, 1988) Inset shows eastern North Island and the documented limits (stippled) of tephra fallout from the AD 1886 Tarawera eruption (Thomas, 1888). Ash fell on a number of ships at sea, the farthest being c.300 km ( Julia Pryce) and c.1000 km (S.S. Waimea) from North Island (Keam, 1988)

All but seven of the 108 known fatalities arising from the Tarawera eruption were Maori (the true number of deaths may have been c. 120: Lowe et al. 2001). The majority of deaths were the result of the Rotomahana explosions, especially the lethal, scorching pyroclastic surges and blasts (Fig. 8.6). Clearly the event had a profound impact on Maori (and others) in the Te Wairoa and Rotomahana area especially, but trauma was felt throughout the extensive fallout zone in the Bay of

VOLCANISM AND EARLY MAORI SOCIETY

143

Figure 8.7 Buried meeting house (wharenui) (‘Hinemihi’) and smaller houses (whare) at Te Wairoa (Fig. 8.6) after the AD 1886 Tarawera eruption Note: The meeting house, although mud-covered, remained intact and provided shelter for many survivors Photograph: A.A. Ryan (photo no.15), 13 June 1886, by permission of the Museum of New Zealand Te Papa Tongarewa, Burton Brothers Collection C.10706

Plenty and eastern North Island (Keam, 1988). For example, some groups of Maori in the region of the Rangitaiki and Tarawera rivers, north of Tarawera, became refugees at Matata (Fig. 8.6). Although they had escaped with their lives and without serious injury, their possessions were buried by c. 15–30 cm of tephra (some were retrievable by excavation), many potato pits were lost and those with livestock had no feed for them and so many starved (Keam, 1988). These people were eventually resettled in 1903–05. The plight of these and other Maori seem minor in comparison with the difficulties of those from Te Wairoa-Rotomahana: apart from the lives lost, all possessions had been buried and many crushed. Among livestock, most smaller animals were killed, but dogs, pigs, cattle and horses that survived wandered loose and starving. The main livelihood of the region, tourism, had been destroyed, literally overnight. (Whilst Maori continued to participate in the tourist trade, its control effectively moved into European hands from 1894 with the opening of the railway line to Rotorua: McKinnon, 1997.) But perhaps the biggest societal impact, according to Keam (1988), was the loss of land. For 30 years, Maori groups in the region had been generally secure in possession of their land and property. In previous times, under the old order, the prospect had always existed

144

D.J. LOWE ET AL.

that a group might lose homes and land through warfare, but by the time of the Tarawera eruption, the people, long-established traders with European settlers, had become accustomed to a new-found security. The eruption rather than warfare (against which there could at least be retaliatory or conciliatory action to make-good losses) had destroyed that security and dispossessed the people of the land which they had prized most. Offers of resettlement for the surviving group, mainly the Tuhourangi subtribe or clan (hapu), were received from various parts of central and eastern North Island and beyond, but most settled at Whakarewarewa and Ngapuna, both near Rotorua. Eventually gifts of land were formally ratified and provided a home for most of the Tuhourangi people (Keam, 1988). Other Tuhourangi settled for a time in the Bay of Plenty and Coromandel. After 30–50 years almost all the refugees or their descendants had returned to Whakarewarewa or Ngapuna and the gifted land was returned to the donors (Keam, 1988).

BENEFITS OF ERUPTIONS The Tarawera eruption exemplifies the volcanic threat faced by early Maori, but numerous benefits were also provided by the wide range of volcanic features and products in much of the North Island (Table 8.3). Most benefits were practical and derived from physical resources. The use of certain ‘sacred’ soils and pigments also held spiritual significance (e.g. Gray, 1996). The exploitation of most of these features is generally well documented (e.g. Davidson, 1984; Jones, 1987; McKinnon, 1997; Stokes, 2000; cf. Table 8.3). Two examples of direct and indirect benefits of volcanism are described here. One of the most important volcanic resources for early Maori was the glassy rock, obsidian (tuhua), which was used extensively for making tools and traded. Sources are found in central and northern North Island (McKinnon, 1997) but Mayor Island, or Tuhua, was by far the most important (Fig. 8.8). Innumerable archaeological sites throughout New Zealand contain obsidian from Major Island. Compositionally distinctive, it occurs also in fourteenth century AD archaeological sites in the Kermadec Islands (e.g. Raoul Island), which are about 1000 km northeast of New Zealand (Anderson and McFadgen, 1990; Higham and Johnson, 1996). A benefit arising indirectly from volcanism is the formation of secondary iron oxide minerals, chiefly ferrihydrite, which occur typically as orange–brown volcanogenic seepages. Ferrihydrite occurs also in small quantities in many types of soils, but is most abundant in those formed from more basic materials such as scoria which are reddish in colour – onerua. Such seepages are the result of ferrous iron, derived ultimately from the dissolution of iron-bearing minerals (characteristic of volcanic and pyroclastic deposits and their derivatives) being oxidised and precipitated rapidly on contact with air and/or via bacterial action. The oxides provided very useful pigments for a range of functional and important ceremonial purposes (Table 8.3). Such deposits were also used for both practical

VOLCANISM AND EARLY MAORI SOCIETY

Table 8.3

145

Beneficial volcanic features and products exploited by prehistoric Maori

Features/products

Benefits

Elevated sites for fortified villages (pa)1 Friable soils (e.g. onetea or onemata) dominated by allophane and/or ferrihydrite and used for horticulture (northern–central North Island)2 Tubes/caves/clefts in lava, ignimbrite Interment of dead, storage (e.g. of fishing nets)3 Volcanic lakes Water and food supply, transport route Volcanic or pyroclastic edifices Mythological figures (e.g. ‘giant stone warrior’ (bluffs, outcrops, etc.) ignimbrite inselbergs on Mamaku Plateau), petroglyphs4 Geothermal activity Hot-water supply for cooking, bathing, medical treatment, etc.5 Volcanic cones, lahar mounds Scoria and volcanic-ash-derived soils

Volcanogenic iron oxides Ferrihydrite seepages

Red scoriaceous soils (haematite rich) Volcanic rock materials Boulders, cobbles

Obsidian, basalt, fine-grained andesite Pumice Volcanogenic sediment 1 2

3 4 5 6 7

8 9 10

Yellowish–reddish pigments (kokowai) for functional and ceremonial purposes (e.g. facial or body decoration, paint for buildings/canoes, insect repellent)6 Pigments for facial decoration, especially on high-ranking chief, etc. (one tapu, pito one, onerua, onekura)7 Stonefield (basalt) structures (e.g. dry stone walls dividing land or garden plots, retaining walls, shelters, pa defence walls), anchors, hammerstones, cooking stones (umu)8 Tools (adzes, beaters, drill points, flake tools, e.g. knives, flax scrapers), weapons9 Abrader, carved figurines/heads, fishing-net floats, bowls, gourd stoppers, children’s toys, etc.9 Additive to garden soils (e.g. for growing sweet potato), backfill10

Davidson (1993), Bulmer (1996), Cameron et al. (1997). Best (1925), Bulmer (1996), McKinnon (1997), Gumbley et al. (in press). Allophane is a shortrange order aluminosilicate clay mineral (e.g. Lowe, 1995). Davidson (1984). Stafford (1999). Stokes (2000). See Table 8.4 and text; Childs et al. (1986), Stafford (1999). One tapu, pito one: sacred soil; onerua: reddish soil; onekura: poor reddish soil. Kokowai would be used on the face of a deceased paramount chief, onerua on that of a lesser chief. Both were generally tapu (Gray, 1996). Davidson (1984), Cameron et al. (1997). Moore (1980), Davidson (1984), Jones (1987). Best (1925), McFadgen (1980), Gumbley et al. (in press).

and ritualistic purposes by Japanese cultures since c. 10,000 BC from the Jomon to Kofun periods (Kamijoh, 1997). One of the most extensive ferrihydrite deposits in the North Island occurs at springs issuing beneath andesite lavas on the northern slopes of Mt Taranaki at Kokowai Springs (Fig. 8.1). The deposits were referred to as kokowai, meaning

146

D.J. LOWE ET AL.

Figure 8.8 Mayor Island or Tuhua, a rift-related peralkaline rhyolite caldera volcano located c.30 km off the western Bay of Plenty coast (Fig. 8.1), was the pre-eminent source of obsidian (tuhua) for prehistoric Maori Photograph: White’s Aviation

‘earth from which red ochre is procured by burning’ or simply ‘red ochre’, by early Maori (Childs et al., 1986). The strong colour intensity and high surface area (200– 400 m2/g), and the powdery nature and chemical stability when dried, help make the ochre a good colouring agent (Table 8.4). The material reddens on heating, giving a range of colours, with maximum redness corresponding to the formation of the iron oxide mineral haematite by 750 ºC (Childs et al., 1986). The use of kokowai by Maori was recorded by Dr Ernst Dieffenbach in his account of his ascent of Mt Taranaki in 1839 They used it for many purposes; when mixed with shark’s oil, it forms a durable paint for their houses, canoes and burying places; it is also universally in request to rub into their faces and bodies . . . the custom of covering themselves with a thick coating of this substance at the death of a relation or friend may have a symbolic meaning, reminding them of the earth from which they have sprung . . . The New Zealander also regards this pigment as a good defence against the troublesome sandflies and musquitos [sic]. (B. Wells in Childs et al., 1986: 86–7)

VOLCANISM AND EARLY MAORI SOCIETY

147

Table 8.4 Physical and elemental properties of volcanogenic red ochre1 (kokowai) from Kokowai Springs, Mt Taranaki Sample2

Density (g cm–3)

Surface area3 (m2 g–1)

Colour (Munsell) 4 (air dry)

1 3.56 225 2 3.29 n.d. 3 3.63 n.d. 4 3.71 380 Mean chemical analysis (N = 5)5 Fe

Mn Ti

Ca

Reddish yellow (7.5YR 6/8) Reddish-brownish yellow (7.5YR 6/8–10 YR 6/6) Strong brown (7.5YR 5/8) Reddish yellow (7.5YR 6/8) K

P

Si

Al

Mg Na H2O(+)* H2O(–)† C

Mean wt % 49.5 0.16 0.03 0.35 0.07 0.23 6.51 0.39 0.08 0.13 12.21 1 sd 6.0 0.23 0.02 0.19 0.11 0.18 1.95 0.36 0.03 0.10 3.68

20.15 1.10 2.85 0.97

1

Primarily ferrihydrite, a short-range order iron oxide (after Childs et al., 1986). Similar deposits occur on Mt Ruapehu (Childs et al., 1982) and elsewhere, especially in volcanogenic terrains (Lowe and Percival, 1993). 2 Sample 1, thick, extensive deposits at main spring seepage vent beneath a lava flow; 2, thin coating (0.01 mm) on boulders downstream from vent; 3, deposit below a spring vent dry at time of sampling; 4, thick layer (>1 mm) in extensive seepage area. 3 n.d., not determined. 4 On heating to 750 ºC, the ochre forms the crystalline iron oxide haematite and attains a maximum redness of 10R 4/8–2.5YR 4/8 (red). 5 Analyses (by X-ray fluorescence, gravimetric analysis, CO2 loss) of oven-dry material. Mean of analyses of samples 1–4 (Kokowai Springs) plus sample of thick ferrihydrite seepage in volcanogenic sediments at the ‘Ferry Bank’, Waikato River, Hamilton (Lowe and Percival, 1993). * Weight loss on ignition at 1,000 ºC for 1 h (loss of organic matter and structural hydroxyls). † Weight loss between air-dry and oven-dry states (absorbed water).

MYTHOLOGICAL AND SPIRITUAL RELATIONSHIPS Maori legends had an important role in communicating knowledge. The sources of some can be traced back for about 2,000 years to the time when Samoan explorers sailed out into the Pacific Ocean (Orbell, 1995) and hence many Maori legends are common to other peoples of Polynesia. The Maori have a large number unique to New Zealand, many relating to the natural environment (Reed, 1977; McCraw, 1990, 1993a, 1993b, 1994, 1995). This emphasis was closely tied to the religion of early Maori because nature, being outside their control,was part of the supernatural. Scores of legends purport to explain how the physical features of the landscape came into being. It is likely, however, that some of these legends had additional, subtler purposes, as illustrated below, and so it is probably a mistake to try to interpret them too literally. Nevertheless, a few show a remarkable parallel to geoscientific explanations for some phenomena. For example, legend records that when the Waikato River was first formed, it followed the wrong course to the sea and so the Maori volcano god, Ruaumoko, was asked to rectify the situation. Ruaumoko caused a great convulsion that blocked the river’s path at Piarere and forced it to follow the correct course (Fig. 8.9). A geoscientific explanation for the abrupt disjunction is nearly the same: the voluminous Kawakawa (Oruanui)

148

D.J. LOWE ET AL.

Figure 8.9 Locations of volcanic mountains and other features in North Island that are referred to in early Maori oral history Note:

TVZ, Taupo Volcanic Zone

eruption of Lake Taupo c. 26,500 calendar years ago (Wilson, 2001) so overloaded the river valley with volcanogenic debris that it overflowed its banks and adopted a new course (Selby and Lowe, 1992). Legends relating to volcanic features Each tribal group had its own sacred mountain, its size and prominence symbolising the importance of the tribe. Volcanic cones, if present in the tribal territory, were favourites for this role. Sacred mountains were regarded as ancestors and were given human attributes. Groups of mountains were often seen as an extended family and the legends tell how they squabbled, fell in love, had

VOLCANISM AND EARLY MAORI SOCIETY

149

offspring and so on. For example, a legend about the northwest-trending line of isolated, eroded andesite-basalt cones near Te Awamutu (Fig. 8.9), which are members of the Plio-Pleistocene Alexandra Volcanics (Briggs et al., 1989, 1997; Goles et al., 1996), tells how in former times there were more members of this family group. Fighting between male rivals over the female, Te Kawa, lead to newcomer Kakepuku defeating Puketarata, who retreated south, and Karewa, who eventually fled west and out to sea (Reed, 1977). It has been suggested that the main purpose of this legend is to make it clear that Karewa (Gannet Island) was once on land and so continues to mark the limits of tribal territory. Te Heuheu, chief of the Ngati Tuwharetoa tribe of Taupo district, explained to Dr Ferdinand von Hochstetter, an Austrian geologist visiting in 1859 during the Novara expedition, that fire was sent from Hawaiiki (the mythical Maori homeland) in response to a call from the high priest, Ngatoroirangi. The latter had been exploring the interior of the North Island and was now perishing from the cold on the summit of Tongariro (Fig. 8.9). The fire came under the sea to Whakaari and from there, according to Te Heuheu, travelled underground towards Tongariro. On the way it came to the surface in various places, giving rise to hot springs, geysers and other hydrothermal features. Finally, the fire burst out on the summit of Tongariro and saved Ngatoroirangi’s life. Hochstetter was impressed by this explanation for the locus of hydrothermal activity in what is now called the TVZ (Figs 8.1, 8.9). He concluded that the Maori had grasped the connection between volcanism and hydrothermal activity (von Hochstetter, 1959; Gregg, 1960). The same legend also tells how Ngatoroirangi eliminated a rival by calling on Ruaumoko to direct clouds of ash on to him. This is possibly a reference to the frequent tephra eruptions (including pyroclastic flows) of Ngauruhoe. Perhaps the best known of all Maori legends tells of a group of volcanoes that once stood in the centre of the North Island near Lake Taupo. They began fighting over the only female, Pihanga (Fig. 8.9), and a great battle ensued. According to the Tuwharetoa version, their sacred mountain, Tongariro, was victorious and at least three defeated volcanoes were put to flight, leaving Lake Rotoaira and the Tama Lakes where they had stood. According to folklore, mountains can move only at night, so when dawn came, and they became frozen in place, Taranaki had reached the west coast and Putauaki (Mt Edgecumbe) the Bay of Plenty region, but Tauhara, still besotted with Pihanga, had dawdled and was stranded near (present-day) Taupo township (Fig. 8.9). In this way early Maori explained the distribution of volcanoes (Grace, 1959). Perhaps an eruption from Tongariro, closely followed by one from Taranaki (Fig. 8.2), was interpreted as volcanic activity moving from one mountain to the other. In the manner of legends, it was but a short step to the mountain itself moving. An informant from the Ngati Awa tribe of the Bay of Plenty added to this story by telling how Putauaki (that tribe’s sacred mountain) had married Mt Tarawera but hankered for Whakaari (White Island). Tarawera became so jealous that she exploded and wept tears, forming Lake Tarawera (Fig. 8.9). Evidently this legend is much older than the historic Tarawera eruption of AD 1886. If so, it could be

150

D.J. LOWE ET AL.

derived from the tribal memory of the Kaharoa eruption of c. AD 1300, as described earlier, and may explain Tarawera’s strict tapu status. Pearce (1977) and Keam (1988) documented several other possible stories that may relate to the Kaharoa eruption, but these are open to various alternative interpretations. Legends associated with volcanic hazards There are few legends about the hazards or disasters associated with volcanic activity. One example from Auckland explains the explosive formation of the Auckland Volcanic Field (Fig. 8.1), when powerful incantations from tohunga (priests) were called upon by one group to destroy an opposition war party (Simmons, 1987). It is intriguing, however, that although the c. AD 1400 eruption of Rangitoto Island was undoubtedly witnessed by early Maori, there are no legends clearly referring to this event. The name ‘Rangitoto’, which can be translated as ‘blood sky’, was formerly thought to refer to an eruption, but it is now accepted as referring to ‘blood-stained rocks’ through an injury received by a chief (Woolnough, 1984). Indeed, a legend explaining the formation of Rangitoto attributes it to a giant or a god lifting the mountain from the mainland where it was blocking the view of an important chief and dumping it in the Waitemata Harbour, leaving Lake Pupuke (a basaltic maar) to mark its original position (Fig. 8.9, inset A; Lowe and Green, 1992). There are several explanations for the lack of legends referring to this eruption. One is that the eruption, although a dramatic physical event, may not have affected the mana (prestige) of the tribe to any extent because no important chief was killed or injured, nor did such a person make comment that was worthy of preservation by a legend. Another reason is that the ancient people who witnessed the eruption were eliminated by later invaders and any memories of the eruption perished with them. Despite many legends describing eruptions as angry mountains fighting each other with much rumbling, the ejection of boulders, and ‘fiery glowing’, none refers to flowing lava. This suggests that Maori were not familiar with the process or material. Apart from Rangitoto Island, the only other places where lava streams would be seen were close to the active vents on Mt Tongariro such as Ngauruhoe and near the summit of Mt Taranaki (Fig. 8.2). These areas came to be designated tapu, and so presumably visitors would not have been able to approach closely enough to recognise the slow-flowing andesitic lava for what it was. Alternatively, if Maori did venture to the upper slopes, as shown by the buried umu on Mt Taranaki (Fig. 8.3), their visits did not coincide with any eruption of lava. Similarly, a legend from Auckland about the formation of an old, 10 km long basalt lava flow derived from Three Kings volcano (Kermode, 1992) and extending part-way across Waitemata Harbour forming Te Tokaroa (Meola) Reef (Fig. 8.9, inset) gives no hint that this feature was once molten rock or that it came from a volcano. The c. AD 200 Taupo eruption from Taupo caldera, Lake Taupo (Fig. 8.1), is the world’s most powerful known eruption for the past 5,000 years (Walker, 1980; Wilson and Walker, 1985; Wilson, 1993; Smith and Houghton, 1995). It involved five phases of plinian (including ‘ultraplinian’) and phreatomagmatic fall

VOLCANISM AND EARLY MAORI SOCIETY

151

activity, generating widespread tephra fallout, and a climactic sixth phase resulted in the violent emplacement of Taupo Ignimbrite over c.20,000 km2 of central North Island (Wilson, 1985, 1993; Wilson and Walker, 1985). The extreme violence and energy release (150 ± 50 megatons of explosive yield) of the ignimbrite-emplacement phase probably generated a global tsunami (Lowe and de Lange, 2000). However, the devastating eruption, which had enormous environmental impacts throughout North Island and beyond (e.g. Wilmshurst and McGlone, 1996), finds no mention in Maori legends. To summarise, early Maori society maintained a strong physical and spiritual relationship with the land, which is reflected in the wealth of legends about natural phenomena including the origin of landforms. Maori readily took advantage of the benefits supplied by volcanic activity as described previously but were well aware of the dangers and took care to placate the volcano god (Simmons, 1987). Restrictions were placed on visiting the frequently active andesitic volcanic centres of Tongariro and Taranaki. There is, however, no strong evidence from their legends that early Maori had suffered catastrophically from volcanic eruptions and associated events until the Tarawera event of AD 1886.

DISCUSSION AND CONCLUSIONS (1) The level and extent of the impacts of volcanism on early Maori in New Zealand are not well known. This is because New Zealand’s prehistory in the pre-European period was exceptionally brief, beginning c. AD 1250–1300. In addition, the historical (written) period also has been very short, effectively covering barely the past 200 years. Consequently, our chapter has relied mainly on interpreting volcanological, palaeoenvironmental and archaeological data, which in turn have been constrained mainly by radiocarbon and other dating techniques and by tephrochronology. (2) Since c. AD 1250–1300 it is likely that Maori would have witnessed only one rhyolitic eruption (Kaharoa, c. AD 1300), two basaltic eruptions (Rangitoto, c. AD 1400; Tarawera, AD 1886), and numerous andesitic eruptions from the very frequently active volcanoes of Tongariro Volcanic Centre, White Island, and Taranaki volcano (Fig. 8.2). Eruptions from Tongariro, Ngauruhoe and Ruapehu, and from White Island probably had relatively little direct impact because there few or no people living near them. In contrast, minor or short-lived impacts on more distant communities within range of tephra fallout, especially in eastern North Island (e.g. Hawke’s Bay, Bay of Plenty), would have been relatively common. The human witnesses to the Rangitoto eruption who were undertaking gardening in the freshly fallen basaltic ash seem to have displayed a remarkable level of insouciance towards the event. Many small-scale, preEuropean, basaltic eruptions in Taveuni, Fiji, similarly had apparently little impact on inhabitants because rapid resettlement or continued occupation occurred in nearby areas (Cronin and Neall, 2000). On the slopes of Mt Taranaki,

152

D.J. LOWE ET AL.

the existence of umu buried within tephra deposits indicates that those who used them did not feel unduly threatened, or they felt that the risk was worthwhile. Later Maori have come to regard the upper slopes of the mountain as tapu (meaning out-of-bounds, sacred) and the original designation of this status was possibly a response to minimise the risk for future generations. (3) Several eruptions, notably the Kaharoa event, the largest eruption in prehistory, and some of the Mt Taranaki events, including the Newall and Burrell series of eruptions, potentially had devastating consequences for relatively few people. Impacts from these eruptions, especially from the blasts, pyroclastic flows and surges, would have been catastrophic near source, as has been observed during historic times in more populous volcanic terrains, such as in Japan or the West Indies (e.g. Machida, 1990; Soda, 1996; Nakada, 2000). Lesser effects, although tending to become minor in distant areas, would have been felt for considerable distances away from the vent areas, probably over most of the North Island in the case of the Kaharoa eruption. It is possible that post-eruptive events, such as landslides, debris avalanches, flooding or lahar emplacement, may have caused significant damage or resulted in fatalities, depending on population levels and geographic circumstances (e.g. proximity of people to drainage channels). Moreover, these sorts of destructive events may have been prolonged (months to years) or occurred decades after the initial eruptive episode and then impacted with little warning on distant areas far from the volcano (cf. Scott et al., 1995; Soda, 1996; Scott and Nairn, 1998; Vallance, 2000). In Tables 8.1 and 8.2 we have summarised the different types of hazards seen or experienced by early Maori, and their potential effects and extent of impacts. (4) The Tarawera eruption of 10 June 1886 provided an insight into the level of impact and trauma that even a very powerful basaltic eruption is capable of generating: at least 108 fatalities, total ruination of villages and other possessions, loss of the main livelihood, dispossession of land, and consequent need for resettlement. The societal impacts of this eruption on local Maori were perhaps worsened by its occurrence during a period when European settlement was exerting significant pressures on Maori land ownership and economic activity. (5) Some secondary consequences of volcanism, such as the possible development of a ‘disaster culture’ by Maori society, may have occurred. Harada (1993b) suggested that, as in other prehistoric societies, early Maori may have developed a response mechanism to avoid the effects of future natural disasters initially by placement of a rahui, meaning ‘prohibited access’, on a devastated area. Subsequently, a more religious or superstitious restriction, or tapu, would be applied. Any violation of the tapu status, or sin (hara), was likely to bring upon a calamity. In contrast, other sacred areas were designated as accessible places of refuge or sanctuaries for all citizens (e.g. a marae, a ceremonial gathering place). This interpretation has some similarities with Japan, where Shinto shrines and their surrounds, which are sacred and inviolate areas, represent religious places

VOLCANISM AND EARLY MAORI SOCIETY

153

both of worship and refuge that may have been initially established in safe zones in response to earlier natural disasters (Harada, 1993b, 1999). (6) Given the extent and frequency of volcanic activity in the North Island for the past c.2000 years (Fig. 8.2), it is notable that the disastrous consequences of volcanism do not figure more prominently in Maori culture, oral history and mythology. This contrasts with findings from Taveuni island in Fiji, for example, where local legends describing past eruptions have persisted since c. AD 120–320, and where such stories and relict place names can be readily related to independent volcanological and archaeological evidence (Cronin and Neall, 2000). Similar firm linkages between oral tradition and geoscientific evidence for volcanic impacts can be found in some other prehistoric societies (e.g. Blong, 1982; Chapters 9–11). The apparent incongruity for New Zealand, however, is reconcilable by the lateness of its settlement (and sparse population), which meant that there were fewer opportunities for substantial populations to witness very destructive eruptions, and probably because most of the frequent eruptions since settlement occurred in comparative isolation or generally had minor (i.e. forgettable) effects. The AD 1886 eruption of Mt Tarawera had catastrophic impacts on some Maori, yet this event was smaller by an order of magnitude than the previous eruption of Mt Tarawera (the Kaharoa event), which in turn is dwarfed in terms of magnitude and impacts by the c. AD 200 Taupo eruption. The c. AD 1300 Kaharoa eruption occurred very early in prehistory and the most hazardous areas close to Mt Tarawera are unlikely to have had dense occupation, if settled at all. At Lake Waikaremoana, situated in remote uplands (Fig. 8.5), the earliest unambiguous evidence for local human settlement in the pollen record occurs well after the 12 cm thick layer of Kaharoa Tephra was deposited (Newnham et al., 1998b). However, those pollen records indicating that initial human settlement occurred just before the deposition of Kaharoa Tephra are obtained mostly from coastal sites where the Kaharoa Tephra layer is much thinner (Newnham et al., 1998a; Fig. 8.9). As a consequence, populations at these sites are unlikely to have developed a disaster culture because they did not need one. The c. AD 200 Taupo eruption, however, obliterated a much more extensive area of the central North Island and would surely have forged an indelible cultural impression on any people who survived beyond the disaster zone, had they been there. That no cultural recognition of this eruption exists is consistent with the archaeological and palaeoenvironmental evidence, which demonstrates that no people were present in New Zealand at the time of the Taupo eruption. (7) As well as bringing discomfort and some destruction to early Maori, volcanism also brought considerable physical and some spiritual benefits in many varied forms, as listed in Table 8.3. (8) Volcanism in New Zealand, as in similar terrains elsewhere, has been a boon for archaeology and palaeoecology (Harris, 2000). Tephrochronologists have been able to use tephras to help establish reliable linkages between archaeological

154

D.J. LOWE ET AL.

and palaeoenvironmental studies, a bridging of disciplines favoured by Edwards and Sadler (1999) and emphasised by Lowe et al. (in press). The work has helped to determine the timing of earliest occupation of archaeological sites (c. AD 1250– 1300) and to determine when the first environmental impacts of early Maori on the landscape were registered (c. AD 1280; Fig. 8.9). These findings, based around the critical ‘settlement layer’ datum represented by the c. AD 1300 Kaharoa Tephra, are in close agreement with other archaeological and palynological data in New Zealand (Newnham et al., 1998a; Lowe et al., 2000).

ACKNOWLEDGEMENTS This chapter proved to be a considerable challenge to write because of the dearth of substantive information and because data are widely scattered across many disciplines and sources. The result has been a pleasant surprise to us. We therefore thank John Grattan for inviting us to write the chapter and for his encouragement in completing it. We are also grateful to Janet Davidson, Willem de Lange, Kenichi Harada, Barbara Hobden, Tomohiro Kamijoh, Ron Keam, Gary Law, Daphne Lee, Vince Neall and Reg Nichol for providing advice or information, and especially to Tom Higham, Roger Briggs and Ron Keam for reviewing the text. Hiroshi Moriwaki kindly provided useful Japanese–English translations. Brent Alloway, Warren Gumbley, the Museum of New Zealand Te Papa Tongarewa, and Air Logistics are thanked for providing photos, and Betty-Ann Kamp for drafting the figures. DJL acknowledges an Invitation Fellowship from the Japan Society for the Promotion of Science that allowed him to make a helpful visit to Japan in October–November 2000.

REFERENCES Allen, S.R. and Smith, I.E.M. (1994) Eruption styles and volcanic hazard in the Auckland Volcanic Field. Geoscience Reports of Shizuoka University 20: 5–14. Alloway, B.V., Fyfe, R. and Neall, V.E. (1990) Further evidence of early Maori occupation on the flanks of Egmont volcano. Archaeology in New Zealand 33: 40–8. Anderson, A.J. (1991) The chronology of colonisation in New Zealand. Antiquity 65: 767–95. Anderson, A.J. (1995) Current approaches in East Polynesian colonisation research. Journal of the Polynesian Society 104: 110–32. Anderson, A.J. (1996) Was Rattus exulans in New Zealand 2000 years ago? AMS Radiocarbon ages from Shag River Mouth. Archaeology in Oceania 31: 178–84. Anderson, A.J. (2000) Differential reliability of 14C AMS ages of Rattus exulans bone gelatin in south Pacific prehistory. Journal of the Royal Society of New Zealand 30: 243– 61. Anderson, A.J. and McFadgen, B.G. (1990) Prehistoric two-way voyaging between New Zealand and East Polynesia: Mayor Island obsidian on Raoul Island and possible Raoul Island obsidian in New Zealand. Archaeology in Oceania 25: 37–42. Ansell, R. and Taber, J. (1996) Caught in the Crunch: Earthquakes and Volcanoes in New Zealand. Auckland: HarperCollins.

VOLCANISM AND EARLY MAORI SOCIETY

155

Best, E. (1925) Maori Agriculture. New Zealand Dominion Museum Bulletin 9. Wellington. Blong, R.J. (1982) The Time of Darkness: Local Legends and Volcanic Reality in Papau New Guinea. Canberra: Australian National University Press. Briggs, R.M., Itaya, T., Lowe, D.J. and Keane, A.J. (1989) Ages of the Pliocene– Pleistocene Alexandra and Ngatutura Volcanics, western North Island, New Zealand, and some geological implications. New Zealand Journal of Geology and Geophysics 32: 417–27. Briggs, R.M., Rosenberg, M., de Lange, P.J., Itaya, T., King, P.R. and Price, R.C. (1997) Geology and geochemistry of Gannet (Karewa) Island, Tasman Sea: a rift-related nephelenite tuff ring. New Zealand Journal of Geology and Geophysics 40: 261–71. Brook, F.J. (2000) Prehistoric predation of the landsnail Placostylus ambagiosus Suter (Stylommatophora: Bulimulidae), and evidence for the timing of establishment of rats in northernmost New Zealand. Journal of the Royal Society of New Zealand 30: 227–41. Bulmer, S. (1996) Settlement patterns in Tamaki-Makau-Rau revisited. In J.M. Davidson, G. Irwin, B.F. Leach, A. Pawley and D. Brown (eds) Oceanic Culture History: Essays in Honour of Roger Green: 641–55. Dunedin: New Zealand Journal of Archaeology Special Publication. Cameron, E., Hayward, B. and Murdoch, G. (1997) A Field Guide to Auckland. Auckland: Godwit Press. Cassidy, J., Locke, C.A., Miller, C.A. and Rout, D.J. (1999) The Auckland volcanic field, New Zealand: geophysical evidence for its eruption history. Geological Society [London] Special Publication 161: 1–10. Childs, C.W., Downs, C.J. and Wells, N. (1982) Hydrous iron oxide minerals with short range order deposited in a spring/stream system, Tongariro National Park, New Zealand. Australian Journal of Soil Research 20: 119–29. Childs, C.W., Wells, N. and Downs, C.J. (1986) Kokowai Springs, Mount Egmont, New Zealand: chemistry and mineralogy of the ochre (ferrihydrite) deposit and analysis of the waters. Journal of the Royal Society of New Zealand 16: 85–99. Cole, J.W. (1970) Description and correlation of Holocene volcanic formations in the Tarawera-Rerewhakaaitu region. Transactions of the Royal Society of New Zealand (Earth Sciences) 8: 93–108. Cole, J.W. and Nairn, I.A. (1975) Catalogue of the active volcanoes and solfatara fields of New Zealand. International Association of Volcanology and Chemistry of the Earth’s Interior XXII. Cole, J.W., Graham, I.J., Hackett, W.R. and Houghton, B.F. (1986) Volcanology and petrology of the Quaternary composite volcanoes of Tongariro Volcanic Centre, Taupo Volcanic Zone. Royal Society of New Zealand Bulletin 23: 224–50. Cronin, S.J.and Neall, V.E. (1997) A late Quaternary stratigraphic framework for the northeastern Ruapehu and eastern Tongariro ring plains. New Zealand Journal of Geology and Geophysics 40: 185–97. Cronin, S.J. and Neall, V.E. (2000) Impacts of volcanism on pre-European inhabitants of Taveuni, Fiji. Bulletin of Volcanology 62: 199–213. Cronin, S.J., Neall, V.E. and Palmer, S.J. (1997) Lahar history and hazard of the Tongariro River, northeastern Tongariro Volcanic Centre, New Zealand. New Zealand Journal of Geology and Geophysics 40: 383–93. Davidson, J.M. (1978) The prehistory of Motutapu Island, New Zealand: five centuries of Polynesian occupation in a changing landscape. Journal of the Polynesian Society 87: 327–37. Davidson, J.M. (1984) The Prehistory of New Zealand. Auckland: Longman Paul. Davidson, J.M. (1993) The chronology of occupation on Maungarei (Mount Wellington): a large volcanic cone pa in Auckland. New Zealand Journal of Archaeology 15: 39–55. de Lange, W.P. and Fraser, R. (1999) Overview of tsunami hazard in New Zealand. Tephra 17: 3–9.

156

D.J. LOWE ET AL.

Donoghue, S.L. and Neall, V.E. (2001) Late Quaternary constructional history of the southeastern Ruapehu ring plain, New Zealand. New Zealand Journal of Geology and Geophysics 44: 439–66. Donoghue, S.L., Neall, V.E. and Palmer, A.S. (1995) Stratigraphy and chronology of late Quaternary andesitic tephra deposits, Tongariro Volcanic Centre, New Zealand. Journal of the Royal Society of New Zealand 25: 115–206. Donoghue, S.L., Neall, V.E., Palmer, A.S. and Stewart, R.B. (1997) The volcanic history of Ruapehu during the past 2 millennia based on the record of Tufa Trig tephras. Bulletin of Volcanology 59: 136–46. Eden, D.N. and Froggatt, P.C. (1996) A 6,500-year-old history of tephra deposition recorded in the sediments of Lake Tutria, eastern North Island, New Zealand. Quaternary International 34–6, 55–64. Edwards, K.J. and Sadler, J.P. (1999) Introduction: striving for an environment–human consensus. Quaternary Proceedings 7: 5–8. Elliot, M.B., Striewski, B., Flenley, J.R., Kirkman, J.H. and Sutton, D.G. (1997) A 4,300-year palynological record of environmental change and human impact from Wharau Road swamp, Northland, New Zealand. Journal of the Royal Society of New Zealand 27: 401–18. Evans, J. (1998). The Discovery of Aotearoa. Auckland: Reed. Freundt, A., Wilson, C.J.N. and Carey, S.N. (2000) Ignimbrites and block-and-ash flow deposits. In H. Sigurdsson (ed.-in-chief) Encyclopedia of Volcanoes: 581–99. San Diego: Academic Press. Froggatt, P.C. and Lowe, D.J. (1990) A review of late Quaternary silicic and some other tephra formations from New Zealand: their stratigraphy, nomenclature, distribution, volume, and age. New Zealand Journal of Geology and Geophysics 33: 88–99. Froggatt, P.C. and Rogers, G.M. (1990) Tephrostratigraphy of high-altitude peat bogs along the axial ranges, North Island, New Zealand. New Zealand Journal of Geology and Geophysics 33: 111–24. Giles, T.M., Newnham, R.M., Lowe, D.J. and Munro, A.J. (1999) Impact of tephra fall and environmental change: a 1,000-year record from Matakana Island, Bay of Plenty, North Island, New Zealand. Geological Society [London] Special Publication 161: 11–26. Gillespie, J.L., Lowe, D.J. and Nelson, C.S. (1999) Large eruptions from White Island volcano: the Holocene record in Bay of Plenty marine cores. Geological Society of New Zealand Miscellaneous Publication 107A: 49. Goff, J.R. and McFadgen, B.G. (2000) Catastrophic seismic-related events and their impact on prehistoric human occupation in coastal New Zealand. Antiquity 75: 155–62. Goles, G.G., Briggs, R.M. and Rosenberg, M.D. (1996) Late Pliocene stratigraphic succession and volcanic evolution of Karioi volcano, western North Island, New Zealand. New Zealand Journal of Geology and Geophysics 39: 283–94. Grace, J.T.H. (1959) Tuwharetoa. Wellington: Reed. Gray, M.M. (1996) The origins of soil and its importance as characterised in Maori mythology and tradition. New Zealand Soil News 44: 64–9. Gregg, D.R. (1960) Volcanoes of Tongariro National Park. New Zealand Geological Survey Information Series 28. Gumbley, W., Higham, T.F.G. and Lowe, D.J. (in press) Prehistoric horticultural adaptation of soils in the central Waikato Basin with particular reference to site S14/ 201, Hamilton. New Zealand Journal of Archaeology. Hackett, W.R. and Houghton, B.F. (1986) Active composite volcanoes of Taupo Volcanic Zone. New Zealand Geological Survey Record 13: 61–114. Harada, K. (1993a) Taboos about oceans and conservation of natural environment. In I. Koizumi and K. Tanaka (eds) Ocean and Civilisation, 155–60. Civilisation and Environment Book Series 10. Tokyo: Asakura. Harada, K. (1993b) Disaster culture as societal technique. In N. Shuto (ed.) The Study of Disaster Culture in Regions Frequently Subject to Natural Disaster, 5–27. Report for Grantin-Aid for Scientific Research Project 04201110.

VOLCANISM AND EARLY MAORI SOCIETY

157

Harada, K. (1999) Shinto shrines as an example of disaster culture in Japan. Abstract, 1st Conference on Advances on Natural Hazards Mitigation Experiences from Europe and Japan, November 1999, Athens–Santorini, Greece. Harris, S.L. (2000) Archaeology and volcanism. In H. Sigurdsson (ed.-in-chief) Encyclopedia of Volcanoes, 1301–14. San Diego: Academic Press. Hedenquist, J.W. (1986) Geothermal systems in the Taupo Volcanic Zone: their characteristics and relation to volcanism and mineralisation. Royal Society of New Zealand Bulletin 23: 134–68. Hellstrom, J., McCulloch, M. and Stone, J. (1998) A detailed 31,000-year record of climate and vegetation change from the isotope geochemistry of two New Zealand speleothems. Quaternary Research 50: 167–78. Higham, T.F.G. and Hogg, A.G. (1997) Evidence for late Polynesian colonisation of New Zealand: University of Waikato radiocarbon measurements. Radiocarbon 39: 149– 92. Higham, T.F.G. and Johnson, L. (1996) The prehistoric chronology of Raoul Island, the Kermadec Group. Archaeology in Oceania 31: 207–13. Higham, T.F.G. and Lowe, D.J. (1998) Significance of an early date from an adzed tree stump in Northland. Archaeology in New Zealand 41: 212–19. Higham, T.F.G., Anderson, A.J. and Jacomb, C. (1999) Dating the first New Zealanders: the chronology of Wairau Bar. Antiquity 73: 420–7. Hobden, B.J., Houghton, B.F., Davidson, J.P. and Weaver, S.D. (1999) Small and shortlived magma batches at composite volcanoes: time windows at Tongariro volcano, New Zealand. Journal of the Geological Society, London 156: 865–8. Hodgson, K.A., Houghton, B.F. and White, J.D.L. (1997) What exactly did happen in the Tarawera River following the 700 14C Kaharoa eruption of Tarawera? Geological Society of New Zealand Miscellaneous Publication 95A: 81. Hogg, A.G., Higham, T.F.G., Lowe, D.J., Palmer, J.G., Reimer, P.J. and Newnham, R.M. (in press) Wiggle-match date for Polynesian settlement layer (Kaharoa Tephra) of New Zealand. Antiquity. Holdaway, R.N. (1996) Arrival of rats in New Zealand. Nature 384: 225–6. Holdaway, R.N. (1999) A spatio-temporal model for the invasion of the New Zealand archipelago by the Pacific rat Rattus exulans. Journal of the Royal Society of New Zealand 29: 91–105. Horrocks, M., Ogden, J., Nichol, S., Alloway, B.V. and Sutton, D.G. (1999) The palynology and sedimentology of a coastal swamp at Awana, Great Barrier Island, New Zealand, from c. 7,000 years BP to present. Journal of the Royal Society of New Zealand 29: 213–33. Horrocks, M., Ogden, J., Nichol, S., Alloway, B.V. and Sutton, D.G. (2000) Palynology, sedimentology and environmental significance of Holocene swamps at northern Kaitoke, Great Barrier Island, New Zealand. Journal of the Royal Society of New Zealand 30: 27–47. Houghton, B.F. and Wilson, C.J.N. (1986) Explosive rhyolite volcanism: the case studies of Mayor Island and Taupo volcanoes. New Zealand Geological Survey Record 12: 33– 100. Houghton, B.F., Weaver, S.D., Wilson, C.J.N. and Lanphere, M.A. (1992) Evolution of a Quaternary peralkaline volcano: Mayor Island, New Zealand. Journal of Volcanology and Geothermal Research 51: 217–36. Houghton, B.F., Wilson, C.J.N. and Pyle, D.M. (2000) Pyroclastic fall deposits. In H. Sigurdsson (ed.-in-chief) Encyclopedia of Volcanoes, 555–71. San Diego: Academic Press. Hunt, J.B. and Lowe, D.J. (in press) Tephra nomenclatura. Journal of Archaeological Science. Jones, K.L. (1987) Skill with stone and wood – Maori technology. In J. Wilson (ed.) From the Beginning: The Archaeology of the Maori 57–72. Auckland: Penguin. Kamijoh, T. (1997) Research on red pigments: pipe-shaped substances contained in base material. Bulletin of the Tokyo Metropolitan Archaeological Center 26: 115–22.

158

D.J. LOWE ET AL.

Keam, R.F. (1988) Tarawera: The Volcanic Eruption of 10 June 1886. Auckland: R. F. Keam. Kermode, L. (1992) Geology of the Auckland Urban Area 1:50,000. Lower Hutt: New Zealand: Institute of Geological and Nuclear Sciences Map 2. Lecointre, J.A., Neall, V.E. and Palmer, A.S. (1998) Quaternary lahar stratigraphy of the western Ruapehu ring plain, New Zealand. New Zealand Journal of Geology and Geophysics 41: 225–45. Lees, C.M. and Neall, V.E. (1993) Vegetation response to volcanic eruptions on Egmont volcano, New Zealand, during the last 1,500 years. Journal of the Royal Society of New Zealand 23: 91–127. Leonard, G.S., Cole, J.W., Nairn, I.A. and Self, S. (2002) Basalt triggering of the c. 1305 AD Kaharoa rhyolite eruption, Tarawera Volcanic Complex, New Zealand. Journal of Volcanology and Geothermal Research. Lowe, D.J. (1995) Teaching clays: from ashes to allophane. In G.J. Churchman, R.W. Fitzpatrick and R.A. Eggleton (eds) Clays: Controlling the Environment. Proceedings 10th International Clay Conference, Adelaide, Australia (1993), 19–23. Melbourne: CSIRO Publishing. Lowe, D.J. and de Lange, W.P. (2000) Volcano-meteorological tsunamis, the c. 200 AD Taupo eruption (New Zealand) and the possibility of a global tsunami. The Holocene 10: 401–7. Lowe, D.J. and Hunt, J.B. (2001) A summary of terminology used in tephra-related studies. In E.T. Juvigné and J.-P. Raynal (eds) Tephras: Chronology, Archaeology, 17–22. CDERAD editeurs, Goudet: Les dossiers de l’Archaéo-Logis no. 1. Lowe, D.J. and Green J.D. (1992) Lakes. In J.M. Soons and M.J. Selby (eds) Landforms of New Zealand: 2nd edition, 107–43. Auckland: Longman. Lowe, D.J. and Percival, H.J. (1993) Clay Mineralogy of Tephras and Associated Paleosols and Soils, and Hydrothermal Deposits, North Island. Guide Book for New Zealand PreConference Field Trip F.1. 10th International Clay Conference, Adelaide, Australia. Lowe, D.J., McFadgen, B.G., Higham, T.F.G., Hogg, A.G., Froggatt, P.C. and Nairn, I.A. (1998) Radiocarbon age of the Kaharoa Tephra, a key marker for late-Holocene stratigraphy and archaeology in New Zealand. The Holocene 8: 499–507. Lowe, D.J., Newnham, R.M., McFadgen, B.G. and Higham, T.F.G. (2000) Tephras and New Zealand archaeology. Journal of Archaeological Science 27: 859–70. Lowe, D.J., Keam, R. and Lee, D. (2001) How many deaths were caused by the Tarawera eruption? Geological Society of New Zealand Newsletter 126: 18–20. Lowe, D.J., Newnham, R.M., Higham, T.F.G., Wilmshurst, J.M., McGlone M.S. and Hogg, A.G. (in press) Dating earliest human impact and settlement in New Zealand. In D.G. Sutton (ed.) Origins of the First New Zealanders, 2nd edition. Auckland: Auckland University Press. Luke, J. (1959) History. In W.M. Hamilton and I.L. Baumgart (eds) White Island, New Zealand Department of Scientific and Industrial Research Bulletin 127: 14–24. Machida, H. (1990) Frequency and magnitude of catastrophic explosive volcanism in the Japan region during the past 130 ka: implications for human occupance of volcanic regions. Geological Society of Australia Symposium Proceedings 1: 27–36. Manville, V., White, J.D.L., Houghton, B.F. and Wilson, C.J.N. (1999) Paleohydrology and sedimentology of a post-1.8 ka breakout flood from intracaldera Lake Taupo, North Island, New Zealand. Geological Society of America Bulletin 111: 1435–47. McCraw, J.D. (1990) Maori legends as an aid in teaching Earth sciences. New Zealand Science Teacher 65: 45–7. McCraw, J.D. (1993a) Maori legends as an aid in teaching Earth sciences. 2. Legends about the origins of rivers. New Zealand Science Teacher 73: 14–17. McCraw, J.D. (1993b) Maori legends as an aid in teaching Earth sciences. 3. Legends about the origins of lakes. New Zealand Science Teacher 74: 15–18. McCraw, J.D. (1994). Maori legends as an aid in teaching Earth sciences. 4. Legends about mountains. New Zealand Science Teacher 76: 23–76.

VOLCANISM AND EARLY MAORI SOCIETY

159

McCraw, J.D. (1995). Maori legends as an aid in teaching Earth sciences. 5. Concepts of erosion and volcanic activity. New Zealand Science Teacher 77: 52–6. McFadgen, B.G. (1980). Maori plaggen soils in New Zealand, their origin and properties. Journal of the Royal Society of New Zealand 10: 3–19. McFadgen, B.G., Knox, F.B. and Cole, T.R.L. (1994) 14C calibration curve variations and their implications for the interpretation of New Zealand prehistory. Radiocarbon 36: 221–36. McGlone, M.S. (1989) The Polynesian settlement of New Zealand in relation to environmental and biotic changes. New Zealand Journal of Ecology (Supplement) 12: 115–29. McGlone, M.S. and Wilmshurst, J.M. (1999) Dating initial Maori environmental impact in New Zealand. Quaternary International 59: 5–16. McGlone, M.S., Neall, V.E. and Clarkson, B.D. (1988) The effect of recent volcanic events and climatic changes on the vegetation of Mt Egmont (Mt Taranaki), New Zealand. New Zealand Journal of Botany 26: 123–44. McKinnon, M. (ed.-in-chief) (1997) New Zealand Historic Atlas. Auckland: Bateman and Department of Internal Affairs. Moore, P. (1980) Prehistoric stone quarries of Tahanga. Historical Review 30: 32–42. Nairn, I.A. (1979) Rotomahana-Waimaungu eruption, 1886: base surge and basalt magma. New Zealand Journal of Geology and Geophysics 22: 363–78. Nairn, I.A. (1989) Sheet V16AC: Tarawera. Geological Map of New Zealand 1:50 000. Wellington: New Zealand Department of Scientific and Industrial Research. Nairn, I.A. (1991) Volcanic hazards of Okataina Volcanic Centre. New Zealand Ministry of Civil Defence, Volcanic Hazards Information Series 2. Nairn, I.A. and Self, S. (1978) Explosive eruptions and pyroclastic avalanches from Ngauruhoe in February 1975. Journal of Volcanology and Geothermal Research 3: 39–60. Nairn, I.A., Self, S., Cole, J.W., Leonard, G.S. and Scutter, C. (2001) Distribution, stratigraphy, and history of proximal deposits from the c. 1305 AD Kaharoa eruptive episode at Tarawera Volcano, New Zealand. New Zealand Journal of Geology and Geophysics 44: 467–84. Nakada, S. (2000) Hazards from pyroclastic flows and surges. In H. Sigurdsson (ed.-inchief ) Encyclopedia of Volcanoes, 945–55. San Diego: Academic Press. Neall, V.E. (1972) Tephrochonology and tephrostratigraphy of western Taranaki (N108– 109), New Zealand. New Zealand Journal of Geology and Geophysics 15: 507–57. Neall, V.E. (1976) Volcanology of Egmont National Park. In J.H. Fullarton (ed.) Egmont National Park: 3rd edition, 56–62. New Plymouth: Egmont National Park Board. Neall, V.E. (1979) Sheets P19, P20 and P21: New Plymouth, Egmont and Manaia. Geological Map of New Zealand 1:50 000. Wellington: New Zealand Department of Scientific and Industrial Research. Neall, V.E. and Alloway, B.V. (1986) Quaternary volcaniclastics and volcanic hazards of Taranaki. New Zealand Geological Survey Record 12: 101–37. Neall, V.E. and Alloway, B.V. (1991) Volcanic hazards at Egmont volcano. New Zealand Ministry of Civil Defence, Volcanic Hazards Information Series 1. Neall, V.E., Stewart, R.B. and Smith, I.E.M. (1986) History and petrology of the Taranaki volcanoes. The Royal Society of New Zealand Bulletin 23: 251–63. Newnham, R.M., Lowe, D.J., McGlone, M.S., Wilmshurst, J.M. and Higham, T.F.G. (1998a) The Kaharoa Tephra as a critical datum for earliest human impact in northern New Zealand. Journal of Archaeological Science 25: 533–44. Newnham, R.M., Lowe, D.J. and Matthews, B.W. (1998b) A late Holocene and prehistoric record of environmental change from Lake Waikaremoana, New Zealand. The Holocene 8: 443–54. Newnham, R.M., Lowe, D.J. and Williams, P.W. (1999a) Quaternary environmental change in New Zealand: a review. Progress in Physical Geography 23: 567–610. Newnham, R.M., Lowe, D.J. and Alloway, B.V. (1999b) Volcanic hazards in Auckland, New Zealand: a preliminary assessment of the threat posed by central North Island

160

D.J. LOWE ET AL.

silicic volcanism based on the Quaternary tephrostratigraphical record. Geological Society [London] Special Publication 161: 27–45. Nichol, R. (1981) Preliminary report on excavations at the Sunde site, N38/24, Motutapu Island. New Zealand Archaeological Association Newsletter 24: 237–56. Nichol, R. (1982) Fossilised human footprints in Rangitoto Ash on Motutapu Island. Geological Society of New Zealand Newsletter 51: 11–13. Nichol, R. (1988) Tipping the feather against a scale: archaeozoology from the tail of the fish. Ph.D. Thesis, University of Auckland, New Zealand. Nichol, R. (1992) The eruption history of Rangitoto: reappraisal of a small New Zealand myth. Journal of the Royal Society of New Zealand 22: 159–80. Ogden, J., Basher, L.R. and McGlone, M.S. (1998) Fire, forest regeneration and links with early human habitation: evidence from New Zealand. Annals of Botany 81: 687– 96. Oliver, W.R.B. (1931) An ancient Maori oven on Mount Egmont. Journal of the Polynesian Society 40: 73–9. Orbell, M. (1995) The Illustrated Encyclopedia of Maori Myth and Legend. Christchurch: Canterbury University Press. Pearce, G.L. (1977) The Story of New Zealand Volcanoes. Auckland: Collins. Petchey, F.J. and Higham, T.F.G. (2000) Bone diagenesis and radiocarbon dating of fish bones at the Shag River Mouth site, New Zealand. Journal of Archaeological Science 27: 135–50. Price, R.C., Gamble, J.A. and Hobden, B.J. (2000) State of the Arc 2000: Guidebook for Field Excursions on Ruapehu and Tongariro Volcanoes. Wellington: Royal Society of New Zealand. Reed, A.W. (1977) Treasury of Maori Exploration. Auckland: Reed. Rodolfo, K.S. (2000) The hazards from lahars and Jökulhlaups. In H. Sigurdsson (ed.-inchief ) Encyclopedia of Volcanoes, 973–95. San Diego: Academic Press. Sase, T. and Hosono, M. (1996) Vegetation histories of Holocene volcanic ash soils in Japan and New Zealand – relationship between genesis of melanic volcanic ash soils and human impact. Earth Science (Chikyu Kagaku) 50: 466–82. Scott, B.J. and Nairn, I.A. (1998) Volcanic Hazard Map of Okataina Volcanic Centre. Environment Bay of Plenty Resource Planning Publication 97/4. Scott, B.J., Houghton, B.F. and Wilson, C.J.N. (1995) Surveillance of New Zealand’s volcanoes. Tephra 14: 12–17. Selby, M.J. and Lowe, D.J. (1992) The middle Waikato Basin and hills. In J.M. Soons and M.J. Selby (eds) Landforms of New Zealand, 233–55. Auckland: Longman. Sigurdsson, H. (ed.-in-chief) (2000) Encyclopedia of Volcanoes. San Diego: Academic Press. Simmons, D. (1987) Maori Auckland. Auckland: Bush Press. Smith, S.P. (1886) Preliminary report on the volcanic eruption at Tarawera. Appendices to the Journal of the House of Representatives of New Zealand H-26, 1–4 (+ maps). Smith, R.T. and Houghton, B.F. (1995) Vent migration and changing eruptive style during the 1800a Taupo eruption: new evidence from the Hatepe and Rotongaio phreatoplinian ashes. Bulletin of Volcanology 57: 432–9. Soda, T. (1996) Explosive activities of Haruna volcano and their impacts on human life in the sixth century AD. Geographical Reports of Tokyo Metropolitan University 31: 37–52. Stafford, D. (1999) Pakiwaitara: Te Arawa Stories of Rotorua as Told to Don Stafford. Auckland: Reed. Stevenson, C.M., Sheppard, P.J. and Sutton, D.G. (1996) Advances in the hydration dating of New Zealand obsidian. Journal of Archaeological Science 23: 233–42. Stokes, E.M. (2000) The Legacy of Ngatoroirangi: Maori Customary Use of Geothermal Resources. Hamilton: Department of Geography, University of Waikato. Sutton, D.G. (1987) A paradigmatic shift in Polynesian prehistory: implications for New Zealand. New Zealand Journal of Archaeology 9: 135–55.

VOLCANISM AND EARLY MAORI SOCIETY

161

Sutton, D.G (ed.) (1994a) Origins of the First New Zealanders. Auckland: Auckland University Press. Sutton, D.G. (ed.) (1994b) Conclusions: origins. In D.G. Sutton (ed.) Origins of the First New Zealanders, 243–58. Auckland: Auckland University Press. Thomas, A.P.W. (1888) Report on the Eruption of Tarawera and Rotomahana, New Zealand. Wellington: Government Printer. Tonkin, P.J. (1970) The soils of the southeastern sector of Egmont National Park. Earth Science Journal 4: 36–57. Topping, W.W. (1972) Burrell Lapilli eruptives, Mount Egmont, New Zealand. New Zealand Journal of Geology and Geophysics 15: 476–90. Topping, W.W. (1974) An AD 1480 Maori oven from Mount Egmont, New Zealand. New Zealand Journal of Science 17: 119–22. Ui, T., Takarada, S. and Yoshimoto, M. (2000) Debris avalanches. In H. Sigurdsson (ed.in-chief ) Encyclopedia of Volcanoes, 617–26. San Diego: Academic Press. Vallance, J.W. (2000) Lahars. In H. Sigurdsson (ed.-in-chief ) Encyclopedia of Volcanoes, 601–16. San Diego: Academic Press. Valentine, G.A. and Fisher, R.V. (2000) Pyroclastic surges and blasts. In H. Sigurdsson (ed.-in-chief ) Encyclopedia of Volcanoes, 571–80. San Diego: Academic Press. von Hochstetter, F. (1959) Geology of New Zealand. Translated from German and edited by C.A. Fleming. Wellington: Government Printer. Walker, G.P.L. (1980) The Taupo plinian pumice: product of the most powerful known (ultraplinian) eruption? Journal of Volcanology and Geothermal Research 8: 69–84. Walker, G.P.L., Self, S. and Wilson, L. (1984) Tarawera 1886 New Zealand – a basaltic plinian fissure eruption. Journal of Volcanology and Geothermal Research 21: 61–78. White, J.D.L., Houghton, B.F., Hodgeson, K.A. and Wilson, C.J.N. (1997) Delayed sedimentary response to the 1886 AD eruption of Tarawera, New Zealand. Geology 25: 459–62. Wilmshurst, J.M. (1997) The impact of human settlement on vegetation and soil stability in Hawke’s Bay, New Zealand. New Zealand Journal of Botany 35: 97–111. Wilmshurst, J.M. and McGlone, M.S. (1996) Forest disturbance in the central North Island, New Zealand, following the 1850 BP Taupo eruption. The Holocene 6: 399– 411. Wilmshurst, J.M., Eden, D.N. and Froggatt, P.C. (1999) Late Holocene forest disturbance in Gisborne, New Zealand: a comparison of terrestrial and marine pollen records. New Zealand Journal of Botany 37: 523–40. Wilson, C.J.N. (1985) The Taupo eruption II. The Taupo ignimbrite. Philosophical Transactions of the Royal Society, London A314: 229–310. Wilson, C.J.N. (1993) Stratigraphy, chronology, styles and dynamics of late Quaternary eruptions from Taupo volcano, New Zealand. Philosopical Transactions of the Royal Society London A343: 205–306. Wilson, C.J.N. (2001) The 26.5 ka Oruanui eruption, New Zealand: an introduction and overview. Journal of Volcanological and Geothermal Research 112: 133–74. Wilson, C.J.N. and Walker, G.P.L. (1985) The Taupo eruption, New Zealand I. General aspects. Philosophical Transactions of the Royal Society, London A314: 199–228. Wilson, C.J.N., Houghton, B.F., McWilliams, M.O., Lanphere, M.A., Weaver, S.D. and Briggs, R.M. (1995) Volcanic and structural evolution of Taupo Volcanic Zone, New Zealand: a review. Journal of Volcanology and Geothermal Research 68: 1–28. Woolnough, A. (1984) Rangitoto. Auckland: A. Woolnough.

9

Under the volcano: Ni-Vanuatu and their environment JEAN-CHRISTOPHE GALIPAUD

INTRODUCTION Studying the relations that people have developed with their environment, including how they have dealt with unexpected natural catastrophes, raises difficult and complex issues, which are important for understanding the social and cultural evolution of human societies. The careful study of the impact of recent hazards in underdeveloped areas can help to gain some idea, but these observations are partly biased by the level of disaster prevention and the relief programmes that are currently available for even the most remote place on earth. In the following chapter I will attempt to show that much useful information can be learned about human responses and/or adaptation to disasters in the past by combining archaeological work with a careful analysis of oral history. As examples I will use two cases of natural hazards which have been recorded in oral history and also confirmed and dated by archaeology. They show that in Vanuatu natural disasters are perceived as social rather than natural events. These events are not feared but respected, and the environmental and physical risks are continually weighed and socially controlled.

THE PHYSICAL CONTEXT The Melanesian tectonic arc contains many risks for permanent human settlement. The islands where people live are subject to violent seismicity and volcanism resulting from tectonic plate movements. This ‘belt of fire’ is also affected from time to time by related catastrophic events such as the tsunami which devastated the northeast coast of Papua New Guinea in July 1998 (cf. Chapter 3). Within this wider region, the Vanuatu (New Hebrides) Archipelago, located at 16º S and 167º E between the Solomon and New Caledonian Islands chain (Fig. 9.1) extends over 1,000 km from north to south and comprises more than 80 islands totalling over 14,760 km2 in land area. Volcanic activity is important today: there are 12 active volcanoes, including 4 submarine ones. Seismicity, especially

Figure 9.1 The New Hebrides island arc showing active volcanoes and the location of the sites discussed in the text

164

J.-C. GALIPAUD

in the northern islands, causes uplift, which in some places has a rate of 6–10 mm/ year (Pineda and Galipaud, 1998). The island arc is Y-shaped and encompasses the results of recent volcanism to the east (Ambae, Maewo, Ambrym), the Banks Islands in the northeast, Tanna and Aneytum in the south, and more ancient formations to the west (Malakula, Santo). The Torres Islands, far to the north, and Futuna and Aniwa to the east are recently emerged islands with a coral cap. The initial settlement of the Vanuatu Archipelago follows closely the model set up for remote Oceania with discovery around 3,000 BP and progressive settlement of coastal areas on low islands and then on high islands, which are all settled by 2,000 BP. During the second millennium BP, coastal areas were abandoned and settlement shifted towards the mid-altitude forested zones. The coast was settled again shortly before the arrival of the Europeans.

ORAL HISTORY In non-literate societies, oral history (sometimes also referred to as myths or legends) is the only means by which important rules and historical events can be kept in the memory and passed to the younger generation (e.g. Feil, 1997: esp. 74–5). The process of making and telling history is an important part of the life of those societies and is left to a few men who are responsible for the teaching of younger generations. Oral traditions relating to tribal descent, social organisation and land are often kept secret and, when told, make use of a specific language to avoid misuse of the information by limiting the understanding to knowledgeable members of the group. Many stories also have a more understandable version meant for children. Those simpler stories, revised and simplified versions of the former, help educate the children by slowly giving them the tools which will eventually allow them to understand the social and cultural message behind the words. These latter stories are naturally often those that are told to foreigners such as researchers, and it is important when studying them to keep their role in society in mind. Although oral history is very important for the transmission of knowledge, stories specifically referring to disasters are not very common in Vanuatu. Disasters are sometimes described in myths and legends and are usually attributed to demons or spirits who wish to punish a breach of a social or cultural taboo. These stories are probably based on natural events but do not provide information on the social and temporal context of the natural hazard and are thus difficult to use. They do, however, reveal the social dimension of the natural world. Oral histories referring specifically to natural hazards which happened before the arrival of Europeans are very few and were probably transmitted because of the importance of the natural phenomenon in question. More recent stories on volcanism in Vanuatu show that most of the islands have been directly or indirectly affected over the past 300 years by some catastrophic events. Among the best remembered is the sudden disappearance of a small island offshore from Aoba some three centuries ago and the destruction of several villages on the west coast of Aoba

PERCEPTION OF DISASTERS BY NI-VANUATU

165

around the same time (Bonnemaison, 1996: 69). The only older natural event that is well remembered in the oral tradition is the eruption that completely destroyed the big island of Kuwae, in the centre of the Vanuatu Archipelago around AD 1450 (Monzier et al., 1994). One might think that people living in such an environment would have a rich tradition concerning events that affected their life and history in the past. It is surprising to find that major well-known and well-dated volcanic eruptions, such as Kuwae, are only locally recorded. This could suggest either that those events had no profound effect on the lives of people or that the memory of the events died out with the inhabitants themselves. Other traditions, seemingly older, indirectly attest extreme natural events that affected some islands in the past. To illuminate what inhabitants of this small volcanic island world (the NiVanuatu) may have thought and how they possibly reacted to the permanent threat posed by their environment, I have chosen two traditions which can both be confirmed by archaeological work. The first one is about a tsunami in the Torres Islands, possibly 2,000 years old. The second tale provides an account of the large volcanic eruption which destroyed the island of Kuwae around AD 1450.

THE OLD MAN WHO COULD RAISE TSUNAMI The following history provides us with one of the few accounts of the destruction of a village by a tsunami. I recorded it in 1993 in the village of Litau on the southeast coast of Toga, one of the islands of the Torres Archipelago in northern Vanuatu. Titus Joel, fieldworker of the Port Vila Cultural Centre, who received it from his father, the former chief of the village, recounted it to me. The story is interesting for several reasons. First, it shows that people in Vanuatu do not conceive what we consider to be ‘natural’ disasters in the same way. They think they are humanly caused. Second, it shows that some oral histories have a factual basis. Finally, the date of the event (c. 2,450 BP, as far as could be determined through archaeological excavations) indicates that some events recorded through oral history can be remembered for a very long time, in this case for more than 100 generations. All the local people knew of the story and of the place where the ancient village was located. The story is summarised below. An old man named Raherir was living a very long time ago in a village named Kurvot, behind the actual village of Litau. He was extremely old and could hardly move. A strong spell kept him alive against his will. Tired of this very miserable life, he decided to kill himself using his magical ability. One day, he called his two sons and explained to them a spell to raise a tsunami that would kill him. Five days before the event, all the inhabitants of the village were led to the nearby plateau with all their belongings and waited for the wave. When announced, a

166

J.-C. GALIPAUD

huge wave raced to the shore. The old man, afraid of the power of what he had triggered, tried to escape and crawled slowly towards the plateau. Too late. The huge swell took him and buried him among the ruins of his village. In 1996 and 1998, during archaeological excavations on the old Kurvot village itself, I was able to show that this tsunami was not a legend and that it had had an important impact on the local society (Galipaud, 1998). The stratigraphy of the Kurvot archaeological site clearly indicates a rapid deposition of sterile marine sand above an older occupation layer and thus gives some credibility to the tradition that some rapid marine event affected this coast in the past. In some areas, the sand is up to 90 cm in thickness. Three dates on charcoal from the top, middle and basis of this old occupation layer gave statistically identical results. These place the occupation at c. 2,450 BP (Table 9.1). Levels 3–6 at the site currently represent the oldest known site in the Torres Islands. A pottery type known as ‘Plain Ware’, which occurs throughout the remote Western Pacific shortly after the Lapita period, characterises it. The site seems to have been disturbed by later gardening activity (level 3) before being buried in sand. Above the sterile sand, a late occupation has been dated to the first millennium BP. Although no exact date can be fixed for the time of the catastrophe, the stratigraphy suggests that it happened shortly after the initial occupation of Kurvot, somewhere around 2,000 years ago. Evidence for early occupations in Vanuatu occur along the coast. Recently acquired dates in Malo suggest, as in the Torres, that these occupations were sporadic beach settlements rather than long-term settlements. During most of the first millennium AD, all the previously occupied coastal locations lack any evidence for settlement, suggesting that those places became unsuitable to live in. The natural disaster may have had a long-term effect on settlement since the evidence from the Kurvot archaeological site indicates that there was a shift inland shortly after the date of the tsunami. Resettlement of the coastal fringe does not seem to have occurred before the seventeenth or eighteenth century. The Kurvot example shows that since some oral histories have a basis in fact we can use them to study past disasters. Unfortunately, the link between oral history and archaeology is not always so straightforward, and the next example will show that oral traditions can sometimes be altered to match precise social or political strategies.

Table 9.1

Radiocarbon dates from the Kurvot site on Toga, Vanuatu

Lab. code

Stratigraphic context

Age C14

C13/C12

Age calib.

Beta-133974 Beta-118605 Beta-118606

Kurvot, spit 3 level 3 Kurvot, spit 3 level 4 Kurvot, spit 3 level 6

2490±100 2470±40 2420±70

–26.0‰ –25.7‰ –22.8‰

2490±100 BP 2460±40 BP 2450±70 BP

PERCEPTION OF DISASTERS BY NI-VANUATU

167

KUWAE, THE LOST ISLAND My second example concerns Kuwae, the so-called ‘lost island’. Around AD 1452, the island of Kuwae, one of the largest of the Vanuatu archipelago, partially exploded, killing a large proportion of its population and sending a huge quantity of ash into the atmosphere. The remains of the island now form the islands of Epi in the north and Tongoa in the Shepperd group. The eruption led to the formation of an oval-shaped submarine caldera 12 km long and 6 km wide, with two distinct basins and a total area of around 60 km2 at the level of the rim. The caldera has been proved to be the result of a single event of short duration, which occurred in the first half of the fifteenth century (Monzier et al., 1994). A study of the caldera and its volcanic layers has allowed a precise reconstruction of the eruption. The main phase was preceded by a moderate volcanic activity with some basalt magma flows during several months. Two dacitic layers followed by 10 m of plinian ash and pumice falls testify to an active phase which probably did not exceed a few days in duration. The volume of magma that was emitted during the main eruptive phase is estimated at around 32 to 39 km3. The dating of burned wood found in pumice layers has given an estimate of AD 1420–30. A sulphur and chlorine anomaly in ice cores from Antarctica in 1452 and from Greenland in 1453 suggests that the paroxysm of the eruption happened at this date (Monzier et al., 1994). The formation of the caldera is ‘one of the 7 most important explosive eruptions in the last ten millennia and is comparable in strength to the Santorini one in Greece (3,600 BP) or the Mt Mazama one in the United States (6,845 BP)’ (Eissen et al., 1994: 1200). This eruption probably caused the death of several thousand people, although the increased seismic and volcanic activity, which probably started a few years before the main eruption, would have encouraged some people to flee to nearby islands before the collapse. In geological terms the Kuwae eruption was clearly a very major event with catastrophic consequences for the local environment. In order to understand the impact on humans of the Kuwae eruption, one can analyse two forms of information. I will begin by examining the oral history and will then turn to the archaeological evidence. In examining the oral history, one must be aware that the story is used by people now to justify their social status and ownership of property. Nevertheless, it reveals a great deal about the way that survivors of the volcanic eruption viewed the disaster. Several version of the story are known. I present a synthesis. A man whose mother was from Ambrym [another active volcano] lived in the village named Tanomala, near the Karua volcano on the island of Kuwae. This now-lost island used to encompass most of the Shepherd group as well as Epi. The man was tricked into an incestuous affair with his mother, whom he recognised too late by her tattoo. To take revenge on those who had led him to the incest, he went to Ambrym, where his uncle gave him a magical power over volcanoes. Once back in his village, he prepared six pigs’ bladders and,

168

J.-C. GALIPAUD

following the sorcerer’s instructions blasted them one after the other, each time choking the island, until the final one burst. While performing his magic, he summoned the inhabitants to leave the place. A few fled to nearby Efate but all others died, apart from a young boy who had hidden in a ceremonial slit-drum and a girl who sheltered in a cave near the shore. The young man, named Asingmet, came back a few years later to reclaim the remaining land and is now known by the chiefly name of Ti Tonga Liseiriki. The legend illustrates a number of important points about Ni-Vanuatu society and its reaction to disasters. Incest and magical power In Vanuatu, every natural event must have a rational explanation. As in the previous example of Kurvot, a socially unacceptable situation is the reason for the use of a magical power, which triggers the natural forces of the volcano. These natural forces are not seen as wild, so it is not unusual to stay in their proximity and not run away. Volcanic catastrophes are conceived as social events rather than natural, unexpected and feared ones. The six bladders In all versions of the myth, the six bladders that burst one after the other show the rise in intensity of the cataclysm. It confirms, in reality, that the final, most destructive eruption was preceded by several smaller events, which must have warned the population and allowed some of them to leave the island before it was too late. The use of the six bladders indicates that people were well aware of the behaviour of volcanoes and used warning signs to flee. The flight towards Efate This part of the myth points towards alliances. The inhabitants of Kuwae were not isolated. They had traditional relationships with people in nearby islands (Ambrym, Lopevi, Efate) and they certainly went to stay with their remote family when life became too difficult on Kuwae. It is difficult to estimate the death toll among those who remained on Kuwae. In these islands, society is hierarchical and structured under the rule of bearers of ‘titles’ which are transmitted from father to son. It seems that people representing the ‘titles’ fled very early to Efate to organise the retreat and preserve the social structure of the traditional Kuwae population. Ordinary people, however, whose survival depended directly on the exploitation of natural resources, probably stayed until the last minute and therefore died. There is surprisingly little information about alliances with islands to the north. Epi in particular was an important part of Kuwae and so we might expect it to now have story connections with the Shepherd Islands, as Efate does. So too with Ambrym, which was also part of the system, as is shown in the Kuwae eruption story. Most other oral traditions referring to later social changes and warfare also

PERCEPTION OF DISASTERS BY NI-VANUATU

169

focus on Efate. The best-known and historically most important one is the story of Roy Mata, which was the subject of thorough archaeological research in the late 1960s (Garanger, 1972).

THE LONG-TERM EFFECTS: ROY MATA Some versions of the Kuwae eruption story explain how the Shepherd Islands, especially the island of Tongoa, which is now the main remaining part of Kuwae in the area, were reoccupied. The population came back to Tongoa as soon as the first plant (a shrub named ‘tongoa’) and birds resettled the place. It is said to have happened about six years after the cataclysm. An island which had lost two-thirds of its original area must have been difficult to resettle and this certainly explains why people were so keen to get back as soon as possible to the place they owned or to claim a new area if their own had been lost. It is clear that the discussions on land ownership were not easy to clarify, and led to a prolonged social instability. This in turn might be the main cause for later socio-cultural transformations, attested by several oral history stories, the best known of which concerns Roy Mata. The legend of Roy Mata, a mythical hero who installed a new peace in Efate and the Shepherd Islands after a long period of fighting and who set up a new social system known as naflak, where allied chiefs organised themselves in a large network, might represent such a change. The French archaeologist José Garanger excavated sites in Efate and the Shepherd Islands (Garanger, 1972) associated with Roy Mata. He tentatively dated the social transformation to 700 BP. More recent work and new dating by Spriggs suggest that Roy Mata lived around 400 BP, that is, just after the Kuwae cataclysm (Bedford et al., 1998). New artefacts, in particular the Terebra shell adze, appear at the same time as Roy Mata, indicating some cultural affinities with Western Polynesia. We are left here with two hypotheses. 1 Roy Mata, a newcomer probably from Western Polynesia, arrived at the beginning of the first millennium BP and started to transform the ancient social system. The Kuwae eruption, which happened shortly afterwards, allowed his followers completely to overrule the ancient chiefs. 2 People arriving from Western Polynesia shortly before or after the Kuwae eruption used the social disruption caused by the cataclysm to get hold of the area and impose on the remaining inhabitants a new social order which was later to become predominant and totally replaced the older social system. In both cases, we can infer that the apparent continuity emphasised in the oral tradition of the Kuwae cataclysm might attest to the successful overruling of local society after the cataclysm.

170

J.-C. GALIPAUD

DISCUSSION These two examples, dating respectively from the first and the second millennium BP, raise some interesting points. First, it is clear that the proximity in time of the Kuwae event allows a better evaluation of the effects of the catastrophe upon the local society than does the much older Kurvot tsunami. Because of the very long period since the tsunami, it is difficult to evaluate the cultural consequences of the event. On archaeological facts alone, one might suggest that it caused the end of the coastally based society in the Torres and possibly in other islands of the same region. The oral history, however, suggests that this was not the case. The oral history is quite useful because it indicates that there was a minor change in settlement from the coastal fringe to the first uplifted terrace. Second, no matter how precisely the oral tradition depicts the Kuwae eruption, it gives us one side of the event only: the story that favours the survivors and supports their claims. Indeed, we know very little about the people in the Shepherd or other islands of the Vanuatu Archipelago who suffered from the eruption, the ashfalls and the darkness which must have affected the area for some time and must have caused much harm to their gardens. It is surprising how scarce oral history concerning the Kuwae catastrophe is in these other islands. The information which has been kept seems to have been retained on purpose to strengthen the social choices made and this can be seen as a testimony to the social imbalance caused by the cataclysm. Third, we should also consider the role of demography. The density of population in the central islands of Vanuatu might have been much higher than during the earlier disaster in the Torres Islands. The increasing population in islands where resources are limited might have become a factor in increasing social instability after natural events which had an influence on food supplies. This helps explain why the Roy Mata stories are mainly about controlling ownership over land.

CONCLUSIONS Are cataclysms really so bad? This somewhat pointed question reflects the reality of the Melanesian world. For those island societies, what might seem to us an extreme threat is conceived as a normal part of life and accepted in all its manifestations, the good and the bad alike. People regard these phenomena as part of life, know how to cope with them and accept the inconvenience and risk that they face. Low population density and well-organised alliances allow populations facing natural disasters to seek help among their allies and so limit the consequences of natural hazards. Since they are conceived of as ‘social events’, disasters are not feared and can be used to regulate and adjust the society. In this regard it is interesting to note that in both traditions, disasters were artificially monitored with the help of magical powers. In this chapter I have tried to show that oral traditions might have an ancient

PERCEPTION OF DISASTERS BY NI-VANUATU

171

background and that these traditions could, with the support of archaeology, tell us something about the attitude of early inhabitants of the Vanuatu Archipelago towards their volcanically active natural environment. In the tsunami example, an archaeological approach alone could lead to the conclusion that early coastal societies of the Torres Islands disappeared rapidly after some natural event. The consideration of the associated tradition shows, however, that the abandonment of the settlements is not associated with radical social changes. In the volcano example, there appear to have been very few effects, since people immediately reoccupied the land. In contrast, archaeological investigations suggest the volcanic eruptions might have caused major social changes in the long term. These two examples demonstrate the complexity of the relationship between people and their natural environment. When studying the impact of large natural events on local societies, one must bear in mind that the state of balance of the society at the time of the event is perhaps a more important criterion than the destructive level of the catastrophe.

REFERENCES Bedford, S., Spriggs, M., Wilson, M. and Regenvanu, R. (1998) The Australian National University–National Museum of Vanuatu Archaeological Project 1994–7: a preliminary report on the establishment of cultural sequence and rock art research. Asian Perspectives 37(2): 165–93. Bonnemaison, J. (1996) Les Fondements Géographiques d’une Identité. L’archipel du Vanuatu. Essai de géographie culturelle. Gens de pirogue et gens de la terre. Paris: Editions de l’ORSTOM. Eissen, J.P., Monzier, M. and Robin, C. (1994) Kuwae, l’éruption volcanique oubliée. La Recherche 270: 1200–02. Feil, D.K. (1997) Enga Genesis. Journal de la Société des Océanistes 104(1): 67–78. Galipaud, J.C. (1998) Recherches archéologiques aux îles Torres. Journal de la Société des Océanistes 107(2): 159–68. Garanger, J. (1972) Archéologie des Nouvelles-Hébrides: contribution à la connaissance des îles du Centre. Paris: ORSTOM. Monzier, M., Robin, C. and Eissen, J.P. (1994) Kuwae (c. 1425): the forgotten caldera. Journal of Volcanology and Geothermic Research 59: 207–18. Pineda, R. and Galipaud, J.C. (1998) Evidences archéologiques d’une surrection différentielle de l’île de Malo (Archipel du Vanuatu) au cours de l’holocène récent. Comptes-Rendus de l’Académie des Sciences 327: 777–9.

10

Earthquakes, subsidence, prehistoric site attrition and the archaeological record: a view from the Settlement Point site, Kodiak Archipelago, Alaska PATRICK SALTONSTALL AND GARY A. CARVER

INTRODUCTION This chapter focuses on the impact of great subduction earthquakes on both the archaeological record and the cultural history of the Kodiak Archipelago, Alaska. The Kodiak Islands are situated on the seismically active Aleutian–Alaskan subduction zone, the source of some of the largest earthquakes in the world. From 1994 to 1997 we undertook a detailed archaeological and geologic investigation at Settlement Point, a prehistoric village site on Afognak Island, to study the effects of subduction earthquakes on the archaeological record of the region. This chapter summarises our investigation of prehistoric occupation at Settlement Point and our inferences and conclusions about the relationship between Kodiak cultural history and great subduction earthquakes. There have been three major earthquakes in the Gulf of Alaska in the last 1,000 years. The most recent event was in 1964. The earliest, which occurred c. AD 1150, is associated with a gap in the archaeological record on the Kodiak Archipelago. The second and somewhat smaller earthquake occurred c. AD 1550. This event seems to have had no marked effect on the archaeological record. Certainly both were major disasters, but while the AD 1150 event has been linked with major cultural changes, archaeologists have previously ignored the AD 1550 event. It is our contention that these earthquakes did not have a long-term impact on the cultural trajectory of the Native people, the Alutiiq, because their society was geared to withstand such disasters. We have found no evidence indicating regional depopulation from the earthquakes. Excellent maritime skills, large societal territories and strong social relationships between villages created residential flexibility that allowed communities to disperse and coalesce in response to disasters. The Alutiiq also maintained trade networks that extended far beyond the Kodiak Archipelago – to the Aleutians, Alaska and Kenai Peninsulas and beyond. These networks expanded their social and economic universe, and mitigated the effects of local disasters. Situated in the western Gulf of Alaska on the eastern part of the Aleutian– Alaskan subduction zone, the Kodiak Archipelago is one of the world’s most

EARTHQUAKES AND THE SETTLEMENT POINT SITE

173

active sources of great earthquakes and tsunamis in historical times. During this century three of the world’s ten largest earthquakes, each in the magnitude range of 9 (Richter scale) or larger, have occurred along this subduction zone and each has generated large locally destructive tsunamis. The largest of these, the 1964 Prince William Sound earthquake, moment magnitude 9.2, ruptured about 850 km of the subduction zone at its eastern end. This caused coseismic subsidence of up to 2 m from the north end of Cook Inlet to the south end of Kodiak Island (Plafker, 1969). Paleoseismic evidence indicates that similar earthquakes, accompanied by substantial coastal subsidence and large local tsunamis, have recurred at roughly 400- to 700-year intervals throughout the Holocene (Plafker et al., 1992; Gilpin, 1995; Crowell and Mann, 1996). The Kodiak Archipelago includes 16 major islands, encompassing approximately 5,000 square miles (Buck et al., 1975). The major islands include 1,274 miles of coastline (Campbell, 1992) dominated by deep, fjord-like bays separated by rocky headlands and a mountainous and rugged interior. During the prehistoric period the islands were treeless and covered with maritime tundra vegetation. Only during the last few hundred years have trees, principally sitka spruce, become established on the islands and only on the northernmost part of the archipelago. Archaeological evidence from the Kodiak Archipelago shows that the first human settlements are about 7,500 BP (Fitzhugh, 1996; Steffian and Saltonstall, in prep.) and that for at least the past 6,000 years the islands have supported an increasingly large population of maritime-adapted people. Kodiak is home to the Alutiiq, an Esk–Aleut people who at the time of Russian contact (1784) had a socially complex society (Townsend, 1980). Communities were composed of large, multi-roomed, semi-subterranean houses built from grass sods, driftwood timbers and earth. A Russian census from 1805 (Lisiansky, [1814] 1968: 193) reports 4,000 Alutiiq people on Kodiak. When fur traders first arrived 30 years earlier, the population was more than twice as large. Hostile action by the Russians, epidemics and a rapid disruption of Alutiiq lifestyles were responsible for a precipitous decline in Native population during the early decades of Russian colonisation. At the time of Russian contact, the population density was greater than one person per square mile. However, the effective population density was even higher because the Alutiiq made only modest use of Kodiak’s rugged interior and so the population was scattered unevenly along the coastline. Most of the Alutiiq lived in villages located near river mouths and at the mouths of bays. In the early historic era, Lisiansky [1814] (1968) reported that the most densely populated region was near the mouth of the Karluk River, where about 600 people lived. The richness of the archipelago’s marine resources encouraged such population aggregates. Kodiak is noteworthy for its rivers teeming with anadromous fish. In the archipelago’s largest river, the Karluk, three million salmon were harvested in 1889 (Bean, 1892: 20). Salmon and other near-shore marine and freshwater fish provided the Alutiiq with a substantial part of their diet. The Alutiiq also hunted whales, sea lions, seals and other marine mammals; harvested

174

P. SALTONSTALL AND G.A. CARVER

birds; and collected invertebrates and other foods from local beaches. This reliance on marine resources contributed strongly to the pattern of coastal settlement. Geologic constraints also limited the number of suitable sites for the construction of semi-subterranean houses. Late Pleistocene glaciation over most of the archipelago scoured inland areas to bedrock and left a blanket of clay-rich till over lower elevations (Buck et al., 1975). These glaciated landscapes are poorly drained, precluding the construction of semisubterranean houses in most places. Sea-level changes during the mid and late Holocene, in part produced by subduction earthquakes, built raised beaches and prograded beach–berm complexes at many places along the coast. These beach deposits are well drained and provide one of the few widespread geologic settings suitable for house construction. Many of the archaeological sites in the archipelago are situated on these beach and berm complexes. A preference for beach ridge occupation made ancient settlements highly susceptible to the effects of subduction earthquakes, particularly land-level changes and tsunamis. This vulnerability is well documented. In 1964 a tsunami and subsidence event devastated three of the native villages in the archipelago: Old Harbor, Afognak and Kaguyak. Two were rendered uninhabitable (Plafker and Kachadoorian, 1966; Davis, 1971). Old Harbor was rebuilt, but both Afognak and Kaguyak were vacated and the survivors moved to other communities. In subsequent years, rapid coastal erosion spurred by subsidence resulted in the retreat of shorelines by up to 100 m. This process was particularly pronounced where soft sediments, including beach berm complexes, bordered the ocean. The shoreline retreat eroded significant evidence of both historic and prehistoric settlements. We believe that the 1964 earthquake and subsequent erosional episodes are not unique historical events. The same type of event and its consequent effects have been played out repeatedly in the past. Each event has probably been associated with social and economic disruption, but little loss of life. The long-standing effect of these events has been the erosional episodes, which have truncated portions of the region’s archaeological record.

KODIAK PREHISTORY There is some debate as to whether the earliest sites on Kodiak, dating to around 7,500 BP (Fitzhugh, 1996; Steffian and Saltonstall, in prep.), represent the first colonisation of the archipelago (Fitzhugh, 1996) or are the end of a long marine transgression that had eroded away the region’s earlier sites. Perhaps they are merely an indication that the archipelago’s coastline had begun to stabilise at that time. Whatever the case, the inhabitants of these sites were a well-adapted maritime people (Hausler-Knecht, 1993; Clark, 1979) who subsisted primarily by hunting marine mammals, birds and fish (Hausler-Knecht, 1993). Their way of life, termed the Ocean Bay Tradition, continued with minor changes for over 3,500 years (7,500 BP to 4,000 BP). These people were not nomadic seafarers.

EARTHQUAKES AND THE SETTLEMENT POINT SITE

175

Instead, they harvested seasonally available resources from local territories (Steffian and Saltonstall, 1999, in prep.). About 4,000 years ago there was a dramatic shift in subsistence emphasis towards fishing. Netsinkers and ground slate ulus first appear in archaeological sites of this age, signalling a shift to mass capture technologies and substantial long-term food storage. Within this new lifestyle, termed the Kachemak tradition (4,000 BP to 1,000 BP), people lived in large villages comprising single-roomed, semi-subterranean houses. They begin to signal social identity with labrets (a form of lip jewellery), which indicates the formation of corporate groups (Steffian and Saltonstall, 1995). Sites from the end of the Kachemak tradition era also yield intricately carved artwork, evidence of elaborate burial practices, an expanded inventory of tools, and a variety of exotic trade goods (Steffian, 1992a; Clark, 1984). The cultural changes that mark the end of the Kachemak tradition have been the subject of great debate. Some archaeologists (Dumond, 1988, 1994, 1998; Maschner, 1995) believe the succeeding Koniag tradition is the result of a largescale migration of Eskimo people to the archipelago from western Alaska. Others believe the changes reflect local, in-situ development ( Jordan and Knecht, 1988; Knecht, 1995) with perhaps a ‘boatload or two’ of immigrants providing outside influence (Clark, 1988, 1992). Whatever the case, the changes were dramatic. Large, multi-roomed houses with interior storage and cooking features replaced the much smaller and simpler dwellings of the Kachemak people (Saltonstall, 1997), and a dramatic increase in feasting and ritual is evident (Donta, 1993). This reflects the emergence of the ranked society encountered by the Russians at European contact. Determining the nature of the Kachemak–Koniag transition is difficult because relatively few sites from the precise transitional period have been found or excavated (Mills, 1993). The general timing of the transition coincides with a subduction earthquake and tsunami similar in magnitude to the 1964 event (Gilpin, 1995). Maschner (1995) suggests that the earthquake produced mass casualties that largely depopulated the islands and that the lack of transitional sites reflects this population crash. He interprets the subsequent Koniag tradition as the result of a large-scale migration to the recently depopulated archipelago. In contrast, Jordan and Knecht (1988; Knecht, 1995) and Fitzhugh (1996) argue that there is a great deal of evidence for cultural continuity across the Kachemak–Koniag transition. For instance, many of the tools from the two traditions are identical ( Jordan and Knecht, 1988), labret styles remained the same in the same regions (Steffian and Saltonstall, 1995), and the general way of life did not change. What did change were house forms – from small single-room structures to large multi-roomed constructions (Clark, 1984). There was an increased emphasis on ceremonialism (Donta, 1993; Saltonstall, 1998), warfare (Knecht, 1995) and storage, with much more storage space inside houses (Saltonstall, 1997, 1998). And, finally, villages were larger and tended to be situated along major salmon streams. These changes seem to be the culmination of trends already apparent in the earlier Kachemak tradition.

176

P. SALTONSTALL AND G.A. CARVER

Cultural changes of this sort are generally associated with the emergence of ranked societies (see Price and Feinman, 1995). One possible explanation for these dramatic changes is that Alutiiq society became socially stratified, a process that may have been accelerated by cultural adaptations to the colder climate of the Little Ice Age (Knecht, 1995). An episode of widespread and severe coastal erosional triggered by an earthquake such that the land subsided provides an alternative explanation for the lack of sites and seemingly abrupt change from the Kachemak to Koniag traditions. Coastal erosion removed much of the physical evidence of Late Kachemak occupation and so the lack of sites makes the cultural changes seem more abrupt. The third proposal for abrupt changes in this region is supported by the recent discovery of additional sites from this time period. On the southeast corner of the archipelago, a region not greatly affected by seismic subsidence, Fitzhugh (1996) found a continuous sequence of sites with radiocarbon dates covering the period in question. Jordan and Knecht (1988) also noted a continuous sequence of dates from two sites on the Karluk lagoon, another area not greatly affected by subsidence. Further archaeological investigation of these sites could yield more definitive answers on the nature of the Kachemak–Koniag transition. Meanwhile, archaeologically, little is known about this time period and in areas affected by subsidence in the AD 1150 earthquake the gap remains.

TECTONICS OF THE KODIAK REGION The Kodiak Archipelago is located on the edge of the North American plate above the eastern part of the Aleutian–Alaskan subduction zone where the Pacific plate is actively slipping beneath the North American plate. The subduction process on this plate margin is characterized by infrequent, large-displacement coseismic slip events that have historically generated great earthquakes and extensive upper plate deformation. The 23 March 1964 Prince William Sound earthquake, moment magnitude 9.2, resulted from rupture of a more than 850 km long segment of the subduction zone interface extending from the northern end of Prince William Sound southwest almost to the south end of Kodiak Island. This earthquake was generated by as much as 20 m of slip on the plate interface, displacing the North American plate margin southeastward over the Pacific plate. The plate interface dips at a very shallow angle, about 8 degrees, and the coseismic rupture was very wide, about 150 km at Kodiak, with the down-dip limit of coseismic slip located along the eastern side of the archipelago. Coseismic land-level changes resulted from this displacement on the subduction megathrust fault. The apparent associated slip on the accretionary margin fold and thrust belt faults in the upper plate offshore of the Kodiak Islands elevated the sea floor adjacent to the archipelago and generated large local tsunamis (Plafker, 1969;

EARTHQUAKES AND THE SETTLEMENT POINT SITE

177

Plafker and Carver, 1999). Tsunamis with run-up heights of 10 m at exposed headlands, and more than 6 m in bays and fjords on the eastern side of the islands, were produced by the sea-floor uplift. Coseismic subsidence lowered the archipelago as much as 2 m, with the axis of maximum subsidence extending through the southeastern part of the islands. The northeastern segment of the archipelago was characterised by the greatest amount of subsidence. Paleoseismic investigations of the Kodiak region have found evidence for similar subduction earthquakes and associated tsunamis during the Late Holocene (Gilpin, 1995; Gilpin and Carver, 1993; Carver et al., 1994). The principal data come from the analysis of shoreline stratigraphy in places where terrestrial sediments are abruptly overlain by intertidal mud. This stratigraphy is interpreted by geologists to reflect sudden coseismic subsidence and is analogous to sediment patterns resulting from the 1964 earthquake. In some places these sediments include sheets of marine sand at the coseismic horizon, which are probably deposits from large run-up-height tsunamis. Geological evidence documenting the two most recent pre-1964 subduction earthquakes, each with an associated tsunami, has been found at many localities around the archipelago. These earthquakes have been dated to c. AD 1550 and AD 1150 respectively (Gilpin, 1995). Less well dated are several additional paleoseismic horizons: one at c. AD 650 and another at c.300 BC. Thus it appears that the Kodiak Archipelago has been subjected to great earthquakes characterised by widespread coseismic subsidence and large run-up tsunamis, which occurred about twice per millennium during the Late Holocene. The coastal subsidence caused by the 1964 earthquake triggered a period of very rapid coastal erosion and shoreline retreat that strongly affected shorelines composed of soft unconsolidated sediments, especially beach berm and dune complexes (Crowell and Mann, 1996). The rapid regression of the shoreline by tens of metres occurred at many places, especially on the northeastern side of the island, immediately following the earthquake. The erosion was caused by increased wave action due to a relatively higher sea level. Distinctive beach ridges and berms composed of cobbles and boulders were redeposited along the new shorelines, burying older sediments and soils. Since the 1964 earthquake, the area that subsided has been rapidly rebounding and the drowned shorelines have been re-emerging (Gilpin et al., 1994a, 1994b; Gilpin, 1995). This uplift is attributed to steady-state slip of the down-dip part of the plate interface below the lower limit of coseismic rupture that is compensating for the slip deficit on the deeper part of the plate interface. The post-1964 uplift has caused the subsided shorelines along the northeastern part of the archipelago to re-emerge, slowing at some locations and stopping the coastal erosion as well as promoting shoreline progradation of new beaches and progradational berms. The new prograding beaches are commonly finer grained and have formed seaward of the cobble-and-boulder berms produced during the immediate post earthquake erosional period. Where the shoreline is now prograding, the cobble-and-boulder berms have stabilised and are becoming vegetated.

178

P. SALTONSTALL AND G.A. CARVER

THE SETTLEMENT POINT SITE The Settlement Point site, situated at an exposed coastal location on the southeastern shore of Afognak Island, was the subject of four field seasons of excavation (1994–98) as part of the Afognak Native Corporation’s ‘Dig Afognak’ programme (Fig. 10.1). The site consists of seven Koniag-era house depressions and an associated shell midden (Fig. 10.2), of which 223 m2 of the site were

Figure 10.1 View of the Settlement Point site with house one excavation in the foreground Note: The dead spruce trees on the far side of beaver pond were killed after their roots were flooded with salt water following subsidence during the 1964 earthquake

EARTHQUAKES AND THE SETTLEMENT POINT SITE

Figure 10.2

179

Settlement Point site

excavated. Parts of six houses were sampled, one house was exposed completely, and 36 m2 of the midden outside the houses were excavated. A suite of a dozen radiocarbon dates from charcoal collected from the house hearths to represent the latest date of occupation shows that they were occupied from c. AD 1200 until c. AD 1550 (Table 10.1). In addition to the archaeological work, the geology of the site was mapped in detail and the late Holocene palaeoenvironment of the area reconstructed. The surficial geology at the site consists predominately of unconsolidated sandy beach ridges of Late Holocene age (Fig. 10.3). During the 1964 earthquake, Settlement Point was lowered about 1.5 m and more than 30 m of shoreline retreat occurred in the following decade. A prominent cobble-and-boulder beach ridge that has buried older sandy beach berms and the bases of trees growing on the older beach sediments marks the landward extent of the post-earthquake coastal erosion. The cobble-and-boulder beach ridge associated with the 1964 subsidence stands 3.64 m above the local high tide level (MHHW) and is now

180

P. SALTONSTALL AND G.A. CARVER

Table 10.1

Radiocarbon dates from the Settlement Point site

Sample ✝

Provenance

Uncalibrated Calibrated with 2 sigma*

High probability with 2 sigma

B101551 B118300 B101552 B114097 B114203 B114202 B114205 B114204 B114096 B114098 B101912 B101913

House 1 hearth House 1 floor House 2 hearth House 3 charcoal filled pit House 4 charcoal filled pit House 5 charcoal filled pit House 6 charcoal filled pit House 7 charcoal filled pit Midden L1 Midden L2G Midden bottom L2 Midden L2D/L2E contact

620 ± 50 BP 570 ± 60 BP 300 ± 50 BP 350 ± 70 BP 330 ± 60 BP 440 ± 60 BP 300 ± 50 BP 450 ± 50 BP 370 ± 80 BP 340 ± 60 BP 440 ± 50 BP 390 ± 50 BP

AD AD AD AD AD AD AD AD AD AD AD AD

✝ *

AD AD AD AD AD AD AD AD AD AD AD AD

1220–1449 1224–1631 1427–1888 1321–1955 1405–1894 1293–1672 1427–1888 1301–1656 1291–1955 1399–1888 1304–1660 1331–1674

1214–1455 (0.98) 1255–1531 (0.90) 1433–1705 (0.75) 1393–1824 (0.89) 1410–1710 (0.77) 1290–1675 (0.97) 1433–1705 (0.75) 1383–1654 (0.90) 1386–1824 (0.86) 1406–1710 (0.79) 1384–1661 (0.92) 1391–1679 (0.94)

Dates from Beta Analytic Laboratory. Calibration using Stuiver and Reimer (1993) with a lab. error multipier of 2.

stable and vegetated. A narrow sandy beach ridge which prograded seaward of the post-1964 earthquake berm is also becoming stabilised and beach sand is accumulating on the presently active beach. These processes reflect the rapid rebound of the area since the earthquake. Landward of the post-1964 beach deposits is an older sequence of sandy prograded beach ridges and behind them is another cobble-and-boulder berm. The sandy prograded beach ridges range from 2.15 to 3.10 m above sea level, with the highest located furthest inland. The crest elevations progressively decrease seaward. The cobble-and-boulder berm behind this sequence of prograded beach ridges is 3.65 m high. These deposits are interpreted to reflect the product of an earlier subduction earthquake cycle. The deposits correlate with similar shoreline features found at many localities along the northeastern side of the Kodiak Islands (Gilpin, 1995). Radiocarbon dates taken from multiple localities in the region place the earthquake at c. AD 1150. Similar evidence for a large subduction earthquake that affected the northern part of Cook Inlet, Prince William Sound, and the Copper River delta, suggest that the AD 1150 earthquake was similar in extent and size to the 1964 earthquake (Plafker, 1969; Plafker et al., 1992; Combellick, 1993; Gilpin, 1995). The earlier earthquake cycle ended c. AD 1550, when there was another large earthquake (Gilpin, 1995). At Settlement Point this earthquake, unlike the AD 1150 and AD 1964 events, is not associated with a cobble-and-boulder berm formed from redeposited eroded materials. Evidence for this subsidence event included a tsunami deposit and palaeo-erosional episode. The final earthquake cycle, dating to c. AD 1550–1964, is associated with the progradational berms immediately adjacent to the modern beach. The Settlement Point village was built on the beach ridges produced in response to the land-level changes from the AD 1150 earthquake and its post-seismic

Figure 10.3 Beach ridges at the Settlement Point site Note: The surface geology of the Settlement Point site is dominated by a sequence of progradational beach ridges that formed across the mouth of a small stream valley in response to land-level changes from three late Holocene subduction earthquakes. The beach ridges include a landward-most cobble berm (B4) produced during a short but intense period of subsidence-induced coastal erosion immediately after the AD 1150 earthquake and a sequence of sand berms (B3) with progressively lower crest elevations which were deposited as the land rebounded following the earthquake. The 1964 earthquake also resulted in subsidence-induced coastal erosion and generated another cobble berm (B2) which is now stable and fronted by a narrow progradational sand berm (B1) and an active sandy beach (S). Erosion of the Settlement Point site (A) by the small tidal stream was caused by the subsidence during the AD 1550 earthquake. The subsequent post-seismic uplift produced a small, elevated terrace (T) along the lower reach of the stream

182

P. SALTONSTALL AND G.A. CARVER

rebound. The earliest house was constructed on top of a newly prograded beach ridge located on the seaward margin of the post-AD 1150 cobble-and-boulder berm. The lack of soil, tephra, and remains of vegetation between the midden associated with this house and the beach sand of the progradational berm show that it was built soon after the berm-and-beach ridge stabilised. Construction of later houses followed at intervals sufficient to allow weak soil development and a thin vegetation mat to form on the berms. Today the site is situated next to a beaver-dammed pond near the mouth of a small tidal stream. A series of younger beach berms separates it from the modern beach. But when people first moved to the site, 700 years ago, they built their houses adjacent to what was then the beach. During the 300 years of Alutiiq habitation, the beach continued to rise slowly above sea level as the subducting Pacific plate deformed the leading edge of the North American plate. We empirically verified the emergence of the site during its occupation, as well as its later submergence following the subsequent subduction earthquake in AD 1550 by carefully measuring the elevation of each house’s floor and its deepest floor feature (Fig. 10.4). A layer of mud covered the floors of two subfloor tunnels that lead to side rooms in the earliest house on the site. The inhabitants had placed slate slabs on top of the mud in an attempt to raise the tunnel floor and mitigate the flooding. This evidence for flooding indicates that the house was built just below the level of the highest high tides since the highly permeable, well-drained beach gravel under the site precludes any other cause of flooding.

Figure 10.4 House floor elevation data from the Settlement Point site in relation to mean high water (MHHW) in 1995 and extreme high water (EHHW) at various points in time Note: The top elevation for each house is the top of the house’s floor while the bottom elevation represents the bottom of the deepest feature: storage pit, side-room tunnel or hearth pit. Because of the extremely permeable beach-gravel substratum, the site is flooded up to the level of extreme high water several times a year. On an average day it is flooded up to the mean high water mark. After the earthquake in AD 1550 every house at the site would have experienced flooding at extreme high tides

EARTHQUAKES AND THE SETTLEMENT POINT SITE

183

Some of the later houses, built at lower levels, do not contain mud deposits. This is evidence the land was uplifted before they were built and that their floors were not flooded by high tides during occupation. The floor in one of the last houses built at the site (house seven) is 30 to 50 cm lower than the floor of the oldest house (house one). Today the floor of house seven is flooded by extreme high tides, as are the subfloor features and tunnels leading to the side rooms of house one. The present tidal flooding of these houses reflects subsidence from the 1964 earthquake. The elevation of house floors and subfloor features at the site relative to sea level was reconstructed for the last three seismic cycles from geomorphic and archaeological evidence (Fig. 10.5). Present sea level relative to the site’s datum is known. The earthquake-generated berm behind the site provides a general proxy for sea level immediately after the AD 1150 earthquake and indicates that at

Figure 10.5 The elevation history of the Settlement Point site for the past 1,000 years includes land-level changes resulting from three large earthquakes Note: About 1 to 1.5 m of coseismic subsidence was associated with each earthquake. During the interseismic part of each cycle, the relatively short interval of rapid post-seismic rebound was followed by a long period of gradual uplift that generated net uplift of the site. The net uplift slightly exceeded the amount of subsidence for each cycle and slightly outpaced the rate of late Holocene sea-level rise. House one was constructed on the first emerging beach ridge a few decades after the AD 1150 earthquake. The timing of the construction of other houses is uncertain but was later than house one. During the AD 1550 earthquake house seven dropped below high tide level and was severely eroded. House seven again subsided below high tide level in 1964 and is presently reemerging

184

P. SALTONSTALL AND G.A. CARVER

that time the entire site would have been submerged. Sea level at the time of construction of house one (probably c. AD 1200) is also known because the house was flooded by extreme high tides while it was occupied. The elevation of the floor of house seven indicates how far the site rebounded above sea level just before the AD 1550 earthquake dropped it into the sea again. During the AD 1550 earthquake, the site subsided about 1 m and was subjected to post-seismic coastal and stream erosion. It is not known how much of the site was lost to erosion, but the extensive midden associated with the village was truncated along its entire seaward margin, destroying most of house seven. Excavation at the site revealed both a palaeo-erosion face cut through house seven, and a stream-cut platform graded to the lower post-seismic sea level, both of which provide a measure of the amount of subsidence and the post-earthquake sea level. We measured this elevation at two points along the seaward side of the site. One measurement was determined at the palaeo-erosion face in the eroded house (house seven) while the other was made in a test pit excavated at the base of the palaeo-erosion face of the shell midden where it abutted the stream-cut platform. The AD 1550 earthquake generated a large tsunami that deposited a fine layer of sand in at least two of the houses (numbers two and six). Analysis of this sand revealed marine diatoms, which supports our interpretation of the sand as tsunami deposits (Carver et al., 1994). The tsunami sand was also found in test pits on the village site and at many other places in the Settlement Point area. A similar layer of fine sand was deposited over the site by the tsunami generated by the 1964 earthquake. A major consequence of the subsidence caused by the AD 1550 earthquake was a raised water table. The extremely permeable gravel-and-sand substratum ensures that the water level in the site is the same as sea level. After subsidence, every house pit at the site was flooded by seepage during high tides. This alone rendered the village uninhabitable. No cultural debris was found above the AD 1550 tsunami deposits and the layer of fine sand was not disturbed at the site. This indicates that the village was abandoned and the people moved elsewhere following the earthquake. Several clusters of late Koniag house pits, which are located near the site, may represent new settlements developed by the refugees from the Settlement Point village. One of these sites, AFG 012, was built in an extremely unfavorable location on top of a glacial till substratum. Excavation of the houses by Megan Partlow in 1997 revealed extensive drainage ditches and one can imagine that people only moved to the site because it was one of the few flat coastal locations in the area high enough above sea level to be habitable after the earthquake. Some of these late Koniag village sites were occupied at the time of the Russians’ arrival in the early nineteenth century and their use persisted into the historic era. Intermittent occupation in the Settlement Point area continued into the twentieth century, but there is no oral history about the site from the time of the AD 1550 earthquake.

EARTHQUAKES AND THE SETTLEMENT POINT SITE

185

AFOGNAK VILLAGE AND THE 1964 EARTHQUAKE Afognak Village, located 5 km across Afognak Bay from Settlement Point, was vacated after the 1964 earthquake. This well-documented event has many parallels with the prehistoric abandonment of Settlement Point. Like Settlement Point in AD 1550, Afognak village was devastated by a tsunami in 1964. Many houses were destroyed and some were even displaced tens of metres from their original foundations. However, not one of the 190 inhabitants died during the episode. Only 18 people lost their lives around the archipelago. Those who died were mostly fishermen at sea, but in shallow water when the tsunami overtook them (Kachadoorian and Plafker, 1967). After the earthquake, the first small tsunami alerted the Afognak villagers and most people headed for high ground, avoiding the larger subsequent tsunamis that destroyed the village. Afognak Village was not rebuilt after the earthquake because the entire area was flooded as a result of nearly 2 m of subsidence (Kachadoorian and Plafker, 1967). The ground was swampy, basements flooded, and wells were contaminated with salt water. The former residents built a new village, Port Lions, 15 km away on the shores of Kizhuyak Bay. Another result of the subsidence at Afognak Village was massive erosion. By 1965 the beach in front of the village had been cut back 3 to 10 m. In places, a large ‘earthquake’ berm had been deposited on top of intact shoreline deposits. If residents had been living directly on the beach, like their prehistoric ancestors, there would have been no evidence of their 100-year-old modern settlement. Since 1964, the Afognak Village area has rebounded almost a metre and a few people have moved back. The events at Afognak Village in 1964 mirror those across the bay at Settlement Point 400 years earlier. This process has occurred repeatedly in the past and will probably occur again in the future.

CULTURAL MITIGATION OF NATURAL DISASTERS Both environmental conditions and societal customs mitigated natural disasters for the Alutiiq people. As a maritime people they were proficient mariners. Large open boats that carried up to 70 people, angyaqs, were used for warfare and long voyages (Lisiansky, [1814] 1968). An early Russian account (Gideon, 1989: 43) documents that 30 angyaqs, each carrying 20 warriors, would go on long-distance raids from Kodiak to the Alaska and Kenai peninsulas or even Prince William Sound. They were so effective at moving large groups of people, even whole villages, that when the Russians first arrived, they had all the angyaqs destroyed to limit Alutiiq resistance to colonisation (Bancroft, 1886). Indeed the Russians found angyaqs ‘superior to their own clumsy boats for trading purposes, and acquired them, . . . as fast as the natives could build them’ (ibid.: 236). There are many indications that the Alutiiq were extremely mobile. Late prehistoric sites contain many off-island trade goods (Knecht, 1995; Steffian, 1992b) such as antler, coal, basalt, chalcedony and ivory that did not occur

186

P. SALTONSTALL AND G.A. CARVER

naturally on Kodiak. Early Russian accounts are full of references to extensive trade between Kodiak and the mainland areas of Alaska in less durable goods such as caribou products, spruce roots, meat and beaver incisors (Lisiansky, [1814] 1968; Davydov, 1977). Furthermore, the distribution of labret styles in the late prehistoric archaeological record indicates that the Alutiiq might have maintained kin relations on either side of Shelikof Strait, the body of water which separates Kodiak from the Alaska Peninsula (Steffian and Saltonstall, 1995). Another consequence of flexible travel is that the Alutiiq did not need to depend on geographic proximity to maintain cohesive social networks. Alutiiq society was spread out over a large area and yet individuals were probably in close contact with relatives who lived many kilometres away. When a disaster affected an Alutiiq village, its inhabitants most probably had connections with relatives who lived in areas less affected by the disaster. Furthermore, since the Alutiiq lived in largely self-sufficient villages, they were not dependent upon an infrastructure that could break down in the event of a large disaster. In this regard, they differ from complex chiefdoms and state-level societies that often experience societal breakdowns after a large natural disaster (see Sheets and Grayson, 1979). The environment of the Kodiak Archipelago also helped to mitigate the effects of natural disasters. Kodiak’s mountainous topography, with its glacially scoured fjords and indented coastline, limited the number of places where people could build villages. Consequently, they lived in population aggregates unevenly distributed along the coastline. Natural disasters would not have affected every village in the same way because villages were widely separated and situated in a diverse array of topographical and environmental settings. During the 1964 earthquake, even in areas that experienced a great deal of subsidence, the tsunamis affected each bay differently due to differences in shoreline topography and near-shore bathymetry (Kachadoorian and Plafker, 1967). The heterogeneous distribution of resources, itself a reflection of the diverse topography, also helped mitigate disasters. A variety of productive and diverse resources created a large suite of alternative subsistence options. In the event of a natural disaster the Alutiiq had recourse to a variety of such options to overcome its effects.

ALUTIIQ DISASTER FOLKLORE In preliterate societies, artwork and folklore were effective means of passing information between generations. This also helped to mitigate the effects of natural disasters (Minc, 1985). The Alutiiq incorporated tales of natural disasters into their folklore (Davis, 1971; Lantis, 1938). Concerning earthquakes, the Alutiiq have a myth in which a shaman loses a son. ‘In his grief he said that the earth [should] quake and the earth quaked’ (Lantis, 1938: 139). Another myth ties earthquakes to Hlam Shua, the highest being, who had a favourite animal that shook the earth during childbirth (ibid.: 139). They had at least two myths about volcanic eruptions (ibid.: 138). In one, invisible men who live inside the earth cause volcanoes to smoke and fire when they were at war. In the other, strong

EARTHQUAKES AND THE SETTLEMENT POINT SITE

187

men who live inside volcanoes cause them to smoke when they cooked their food or heated their baths. There is also archaeological evidence that the Alutiiq recorded stories about disasters. In 1987 archaeologists recovered a painted box panel that depicts a prehistoric volcanic eruption, probably of nearby Augustine volcano from the Karluk One site, a Koniag-era village on the southwestern side of Kodiak Island (Steffian et al., 1996). It also seems to illustrate an associated tsunami that might have been caused by eruption-generated debris avalanches sliding into the ocean. None the less, in comparison with the Aleutians or Cascadia, where large earthquakes are more common, Kodiak does not have a large body of earthquakerelated folklore. Major earthquakes only affected the Alutiiq approximately every 500 years, and tephra stratigraphy shows that there have only been three major ash falls on the archipelago in the last 10,000 years, the most recent and largest in 1912. Alutiiq myths probably relate to the relatively common minor events. None of the myths seems to relate to a particular event. After the 1964 earthquake, interviews with Alutiiq elders did not produce accounts of traditional knowledge that foretold what to do in such a situation. Common sense seems to have prevailed. Villagers headed for high ground after radio warnings, the sea receded, and the first minor tsunami arrived. In sum, while traditional knowledge might not have helped mitigate the effects of a subsidence event, mobility, topography, a diversity of resources, and societal networks spread out over a large region surely did help the Alutiiq weather prehistoric natural disasters.

EFFECT OF EARTHQUAKES ON THE ARCHAEOLOGICAL RECORD In the decade immediately following the 1964 earthquake many of the archaeological sites along the coast in the Kodiak Archipelago were subjected to coastal erosion and some were completely destroyed. At several places beaches were littered with artefacts washed from eroded sites. At Mill Bay, near the town of Kodiak, a very large site has disappeared completely (Donald Clark, pers. comm. 1999). Today, all that remains is a driftwood-covered storm berm. Three sites near Settlement Point (AFG 012, AFG 017 and AFG 018) were cut back as much as 10 to 30 m. Although there has been no systematic survey of site attrition resulting from the erosion associated with the 1964 earthquake, it is apparent that a significant portion of the late prehistoric archaeological record on Kodiak has been eradicated. Across the archipelago, the distribution and severity of coastal erosion were strongly influenced by the amount of subsidence. The most severe erosion occurred along the northeastern side of the islands where the maximum amount of subsidence was centred. The eastern side of the archipelago, where subsidence was minimal, and the far eastern capes on Kodiak and Sitkalidak Island, which lie

188

P. SALTONSTALL AND G.A. CARVER

trenchward of the zero isobase and experienced uplift during the earthquake, were not affected by shoreline retreat. The pattern of subsidence and resulting coastal erosion generated in 1964 is representative of the seismically related processes in AD 1150 and to a lesser extent AD 1550. Afognak Island and the northeastern parts of the Kodiak Archipelago are characterised by truncated sequences and single component sites while other regions such as Sitkalidak Island are not. For example, Fitzhugh (1996) found a continuous sequence of radiocarbon dates from archaeological sites on Sitkalidak Island spanning the last 3,000 years. After each earthquake, people living on Afognak Island and in other areas affected by subsidence had to relocate their villages and, each time, erosion would have destroyed all or part of their old settlement. It is understandable that there are few twelfth-century radiocarbon dates from the region. Because the region has been a focus of Kodiak’s archaeological research, this gap in the record has been over-emphasised in interpretations of cultural development. As archaeological investigation of Kodiak covers more regions, we predict that more sites from this time period will be found. At Settlement Point, the AD 1550 subsidence event triggered erosion that removed a portion of the site, but did not destroy it. Similar limited erosion probably characterised this event across much of the northeast part of the archipelago. That is why there is not as much of an erosional gap associated with this event as there is with the AD 1150 earthquake. Thus the AD 1550 earthquake has not significantly altered the archaeological reconstruction of the region’s prehistory.

CONCLUSION Extremely large subduction zone earthquakes have occurred roughly every 500 years in the Kodiak region. Although these earthquakes and accompanying tsunamis were locally destructive, it appears that relatively few people were killed. During the 1964 earthquake relatively few people died in the island’s villages, even though several villages were completely destroyed. There is no evidence that the earthquakes and tsunamis have directly affected the cultural trajectory of the Alutiiq people. The Settlement Point site shows that the major effect of the subduction earthquakes was to trigger abandonment of some coastal villages and promote resettlement to new sites nearby on the remodelled shoreline. The earthquakes have, however, affected the archaeological record. The change from Kachemak to transitional Koniag traditions reflects a hiatus in the record. This unconformity in the cultural stratigraphy results from the loss of sites to major coastal erosion triggered by the AD 1150 subduction earthquake. Because the shoreline was preferred for settlement, sites occupied immediately before the earthquake and located near the shoreline at the time of the earthquake were particularly affected by the subsidence-generated erosion. Thus evidence of possible incremental change over several centuries before the earthquake is sparse and cultural change appears abrupt in the archaeological record. The lack of a

EARTHQUAKES AND THE SETTLEMENT POINT SITE

189

break in the cultural trajectory associated with the smaller AD 1550 subduction earthquake supports this interpretation. The AD 1550 earthquake also caused flooding due to subsidence and inundation by a tsunami at Settlement Point. We believe that the immediate impact of the two earthquakes was roughly comparable, but that the greater subsidence in AD 1150 caused a more severe erosional episode. The difference between the amount of erosion after each earthquake and not the direct effects of either earthquake explains the archaeological gap associated with the AD 1150 event and the absence of a gap associated with the AD 1550 event. The cultural sequence of the Kodiak Archipelago was not modified by the earthquakes that have repeatedly affected the region because the Alutiiq were well suited to survive their effects. The Alutiiq were a highly mobile maritime people dispersed among many small villages. Their social contacts extended through multiple settlements spread out over a large area. This allowed local relocation from strongly affected sites to others that experienced less severe effects. These contacts helped the Alutiiq to maintain cultural continuity following largescale natural disasters. ACKNOWLEDGEMENTS This work would not have been possible without the generous support of the Afognak Native Corporation. Their ‘Dig Afognak’ ecotourism programme encourages the highest-quality collaborative research. Special thanks also to Polly Saltonstall for her help in editing this chapter; to Rick Knecht who initially helped to formulate the ideas contained in it; and, finally, to Amy Steffian for her insightful comments and editing skills. REFERENCES Bancroft, H.H. (1886) The Works of Hubert Howe Bancroft, Volume XXXIII, History of Alaska 1730–1885. San Francisco: A.L. Bancroft & Company. Bean, T.H. (1892) Report on the Salmon and Salmon Rivers of Alaska. Washington, DC: Government Printing Office. Buck, E.H., Wilson, W.J., Lau, L.S., Liburd, C. and Searby, H.W. (1975) Kadyak a Background for Living. Anchorage: Arctic Environmental Information and Data Center, University of Alaska. Campbell, L.J. (1992) Kodiak. Alaska Geographic 19(3). Carver, G.A., Gilpin, L.M. and Boer, J. (1994) Tsunami deposits from the 1964 Alaskan earthquake on N. E. Kodiak Island, Alaska. Geological Society of America Abstracts with Programs, Annual Meeting 26(7): 529. Clark, D.W. (1979) Ocean Bay: an early north Pacific maritime culture. National Museum of Man, Mercury Series Archaeological Survey of Canada Paper No. 86. Ottawa: National Museum of Man. Clark, D.W. (1984) Prehistory of the Pacific Eskimo region. In D. Dumas (ed.) Handbook of North American Indians, Vol. 5, Arctic, 136–48. Washington, DC: Smithsonian Institution Press.

190

P. SALTONSTALL AND G.A. CARVER

Clark, D.W. (1988) Pacific Eskimo encoded precontact history. In R.D. Shaw, R.K. Harritt and D.E. Dumond (eds) Late Prehistoric Development of Alaska’s Native People, 211–23. Aurora IV. Anchorage: Alaska Anthropological Association. Clark, D.W. (1992) ‘Only a skin boat load or two’: the role of migration in Kodiak prehistory. Arctic Anthropology 29(1): 2–17. Combellick, R. (1993) The penultimate earthquake in southcentral Alaska: evidence from a buried forest near Girdwood, Alaska. Division of Geological and Geophysical Surveys Professional Report 113: 7–15. Crowell, A.L. and Mann, D.H. (1996) Sea level dynamics, glaciers, and archaeology along the central gulf of Alaska Coast. Arctic Anthropology 33(2): 16–37. Davis, N.Y. (1971) The effects of the 1964 Alaska earthquake, tsunami, and resettlement on two Koniag eskimo villages. Unpublished Ph.D. dissertation, Department of Anthropology, University of Washington, Seattle. Davydov, G.I. (1977) Two Voyages to Russian America, 1802–1807. Trans. Colin Bearne and ed. Richard A. Pierce. Ontario: Limestone Press. Donta, C. (1993) Koniag Ceremonialism: An Archaeological and Ethnographic Analysis of Sociopolitical Complexity and Ritual among the Pacific Eskimo. Unpublished Ph.D. dissertation, Department of Anthropology, Bryn Mawr College, Bryn Mawr, Pennsylvania. Dumond, D.E. (1988) The Alaska peninsula as superhighway: a comment. In R.D. Shaw, R.K. Harritt and D.E. Dumond (eds) Late Prehistoric Development of Alaska’s Native People, 379–88. Aurora IV. Anchorage: Alaska Anthropological Association. Dumond, D.E. (1994) The Uyak site in prehistory. In T.L. Bray and T.W. Killian (eds) Reckoning with the Dead: the Larsen Bay Repatriation and the Smithsonian Institution, 43– 53. Washington, DC: Smithsonian Institution Press. Dumond, D.E. (1998) The archaeology of migrations: following the fainter footprints. Arctic Anthropology 35(2): 59–76. Fitzhugh, B. (1996) The evolution of complex hunter-gatherers in the North Pacific: an archaeological case study from Kodiak Island, Alaska. Unpublished Ph.D. dissertation, Department of Anthropology, University of Michigan, Ann Arbor, Michigan. Gideon, H. (1989) The Round the World Voyage of Hieromonk Gideon 1803–1809. Trans. with an introduction and notes by Lydia T. Black, ed. Richard A. Pierce. Kingston, Ontario: University of Alaska-Fairbanks and Limestone Press. Gilpin, L.M. (1995) Holocene paleoseismicity and coastal tectonics of the Kodiak Islands, Alaska. Unpublished Ph.D. dissertation, Department of Geological Sciences, University of California, Santa Cruz, California. Gilpin, L., and Carver, G. (1993) Paleoseismicity of the SW extent of the 1964 Alaskan Earthquake Rupture Zone, Eastern Aleutian Arc, Kodiak Islands, Alaska, (abs). EOS, Transactions of the American Geophysical Union 74(43): 402. Gilpin, L.M., Ward, S., Anderson, R., Moore, J.C. and Carver, G.A. (1994a) Holocene interseismic deformation and stratigraphic modeling of the earthquake cycle, Kodiak Islands, Alaska. U.S. Geological Survey Open File Report 94–176P: 339–44. Gilpin, L.M., Carver, G.A., Ward, S. and Anderson, R.S. (1994b) Tidal benchmark readings and post-seismic rebound of the Kodiak Islands, SW extent of the 1964 Great Alaskan Earthquake rupture. Seismological Research Letters 65(1): 68. Hausler-Knecht, P. (1993) The origins, development, and spread of north Pacific maritime cultures. Paper presented at NSF-JSPS Seminar, Honolulu. Jordan, R.H. and Knecht, R.A. (1988) Archaeological research on western Kodiak Island: the development of Koniag Culture. In R.D. Shaw, R.K. Harritt and D.E. Dumond (eds) Late Prehistoric Development of Alaska’s Native People, 225–306. Aurora IV. Anchorage: Alaska Anthropological Association. Kachadoorian, R. and Plafker, G. (1967) Effects of the earthquake of March 27, 1964 on the communities of Kodiak and nearby islands. U.S. Geological Survey Professional Paper 542–F, F1–F41.

EARTHQUAKES AND THE SETTLEMENT POINT SITE

191

Knecht, R.A. (1995) The Late Prehistory of the Alutiiq People: Culture Change on the Kodiak Archipelago from 1200–1750 AD. Unpublished Ph.D. dissertation, Department of Anthropology, Bryn Mawr College, Bryn Mawr, Pennsylvania. Lantis, M. (1938) The mythology of Kodiak Island, Alaska. The Journal of American FolkLore 51(200): 123–72. Lisiansky, U.F.[1814] (1968) Voyage Around the World in the Years 1803, 1804, 1805, and 1806. New York: De Capo Press. Maschner, H.D.G. (1995) A comment on the Kachemak to Koniag transition on Kodiak Island. Paper presented at the annual meeting of the Alaska Anthropological Association, Anchorage, Alaska. Mills, R.O. (1993) Radiocarbon calibration of archaeological dates from the central gulf of Alaska. Arctic Anthropology 31(1): 126–49. Minc, L.D. (1985) Scarcity and survival: the role of oral tradition in mediating subsistence crises. Journal of Anthropological Archaeology 5: 39–113. Plafker, G. (1969) Tectonics of the March 27, 1964 Alaska earthquake. U.S. Geological Survey Professional Paper 543(I). Plafker, G. and Carver, G.A. (1999) Seismotectonics of the eastern Aleutian subduction zone: an analog for great tsunamigenic earthquakes in southern Cascadia? (abs). Seismological Society of America, Seismological Research Letters 70(1): 245. Plafker, G. and Kachadoorian, R. (1966) Geologic effects of the March 1964 earthquake and associated seismic sea waves on Kodiak and nearby Islands, Alaska. U.S. Geological Survey Professional Paper 543(D): D1–D46. Plafker, G., LaJoie, K.R. and Rubin, M. (1992) Determining recurrence intervals of great subduction zone earthquakes in southern Alaska by radiocarbon dating. In R.E. Taylor, A. Long and R.S. Kra (eds) Radiocarbon after Four Decades: An Interdisciplinary Perspective, 436–53. New York: Springer-Verlag. Price, T.D. and Feinman, G.M. (eds) (1995) Foundations of Social Inequality. New York: Plenum Press. Saltonstall, P. (1997) Archaeology at Settlement Point: 1997 Preliminary Report. Report prepared for the Afognak Native Corporation, Kodiak, Alaska. Saltonstall, P. (1998) Cooking and storage in the Early Koniag period: a view from Settlement Point, Afognak Island. Paper presented at annual meeting of the Alaska Anthropological Association, Anchorage, Alaska. Sheets, P.D. and Grayson, D. (eds) (1979) Volcanic Activity and Human Ecology. Toronto: Academic Press. Steffian, A.F. (1992a) Fifty years after Hrdlicka: further excavation of the Uyak Site, Kodiak Island, Alaska. In R.H. Jordan, F. de Laguna and A.F. Steffian (eds) Contributions to the Anthropology of Southcentral and Southwestern Alaska. Anthropological Papers of the University of Alaska 24: 141–64. Steffian, A.F. (1992b) Archaeological coal in the Gulf of Alaska: a view from Kodiak. Arctic Anthropology 29(2): 111–29. Steffian, A.F. and Saltonstall, P. (1995) Markers of identity: labrets and social evolution on Kodiak Island, Alaska. Paper presented at the 60th Annual Meeting of the Society for American Archaeology, Minneapolis. Steffian, A.F. and Saltonstall, P. (1999) Early prehistoric settlement in Chiniak Bay: a view from Zaimka Mound. Paper presented at annual meeting of the Alaska Anthropological Association, Fairbanks, Alaska. Steffian, A.F. and Saltonstall, P. (in prep.) Archaeology of Zaimka Mound: Early Prehistoric Occupations in Chiniak Bay, Kodiak Archipelago. Technical report prepared for the Lesnoi Native Corporation. Alutiiq Museum, Kodiak. Steffian, A.F., Begét, J. and Saltonstall, P.G. (1996) Prehistoric Alutiiq artifact from Kodiak Island provides oldest documentary record of ancient volcanic eruptions in Alaska. Alaska Volcano Observatory Bimonthly Report 8(2): 13–14.

192

P. SALTONSTALL AND G.A. CARVER

Stuiver, M. and Reimer, P.J. (1993) Extended 14C data base and revised CALIB 3.0 14C age calibration program. Radiocarbon 35: 215–30. Townsend, J.B. (1980) Ranked societies of the Pacific Rim. In Y. Kotani and W.B. Workman (eds) Alaska Native Culture and History, 123–56. Senri Ethnological Series 4. Osaka: National Museum of Ethnology.

11

Natural disasters and culture change in the Shumagin Islands LUCILLE LEWIS JOHNSON

INTRODUCTION Based on the case study reported here, my short answer to the question ‘What effect might a natural disaster have on culture change?’ is ‘Not much’. Catastrophic natural events may include meteor impacts, volcanoes and earthquakes, with their often attendant tsunamis, and cyclonic storms. These are listed in the order of the potential areal disruption they can cause. Thus large meteors and volcanoes can have worldwide environmental effects, earthquakes immediately disrupt no more than several hundred linear miles, while tsunamis caused by undersea earthquakes can affect an entire ocean basin, and cyclonic storms may impact along a narrow path. The effects of natural catastrophes on modern society live up to their name in the short run: Pompeii was destroyed (Allison, Chapter 7), the 1964 Good Friday earthquake in Alaska and its associated tsunami were responsible for 125 deaths and 311 million dollars-worth of damage (USGS, 1999). However, Roman society did not collapse after Pompeii and people are building on the 1964 fault surface in Anchorage. Major seismic events do have important cultural effects in that they have stayed in people’s minds and imaginations – for instance, Pompeii is an important tourist destination and icon of disaster and there are at least 27,500 sites on the Web which refer to the Alaska earthquake of 1964 (Google, 19 July 2000). When does a disaster have a significant effect? If my house is destroyed in an earthquake, it is significant for me and I may lose artefacts I treasure, but I can rebuild. If an Aleutian village is destroyed by an earthquake-induced tsunami, the people who survive will certainly remember it and have to put in many hours of work rebuilding their houses and replacing their tools. The question becomes, what is the cultural salience of these effects and can they be observed in the archaeological record? Certainly, surviving a natural disaster will probably have an important effect on individual psyches and on the mental and emotional states of affected communities (e.g. Chapters 1, 16 and 18), but do these effects translate into changes in such things as artefacts, settlement patterns or land-use practices which will be recognisable in the archaeological record?

194

L.L. JOHNSON

I will explore these issues as they apply to the Shumagin Islands in the Northeast Pacific Ocean, where I have carried out extensive archaeological research (Fig. 11.1). I will argue that in this case the major archaeologically visible effects of seismic catastrophes are positive ones, in that new land is often exposed for settlement. Negative psychic effects are recorded in the folklore, but earthquakes and tsunamis do not seem to have changed the trajectory of cultural development in the Shumagin Islands, although, of course, we cannot know what might have been had these disruptions not happened.

THE ALEUTIAN ISLANDS The Aleutian Islands, the easternmost of which are the Shumagin Islands, form a necklace in the North Pacific Ocean. They link the Alaska Peninsula in the east to the Kamchatka Peninsula in the west and separate the Pacific Ocean from the Bering Sea. Their cool maritime climate is characterised by fog, drizzle and fierce storms. The island vegetation is marine tundra, which contains very few resources of value to human populations. Land foods are limited to a small number of edible roots, shoots and berries and the eggs of marine birds. Not surprisingly, the inhabitants of the Aleutian Islands have always been maritime hunter-gatherers. Major subsistence resources include large and small sea mammals, including whales, sea lions, seals and sea otters; fish, cod being the most important, but halibut, salmon, sea bass and other bottom fish also contributing significantly to the diet; sea birds and their eggs; and shellfish. Other critical

St Lawrence I.

Arctic Ocean

own on

Pribilof Is.

map

as

Area sh

Port M o

Kamchatka Commander Islands

Pacific Ocean

Aleutian Islands

Al

Kamchatka

ka

ller

Bering Sea

Pe n

in

su

Alaska

la

Siberia

Canada

180°

s

The Aleutian Chain Near Islands

Un S im an Shumagin ak ak Islands I. I.

k na

ka as

al

Um

Delarof Islands Andreanof Islands

Kupreanof Pt.

Is x Fo

Kodiak I.

Un

Rat Islands

Islands of the Four Mountains

d lan

I.

I.

0

100

200

300 km

Figure 11.1 The Aleutian Islands chain and location of the Shumagin Islands

DISASTERS AND CULTURAL CHANGE IN THE SHUMAGIN ISLANDS

195

resources include driftwood, Elymus grass for nets and baskets, and terraces into which semisubterranean houses can be dug. The most important of these terrestrial resources is the land itself, specifically beaches on which to land kayaks. Food is only available if hunters and fishermen can get off an island and back on again, and the ability to construct semi-subterranean houses is critical to survival in this land in which there are long, cold winters with substantial storms. The Aleuts used driftwood or whale bones for their rafters and roofed the houses with matting, grass and the earth and sod dug out from the interior. Within this North Pacific region storms are a frequent and dangerous menace. While storms may not qualify as one-off disasters, they are highly dangerous and are perceived as such by the Aleuts. Given their obligatory maritime focus, the Aleut spend much of their time at sea. If the weather prevents them from leaving shore, the whole community starves. If the hunters cannot return with their prey or they get lost or capsized, the people on shore may starve. Writing in 1910, Aleksey Yachmenev told of a trip by kayak from Sanak Island to Unalaska Village, which took place at the end of the annual sea otter hunt. The kayaks ran into storms and fog, the party got separated, and three kayaks with six persons were lost from among a ‘large number’ of kayaks (Bergsland and Dirks, 1990: 304–13). Storms also erode beaches and may degrade landing spots. On the positive side, storms may deposit various items of flotsam and jetsam on beaches, including valuable driftwood and killed whales. In addition to storms, volcanoes and earthquakes are common along the Aleutian Chain. From a human point of view, earthquakes are extremely rapid in comparison to volcanoes. Volcanoes usually rumble and smoke for some period of time before exploding, thus alerting the populace to their unease. Once a volcano erupts, it may continue spewing for days, weeks, months, years or decades. Volcanoes are also visually striking with smoke and fire erupting from their cones and lava running inexorably down their sides. Even when dormant, they stand above the countryside in silent menace. In contrast, an earthquake strikes suddenly and is over in seconds, or minutes at the most. Aftershocks may accompany a large earthquake, but they are gone in a month or so and then the earth is quiescent, although its configuration may have been altered. In terms of their effects on the Aleuts, volcanoes can bury settlements and blanket vegetation in ash, causing their demise. Those that erupt under water or that flow into the oceans can also destroy fish or disrupt their feeding patterns: The causes of the decrease of sea fish are totally unknown. It seems that it must be ascribed to the action and influence of subterranean fire. For example, in the winter of 1825, prior to the eruption of the Unimak Range, cod . . . were floating half dead on the surface of the sea and in great numbers; after the explosion this fish was hardly caught at all even up to 1827. (Veniaminov, [1840] 1984: 39–40) On the other hand, the hot springs associated with volcanoes are used for bathing and cooking, and the decline in plant fertility is temporary, with the final effect

196

L.L. JOHNSON

being enhanced vegetation: ‘in 1826 all crowberry patches were covered with ash, but in 1829 they began to shoot up again and were much improved’ (Veniaminov, [1840] 1984: 109). Earthquakes cause landslides which can bury settlements, cause houses to collapse, and may be associated with tsunamis. The movement of sediment and changes in submarine elevation can disrupt the marine food chain. On the other hand, since the Aleutian Islands are uplifting, earthquakes are also associated with the creation of new land for settlement in the form of uplifted terraces.

CULTURAL SALIENCE OF DISASTER The vast majority of the Aleut tales and narratives collected in 1909–10 by Waldemar Jochelson (Bergsland and Dirks, 1990) concern only human behaviour: people fighting, killing or loving, and transforming themselves magically into various animals in order to undertake the aforementioned activities. Among the 87 narratives, there are only eight that mention natural disasters. In addition to the real storm mentioned above, in one tale a man’s kayak capsizes and kills him in a storm to punish him for not treating it correctly, while another storm is called by a dead man to punish the people of his village. In a variant of the second tale, the dead man causes an earthquake which collapses a cliff on the boats of the village and sinks them. The other tale involving an earthquake relates the story of a man given hunting luck by an octopus. He disobeys the octopus’s rules and the octopus then brings an earthquake and submarine volcano to destroy both him and his entire village. A second volcano tale occurs during a long narrative in which two Tigalda men go to get wives and glory in Koniag country. At the climax of their adventures, two women with vagina dentatas give birth to an army of people whom they proceed to send two by two to be crushed and killed between ‘two mountains that, whenever they knocked against one another, made fire’. Using dead-man’s fat, an important item in Aleut magic, the older brother stops the mountains from erupting and they then collapse. The brothers then kill the two women, claim their wives and return home. In the last two tales the protagonists are the spirits of volcanoes, but they don’t do anything particularly explosive in the tales (Bergsland and Dirks, 1990: narratives 2, 16, 17, 28, 40, 47, 69, 71). There is one major volcano legend reported by Veniaminov ([1840] 1984: 300) in which the volcanoes of Unalaska and Umnak Islands got into a quarrel about who was most powerful and proceeded to blow each other up, hurling fire and rocks at each other. At last only Makushin on Unalaska and Rechesnaia on Umnak were left. The latter lost in single combat and blew itself up, the former calmed down, having no one left to fight. While all the animals in the area of this final battle died, there is no indication that people were hurt. Thus, for the Aleuts, their active and dangerous environment does not seem to have much cultural salience: the evil doings of other humans are much more important to understand and discuss in myth.

DISASTERS AND CULTURAL CHANGE IN THE SHUMAGIN ISLANDS

197

THE SHUMAGIN ISLANDS I would now like to turn to the specific case of earthquakes and their effects on the Shumagin Islands. The Shumagin Islands are the easternmost of the Aleutian Islands and are located south of the Alaska Peninsula and west of the Kodiak Archipelago (Fig. 11.1). This group of small islands stretches southeast from the peninsula towards the Aleutian Trench, where the Pacific plate underthrusts the North American plate. Over the 4,000 years during which the islands are known to have been occupied, they have been subjected to sea-level rise, to gradual shifts in elevation between earthquakes and to substantial uplift during major earthquakes. The overall rate of uplift varies from southeast to northwest, with Unga, the largest island, experiencing an average uplift of 8 metres/millennium since the late Pleistocene period, while the uplift on Chernabura Island to the southeast has barely been sufficient to outpace sea-level rise (Winslow and Johnson, 1989; Johnson and Winslow, 1991). Shumagin seismicity Major earthquakes are quite common in the Shumagin region. The Pacific plate continuously underthrusts the North American plate. However, at some times the plates bind and movement is slowed or stopped. When the bind snaps there is an earthquake, whose magnitude depends upon how much stress has built up during the bound period. In the Shumagin Islands, aseismic periods (i.e. between earthquakes) are characterised by gradual buckling of the islands, resulting in both sinking and rising shorelines. Earthquakes result in uplift throughout the affected area, with large earthquakes causing up to one metre of uplift, while great earthquakes can produce both considerably more uplift and tsunamis (Winslow and Johnson, 1989). Ivan Veniaminov, the first ethnographer of the Aleuts, reports in one record I saw, it says that, ‘On the 11th of July 1788 . . . also on Unga (the largest and northernmost Shumagin Island) there was so strong an earthquake that one could not keep on one’s feet. Many mountains crumbled. Some time after this event there was a terrible flood.’ (Veniaminov, [1840] 1984: 16) Unfortunately, this report does not indicate how much uplift resulted from this event. Margaret Winslow’s analysis of uplifted terraces in the Shumagin Islands has provided a general chronology of earthquakes in the area. A number of events took place between 3,740–3,540 BP. Following this, a series of very large events began about 3,170 BP and continued until 2,980 BP. The area was then quiet until 2,600–2,400 BP, when uplift occurred on several islands. The last prehistoric period of major uplift began about 2,150 BP. Historically recorded earthquakes include the great earthquake of July 1788, referred to above, and large earthquakes in 1857, 1917 and 1946 (Winslow and Johnson, 1989; Winslow, 1991).

198

L.L. JOHNSON

In addition to the landslides and tsunamis mentioned by Veniaminov, earthquakes can also have serious effects on marine productivity, both immediately and over a longer time period. Landslides that reach the sea can result in increased turbidity, which will disrupt shellfish filter feeding, as well as bury shellfish beds outright. The uplift which accompanies earthquakes can move shellfish beds above their acceptable habitation zone. As much of the oceanic food chain depends upon the invertebrates, this loss can have repercussions throughout the system. In addition, fish and sea mammal habitats can also be directly affected. Recovery time for these resources will vary depending upon the severity of the earthquake, the amount of uplift and the presence or absence of local populations to replenish them. On land, the immediate effects of earthquakes are disruptive, with structures collapsing and landslides. However, due to the Shumagin Islands’ position relative to the Alaska subduction zone, the ultimate result is more land because ocean bottom is uplifted to become subaerial terraces and bays become valleys. Archaeological evidence The Shumagin Island archaeological and historical record extends over the past 4,000 years. The Outer Shumagin Islands, from Nagai Island south, have been completely surveyed for archaeological sites, while the Inner Islands, closer to the Alaska Peninsula to the north, have been partially surveyed. Seventy-three archaeological sites have been found and 39 have been radiocarbon dated (Johnson, 1988, 1994; Winslow and Johnson, 1989; Johnson and Winslow, 1991). Extensive excavations have taken place on two of the 39 sites, both of which are located on Chernabura Island, the southernmost of the group. The overall set of dates (Fig. 11.2) falls into two series, which initially appear to demonstrate a hiatus in the occupation of the islands between c. 1,800–2,400 BP. Shumagin Island Unexcavated Site Dates 4500

Radiocarbon Years B.P.

4000 3500 3000 2500 2000 1500 1000 500

XS I -0 40 X SI -0 40 X SI -0 X 40 SI -0 0 XS 6 I -0 4 XS 0 I -0 XS 40 B0 X 14 SI -0 43 X X SI SB -0 0 -0 29 XS 9 ,u Ipp 0 38 er la y XS er I -0 X 14 SI -0 31 X SI -0 46

0

Sites

Figure 11.2 Radiocarbon dates from lowest levels and therefore earliest habitation of sites in the Shumagin Islands

DISASTERS AND CULTURAL CHANGE IN THE SHUMAGIN ISLANDS

199

However, the methods used to sample sites may be responsible for this pattern. Two procedures were used to secure samples for radiocarbon dating. Wherever possible, samples were selected from the lowest level of cultural material exposed in an erosion face. These samples, therefore, reflect the earliest occupation of the site. When the site was not eroded, a 50 cm pit was excavated in the top of a house pit and charcoal was collected from the first cultural layer which appeared. Therefore, these samples reflect the last occupation of the roof of the structure. If these two sets of dates are disaggregated, a very different picture emerges. The earliest inhabitation dates (Fig. 11.3) range from 4,000 to 2,500 BP, while the latest construction dates (Fig. 11.4) run from about 1,700 to 100 BP. Excavated site dates support the 4–3,000 BP initial habitation of the islands, but there is continuity of occupation, particularly at XSI-007, and, therefore, no hiatus in the occupation of Chernabura Island (Fig. 11.5). A series of dates from the basal levels of both

4000 3500 3000 2500 2000 1500 1000 500 0 XS I-0 XP 24 M -0 6 XS 5 I-0 3 XS 0 I-0 XP 30 M -0 XS 40 B02 XS 8 I-0 XS 27 B0 XP 11 M -0 3 XS 8 I-0 0 XS 8 I-0 XS 03 B0 2 XS N B9, 0 lo we 22 rl ay er

Date of Earliest Terrace Exposure

Site Inhabitation Dates 4500

Sites

Figure 11.3 Radiocarbon dates for the basal levels of barabaras (houses) at site XSI-040 represent the earliest dates for house construction

XS I-0 XS 40 I-0 XS 01 I-0 XS 40 I-0 XS 40 I-0 XS 40 I-0 XS 40 IXS 006 B0 XS 34 I-0 XS 40 I-0 XS X 11 I-0 SI11 040 -1 XS 984 B0 XS 14 B0 XS 12 I-0 43

Date of Barabara Roof

End Construction Dates 1800 1600 1400 1200 1000 800 600 400 200 0

Sites

Figure 11.4 Radiocarbon dates from uppermost levels and therefore earliest habitation of sites in the Shumagin Islands

200

L.L. JOHNSON

excavated sites and from both barabaras (houses) of XSI-040 indicate that these sites were also first inhabited in the fourth millennium BP. However, the end construction dates, while they do indicate that a site was occupied at the indicated time, do not provide any necessary indication of the earliest or latest occupation of the site as a whole. This can be seen by examining the set of dates from site XSI-040, one of the excavated sites on Chernabura Island, at which samples were dated from 15 barabara roofs (Fig. 11.6). The last construction dates of barabaras from this site range over 1,400 years, from 100 to 1,500 BP, while excavation data show that the site was first inhabited 3,600 years ago, although there does seem to be a gap in the occupation of XSI-040 between 2,870–1,500 BP. Thus the information currently available suggests a sudden explosion of people into the Shumagin Islands 4,000–3,000 BP and a continuing population presence, at least on Chernabura Island, for the remainder of prehistory. However, there is some evidence that the population might have declined, with some sites being abandoned, between 2,870–1,500 BP. Dates Within Excavated Sites 4000

Radiocarbon Years B.P.

3500 3000 2500 2000 1500 1000 500

I4 0B 26 I4 0B 26 I4 0B 12 I4 0B 1 XS 2 I-0 0 XS 7 I-0 0 XS 7 I-0 07 I4 0B 12 I4 0B 1 XS 2 I-0 07 I4 0B 12 I4 0B 12 I4 0B 12 I4 0B 26

0

Excavation Units

Summary of radiocarbon dates for sites in the Shumagin Islands Radiocarbon Years B.P.

Figure 11.5

XSI-040 Individual Barabara Dates 1600 1400 1200 1000 800 600 400 200 0 1

2

3

4

5

6

7 8 9 Barabaras

10

11

12

13

14

15

Figure 11.6 Radiocarbon dates for roofs of barabaras (houses) at site XSI-040 represent the latest dates for house construction

DISASTERS AND CULTURAL CHANGE IN THE SHUMAGIN ISLANDS

201

Influence of earthquakes on the Shumagin Islanders Having reviewed the seismic and archaeological evidence, what can be said about the influence of earthquakes on human populations in the Shumagin Islands? There clearly was a rapid immigration into the Shumagin Islands, with 21 sites being occupied in the period between 3,700–2,870 BP. The start of this period is roughly coincident with the major uplift episode of 3,740–3,540 BP, suggesting that the effect of this uplift was to substantially increase the available terrace living spaces in the Shumagin Islands. Given the rapid establishment of settlements thereafter, it would seem that there was a reservoir of people outside the Shumagin Islands who were eager to move into this new real estate, either because their initial occupation locales were overcrowded or because these locales had been disrupted rather than enhanced by the seismic activity. The uplift events beginning in 3,170 BP do not seem to have slowed this population influx. The uplift episodes of 2,500–2,400 BP and 2,150 BP occur, unfortunately, in the hiatus in dated sites between those with dates on the first cultural layer and those with dates on barabara roofs. One might hypothesise that these earthquakes were responsible for initiating the erosion of the earlier inhabited sites since later sites are not eroded. The one site which dates within the hiatus has an inhabitation date of 2,390 BP, suggesting that its terrace might have been created during the 2,500–2,400 BP uplift events. As far as villages already settled before these earthquakes are concerned, the excavation evidence indicates that one of the two excavated sites, XSI-007, continued to be occupied throughout these episodes, while the other site, XSI-040, does not have any radiocarbon dates falling between 2,870–1,500 BP. The desertion of XSI-040 may be related to the earthquake cycle. However, this is a period of depopulation on the western Alaska Peninsula as well, suggesting a large-scale regional depopulation in the area. Given the linearity of earthquakes, they are probably not entirely responsible for the dearth of sites during this time, but they may have contributed. Europeans first encountered Shumagin Islanders in August of 1741, when the Russian exploratory vessel St Peter observed three small sites in the outer Shumagin Islands (Stellar, [1743] 1988: 88, 97–107; Golder, 1922: 1: 140–9, 273–5). By the time of the writing of the earliest ethnography of the Aleuts, in the middle of the nineteenth century, the Shumagin Islands were almost entirely depopulated. According to Veniaminov, Previously there were reckoned to be twelve villages here, which were disposed on six islands. As time passed, little by little, the villages dwindled, partly from losses due to civil strife, partly because of the Russians, but most of all from the [ravages of the] Koniags or Kad’iak Islanders, their bitterest enemies. Nowadays only the one island of Unga is inhabited. (Veniaminov, [1840] 1984: 127–8) Unga Village, which was the focus of Shumagin population during Veniaminov’s time and was inhabited up until the 1950s, was rendered economically non-viable following the large earthquake of 1946, which rendered its harbour completely

202

L.L. JOHNSON

unnavigable. At low tide today, the inner harbour, in which ships anchored during the Russian period, is a mud flat. The Chernabura Spit, where both of the excavated sites are located, was a particularly advantageous location for Aleut settlement. First, the spit itself, actually an isthmus connecting the main part of Chernabura Island to a small island to its north, is so situated that it can be approached from almost all points of the compass and, therefore, is accessible for kayak landings whatever the wind direction. Second, both the spit (before cattle destroyed its ground cover) and the surrounding shores have extensive terraces into which barabaras can be dug. Finally, Chernabura is the closest of the Shumagin Islands to the deep waters of the Aleutian Trench. While the resources both of the trench and of the inland sea to the north of Chernabura would be disrupted by earthquakes, the cold water resources, such as cod, would probably be less disturbed and, therefore, should recover more rapidly, giving the Chernabura Islanders an advantage over the residents of the more northerly islands. Late in the prehistoric period Chernabura lost the majority of its population, a trend which continued in the historic period: ‘Formerly silver foxes lived on it and there were two villages. Now, however, there is neither the one nor the other. In summer the sea otter hunters, who voyage out to sea after the sea otters, live here temporarily’ (Veniaminov, [1840] 1984: 132). The reasons for the desertion of the Chernabura Spit mirror those for its former popularity. First, in the late prehistoric period, as noted above, active warfare and raiding broke out along the North Pacific Rim, with particular enmity being felt between Unalaska, to the west of the Shumagins, and Kodiak Island to the east. Neither the Unalaskans nor the Koniag were averse to raiding the Shumagins on their way through, and the position and accessibility of the Chernabura Spit would have made it a prime target. Second, in the historic period, the deep bays of Unga, particularly Unga Harbour until it lifted up and became silted in, were more attractive to deep-draft Russian ships than the open shallow shelf around the Chernabura Spit. Today, the population of the Shumagin Islands is focused around an artificially enhanced harbour at Sand Point on Popof Island.

CONCLUSION In conclusion, the effects of major earthquakes on Shumagin populations and the development of Shumagin societies appear to have been variable. It can be assumed that all Shumagin Islanders would have felt immediate negative effects following large earthquakes, due to the land and seascape changes and the disruption of the food supply. Unfortunately, effects on this time scale are rarely visible in the archaeological record. One might look for sites buried by landslides, but, even if such sites could be located, excavating them would prove difficult. On a longer time scale, both positive and negative effects of earthquakes have been noted. In the outer Shumagins, the earthquakes of 3,740–3,540 BP appear to have lifted terraces around all the islands, providing living space for people who

DISASTERS AND CULTURAL CHANGE IN THE SHUMAGIN ISLANDS

203

rapidly emigrated to inhabit it. On Unga Island in the inner Shumagins, early coastal sites have been lifted well above sea level and thus rendered uninhabitable, while the best harbour of the early historic period has also been deserted due to uplift. For maritime hunter-gatherers like the Shumagin Islanders, earthquakes are terrifying occurrences which may slow cultural development and force populations to move their settlements, but they do not seem to have a major effect on cultural development. Shumagin history, like ours, is more affected by human events than by natural events, however catastrophic. In a larger frame, should we expect one-off catastrophes to have more serious consequences for hunting and gathering societies or for more settled agricultural or industrial societies? While earthquakes and volcanoes can be locally devastating, there is no more reason to expect long-term disturbance to hunter-gatherers than to modern societies. In fact, we might expect the effects to be less severe given the natural mobility of hunter-gatherer societies. Thus, if modern residents of Anchorage, knowing that they live in an earthquake-prone region, still go ahead and rebuild on a 40-year-old earthquake scarp, we should not be surprised that the prehistoric Shumagin Islanders, maritime hunter-gatherers whose major focus was the sea, continued to dig their semi-subterranean houses into appropriate terraces without regard to the potential dangers of their location.

REFERENCES Bergsland, K. and Dirks, M.L. (eds) (1990) Aleut Tales and Narratives, Collected 1901– 1910 by Waldemar Jochelson. Fairbanks: University of Alaska. Golder, F.A. (1922 and 1925) Bering’s Voyages, 2 vols. American Geographical Society Research Series No. 1. New York: American Geographical Society. Google, 19 July 2000. ‘Good Friday Earthquake’: http://www.google.com/search?q= Good+Friday+Earthquake&meta=lr%3D%26hl%3Den Johnson, L.L. (1988) Archaeological surveys of the Outer Shumagin Islands, Alaska, 1984 and 1986. Arctic Anthropology 25(2): 139–70. Johnson, L.L. (1994) Prehistoric settlement patterns and population in the Shumagin Islands. In R.H. Jordan, F. de Laguna and A. Steffian (eds) Contributions to the Anthropology of South Central and Southwestern Alaska, 73–88. Anthropological Papers of the University of Alaska 24(1–2). Anchorage: University of Alaksa. Johnson, L.L. and Winslow, M.A. (1991) Paleoshorelines and prehistoric settlement in the Outer Shumagin Islands, Alaska. In L.L. Johnson (ed.) Paleoshorelines and Prehistory, 171–86. Boca Raton: CRC Press. Stellar, G.W. ([1743] 1988) Journal of a Voyage with Bering, 1741–1742. Ed. O.W. Frost, trans. M.A. Engel and O.W. Frost. Palo Alto: Stanford University Press. USGS (1999) Largest earthquakes in the United States: Prince William Sound, Alaska 1964.14. http://wwwneic.dcr.usgs.gov/neis/eqlists/USA/1964_03_28.html Veniaminov, I. ([1840] 1984) Notes on the Islands of the Unalashka District. Ed. R. Pierce, trans. L. Black and R.H. Geoghegan. Ontario: Limestone. Winslow, M.A. (1991) Modeling paleoshorelines of seismically active coasts. In L.L. Johnson (ed.) Paleoshorelines and Prehistory, 151–69. Boca Raton: CRC Press. Winslow, M.A. and Johnson, L.L. (1989) Human settlement patterns in a tectonically unstable environment: Eastern Aleutian Islands, Alaska. Geoarchaeology 4: 297–318.

12

Horsemen of the Apocalypse: the relationship between severe environmental perturbations and culture change on the north coast of Peru KIMBERLY D. KORNBACHER

THE FOUR ‘NATURAL’ HORSEMEN OF THE APOCALYPSE. Cultural development in the Central Andes cannot be interpreted without correlating the cultural record with geological and climatic change. This region is subject to many of the natural catastrophes that assail other areas of the world, but nowhere else do they occur in such profusion. The disasters range from localized avalanches, or huaycos (floods of liquid mud, usually transporting large boulders), to volcanic eruptions, earthquakes, and El Niño rainfall and drought events that bring severe devastation to vast areas of the Central Andes. (Richardson, 1994: 18–19)

INTRODUCTION Extreme environmental events and processes have not traditionally played a prominent role in our understanding of evolution, and the process itself has generally been considered slow and gradual (Hoffman and Parsons, 1997: 16). Recent studies of non-human organisms have shown, however, that catastrophic events and extreme environmental processes may intensify the effects of natural selection and cause rapid evolutionary change (e.g. Grant, 1986; Gibbs and Grant, 1987; Benton and Grant, 1996). As the number of evolutionary biologists documenting and recognising the importance of extreme environmental change in precipitating extensive evolutionary change increases, the traditional ‘slow and steady’ perspective of evolution is shifting (Gould and Eldridge, 1993; Hoffman and Parsons, 1997). Unfortunately, similar gains are lacking in studies of human prehistory. Perhaps due to our unwillingness to examine the ongoing evolution of our own species, or maybe because of the complexities involved in doing so, the findings of evolutionary biology have not been widely applied to studies of ourselves. Yet archaeologists studying cultural change over a long time scale are uniquely situated to research the nature of evolutionary responses to extreme environmental change in humans (Torrence and Grattan, this volume, Chapter

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 205

1), and such studies will play a major role, I maintain, in the development of our understanding of how and why cultures change. Archaeologists have long noted that the archaeological and geologic record of Peru’s northern north coast (Fig. 12.1) holds evidence of a variety of large-scale disasters such as massive flooding and erosion, dune incursion, tectonic uplift, mass wasting, and other catastrophic events. A number of researchers have even tried to document the general impact of such events on prehistoric culture change in this part of the world (e.g. Paulsen, 1976; Osborn, 1977; Isbell, 1978; Lischka, 1983; Thompson et al., 1994; Thompson, 1995; Fagan, 1999). Some have focused more specifically on certain time periods and attempted to explain

Figure 12.1 Map of the north coast of Peru showing location of river valleys and major Moche archaeological sites

206

K.D. KORNBACHER

particular aspects of the archaeological record of Peru’s north coast (e.g. Moseley, 1978, 1983; Moseley et al., 1981, 1983; Moseley and Deeds, 1982; Schaaf, 1988; Moore, 1991; Shimada et al., 1991; Shimada, 1994a, 1994b). Of particular interest to many of these scholars is the Moche cultural tradition (c. AD 1–700). In the general culture historical scheme of Peru (Rowe, 1962), the span of time encompassed by the Moche cultural remains falls within the Early Intermediate Period and extends into the Middle Horizon. In the local sequence it is divided into Moche Phases I–V (Larco Hoyle, 1946). Now generally believed to be an in-situ local development from earlier traditions (Bawden, 1996; Shimada, 1994b), the prehistoric Moche culture is recognised in the archaeological record of north coastal Peru by a remarkable artistic tradition expressed in a distinctive body of ceramics, textiles, metalwork, and architecture. From c. AD 1 through c. AD 550 (Moche I–IV), changes in settlement, subsistence and technology have been amply documented. Territorial expansion, large-scale trade networks, elaborate funerary activities, and monumental projects that span several centuries reflect a powerful, evolving population. Despite its dynamic aspect, this record does not begin to anticipate the wholesale changes that culminate in the period referred to as the Moche IV–V transformation (c. AD 550), and the subsequent ‘collapse’ of the entire Moche culture c. AD 700. The Moche IV–V transition is characterised by population restructuring (movement of people from the southern valleys), abandonment of vast areas of previously cultivated land in the northern and central valleys, settlement change (aggregation of people in inland rather than coastal settlements), cessation of building in one of the main political centres (Cerro Blanco), and shifts in architecture, iconography and social organisation (Bawden, 1996: 263). Why archaeologists interpret this as a continuous lineage and refer to the post-Moche IV record as ‘Moche V’ rather than assigning a different appellation altogether is primarily due to the persistence of the hallmark artistic tradition. Although truly extensive changes are documented, the continuity of stylistic aspects of the material record, particularly in the ceramic and textile records, is unequivocal – these are the same people (and/or related descendants) but doing things very differently. Thus the Moche archaeological record provides an important opportunity to study human responses to environmental stress over time within a single population or lineage. Documenting the fact that disastrous perturbations occurred and had some effect on the culture-historical trajectory is a crucial first step. We may be unable to progress beyond that point, however, to gain any understanding of the role such events exert on culture change unless we cast the question in theoretical terms and work within the framework of testable evolutionary models. Disparate theoretical perspectives have hindered our ability to generate and incorporate data within an explanatory foundation upon which we can build a cumulative understanding of why cultures change. This chapter is a preliminary attempt to increase our understanding of the effect of catastrophic events and climatic instability on culture change, focusing on the

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 207

north coast of Peru. Beginning with an evolutionary model, I identify a series of testable hypotheses about energy expenditure, reproductive success, and the relation of these factors to environmental variability. After reviewing pertinent aspects of the Moche environmental and archaeological records, I compare the data to the empirical expectations of the hypotheses to assess whether they are consistent, consider any inconsistence, and suggest directions for future application of the evolutionary model to studies of human culture change.

BET-HEDGING: AN EVOLUTIONARY MODEL A great deal of research in evolutionary biology (e.g. Giesel, 1976; Seger and Brockmann, 1987; Charlesworth, 1994; Roff, 1992; Nilsson et al., 1996; McNamara, 1997) indicates that populations exhibiting behaviours that minimise the interannual variance in the survival of offspring have a selective advantage in environments that fluctuate in an extreme and unpredictable manner. Under such conditions, researchers identify a life history trade-off between offspring number and offspring survival. Those individuals that have fewer offspring (lower arithmetic fitness) but – due to extreme and unpredictable perturbations in resource availability – have lower variance in the number of offspring that actually survive to reproductive age will have greater fitness over the long term (higher geometric mean fitness). The effect is termed ‘bet-hedging’ (Gillespie, 1977) and has been documented in a variety of contexts involving both humans (e.g. Westendorp and Kirkwood, 1998; Sterling, 1999) and non-humans (e.g. Boyce and Perrins, 1987). Imagine populations with two different reproductive strategies, A and B, in an environment characterised by severe and unpredictable environmental fluctuations (see Fig. 12.2). Strategy A maximises birth rate, and Strategy B diverts energy from reproduction, resulting in lower fecundity. In a relatively constant and predictable environment or series of years, the strategy that produces the most offspring (Strategy A) will result in the highest rate of reproduction or greatest reproductive fitness. Strategy A will be selectively advantageous under these conditions (Seger and Brockmann, 1987; Madsen et al., 1999), other things being equal. But in a highly variable and uncertain environment, resources may be abundant one year and parents may be unable to feed all of their offspring the next. In such a situation, infant mortality is in large part a function of the number of offspring – the more dependent mouths there are to feed, the higher the mortality rate during times of resource shortfalls. In a variable environment, Strategy B is favoured by selection, since in the long run a reproductive strategy that reduces the variance in numbers of surviving offspring will serve to increase survivorship and allow growth to occur at a faster rate (Fig. 12.2). As noted by Madsen et al. (1999), amplitude and frequency of perturbations are important variables that affect the strength of selection for bet-hedging in a given context. The evolutionary trade-off between the number of offspring produced and the number that survive to reproduce is especially pronounced in

208

K.D. KORNBACHER

Figure 12.2 Bet-hedging model Note: Diagram illustrates the geometric and arithmetic means. Two reproductive strategies are shown, Strategy A with high birth rate (dashed line), and Strategy B with lower birth rate (solid line). The two strategies are displayed over a number of ‘good’ and ‘bad’ years. During good years Strategy A (high birth rate) is more successful (increases at a faster rate) than Strategy B with the lower birth rate. In a random mix of good and bad years (temporally variable environment), the strategy with the high birth rate (A) experiences greater variance in the rate at which it spreads through the population. Compare geometric means of both strategies to see that in a temporally variable environment, Strategy B (low birth rate) has a higher rate of increase over the long term Source:

Redrawn from Madsen et al. (1999: 259)

environments characterised by severe and unpredictable perturbations. Selection for bet-hedging is heightened if extreme perturbations occur irregularly, yet frequently enough to reduce the probability of giving birth and successfully rearing offspring through the most vulnerable years before another event of serious magnitude. The applicability of the bet-hedging model to humans has recently been explored in a variety of prehistoric contexts (Aranyosi, 1999; Dunnell and Greenlee, 1999; Hamilton, 1999; Kornbacher, 1999; Sterling, 1999). As first noted by Robert Dunnell (Dunnell and Wenke, 1980; Dunnell, 1989, 1999), certain activities characteristic of some human populations are energetically ‘wasteful’ in the sense that they do not result in generating food or offspring. Well-known and persistent material consequences of such activities are monumental architecture and the remains of elaborate funerary activities, features often used by archaeologists as proxy measures of social complexity. The decline in frequency or disappearance of archaeological evidence of these activities (commonly regarded as cultural ‘collapse’ or ‘devolution’) frequently inspires archaeologists to invoke severe environmental conditions as ‘explanations’ for culture change. But the mere existence of cultural elaborations is puzzling from an evolutionary standpoint, since the activities that produce them divert energy away from subsistence and reproduction. Such activities are thus apparently contrary to the ‘prime directive’ of evolution, that of increasing reproductive fitness. Put another way, if all the time and energy expended building pyramid mounds or mummifying the dead were

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 209

spent increasing the food supply and reproducing, the rate of population growth would theoretically increase, thereby maximising reproductive fitness. The bethedging model provides a potential explanation for understanding why and under what kind of conditions such ‘non-reproductive’ or ‘wasteful’ behaviours are favoured by natural selection. Apart from reducing variance in reproductive rates over time, such nonreproductive or wasteful (in the evolutionary sense) activities may also be advantageous for reasons involving human motivation and decision-making (e.g. Burger, 1992; Shimada, 1994b; Trigger, 1990). Although multiple mechanisms may be at work in a given situation (Dunnell, 1999: 247), the evolutionary bethedging model of selection for reduced variance in reproductive fitness is the focus here because it does not view human motivation as causal and is potentially testable on an archaeological scale. Some relevant testable hypotheses can be derived from the bet-hedging model. Perhaps most basically, the degree to which energy is diverted from reproductive and subsistence-oriented activities is expected to be high in variable environments when population growth rate is high relative to the number of people that can be supported in a given location with a given technology (carrying capacity). In direct contrast to the pervasive idea that non-reproductive cultural elaborations (e.g. monument-building, burial elaboration) develop and persist in times of plenty, the model predicts that in more stable, predictable environments or with decreases in population relative to carrying capacity (due to population decline, colonisation of a new environment, or development of a new and more productive subsistence technology), diverting energy into non-reproductive activities no longer confers a selective advantage. Decreases in bet-hedging or energetically wasteful behaviours are the expected result (e.g. Dunnell and Greenlee, 1999), since these conditions tend to result in a larger differential between carrying capacity and population. Recent work aimed at further theoretical development of bet-hedging for application to human phenomena (Madsen et al., 1999) uses simple mathematical models and agent-based simulation to derive a number of more specific archaeological expectations pertaining to settlement, subsistence and demographic data. The implications for settlement involve environmental circumscription and the ability (or inability) to migrate. Selection for bet-hedging is expected to be stronger in more highly circumscribed areas in which the probability that a population can ameliorate the effects of shortfalls simply by moving is lower. In terms of subsistence, selection is expected to favor bet-hedging behaviours if subsistence is focused on a few resources and/or if the potential impact of environmental perturbations on the food supply is large and unpredictable. Specialised agricultural subsistence systems tend to create conditions under which archaeologically visible bet-hedging behaviours are selectively advantageous. Demographic expectations derived from the model may be the most direct way to test for the bet-hedging effect. For instance, if bet-hedging behaviours are occurring, mortality profiles of age distributions should display a lower rate of population growth indicated by lower infant mortality and longer life-spans – a

210

K.D. KORNBACHER

more equal ratio of adults to juveniles in the population (Dunnell and Greenlee, 1999: 386; Madsen et al., 1999: 275; Sterling, 1999: 335). Some hypotheses about conditions under which selection is expected to favour the bet-hedging effect are difficult to test archaeologically. In extraordinary circumstances, organisms might divert energy away from ‘wasteful’ activities and from reproduction – to focus exclusively on subsistence and increase or acquire resources necessary for survival (Boag and Grant, 1981, 1984; Grant and Boag, 1980; Grant, 1986; Hoffman and Parsons, 1997). This will be affected by factors such as the percentage of population loss, the productivity of the newly colonised environment if relocation occurred, the degree of population aggregation, the efficiency of the technology, etc. A shift in the amount of energy being expended on non-reproductive activities is certainly expected, but without access to information about these factors (e.g. percentage of population loss) for a prehistoric population, we cannot estimate the magnitude, or in some cases even the direction, of the shift. Due to the aggregate nature of most archaeological remains, we are rarely able to examine the immediate effects of such long-term and/or extreme perturbations. But we can focus on the impacts on the population on a generational scale, and we can start by posing the relevant questions and identifying the information necessary to answer them. To summarise, bet-hedging is an observed consequence of evolution in uncertain, temporally fluctuating environments, the effect of which is a reduction in the variance of offspring survival over time and an increase in geometric mean fitness. Among humans this is accomplished through the diversion of energy into non-reproductive activities that may, under certain conditions, include the production of archaeologically visible monuments and elaborate burial offerings. Demographic consequences of bet-hedging behaviours should be observable in archaeological mortality profiles as increased adult longevity, and a reduction in infant mortality. It is expected that large-scale disasters resulting in population reduction, relocation and/or migration will result in diversion of energy back into reproductive and subsistence-oriented activities and an initial reduction in energy expended on non-reproductive or energetically ‘wasteful’ activities.

THE ENVIRONMENT OF PERU’S NORTH COAST Since I am using bet-hedging to construct a framework for understanding culture change and the model is a source of hypotheses about life-history trade-offs in particular kinds of environments, a crucial aspect of the investigation is the documentation of the environment. Is the Peruvian north coast environment characterised by extreme and severe perturbations? If so, are the perturbations predictable? How extreme are they and how often do they occur? What is the relation between modern conditions and those of the past? Is there evidence of catastrophic events during the prehistoric period of interest? All these questions are addressed in the following discussion, setting the stage for the subsequent analysis of the Moche archaeological record.

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 211

Modern conditions Though most of coastal Peru is dry desert, over 50 rivers flow to the Pacific from the western cordillera of the Andes, forming a series of valley ‘oases’ that punctuate the dry desert terrain (Fig. 12.1). Within the wedge-shaped valleys of variable size and geomorphology, resource availability and quantity of arable land are structured by abrupt changes in altitude (Cardich, 1985; Pulgar Vidal, 1987; Shimada, 1994b). In general, the central and southern coastal valleys are more highly circumscribed than those of the north, with the western slope of the Andes increasing in gradient, elevation and proximity to the ocean as one proceeds from north to south. These differences are quite pronounced even between 7° and 9° S (Burger, 1992: 20; Pulgar Vidal, 1987: 25; Shimada, 1994b: 40; Silverman, 1996). In the highlands, rainfall is strongly seasonal, and occurs mainly between November and April. This is an important limiting variable for agriculturally based subsistence systems on the coast, since it greatly affects the volume of discharge of coastal rivers and the amount of land that can be irrigated. Although flooding may occur in coastal valleys as a result of heavy rainfall in the highlands, a decade or more may pass with no measurable precipitation on the coast. The exception is during El Niño years. El Niño is an anomalous warming phenomenon that directly and most severely affects the coasts of southern Ecuador and northern Peru. El Niño and its atmospheric counterpart, the Southern Oscillation, together known as ENSO (Diaz and Markgraf, 1992; Enfield, 1992) are the cause of periodic climatic perturbations with global-scale effects. Perhaps the most famous aspect of the El Niño component of ENSO is a drastic warming of the normally cool sea-surface temperatures caused by a decline and/ or directional shift in the east–west trade winds. Since the associated increase in sea-surface temperature causes a cessation of the normal cool upwelling conditions and the consequent death of phytoplankton upon which the anchovies and other marine organisms depend, human populations reliant upon marine resources are greatly affected (Arntz et al., 1985; Glynn, 1990). The shifting trade winds also drastically affect local and global climatic regimes. Typically desiccated landscapes are deluged and normally wet areas experience severe drought. Human populations of coastal Ecuador and Peru dependent on agriculture are disastrously affected by El Niño rains that wash out or cut off irrigation canals, drown or bury crop lands and cause massive and catastrophic erosion of the normally dry, unconsolidated sediments. At the time of this writing, inhabitants of coastal Ecuador and Peru are in the process of recovering from the effects of the 1997–98 El Niño, among the most severe in recorded history. Although early warning systems developed after the 1982–83 ENSO recorded elevated sea-surface temperatures well in advance, the strength of the anomaly was underestimated by a factor of three (McPhaden, 1998). Hundreds died, countless acres of farmland were destroyed and thousands of homes lost. Damage far exceeded the one billion dollars recorded in Peru alone for the 1982–83 event. ENSO disturbances are only one source of severe perturbations experienced by inhabitants of the Central Andes. Situated in the subduction zone between the

212

K.D. KORNBACHER

Nazca oceanic plate and the South American continental plate, the Peruvian coastline and adjacent Andes is one of the most seismically active areas of the world (Moseley et al., 1981; Sandweiss, 1986). The 1970 earthquake in Yungay, Peru provides a dramatic modern example. In this single earthquake and associated Huascarán mud avalanche, nearly 70,000 people lost their lives and 500,000 were left homeless. The destruction of 152 highland and coastal towns and cities and over 1,500 smaller villages is attributed to this single catastrophic event (Moseley et al., 1981; Richardson, 1994). Moseley et al. (1981) coined the term REAC (Radical Environmental Alteration Cycles) to describe events that occur on Peru’s north coast as a result of the interaction of these independent tectonic and climatological systems: The tectonic regime is destabilizing, but its effects are generally additive, with uplift expanding the landscape through exposure of bedrock and near-shore sea floors that feed dune fields producing topographic infilling. The climatological regime, specifically El Niño rain, is subtractive, with periodic desert deluges stripping and redistributing soft sediments. Man suffers because the evolutionary pendulum swings sporadically but swiftly from one regime to the other. (Moseley et al., 1981: 237) Historic records indicate that if a severe El Niño event is preceded by a seismic event that results in uplift, the effect is radical landscape alteration and potential impact to human populations that exceeds ‘normal’ severity (Moseley et al., 1981). Rainfall records provide another example of the unpredictable nature and large scale of environmental variability in this region. The average annual rainfall recorded between the years of 1943 and 1970 was 45 mm, with an annual average of 1.7 mm. However, during the 1977 El Niño, 226 mm of rain fell in a three-day period (Nials et al., 1979: 7). In order to understand the impact of such variability on cultural systems, we need to know something about the severity of the perturbations and how often they occur. Amplitude (strength) and frequency of perturbations are the important variables structuring life-history trade-offs in humans and other organisms living in variable environments (Hoffman and Parsons, 1997; Madsen et al., 1999). Moseley et al. (1992) integrated recent El Niño records with seismic data for the period between 1940 and 1983. If the very strong El Niño event of 1997–98 is added to these records, it is apparent that five El Niño events of strong or very strong (Quinn, 1992; Quinn et al., 1987) magnitude and at least three seismic events measuring above 6.5 on the Richter scale occurred between 1940 and 2000. The frequency of disastrous natural events in the north coast region documented within this particular period indicates that a person with a life expectancy of approximately 60 years might experience as many as 12 such perturbations in a lifetime. Even if seismic events are not considered, the frequency of strong El Niño events is quite high. Of course, these events do not occur in a regularly distributed fashion throughout a given time period, sometimes

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 213

occurring in consecutive years and sometimes not for decades. Scientists do not fully understand the oscillatory nature of ENSO occurrences (Enfield, 1989; Anderson, 1992; Cane et al., 1995; Rasmusson et al., 1995; Sandweiss, 1999a). For the single 60-year period considered, the data indicate that modern coastal Peru experiences frequent severe perturbations that are also highly unpredictable. But can the modern history of coastal conditions and catastrophes be extended back to describe prehistoric environmental conditions? The paleoenvironmental record Geological, archaeological and historic records contain ample evidence of largescale environmental perturbations on the north coast of Peru extending back at least to mid-Holocene times (e.g. Rollins et al., 1986; Sandweiss, 1986; DeVries, 1987; Hansen and Rodbell, 1995; Thompson et al., 1995; DeVries et al., 1997; Keefer et al., 1998; Sandweiss et al., 1996, 1997; Rodbell, 1999). The impact of such events on human populations is also indicated and has been the subject of a great deal of research (e.g. Moseley, 1978, 1987; Nials et al., 1979; Moseley et al., 1981, 1983, 1992; Moseley and Deeds, 1982; Ortloff et al., 1982, 1983, 1985; Richardson, 1983; Sandweiss et al., 1983; Craig and Shimada, 1986; Sandweiss, 1986; Moore, 1991; Shimada et al., 1991; Shimada, 1994a, among others). In some cases researchers have been able to document particular events that have rapidly altered the landscape and had devastating effects on irrigation systems and other aspects of human settlement. The paleoenvironmental data most relevant to this inquiry are derived from geomorphologic features, high-resolution proxy records, historic records and the archaeological record. In an innovative piece of work Moseley et al. (1992) used space shuttle and other high-altitude imagery to examine the extremely dynamic nature of the north coast landscape. They demonstrate that the recent formation of beach ridges and dune fields in the Santa Valley region of the north coast (see Fig. 12.1) is the result of the complex interaction of a number of continuous processes of sedimentation (aeolian, marine and fluvial) coupled with the stochastic occurrence of three kinds of natural disasters: earthquakes, El Niño-related flooding and erosion, and dune incursions. Many of Moseley et al.’s (1992) observations about the Santa Valley can be applied to neighbouring valleys as well. For instance, dune incursion and inundation are more likely to occur on southern sides of the north coast rivers. The coastal plains are wider on the south sides, which are more amenable to large-scale irrigation and farming. Human settlements on similarly configured southern alluvial plains in other valleys are thus more likely to be greatly affected by exposure and movement of large sand sheets. Not surprisingly, evidence of large-scale sand movement is indeed well documented on the southern plain of the Moche Valley to the north of Santa (e.g. Moseley and Deeds, 1982; Moseley, 1992). Moseley et al.’s (1992) study of the changing coastline and geomorphology of the Santa Valley reveals the extremely complex nature of the processes operative throughout much of the Holocene occupation of the north coast and continuing

214

K.D. KORNBACHER

to the present day. Other researchers (e.g. Craig and Shimada, 1986; Huckleberry, 1998) have also documented floods interpreted as possible El Niños in the Lambayeque and Moche Valleys. While these findings demonstrate that individual catastrophic events can be identified in the geologic record of north coast Peru, the chronological resolution is quite coarse. The long-term record is also difficult to decipher since much of the empirical evidence for one El Niño flood may be obliterated by the next. In order to evaluate the impact of severe perturbations on past human populations in more definitive terms, a higher resolution record is needed. In recent years advances in method and technique and an increased interest in understanding the historical record of ENSO have resulted in a remarkable and diverse body of paleoenvironmental data (e.g. Thompson et al., 1984, 1985; Baumgartner et al., 1989; Michaelson, 1989; Diaz and Markgraf, 1992 and references therein; Steinitz-Kannan et al., 1992; Rodbell et al., 1999). The ice-core data generated by Lonnie Thompson and colleagues are particularly applicable (e.g. Thompson et al., 1984, 1985, 1986, 1995; Thompson and MosleyThompson, 1987; Michaelson and Thompson, 1992; Thompson, 1995). The ice cores record annual and seasonal variations in chemical and physical constituents, making the precise dating of annual precipitation layers possible. Most work attempting to link the changes in the ice core record with archaeological phenomena has focused on the data from the Quelccaya ice cap (Schaaf, 1988; Shimada et al., 1991; Shimada, 1994a, 1994b; Thompson et al., 1994). Recently, changes in the core data from the Huascarán ice cap in the Cordillera Blanca have also been examined (Davis et al., 1995; Thompson, 1995). Researchers caution that specific inferences about temperature and moisture must be made cautiously (Thompson et al., 1986), but in general, low ice accumulation, high total particle concentration and conductivity, and lower negative oxygen isotope ratios are taken as signatures of El Niño in the ice-core record. Schaaf (1988) provides an annual record of ice accumulation, oxygen isotope ratios, and total particles for the Quelccaya data for a 150-year period between AD 500 and AD 650 (cf. Shimada et al., 1991: 260; Thompson et al., 1992: 316; Shimada, 1994a: 381, 1994b: 125). The years identified as severe El Niño years (in this reconstruction) are AD 600, 612, 650 and 681. Years identified with less certainty are AD 511–12, 546 and 576 (Schaaf, 1988). One of the most extreme climatic perturbations documented in the Quelccaya and Huascarán cores is not an El Niño flood, but a severe and long-term drought. Identified by an increase in total particle concentration (which is correlated with a period of subnormal precipitation), this perturbation spanned over 30 years, from AD 562 to AD 594. Other, less protracted episodes of subnormal precipitation have been identified in the reconstructed annual precipitation curve for Quelccaya (Shimada, 1994b: 125), and can also be identified less precisely in the 10-year moving average data provided in Thompson (1995: 627). Other less severe but notable droughts extended from AD 524 to 540 and from AD 506 to 512. Although the data have a lower chronological resolution, studies of fossil diatoms recovered from lake cores in highland Ecuador support the Quelccaya

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 215

findings (Steinitz-Kannan et al., 1992). Changes in the frequency distribution of two diatom taxa affected in disparate ways by changes in precipitation (measured as changes in conductivity, salinity, and water level) are plotted for the past 2,500 years. Shimada (1994b: 126) provides extrapolated dates for the sequence from Steinitz-Kannan et al.’s (1992) radiocarbon dates. Based on these, the diatom frequency distributions provide an independent line of evidence for a severe drought in the mid–late sixth century. Systematic work focused on the identification of El Niño floods in the archaeological record began with the Chan Chan–Moche Valley Project’s survey and excavations in the late 1960s to mid-1970s. Working with geologists specialising in hydrology and geomorphology, Michael Moseley and his team surveyed and excavated a large number of structures in the Moche Valley (Nials et al., 1979). Through a combination of fieldwork and analyses of aerial photo series, they soon encountered dramatic evidence of flooding that resulted from the severe 1925 El Niño. Realising the potential for understanding past processes and also for mitigating future disastrous impacts on contemporary populations, the team began documenting individual flooding events. Erosional and depositional features and layers of fused mud brick on the largest adobe monument in the Moche Valley have aided the archaeologists in identifying catastrophic El Niño floods and anchoring them (approximately) in time. Work along these lines has continued and recently other major floods have been documented from additional structures in the valley (e.g. Uceda, 1992). Some of the most promising records of catastrophic events have been obtained from archaeological features that hold evidence of seismic occurrences since these records provide more specific information about how the prehistoric human population was affected. Seismic damage to prehistoric structures (Marcus, 1987) and canal systems (Ortloff et al., 1982) correlated with diagnostic ceramics can provide a rough chronology of events. However, as Moseley et al. (1992: 232) point out, unless structures are abandoned and not rebuilt, it is difficult to identify seismic events since rebuilding and subsequent use obscures previous damage, and tectonic events that did not result in uplift may go unrecognised. Moseley and Deeds (1982: 39) have documented a massive invasion of aeolian sand in the Moche Valley sometime in the latter part of the Moche IV occupation (c. AD 550–600). An adobe mud brick platform between the two gigantic huacas in the Moche (Cerro Blanco) site was buried by aeolian sand deposits and capped by 1–2 m of water-deposited silt. Water damage can be clearly seen on one of the gigantic monumental structures, Huaca del Sol, where layers of mud brick are fused together. A residential structure to the southeast of the huge huaca is ‘completely filled with clean aeolian sand’ (Moseley and Deeds, 1982: 37). Moseley and Deeds use burial evidence to argue that the sand accumulated during Phase IV occupation. Historic records are the final source of palaeoenvironmental data considered. While there are no written records until after the Spanish Conquest, we do have access to an alternative body of historic data that can provide information on the occurrence of pre-Columbian ENSO events. William Quinn (Quinn et al., 1987;

216

K.D. KORNBACHER

Quinn, 1992) has used various historic sources, including maximum annual Nile flood data, to reconstruct a history of ENSO events that extends from AD 622 to AD 1522. Each record is accompanied by a strength rating and a confidence rating that is based on the quantity and quality of information ranging from minimal (1) to complete (5). The records reveal events of strength 5 (strong) and confidence rating of 5 in AD 688–9, and AD 694–6. A strong event is recorded for AD 650 at a lower confidence level (3). Although data are lacking for the early part of the time period of interest, the records indicate that the latter part of the seventh century was characterised by extreme climatic perturbations that resulted in large-scale flooding on the north coast of Peru. Based on the varied sources of data discussed, it is apparent that evidence of past floods, drought and seismic events, as well as episodes of dune formation and incursion, can be identified in Peru during the sixth and seventh centuries AD. Geologists can identify individual flooding and seismic events, but these lack precise chronological control. Techniques with increasingly fine (annual) resolution are improving our ability to identify and date specific events with greater precision. Archaeological and historic records, especially used in combination with high-resolution proxy records and geological data, can be a powerful source of information on past perturbations and catastrophic events. Several disparate sources of data (e.g. ice-core and diatom records, erosional and depositional surfaces, buried and damaged archaeological features) indicate that during the latter half of the sixth century to the mid-seventh century AD a series of severe climatic perturbations occurred, including a minimum of six major ENSO events, two large-scale floods (probably associated with two of the ENSO events documented), one major drought, two smaller droughts, and at least one major episode of dune formation/exposure and incursion. There is evidence that at least three large floods occurred in the Moche Valley during the latter part of the seventh century AD. These floods may be related to three ENSO events documented in historic records for that period.

THE ARCHAEOLOGICAL RECORD OF CULTURE CHANGE The environmental data presented have established that the north coast Peru environment is – and was – characterised by extreme and unpredictable perturbations. Normal north coast environmental conditions are consistent with those in which wasteful, non-reproductive behaviours are expected to be selectively advantageous according to hypotheses generated from the bet-hedging model. Within the Moche cultural context the archaeological record is expected to exhibit evidence of this in the form of large monumental constructions, burial elaborations, and other manifestations of non-reproductive energy expenditure. The Moche archaeological record is extremely complex and varied and I necessarily limit the following discussion to evaluating the hypotheses derived from the bet-hedging model. Thus the emphasis is on the general nature of subsistence and subsistence-related technologies, shifts in the energy investment

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 217

represented by monumental architecture and the manufacture and burial of nonutilitarian objects, and changes in demographic mortality profiles. The analysis emphasises the latter part of Moche III, Moche IV and Moche V to ensure adequate coverage of the transitional period of interest. Moche Phases I–IV From at least Salinar (Moseley and Deeds, 1982: 35) and probably earlier Cupinisque times (Shimada, 1994b: 8), populations inhabiting the north coast valleys engaged in large-scale irrigation agriculture of maize, beans, squash, and cotton and other plants. Shellfish, fish and other marine resources supplemented the early agricultural diet, although evidence indicates that the importance of marine resources declined over time (Pozorski, 1976). Pozorski’s (1976) analysis of the Moche I–IV plant remains at Cerro Blanco in the Moche Valley are summarised in Table 12.1 and indicate that three cultigens (maize, squash and lucuma) provided about 88 per cent of the plant diet. Domesticated llama supplied a large proportion of the animal protein consumed from at least Moche I times (Pozorski, 1982: 180). Evidence of new irrigation canal construction, as well as expansion of existing canals and cultivation of larger areas of land, has been documented from archaeological sites in both northern and southern regions of the north coast throughout Moche times until late Moche IV (Nolan, 1980; Moseley and Deeds, 1982; Moseley, 1983; Eling, 1987; Shimada, 1994b). Increasing agricultural productivity is inferred. Researchers estimate that the Mochica cultivated about 25 to 30 per cent more area during Moche III and IV times than is under cultivation today (Kosok, 1965; Nials et al., 1979; Moseley and Deeds, 1982). These subsistence data, coupled with information on canal construction and expansion, are strongly indicative of a developing agroeconomy from Moche I through early Moche IV times (c. AD 1–AD 500) and an increasing dependence on irrigation. In attempting to assess the importance of prolonged drought to the Mochica inhabitants of the north coastal valleys, Shimada (1994b: 129) considers the effect of a reduction in water supply on a modern agricultural community in the Chicama Valley just to the north of Moche (see Fig. 12.1) (Table 12.1). With more land under cultivation during Moche Phase III and IV times than today, he effectively points out that even a small decline in the amount of water reaching the fields would have resulted in a reduction in the amount of food available. As noted by Madsen et al. (1999: 277), selection for ‘wasteful’ traits is expected to be particularly strong under conditions in which small changes in productivity have large impacts. Cerro Blanco (located in the Moche Valley and also known as ‘Moche’) is generally regarded as the capital, as it is the largest and arguably most complex of the contemporaneous sites located within the Moche ‘heartland’ (see Shimada 1994b for a discussion of the recently postulated existence of two semiautonomous regions – a northern and southern sector). Cerro Blanco was a large nucleated settlement, with ceramics spanning Phases I–IV (Larco Hoyle, 1946). Located on the southern margin of the valley, approximately 5.75 km inland from the ocean, the site covers an area of at least 2 km2.

218

K.D. KORNBACHER

Table 12.1 Subsistence data for Moche I–IV (Cerro Blanco) and Moche V (Galindo) occupations in the Moche Valley (compiled from Pozorski, 1976) Cerro Blanco

Zea mays (maize) Phaseolus vulgaris (common bean) Phaseolus lunatus (lima bean) Capsicum sp. (pepper, aji) Cucurbita sp. (squash, calabaza) Persea americana (avocado) Psidium quajava (guava) Lucuma obvata (lucuma) Arachis hypogaea (peanut, mani) Bunchosia armeniaca (cansaboca) Inga feuillei (pacae) Gossypium barbadense (cotton) Lagenaria siceraria (gourd) TOTAL

Galindo

Food volume (cm3)

Contribution to diet (%)

Food volume (cm3)

Contribution to diet (%)

3,756.50 152.00 0.00 540.00 4,000.00 500.00 132.00 2,062.50 22.00 20.00 Present Present Present

33.60 1.40 0.00 4.80 35.80 4.50 1.20 18.40 <0.01 <0.01 Negligible Negligible Negligible

37,195.00 15.00 0.70 0.00 1,400.00 62.50 Present 562.50 1.50 10.00 2.50 Present Present

94.77 0.04 <0.01 0.00 3.57 0.16 Negligible 1.43 <0.01 0.03 <0.01 Negligible Negligible

11,185.00

99.70

39,249.70

100.00

The dominant features of Cerro Blanco are two gigantic adobe platform mounds, or huacas (Huaca del Sol and Huaca de la Luna), which stand 500 m apart and tower over the valley. The site was believed for many years to have been exclusively ceremonial. In the early 1970s, however, researchers of the Chan Chan–Moche Valley Project discovered a large number of residential structures buried beneath the sand in between the two huacas (Moseley and Deeds, 1982; Moseley et al., 1981, 1983). Stratified midden deposits are now known to extend over 6 m below the surface and complement structural evidence of densely packed habitation areas. The largest of the two monuments, Huaca del Sol, is arguably the most massive pre-Columbian solid adobe structure built in the New World (Wilson, 1988: 333), measuring 340 × 160 m and standing over 40 m high (Moseley, 1992: 167). Archaeologists estimate that over 140 million mould-made adobe bricks went into the construction of this truly massive monument (Hastings and Moseley, 1975; Moseley, 1975). Monument construction activities are clearly documented to late Moche IV times (AD 500). Although Moche burial data are known from a number of areas, the majority (nearly 80 per cent) are from just two sites: Cerro Blanco and Pacatnamú (Donnan, 1995: 120). Most of the Moche IV burials excavated so far were found in rectangular chambers created by removing mud bricks from a platform mound or other structure. Alternatively, a rectangular-shaped pit was created and the walls were lined with rock or mud brick set in mud mortar. Burials were almost always extended and included from one to 55 or more ceramic vessels, llamas and various other inclusions. A double burial with 37 ceramic vessels associated was found incorporated within the Huaca del Sol near the summit (Donnan and Mackey, 1978: 92–3).

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 219

Within this environment (characterised by extreme and unpredictable fluctuations) and, given a subsistence system reliant on irrigation agriculture, we expect a significant diversion of energy into non-reproductive activities to be selectively advantageous. The large amounts of energy indicated by the labourintensive construction of solid adobe monuments and the existence of burials that required concerted construction effort and a substantial outlay of energy for the manufacture and inclusion of ceramic vessels and other burial accompaniments, are consistent with the hypotheses derived from the bet-hedging model. The next task is to examine available demographic data and evaluate whether or not they are consistent with the model expectations. To date, the most promising north coast mortuary data for assessing population demographics during Moche III–IV are from the site of Pacatnamú in the Jequetepeque Valley (Donnan and McClelland, 1997; Verano, 1997). Since the sample (Fig. 12.3a) is derived from one cemetery (H45CM1) and appears to represent a single population of Moche III–IV people buried over a relatively short period of time, this is an important assemblage. Whether this sample is actually representative of the entire Pacatnamú population (e.g. random across functionally differentiated groups) is not known, but all age groups and both sexes are represented. Sixty-one individuals make up the Pacatnamú H45CM1 assemblage. Nearly all of these have ceramics and textiles included in the burial, and many have other objects, such as copper, llama, gourds, ornaments and spindle whorls. The results displayed in Fig. 12.3a indicate lower infant–toddler mortality relative to the number of individuals surviving into adulthood and a substantial proportion of individuals who lived to 50 and beyond. ‘Based on its overall age and sex distribution, H45CM1 can be considered a representative mortuary population, although one might expect a larger number of individuals in the 0–0.9 year group’ (Verano, 1997: 191). From an evolutionary perspective the result is not surprising. The profile is basically U-shaped with an approximately equal proportion of infants–toddlers and older adults represented (ibid.: 192). This is entirely consistent with the demographic trends expected if bethedging behaviours are being favoured by natural selection (Madsen et al., 1999; Sterling, 1999). The Moche IV–V ‘cultural transformation’ Late in Moche IV, ‘there was a decline in the technical and artistic qualities of Moche ceramics and the energy and time devoted to individual objects seems to have been reduced’ (Shimada, 1994b: 117). Yet this shift in the quality and character of ceramic vessels ‘does not adequately prepare us for the breadth and depth of cultural changes that transpired within a relatively brief time span. . . .’ (ibid.: 117). Among the changes noted by archaeologists are the abandonment of the southern sector of the Moche territory (from the Cerro Blanco area of the Moche Valley south and including the Nepeña Valley), the virtual abandonment of presumed capital site of Cerro Blanco in the Moche Valley, an unprecedented population aggregation into large urban settlements located at valley necks rather than valley bottoms, and a dramatic change in monument

220

K.D. KORNBACHER

(a)

(b)

Figure 12.3 Mortality profiles showing age at death for different archaeological burial assemblages. (a) a Moche III–IV sample (n = 61) from Pacatnamú in the Jequetepeque Valley; (b) a Moche V sample (n = 15) from Pacatnamú Source:

Based on data presented in Donnan and McClelland (1997); Verano (1997)

character and construction technique (Bawden, 1996; Donnan and Mackey, 1978; Moseley, 1992; Moseley and Deeds, 1982; Shimada, 1994b). Available radiocarbon dates place these changes between AD 500 and AD 600 (Shimada, 1994b), indicating broad temporal correlation between the Moche IV–V transformation and abnormally severe environmental perturbations that occurred throughout most of the sixth century. While radiometric dating of the archaeological evidence of change is still too coarse-grained to allow a firm statement of cause and effect, available data support the hypothesis that climatic factors played a causal role in the Moche IV–V transformation. As Shimada (1994b: 134) notes, it ‘stands on a wide range of internally consistent lines of

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 221

evidence’. Although local effects and responses were varied, the documented climatic perturbations (perhaps the most severe of which is the drought inferred from ice-core records and fossil diatoms to have spanned the period between AD 562 and AD 594) would have affected a wide area indeed. And certainly the changes we know as Moche V were very far-reaching and not confined to a single valley. However, examination of data from a single valley is warranted for direct comparison of pre- and post-disaster adaptations. Focusing again on the Moche Valley, available geoarchaeological evidence suggests a fairly abrupt end to the large and previously dominant Cerro Blanco settlement: Between 500 and 600 AD, when the prosperous [Moche] metropolis was at its height, it experienced an extremely calamitous rainfall and large-scale flooding attributable to El Niño precipitation. Flood waters stripped as much as 4 m of deposit from some occupation surfaces. The scale of erosion was unprecedented in the long occupational and post occupational history of the site. It is unclear whether the unusual magnitude of destruction reflects more than one El Niño event, or if the severity of erosion was exacerbated by prior tectonic activity. Whatever the reason, the surviving population rebuilt the capital. Yet abandonment ensued shortly due to sand encroachment. As dunes migrated into the city, people removed the roofs and the valued contents from the buildings in front of the advancing drifts. The adobe structures are in otherwise pristine condition due to their burial by yellow sands. By 600 AD, all standing architecture was interred, except for the two enormous Huacas. Nearby canal systems and agricultural lands were also engulfed. Much of the lower valley south of the river was deserted as populations resettled inland and across the river to avoid the eolian drifts. (Moseley et al., 1992: 217–18) Based on this description, it is apparent that by c. AD 600, the site of Cerro Blanco was uninhabitable for some period of time. Apart from the documented internment of structures, evidence of the cessation of use of Cerro Blanco is negative: primarily a lack of habitation refuse dating to Moche V (although recently a number Moche V ceramics were recovered from deposits in and around Huaca de la Luna [Uceda et al., 1994]). In addition, the last construction phases of both the Huaca del Sol and the Huaca de la Luna are dated to the end of Moche IV or the beginning of Moche V, an assessment based on changing frequencies (seriated) of adobe brick markings (Hastings and Moseley, 1975; Mackey and Hastings, 1982). If people ceased living at Cerro Blanco c. AD 550–600, a limited number of possibilities exists. The population was either decimated, migrated to other established settlements in the northern valleys (since the southern sector sites show virtually no evidence of Moche V occupation), or colonised new areas within the Moche Valley. There is clear evidence of population relocation and aggregation at the site of Galindo within the Moche Valley beginning c. AD 600.

222

K.D. KORNBACHER

It is important to note there is no evidence to support or refute the hypothesis that the Galindo occupants were the same individuals (or immediate descendants) as those previously living at Cerro Blanco (cf. Shimada, 1994b: 127). The population aggregation at Galindo (together with Pampa Grande to the north) c. AD 570 appears to represent the resettlement of people formerly living not only at Moche, but in the southern valleys of Culebras, Casma, Nepena, Santa, Chao and Viru. The archaeological record of these valleys is characterised by a great deal of Moche I–mid Moche IV ceramic and other material and virtually none dating to late Moche IV and Moche V times (Bawden, 1996). Recall that the differential between carrying capacity and population is hypothesised to be great in a newly colonised area and post-disaster observations of other organisms show major increases in the rate of reproduction under such conditions (Gibbs et al., 1984; Grant, 1986). Thus the derived expectation is that the initial record of a newly established settlement should exhibit little evidence of energy being expended on non-reproductive ‘wasteful’ activities and much greater energy invested in ‘non-wasteful’ reproduction and subsistence-oriented activities, especially if there is evidence of technological advances that might result in increased subsistence productivity. Within the Moche cultural context, the archaeological record is expected to exhibit a decrease in energy expended on monument construction and burial elaboration and an increase in food production activities during the initial post-disaster occupation. Where available, mortality data should indicate an increase in infant mortality and decrease in adult life-span, as selection would favour increased reproductive investment under these conditions. As the principal Moche V settlement within the Moche Valley, the settlement of Galindo is located ~ 25 km inland, 18 km from the site of Moche on the north side of the river. The site is situated at the juncture between the upper and lower courses of the River Moche and covers ~ 6 km2 in surface area. It is roughly three times the size of the Cerro Blanco settlement (Moseley and Deeds, 1982; Wilson, 1988). Radiocarbon dates for Galindo occupation are not abundant (Shimada, 1994b: 4), but the site is known to date almost entirely to Moche Phase V on the basis of extensive ceramic surveys (Bawden, 1982: 289). Sheila Pozorski studied prehistoric subsistence at Galindo and her research indicates that: subsistence patterns closely parallel patterns reconstructed for the earlier Moche center. Again the domesticated llama furnished virtually all the animal protein consumed. The population of Galindo largely consisted of agriculturalists concerned with irrigating and farming local fields. Field crop cultigens were far more abundant than fruits. Of the crop cultigens, the common bean and especially maize were most frequent; therefore agricultural production at Galindo may have focused on these two storable species. (Pozorski, 1982: 181) Increasing specialization is suggested by this statement and by Pozorski’s (1976) data (summarised in Table 12.1). As Table 12.1 indicates, maize accounts for

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 223

nearly 95 per cent of the dietary plant material recovered at Galindo. An increasingly specialised and productive agricultural system developed between the occupation of Cerro Blanco and the settlement of Galindo. This may be an outgrowth of the shift to the valley neck location that characterises other Moche V sites as well. The valley neck location is determined, at least in part, by the increased water control it affords (see Shimada, 1994b: 128 for a similar argument about selection criteria for the location of Pampa Grande), since it is located on a broad alluvial fan just below the valley neck on the north side of the Moche Valley. The drainage basin behind waters the valley and has only the single small exit channel or throat in front of which Galindo is located. Moseley et al. (1981) have recorded a great deal of evidence of flooding, erosion and rebuilding at Galindo, illustrating that ‘normal’ severe and unpredictable conditions prevailed after the establishment of Galindo. There are several distinct topographical areas that comprise the Galindo settlement. Bawden (1982) and Moseley et al. (1981) divide it into two natural terraces or plains (Plain A and Plain B), and two hillside slopes, Cerro Galindo (Hillside A) and Cerro Muerto (Hillside B). Both Hillside A and B and Plain A are large areas of residential occupation. Architectural forms apart from residential structures include both adobe-walled enclosures and huacas. An interesting addition to the domestic storage areas, bins and niches associated with earlier residences documented at Cerro Blanco (Topic, 1982) is a large storage area (interpreted as corporate-controlled). It was built on terraces and ‘protected from general access by flanking gullies and a succession of high walls’ (Bawden, 1996: 291). Enclosures inferred to be llama corrals on the basis of an absence of internal architectural features and the presence of layers of llama manure, pieces of rope, gourd, and cane walling materials are also reported (Bawden, 1982: 309). The main huacas at Galindo are referred to as Platforms A–D. Although some domestic refuse has been documented at Platform A that is thought to be incorporated into an élite or high-ranking residence, no domestic material was recovered from Platforms B, C and D, and the function of these structures is thought to be similar, if more secularly oriented (Bawden, 1977, 1996) to the huacas at Cerro Blanco (Moche). Researchers have commented on the lack of cultural elaboration apparent at Galindo (especially relative to that of the earlier Cerro Blanco), as the following passage illustrates: When the huacas of Galindo are compared with their predecessors in Moche culture, it becomes clear that significant changes have occurred in this architectural form. Earlier huacas generally dominated the countryside in which they stood. In their great size they represented major enterprises, requiring the mobilization of very large work forces for their construction. By contrast, the Moche V huacas at Galindo appear as mere rudimentary forms; even the larger examples, platforms A and B, are minuscule in comparison with the majority of their antecedents. (Bawden, 1982: 295)

224

K.D. KORNBACHER

Though Galindo is approximately three times larger than the earlier settlement at Cerro Blanco, the energy expended on monument construction appears to have been orders of magnitude less than that required to build a single huaca at the earlier site. A cursory calculation of the maximum number of bricks required to build Platforms A–D (Table 12.2) reveals that at least 12 times as many bricks were required to build the Cerro Blanco huacas. As noted by Bawden (1982: 295), ‘The body of the huaca, so massive at Huaca del Sol at Moche, has, in this structure [Platform D at Galindo], shrunk to the bare minimum necessary to achieve the shape required by this architectural form.’ Monumental architecture is but one elaborate aspect of the Moche V occupation at Galindo for which a decrease in energy investment is noted. Only four burials are described in Bawden’s (1977) work, and these are the only Galindo examples listed in Donnan’s (1995: 115) comprehensive list of Moche burials. Generally speaking, Moche V burials are much less common everywhere on the north coast than those from Moche III and IV times (Donnan, 1995). It is reasonable to infer that fewer burials means less energy went into burial ceremonialism during Moche V times, although the lack of burials does not absolutely rule out the possibility that Moche V populations spent a great deal of time and energy on burial ceremonialism that left no physical trace. However, the available evidence thus far supports a hypothesis of reduced energy expended on burial elaboration during Moche V. The paltry number of Moche V burials makes a comparison of degree of energy invested in burial elaboration over time difficult, and the difficulties are only compounded when attempting to understand demographic trends over time. Again we must rely on the Moche V burial data (Fig. 12.3b) from Pacatnamú to provide a comparison to the data in Fig. 12.3a. Only 15 Moche V burials are reported with age estimates from which demographic data can be generated (Donnan and McClelland, 1997). Unfortunately, these are not derived from a single cemetery as are the earlier Moche burials from this site (Donnan and Table 12.2 Estimate of the maximum number of mud bricks required to construct Huacas A–D at Galindo* Huaca

Huaca area ✝

Maximum number of bricks required **

Huaca A Huaca B Huaca C Huaca D TOTAL

50m × 50m × 1.8m 70m × 50m × 1.5m 10m × 10m × 1.3m 40m × 35m × 1.5m

5,332,800 4,666,200 80,040 559,920 10,638,960

* To ensure that the difference between numbers of bricks required for the Moche I–IV (Cerro Blanco) and Moche V (Galindo) is not exaggerated, the size of brick (25 cm × 15 cm × 10 cm) used in the calculation above is derived from the low end of the range of documented Moche brick sizes (McClelland, 1986: 28). ✝ Huaca areas from Bawden (1982). ** Calculation over estimates actual number since it is based on an assumption of straight-sided structures that are constructed on level ground. Actual structures are terraced platform mounds (which decrease in area from bottom to top) and are constructed on a hill slope.

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 225

McClelland, 1997; Verano, 1997), but instead were found in various locations within 1 km of each other. They are considered reliably dated by associations with Moche Phase V ceramics. A single radiocarbon date (1350 ± 80 BP, calibrated AD 600 [Beta-89546]) falls within the expected time period. The mortality profile in Fig. 12.3b indicates a higher proportion of younger individuals than indicated in the Moche III and IV profiles (Fig. 12.3a). Higher infant mortality and shorter life spans appear to characterise the Moche V situation. This is consistent with the hypothesis that population growth rate was higher during this time period and that relatively more energy was expended on reproduction and related activities than during the previous Moche occupation (Westendorp and Kirkwood, 1998; Dunnell and Greenlee, 1999; Madsen et al., 1999; Sterling, 1999). Since the sample is so small and was not obtained from a single cemetery, the distribution should be regarded as suggestive, but inconclusive. More burial data are needed to ‘flesh out’ the Moche V demographic picture and create a mortality profile that would enable a conclusive test of the hypotheses derived from the bet-hedging model.

SUMMARY AND DISCUSSION During Moche I–IV times when environmental perturbations appear to have been within the ‘normal’ range for this region – severe and unpredictable but not catastrophic – information on monument construction, burial elaboration and mortality profiles derived from the archaeological remains at Cerro Blanco in the Moche Valley and from the cemetery at Pacatnamú is consistent with that expected if the diversion of energy from subsistence and reproduction into nonreproductive or ‘wasteful’ activities were favoured by selection. A life-history trade-off involving slowed reproductive rates and lower variance in fecundity over time is indicated. Correlated in time with a series of extreme perturbations that occurred on both local and regional scales during the sixth century AD, is the suite of cultural changes archaeologists refer to as the Moche IV–V cultural transformation. On a local level, evidence indicates that the important Moche site of Cerro Blanco was rendered uninhabitable for an undetermined period of time, most probably as a result of the cumulative effect of the series of catastrophic perturbations indicated by palaeoenvironmental records. Although we currently have no means of assessing the loss of life that occurred during this time, there is evidence of relocation of a number of Moche III–IV sites (including Cerro Blanco) in the archaeological record. As in the valleys in the northern Lambayeque region of the Moche V distribution, the population aggregated in large urban settlements located in inland valley necks. This is the case in the Moche Valley, where the Moche V site of Galindo shows evidence of a post-disaster inception. Whether a portion of a population is decimated or migrates and colonises a new area, the expected result of the increased differential between population (lower) and carrying capacity (higher) from an evolutionary perspective is a

226

K.D. KORNBACHER

decrease in energy expended on non-reproductive or ‘wasteful’ pursuits and an increase in energy going into reproductive and subsistence-oriented activities. Thus a marked decrease in the quantity of energy expended on monumental construction and burial elaboration is hypothesised, and indeed, noted at Galindo and for the Moche V time period in general. Although the available Moche V mortuary data from Pacatnamú (Fig. 12.3b) are suggestive of higher population growth rate (as indicated by higher infant mortality) and shorter life-spans overall, the sample is too small to assess whether it is representative of all age classes and thus cannot confirm or falsify the demographic hypotheses derived from the bet-hedging model. Even for the Moche III–IV Pacatnamú assemblage derived from a single cemetery (Fig. 12.3a) and yielding the best data currently available (Verano, 1997), it is not possible to assert that the sample is random across age classes. Although the low numbers of infants and toddlers are consistent with the hypothesis of lower infant mortality during this time period, the investigators suggest that there may be an area separate from the adult cemetery where infants were buried. Thus I cannot assert that the 61 individuals recovered (Fig. 12.3a) are actually a random sample of the population across all age classes. Ideally, I would have compared mortality profiles between Cerro Blanco and Galindo occupations as the Moche Valley was the location upon which most of this analysis was focused. Unfortunately, despite the large number of burials excavated from Cerro Blanco (Donnan and Mackey, 1978; Donnan, 1995) most are from diverse locations and functional contexts, excavated by different archaeologists and lacking chronological control in many instances. Whether the sample is random with respect to age class or is representative of the Moche III–IV population structure is not known. Even more problematic is the fact that only four Moche V burials were recovered from Galindo (Bawden, 1977; Donnan, 1995), hardly a sound basis for comparison with the earlier Cerro Blanco assemblage. Thus far, then, the available data are consistent with the hypotheses generated from the bet-hedging model, although the mortuary data are problematic. Lest the reader walk away with the notion of a uniform cultural response to catastrophe, however, the variable nature of the north coast archaeological record must be briefly considered. The north coast Peruvian archaeological record provides a remarkable example of variable responses to extreme conditions if one compares two well-studied Moche V sites: the Moche Valley site of Galindo discussed here and Pampa Grande in the Lambayeque region. At the same time people at Galindo were constructing their diminutive huacas and walled residential structures, the residents of Pampa Grande (cf. Shimada, 1994b), arguably the largest nucleation of people known in north coast prehistory, were engaged in a massive construction project that resulted in Huaca Forteleza. The three-tiered huaca is situated within a huge walled enclosure measuring ~600 m by 400 m, distinguished further by a 290 m access ramp leading to the first terrace (Shimada and Cavallaro, 1986: 47; Shimada, 1994b: 145–6). A number of smaller monuments, most of which are larger than any documented at Galindo, were also constructed during the initial occupation.

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 227

Pampa Grande is widely regarded as the Moche V capital (Shimada, 1994b documents the evidence upon which this assessment is based) and appears to be inconsistent with the expectation that non-reproductive energy expenditure should be relatively minimal during this post-catastrophe occupation. An interesting aspect of Huaca Forteleza in this regard, however, is that it is not constructed of solid adobe. Using a technique called ‘chamber-and-fill’ construction, large adobe-walled chambers (sometimes built without mortar) were constructed to form the internal structure of the platform. The chambers were then filled with a mixture of earth, boulders, adobe and clay mortar fragments, domestic refuse, camelid dung and other debris. Evidence indicates Huaca Forteleza is the result of a single construction episode of perhaps as little as 50 years and that nearly 60 per cent of the volume is fill. The level of organisation required to mobilise the large number of people needed to construct Huaca Forteleza in the short amount of time indicated is believed to be greater than that required for the construction of the Moche Valley monuments (Shimada and Cavallaro, 1986: 58). Given the fact that the Cerro Blanco monuments are solid adobe constructions that exhibit at least eight construction episodes over several hundred years, however, the overall investment of time and energy was probably less at Pampa Grande, certainly in terms of the number of generations affected. Thus the construction of Huaca Forteleza indicates a continuation of the cultural tradition of huaca building but lower total energy expenditure than is represented by the early solid adobe monuments at Cerro Blanco. Clearly, however, a far greater amount of non-reproductive energy was expended at Pampa Grande than during the contemporaneous occupation at Galindo. These are fascinating phenotypic variants in evolutionary terms and definitely merit investigation in that context. Time and space limitations unfortunately preclude probing beyond this cursory description. However, the main point has been made: the Galindo, Cerro Blanco and Pacatnamú data presented here only document a small part of the variation represented in the north coast archaeological record during this fascinating and dynamic period of time.

CONCLUSION A rich and varied palaeoenvironmental record documenting a long history of severe and unpredictable perturbations, combined with a distinctive art style (as documented mainly on ceramics, metalwork and textiles) and a plethora of other attributes that reflect historical relatedness, make the north coast Peruvian archaeological record a promising setting in which human responses to disasters can be studied. Evolutionary biologists have established that the rate of evolution of non-human organisms may be very rapid in extremely variable environments. Unfortunately, documentation of this in humans has rarely been attempted or investigated. I contend that before advancing this sort of inquiry, our questions and explanations must be constructed in evolutionary terms, and I have provided an example of how this might be accomplished. Testing hypotheses derived from the bet-hedging model of life-history

228

K.D. KORNBACHER

trade-offs in uncertain environments involves documenting energy expenditure and changes in reproductive rates. It is possible to test such hypotheses with a prehistoric case, but this requires the availability of certain kinds of archaeological data. Subsistence data, records of energy expenditure on non-reproductive activities such as monument construction and burial elaboration, and demographic data derived from mortality profiles have been examined within this evolutionary framework in an attempt to demonstrate the explanatory potential of the approach. But most of these data are lacking in certain respects. In terms of documenting energy expenditure on activities like burial elaboration and monument construction, quantitative data are sorely needed. For example, it is obvious that less energy is spent on burial elaboration during the Galindo Moche V occupation than during the previous Moche phase occupations, but how much less? This is difficult to quantify in terms of the several important variables involved (such as preparation of the corpse, kind of body covering, container or wrapping for the corpse, excavation and preparation of the burial chamber, etc.) since many of the activities involved leave no physical trace or the traces have been obliterated over time. Focusing on a single variable such as ‘burial accompaniments’ and counting numbers or weights of different materials interred with the dead is a possibility, although many perishable items (e.g. feathers) may leave no record. In coastal Peru looters have targeted burials in particular and destroyed much of this aspect of the archaeological record. Energy expended on monument construction is easier to document, especially when the monuments are constructed from bricks or blocks, or other easily quantifiable material (e.g. Sterling, 1999). The calculations in Table 12.2 are an attempt to illustrate how this might be done. These particular figures only yield ordinal information, however, since they are calculated as if the huacas were rectangular blocks. Although the figures provide some index for comparison, ratio-level data are needed to provide a more robust test of the hypotheses. The mortality profiles required to test the expected effect of bet-hedging behaviour on reproductive growth rate (fitness) over time are difficult to generate due to the problems noted above with looting and preservation. In addition, most ‘assemblages’ are composed of burials from a range of locations and may cross-cut functional contexts and even occupations. To assess whether a burial assemblage is representative of a living population in terms of age structure requires establishing that all age classes are equally represented, or more precisely, that the sample is random across all age classes. This is difficult to accomplish with small samples and is often confounded by the fact that infant and toddler skeletal remains are far more susceptible to diagenetic processes than the denser, more robust bones of adults. A final point of note involves the hypothesis of decreasing non-reproductive energy expenditure following a catastrophic perturbation. The expectation is based on an assumption of an increased differential between carrying capacity and population growth rate immediately following a catastrophic event, whether the population was decimated or migrated and recolonised a new area. This assumption is simply a logical extension of the bet-hedging model rather than a

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 229

mathematically based formulation and has not been adequately developed. Observations of the breeding behaviour of non-human organisms (Boag and Grant, 1984; Grant, 1986) provide support for this, but observations of human behaviour in similar situations are lacking. Historic and/or contemporary observations about energy expenditure and rate of reproduction following a disaster might provide a rich source of data from which the hypothesis could be more rigorously tested. Evaluation of the potential of evolutionary models for increasing our understanding of the relationship between severe environmental perturbations and culture change is a wider goal of the volume and a particular objective of this chapter. The fact that the archaeological data are consistent with the hypotheses derived from the bet-hedging model in all aspects examined establishes and underscores the potential of such models. It is clear, however, that the data requirements are substantial for this sort of inquiry. Data quality must be improved to enable more robust tests and significantly advance our knowledge of the relationship between catastrophic environmental events and culture change.

REFERENCES Anderson, R.Y. (1992) Long-term changes in the frequency of occurrence of El Niño Events. In H.F. Diaz and H.F. Markgraf (eds) El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation, 193–200. Cambridge: Cambridge University Press. Aranyosi, E.F. (1999) Wasteful advertising and variance reduction: Darwinian models for the significance of nonutilitarian architecture. Journal of Anthropological Archaeology 18: 356–75. Arntz, W., Landa, A. and Tarazona, J. (1985) El Niño – Su impacto en la fauna marina. Boletin del Instituto del Mar del Peru, Volumen Extraordinario. Baumgartner, T.R., Michaelsen, J., Thompson, L.G., Shen, G.T., Soutar, A. and Casey, R.E. (1989) The recording of interannual climatic change by high-resolution natural systems: tree-rings, coral bands, glacial ice layers, and marine varves. In D. Peterson (ed.) Aspects of Climate Variability in the Pacific and Western Americas, Geophysical Monograph vol. 55, 1–15. Washington, DC: American Geophysical Union. Bawden, G.L. (1977) Galindo and the nature of the Middle Horizon in northern coastal Peru. Ph.D. dissertation, Department of Anthropology, Cambridge: Harvard University. Bawden, G.L. (1982) Galindo: a study in cultural transition during the Middle Horizon. In M.E. Moseley and K.C. Day (eds) Chan Chan: Andean Desert City, 285–320. Albuquerque: University of New Mexico Press. Bawden, G. (1996) The Moche. Cambridge: Blackwell. Benton, T.G. and Grant, A. (1996) How to keep fit in the real world: elasticity analyses and selection pressures on life histories in a variable environment. American Naturalist 147: 115–39. Boag, P.T. and Grant, P.R. (1981) Intense natural selection in a population of Darwin’s finches (Geospizinae) in the Galapagos. Science 214: 82–5. Boag, P.T. and Grant, P.R. (1984) Darwin’s finches (Geospiza) on Isla Daphne Major, Galapagos: breeding and feeding ecology in a climatically variable environment. Ecological Monographs 54: 463–89. Boyce, M.S. and Perrins, C.M. (1987) Optimizing great tit clutch size in a fluctuating environment. Ecology 68, 142–53.

230

K.D. KORNBACHER

Burger, R.L. (1992) Chavin and the Origins of Andean Civilizations. London: Thames and Hudson. Cane, M.A., Zebiak, S.E. and Xue, Y. (1995) Model studies of the long-term behavior of ENSO. In N.R. Council (ed.) Natural Climate Variability on Decade-to-Century Time Scales, 442–56. Washington, DC: National Academy Press. Cardich, A. (1985) The fluctuating upper limits of cultivation in the central Andes and their impact on Peruvian prehistory. Advances in World Archaeology 4: 293–333. Charlesworth, B. (1994) Evolution in Age Structured Populations, 2nd edn. Cambridge: Cambridge University Press. Craig, A.K. and Shimada, I. (1986) El Niño flood deposits at Batán Grande, northern Peru. Geoarchaeology 1: 29–38. Davis, M.E., Thompson, L.G., Mosley-Thompson, E., Lin, P.N., Mikhalenko, N. and Dai, J. (1995) Recent ice-core climate records from the Cordillera Blanca, Peru. Annals of Glaciology 21: 225–30. DeVries, T.J. (1987) A review of geological evidence for ancient El Niño activity in Peru. Journal of Geophysical Research 92: 14, 471–9. DeVries, T.J., Ortlieb, L., Diaz, A., Wells, L. and Hillaire-Marce, C. (1997) Determining the early history of El Niño. Science 276: 965–6. Diaz, H.F. and Markgraf, V. (eds) (1992) El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation. Cambridge: Cambridge University Press. Donnan, C.B. (1995) Moche funerary practice. In T.D. Dillehay (ed.) Tombs for the Living: Andean Mortuary Practices, 111–59. Washington, DC: Dumbarton Oaks. Donnan, C.B. and Mackey, C. (1978) Ancient Burial Patterns of the Moche Valley Peru. Austin: University of Texas Press. Donnan, C.B. and McClelland, D. (1997) Moche burials at Pacatnamú. In C. B. Donnan and G.A. Cock (eds) The Pacatnamú Papers Volume 2: The Moche Occupation, 17–187. Los Angeles: Fowler Museum of Cultural History, University of California. Dunnell, R.C. (1989) Aspects of the application of evolutionary thought in archaeology. In C.C. Lamberg-Karlovsky (ed.) Archaeological Thought in America, 35–49. New York: Cambridge University Press. Dunnell, R.C. (1999) The concept of waste in an evolutionary archaeology. Journal of Anthropological Archaeology 18: 243–50. Dunnell, R.C. and Greenlee, D.M. (1999) Late Woodland Period ‘waste’ reduction in the Ohio River Valley. Journal of Anthropological Archaeology 18: 376–95. Dunnell, R.C. and Wenke, R.J. (1980) An evolutionary model of the development of complex societies. Paper presented at the 1979 annual meeting of the American Association for the Advancement of Science, San Francisco. Eling, H. (1987) The role of irrigation networks in emerging societal complexity during late prehispanic times, Jequetepeque Valley, North Coast, Peru. Ph.D. dissertation, Department of Anthropology, University of Texas, Austin. University Microfilms International, Ann Arbor. Enfield, D.B. (1989) El Niño past and present. Reviews of Geophysics 27: 159–87. Enfield, D.B. (1992) Historical and prehistorical overview of El Niño/Southern Oscillation. In H.F. Diaz and V. Markgraf (eds) El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation, 95–117. Cambridge: Cambridge University Press. Fagan, B. (1999) Floods, Famines and Empires. New York: Basic Books. Gibbs, H.L., Grant, P.R. and Weiland, J. (1984) Breeding of Darwin’s finches at an unusually early age in an El Niño year. Auk 101: 873–4. Gibbs, H.L. and Grant, P.R. (1987) Adult survivorship in Darwin’s ground finch (Geospiza) populations in a variable environment. Journal of Animal Ecology 56: 797–813. Giesel, J.T. (1976) Reproductive strategies as adaptations to life in temporally heterogeneous environments. Annual Review of Ecology and Systematics 7: 57–79. Gillespie, J.H. (1977) Natural selection for variance in offspring numbers: a new evolutionary principle. American Naturalist 111: 1010–14.

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 231

Glynn, P.W. (ed.) (1990) Global Consequences of the 1982–1983 El Niño–Southern Oscillation. Amsterdam: Elsevier. Gould, S.J. and Eldridge, N. (1993) Punctuated equilibrium comes of age. Nature 366: 223–7. Grant, P.R. (1986) Ecology and Evolution of Darwin’s Finches. Princeton: Princeton University Press. Grant, P.R. and Boag, P.T. (1980) Rainfall on the Galapagos and the demography of Darwin’s finches. Auk 97: 227–44. Hamilton, F.E. (1999) Southeastern mounds: examples of evolution in a temporally fluctuating environment? Journal of Anthropological Archaeology 18: 344–55. Hansen, B.C.S. and Rodbell, D.T. (1995) A Late-Glacial/Holocene pollen record from the eastern Andes of northern Peru. Quaternary Research 44: 216–27. Hastings, C.M. and. Moseley, M.E. (1975) The adobes of Huaca del Sol and Huaca de la Luna. American Antiquity 40: 196–203. Hoffmann, A.A. and Parsons, P.A. (1997) Extreme Environmental Change and Evolution. Cambridge: Cambridge University Press. Huckleberry, G. (1998) Evidence for prehistoric flooding in the Moche Valley. Paper presented at the 63rd annual meeting of the Society for American Archaeology, Seattle. Isbell, W.H. (1978) Environmental perturbations and the origin of the Andean state. In C.L. Redman, M.J. Berman, E.V. Curtin, W.T.J. Langhorne, N.W. Versaggi and J.C. Wanser (eds) Social Archaeology: Beyond Subsistence and Dating, 303–13. New York: Academic Press. Keefer D.K., deFrance, S.D., Moseley, M.E., Richardson III, J.B., Satterlee, D.R. and. Day-Lewis, A. (1998) Early maritime economy and El Niño events at Quebrada Tacahuay, Peru. Science 281: 1833–5. Kornbacher, K.D. (1999) Cultural elaboration in prehistoric coastal Peru: an example of evolution in a temporally variable environment. Journal of Anthropological Archaeology 18: 282–318. Kosok, P. (1965) Life, Land, and Water in Ancient Peru. New York: Long Island University Press. Larco Hoyle, R. (1946) A culture sequence for the north coast of Peru. In J.H. Steward (ed.) Handbook of South American Indians, vol. 2, 149–75. Washington, DC: Bureau of American Ethnology, Bulletin 143. Lischka, J.J. (1983) The Niño as a natural hazard; its role in the development of cultural complexity on the Peruvian coast. Natural Hazard Research, Working Paper no. 48. Mackey, C.J. and Hastings, C.M. (1982) Moche murals from the Huaca de la Luna. In M.J. Cordy-Collins (ed.) Pre-Columbian Art History: Selected Readings, 293–312. Palo Alto: Peek Publications. Madsen, M.E., Lipo, C. and Cannon, M. (1999) Fitness and reproductive trade-offs in uncertain environments: Explaining the evolution of cultural elaboration. Journal of Anthropological Archaeology 18: 251–81. Marcus, J. (1987) Late Intermediate Occupation at Cerro Azul, Peru: A Preliminary Report. University of Michigan, Ann Arbor Museum of Anthropology Technical Report No. 20. McClelland, D. (1986) Brick Seriation at Pacatnamú. In C.B. Donnan and G.A. Cock (eds)The Pacatnamú Papers Volume I, 27–46. Los Angeles: Museum of Cultural History, University of California. McNamara, J.M. (1997) Optimal life histories for structured populations in fluctuating environments. Theoretical Population Biology 51: 94–108. McPhaden, M. (1998) Quoted in Advance Warning by Mark Hoover. Tracking El Nino (NOVA Online) produced for PBS by WGBH, Boston. Michaelson, J. (1989) Long-period fluctuations in El Niño amplitude and frequency reconstructed from tree-rings. In D.H. Peterson (ed.) Aspects of Climate Variability in the

232

K.D. KORNBACHER

Pacific and the Western Americas (Geophysical Monograph vol. 55), 69–74. Washington, DC: American Geophysical Union. Michaelson, J. and Thompson, L.G. (1992) A comparison of proxy records of El Niño/ Southern Oscillation. In H.F. Diaz and V. Markgraf (eds) El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation, 323–48. Cambridge: Cambridge University Press. Moore, J.D. (1991) Cultural responses to environmental catastrophes: post-El Niño subsistence on the prehistoric north coast of Peru. American Antiquity 2: 27–47. Moseley, M.E. (1975) Prehistoric principles of labor organization in the Moche Valley, Peru. American Antiquity 40: 191–6. Moseley, M. (1978) An empirical approach to prehistoric agrarian collapse: the case of the Moche Valley, Peru. In N.L. Gonzalez (ed.) Social and Technological Management in Dry Lands, 9–43. Boulder: Westview Press. Moseley, M.E. (1983) The good old days were better: agrarian collapse and tectonics. American Anthropologist 85: 773–99. Moseley, M.E. (1987) Punctuated equilibrium: searching the ancient record for El Niño. The Quarterly Review of Archaeology 8: 7–10. Moseley, M.E. (1992) The Incas and Their Ancestors. London: Thames and Hudson. Moseley, M.E. and Deeds, E. (1982) The land in front of Chan Chan: agrarian expansion, reform, and collapse in the Moche Valley. In M.E. Moseley and K.C. Day (eds) Chan Chan: Andean Desert City, 25–53. Albuquerque: University of New Mexico Press. Moseley, M.E., Feldman, R.A. and Ortloff, C.R. (1981) Living with crises: human perception of process and time. In M. Nitecki (ed.) Biotic Crises in Ecological and Evolutionary Times, 231–67. New York: Academic Press. Moseley, M.E., Feldman, R.A., Ortloff, C.R. and Narvaez, A. (1983) Principles of agrarian collapse in the Cordillera Negra, Peru. Annals of the Carnegie Museum 52: 299–327. Moseley, M.E., Wagner, D. and Richardson, J.B. III (1992) Space shuttle imagery of recent catastrophic change along the arid Andean coast. In L.L. Johnson (ed) Paleoshorelines and Prehistory: An Investigation of Method, 215–35. Boca Raton: CRC Press. Nials, F.L., Deeds, E.E., Moseley, M.E., Pozorski, S.G., Pozorski, T. and Feldman, R.A. (1979) El Niño: the catastrophic flooding of coastal Peru. Field Museum of Natural History Bulletin 50: 4–14. Nilsson, P., Tuomi, J. and Astrom, M. (1996) Bud dormancy as a bet-hedging strategy. American Naturalist 147: 269–81. Nolan, J.L. (1980) Pre-Hispanic irrigation and polity in the Lambayeque sphere, Peru. Ph.D. dissertation, Department of Anthropology, Columbia University, New York. University Microfilms International, Ann Arbor. Ortloff, C.R., Feldman, R.A. and Moseley, M.E. (1985) Hydraulic engineering and historical aspects of the Pre-Columbian intravalley canal systems of the Moche Valley, Peru. Journal of Field Archaeology 12: 77–98. Ortloff, C.R., Moseley, M.E., and Feldman, R.A. (1982) Hydraulic engineering aspects of the Chimu Chicama–Moche intervalley canal. American Antiquity 47: 572–95. Ortloff, C.R., Moseley, M.E. and Feldman, R.A. (1983) The Chicama–Moche intervalley canal: social explanations and physical paradigms. American Antiquity 48: 375–89. Osborn, A.J. (1977) Strandloopers, mermaids, and other fairy tales: ecological determinants of marine resource utilization - the Peruvian case. In L.R. Binford (ed.) For Theory Building in Archaeology, 157–243. New York: Academic Press. Paulsen, A.C. (1976) Environment and empire: climatic factors in prehistoric Andean culture change. World Archaeology 8: 121–32. Pozorski, S. (1976) Prehistoric subsistence patterns and site economics in the Moche Valley, Peru. Ph.D. dissertation, Department of Anthropology, University of Texas, Austin. University Microfilms International, Ann Arbor. Pozorski, S. (1982) Subsistence systems in the Chimu state. In M.E. Moseley and K. Day (eds) Chan Chan: Andean desert city, 177–96. Albuquerque: University of New Mexico Press.

ENVIRONMENTAL PERTURBATIONS ON THE NORTH COAST OF PERU 233

Pulgar Vidal, J. (1987) Geografia del Peru. Lima: Promoción Editorial Inca. Quinn, W.H. (1992) A study of Southern Oscillation-related climatic activity for AD 622–1990 incorporating Nile River flood data. In H.F. Diaz and V. Markgraf (eds) El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation, 119–49. Cambridge: Cambridge University Press. Quinn, W.H., Neal, V.T. and Antunez de Mayolo, S.E. (1987) El Niño occurrences over the past four and a half centuries. Journal of Geophysical Research 92: 14, 449–61. Rasmusson, E.M., Wang, X. and Ropelewski, C.F. (1995) Secular variability of the ENSO cycle. In N.R. Council (ed.) Natural Climate Variability on Decade-to-Century Time Scales, 458–70. Washington, DC: National Academy Press. Richardson, J.B., III (1983) The Chira beach ridges, sea level change, and the origins of maritime economies on the Peruvian coast. Annals of the Carnegie Museum 52: 265–76. Richardson, J.B., III (1994) People of the Andes. Montreal and Washington, DC: St Remy Media and the Smithsonian Institution. Rodbell, D.T., Seltzer, G.O., Anderson, D.M., Abbott, M.B., Enfield, D.B. and Newman, J.H. (1999) A ~15,000-year record of El Niño-driven alluviation in southwestern Ecuador. Science 283: 516–20. Roff, D. (1992) The Evolution of Life Histories. London: Chapman and Hall. Rollins, H.B., Sandweiss, D. and Richardson, J.B. III (1986) The birth of El Niño: geoarchaeological evidence and implications. Geoarchaeology 1: 3–15. Rowe, J.H. (1962) Stages and periods in archaeological interpretation. Southwestern Journal of Anthropology 18: 40–54. Sandweiss, D.H. (1986) The beach ridges at Santa, Peru: El Niño, uplift, and prehistory. Geoarchaeology 1: 17–28. Sandweiss, D. (1999) El Niño recurrence intervals and monumental architecture on the Peruvian coast. Paper presented at the 64th annual meeting of the Society for American Archaeology, Chicago. Sandweiss, D., Rollins, H.B. and Richardson, J.B. III (1983) Landscape alteration and prehistoric human occupation on the north coast of Peru. Annals of the Carnegie Museum 52: 277–98. Sandweiss, D.H., Richardson, J.B. III, Reitz, E.J., Rollins, H.B. and Maasch, K.A. (1996) Geoarchaeological evidence from Peru for a 5000-year BP onset of El Niño. Science 273: 1531–3. Sandweiss, D.H., Richardson, J.B. III, Reitz, E., Rollins, H.B. and Maasch, K.A. (1997) Determining the early history El Niño. Science 276: 966–7. Schaaf, C.B. (1988) Establishment and demise of Moche V: assessment of the climatic impact. Unpublished M.A. thesis, Harvard University Extension School, Cambridge: Harvard University. Seger, J. and Brockmann, H.J. (1987) What is bet-hedging? In H.J. Harvey and L. Partridge (eds) Oxford Surveys in Evolutionary Biology, 182–211. Oxford: Oxford University Press. Shimada, I. (1994a) Los modelos de organización sociopolitica de la cultura Moche: nuevos datos y perspectiva. In S. Uceda and E. Mujica (eds) Moche Propuestas y Perspectivas, 359–87. Lima: Universidad Nacional de La Libertad-Trujillo. Shimada, I. (1994b) Pampa Grande and the Mochica Culture. Austin: University of Texas Press. Shimada, I. and Cavallaro, R. (1986) Monumental adobe architecture of the late preHispanic northern north coast of Peru. Journal de la Societe des Americanistes 71: 41–78. Shimada, I., Schaaf, C.B., Thompson, L.G. and Mosley-Thompson, E. (1991) Cultural impacts of severe droughts in the prehistoric Andes: application of a 1,500–year ice core precipitation record. World Archaeology 22: 247–70. Silverman, H. (1996) The Formative Period on the south coast of Peru: a critical review. Journal of World Prehistory 10: 95–146. Steinitz-Kannan, M., Nienaber, M.A. and Riedinger, M.A. (1992) The fossil diatoms of

234

K.D. KORNBACHER

Lake Yambo, Ecuador: a 2500–year record on intense El Niño events. In L. Ortlieb and J. Machare (eds) Paleo ENSO Records, International Symposium: Extended Abstracts, 295–8. Lima: ORSTOM and CONCYTEC. Sterling, S.L. (1999) Mortuary profiles as indicators of slowed reproductive rates: evidence from ancient Egypt. Journal of Anthropological Archaeology 18: 319–43. Thompson, L.G. (1995) Late Holocene ice core records of climate and environment from the tropical Andes, Peru. Bulletin de l’Institut Français des Études Andines 24(3): 619–29. Thompson, L.G., Davis, M.E. and Mosley-Thompson, E. (1994) Glacial records of global climate: a 1500-year tropical ice core record of climate. Human Ecology 22: 83–95. Thompson, L.G. and Mosley-Thompson. E. (1987) Evidence of abrupt climatic change during the last 1,500 years recorded in ice cores from the tropical Quelccaya ice cap, Peru. In E. Berger and L.D. Labeyrie (eds) Abrupt Climatic Change: Evidence and Implications, 99–110. Dordrecht: D. Reidel Publishing Co. Thompson, L.G., Mosley-Thompson, E. and Arnao, B.M. (1984) El Niño-Southern Oscillation events recorded in the stratigraphy of the tropical Quelccaya ice cap, Peru. Science 226: 50–3. Thompson, L.G., Mosley-Thompson, E., Bolzan, J.F. and Koci, B.R. (1985) A 1500year record of tropical precipitation in ice cores from the Quelccaya ice cap, Peru. Science 229: 971–3. Thompson, L.G., Moseley-Thompson, E., Dansgaard, W. and Grootes, P.M. (1986) The ‘Little Ice Age’ as recorded in the stratigraphy of the tropical Quelccaya ice cap. Science 234: 361–4. Thompson, L.G., Mosley-Thompson, E., Davis, M.E., Lin, P.N., Henderson, K.A., Cole-Dai, J., Bolzan, J.F. and Liu, K.B. (1995) Late glacial stage and Holocene tropical ice core records from Huascarán, Peru. Science 269: 46–50. Thompson, L.G., Mosley-Thompson, E. and Thompson, P.A. (1992) Reconstructing interannual climate variability from tropical and subtropical ice-core records. In H.F. Diaz and V. Markgraf (eds) El Niño: Historical and Paleoclimatic Aspects of the Southern Oscillation, 295–322. Cambridge: Cambridge University Press. Topic, T. (1982) The Early Intermediate Period and its legacy. In M.E. Moseley and K.C. Day (eds) Chan Chan: Andean Desert City, 255–84. Albuquerque: University of New Mexico Press. Trigger, B. (1990) Monumental architecture: a thermodynamic explanation of symbolic behaviour. World Archaeology 22: 119–32. Uceda, S. (1992) Evidencias de grandes precipitaciones en diversas etapas constructivas de la Huaca de la Luna, costa norte de Perú. In L. Ortlieb and J. Macharé (eds) PaleoENSO Records, International Symposium: Extended Abstracts, 315–18. Lima: ORSTOM and CONCYTEC. Uceda, S.R., Morales, G., Canziani, J.A. and Montoya, M.V. (1994) Investigaciones sobre la arquitectura y relieves policromos en la Huaca de la Luna, valle de Moche. In S. Uceda and E. Mujica (eds) Moche: Propuestas y Perspectivas, 251–305. Trujillo: Universidad Nacional de la Libertad. Verano, J.W. (1997) Physical characteristics and skeletal biology of the Moche population at Pacatnamú. In C.B. Donnan and G.A. Cock (eds) The Pacatnamú Papers Volume 2: The Moche Occupation, 189–214. Los Angeles: Fowler Museum of Cultural History, University of California. Westendorp, R.G.J. and Kirkwood, T.B.L. (1998) Human longevity at the cost of reproductive success. Nature 396: 743–6. Wilson, D. (1988) Prehispanic Settlement Patterns in the Lower Santa Valley, Peru: A Regional Perspective on the Origins and Development of Complex North Coast Society. Washington, DC: Smithsonian Institution Press.

13

Climatic change, flooding and occupational hiatus in the lake-dwelling central European Bronze Age FRANCESCO MENOTTI

INTRODUCTION Although lacustrine settlements began to appear in the Alpine region towards the end of the fifth millennium BC, and lasted until the beginning of the Iron Age (seventh century BC), the lake-dwelling phenomenon is not regarded as continuous. One of the most intriguing gaps in the occupation of lake settlements occurred in the northern Alpine region during the Middle Bronze Age (MBA), which lasted from the fifteenth to the twelfth century BC, and has prompted research to determine whether this was caused by cultural or natural phenomena. Lack of archaeological evidence around all the MBA lacustrine environments north of the Alps argues for an evident cultural change within the lake-dwelling communities. Lake villages disappeared from the MBA lake shores, and settlements started to appear inland in drier locations. After an initial controversy between scholars in the 1970s, who argued for the cultural cause and those in favour of the natural event explanation (climatic change), the results of various multidisciplinary studies led to the hypothesis that the changes were caused by a natural disaster. It is now certain that the climate in the northern Alpine region changed drastically towards the end of the sixteenth century BC, altered the hydrological balance of most of the lakes, and caused their level to rise and transgress on to surrounding shorelines. As a result, a number of lake dwellings became flooded, forcing the villagers to abandon their houses in search of a drier and safer environment. Two of the most relevant sites, which clearly show this process of abandonment, are to be found on Lake Constance. One is the Early Bronze Age (EBA) site of Arbon-Bleiche 2 situated on the southern shore of the lake (in Switzerland), and the second is the EBA village of Bodman-Schachen 1, which lies on the northwestern edge of Lake Constance (in Germany) (also known in German as Lake Überlinger) (Fig. 13.1). Not only were these two sites occupied in the same period, had a similar chronology of occupation and shared a similar environment, but, most importantly, they were also abandoned in the same time span and apparently for the same reason. In fact, archaeological as well as other scientific

236

F. MENOTTI The Alpine Region

The Bodman Bay

Germany France Austria Switzerland

Germany

Switzerland

N

Italy

Slovenia

Co Lake ns tan ce

The Arbon Bay 0

4

8

12 km

Austria Figure 13.1 (Germany)

Lake Constance: locations of Arbon Bay (Switzerland) and Bodman Bay

analyses argue for a major transgression of Lake Constance occurring on both sites just before their abandonment: 1508 BC for Arbon-Bleiche 2 and 1503 for Bodman-Schachen 1. The importance of these lacustrine sites is also underlined by the fact that they were the very last settlements to be abandoned on Lake Constance before the MBA occupational hiatus. A second relevant point is that the natural disaster (a drastic climatic change) took place over a fairly long time in comparison to most other types of natural disaster, such as the volcanic eruptions and tsunamis considered in this volume. Here the flooding occurred over several decades and cultural material associated with the abandonment argues against a rapid-onset event. One can instead consider it a ‘long-term’ natural disaster. The first objective of the chapter is to discuss the similarities as well as the differences between the two above-mentioned EBA lacustrine sites and show graphically, through GIS computer simulations, the impact of the transgressive waters on the settlements and their surroundings. A second goal is to identify the impact of the disaster and the cultural transformations which followed the natural event – in other words, how the flooding actually affected lives. While the cultural change caused by the climatic deterioration resulted in a ‘inland’ shift of the lake dwellings, this was not the only response to the environmental change. The process of adaptation to a different environment also caused a cultural transformation, which affected both social and economic aspects of the former lake villagers.

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

237

HISTORY OF ARCHAEOLOGICAL RESEARCH Arbon-Bleiche 2 (Switzerland) The first evidence of lacustrine villages around the Arbon Bay was found by Jacob Messikommer (1885) in 1882. The discovery, which consisted of wooden piles, an axe and pieces of pottery, was made near the Bär Hotel on the lake shore. The main lake dwellings of Arbon, also known as Arbon-Bleiche, came to light because of water canalisation works carried out in 1885 (Keller-Tarnuzzer, 1925). The site (Arbon-Bleiche 1) was excavated by Messikommer the same year and there was no further research until the second excavation campaign in 1925 (Keller-Tarnuzzer, 1925). Further water drainage works brought about the fortunate discovery of Arbon-Bleiche 2 EBA village in 1944. The site, excavated by Keller-Tarnuzzer (1945), turned out to be the most interesting one of the whole Arbon-Bleiche lake-settlement complex. Numerous objects were found and the high quantity of well-preserved wood enabled fairly accurate dendrochronological dates to be established. A second excavation was carried out in 1990, which broadly confirmed the results obtained almost half a century before (Hochuli 1994). A further site, Arbon-Bleiche 3, was discovered about 50 m west of Arbon-Bleiche 1 (Leuzinger, 1997). There are four other sites in the area known as Arbon-Bleiche 4, 5, 6 and 7, which were discovered in 1944, but these have not yet been systematically excavated. Bodman-Schachen 1 (Germany) Four prehistoric lake-dwelling sites, namely Bodman-Schachen 1, BodmanLöchle, Bodman-Weiler 1 and Bodman-Weiler 2, had been discovered in the Bodman-Schachen area by the end of the second half of the nineteenth century (Köninger, 1993). Two further minor sites are known today: one is located near Ludwigshafen and the other in the Bodman-Blissenhalde area on the southeastern shore of Lake Überlinger. Recorded discoveries began in 1854 (Keller, 1858), when Ley (1866) found some prehistoric lake-dwelling artefacts in his backyard. That year the water level of Lake Constance dropped drastically and wooden piles as well as artefacts became visible in the shallow water. Although serious archaeological studies may be said to have begun with Dehoff’s report in the fifth Pfahlbaubericht (Keller, 1863), scientific research was not conducted immediately, and no stratigraphic recording was done. Even during the first decades of the twentieth century, work carried out by Reinerth (1937) in the 1930s and Maier (1955) in the 1950s was essentially based on artefact analyses without any systematic excavation. For some reason the lake-dwelling sites in the Bodman area remained poorly researched until the beginning of underwater expeditions carried out by the private antiquity-collector Menzel in the early 1970s. This was the starting-point for underwater archaeology, which considerably developed with the PBO (Project Bodensee Oberschwaben) during the 1980s. The last two excavation campaigns, with detailed multidisciplinary scientific research, were carried out by Joachim Köninger in 1982–84 and 1986 (Köninger, 1993, 1996).

238

F. MENOTTI

CHRONOLOGY AND PREHISTORIC SETTLEMENT CHANGES Arbon-Bleiche 2 The chronology of Arbon-Bleiche 2 is based on sedimentological, dendrochronological and typological data. Two phases of occupation can be identified. The first phase took place in the Early Bronze Age between 1650 and 1590 BC (Hochuli, 1994). The second and last occupation started around 1558 and ended at the brink of the fifteenth century BC, just before the beginning of the MBA lakedwelling hiatus (fifteenth–twelfth centuries BC) (Hochuli, 1994). The time span of the initial occupational phase covers more than 60 years and incorporates two subphases of tree felling and of house construction between 1640 and 1590 BC (Hochuli, 1994). Bodman-Schachen 1 Similar to Arbon-Bleiche 2, the chronology of Bodman-Schachen 1 is also based on sedimentological, dendrochronological and typological data, and three phases of occupation can be identified. The first phase occurred in the Early Bronze Age during the nineteenth century BC. The second took place between 1640 and 1610 BC and finally, the third occupation developed just before the MBA lakesettlement occupational gap (Köninger, 1993). The initial occupation period of the first phase is reflected by an anthropogenic stratum, which is both preceded and followed by natural lake sediments with no traces of human industry; this is clear evidence of abandonment. The time span of the second occupational phase covers more than 40 years and has two sub-phases of tree felling and three of house construction, all between 1640 and 1590 BC (Köninger, 1995). There are three layers of cultural debris (Köninger, 1993). Phase three registers only a short period of occupation (less than two decades), but as it was deposited in 1503 BC, just before the MBA ‘missing period’, it is of crucial importance for the understanding of this general phenomenon of abandonment (Köninger, 1996, 1997).

TOPOGRAPHICAL AND GEOLOGICAL ASPECTS Arbon-Bleiche 2 The lacustrine settlement of Arbon-Bleiche 2 is located on the Swiss shore of Lake Constance (Fig. 13.1). The site is found on an extensive flat area formed by alluvial and colluvial activity during the post-glacial period. It is surrounded to the north by gentle morainic slopes, and to the west and south by moderately high slopes. The area is crossed by a number of small rivers, namely the Salbach, the Aach and the Steinach. Because the EBA level of Lake Constance was at c. 392–3 m asl and the prehistoric dwellings of Arbon-Bleiche 2 were found at an altitude which ranged between 396.00 and 396.90 m asl (Hochuli, 1994), it appears that the latter were never influenced by seasonal lake transgressions, which were, and still are, usually less than 3 m in range.

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

239

Bodman-Schachen 1 The EBA settlement of Bodman-Schachen 1 is situated on the peninsula called Schachenhorn, which was formed by fluvial accumulation. The surviving wooden posts lie in a shallow-water environment at c. 120–60 m from the present shoreline at 395.5 m asl. This shallow-water zone extends down to 392 m asl. Since the prehistoric dwelling is situated at 393 m asl, it is at present submerged all the year around. The settlement is located on the extreme northwestern shore of Lake Constance (Fig. 13.1). The environment surrounding the site consists of an extensive plain called Espasinger Niederung. The plain was part of Lake Constance during the Holocene. The soil mainly consists of an accumulation of calcareous lacustrine deposits. The settlement area is surrounded to the north, east and south by modest hills. The plain is cut through by two main rivers: the River Stockacher Ach flowing west to east into Lake Constance, and the River Dettelbach coming from the Bodanrück Hills and connecting with the Stockacher Ach at the point called the Grosse Ried. The delta of the Stockacher Ach on the western shore of Lake Überlinger does not now occupy the same position as it did in the past: in prehistoric time it was situated c. 700 m northwards (Erb et al., 1961) and there have been moderate changes to the morphology of the landscape (Schlichtherle, 1985a). .

THE LAKE-LEVEL FLUCTUATION HYPOTHESIS: GIS SIMULATIONS Palaeoclimatological studies based on dendrochronology (Bortenschlagen, 1977; Furrer, 1977; Renner, 1982), pollen analyses (Burga, 1987, 1988, 1991) and sedimentology (Joos, 1982, 1991; Magny, 1980, 1992) have shown that increases in humidity and precipitation within the catchment area of a lake are directly reflected in lake levels. These lake-level transgressions are believed to be the main factor that forced the EBA lake dwellers of the Lake Constance region to abandon the lake shores and settle further inland at the end of the sixteenth century BC. Research carried out by Gamper and Suter (1982) and Jacomet (1985) has shown that within the Lake Zurich region lake levels fluctuated considerably from 2500 BC to the present (Fig.13.2). These water-level variations were mainly related to changes in climate. Towards the end of the sixteenth century BC, precipitation increased and initiated lacustrine transgression. However, as Joos (1982) argues, climatic factors are not the only reasons for the lake-level changes. Land use and deforestation, for example, can easily alter the hydrological balance of lakes and rivers. The process of forest clearance associated with agricultural activity reduces the permeability of the ground and consequently more water reaches the lakes, causing them to rise (Gross-Klee and Ritzmann, 1990). Since Lake Constance has a similar climate and comparable setting to Lake Zurich, it is reasonable to propose that changes in lake level observed in one were mirrored in

240

F. MENOTTI

Figure 13.2 Source:

Lake-level fluctuations on Lake Zurich in the past 4,500 years

Modified from Gross-Klee and Ritzmann (1990: 168)

the other. Dendrochronological data support this; all the studied EBA lake villages of the region, Arbon-Bleiche 2, Bodman-Schachen 1, Lake Constance and ZHMozartstrasse, Lake Zurich (Menotti, 1999b) were abandoned in the same decade, namely 1510–1500 BC. If we compare this date with Fig. 13.2, it can be seen that the two sites considered in detail here were occupied immediately before a significant rise in lake levels.

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

241

The hydrology of Lake Constance Palaeoclimatological and sedimentological analyses (Liese-Kleiber, 1985; Joos, 1982) show that Lake Constance was much larger after the Würm glaciation, with its water level reaching 403–5 m asl. The level dropped to 400 m asl in the Mesolithic, 398 m asl at the beginning of the Neolithic and stabilised to 392 m asl during the EBA. Unlike Lake Zurich, Lake Constance has never been artificially regulated, and therefore the present seasonal-level fluctuations should have more resemblance to those in prehistoric times (Schlichtherle, 1985a,b). Because of its size and the vastness of its catchment area, seasonal-level variations on Lake Constance are greater than those on any other Alpine lakes. Under normal climatic conditions, the winter level is 2–2.5 m lower than that in the summer, and in extreme cases the level might reach 4 m difference (Köninger, 1993). One of the highest levels in historic times was registered in the Arbon Bay in 1890 (398 m asl). Similar transgressions have been observed during the twentieth century: e.g. 397.80 m asl in 1910, 397.77 m in 1926 and finally the most recent one, 397.88 m in May 1999 (Thurgauer Zeitung, 25 May 1999). MBA lake-level fluctuations on Lake Constance: GIS computer simulations How much did the level of Lake Constance in the Arbon and Bodman bays fluctuate during the MBA, and what were the effects of the flooding waters on the sites’ surroundings? Unfortunately there is no straight answer to this question. What can be done instead is to present some computer simulations to show graphically the impact of the transgressive lake levels on the landscape. The four maps (Figs 13.3, 13.4, 13.5 and 13.6), based on elevation contour lines, have been digitised using AutoCAD and exported in GIS Arch.View 3.2

Figure 13.3 GIS computer simulation of the inferred EBA Lake Constance level in Arbon Bay (392 m asl)

242

F. MENOTTI

Figure 13.4 GIS computer simulation of the inferred MBA Lake Constance level in Arbon Bay (400 m asl)

Figure 13.5 GIS computer simulation of the inferred EBA Lake Constance level in Bodman Bay (392 m asl)

and elaborated into a DTM (Digital Terrain Model) using the TIN (Triangulated Irregular Network) method. Before explaining the computer simulations based on the rising of lake levels, it has to be pointed out that the maps’ digitised contours have been reconstructed in

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

243

Figure 13.6 GIS computer simulation of the inferred MBA Lake Constance level in Bodman Bay (400 m asl)

relation to geomorphological as well as topographical maps, archaeological excavations and core sampling. The maps cannot therefore be 100 per cent accurate as representations of the original conditions. Arbon-Bleiche 2 and the Arbon Bay With a lake level of 392 m asl, which is thought to have been the level of Lake Constance in the Arbon bay during the EBA, it can be seen that the EBA village of Arbon-Bleiche 2 was situated at c. 1.2 km inland from the coast. In this case, even a severe seasonal transgression would not have influenced the settlement (Fig. 13.3). If the lake level is raised by 2 m the distance of the village from the lake does not decrease significantly, but a narrow depression near the settlement begins to be inundated. Two more metres of increase (396 m asl) and the lake water would have partially flooded (perhaps not permanently) the EBA village situated at an altitude of 396 m asl. At the 398 m contour line, which is 2 m below the maximum flood extension of Lake Constance (Schlichtherle, 1995a), the village is inundated by several metres of water. Finally, when the lake level reaches the 400 m contour line the settlement is well under water and the flooded area around the EBA site is quite considerable (Fig. 13.4). It can be argued that 5–6 m of lake-level increment (from the EBA level of 392 m asl) should have been enough to flood a fairly extensive area of adjacent shorelines, affecting not only the settlement, but also its agricultural economy.

244

F. MENOTTI

Bodman-Schachen 1 and the Bodman Bay The EBA lacustrine site of Bodman-Schachen 1 was situated on the Schachenhorn peninsula at an altitude of ~393–393.5 m asl. Since the level of Lake Constance in the Bodman coastal area was c. 392 m asl, the first simulation represents the lake level before it began transgressing in the MBA (Fig. 13.5). Increasing the lake level by only 2 m, the EBA village would have been inundated by 0.5 m of water. It is hard to tell whether the flood would have persisted all year around, but, even if it had only been seasonal, life in the settlement would have been seriously disrupted. At today’s level of 395.5–396 m asl the site would have been completely submerged by water and would have lain at c. 150 m from the shore. Raising the lake level by up to 397 m asl would have doubled the distance of the settlement from the shore and there would have been a high loss of cultivable land. Because of the extremely flat land morphology, another 1 m increase (398 m asl) would have left the EBA lacustrine village almost in the middle of the lake, more than 600 m from the coastline. Finally, at a lake level of 400 m asl (the maximum) almost the entire Espasinger plain would have been covered in water (Menotti, 1999a), leaving a very limited area of tillable land towards the west (Fig. 13.6). It has been calculated that if the lake level had reached the 400 m asl contour line, c. 570 ha of cultivable land would have been flooded (almost three-quarters of the total Espasinger plain).

MOVING INLAND: MBA SETTLEMENTS AROUND LAKE CONSTANCE A striking feature of the chronological chart of lake dwellings is the complete lack of archaeological evidence of human occupation along all the northern Alpine region lake shores during the MBA (fifteenth to twelfth centuries BC). Despite the severity of the natural disaster, it is obvious that all those northern Alpine lacustrine populations did not vanish without trace. Although it is too hazardous to state with certainty that the exodus from the lake shores was homogeneous and sudden all over the region, there is evidence for a landwards movement of EBA lake dwellers. MBA land settlements north of the Alps have always been known, but recent discoveries of some MBA sites in close proximity to the lakes have reopened the issue of whether or not these settlements should be regarded as lake dwellings, or at least of lacustrine tradition. The MBA land settlement of Bodman-Breite, for example, was discovered within Bodman bay in 1994. The only traces of archaeological structures are a living floor (the characteristic MBA cobbled floor) measuring 9 m long by 2 m wide, and a number of pottery sherds belonging to the MBA horizon (fifteenth century BC onwards) (Schlichtherle, 1995b). This site is probably the most important among all the MBA settlements in the Lake Constance region for our understanding of the MBA lacustrine hiatus in the northern Alpine foreland. In fact, being situated at c. 403.5–404.5 m asl, Bodman-Breite is one of the closest to the shore MBA land sites ever discovered around Lake Constance. At today’s lake

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

245

level of c. 395.5 m asl, the settlement lies c. 450–500 m away from the shore, but at a possible maximum lake level of 400 m asl the site would lie very close to the shore (c. 100 m). This suggests that in some cases the EBA lake dwellers of the northern Alpine foreland did not move very far away from the shores, and settled slightly further inland for environmental security reasons. This is also confirmed by the recent discoveries of a few EBA–MBA sites near the city of Constance which, because they were located at an altitude above the 400 m contour line, were not influenced by the fluctuations of the Lake Constance level and continued to be occupied throughout the MBA (see Tägerwilen-Hochstrasse below). An eloquent proof of the MBA lake dwellers’ inland movement in the northern Alpine region is provided by the recent discoveries made on the Swiss side of the city of Constance area where, following an archaeological survey which preceded the construction of the motorway N7, new MBA sites have come to light. The site of Kreuzlingen-Töbeli was also discovered thanks to the construction of a road (Rigert, 1999). The settlement lay at an altitude of 400 m asl and was occupied from the end of the EBA to the beginning of the MBA. This suggests that the level of Lake Constance may have reached the 400 m contour line in that period, so forcing the EBA lacustrine groups to move further inland. Further confirmation of the influence of lake level upon settlement continuity comes from the site of Kreuzlingen-Töbeli which, lying at 408 m asl, was not influenced by the lake-level fluctuations and was therefore occupied from the EBA until the end of the MBA (Rigert, 1998, 1999). MBA land settlements which may be related to the inland movement caused by the displacement of lake dwellings during the gap fifteenth–twelfth centuries BC are not only present in the Lake Constance region. A few important sites such as Erlenbach-im-Grund, Erlenbach-Obstgartenstrasse, Dietikon-Vorstadtstrasse 32, Küsnacht-Itschnach (Zumikerstrasse-Schürackerstrasse) and Küsnacht-Itschnach (Neuwies) have been discovered in the Lake Zurich region; Fällanden-Wigartenstrasse on Lake Greifen; and Pfäffikon-Hotzenweid and Pfäffikon-Steinacker on Lake Pfäffikon (Fischer, 1997). The Lake Zug area is also an exceptionally good example because almost all the MBA settlements were constructed far from the lake shore and were built from the fifteenth century BC onward, exactly when the lacustrine exodus began (Gnepf, 1995; Hochuli, 1995; Gnepf et al., 1996). It can therefore be argued that these MBA land settlements located near the lake were constructed by the lake dwellers who were abandoning the lake shores.

DISCUSSION The MBA lacustrine ‘missing period’ on Lake Constance is a good example of how a slow-onset natural hazard may affect societies and eventually lead to adjustments to the social and economic practices of the communities exposed to the event. The cultural implications of the lake-level rises for the EBA lacustrine societies were fairly marked. Not only did the flooding cause the inland shift of

246

F. MENOTTI

the EBA lake dwellers, but it also forced them into an intense contact with other land populations, which led to a process of acculturation. Both the social and economic aspects of the former lake villagers changed. The economy in particular underwent a substantial transformation. Agriculture continued to be an important aspect of the people’s everyday life, but animal breeding (cattle and goats) became more and more part of the groups’ subsistence. Because the research on recently discovered MBA land settlements is still in progress, a clear picture of potential changes in the social organisation has not as yet been constructed. It is, however, vital to highlight the fact that this kind of slow-onset natural hazard will engender the development of different adaptation strategies from those called for in response to rapid-onset events. As pointed out at the beginning of the chapter, the flooding occurred over a short period of time (but not instantly), in some cases within a few years and in others over one or two generations. This issue of timescale is very important when we consider how cultures adapt to natural disasters. The lacustrine cultures had time to devise appropriate strategies in response to the deterioration in their environment. But while the shift of settlement and change in subsistence strategies were inevitable, the process of acculturation with the neighbouring terrestrial societies also transformed (although not totally) the social and economic aspects of those groups affected by the natural disaster.

CONCLUSIONS The climatic deterioration which occurred in the northern Alpine region towards the end of the sixteenth century BC altered the hydrological balance of most of the Alpine foreland lakes, causing their levels to rise. As a result, the various lacustrine populations which occupied the lake shores were forced to abandon their existing settlements. Lake Constance is particularly important for the study of the northern Alpine MBA lake-dwelling occupational hiatus because two of the last EBA sites to be abandoned before the hiatus were indeed located on its shores. In fact, Arbon-Bleiche 2 (the southern shores of the lake, in Switzerland) and Bodman-Schachen 1 (the lake’s northwestern extreme, in Germany) were both abandoned in the last decade of the sixteenth century BC. The similarities of these sites, the way they were both abandoned, and the morphological aspect of their surroundings, have allowed us to make a comparative analysis of the settlements and create palaeo-maps whereby the transgressions of the lake levels have been simulated and graphically displayed through GIS computer analyses. A study of the immediate vicinities of Lake Constance has subsequently proved that the lacustrine communities forced by the natural disaster to abandon the lake shores did not vanish. A number of MBA ‘land’ settlements located near the lake were in fact built after the exodus from the lake shores and a careful examination of the archaeological record has shown a link to the former lake-dwelling communities. The climatic variability, which caused a general inland movement and transformed the cultural make-up of the former lacustrine groups, offers an

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

247

example of cultural change and cultural continuity at the same time. In fact, the MBA former lake dwellers maintained a connection to the lacustrine environment even during those centuries ‘away’ from the lakes and they started to move back to their ancestors’ settlements once the lake level decreased and the shores became safe for settlement once again. Natural disasters always convey the idea of a quick phenomenon, which affects people’s lives within a very brief period of time. The MBA lacustrine flooding in the northern Alpine region is an eloquent example of how natural disasters also occur on a ‘long-term’ basis and shows that the implications for a society’s cultural organisation can be as severe as those brought about by an abrupt and extremely rapid natural event. ACKNOWLEDGEMENTS I wish to thank the following people for their invaluable help: Dr Andrew Sherratt, Ashmolean Museum Oxford (UK); Dr Helmut Schlichtherle, Landesdenkmalamt Baden-Württemberg (Germany); Dr Joachim Köninger, Archäologische Dienstleistungen Freiburg (Germany); Dr Urs Leuzinger, Amt für Archäologie Kanton Thurgau (Switzerland); Erwin Rigert, Amt für Archäologie Kanton Thurgau (Switzerland); Patrick Daly and Andre Tschan, St Cross College Oxford (UK); Tyler Bell, Queen’s College Oxford (UK). REFERENCES Bortenschlagen, S. (1977) Ursachen und Ausmass postglazialer Waldgrenzschwankungen in den Ostalpen. In B. Frenzel (ed.) Dendrochronologie und Klimaschwankungen in Europe, 260–66. Wiesbaden: Steiner Verlag. Burga, C. (1987) Vegetazionsgeschichte seit der Späteiszeit. Geographica Helvetica 2: 71– 80. Burga, C. (1988) Swiss vegetation history during the last 18,000 years. New Phytologist 110: 581–602. Burga, C. (1991) Vegetation history and palaeoclimatology of the Middle Holocene: pollen analysis of alpine peat bog sediment, covered formerly by the Rutor Glacier 2510m (Aosta Valley, Italy). Global Ecology and Biogeography Letters 1: 143–50. Erb, L., Haus, H.A. and Rutte, W. (1961) Geologische Karte von Baden-Württemberg 1:25,000, Erläuterungen zu Blatt Stockach L 8120. Stuttgart: Vermessungamt. Fischer, C. (1997) Innovation und Tradition in der Mittel- und Spätbronzezeit: Gräber und Siedlungen in Neftenbach, Fällanden, Dietikon, Pfäffikon und Erlenbach. Zurich: Monographien der Kantonsarchäologie Zürich 28. Furrer, G. (1977) Klimaschwankungen im Postglazial im Spiegel fossiler Böden: Ein Versuch im Schweizerischen Nationalpark. Erdwissenschaftliche Forschung 13: 267–70. Gamper, M. and Suter, J. 1982. Postglaziale Klimageschichte der Schweizer Alpen. Geographica Helvetica 2: 105–14. Gnepf, U. (1995) Acht neue prähistorische Fundstellen aus dem Kanton Zug. Tugium 11, 60–73. Gnepf, U., Moser, P. and Weiss, J. (1996) Morastige Wege und stattliche Häuser im mittelbronzezeitlichen Cham. Archäologie der Schweiz 19: 64–7.

248

F. MENOTTI

Gross-Klee, E. and Ritzmann, C. (1990) Die neolithische und bronzezeitliche Siedlungen im Züricher Seefeld. In Schweizerisches Landesmuseum (ed.) Die erste Bauern (Band 1), 161–76. Zürich: Schweizerisches Landesmuseum. Hochuli, S. (1994) Arbon-Bleiche. Die neolithischen und bronzezeitlichen Seeufersiedlungen: Ausgrabungen 1985–1991. Frauenfeld: Amtes für Archäologie des Kantons Thurgau. Hochuli, S. (1995) Die frühe und mittlere Bronzezeit im Kanton Zug. Tugium 11: 74– 96. Jacomet, S. (1985) Botanische Makroreste aus den Sedimenten des neolithischen Siedlungsplatzes AKAD-Seehofstrasse am unteren Zürichsee. Zürich: Juris Verlag. Joos, M. (1982) Swiss Midland-lakes and climatic changes. In A.F. Harding (ed.) Climatic Change in Later Prehistory, 44–51. Edinburgh: Edinburgh University Press. Joos, M. (1991) Zur Bedeutung der Steinhaufen (Ténevières) von Yverdon VD-Avenue des Sports. Jahrbuch der Schweizerischen Gesellschaft für Ur- und Frühgeschichte 74: 195–9. Keller, F. (1858) Pfahlbauten (2. Bericht). Mitteilungen der antiquarischen Gesellschaft in Zürich 2: 111–55. Keller, F. (1863) Pfahlbauten (5. Bericht). Mitteilungen der antiquarischen Gesellschaft in Zürich 5: 127–88. Keller-Tarnuzzer, K. (1925) Pfahlbauten-Ausgrabungen in der Bleiche bei Arbon. Jahrbuch der Schweizerische Gesellschaft für Ur- und Frühgeschichte 17: 35–42. Keller-Tarnuzzer, K. (1945) Pfahlbauten Arbon-Bleiche. Jahrbuch der Schweizerische Gesellschaft für Ur- und Frühgeschichte 36: 19–26. Köninger, J. (1993) Die frühbronzezeitlichen Ufersiedlungen von Bodman-Schachen 1: Befunde und Funde aus den Tauchsondagen 1982–1984 und 1986. Unpublished Ph.D. dissertation, University of Freiburg. Köninger, J. (1995) Die Tauchsondagen in den Ufersiedlungen von Bodman-Schachen 1. In Landesdenkmalamt Baden-Württemberg (ed.) Archäologie unter Wasser 1, 43–50. Stuttgart: Konrad Theiss Verlag. Köninger, J. (1996) La stratigraphie de Bodman-Schachen 1 dans le contexte Bronze ancien du sud de l’Allemagne. In C. Mordant and O. Gaiffe (eds) Cultures et Sociétés du Bronze Ancien en Europe, 239–50. Paris: Comité des Travaux Historiques et Scientifiques Köninger, J. (1997) Ufersiedlungen der frühen Bronzezeit am Bodensee. In H. Schlichtherle (ed.) Pfahlbauten rund um die Alpen, 29–35. Stuttgart: Konrad Theiss Verlag. Liese-Kleiber, H. (1985) Pollenanalysen in urgeschichtlichen Ufersiedlungen – Vergleich von Untersuchungen am westlichen Bodensee und Neuenburger See. In Landesdenkmalamt Baden-Württemberg (ed.) Berichte zu Ufer- und Moorsiedlungen Südwestdeutschlands 2, 200–40. Stuttgart: Konrad Theiss Verlag. Leuzinger, U. (1997) Schmuck und Zier in der jungneolithischen Seeufersiedlung Arbon TG Bleiche 3. Plattform 5–6: 67–75. Ley, F. (1866) Stein- und Bronzepfahlbauten zu Bodman am Überlinger See. Pfahlbauten. Mitteilungen der antiquarischen Gesellschaft in Zürich (Bericht 6) 6: 33–45. Magny, M. (1980) Fluctuations lacustres et paléoclimatologie postglaciaire. Bulletin de l’Association Française pour l’Etude du Quaternaire 1–2: 57–60. Magny, M. (1992) Holocene lake-level fluctuations in Jura and the northern subalpine ranges, France: regional pattern and climatic implications. Boreas 21: 319–34. Maier, R.A. (1955) Keramik der Badener Kultur aus Ufersiedlungen des Bodensees. Germania 33: 155–78. Menotti, F. (1999a) The abandonment of the Early Bronze Age lake-settlement of Bodman-Schachen 1: A CAD and GIS approach to the lake-level fluctuation hypothesis. In J.A. Barceló, I. Briz and A. Vila (eds) New Techniques for Old Times – CAA98: Computer Applications and Quantitative Methods in Archaeology, 265–69. Oxford: BAR International Series 757. Menotti, F. (1999b) The abandonment of the ZH-Mozartstrasse Early Bronze Age lake-

CLIMATIC CHANGE IN THE CENTRAL EUROPEAN BRONZE AGE

249

settlement: GIS computer simulations of the lake-level fluctuation hypothesis. Oxford Journal of Archaeology 18(2): 143–55. Messikommer, J. (1885) Der neue entdeckte Pfahlbau Bleiche-Arbon. Antiqua 11: 22–8. Reinerth, H. (1937) Das Federseemoor als Siedlungsland der Vorzeitmenschen. Leipzig: Barth. Renner, F. (1982) Beiträge zur Gletschergeschichte des Gottardgebietes und dendroklimatologische Untersuchungen an fossilen Hölzern. Unpublished Ph.D. dissertation, University of Zurich. Rigert, E. (1998) Fundbericht 1997. Jahrbuch der Schweizerischen Gesellschaft für Ur- und Frühgeschichte 81: 266–79. Rigert, E. (1999) Fundberichte 1998. Jahrbuch der Schweizerischen Gesellschaft für Ur -und Frühgeschichte 82: 250–90. Schlichtherle, H. (1985a) Probleme der archäologischen Denkmalpflege in den Seen und Mooren Baden-Württembergs. Denkmalpflege in Baden-Württemberg 14: 69–82. Schlichtherle, H. (1985b) Prähistorische Ufersiedlungen am Bodensee – Eine Einführung in naturräumliche Gegebenheiten und archäologische Quellen. In Landesdenkmalamt Baden-Württemberg (ed.) Berichte zu Ufer- und Moorsiedlungen Südwestdeutschlands 2, 9– 42. Stuttgart: Konrad Theiss Verlag. Schlichtherle, H. (1995a) Bemerkungen zur Siedlungsstruktur der Feuchtbodensiedlungen im südwestdeutchen Alpenvorland. In A. Aspes (ed.) Modelli Insediativi: Tra le Alpi e Mar Nero dal Quinto al Secondo Millennio A.C. (Verona-Lazise 1992), 22–8. Verona: Museo Civico di Storia Naturale. Schlichtherle, H. (1995b) Eine Mineralbodensiedlung der Mittelbronzezeit in Bodman, Gde. Bodman-Ludwigshafen, Kreis Konstanz. Archäologische Ausgrabungen in BadenWürttemberg 1994: 61–5.

14

Towards an archaeology of crisis: defining the long-term impact of the Bronze Age Santorini eruption JAN DRIESSEN

INTRODUCTION: DISASTERS AS SOCIAL PHENOMENA Recent anthropological research has given increased attention to disasters as social phenomena following the realisation that they not only constitute one of the basic elements of many environments, but they also form part of the constructed features of human systems (Oliver-Smith, 1996). Impact and response are central themes in the discussions, especially where natural-resource-dependent communities (NRCs) are concerned. More importantly, disasters have been shown to act as significant accelerators in political, social, cultural and economic situations of instability. Disasters can and have changed history (e.g. Prince, 1920; Sheets and Grayson, 1979; Tainter, 1988; Yoffee and Cowgill, 1988; Whittow, 1980; Ward and Joukowsky, 1992; Weiss et al., 1993). At a regional scale, they can be seen to influence economic strategies (Andreau, 1973; Widemann, 1986). Essentially, a disaster is an event that involves a ‘combination of a potentially destructive agent(s) from the natural and/or technological environment and a population in a socially and technologically produced condition of environmental vulnerability’ (Oliver-Smith, 1996: 305). This combination leads to damage of the major social organisational elements and physical facilities of a community to such a degree that the essential functions of the society are interrupted or destroyed. A result is individual and group stress combined with social disorganisation of varying degrees of severity. Disasters, therefore, tend to affect most aspects of community life. The impact of disasters on human societies depends on the magnitude, duration and frequency of the phenomenon, on its impact on natural resources, on the pre-existing adaptive strategies of the affected human population and on the size and distribution of the groups.

THE SANTORINI ERUPTION For antiquity, we have some information on the impact of earthquakes and natural catastrophes in Anatolia and the Levant (Ünal, 1977; Fadhil, 1993).

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 251

Leaving aside the Bible, the Ahmose Stele, the Talos myth, the Atlantis legend and some other even more dubious sources (Foster and Ritner, 1996), we lack eyewitness accounts of the Bronze Age eruption of the volcano on the Cycladic Island of Santorini (Thera), possibly one of the largest natural catastrophes in human history. We know that this happened in the second half of the second millennium BC, when the Aegean witnessed the heyday of Minoan culture, which was centred on the island of Crete. Its civilisation, the Linear A script, which still defies decipherment, evidence for a complex exchange network, and a stratified social system all show that Minoan Crete had acquired a sophistication equalling that of Pharaonic Egypt. It has been suggested that Knossos had become if not the political then undoubtedly the cosmological capital of the island at the beginning of the Late Bronze Age (Schoep, 1999), with several secondary centres in agriculturally favoured areas. Minoan Crete also influenced the culture of the nearby Cycladic and Dodecanese islands where colonies or trading enclaves were established (Wiener, 1990). Recent studies (Hamilakis, 1996; Haggis, in press), however, have stressed that, in essence, Minoan palatial society was highly unstable, partly due to the complexity of its hierarchical structure and partly because of its dependence on agricultural specialisation and extensive cropping. Indeed, around 1450 BC, at the end of the period labelled Late Minoan IB, this civilisation collapsed when almost all settlements on the island were burned down. In the subsequent periods Crete increasingly underwent influence from Mycenaean or Mainland Greece.

ASSIGNING THE BLAME From 1939 onwards, it was assumed that the Santorini eruption had to be blamed for the demise of the Minoans (Marinatos, 1939). However, the recognition that Late Minoan IA pottery styles were present in the destruction layers of the site of Akrotiri on Santorini, while Late Minoan IB pottery was present in the destruction contexts on Crete, made such an assumption untenable. Scholars still argue about the precise time lapse between the two events. In any case, other explanations had to be sought for the Minoan collapse. The most popular one at present is a series of massive earthquakes on Crete (Pichler and Schiering, 1980; Warren, 1991a), but human aggression is also a widely accepted hypothesis (Hood, 1985). Evidence has been collected for the occasional occurrence of island-wide tremors or possibly even pan-Aegean earthquakes during the Bronze Age (Warren, 1991b). Such an explanation, however, does not agree with some of the destruction-related phenomena identifiable in the archaeological record of LM IB sites. These phenomena are as follows: 1 Preferential treatment. In some cases (e.g. Petras, Myrtos-Pyrgos), only the central buildings were burned down whereas the settlement itself escaped destruction but was nevertheless abandoned.

252

J. DRIESSEN

2 Burning. In LM IB, fire is the only destruction agent whose identity is beyond doubt. Although earthquakes occasionally also cause fires, destruction caused by tectonic activity remains difficult to identify in the archaeological record (Stiros, 1996). The best indication for earthquake destruction on Crete is the discovery of an extensive rubbish deposit reflecting the clearing out of debris and the re-use of the structure. This did not happen after the LM IB destructions. 3 Plunder and malicious destruction. Certain destruction deposits illustrate the smashing of prestigious art objects. Other buildings entirely lack valuables. 4 Lack of reoccupation after the destruction. Because of its geographical location, earthquakes and tremors often bother Crete. During the Minoan period, extensive rebuilding always followed earthquake destructions. Since this only happened on a very limited scale after the LM IB collapse, it seems to suggest that an earthquake was not the cause of the destruction. 5 Pre-destruction features, such as the protection of access, water supplies and livestock, and the hiding of valuables, seem to reflect emergency procedures which were reactions to an imminent danger. These features suggest that humans caused the LM IB destructions. The preferential treatment and the destruction of prestigious items may even imply that the aggression was especially directed towards symbols of authority, such as a class or an élite. If indeed this was the case, then we should look for a social explanation for the demise of the Minoan palace culture.

A ROLE FOR THE SANTORINI ERUPTION? If humans are indeed to blame for the collapse of the Minoan civilisation, does this imply that the potential of the Santorini eruption to explain the history of the initial Late Bronze Age Aegean is limited? Was the eruption no more than a mere nuisance (see Chapter 15), without further consequences? Recently, the archaeological record of the Late Minoan I phase was analysed in terms of disaster studies with a focus especially on the identification of archaeological correlates of post-traumatic stress syndromes (Driessen and Macdonald, 1997). This research allowed the formulation of a working hypothesis in which the eruption served as a catalyst provoking changes that drastically altered the face of both Minoan Crete and the Bronze Age Aegean, and ultimately paved the way for the Hellenic civilisation. We argued that the eruption was indirectly responsible for the collapse of the Minoan civilisation because it triggered a series of societal changes that caused the fragmentation of the political landscape, accompanied by a process of internal disintegration. In discussing the role of the Santorini eruption, several factors should be distinguished: (1) the eruption itself; (2) the identification of its immediate, physical effects on Minoan Crete; and (3) the recognition of its potential longterm effects in the archaeological record.

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 253

The eruption In terms of the precise chronology of the Santorini eruption, we know that the volcano erupted during a ceramic phase labelled Late Minoan IA (henceforth LM IA), which is characterised by pottery decorated with floral motifs, widely distributed on Crete and the islands of the Aegean. Most of the destruction of buildings by fire associated with the collapse of Minoan society happened slightly later, in a period when Late Minoan IB, especially Marine Style, pottery was widely used. The absolute date of the eruption remains a topic of dispute with a High Chronology, around 1628 BC (Manning, 1995: 214; cf. Chapter 15), and a Low Chronology, around 1520 BC (Warren, 1996: 288). Proponents of the High Chronology refer to dendrochronological evidence from Turkey (Kuniholm et al., 1996) and to Greenland ice cores, but Zielinski and Germani (1998) have recently shown that the composition of the tephra situated at a date of 1628/1627 does not match that of Santorini, although this conclusion has been questioned recently (Manning, pers. comm.; see Chapter 15). Incidentally, the dates for the LM IB destructions on Crete also linger between a high date, around 1490 (Manning, 1999) and a low date, around 1470 (Warren and Hankey, 1989). If we assume that the eruption happened at the very end of the LM IA period, as some have suggested (Soles et al., 1995) but not all accept (Manning, 1999; Warren, 1999), the LM IB period would have lasted between 50 and 100 years. Arthistorical studies likewise suggest a time lapse of three generations (Popham, 1990). The excavations at the site of Akrotiri on Santorini and three conferences devoted to its contextual relevance reconstructed the following process for the eruption (cf. Friedrich, 1994; Forsyth, 1997; Driessen and Macdonald, 1997; cf. Druitt et al., 1999 for the most recent geological and archaeological reviews): 1 Some time before the eruption, a serious earthquake caused substantial damage at Akrotiri and also on Crete (Palaikastro, Mochlos) and some islands of the Dodecanese (Kos, Rhodes). Blot (1978) suggested a causal link between intermediate-depth earthquakes and volcanic eruptions and, in the case of Santorini, a delay of two to five years is assumed. 2 Repairs were undertaken at Akrotiri, including the demolition of dangerous walls, the shovelling aside of earthquake debris and the construction of squatter habitation. Rehabilitation work was interrupted when the volcano started to erupt. The emission of gases, smoke and ash forced the inhabitants to flee. 3 A thin layer of phreatic ash covered the island before the main strata of pumice were deposited. 4 Finally, the volcano ejected a mass of tephra along with enormous boulders. There is disagreement about the size of the caldera, and the magnitude and nature of the sulphur discharge of the eruption – issues that influence its potential global effects and, in particular, its climatological consequences. It has been suggested, for instance, that the Santorini eruption was five times more powerful than the Krakatau explosion, which released 26 times the energy of the largest hydrogen bomb ever detonated (Whittow, 1980: 91). The sheer size of the tephra and

254

J. DRIESSEN

pumice deposition on the Bronze Age ruins of Akrotiri leave no doubt about one issue: the eruption was a very large one. Physical effects of the eruption Whereas many natural disasters (earthquakes, floods, tornadoes) mostly occur in isolation, eruptions tend to be accompanied by a number of severe events, including earthquakes, flood, ash fall, tsunami, etc. It is the combined effect of these that is especially devastating. It is doubtful that direct volcanic action would have affected the island of Crete, situated at more than 100 km to the south. Akrotiri and other settlements on the island of Santorini itself were annihilated, however, and the blast of the explosion would have been heard and seen from the Cretan shores. The psychological effects should therefore not be underestimated. Volcanic tremors (as opposed to tectonic earthquakes) are usually of minor importance. At the moment we are archaeologically unable to distinguish between the tremors heralding and possibly provoking the eruption and those that accompanied the eruption itself. We only know that buildings were damaged shortly before the ash fell and that at Akrotiri, several buildings sustained damage during the eruption, partly by the impact of volcanic bombs and the weight of ash, and partly, perhaps, by additional seismic activity. The eruption and collapse of the caldera undoubtedly caused a tsunami or massive shock-generated sea-wave. Such a tsunami, dating to the second millennium BC, has been identified in the seabed between Crete and Santorini (Kastens and Cita, 1981; McCoy and Heiken, 2000). S. Bottema and G. Marinos (pers. comm.) and press reports (Patris, 1998) speak of recently identified tsunamiassociated layers in West and Central Crete. A computer simulation of a tsunami in the Cretan Gulf of Mirabello by Monaghan et al. (1994) calculated a wave c. 40 m high. In several riparian areas on Crete, pumice layers, possibly derived from a tsunami, have been found inland, occasionally even on top of Minoan ruins (Driessen and Macdonald, 2000). The tsunami may also have caused the temporary salinisation of wells and of coastal agriculture. The ejection of volcanic ash or tephra can probably be regarded as the worst physical effect of an eruption because it has both short- and long-term consequences. As short-term effects, the hiding of the sun and moon and the darkness may have had a profound psychological impact, causing panic (see Chapter 6). As regards ash fall, we know from deep-sea sediment cores that an average of at least 5 cm must have been deposited on the southeastern Mediterranean region (Watkins et al., 1978). Ash layers up to 1 m thickness have been identified on Rhodes (Doumas and Papazoglou, 1980) and Santorini ash has been traced all over Crete, in the Black Sea, in inland Anatolia and in the Nile Valley, Syria and Israel (Vitaliano and Vitaliano, 1974; Sullivan, 1993; Gulchard et al., 1993; Stanley and Sheng, 1986). Ash layers of 5 to 12 cm have been excavated in some East Cretan sites such as Mochlos and Palaikastro (Soles et al., 1995; MacGillivray et al., 1998: 242). Taking into account bioturbation and dispersal after wind and rain, the package of freshly fallen tephra may have been considerably thicker (in the order of three times). The relatively high fluorine content of Santorini tephra may well

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 255

have been poisonous and polluted water supplies, killed animals and destroyed agriculture, but little research has been carried out on these potential effects. If the sulphuric acid aerosol was considerable, climatological anomalies may have resulted (Rampino et al., 1988). From the Tambora eruptions we know that large eruptions may modify climate, with consequent impacts on agriculture and society (Flohn, 1981; Grattan and Sadler, 1999, Harington, 1992). The Santorini eruption hence undoubtedly produced some immediate effects in the Aegean and Mediterranean world. The havoc caused by tsunamis, ashfall and climatological anomalies may have affected agriculture, livestock, water supplies and traffic. Potential long-term consequences A proper understanding of the impact of the Santorini eruption is only possible if this event is considered against the general background of disaster studies. For instance, it has been documented that, whereas the direct impact of a disaster on a human group is more or less egalitarian, affecting all levels of society in similar ways, the recovery process is almost always discriminatory, intensifying preexisting social, political and economic differences (Grayson and Sheets, 1979: 626–7; Whittow, 1980: 375). Moreover, in disaster relief operations, one usually distinguishes between the emergency phase, which is associated with the immediate effects and recovery after a disaster, and the rehabilitation phase, which is characterised by attempts to bring the community back to its former level of existence. Violence, antisocial behaviour and community conflicts are apparently quite rare in the emergency period but quite frequent during the rehabilitation period precisely because of the discriminatory recovery process, when underprivileged minority groups unleash their grievances (Torry, 1978a; 1978b; Quarantelli and Dynes, 1976: 140; Whittow, 1980: 393). Conflict arises because of problems with the allocation of resources for rehabilitation and the assignment of blame. Where the allocation of resources is concerned, one should take into account that ash fall and climatological anomalies that may affect harvest yields may accompany volcanic eruptions. Reduced yields in unstable NCRs rapidly lead to famine. Several authors (Baehrel, 1952; Dirks, 1980; Arnold, 1988; Dodghson et al., 2000) have underlined the role of famine as an accelerator of historical change, with effects on demography, economy, politics, society and culture. Frequent cultural responses to reduced harvest yields are a reduction in population size, a change in the distribution of human groups (including their mobility patterns), a diversification of production and the conversion of food into direct and indirect storage (Halstead, 1992: 111–14; Post, 1985). Often, however, it is not the lack of resources but their distribution that is problematic since disasters may have disrupted the infrastructure. The tendency for hoarding also disrupts the exchange system. As to the social consequences of a disaster and the assignment of blame, Adams and Adams (1984: 255) describe how during the seven months after the Mt St Helens eruption, the following increased in the area where the ash had fallen: 18.6 per cent in death rate; 21 per cent in emergency room visits; 198 per cent in stress-related illnesses; 235 per cent in mental illnesses; 25.5 per cent in sick-leave;

256

J. DRIESSEN

45.5 per cent in domestic violence; and 37.5 per cent in aggression. Clearly, then, disasters have great psychological consequences and may cause unusual behaviour (Wenger et al., 1975). Dynes (1975: 22–4) has also shown that, following a disaster, greater cohesion in local groups is often the case, which in political terms is sometimes translated in moves to regionalism, decentralisation and the formation of new groups. ‘Acts of God’ may be an easy way out for modern-day insurance companies, but disasters in NRCs are closely connected to religion and ritual, and someone must take the blame (cf. Fadhil, 1993: 276–7). Victimisation and scapegoating after a disaster is a normal human reaction. This is because there is a widespread tendency to explain disasters in terms of the sins of the people or the leaders (Mackay, 1981: 367). Baum (1987: 35) has shown that such blame is often directed against individual groups of people, especially during the recovery process. Ancient Egypt presents some good examples since some pharaohs were executed after natural catastrophes (Bell, 1971). Often crises are accompanied by the appearance of extreme ritual reactions or, as La Barre (1971) calls them, ‘crisis cults’ or revitalisation movements (Wallace, 1956). A fine example of such a cult, immediately connected to a volcanic eruption, was recently discovered on the flanks of Mt Popocatépetl, where the courtyards of some huts had miniature volcano shrines provided with chimneys from which smoke would escape from charcoal to appease the volcano gods. These failed, however, since a pumice and lava carpet covered the settlement during Popocatépetl’s AD 80 eruption (Plunket and Uruñuela, 1998). Whittow (1980: 392) has further illustrated his point that following some disasters ‘previously unaccepted practices such as theft, prostitution or even cannibalism become acceptable’. In summary, disasters produce long-term effects which act as catalysts for political, economical, social and psychological actions (Baum, 1987: 12).

AN ARCHAEOLOGY OF CRISIS As I have illustrated, disaster studies offer a framework against which we may reconsider the archaeological evidence of Late Minoan I Crete. Since Minoan society may have been inherently unstable because of its hierarchical complexity and its agricultural overspecialisation (Hamilakis, 1996; Haggis, in press), I assume that following the Santorini eruption it must have encountered problems concerning the allocation of resources and the assignment of blame. What is therefore required is an ‘archaeology of crisis’ which can identify adaptations or changes in different domains of material culture caused by stress situations (cf. Chapter 5). Some data, such as demographic curves, settlement patterns, subsistence activities, ritual actions and perhaps élite symbolism, may manifest changes. The simple observation of such change in the archaeological record is, of course, not sufficient to assume a crisis or, as La Barre (1971: 11) calls it, ‘a basic problem with which routine methods, secular or sacred, cannot cope’. The cumulative result of changes in a wide area of cultural domains, however, may be suggestive of a period of stress (Driessen, 1995; Driessen and Macdonald, 1997).

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 257

Thanks to a large number of excavations dealing with precisely this period, the Minoan archaeological record provides abundant data that, seen against this hypothesis, may be explained as stress-induced phenomena. In a recent book Driessen and Macdonald (1997) provide a full discussion of the various changes observed in the course of the Late Minoan I period. In the following, I briefly highlight the most important ones which relate to settlement patterns, architecture, economy, ritual, funerary practices, administration and international relations. From the mature Late Minoan IA period onwards, destructions and abandonment of settlements occur. Although these phenomena concentrate respectively at the end of the LM IA and at the end of LM IB phase, several need to be placed in the course of these two periods. One should therefore consider the events as part of a process or a continuum of settlement changes. As Driessen and Macdonald (1997: 35–9) have shown, of the 54 settlements occupied during the LM IA period, only 32 remained in use in the next phase, LM IB, and only 10 of these were present in the following LM II period. Large settlements usually remained occupied but special function sites (rural and ritual establishments) and minor settlements were mostly given up. This development should be seen against the general trend of site creation: at the beginning of or during LM IA, many new sites were established. Not a single settlement, however, seems to have been founded during LM IB, even if some existing settlements, such as Mochlos and Vathypetro, appear to have extended their special function zones during this phase. Despite these two examples, the occupied area within the existing settlements decreased progressively in the course of the overall LM I period. In those sites occupied in LM I and LM II, we observe that on average 70 per cent of the site area remained occupied after LM IA but only 20 per cent after LM IB (Driessen and Macdonald, 1997: 39, fig. 4.6). The combination of abandonment and destruction of sites in the course of this period, together with the reduction of occupied area, implies a serious decline in population which, in its turn, requires a prime mover such as famine, plague, war, natural catastrophe or a combination of these (Cowgill, 1975). The gradual reduction in site numbers and site size can be paralleled with a scarcity of building activity during LM IB. This is surprising since a vast series of major construction programmes was undertaken at the beginning of and during LM IA. During LM IB, construction was limited to repairs in cheap materials (cf. Adam, 1986) and only occasionally to the veneering of existing élite dwellings. There was, however, a tendency to increase the security of existing buildings through the construction of enclosure walls, the narrowing of doorways, and the complicating of access and circulation patterns – modifications especially noticeable in the central building of the settlements (Driessen, 1997). Despite a difference in scale and architectural elaboration, there were at least half a dozen settlements on the island before the LM IB destruction that comprise a central court building, usually labelled ‘palace’, each probably administering its own polity: Knossos, Phaistos, Malia, Zakros, Gournia, Kommos, Petras and Galatas. We have shown that the number of polities drops to one (Knossos) afterwards and that, taking into account the lack of finds and the blocking of doorways, most of

258

J. DRIESSEN

these central court buildings were no longer operational by the end of the LM IB period (Driessen and Macdonald, 1997: 44). Other features that can be observed in buildings that were not entirely or partly given up after LM IA is their subdivision into different households and a change in the function of several rooms. There is a clear trend in LM IB to use original, fancy rooms for storage or food processing. Even more dramatic is the almost complete disappearance of private élite architecture after the LM IB destructions, implying a drastic development in social relations between central authorities and élites. A centralisation of industrial production (especially pottery, metal, stone vases, textiles, wine, ivory) within the confines of LM IB settlements and central buildings also occurred. In some cases, there may even be evidence for mass production of pottery. The aforementioned construction of enclosures also served to protect livestock. There was also an increase of storage areas (particularly community and domestic), whether newly constructed or modified (cf. Smyth, 1991). A high number of hoards comprising bronze objects or other valuables occurred. The widespread intentional deposition of bronze seems to reflect a socio-economic disruption, in part because it implies the collapse of the redistributive system (cf. Knapp, 1988). Finally, several freshwater wells were given up and new ones excavated, some of which are of large dimensions and hence possibly for public use (e.g. Palaikastro – Area 6). Traditional non-palatial ritual areas such as peak sanctuaries in the countryside were abandoned before LM IB. Of the original 24 sacred mountain sites, only that of Iuktas near Knossos attests some minor use. Likewise, ‘lustral basins’, ritual sunken areas accessible via a staircase, are filled in and often lost their function. Major monumental rural shrines (such as Symi) were modified and lost their grandeur. Cave sanctuaries became more important, illustrating a chthonic aspect of religion. The appearance of ritual equipment (including bronze figurines, inscribed stone vessels) in settlement contexts and especially in wealthy households also suggests that urban community shrines and domestic cults gained in importance. In addition, two- and three-dimensional anthropomorphic representations became more important during LM IB. Finally, it is striking that there are half a dozen cases where volcanic pumice is used in rituals during the mature Late Minoan I period, for example, as part of a foundation deposit beneath a threshold at Nirou Chani, in an offering in a well at Zakros and before the collapsed entrance of the Arkalochori Cave. Finally, from Knossos comes an LM IB ritual deposit probably indicative of cannibalism. Pumice as well as cannibalism may suggest the existence of ‘crisis cults’. From the beginning of the Early Bronze Age up to a mature phase of the LM IA period, we have a continuous record of Minoan funerary practices which have a clear preference for collective burial. There is, however, a remarkable absence of LM IB burial contexts, apart from half a dozen chamber tombs excavated at Poros, the harbour town of Knossos (Muhly, 1992). From LM II onwards, however, individual burial was the norm and the grave goods reflect an obvious stress on status and military matters. Both the change from collective to individual burial and the new funerary assemblages reflect a social evolution during

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 259

LM IB, a period during which alternative ways of body disposal must have been in use. Most of the Late Minoan I central court buildings yielded some evidence of administrative practices in the form of tablets and sealings written in the Linear A script. At Knossos, Malia and Phaistos, the three major Minoan ‘palaces’, this evidence dates to LM IA, despite the LM IB contexts destroyed by fire, which could have preserved the clay documents. In contrast, many LM IB regional centres (Chania, Hagia Triada, Sklavokampos) and minor ‘palaces’ (Zakros, Gournia, Petras) preserve some signs of administration. From the mature LM IA period onwards, several important coastal and inland sites on Crete were either abandoned or destroyed. This must have hampered maritime and overland trade and communication. Moreover, the disappearance of the Minoanised trading post and Minoan bridgehead at Akrotiri on Santorini meant that Crete’s relations with the Cyclades were seriously jeopardised. A society such as the Minoan one, which depended heavily on trade for the import of raw and luxury materials, must necessarily have suffered a serious setback. From the LMIB period, the Mycenaeans of the Greek mainland became the strongest sea power and gradually established themselves in places that were previously Minoan colonies. Several of the features discussed above imply an unstable situation. The tailing off in construction activity, the attention to reduced access and security through the construction of enclosure walls and even the decline of ritual activity in the countryside and its increase within urban environments suggest conditions of insecurity outside of towns. On the other hand, it also appears that local authorities saw their power expanding. This is not only suggested by the increase of local food and specialist craft production as well as storage, but also by the setting up of local administrations and the increase of cult symbolism such as horns of consecration, double axes and sacred knots in élite assemblages. The centralisation of production and storage obviously increased the power base of the local élites, allowing autonomy. Whether this should be seen as a remedial and preventive measure, which reflected an adaptive process because the ‘palaces’ failed to provide, or represented decentralisation organised by the central authorities, it does imply a weakening of the palace authorities and a fragmentation of the political landscape, with an increased use of ritual in local political propaganda. The sudden appearance of Marine Style pottery in Late Minoan IB is a case in point. Although its symbolism was most probably influenced by the natural phenomena (tsunami) accompanying the Santorini eruption (Bicknell, 2000), it also has élite and ritual connotations at the same time as filling a sudden gap in propaganda-laden art objects in other materials such as ivory and precious stones.

CONCLUSIONS I propose that a combination of natural catastrophes, namely first a massive earthquake followed soon after by the Santorini eruption, inflicted a serious blow

260

J. DRIESSEN

on Minoan society. The Santorini eruption served as a catalyst, which, through the considerable damage it caused, triggered off a disintegration process that lasted several generations. The archaeological evidence suggests a severe economic dislocation following the eruption, forcing local centres to adapt and to assume greater independence from the palace centres. This fragmentation seems to have led to internal Cretan conflict with increasing competition between élites and settlements in the acquisition of resources. The massive wave of places destroyed by fire may reflect a state of anarchy at the end of the LM IB period. The disappearance of the palace societies opened the way for the gradual Mycenaeanisation of the island and hence its absorption into the Greek world. ACKNOWLEDGEMENTS I would like to thank C.F. Macdonald, T. Cunningham and S. Bottema for stimulating discussions and D. Haggis for letting me read his paper in advance of publication. REFERENCES Adam, J.-P. (1986) Observations techniques sur les suites du séisme de 62 à Pompéii. In C. Albore Livadie (ed.) Tremblements de Terre, Éruptions Volcaniques et Vie des Hommes dans la Campanie Antique (Bibliothéque de l’Institut français de Naples 2: VII), 67–87. Naples: Centre Jean Bérard. Adams, P.R. and Adams, G.R. (1984) Mount Saint Helen’s ash fall. Evidence for disaster stress reaction. American Psychologist 39: 252–63. Andreau, J. (1973) Histoire des séismes et histoire économique. Le tremblement de terre de Pompéi (62 ap. J.-C.). Annales Économies. Sociétés. Civilisations 28: 369–95. Arnold, D. (1988) Famine, Social Crisis and Historical Change. Oxford: Blackwell. Baehrel, R. (1952) La haine de classe en temps d’épidémie. Annales Économies. Sociétés. Civilisations 7: 351–60. Baum, A. (1987) Toxins, technology, and natural disasters. In G.R. Vandenbos and B.K. Bryant (eds) Cataclysms, Crises and Catastrophes: Psychology in Action (Master Lectures 6), 5–53. Washington, DC: American Psychological Association. Bell, B. (1971) The Dark Ages in ancient history. 1. The first Dark Age in Egypt. American Journal of Archaeology 75: 1–26. Bicknell, P. (2000) Late Minoan IB marine ware, the marine environment of the Aegean, and the Bronze Age eruption of the Thera volcano. In W.J. McGuire, D.R. Griffiths, P.L. Hancock and I.S. Stewart (eds) The Archaeology of Geological Catastrophes. London: Geological Society Special Publication, 171: 95–104. Blot, C. (1978) Volcanism and seismicity in Mediterranean island arcs. In C. Doumas and H.C. Puchelt (eds) Thera and the Aegean World, II, 33–44. London: The Thera Foundation. Cowgill, G.L. (1975) Causes and consequences of ancient and modern population changes. American Anthropologist 77: 505–25. Dirks, R. (1980) Social responses during severe food shortages and famine. Current Anthropology 21: 21–44. Dodgshon, R.A, Gilbertson, D.D. and Grattan, J.P. (2000) Endemic stress, farming communities and the influence volcanic eruptions in the Scottish Highlands. In W.J.

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 261

McGuire, D.R. Griffiths, P.L. Hancock and I.S. Stewart (eds) The Archaeology of Geological Catastrophes. London: Geological Society Special Publication, 171: 267–80. Doumas, C. and Papazoglou, L. (1980) Santorini tephra from Rhodes. Nature 287: 322–4. Driessen, J. (1995) ‘Crisis Architecture’? Some observations on architectural adaptations as immediate responses to changing socio-cultural conditions. Topoi 5: 63–88. Driessen, J. (1997) Observations on the modification of the access systems of Minoan palaces. Aegean Archaeology 2: 67–85. Driessen, J. and Macdonald, C.F. (1997) The Troubled Island. Minoan Crete before and after the Santorini Eruption (Aegaeum 17). Liège and Austin: Université de Liège. Driessen, J. and Macdonald, C.F. (2000) The eruption of the Santorini volcano and its effects on Minoan Crete. In W.J. McGuire, D.R. Griffiths, P.L. Hancock and I.S. Stewart (eds) The Archaeology of Geological Catastrophes. London: Geological Society Special Publication, 171: 81–93. Druitt, T.H., Edwards, L., Mellors, R.M., Pyle, D.M., Sparks, R.S.J., Lanphere, M., Davies, M. and Barriero, B. (1999) Santorini Volcano (Geological Society Memoir 19). London: The Geological Society. Dynes, R.D. (1975) The comparative study of disaster: a social organizational approach. Mass Emergencies 1: 21–31. Fadhil, A. (1993) Erdbeben im Alten Orient. Baghdader Mitteilungen 24: 271–8. Flohn, H. (1981) Short-term climatic fluctuations and their economic role. In T.M. Wigley, M.J. Ingram and C. Farmer (eds) Climate and History: Studies in Past Climates and their Impact on Man, 310–18. Cambridge: Cambridge University Press. Forsyth, P.Y. (1997) Thera in the Bronze Age. New York: Peter Lang. Foster, K.P. and Ritner, R.K. (1996) Texts, storms, and the Thera eruption. Journal of Near Eastern Studies 55: 1–14. Friedrich, W.L. (1994) Feuer im Meer. Vulkanismus und die Naturgeschichte der Insel Santorin. Heidelberg–Berlin–Oxford: Spektrum Akademischer Verlag. Grattan, J. and Sadler, J. (1999) Regional warming of the lower atmosphere in the wake of volcanic eruptions: the role of the Laki fissure eruption in the hot summer of 1783. Geological Society Special Publication, 161: 161–72. Grayson, D.K. and Sheets, P.D. (1979) Volcanic disasters and the archaeological record. In P.D. Sheets and D.K. Grayson (eds) Volcanic Activity and Human Ecology, 623–32. New York: Academic Press. Gulchard, F., Carey, S., Arthur, M.A., Sigurdsson, H. and Arnold, M. (1993) Tephra from the Minoan eruption in sediments of the Black Sea. Nature 363: 610–12. Haggis, D. (in press) Integration and complexity in the Late Prepalatial Period: a view from the countryside in Eastern Crete. In Y. Hamilakis (ed.) Labyrinth Revisited: Rethinking Minoan Archaeology. Oxford: Oxbow Books. Halstead, P. (1992) Agriculture in the Bronze Age Aegean. In B. Wells (ed.) Agriculture in Ancient Greece. Proceedings of the 7th International Symposium at the Swedish Institute in Athens, 16–17 May 1990, 105–16. Stockholm: Svenska Institutet I Athens. Hamilakis, Y. (1996) Wine, oil, and the dialectics of power in Bronze Age Crete: a review of the evidence. Oxford Journal of Archaeology 15: 1–32. Harington, C.R. (ed) (1992) The Year Without a Summer. Ottawa: Canadian Museum of Nature. Hood, S. (1985) Warlike destruction in Crete c. 1450 BC. In Proceedings of the International Conference of Cretan Studies held at Agios Nikolaos in 1981, 170–8. Heraklion: Philological Society. Kastens, K.A. and Cita, M.B. (1981) Tsunami-induced sediment transport in the abyssal Mediterranean Sea. Bulletin of the Geological Society of America 92: 845–57. Knapp, B.A. (1988) Hoards d’oeuvres: of metals and men on Bronze Age Cyprus. Oxford Journal of Archaeology 7: 147–76. Kuniholm, P., Kromer, B., Manning, S.W., Newton, M., Latini, C.F. and Bruce, M.J.

262

J. DRIESSEN

(1996) Anatolian tree rings and the absolute chronology of the Eastern Mediterranean, 2220–718 BC. Nature 381: 780–3. La Barre, W. (1971) Materials for a history of studies of crisis cults: a bibliographic essay. Current Anthropology 12: 3–44. MacGillivray, J.A., Sackett, L.H. and Driessen, J. (1998) Excavations at Palaikastro, 1994 and 1996. Annual of the British School at Athens 93: 221–68. Mackay, A. (1981) Climate and popular unrest in Late Medieval Castile. In T.M. Wigley, M.J. Ingram and C. Farmer (eds) Climate and History. Studies in Past Climates and their Impact on Man, 356–76. Cambridge: Cambridge University Press. Manning, S.W. (1995) The Absolute Chronology of the Aegean Early Bronze Age (Monographs in Mediterranean Archaeology 1). Sheffield: Academic Press. Manning, S.W. (1999) A Test of Time: the Thera (Santorini) Eruption and the Absolute Chronology of the Aegean Bronze Age. Oxford: Oxbow Books. Marinatos, S. (1939) The volcanic destruction of Minoan Crete. Antiquity 13: 425–39. McCoy, F.W. and Heiken, G. (2000) Late-Bronze Age explosive eruption of Thera (Santorini), Greece. Regional and local effects. Volcanic Hazards and Disasters in Human Antiquity. Geological Society of America Special Paper 345: 43–70. Monaghan, J.J., Bicknell, P.J. and Humble, R.J. (1994) Volcanoes, tsunamis and the demise of the Minoans. Physica D 77: 217–28. Muhly, P. (1992) Minoïkos laxeftos tafos ston Poro Irakliou. Athens: Archaeological Society. Oliver-Smith, A. (1996) Anthropological research on hazards and disasters. Annual Review of Anthropology 25: 303–28. Patris (1998) 22 May. Athens. Pichler, H. and Schiering, W. (1980) Der spätbronzezeitliche Ausbruch des TheraVulkans und seine Auswirkungen auf Kreta. Archäologischer Anzeiger: 1–37. Plunket, P. and Uruñuela, G. (1998) Appeasing the volcano gods. Archaeology 51: 36–42. Popham, M.R. (1990) Pottery styles and chronology. In D. Hardy (ed.) Thera and the Aegean World, III: 1, 27–8. London: The Thera Foundation. Post, J.D. (1985) Food Shortage, Climatic Variability, and Epidemic Disease in Preindustrial Europe. Ithaca: Cornell University Press. Prince, S.H. (1920) Catastrophe and Social Change Based upon a Sociological Study of the Halifax Disaster. New York: Columbia University Press. Quarantelli, E.L. and Dynes, R.R. (1976) Community conflict: its absence and its presence in natural disasters. Mass Emergencies 1: 139–52. Rampino, M.R., Self, S. and Stothers, R.B. (1988) Volcanic winters. Annual Review of Earth and Planetary Sciences 16: 73–99. Schoep, I. (1999) Tablets and territories? Reconstructing Late Minoan IB political geography through undeciphered documents. American Journal of Archaeology 103: 201–22. Sheets, P.D. and Grayson, D.K. (eds) (1979) Volcanic Activity and Human Ecology. New York: Academic Press. Smyth, M.P. (1991) Domestic storage behavior in Mesoamerica. An ethnoarchaeological approach. In M.B. Schiffer (ed.) Archaeological Method and Theory I, 89–138. Tucson: University of Arizona Press. Soles, J.S., Taylor, S.R. and Vitaliano, C. (1995) Tephra samples from Mochlos and their chronological implications for Neopalatial Crete. Archaeometry 37: 385–93. Stanley, D.J. and Sheng, H. (1986) Volcanic shards from Santorini (Upper Minoan Ash) in the Nile Delta, Egypt. Nature 320: 733–5. Stiros, S.C. (1996) Identification of earthquakes from archaeological data: methodology, criteria and limitations. In S. Stiros and R.E. Jones (eds) Archaeoseismology (Fitch Laboratory Occasional Paper 7), 129–52. Athens: British School. Sullivan, D.G. (1993) Effects of the Santorini eruption on the Bronze Age settlement in Aegean Turkey. American Journal of Achaeology 97: 330. Tainter, J.A. (1988) The Collapse of Complex Societies. Cambridge: Cambridge University Press.

CRISIS DEFINITION ON LONG-TERM IMPACT OF SANTORINI ERUPTION 263

Torry, W.I. (1978a) Bureaucracy, community and natural disaster. Human Organization 37: 302–28. Torry, W.I., (1978b) Natural disasters, social structure and change in traditional societies. Journal of Asian and African Studies 13: 267–83. Ünal, A. (1977) Naturkatastrophen in Anatolien im 2. Jahrtausend v. Chr. Belleten 41: 447–72. Vitaliano, C.J. and Vitaliano, D.B. (1974) Volcanic tephra on Crete. American Journal of Achaeology 78: 19–24. Wallace, A.F.C. (1956) Revitalization movements. American Anthropologist 58: 204–83. Ward, W.A. and Joukowsky, M.S. (eds) (1992) The Crisis Years. The 12th Century B.C. Dubuque: Kendall-Hunt Publishing Company. Warren, P.M. (1991a) The Minoan civilisation of Crete and the volcano of Thera. Journal of the Ancient Chronology Forum 4: 29–39. Warren, P.M. (1991b) A new Minoan deposit from Knossos, c. 1600 BC, and its wider relations. Annual of the British School at Athens 86: 318–40. Warren, P.M. (1996) The Aegean and the limits of radiocarbon dating. In K. Randsborg (ed.) Absolute Chronology. Archaeological Europe 2500–500 BC (Acta Archaeologica 67), 283–90. Munksgaard: International Publishers. Warren, P.M. (1999) Aspects of Minoan chronology. In P.P. Betancourt, V. Karageorghis, R. Laffineur and W.-D. Niemeier (eds) MELETEMATA. Studies in Aegean Archaeology Presented to Malcolm H. Wiener as he Enters his 65th Year (Aegaeum 20), 893– 903. Liège and Austin: Université de Liège. Warren, P.M. and Hankey, V. (1989) Aegean Bronze Age Chronology. Bristol: University Press. Watkins, N.D., Sparks, R.S.J., Sigurdsson, H., Huang, T.C., Federman, A, Carey, S. and Ninkovich, D. (1978) Volume and extent of the Minoan tephra from Santorini volcano: new evidence from deep-sea sediment cores. Nature 27: 122–6. Weiss, H., Courty, M.A., Wetterstrom, W., Guichard, F., Senior, L., Meadow, R. and Curnow, A. (1993) The genesis and collapse of 3rd millennium north Mesopotamian civilisation. Science 261: 995–1004. Wenger, D.E., Dykes, J.D. and Sebok, T.D. (1975) It’s a matter of myths: An empirical examination of individual insight into disaster response. Mass Emergencies 1: 33–46. Whittow, J. (1980) Disasters: The Anatomy of Environmental Hazards. Harmondsworth: Penguin Books. Widemann, F. (1986) Les effets économiques de l’éruption de 79. Nouvelles données et nouvelle approche. In C. Albore Livadie (ed.) Tremblements de Terre, Éruptions Volcaniques et Vie des Hommes dans la Campanie Antique (Bibliothéque de l’Institut français de Naples 2: VII), 107–12. Naples: Centre Jean Bérard. Wiener, M. (1990) The isles of Crete? In D. Hardy (ed.) Thera and the Aegean World, III: 1, 128–61. London: The Thera Foundation. Yoffee, N. and Cowgill, G.L. (eds) (1988) The Collapse of Ancient States and Civilizations. Tucson: University of Arizona Press. Zielinski, G.A. and Germani, M.S. (1998) New ice-core evidence challenges the 1620s BC age for the Santorini (Minoan) eruption. Journal of Archaeological Science 25: 279–89.

15

Volcanoes and history: a significant relationship? The case of Santorini STURT W. MANNING AND DAVID A. SEWELL

INTRODUCTION If you are in the immediate vicinity of an explosive volcanic eruption you face death or at least significant damage to person and property. People may die both from the immediate eruption and its long-term impacts on the environment (ranging from tsunamis to starvation through destruction of the landscape). Such tragedies happened to c.30,000 people in the city of St Pierre in AD 1902 when Mt Pelée erupted, to an unknown number of people, possibly many tens of thousands, due to the eruptions of Mt Krakatoa in AD 1883 and Mt Tambora in AD 1815, or, infamously, to the several thousand inhabitants of Pompeii and Herculaneum when Vesuvius erupted in AD 79 (cf. Blong, 1984; Scarth, 1999). Occasionally, a large eruption proximate to a major region of civilisation may even, through the destruction of the wider arable environment, force demographic shifts, which in turn may have significant historical impacts within a region. This has been argued to be the case for the Ilopango volcano with regard to Mayan civilisation (Sheets, 1979). Considering such examples, there is no denying the awesome and ruthless majesty of a volcanic eruption. There is an innate tendency to assume that volcanic eruptions are therefore historically important. Over the years, it has been proposed that a number of volcanic eruptions have had significant impacts on a local or regional human population. This is hardly surprising: volcanic eruptions are relatively common events. On average, three moderate (or larger) eruptions occur somewhere on the earth each decade (Simkin and Siebert, 1994; Simkin, 1994; Pyle, 1998). However, from a global or long-term historical perspective, most volcanic eruptions are very minor events of purely specific and local significance. Some claim that the largest eruption of the last two centuries, Tambora, may have caused a year without a summer in AD 1816 (Rampino et al., 1988: 83–5; Harington, 1992; Arnold, 1988: 30; Post, 1977). But no other eruption in recent times (AD 1700 onwards) can make such a claim and the experts debate whether Tambora was the primary cause of the poor climate of this period, or merely contributed to a

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

265

coincidental global cooling c.1800–30 (cf. Sadler and Grattan, 1999: 184–7). There is good evidence showing that larger volcanic eruptions have modified world temperatures (and so climate) on a short-term basis (Briffa et al., 1998). The seventeenth century AD, for example, stands out in the last 600 years as a period of exceptional volcanic activity (Briffa et al., 1998: fig. 1, 452–3). However, such variations are both short-lived and usually close to normal seasonal variations and are thus of debatable historical significance (Sadler and Grattan, 1999). No one, of course, disputes the relevance of climate change to long-term human development and history (e.g. Sherratt, 1997) and even moderate short-term climate events, such as one c.2200 BC, can be argued to have had a significant negative impact on human civilisation (Weiss et al., 1993; Dalfes et al., 1997; Cullen et al., 2000), but historical records provide little compelling evidence for volcanic-forced impacts of sufficient magnitude to affect human civilisation. Instead, the established minor, short-term (one to a couple of years) climate effects of volcanic eruptions such as Tambora may be argued to have had limited agricultural and economic effects in some regions (Flohn, 1981), but these are no greater than those caused by many other non-volcanic factors in these or other years; nor is volcanic forcing necessarily the only causal agent in many such cases. In summary, based on observations of the recent past, if you happen to live close to a volcano, then it will play a role in life, history and mythology, sometimes decisively (Blong, 1982). It must also be remembered that to those living in their shadow, volcanoes are not merely a dark force; tephra falls can act as a boon over the longer term by creating a fertile farming environment. However, in wider historical (and in regional to pan-regional) terms, a significant role for volcanoes over the last few thousand years can be questioned (e.g. Buckland et al., 1997). The nineteenth century AD world carried on (indeed thrived) despite Tambora and Krakatau; Europe forgot the toxic fog emitted by the Laki Fissure (cf. Chapter 6); Rome barely noticed the eruption of Vesuvius (cf. Chapter 7) and many Japanese cultures adjusted rather than collapsed in the face of volcanic eruptions (cf. Chapter 18). But is this a complacent view, based simply on the lucky chance that no really big eruption has occurred proximate to a centre of western civilisation? The last few hundred years are notable for the low to moderate number of large explosive volcanic eruptions (Decker, 1990). On a longer timescale, we cannot blithely dismiss all volcanic activity as of no risk to global weather, culture and ecology. Rare but massive volcanic events are observed in the geological record of the earth, and elsewhere in the solar system (Frankel, 1996), and such ‘supereruptions’ could have had a devastating impact on the planet’s flora and fauna (Rampino et al., 1988). Indeed there is a suspicious (but not proven) correlation with several important mass biotic extinctions over the last 250 million years (Rampino and Stothers, 1988; Stothers, 1993; Courtillot, 1994, 1999; Courtillot et al., 1996; Pálfy and Smith, 2000; Rampino and Self, 2000). Theoretically, events of this magnitude may block out nearly all sunlight for several years, causing a ‘volcanic winter’ with dire consequences for flora and fauna (Rampino et al., 1988). The only super-eruption of the Quaternary was the explosive Toba

266

S.W. MANNING AND D.A. SEWELL

eruption (c.75000 BP), which was some twenty (plus) times the scale of the AD 1815 Tambora eruption (Rampino and Self, 1992, 1993; Zielinski et al., 1996a). Average temperatures could have been lowered by a catastrophic c.3–5 °C for several years, and life-critical photochemical reactions in the atmosphere could also have been disrupted (Yang et al., 1996). It has been argued that the consequence of near extinction of many plants and animals (including humans) on a worldwide basis had a dramatic impact on subsequent human evolution (Ambrose, 1998), but this hypothesis is still open to debate (see Oppenheimer, 2002). Post-Toba, the entire history of modern humans has occurred in a period without any super-eruptions. More particularly, the post-Ice Age world, the Holocene, with the invention of agriculture, urban civilisation, and so on, lies in an era of relative climate stability and largely modest global volcanic activity. Thus do we dismiss volcanic eruptions as significant to human history since the Pleistocene? Or have there been occasional events in between the almost irrelevant, in global terms, fireworks displays of the last two centuries, and the Armageddon of Toba? In this chapter we will consider this question in terms of the last 4,000 years, where there is a combination of detailed environmental and archaeological and/or historical data, and any impact can be properly assessed. When we consider the archaeological record and recent human history for volcanic influences, the most likely candidates are sulphur-rich volcanic events on the scale of the Tambora eruption in AD 1815, or larger, or about VEI = 7, using the Volcanic Explosivity Index of Newhall and Self (1982).

CANDIDATE POINTS IN HISTORY? How might we identify potential instances of historically significant past volcanic eruptions – if they exist? Despite much modern study, volcanological records are poor before the nineteenth century AD. The tendency is for one or two wellknown cases to be cited frequently, rather than the world picture to be considered as a whole. None the less, where a very large eruption has been identified, we can ask whether or not it had a wide-ranging, historically significant, impact. One such event is the Minoan (Bo) eruption of the Santorini (or Thera) volcano in the Aegean (Friedrich, 2000; Forsyth, 1998; Fouqué, 1999; McCoy and Heiken, 2000a): discussed below. The alternative approach is to invert the question. If a historically significant volcanic impact is the quest, then when did such impacts occur and can these episodes be associated with a major volcanic eruption? Needless to say, the former is a great deal easier to establish than the latter. If there were only the historical and/or archaeological record, then scant real progress could be made. There is in effect little in the way of a solid, detailed, observational record back for more than a century or so, with occasional exceptions such as the account in two letters written by Pliny the Younger (Letters of Pliny, 6.16, 20). Various periods do offer apparent evidence of records of unusual atmospheric effects potentially consistent with a major volcanic eruption or mention of climate change, famine and so on, but in no case could a volcanic cause

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

267

be more than an unproven hypothesis if this were all the evidence available. However, polar ice offers a well-dated and detailed long-term record of past volcanism (Hammer et al., 1980; Zielinski et al., 1994, 1996b; Clausen et al., 1997; Hammer et al., 1997). Tree-rings can also offer well-dated records of past temperatures and climate trends and these can be linked with volcanic impacts with good confidence in rigorous work based on large samples and secure time-series and proxy-temperature reconstruction (e.g. Briffa et al., 1998), but with less certainty where samples are restricted and the nature of the tree response is an important variable and the linkage is hypothesised at best (e.g. Baillie, 1995, 1998; cf. Buckland et al., 1997, 1998). The question then becomes how to bring these various types of data together in a rigorous way in the pre-modern period.

THE PROBLEM OF PRECISE DATES Work in the 1980s to early 1990s identified perhaps five episodes in the last c.4,000 years when a package of historical and/or archaeological data, combined with ice-core evidence for major volcanism and absolutely dated tree-ring evidence indicative of serious climatic upheaval, seemed to suggest that there may have been instances of historically significant volcanism and/or some other relatively dramatic scenario of supra-regional climate impact. These episodes, giving the tree-ring dates first and then the ice-core date, can be summarised as follows (see e.g. Stothers and Rampino, 1983a, 1983b; LaMarche and Hirschboeck, 1984; Rampino et al., 1988; Pang and Chou, 1985; Pang et al., 1988, 1989; Baillie, 1991, 1995: 73–9; Baillie and Munro, 1988; Warner, 1990): (i) (ii) (iii) (iv) (v)

AD 536/541 or AD 540 ± 10; 44/42 BC or 50 ± 30 BC; 207/206 BC or 210 ± 30 BC; 1159 BC or 1120 ± 50 BC; 1628/1627 BC or 1644 ± <20 BC.

In each of these cases, there appeared to be a coincidence within dating errors of major volcanic eruption evidence as found in the ice-core record and the evidence of notable growth anomalies absolutely dated in trees, and this was potentially consistent with historical and/or archaeological data. Therefore, these five episodes could represent instances of major historically significant volcanism with periodicity once every several hundreds of years. They require further investigation. The ice-core evidence was critical. This alone provided a direct volcanic association (via identification of acid signatures in ice-layers which could only be the result of major volcanism: e.g. Hammer et al., 1987; Calusen et al., 1997) with these ‘packages’ of environmental anomalies. The quality of this evidence varied. The AD 540 date for Event (i) was a ‘preliminarily dated’ eruption signal from ‘a new ice core from South Greenland’ (subsequently the Dye 3 ice core) mentioned in a ‘note added in proof ’ to Hammer et al. (1980: 235). Events (ii)–(iv) were from

268

S.W. MANNING AND D.A. SEWELL

the Camp Century ice core. The 1644 BC date for Event (v) was obtained from the Dye 3 ice core (Hammer et al., 1987). While coincidences do not make a correlation, there did seem a compelling case at the time this evidence was first assembled, for example by Baillie and Munro (1988). However, since then, new and better-dated ice-core evidence has become available from Greenland: the Dye 3 and GRIP ice cores (Clausen et al., 1997; Johnsen et al., 1992), and the GISP2 ice core (Zielinski et al., 1994, 1996b; but cf. Southon (2002) with regard to the second millennium BC). The dating precision for the Dye 3 and GRIP ice cores is much better than for the earlier Camp Century ice core (Hammer, 1984), and both offer strong replication over the last 4,000 years (Clausen et al., 1997). A review of the evidence from these modern ice cores casts serious doubt on the volcanic association for several of the events listed above. Event (i) The preliminary date of AD 540 ± 10 for a major volcanic event was refined to AD 516 ±
VOLCANOES AND HISTORY: THE CASE OF SANTORINI

269

should be 43 BC due to a year 0 in their chronology), and a growth anomaly 44–41 BC in foxtail pines (Scuderi, 1990). On face value, the stated dating errors for the ice-core evidence cannot rule out this possibility. The dates are either very close or just within the error margins. However, Zielinski (1995: 20948) points to another small eruption signal in the GISP2 core at 43 BC, and, since the GISP2 core appears to be fairly accurately and precisely dated only a century earlier (Vesuvius at AD 79),1 he argues that the 53 BC event is in fact distinct from the 44–42 BC historical data and frost damage event. The Roman literary evidence indicates that an eruption of Etna occurred in about 44 BC. However, the historical evidence for 44–42 BC from the RomanMediterranean world, and importantly also from China on the other side of the globe, indicates both dramatic mid-latitude atmospheric anomalies/effects and significant climate modification and associated agricultural problems, in the year or two following (see Stothers and Rampino, 1983b: 6358–60; Pang et al., 1986; Forsyth, 1988; Bicknell, 1993; Ramsey and Licht, 1997: 99–107). The tree-ring evidence from 44–41 BC from the USA likewise indicates significant hemispherewide effects. Thus the question arises whether a supposedly small to moderate eruption in Sicily could have caused this widespread package of unusual events and yet be such a small signal in the Greenland ice. For example, the proposed 43 BC signal was less than one third of the scale of the regionally proximate AD 79 Vesuvius signal which did not cause such effects (Zielinski, 1995: table 4). Ramsey and Licht (1997: 105, 107), who follow Zielinski (1995), seek to explain away this somewhat discordant evidence by noting Etna’s highly sulphurrich nature. But both the sulphur input necessary to cause the relatively significant atmospheric and climatic effects and impact over 1–3 years, and the indication that there must have been a long-lasting (1–3 years) and thus stratospheric sulphurdioxide aerosol, run against association with the modest 43 BC ice-core signal. Overall, no firm resolution is possible at present and the data available are clearly not decisive, but it none the less seems fair to conclude that the linking of the widely reported atmospheric effects and climate impacts of 44–42 BC with a modest eruption (e.g. the c.43 BC signal in the GISP2 ice core) is debatable. This leaves two courses: (i) If a volcanic eruption is involved (as seems likely given some of the historical evidence), it would appear possible, and indeed perhaps more likely, that it is the major 50/49 BC (Dye 3/GRIP), or 53 BC (GISP2), volcanic signal that might better account for the unusual and significant atmospheric and climate anomalies recorded from 44–42 BC? This scenario is more or less possible within stated dating errors, but does require all ice cores to very slightly overestimate the age of this volcanic signal, something which is not likely with the GRIP, Dye 3 and GISP2 ice cores all offering good agreement until the end of the first millennium BC (Southon, 2002). Whether the volcano responsible was Etna, and/or another large volcanic eruption somewhere else in the world at about the same time, is not relevant for present purposes (and we may note that there are many unknown eruptions in the ice-core

270

S.W. MANNING AND D.A. SEWELL

record and it is also perfectly reasonable that one year’s ice-core acidity is derived from the emissions of more than one eruption). Hence this particular package of volcanic event and associated impacts might be allowed to stand (for the time being). (ii) No major volcanic eruption was involved in the 44–42 BC events. The Etna eruption described in historical sources is thus irrelevant in wider terms, and this event would highlight the dangers in adopting too deterministic an approach to correlating ice-core acidity with known volcanic eruptions. This second scenario is necessary if the ice-core evidence is considered securely dated, and not capable of moving to the limits of the dating errors stated. The main objection surrounds the historical evidence indicating that the Etna eruption was quite significant and that there was a widely observed volcanic dust veil and aerosol. Event (iii) The 210 ± 30 BC volcanic signal in the Camp Century record simply vanishes, or at best is a less than enormous signal c.180 BC in the GISP2 core. The Dye 3 and GRIP cores offer no association with tree-ring evidence for environmental stress at 207 BC, nor the other similarly dated cultural evidence cited by Baillie (e.g. 1995: 89, 1999: 72, 78). The GISP2 event at c.180 BC could only be relevant at the very limits of its dating error and does not appear of sufficient magnitude anyway. Event (iv) The Camp Century ice-core volcanic event at 1120 ± 50 BC also either vanishes or could be placed 1074/1073 ±
VOLCANOES AND HISTORY: THE CASE OF SANTORINI

271

reconsidered’. Zielinski and Germani (1998) state a dating error of ± c.36 years for the mid-seventeenth century BC for the GISP2 ice core, but contrasting opinion holds that the error could be somewhat larger. Possible large volcanic signals of relevance for the Dye 3/GRIP event at 1644/1636 BC could thus include the c.1623, 1669 or 1695 BC signals in the GISP2 core, among others. The other alternative, suggested by Hammer (2000: 36), is that the relevant Dye 3/GRIP 1644/1636 BC signal may in fact be missing in the GISP2 ice core. So, ignoring the GISP2 core, the key question for this discussion is whether it is possible that the Dye 3/GRIP volcanic signal is chronologically compatible with the tree-ring event precisely dated from 1628/1627 BC. The 1644 BC date was confusingly given an estimated one standard deviation error of ± 7 years, and an error limit of ± 20 years. The former would exclude the tree-ring event; the latter could just include it. The 1636 BC date would appear even more likely to be potentially compatible. Given the difficulties involved in the precise dating of individual ice cores (Alley et al., 1997) versus the multi-sample and highly replicated absolute nature of the tree-ring timescale from several distinct areas, it might appear incorrect on current evidence to rule out the possibility of the 1628/1627 BC treering event and the 1636/1644 BC ice-core event being coeval (Hughes, 1988; cf. Hammer and Clausen, 1990). However, with the recent further replication of the c.1645 BC volcanic signal in the North GRIP ice core (Hammer, 2000), and, as of late AD 2001, the current best-dating of the Aegean tree-ring record now placing its unique and extraordinary tree-growth anomaly – perhaps explained by the impact of a major proximate volcanic eruption – c.1650 +4/–7 BC (Manning et al., 2001), it seems increasingly likely that the 1650/1645 BC date in fact cannot be reconciled with the 1628 BC date, and the two thus represent different events (with the earlier most likely Thera – see further below; Manning et al., 2002a). In summary, at present the instances in the last 4,000 years of a sustainable linkage of ice-core evidence of major volcanism, tree-ring evidence of a significant growth anomaly, and archaeological/historical evidence consistent with a large volcanic eruption and significant cultural impact are very few. Only the 53/50/49 BC and 1644/1636 BC ice-core volcanic signals survive even to offer possible correlations within stated dating limits with the tree-ring growth anomalies identified respectively at 43/42 BC and 1628/1627 BC (and, as noted at the end of the previous paragraph, the latter appears now also to be excluded on the latest evidence). Clearly, if the high precision claimed for the Dye 3/GRIP and GISP2 ice-core records is valid, there may be no correlation at all. Is a total non-correlation scenario likely? Here we must move from strict data to general observation and common sense. It is notable that in the period 1–2000 BC there are only two instances when several sets of northern hemisphere tree-ring data offer evidence potentially consistent with the impact of a major volcanic eruption: 44/43–41 BC and 1628/1627 BC. The only proximate major volcanic eruptions in the well-dated Dye 3/GRIP ice-cores are dated at 50/49 BC and 1644/1636 BC. The very near to possible temporal overlaps between the pairs of events in each independent record are striking and appear to form a remarkable coincidence if in fact due to random chance.

272

S.W. MANNING AND D.A. SEWELL

The historical and archaeological evidence perhaps consistent with volcanism or a climate anomaly for the seventeenth century BC is not itself precisely dated and cannot assist us further (e.g. Baillie, 1995: 89), but, as noted above, this is not the case for the 53/50/49 BC ice-core signal. Hammer et al. (1980: 233) observed when first discussing their c.50 ± 30 BC signal that Roman writers clearly describe what is most likely a significant volcanic aerosol effect in 44–43 BC, along with climate-related problems, and there is similar evidence from 43–42 BC from China. It seems intrinsically very likely that these events, not mentioned for any other years around about, must equate with the large volcanic signal registered at almost this exact time in all the ice cores (as discussed above and contrary to Zielinski 1995: 20948). Thus 44 BC tree-ring date and history plausibly might equate with the 53/50/49 BC ice-core signals, and, if so in this case, then perhaps one might think also in 1628/1627 BC? Until recently perhaps, but very recent evidence seems to weigh against such a possibility being likely (see above). Notwithstanding, Baillie, in particular, has noted these apparent coincidences in the tree-ring and ice-core records, and has argued that it does seem tempting to try to associate the above pair of very similarly dated records (1994: 215–16). Of course, not all major volcanic eruptions show up in the tree-ring record, nor in every ice-core record, and there is also marked variation in recorded signals in ice cores due to local depositional, meteorological and other factors. None the less, the circumstantial case could lead to the speculation that there is a very minor and gradual ‘inflation’ of calculated ages operating at times in the ice-core chronologies, such that a + 5/6 years overdating at 44 BC has become + 8/16 years at 1628 BC. This process would not seem to be systematic, but probably reflects a few specific core problems over the timescale (e.g. one might speculate about a few possible false annual layers caused by temperature/precipitation anomalies in rare years which have led to the recording of two apparent years in just one real year, or problems interpreting sections of brittle ice, and so on). The next, and again only, major frost damage event in the LaMarche and Hirschboeck (1984) dataset occurs at 2036 BC (correcting erroneous year 0). The only and again proximate major volcanic signal identified around this time in both the Dye 3 and GRIP ice cores lies at 2053/2045 BC (Clausen et al., 1997). If we were to associate these events, then the age ‘inflation’ has remained almost stable from the seventeenth to the twenty-first centuries BC, with just a one-year increase to a + 9/17 years overdating by the ice cores. The existence and coincidence of this ‘triplet’ of events over a 2,000-year period is striking. Until recently it appeared possible that this was more than random chance. Thus we might have permitted Events (ii) and (v) to remain at the plausible hypothesis level (supposed ice-core dating precision notwithstanding). If so, there has been a very minor, gradual and non-systematic, inflation of ages in the Dye 3/ GRIP ice-core timescale. But the best most recent evidence to hand (AD 2000–02: Hammer, 2000; Manning et al., 2001, 2002a) in fact indicates that the c.1650 +4/– 7 BC Anatolian tree-ring growth anomaly and the c.1645 ±7 BC ice-core evidence for a major volcanic eruption are relatively robustly dated, and so distinct from the

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

273

absolutely dated tree-growth anomaly 1628/1627 BC, which is widely attested in mid-high latitude Northern Hemisphere trees.

HISTORICAL IMPACT? THE SANTORINI CASE The 53/50/49 BC or 44–42 BC event is typical of the apparent relative irrelevance of volcanoes for macro-history or the longue durée. It was one of the largest volcanic acid signals in the Greenland ice for the last 4,000 years. Yes, there are records of atmospheric effects, poor harvests and famine in China, from 44 to 42 BC, but these were short-term and did not have long-lasting significance. Famine constantly stalked the ancient to pre-modern world (e.g. Garnsey, 1988, 1992; Arnold, 1988); it was important, and volcanically driven incidents appear no more important than many others caused by other factors such as variations in solar output. Apart from strange atmospheric conditions, nothing occurred which does not occur because of the many other forces operating in nature. It is difficult to find any major historical change in the few years around 50/49 or 44–42 BC that could be attributed to the effects of this eruption. The situation is not unlike that in AD 1816. Some poor harvests and problems associated with cooler temperatures and climate disturbance may be noted (Harington, 1992), but no long-lasting impact or historical change. There were just a few poor years. People on the margin will undoubtedly have suffered the most, as they did in North America in AD 1816. This is not to minimise the 44–42 BC crises. They were undoubtedly serious at the time to a great many people. Six consecutive grain harvests over three years failed in China. This would have been an extremely grave subsistence crisis. The event at 44 BC might very well serve to offer an example of what we are looking for: a serious volcanic impact that only occurs once every several hundred (or more) years. Even so, more information is needed, and the dating and correlation of the specific volcanic, environmental and historical records remains subject to significant debate. What about the seventeenth-century BC event? There are other volcanic candidate eruptions, but many believe it to have been caused by the Minoan eruption of Santorini/Thera in the Aegean (e.g. Manning, 1999; Manning et al., 2001, 2002a) and, indeed both new radiocarbon evidence (Manning et al., 2002b), and analysis of tephra particles from the GRIP ice core will shortly offer support for this belief (Hammer, 2000: 37, n.1; Hammer et al., 2001; C.U. Hammer, pers. comms, 2000–01 and paper in preparation). As already noted, the ice-core acidity levels suggest that the event was one of the largest eruptions of the last several thousand years. In 1939 Marinatos argued that the eruption destroyed the important Minoan civilisation of Crete and this simple historical scenario of regional and historical destruction is repeated to this day in numerous popular books and general geological writings. The volcano has also long been linked with the cataclysmic Atlantis legend (see Friedrich, 2000: 147–57), and events of the Exodus (Bruins and van der Plicht, 1996; Baillie, 1999). Recently, Driessen and

274

S.W. MANNING AND D.A. SEWELL

Macdonald (1997, 2000; cf. Chapter 14), have reasserted the role of the Santorini eruption in the demise of Minoan civilisation. This may therefore be a volcanic eruption that has had a historical impact across a large region (for a recent review of the local and regional effects of the eruption, see McCoy and Heiken, 2000a). This eruption is therefore an instance that could usefully be used to model in detail the impact of volcanic eruptions on cultures and environments. However, considered in detail, the hypothesis that the eruption was responsible for the fall of Minoan civilisation appears too deterministic in approach. Solid evidence indicates otherwise. The critical flaw in the original Marinatos (1939) theory, and its restatement a generation later by Page (1970), was that archaeological evidence from the late 1960s onwards clearly demonstrated that the eruption occurred in the Late Minoan IA ceramic phase, whereas the destruction of the Minoan civilisation did not occur until later, at the close of the subsequent Late Minoan IB period. This left a difficult-to-bridge temporal gap of about 50 years between eruption and destructions according to the then standard archaeological chronology (Renfrew, 1979). During the 1970s–1980s two principal attempts to try to overcome the time interval were advanced. 1 The eruption might have occurred in the Late Minoan IB period, following arguments that there was an interval of c.20 + years between the final preeruption earthquake and/or some very minor initial eruptive event and the abandonment of settlements on Santorini at the close of the Late Minoan IA period, and the great eruption was thus in the Late Minoan IB period (e.g. Page, 1971; Warren, 1984; Manning, 1987; cf. Driessen and Macdonald, 1997: 107), or that the ceramic styles of Santorini found buried under the volcanic debris were provincial and lacked the latest style current on Crete (e.g. Luce, 1976; Bolton, 1976; and others listed by Driessen and Macdonald, 1997: 107). 2 Although within the Late Minoan IA period or at its very end, the eruption none the less occurred only a few decades before the Late Minoan IB destructions. In this case, the eruption impact could be argued to have acted as a trigger for subsequent cultural decline and collapse (e.g. Manning, 1987: 84, 85, postscript no.3). At the time they were written these arguments were valid, potentially. Indeed, the discovery of soil formation at the town of Akrotiri on Santorini after its abandonment by humans, but before the eruption (Money, 1973; cf. Limbrey, 1990: 380), seemed to offer some support for scenario (1) above, as did publication of archaeomagnetic evidence, which apparently indicated the approximate synchronicity of the eruption and the Cretan Late Minoan IB destructions (Downey and Tarling, 1984; but cf. Manning, 1999: 17–18; Driessen and Macdonald, 1997: 106, n.1). However, the recent finds of in situ Santorini airfall tephra at several sites on Crete has established without doubt that the eruption occurred in the later Late Minoan IA period and not the Late Minoan IB period. The temporal gap was real (further details in Manning, 1999: 13–18, 69–75; MacGillivray et al., 1998: 242; Soles et al., 1995).

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

275

Driessen and Macdonald (1997, 2000), in support of argument (2) above, proposed that the Santorini eruption triggered and caused a long-term decline process in the Aegean, which in due course led to the collapse of Minoan Crete. They cite anthropological parallels and models of how crises can undermine societies. However, nearly all the time-quantified analogues cited and the logic in the anthropological review of Oliver-Smith (1996) suggest a response should occur within a few years to one generation after the forcing event, and not significantly longer. Is such a timescale possible in the Cretan case? Although they note an alternative ‘high’ Aegean chronology, Driessen and Macdonald (1997: 23) adopt a traditional chronology for the Late Minoan ceramic periods (based on linkages with Egypt), which places the eruption c.1550–1530 BC and the collapse of Crete within the Late Minoan IB period, c.1480–1425 BC. This chronology indicates a 50–100 + year interval between eruption and collapse. The most recent statements by the leading scholars associated with the traditional Aegean chronology now suggest an eruption date c.1520 BC (or even c.1520/1500 BC) and the end of Late Minoan IB Crete c.1430/1425 BC – interval thus c.70–95 years (Warren, 1999, 2001). This temporal interval, of several generations, might strike most as rather too long for the subsequent hypothesis of Driessen and Macdonald. However, as noted above, the ice-core evidence, and a large body of other data, now appear to provide a strong case for dating the Minoan eruption of Santorini in the seventeenth century BC. A reassessment of the archaeological data, and recent science-based dating results, might also raise the close of the Late Minoan IB period by 50 years or so (Housley et al., 1999; Manning, 1999; Manning et al., 2002b). Does this affect the situation? No. The temporal gap between the Santorini eruption and the collapse of Crete appears slightly longer, some 120+ years, following this new chronology. The actual nature of the volcanic impact caused by the eruption on Crete also needs critical assessment. The effects of earthquakes associated with the eruption may well have been moderate to severe in the region. But it is difficult to envisage such a regular occurrence in the Aegean destroying a whole civilisation. Civilisation continued in the subsequent ceramic period in areas not physically destroyed by the eruption. Tephra falls on Crete, in contrast to the islands of the east Aegean such as Rhodes, were quite light: 0–5 cm (Watkins et al., 1978; Pyle, 1990a; McCoy and Heiken, 2000a: 57–8). Only in east Crete could the falls even have been annoying, and nowhere on Crete do they seem likely to have been sufficient to cause serious long-lasting damage, even if extant airfall traces were increased by an average of c.70 per cent to represent fresh tephra falls, as Thórarinsson (1971: 268) suggested, or even by as much as a factor of three, as Thórarinsson (1979: 134) observed for the AD 1693 Hekla eruption (Blong, 1980, 1984). Most tephra falls went east, with at most 1–5 cm falling in far eastern Crete, and only then if the eruption was in summer (Figs 15.1 and 15.2).2 Sub-1 cm ‘distal’ falls of very fine ash will have been much more widespread, but will not have caused any real impact. Occasional reports of slightly thicker finds in east Crete, of 5–12 cm, are not from in-situ deposits, but are usually either anthropogenic collections or natural, redeposited contexts. Significant or long-lasting tephra-fall

CRETE

N

CRETE

N

➞

➞

CRETE

N

➞

Notes: (i) this model does not include the initial plinian phase (cf. Fig. 15.2); (ii) the model as presented here employs a single point source for the eruption; and (iii) no allowance is made for changes from average wind directions during the course of the eruption

Note: Contour lines represent tephra depths of 1 cm, 5 cm, 10 cm, 50 cm and 1 m. The results indicate that the wind speeds for the winter months are very strong and that the ash is swept almost directly east. In the summer the winds are more variable with the pattern for June–September producing the best fit for the distribution of ash as revealed in deep-sea cores and land excavations

Figure 15.1 Simulation of co-ignimbrite phase Santorini tephra fall for (a) March, (b) June, (c) September and (d) December

CRETE

N

➞

278

S.W. MANNING AND D.A. SEWELL

Figure 15.2 Simulation showing the total distal distribution of Santorini tephra for the plinian phase and the co-ignimbrite phase in June. For the final development of this model, and a proposed correlation with observed data, see Sewell (2001: chapter 7 and esp. Fig. 7.28) Note: An adaptation of the wind vectors for June was employed with the maximum southerly wind vectors for the column heights between 1 km and 10 km. This produces a complex distribution of contours, but one which fits well in approximate terms with available data from marine and terrestrial finds of Santorini tephra (e.g. Sparks et al., 1983/1984). By deliberately emphasising the maximum southerly distribution in order to get maximum tephra fall on Crete, the distribution over Anatolia is dragged south. If average values were used instead, there would be less tephra fall to the south, and more into southwest Anatolia (as known finds from Anatolia clearly indicate: cf. Sullivan, 1988, 1990; Eastwood et al., 1998). In reality, wind directions may have changed during the eruption, which may account for the widespread observed finds of tephra

damage may therefore be ruled out for Crete. Instead, a once-off agricultural disaster with follow-on effects for livestock is the worst-case scenario for far east Crete only. This could occur as a result of 1–5 cm falls if crops were at a susceptible stage (see Blong, 1980, 1984). Suggestions of destructive nuées ardentes may be discounted (Sparks, 1986) and there is no positive evidence for the relevance of poison gases, although fluorosis from fluorine adhering to tephra is a possibility (Driessen and Macdonald, 1997: 93–4). The timing of the tephra fall is of course very relevant. Previous indications suggested perhaps a spring–summer timing.3 The computer modelling of Santorini tephra distribution according to modern monthly mean wind vector data (Fig. 15.2) indicates that only during the summer months (June, July, August, September) does the computed tephra distribution even approximately offer a fallout pattern which is consistent with the known terrestrial and seabed finds of Santorini tephra (Sewell, 2001). Only in the summer months do wind directions

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

279

vary significantly – from simply eastwards – at various heights in the local atmosphere as necessary to enable some small amount of tephra to fall on far eastern Crete and for tephra to fall in southwestern Anatolia (Turkey) (Fig. 15.3). The summer distribution of winds can then help account for the spread of Santorini tephra over Anatolia (Sullivan, 1988, 1990, 1993; Eastwood et al., 1998), and a distal fraction even as far north as the Black Sea (Guichard et al., 1993). At all other times of the year the wind directions are more or less entirely eastwards. Of course, the key assumption is that wind vector data have remained broadly similar for the last 3,600 years. But, given the basic stability of Holocene climate over this time, there is no particular reason to question this working assumption. In this case, what little tephra that fell on eastern Crete did so in summer. This may have affected any crops standing in fields awaiting harvest – harvest time of subsistence crops in Greece ranges today (FAO data, 1959) from May to October, with cereals harvested May to August and especially June to July. Thus this could have led to a possible significant impact. Autumn rains will have washed most of the light tephra fall away and so the impact on the subsequent year’s crop would have been very slight. Driessen and Macdonald (1997: 91–2) suggested a potentially severe climate impact as a result of the Santorini eruption.4 However, the available data do not support such a dramatic scenario. The likely climatic effect of the volcanic aerosol

Figure 15.3 Average wind speeds for Santorini area, 25.4 ºE and 36.5 ºN, for June ( J) and December (D) at 1 km, 10 km, 20 km and 30 km heights for the period 1992 to 1999 Note: Wind vectors represent the direction the wind is blowing and lengths are proportional to magnitude (see scale in centre of diagram). December values vary little in direction but are variable in strength. June values change in strength and direction, turning from nearly directly west through northeast to a southeasterly direction at the bottom of the column. This supports the distribution for the plinian phase of the eruption as indicated in Bond and Sparks (1976: fig1.b)

280

S.W. MANNING AND D.A. SEWELL

from the Minoan eruption of Santorini may have been in the range c.0.5–1.0 °C average temperature reduction in the northern hemisphere based on Tambora and Pinatubo analogues (Pinto et al., 1989; McCormick et al., 1995; Self et al., 1996), which is within the normal fluctuation range. Further, volcanological theory, icecore evidence (Hammer et al., 1987), and likely relevant Anatolian tree-ring evidence (Kuniholm et al., 1996 – but dated c.1650 +4/–7 BC: Manning et al., 2001) show that contemporary weather anomalies were limited to no more than a few years at most (5 years). Further, such conditions (typically cooler, wetter – but with local variations depending on hemisphere location and proximate landmass: see Sadler and Grattan, 1999; Pyle, 1998) may not have had a necessarily adverse effect on agriculture on Crete. In contrast, the only volcanic impact which appears potentially significant for Crete is tsunamis, which may have affected the northern coast and were perhaps especially directed to the west of Crete, immediately following the eruption (Kastens and Cita, 1981; Cita et al., 1996; Cita and Rimoldi, 1997; Cita and Aloisi, 2000; Monaghan et al., 1994; Francaviglia, 1990; McCoy and Heiken, 2000b; Minoura et al., 2000; Hieke and Werner, 2000; Driessen, Chapter 14, this volume). Although one recent study found no strong evidence for major Santorini period tsunamis affecting Crete (Dominey-Howes, 1996), and, in general, the identification of ancient tsunamis from terrestrial deposits is often problematic (cf. Goff et al., 1998), evidence possibly relevant to such events from Santorini itself, from seabed deposits, and from western Anatolia appears to be mounting (Sewell, 2001: chapter 5; and references above – although satisfactory dating evidence is not always available to provide a Santorini association, e.g. Minoura et al., 2000: 60–2). Such an event may have been devastating in susceptible areas. It is also possible that the Marine Style ceramics of the subsequent Late Minoan IB period may in some way reflect this event (and/or tephrapoisoning of shallow marine environments), prompted by sea creatures being deposited en masse on land, or the sudden disappearance of such creatures from the sea (Bicknell, 2000). But it appears unlikely that such a tsunami event, occurring many years to over a century before the collapse of the Minoan civilisation, could be held responsible for this collapse. Nor can it explain destruction and collapse in southern and non-coastal Crete. Specific impact damage aside, the key underpinning logic for the Driessen and Macdonald (1997: 15, 110) case is their argument that the Santorini eruption ‘triggered off ’ the Late Minoan IB period and this period was one of decline and the civilisation as a whole was ‘tired’. They thus see the eruption causing a downwards spiral that led eventually to collapse. Recent evidence runs against this interpretation. First, as Warren (2001) highlights, there is abundant evidence from Crete of a prosperous and vibrant Late Minoan IB period. Second, the timing is problematic. Soles et al. (1995: 391) argued that the finds of Santorini tephra at the site of Mochlos on Crete show that the eruption occurred at the very end of the Late Minoan IA period. This analysis was employed by Driessen and Macdonald (1997: 15) to support the idea of the eruption triggering the Late Minoan IB period. However, the latest evidence from Crete and elsewhere in the Aegean

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

281

appears to indicate that the eruption was not right at the end of the overall Cretewide Late Minoan IA period and that there is a discernible post-Santorini-eruption final phase of Late Minoan IA before the Late Minoan IB period (see Manning, 1999: 69–75).5 In particular, work at Kommos in south central Crete has recently identified such a phase among the sub-phases of Late Minoan IA at the site (Shaw et al., 1997, 2001). Moreover, the evidence from Mochlos presented by Soles et al. (1995) is, arguably, less than decisive on critical examination (Manning, 1999: 73– 5) – although it may none the less be the case that in east Crete, where there was a greater eruption impact (and also the east Aegean), the post-eruption recovery took longer, and hence a clear post-Santorini-eruption final phase of Late Minoan IA as visible elsewhere (areas of Crete and the Aegean less affected) is not present or easily discerned. But its existence elsewhere further distances the eruption from a significant role in the ‘creation’ of the Late Minoan IB period, and more or less totally removes its relevance to the collapse of Late Minoan IB Crete a century or more later. In sum, the environmental impact of the Minoan eruption of the Santorini volcano may have been severe in the east Aegean, and moderate to minor elsewhere in the region, and it may have had, at worst, moderate climate impact for 1–3 years. If the time interval between this eruption impact and the collapse of Minoan Crete was short, from a few years to a few decades, it would seem plausible to attempt to argue for a causal linkage. The attraction is obvious. However, the evidence and its chronology simply do not permit this catastrophe/crisis model. The collapse of Minoan Crete was around a century or more later. Crete and the Aegean recovered, reorientated after the loss of Santorini itself and the more serious instances of damage elsewhere in the east Aegean, and carried on. Of course, the impact of the eruption may have in some way weakened the prior status quo (and certainly any trading system centred on the east Aegean or Santorini itself), and formed an element in the structuration, background and mythology (e.g. Hiller, 1978) of the society (as all history forms the context for the dynamic present), but it did not in a significant, direct, immediate or quantifiable way have anything to do with the destruction of the Minoan palaces some four (or more) generations later. No anthropological or historical analogues exist to suggest otherwise. CONCLUSIONS On the basis of the last 4,000 years, instances of supra-regional, or global, significant volcanic impacts which have in some way directly changed or affected broadscale human history appear to have been very rare. It is very unusual that a volcanic impact exceeds those of other regular natural climate perturbations (e.g. the El Niño oscillation). One event in c.44–42 BC is the only serious candidate since the end of prehistory, and even this case is disputed. Further research into the period would be worthwhile. The very large eruption of the Santorini volcano, often cited as having a historical impact, most certainly did not cause the collapse of the

282

S.W. MANNING AND D.A. SEWELL

Minoan civilisation of Crete as has been argued by various scholars over the years. Its climatic impact may have been significant for a few years in the seventeenth century BC, but we lack closely dated historical or archaeological correlates at present. Again, further research is called for. Otherwise, direct, significant volcanic impacts over the last few thousand years are limited to the immediate region of the eruption. With very few exceptions, major eruptions have not directly affected core regions or primate urban centres of the main civilisations of the last several thousand years. Since Toba, humans and volcanoes have in the main managed to avoid each other for 75,000 years. Our luck will not last forever. ACKNOWLEDGEMENTS We thank Constanza Bonadonna, University of Bristol, for supplying the original computer model for inferring tephra distribution, which has been subsequently developed by Sewell. Sewell thanks Geoff Wadge for his assistance. We are grateful to John Grattan and Robin Torrence for the invitation to contribute to this volume and their subsequent patience.

NOTES 1 Zielinski (1995: 20948) writes that ‘the calibration to the A.D. 79 Vesuvius event suggests that the 53 and 43 B.C. ages for these signals are correct’. However, much could have occurred in the intervening 120 or so years to account for a minor dating error. Moreover, while the one clear signal in the first century AD (see Zielinski, 1995: fig. 3) may very well be Vesuvius, and it is claimed as the oldest securely dated event in the GISP2 ice core, it should be remembered that no positive confirmation has been established for this assumption, since no volcanic glass shards could be identified in this layer (Meese et al., 1997: 26413). The logic is merely that there is a large acid signal, and no other large volcanic eruption at this approximate time is known (a logic now widely criticised for the Santorini debate). Before Vesuvius, the oldest ‘secure’ eruption date cited by the GISP2 team is some nine centuries later: Eldgjá in Iceland. Analysis of the GISP2 core dates this signal (and positively identified tephra) c. AD 938. Clausen et al. (1997: 26721) report a consonant date of c. AD 934 for the same signal (Meese et al., 1997: 26413; Zielinski et al., 1995). 2 The tephra distribution model is a development from that outlined in Armienti et al. (1988) and Macedonio et al. (1988). Figures 15.1 and 15.2 represent earlier versions of the model, as of AD 2000. For a more complete, and further developed, discussion and model, see Sewell (2001). Wind vector data are derived from UKMO-UARS assimilated data (http:// www.badc.rl.ac.uk). Mean monthly data are averaged over the years 1992 to 1999. Sources of data employed are as follows: grain-size distribution for the plinian phase (Woods and Bursik, 1991); grain-size distribution for the co-ignimbrite phase (Sparks and Huang, 1980); terminal velocities for particles (Armienti et al., 1988; Walker et al., 1971); mass of 4 × 1012 kg for the plinian phase and 3.8 × 1013 kg for the co-ignimbrite phase (Sigurdsson et al., 1990); density of 500 kg/m3 for the plinian phase (Pyle, 1990a); density of 1,000 kg/m3 for the co-ignimbrite phase (Watkins et al., 1978); height of column for the plinian phase of 36 km (Sigurdsson et al., 1990); height used for the co-ignimbrite phase also 36 km based on Sparks and Huang (1980) and Sigurdsson et al. (1990: 104). Distribution within the column employs a Gaussian distribution as outlined in Suzuki (1983). 3 Previous suggestions of a spring–summer eruption timing were derived from apparent wind

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

283

direction at the time of the plinian phase of the eruption (Heiken and McCoy, 1984: 8451) and the general eastwards tephra distribution, confirmed especially by finds in Anatolia (see e.g. Doumas, 1990: 49; Sullivan, 1988). On the assumption that the ice-core evidence of 1644 BC in the Dye 3 ice core reflected the Santorini eruption, Hammer et al. (1987: 519) observed that the increased acidity began as the δ18O values changed to warmer summer temperatures, so this too was consistent with a spring to summer eruption timing. 4 The climatic impact of the Santorini eruption depends largely on the scale of its sulphur-dioxide loading of the atmosphere–stratosphere. In reverse, the scale of the aerosol also determines whether or not it is plausible to associate the eruption with specific volcanic-acid-derived signals in the Greenland ice, or widespread tree-ring-growth anomalies. The sulphur production of the Minoan eruption of Santorini has been the subject of much discussion for over a decade. Pyle (1989, 1990b), Sigurdsson et al. (1990), and others have argued that estimates for Santorini’s sulphur output from the petrologic method (Devine et al., 1984; Sigurdsson et al., 1990: 105–6) were modest and certainly far too small to account for the seventeenth-century BC ice-core signal associated with Santorini by Hammer et al. (1987). The estimated sulphur output was thus regarded as insufficient to account for the climatic anomalies indicated by both the icecore signal c.1644/1636 BC, and the tree-ring evidence from 1628–1627 BC. However, other scholars observed that the petrologic technique can seriously underestimate the atmospheric– stratospheric loading of an eruption (Rampino and Self, 1984; Gerlach et al., 1994; Manning, 1992, 1999: 269–72) – and indeed it now seems likely that a discrete bubble phase at depth, allowing the presence of volatiles far in excess of saturation and so far in excess of any estimate made by petrologic analyses of erupted products, is a feature of many and perhaps all large explosive volcanic eruptions (Newhall et al., 2002: 1241). In addition, it appears that studies of ice-core records typically lead to over-estimates of atmospheric–stratospheric loading (Zielinski, 1995: 20944). Thus the true scale of the sulphur-dioxide loading is not clear, but may well be rather larger than previously thought (especially as recent observations lead some geologists to revise upwards the volume of the overall eruption: R.S.J. Sparks, pers. comm.; F.W. McCoy and S.E. Dunn abstract 2002 Chapman Conference on Volcanism and Earth’s Atmosphere, Thera, Greece, 17–21 June 2002). Michaud et al. (2000) analysed mafic (iron-rich) inclusions from the Minoan eruption magma. They argue that this material, although it comprises only a minute fraction of the erupted magma (see Druitt et al., 1999), indicates that there was a large amount of non-erupted sulphur-rich basalt involved in the eruption cycle, and, in turn, that the previous very small estimates of Santorini’s volatile release are potentially massive underestimations. Michaud et al. (2000) calculate a sulphur release for the Minoan eruption of Santorini at 1.8–2.7 × 1011 kg, some c.33– 49 times larger than the estimates of Sigurdsson et al. (1990)! However, there are some obvious caveats. The existence of very small amounts of mafic material is not disputed. This can support high sulphur concentrations. The question is the significance of this material. Michaud et al. (2000) argue that it indicates there was a large amount of non-erupted sulphur-rich basalt involved in the eruption process, and thus there may have been a sulphur-rich vapour phase present in the pre-eruption magma chamber (the mechanism believed to explain large sulphur releases not detected by petrologic analysis for several recent volcanic eruptions, such as Pinatubo: Gerlach et al., 1996). However, while this is possible, and the mafic material may indeed suggest that a sulphur-rich basalt magma was involved in the formation of the Minoan magma chamber, it is important to remember that this process occurred over several thousand years (up to about 17,000 years following Druitt et al., 1999). It is therefore likely that some, to much, of this sulphur will have leaked out and not have been retained until the eruption. Therefore, the extremely large sulphur yield calculated by Michaud et al. (2000) is very likely to be a significant overestimate. None the less, their data do point clearly to the likelihood of the presence of a pre-eruption sulphur-rich vapour in the magma chamber, and therefore to a larger total sulphur release than what now must be seen as the extreme minima calculated previously solely from the Minoan magma. In turn, the arguments against the correlation of the Minoan eruption of Santorini with the significant volcanic sulphur-derived signals in the Arctic ice based on a supposed very low sulphur release for Santorini no longer necessarily apply. The Minoan

284

S.W. MANNING AND D.A. SEWELL

eruption of Santorini may therefore represent the major volcanic signal in the seventeenthcentury BC ice-core record, and perhaps be responsible for the tree-ring growth anomalies of 1628–1627 BC (see further in Manning et al., 2002a). 5 Manning (1999: 72, fig. 21) lacks indication of tephra depths in cm: for corrected figure, see http://www.rdg.ac.uk/~lasmanng/testoftime.html

REFERENCES Alley, R.B., Shuman, C.A., Meese, D.A., Gow, A.J., Taylor, K.C., Cuffey, K.M., Fitzpatrick, J.J., Grootes, P.M., Zielinski, G.A., Ram, M., Spinelli, G. and Elder, B. (1997) Visual-stratigraphic dating of the GISP2 ice core: basis, reproducibility, and application. Journal of Geophysical Research 102: 26367–81. Ambrose, S.H. (1998) Late Pleistocene human population bottlenecks, volcanic winter, and differentiation of modern humans. Journal of Human Evolution 34: 623–51. Armienti, P., Macedonio, G. and Pareschi, M.T. (1988) A numerical-model for simulation of tephra transport and deposition – applications to May 18, 1980, MountSt-Helens Eruption. Journal of Geophysical Research 93: 6463–76. Arnold, D. (1988) Famine, Social Crisis and Historical Change. Oxford: Blackwell. Baillie, M.G.L. (1990) Irish tree rings and an event in 1628 BC. In D.A. Hardy and A.C. Renfrew (eds) Thera and the Aegean world III. Volume Three: Chronology, 160–6. London: The Thera Foundation. Baillie, M.G.L. (1991) Marking in marker dates: towards an archaeology with historical precision. World Archaeology 23: 233–43. Baillie, M.G.L. (1994) Dendrochronology raises questions about the nature of the AD 536 dust-veil event. The Holocene 4: 212–17. Baillie, M.G.L. (1995) A Slice through Time: Dendrochronology and Precision Dating. London: Batsford. Baillie, M.G.L. (1998) Bronze Age myths expose archaeological shortcomings? A reply to Buckland et al. 1997. Antiquity 72: 425–7. Baillie, M.G.L. (1999) Exodus to Arthur: Catastrophic Encounters with Comets. London: Batsford. Baillie, M.G.L. and Munro, M.A.R. (1988) Irish tree rings, Santorini and volcanic dust veils. Nature 332: 344–6. Bicknell, P. (1993) Blue suns, the son of heaven, and the chronology of the volcanic veil of the 40s BC. Ancient History Bulletin 7: 2–11. Bicknell, P. (2000) Late Minoan IB marine ware, the marine environment of the Aegean, and the Bronze Age eruption of the Thera volcano. In W.J. McGuire, D.R. Griffiths, P.L. Hancock and I.S. Stewart (eds) The Archaeology of Geological Catastrophes, 95–103. London: Geological Society Special Publication 171. Blong, R.J. (1980) The possible effects of Santorini tephra fall on Minoan Crete. In C. Doumas (ed.) Thera and the Aegean World II, 217–26. London: The Thera Foundation. Blong, R.J. (1982) The Time of Darkness: Local Legends and Volcanic Reality in Papua New Guinea. Seattle: University of Washington Press. Blong, R.J. (1984) Volcanic Hazards: A Sourcebook on the Effects of Eruptions. Orlando: Academic Press. Bolton, K. (1976) Addendum to J.V. Luce’s article: ‘Thera and the devastation of Minoan Crete: a new interpretation of the evidence’. American Journal of Archaeology 80: 17–18. Bond, A. and Sparks, R.S.J. (1976) The Minoan eruption of Santorini, Greece. Journal of the Geological Society 132: 1–16. Briffa, K.R., Jones, P.D., Schweingruber, F.H. and Osborne, T.J. (1998) Influence of volcanic eruptions on Northern Hemisphere summer temperatures over the past 600 years. Nature 393: 450–5.

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

285

Bruins, H.J. and van der Plicht, J. (1996) The Exodus enigma. Nature 382: 213–14. Buckland, P.C., Dugmore, A.J. and Edwards, K.J. (1997) Bronze Age myths? Volcanic activity and human response in the Mediterranean and North Atlantic regions. Antiquity 71: 581–93. Buckland, P.C., Dugmore, A.J. and Edwards, K.J. (1998) Comment. Antiquity 72: 431– 2. Cita, M.B. and Aloisi, G. (2000) Deep-sea tsunami deposits triggered by the explosion of Santorini (3500 y BP), eastern Mediterranean. Sedimentary Geology 135: 181–203. Cita, M.B. and Rimoldi, B. (1997) Geological and geophysical evidence for a Holocene tsunami deposit in the eastern Mediterranean deep-sea record. Journal of Geodynamics 24: 293–304. Cita, M.B., Camerlenghi, A. and Rimoldi, B. (1996) Deep-sea tsunami deposits in the eastern Mediterranean: new evidence and depositional models. Sedimentary Geology 104: 155–73. Clausen, H.B., Hammer, C.U., Hvidberg, C.S., Dahl-Jensen, D., Steffensen, J.P., Kipfstuhl, J. and Legrand, M. (1997) A comparison of the volcanic records over the past 4000 years from the Greenland Ice Core Project and Dye 3 Greenland ice cores. Journal of Geophysical Research 102: 26707–23. Cole-Dai, J., Mosley-Thompson, E., Wight, S.P. and Thompson, L.C. (2000) A 4100year record of explosive volcanism from an east Antarctica ice core. Journal of Geophysical Research 105: 24431–24442. Courtillot, V. (1994) Mass extinctions in the last 300 million years: one impact and seven flood basalts? Israel Journal of Earth Sciences 43: 255–66. Courtillot, V. (1999) Evolutionary Catastrophes: The Science of Mass Extinction. Translated by J. McClinton. Cambridge: Cambridge University Press. Courtillot, V., Jaeger, J.-J., Yang, Z., Feraud, G. and Hofmann, C. (1996) The influence of continental flood basalts on mass extinctions: where do we stand? Geological Society of America Special Paper 307: 513–25. Cullen, H.M., deMenocal, P.B., Hemming, S., Hemming, G., Brown, F.H., Guilderson, T. and Sirocko, F. (2000) Climate change and the collapse of the Akkadian empire: evidence from the deep sea. Geology 28: 379–82. Dalfes, H.N., Kukla, G. and Weiss, H. (eds) (1997) Third Millennium BC Climate Change and Old World Collapse. NATO ASI Series I, vol. 49. Berlin: Springer-Verlag. Decker, R.W. (1990) How often does a Minoan eruption occur? In D.A. Hardy, J. Keller, V.P. Galanopoulos, N.C. Flemming and T.H. Druitt (eds) Thera and the Aegean World III. Volume Two: Earth Sciences, 444–52. London: The Thera Foundation. Devine, J.D., Sigurdsson, H., Davis, A.N. and Self, S. (1984) Estimates of sulfur and chlorine yield to the atmosphere from volcanic eruptions and potential climatic effects. Journal of Geophysical Research 89: 6309–25. Dominey-Howes, D.T.M. (1996) The geomorphology and sedimentology of five tsunamis in the Aegean Sea region, Greece. Ph.D. dissertation, Coventry University, UK. Doumas, C. (1990) Archaeological observations at Akrotiri relating to the volcanic destruction. In D.A. Hardy and A.C. Renfrew (eds) Thera and the Aegean World III. Volume Three: Chronology, 48–9. London: The Thera Foundation. Downey, W.S. and Tarling, D.H. (1984) Archaeomagnetic dating of Santorini volcanic eruptions and fired destruction levels of Late Minoan civilisation. Nature 309: 519–23. Driessen, J. and Macdonald, C.F. (1997) The Troubled Island: Minoan Crete Before and After the Santorini Eruption. Aegaeum 17. Liège: Université de Liège, Histoire de l’art et archéologie de la Grèce antique. Driessen, J. and Macdonald, C.F. (2000) The eruption of the Santorini volcano and its effects on Minoan Crete. In W.J. McGuire, D.R. Griffiths, P.L. Hancock and I.S. Stewart (eds) The Archaeology of Geological Catastrophes, 81–93. London: Geological Society Special Publication 171.

286

S.W. MANNING AND D.A. SEWELL

Druitt, T.H., Edwards, L., Mellors, R.M., Pyle, D.M., Sparks, R.S.J., Lanphere, M., Davies, M. and Barreirio, B. (1999) Santorini Volcano. Memoir of the Geological Society 19. London: Geological Society. Eastwood, W.J., Pearce, N.J.G., Westgate, J.A. and Perkins, W.T. (1998) Recognition of Santorini (Minoan) tephra in lake sediments from Gölhisar Gölü, southwest Turkey by laser ablation ICP-MS. Journal of Archaeological Science 25: 677–87. Flohn, H. (1981) Short-term climatic fluctuations and their economic role. In T.M. Wigley, M.J. Ingram and C. Farmer (eds) Studies in Past Climates and their Impact on Man, 310–18. Cambridge: Cambridge University Press. Forsyth, P.Y. (1988) In the wake of Etna, 44 BC. Classical Antiquity 7: 49–57. Forsyth, P.Y. (1998) Thera in the Bronze Age. New York: Peter Lang. Fouqué, F.A. (1999) Santorini and its Eruptions. Translated and annotated by A.R. McBirney. Baltimore: The Johns Hopkins University Press. Francaviglia, V. (1990) Sea-borne pumice deposits of archaeological interest on Aegean and eastern Mediterranean beaches. In D.A. Hardy and A.C. Renfrew (eds) Thera and the Aegean World III. Volume Three: Chronology, 127–34. London: The Thera Foundation. Frankel, C. (1996) Volcanoes of the Solar System. Cambridge: Cambridge University Press. Friedrich, W.L. (2000) Fire in the Sea: the Santorini Volcano: Natural History and the Legend of Atlantis. Translated by A.R. McBirney. Cambridge: Cambridge University Press. Garnsey, P. (1988) Famine and Food Supply in the Graeco-Roman World: Responses to Risk and Crisis. Cambridge: Cambridge University Press. Garnsey, P. (1992) Famine in history. In J. Bourriau (ed.) Understanding Catastrophe: Its Impact on Life on Earth, 145–78. Cambridge: Cambridge University Press. Gerlach, T.M., Westrich, H.R., Casadevall, T.J. and Finnegan, D.L. (1994) Vapor saturation and accumulation in magmas of the 1989–1990 eruption of Redoubt Volcano, Alaska. Journal of Volcanology and Geothermal Research 62: 317–37. Gerlach, T.M., Westrich, H.R. and Symonds, R.B. (1996) Preeruption vapor in magma of the climactic Mount Pinatubo eruption: source of the giant stratospheric sulphur dioxide cloud. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruptions and Lahars of Mount Pinatubo, Philippines, 415–33. Seattle: University of Washington Press. Goff, J., Crozier, M., Sutherland, V., Cochran, U. and Shane, P. (1998) Possible tsunami deposits from the 1855 earthquake, North Island, New Zealand. In I.S. Stewart and C. Vita-Finzi (eds) Coastal Tectonics, 353–74. London: Geological Society Special Publication 146. Grudd, H., Briffa, K.R., Gunnarson, B.E. and Linderholm, H.W. (2000) Swedish tree rings provide new evidence in support of a major, widespread environmental disruption in 1628 BC. Geophysical Research Letters 27: 2957–60. Guichard, F., Carey, S., Arthur, M.A., Sigurdsson, S. and Arnold, M. (1993) Tephra from the Minoan eruption of Santorini in sediments of the Black Sea. Nature 363: 610– 12. Gunn, J.D. (ed.) (2000) The Years Without Summer: Tracing AD 536 and its Aftermath. BAR S872. Oxford: Archaeopress. Hammer, C.U. (1984) Traces of Icelandic eruptions in the Greenland ice sheet. Jökull 34: 51–65. Hammer, C.U. (2000) What can Greenland ice core data say about the Thera eruption in the second millenium BC? In M. Bietak (ed.) The Synchronisation of Civilisations in the Eastern Mediterranean in the Second Millennium BC. Proceedings of an international symposium at Schloß Haindorf, 15–17 November 1996 and at the Austrian Academy, Vienna, 11–12 May 1998, 35–7. Contributions to the Chronology of the Eastern Mediterranean I. Wien: Österreichischen Akademie der Wissenschaften. Hammer, C.U. and Clausen, H.B. (1990) The precision of ice-core dating. In D.A. Hardy and A.C. Renfrew (eds) Thera and the Aegean World III. Volume Three: Chronology, 174–8. London: The Thera Foundation.

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

287

Hammer, C.U., Clausen, H.B. and Dansgaard, W. (1980) Greenland ice sheet evidence of post-glacial volcanism and its climatic impact. Nature 288: 230–5. Hammer, C.U., Clausen, H.B., Friedrich, W.L. and Tauber, H. (1987) The Minoan eruption of Santorini in Greece dated to 1645 BC? Nature 328: 517–19. Hammer, C.U., Clausen, H.B. and Langway, C.C. (1997) 50,000 years of recorded global volcanism. Climate Change 35: 1–15. Hammer, C.U., Kurat, G., Hoppe, P. and Clausen, H.B. (2001) Recent ice core analysis strengthens the arguments for a mid-seventeenth century BC eruption of Thera. Extended abstract for the Haindorf Castle SCIEM Conference April [in fact 2–7 May], Haindorf, Lower Austria, 2001. Harington, C.R. (ed.) (1992). The Year Without a Summer? World Climate in 1816. Ottawa: Canadian Museum of Nature. Heiken, G. and McCoy, F.W. (1984) Caldera development during the Minoan eruption, Thira, Cyclades, Greece. Journal of Geophysical Research 89: 8441–62. Hieke, W. and Werner, F. (2000) The Augias megaturbadite in the central Ionian Sea (central Mediterranean) and its relation to the Holocene Santorini event. Sedimentary Geology 135: 205–18. Hiller, S. (1978) Minoan Qe-Ra-Si-Ja. The religious impact of the Thera volcano on Minoan Crete. In C. Doumas (ed.) Thera and the Aegean World I, 675–9. London: The Thera Foundation. Housley, R.A., Manning, S.W., Cadogan, G., Jones, R.E. and Hedges, R.E.M. (1999) Radiocarbon, calibration, and the chronology of the Late Minoan IB phase. Journal of Archaeological Science 26: 159–71. Hughes, M.K. (1988) Ice layer dating of the eruption of Santorini. Nature 335: 211–12. Johnsen, S.J., Clausen, H.B., Dansgaard, W., Fuhrer, K., Gundestrup, N., Hammer, C.U., Iversen, P., Jouzel, J., Stauffer, B. and Steffensen, J.P. (1992) Irregular glacial interstadials recorded in a new Greenland ice core. Nature 359: 311–13. Kastens, K.A. and Cita, M.B. (1981) Tsunami-induced sediment transport in the abyssal Mediterranean Sea. Geological Society of America Bulletin 92: 845–57. Keys, D. (1999) Catastrophe: An Investigation into the Origins of the Modern World. London: Century. Kuniholm, P.I., Kromer, B., Manning, S.W., Newton, M., Latini, C.E. and Bruce, M.J. (1996) Anatolian tree-rings and the absolute chronology of the east Mediterranean 2220–718 BC. Nature 381: 780–3. LaMarche, V.C. Jr and Hirschboeck, K.K. (1984) Frost rings in trees as records of major volcanic eruptions. Nature 307: 121–6. Limbrey, S. (1990) Soil studies at Akrotiri. In D.A. Hardy, J. Keller, V.P. Galanopoulos, N.C. Flemming and T.H. Druitt (eds) Thera and the Aegean World III. Volume Two: Earth Sciences, 377–82. London: The Thera Foundation. Luce, J.V. (1976) Thera and the devastation of Minoan Crete: a new interpretation of the evidence. American Journal of Archaeology 80: 9–16. McCormick, M.P., Thomason, L.W. and Trepte, C.R. (1995) Atmospheric effects of the Mt Pinatubo eruption. Nature 373: 399–404. McCoy, F.W. and Heiken, G. (2000a) The Late-Bronze Age explosive eruption of Thera (Santorini), Greece: regional and local effects. In F.W. McCoy and G. Heiken (eds) Volcanic Hazards and Disasters in Human Antiquity, 43–70. Boulder, CO: Geological Society of America Special Paper 345. McCoy, F.W. and Heiken, G. (2000b) Tsunami generated by the Late Bronze Age eruption of Thera (Santorini), Greece. Pure and Applied Geophysics 157: 1227–56. MacGillivray, J.A., Sackett, L.H. and Driessen, J.M. (1998) Excavations at Palaikastro, 1994 and 1996. Annual of the British School at Athens 93: 221–68. Macedonio, G., Pareschi, M.T. and Santacroce, R. (1988) A numerical simulation of the plinian fall phase of 79 AD eruption of Vesuvius. Journal of Geophysical Research 93: 14817–27.

288

S.W. MANNING AND D.A. SEWELL

Manning, S.W. (1987) The volcano of Thera and the destruction of Minoan Crete. Kritikon Chronikon Tomos KZ: 59–85. Manning, S.W. (1992) Thera, sulphur, and climatic anomalies. Oxford Journal of Archaeology 11: 245–53. Manning, S.W. (1999) A Test of Time: The Volcano of Thera and the Chronology and History of the Aegean and East Mediterranean in the Mid Second Millennium BC. Oxford: Oxbow Books. Manning, S.W., Kromer, B., Kuniholm, P.I. and Newton, M.W. (2001) Anatolian treerings and a new chronology for the east Mediterranean Bronze–Iron Ages. Science 294: 2532–5. Manning, S.W., Sewell, D.A. and Herscher, E. (2002a) Late Cypriot IA maritime trade in action: underwater survey at Maroni–Tsaroukkas and the contemporary east Mediterranean trading system. Annual of the British School at Athens 97 in press. Manning, S.W., Bronk Ramsey, C., Doumas, C., Marketou, T., Cadogan, G. and Pearson, C.L. (2002b) New evidence for an early date for the Aegean Late Bronze Age and Thera eruption. Antiquity 76: in press. Marinatos, S. (1939) The volcanic destruction of Minoan Crete. Antiquity 13: 425–39. Meese, D.A., Gow, A.J., Alley, R.B., Zielinski, G.A., Grootes, P.M., Ram, M., Taylor, K.C., Mayewski, P.A. and Bolzan, J.F. (1997) The Greenland Ice Sheet Project 2 depth-age scale: methods and results. Journal of Geophysical Research 102: 26411–23. Michaud, V., Clocchiatti, R. and Sbrana, S. (2000) The Minoan and post-Minoan eruptions, Santorini (Greece), in the light of melt inclusions: chlorine and sulphur behaviour. Journal of Volcanology and Geothermal Research 99: 195–214. Minoura, K., Imamura, F., Kuran, U., Nakamura, T., Papadopoulos, G.A., Takahashi, T. and Yalciner, A.C. (2000) Discovery of Minoan tsunami deposits. Geology 28: 59–62. Monaghan, J.J., Bicknell, P.J., Humble, R.J. (1994) Volcanoes, tsunamis and the demise of the Minoans. Physica D 77: 217–28. Money, J. (1973) The destruction of Acrotiri. Antiquity 47: 50–3. Newhall, C.G. and Self, S. (1982) The volcanic explosivity index (VEI): an estimate of explosive magnitude for historical volcanism. Journal of Geophysical Research 81: 1231–8. Newhall, C.G., Power, J.A. and Punongbayan, R.S. (2002) Perspectives: Pinatubo eruption. “To Make Grow”. Science 295: 1241–2. Oliver-Smith, A. (1996) Anthropological research on hazards and disasters. Annual Review of Anthropology 25: 303–28. Oppenheimer, C. (2002) Limited global change due to the largest known Quaternary eruption, Toba ≈ 74kyr BP? Quaternary Science Reviews: in press. Page, D.L. (1970) The Santorini Volcano and the Desolation of Minoan Crete. London: Society for the Promotion of Hellenic Studies. Page, D.L. (1971) The volcano at Santorini and the devastation of Minoan Crete: an introduction to the historical and archaeological problem. In S. Marinatos and D. Ninkovich (eds) Acta of the 1st International Scientific Congress on the Volcano of Thera, 371–6. Athens: Archaeological Services of Greece. Pálfy, J. and Smith, P.L. (2000) Synchrony between Early Jurassic extinction, oceanic anoxic event, and the Karoo-Ferrar flood basalt volcanism. Geology 28: 747–50. Pang, K.D. and Chou, H.-H. (1985) Three very large volcanic eruptions in antiquity and their effects on the climate of the ancient world. EOS. Transactions, American Geophysical Union 66: 816. Pang, K.D., Pieri, D. and Chou, H.-H. (1986) Climatic impacts of the 44–42 BC eruptions of Etna, reconstructed from ice core and historical records. EOS. Transactions, American Geophysical Union 67: 880–1. Pang, K.D., Srivastava, S.K. and Chou, H.-H. (1988) Climatic impacts of past volcanic eruptions: inferences from ice core, tree ring and historical data. EOS. Transactions, American Geophysical Union 69: 1062. Pang, K.D., Keston, R., Srivastava, S.K. and Chou, H.-H. (1989) Climatic and hydro-

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

289

logic extremes in early Chinese history: possible causes and dates. EOS. Transactions, American Geophysical Union 70: 1095. Pinto, J.P., Turco, R.P. and Toon, O.B. (1989) Self-limiting physical and chemical effects in volcanic clouds. Journal of Geophysical Research 94: 11165–74. Post, J.D. (1977) The Last Great Subsistence Crisis in the Western World. Baltimore: Johns Hopkins University Press. Pyle, D.M. (1989) Ice-core acidity peaks, retarded tree growth and putative eruptions. Archaeometry 31: 88–91. Pyle, D.M. (1990a) New estimates for the volume of the Minoan eruption. In D.A. Hardy, J. Keller, V.P. Galanopoulos, N.C. Flemming and T.H. Druitt (eds) Thera and the Aegean World III. Volume Two: Earth Sciences, 113–21. London: The Thera Foundation. Pyle, D.M. (1990b) The application of tree-ring and ice-core studies to the dating of the Minoan eruption. In D.A. Hardy and A.C. Renfrew (eds) Thera and the Aegean World III. Volume Three: Chronology, 167–73. London: The Thera Foundation. Pyle, D. (1998) How did the summer go? Nature 393: 415–17. Rampino, M.R. and Self, S. (1984) Sulfur-rich volcanic eruptions and stratospheric aerosols. Nature 310: 677–9. Rampino, M.R. and Self, S. (1992) Volcanic winter and accelerated glaciation following the Toba super-eruption. Nature 359: 50–2. Rampino, M.R. and Self, S. (1993) Climate–volcanism feedback and the Toba eruption of ~74,000 years ago. Quaternary Research 40: 269–80. Rampino, M.R. and Self, S. (2000) Volcanism and biotic extinctions. In H. Sigurdsson (ed.) Encyclopedia of Volcanoes, 1083–91. San Diego: Academic Press. Rampino, M.R. and Stothers, R.B. (1988) Flood basalt volcanism during the last 250 million years. Science 241: 663–7. Rampino, M.R., Self, S. and Fairbridge, R.W. (1988) Volcanic winters. Annual Review of Earth and Planetary Sciences 16: 73–99. Ramsey, J.T. and Licht, A.L. (1997) The Comet of 44 BC and Caesar’s Funeral Games. Atlanta: Scholars Press. Renfrew, C. (1979) The eruption of Thera and Minoan Crete. In P.D. Sheets and D.K. Grayson (eds) Volcanic Activity and Human Ecology, 565–85. New York: Academic Press. Sadler, J.P. and Grattan, J.P. (1999) Volcanoes as agents of past environmental change. Global and Planetary Change 21: 181–96. Scarth, A. (1999) Vulcan’s Fury: Man Against the Volcano. New Haven: Yale University Press. Scuderi, L.A. (1990) Tree-ring evidence for climatically effective volcanic eruptions. Quaternary Research 34: 67–85. Self, S., Zhao, J.-X., Holasek, R.E., Torres, R.C. and King, A.J. (1996) The atmospheric impact of the 1991 Mount Pinatubo eruption. In C.G. Newhall and R.S. Punongbayan (eds) Fire and Mud: Eruptions and Lahars of Mount Pinatubo, Philippines, 1089–115. Seattle: University of Washington. Sewell, D.A. (2001) Earth, air, fire and water. An elemental analysis of the Minoan eruption of the Santorini volcano. Ph.D. dissertation, University of Reading UK. Shaw, J.W., van de Moortel, A., Day, P.M. and Kilikoglou, V. (1997) A LMIA pottery kiln at Kommos, Crete. Aegaeum 16: 323–31. Shaw, J.W., van de Moortel, A., Day, P.M. and Kilikoglou, V. (2001) A LMIA Ceramic Kiln in South Central Crete: Function and Pottery Production. Hesperia Supplement 30. Princeton: American School of Classical Studies at Athens. Sheets, P.D. (1979) Environmental and cultural effects of the Ilopango eruption in Central America. In P.D. Sheets and D.K. Grayson (eds) Volcanic Activity and Human Ecology, 525–64. New York: Academic Press. Sherratt, A. (1997) Climatic cycles and behavioural revolutions: the emergence of modern humans and the beginning of farming. Antiquity 71: 271–87.

290

S.W. MANNING AND D.A. SEWELL

Sigurdsson, H., Carey, S. and Devine, J.D. (1990) Assessment of mass, dynamics and environmental effects of the Minoan eruption of Santorini volcano. In D.A. Hardy, J. Keller, V.P. Galanopoulos, N.C. Flemming and T.H. Druitt (eds) Thera and the Aegean World III. Volume Two: Earth Sciences, 100–12. London: The Thera Foundation. Simkin, T. and Siebert, L. (1994) Volcanoes of the World. Tucson: Geoscience Press. Simkin, T. (1994) Distant effects of volcanism – how big and how often? Science 264: 913–14. Soles, J.S., Taylor, S.R. and Vitaliano, C.J. (1995) Tephra samples from Mochlos and their chronological implications for Neopalatial Crete. Archaeometry 37: 385–93. Southon, J. (2002) A first step to reconciling the GRIP and GISP2 ice-core chronologies 0–14,500 yr BP. Quaternary Research 57: 32–7. Sparks, R.S.J. (1986) Comment on ‘The volcanic eruption of Thera and its effects on the Mycenaean and Minoan civilizations’ by I.G. Nixon. Journal of Archaeological Science 13: 289–90. Sparks, R. and Huang, T.C. (1980) The volcanological significance of deep-sea ash layers associated with ignimbrites. The Geological Magazine 117: 425–36. Sparks, R.S.J., Brazier, S., Huang, T.C. and Muerdter, D. (1983/1984) Sedimentology of the Minoan deep-sea tephra layer in the Aegean and eastern Mediterranean. Marine Geology 54: 131–67. Stothers, R.B. (1993) Flood basalts and extinction events. Geophysical Research Letters 20: 1399–402. Stothers, R.B. and Rampino, M.R. (1983a) Historic volcanism, European dry fogs, and Greenland acid precipitation, 1500 BC to AD 1500. Science 222: 411–13. Stothers, R.B. and Rampino, M.R. (1983b) Volcanic eruptions in the Mediterranean before AD 630 from written and archaeological sources. Journal of Geophysical Research 88: 6357–71. Sullivan, D.G. (1988) The discovery of Santorini Minoan tephra in western Turkey. Nature 333: 552–4. Sullivan, D.G. (1990) Minoan tephra in lake sediments in western Turkey: dating the eruption and assessing the atmospheric dispersal of the ash. In D.A. Hardy and A.C. Renfrew (eds) Thera and the Aegean World III. Volume Three: Chronology, 114–19. London: The Thera Foundation. Sullivan, D.G. (1993) Effects of the Santorini eruption on Bronze Age settlement in Aegean Turkey. American Journal of Archaeology 97: 330. Suzuki, T. (1983) A theoretical model for dispersion of tephra. In D. Shimozuru and I. Yokoyama (eds) Arc Volcanism: Physics and Tectonics, 95–113. Tokyo: Terra Scientific Publishing Company. Thórarsinsson, S. (1971) Some comments on the Minoan eruption of Santorini. In D. Ninkovich and S. Marinatos (eds), Acta of the 1st International Scientific Congress on the Volcano of Thera, 263–75. Athens: Archaeological Services of Greece. Thórarinsson, S. (1979) On the damage caused by volcanic eruptions with special reference to tephra and gases. In P.D. Sheets and D.K. Grayson (eds) Volcanic Activity and Human Ecology, 124–59. New York: Academic Press. Walker, G.P.L., Wilson, L. and Bowell, E.L.G. (1971) Explosive volcanic eruptions – I. The rate of fall of pyroclasts. Geophysical Journal of the Royal Astronomical Society 22: 377–83. Warner, R. (1990) The ‘prehistoric’ Irish annals: fable or history? Archaeology Ireland 4: 30–3. Warren, P. (1984) Absolute dating of the Bronze Age eruption of Thera (Santorini). Nature 308: 492–3. Warren, P. (1999) LMIA: Knossos, Thera, Gournia. In P.P. Betancourt, V. Karageorghis, R. Laffineur and W.-D. Niemeier (eds) Meletemata: Studies in Aegean Archaeology Presented to Malcolm H. Wiener as he Enters his 65th Year, 893–903. Aegaeum 20. Liège and Austin: Université de Liège, Service d’Histoire de l’art et archéologie de la Grèce

VOLCANOES AND HISTORY: THE CASE OF SANTORINI

291

antique, and Program in Aegean Scripts and Prehistory, The University of Texas at Austin. Warren, P. (2001) Review of The Troubled Island: Minoan Crete Before and After the Santorini Eruption by Jan Driessen and Colin MacDonald. American Journal of Archaeology 105: 115–18. Watkins, N.D., Sparks, R.S.J., Sigurdsson, H., Huang, T.C., Federman, A., Carey, S. and Ninkovich, D. (1978) Volume and extent of the Minoan tephra from Santorini volcano: new evidence from deep sea sediment cores. Nature 271: 122–6. Weiss, H., Courty, M.-A., Wetterstrom, W., Guichard, F., Senior, L., Meadow, R. and Curnow, A. (1993) The genesis and collapse of third millennium north Mesopotamian civilization. Science 261: 995–1004. Woods, A.W. and Bursik, M.I. (1991) Particle fallout, thermal disequilibrium and volcanic plumes. Bulletin of Volcanology 53: 559–70. Yang, Q., Mayewski, P.A., Zielinski, G.A., Twickler, M. and Taylor, K.C. (1996) Depletion of atmospheric nitrate and chloride as a consequence of the Toba volcanic eruption. Geophysical Research Letters 23: 2513–16. Zielinski, G.A. (1995) Stratospheric loading and optical depth estimates of explosive volcanism over the last 2100 years derived from the Greenland Ice Sheet Project 2 ice core. Journal of Geophysical Research 100: 20937–55. Zielinski, G.A. and Germani, M.S. (1998) New ice-core evidence challenges the 1620s BC age for the Santorini (Minoan) eruption. Journal of Archaeological Science 25: 279–89. Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S., Twickler, M.S., Morrison, M., Meese, D.A., Gow, A.J. and Alley, R.B. (1994) Record of volcanism since 7000 BC from the GISP2 Greenland ice core and implications for the volcano-climate system. Science 264: 948–52. Zielinski, G.A., Germani, M.S., Larsen, G., Baillie, M.G.L., Whitlow, S., Twickler, M.S. and Taylor, K. (1995) Evidence of the Eldgjá (Iceland) eruption in the GISP2 Greenland ice core: relationship to eruption processes and climatic conditions in the tenth century. The Holocene 5: 129–40. Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S. and Twickler, M.S. (1996a) Potential atmospheric impact of the Toba mega-eruption ~71,000 years ago. Geophysical Research Letters 23: 837–40. Zielinski, G.A., Mayewski, P.A., Meeker, L.D., Whitlow, S. and Twickler, M.S. (1996b) A 110,000-year record of explosive volcanism from the GISP2 (Greenland) ice core. Quaternary Research 45: 109–18.

16

What makes a disaster? A long-term view of volcanic eruptions and human responses in Papua New Guinea ROBIN TORRENCE

VULNERABLE OR RESILIENT? Modern scholars of natural hazards predict that in developed countries with complex, state-level social organisations, disasters can be mitigated through aid programmes, redistribution of resources, etc., whereas egalitarian societies with smaller social networks have a wide range of responses, but are more likely to fail to adapt and will abandon their homeland (e.g. Chester, 1993: table 8.4). Anthropologists have often painted quite a different picture in which traditional societies are and have been very resilient and adaptable in the face of extreme climatic events. In their view modernisation has often undermined capable traditional means for coping with hazards. They therefore argue that the vulnerability of indigenous groups witnessed in recent disasters is a result of their marginalisation through globalisation and externally forced changes (e.g. OliverSmith, 1996: 312–14). A comparison by Sheets et al. (1991; cf. Sheets and McKee, 1994) of the prehistory of the Arenal Valley in Costa Rica with that of El Salvador and Panama substantiates the wider anthropological view that simple societies are quite resilient or even adapted to environmental hazards. Despite the occurrence of ten volcanic eruptions in Costa Rica during a period of 4,000 years, the archaeologists have reconstructed a quite remarkable picture of cultural stability. In contrast, they concluded that a major eruption of Ilopango volcano in El Salvador had disastrous consequences for Mayan civilisation and the Baru volcano severely undermined the prehistoric Bariles chiefdom society in Panama. According to Sheets et al. (1991: 446), simpler societies ‘appear to be more resilient in the aftermath of explosive eruptions’ than more complex societies because the latter are dependent on a built environment and economies characterised by ‘occupational specialisation, redistribution, and long distance trade routes’. These conclusions are limited, however, because the authors do not adequately account for differences in the severity of the events they have compared. One of the problems with trying to develop a general understanding of the

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 293

impact of extreme events on human societies is that scholars often focus on particular cases which we call ‘disasters’. Since a disaster is defined as a situation for which there is a failure to cope, only one possible outcome is considered. Consequently, we know far too little about the whole range of situations, including those in which there is little if any impact. In addition, the focus on the disaster itself has drawn attention away from how societies recover and the very longterm, follow-on effects of the events. The focus only on hazards that have had catastrophic consequences for societies therefore makes it difficult to understand the broader relationships between the nature of extreme climatic events and human responses. Furthermore, this approach has led to an overemphasis on the vulnerability of societies as the most important variable and has ignored differences in the character of the natural events that trigger disasters. I think that this one-sided view combined with the inattention to variability has led to the paucity of theory to explain why disasters take place in particular social contexts and not in others (cf. Oliver-Smith, 1996: 320). In this chapter I propose that archaeological case studies can make an important contribution to a broader understanding of what does and does not make a disaster. They can also provide data that enable a more careful consideration of the different kinds of consequences that follow from variations in social and environmental contexts. By examining the history of an area which has experienced a number of natural events with differing levels of severity and that have occurred over a relatively long time period, one can compare and contrast the impacts on a range of societies within the same general environmental setting. In this approach the emphasis on ‘disaster’ as the only outcome is diverted to a more detailed assessment of variation in the nature of the impacts on human groups. I illustrate the value of research of this type by summarising the findings of an archaeological investigation into the effects of a series of volcanic eruptions over at least 6,000 years in the province of West New Britain in Papua New Guinea (Fig. 16.1). The archaeological case study also raises some important points about the theory and methodology of disaster research. It is shown that problems confronted in the archaeological analysis, particularly in terms of the methods for measuring the severity of events and for monitoring impacts on societies, are as yet not fully developed and need more careful consideration. Second, I argue that we need to look at a wider range of factors when evaluating why some events led to disasters and others did not. Not only are social variables important, as stressed by modern hazard managers and Sheets et al. (1991), but more care needs to be taken to understand the impacts on the natural environment and the follow-on effects over the longer term. Finally, the difficult issue of choosing the appropriate temporal and spatial scales both for monitoring impacts and for understanding how societies recover from disasters is discussed. Although solutions are not found for all the limitations identified, the long-term perspective derived from archaeological case studies is clearly shown to provide new insights into how disasters can be defined and monitored.

294

R. TORRENCE

Figure 16.1 Location of volcanoes and relevant archaeological sites in West New Britain

WEST NEW BRITAIN CASE STUDY A collaborative archaeological research project based in the province of West New Britain in Papua New Guinea (Fig. 16.1) has examined the long-term consequences of disasters from volcanic eruptions on various forms of social and economic organisation. The research illustrates the benefits and limitations of studying the effects of natural disasters on the prehistory of a region. A more detailed presentation of our results is presented in Torrence et al. (2000) and Torrence et al. (1999) (also cf. Torrence, 1993; Torrence and Webb, 1992; Torrence and Boyd, 1996, 1997). The West New Britain case study is important because it raises a number of issues concerning the relevant variables and scales of analysis that are necessary for understanding why some events become disasters. Research by Sheets and his colleagues (Sheets et al., 1991; Sheets and McKee, 1994) in the Arenal Valley, Costa Rica, sets the scene for my discussion of volcanic activity in West New Britain, Papua New Guinea. Their work has admirably demonstrated how archaeological studies of disasters can benefit a broader understanding of human and environmental interactions. By examining a very long period of time within a single area, one can compare and contrast the effects of a series of natural events on societies. In case studies of this type some variables can be considered as constant since the groups would have inhabited a

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 295

roughly similar environment and are likely to have shared cultural traditions. Sheets et al. (1991) were unable to link the incidence of volcanic events to major cultural changes and therefore concluded that in this case the societies were extremely resilient. Although they may be correct, I feel that they have not adequately considered the relatively low severity of the eruptions they have studied. If the impacts of the Arenal eruptions were not very serious, then there would have been no reason for change, regardless of the nature of the society at the time. Indeed, this seems to have been the case. The volume of downwind tephra for the nine eruptions averaged 0.16 km3 and was never larger than 0.3 km3 (Melson, 1994: 39) and the stresses on vegetation are assumed to have ‘become small at 20 km to 30 km’ from the volcano (Sheets, 1994: 325). In contrast, Sheets et al. (1991) propose that the Ilopango eruption had a greater impact because complex societies are more susceptible to disasters. However, the Ilopango eruption, with a volume of 30 km3, was many orders of magnitude larger than the Arenal eruptions. It had a very great environmental impact and destroyed vegetation and human constructions over thousands of square kilometres (Sheets, 1979). It therefore seems likely that the relative scales of the natural hazards discussed by Sheets et al. (1991) played at least as much a part in causing the different outcomes as did the variation in the social organisation of the groups affected by the volcanic eruptions. The West New Britain study has also compared the social and cultural effects of events with varying degrees of severity, although all the eruptions were considerably larger than those of Arenal. As in Costa Rica the fairly egalitarian societies of Papua New Guinea were also very resistant to major change despite facing potentially much more serious catastrophes, but there was nevertheless considerable variability in how these societies responded to the different events. In some cases the volcanic eruptions caused serious human disasters, whereas for others the effects were either less problematic or, when viewed over a long time span, were not disasters at all. Although the data demonstrate that vulnerability is important for understanding disasters, the long-term archaeological studies illustrate that the nature of the triggering factor also causes variation in human responses. This seemingly obvious fact appears to have been overlooked in some studies of disasters. As illustrated below, then, the first step in analysing the impacts of a severe natural event is to assess its severity independent of its effects on cultural processes. Severity of eruptions Whereas volcanic eruptions may have had disastrous effects on human populations, they can have enormous benefits for archaeological research. The tephras from these short-lived, specific events provide distinctive marker beds to establish a relative chronology and can often be applied to a very large region (e.g. Machida and Sugiyama, Chapter 17, this volume; Dugmore, 1989; Sheets and Grayson, 1979; Pilcher and Hall, 1996; Sheets et al., 1991; Sheets and McKee, 1994; Zeidler and Isaacson, 1999). Using a combination of stratigraphy, macroscopic properties, refractive indices, and SEM microprobe geochemical analysis, it has

296

R. TORRENCE

been possible to match a number of tephras found in archaeological contexts in West New Britain with their source volcano and specific eruption (Machida et al., 1996; Torrence et al., 2000). We can therefore directly correlate the stratigraphy of sites on the north coast in the Willaumez Peninsula area (e.g. Torrence et al., 1990; Specht et al., 1991; Torrence et al., 1999) with those in the interior (e.g. Yombon) and south coast of the island (e.g. Lolmo cave) up to 100 km away (Gosden et al., 1994; Pavlides, 1993; Pavlides and Gosden, 1994) (Fig. 16.1). In addition, a combination of geological and archaeological research has yielded a relatively large number of radiocarbon assays which can be used to establish fairly accurate dates for many of these eruptions (cf. Torrence et al., 2000; Machida et al., 1996). To date our research has concentrated on five major eruptions from the two volcanic centres of Witori and Dakataua (Fig. 16.1). Their characteristics are summarised in Tables 16.1–16.3. In the northern part of the Willaumez Peninsula at Garua Island and Bitokara, Witori tephras from W-K1 at 5900 BP, W-K2 at Table 16.1 Summary of major Holocene volcanic events in West New Britain (based on Machida et al., 1996; Torrence et al., 2000) Eruption

Date BP (Cal. years)

Volume (km3)

Spatial scale (Rank, 1 is highest)

Type

Dakataua volcano Dk

1,000

10

4

Phreatomagmatic, plinian, ignimbrite forming

Witori volcano W-K4

1,400

6

3

Phreatomagmatic, plinian, ignimbrite forming

W-K3

1,700

6

3

Plinian

W-K2

3,600

30

1

Phreatomagmatic, plinian, ignimbrite forming

W-K1

5,900

10

2

Plinian, ignimbrite forming

Table 16.2 Average thickness (m) of airfall tephra in archaeological sites Locality Eruption

Numundo

Garu

Garua/Bitokara

Yombon

Dakataua W-K3/4 W-K4 W-K3 W-K2 W-K1

Absent

Absent

0.75 0.20

Absent 0.15

0.35 0.50 1.00 0.30

0.35 0.30 0.80 0.40

0.50 0.50

0.35 0.35

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 297

Table 16.3

Comparison between severity of event and cultural response

Severity

Eruption

Abandonment (years)

1 2 2 3 3

W-K2 W-K1 Dakataua W-K3 W-K4

250–800 1,000–1,600 0

}

0–300

3600 BP, and a mixed layer containing tephras from W-K3 and W-K4 dating to around 1700–1400 BP are well preserved. At Numundo and Garu plantations farther south and nearer the source volcano the full W-K (1–4) sequence plus 2–3 tephras belonging to the later W–H series from Witori is well preserved. These younger tephras are not included in this discussion because they have only recently been recognised in the archaeological record. Although in some cases they may have had significant impacts on human settlement, only a small part of the study area was affected by these eruptions (Torrence et al., 1999). In contrast to the Witori events, tephra from the Dakataua eruption of c.1100 BP is restricted to the northern part of the Willaumez Peninsula. Using a number of indices, including total volume of material ejected (Table 16.1), average depth of tephra in sites in the Willaumez Peninsula and Yombon regions (c.80 km to the south) (Table 16.2), and the spatial extent of the tephra (relative ranking with ‘1’ as the largest are given in Table 16.1; cf. isopach maps in Machida et al., 1996; Boyd et al., in press), a relative ranking of the overall severity of these volcanic events can be obtained. These are presented in Table 16.3, where ‘1’ represents the most severe case. The index of severity provides a general measure for comparing the postulated effects of each eruption on the prehistoric human population. In general terms the W-K2 eruption was by far the largest and would have destroyed virtually all the vegetation within most of the eastern twothirds of West New Britain. The W-K1 event would also have devastated a very large area, although it was emplaced during the wet season and was washed off much of the landscape shortly after the eruption. The highly variable thicknesses for the W-K1 tephra in Table 16.2 reflect the fact that most of the deposits we observed were reworked and redeposited. In contrast, the W-K3 and W-K4 eruptions produced smaller amounts of tephra that would have had more localised effects. The Dakataua eruption produced a great deal of material but its impacts would have been quite restricted in space. Assessing impacts The next stage of the analysis is to compare severity (in the West New Britain case using the severity index in Table 16.3) with some measure of how these events affected human settlement. There are several ways to conceptualise the impacts on humans. For example, Sheets et al. (1991) compare the timing of cultural changes in terms of phase or period boundaries with the date of eruptions. After both the Ilopango and Baru eruptions, the groups which reoccupied the areas

298

R. TORRENCE

were sufficiently different to warrant the assignment of a new cultural phase, but in the Arenal area the phases are not delimited by a volcanic event. They therefore conclude that the Arenal societies were more resistant to change. Changes in material culture in West New Britain are summarised in Table 16.4. In this case there is one major cultural boundary which is associated with a volcanic eruption and one that is not. The clearest association with material culture change and volcanism is the W-K2 eruption. After this event a very distinctive type of chipped stone tool disappeared and pottery appeared for the first time. In contrast, the other major change in material culture, the disappearance of pottery c.2,000 BP, is not associated with a particular eruption but is probably due to internally generated social change (cf. Specht and Gosden, 1997: 190). One problem with phase boundaries is that they are largely established by the nature of material culture: e.g. presence or absence of key artefact types or styles. These may not adequately reflect changes which took place in cultures like those in West New Britain, for which only a relatively simple repertoire of durable material culture has been preserved in the archaeological record. In evaluating the nature of impacts, one should probably consider a wide range of behaviour. For example, in the West New Britain case settlement and subsistence patterns changed in subtle ways throughout the past 6,000 years (Torrence et al., 2000; Torrence and Stevenson, 2000), but this type of evidence is rarely used to define cultural boundaries. Another measure of impact is the length of time that a region is abandoned after an extreme natural event. Abandonment is an important measure because it signifies a total failure of the cultural group to adapt to the situation. It is therefore a good reflection of the seriousness of the impact. In many cases the reason why there is a cultural phase boundary after a major environmental event is that people left the area and, after some period of time, a much altered or completely different group brought a new material culture to the area. When severity of the volcanic eruptions in West New Britain is compared with human impact defined as length of abandonment (Table 16.3), there are some surprising results. The data show that the length of abandonment decreased through time irrespective of the size of the volcanic explosion that caused it. For example, the most severe event, the W-K2 eruption, is not associated with the Table 16.4

Temporal patterning in material culture in West New Britain

Date BP

Material culture

?>5,900–3,600

Stemmed obsidian and chert chipped stone artefacts

3,600–2,000

Lapita style pottery, unretouched obsidian flakes and flake cores, small groundstone axes (extremely rare)

2,000–present

Imported pottery (extremely rare) unretouched obsidian flakes and flake cores, large groundstone axes

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 299

longest period of abandonment. Second, although the Dakataua eruption must have been a catastrophic event, since these are the deepest tephra deposits in the northern part of the Willaumez Peninsula, radiocarbon dates from beneath and above the tephra are indistinguisable (Torrence et al, 2000: table 2). Viewed from an archaeological perspective, then, Dakataua did not cause a disaster. The data for W-K3 and W-K4 are harder to interpret since these tephras have been mixed together in a relatively thin stratum on Garua Island and at Yombon and the dates for these eruptions (1700 BP, 1400 BP), particularly W-K4, are based on very few radiocarbon determinations (Machida et al., 1996). Nevertheless, the eruptions seem to have had minimal effect over much of West New Britain. Sites appear to have been abandoned at Yombon several hundred years before W-K3 and were not occupied until shortly after W-K4 (Torrence et al., 2000). On Garua Island there is a change in land use beginning about 2000 BP such that artefacts were deposited in fewer places than previously and after 1600 BP only two hilltops have evidence of use. Although parts of the island may have been abandoned, the site of FSZ continued to be actively used until the Dakataua eruption. Furthermore, at FSZ the W-K3 and W-K4 mixed tephras found elsewhere have been removed by human activities. Seemingly, then, W-K3 and W-K4 were not disasters on Garua Island. It seems likely that the people who left Garua resettled on the nearby mainland since Bitokara is reoccupied about the same time as people left Garua and use of this place expands greatly after the time of the W-K4 event (Fig. 16.1). A larger set of chronometric dates is needed to be completely certain that the changes in settlement in the interior at Yombon and on the Willaumez Peninsula were not causally related to the W-K3 and W-K4 events. It seems more likely, however, that the population movements reflect the expected pattern for people using a low-intensity system of shifting cultivation (Torrence and Stevenson, 2000). A very different pattern of impacts has been recorded at Numundo and Garu plantations further to the south and nearer the source volcano (Fig. 16.1). The results of limited test pitting indicate that this region was abandoned after the WK3 eruption and, with the exception of very rare sparse artefact scatters, there is no evidence of reoccupation before the W-K4 eruption approximately 300 years later. In other words, the impact of the W-K3 eruption was as almost as severe in these areas as was the considerably larger W-K2 event. What is also interesting is that there is no evidence in the form of a noticeable increase in number of places used or density of artefacts that the people fleeing this disaster at Numundo and Garu sought refuge in the interior or further up the peninsula, where the impact appears to have been minimal. In contrast, after the W-K4 eruption there was a significant increase in the number of people and the intensity of land use at Numundo and Garu. Unfortunately, we do not have dates to assess exactly when the reoccupation after the W-K4 event occurred. This is important because the considerable intensity of land use recorded at Numundo and Garu may indicate the presence of refugees from the Dakataua eruption, which would have caused serious problems for these populations (Torrence et al., 1999).

300

R. TORRENCE

INTERPRETING IMPACTS The brief summary of the Witori and Dakataua events demonstrates that there is no simple, direct relationship between the severity of the volcanic eruptions and the cultural responses in West New Britain. If we accept Sheet et al.’s (1991) findings, then the explanation for variation in responses should be accounted for by differences in social organisation. The data discussed below illustrate that the situation is much more complex. In my view one needs to take a more comprehensive look and consider a much broader range of factors that would have affected the way human groups were able to respond to the volcanic eruption. To begin with, environmental damage needs to be assessed. In the following discussion I consider gross changes in topography and physiography as well as damage to vegetation and processes of forest regeneration. Second, in order to assess cultural vulnerability, variation in economic and social organisation is considered. Landscape changes Volcanic eruptions can cause a total alteration of the landscape, especially near the source. Away from the centre of activity the major changes in topography are caused by airfall tephra. The emplacement of tephra up to 50 cm thick over a landscape will significantly alter the overall topography (i.e. by smoothing over rough surfaces, decreasing slopes, infilling valleys, changing of drainage patterns etc.), which has implications for human settlement, especially when combined with the erosional processes that follow (Boyd and Torrence, 1996). In addition to impacts on raw material extraction (Torrence et al., 1996) and subsistence patterns, these changes can markedly affect the social and sacred landscapes so that, for example, ‘special’ places, trails and territorial markers are no longer recognisable and/or lose their meaning. The main changes in behaviour due to landscape alteration that we have traced relate to the creation of beaches and in places an extensive coastal plain (cf. Gosden and Webb, 1994). On Garua Island the emplacement of the W-K2 tephra created a new strip of land surrounding parts of the island and subsequent settlement focused on the new coastal plain before moving into the uplands (Torrence and Boyd, 1997; Torrence and Stevenson, 2000). At Numundo Plantation changes in coastal topography had more profound effects on settlement. From our very preliminary survey work, we can track the evolution of the current coastal plain during the past 6,000 years from shallow water to dry land as a result of infilling from the Witori W-K1–4 airfall tephras and deposition of additional tephras eroded from the surrounding hillsides. As the plain filled in, the number of ecozones (e.g. shallow water, reef, mangroves, swamp, lakes, etc.) expanded after W-K2 and then declined after W-K3. As a consequence, the intensity of human use shifted from dispersed over the whole area (pre-W-K2), to concentrated near the coast (post-W-K2), and then to a mixed use of foothills for settlement and coastal plain for agriculture (post-WK4). It seems likely that the long abandonment period after the W-K3 eruption in

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 301

this region may have resulted from the destruction of many of the productive shallow-water swamps and mangroves. In contrast, one factor in the marked increase in intensity of land use after W-K4 might have been the expansion of dry, low-lying areas suitable for agriculture as a consequence of the emplacement of the tephra (cf. Torrence et al., 1999). Regeneration The speed and nature of forest regeneration should be critical in determining the potential for human recolonisation of an area after a volcanic eruption. The length of time for complete regeneration depends on the extent of destruction, which in turn is caused by the chemical composition, physical properties (e.g. grain size, porosity) and depth of the tephra. Airfall tephra with depths up to 50 cm, as in the case of the W-K1, W-K2 and Dakataua events, would have totally annihilated the ground cover and stripped most, if not all, of the canopy of its leaves and branches (Blong, 1984: 316–25). A recent case study of the 1994 Rabaul eruption found that in depths less than 20 cm, trees were defoliated but many plants were able to regenerate from suckers and roots, whereas tephra greater than 50 cm caused severe problems. Replanting was also difficult in areas of thick tephra deposits because of ‘calcium, phosphorous and potassium deficiencies, high salt concentrations in the tephra, low nutrient holding capacity, and absence of a balanced soil ecology’ (Lentfer et al., in press). In places where the tephra fall was 1–15 cm there was good recovery of gardens once the rains had come. In some cases the regrowth was extremely lush because of the added nutrients derived from the decayed plant matter buried under the ash (ibid.). On the whole, one can predict that the length of time for recovery would be directly correlated with the severity index I have used to compare the West New Britain eruptions, since it was calculated in terms of both the depth and areal extent of the tephra. Spatial extent of the destruction is critical because recolonisation of vegetation depends on sea, water and animal transport of plant propagules. Given the history of recolonisation on Krakatau – such that a rainforest currently stands on islands where all living matter had been completely obliterated in 1883 (Thornton, 1996), one would expect that 100 years would have been adequate for full forest cover even after the most severe eruptions, although the inland regions might have required a bit longer because of the increased distance from seed banks. Although tropical forests can regenerate quite quickly, the types and variety of species present may vary widely because succession has a very significant random element to it. For example, Krakatau may be totally reforested today, but the number of species is quite low compared to before the eruption and, due to the vagaries of history, a very different suite of both animal and plant species has recolonised the various island remnants of the original caldera (Thornton, 1996). In contrast, in their study of revegetation on Vulcan volcano at Rabaul, Lentfer et al. (in press) found that the composition of the early plant colonisers was very similar to that recorded for Krakatau, suggesting some measure of determinism in the early stages of the regeneration process. Research on how vegetation recovery

302

R. TORRENCE

took place in West New Britain has recently been initiated using the analysis of phytoliths (Boyd et al., 1998), starch grains (Therin et al., 1999) and pollen (Torrence et al., 1999). Species composition would have been extremely important to huntergatherers or forest managers (i.e. low-intensity, shifting agriculturalists). If, as a consequence of random events, the post-eruption forest contained a larger number of fruit or nut trees or tubers, then it might have lent itself to intensive culling and/or management and permitted higher population densities than previously. Obviously the reverse is also true. Furthermore, food plants tend to be more common in the early stages of forest succession, and some sources of food such as fruits and nuts are much easier to harvest before the full canopy develops. Returning sooner, rather than later, in the sequence of forest succession could therefore have had advantages in terms of harvesting or managing of wild plant resources. Part of the observed temporal trend in the length of abandonment may therefore have been caused by differences in the way reforestation occurred, but it is difficult to argue that the extremely long periods of abandonment after the first two eruptions, 1,000 and 250 years respectively, can be satisfactorily explained by environmental factors alone. Furthermore, the human response to these environmental factors would not have been direct, but would have been mediated by cultural factors. For example, at a very simple level it is quite clear that societies practising agriculture could have recolonised the region at different speeds and in different ways than would be the pattern with hunter-gatherers. Subsistence and settlement The resilience of the various societies in the face of the volcanic eruptions and their ability to recolonise at different speeds would have depended at least in part on the patterns of subsistence and settlement which they practised. Archaeological research in the province has revealed that despite the punctuated environmental record, the nature of subsistence and settlement changed in a directional manner. A gradual shift through time from a focus on the production of multi-purpose, retouched, chipped stone artefacts (‘stemmed tools’) to the production and use of expediently manufactured, unretouched flakes has been proposed as the result of a decrease in mobility (Torrence et al., 2000; Torrence, 1992, 1994; Fullagar, 1992; Kealhofer et al., 1999). Data on changes in the distribution of artefacts across space on Garua Island (Torrence et al., 2000) and Yombon (Pavlides, 1999) from dispersed to clustered appear to indicate there was a gradual reduction in mobility through time, paralleling the shifts in technological organisation of stone tools. Studies of starch grain assemblages recovered from sites on Garua Island and at Bitokara have also been interpreted to indicate a higher degree of sedentism through time (Therin et al., 1999). I have proposed that these changes in lithics and settlement patterns are part of a more general process involving a gradual intensification in the management of plant resources through time (Torrence, 1992, 1994; Torrence et al., 2000). If, for whatever reasons, people were increasingly managing their resources so

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 303

as to get higher returns per unit area, it seems likely that they would have been able to colonise a new region more quickly. Along similar lines, if they were gardening tubers and could clear the ash from their plants or had an alternative source of new tubers, then they could return permanently within a year after an eruption – if the ash was less than 50 cm – and rather more slowly in worseaffected areas which require the addition of nutrients and the restoration of particular soil properties (cf. Lentfer et al., in press). It therefore seems likely that the linear change in subsistence pattern, which is inferred from the lithics and settlement data, took place independently of the history of eruptions but was critical in accounting for the temporal trend toward increasingly rapid recolonisation. Social organisation Although Sheets et al. (1991) were correct to recognise the importance of social organisation as a factor in how disasters are caused and managed, in the case of West New Britain, their prediction that complex social organisations are less sensitive to environmental perturbation than simpler societies fails to explain the observed pattern. Throughout the time period considered the social groups appear to have been relatively egalitarian societies. There is little evidence for major changes in the complexity of the social organisations in this region during prehistory, although the nature and role of prestige goods were not stable. Social differences are nevertheless implicated in the long-term trend toward smaller periods of abandonment after the volcanic eruptions because the nature and speed of recolonisation cannot be explained solely by the severity of the event in terms of damage to the environment. The most important social implication for understanding variation in the nature of recolonisation is that it cannot be understood as a purely local process. After an extreme event people need social ties that allow them to take refuge elsewhere, even if only temporarily. If not, they will perish. Part of the variation in recolonisation witnessed in the West New Britain data must be due to the presence or absence of and/or differences in the social linkages between the victims and their neighbours or with more distant people with whom they could shelter. Where they could have sheltered and for how long would have depended on the nature of their social ties with groups outside the affected area, whether the immigrants could have been incorporated within the social practices and land tenure arrangements of their allies, and the potential of the resources in these places to support increased population. Whereas finding temporary refuges might have been relatively easy, maintaining a satisfactory existence away from the homeland on a very long-term basis might not have been possible. In all the cases that we have monitored in West New Britain, the thickness of the tephra was such that sources of food would have been destroyed and fresh water would also have been badly affected. If the event occurred in the dry season, as was the case in the Rabaul 1994 eruption, people might have had to wait as long as six months for rains leading to natural regeneration or the revitalisation of gardens (cf. Lentfer et al., in press). For the W-K1, W-K2 and Dakataua events

304

R. TORRENCE

throughout the region and for W-K3 and W-K4 at Numundo and Garu, the environmental damage would have been so extensive that people could not have recolonised for some years even if they were able to replant their gardens. Survival in these cases would have been completely contingent on having social relations elsewhere. Since the spatial scale of the eruptions differed, however, the scale and nature of the social linkages required to enable safe refuges would have varied. So, for example, groups surviving the W-K1 or W-K2 disaster would have needed friends and relations at much greater distances than those experiencing the impacts of the less spatially extensive Dakataua, W-K3 or W-K4 eruptions. It seems very likely that the much longer times for reoccupation after W-K1 and W-K2 indicate that local groups were not able to seek refuge and did not survive. In contrast, the rapid resettlement after Dakataua implies that the original residents may have been responsible for the recolonisation. It is difficult to obtain accurate comparisons toward the length of social ties at various times in the prehistory of West New Britain given current data. Obsidian from the Bismarck Archipelago outcrops has been recovered from sites located in many areas of Melanesia and dating to all the periods we have studied (e.g. White, 1996). After the W-K2 eruption, obsidian is found over a larger area than in previous periods. This pattern suggests the presence of wider cultural links. At this same time people were decorating their pottery with motifs that are also shared over a very large region extending as far as Tonga and Samoa (e.g. Summerhayes, 2000). The wider regional social ties could help explain the shorter length of depopulation when compared to the W-K1 eruption. Kirch (1988) has suggested that the fairly rapid speed of the colonisation throughout Near Oceania of people bearing the Lapita decorated pottery was facilitated by social links between communities such as are reflected in the distribution of obsidian from West New Britain. From about 2,000 years ago onward, obsidian movement is more restricted in space, but there is an expansion into newer areas, such as the Papua New Guinea Highlands region (White, 1996: 201). Clearly, there was a reorientation in the location of social links, but currently we are unable to assess precisely how these data relate to differences in the length and strength of social networks. Although we can infer that variations in the social make-up of the societies involved must have greatly affected rates of recolonisation in West New Britain, it is difficult to accurately monitor these differences with current archaeological data, although this should be pursued in future work. We also need to invoke social differences in explaining the long-term trend for a reduction in the length of abandonment because the ability to bud off from a larger social group or to organise a mass movement of people is an explicitly social process. Relatively rapid recolonisation of an area which has been totally decimated would also have required means for buffering failures, such as ways to secure links with people remaining behind and/or efficient means for redistributing resources. Small groups with fewer social ties are likely to have been less successful at supporting themselves at an early stage of natural regeneration, but could have recolonised later in the succession when food supplies were more abundant. In summary, understanding the impacts of the various eruptions requires a

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 305

great deal of geological, environmental and archaeological data. At this stage much of this is not yet available in adequate detail to assess fully the relative importance of the environmental changes brought by the volcanic events in West New Britain versus the vulnerability of the societies, as monitored by their economic and social organisation. Nevertheless, the case study demonstrates the importance of considering a wide number of factors rather than concentrating solely on the vulnerability of the societies themselves. Furthermore, given the small amount of variation in the social organisations present in the study region, it is obvious that the severity of the event was a critical variable. Finally, a number of methodological problems have been faced in trying to sort out the various impacts, and these deserve more serious discussion.

METHODOLOGICAL ISSUES The most serious methodological issues identified by the West New Britain case study are limitations in the way impacts are assessed by archeologists and problems with selecting appropriate spatial and temporal scales for monitoring the effects of the eruptions. For the West New Britain case study I have used length of abandonment as a rough measure of the impact of volcanic events on the contemporary human societies. As a first stage, this is a reasonable way to compare disasters occurring in the same area, but detailed comparison between regions in other parts of the world would be more difficult because potential rates of recovery are necessarily linked to local environmental conditions. This measure is also limited because the character and speed of landscape recovery is unlikely to be identical after different events due to landscape changes caused by lavas and airfall tephras and the stochastic nature of forest regeneration. An alternative measure of impact is the degree to which the trajectory of cultural change parallels that of disasters. So, for example, Sheets et al. (1991) consider the chronological relationship between cultural phase boundaries and volcanic eruptions. The prehistory of West New Britain illustrates a difficulty with using this approach on its own. In the case of the W-K1 event there was no appreciable change in the artefact assemblages before and after the eruption, but a very large area appears to have been abandoned for 1,000 years afterwards. Obviously, the local population was seriously affected by this event. On the other hand, the wider region, which was the source of the populations who recolonised the affected area, probably did not face a disaster and maintained the same technological system outside the affected region. The opposite took place after the W-K2 eruption. Major cultural changes took place outside West New Britain during the time the area was abandoned and so the colonisers brought with them a very different set of material culture items. Whether the impact of people fleeing the region and taking refuge elsewhere had a role to play in the development of these new cultural items or the process was completely external to the volcanic eruptions cannot be determined solely from a study of West New Britain itself, but is a research topic worth pursuing.

306

R. TORRENCE

Although each provides a gross index of how volcanic events have affected populations, neither the cultural phase nor length of abandonment approach to monitoring cultural responses is completely satisfactory, mainly because each focuses on very large-scale changes and glosses over the complexity in the data. New approaches to assessing impacts are needed that examine a wider range of effects. Possibly the most important methodological issue to be faced when assessing the nature of disasters is choosing the spatial scale of analysis. First, one needs a way to determine the size of the area directly affected by the volcanic eruption. For this one could use a simple measure such as depth of tephra. The second difficulty is locating the areas which suffered indirect effects. Surrounding populations could have been affected because they provided food and shelter for refugees and may have made adjustments to enable settlement of refugees on a long-term basis. Even further afield, other areas may have suffered because trading links with an affected area were disrupted. In addition, some places may even have benefited through loss of competition or disruption of trading networks (cf. Sheets, 1979). A second methodological issue is the selection of the spatial scale for analysis. The West New Britain study shows that spatial scales are particularly important when trying to understand the process of recolonisation. In the case of the W-K1 and W-K2 eruptions, when the region was abandoned for many generations, the recolonising population must have comprised a different group of people from the former inhabitants and one that may have had tenuous connections to this area. The case for population replacement is less clear for recolonisation after the W-K3 and W-K4 events, but the timing and nature of reoccupation must nevertheless have been determined by the social organisation, subsistence and settlement patterns of the populations in the surrounding regions. If these had reached the culturally determined carrying capacity, then the refugee population may have been encouraged to return more quickly than if there had been opportunities for them to be incorporated and so on. A number of scenarios can be imagined, but all involve consideration of the region (of unknown size and different for each event) outside the affected area. Since the prehistory of West New Britain can be assumed to have comprised a set of recolonisations by different populations, most likely coming from different sources, the pattern provided by the archaeological data for a gradual decrease in the length of abandonment and an increase in the intensification of land use is very puzzling. One would not have expected a punctuated history of disasters to be correlated with stability or a slow, gradual cultural change. In particular, the continuity in lithic technology before and after the W-K1 disaster is remarkable. How is it that people returned 1,000 years later to the same lithic sources, made distinctive tools that were the same as before, and generally took up almost where they left off? In contrast, abrupt changes, such as the arrival of pottery after the W-K2 event, are quite predictable since the populations may have been quite different. To understand these puzzles, one would have to look outside West New Britain to the regions where the colonisers originated.

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 307

Even without the benefit of having studied a large enough region, some general predictions to guide future research can be derived from considering the process of recolonisation in West New Britain on an appropriately large spatial scale. First, I expect that in the period from c.6,000 to around 5,000 BP, when the affected area appears to have remained abandoned, other regions did not experience population stress. Either local solutions were found to cope with rising populations, e.g. subsistence intensification took place, or population size was relatively stable and there was no stimulus for moving out and colonising new territory. Furthermore, social and ideological factors may not have encouraged movement into the empty region. The timing of repopulation c.3,300 BP, several hundred years after the W-K2 event, must also be explained by processes operating outside of West New Britain. Finally, recolonisation after the Dakataua event appears to have been so rapid that, apart from the fact that this was a more localised catastrophe, the process must have differed markedly from previous episodes of recolonisation. Although in theory one could describe the attributes of the relevant region that should be studied, in practice it is difficult to define adequately the nature and level of the indirect effects to be studied and to predict exactly where the spatial boundaries of the affected area would have been. Furthermore, as in the case of the West New Britain eruptions, where the region that was indirectly affected would probably lie outside the island itself, the size of the appropriate study area would be beyond the capabilities of a single research project. In addition to selecting the best spatial scale of analysis, the choice of which chronological scales to study is very important and is also not straightforward. One of the advantages of archaeological research is that it enables one to track the follow-on effects of a extreme natural hazard over the period anywhere from a few decades to a millennium, depending on the chronological resolution that is possible (cf. Sheets, 1979: 558; Grayson and Sheets, 1979). Immediately after a disaster people are usually forced to abandon a region, creating a very major shortterm impact on their lives. The effect, however, may not be a long-lasting one and the population may return, clean up the mess, and carry on. Archaeology, which operates over scales rarely less than several generations, may not be able to monitor adequately the very short-term effects and may be tempted to conclude that the event had no impact. On the other hand, a group may return quickly but be forced to adopt new patterns of behaviour in order to adapt to its altered environment. Alternatively, groups may have to leave their homes for a very long time period, in which case the people who recolonise the region may bring many new cultural practices with them. In both these cases the one-off natural event had very major long-term effects, but these occurred at different time scales. From a methodological point of view, it is therefore quite important to analyse the impact of these short-term events over a number of chronological scales. This point is well demonstrated by the wide range of temporal impacts witnessed in prehistoric West New Britain. With the exception of the W-K3 and W-K4 eruptions as experienced on Garua Island, all the volcanic events in West New Britain that we have studied

308

R. TORRENCE

must have had catastrophic immediate impacts, but there was considerable variation in the way societies were affected when considered at other time scales. In terms of as little as several generations the Dakataua eruption seems to have made very little difference to population levels in the Willaumez Peninsula. In contrast, over the very long term of several centuries, the W-K2 event made quite a significant impact on the prehistory of the region. Finally, when viewed over several centuries, the W-K1 event caused a complete depopulation of a large region, but considered a millennium later, it seems to have had very little effect on the cultural practices in the affected area. Although archaeological data are generally amenable to the study of long-term processes, one of the most serious problems is getting good enough chronological control to adequately assess the nature and scale of impacts. I have argued previously that studies of volcanic disasters are much aided by the presence of airfall tephras that can be linked to single events. Although sourced tephras are invaluable for establishing a relative chronology over sometimes large regions, their presence does not solve the need for accurate, short-term dates. For example, material underlying a dated tephra may have no relevant chronological association with it because it might have been abandoned some considerable time before the eruption. If one wants to relate the buried material directly to the eruption, then independent chronometric control is required. Similarly, monitoring the nature and rates of recolonisation requires excellent chronological control. The history of recolonisation in West New Britain presented in Table 16.3 is only based on a limited sample of radiocarbon dates (Torrence et al., 2000) and so the patterns which we have detected could be inaccurate. Another difficulty archaeologists face is determining whether a population remained or left the area for relatively short periods that may nevertheless have been significant for cultural processes, e.g. several generations, but are too small to be detected given the limitations of radiocarbon dating. Archaeologists need to think hard about what types of questions concerning the impacts of natural hazards can be answered satisfactorily given the limitations of their data. As a first step, we should continue to develop better means for measuring impacts than are commonly used at present. Second, examining impacts on a range of spatial and temporal scales is very important for understanding the wide range of processes that result when an area experiences a severe climatic event. Finally, we must accept that there are major difficulties in assessing the degree of short-term impacts on human societies and in tracing indirect effects over very large regions, particularly if they are as poorly researched as is the case in Papua New Guinea.

IMPORTANCE OF DISASTER STUDIES Since the analysis of disasters can make a significant contribution to theories about human social and cultural change, it is important to overcome the methodological difficulties raised above. For example, the long-term history of recolonisation in

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 309

West New Britain, which is characterised by a series of short-term events that had a variety of impacts on cultural process, has far-reaching implications for current theories of culture change. First, the lack of correlation between the severity of volcanic events and the impact on societies as measured by the length of abandonment illustrates the importance of historical contingencies and stochastic elements in long-term cultural processes. Second, environmental factors, such as ecological succession and forest regeneration, are unlikely to fully explain the patterns observed – especially the exceedingly long periods of abandonment in the early periods and the gradual trend toward increasingly intensified systems of land management. Third, the pattern of subsistence intensification within West New Britain must be explained largely in terms of complex interactions between social processes operating outside the region and the environment of the affected region. The decrease in the length of abandonment over time might be assumed to represent some form of long-term ‘adaptation’ to volcanic disasters, but this explanation is untenable since the changes in culture reflected in the archaeological record took place outside the region directly affected by the volcanic events and were then reintroduced into the affected area. If there was any form of adaptation to the volcanic history of West New Britain, then it took place within the wider region which experienced indirect effects of the eruptions. Researching this hypothesis demands further research aimed at identifying the homelands of the people who recolonised the region and a comparison of the cultural history of these regions with that of West New Britain. Disaster research also has a significant role to play in archaeology because it focuses on the fundamental interactions between externally generated environmental factors and the way these are perceived and manipulated by human groups. I have argued that although the study of disasters provides an excellent opportunity for archaeologists to seriously examine important questions concerning the cause, speed and direction of culture change, there are a number of methodological issues concerning the measurement of impact and spatial and chronological scales that require further study. Considering the question raised in the title of this chapter – what makes a disaster? – the data from West New Britain illustrate that an answer can be framed on several levels. First, the severity of the natural hazard is very important. Smaller eruptions had less impact on societies than did the larger ones. Second, the impacts are felt at differing spatial scales depending on the social ties of the group affected and the nature of social process in the region where people took refuge. Third, disasters need to be defined at different time scales. A disaster may have occurred over a short time period when a region was abandoned immediately following the event, but the picture over the long term may have been more complex since various areas were abandoned for different lengths of time. If the group returned after a short time interval, then it might not be considered to be as large a disaster as if the region were depopulated for many generations. Evaluating when and how serious a disaster was is not a simple procedure. For example, in the West New Britain case major changes in material culture took place with recolonisation after the W-K2 event but not after the other eruptions. In contrast,

310

R. TORRENCE

there do not appear to have been any major changes once people returned after the W-K1 disaster. Furthermore, the slow, gradual intensification in land use which took place over about 6,000 years may have had little to do with the catastrophic events, since subsistence and settlement patterns were established by social systems that were introduced into the region through colonisation, often long after the volcanic eruptions. As yet we do not have an adequate framework for distinguishing impacts that occur at different temporal scales to allow us to make adequate comparisons of different natural hazards. Finally, the value of focusing on disasters as a way of developing new theories of cultural change is highlighted by the study. Without the emphasis on searching for the impacts of volcanic eruptions, the punctuated history of abandonment and reoccupation is likely to have been overlooked in favour of the gradual changes in lithic technology and settlement which would probably have been assumed to have been internally generated. We now know, however, that changes in material culture, subsistence and economic systems were at least in part a consequence of processes occurring elsewhere. These new insights are very important because they have led to a number of new hypotheses that will guide further research. ACKNOWLEDGEMENTS This research has been supported by an Australian Research Council Senior Fellowship and grants from the Australian Research Council, Australian Museum, Australia and Pacific Foundation, Australian Institute for Nuclear Science and Engineering, New Britain Palm Oil Ltd and the Earthwatch Foundation. Support from the following also aided my fieldwork: National Research Institute, National Museum of Papua New Guinea, University of Papua New Guinea, West New Britain Provincial Government,West New Britain Provincial Cultural Centre, Walindi Plantation, Garua Plantation, Kimbe Bay Shipping Company, New Britain Palm Oil Ltd. I am especially grateful to my field crews and local volunteers for all their hard work and also the following colleagues, whose collaborative work and discussions have played an important part in the research reported here: Hiroshi Machida, Jim Specht, Richard Fullagar, Christina Pavlides, Bill Boyd, Carol Lentfer, Glenn Summerhayes, Lisa Kealhofer and Peter White. REFERENCES Blong, R.J. (1984) Volcanic Hazards: A Sourcebook on the Effects of Eruptions. Sydney: Academic Press. Boyd, W. and Torrence, R. (1996) Periodic erosion and human land-use on Garua Island, PNG: a progress report. In S. Ulm, I. Lilley and A. Ross (eds) Australian Archaeology ’95, Tempus 6: 265–74. Boyd, W., Lentfer, C. and Luker, C. (in press) Environmental impacts of major catastrophic Holocene volcanic eruptions in New Britain, P.N.G.: a preliminary model for palaeoenvironmental change. Geodiversity:Proceedings of the IAG Conference, 1998. Boyd, W., Lentfer, C. and Torrence, R. (1998) Phytolith analysis for a wet tropics

VOLCANIC ERUPTIONS AND HUMAN RESPONSES, PAPUA NEW GUINEA 311

environment: methodological issues and implications for the archaeology of Garua Island, West New Britain, Papua New Guinea. Palynology 22: 213–28. Chester, D. (1993) Volcanoes and Society. London, Arnold. Dugmore, A. (1989) Icelandic volcanic ash in Scotland. Scottish Geographical Magazine 105: 168–72. Fullagar, R. (1992) Lithically Lapita. Functional analysis of flaked stone assemblages from West New Britain Province, Papua New Guinea. In J.-C. Galipaud (ed.) Poterie Lapita et Peuplement, 135–43. Noumea: ORSTOM. Gosden, C. and Webb, J. (1994) The creation of a Papua New Guinean landscape: archaeological and geomorphological evidence. Journal of Field Archaeology 21: 29–51. Gosden, C., Webb, J., Marshall, B. and Summerhayes, G. (1994) Lolmo Cave: a mid-to late Holocene site, the Arawe Islands, West New Britain province, Papua New Guinea. Asian Perspectives 33: 97–119. Grayson, D. and Sheets, P. (1979) Volcanic disasters and the archaeological record. In P. Sheets and D. Grayson (eds) Volcanic Activity and Human Ecology, 623–32. New York: Academic Press. Kealhofer, L. Torrence, R. and Fullagar, R. (1999) Integrating phytoliths within usewear studies of stone tools. Journal of Archaeological Science 26: 527–46. Kirch, P. (1988) Long-distance exchange and island colonization: the Lapita case. Norwegian Archaeological Review 21: 103–17. Lentfer, C., Boyd, B. and Sasambie, S. (in press) Effects of Volcanic Eruptions on People and the Environment: The 1994 Rabaul Eruptions, a Case Study. Lismore: Southern Cross University Press. Machida, H., Blong, R., Moriwaki, H., Hayakawa, Y., Talai, B., Lolock, D., Specht, J., Torrence, R. and Pain, C. (1996) Holocene explosive eruptions of Witori and Dakataua volcanoes in West New Britain, Papua New Guinea and their possible impact on human environment. Quaternary International 35/36: 65–78. Melson, W. (1994) The eruption of 1968 and tephra stratigraphy of Arenal volcano. In P. Sheets and B. McKee (eds) Archaeology, Volcanism, and Remote Sensing in the Arenal Region, Costa Rica, 24–45. Austin: University of Texas Press. Oliver-Smith, A. (1996) Anthropological research on hazards and disasters. Annual Review of Anthropology 25: 303–28. Pavlides, C. (1993) New archaeological research at Yombon, West New Britain, Papua New Guinea. Archaeology in Oceania 28: 55–9. Pavlides, C. (1999) The Story of Imlo: the organization of flaked stone technologies from the lowland tropical rainforests of West New Britain, Papua New Guinea. Unpublished Ph.D. Dissertation, La Trobe University. Pavlides, C. and Gosden, C. (1994) 35,000-year-old sites in the rainforests of West New Britain, Papua New Guinea. Antiquity 68: 604–10. Pilcher, J. and Hall, V. (1996) Tephrochronological studies in northern England. The Holocene 6: 100–5. Sheets, P. (1979) Environmental and cultural effects of the Ilopango eruption in Central America. In P. Sheets and D. Grayson (eds) Volcanic Activity and Human Ecology, 525– 64. New York: Academic Press. Sheets, P. (1994) Summary and conclusions. In P. Sheets and B. McKee (eds) Archaeology, Volcanism, and Remote Sensing in the Arenal Region, Costa Rica, 312–25. Austin: University of Texas Press. Sheets, P. and Grayson, D. (eds) (1979) Volcanic Activity and Human Ecology. New York, Academic Press. Sheets, P. and McKee, B. (eds) (1994) Archaeology, Volcanism, and Remote Sensing in the Arenal Region, Costa Rica. Austin: University of Texas Press. Sheets, P., Hoopes, J., Melson, W., McKee, B., Sever, T., Mueller, M. Cheanult, M. and Bradley, J. (1991) Prehistory and volcanism in the Arenal area, Costa Rica. Journal of Field Archaeology 18: 445–65.

312

R. TORRENCE

Specht, J. and Gosden, C. (1997) Dating Lapita pottery in the Bismarck Archipelago, Papua New Guinea. Asian Perspectives 36: 175–99. Specht, J., Fullagar, R. and Torrence, R. (1991) What was the importance of Lapita pottery at Talasea? Bulletin of the Indo-Pacific Prehistory Association 11: 281–94. Summerhayes, G. (2000) Lapita Interaction. Terra Australis 15. Canberra: ANH Publications and the Centre for Archaeological Research, The Australian National University. Therin, M., Fullagar, R. and Torrence, R. (1999) Starch in sediments: a new approach to the study of subsistence and land use in Papua New Guinea. In C. Gosden and J. Hather (eds) Food in Prehistory, 438–62. London: Routledge. Thornton, I. (1996) Krakatau: The Destruction and Reassembly of an Island Ecosystem. Cambridge: Harvard University Press. Torrence, R. (1992) What is Lapita about obsidian? A view from the Talasea sources. In J.-C. Galipaud (ed.) Poterie Lapita et Peuplement, 111–26. Noumea: ORSTOM. Torrence, R. (1993) Archaeological research on Garua Island, West New Britain Province, Papua New Guinea, June–July 1993. Report submitted to official organisations within Papua New Guinea. Torrence, R. (1994) Strategies for moving on in lithic studies. In P. Carr (ed.) The Organization of Technology, 123–31. Ann Arbor: University of Michigan Press. Torrence, R. and Boyd, B. (1996) Archaeological fieldwork on Garua Island, West New Britain, Papua New Guinea, June–August 1996. Report submitted to official organisations within Papua New Guinea. Torrence, R. and Boyd, B. (1997) Archaeological fieldwork in West New Britain, Papua New Guinea, June–August 1997. Report submitted to official organisations within Papua New Guinea. Torrence, R. and Stevenson, C. (2000) Beyond the beach: changing Lapita landscapes on Garua Island. In T. Murray and A. Anderson (eds) Papers for Jim Allen. Canberra: Archaeology and Natural History, Research School of Pacific and Asian Studies, Australian National University. Torrence, R. and Webb, J. (1992) Report on archaeological research on Garua Island, West New Britain Province, Papua New Guinea, July–August 1992. Report submitted to official organisations within Papua New Guinea. Torrence, R., Pavlides, C., Jackson, P. and Webb, J. (2000) Volcanic disasters and cultural discontinuities in the Holocene of West New Britain, Papua New Guinea. In B. McGuire, D. Griffiths and I. Stewart (eds) The Archaeology of Geological Catastrophes, 225–44. London: Geological Society Special Publication 171. Torrence, R., Specht, J. and Boyd, B. (1999) Archaeological fieldwork on Numundo and Garu Plantations, West New Britain, PNG. Report submitted to official organisations within Papua New Guinea. Torrence, R., Specht, J. and Fullagar, R. (1990) Pompeiis in the Pacific. Australian Natural History 23: 3–16. Torrence, R., Specht, J., Fullagar, R. and Summerhayes, G. (1996) Which obsidian is worth it? A view from the West New Britain sources. In G. Irwin, J. Davidson, A. Pawley and D. Brown (eds) Oceanic cultural history: essays in honour of Roger Green, 211–24. Wellington: New Zealand Journal of Archaeology Special Publication. White, J.P. (1996) Rocks in the head: thinking about the distribution of obsidian in Near Oceania. In G. Irwin, J. Davidson, A. Pawley and D. Brown (eds) Oceanic cultural history: essays in honour of Roger Green, 199–209. Wellington, New Zealand Journal of Archaeology Special Publication. Zeidler, J. and Isaacson, J. (1999) Volcanic disasters and historical contingency: the prehistoric record of differential response to volcanic eruptions in Western Ecuador. Precirculated paper for the World Archaeological Congress 4, Capetown.

17

The impact of the Kikai-Akahoya explosive eruptions on human societies HIROSHI MACHIDA AND SHINJI SUGIYAMA

INTRODUCTION This chapter addresses in detail the precise mechanisms by which a large volcanic eruption in Japan generated significant cultural change. The principal mechanism concerned appears to be that of landscape change, which made large-scale adaptation and readjustment necessary. The chapter illustrates the wide range of research tools which may be adopted to determine the relationship between an archaeological site and the hazards present in the natural environment. Japan is a volcanically active area where explosive activities have distributed tephra deposits over extensive areas on numerous occasions. Many of the tephra layers play important roles as time-markers in establishing a chronology for the Quaternary. Artefacts and archaeological remains have been excavated from soils sandwiched between distinctive tephra layers. These marker layers are useful for dating and correlating deposits found in different areas, especially when the relative dating is confirmed by radiometric methods. In addition, the presence of the tephra layers suggests that explosive volcanism may have had a significant impact on physical environment, human societies and culture change in Japan. Among the many large-scale eruptions during the Japanese Holocene period, the Kikai-Akahoya eruption was the biggest. The eruptive products from it are called the Kikai-Akahoya tephra (K-Ah) (Machida and Arai, 1978). The KikaiAkahoya event has a Volcanic Explosive Index of 7 (Newhall and Self, 1982). In Japan and adjacent areas volcanic eruptions with VEIs of 4 and 5 are known for the historic period. However, with the one exception of the Changbaishan (Baegdusan) eruption, which took place on the northern part of Korean Peninsula during the latest period of the ninth century or the earliest of the tenth century (Machida et al., 1990), very large-scale activity with VEIs similar to the KikaiAkahoya event are unknown. In this chapter we summarise the processes that took place during the Kikai-Akahoya tephra-forming eruption and assess the impact of the emplacement of the tephra and associated environmental changes on human societies. General features of this volcanic hazard and the resulting human disasters have been reported previously (Machida, 1984) and so the major

314

H. MACHIDA AND S. SUGIYAMA

purpose of this chapter is to discuss new information concerning the impact of the Kikai-Akahoya eruption on landscape change and vegetation. The enormity of the environmental change indicated by the new studies is more than adequate to explain the cultural changes in the archaeological record that were previously identified.

KIKAI CALDERA AND TEPHRAS South Kyushu is one of the most active volcanic areas in Japan. In this region many large-scale ignimbrite sheets form extensive plateaux and fallout tephras occur around several large calderas. At least four gigantic calderas with approximate diameters of 20 km lie in the Kagoshima volcanic depression which extends 150 km from north to south. The Sakurajima, Kirishima, Kaimondake and SatsumaIwodake volcanoes, which are located here, are all active and have caused localised but severe volcanic hazards in historic times. The Kikai caldera occupies the southern part of this volcanic zone and is mainly submerged in the sea (Fig. 17.1). Submarine topography shows that this caldera is composed of double depressions with a maximum diameter of 20 km. Many submerged cones and a moat nearly 600 m deep are present within the caldera. The two islands, Takeshima and Satsuma-Iwo-jima, and some volcanic reefs form the land portions of the caldera rim, which is traced to undersea portions and separates into two in the southeast. There are ignimbrite plateaus on Takeshima and the western half of Satsuma-Iwo-jima (Fig. 17.1). The date for the beginning of the volcanic activity from this centre is not yet clear, but the oldest known major tephra, the Koseda ignimbrite, which is found in the islands of Tanegashima and Yakushima, has an isothermal plateau fission track age of c.580,000 BP (Westgate and Moriwaki, pers. comm.). The volcanic centre also ejected at least two large-scale tephra units in the late Pleistocene, namely the Ko-abi yama tephra which has dates of c.130,000 + 20,000 BP and 140,000 + 20,000 BP (K-Ar age, Machida and Fujiwara, pers. comm.) and the Kikai-Tozurahara tephra of c.95,000 BP (Machida, 1999). The main subject of this chapter, the Kikai-Akahoya (K-Ah) Holocene tephra, is the youngest major product of this caldera. It has several local names because of its widespread occurrence and various types of ejecta: i.e. Takeshima fall and flow deposits on the caldera rim (Ono et al., 1982) and Koya fall and flow deposits (K-Ky) in the southern part of Kyushu mainland (Ui, 1973). In south to central Kyushu local farmers and pedologists have named the distal fine-grained ash Akahoya after its reddish colour and low agricultural potential.

KIKAI-AKAHOYA TEPHRA Volcano-stratigraphical studies of the Kikai-Akahoya tephra show that paroxysmal activity of the Kikai caldera started with a plinian eruption, which produced the

THE IMPACT OF KIKAI-AKAHOYA ERUPTIONS

315

Figure 17.1 Geographical setting of the caldera volcanoes in South Kyushu

pumice-falls. This was followed by two ignimbrite-forming eruptions (Figs 17.2 and 17.3). The last and the most extensive pyroclastic flow is called Takeshima. It spread in a concentric pattern as a low-aspect ratio ignimbrite (Walker et al., 1980) as much as 100 km from the source (Fig. 17.4). The basal layer of the Takeshima pyroclastic flow deposits is characterised by coarse-grained, rounded boulders and gravels, which must have been deposited on the shallow sea bottom before the eruption and were later ejected into the air as a consequence of the eruption. The K-Ah ash is a fine-grained vitric fallout ash with much accretionary lapilli at its bottom layer. The distal ash, K-Ah, was formed as a coignimbrite ash of the Takeshima–Koya pyroclastic flows, which were generated by a collapse of a large and highly mobile eruption. A considerable part of the ash would have fallen as saturated aggregates. It is composed of bubble-walled glass shards and small

316

H. MACHIDA AND S. SUGIYAMA

Figure 17.2 Schematic diagram showing the sequence of the Kikai-Akahoya eruptions Source:

Machida (1987)

pumice lumps of rhyodacitic composition and small amounts of the opaque mafic minerals orthopyroxene and clinopyroxene. On the basis of the petrographic characters of the constituent materials of this ash, it is correlated and identified at many localities. The fallout area as currently known is shown in Fig. 17.4. It occurs in extensive areas from Kyushu to the southern part of northeast Honshu and has been identified in several piston cores from the southern part of the Sea of Japan, the northwest Pacific Ocean and the East China Sea. The bulk volume of the deposits associated with this ash is estimated to be more than 150 km3 (Machida and Arai, 1978). The K-Ah tephra forms a good marker for the hypsithermal stage because it is found in the marine deposits formed during the culmination stage of the Holocene transgression. More than 50 radiocarbon ages obtained from various localities in Japan indicate that the great Kikai eruption occurred at around 6,300 BP in the conventional radiocarbon timescale (Machida and Arai, 1978), which calibrates to c.7,300 cal. BP (Stuiver and Reimer, 1993). The calibrated date has been given considerable support from recent chronological studies of laminated lake sediments in the brackish lake of Suigetsu, north of Kyoto. A detailed counting of annual layers suggests that the eruption occurred at 7,280 BP (Fukusawa, 1995).

THE IMPACT OF KIKAI-AKAHOYA ERUPTIONS

317

Figure 17.3 Eruptive formations of the Kikai-Akahoya tephra outcropping at the harbour on Takeshima Island

VOLCANIC IMPACT ON THE NATURAL ENVIRONMENT In addition to identifying a correlation between significant cultural change and the Kikai eruption, new research has discovered environmental changes associated with the eruption. Along with the immediate impacts of the eruption, environmental changes that were caused by tectonic activity and the large falls of tephra must have had a strong impact on the landscape in general and human life in particular. Our recent studies have found evidence for (1) vegetation changes, (2) swarms of landslides, (3) flooding, and (4) a co-volcanic earthquake and tsunami. Devastation of forests Phytolith and pollen analyses for a soil section at Onejime in Kyushu suggest that before the great Kikai eruption, this region was covered by temperate evergreen forest and bamboo grass (sasa) bush very similar to that of the present day. A comparison of the floral assemblage reconstructed by the phytolith analysis

318

H. MACHIDA AND S. SUGIYAMA

Figure 17.4 Isopach map of the Kikai-Akahoya ash (thickness in cm) and the approximate distribution of the Koya pyroclastic flows Source:

revised from Machida and Arai (1978)

beneath and above the Kikai-Akahoya indicates a drastic vegetation change (Fig. 17.5). Just below the K-Ah tephra there are abundant phytoliths of Lucidophylus origin, indicating a broad-leafed evergreen forest. They are lacking, however, in the soil just above the K-Ah tephra and below the Ikedako tephra. Fig. 17.6a shows localities where abundant Lucidophylus phytoliths can be identified before the Kikai eruption. In contrast, Fig. 17.6b shows that, particularly in the area covered by the Ky pyroclastic flow deposits, Lucidophylus phytolith grains are drastically lacking, with the exception of one locality. A decline in Lucidophylus vegetation would have continued for around 800–900 years after the Kikai eruption in the area covered by the subsequent Ikedako tephra layers. Landslides As yet there are few archaeological data which can be used to assess the volcanic impact in the Shikoku and Kii areas, which are 300 to 600 km to the east of Kikai. However, some geological observations indicate that the effect of the K-Ah ash fallout was very serious. The K-Ah ash is a product of phreatoplinian eruptions

THE IMPACT OF KIKAI-AKAHOYA ERUPTIONS

319

Figure 17.5 Changes in phytolith assemblages at Onejime in the southernmost part of Kyushu Source:

Sugiyama (1999)

(Self and Sparks, 1978) caused by reactions between magma and water that result in the production of a large amount of fine-grained vitric ashes. Ash of this type is likely to be wet when it falls. It becomes sticky after deposition due to absorbed moisture and then hardens. In association with the torrential mud-rain, eruptioninduced storms might also have taken place. Not surprisingly, the occurrence of the K-Ah ash induced severe landslides or slope erosion in every mountainous area. There are patches of the K-Ah ash scattered within debris avalanche deposits in the Shikoku and Kii Mountains. It seems likely that the forest would have been destroyed by the combined actions of the wet ash and associated landslides. Flooding Many localities provide evidence for the impact of mudflows or river floods in the delta plains following the Kikai-Akahoya eruption. In South Kyushu and Shikoku in the Holocene sediments immediately above the marine transgression, flood deposits are mainly composed of a thick layer of reworked vitric ash which is identified as the K-Ah tephra. These flood contexts also contain a large quantity of logs, which suggests that the mountain slopes were deforested as a consequence of this volcanic event.

320

H. MACHIDA AND S. SUGIYAMA

(a)

(b)

Figure 17.6 Changes in Lucidophylous forest inferred from phytolith analysis (a) before the Kikai-Akahoya eruption and (b) after the Kikai-Akahoya eruption

The Miyazaki coastal plain, one of the largest in South Kyushu, is composed of several coastal bars and dunes interspersed with back dune swamps and marsh. Archaeological and tephrochronology data indicate that the largest bars and dunes were formed at the time of the Kikai eruption. These data clearly demonstrate that a large amount of sand of volcanic origin was supplied from the eruption-induced flood. Consequently, all the shorelines advanced significantly. Archaeological excavations from shell mound sites on Chita Peninsula, central Honshu (c.750 km from Kikai), have revealed that there is a conspicuous decline in the size and number of shells immediately after the Kikai eruption (Yamashita, 1980). It therefore seems likely that extensive devastation and decline of marine productivity was caused by the deposition of ash, associated flooding and coastal change. Co-volcanic earthquake and tsunami Another very interesting impact on the environment is the evidence for a big earthquake, which was closely connected with the Kikai-Akahoya eruption (Naruo, 1996). Fig. 17.7 shows a clastic dike, which was formed by fluidisation of

THE IMPACT OF KIKAI-AKAHOYA ERUPTIONS

321

Figure 17.7 Clastic dike formed by the co-volcanic earthquake of the Kikai-Akahoya eruption outcropping at Aira town, Osumi Peninsula

underlying Pleistocene pyroclastic flow deposits (Aira-Ito) as a consequence of a large earthquake. An examination of the parade of dikes with lateral intervals of c.10 m around the coastal areas facing the Kikai caldera (i.e. Yakushima, Tanegashima, Kuchierabu-shima) and the Osumi and Satsuma Peninsulas, South Kyushu shows clearly that this clastic dike penetrates into the Kikai plinian pumice layer of the first stage of the eruption; that it is mixed with or covered by the pyroclastic flow deposits (Takeshima-Koya unit); and that it is covered by the Akahoya fallout ashes. It seems reasonable to suppose, therefore, that the big earthquake which created this pattern occurred simultaneously with the cataclysmic eruption of Kikai. It might have been triggered or associated with the collapse of the Kikai caldera. The distribution of this feature clearly shows that the epicentre might have been located in or very close to the present Kikai caldera. On Mageshima Island, about 50 km east of the central part of the Kikai caldera, many boulders of coral limestone are scattered on the surface of the Pleistoceneage terrace, which is about 20 m above current sea level. It seems very likely that a great tsunami removed the boulders from the coral reef and tossed them on to the shore. We have not yet substantiated that the tsunami deposits are the product of the co-volcanic earthquake, because radiocarbon ages obtained from them are slightly younger than the Kikai event, but this still seems a likely hypothesis.

322

H. MACHIDA AND S. SUGIYAMA

CULTURAL DISCONTINUITY IN THE JOMON CERAMIC CULTURE Given the major environmental changes caused by the tephra falls, landslides, floods, earthquake and tsunami, it is not surprising that the phytolith data show that the forest was largely destroyed following the Kikai volcanic event. It seems likely that the massive changes in vegetation and landscape would have caused great hardships for the human population. At the time of the Kikai eruption, people belonging to the Neolithic Jomon culture occupied most of the Japanese islands, as shown in Fig. 17.8. The cultural sequences for this period have been particularly well studied in Kyushi. On the basis of stratigraphical and archaeological studies, Shinto (1978) pointed out a prominent lack of cultural discontinuity between the pre-K-Ah tephra, earliest Jomon phase of culture and the post-tephra, early Jomon phase. In contrast, Machida (1984) provided data to support the opposite conclusion, that the Kikai event had a very major impact on human societies and caused significant cultural change. This latter view has been substantiated by more recent archaeological data, which are discussed below. At the time of the Kikai eruption, Kyushu was occupied by an indigenous ceramic culture represented by the Kyushu-Kaigara-mon style pottery (identified

Figure 17.8 Geographical extension of the regional phases of the earliest Jomon ceramic culture immediately before the Kikai-Akahoya eruption Source:

Machida (1984)

THE IMPACT OF KIKAI-AKAHOYA ERUPTIONS

323

by the use of shell scraping as illustrated in Fig. 17.9) and by the Oshigata-mon style pottery (distinguished by a roller pattern), which was imported from the Honshu mainland. The Senokan style pottery is considered to represent the most recent native ceramic culture in Kyushu. It is significant that these ceramic types are not found in the soil above the K-Ah tephra and therefore may have been terminated as a consequence of the eruption. Based upon the style of pottery and associated artefacts from the soil above the K-Ah ash, the spatial distribution of Jomon ceramic culture after the Kikai eruption is shown in Fig. 17.9. The marked contrast between the pre-tephra and post-tephra situations, as shown in Figs 17.8 and 17.9, indicates a significant cultural hiatus. Archaeological research suggests that the post-ash pottery styles were derived from outside the region. The Jokon-mon style ceramic culture, called the Todoroki style in Kyushu, had its origin in East and/or Central Honshu and therefore immigrated into the area affected by the K-Ah tephra. In addition, the Sobata style pottey, which possibly originated in the Korean Peninsula or Northwest Kyushu, also moved southward after the Kikai eruption. Recent archaeological work has substantiated and expanded on Machida’s (1984) conclusion that the K-Ah tephra seriously disrupted human societies. First, it seems probable that the Jokon-mon style pottery occurs in the paleosoil below

Figure 17.9 Geographical extension of the regional phases of the early Jomon ceramic culture after the Kikai-Akahoya eruption Source:

revised from Machida (1984)

324

H. MACHIDA AND S. SUGIYAMA

the K-Ah tephra, but in very minor amounts (Kuwahata, 1996; Higashi, 1996). Second, no Senokan style pottery has been recovered from above the K-Ah tephra as Shinto (1978) suggested. At the Ishiki site in Kanoya, South Kyushu, very small quantities of the Sho type of pottery lie immediately above the K-Ah and below the Ikedako tephra, which dates to approximately 6,400 cal. BP (Okuno, 1996), i.e. about 900 years after the Kikai event. The occurrence of the pottery suggests that reoccupation in the highly devastated area took place several hundred years after the Kikai eruption. The still later Todoroki style pottery, which is recovered from above the Ikedako tephra, represents yet another attempt to resettle the region following abandonment as a consequence of a volcanic eruption. Third, it seems possible that a similar cultural discontinuity occurred not only in South Kyushu, but also in Central to North Kyushu and the western part of the Shikoku and Chugoku areas (Shinto, 1994). The relationship between tephra layers and artefacts therefore indicates that following the great Kikai eruption, people either totally perished or dispersed to the north, where the volcanic hazard was not so severe. The archaeological data also suggest that South Kyushu was abandoned temporarily after this volcanic event and was then reoccupied by bearers of a different cultural tradition. Migration into South Kyushu would have started several hundred years after the Kikai eruption and just before the Ikedako eruption. The presence of the Todoroki and Sobata style pottery above these tephra layers in many excavations indicates that permanent reoccupation only started after the Ikedako eruption (c.6,400 BP).

CONCLUSIONS Tephrochronological studies show that because of its very widespread distribution, the Kikai eruption of c.7,300 was of exceptional magnitude. Many forms of evidence indicate that the extensive burial and destruction of the forest, flooding, and an earthquake of extraordinary magnitude occurred in South Kyushu and adjacent areas as a consequence of the Kikai volcanic eruptions. Certainly the occurrence of such a major volcanic event would have had severe consequences for human societies at that time. Not only was the landscape radically changed, but the vegetation was also destroyed. Archaeological evidence provides further support for the proposition that the eruptions caused significant impacts on human cultures over a vast area. Particular styles of pottery indigenous to the region disappeared altogether and were replaced by new styles which had been developed elsewhere. People either perished as a result of the K-Ah tephra or moved elsewhere, and only many generations later did groups with a different cultural repertoire resettle the region. The particular case of the Kikai-Akahoya event demonstrates that explosive volcanic eruptions of large magnitude have had disastrous consequences for past human populations in Japan.

THE IMPACT OF KIKAI-AKAHOYA ERUPTIONS

325

REFERENCES Fukusawa, H. (1995) Non-glacial varved lake sediment as a natural timekeeper and detector for environmental changes. Quaternary Research (Japan) 34: 135–49. Higashi, K. (1996) Volcanic ash layers and types of potteries above Akahoya layer in South Kyushu. Summaries of Researches Using AMS at Nagoya University VII: 48–59. Kuwahata, M. (1996) Volcanic ash layers and types of potteries below Akahoya layer in South Kyushu. Summaries of Researches Using AMS at Nagoya University VII: 39–47. Machida, H. (1984) The significance of explosive volcanism in the prehistory of Japan. Geological Survey of Japan Report 263: 301–13. Machida H. (1999) The stratigraphy, chronology and distribution of distal marker-tephras in and around Japan. Global and Planetary Change 21: 71–94. Machida, H. and Arai, F. (1978) Akahoya ash: a widespread tephra erupted from the Kikai caldera, southern Kyushu, Japan. Quaternary Research (Japan) 17: 143–63. Machida, H., Moriwaki, H. and Zhao, D.C. (1990) The recent major eruption of Changbai Volcano and its environmental effects. Geographical Reports, Tokyo Metropolitan University 25: 1–20. Naruo, H. (1996) A clastic dyke formed by fluidization at the time of Kikai-Ah eruption in the central Osumi Peninsula. In K. Endo, H. Kumai, H. Machida, K. Okumara, C. Shimizu, K. Suzuki, M. Wantanbe, A. Uesugi, Y. Ikeda and S. Tsuju (eds) Inventory of Quaternary Outcrops Tephras in Japan, 311. Tokyo: Japan Association for Quaternary Research. Newhall, C.G. and Self, S. (1982) The Volcanic Explosivity Index (VEI): an estimate of explosive magnitude for historical volcanism. Journal of Geophysical Research (Oceans & Atmosphere) 87: 1231–8. Okuno, M. (1996) AMS-14C ages of tephra layers distributed on southern Kyushu, Japan. Summaries of Researches Using AMS at Nagoya University VII: 89–109. Ono, K., Soya, T. and Hosono, T. (1982) Geology of the Satsuma-Io-Jima District. Quadrangle Series, Scale 1:50,000. Tskuba: Geological Survey of Japan. Self, S. and Sparks, R.S.J. (1978) Characteristics of widespread pyroclastic deposits formed by the interaction of silicic magma and water. Bulletin Volcanology 41: 196– 212. Shinto, K. (1978) Volcanic ashes in South Kyushu with special reference to archaeology. Dorumen 19: 40–54. Shinto, K. (1994) South Kyushu originated potteries out of South Kyushu in the Early Jomon period. Newsletter, Jomon Study in South Kyushu 8: 56–64. Stuiver, M. and Reimer, P.J. (1993) Extended 14C database and revised CALIB 3.0 14C age calibration program. Radiocarbon 35: 215–30. Sugiyama, S. (1999) Lucidophyllous forest development since the Last Glacial age in Southern Kyushu, Japan, clarified by phytolith studies. Quaternary Research ( Japan) 38: 109–23. Ui, T. (1973) Koya pyroclastic flow – a discovery of unusually widespread and thin pyroclastic deposits. Bulletin of the Volcanology Society of Japan 18: 153–68. Walker, G.P.L., Heming, R.F. and Wilson, C.J.N. (1980) Low-aspect ratio ignimbrites. Nature 283: 286–7. Yamashita, K. (1980) The Otofuku-dani site. In Mazukari Shell Mound Site, 121–33. Board of Education: Minami Chita, Aichi Prefecture, Japan.

18

Volcanic disasters and archaeological sites in Southern Kyushu, Japan SATORU SHIMOYAMA

INTRODUCTION The role of volcanism in archaeology and cultural change in Japan is graphically demonstrated in Ibusuki, described by the excavator (Hamada, 1921) as the ‘Pompeii or Santorini of the prehistoric era in Japan’. In Ibusuki, Hamada demonstrated that the Jomon pottery types in use before the eruption were older and distinctly different from the Yayoi pottery found in the post-disaster stratigraphy. The volcanic disaster is therefore associated with a distinct change in the cultural assemblage. Disaster archaeology in Japan can be broadly divided into four categories. The first category uses tephra to date cultural artifacts (Arai, 1971; Machida and Arai, 1992). The second category ignores tephra isochrones and develops typologies of cultural material retrieved from archaeological contexts and explores the cultural developments that may be apparent (Shinto, 1978). The third category focuses on disasters themselves and explores the nature of disaster events by examining related features such as villages buried by tephra deposits (Noto, 1983). The fourth category seeks to develop a theoretical understanding of the influences and processes of disasters by simulation. These methodological approaches, when used in a complementary fashion rather than in isolation, may reveal the essence of the volcanic disaster. These approaches are illustrated by a study of archaeology and tephras from the Jomon period in Southern Kyushu. There have been many active volcanoes on Southern Kyushu, some of which are still active, such as Sakurajima volcano, and there are many distinct tephra isochrones that are of great value in archaeological research (Fig. 18.1). The ‘Akahoya’ tephra, erupted from the Kikai caldera, has helped establish the chronology of Jomon potteries on Kyushu. The cylinder-shape potteries of the first Jomon stage, while ubiquitous below, are not found above the tephra, whilst the Todoroki and Sobata type potteries are only found above the tephra (Machida and Arai, 1978; Shinto, 1978). The Akahoya tephra isochrone is widely distributed across Japanese islands, which has allowed the comparison of potteries from widely separated areas (Maizou-bunkazai-kenkyukai, 1987).

Figure 18.1 Distribution of tephras and principal archaeological sites during the Jomon period in Southern Kyushu

328

S. SHIMOYAMA

In Japan, there are several examples of archaeological sites which have suffered a great deal of direct damage from tephra fall. In these cases it is possible to move beyond questions of the chronological succession of cultural artefacts and consider the impact of the volcanic event upon the natural environment. Such studies are concerned with the third and fourth methodological approaches outlined above.

CASE STUDIES Sakurajima tephra, 11,500 BP The Sakurajima tephra (Sz-S, P14) erupted during the Kitadake stage of the Sakurajima volcano (Okuno et al., 1997) and has been used to establish the relative chronology of Ryutaimon type pottery and the Iwamoto cylinder-shape pottery that used shell-crafted tools. Ryutaimon type pottery has been discovered under the P14 tephra at the Soujiyama site in Kagoshima City and the Kakoinohara site in Kaseda City. The earthen potteries found above the tephra consist mainly of Iwamoto type pottery, which appears to be the oldest form of the cylinder-shape style, foreign to other regions (Fig. 18.2). The Ryutaimon type pottery disappeared within the thickest isopach of the P14 tephra, and was replaced by the Iwamoto cylinder shape, which suggests that the latter was developed outside the area and imported, or brought in by new settlers (Shinto, 1997; Fig. 18.3). If this hypothesis can be proven, the results will be useful for the first, second and fourth approaches outlined above. In future studies it will be necessary to collect concrete data about the correlations between the volcanic eruptions, tephra distribution and isopachs, damaged cultural material and the alternation of pottery styles. Sakurajima tephra, 9,400 BP The Sakurajima tephra (Sz-Tk3, P13) was formed during a later eruption of the Sakurajima volcano (Moriwaki, 1994; Okuno et al., 1997). In several historic sites, such as Uenohara in Kagoshima prefecture, Kokubu city, it was recognised that the P13 tephra eruption corresponds to the period of the Maebira type pottery, which belongs to the early half of the second Jomon period. A detailed chronology of the remains of the ancient houses was constructed using the P13 tephra. This study combines the first and third methodologies outlined above. Yonemaru and Sueyoshi tephras, 7,500 BP The Yonemaru tephra, which originated from a maar volcano in the Aira district, has been found across the Kagoshima Bay area, in the northern Oosumi Peninsula and in Miyakonojo City (Moriwaki, 1994; Okuno et al., 1997). This tephra is useful for the distinction of Senokan type pottery and the Hiragakoi type pottery which belongs to the latter half of the second Jomon stage (c.10,000 BP–6,000 BP) in Southern Kyushu (Fig. 18.4) and is found over much of Kyushu (Maizoubunkazai-kenkyukai, 1987).

VOLCANIC DISASTERS AND SITES IN SOUTHERN KYUSHU

329

Figure 18.2 The sequence of Ryutaimon type pottery in Southern Kyushu

Senokan type pottery is found between the Yonemaru tephra and the Sueyoshi (Sz-Sy, P11) tephras (Kuwahata, 1996). In the immediate future, these tephras will contribute potentially to the detailed classification of the Senokan type, and on the understanding of the direction of the Senokan type change. This approach applies comes under the first and the second categories of disaster research.

Sites of Douchinishitype discovered (early cylinder-shaped pottery) Sites of Ryutaimon-type discovered

Sphere of influence by P14

Kikai caldera

Stage 1 The later stage of the incipient Jomon period (12,000 BP Ryutaimon-type)

Stage 2 The eruption occurred (11,400 BP Ryutaimon-type disappeared)

Sites of cylinder-shaped pottery discovered

Stage

Stage 3 The second Jomon period (11,400 BP Iwamoto-type, cylinder-shaped pottery appeared)

Period

Range of pottery type

1

The incipient the Jomon period (Jomon I)

Ryutaimon-type (pottery with heavy cray strings)

2

Eruption of Sastuma P14

Douchinishi-type (pottery with heavy cray strings and nails stamp)

3

Early Jomon period (Jomon II)

Iwamoto-type (cylinder shaped pottery with traces of shell)

* This figure is quoted from Shintou Kouichi/ (1997), and a part of this figure is annotation added by this author

Figure 18.3 The spread of the potteries above and below P14

* This figure is quoted from Kuwahata Mitsuhio/ (1996). and the figure is arranged by this author

Figure 18.4 Tephras and types of pottery on Satsuma and Osumi Peninsula

332

S. SHIMOYAMA

Akahoya tephra, 6,500 BP The Akahoya tephra (K-Ah) from the Kikai caldera is found extensively across the Japanese islands and confirms the chronological typology of the Jomon pottery devised independently by archaeologists (Kawaguchi, 1972; Machida and Arai, 1978). The Akahoya tephra has been used to separate pottery of the second and third Jomon groups. Senokan pottery has been discovered under the Akahoya volcanic ash whilst Todoroki type has been discovered above it. Shinto (1978) speculated that the 6,500 BP eruption of the Kikai caldera caused an unspecified disaster, which effectively eradicated the Senokan culture. This allowed the culture using Todoroki type pottery to expand in Southern Kyushu. This hypothesis is only plausible if the eruption brought about considerable destruction to Southern Kyushu (Shinto, 1978; Fig. 18.5), but the mechanism by which an eruption could destroy one culture while allowing another to expand remains unclear. However, it has also been suggested that the change from Senokan pottery to Todoroki pottery only occurred within a limited area of Kyushu where the direct impact of the eruption is unequivocal (Kawaguchi, 1986). The contrasting nature of these theories illustrates the current debate within archaeology as to the nature of such changes and whether the forces that cause them are internal or external. The uncritical use of external forcing factors in the past has led to quite reasonable resistance to the application of environmental determinism in many Japanese archaeological studies and such changes are now normally explained in terms of internal forcing factors. However, recent developments within archaeological theory may rehabilitate the use of external forcing factors and volcanic eruptions to explain cultural change (cf. Chapter 1, this volume). Past discussions have been based on a priori thinking, where the scale of impact of the external forcing mechanism has been assumed to be sufficient to bring about the change observed in the archaeological record, without real regard to the scale and magnitude of the external mechanism itself. Future studies that invoke external forcing mechanisms need to be conducted with a realistic understanding of such mechanisms. To be effective such studies need to be holistic in nature and not rely solely on pottery typologies. It seems inevitable that the results of such studies will pose theoretical questions as to the effects of a volcanic eruption or other forcing mechanism. This research question in particular is one that could effectively be modelled and should develop to the fourth category of studies. Ikeda tephra, 5,500 BP In the Ibusuki region several tephras, which were erupted 5,500 years ago from the Ikeda caldera, are found (Naruo and Kobayashi, 1984; Okuna, 1997). These are the tephra fall (Ik-afa), the pyroclastic flow deposit (Ik-pfl), the pumice fall (Ik-pfa), the Yamagawa base-surge (Ym-bs), the Osagari scoria fall (Ik-Os) and the Ikezaki volcanic ash (Ik-Ik). The discovery of the Ataka type pottery at a high point of the Yamagawa base-surge (Ym-bs), which erupted at the same time as the Ikeda tephra was formed, places this pottery in the middle Jomon period (Kuwashiro, 1967; Naruo, 1997).

Figure 18.5 The movement of Jomon potteries under and above the eruption of Kikai caldera

334

S. SHIMOYAMA

Todoroki type pottery was discovered below the Ik-afa tephra at the Ishiki site in Kanoya city and above it at the Chinjugasako site, which has resulted in a subdivision of the Todoroki type pottery. Such studies belong to the second category of disaster research outlined above. But it is interesting to note that the survival of the pottery style across the tephra layer suggests that the cultural impact of the eruption may have been minor. Miike pumice, 4,200 BP The Miike maar (Kr-MiP) is one of more than 20 volcanoes in Kirishima (Okuno, 1997), and the pumice of the 4,200 BP eruption is distributed throughout the northeastern region of Southern Kyushu. At the Maetani excavation the pumice fall overlies Kasuga type pottery of the fourth Jomon period. Elsewhere at the Iwatate excavation, Ohira and Ataka type pottery has been discovered above the tephra horizon and is therefore younger than the tephra fall. The association of these cultural artefacts with the Miike Maar places the eruption in the latter half of the middle Jomon period (the fourth Jomon period). This volcanic event may fall into the third category outlined above, where the tephra is intimately related with domestic and cultural material. A dwelling pit, covered with a thin soil layer, was found directly under the tephra during the excavation of the Isetani site in Miyakonojo city (Yokoyama, 1998). A house in the dwelling pit was buried by the tephra shortly after it had been abandoned. Further excavation may reveal more of the settlement affected by the pumice fall. Kikora tephra, 4,000 BP The Kikora deposit (Km-1; YK: Yellow-Kora), which is distributed throughout Ibusuki region, is a scoria fall initially ejected from the 4,000 BP eruption of Kaimondake (Naruo, 1984; Fujino and Kobayashi, 1997). The tephra layer is compacted and occasionally incorporates traces of plants that suffered during the disaster. Because the Ibusuki type pottery belonging to the fifth Jomon period was directly covered with the tephra at the Narikawa excavation in Yamagawa town in Kagoshima prefecture, the eruption clearly corresponds to the later Jomon period (the fifth period). The eruption coincides with the change from Ibusuki type pottery to Ichiki type pottery. It is clear that where the tephra accumulated Ibusuki type pottery fell out of use, to be replaced in the south by Ichiki type pottery. It is easy to propose the hypothesis that the change of pottery styles was caused by a cultural or economic disaster of volcanic origin, but to date a sustainable forcing mechanism is lacking and a dependent relationship between the two events has not been established. Haikora tephra, 3,200 BP The Haikora tephra (Km-4, GK: Gray Kora) is a scoria fall from the 3,200 BP eruption of Mt Kaimondake, which was distributed in a northwesterly direction from its source (Fujino and Kobayashi, 1990). It is thin in the Ibusuki City area, which is 10 km away from Kaimondake. At the Shinbansho excavation, in Ibusuki City, Kurokawa type pottery, which was in use during the final Jomon

VOLCANIC DISASTERS AND SITES IN SOUTHERN KYUSHU

335

period, was discovered directly below the Gk tephra. In short, this tephra covered the pottery group just before the first stage of the Yayoi period. In contrast, at the Kunden excavation in Karatsu City in Saga prefecture, Kurokawa type pottery is placed in an older period than suggested by radiocarbon dating 290 ± 50 BC, which corresponds to the first Yayoi period. In Karatsu City at the Nabatake site, materials of the latest Jomon period are radiocarbon dated to 680 ± 30 BC and 670 ± 60 BC by radiocarbon dating (Nakajima, 1982). In addition, the radiocarbon dating of carbonised rice in Korea at the Kinganri site, 2,980 ± 70 BP (Okazaki, 1982) determined that Kouretsumon pottery was contemporary with the use of Kurokawa type pottery in Japan. It is clear that more work is necessary to firmly establish the joint chronology of Japanese and Korean cultures. Research that combines the first and second approaches outlined above will help resolve these issues.

TEPHRA FALLS AND CULTURAL RESPONSE: ERUPTIONS OF MT KAIMONDAKE Tephrochronology was used successfully to construct a chronology of Jomon pottery in Southern Kyushu and was a great aid to archaeological investigation in the region. However, it is clear that where a tephra deposit coincided with a change in pottery style the archaeologists readily assumed that the eruption caused a cultural catastrophe, regardless of the scale of the eruption or the nature of the tephra deposit itself. This approach is unsatisfactory. It is necessary to demonstrate the effects of the eruption and to reconstruct the process of the disaster in detail. In particular it is essential to clarify the precise mechanism by which the often distant volcanic eruption was able to wield such a profound impact upon the exposed culture. The problem inherent in adopting a deterministic approach towards establishing a relationship between cultural change and tephra falls in the Jomon period is well illustrated by a brief examination of volcanic eruptions and the response of affected cultures during the historical period at the Hashimuregawa archaeological site (sixth to ninth century AD). These appear to show that cultural change was brought about by the expansion of the Ritsuryo system rather than by any of the frequent volcanic eruptions that occurred during that period. A village that suffered directly from the Kaimondake eruption was discovered in 1988 at the Hashimuregawa site, which is located on a volcanic alluvial fan. Several tephra layers are found: Ik-pfa (5,500 BP: Ikeda caldera), Km-1, Yk (4,000 BP: Kaimondake), Km-4, Gk (3,200 BP: Kaimondake), Km-9 (2,000 BP: Kaimondake), Km-11c, Ak (1,300 BP: Kaimondake), Km-12a, Mk (1,100 BP: Kaimondake, AD 874 based on document history and archaeological dating). The Ak and Mk volcanic events will be considered in detail here. The Ak eruption appears to have been a two-stage event. A humus layer formed upon the initial scoria deposit before it in turn was buried by subsequent tephra fall. In this discussion the initial scoria deposit is designated Ak-1 and the

336

S. SHIMOYAMA

later tephra fall Ak-2. Volcanological opinion is divided as to whether these deposits represent a single eruption cycle or were ejected in separate events. However, as each deposit appears to have had a distinct impact upon the underlying archaeology they will be considered as separate events here. There are conflicting interpretations of the eruptive activity recorded by the Ak tephras: one is that Ak-1 and Ak-2 are stages within a distinct cycle of volcanic eruptions, and the other is that they originate from different eruptive cycles. Because it is apparent that each set of ejected materials is associated with different cultural responses, they will be considered here as two separate disasters. The damaged settlements and the distribution of the Ak-2 and Mk tephras are shown in Figs 18.6 and 18.7. It is thought that the Ak-1 scoria was deposited between the end of the sixth century and the first half of the seventh century AD. The Ak-2 tephra fall dates to the last quarter of the seventh century AD. (Shimoyama, 1997). In the case of the Mk deposit absolute dates exist. The date of the eruption is given as 25 March AD 874 in a historical document ‘NIHON SANDAI JITSUROKU’ compiled in the tenth century, which also broadly confirms the conclusions reached by geological investigations of the eruption deposits (Nagayama, 1992; Naruo, 1992). It is clear from excavation that both the Ak and Mk events had direct impacts upon roads and paths and damaged fields, paddy fields, dwelling houses and rivers. The impact of the eruption upon the culture and environment was determined by documentary research, geological study and archaeological excavation. Using this combination of techniques it was possible to assess the ways in which the culture adapted to the volcanic eruption. The structures of the village were discovered both on and under the 5 cm thick Ak-1 tephra layer, which suggests that life in this ancient village continued regardless of the tephra fall. It is clear that its apparent victims did not perceive this event as a serious threat. The Ak-2 tephra layer is 20 cm thick and it appears that, in contrast to Ak-1, this event seriously affected all the facilities of the ancient village. The attack factors identified included burial, pressure and compaction, pressure and hardening. Archaeological excavation revealed direct damage to the river, roads, shellmounds and ditches. Direct damage to domestic houses has not been identified, settlement continued and several houses, which had accumulated tephra, were cleaned out and continued in use. Even the shell-mounds continued in use in the same location as those buried beneath the tephra, a clear indication that the preexisting community continued in being. There are no traces of the resumption of food production in the fields, and it appears that the community depended upon marine resources for food. There is also evidence of spiritual activity in response to the eruption. There is an example of a pot being offered during the accumulation of the Ak-2, seemingly to propitiate the disaster by spiritual means. This is judged as one of the emotional adaptation methods used in response to a disaster. Further adaptation is evident after the tephra fall, where compacted tephra was used as a roadway; people clearly understood the inherent properties of the material. The Mk tephra, which is 30 cm or more in thickness, had a more profound impact upon the settlement. All facilities in the village received serious damage

Figure 18.6 Thickness of Ak-2 (km-11) and archaeological sites around Ibusuki City

Figure 18.7 Thickness of Mk (km-12) and archaeological sites around Ibusuki City

VOLCANIC DISASTERS AND SITES IN SOUTHERN KYUSHU

339

from the burial, adhesion, compaction and hardening of tephra. There is no evidence that the community adapted to the tephra fall: shell-mounds did not continue in use and it appears that the settlement was abandoned. However, as there is no evidence of any person killed during the disaster it appears that abandonment was a strategy consciously adopted by the community. Together these case studies provide us with useful paradigms for the responses of communities to volcanic disasters. In the case of the Ak-1 tephra, the influence of disaster on the community and its environment was slight, whilst in the case of Ak-2 the community adapted to their changed environment; in the case of the Mk event they selected abandonment. Cultural change did occur at Hashimuregawa during the period under consideration, but this mainly took place between the Ak-2 and the Mk eruptions, not during the periods when the influence of the volcanic eruption was at its greatest. Indeed the greatest change observed appears to have occurred with the introduction of the Ritsuryo system of government into this area. The Narikawa type pottery disappeared synchronously with this historical event. The village buried by the Mk tephra followed the normal village pattern of the Ritsuryo system, such as standardised roads, dwellings and farmland. The extension of the Ritsuryo system and the resultant cultural contact had a greater influence upon the cultural items of local communities than the volcanic eruptions.

DISCUSSION The problems inherent in many archaeological approaches to the role of volcanism in cultural change are illustrated by a series of examples drawn from the Neolithic Jomon period in Southern Kyushu. It is clear from these examples that if archaeologists are to invoke a volcanic eruption as the mechanism responsible for cultural change, it is necessary to construct a sustainable model whereby a specific eruption may have brought about the changes observed. In short, it is necessary to define exactly how a particular volcanic eruption was able to exert an influence upon the culture and verify this in the archaeological record. Pottery style may be observed to change either side of a tephra horizon, but were fields abandoned, were water courses choked, did people continue to use the same areas to dispose of waste, did buildings continue in use? When such factors are considered it can be seen that in most cases the cultures that have been studied adapted to the volcanic event in a manner that suggests familiarity with the problem. In summary, from these examples it can be seen that the extent to which cultural items change depends on the degree to which the culture was able to adapt to the external forcing, in this case volcanic. Therefore, we may classify the effects of a disaster into three categories, which measure the impact spatially and temporally: 1 A cultural item disappears or is abandoned in the short term, and a culture of the same system continues after a disaster (the case of Mk; see also similar

340

S. SHIMOYAMA

points made in Chapter 16). In the case of a major of volcanic disaster, it is necessary to consider the cultural inflow of another system from areas unaffected by the disaster (the case of Ak). 2 The cultural item is changed because people adapted to the changes in their environment. There are also cases where people invent a new cultural item (the case of Ak-2). 3 There is no direct effect but, as demonstrated in the case of Ak-1, affected people may emigrate from the region. It is clear from the discussion above that it is necessary to understand the physical nature of the volcanic event, the magnitude of the eruption, the spatial distribution of the tephra and its depth; other factors could also be added, such as the chemical impact of adsorbed volatiles, the post-depositional behaviour of the tephra itself, the nature of compaction, etc. Only by establishing these factors independently of any archaeological influence is it possible to determine the potential of the volcanic event for cultural and material disruption. As an independent process, the prosperity and decay of cultural items obtained from excavation may also be established. Once both these factors have been determined it is possible to explore realistically the relationship, if any, between the volcanic event and the cultural material. In prehistoric contexts, where documents are not in existence, the nature of the volcanic event and the cultural response must be established after carefully planned, precise excavations. The relationship between volcanic events and culture in Southern Kyushu has been established through interdisciplinary studies, which involve not only archaeologists, but also volcanologists, historians, botanists and zoologists. Finally, to develop coherent theories to explain cultural response to volcanic events it is necessary to accumulate case studies from more than one region. Different cultures may have alternative strategies by which to cope with volcanic events and these must be established before a comprehensive theory of volcanic events and cultural change can be established.

REFERENCES Arai, F. (1971) Lithic artifact-bearing layers in the Kanto Loam in North Kanto, Japan problems on the so-called Early Palaeolithic in geology –, Japan. Quaternary Research (Japan) 10: 317. Fujino, N. and Kobayashi, T. (1997) Eruptive history of Kaimondake volcano, Southern Kyshu, Japan. Volcanological Journal of Japan, Ser. 3, 42: 195–211. Hamada, K. (1921) A prehistoric site at Ibusuki in the province of Satsuma and pottery found in it. Report upon Archaeological Research, Department of Literature, Kyoto University, Japan. 6: 45. Kawaguchi S. (1972) Senokanshiki-doki (Senokan type pottery). Kagoshima-Kouko, Kagoshima-koukogakkai 6: 1–35. Kawaguchi, S. (1986) Sousetsu. Kagoshima-Kouko, Kagoshima-koukogakkai 20: 1–10. Kuwahata, M. (1996) Volcanic ash layers and types of potteries below the Akahoya layer in South Kyushu. Summaries of Researches Using AMS at Nagoya University VII: 39–47.

VOLCANIC DISASTERS AND SITES IN SOUTHERN KYUSHU

341

Kuwashiro, H. (1967) Kaimondake funshutsubutsu ni tsuite (volcanic materials from Kaimondake Volcano). Chiran Bunka 4: 42. Machida, H. and Arai, F. (1992) Atlas of Tephra in and around Japan. Tokyo: University of Tokyo Press. Machida, H. and Arai, F. (1978) Akahoya ash – a widespread tephra erupted from the Kikai caldera, southern Kyushu, Japan. Quaternary Research (Japan) 17: 143–63. Maizou-bunkazai-kenkyukai (1987) Kazanbai to Koukogaku o meguru Shomondai, Shiryoshu (comprehensive compilation of cases), Maizou Bunkazai Kenkyukai Kagoshima. Japan: Jikkouiinkai. Moriwaki, H. (1994) Stratigraphy and distribution of the Sakurajima tephra group. Unpublished report of Grant-in-aid for research. Ministry of Education and Culture (in Japanese). Nagayama, S. (1992) On the evidence of the volcanic activities of Mt Kaimon obtained from the ‘Nihon Sandai Jitsuroku’. Report of the Hashimuregawa Archaeological Site 3: 501–10. Nakajima, N. (1982) ‘Nabatake’ the report of the archaeological investigation. Karatsu City 5: 572–4. Naruo, H. (1997) Origins and dating of tephra in Ibusuki area. The Society of Human History 9: 148–58. Naruo H. (1992) On the geological features of the Hashimuregawa Archaeological site. Report of the Hashimuregawa Archaeological Site 3: 511–22. Naruo, H. and Kobayashi, T. (1984) Pyroclastic fall deposits from the Ikeda caldera. Japanese Bulletin of Volcanology 29: 148. Noto, T. (1983) Gunmaken-ka niokeru Maibotsu Denhata-chousa no Genjyo to Kadai – Kazansaigai eno Koukogakuteki approach. Gunmakenshi- Kenkyu 17: 14–51. Okazaki, T. (1982) Jomon Banki oyobi Yayoi Zenki no Kome, ‘Nabatake. Karatsu City 5: 565–71. Okuno, M. (1997) Accelerator mass spectrometric radiocarbon chronology during the last 30,000 years of the Aira caldera, Southern Kyushu, Japan. Unpublished report of Grant-in-aid for research, Ministry of Education, Science and Culture, 2051. Okuno, M., Nakamura, T., Mariwaki, H. and Kobayashi, T. (1997) AMS radiocarbon dating of the Sakurajima tephra group, Southern Kyushu, Japan. Nuclear Instruments and Method in Physics, Japan, 123: 470–4. Shimoyama, S. (1997) On the range of disaster archaeology – assessing the effects of disasters. Hominids – Reconstructing Archaeology 1: 83–103. Shinto, K. (1978) Kazanbai to Dokikeishiki. Dorumen 19: 40–54. Shinto, K. (1997) Impact of the eruption of Sakurajima on the incipient Jomon culture in Southern Kyushu. Journal of the Society of Human History 9: 95–103. Yokoyama, T. (1998) Miyazakiken, Miyakonojo-shi, Isetani-iski nitsuite. Kazansaigai to Jinrui no Tekiou, 50. Japan: Senshigaku Kenkyukai.

Index

Adams, R. 117 Afognak Island: abandonment of 174, 185; Afognak Native Corporation 178; Dig Afognak programme 178; erosion 185; tsunami 185; Settlement Point 172, 178–180, 182–184; see also Alutiiq Ahmose Stele 251 Aitape disaster 28, 30, 31–35, 162; Arop 32, 34, 35, 39; casualty rates 35; cultural impact 35, 36, 39, 41; Lemieng 39; Malo mission station 35, 36; Malol 32, 34, 35; Pou 35, 36; Ramou 35, 36; re-settlement of survivors 35–36; Rowoi 35, 36; Sera 39; settlement pattern, impact on 35–36, 39; Sissano 32, 34, 35, 39; social consequences of 36; tsunami sediments 40; victims’ experiences 33, 34; Warapu 32, 34, 35, 39 Akahoya-Kikai: see Kikai-Akahoya Akrotiri 253, 254; abandonment of 274 Aleutian Islands: economy of 194–195; maritime climate 194; seismic activity 187, 195; storms, danger of 195; Umnak Islands 196; Unalaska Islands 196; volcanism 195 Aleuts: earthquakes, cultural impact of 201, 203; earthquakes, oral traditions of 10, 13, 194, 195, 196; maritime economy 195; natural disasters, oral traditions of 10, 13, 194, 195, 196; volcanism, cultural impact 195–196; volcanoes, oral traditions of 196 Alutiiq: Afognak, abandonment of 174, 185; angyaqs 185; earthquakes, cultural impact 14, 172, 174, 176, 184, 185, 188–189; earthquakes, oral traditions of

186–187; economy 173, 174; housing 173, 174; Kaguyak, abandonment of 174; maritime economy 185; natural disasters, cultural resilience to 172, 185–186; natural disasters, oral traditions of 186–187; Old Harbor, destruction of 174; population density 173; Russian colonisation 173, 185; social stratification, emergence of 175, 176; trading network 172, 185–186; tsunamis, settlement destruction 174; volcanoes, oral traditions of 186–187; see also Afognak Island, Kodiak Archipelago, Settlement Point Aoba: destruction of 164–165 Arbon-Bleiche: abandonment of 235, 236, 240, 246; inundation of 236, 243; chronology 238; discovery of 237; excavation of 237; settlement archaeology 238; topography 238–239; see also Bodman-Schachen, lake villages Arenal eruptions: cultural impact 292, 294, 295, 298 Arkalochori Cave 258 Arop 32, 34, 35, 39; Aitape tsunami, devastating effect of 34, 35; casualty rates 35; relocation of 39 Ataka pottery type 332, 334 Atlantis: legend of 251, 273 Auckland Volcanic Field 128, 132, 134, 139–140, 151; cultural impact 139–140; oral traditions of 150 Augustus Caesar 111, 115 Bacolor: abandonment of 43, 44; attachment to place 60, 61; barangays, abandonment of 53, 54, 60; barangays,

344

INDEX

destruction of 51–53, 56, 60, 61; casualty rates 52, 53; churches, rehabilitation of 57, 61; colonial architecture 49–50; commerce, loss of 59, 60; drainage network 59; failure to evacuate 51–53; flood defences 48; history of 48, 49; houses, raising of 54, 55, 56; Pinatubo eruption, destruction caused by 43, 50, 51, 52, 53–61, 63; plaza 49; refugees 52; resettlement programme 53, 54, 60; San Guillermo church 51, 52, 57, 61; San Vicente chapel 57; Santa Ines chapel 57; schools, rebuilding of 57–59; volcano defences 51, 52, 60; see also Pinatubo Bali: natural disasters, concealment of 10 Bariles chiefdom: Baru eruptions, cultural impact of 292, 297–298 Baru eruptions: cultural impact 292, 297–298 Batavia: abandonment of 74–75; discovery of 69, 75, 79; jettisoning of equipment 73, 75; massacre of survivors 69, 76, 77; mutiny 68, 69, 76, 77, 78; participant accounts 78, 79; Pelsaert, Francisco 70, 72–78; rescue voyage 69, 75, 76, 77–78, 79; rioting 73; salvage of 75, 76, 78; survivor camps 76, 77, 80, 81; wreck of 67, 69–73 bet-hedging 207–210, 226, 227–228 Bitokara: sedentism, development of 302; volcanism, impact of 296, 299; see also West New Britain Bodman-Schachen: abandonment of 235, 236, 240, 246; Bodman-Breite 244–245; Bodman-Löchle 237; Bodman-Weiler 237; chronology 238; excavation of 237; inundation of 236, 244; settlement archaeology 238, 239; topography 239; see also ArbonBleiche, lake villages Boscoreale villa 118; see also Pompeii Cabalantian: destruction of 52, 53, 56, 60, 61; see also Bacolor Cabetican: chapel, rehabilitation of 57; school, rehabilitation of 58; see also Bacolor Camp Century ice core 267, 270 Campania: agricultural economy 110–111; archaeological significance 117–119; aristocratic residences 111, 114, 115; commerce 111, 115, 116, 118; earthquakes 112, 113; economic

decline 114–115; imperial properties 111; maritime importance 111; Roman economy 110–111; Vesuvius eruption, impact of 114–116; wealth of 111–112; wine trade 115, 116, 118; see also Herculaneum, Pompeii, Vesuvius Capampangan: history of 48; volcanism, experience of 48, 49; see also Bacolor Capua: Roman economy of 115 Cassius, Dio 110, 113 Cerro Blanco: abandonment of 219, 221; architecture 218–219, 221, 223, 224; burials 218, 219; environmental fluctuations, cultural impact of 225; Huaca del Luna 218, 221; Huaca del Sol 218, 221; huacas 218, 221, 223, 227; Moche I–IV economy 217; see also Moche culture Chan Chan Valley 215, 218 Changbaishan 313; see also Kikai-Akahoya Chania villa 259; see Minoan civilisation Chernabura Island: settlement history 198, 200, 202 China: volcanism, environmental impact of 269, 271, 273 Chita Peninsular 320 Conway, Margaret 34 Cook, James 126 Crete: see Minoan civilisation crisis cults 256 culture: earthquakes, cultural impact 1, 14, 172, 176, 184, 185, 188–189, 194, 201, 203, 251, 252, 259, 260; natural disasters, cultural impact 1, 2, 3, 4, 7, 8, 9, 14, 15, 23, 25, 26, 35–36, 38, 39, 43, 87–88, 98–99, 100, 134, 137–144, 151, 172, 174, 176, 184, 185, 188–189, 193, 195–196, 201, 203–206, 245, 246, 247, 250, 251, 252, 255, 259–260, 264, 265, 266, 273, 274, 292, 294, 295, 297–303, 322–324, 332; natural disasters, cultural resilience to 172, 185–186, 292, 295, 302–303, 309–310; tsunamis, cultural impact 35, 36, 174, 192, 193, 254, 259, 278–279; volcanoes, cultural impact 1, 2, 7, 8, 9, 14, 15, 23, 25, 38, 87–88, 98–99, 100, 134–135, 137–144, 151, 153, 195–196, 252, 259–260, 264, 265, 266, 273, 275, 280, 292, 294, 295, 297, 300, 301, 303, 322–324, 332; see also culture change culture change: environment, relationship to 1, 2, 3, 10, 15, 23, 220–221, 225,

INDEX

245, 246, 247; natural disasters, relationship to 1, 2, 7, 8, 9, 15, 20, 23, 25, 26, 308–310, 313, 314, 326, 332, 334, 335, 336, 339–340; theory of 3–4; volcanism, relationship to 1, 7, 8, 9, 15, 308–310, 313, 314, 325, 332, 334, 335, 336, 339–340 Dakataua volcanic centre 296, 297, 299, 300, 301, 303, 304, 307, 308; see also West New Britain disasters: definition of 5, 6; disaster research 19–20, 26; human response to, structure of 67–78, 80–81, 84; psychological impact of 66, 74, 76; see also natural disasters Dye 3 ice core 267, 268, 269, 270, 271, 272 earthquakes: cultural impact 1, 14, 172, 176, 184, 185, 188–189, 194, 201, 203, 251, 252, 259, 260; Minoans, impact on 259, 260; oral traditions of 10, 13, 186–187, 194, 195, 196; Prince William Sound 173, 174, 176, 177, 179, 180, 184, 185, 186, 187, 188, 193, 201 Egypt: pharaohs, execution of 256 El Niño 6, 12, 211–215, 280; Ecuador, effect on 211; ENSO 211, 213, 214, 216; Moche culture, impact on 221; Peru, effect on 210–216 Etna Volcano 269, 270, 271, 273 evolution: bet-hedging 207–210, 227–228; natural selection 204 Exodus: myth of 273 Fiji: Taveuni Island eruption, cultural impact of 151, 153; Taveuni Island eruption, oral tradition of 153 Flavian dynasty 115 flooding: cultural impact 1, 8, 10, 15, 23, 245, 246, 247; see also lake villages France: Laki Fissure eruption, environmental impact 90–91, 93–94, 96, 98 Freud, Sigmund 116 Galatas 257; see also Minoan civilisation Galindo 221–223; architecture of 223, 224; cemetery 224, 226; economy 222–223; huacas 223–224, 226; Moche V archaeology 227, 228; see also Moche culture

345

Garau Island: sedentism, development of 302; volcanism, impact of 296, 299, 300, 307; see also West New Britain Garu plantation: volcanism, impact of 297, 299, 304; see also West New Britain Gell, W. 117 Germany: Gleichberg ‘eruption’ 101; Laki Fissure eruption, environmental impact 91, 95 GISP2 ice core 268, 269, 270, 271 Gleichberg ‘eruption’ 101 Goethe, J. 116 GRIP ice core 268, 269, 270, 271, 272, 273 Guague: flood defences 48; see also Pinatubo Haghia Triada villa 259; see also Minoan civilisation Haikora tephra 334–335 Hamilton, William 116 Hashimuregawa: abandonment of 25; damage, preservation of 23; volcanism, impact of 335–336, 339 Hekla tephra 88 Herculaneum 10, 107, 110, 111, 113, 114, 116, 117–119, 264; abandonment of 114; archaeological significance 117–119; aristocratic residences 111; excavation of 107, 118, 119; preservation of 110; restoration of 113, 114; tourism 116, 117; see also Pompeii, Vesuvius Heuheu, Te 149 Hiragakoi pottery 328 Hochstetter, F. von 149 Hogarth, W. 117 huacas 218, 221, 223, 224, 226–227; see also Moche culture Ibusuki 326, 332, 334 Iceland: depopulation of 87; Hekla tephra 88; Laki Fissure eruption, environmental impact 87, 88, 92–93, 102 Ichiki pottery 334 Ikeda caldera 332, 334; cultural impact 334 Ikedako tephra 317, 323, 324 Ikezaki: volcanic ash 332 Ilopongo eruption: cultural impact 264, 292 International Decade for Disaster Reduction 2, 5 Italy: Laki Fissure eruption, environmental impact 91–92, 95

346

INDEX

Itatsuke: damage, preservation of 23 Iuktas 258 Iwamoto pottery 328 Japan: ceramic chronology 323, 324, 326, 328, 329, 332, 334–335; culture change and volcanism 313, 314, 326, 332, 334, 335, 336, 339–340; disaster archaeology 326; Haikoro tephra 334–335; Ikeda eruption 332, 334; Ikedako eruption 317, 323, 324; Ikezaki ash 332; Jomon culture 322–324, 326, 328, 332, 334, 335, 339; Kaimondake eruptions 20, 25, 334, 335–336; Kaya fall 314; Kikai-Akahoya eruption 313, 314, 315, 316, 317, 318, 319, 320, 321, 322–324, 326, 332; Kikai-Tozurahara 314; Kikora tephra 334; Kirishma volcano 314; Ko-abi yama tephra 314; Koseda 314; Kyushu Island 313–324, 326–340; Miike pumice 334; Osagari scoria fall 332; Ritsuryo system 335, 339; Sakurajima, eruption of 314, 326, 328; SatsumaIwodako volcano 314; Shinto shrines 152; Sueyoshi tephra 329; Takeshima fall 314, 315; tephrochronology 323, 324, 326, 328, 329, 332, 334–335; volcanism, cultural impact 15, 23, 25, 265, 322–324, 332, 334; Yamagawa base-surge 332; Yayoi culture 326, 335; Yonemaru tephra 328, 329 Jomon culture: 322–324, 326, 328, 332, 334, 335, 339; Kikai-Akahoya eruption, cultural impact 322–324 Julio-Claudian dynasty 111, 114 Kachemak tradition 175, 176, 188 Kaguyak: abandonment of 174 Kaharoa eruption 134, 140, 151, 152, 153; oral tradition of 150; tephra 128, 134, 153, 154 Kaimondake volcano 20, 25, 334, 335, 336 Kasuga pottery 334 Kawakawa eruption 147–148 Kaya tephra fall 314 Kenai Peninsulas 172 Kikai-Akahoya eruption 313, 314, 315, 320, 326, 332; ash fall 318–319; chronology 316, 324; clastic dikes 320–321; cultural impact 322–324, 332; deforestation 317–318, 319, 322, 324; earthquake 320–321, 322, 324; environmental impact 317–321, 322,

324; flooding 319–320, 324; landslides 318–319, 322; severity of 313, 314; tephra 314–316, 318, 319, 323, 324, 326, 332 Kikai-Tozurahara tephra 314 Kikora tephra 334 Kirishma volcano 314 Knossos 251, 257; cannibalism, evidence of 258; destruction of 259 Ko-abi yama tephra 314 Kodiak archipelago: Afognak Native Corporation 178; coastal erosion 187–188; Dig Afognak programme 178; earthquakes, cultural impact 176, 184, 185, 188–189; geology 174; initial settlement 173, 174; Kachemak tradition 175, 176, 188; Koniak Tradition 175, 176, 178, 179, 184, 188; maritime economy 173, 174–175; Mill Bay, erosion of 187; natural disasters, oral traditions of 186–187; Ocean Bay Tradition 174–175; Prince William Sound earthquake 173, 174, 176, 177, 179, 180, 184, 185, 186, 187, 188, 193, 201; Russian colonisation 173, 185; salmon harvesting 173; seismic activity 172, 173, 176–177, 180, 184, 188; Sitkalidak Island 188; social stratification, emergence of 175, 176; tsunamis 173, 177; see also Alutiiq Kommos 257 Koniak tradition 175, 176, 178, 179, 184, 188 Koseda ignimbrite 314 Kouretsumon pottery 335 Krakatau volcano 253, 264, 265; recolonisation of 301 Kurokawa pottery 334, 335 Kurvot tsunami: oral tradition of 165–166, 170, 171 Kurvot village: excavation of 166 Kuwae Island 165, 167, 169; oral histories of destruction 165, 167–170 Kyushu island 313–324, 326–340 Lake Constance: Arbon-Bleiche 235, 236, 237, 238, 239, 240, 246; BodmanBreite 244–245; Bodman-Löchle 237; Bodman-Schachen 235, 236, 237, 238, 239, 240, 246; Bodman-Weiler 237; Espasinger Niederung plain 239, 244; flooding 241; hydrology of 241; lake levels, GIS simulation of 241–243; lake villages, discovery of 237; lake villages,

INDEX

inward drift of 244, 245; lake-level fluctuations 239, 240, 241, 245; rainfall patterns 239; rising lake levels, cultural impact of 10, 15, 245–247; Stockacher Ach 239; Tägerwilen-Hochstrasse 245 Lake Griefen: Fällanden-Wigartenstrasse 245 Lake Nyos disaster 13 Lake Pfaffikon: Pfaffikon-Hotzenweid 245; Pfaffikon-Steinacker 245 lake villages: abandonment of 235, 240, 244, 246; discovery of 237; lake-level fluctuations, cultural impact of 10, 15, 245–247; settlement pattern, inward drift of 244, 245; see also ArbonBleiche, Bodman-Schachen Lake Zug 245 Lake Zurich: Dietikon-Vorstadtstrasse 245; Erlenbach-im-Grund 245; Erlenbach-Obstgartenstrasse 245; Küsnacht-Itschnach 245; lake-level fluctuations 239, 240, 241 Laki Fissure 11–12, 87–101; Britain, environmental impact 88, 89, 90, 93, 96, 98, 99; crops, destruction of 87, 92, 93, 94, 95, 100; cultural impact 98–99, 100; dry fog 11–12, 88, 89–100, 265; folk memory, impact on 100, 101; France, environmental impact 90–91, 93–94, 96, 98; Germany, environmental impact 91, 95; Iceland, environmental impact 87, 88, 92–93, 102; Italy, environmental impact 91–92, 95; mortality rates 96, 98, 100; Netherlands, environmental impact 91, 94–95; North Africa, environmental impact 92; respiratory disorders, relationship to 96, 98; sulphur emissions 11–12, 88, 89–100, 265; Switzerland, environmental impact 92 Lambayeque valley 214, 226 Lapita culture 166, 304 Lemieng village: relocation of 39 Long Island eruption: oral tradition of 49 Lytton, B. 117 Mabalacat: lahar damage 61 Madang-Bogia tsunami 30, 37, 41 Maebira pottery 328 Maetani excavation 334 Mageshima Island 321 Malia 257; destruction of 259 Malo mission station: abandonment of 35, 36

347

Malol 32; Aitape tsunami, impact of 34, 35; casualty rates 35; tsunami sediments 40 Maoris: Auckland Volcanic Field, oral traditions of 150; disaster culture, development of 152–153; homeland of 127, 149; initial settlement of New Zealand 126, 127, 128, 134, 141, 154; Kaharoa eruption, oral tradition of 150; lava, oral tradition of 150; mythology 147–151; natural disasters, cultural perception of 13; Ngatoroirangi, legend of 149; obsidian, use of 144; ochre, use of 145–146; Pihanga, legend of 149; Rangitoto Island, oral tradition of 150; Ruaumoko, myth of 147; Taranaki eruptions, impact of 135, 137–138; Taranaki eruptions, oral traditions of 149; Tarawera, legend of 149–150; Taupo eruption, oral tradition of 149, 151; Three Kings volcano, oral tradition of 150; Tongariro eruption, oral tradition of 19, 150; volcanism, benefits of 144–146, 153; volcanism, cultural impact of 14, 15, 134–135, 137–144, 150, 152, 153; volcanism, oral traditions of 138, 147–151, 153; volcanism, social impact of 134–135, 137–144; volcanism, witnessing of 134, 138, 139, 140, 149, 150, 151; volcanoes, restricted access to 150, 151, 152; see also New Zealand Marine Style pottery 253, 259, 280 Mata, Roy: legend of 169–170 Maya: Ilopongo eruption, cultural impact of 264, 292, 295, 297, 298 Mayon volcano 44, 62 Mazama volcano 167 Messikommer, J. 237 Miike pumice 334 Mill Bay: erosion of 187 Minoan civilisation: cannibalism, evidence of 258; cave sanctuaries 258; central buildings, destruction of 251; centralisation 257–259; chamber tombs 258; collapse of 251, 252, 259, 273–275, 280; earthquakes, impact of 251, 252, 253, 259, 260; fire destruction 252; funerary practices 258–259; Linear A 251; Late Minoan IA phase 253, 257; Late Minoan IB phase 251, 252, 253, 257; looting 252; Marine Style pottery 253, 259, 279; maritime trade, disruption of 259;

348

INDEX

palaces 251, 257–259; palatial society 251, 256–258; peak sanctuaries 258; political restructuring 257–258; ritual sites 258; social structure 257–258; Santorini, environmental impact 254, 256–259, 260; Santorini, relationship to the collapse of 251–252, 273–275, 279, 280; settlement abandonment 257; social causes of collapse of 252; social organisation 251, 256–258; tsunami, cultural impact 254, 259, 280–282; villas 259; see also Santorini Miyazaki coastal plain 320 Moche culture 12; agricultural productivity 217; bet-hedging 226, 227–229; cemeteries 218–219, 224, 225, 226, 227; Cerro Blanco 217–227; chronology 206; collapse of 206; demography 206, 219, 221–222, 224–226; El Niño, cultural impact 221; environmental fluctuations, cultural impact 220–221, 225, 227–229; flooding, impact of 213, 215; Galindo 221–224, 227, 228; Huaca Forteleza 226–227; huacas 218, 221, 223, 224, 226–227; irrigation agriculture 217–219; Moche I–IV economy 217; Moche IV 216; Moche IV–V abandonment of territory 219; Moche IV–V ceramics 219; Moche IV–V monument construction 219, 220, 224; Moche IV–V transition 206, 219–222; Moche V 206, 222–225; Moche V monument construction 223–224; Pacatnamú cemetery 219, 224, 225, 227; Pampa Grande 222, 226–227 Moche valley: Project 215, 218; flooding of 214, 215 Mochlos: earthquake damage 253 Mt Pelée 264 Mt St Helens 256 Mycenae 259, 260 myth: see oral history natural disasters: characteristics of 20, 21, 23, 25; cultural impact of 1, 2, 3, 4, 7, 8, 9, 14, 15, 23, 25, 26, 35–36, 38, 39, 43, 87–88, 98–99, 100, 134–135, 137–144, 151, 172, 174, 176, 184, 185, 188–189, 193, 195–196, 201, 203, 204, 205, 206, 245, 246, 247, 250, 251, 252, 255, 259–260, 264, 265, 266, 274, 275, 280, 292, 294, 295, 297, 298, 299, 300, 301, 303, 322–324, 332; cultural

resilience to 172, 185–186, 292, 295, 302–303, 309–310; culture change, relationship to 1, 2, 7, 8, 9, 15, 20, 23, 25, 26, 308–310, 313, 314, 326, 332, 334, 335, 336, 339–340; earthquakes, cultural impact of 14, 172, 174, 176, 184, 185, 188–189; magnitude of 11, 12; oral traditions of 10, 13, 28, 30, 49, 138, 147–151, 153, 162, 164–170, 171, 186, 187, 194, 195, 196; psychological impact of 255–256; tsunamis, cultural impact 35, 36, 174, 192, 193, 254, 259, 278–279; volcanoes, cultural impact 1, 2, 7, 8, 9, 14, 15, 23, 25, 38, 87–88, 98–99, 100, 134–135, 137–144, 151, 153, 195–196, 252, 259–260, 264, 265, 266, 274, 275, 280, 292, 294, 295, 297, 298, 299, 300, 301, 303, 322–324, 332 Nero Caesar 114 Netherlands: Batavia 69, 70–81; Laki Fissure eruption, environmental impact 91, 94–95 New Zealand: Auckland Volcanic Field 128, 132, 134, 139–140, 150, 151; deforestation 134; ferrihydrite deposits 144–146; initial Maori settlement 126, 127–128, 134, 141, 154; Kaharoa eruption 128, 134, 140, 150, 151, 152, 153; Kawakawa eruption 147–148; Ngauruhoe 151; Novara expedition 149; Okataina volcano 128, 132, 140–144; Rangitoto Island 128, 132, 134, 139–140, 150, 151; Ruapehu volcano 132, 151; Taranaki volcano 128, 132, 134, 135–138, 149, 150, 151, 152; Tarawera eruption 127, 132, 143, 140–144, 151, 152, 153; Taupo Volcanic Zone 128, 134, 149, 150–151, 153; Te Awamutu 149; tephrochronology 132, 134, 153–154; Three Kings volcano 150; Tongariro Volcanic Centre 128, 134, 138–139, 149, 150, 151; Tuhua Volcanic Centre 128; volcanism, environmental impact 134; White Island volcano 128, 139, 151; see also Maoris Ngatoroirangi, legend of 149 Ngauruhoe eruption 151 Numundo plantation: volcanism, impact of 297, 299, 300, 304; see also West New Britain Ocean Bay Tradition 174–175 Ohira pottery 334

INDEX

Okataina volcano 128, 132, 140–144 Old Harbor: destruction of 174 Onejima: soil section 317 oral history: Aleutian 10, 13, 194, 195, 196; Alutiiq 186–187; earthquakes, oral traditions of 10, 13, 186–187, 194, 195, 196; Fiji 153; Maori 49, 138, 147–151, 153, 164–170, 187, 196; natural disasters, oral traditions of 10, 13, 28, 30, 49, 138, 147–151, 153, 162, 164–170, 171, 186, 187, 194, 195, 196; Torres Island 165–166, 170, 171; tsunamis, oral traditions of 28, 30, 37, 41, 165–166, 170, 171; Vanuatu 162, 164–168, 170, 171; volcanoes, oral traditions of 49, 138, 147–151, 153, 164–170, 186–187, 196 Osagari scoria fall 332 Oshigata-man pottery 323 Ostia harbour 115, 116 Osumi Peninsular 321 Pacatnamú cemetery 219, 224, 225, 227 Pacific Tsunami Warning Centre 31 Palaikastro: earthquake damage 253 Pamapa Grande 222; huaca 226–227; see also Moche culture Papua New Guinea: Aitape disaster 28, 30, 31–35, 39, 40, 41, 162; Long Island eruption 49; Madang-Bogia tsunami 30, 37, 41; Parker eruption 49; Rabaul eruption 301, 303; Ritter Island tsunami 30, 37, 40, 41; survivor mentality 13; tsunamis, cultural perception of 13, 36, 37; tsunamis, frequency of 37, 40, 41; tsunamis, oral traditions of 37; tsunamis, public awareness of 36–38, 41; tsunamis, settlement relocation 35–36, 39; Umboi Island 30; volcanism, cultural impact of 14; volcanism, settlement abandonment 9, 15; see also West New Britain Parker eruption 49 Pasig-Potrero River: Pinatubo eruption, impact on 45, 48, 50, 51 Pelsaert, Francisco 70–78 Peru: Casma valley 222; Cerro Blanco 217–227; Chan Chan valley 215; Chan Chan-Moche Valley Project 215, 218; Chao valley 222; Culebras valley 222; Cupinisque period 217; droughts 214; earthquakes, prehistoric evidence of 215–216; El Niño 211–215;

349

environment 211–213; Lambayeque valley 214, 226; Moche valley 213–215; natural disasters, cultural impact of 204–206; Nepena valley 222; palaeoenvironment 213–216; rainfall patterns 211–213; REAC 212; Santa Valley 213–214, 222; Shalinar period 217; Spanish conquest 215; Viru valley 222; see also Moche culture Petras 251, 257, 259; central buildings, destruction of 251 Phaistos 257; destruction of 259 Philippines: Capampangan 48–50; colonial history 48–50; Mayon eruption 44, 62; natural disasters, concealment of 10; Pinatubo eruption 9, 43, 44, 45, 50, 51, 53, 54, 58, 59, 61, 62; population explosion 48; volcanism, experience of 48, 49; see also Bacolor Pihanga: legend of 149 Pinatubo eruption 9, 43, 44–45, 50, 51, 52, 53–61, 62; Bacolor, impact on 43, 50, 51, 52, 53–61, 62, 63; casualty rates 52, 53; destruction caused by 43, 50, 51, 52, 53–61, 63; lahars 43, 45, 50–54, 58–62; Pasig-Potrero River, impact on 45, 48, 50, 51; settlements, destruction of 51–53; see also Bacolor Piranesi, G. 117 Pliny the Elder 109, 115 Pliny the Younger 109–110, 113, 266 Pompeii 10, 23, 87, 107, 193, 264: abandonment of 114; archaeological significance 117–119; aristocratic residences 111, 114; burial of 10; Casa di Julius Polybius 118; damage, preservation of 23; earthquakes, impact of 112; excavation of 107, 114, 116, 118, 119; looting of 114; Oplontis Villa A 114, 118; preservation of 110; restoration of 113, 114; tourism 116, 117, 119–120; Vesuvius, impact of 112, 114, 121; wealth of 111–112; see also Campania Herculaneum, Vesuvius Poobata: destruction of 61 Popocatépetl volcano 256 Poppeae 114 Poros: chamber tombs 258 Postumus, Agrippa 118 Prince William Sound earthquake 173, 174, 176, 177, 179, 180, 184, 185, 186, 187, 188, 193, 201 processualism 3

350

INDEX

Puteoli: growth of 115; trade network 115–116 Rabaul town and volcano 38, 301, 303 Rabaul Volcanological Observatory: 38 Rangitoto Island 128, 132, 134, 139–140, 151; cultural impact 139–140; oral tradition of 150 revitalisation movements 256 risk management 5 Ritsuryo system 335, 339 Ritter Island tsunami 30, 37, 40, 41; sediments of 40; settlement pattern, impact of 40 rock art: neuropsychological model 80 Roman Empire: Ostia 115, 116; wine trade 115; see also Herculaneum, Pompeii Ruapehu volcano 132, 151 Ruaumoko: myth of 147 Ryutaimon pottery 328 Sakurajima volcano 314, 326, 328 San Cristobal tsunami 37, 38 San Fernando: volcano defences 57 San Rafael: destruction of 61 San Vicente: chapel, rehabilitation of 57 Santa Ines: chapel, rehabilitation of 57 Santa Rita: flood defences 48 Santa Valley 222; flooding of 213–214; palaeoenvironment 213 Santorini: 8, 15, 23, 167, 250–256, 266, 273, 274–275; chronology 252, 253, 273–275, 277; environmental impact 254, 256–259, 260, 275, 277, 278, 279, 280; High Chronology 253; Low Chronology 253; Minoan collapse, relationship to 251, 252, 254, 259–260, 273, 274, 279, 280; tephra, distribution of 254, 275, 277; tsunami 254, 278– 279; see also Minoan civilisation Satsuma Peninsular 321 Satsuma-Iwodako 314 Satsuma-Iwo-Jima island 314 Schliemann, H. 117 Senokan pottery 323, 328, 329, 332 Sera village: relocation of 39 Settlement Point 172, 178–180, 182–184; erosion 179, 188; geology 179–180; Koniak tradition 178, 179, 184; seismic activity 180, 184, 188; subsidence 179, 188; village, abandonment of 184; village, construction of 180, 182; village, submergence of 182, 183; see

also Alutiiq Shepherd Islands: resettlement of 169 Sho pottery 324 Shumagin Islands: archaeology of 198–200; Chernabura Island, settlement history 198, 200, 202; earthquakes, cultural impact 194, 201, 203; earthquakes, oral traditions of 194, 195; European contact 201; initial settlement 199–200; population decline 200, 201; seismic activity 197, 198, 201; settlement history 198–200; tsunamis 192, 197, 198; Unga Island 202, 203; Unga village, abandonment of 201, 202 Siculus, Diodorus 109 Sissano village 32; abandonment of 36; Aitape tsunami, impact of 34–35; casualty rates 35; relocation of 39 Sitkalidak Island 188 Sklavokampos villa 259; see also Minoan civilisation Sobate pottery 323, 324, 326 Solomon Islands: San Cristobal tsunami 37, 38 St Pierre: destruction of 264 Stabiae: Vesuvius, impact of 112, 114 Strabo 109 Suetonius 113 Sueyoshi: tephra 329 Switzerland: lake-level fluctuations and culture change 10, 15, 245, 246, 247; Laki Fissure eruption, environmental impact 92; see also lake villages Tacitus 109 Takeshima tephra fall 314, 315 Talos myth 251 Tambora eruption 1816: 254, 264, 265; environmental impact 264, 273 Tanegashima island 314 Taranaki volcano 128, 132, 134, 135, 137, 138, 150, 151, 152; Burrell eruptions 135, 137, 138, 152; cultural impact 135, 137–138; Newall eruptions 135, 137, 138, 152; oral traditions of 149 Tarawera volcano 127, 132, 134, 140–144, 151, 152; cultural impact 140–144, 151, 152, 153; legend of 149–150 Tasman, Abel 126 Taupo Volcanic Zone 128, 150–151, 153; cultural impact 150–151; oral traditions of 149; tephra 128, 134

INDEX

Taveuni Island 151, 153; oral history of 153 Thera: see Santorini Three Kings volcano: oral tradition of 150 Tiberius Caesar 111, 115 Titus Caesar 113 Titus, Joel 165 Toba volcano 265–266, 280 Todoroki pottery 323, 326, 332, 334 Tongariro Volcanic Centre 128, 134, 138, 151; cultural impact 138–139; oral traditions of 149, 150 Torres Island: tsunamis, oral traditions of 165–166, 170, 171 tsunamis: Afognak village, impact on 185; Aitape disaster 28, 30, 31–35, 41, 162; Alutiiq, impact on 174; Buna 40; characteristics of 30, 31; cultural impact 35, 36, 39, 174, 192, 193, 254, 259, 280–281; cultural perception of 13, 36, 37; frequency of 37, 40, 41; MadangBogia tsunami 30, 37, 41; Minoans, cultural impact on 254, 259; oral traditions of 28, 30, 37, 41, 165–166, 170, 171; Pacific Tsunami Warning Centre 31; recurrence intervals 37, 40, 41; Ritter Island tsunami 30, 37, 40, 41; San Cristobal tsunami 37, 38; sediments of 40; Shumagin Islands, impact on 197, 198; warnings of 31 Tuhua Volcanic Centre 128 Tumlee Island: tsunami sediments 40 Twain, Mark 116 Umboi Island: settlement relocation 30; tsunami, impact of 40 Umnak Islands 196 Unalaska Islands 196 Unga Island 202, 203; abandonment of 201, 202 United Kingdom: Laki Fissure eruption, environmental impact 88, 89, 90, 93, 96, 98, 99 United Nations: International Decade for Disaster Reduction 2, 5 Vanuatu: Aoba, destruction of 164–165; initial settlement of 164, 166; Kurvot tsunami, oral tradition of 165–166, 170, 171; Kurvot village, excavation of 166; Kuwae eruption, oral history of 165, 167–170; Lapita culture 166; natural disasters, oral traditions of 10, 13, 162, 164–168, 170, 171; oral history, social

351

importance of 166; Plain Ware 166; Roy Mata, legend of 169–170; Shepherd Islands, resettlement of 169; Torres Island tsunami, oral history of 165–166, 170, 171; volcanic activity 162, 164 Veniaminov, I. 197, 198, 201 Vergulde Draeck: wreck of 84 Vesuvius 10, 87, 107, 108, 109, 110, 112, 114, 141, 264, 265, 269; archaeological significance 117–119; Campania, impact on 114–116; commercial impact 107, 108; cultural significance 116, 117; Herculaneum, impact on 112, 114; Pompeii, impact on 112, 114, 121; Stabiae, impact on 112, 114; tourism 119–121; volcanic history 108–109; see also Herculaneum, Pompeii volcanism: benefits of 153; climate change, impact on 88; crops, destruction of 87, 92–95, 100; cultural impact 1, 7, 8, 9, 14, 15, 23, 25, 38, 43, 87–88, 98–99, 100, 134–135, 137–144, 151, 153, 195–196, 251, 252, 259–260, 264–266, 273, 274–275, 279, 292, 294–301, 303, 322–324, 332; cultural resilience to 301–303; culture change, relationship to 7, 8, 9, 308–310, 313, 314, 326, 332–336, 339–340; environmental effects 87–92, 134, 264–266, 269, 271, 272, 273, 275, 278–281, 300–303, 317–321, 322, 324; gases, devastating impact of 87, 89, 90, 91; global temperatures, impact on 264–266; mass extinction, relationship to 265–266; oral traditions of 49, 138, 147–151, 153, 164–170, 186, 187, 196; super eruptions 265–266 Warapu village 32, 34, 35; Aitape tsunami, impact of 34, 35; casualty rates 35; relocation of 39 West New Britain: Dakataua eruption 296, 297, 299, 300, 301, 303, 304, 307, 308; disasters, cultural resilience to 295, 302–303; forest regeneration 301–303; Lapita culture 304; lithic technology 302, 306; obsidian trade 304; pottery trade 304; recolonisation of 302–305, 307–309; resource management 302–303, 309; sedentism, development of 302; settlement abandonment 9, 15, 298–299, 302, 305, 307, 309; trade

352

INDEX

networks 304; volcanic activity 293, 296–298; volcanism, cultural impact 14, 15, 298, 299–303; volcanism, environmental impact 300–303; Witori eruption 296–297, 298, 299, 300, 301, 303, 304, 305, 306, 307, 309–310; see also Bitokara, Garau Island, Garu plantation, Numundo plantation, Willaumez Peninsular White Island 128, 139, 151 Willaumez Peninsular: volcanism, impact on 296, 297, 299, 308; see also West New Britain Winckelman, J. 116 Winslow, M. 197 Witori Volcanic Centre 296–297, 298,

299, 300, 301, 303, 304, 305, 306, 307, 309–310 Yachmenev, A. 195 Yakushima island 314 Yamagawa base-surge 332 Yayoi culture 326, 335 Yombon: sedentism, development of 302; volcanism, impact of 297; see also West New Britain Yonemaru: tephra 328, 329 Zahn, W. 117 Zakros 257, 259 Zeewyck: wreck of 81 Zutdorp: wreck of 84